Diagnosis, medicines, recovery

Diagnosis, medicines, recoveryLSH-FES: Research to close the gap between fundamental science and the patient

‘Diagnosis, medicines, recovery’ reports on

the groundbreaking results achieved by the

six LSH-FES consortia: tEPIS, Cyttron II,

NeuroBasic, Virgo, DCTI and NIRM. The

activities of these consortia were carried out

from 2010 to 2016 and were partly funded

from natural gas revenues.

The fields of activity of these research consortia

were diverse, ranging from image sharing,

bioimaging and brain disorders to viruses, type 1

diabetes and regenerative medicine. Despite

their varied focus, the six consortia all worked

on the often-difficult transition from funda-

mental research to practical applications. After

six years of hard work, the results of these joint

initiatives are now clearly visible – many of the

developments can be applied immediately to

help patients or used in further research that

will directly affect patients.

Diagnosis, m

edicines, recovery

Amsterdam 2016

Diagnosis, medicines, recoveryLSH-FES: RESEARCH TO CLOSE THE GAP BETWEEN FUNDAMENTAL SCIENCE AND THE PATIENT

LSH-FESLSH-FES

NEUROBASIC – The next generation of neuropharmaceutical drugs

Developing medicines for brain diseases 44

NeuroBasic in a nutshell 46

The treatment of schizophrenia begins with understanding 48

Mice with epilepsy 53

Ingenious equipment makes brain research easier 56

‘It is not an automatism; you will be disappointed sometimes’ 62

VIRGO – Viral infections: better understood, better contested

Follow the immune response step by step 66

Immune system model improved by cleverly combining data 68

Virus killers: a great economic success too 70

The sniffling dromedary and the deadly virus 72

Booster for an exhausted immune system 76

Peeking through the keyhole: A cacophony of genetic activity 79

‘Innocent’ measles virus wrecks your immune system 82

Hospital admission or not? Your genes decide 86

Vaccines do not work when they are left on the shelf 88

LSH-FES

Introduction 7

Life Sciences & Health: knowledge development for vital citizens 8

TEPIS – Digital infrastructure for pathologic research

tEPIS in a nutshell 14

Pathologists enter the cloud 16

Comparing 16,000 photos 18

The challenge is to get the analysis right 19

CYTTRON II – Bioimaging technology for better diagnostic tools

Google Earth up into the cell: ‘That is how you look too’ 22

Cyttron II in a nutshell 26

Zooming in to the max without destruction 28

Slides are going digital 30

A molecular map of tissue 35

Cells in 3D 36

‘The challenge is the user, not the technology ’ 39

Content

LSH-FESLSH-FES

The heart still has some secrets to reveal 138

Repairing liver damage with the patient’s own cells 142

Barcoded blood cells 145

Are the regulations ready for regenerative medicine? 146

Partners

Health˜Holland 149

tEPIS 150

Cyttron II 151

NeuroBasic 152

Virgo 153

DCTI 155

NIRM 156

Colophon 159

Content

DCTI – Towards new cures for diabetes type 1

Type 1 diabetes: on its way to being cured 92

DCTI in a nutshell 94

Wanted: cells for insulin production 96

Assignment: culture millions of beta cells 100

Transplanting islets in an artificial pancreas 102

Pump keeps donor pancreases in shape 106

First tests versus the last straw 110

NIRM – Help the body to repair itself

When stem cell research and tissue engineering meet 114

Horses for courses – printed cartilage for joint repair 116

Design your own heart valve 119

A shove in the direction of bone or cartilage 122

Build a tailored test kit 128

Stem cells surviving on synthetic hydrogel 130

Reparation with stem cells 132

Never eating toast again 134

Make your own research material 137

LSH-FES

7

Introduction

Herman Verheij, L S H - F E S S E C T O R C O O R D I N A T O R

Presented here is a book describing the groundbreaking results that have been achieved in

the six LSH-FES consortia: tEPIS, Cyttron II, NeuroBasic, Virgo, the Diabetes Cell Therapy

Initiative (DCTI), and the Netherlands Institute of Regenerative Medicine (NIRM). LSH

stands for ‘Life Sciences & Health’, more precisely for the application of life sciences to

the benefit of our health, and FES for the ‘Fund for Economic Structure Enhancement’,

financed by the so-called natural gas profits.

These public-private partnerships started their activities in 2010, upon the Dutch govern-

ment’s approval of a joint proposal, which was submitted by a combination of more

than a hundred partners. These included many small and large businesses, multinationals,

academic and medical research groups, patient organisations, and the ministries of Health,

Welfare & Sport, Education, Culture & Science, and Economic Affairs.

The FES grant was used to establish public-private partnerships: every single euro invested

by the partners was matched by the government. This money was used to finance projects

that had already shown their strength in the LSH clusters of Diagnosis, Medication, Regen-

erative Medicine, and Technology & Infrastructure.

After six years of hard work the results achieved by these partnerships are clearly visible –

many of the developed technologies and products can be directly applied for the treatment

of patients or used for research that is very close to the patient benefit. The consortia have

thus proven that they have optimally used the research funds to contribute to the improve-

ment of our national public health and boost the economic activities in the Netherlands.

A vast number of these results are presented in this book. The most recent developments in

the various fields will astonish you time and time again. I therefore trust that you will

greatly enjoy reading this book!

LSH-FESLSH-FES

98

The Netherlands has an excellent knowledge base and top scientists. Their collaboration with busi-nesses, health care funds, patients, patients’ associ-ations, the government and other organisations is continually generating new solutions: better diagnostics, smarter medication, ways to help tissue repair itself and new services for the benefit of patients and care. Increasingly, preventive and therapeutic interventions are combinations of such solutions.

The consortia that present their results in this book are examples of public-private partnerships in research and development. Apart from contributions by the partners involved, the projects presented here were co-funded from the Dutch natural gas revenues via the FES Economic Structure Enhancing Fund of the Dutch government, within the framework of the LSH-FES programme. This programme, developed under the leadership of the scientific directors of four top institutes – the Center for Translational

Molecular Medicine (CTMM), Top Institute Pharma (TI Pharma), the Biomedical Materials Programme (BMM) and the Netherlands Genomics Initiative (NGI) – has helped build solid foundations for Life Sciences & Health as a recognised economic priority sector: a “top sector”.

The programme was carefully designed in 2009 on the basis of the idea that the Netherlands was ever so close to making the notoriously difficult step from research to patients in various healthcare disciplines, and that a funding programme would be just the right incentive to make this step. At the same time it would significantly strengthen collaboration between scientific researchers and the business community.

In the field of diagnostics, there are two consortia that have capitalised on public-private collaboration: tEPIS and Cyttron II. Both research programmes aspire to simplify diagnostics through

Life Sciences & Health: knowledge developmentfor vital citizens

The breakthroughs in Life Sciences & Health (LSH) are achieved in

increasingly rapid succession. No wonder, then, that this century is

already being referred to as “the era of life sciences”. Life sciences

greatly contribute to the improvement of people’s health.

story: Herman Verheij and René Rector

Introduction

LSH-FESLSH-FES

1110

It has generated knowledge-intensive jobs and new businesses, has boosted import and export and encouraged major investments by foreign companies. The combination of social, scientific and economic interests makes it more important to forge strong public-private collaborative partner-ships where these interests can be joined.

Innovative powers The Life Sciences & Health top sector partly builds on the successes of the LSH-FES programme under the flag of Health~Holland. One of the greatest social challenges that Health~Holland focuses on is how to achieve a future-proof healthcare system which ensures increasingly better health and health care, provides solutions to deal with the growing care demand and will still be affordable in terms of human resources and funds. The Life Sciences & Health top sector is an area where the innovative powers of the research community, the business community and the Dutch government combine to produce smart solutions that contribute to a single goal: vital citizens in a healthy economy. Examples

of such innovations described in this book include steps needed to eventually allow transplantation of the islets of Langerhans (for patients with diabetes type I), the identification of the MERS virus and the accompanying development of a vaccine.

Apart from public-private research projects, Health~Holland also focuses on important cross-sectoral and inter-sectoral themes. One example is the human capital agenda, which is to contribute to a future-proof workforce of well- educated and enterprising workers to meet future demand. Another example is the collaboration with other top sectors like Agrofood, Chemistry, Energy, High-Tech Systems and Materials, Logistics and IT, as well as with the Creative Industry (CLICK NL). Together with the latter, for example, Life Sciences & Health is working towards solutions for older people with dementia that enable them to continue to live independently in their home environment for longer.

