.
O P E N A C C E S S S O U R C E : Cell
This year’s Gairdner Foundation Award for Biomedical Research goes to Roel Nusse for his pioneering work on the Wnt signaling pathway and its many roles in development, cancer, and stem cells.
All multicellular animals face the daunting task of properly organizing their cells into complex, specialized tissues. They do not only have to build these structures during embryonic development but also actively maintain them to preserve homeostasis later in life. A tissue, therefore, is a dynamic entity in which cells actively turn over. The inability to balance cell proliferation and differentiation in this context is a recipe for disaster, resulting in either unbridled cell division (risking tumor formation) or loss of tissue integrity (contributing to degenerative diseases and aging).
In the past 40 years, the Wnt signal transduction pathway has emerged as a cell-to-cell communication pathway with an important role in each of the aforementioned processes. Conserved in all multicellular animals, it is one of the oldest developmental signaling pathways, controlling robust pattern formation in the early embryo in myriad species. Interest in Wnt signaling stretches far beyond the developmental and evolutionary biology communities, however, and research in the field has impacted on many areas of biomedical research. None of this would have been possible without the contributions of Roel Nusse, who has, in many ways, shaped the field itself from the very start (Figure 1).
Figure 1 Breakthroughs in Wnt Signaling Research
(Top left) The principle behind insertional mutagenesis as a method for oncogene identification using slow-transforming retroviruses such as MMTV. Proviral integrations occur randomly throughout the genome. If a proviral insertion in the vicinity of a given gene yields a selective advantage, for instance by resulting in overexpression of a growth-promoting gene, it can be identified as a prominent or clonal mutation in the resulting tumor. By searching for so-called common integration sites (i.e., integrations of the provirus in the same genetic locus across multiple independent tumors), new cellular oncogenes can be identified.
(Top right) Identification of the wingless mutant as the Drosophila int-1 homolog (Dint-1) using classical mapping on polytene chromosomes. The wingless protein (wg) is a classical morphogen that forms gradients in a tissue. Cells close to the site of wg production encounter high levels of the protein (resulting in the induction of high-threshold target genes), whereas cells further away from the source face much lower levels (resulting in the induction of low-threshold target genes). Examples from Drosophila development are shown, such as the wingless/engrailed (wg/en) circuitry in establishing segment polarity and the induction of senseless (sens) and distalless (dll) in the wing imaginal disc.
(Bottom right) The power of Drosophila genetics resulted in an early map of the Wnt pathway and its core signaling mechanism. It recognized an early role for porcupine (PORCN) in the Wnt-producing cell and revealed the order of signaling events in the receiving cells with frizzled2 (FZD), dishevelled (DVL), zeste-white 3 (GSK3), and armadillo (CTNNB1) transducing the signal. Current Wnt-secretion inhibitors block the activities of mammalian PORCN. The precise mode of interaction between Wnt and its receptor was not revealed until the Wnt/Fzd crystal structure was resolved by Claudia Janda and Chris Garcia in 2012.
(Bottom left) Successful purification of active WNT3A protein allowed direct testing of a role for Wnt proteins in balancing stem cell self-renewal and differentiation. Nowadays, Wnt proteins are well recognized as self-renewal factors for many different populations of stem cells.
Unbeknownst to him at the time, Nusse stumbled upon the Wnt pathway in the late 1970s, when he was a PhD student in what was then the Division of Virology at the Netherlands Cancer Institute in Amsterdam. His research focused on the mouse mammary tumor virus (MMTV), a retrovirus that had long been known to cause mammary tumor formation in certain inbred mouse strains. The mechanism, however, remained entirely unknown. His was another project in a long tradition of MMTV studies that had been ongoing at the Netherlands Cancer Institute since the 1930s.
