Do you always use the same intensity of stimulation? I suppose it changes, beacuse it is proportional to the maximum fEPSP amplitude you have. My idea is that a value proportional to the maximum doesn't mean an equal number of dendrites to reclute.

Anyway I have experience of PPF and PPD only in-vivo animals, where I'm used to a bigger varibility. I am afraid it is someway different in slices.

How did you calculate the PPF, amplitude or slope?

if it is the slope, how did you calculate it?

What is the frequency of stimulation between two PPFs?

What is the age of the animal?

You mention that you are stimulating at 50% of maximum, but what is the range of current that you are using to stimulate. From the traces you have shown it looks like there is a big difference in the amplitude of the baseline responses. In your first data set with the lowest PPF ratio, it seems as that a pop spike is visible in the first trace, which would suggest that your are not using a sub-threshold stimulus. This would also explain why the first baseline response is approximately 1 volt where as the baseline response in the third data set is only 300 mV (not sure about the units). In other words I believe must of the variability in your PPF is due to the variability in the baseline response and presumably to the amplitude of the stimulus you are using.

It seems to me from your traces that you're evoking a population spike. It would be good to avoid it because it changes values of amplitude or slope of the response. So, maybe try smaller stimulation intensities. Then, for field experiments slope is more reliable (again, it changes less if there is a pop-spike, which might appear in the second response) - maybe if you're measuring only the amplitude, it would be good idea to measure the slope as well.

If you record in the interface chamber, the increase of the response (usually it comes together with the drift in the basal voltage) may indicate that your slice is drying out. If it is the case, it may help to check the quality of chamber and perfusion rate.

Hope it will help! Good luck!

Do a correlation between the slope of the initial event and the magnitude of PPF. I bet you'll find something interesting. Then think about what might be causing that. Hint: See if it comes more reproducible with/without bicuculline.

I see population spike (PS) components (upgoing wave) in the middle of the fEPSPs. Especially the top two traces, they are quite obvious. (Also, the second field potentials in these four traces are all distorted by PS, too.) They may be screwing up your EPSP shape as well as the PPF ratio. I recommend you to reduce stimuli amplitude in order to avoid population spike induction.

One more advice is to average the traces. That will make PPF ratio more stable.

What about the membrane voltage? The Y axis scales vary a lot from one recording to the others. As, NMDAR are voltage-dépendent, Vm will influence NMDARs and comsequently PPF.

Just to carry on with Stephen's comments

I have quite some experience with PPF/PPD

The first thing that hits the eye is the pathological state of the slice.

The fEPSP shows multiple spikes, a sign of disinhibition of the slice (traces 3 and 4). Did you perfuse the slice with bicu or PTX? If you did not, then there are problems with the slicing procedure.

This is the reason why I asked about age. Which slicer did you use? Are you submerged or interface? What is the perfusion speed?

In any case, you need to stimulate every 30s (max every 20s) to keep a constant base line.

The problem with your fields is the contamination by population spikes. This dramatically affects the slope and the measure. The first displayed response clearly shows a popEPSP followed by a pop spike. This can be due to suprathreshold stim intensity, or more likely a problem with slice quality.

The location of the recording and stim electrodes is also central. Where are they?

First of all, thank you all for your comments, they are being very useful for me in terms of guessing what I am doing wrong.

Elisa and Stephen, usually I don't pay attention to the intensity I use, but normally after setting at 50% of maximum, the intensity is 10 ~mV. I am aware I have differences in the baseline between experiments, but as I was always considering percentages, I've never given importance to this, unless the responses are too low (less than 250 mV). Do you think I should focus more in the intensity (for example, subthreshold) than the percentage of the maximal response?

Svetlana, in almost all of my experiments I see a pSpike in the second of pairs of responses. I've never tried to measure the slope. What parameters should I use for measuring the slope?

I am working with the slice summerged.

William, I will try this correlation. I guess I'll find out that the greater slope the smaller PPF? Should I look for setting the same slope/amplitude for the baseline in all my experiments instead of 50% of maximal response? I've never tried to use any blocker. My ACSF composition during the recording is (mM): D-glucose 10, NaHCO3 25.6, NaCl 124, KCl 2.45, KH2PO4 1.2, CaCl2 2.5 and MgSO4 0,8.