Life Sciences & Health: knowledge development for vital citizens

technological innovation. tEPIS focuses on the infrastructure for integration of pathological data, Cyttron II on bioimaging techniques. One step fur-ther in the treatment process, doctors are con-fronted with the lack of good remedies for patients with various disorders. NeuroBasic PharmaPhenom-ics builds on the knowledge acquired in the prede-cessor of the research programme, NeuroBsik. In that programme, researchers came up with a method to actually test medicines for a wide range of brain disorders in models – which had been virtu-ally impossible until then. In PharmaPhenomics, the consortium has taken a further step, so that various very promising medicines against numerous brain disorders are now ready for clinical testing. Research consortium Virgo aims to further unravel the secrets of the immune system in order to identify exactly what happens upon a viral infection, and develop effective intervention strategies and vaccines.

For some disorders, the development of a good ther-apy in the conventional way remains problematic. These disorders require a revolutionary new way

of practising medicine. The Diabetes Cell Therapy Initiative (DCTI) and the Netherlands Institute for Regenerative Medicine (NIRM) are looking for alter-native therapies to treat existing disorders. Diabetes type 1 patients have nothing else to turn to but insulin injections. Through its attempts to improve pancreas transplants and the cultivation and trans-plantation of the islets of Langerhans, DCTI aims to boost these patients’ quality of life. NIRM, for its part, focuses on the cultivation of autologous cell material outside the body in order to provide solutions in situations where the body does not automatically repair itself. New businesses By identifying Life Sciences & Health as a “top sector”, the Dutch government has taken an essential step in the recognition of this sector’s contributions to both the economy and society. It is a unique sector which, on the one hand, contributes to new solutions for something that concerns us all – our health – and, on the other, has proved to be a sector of huge economic value.

Digital infrastructure for pathologic researchTEPIS

Pathologists examine tissue under a microscope. When they do

so together, large-scale research is possible. tEPIS takes away

the greatest hurdle in the collaboration: the logistics.

+

-

+

-

TEPISTEPIS Infographic

1514

3 A pathologist examines the tissue under a microscope. Others who also want to look at the slide (for instance, for consultation about a diagnosis or for a scientific study) have to rely on photos (which are fixed images) or a double microscope.

3 Using a scanner, a specimen is digitalised resulting in a 5 GB photo file.

4 The researcher now has a digitalised histological section. For remote collaboration, these images will need to be exchanged over the Internet.

5 tEPIS was to tackle two challenging problems: security and the exchange of the huge scan files over the Internet.

4 Examining together means: taking the slides with you. This practice is time-consuming and involves the risk of dama-ge to the research material; in large multi-institutional studies the logistics are a nightmare.

5 Storage obstructs the accessibility of the study material. Results from a study are usually limited to Excel files and photographic images.

story: René Rector image: Parkers

tEPIS in a nutshellPathologists examine wafer thin slices of tissue under a microscope.

They diagnose the patient on the basis of their observations. As long

as they do not have to consult other specialists this works out fine,

but for large-scale research or a second opinion, pathologists had to

rely on inefficient and cumbersome procedures.

1 A doctor takes a biopsy from a patient and sends it to the department of Pathology.

2 The biopsy is cut into wafer thin slices known as histologic sections. These end up on microscope slides and are often treated with dyes in order to enhance the resolution of certain cells or cell components.

New

Old

TEPISTEPIS

1716

technological development: the possibility of scan-ning microscope slides. Preparations are often smaller than a square centimetre, but new scanners (such as the one developed by the Cyttron II consortium; see Slides are going digital) are able to scan this square centimetre as a dazzling three-thousand-Megapixel image. Sample scanners render digital images which, unlike slides, cannot break, do not deteriorate and can be copied. But these image files are not only highly detailed – they are also extremely large.

Medical data“You cannot share this type of images over the Web in the conventional way”, Stathonikos explains. “What-ever the speed of your connection, quickly scrolling through a preparation is impossible.” The solution lies in digital technology, which makes it possible to only download the part of the image you want to see onto your computer, while the rest of the scan stays on the server. The tEPIS software calculates which part of the scan your computer should show, and at which mag-nification. Jan-Willem Boiten, project coordinator of the TraIT consortium responsible for tEPIS: “It’s more or less similar to Google Earth. If you start Google

Earth, you see your own country, depending on your location. If you zoom in, Google sends you increas-ingly detailed information. If you would like to have a better overview, you get a different set of data.”

Where the exchange of large files was mostly a technical problem, researchers also need the (anony-mised) detailed description of the corresponding patient. Because this entails some risks to patient privacy, clinical information is stored separately from the scans themselves. In this way, the risk that a hacker would be able to steal Mr Smith’s medical data from the cloud is reduced to a minimum. Stathonikos: “Whatever is uploaded to tEPIS cannot be traced back to patients – it is anonymized. How-ever, we inevitably had to come up with solutions for a wide range of security issues.”

IMA

GE

: P

HIL

IPS

Left: Storage according to conventional methods: in rows and rows of filing cabinets. Right: Preparations are now scanned using a hyper-precise image scanner.

“ In a practical sense, collaboration on a large scale is pretty difficult for pathologists.”

Traditionally, looking at a preparation together with a colleague was only possible using a multi-head microscope: a microscope with two sets of looking glasses. Quickly emailing someone a digital scan was not an option: until a few years ago, there were no scanners that could scan an entire preparation in suf-ficient detail. A photo of a preparation could easily be 5 GBs, which took ages to attach. Besides, everything connected to the Internet is vulnerable to information leaks, and the information in tissue samples is highly privacy-sensitive.

3,000 Megapixels“Security and file size were our greatest challenges”, says Nikolaos Stathonikos, IT specialist at the pathol-ogy department of University Medical Center Utrecht and project leader for the tEPIS project. The project, a joint venture of the pathology departments of six university medical centres and Philips, aspires to achieve what was impossible until now: examining preparations from a distance, sharing information and in this way greatly simplifying the logistics of large-scale pathology research. tEPIS, the TraIT Enhanced Pathology Image Sharing system, is based on a recent

story: René Rector image: Gwen Dackus

Introduction

Pathologists enter the cloud

In a practical sense, collaboration on a large scale is pretty

difficult for pathologists. The objects of their examinations

are wafer-thin slices of tissue under a microscope. If you

wanted to look at them together, you had to sit side by

side – until recently. Pathologist can now enter the cloud. IMA

GE

: W

ILL

Y V

AN

BR

AG

T

TEPISTEPIS

1918

How can you get a computer to recognise whether tissue

under the microscope contains cancer cells? Researchers in

“digital pathology” have a solution: they develop algorithms

so that the computer knows what to look for. The trouble is

that the different algorithms are difficult to compare.

It certainly is a problem in pathology: in order to make cells

or parts of cells visible in a tissue sample, you colour them

with special dyes that attach to specific structures – but

different pathologists colour their samples in different ways.

This is fine for diagnostic purposes, but if you want to com-

pare preparations, for example for large-scale research, this

presents a problem: features coloured a deep purple by one

pathologist are coloured a light shade of pink by the next.

The trained human eye has no difficulty reading slides that

have been coloured differently, but automated interpretation

of images is an important development in pathology, and

computers are not as good at reading different slides. All the

same, biomedical image analysis research has seen plenty of

publications following each other in rapid succession in which

researchers claim to have developed an algorithm that knows

how to properly interpret microscopic images. “Once verified

by other research groups, the results of many of these publi-

cations turn out to be highly dependent on the treatment of

the preparation”, says Jeroen van der Laak from Radboud

University Medical Centre, “whereas you actually want an

algorithm for general use.” That is why, in the pathology

field, so-called Grand Challenges are organised for samples

of different tissues: challenges that dare the participants to

write an algorithm for a given set of samples that is both

reliable and efficient. The set of samples is assembled from

different sources. It sounds like a game, but it is very useful:

this way, the best algorithms come to the fore – which takes

the field a whole lot further towards a solution.

Van der Laak’s group started a special challenge in November

2015. The goal of this challenge is to develop an algorithm

that, on the basis of lymph node tissue from breast cancer

patients, can indicate to pathologists where the cancerous

cells are located. What makes this challenge special is that it

is the first to provide fully scanned microscopic preparations

by offering all participants direct access to the images stored

in tEPIS.