Let us pause for a moment to sketch the biomedical research landscape at the time. In 1975, the potential hazards of recombinant DNA technology (i.e., the possibility to cut and paste together pieces of DNA from different species for the first time in history, something referred to as “DNA cloning” and nowadays one of the most basic techniques in any molecular biology lab) had been discussed at the famous Asilomar conference. This had resulted in a set of guidelines under which experiments using the technology were judged safe to continue. Up until then, it had been virtually impossible for researchers to get their hands on specific DNA sequences—let alone use them for experimentation. Around the same time, the question of what caused cancer was also a hot topic of debate. Cumulative research had just revealed that the fast-transforming Rous sarcomavirus carried a transforming oncogene, later to be known as v-src. Other, slow-transforming viruses, including MMTV, seemed devoid of such a load. Then came the Nobel prize-winning work of Harold Varmus and J. Michael Bishop, which showed that v-src had a non-viral origin: somehow, the virus had at one point hijacked a cellular gene (c-src). This so-called proto-oncogene was found to be present in the genome of many different species, suggesting that cancer too might have a cellular origin and that, perhaps, slow-transforming retroviruses could induce cancer by activating these proto-oncogenes.
Roel Nusse initially studied the infection mechanism of MMTV as well as the characteristics of some of its proteins. But in 1980, researchers at the institute had gotten their hands on recombinant DNA technology, and the expectation was that this would help reveal what changes were brought about by integration of MMTV into mammary epithelial cells. By then, Nusse had also come into contact with Harold Varmus, who was equally intrigued by the question of how MMTV caused mammary tumors to form. Nusse joined Varmus for a postdoc at the University of California in San Francisco. The offer letter (archived by the US National Library of Medicine) sketches a long list of interesting projects, mentioning what would later become groundbreaking work as almost an afterthought in the P.S.: “I should have mentioned the possibility of looking directly at the ‘dominant’ proviruses in tumor DNA.” Together, Nusse and Varmus indeed set out to find a tumor with a clonal MMTV insertion in which they could ultimately map the precise proviral integration site: in this particular locus (int-1), the virus had landed in the vicinity of an unknown gene, inducing its enhanced expression via the strong enhancer sequences in the viral LTRs. Many other, independent tumors also harbored a proviral integration in int-1, suggesting a causal role for this cellular gene product in mammary tumor formation. One year later, the structure of the gene had been mapped. Barring any homology to known oncogenes, this represented the discovery of a novel cellular proto-oncogene (Nusse and Varmus, 1982).
After returning to the Netherlands Cancer Institute, Nusse and coworkers spent the remainder of the decade trying to figure out the identity and function of the int-1 gene. As a sign of the times, the Division of Virology had been renamed to the Division of Molecular Biology, and Nusse and colleagues were eager to apply the appropriate methods to further investigate int-1's biological function and oncogenic properties. This was not immediately successful: sequencing of the gene did not reveal similarities to other known sequences (although it should be noted that the database at the University of California in San Diego contained only 2,000 of such sequences at the time). It did uncover a stretch of hydrophobic amino acids at the N terminus—now known to be the signal peptide for Wnt protein secretion. Attempts to raise antisera against an int-1 fusion protein expressed in bacteria did not yield antibodies capable of recognizing the native protein. Transcripts of the int-1 gene could be detected in embryos, but not in adult animals. An early effort to generate int-1 transgenic mice (in collaboration with molecular geneticist Anton Berns, who had recently introduced genetically engineered mouse models at the institute) also failed to provide any clues, as the transgenic animals all died prior to weaning.
In 1986, because of the restricted embryonic expression pattern and given the fact that int-1 was highly conserved in evolution, the idea arose to clone the Drosophila homolog, which could then be used to screen for developmental mutants. This turned out to be a lucky decision: the fly homolog mapped to the same chromosomal position as the segment polarity gene wingless (wg). Just like that, int-1 had become the first mammalian oncogene with a known developmental mutant in Drosophila and, with that, the first example of an oncogene with a critical role in normal development (Rijsewijk et al., 1987).