Ryuichi, I will think seriously about avoiding this pSpikes in my signals. I also think that it could be a problem when I am performing the I/O to infer the 50% of maximal response, because I increase at 5 mV intervals until the amplitude stops increasing, but I don’t pay attention whether there are or not pSpikes during this I/O procedure. The PPF ratio values are the average of five measurements.

Erik, as far as I know I can’t manipulate the membrane voltage because I am working with field recordings, but my experience is not very large. Do you think I should consider blocking NMDA receptors?

Christophe, the animals are all males between 6-7 weeks. I calculate the PPF always using the amplitude (EPSP2/EPSP1). My protocol is with single recordings at 0.033 Hz during 30 minutes, and then I assay PPF by evoking five paired recordings, separated 3 minutes one to the other, with single stimulations at the same 0.033 Hz during these three minutes between PPFs. Do you think I should evoke paired responses during the whole recording and use the first response as a measure of the baseline and the second one for the PPF ratio? Do you think that my protocol can be causing that in some of my experiments, after eliciting this set of five PPF, the baseline starts slowly increasing (see figure)?

I started my training in slicing hippocampus from rat almost ten months ago, so I have to consider the posibility that my slices are not specially healhly, but people with experience have said to me that normally they are ok. In any case, I never use bicuculline nor picrotoxin. I make the slices with Vibratome. I use the same ACSF composition I stated before for the recording, except for magnesium (8 mM). Once the slice is cut, I transfer it to a holding chamber containing the same high magnesium ACSF bubbled with 95% O2/5% CO2 at room temperature, where I let them to recover for around 1.5 h. Then I transfer the slice to the recording chamber and it takes it around 1.5 h. to adquire a stable baseline response (recording is also at room temperature). The perfussion speed is 2.5 ml/min. The stimulating electrode with broken tip is positioned in Schaffer collateral-commisural fibers and the recording is placed in apical dendritic layer of CA1. I attach a picture with some examples of what my slices with the electrodes placed look like, and a graph showing this potentiation after PPF which I mentioned above.

Again, many thanks everybody for your invaluable help.

All your traces are showing a hint of population spikes. You probably should move your recording electrode further away from the pyramidal cell body layer. A really nice field EPSP decays with an approximately single exponential component.

In aggreement with both Christophe's assessment and many others in the thread, I think the first thing that comes to mind is that you have pop spike contamination. Furthermore, the appearance of muliple pop spikes in your recordings suggest unhealthy preparations. Are you conducting these experiments in a submerged or interface configuration? I have found that if using a submerged configuration, bath depth and volume is important to the health of your slice during an experiment.

It is my rule of thumb that bath volume should be theoretically close to full exchange within one minute. Thus, if you are flowing at 2.5ml/min bath volume should be close to 2.5ml max. Moreover, what is the deadspace volume between resevior and slice bath? If it is more that 2.5ml there is a chance you are de-gassing your aCSF if you are using Tygon tubing even before your solution hits the bath.

Finally, there appears to be some run down in the traces you posted in your second response (especially in the blue trace) and even before you get to PPF. Have you looked at the stability of your fEPSPs over the window of time these you are conducting these experiments?

Hi Diego.

Since you are using old rats, the slicing procedure is critical. To ensure that you get good slices, I recommend perfusing animals before doing experiments with an ice cold modified ACSF (like sucrose-based). After brain extraction, you cut the hippo in the same ice cold modified ACSF. Then, you transfer them is your holding chamber.

Regarding PP experiments, I would perform them all at physiological temperature as paired pulse phenomena are critically temperature-dependent.

Always manipulate the slices gently.

I could increase the perfusion speed up to 8 ml/min (this is what we are using - see N. Hajos paper in J Neuroscience). btw, if your temperature change during the experiment, the PP ratio will change. You need a feedback control loop of the temperature.

Your electrodes appear to be appropriately placed.

If you can put a second recording electrode in the PC layer, you could build input/output curves, which give a lot of information.

You should not use the amplitude to calculate PP ratio; only the 10-90% rise time of the fEPSP. You can check my early work of this: Bernard & Wheal.

Your main problem is the viability of the slice; this may be the main reason for instability.

Don't hesitate to contact me direclty if you need help.