Van der Laak: “It would be possible do without, of course,

but besides a logistical advantage we now also have a

good interface. This simplifies software writing considerably,

especially if you don’t know what pathological images look

like. This is particularly useful for the IT specialists whose

help you need in a challenge like this one.”

Pathologists enter the cloud

IMA

GE

S:

WIL

LY

VA

N B

RA

GT

Left: You can still collaborate sitting behind the computer together, but you might just as well be a thousand miles away.Right: Collaborating according to conventional methods.

The challenge is to get the analysis right

If breast cancer is diagnosed in young women, the diagnosis is

usually followed by surgery, radiation therapy and additional

treatment, often chemotherapy. But is the latter always useful?

The Young Breast Cancer Project (YBCP) is seeking an answer

to this question. “This would be a logistic nightmare if it hadn’t

been for tEPIS.”

Chemotherapy administered after the removal of a breast

tumour has nasty side effects, such as osteoporosis and

menopausal symptoms. “These side effects are extra dismal for

young women. They often still have a desire to have children,

and if they had not had this disease, they would still have had a

life expectancy of at least another 45 years”, says Gwen Dackus

from the Netherlands Cancer Institute. She is coordinating a

large-scale study in which pathologists are trying to find out if

there is another way.

“In the old days, twenty per cent of breast cancer patients used

to receive follow-up treatment with chemotherapy. The protocol

changed in the 1990s, and eighty per cent of patients is treated

with chemo ever since. But the mortality rate of patients as a

consequence of their disease has fallen only very slightly.” Dac-

kus thinks this decrease is so small because many women do

not benefit from chemo after the tumour has been removed.

These patients are unnecessarily burdened with side effects.

“We would rather give them tailor-made chemotherapy.”

The side effects are so serious that it is worth looking into. The

study cannot include patients that are currently undergoing

treatment: they virtually all receive chemotherapy, whereas the

researchers would like to know what happens if they do not. For

that reason, the YBCP works with stored tissue samples from

2,500 patients who were diagnosed with breast cancer between

1989 and 2000 and who were younger than forty at the time.

Five to ten new preparations were made out of each sample:

16,000 slides in total. “This would be a logistical nightmare if it

hadn’t been for tEPIS. If you wanted to make 16,000 slides

logistically manageable, you would have to analyse them with a

very small team, as you would have to physically send them the

slides. This study is not part of the pathologists’ daily work; it

comes on top of their daily workload. No one can work on it

full-time, so the analysis would normally take years”, Dackus

continues. However, large-scale research is now possible thanks

to the new digital technology. The samples have been digitalised

and can be viewed online with tEPIS technology. Dackus is now

working together with almost twenty pathologists, from Italy to

regional hospitals, who analyse the digitalised samples accor-

ding to the time they have available.

Follow-up treatmentDackus also collates the results of the analysis digitally. Once

the analysis will be completed, insight into the needs for

patients to be treated with chemotherapy will emerge. Dackus:

“Medically speaking we know what happened to these women.

Whose cancer metastasised or returned? Whose did not? We

want to know the difference between these two groups, both

pathologically and genetically. This way we will be able to pre-

vent overtreatment and reserve chemotherapy for women who

will benefit from it.”The pathologist uses colours to annotate whatever he sees on the screen.

IMA

GE

: G

WE

N D

AC

KU

S

Comparing 16,000 photos

Many deseases are caused by a defect at the molecular level.

Therefore, improved imaging at each level – from molecules

to cells to tissue to organs – is essential for healing.

Bioimaging technology for better diagnostic toolsCYTTRON II

CYTTRON IICYTTRON II

2322

The collaboration between LUMC and Naturalis was a project of the Cyttron II consortium. It wants to gain insight as to how molecular changes affect body and health in order to be able to develop new diagnostics and treatments. The data of different imaging tech-niques are brought together for this purpose into one single database that can then provide a full 3D over-view at the molecular level via the cell to the organ and the entire organism. The Cell Zoomer is an illustrative example of this quest for the “complete picture”.

Human, mouse or fishWhy is Naturalis exhibiting the Cell Zoomer? “This is the ‘Research in progress’ gallery. We wanted to introduce the micro world to the general public”, Van Zeeland explains. “The deeper you zoom in, the more all organisms look the same. ‘This is how you look too’, is what we want our visitors to know. It does not make a difference whether you put a human being, mouse, fish or any other animal under the

microscope. Our cells look almost identical. Making it physical by zooming in – a principle that we all know – we want to create understanding as to how small it all is.”

Light-sensitive barsYou can zoom in on any random spot in the fish. A few blue dots – “hotspots” – indicate the things for which clarification is provided. Van Zeeland zooms in on the light-sensitive bars in the eye. “… you also have such cells in your eyes”, the voice from the screen says. According to Van Zeeland, the Cell Zoomer is also very suitable for education. “In medical and biology textbooks everything is usually arranged by organ. Here, you can pay an anatomical visit to, for instance, the area of the brain and then zoom in. If you ask me, every university should have something like this for their biology courses.” LUMC is developing a variant of the Cell Zoomer that could be used in practicals.

Zoom in on an adult zebrafish.

“Touch me”, the screen suggests. If you do so, all those fish make way for one zebrafish and an embryo. You can then, by hand, zoom in further and further on a cross-cut of the embryo. It will soon fill up the entire screen. You can see its organs, the brain, muscles, the eye – all in black and white. You can even see a bacterium as big as a marble in the intestine. In comparison with the bacteria, the intestinal cell is colossal. If you zoom in even further, you can have a peak deep inside these cells. You can see the mitochondria (the energy suppliers), the nucleus with DNA, and the small balls called ribosomes that doctor together the proteins.

44 football fieldsAn illuminated, round window next to the screen will show you zebrafish embryos: black dots with a transparent tail, not even 1.5 millimetre in length. “In order to look into cells this deep, you have to enlarge these embryos five hundred thousand

times”, Ilse van Zeeland, content developer for the exhibition, explains. “In this maximum enlarge-ment, you have zoomed into a photo of 800 to 400 metres, about the size of 44 football fields.”

Thirty thousand photosThe photo itself was made at the Leiden University Medical Center (LUMC). The department of Electron Microscopy put the zebrafish in a moveable container under the electron microscope. In an automated process, they made thirty thousand photographs from head to tail in 4.5 days, while tiny precision engines shifted the container with the fish in two directions before taking the next picture. Software with a “stitching” algorithm developed within LUMC carefully pasted the thirty thousand images together to make up one digital mega photo. The main touch screen and the zoom software were developed at Naturalis.

story: Joost van der Gevel image: Naturalis

Introduction

Google Earth up into the cell: ‘That is how you look too’

On a wall on the top floor of the Leiden Naturalis Biodiversity Centre – accompa-

nied by the wolf of Luttelgeest, Herman the bull, a dodo, and the skeleton of the

Plateosaurus – is a huge television screen with calmly swimming fish. This is the

Cell Zoomer. It explaines in a nutshell the activities of the Cyttron II consortium

in past few years.

CYTTRON II

d

c

CYTTRON II

a

b

e f hg

i j lk

2524

a Swimming zebrafishes. b Zebrafish embryo. c Zebrafish embryo. d Light-sensitive cells in the eye.

e Brain cells. f Cartilage cell. The tiny black ball is the cell core. g Intestinal cells, with finger-shaped extensions. h A bacteria in the intestine.

i Muscle cells. j This also looks like a cell, but what you see is merely k The small balls are ribosomes. l Mitochondria. a part of the cell: the cell core with DNA.

Google Earth up into the cell: ‘That is how you look too’

From big to small

CYTTRON IICYTTRON II Infographic

2726

“Zooming in”It would be ideal if you could zoom in on organs up to molecules

in a patient’s tissue, but the required technology has yet to be

invented. What you can do, however, is

combine software so that different

techniques together give you a

coherent collection of images.

AutomationSome of a pathologist’s duties

are old-fashioned craftsmanship.

A sample machine that directly

scans, colours and begins an

analysis of the samples, makes

the work considerably more

efficient.

Combining images Not everything that you would

like to see can be captured in

a photograph. If a mass spec-

trometer makes an analysis and

the images are combined at the

same time, researchers will acquire

far more information about the object:

they can see where cells are dividing,

for instance.


Looking at detailsX-ray diffraction focuses on

molecules’ atomic structures,

which you first have to

crystallise. This technique

does not work, however, for nanocrys-

tals (smaller than one thousandth of

a millimetre). One Cyttron II result is

that the structure of nanocrystals can

now be made visible through electron

diffraction.