With their hands on a fly homolog and access to developmental mutants, exciting new opportunities arose. Attempts to generate antisera against the wg protein were successful. Earlier work on wg had shown that the mutation acted non-cell autonomously. This, in combination with the protein sequence, suggested that the wg gene product could be secreted. Together with Peter Lawrence at Cambridge University, who studied pattern formation in the Drosophila embryo, Nusse was able to show the specific location of wg in developing larva. Electron microscopy studies clearly revealed the presence of wg protein inside small vesicles, as well as in the extracellular space between wg-producing cells and neighboring engrailed (en)-expressing cells. The latter also seemed to have taken up wg protein, leaving the authors to conclude that “it seems likely that the wingless gene product functions as a paracrine factor that binds to a receptor” (van den Heuvel et al., 1989).
In 1989, Roel Nusse was whisked away by Stanford University to join the newly formed Department of Developmental Biology in the Beckman Center for Molecular and Genetic Medicine and, together with most of his group, left the Netherlands Cancer Institute—something that then-director Piet Borst described as “a major blow,” highlighting his “broad knowledge, nose for quality and ability to promote collaborations.”
The generation of wg transgenic flies turned out more successful than the earlier attempts in mice: the flies presented with an embryonic phenotype that made it possible to screen for suppressor mutations and thereby map the factors responsible for transducing the wg signal. Multiple groups used similar approaches at the time, harnessing the full power of Drosophila developmental genetics that had been made possible by the Nobel prize-winning work of Christiane Nüsslein-Volhard and Eric Wieschaus. In the early 1980s, they had managed to identify, and then classify, many different segmentation mutants. The wg mutant belonged to one of these classes—that of the segment polarity genes (Nüsslein-Volhard and Wieschaus, 1980). Mark Peifer and Eric Wieschaus first proposed that another one of these genes, armadillo (arm), interacted with the wg pathway to control pattern formation in the developing and adult fly (Peifer et al., 1991). By combining different mutants, genetic epistasis experiments soon revealed many more components involved in the wg pathway and, more importantly, the order in which they acted. Together, work from the Nusse, Peifer, Perrimon, and Wieschaus labs thus revealed the core working mechanism of what was, by then, becoming known as the Wnt pathway (Clevers and Nusse, 2012, Nusse and Varmus, 2012).
In a savvy display of forward thinking, the few scientists working on int-1/wingless-related proteins in vertebrates jointly proposed a new nomenclature for int-1 and its rapidly expanding gene family. Among them were Andrew McMahon and Randall Moon, who had by then begun to identify a large family of int-1-related genes in mice and frogs. They had also just introduced a powerful technique to study the activity of int-1 in early vertebrate development: the famous Xenopus axis duplication assay (McMahon and Moon, 1989). The name Wnt (pronounced “wint”) was proposed “for the wingless-type MMTV integration site that founded the gene family” (Nusse et al., 1991). It has often been said that behind every great man there’s a great woman, and rumor has it that it was in fact Roel Nusse’s wife, Betsy, who came up with the suggestion for that name (Nusse and Varmus, 2012).
With a large part of the Wnt signal transduction pathway now mapped out, two main question marks remained. What was the identity of the cell-surface receptor that made cells competent to respond to Wnt proteins, and what happened once the signal reached arm? Drosophila would again provide one of the answers, when fz2 was identified as a wg receptor in a collaborative effort between the labs of Roel Nusse and Jeremy Nathans (Bhanot et al., 1996). Light on the other question was to be shed from a different corner, representing one of the few holes in the Wnt pathway that were plugged without Roel Nusse’s apparent direct involvement.