Dear Diego,

As many of commentators agreed, it would be better to measure the slope of the response instead of amplitude. For that, you can do the linear approximation of the decaying part of the responce and take the slope of this line. In this case you should avoid the part of the response with pop-spike. Many programs have a build-in calculation protocol for the slope, for example WinLTP (www.winltp.com/). Here is the good example of the slope measurement (fig. 1 in the paper): http://www.sciencedirect.com/science/article/pii/S0165027007000027

One more thing to advise: it has already been discussed here that the stim intensity may be important. For most of the stim boxes you can not rely on the stim voltage because it will vary between electrodes. In your case I would try to have more or less stable amplitude of the first response between recordings. And for this, I would not use 50 % of the max peak amplitude, but go for much smaller values. With the stimulation too high you might start depleting readily releasable vesicle pool, and in this case the ratio of the paired response will shift to PPD.

It is normal that you have pop spikes in the second response, it just means that you really have PPF :) The thing is that if you have pop-spike, you can not use amplitude. So, use slope.

Good luck!

Hi Diego,

Something else to consider is the slice orientation. You do not say whether you are cutting horizontal, saggital, coronal or transverse. This is important since different slice orientations will sample different portions of the hippocampus along the dorso-ventral axis (or in the case of transverse will sample all along the DV axis). The location of the slice along the DV axis has very large effects on PPF - ventral slices will have much lower PPF than dorsal slices. For details see: http://www.ncbi.nlm.nih.gov/pubmed/11811659

That said I also agree with other commenters - in these examples your slices look fairly disinhibited. Plus, I agree with Christophe, that PPF experiments (indeed all brain slice experiments) should be done at physiological temperature.

Jon

Many thanks again for all the answers. They are being really helpful.

Today I did new experiments and I got signals (most) without pSpikes, very clean signals. The difference was that I broke much less the tip of the stimulating electrode and I replaced it further of the pyramidal layer. Also the recording electrode was further of the pyramidal layer in CA1. However none of the slices I tried today showed PPF. In despite of showing such nice responses (see picture).

Do you know if there is some variability between animals, in a way that maybe the animal I used today “didn’t have PPF”?

Moreover, can anyone confirm me the viability of my protocol? As I showed before I am performing single responses every 30 seconds during 30 minutes (baseline) and then I assess PPF by eliciting a total of five paired stimulations (50 ms), every pair is separated to each other by three minutes. During these three minutes between pairs I still record baseline responses (single stimulations) every 30 seconds. Is this correct? Should I elicit paired responses continuosly during the whole experiment, every 30 seconds?

Alfredo, my flow rate is 2.5 ml/min. I will recheck the deadspace again though. Thank you. We already tried the stability of the experiment for up to 5 hours, but it is true that sometimes this rundown occurs and when it happens after the first 45 minutes (Baseline + PPF in control) I can’t discrepate between a rundown or a drug effect…

Fang, Christophe, Svetlana and Jon, thank you very much for all your comments. I will consider the aspect of temperature having in mind that is important. Also I will try to set the software in order to measure slope. Regarding the preparation, I make horizontal cuts in each hemisphere. I do each separately. Normally it takes me 10 minutes to cut every hemisphere; during this time the other hemisphere is “waiting” in the same cold and bubbled ACSF. I read something before about these differences in PPF among dorsal and ventral slices. Regarding the temperature, can I perform these experiments at room temperature, just assuminig that the PPF ratio will be different? With room temperature, would you expect greater ratios than with physiological temperature?

Thank you very much for your help and patience..

When we cut horizontal slices we get much less PPF than when we cut coronal or parasaggital.

If you are cutting horizontally, I can't see any reason why you can't cut both hemispheres simultaneously. In fact what we do when we cut horizontal slices is mount and section the whole brain (not hemisectioned) and then trim out the hippocampal slices after slicing in a petri dish filled with carbogenated aCSF. This will improve slice quality certainly in the half that is normally waiting (dying!) for 10 min prior to cutting.

I would also recommend slicing in a sucrose-based aCSF before transferring to your normal aCSF for recovery/recording. Even if you don't want to go down the route of perfusion (which we don't), the sucrose-based cutting solution is thought to better preserve inhibition. This would help with your hyperexcitable slices...

Hi,

I agree with the others about the contamination of population spikes. In any case you can find a variability in the PPF, because there are synapses with different probability of release (it is not strange to find also PPD). In fact, this is not the best technique to study PPR and some scientists do no believe to any conclusion from this type of measurement. In any case if you apply some drugs and you have the internal control before I think that it is informative.

Ciao

laura

A simple comment...Are you sure that you are on steady state? You may exted the time between the pulses,,,

In bocca la Lupo

Ciao

OK, my last comment on this!