Seeing more with less “light”Biological molecules are visible under an electron microscope.

You are “looking” by way of a bundle of electrons. These

electrons, however, damage whatever you are looking at.

The solution lies in a new detector that cuts down

on the number of electrons that you need

to create an image.

3DTissue samples are usually seen in

two dimensions. A new scanning

technique makes it possible to also

make a 3D scan at the tissue level.

Cyttron II in a nutshellThe cause of diseases can often be found at the molecular level.

The ability to zoom in on different structures - molecules, cell

organelles, cells, tissue, and organs – is extremely helpful in

setting a diagnosis. Unfortunately… there is no such thing as a

megazoomer. Taking a better look, however... that is possible.

CYTTRON IICYTTRON II Case study

2928

quite easily be solved by cryo-electron microscopy: freeze the object until it is frozen in place. If your sample cools down to -196 degrees Celsius quick enough, you will not get any annoying ice crystals that destroy your sample. The second problem is more difficult to solve. Biological structures are extremely fragile, so it is best to use as low a dosage of electrons as possible but enough to collect all the information you need.

In order to solve this problem, Cyttron II research partner FEI developed in cooperation with Leiden University a new camera as an additional fixture to the Titan cryo-electron microscope. The camera, named Falcon, creates a better image of the object without any need for an increased dosage of elec-trons. “It basically works the same as any other camera – but with electrons rather than light”, De Jong says. Electrons that travel through the specimen land on a chip in the camera; the chip is divided into pixels. Normally, this would cause quite some interference; electrons, which are full of energy, are difficult to bring to a halt. They scatter. If they do, they might not mark just one, but multi-ple pixels. The researchers have solved this problem in the Falcon with a very thin chip. De Jong explains: “Electrons pass through it and leave a sign that lets us know that they have been there. This way, they scatter far less.” Moreover, the camera is cooled, which takes away a considerable amount of interference.

Movie-modusThe camera is quite an asset to researchers, so Marin van Heel, Professor of Data Analysis for Cryo-Elec-tron Microscopy at Leiden University, confirms. “With the camera’s high sensitivity we need four times less electrons to get a good image. Another advantage is the ‘movie mode’: you can make ten pictures per second of your object. If you play these images in sequence, you get a far better view.” Moreover, it allows you to see “the whole

soup”, Van Heel says. “The molecular machines in your specimen have different functional states. An enzyme, for instance, may have bound itself to a substrate in order to accelerate a response, but it might also have already returned to its original state. You want to be able to see all these different statuses in 3D.”

A direct application of this new technology is the development of medicines. Van Heel: “Take antibiotics. You can precisely see how a substance blocks something in a bacterium. This makes it possible to develop far more specific medication.” Falcon can also help to explore a virus like HIV in greater detail, so it will also add to our knowledge of how such viruses behave.

IMA

GE

: JU

LIA

N V

ON

DE

R E

CK

EN

/N

ATU

RE

Titan electron microscope with a built-in Falcon camera

Reconstruction of an actin microfilament, a modular structure of proteins that forms the cytoskeleton of a cell. The image is based on data collected by a Titan electron microscope with a Falcon camera.

Structural biologists are working on unravelling protein construction and other complicated struc-tures like ribosomes or viruses. After all, the role of a substance is in its design. Proteins for instance, the workhorses of cells, are ingeniously folded in a way that matches their duty. In order to understand their specific function, you will have to know precisely what they look like. The same goes for viruses – and even medication.

Thick aluminium foilThese substances cannot be seen with the naked eye. They are also too small for a light microscope. Using X-ray radiation would be an option, but that requires a crystal structure of the object that you want to study. Not every material lends itself for it. Another option is the use of an electron microscope. This apparatus uses electrons to make a projection of an object. A bundle of accelerated electrons is fired at the object in a vacuum tube. Electrons that pass through the specimen are greatly enlarged, like slides in a projector, and subsequently registered and recorded into an image by a camera. This renders a 3D image of an object, which is essential

if you want to see the folding of a protein or the wrapping of a virus.

In recent decades, electron microscopes have made examining our smallest building blocks quite a lot easier. Technology is making rapid progress: the smallest detail that you could see just ten years ago was ten times as big as it is now. Nowadays, an electron microscope allows you to examine an object up to a detail of a few tenths of a nanometre; one nanometre is the size of five atoms. For compar-ison: a bacterium measures some two thousand nanometres, and aluminium foil is ten thousand nanometres thick.

Thin chipsHowever, the use of an electron microscope for the examination of biological structures does have two downsides. “First, the molecules move in all directions at body or room temperature. Secondly, the objects are easily shot to pieces by the elec-trons”, says Frank de Jong, Director of Partnerships at microscopy engineering company FEI (Field Electron and Ion Company). The first problem can

story: Rineke Voogt

Zooming in to the max without destruction

If you want to zoom in on proteins, DNA or any other small structures, you will

have to use an electron microscope. Electron microscopes are used to observe

in detail. However, there is a problem: the more detail you want, the more

electrons you will have to fire at your specimen – and the greater the chance

of these electrons destroying the structures that you are about to inspect.

CYTTRON IICYTTRON II Interview

30 31

Slides are going digital

The entire world is going digital. Super-fast computers and other smart

devices are the order of the day, but that is not the case at the pathology

department of a hospital. The pathologist types his report on the computer,

but other than that, his desk is filled with a microscope and a great many

slides with preparations that need to be checked.

story: Elles Lalieu

The Philips IntelliSite system: a slide scanner (photo) with an image management system and software.


3332


to very precisely compare two different dyes of the same area. Verhagen demonstrates. On the first slide, the cell walls have been dyed pink and their cores are bluish purple. “Such dyes reveal whether you are dealing with cancer”, he explains, “for if so, the tissue shows a proliferation of cells rather than a nice arrangement. Areas that seem to colour bluish purple overall mark the tumour tissue.”

On the second slide, a brown dye was used to reveal whether a patient is receptive to hormone treat-ment. In this example, the bluish purple areas on the first slide are now brown. This means that this patient has tumour cells that are receptive to hor-mone treatment. Verhagen: “This is important infor-mation, for such chemotherapy (with Herceptin, for instance) may have consequences for the func-

tioning of the heart, so you do not want to just administer it to a patient who does not benefit from it.”

Central scanningComparing and connecting images with seperates slides is not possible. This is one up for the digital pathology system. It also has another important plus: it is easy to share the images. “There are microscopes that allow you to check the same preparation together with up to ten persons at the same time, but this still requires everyone to sit around the same table”, Verhagen says. Part of the LSH-FES research project tEPIS involved the quest for a solution that made it possible to access stored images without physical boundaries. The only thing required now, is an internet browser.

Screenshot of a cell with Herceptin colouring. When they compare this image with a standard dyed cell, pathologists are able to establish immediately whether a patient’s tumour cells are receptive to hormone treatment.

Pathology plays a vital role in establishing a clinical picture, especially in diagnosing cancer. Suspect tissue samples are examined with a microscope in order to determine whether the tumour tissue is malignant or not. The outcome of this examination is the basis for the pathologist’s subsequent recom-mendations for the patient’s treatment.

“The microscope is, and will always be, a great piece of equipment, but there are downsides to working with slides on glass”, Dirk Verhagen, researcher and developer at Philips explains. “You can lose images; slides break or preparations get lost when you send them to a specialist for consultation.” In order to make the storage and sharing of patho-logical images easier, Philips came up with the IntelliSite digital pathology solution, developed within the Cyttron II research program. The system comprises a slide scanner, an image management system and software whereby the pathologist can share and assess the images.

Respectable data volumesAt the Philips visitor centre in Best, Verhagen demonstrates how the system works. He places a slide in the scanner, closes the door and presses “start”. About sixty seconds later, the scan is ready. It took the first scanners thirty minutes to scan one slide. This scanner can process three hundred slides within five to eight hours. “This makes scanning for pathological examination more attractive”, Verhagen believes.

Not only the duration of the work itself has fallen dramatically, the storage of the data that comes out is now also affordable. “The data files vary from 0.5 to 4 or 5 gigabyte (GB), depending on your

object. That is quite a lot of data that has to be stored somewhere. This storage is now reasonably priced. Sometimes it is even cheaper than storing slides in an archive, as they do now”, Verhagen emphasises.

