During the 2012 AACR Annual Meeting at the beginning of the month, I had the opportunity to meet with AACR speaker Professor Hans Clevers, Director of the Hubrecht Institute for Developmental Biology and Stem Cell Research, Utrecht, Netherlands. Catching Clevers’ talk at the meeting, I was keen to ask him about every aspect of his lab’s work up to that point covering the journey from T cells to growing “organoids” in culture. This article covers the first half of our discussion, during which I found myself even questioning the necessity of hypotheses in research…
Professor Hans Clevers is well known for his lab’s work encompassing Lgr5+ stem cells in the gut; work that has now culminated in a technique to grow “organoids” – structures that resemble organs - in culture, independently of any host organism.
The implications of this body of research are potentially numerous and diverse, and the lab’s journey to this present state of scientific capability has been just as varied. It began with Clevers’ origins in immunology and traversed the study of model organisms for a while, before essentially starting again from scratch with the intestine. Since recent times, several more organ tissues have been under the scrutiny of the Clevers lab: pancreas, liver, lung, kidney, the list goes on. Unsurprisingly, I faced a genuine problem upon beginning my discussion with Hans – knowing where to start.
Coincidentally, the subject of deciding where to begin became thematic with my opening question (a rather generic: “What was your experience of the AACR Annual Meeting this year?”) I found my starting point in Professor Clevers’ description of his “Meet the Research Pioneer” session – a forum organized by the AACR committee to facilitate career discussions between prominent researchers and early-career scientists.
Amongst answering questions in the session such as “How do you make career decisions?” and “Do you fight journal editors over rejections?” Clevers had described to those present his general approach to scientific practice and how he locates starting points in research; I found this fascinating and totally opposed to how I was taught to go about science during my graduate studies….
“I try to discourage people from formulating hypotheses: it is not how science should be learned. This idea that you should first formulate a question – your hypothesis – then test it, often goes wrong I find.” He explained further: “Once you formulate your hypothesis, it’s very natural to then prove that you are right rather than to do the opposite; hypotheses are often formulated in such a way that it takes longer to find out if you are wrong. Another negative aspect of this approach is that you rarely learn much more than one answer to one question – typically, you do not look at other possibilities.”
“I’m strongly convinced this is how it should work when you really want to discover novel things – you cannot hypothesize how something works.”
So perhaps not having a solid starting question was not such a bad thing for this interview? After all, it is a tactic that seems to have worked for Clevers: “The hypothesis-free approach has worked very well for me, and for others, but…” He admits, “There is a significant hurdle to it: It is very difficult to write grants when you have no hypothesis. We now rely on previous studies to generate grants for new ones.” And there was indeed a time when the Clevers lab found it difficult to secure funding to approach science in this way…
After finishing postdoctoral work at the Dana-Farber institute at Harvard – where he cloned the T cell gene CD3 epsilon – Clevers set up his lab in Utrecht, the Netherlands, primed to chase down a discovery made during his postdoctoral studies. He had found a T cell-specific enhancer back in Boston and, shortly after being established, his lab managed to clone its gene: T cell factor 1 (TCF1). But then a challenge presented itself:
“We cloned TCF1 right away, but then it took 6-7 years to figure out what it did. We found there were three more TCFs in that time too: LEF, TCF3 and TCF4. They all bind DNA beautifully but we were without explanation to their exact function as they couldn’t regulate transcription – so it was not easy to get grants back then.”
The advent of two-hybrid screening technology ushered in an experiment in the lab that was to identify the already famous β-catenin as one of several cofactors binding the Tcf protein: “The moment we pulled out β-catenin from the two-hybrid screen we realized Tcf alone only binds DNA – it requires β-catenin to activate transcription.” But completing the picture of TCF function was not the only conclusion of this work – β-catenin was already famous for its role in the Wnt signaling cascade:
“We essentially found a missing link in the Wnt pathway.”
“Before the two-hybrid screen, nobody knew that β-catenin could control gene expression – we essentially found a missing link in the Wnt pathway.” Something they had not set out to do initially – the hypothesis-free approach was working well.