Your PPF may appear lower than it actually is because of the way you are filtering the data - correct me if I'm wrong but you appear to be high pass filtering the data which means you are getting a slow positive going response after the first EPSP, which your second EPSP is riding on. This is likely to be an artefact from the HP filter. Depending on where you measure the amplitude of the 2nd response from (i.e. baseline prior to 1st response or prior to 2nd) this will distort the amplitude of the second EPSP.

Thank you Jon, Laura and Paolo for your comments.

If I set the first response being sure that it doesn't have any pSpike, but the second response elicits pSpike, how reliable can be measuring the slope? It will be also affected by the pSpike, won't it?

We are working with 50 ms intervals (often used in literature), I think it is enough to get a steady state in most of my experiments. We are measuring the amplitude as an absolute negative valour from zero.

Is it ok to evoke paired pulse responses (one every 3 minutes), just during 15 minutes after 30 minutes of baseline (single responses at 0.033 Hz)?

The point of measuring the slope is that in general it is less contaminated by popspikes, which usually occur around or just after the peak of the EPSP, not during the rising phase.

50ms should be fine - we routinely use 50ms pulses every 30s with no run up or down in basal responses.

If you are measuring amplitude as absolute deflection from zero, this will be affected by a high pass filter if you are doing this. reducing or eliminating your HP filter will may help.

Other than the many factors already stated, one factor to emphasize is that you have to control the stimulus intensity. We found that PPF as measured by the ratio of initial slopes of the field EPSP (measured at least 0.5 ms before the onset of the population spike) decreased with stimulus intensity. However, when slopes were used to estimate PPF, using

Stimulus strength can be an important factor, but there is a common misconception as the why this can be the case. One needs to understand that the stimulus strength does not matter at all to transmitter release of an individual synapse within your recorded population. The action potential that is evoked by your stimulus is all-or-none, and it doesn't matter how 'hard' you have to stimulate to evoke it. You don't get any change in release probability because an individual presynaptic action potential was evoked with a stronger stimulus. Since the field potentials that are typically recorded for PPF and other measures are population responses, turning up the stimulus has the main effect of increasing the size of the population of axons that are stimulated, and thus increasing the NUMBER of synapses activated. In a homogeneous population of synapses, this will have no effect on PPF directly. If turning up the stimulus strength is having an apparent effect of PPF, then this is likely coming from an indirect effect. Possibilities include evoking larger EPSPs (with a greater number of axons) that more closely approach the reversal potential (and thus have less 'head-room' to get larger, or increased synaptic inhibition that may influence the response to the second stimulus. Both of these factors can be minimized by recording only the initial slope and only in the dendritic layer (EPSPs recorded in s. pyramidale are more susceptible to influence by synaptic inhibition, and besides, are an indirect reflection of the synaptic current). I would upgrade the advice that many have given about recording only the initial slope of the EPSP from 'important' to 'essential'. The population spike is NOT a synaptic potential, and besides, it behaves quite differently then the fEPSP with regards to PPF and other factors.

When I first offered an answer to this question, I was commenting more on the previous answers than I was the original question. I hadn't actually looked at the figure before today. I was surprised by the very high PP ratio numbers. I never get anything above about 1.5, and at the interpulse interval they are using, I usually get depression.

Having now looked at the figure supplied by the original questioner, I feel fairly certain that they have misinterpreted the amount of PPF in 3 out 4 of their illustrated experiments. I think that only their first set of field potentials (green), and maybe the second (blue) are interpretable with regard to PPF. In the top trace, I think it is measured correctly, as long as they took their measurement in the initial slope before 'notch' near the peak in the negative slope. The PPF reported in the top trace seems about right (by eye), and is in line with what I see in my own experiments. The second (blue) trace, I would actually interpret as PP-depression (perhaps, hard to tell by eye - it is pretty close to a ratio of '1'), BUT, because they are reporting a very high PP ratio, it appears that they have measured the response to their second stimulus in the very steeply negative-going region after the positive-going notch. I would interpret this steep region after the positive-going notch as being heavily contaminated by non-synaptic active conductances (likely a mixture of a reflected population spike and dendritic action potentials). Thus it is not of much use in determining how much PPF one has. The same is likely the case in bottom two traces. This is a common mistake in measuring field potentials - mistaking active postsynaptic conductances for 'synaptic' responses. They are driven by the synaptic response, but they are not themselves synaptic. The simplest example of this is the population spike. The population spike is driven by the synaptic potential, but it is not itself a synaptic potential.

