Skip to main content

Together we are beating cancer

Donate now
  • Science & Technology

Sticky cells, blood vessels and cancer – the paradox of Claudin-14

by Marianne Baker | Analysis

14 June 2013

1 comment 1 comment

Dr Marianne Baker

Dr Marianne Baker, in the lab

How do our bodies form new blood vessels? This is a key question in cancer research, as tumours need to develop a new blood supply to grow.

Last summer, Dr Marianne Baker finished her Cancer Research UK-funded PhD in the Centre for Tumour Biology at the Barts Cancer Research UK Centre in London, focusing on a small but potentially significant part of this question.

In May this year, her findings – focusing on some of the molecules involved in new blood vessel growth – were published in the journal PLoS ONE.

In this guest post, she explains what she did, what she found, what it means, and what happens next.

A sticky subject

Staying in one piece is very important for any organism that consists of more than one cell (that’s everything except bacteria and things like amoebas). Our cells obviously have to stick to each other and to the non-cell parts of the body (bones and all the other stuff) or we’d simply fall apart.

Researchers call this ‘cell adhesion’, and it was this that was the focus of the lab in which I did my PhD, under the supervision of Professor Kairbaan Hodivala-Dilke.

Studying cell adhesion is, basically, working out exactly how cells stick to and interact with each other and their environment. To do that, cells need their surfaces to be ‘sticky’ in some way – just like a smooth, round ball won’t stick to anything unless you cover it in Velcro and throw it at a receptive surface.

Cells’ stickiness comes from proteins that sit on their surface, often spanning each cell’s surrounding membrane like a sea monster loops in and out of the water. These proteins can make contact with similar proteins on other cells, holding together and keeping our bodies and organs in the right shape, and place.


‘Sticky’ adhesion proteins weave in and out of a cell’s surface like a sea monster

Check the pipework

But as well as giving our bodies shape and form, this stickiness is also important for our internal ‘plumbing’ – the circulatory system, which carries blood to our organs, providing life-sustaining oxygen and nutrients.

Studying exactly how tumours subvert the body’s plumbing is a key focus for researchers, because cancers also require their own blood vessels to get nutrients and grow (you can read more about this process – known as angiogenesis – in a previous post I wrote).

Researchers have known for decades that blood vessels in a tumour are abnormal and leaky, and that this can be both good and bad for the tumour. On the one hand, it can get more oxygen and nutrients in some regions of the tumour. But on the other, it can lead to areas of poor blood supply – leading to regions with a lot of cell death (called necrosis).

This can be good and bad for patients, too – good, because more chemotherapy drugs can leak out in the tumour, but bad because the drugs don’t penetrate the whole of it due to the poor blood flow. Low-oxygen areas are also more resistant to radiotherapy, because the treatment needs oxygen to work.

So, understanding the intricate details of what affects blood vessel growth, and the integrity (or permeability) of blood vessel walls can actually help us design different types of treatment that might be more effective.

This is a bigger job than we thought

The blood vessels in our bodies are lined by specialised cells called endothelial cells, which stick to each other via cell-cell contacts of different types, depending on which “sticky proteins” are involved.

One type of endothelial cell-cell contact is called a tight junction – and that’s what I looked at in the latter part of my PhD, and my results are reported in the paper that we published last month.

Tight junctions are made up of several different types of proteins, including Claudins, which are transmembrane surface proteins (the sea monster-type ones).

The type of claudin you find in the junctions depends on what type of cell you look at, and until recently it was thought that Claudin-5 was the main Claudin sticking our endothelial cells together.

However, there are members of the Claudin protein family whose functions are still shrouded in mystery – and if we’re going to properly understand how to target blood vessels to treat cancer, we need to understand the proteins involved.

So for part of my PhD, I focused on a little-understood Claudin called Claudin-14, which is known to be involved in a certain form of deafness.

But, we wondered, is it involved in blood vessel growth too? So far the only other clue came from research linking higher levels of Claudin-14 protein to the shape endothelial cells grew in.

Could we find out more?

