Large-scale growth of blood vessels, what biologists call angiogenesis, is usually a bad sign. The unstructured growth and differentiation of new blood vessels often signifies the presence of a tumor, which requires a tremendous blood supply to bring it nutrients and oxygen.
Some cancer treatments work by blocking angiogenesis and subsequently starving the malignant tissue, an idea first introduced in the 1970s. But how does one halt the juggernaut of gargantuan, spontaneous vascular networks whose sole purpose for existence is feeding a cancerous tumor?
Hopkins researchers Emmanouil Karagiannis and Aleksander Popel, both from the biomedical engineering department, have addressed that question in a recent paper published in the Proceedings of the National Academy of Sciences on Sept. 16.
Their research focuses on developing methods for the identification of peptides - short proteins - that thwart the creation and mobility of endothelial cells which line blood vessels.
"In the current work, we have applied [a computational algorithm] in order to identify antiangiogenic peptides - peptides that inhibit the proliferation and migration of endothelial cells," Karagiannis said.
"During the last 30 years, approximately 40 peptides that inhibit angiogenesis have been identified. The way that those peptides were identified was through a time-consuming, low-yield and expensive methodology. Whole proteins that the investigators were thinking that they may regulate angiogenesis were processed experimentally with various enzymes, the proteases, that cleave proteins in smaller fragments and later these small fragments were screened for activity."
The methodology addressed by Karagiannis and Popel is computational, meaning it relies on heavy use of data processing rather than work in the laboratory. Part of this method uses information from BLAST, the Basic Local Alignment Search Tool, a free computer program that is available on the National Institutes of Health's website. Two other calculations, called the Smith-Waterman algorithm and a Monte Carlo filter, comb through this data to find bioactive peptides that can be as long as 25 amino acids.
BLAST is a widely used algorithm in bioinformatics that can be used for comparing amino acid sequences in different proteins. The SW algorithm is also used for local sequence alignment, or the determination of similar regions between protein sequences. It works by comparing fragments of all lengths and optimizing the similarity measure.
"We are using as an input a set of sequences, and from there on the algorithm can scan the whole proteome and identify similarities. We combine all of the algorithms in order to identify statistically significant hits of short peptides. We are using the BLAST algorithm with loose search criteria to identify a large number of hits, from there on the SW algorithm is used to remove the 'noise' induced by evolution and the Monte Carlo to filter the statistically significant hits from the outcome of the SW algorithm," Karagiannis said.
They were able to identify another 120 antiangiogenic peptides using the algorithm, thereby quadrupling the number of peptides identified within the last 30 years in just one year. These peptides can now be studied to discover possible therapies to combat angiogenesis in tumors.
It is clear the computational methodology works in real life: The lab is working on optimizing the peptide sequences for in vivo administration, or use in live models. Karagiannis said they have already screened some of the peptides predicted by the model lung and breast cancer samples, and the results so far are good.
"In order to provide proof of principle that the algorithm is using, we have also screened the peptides for activity. Approximately 80 percent of the predicted peptides are working," he said.
The study included the identification of the receptors to which the peptides are binding. This initiates a cell signaling pathway that will leads to decreased proliferation of endothelial cells. Karagiannis notes five different receptors in the paper.
"One important result from our study was that by performing combinatorial screening of the peptides, meaning using combinations of two peptides together that actually target two different receptors, we were able to show synergism, meaning the activity of the combination was greater than the additive effect if each of the peptides was applied separately. And this result is important because tumors acquire resistance if treated by only one drug," Karagiannis said.
This implies that resistance to cancer treatment drugs can be eradicated by using combinations of the 120 identified peptides, which target different receptors. Treatments may also evolve to use peptide combinations in conjunction with a chemotherapeutic agent.
Please note All comments are eligible for publication in The News-Letter.