Hi Justin, you can avoid the spread in different ways.

1) if you have a pico-injector that allows you to apply a slightly negative pressure, once you lower the pipette you won't lose tracer from the tip. 

2) if you load a tiny drop of some neutral unsoluble solution after your tracer (mineral oil) this, even if really low in quantity, will efficiently block the spread.

3) if your injection is low in volume (an injection every minute, for a volume of 20nL ca) for a total of 1-3 injections and if you wait at least 10-15 minutes after the last before extracting the pipette from the tissue, you should be able to avoid the spread that usually occurs at the end of the injection protocol.

to improve specificity and precision some people even start injecting 5 minutes after lowering the pipette, which in this case is a step that takes 1 minute each 200-300 micrometer (usually I am not that slow but this could allow the tissue to come back in shape, specially if you reach a deep region)

relatively to your problem, depending on which BNST nucleus you are targeting you almost can't avoid the ventricles. best would be to enter from the cortex above the CPu with an angled injection (you can easily re-adjust coordinates with pitagora) which would totally avoid the issue, roughly could be 30 degrees. that's what I do. depending on which dye/tracer/virus you inject the spread in the ventricles could also be negligible because it would get diluted. of course most important is to know whether the tracer is anterograde or retrograde so that you know where you have to look for cell bodies or fibers you won't be 100% precise anyway, but testing on different animals will give you the right control, by comparison. 

Which tracer are you using? in any case, as a rule of thumb, use low pressure (10-15psi max) and prolonged microinjection time (~10nl/minute), do not inject more than 20-30nl total volume. Keep the pipette in place for ~10minutes after you are done with the injection. this should solve the issue, at least in theory. As a control, microinject the same volume of  tracer in the ventricles and check whether it labels the adjacent areas. Good luck.

Hi Justin, you can avoid the spread in different ways.

1) if you have a pico-injector that allows you to apply a slightly negative pressure, once you lower the pipette you won't lose tracer from the tip. 

2) if you load a tiny drop of some neutral unsoluble solution after your tracer (mineral oil) this, even if really low in quantity, will efficiently block the spread.

3) if your injection is low in volume (an injection every minute, for a volume of 20nL ca) for a total of 1-3 injections and if you wait at least 10-15 minutes after the last before extracting the pipette from the tissue, you should be able to avoid the spread that usually occurs at the end of the injection protocol.

to improve specificity and precision some people even start injecting 5 minutes after lowering the pipette, which in this case is a step that takes 1 minute each 200-300 micrometer (usually I am not that slow but this could allow the tissue to come back in shape, specially if you reach a deep region)

relatively to your problem, depending on which BNST nucleus you are targeting you almost can't avoid the ventricles. best would be to enter from the cortex above the CPu with an angled injection (you can easily re-adjust coordinates with pitagora) which would totally avoid the issue, roughly could be 30 degrees. that's what I do. depending on which dye/tracer/virus you inject the spread in the ventricles could also be negligible because it would get diluted. of course most important is to know whether the tracer is anterograde or retrograde so that you know where you have to look for cell bodies or fibers you won't be 100% precise anyway, but testing on different animals will give you the right control, by comparison. 

A great question.  Perhaps some sort of binding agent can intentionally be injected into the ventricles at the start of the procedure to inactivate/override the fluorescence? It's been awhile for me since I've had the opportunity to work with fluorescent labeling, but it does seem like your concern is valid and that you need a protocol to observe the results when that happens.

Happy persisting and best wishes,

Obviously you must make smaller injections and off from the  limits to the ventricle. The best way to check if the tracer is spread into the ventricles is to inject an animal and perfuse it immediately after the injection. Finally the seccioens will give you the key to whether or not there was contamination, the connections of the area are well known. luck and regards

Miguel

As Livio suggest, I would try to do an angled injection.

You should always wait before the start of the injection and after it. You can also exit just for 50-100µm and wait a few minutes again.

Best wishes!

