In article <3q1v0m$m3d at jhunix1.hcf.jhu.edu>, ejf at welchlink.uoregon.edu
(Ephraim Fuchs) wrote:
> Dr. Kirberg,
please, just say jorg as I said at the end of my mail; I still havent got
my PhD. ((hopefully) I get it within the next months)
>> Since you are at Basel, you may wish to talk to Franca Ronchese (if she
> is still there; I do not know).
unfortionally she left already from Basel, and before, I did not talk
too much with her. Maybe I should have better done. So for now we should
discuss about the published data of her only (J Exp Med 177(1993):679).
>She did a neat experiment where she
> adoptively transferred F1 (b x d) T cells and B cells into scid (H-2d)
> mice, followed by priming with soluble antigen. While this priming was
> accompanied by vigorous antibody production, the primed T cells were
> restricted only to H-2d and not to H-2b. Obviously, because the F1 B
> cells were making antibody, they were activated and interacting with T
> cells. Yet no H-2b restricted helper T cells were activated.
Up to here its correct.(with one important exception, see later)
>H-2b
> restricted T cells were generated if the adoptive transfer also included
> F1 professional APC.
This is not in that paper. H-2b restricted cells were detected when she
transferred primed T and primed B cells. (As data not shown she says
that H-2b restricted cells were detected as well following transfer
of primed T and naive B cells.) Anyhow, with my view, such a result would
not be unexpected.
So let me show how I would interpret this story: I think this paper
shows very nicely that professional APCs are much better in activating/
priming of naive T cells. However, I do not think these experiments
can finally show that B cells are unable to prime naive T cells. In
the scid host, all the professional APCs are H-2d, therefore H-2d
restricted cells will be favored tremendously, precluding detection
of H-2b restricted cells that can only be activated on B cells in this system
(were either naive or primed B cells gave the same result).
[One should remember here, that other problems were adressed in similar
way; diluting T cells enormously gave the (wrong !) impression that
antigen is required for keeping memory (Gray and Matzinger
Exp Med 167, 805 (1988))]
But lets continue with Francas paper:
Because of your posting I had to read her paper again. Nicely, in the last
sentences she cites some data which she did not show, but that actually
give strong support for my notion. Allow me to cite this part:
"....Evidence that such a mechanism may indeed take place in vivo was
obtained in experiments in which SCID mice were reconstituted with high
density [that are prediminantly naive cells] splenic T+B cells from
nonimmunized mice, and found to be able to generate low numbers of H-2b-
restricted T cells after Ag immunization (not shown). ..."
So there were some few H-2b restricted cells ! To be fair, one has to note
that this sentence came in the context of crossreactive T cells. However,
if there are few H-2b restricted cells, the notion of B cells not priming
T cells in this case might simply due to the very low frequency of such T
cells in this assay system, that might frequently not be detectable.
>So it appears that activated B cells are unable to
> prime the precursors of CD4+ T cells that give help in antibody
> production.
There is still no proof for this at all. And in the experiment of
Fuchs & Matzinger Science 258, 1156-9 (1992), one might still argue
that this is due to lack of help (as I did in my previous posting).
I wonder, if you ever made the same experiment using male (B6xbm12)F1
or male (BALB/bxB6)F1 B cells. I assume that these cells will prime a
male specific CTL response.
>> Maybe we are agreeing, and this is just semantics. However, the bottom
> line, I believe is that no B cell, resting or activated, can prime naive
> T cells, but they can reactivate memory T cells.
well, its somehow semantics if the fianal outcome is tolerance. To put
it more simply, B cells as any other cell will activate CD8+ T cells.
However, the circumstances decide the final outcome: e.g., absence of help
will exhaust (or what ever you call it) these T cells, resulting in deletion.
>> By the way, I don't trust transgenic T cells.
Transgenic mice are the only way to proof that point (since recently
maybe not the only, but the easier way, McHeyzer-Williams & Davis Science
268, 106-11 (1995))
But I agree, that there are some risks with these mice. It is wrong to
assume that TCR transgenic T cells are naive, as has been done frequently.
E.g., the Schmitt-Verhulst mice are not naive, as T cells label with
BrdU extensively (I dont remember the reference now, I think it was
Intl Immunol, relatively recently). In most other (but one, guess witch !),
TCR transgenic mice, BrdU studies have not been done.
Swain, Immunity 1, 543-52 (1994) even showed expansion after transfer. How
she cound convince referees that the transgenic cells were naive is a
myrical to me.
>They have more than one
> species of antigen-specific receptor on their surface and so can acquire
> memory from a pairing of endogenous alpha chain and transgenic beta chain.
Some mice seem to have less problems with this than others. If you
show that your cells never divide without giving the antigen, I think that
one should feel confident that the cells are naive.
Anyhow, there is the possibility to breed the mice to RAG-/-. Apparently,
you must stop arguing than ...
Anyhow, saying "I don't trust transgenic T cells" will not solve any
problems.
>> The reason that all this is important to the original discussion is that
> if B cells fail to activate naive T cells, then the immune system CANNOT
> discriminate self from nonself because any antigen, self or foreign,
> presented exclusively by B cells will tolerize naive T cells specific for it.
Any antigen will activate a T cell regartless where its presented. Some cells
moght be better with this than others. [this is rather ment to be
provocative; maybe not all cells, but I include B cells to those that
do activate naive T cells]. I agree
that than there is a need to explain how self/nonself discrimination
is achieved. Up to know, lack of help, deviation (Th1/Th2), exhaustion (some
times overlapping to lack of help?), and ignorance (niches ?) have been
documented or are discusses. Making everything more complicated with
an alarm factor is simply not required.
> Thus one must look for a new paradigm of immune function.
?
