Cell Adhesion

Richard O. Hynes
Daniel K. Ludwig Professor for Cancer Research, David O. Koch Institute for Integrative Cancer Research, Massachusetts Institute of Technology, Cambridge, MA USA; also Howard Hughes Medical Institute Investigator
Erkki Ruoslahti
Distinguished Professor, Center for Nanomedicine, Sanford-Burnham Medical Research Institute, La Jolla, CA USA
Masatoshi Takeichi
Director, RIKEN Center for Developmental Biology, Kobe, Japan

Hynes, Ruoslahti, and Takeichi have been suggested as possible Nobel Prize winners “for pioneering discoveries of cell adhesion molecules, Hynes and Ruoslahti for integrins and Takeichi for cadherins”

Richard Hynes and Erkki Ruoslahti were pursuing the same molecules at the same time, and both made fundamental discoveries about how cells stick to one another and move. Takeichi’s quarry was slightly different, but plays the same role, as the glue that makes multicellular bodies possible.

Hynes and Ruoslahti independently discovered a large protein molecule that was present on the surface of normal fibroblast cells, but absent in transformed tumorous fibroblasts. Hynes called it the large external transformation-sensitive, or LERTS, protein. Ruoslahti called it fibronectin, and that name stuck. The big question then was how exactly did this relatively huge molecule help cells to adhere to one another and to extracellular components such as bone and tendons. Circumstantial evidence began to accumulate that there was a physical link between fibronectin on the outside and actin filaments that make up the cell’s cytoskeleton, and the link even appeared as an unknown element in diagrams of cell adhesion. Hynes and Ruoslahti sought that link in different ways.

Hynes learned of a monoclonal antibody that could be seen in fluorescence micrographs connecting fibronectin outside the cell to actin within. Via a complex series of steps, his group eventually cloned the protein. It turned out to be a transmembrane protein that bound to fibronectin and actin. Hynes called it integrin, reflecting that this integral membrane protein complex maintained the integrity of cell and body structures and had an integrating function.

Ruoslahti purified the separate domains of fibronectin and zeroed in on a small fragment — just 108 amino acids of more than 2,500 — that promoted cellular adhesion. That smaller fragment contains a sequence of three amino acids, known as the RGD peptide, which is the crucial recognition sequence for the entire family of integrins in every known multicellular organism examined to date. Indeed, the molecular mechanism of cell adhesion predates the Cambrian explosion 530 million years ago and in some sense permitted the evolution of multicellular organisms.

Well before Hynes and Ruoslahti identified integrins, researchers knew that trypsin, an enzyme that breaks down proteins, loosened the bonds between cells. Usually, the disruption is temporary, and the cells reaggregate once the trypsin is removed. Masatoshi Takeichi knew that from his doctoral work in Kyoto. But when he moved to the Carnegie Institution for a post-doc, he discovered that dissociated cells stayed dissociated. The difference was that at the Carnegie Institution the trypsin solution also contained EDTA, which effectively removed positive ions from solution. Takeichi quickly established that the cell adhesion he was working on depended on the presence of calcium ions. Without calcium, dissociated cells remained separate. Add calcium, and they once again stick to each other. Back in Japan, his graduate student Chikako Yoshida proposed the name cadherins, from “calcium” and “adhere.”

Cadherins and integrins are absolutely fundamental to the development and maintenance of multicellular organisms. They tell cells where to go in development and which other cells to associate with. They are fundamental in wound healing; activated integrins stick blood platelets together to form a clot and engage white blood cells to stay close by and fight infections. And, of course, they can go wrong. The original observation that transformed cells lack fibronectin explains why cancer cells are often a different shape and why they metastasize and go off to invade secondary tissues. Therapeutic applications have followed, based on either keeping cells in place, to promote wound healing or slow the spread of tumors, or loosening their grip, to reduce the inflammation caused by white blood cells that have attached in the wrong place. The RGD peptide is proving useful to guide drugs to specific target cells, and analogs are proving more potent and more target specific.

Commentary on the Medicine Laureates by Jeremy Cherfas, Biology correspondent, ScienceWatch

Interview with Erkki Ruoslahti, Distinguished Professor, Center for Nanomedicine, Sanford-Burnham Medical Research Institute at University of California Santa Barbara.

For pioneering discoveries of cell adhesion molecules

QPlease provide a brief overview of your field of research and explain what led you to focus in this area?

AMy work deals with cell adhesion, tumor metastasis, and more recently, molecular specialization within the vasculature (“vascular zip codes”). When I was a postdoctoral fellow at California Institute of Technology many years ago, my supervisor Bill Dreyer introduced us to Nobel laureate Roger Sperry’s ideas on axonal growth. Sperry hypothesized that specific recognition molecules at the cell surface guided axonal processes of neurons to the appropriate place in the brain during development, and that perhaps similar mechanisms operated in all cells. We were already, at that time, talking about area codes and zip codes for cells to find the right place in the body and keep them there. My main interest was cancer, and it seemed to me that if such a guidance system existed there would have to be something wrong with it in cancer, because cancer cells do not stay where they are supposed to be. This was all entirely hypothetical, no molecules with such activities were known, but I decided that it was what I wanted to study when I set up my own laboratory.

QWhat did you want to accomplish when you began your research?

ABack in Finland and with my own laboratory, I wanted to find guidance molecules that kept normal cells in their appropriate place and were altered in metastasizing cancer.

