New Vessels Take Direction from Vascular Cell Signals

Endothelial Cell Enzymes Can Stimulate Either Growth or Regression

In 1971, when Judah Folkman, the Julia Dyckman Andrus professor of pediatric surgery at Children’s Hospital Boston, first published evidence in The New England Journal of Medicine that angiogenesis underlies tumor growth, many scientists eyed his claims with skepticism. But by 1994, the Angiogenesis Foundation had labeled angiogenesis a “disease common denominator” because both excessive and insufficient blood vessel growth have been found to contribute to diverse diseases, including cancer, macular degeneration, rheumatoid arthritis, and coronary artery disease.

Photo by Graham Ramsay

Eunok Im (left) and Andrius Kazlauskas use the eye as a model for studying angio-genesis because the growth of new blood vessels in the retina, cornea, and other vascular beds is well understood. “We can watch vessels grow and retract in an ordered and predictable way,” said Kazlauskas, “something that is much harder to do in tumors.”

Today, several FDA-approved drug therapies inhibit or promote blood vessel growth, predominantly by exploiting the extensively studied angiogenic growth factors outside endothelial cells. “What is less known are the signaling pathways within the vascular endothelial cell that can govern angiogenic behavior,” said William Li, medical director of the Angiogenesis Foundation. Andrius Kazlauskas, HMS associate professor of ophthalmology at the Schepens Eye Research Institute, delves into this uncharted territory in a new study reported in the April 20 online EMBO Journal and in print on May 17.

He and first author Eunok Im, HMS research fellow at Schepens, describe an intracellular signaling mechanism controlling in vitro vascular tube formation. They show that two enzymes vie for the same membrane lipid as they work to control the opposing activities of tube formation and regression. A tip in the balance between these competing reactions in one direction results in tube formation; in the other, regression of newly formed vessels. “This work is important because it reveals, for the first time, a control switchboard within the endothelial cell for angiogenesis, ” said Li.

This switchboard is of extremely high interest to Kazlauskas, who specializes in eye diseases such as proliferative diabetic retinopathy and wet age-related macular degeneration. Both cause vision loss and result from a pathological overproduction of blood vessels. And for drug developers concentrating on angiogenesis, this study presents intriguing new molecular targets to investigate.

Enzymes Going at It
Normally, blood vessels are stable and endothelial cells dormant. In response to injury, however, a variety of factors activate endothelial cells, and each cell absorbs many signals that direct a collective rearrangement into more or fewer vessels. Kazlauskas and Im want to understand how these signals coalesce in a cell to yield appropriate angiogenesis. “The endothelial cell integrates lots of information,” said Kazlauskas. “We want to study the [cell’s] signaling enzymes in order to learn how they control the cell, and then learn how the environment controls the signaling.”

Kazlauskas and Im investigated phosphoinositide 3 kinase (PI3K) and phospholipase C-gamma (PLC-gamma), two enzymes identified by previous research to be linked to angiogenesis. Im tested them in the presence of growth factors in vitro to determine their specific roles in angiogenesis. She compared cells expressing the wild type vascular endothelial growth factor receptor2 (VEGFR2), which naturally activates PLC-gamma, with cells expressing a genetically modified form of VEGFR2 that does not. The wild type VEGFR2-expressing cells formed blood vessel tubes that regressed spontaneously. Yet tubes assembled from cells expressing modified VEGFR2 did not regress, providing evidence linking PLC-gamma activation with tube regression.

Pathological signals. In the case of diabetic retinopathy, blood vessels in the eye collapse, creating a hypoxic environment that triggers angiogenesis. When vessels are destabilized, it appears that the signaling of endothelial cells (red rectangles) may determine whether the vessels grow or regress. A better understanding of the process may explain the pathological overgrowth of vessels that impairs vision.

Image adapted by Rachel Eastwood from original courtesy of Andrius Kazlauskus

Im then examined the interactions of PI3K and PLC-gamma in an experiment involving four receptor mutants of another growth factor, each of the four activating one combination of the two enzymes in on or off states. The mutants unable to activate PI3K failed to form tubes at all, indicating that tube formation requires PI3K activity. The mutants unable to activate PLC-gamma but able to activate PI3K formed tubes that did not regress. A time-lapse analysis of the progression of angiogenesis in the permutation activating both PI3K and PLC-gamma showed that tubes first formed then regressed, confirming that PLC-gamma does not prevent tube formation. Rather, it induces regression after tube construction.

Kazlauskas and Im found that both PI3K and PLC-gamma compete for the same lipid substrate, PtdIns-4,5-P2 and hypothesized that PI3K initially promotes tube creation, then PLC-gamma drains the lipid supply and induces regression. But if both enzymes become active right away, the researchers wondered, why doesn’t PLC-gamma interfere with tube formation in the first place?

They found evidence that dynamic signaling sways the angiogenesis program’s balance over time. Though both tube formation and stabilization require PI3K activity, stabilization requires more activity later in the process than during early tube formation. If PLC-gamma has already depleted the lipid supply by the time the tubes are ready for stabilization, the tubes regress instead.

Parsing Signals
Kazlauskas and Im speculate that VEGFR2 signaling integrates, in part, the many environmental cues that instruct a vessel to stabilize or regress. “You need the right quantity and quality of VEGF-dependent signaling to drive nice angiogenesis,” said Im. They suspect that with further investigation they will find VEGF signaling dynamics that are, said Kazlauskas, “more complicated than an on–off switch, and more elegant.”

They are testing this theory now, along with exploring the signaling dynamics of regression. “Some think the [endothelial] cells die and that causes regression. But we think the tubes regress first, then the endothelial cells have nothing to do, so they die,” said Im.

Kazlauskas added, “When [endothelial cells] are in the tube, they generate information from being in that tube, from touching other cells, from the extracellular matrix, from being stretched out as opposed to balled up. All of these factors provide information that signals the cell not to die. Taking the cell out of the tube may be just enough to cause it to die.”

While Im and Kazlauskas strive to understand the big picture, this study’s specific results may have more immediate implications for angiogenic drug development. Many of today’s anti-angiogenesis drugs combat tumor growth by suppressing new blood vessel formation. But pre-existing vessels may not be affected, so drug developers are looking for ways to dismantle and regress them. “The switching mechanism described in this paper raises the intriguing possibility,” said Li, “that targeted drugs could be designed to antagonize the tube-forming enzyme or augment the enzyme governing regression or compete for PtdIns as a way of tipping the balance in favor of vascular regression.”

—Elizabeth Dougherty

top