Bottom line; I would throw out the bottom two traces, and only measure the second one before the notch. The first (top) one is the only clean recording. When I see things like this, I usually try to re-position my electrodes to get a cleaner synaptic response.

Indeed, as a general rule of thumb, only measure initial slope of EPSP field, and anything after a 'notch' is suspicious - i.e. there is a good probability that everything after that notch is contaminated by non-synaptic elements. Changes in the slope of the rising phase of the EPSP usually indicate the additional of non-synaptic elements (it is common to see a smooth initial slope, and then a sudden increase in that slope - this is suspicious). If you are recording in the dendrites, anything after a positive-going region is likely not purely synaptic. By 'suspicious', I mean that it is not purely synaptic. Interpreting a non synaptic potential as if it was a synaptic potential will lead to misinterpretation of your results. In this case, they have hugely overestimated the amount of PPF.

This is something I drill into my students: If it's not synaptic, you can't interpret it as synaptic ;-)

OK, sorry, just saw the new traces. They are much better. There are no longer concerns about pop spikes and such contaminating these traces. I don't think there is anything wrong with your protocol - I think if anything, it's overkill - you don't really need to wait 3 min between your PPF. The fact that you now have ratio's around =1 with you current protocol is very much in line with what I generally see and the literature. But you should consider different interpulse intervals. There are at least two processes at play here: 1) Synapse fatigue/vesicle depletion and 2) calcium-mediated facilitation. The depression tends to dominate at shorter inter-pulse intervals and facilitation at longer. We typically see about 30% PP depression at ~20 ms and 20-50% PP facilitation at 100-150 ms interval. The cross-over happens somewhere in the vicinity of 50ms. I would suggest that you collect data at a range of intervals to try to gain more information about the interplay between the depression/facilitation. We typically use an array of intervals at ,20,30,40,50,75,100,150,250,500 and 1000 ms (5 trials each interval). then rather than expressing your result at one interval, you plot a ratio vs interval. We do our ppf at test frequency (i.e. one every 30 s) and don't see any interaction between the stimulus (i.e. rundown or runup). We are working at room temp. Keep in mind also that synaptic inhibition evoked by the first stimulus starts to kick in at 20-40 ms after the first stimulus and continues for 100's of ms. So doing your experiments in the presence of GABA A and B blockers might decrease your variability and make it easier to interpret your results.

Daniel - it's interesting you say that. At (close to) physiological temperatures we almost always see PPF even at very short intervals - PPR peaks around 50 ms. See attached data from mice, transverse slices.

Jon - I suppose it's not surprising that the amount and timing of PPF will change with temperature. Your data looks very nice.

Firts of all, thank you very much again for all your responses. I am realising about many things I didn’t know before, I am learning a lot thank to all of you.

I think I am still missing something in the interpretation of the pSpikes. If I understood well, a pSpike is a non-synaptic current which comes from the syncronised action potentials fired by cells located (in this case) in CA1 pyramidal layer. For this reason, in the beginning I thought that I see the pSpike as an upward deflection supperposed on my fEPSP (second response of the pair in the example A in the figure, with the location of the electrodes in the slice). In this case, should I reject the experiment, or can I still measure the slope just before the pSpike positive-going notch (I think, the slope of the trace just before the blue vertical lines), as you Daniel suggest?

However, I get this kind of responses when I place my recording electrode medial/distal in the str. radiatum, and this upgoing deflection is getting greater as further the recording electrode is from str. pyramidale, i.e., as distal it is in str. radiatum (B in the figure). For me, this is the commonly described pSpike, but my question is: is it actually coming from action potentials in pyramidal cells of CA1? It is difficult for me to understand it because it gets bigger as more distal I place the electrode from these CA1 pyramidal cells. Because of this, I am thinking that maybe this populations spike comes from another population of cells, perhaps interneurons located in the distal str. radiatum or even in str. lacunasum moleculare, which are well described (Lacaille and Schwartzkroin, 1988 J Neurosci; Dvorak-Carbone and Schuman, 1999 J Neurophysiol). What do you think?

Moreover, when I place the electrode in Str. pyramidale, I see a very sharp downward deflection followed by a positive response (C in the Figure), and this deflection gets bigger in the paired response. I think that this downward deflection is also a pSpike, but in this case, a pSpike which actually is elicited by pyramidal neurons in str. pyramidale. Therefore I can not use it to assess PPF of synaptic responses, is it right? Moreover, there is no a fEPSP component in this third example, right?