“With the results in front of us we realized that T cells were not the best model system for furthering our understanding of TCF in the context of Wnt signaling. As such, we decided to work in animal models, and we quite rapidly went on to show that TCF is ubiquitous across the species: frogs, flies and the worm C.elegans.”
So how did the Clevers lab make the transformation from what was – circa 1996 – essentially a developmental biology lab to an organoid growing one, becoming experts in intestinal epithelium renewal and stem cells in between? Why is it that Clevers now frequently appears on the bill of cancer focused forums and conferences when at one time it would make more sense to find him on the speaker list of those dealing with Xenopus development?
Well, the Clevers lab’s involvement in cancer research began with several collaborations with major cancer authority Bert Vogelstein’s lab; inspired by the Vogelstein lab’s work on APC – an upstream regulator of the Wnt pathway – in colon cancer. Together the labs found that colon cancer cells lacking APC have high levels of β-catenin/TCF signaling. Conversely, Clevers then wanted to see what happened if TCF was removed from the equation: “We knocked out TCF4 in mice and we found a phenotype that was essentially the opposite of cancer: we lost the intestinal structures vital to the renewal of gut. This told us that there were two sides to this; one: if you over-activate β-catenin/TCF signaling you get colon cancer; two: this signaling cascade is the normal driver of turnover in the gut.”
“The architecture of the gut…told me this was a great place to begin to locate and study stem cells.”
The intestine then became an intense focus for Clevers’ lab: “So we realized that the gut is incredibly interesting because of this turnover – its epithelium replenishes within 4-5 days. We knew that the signaling program activated by TCF in this tissue caused cancer, but was also responsible for normal cell renewal – we then began to consider that finding whatever genes might be activated by TCF as the secret to finding adult stem cells. Plus, the architecture of the gut – the villi and their crypts – told me this was a great place to begin to locate and study stem cells.”
However, this change in direction was not such a smooth transition for the lab: “It was tough. You cannot just start something new like that – there are no funds available. So it was a slow progression, it took a few years overlapping with the model organism and T cell work. “There was no technology available at the time (around 1998-99) to support a hypothesis-free approach to finding TCF’s target genes; there was no such thing as a microarray just yet. There were all sorts of complicated cloning strategies to begin to ask what genes were controlled by TCF – but we had to take them one at a time, individually.” Technology however, was catching up…
“We were probably among the first labs to collaborate with Pat Brown – who had just invented DNA microarraying at Stanford. So in 2000 we constructed colon cancer cell lines (where the Wnt pathway is very active) in such a way where we could block their signaling and then do microarrays to assess many genes at a time and see which ones were ‘switched off’. We did an experiment with four microarray chips and found maybe 60-70 genes that were active in colon cancer that were activated by β-catenin/TCF signaling.” And Lgr5, that would become so significant not long after, was one of them.
The next step for the lab was to conduct expression studies on each of these sixty-plus genes – every member of the lab set about working on the list. Nick Barker, a postdoctoral researcher in Clevers lab at the time – who had been a major player in establishing a link between TCF signaling and cancer through the Vogelstein lab collaboration – noticed that one gene in particular had unique expression pattern: “Nick looked at a few genes – expressing them with GFP tags in mice – and realized Lgr5 had this very unusual expression pattern. It basically marked these very small cells in the crypts of the gut that we had never seen before. Immediately upon seeing these cells I knew we had found something. Now when I look at an intestinal crypt I see them immediately – in every crypt you see this cell.”
That cell type came to be known as the Lgr5+ cell, and later studies began to show their proliferative capabilities as well as their other rather unique and interesting traits. It was the identification of these cells that formed foundations for the work that now enables Clevers and his collaborators to grow organ-like structures in a culture dish – completely independently of any complex system or organism. How that leap was made however remains for part two of the interview – which you can catch in an upcoming blog post. A conclusion for the moment though: it seems, despite the difficulties associated with it, there is certainly something to be said for a hypothesis-free approach to conducting science, and Professor Clevers would support that notion: “I’m strongly convinced this is how it should work when you really want to discover novel things – you cannot hypothesize how something works.”