QWhat notable problems, challenges, or obstacles did you face? Conversely, have there been particular sources of enjoyment, satisfaction, or pride?

AI was trained as an immunochemist and protein chemist and was not comfortable with cells at all.  Fortunately, my cell biologist colleague Antti Vaheri had also just returned to Helsinki from a postdoctoral stint in the US and was interested in similar problems.  We decided to join forces and had a great 5 years of collaboration before I left for the US to stay.  We made antibodies against proteins.  We peeled off the surface of cultured fibroblasts thinking that we would then sort out if any of the many antibodies we expected to get were against cell adhesion molecules. Much to our surprise, the antiserum was essentially specific for one protein, a protein we later on, together with Deane Mosher, named fibronectin. Our colleagues and competitors who had independently identified the same protein using other approaches weren't all too happy about the name, but it stuck. (A decade later, one of these competitors, my friend Richard Hynes, had his revenge; he gave the adhesion receptors we all were then working on the name integrins). 

One other problem in hindsight was that I was not single-minded enough about pursuing the original zip code molecule hypothesis; we worked on many other things, circulating tumor markers in particular that produced a lot of papers, but that did not turn out to be as important as the cell adhesion work. The loss of cell surface fibronectin in malignantly transformed cells created a tremendous amount of excitement at that time because it suggested an important role for fibronectin in cancer. My great disappointment in those days was missing the discovery of the cell attachment activity of fibronectin, which became another crucial piece of information suggesting that fibronectin was part of a cell recognition system. It was all the more disappointing because we had some initial results suggesting cell attachment activity of fibronectin.

After I had moved my laboratory to California, a great step forward was the discovery by Eva Engvall of the binding of fibronectin to denatured collagen (gelatin). It allowed us, and once we published the finding, everyone else, to produce essentially unlimited amounts of fibronectin for biochemical work. We set out to map the various binding sites in fibronectin, and soon focused on the cell attachment site. We zeroed in on it by identifying smaller and smaller fragments with this activity, until we had a tetrapeptide RGDS. Actually, it was a tripeptide, RGD, because using synthetic peptides, we showed that the residue, serine, could be almost any amino acid.  That this small peptide motif in the huge fibronectin polypeptide could have so profound an activity was a surprising finding.  It didn’t end there; we quickly realized that RGD was the attachment motif in a number of other adhesive proteins. One of the best decisions I made at that time, in the early 1980’s, was to not limit our studies to fibronectin, but to compare and contrast fibronectin with another adhesion protein with a similar activity, vitronectin. When we cloned vitronectin, we found an RGD in its cell attachment site.

The RGD peptides gave us a new tool to use in identifying cell adhesion receptors, which became our next goal. Using affinity chromatography, we purified receptors for fibronectin and vitronectin (now integrins α5β1 and αvβ3) and conclusively demonstrated the activity of the purified receptors by reconstituting a cell adhesion system in liposomes.  The results showed that both receptors were heterodimeric and RGD-dependent, but with different subunits and subtly different specificities.  These results established the existence of a cell recognition system very much like what I had set out to find. I remember having been really excited about that; I told everyone who would listen that we had found an important cell recognition system. It took 15 years, but there it was.

My recent work on vascular zip codes was inspired by a hypothesis proposing that tissue-specific tumor metastasis might be explained on the basis of circulating tumor cells having a specific affinity for the blood vessels in the favored target tissue, which would help the tumor cells lodge in that tissue and grow into a metastatic colony. For this to be the case, the blood vessels of individual tissues would have to be different. It occurred to me that we might be able to explore molecular specialization of blood vessels by screening phage-displayed peptide libraries in live mice. The results have revealed extensive molecular specialization of the endothelium in different organs and tissues, and we also documented an example of a vascular zip code identified in this manner actually being responsible for tissue-specific metastasis.  We then turned to using in vivo phage display to analyze tumor vasculature. We have created a large panel of tumor-specific homing peptides and have used these peptides to deliver drugs into tumors.  The tumor-homing peptides are particularly well suited for the targeting of nanoparticles, a recent major focus of my laboratory.  Remarkably, the most commonly encountered peptide motif in the phage screens is RGD. The reason is that RGD-directed integrins are specifically expressed in angiogenic blood vessels (and tumor cells). Quite recently, we identified an RGD-containing peptide with unique tumor-penetrating properties. Much of the current work in my laboratory is directed at understanding the basis of this tumor penetrating activity and making use of it in drug delivery.

QHow would you assess the importance and influence of your work?

QIt is now clear that cell-matrix interactions play major roles in cell and developmental biology.  The establishment and maintenance of tissue architecture, and cell differentiation, migration and survival are examples. Anchorage dependence, which metastatic cells must circumvent to be able to travel to distant sites in the body, is another manifestation of cell-matrix interaction. I am proud of our contributions to this field. The vascular zip code work is still too recent for a retrospective analysis, but holds substantial promise, particularly in nanomedicine. The trans-tissue transport pathway revealed by the tumor-penetrating RGD peptide may have profound physiological significance that remains to be explored.

QHas this research found wider application in industrial or commercial areas? If so, what are some possible future developments?

AInhibiting integrin activity with RGD peptides and their mimics has resulted in a number of drugs that are in the clinic or clinical trials. RGD peptides have also found application in the design of surfaces that promote cell adhesion. One of our first generation tumor-homing peptides different from RGD is in late stage clinical trials as a homing device for tumor necrosis factor α, and there is considerable commercial development activi4ty around our newer homing peptides.