I always try to place the electrodes in the same position, like in figure A, but I am afraid that I am measuring pSpikes very often without realising that they are actually it, and mistaking them as fEPSPs. In this sense, in the figure, I wonder what kind of response is every component I see in the signal. Can you give me your opinion? I numberer them from 1 to 6, to make it easier to refer to each one. It would be really helpful if I know whether I am right or not with my interpreation of these components.

Daniel, I already performed this experiment for different ISIs some months ago, and I got the greater PPF ratio for 50 ms. But today, I am not sure whether this PPF that I measured is from fEPSPs or actually from a kind of population spike facilitation…

Finally, I work with broken stimulation electrodes. Because they are never broken exactly the same between experiments, do you think that this can affect my measurments? How much broken should the electrode be? Normally mine are like the one showed in the slice C (I emphasized it).

Sorry for having so many questions… but as more I know, as more confused I get! :S

Thank you again for your fantastic help.

1) I think it's fine to measure the slope before the population spike commences. The only issue is whether the latency of the population spike moves so far forward that you have an insufficient duration of uncontaminated EPSP to get an accurate slope measure. This doesn't look to be a problem in your Fig A.

2) So, going back to fundamentals, a positive-going field potential is a current source - and so most likely represents a population outward current in your recorded cells. This makes it unlikely that this initial positivity you see in LM is from local somata (because action potentials from these cells would be an inward current and thus a current sink). The reason that the population spike is positive-going in the SR is because the inward current in the soma travels down the dendritic lumen and then sources out the dendrite. What you see in the LM is hard to interpret, but my best guess is that the sourcing of the current that has traveled from the somatic action potential (in SP) is simply more dense here. I would tend to think that the total surface area of the pyramidal cell dendrite is higher in LM, and thus a more favorable total conductance for sourcing across the membrane. Basically, keep your electrode in SR.

3) The immediate downward deflection you see in SP is certainly a population spike. The fact that is is coming before you see any positive-going source (the source corresponding to the synaptic sink in the dendrite) worries me a bit. It is either that your cells are so excitable that they are firing immediately, or more potentially problematic, you are activating your 'postsynaptic' pyramidal neurons directly by the stimulators (i.e. not synaptically- the current from your stimulating electrode is directly activating the pyramidal cells). Looking at your picture, it looks like you have the stimulating electrode far-enough away that I wouldn't expect this, but on the other hand, I'm not familiar with the behavior of monopolar stimulating electrodes. It would be expected that the stimulus current from a monopolar electrode will spread farther than it would from a bipolar electrode. I would suggest trying a bipolar electrode, even better a concentric bipolar electrode (Frederick Haer), and be sure to attach it to a stimulus isolation unit (otherwise you are stimulating to ground, which also increases stimulus spread).

4) Regarding what you can and can't do with pop spikes: It's not so much that you 'can't' use a pop spike for a particular purpose, it's just that you should be aware of what you are actually looking at when you interpret it. The EPSP field in the dendrite is a fairly pure look at the EPSP (at least the part before the reflected pop spike). The Population spike includes the EPSP, everything that happened in the dendrite while the EPSP was travelling to the soma, and the characteristics of the somatic action potential. So, for instance, if you wanted to know the effect of paired pulses on pyramidal cell excitability, then population spike would be an appropriate measure. If, however, you want to know the effect of paired pulses on synaptic transmission, then it's not.

5) In your 3rd figure, the positivity that follows your population spike is probably a mixture of the EPSP source (reflecting the EPSP sink in the dendrite) and to a lesser extent, the outward currents associated with the falling phase of the action potential. The duration of the EPSP outlasts the duration of the population spike, so they are commonly followed by a positivity. Incidentally, there appears to be a second population spike in your first response of Fig C. This is generally an indication of impaired synaptic inhibition in your slice. It is more often seen in older or more damaged slices, and, of course, in cases where GABA receptors are blocked.

Thank you very much again Daniel. I was reading this weekend about this basics you mention (currents and sinks in extracellular recordings) and I think that now I perfectly understand what you mean. Everything is clear for me: the origin of this population spike, how it looks from SRad or SPyr, what a pure and noncontaminated fEPSP should look like.