“Sliding” through tissuePhilips’ imaging specialist Jelte Vink was responsible for the developments done to the digital pathology system. “The device already existed. What we have primarily done during the past four years was updat-ing the software in order to make it possible to put images against one another or to inspect different colourings at exactly the same spot of a prepara-tion”, he explains. Verhagen can demonstrate it immediately with images that were saved on the computer. Suppose the doctor has taken a sample of tissue from a patient and would like to know whether it is a tumour and, if so, how this tumour will grow. The pathologist cuts the tissue sample in very thin slices and prepares each of these on separates slides. When he examines these slides one by one, he will be able to tell more about the tumour’s location – whether it is located at the surface, for instance, or deep in the tissue.

“Once the slides have been scanned, you can put the images next to one another and also make con-nections between them”, Verhagen continues. “You can virtually paste the different slices back together again and so ‘travel’ through the tissue. You can thus tell with far greater precision if a tumour is growing deeper, for instance.”

Sensitive tumour cellsIn the preparation of slides, each tissue slice gets its own dye. Each of these dyes helps establish different things. The computer now allows you

CYTTRON II Case studyCYTTRON II

34 35

?

You can identify molecules with a mass spectrometer. You can see structures in the tissue with a light microscope. When you combine the two, you can make a molecular map of the tissue that exactly shows where certain molecules are located and how they are affected by diseases.

This technique is called mass spectrometry imaging or MSI. It is a useful technique for all disorders involving molecular changes. You can see the location of neurotransmitters in mice brains, for example, or the distribution of certain molecules that could serve as medication. It also offers more diagnostic possibilities - for cancer, for instance. At first glance, a dividing cancer cell looks similar to an almost dying cancer cell, but their chemical contents are quite different.

“Our system analyses the chemical content of tissue with a mass spectrometer and couples this analysis with microscopic images of the same tissue”, Liam McDonnell from the University Medical Centre Leiden explains. “This gives us good insight into the exact location of molecules in the tissue.”

Within the scope of Cyttron II, McDonnell and his team primarily worked towards the automation of the scanning process. The major advantage of the robot that they developed is its speed. “A few years ago, you had to load every patient’s material into the mass spectrometer by hand. The mass spec-trometer is now equipped with a ‘hotel’ where ninety samples can wait their turn. The technology is therefore available to a great many patients”, McDonnell says.

This is important, for molecular changes can be helpful in setting diagnoses or determining prog-noses. By examining molecular compositions in tissue individually, the treatment can be better attuned to each patient. This makes this combined scanning technique considerably easier to apply in practice.

A molecular map of tissue

Cross-section of the human body by mass spectrometry imaging.

story: Elles Lalieu image: Imabiotech


When in doubt about a preparation, you can directly consult a pathologist on the other side of the country. That is quite revolutionary.

If you can share the images, the preparation of slides can also be centralised. When the images can be accessed everywhere, there is no need for each and everyone have their own slides. According to Verhagen, this centralisation is cost-effective; it could save tons or millions of euros. “The lab scans them all and a pathologist can be allocated per case. There are 15 to 25 specialisations within the field of pathology, but it is impossible for any hospital to house them all. If the images can immediately be sent to the right specialist, you will get sharper diagnoses. You can see this already happening in veterinary pathology”, he adds.

Philips has already sold quite a few digital pathology systems. They are being used in the Netherlands, Belgium, Germany, Great Britain, Austria, the United States, and Singapore – for both diagnostics and research purposes. “A good example of diagnostics is Hengelo”, Verhagen says. “They have a lab there, where they scan everything, so they now have a number of scanners.” He has noticed an increasing interest in scanning in recent years anyway. “Even people who are devoted to their microscope are now saying: ‘The question is not whether or not it disappears, but rather when’.”

Sorting imagesThe digital pathology system already surpasses the microscope, but the equipment is destined to become even smarter in the future. Researchers are extremely busy digitalising hundreds of thousands of old slides from archives. “We already know what happened to these patients, and computers can learn from that”, Verhagen explains. “Who knows, it might even turn out that adipose tissue grows differently for a certain type of tumour. The pathologist will probably have never looked at it in this way because they have never been able to retrieve any examples of the same type of tumour for comparison.”

The computer can also make a pathologist’s job easier, for instance when they are looking into metastases. Verhagen: “The pathologist removes a piece of a lymph node and divides it over ten to fifty slides. You can check them all but, in principle, one single ‘hit’ suffices to know that there are metastases. We could leave it up to the computer to preselect the images, so that the pathologist is shown the most ‘suspect’ ones first.”

The speed of pathology scanners has tremendously increased. The first scanners took thirty minutes over one slide. Nowadays, they process three hundred slides within five to eight hours.


36 37

Report

Within the field of pathology, scanners are being used nowadays. These are usually scanners that make a 2D image of the tissue samples. 2D images suffice when you are looking at thin slices, but for thicker slices of tissues 3D imaging is an absolute necessity. Philips and the Delft University of Technology are now working on a scanner that can make a 3D image of tissue all in one go.

Even before the start of the Cyttron II research project, Philips developed a special sensor. “This sensor made it possible to record different depths simultaneously”, says Bas Hulsken, digital pathology specialist at Philips. “Existing scanners made a number of 2D pictures and used these to assemble a 3D reconstruction. This new scan-ner makes it possible to make that 3D picture in one go. This makes the machine considerably faster than the scanners that are now available and this is to the benefit of patients who are waiting for a diagnosis.”

The 3D scanner will particularly be helpful within cytopathology: diagnostics at the cell level. Hulsken even uses the word necessity. “In cytopathology, you are often looking at liquid samples, such as entire cells that have been dissolved in a liquid substance. Such samples may be tens of micrometres thick, whereas the slices used in histology are often only five or six micrometres thick. You will get a pretty good focus on such a thin slice with a 2D scanner, altough a 3D scanner will probably also provide a better picture quality here. But as for thick slices, the 2D scanner can only provide a sharp picture of a limited number of cells.”

On the picture: PhD student Mojtaba Shaker (Delft University of Technology) is analysing images made by the 3D scanner.

Cells in 3D

story: Elles Lalieu


3938

Proces report

If you search for “Jaguar” in Google Images, your screen will be filled by pictures of black panthers, spotted felines, cars, airplanes and ships. Whenever someone starts digitalising images, they will be faced with a problem: how can a computer make adistinction between their contents? More important: how does it know what the user is looking for exactly? Fons Verbeek at the Leiden Institute of Advanced Computer Science of the Leiden University sought and found a solution.

story: René Rector

‘The challenge is the user, not the technology ’

Cyttron I

2006-2010

The idea behind the project is that a medical researcher can make more and better connections by linking images at different zoom levels and biochemical informa-tion. It would be even better when its smart algorithms could help point out any connections that researchers might not have made themselves just yet. This is, however, easier said than done: a cell organelle is completely invisible in an X-ray and you can barely see its outline under a light microscope. You will have to use a fluorescence and confocal microscope to get something close to useful. The alternative would be to use an electron microscope, but its magnification is often so strong that you end up looking at just a part of the organ-elle. This is not the only problem, though. “Black” has different

meanings in different cases, whereas “purple” depends on the dyes that have been used. And some microscopes only render black-and-white images. We initially tried to create some order in this using techniques that are similar to those used for facial recognition in photos. Unfortu-nately, this didn’t work out. Of course, if someone brings a box full of glasses with bowel tissue

samples with the same dye, a computer will be able to make smart suggestions after the third specimen. The problem starts when you bring a new box of glasses with an ever so slightly different dye.This had to change. Away with the pixels! We must use language! We realised that we should learn the computer one way or the other what the vague pile-up of purple

IMA

GE

: C

OL

IN B

UR

NE

TT

IMA

GE

: S

AC

SC

OTT

LE

WIS

/M

OD

Cells in 3D

The 3D scanner can also play an important role in special measurements. Hulsken gives the example of fluorescence. It is possible to attach fluorescent markers to genetic material (DNA or RNA). These markers are recognised during the scanning process and the pathologist can then count the number of markers in the picture and thereby find out more about important changes in the cell’s genetic material. “These markers might be on top of one another, in which case you cannot see them all with a 2D scanner. When you use a 3D scanner, however, there is no such problem since you can look at the sample from all angles”, Hulsken explains.

