Ever since DrugBaron studied biochemistry at the end of the 1980s, one question about cancer has always perplexed him: why can we induce almost complete remission and yet the cancer can come back stronger and invariably kill the patient? The answer now, as then, has always been selection of the rare, resistant mutant.
But DrugBaron was always left with a nagging doubt. Two observations seemed to call that obvious and simple answer into question.
Firstly, for the most part that is not what we see with infectious agents. Even viruses, which share with cancer cells a short generation time and a ready ability to mutate, can usually be cured as long as you hit them hard – because the chances of a resistant variant existing in the population is quite low. As long as you don’t give them time to find one over many generations, you can eliminate the infection. Since tumours typically have a much smaller cell population size than an infection, it follows that cancer should be easier to cure than it is.
Secondly, some cancers, such as some leukemias can be cured almost every time, while others, such as the brain cancer glioblastoma, are almost invariably lethal. Yet there is no obvious association between mutation rate or the cancer cell population size and tractability of the cancer type.
It’s hard to put a finger on it, but it has always felt like a key part of the puzzle was missing.
And then, a few years ago, DrugBaron was sitting in a bar in Cambridge, UK, enjoying a drink with my friend and colleague Professor Miro Radman, when in the space of an hour he suggested to me exactly why some tumours were essentially invincible. The missing piece, it turned out, was the ability of cells to form networks.
For sure, I was aware that cells often formed connections with each other. Indeed, skeletal muscle is made up of giant syncytia formed by the complete fusion of many individual cells that give up their autonomy to work together. In other tissues, the cells remain separate but communicate through numerous channels, such as the gap junctions between cardiac cells that ensure the contraction during a heartbeat is carefully co-ordinated. But, like most people, DrugBaron had never considered the full implications of such cell networks before – and certainly never imagined they had anything to do with cancer.
Prof Radman’s concept was logical enough: if cells are highly connected, then they cannot change phenotype very easily. Rather like a snake with two heads, if one wants to go left and the other right, it will continue down the middle. In the same way, networks of cells ensure phenotypic stability in normal tissues. For any cell to act differently from its neighbours, it would first need to disconnect from the local social network.
This phenomenon, which he termed cellular parabiosis, he reasoned, protects long-lived species such as humans against accumulated mutations in the DNA. Even as individual cells suffer different mutations over time, the phenotype of the tissue remains robust and stable thanks to the network effect. The parallel with the internet-based social networks of the 21st Century is very real: rare variant views are rapidly suppressed by the voice of the masses.
But this very same network can also stabilise unhealthy phenotypes once they are established – something Prof Radman called “solidarity among criminals”. By sharing proteins and even whole organelles, the cancer cell network can sustain itself even in the presence of the really rapid mutation rate that typifies cancer.
DrugBaron quickly saw another consequence of such a network: resistance to death. Essentially all anti-cancer agents work by inducing the cancer cells to die – toxic chemotherapeutics, radiotherapy, even modern-day immunotherapy all have the same final common goal: activation of the programmed cell death pathway in the cancer cells. And the common molecular signal is high intracellular levels of calcium ions. Killer T cells, for example, punch holes in the cell membrane allowing calcium ions to flood in and trigger cell death. Irrespective of the trigger, the goal is the same: get calcium levels above a threshold so the cell will die.
But what happens if the cell is heavily networked with its neighbours? It can quickly share the calcium influx across the network and blunt the rise in intracellular calcium. We have called this “calcium capacitance”. In effect, the cell has become harder to kill. If all the cells in the network are subject to the attack, then eventually all would succumb – but critically if the cancer cells can network with normal cells, it can reach the point where the calcium levels can never reach the threshold to trigger death. The cells in the network have become, quite literally, invincible!
But all that was a discussion in a bar. Yes, it was a fascinating possibility, but it was, after all, only a theory. There were little pieces of evidence here and there of cells forming networks, and stabilising neighbouring phenotypes, but hardly proof of the central importance of networks in cancer cell survival.
All that changed in 2015.
