Bad bugs: good for gene therapy?
2 December 2002 17:00 GMT
by Clyde Burnham
 |
Transmission electron
micrograph of
Escherichia coli
O157:H7, CDC/Elizabeth
H. White, M.S. |
A transatlantic research team has transformed the notorious
bacterium Escherichia coli into a promising therapeutic
agent in the fight against cancer. Darren Higgins, a
microbiologist at Harvard Medical School, and colleagues in
England, used the novel method to prevent the growth of melanoma
cells in mice.
The researchers engineered a recombinant strain of E. coli
to express the proteins listeriolysin and ovalbumin. This
recombinant bacterium could then induce dendritic cells to
efficiently present ovalbumin to cytotoxic T cells in mice.
Cytotoxic T cells, the mainstay of so-called cell-mediated
immunity, activated in this way then prevent the growth of
ovalbumin-labeled melanoma cells in mice.
Cancers succeed by evading cell-mediated immune attack, so
Higgins' approach provided a way of preventing that evasion - by
efficiently presenting tumor antigen and activating cytotoxic T
cells to kill tumor cells. Antigens presented to dendritic cells
have to trigger maturation of the dendritic cell, otherwise the
result is tolerance of, rather than immune attack on, the antigen.
Higgins and colleagues used E. coli which, as with all
bacteria, drive dendritic cells to mature. The matured dendritic
cells then present the E. coli antigens to cytotoxic T
cells. This activates the cytotoxic T cells, which will then
attack and lyse cells with those antigens.
Higgins selected a mouse melanoma strain labeled with an
ovalbumin marker antigen as their target cell. He and his
colleagues engineered E. coli to express ovalbumin and
exposed mouse dendritic cells to the recombinant bacteria. The
dendritic cells phagocytosed the E. coli and were driven
into maturation. The matured dendritic cells then presented the
ovalbumin protein to cytotoxic T cells. The cytotoxic T cells thus
produced were only slightly more toxic to the ovalbumin-labeled
melanoma than were naive cytotoxic T cells.
However, when Higgins engineered E. coli to express
ovalbumin and listeriolysin, the outcome changed. Listeriolysin is
a pore-forming cytolysin from the bacteria Listeria
monocytogenes. When E. coli cells containing ovalbumin
and listeriolysin were phagocytosed by dendritic cells the
phagosome was perforated by the listeriolysin, and ovalbumin
flooded the dendritic cell's cytosol. This presentation of antigen
strongly activated cytotoxic T cells against the ovalbumin.
Listeriolysin thus appears to moderate the efficient presentation
of ovalbumin antigen to cytotoxic T cells.
Mice vaccinated with dendritic cells exposed to ovalbumin and
listeriolysin-labeled E. coli mounted a strong immune
response when challenged with ovalbumin-labeled melanoma cells.
Mice vaccinated with dendritic cells exposed to an E. coli-ovalbumin
recombinant without listeriolysin mounted only a minimal immune
response to melanoma cells expressing ovalbumin. The researchers
then vaccinated the mice with the E. coli-ovalbumin-listeriolysin
recombinant without the dendritic cells and saw a stronger immune
response. When they vaccinated mice with paraldehyde-fixed E.
coli-ovalbumin-listeriolysin they saw an even stronger
cell-mediated response against the ovalbumin-labeled melanoma.
William Robinson, professor of medicine and a cutaneous
oncologist at the University of Colorado Health Sciences Center,
finds the use of listeriolysin to stimulate cell-mediated toxicity
an interesting concept. He notes, however, that human melanoma
cells are rapidly mutating and have only minimally antigenic
surface markers. He thinks it a long step from attacking the very
antigenic ovalbumin to the slightly antigenic cells of human
cancer, which are so heterogeneous and unstable.
Ralph Reisfeld, a microbiologist at The Scripps Institute, says
that using listeriolysin to enhance presentation of antigen to
cytotoxic T cells is a welcome advance. He warns that there are
many complications between this interesting work in mice and the
creation of a vaccine effective against any human cancer. The use
of bacteria as a vector is one of his concerns, as the risk of
pathogenicity, either acute or late, exists with any bacterium.
The bacteria chosen for a vaccine must be maximally innocuous. He
also points out that ovalbumin, the marker antigen used by
Higgins, is, in mice, a xeno-antigen. Any human tumor antigen
chosen for a vaccine in humans will not be a xeno-antigen and
additionally must not provoke immune attack on normal cells.
Ronald Levy, professor of medicine at Stanford School of
Medicine, agrees that that this paper offers some interesting
possibilities for creating human cancer vaccines. He expressed
optimism that tumor markers for at least some cancers would prove
to be useful vaccine targets.
Higgins, whose primary interest is in bacterial and viral
infection not oncology, noted in earlier work that listeriolysin
forced macrophages into a much stronger presentation of antigen to
cytotoxic T cells. When discussing this finding with his English
co-authors they resolved to use the approach against cancer.
While gratified by this proof-of-concept work, Higgins says his
group is planning significant follow-up studies using
listeriolysin-based vaccines against bacteria, viruses, and
cancers. His group has experience with E. coli, and hence
used that bacterium for proof of concept. Choosing the bacterial
vector for human vaccines is a major issue, which must be
determined not by ease of use but rather safety, he says. He notes
that choice of the appropriate target antigen for each variety of
cancer is another area that will require much work.