When Claudin-14 supplies are at 50 per cent

To start to understand how Claudin-14 works, we worked with three types of mice:

  1. normal mice, which carry two copies of the gene that makes Claudin-14 (called ‘wild-types’);
  2. “heterozygous” mice (Hets) who have been genetically altered to carry just one copy,
  3. and “Null” mice, who have no copies of the Claudin-14 gene.

Comparing tumour growth and angiogenesis in these mice would allow us to work out what role – if any – Claudin-14 was playing.

First, we looked at the endothelial cell tight junctions in the blood vessels from tumours the mice had developed. Usually where cells stick together you can see nice neat lines of tight junction proteins:

Tight junctions in 'normal' epithelial cells

Tight junctions in ‘normal’ epithelial cells

But in blood vessels from tumours in Het mice, we found a lot of junctions were disrupted:

Tight junctions in endothlial cells lacking a copy of  the gene for Claudin-14

Tight junctions in endothlial cells lacking a copy of the gene for Claudin-14

We also saw that small, drug-sized molecules tended to leak out of the vessels more in Het tumours, and that low-oxygen areas were actually harder to spot in those tumours. That seemed strange, but on closer inspection, there actually seemed to be more blood vessels in the tumours from Het mice.

It was also apparent that a greater proportion of those vessels were closed, or seemed to be under-developed and non-functional. Could it be that there were more endothelial cells, but fewer working blood vessels?

To find out more, I grew some blood vessels from bits of the aorta (the big blood vessel that comes out of the heart) in plastic Petri dishes, and found that the Het mice indeed produced more. And when I grew individual endothelial cells in the lab, they divided more frequently too.

So it looked like only having one copy of Claudin-14 caused problems for the integrity of newly grown blood vessels, and more rapid growth, maybe in a kind of feedback loop.

This clearly shows a role for Claudin-14 in angiogenesis, and suggests that interfering with its function makes tumour blood vessels more ‘leaky’ (although it didn’t affect tumour growth overall).

What about the nulls?

But what about the mice that had no Claudin14? Surely their vessels would be even leakier, their endothelial cells grow even faster?

And here’s the paradox of Claudin-14. Because it turns out that, with everything I looked at, the nulls all looked the same as the wild-types. No effect. Nada. Zilch.

Why this would be the case – seeing an effect when one copy of a gene was missing, but nothing when both were – has left us completely stumped.

We assume it’s due to a phenomenon called compensation, where similar molecules fill in for the role of the one that’s missing (Like when you’ve run out of butter for the cake so have to use some butter-like imitation substance from that tub at the back of the fridge: It does the job, it’s not quite right, but it’ll do).

The go-to candidate for this compensation was Claudin-5, as we knew from the work of others that it could be found in endothelial cells.

But a colleague checked Claudin-5 levels in our samples, and they were the same in all three situations; whether Claudin-14 was present in two gene copies, one, or none.

So we still don’t know – but that’s the nature of scientific research, you put together a few small pieces of a much bigger jigsaw, laying some groundwork that perhaps someone else can build on in the future.

So what next?

I’m no longer in the lab, but now this research has been published, other researchers can access my results, and maybe one day there will be some answers to these “what the hell is going on?!” questions.

It’s nice to think that one day this research could be looked at by another scientist, facing their own challenging questions, and it might help answer them. Or even, perhaps, that this knowledge might one day benefit cancer patients, or even another condition we don’t yet know to be related.

It’s a long pipeline, and there are no guarantees, but even better than satisfying curiosities is the fact that I’ve played a part in the collective efforts of researchers and medical professionals, the likes of which have been key in improving cancer treatments and the lives of cancer patients, and will continue to do so in the future.



  • Baker M., Reynolds L.E., Robinson S.D., Lees D.M., Parsons M., Elia G. & Hodivala-Dilke K. (2013). Stromal Claudin14-Heterozygosity, but Not Deletion, Increases Tumour Blood Leakage without Affecting Tumour Growth., PloS one, PMID:


  • Tony Jones
    14 June 2013

    Tumor blood leakage and low oxygen supply are elements that can be addressed with alternative approaches possibly to enhance life expectancy of current subjects.


  • Tony Jones
    14 June 2013

    Tumor blood leakage and low oxygen supply are elements that can be addressed with alternative approaches possibly to enhance life expectancy of current subjects.