Within Cyttron II, Hulsken is working together with Sjoerd Stallinga, an imaging specialist at the Delft University of Technology. They devoted a considerable amount of time making the scanner suitable for digital pathology. “A scanner normally scans in three colours (red, green and blue) because this suffices for a coloured picture that is to be seen with the human eye”, Stallinga explains. “After all, each colour that one can observe, can be made out of the colours red, green and blue. The machine can now scan up to five colours. The com-puter shows things that are invisible to the human eye. This addition is to provide additional information that is relevant to the patients’ diagnoses.”

The research stage for the 3D scanner is now over. The first series of test models is expected to leave the factory early 2016. Hulsken and Stallinga want to use the rest of the year to get rid of any teething problems. The machine should then, hopefully, be ready to be shipped to pathology departments and research labs in 2017. Hulsken is not focusing on specific disorders. “Our design takes into account that the scanner should be widely useable”, he says, “for we want to eventually digitalise the entire field of pathology.”

“ The computer shows things that are invisible to the human eye.”


4140

‘The challenge is the user, not the technology’

limit the images that appear and thereby also affect the time required for the annotation.The same applies the other way around: if Janssen focuses on stress response in his brain smears, the database recommends images that are also about stress response. That is why we now know that, for instance, stress response in yeasts at the molecular level is virtually similar to our own.

2014

A breakthrough: not so much in the technology as in our delib-erations on adequate entry. Say, researcher Janssen logs into the database. We know that Janssen is conducting research into brain stem cells, as Janssen has indicated so in his personal profile. We also know that Janssen would like to annotate a set of images that has been produced with a light microscope because he has done so before or because he tells the data-base. These two aspects – Janssen, light microscope – are two consid-erably restricting factors for the set of ontological elements that Jans-sen is able to annotate: he cannot see any proteins, for you cannot see them with a light microscope. He doesn’t see muscle cells either, because we know that Janssen is involved in brain research. The imaging technique and the re-searcher – two obvious aspects –

20??

It all sounds so nice: a computer that makes the connections that researchers might overlook. How-ever, there is still a last bastion of the human aspect for us to conquer, and that is “jargon”. If a structural biologist refers to an “atlas”, he means something different than a topographer or an orthopaedist. We are still looking for a way to prevent researchers from annotating according to their own jargon without it becoming clear which jargon it is. In this sense, Cyttron II’s success will be determined by the eventual unambiguity of the annotation. Therefore, even if Cyttron II now comes to its end, its mission has not fully been accomplished.

“ Who would annotate two hundred photos one by one?”

Cyttron II

2011-2012

Not the images themselves, but the tags allocated to these images are therefore leading in the clarifi-cation of their interconnection. Say that you want to study the ribosomes in a tissue sample, the term “ribosomes” will help you zoom in until you are looking at the protein synthesis at the molecular level – or zoom out until the cell content is so vague that it can no longer be viewed in parts.We studied gene ontology, an existing ontology for genetics. We soon found out that this was certainly useful for us, but it also raised a problem: we had to annotate each image separately. It is manageable if there are only a few images, but it gets difficult if you have to translate two hundred photos in an exten-sive ontology. Which researcher would be willing to do that?Our project changed from a technical to a human problem – and, as always, this would be our greatest challenge.

2013

We tried to add a game element – literally. We built the amusement arcade game Frogger into the user interface, whereby hopping to the other side (which was the inten-tion of the Frogger game: help a frog across a river by hopping from log to log) becomes simpler if the terms that serve as these logs have been annotated. We hereby taught the user that annotation is useful and pays off and encouraged researchers to annotate properly by rewarding them with a game afterwards.

at the top left of the image actually represents, so that it can draw a researcher’s attention to the pres-ence of cell festering. We decided to start experimenting with an ontological description: a set of descriptive terms with not only a definition of the meaning of the terms, but also a description as to how these meanings relate to each other. Ontological descriptions are already used for quite some time. We know, for instance, that “leaves” are a fixed item in the collection of “deciduous trees” in the category of “components”, and that “oak”, “elm”, and “ash” are also in the same collection, but in the different category of “species”. Ontologies have clus-ters, collections, and a hierarchical structure.

The next generation of neuropharma-ceutical drugsNEUROBASICPHARMAPHENOMICS

The development of drugs for brain disorders has always fallen short

from definite success. Good animal models and far greater insight in

pathological processes open the door towards better treatment.

NEUROBASIC

€

NEUROBASIC

4544

only country where so many scientists involved in so many different disciplines are so close together. We have people with a more than considerable understanding of, for instance, behaviour, brains, nerves, and the limbic system as well as the software and technology that we need to rapidly screen mice without causing them any stress. There is nowhere else in this world that you can get so many experts together, not in the academic world nor in the phar-maceutical industry.”

It was a start. With sick mice now in hand, the great challenge for NeuroBasic was to develop effective ingredients for medicines. This was followed by tests on lab animals, so that it became clear whether the medicine-to-be was doing what it should. The next step was to gain some insight into why a medicine worked, or did not work, in mice. Some researchers had to go back to the drawing board.

The last step which was taken before the clinical tests began was possibly the most exciting. Being able to treat an epileptic mouse is one thing – whether it also works in humans still remained

the question. “We people have similarities to mice. However, we had already seen that even if a com-pound affects the synthesis of one specific protein and this synthesis is the same in mice as in humans, there are still differences in what the body does with the medicine or the protein”, De Zeeuw clarifies. That is why the researchers will remain in suspense until the very last clinical trial. However, De Zeeuw believes that NeuroBasic’s chances of not only bringing lab animal tests to a new level, but also

actually yielding medicines against autism, neuro- fibromatosis1, tuberous sclerosis2, epilepsy, and schizophrenia are considerable.

1 Neurofribromatosis is the body’s hereditary inability to make neuro- fibromin – with a wide range of symptoms, varying from café au lait spots to learning problems.

2 Tuberous sclerosis is usually a genetic mutation, with wide-ranging consequences such as epilepsy, autism, a mental handicap, skin disorders, and kidney diseases.

Medicine********

Developing medicines for brain diseases

Brain diseases score far below cancer and cardiovascular diseases in the

mortality statistics. Their impact is, therefore, often underestimated. The

effect of illnesses, such as depression, epilepsy, autism, and schizophrenia,

on the quality of life and the costs involved in their treatment and the sick

leave that they cause are, in fact, enormous.

Apart from their underestimated impact, these diseases have something else in common: there is no medication for them. However, if it is up to the NeuroBasic consortium, that is about to change. The consortium developed a method whereby chemicals can be specifically developed for the suppression of epileptic fits or combatting schizophrenia.The main problem with brain diseases was that medi-cines could not be tested in any way. A great number of ethically irresponsible experiments have been conducted throughout history, where people were used as guinea pigs. Pharmacology uses animal lab tests as a rule, but as no one could say whether a mouse suffered from schizophrenia or not, let alone if it could be induced in them, developing a medicine for a brain disease was a perilous undertaking.

“In NeuroBasic’s predecessor, we developed a method to make mice suffer from the very specific genetic defect that leads to a brain disease in humans”, Chris de Zeeuw, Programme Director and Professor

of Neurosciences explains. “For each of the brain diseases, we then very carefully recorded the mice’s aberrant behaviour. Such observation and recording is vital. For diseases such as cancer or cardiovascular disorders, it is often at the cellular level that you can see the effects of your medicine and whatever is going right or wrong. For brain diseases, you will have to deduce it from the lab animal’s behaviour.”

Considerable innovations were required before the mice’s behaviour could actually be recorded: a method to switch genes on and off as necessary, tests wich measure accurately the extend to wich a subject is suffering from the condition which should be caused by the genetic mutation, and a registration system that could record and analyse the behaviour of hundreds of mice at the same time.

NeuroBasic started off with something that no scien-tific researcher had ever had before: a lab animal model for tests. De Zeeuw: “The Netherlands is the

Introduction


NEUROBASICNEUROBASIC

4746

6 If the medicine does not have the desired effect, it will be refined – or the researcher goes back to the drawing board.

8 An effective medicine for a mouse is not automatically effective in humans. The researcher now has to find out if the medicine would work the same in humans.

7 The next step is to de-termine the reason why the medicine is working well.

9 A clinical trial is to be conducted.

10 The medicine is ready.

Infographic


NeuroBasic in a nutshellThe development of medication for brain disorders has always fallen short of definitive

success because their effectiveness will not be known until it comes to the fore in

behaviour. That is why their development has been a matter of trial and error.

