Lessons from Wounded Flies
Scabs aren't pretty, but they're useful. They stop blood from
flowing, a function called hemostasis. Scabs also
inaugurate the critical process of wound healing—a process
that, even when successful, leaves inelastic scar tissue in place of
supple skin. But it doesn't have to be this way—when embryos
are injured, they heal with barely a trace. So what distinguishes
perfect healing in embryos from the successful healing exhibited by
a run-of-the-mill cut from the difficult or nonexistent healing that
accompanies some deep burns and diabetic ulcers?
A new model developed by Michael Galko, a postdoctoral fellow in the
lab of biochemist Mark Krasnow at Stanford, may expedite answers.
Writing in the journal Public Library of Science Biology
(August 2004), he describes a puncture-wound assay in
Drosophila larvae that shares important cellular and
molecular features with vertebrate healing. Even the authors were
surprised that the broad outlines were so similar, given the
differences in tissue architecture between fruit flies and
vertebrate animals. The work has attracted enthusiastic support
among colleagues. According to Paul Martin of the University of
Bristol in the U.K., a leader in the field and author of more than
30 papers on the science of wound healing, such similarities in a
"genetically tractable" organism provide a powerful tool:
"It's got scabs, migratory cells and genetics to boot."
Part of the excitement comes from the potential to pry apart the
cell and molecular biology behind a notoriously intricate
physiological process—"It's bloody complicated,"
Martin quips. Scientists already know of a dozen or so cell types
and several times that number of molecular role players in wound
healing. The complexity has limited the usefulness of mouse models.
Mutations of some genes relating to this process have been lethal in
mice; others are difficult to interpret because of genetic
compensation or redundancy. Consequently, many crucial details about
the sequence and independence of events remain unknown. Which steps
in healing are serial? Which take place in parallel?
A bare summary of wound healing in mammals goes like this: As blood
clots, the clot traps platelets and cells in a net of connective
fibers. The platelets attract immune cells to attack bacteria and
scavenge debris; keratinocytes (skin cells) creep toward the damage,
using the scab as scaffolding. Finally, fibroblasts (other skin
cells) pay out collagen fibers and provide the muscle (literally) to
tug the wound closed as migrating cells cover the site and re-form a
covering of layered skin. The puncture model in Drosophila
larvae follows the same general script, albeit with a few twists.
In Galko's experiment, stabbing the anesthetized larvae with a
minuscule needle only six cells wide caused the loss of some
hemolymph (blood), but the wound quickly plugged with a clot of
debris that became the scab. Within 30 minutes, epidermal cells at
the edge of the wound re-oriented and began to fuse, creating a
syncytium, or cell with many nuclei, around the puncture.
At the same time, the edges of the syncytium spread along and
through the plug. The creeping cell border reestablished epithelial
continuity within eight hours, after which the cluster secreted a
new outer cuticle. Residual debris was engulfed and degraded. Thus,
scab formation, epithelial activation and migration, and a clean-up
phase are all evolutionarily conserved stages in the wound healing
process—stages shared by the simple fly and the complex primate.
Previous studies showed that dorsal closure during embryonic
development (a proxy for the knitting together of wound edges)
requires a signaling molecule called Jun N-terminal kinase, or JNK.
Without JNK, the dorsal cell sheet never properly migrates, and an
embryo fails to develop. Galko found the same requirement in his
puncture model, but he was able to extend the observation. To find
out when and where JNK is present, he used a type of so-called
"reporter" fly that makes the enzyme β-galactosidase
(creating a blue color) in cells with JNK. Against this background,
Galko engineered specific mutations to block JNK, but only in the
larval epidermis. Thus, he was able to satisfy the need for JNK in
development but disrupt it later, in the desired tissue at a
specific time. The resulting flies were normal to all appearances
but had a dramatically different response to the puncture test.
Although scab formation, cell reorientation and syncytium formation
were normal, the cells were unable to reform a continuous epithelium
and could not close the wound.
Galko next analyzed the role of the scab itself using mutants that
fail to form crystal cells, which he found to be necessary for scab
formation. These larvae only muster a diffuse plug at the wound
site, and 85 percent of them die within 24 hours of the procedure
(as opposed to 15 percent of normal larvae). Even when sharper pins
were used to create the wound (a change that minimized mortality),
the area of cellular damage was greater than normal, and the
remaining cells showed slackened, wavy borders. In these flies, the
levels of JNK activation were higher and more widespread, suggesting
that the scab might normally inhibit the activation of the kinase.
However, the same mutant healed perfectly well when pinched instead
of punctured. This type of injury breaks the epidermis but leaves
the overlying cuticle intact, so it doesn't bleed or form a scab.
The normal cell orientation, fusion, migration and JNK activation in
the absence of crystal cells indicate that the scab's primary
function may be to stabilize the wound site so that a wave of
epithelial cells can infiltrate it. Taken together, the
JNK-inhibited and scab-deficient mutants suggest that the components
of wound healing—scab formation, epithelial-cell activation,
epithelial-sheet spreading and clean-up—are more independent
of one another than scientists believed.
The fly model isn't perfect. No one sees syncytia in vertebrate
epithelium, for one thing, and insect wounds don't undergo the same
burst of cellular proliferation and differentiation that mammalian
wounds do. But the lack of both inflammation and an adaptive immune
system in Drosophila may be the most profound differences.
According to Paul Martin, "If you're interested in the
leukocyte [white blood cell] response you're stuffed." He
includes himself in that flip assessment: Martin's mouse model
suggests inflammation may be the key difference between scars in
adults and perfect healing in embryos. But Michael Galko is already
working on the next breakthrough from his model. Unlike a mouse
geneticist, he can easily screen whole-genome libraries of randomly
mutated flies to find other mutants—hence other genes and
proteins—that affect healing. And it's open ground, too.
"Nobody had studied wound healing in a genetic model organism
before," says Galko. He's poised to make some important
strides. As Martin, who is cranking up a zebrafish model of his own,
states, "To get good therapeutic targets you really need a
genetic model. You'd be daft to try this in the mouse."