But still I have a question. In the last figure I submitted, in SRad, and also in the first figures I submitted, the last two recordings (also from Srad)... what is this very sharp downward deflection that occurs after the pSpike and you strongly recommend not to measure. I often get it when recording in Srad. Is this kind of dendritic action potentials?

Thank you

You're welcome. I've been recording these li'l suckers for a long time ;-). Honestly, I don't know exactly is the basis of those sharp downward deflections. As I tell my students: the more milliseconds you get away from the stimulus, the less interpretable the field becomes. I too see those sharp downward deflections in two circumstances: 1) immediately after a population spike, and 2) before the population spike, in the latter phases of the 'EPSP' downward slope.

In the former case - it really doesn't matter what the identity of the downward deflection is - anything after the pop spike is not interpretable as a synaptic potential. I think it's likely to be a mixture of currents associated with the falling phase of the pop spike, and active dendritic conductances (i.e. dendritic action potentials). The latter case is more potentially problematic. It is not uncommon to see a 'sharpening' of the downward slope of the EPSP field somewhere during that downward slope. Sometimes, there is a distinct 'notch' in the rising phase where the slope suddenly gets steeper. You can see a fairly good example of this in the 4th trace of your original figure (response to the second stimulus), where there is an initial slope that suddenly gets steeper. I suspect that this is due to dendritic action potentials, but I don't know for sure. Because I'm uncertain, I simply don't measure "EPSP slope" in that steeper region of the field, because I don't think it's synaptic.

Daniel, thank you for all your contributions. They were so helpful, like those from other people. Now, thanks to all these advice, trying to avoid this pSpikes, at least in the first fEPSP of a pair, and measuring the slope instead of the amplitude, my paired-pulse ratio is much more consistent with a value around 1.2.

I will keep working the same way and hopefully things will improve.

This was my first question posted in RG and I feel so glad and grateful for the assistance I got. Thank all of you for your help.

Kind regards,

Diego Fernández

It is great to learn from this topic. When I did input-output experiment, I got a kind of responses after EPSP if the stimulus reached the maximal. The figure is attached. Some papers indicate that this is a sign of hyperexcitatory slices.

The transverse slices is prepared from 20-24days C57 mice. 100uM picrotoxin was added in recording normal ACSF. And a cut was made between CA3 and CA1 to avoid the epilepsy.

Is this response indicated in attachment a EPSP? And why? Is this a CA1 recurrent input? Thank you.

Great discussion on fEPSPs here...

Moore-your responses are pretty typical of what one sees when the slice is 'disinhibited' by picrotoxin. Without the strong shunting provided by GABAergic inhibition, all sorts of pop spike contamination occurs. Which goes back to Dan Madison's point--you would have to measure the slope in order to have any chance of quantifying the fEPSP.

One other thing that is apparent from the smaller responses (the ones with more of a 'pure' fEPSP) is that the ratio of the fiber volley to fEPSP is a bit high. Generally, a small FV/fEPSP ratio is an indication of a healthier circuit. I don't know about your particular experimental paradigm, slice prep, etc. But something to think about...

Moore Yip: To answer your question, so long as your recording is deep in Radium, you've got your labels a bit mixed up. In Stratum Radiatum, the "reflection" (aka current source) of the population spike is infact a positive going wave. I.e. the population spike is not the negative peak, but the positive going waveform preceding it. Note how it moves earlier and earlier in time as you induce larger EPSPs. The fEPSP continues on, and then you have what is probably a second population spike, again, forcing the waveform up the page. It is broader than the first one, because it lacks synchrony, that is to say, all the neurons generating it are not spiking over a larger time range than the first one. See here for how the population spike "inverts" as you move deeper into Stratum Radiatum.

http://jn.physiology.org/content/jn/86/5/2435/F1.medium.gif

 Great discussion!!! but after read this, i have  one question. If i want record PPF and LTP  in same slice, so how should i do first?  Can i record 30 min PPF frist , use the PPF  data (P1) as the baseline, and if the baseline is stable, then use HFS or TBS induce, then recoding 1hr . is it ok?

Thank you. but i stil do not understand. I mean if i want to record the PPF and LTP,how should i do?  i know i can not record LTP first, then record PPF in the same slice. but if i record PPF first, then can i record the LTP? i mean whether  the PPF  can effect the LTP or not? 

The PPF in fEPSC shows how healthy is your slice, then you can do an I/O curve , and finally do your LTP experiment , I did my experiments in that order , and didn't have any trouble