In a seminal paper, published in Nature, Prof Frank Winkler and his colleagues from the German Cancer Centre DKFZ and the University of Heidelberg, showed that – at least in glioblastoma – exactly what we had proposed was really happening. Using a unique intravital microscopic model of human glioblastoma cells growing in mouse brain, they were able to show that the cells formed functional interconnections – the cellular social network we had imagined. Better still, when they treated with chemotherapy or radiotherapy, they found that the isolated, “out of network” glioblastoma cells died, but the more connected ones survived. These networked cells then quickly expanded to kill the mouse.
And today, with their new paper, also published in Nature, Prof Winkler has completed the picture, showing functional networks between the glioblastoma cells and the normal, healthy brain cells in the mouse. They have even used a fluorescent calcium-sensing dye to see calcium capacitance in action.
Now, then, we have direct evidence that the ability to form networks grants those cancer cells a cloak of invincibility. No matter how good we get at targeting the cancer cells with killer agents, the most connected cells simply will not die – but will expand rapidly and kill the individual, exactly as we observe, sadly, in almost every patient today.
This entirely new area of biology provides an elegant explanation for why some tumour times have proven so intractable against generation after generation of new, and ever more sophisticated, approaches to killing them. But it does much more: it provides us with an approach to curing these most impossible-to-treat cancer types, expressed succinctly by Prof Winkler as “disconnect the bastards”.
And so it was that our new oncology company, Divide & Conquer came into being. The two Professors and I raised £10million from Medicxi, who quickly saw the huge potential for this new approach to curing, rather than delaying the progress, of these cancers, to discover drugs that would indeed disconnect the cells from the network and strip away their cloak of invincibility. And we have been exceptionally busy, in stealth mode, learning how cells regulate the formation of these networks, and designing agents that could break them up.
So, to mark the key publication of Prof Winkler’s latest experiments, Divide & Conquer has today announced itself and its mission to the world.
By studying the way normal cells regulate their networks, we have focused on a master switch: a protein family whose mysterious central role in cell biology has been partly understood for more than 40 years – Protein Kinase C (PKC). A hitherto under-appreciated role of this molecular switch is to disconnect a cell from its neighbours, to grant it the autonomy to differentiate, to proliferate and – for our story – most critically to die.
Divide & Conquer have found a way to throw this switch so that it disconnects the glioblastoma cells from the network, rendering them mortal. This agent, a small molecule drug, is now on a fast-track to the clinic, which is the only place we can really test the final plank in Prof Radman’s grand hypothesis. If we disconnect “the bastards” shortly before conventional chemo- and radiotherapy will we cure, rather than temporally hinder, this most fearsome of tumours?
There is something of a delicious irony to the fact that molecules that activate this PKC switch are well known in the literature as tumour promoters. But that makes perfect sense, at least according to Prof Radman’s cellular parabiosis story. The healthy cell network is protected against the impact of mutations – but if you throw the PKC switch and disconnect them all, then mutations in, for example, tumour suppressor genes are suddenly granted the autonomy to express themselves, and hence tumours appear. Disrupt healthy cell networks and you risk causing cancer; but disrupt the network of cancer cells and you open a window of opportunity to kill the otherwise invincible terrorists.
For Divide & Conquer, then, as for every other oncology company, the key question is whether we can safely disconnect the tumour cell network. The omens are good. Rather like a teenager disconnected from Facebook or Instagram for a week’s holiday, who is grumpy but essentially unharmed, normal cells can, it seems, cope with a short period of isolation. But the cancer cells, who, without the support of their neighbours, are intrinsically very susceptible to our current anti-cancer drugs, may just succumb.
Only time will tell whether this entirely new approach can deliver durable cures for the worst types of cancer. But hopefully we will not have too much longer to wait to find out. DrugBaron will be disappointed if we haven’t treated out first patient before the end of 2020. Remember the name: Divide & Conquer – it might just provide the breakthrough cure for solid tumours that even immunotherapy has, so far, largely failed to deliver.
RxCelerate Ltd is an outsourced drug development platform based near Cambridge, UK. We specialize in delivering an entire road map of drug development services from discovery and medicinal chemistry through to formal preclinical development and clinical up to Phase IIa. In the last five years, we have witnessed dramatic changes in the drug development …