In article <3579tn$7k4 at nntp.Stanford.EDU> ladasky at leland.Stanford.EDU
(John Ladasky) writes:
>> I just finished a presentation of the paper "Cytotoxic T cell
>memory without antigen" (Lau LL et. al., Nature 369:348, 23 June 1994)
>for a face-to-face journal club here at Stanford. The authors claim
>that they have demonstrated the existence of a memory T cell population
k>specific for lymphocytic choriomeningitis virus (LCMV) that can be
>adoptively transfered from one host to another. Previous studies have
Great choice for a JC paper, it covers a number of very interesting issues.
I thought that in case anyone is only peripherally interested in this, I would
try to post a little of the background and context for this paper. (Also, I
haven't yet re-read the papers - the background is as far as I've got.)
Be warned - this is not specifically my field; I'm merely an interested
bystander.
The question that's being addressed in this paper (and it's Nature neighbor) is
the nature of immunological memory. As we all know, exposure to an antigen
induces a number of acute-type responses (acute meaning days to a few weeks
after the initial exposure) including some of the obvious things such as anti-
body production/isotype switch, CTL lysis, etc. On subsequent exposure to the
same antigen, the immune response is much more rapid and much greater. "Sub-
sequent exposure" can mean very subsequent - i.e. years later. This is due to
immunological memory and happens because one of the things the acutely reacting
cells do is to form a population of memory cells.
When a memory response occurs months or years after an initial exposure
to antigen, is this because the same memory T cells are still around, persist-
ing from their intial exposure? - or is it because there has been an intermit-
tent, or continual, trickle of antigen in the intervening period? I had
assumed the former, without a very good reason for it. In the wild, of course,
the experiment is pretty tough to do right: are you immune to measles today
because you had that case of it twenty-five years ago when you were two, or
because you have been exposed to your neighbor's snotty-nosed brats shedding
their measles over you at intervals in the meantime? The problem is also more
complicated than that: even if there is no overt antigen hanging around, there
could be a very long-lived antigen depot - follicular dendritic cells are
frequently mentioned as possible depots. Although I feel uncomfortable
with the idea of cell-surface antigen persisting for months or years, this is
really a matter for my therapist and myself and should not influence the
hypothesis.
The paper that brought this to my attention (it was not the first to
discuss it) was from Zinkernagel's lab: Oehen et al, J Exp Med 176:1273-1281
(1992). They used lymphocytic choriomeningitis virus in mice, a paradigm of
a persistent virus, which induces a very long-lived memory response, and
asked what happened when the continual antigenic reminder - the LCMV - was
removed from the system. They addressed this first by using a different
immunogen - LCMV proteins in a vaccinia vector. Vaccinia doesn't replicate
well in mice and doesn't persist. They found that the vectors induced a short-
lived response, in which no memory was detectable after 30 days. Higher doses
of the vax vector gave longer lived responses (>60 d). However, by doing
adoptive transfer of immune spleen cells to naive recipients, they presumably
got rid of persistent antigen, and they found that the memory response in
the recipients was short lived. They argued that memory cells were an artifact
which were actually continually reactivated cells, constantly meeting new
antigen (which is at least partially consistent with the phenotype of mem-
ory cells: they show a number of activation-type markers.)
Arno Mullbacher (J Exp Med 179:317 - 321 (1994)) begged to differ.
Mice vaccinated with flu were used as donors into irradiated naive mice (I am
simplifying the experiments here, a little bit; he had more controls) and
he looked at the kinetics of the cytolytic response after infection of the
recipients; he argues that lysis 3 days after infection was diagnostic of
a memory reponse (which seems reasonable) and found that by this criterion
a memory response was detectable as much as 25 weeks after transfer - 42
weeks after the initial infection. However, Mullbacher was not really
anal about excluding possible persistent antigen from his transfers; either
(a very low level of) virus could have still been present (not likely,
but possible), or (more plausible) he could have transferred some of the
hypothetical antigen-depot cells along with his splenocytes: he selected
against MHC II expression, but who knows what these hypothetical cells
might have on their hypothetical surfaces.
The papers under discussion arise out of this. There are a couple of
other relevant papers people might be interested in glancing at - one
which I remember seeing but can't turn up - perhaps that issue of J Virol (if
that is where I saw it) is out for binding at the nearest library. This
was a study done on persistence of the immune response to vaccinia. If I
remember it right (and I might not) the population was one which had been
vaccinated for smallpox many years before, and then relatively isolated
from the rest of the world. Of course they hadn't been exposed to smallpox,
and it was unlikely that they had been exposed to vaccinia; but they had
detectable memory repsponses as much as 25 years later (This is all from
memory, and I didn't do much more than skim the article; if someone knows
better, please correct me) - anyway, if so, it argues that a long-lived
memory response doesn't require a continuous source of antigen; it doesn't
prove that antigen depots are not involved, but I really, really feel un-
comfortable with a cell presenting the same antigen for 25 years.
There is also an excellent review by Sprent and Tough: Science
265:1395-1399 (1994) [the Sept. 2 issue]: _Lymphocyte Life-Span and
Memory_ , which nicely discusses the issues in lymphocyte maturation and
lifespan; highly recommended.
Hope some others get involved in this discussion. When I've had a chance to
go through the papers more carefully, in the next couple days, I will try
to discuss John's specific questions.
