Revisiting data from new angles

I’m off to Glasgow in a few days to attend the 12th World Congress on Pain held by the International Association for the Study of Pain. In between preparing my talk and getting my poster ready I’ve been poking around the abstracts to get my itinerary in order for the meeting. I’m not even there yet but I’m noticing a big trend…

A good number of abstracts show that many high profile researchers are revisiting some of their older findings (or findings of others) using some of the fancy new genetic tools that are available for use today. In particular, I’ve noted that there are a large number of abstracts using transgenics (mostly expressing GFP fused receptors or other fluorescent markers expressed in a certain gene locus) to take another look at anatomical findings that previously depended on antibody staining (commonly called immunohistochemistry or IHC). Perhaps not surprisingly, the findings discussed in the abstract are at odds with what we thought we knew from the IHC data. I find this to be simultaneously refreshing and disturbing. I use IHC fairly regularly and I have always found the technique to be useful but fraught with very serious problems. When I use IHC I always try to combine it with in situ hybridization (a method for determining the cellular expression of mRNA) in order to confirm that, at the very least, the mRNA and protein are in the same cells. While this gives you some confidence in your results it is by no means an indication that your antibody staining is specific. Moreover, when looking at fiber tracts or other anatomical structures this combination is often not useful because the mRNA may or may not make it into the fiber (e.g. the mRNA may be restricted to the cell body of the neuron). Another problem is epitope preabsorption. This is a very common control where the epitope (a blocking peptide) that the antibody is made to recognize is used to “preabsorb” antibody binding. Many researchers use this as a method of specificity; however, it is actually just a test to tell you that the antibody binds what it is made to bind. If your antibody staining is preabsorbed by an excess of blocking peptide it does not exclude the possibility that the antibody has non-specific interactions that are unrelated to the preabsorption (which is generally high affinity). So, it is not surprising to me that genetic methods would yield quite different results. What is disturbing is when those results are completely at odds with what we thought was well established (as I’m seeing in a few of these abstracts). It will be fascinating to see what the talk is like around these posters at the meeting. I, for one, am not overly confident that the GFP-fusion approach does not alter expression patterns in and of itself. On the other hand, at least one of these posters lists a number of excellent controls all of which appear to indicate that the previous IHC data were indeed inaccurate. We’ll see…

Another big controversy in the field has been rewiring of the dorsal horn of the spinal cord after a peripheral nerve injury. This process was originally proposed as a mechanism for allodynia after peripheral nerve injury (the first data came from Clifford Woolf’s work). This is likely the idea that Physioprof was referring to in his comment (first comment on the page) on my post a few days ago on the work from John Wood’s lab. The idea was that light touch sensitive fibers moved into the dorsal horn lamina I and II, where the pain processing neurons reside, from there original location somewhere in lamina III. In the decades since this original discovery several other groups found that the tracing methods used in the original paper had some technical problems that only arose after the nerve injury. On the other hand, several other groups used electrophysiology in nerve injured animals to reach the conclusion that while the anatomical tracings may not have been accurate (again because of a technical issue that was unknown at the time) the electrophys data still supported the original conclusion. Since I’ve been in the field this particular theory has remained quite contentious with groups firmly entrenched on either side. There will also be a poster on this topic at the IASP meeting where the authors have used a transgenic approach to trace fiber projections after nerve injury. The abstract states that the original theory holds, at least in these mice, and that light touch fibers do indeed sprout into the outer lamina of the dorsal horn after a peripheral nerve injury. Again, I can’t wait to hear the debate that will surely occur around this particular poster. Should be a grand time!!

So, the controversy is fun and all but the progress is what really gets me excited. I can’t wait to get to Glasgow, looks like there will be plenty of steps ahead to get all hot and bothered about!


7 responses to “Revisiting data from new angles

  1. I’m going to point people to this post when they show me immunostaining results that totally contradict in situ results. Cause damnit it’s been happening a lot lately, and frankly I’m sick of it.

    “But we have the preabsorption control” Sorry, that’s not enough.

    One of the better things I’ve seen regarding this was Charles Bourque’s use of the V1 KO animals to show that the antibody staining he saw in the supraoptic nucleus was really V1 signal. It was gone in the KO, under the same staining conditions. That’s about the only V1 in the brain that I really believe, even though it’s still an incomplete story (how do you get a capsaicin insensitive, osmotically regulated V1?).

    Sounds like the meeting will be very interesting. A little contentiousness should prove exciting. Enjoy it!

  2. That’s about the only V1 in the brain that I really believe, even though it’s still an incomplete story (how do you get a capsaicin insensitive, osmotically regulated V1?).
    Alternative splicing if I remember correctly. Agreed, that is an interesting story but with some considerable holes left to fill…

  3. It seems very likely to be alternative splicing, but the N-term splice that controls this hasn’t been determined. Because not only did they loose capsaicin activation, they lost Ab staining directed against N-term epitopes.

    One thing I like about the whole story is that it’s from a guy who has spent much of his time focused on this particular brain area. Very likely this new aspect of TRPV1 wouldn’t have been discovered (well, it would have, but it probably would have taken a heckuva lot longer) had it not been for someone with his knowledge gained by years of experience.

  4. I work in sub-unit vaccine development, and have a desired to track to sub-cellular localisation of recombinant proteins around the cell.

    We obviously favour native protein structures, but often require a tag for ease of staining. We have recently moved many of our constructs to the FlAsh/ReAsh system. All this requires is a 6 amino terminal fusion, which is far less bulky than conventional flurophores. Once bound to the poly cysteines of the tag, these compounds fluoresces in readily detectable spectra via confocal. Until they bind they are not excitable.

    Using this system should minimise tag-related variation from wildtype, and works well for us.

    Griffin et al. Specific covalent labeling of recombinant protein molecules inside live cells. Science (1998) vol. 281 (5374) pp. 269-72

  5. Thanks for the info Rob! I didn’t know about that particular technique and it looks like it could be quite useful for some applications we are interested in.

  6. This is one of the fundamental problems with indirect detection methods like immunodetection. Antibodies are molecules, and they bind to other molecules, but everybody forgets how many other molecules are around in their preps, how many other VERY SIMILAR MOLECULES are around in cells, and about how much signal amplification can affect WHICH molecules of that mix you are seeing evidence for.

    They are made by a process that is extremely dependent on the quality of the molecule you use to generate them, and SO few of those are made carefully. Even if they are made carefully, there are ALWAYS impurities–particularly from typical peptide synthesis. It really bothers me how many assumptions get made on the basis of an antibody generated against a peptide that was less than 99% pure. And even when peptides are greater than 99% pure, there is just so little validation done to make sure that the immunoreactivity of the antibody really has anything to do with the primary component of the mixture. That amplification mechanism that gives antibodies their desirable sensitivity just totally obliterates any REAL confidence I think people should have in what the immunostaining tells them.

    No understanding and thinking about how methods/processes work and the mechanisms behind data generation really chaps my hide! I picture many a future grad student lamenting the day they asked me to be on their committee.

  7. Pingback: Recent Links Tagged With "angles" - JabberTags

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s