Regards

hi there,

we occur similar problems when placing tracer injections close to the ventricle. as stated by the other comments, you might "lose" some tracer when pulling out the capillary from the tissue (sometimes pretty obvious when you see tracer signal along the injection canal). we keep the capillary in place for at least a few minutes. after that, applying little negative pressure to "suck off" excess tracer which has not been taken up by the tissue could minimize tracer spread. even if your capillary might immediately clog, you avoid negative pressure which automatically builds up when pulling out the capillary.

good luck!

In my experience the picospritzer injections were  unpredictable quite often, it usually pumps  more tracer than that is needed even in cortical areas of little bigger animal brains eg. ferrets. It is more desirable to use iontophoresis for small injections for very minute areas like the one you described. The  iontophoresis may be best the alternative in case the tracer molecule carries electrical charges, example. Cholera toxin B, Fluororuby, microruby etc. 

Sorry to say that, but all the methods reducing the spread of the tracer may help to reduce the danger to get some tracer into the ventricle, but it does not fully prevent it. The only safe way to distinguish artifacts from real connections is to make injections only into the ventricle and compare the results with those of injections into the nucleus. This does not mean the other suggestions are wrong, all of them help to reduce spread, and combined with a control injection into the ventricle, you might come to a projection pattern which is near to "the truth"

As many suggested it is very tricky and one can not easily delineate spreading into ventricles. I have faced similar problems when I was injecting into arcuate nucleus of mice and hamster. Waiting up to five minutes after injection could help the surrounding tissue to absorb the tracer. As Hans Joachim Bischof suggested you can do a pilot study by injecting into ventricles and see if that bothers your experiment.

Good luck

Excellent suggestions and comments all around.  We are using a Cre recombinase-inducible AAV to selectively infect neurons that have Cre driven by an oxytocin receptor promoter to cause expression of eYFP, which then acts as an anterograde tracer.  The literature is conflicting regarding the distribution of these neurons and thus their downstream effects as well.  Thank you to all for the insight and guidance which will undoubtedly improve the accuracy of our study.

Hi Justin,

Regarding a doubt,  a way to check if a pipette is in the ventricular system is when the pipette is placed in the set point, apply a slightly negative pressure to aspirate cerebrospinal fluid in the case that the pipette is in the ventricle. If the pipette is in the right place not appear cerebrospinal fluid.

My favorite solution is the deliberate miss!  In your case fill the ventricles and then subtract that pattern from your 'successful' injections. As usual it is better to know your enemies than try to avoid them! 'Deliberate' mistakes have taught me many things as I tried to localize tracer injections.

You might want to look into injection by iontophoresis, my understanding is that these injections spread less.

Hi Justin,

We have injected ibotenic acid into the BST. We had the same concern: the problem was that if ibotenic acid entered the ventricular system the animal could suffer epileptic seizures. Thus we decided to use an angle in the anteroposterior direction. We introduced the micropipette caudal with some 30º of rostral inclination. This allowed us to avoid the ventricular system. Just make your own calculations, depending on your exact target within the BST and the strain of mouse you use (we used CD1)

Good luck

Fernando

If your concern is that the Cre-AAV virus may be entering the CSF and then infecting neurones throughout the brain, I would say this is unfounded. Viruses require high titres to produce any viable level of expression and any virus that does enter the CSF will be so dilute that I would be very surprised if this produces ectopic expression. Although I agree with what has been said previously regarding leaving the pipette in place for 5 minutes or so after the injection. If any cells within the ventricular system (e.g. ependymal cells) are being labelled, these should be clearly visible and distinct from true BNST cell labelling

Please leave the micropippete in place for 20 min before removal. Additionally and depending the tracer you are using, perhaps you might consider to use iontophoresis delivery instead of picosprtizer.

Hi all,

I know this is a very old thread but still. Is anybody familiar with a good neutral mineral oil for the delivery system. Could I use the mineral oil that Sigma has for mouse cell embryo cultures?

http://www.sigmaaldrich.com/catalog/product/sigma/m8410?lang=en®ion=CH&gclid=CjwKEAjwoLfHBRD_jLW93remyAQSJABIygGpmLkjmok9QLuYy2GwWRGxsptFdT81DNQwCM_EcESgTxoCzlDw_wcB