In article <3op54r$h1g at jhunix1.hcf.jhu.edu>,
RIMA JEAN COUZI <rcouzi at welchlink.uoregon.edu> wrote:
>>Ken,
>>Regarding your question as to why responses to viruses in the skin don't
>evoke anti-skin autoimmunity:
>>It is true, and Polly acknowledges, that a lytic viral infection of
>keratinocytes may indeed release skin-specific antigens which are then
>captured by Langerhans' cells that have also received a specific danger
>signal. These cells will travel back to the lymph node where they will
>act as immunogenic antigen presenting cells to elicit an immune response
>against skin-specific antigens. Consider, however, what happens
>subsequently. Activated effector anti-skin T cells will go out into the
>skin, kill some uninfected skin cells, and then rest down to become
>memory anti-skin T cells. It is true that the CTL-mediated destruction
>of skin cells may be a source of more antigen to be picked up by
>macrophages, but since the skin cells are now dying an apoptotic cell
>death the macrophages do not upregulate costimulatory signals. In fact,
>it is possible that cells dying by apoptosis may release substances
>(?interleukin 1-B?) that specifically tell phagocytic cells NOT to
>activate T cells. Eventually, the memory anti-skin T cells would
>circulate out to the skin, see the skin-specific antigens presented by
>the keratinocytes themselves, and become tolerized. This last idea is
>Polly's; although an attractive hypothesis, I am not absolutely sure that
>the only cell types that can reactivate memory T cells are professional
>APCs and B cells.
>>Ephraim Fuchs
>ejf at welchlink.welch.jhu.edu
But is it that simple? Macrophages and DC's constitutively express B7-2
(according to today's seminar by Gordon Freeman), which is very definitely a
costimulatory molecule. Further, skin cells die every day by necrosis as
well as by apoptosis (abrasions, bruises, other daily insults), and these
should be perfectly capable of providing "skin antigens", even if apoptotic
cells are not (is that known, or just proposed?). Further, unless
keratinocytes express MHC Class II (and I don't think they do), then CD4+
T cells should not be tolerized. Since many autoimmune disorders appear to be
mediated by CD4+ T cells (diabetes, RA, etc.), it seems critical that these
cells be tolerized against "self-antigens" (or "non-danger" antigens). A
general inflammatory response (Th1 mediated) or antibody response (Th2
mediated) can be very dangerous to the organism, even if all of all of the
anti-self CD8+ cells have been put out of commission
I am not trying to shoot down the model - I think it extremely elegant, and
cleverly addresses some of the holes in the "self/non-self" model. I was
just wondering if you have a mechanism for tolerizing CD4+'s, since they
appear to be the CPU for the immune system. As a relative newcomer to
immunology (only since I joined Nilabh Shastri's lab 2-1/2 yrs ago), I have
also been troubled by some of these nagging questions about tolerance, and
my own work on Class II antigen processing leads me naturally to wonder about
autoimmunity and the definitions of "self" and "non-self". I'm really glad
that Dr. Matzinger posted the article - it is nice to read something on this
group other than requests for antibodies.
BioKen
--
Ken Frauwirth (MiSTie #33025) _ _
frauwirt at mendel.berkeley.edu |_) * |/ (_ |\ |
Dept. of Molec. & Cell Bio. |_) | () |\ (_ | \|
Univ. of Cal., Berkeley "Yes, we have second bananas" - Torgo the White
