Artificial Intelligence helps predict cancer

Tadpole melanocyte cells reveal that they can be hyperstimulated into a cancerous state. Research with artificial intelligence programming, however, has unraveled the molecular signaling pathway and physiological circuit that converts "instructor cells" - precursor cells to melanocytes - to become those cancerous pigment cells.

Uncle Joe smokes a pack a day, drinks like a fish and lives into his late 80's. His brother, with a similar lifestyle, dies of cancer at 55. Why do some people develop disease or disorders while others do not? New research fromTufts University and the University of Florida, could help provide some answers. Scientists used an artificial intelligence approach to identify aberrant cells.

Published in Science Signaling, their findings reflect the first time artificial intelligence has explained why some groups of cells deviate from normal during embryogenesis. According to senior author Michael Levin PhD, the Vannevar Bush Professor of Biology at Tufts and director of the Tufts Center for Regenerative and Developmental Biology.

The paper builds on the center's earlier studies applying artificial intelligence to help explain how the flat worm, planaria, regenerates.

"Our methodology can be taken well beyond simple organisms and applied to the physiology of cell behavior in vertebrates."

Michael Levin PhD,Vannevar Bush Professor of Biology at Tufts University, and director of the Tufts Center for Regenerative and Developmental Biology.

In the work, researchers applied artificial intelligence called "evolutionary computation" to pinpoint molecular mechanisms which induce normal pigment cells in African clawed frog (Xenopus laevis) embryos to metastasize. A series of drugs were used to disrupt embryonic cells' normal bioelectrical and serotonin signaling at a crucial stage of development. Serotonin is primarily found in the gastrointestinal tract (GI tract), blood platelets, and the central nervous system (CNS) of animals, including humans.

Even in the absence of DNA damage or exposure to carcinogens, the pigment cells of affected embryos acquired bizarre, branch-like shapes and developed other melanoma-like characteristics, proliferating uncontrollably and invading the frogs' internal organs.

Depending on which protein in the bioelectric pathway was tweaked, only a certain percentage of the frogs developed melanoma, while the rest remained healthy. "There's randomness to this process. It doesn't have the same result in all animals exposed to precisely the same agent, which mimics the variability in human responses to cancer-inducing stimuli," added Levin.

Furthermore, the tadpoles that did develop melanoma developed it in every pigment cell — each frog was either 100 percent metastatic or completely normal. Essentially all pigment cells in a tadpole are part of a single coin, which either flips heads (normal) or tails (cancerous) — said Levin.

"Metastasis appears to be a group dynamic rather than a single-cell decision."

Michael Levin PhD

Maria Lobikin PhD, a recent doctoral graduate from the Levin laboratory and first author on the paper, first identified all the building blocks: receptors, hormones and other signaling proteins, in the serotonin signaling pathway regulating the melanoma-like cell response. Then, the team applied artificial intelligence to mimic evolution and generate a signaling network for a "virtual embryo" which exhibited the same behavior observed in tadpoles.

Like biological evolution, evolutionary computation uses incremental improvement and selection to rapidly reach a conclusion to a hypothesis. Traditional lab experiments must randomly and exhaustively test each possible outcome suggested by experimentation, then retest to confirm.

"The artificial intelligence system evolved a pathway that correctly explains all the existing and very puzzling data. Best of all, it also made correct predictions on data it had never seen."

Michael Levin PhD

The knowledge from these molecular pathways has implications not only for new treatments and targets for tumor prevention, but for understanding many other seemingly random changes found in cells of living organisms.

When enough data is available, researchers could use this AI approach to develop a system to help doctors predict patients' individual genetic responses to environmental factors causing cancer — and individualize treatment as well.

Abstract: Driving melanocyte proliferation and invasion

Melanocytes play key physiological functions; one of the easiest to see is pigmentation. In frogs, the number, distribution, and shape of melanocytes are determined by a subpopulation of cells called “instructor cells,” which are regulated by changes in membrane potential. Forced depolarization of instructor cells can result in excessive melanocyte proliferation, altered melanocyte cell shape, and abnormal migration of melanocytes into multiple tissues, which results in darkly colored tadpoles through a stochastic all-or-none process; the embryos are either normally pigmented or hyperpigmented. Lobikin et al. unraveled the molecular signaling pathway and physiological circuit that mediates this melanocyte conversion process, and they used computational approaches to explain how this all-or-none, stochastic process can occur.

In addition to Levin and Lobikin, paper authors were Douglas J. Blackiston and Elizabeth Tkachenko of the Department of Biology and Center for Regenerative and Developmental Biology, Tufts University; Daniel Lobo, formerly of the Levin laboratory and now at the University of Maryland in Baltimore; and Christopher J. Martyniuk of the Center for Environmental and Human Toxicology and Department of Physiological Sciences, UF Genetics Institute, University of Florida.

M. Lobikin, D. Lobo, D.J. Blackiston, C.J. Martyniuk, E. Tkachenko, M. Levin, "Serotonergic regulation of melanocyte conversion: a bioelectrically regulated network for stochastic all-or-none hyperpigmentation". Sci. Signal.8, ra99 (2015).

This work was supported by the G. Harold and Leila Y. Mathers Charitable Foundation. Computation used a cluster computer awarded by Silicon Mechanics and the Campus Champion Allocation for Tufts University TG-TRA 130003 at the Extreme Science and Engineering Discovery Environment, which is supported by NSF grant ACI-1053575.

Tufts University, located on three Massachusetts campuses in Boston, Medford/Somerville and Grafton, and in Talloires, France, is recognized among the premier research universities in the United States. Tufts enjoy a global reputation for academic excellence and for the preparation of students as leaders in a wide range of professions. A growing number of innovative teaching and research initiatives span all Tufts campuses, and collaboration among the faculty and students in the undergraduate, graduate and professional programs across the university's schools is widely encouraged.

The content of this release is solely the responsibility of the authors and does not necessarily represent the official views of the funders.

Return to top of page

Oct 7, 2015   Fetal Timeline   Maternal Timeline     News     News Archive   

Top image shows a normal tadpole. Bottom images shows a tadpole in which the pigment cells' normal bioelectrical signaling was interrupted, causing the cells to develop melanoma-like characteristics.
Image credit: Courtesy of the Levin Lab/Tufts University



African clawed frogs Female [L]and Male [R]
Image credit: Edmonton, Canada