Armadillo in flies had by then been shown to be homologous to beta-catenin (CTNNB1) in vertebrates, a known component of adherens junctions. While Nusse was occupied with the upstream hunt for the Wnt receptor, scientists over in Europe uncovered an interaction between CTNNB1 and transcription factors of the TCF/LEF family. One of them was Hans Clevers, who had up until then been studying TCF1, a transcription factor critical for T cell development. Another was Walter Birchmeier, whose research had been angled toward the role of E-cadherin, and by association CTNNB1, in suppressing cell invasion and metastasis. Yeast-two-hybrid screens revealed CTNNB1 as a binding partner for human TCF1 and, conversely, led to the discovery of LEF1 as a CTNNB1-interacting protein (Behrens et al., 1996, Molenaar et al., 1996). This not only marked a clear end point for how Wnt signaling was ultimately able to regulate target gene transcription but also gave rise to one of the most robust tools to study WNT/CTNNB1 signaling: the famous TOPFLASH luciferase reporter assay, which contains a multimerized stretch of TCF/LEF binding sites. A couple of years prior, Paul Polakis and others had already shown that CTNNB1 interacted with the tumor suppressor protein APC, thereby firmly consolidating the link between Wnt signaling and cancer. By 1998, the idea of a “destruction complex,” responsible for the controlled turnover of CTNNB1 in the absence of a WNT signal, had become established in the rapidly expanding community of Wnt researchers, tying all of the available genetic evidence together into a biochemical model that still very much holds up today (Clevers and Nusse, 2012, Nusse and Varmus, 2012).
Meanwhile, Nusse revisited the problem of Wnt protein purification. Plans to overproduce and purify a bioactive form of the mammalian Wnt1 protein had been made as early as 1988 and had failed miserably. But in 2003, Karl Willert and Roel Nusse finally succeeded to purify active mouse WNT3A protein (Willert et al., 2003). The “trick,” as it turned out, was to preserve solubility of the Wnt protein—something that is achieved by including serum in the medium while culturing and by including detergent during the purification and fractionation steps. Otherwise, Wnt proteins, which turned out to be highly hydrophobic as a result of a post-translational lipid modification, will aggregate and become nonfunctional. This important achievement chimed in a new chapter in Nusse’s research, which would slowly move away from Drosophila and back to studying mammalian tissues over the next decade. It was becoming clear that in addition to controlling the development of complex animal tissues, WNT/CTNNB1 signaling was also critical for maintaining their integrity, suggesting a role in stem cell biology. Indeed, Nusse’s lab has shown that purified WNT3A is capable of maintaining self-renewal in many types of stem cells, including pluripotent embryonic stem cells and, signaling a brief return of his research to the tissue in which it all began, those of the mammary epithelium (Zeng and Nusse, 2010). Roel Nusse’s recent work has also employed in vivo lineage-tracing strategies, using expression of the negative-feedback target gene Axin2 to mark WNT/CTNNB1-responsive cells, to identify new populations of stem cells in multiple tissues, with a recent focus on hepatocytes and adjacent central vein endothelial cells as a liver stem cell/niche compartment.
It should be evident by now that Roel Nusse is a well-deserving recipient of this year’s Gairdner award. Major other breakthroughs have been made possible as a result of his work, which reverberates across the cancer research and stem cell biology fields. Retroviral insertional mutagenesis was used for decades after the initial MMTV screen by Nusse and Varmus to identify new oncogenes, for instance by the labs of Anton Berns and Neil Copeland. Model organisms from across the animal kingdom, including the starlet sea anemone Nematostella and planarian flatworms, have revealed a fundamental role for Wnt signaling in regeneration. Organoid cultures, branded “method of the year” by Science magazine in 2017, have transformed basic stem cell research and hold great promise for regenerative medicine. Those cultures, started from adult tissue stem cells, almost invariably require active WNT/CTNNB1 signaling for their long-term maintenance.
.../...
T O A C C E S S T H E R E S T O F T H E S T U D Y, P L E A S E V I S I T T H E S O U R C E .
.
Edited by Engadin, 03 May 2020 - 07:46 PM.