1 Sperm and ova are purposefully made to genetically mutate. This causes genetic defects.

2 The mice are screened to find out whether they have the correct defect.

4 A medicine is designed on the basis of the genetic defect.

3 Standardised tests help to ascer-tain each mouse’s specific brain disor-der, such as epilep-sy, schizophrenia or Parkinson’s.

5 The mouse is administered the medicine, after which its effect will be analysed.

AGGTCC

OK?YES

NO

?Epilepsy?

€Medicine

********


4948

Schizophrenia is sometimes commonly called a “split personality” with numerous alter egos. In reality, the disease has nothing to do with multiple personalities but rather with a broken thinking process. Schizophrenia affects the patient’s way of thinking. Their walking, talking, eating or sleeping remains the same, but it is fundamentally difficult for the patient to understand the world around them. Kushner: “We see things around us and can immediately distinguish whatever is important or can be ignored; we see what is real and what is not. This is far more difficult for someone with schizophrenia. If, for instance,

I am on my own, daydreaming about a conversation, I, as a healthy person, know: this is not real, this is a daydream. A schizophrenia patient, on the other hand, can experience the daydream conversation as reality. That makes their life very complicated.”

On top of the unpleasant problems comes the nasty stigma of the disease, Kushner continues: “When a person is ‘talking to themselves’, alarm bells will soon go off in his environment: something’s not right here. We think it scary. We might feel uncomfortable when we see

story: Rineke Voogt images: Michel Mees

The treatment of schizophrenia begins with understanding

Daily life is not easy for people who suffer from

schizophrenia. Medication can help to suppress

the symptoms, but the side effects are rather

harsh. Steven Kushner, professor of Neuro-

biological Psychiatry at the Erasmus University

Medical Center and NeuroBasic researcher

is looking for alternative treatments, but:

“the disorder is difficult to understand and it

is a giant leap from lab animals to humans.”

Interview


5150


A little skin tissue from the participants in the study sufficed to grow IPSCs – firstly for the pur-pose of establishing the differences in the working of their brain cells, and secondly to test the effects of a compound. The researchers studied the mor-phology, gene expression and electrophysiological properties of the brain cells. “For schizophrenia it is important to look at the neuron level, as this is the level where, eventually, something goes wrong”, Kushner explains. “But we have also looked at the other brain cell types like oligoden-drocytes (the cells responsible for the isolation of neurons so that they induce electrical impulses more efficiently; ed.), since the origin of the prob-lem can be elsewhere.” Until now, the tests seem to indicate that it is primarily the oligodendrocytes (from ISPCs) that differ between schizophrenia patients and their healthy family members.

The researchers managed to identify a number of potentially effective compounds with the help of IPSCs. However, even if you know that a compound has the desired effect on a neuron, you still have a long way to go. The next stage was examining as to whether the compound had a good effect on the behaviour and performance of the individual as a whole.

This is where the NeuroBasic working method came back into it. For the treatment of brain disorders, it does not suffice to merely look at the cellular level; it is actually the behaviour that

“ For this type of study, you cannot avoid working with lab animals.”

Steven Kushner: “When a person is ‘talking to themselves’, we think it scary.”

someone injecting insulin, but we immediately know why and find it normal.”

Around one half to one per cent of all adults have schizophrenia. This percentage applies to every population group. The disorder reveals itself from the age of twenty and is chronic: schizophrenia cannot be cured. There is medication available and these make a considerable difference, according to Kushner. “In the times when we had no medi-cines, a patient had to stay virtually continuously in a psychiatric hospital. Fortunately, those days are over. Yet it is still difficult for patients to utilise all their potential and lead a happy, productive life. Medication is quite something.”

Antipsychotics – the type of medicine that most patients take – result in dulled emotions and a never-ending fatigue. Patients are no longer extremely sad, but nor are they extremely happy. Some people are better able to deal with these side-effects than others, but that does not mean that their lives can be compared with living without the schizophrenia. What is more, in the long term, the medication leads to

abnormal movements, comparable with Parkin-son’s disease, because the receptors in which the medicines interfere, are also important to movement.

The room for improvement, therefore, is mostly in diminishing the side effects. “You will hardly ever hear a patient complain: I wish that my medication had a better effect. It is always about the side effects, which can be so nasty that some-one prefers not to take any medication at all”, Kushner says.

The NeuroBasic researchers went looking for ways to improve the medication, step by step. Firstly, you have to understand the disease. Like the majority of disorders, schizophrenia is caused by a combination of genes and, what researchers refer to as, “life experience”: food, lifestyle, accidents, environment, etc. However, which ratio of genes and life experience are determinant, is still to be established. “That is why we started the Genetic Examination into Neuropsychiatric Disorders or Gezin study (Gezin means ‘family’ in Dutch; ed.), a study among families where schizophrenia occurs”, Kushner continues. “The genetic com- ponent strongly prevails there, and so we could identify the genes that go with schizophrenia.” Knowing the specific genes would make it easier to study the disorder in lab animals.

In the quest for better medication, it was necessary to test the working of the different compounds. You can use lab animals for these tests but, fortu-nately, researchers have had other options for a few years. Brain cells like neurons can also be grown from induced pluripotent stem cells (IPSCs), adult skin or blood cells reprogrammed into stem cells. The way these active living brain cells behave is identical to the behaviour of brain cells from the body.

“ It is always the side effects, that the patients complain about.”

NEUROBASIC Case studyNEUROBASIC

52 53

story: Joost van der Gevel images: Ruud Toonen

The baby boy in the video is lying on his back and suddenly makes contracting, spastic movements. He quiets down and then convulses again. This can be traced back in the EEG (brain film). The output is flat for a while, and suddenly there is a peak in activity. These are epileptic seizures. The child has Ohtahara syndrome, a form of epilepsy that goes together with severe mental retardation. Children with this syndrome often come to the hospital during their first few months. The Japanese researcher Ohtahara found the cause in 2008: spontaneous mutations in the MUNC18 protein.

Signal transferThese mutations cause errors in the transfer of signals between the nerve cells in the brains of these children. The MUNC18 protein plays a key role in this transfer. It ensures that the pouches filled with transfer substances, the substances

in the outer ends of the nerve cells, discrete the correct amount at the correct time – and that the next nerve cell absorbs them. This is how nerve cells “talk”. The defects in the protein disrupt this process. This is expressed in epileptic seizures and leads to mental retardation.

MUNC18 is not the only protein involved in the signal transfer of the brain. A number of proteins were already known to lead to brain disorders like schizophrenia, Alzheimer’s, ADHD, and epilepsy if defective. Neurobiologist Ruud Toonen, member of Verhage’s group, focused his search specifically on MUNC18. He is investigating how defects in this protein disrupt the signal transfer and how this relates to Ohtahara syndrome. “This protein is vital for the transfer of signals in the brain”, he emphasises. “No human or animal life is possible without this genome.” Without

Stills from a video of a “jumping” mouse with epilepsie.

Mice with epilepsyMatthijs Verhage’s group within the NeuroBasic consortium is making mouse

models for epilepsy by switching off a protein that is involved in the transfer

of signals between nerve cells in the brain – neural communication. They are

on a quest for medication specifically for brain disorders – with the help of a

mouse with epilepsy.

is determinant: it is behaviour that shows whether the compound has the desired effect. Kushner: “For this type of study, you cannot avoid working with lab animals.”

In a preceding NeuroBasic project, researchers successfully adapted mice in such a way that they can suffer from human diseases. Then, the first step is: check whether the mice do indeed show the associated symptoms. This is quite difficult for schizophrenia. Kushner: “We still do not know whether mice can really become schizophrenic. For people, the diagnosis is set by an interview and is based on someone’s thoughts and experi-ences – you can hardly do so with mice.” Still, the researchers managed to increase the risk of schizophrenia and induce its symptoms in mice through genetic and pharmacological manipula-tions, and are thus able to study the neurological mechanisms that are the cause of changes in the brains of patients with schizophrenia. “We have been quite successful in simulating in mice the neurophysiological changes in the brain as well as the cognitive and social disorders in people with schizophrenia”, Kushner says.

Next, the compounds could be tested in the mice – to find out what they did to their behaviour, motor system, and cognitive functions. In order to assess a compound’s effect on the brain functions of a schizophrenic mouse, you can, for instance, entice a mouse to perform a task that requires memory and cognitive flexibility (see also Ingenious equip-ment makes brain research easier). If a mouse has difficulty in changing a routine that it has been taught, it is an indication of a symptom of schizophrenia: inflexibility.

When it looks like the compound is working and the animal’s behaviour is improving, there is still no reason for a party. After all, the medicine also has to be effective in people. This step from lab animals to humans is gigantic. “A lab animal is not a person; it is the next best option: a model.” Before a compound that has to intervene in poorly functioning neurological circuits can be prescribed to a patient, it has to be absolutely clear that the physiological properties affected by the compound are the same for humans and mice. Even then, success is not guaranteed. Only a clinical study by the end of the programme can provide the true and definite answer.

There are multiple clinical studies into schizophre-nia conducted at the moment, which all focus on improving antipsychotics and additions to the available range. The results are some time coming, but Kushner is positive: “NeuroBasic has booked so much progress when it comes to the treatment of genetic neurological disorders – especially for the diseases that we knew more about when we started. We have now considerably increased our knowledge of schizophrenia. I have no doubt that this knowledge will eventually lead to far better medication.”


“ A lab animal is not a person; it is the next best option.”


–/–

–/–

+/+

+/++/–

54 55

embryonic stem cells are replaced into the young embryo. They will then divide along with the rest while it grows into a mouse.”

Black-and-white mosaicOne part of the stem cells is, therefore, manipulated, but not all. In order to localise them at a later stage in the development of the mouse, the researchers take the stem cells that they manipulate from white mice and inject the mutated cells into a young embryo of a black mouse. Most of the pups that are subsequently born, have black fur, but some consist of both black and white cells: these grow into mosaic mice. “In these mice some of the egg and sperm cells will also be mutated and some won’t”, Toonen explains. “If you cross these mice and a pup is born from one of the mutated sperm cells, the pup will be completely white. It will get its father’s white coat. This pup has the mutation in all the cells on one chromosome, stemming from the father. The chromosome that stems from its mother is normal. The fur colour must be deter-mined by the mutated embryonic stem cell in order to use it as a tell-tale.” This is the mouse model for Ohtahara syndrome.

TranslationThey can change the research setting from a dis-tance and let the mice perform tasks without disturbing them. “We have a test battery where we can observe a great many mice simultaneously. This way, we try to make the translation from the mutations in the genome to their behaviour”, Toonen explains. He shows a video of mice with mutated MUNC18, the lab animal model for the Ohtahara syndrome. On the left of the screen, a mouse is quietly walking around its cage and sud-denly jumps straight up in the air. The mouse in the next cage is sleeping and seems to quiver all of a sudden – and again a few moments later. These are uncontrolled movements. The mice have sei-

zures that resemble epilepsy. “Insofar as their behaviour goes, we see them jump and shake. But it is difficult to know for certain that it is epilepsy, for epilepsy is only defined with EEGs, brain films, in people.” A specialised laboratory in Germany will soon make EEGs in these mice. It will then become clear whether they have “real” epileptic seizures, or not.

MedicationThe next step in the study is the search for com-pounds that can repair the disrupted signal transfer. Any molecule that compensates for the mutation in MUNC18 is a potential medicine. These kinds of compounds are designed by other research groups within NeuroBasic and the phar-maceutical industry. Toonen: “This all starts with a library of some hundred thousand molecules. This collection is reduced to about fifty candidates whose nerve cells we can test in a dish in order to find to what extent they affect the neural signal transfer. If these tests show that they improve it, you could, for instance, test the five most promis-ing candidates in our mouse model. However, that is not until the very last stage – just before they are tested on humans, actually.”

A nerve cell, with the dendrites in red and the synapses, the points of contact, in green.

Mice with epilepsy

MUNC18, the nerve cells cannot communicate. If it has a flaw, the balance in the signal transfer is upset.

Smart medicationThis asks for medication that very accurately intervenes in this disruption of the nerve cells’ communication and repairs the “connection” – a remedy that can repair, or compensate for, the defect in the protein. Toonen: “But we first have to understand how exactly the mutation disrupts the communication in the brain – why the patient becomes ill. As soon as we know this, we can develop medication that is specifically aimed at changing or improving the protein. We want to develop medication that is ‘smarter’ than the existing medicines or works better.”

In order to develop such medication, you need a model in which you can study the defect as well test the new medication at a later stage. For this purpose, NeuroBasic developed a method whereby lab animals are given the exact same genetic defects as the defects that are at the bottom of

Alzheimer’s, schizophrenia and other brain disorders. In the US, in the lab of Thomas Südhof – who was awarded the Nobel Prize in 2013 for his research into signal transfer in the brain, Matthijs Verhage created such a model for the Ohtahara syndrome. The result was a mouse model with a mutation in the MUNC18 protein: a mouse with epilepsy.

Nonsense geneYou can make this type of mouse model with stem cells that you extract from a very young embryo. These stem cells have usually not divided them-selves as yet and can still grow into an entire animal. “Using DNA technology the MUNC18 gene is destroyed or replaced by a nonsense gene”, Toonen explains. With this technology you copy a small particle of a gene and put another particle in its place. When the cells divide, the nonsense gene is copied along. “We do so on one chromo-some, one of the two parts of the gene – you have one of each of your parents. You will then end up with embryonic stem cells with one functional and one non-functional MUNC18. These mutated

Mutant MUNC18 mice, a day before they were to be born. A brain-dead pup with two mutated MUNC18 genes (--), a normal pup with two healthy MUNC18 genes (++), and a mouse with epilepsy, with both a healthy and a mutated MUNC18 gene (+-). The pups in the middle and on the right strongly resemble one another, but the left one looks slightly diffe-rent: it cannot move and is brain-dead. If you insert an electrode into the brain, you will see that, without the protein, there is no signal transfer whatsoever. This proves that if you remove the MUNC18 protein, the nerve cells can no longer transfer signals.


56 57

The research by the NeuroBasic consortium has led to multiple spin-offs. Various companies are specialising in the development of hardware and software that contribute to better brain research – in lab animals and even in humans.

37 stepsA walking test to assess brain functions: that is the idea of the Erasmus ladder. The mouse catwalk was developed by Neurasmus, the neuroscientific company of Erasmus University. Because mouse genes can be adapted so as to simulate brain dis-orders, there is a need of tests that can measure the effects of both these disorders and the rele-vant medication, and this piece of equipment is just the thing for this: a horizontal ladder with 37 steps on each side, between two cages. Each step can change height on command. Mice are trained to walk from one cage to another with a constant speed, and the apparatus’ pressure sensors measure this speed as well as the mice’s stumbles and their jumps. “This way, you can not only thoroughly study a mouse’s walking behaviour, but also the way in which it learns a new walking pattern”, Chris de Zeeuw, Programme Director and Professor of Neurosciences at the Erasmus University Medical Center Rotterdam says. If the position of the steps changes so that an obstacle appears, the mouse has to shift its legs in order to avoid it.

Ingenious equipment makes brain research easier

story: Rineke Voogt image Noldus

Report


58 59

Ingenious equipment makes brain research

image: Sylics

Up to 350 timesThis mouse has the option of three holes to climb through. If it picks the correct one, with the green arrow, it will get a treat. If it chooses the wrong one, nothing happens. It looks simple enough. However, the mouse will only get its nice snack every fifth time that it crawls through the correct gate. An ordinary mouse usually takes an hour or two to learn which hole it should choose. This makes it a suitable task to study Alzheimer’s, for instance, Maarten Loos of bioinformatics com-pany Sylics, the developer of this CognitionWall, explains. “The cognitive performance of mice that develop brain damage like Alzheimer’s is poorer and they take significantly longer to execute this task. They will have to try for up to 350 times before they get the hang of it. So, if you give these mice a substance that should counter Alzheimer’s disease, you can hereby check whether it is effec-tive enough to speed up learning and memorising.” The test is also a useful tool for schizophrenia: if you change the correct entrance after a while, the mouse should be flexible enough to notice.


D

2.9588 cm

60 61

Ingenious equipment makes brain research

image: Sylics

Eons of timeThis line-up of mouse cages looks common enough. Nevertheless, they are ingenious appara-tuses: mouse after mouse can be watched without any need of a hands-on researcher. The lids house a camera that meticulously follows the animals’ every movement and sends it to a computer. The software analyses the video images and provides the researcher with an overview of the endless ramblings of their lab animals. Different colours indicate to where the mouse has stood still for a long time, which corner of its cage it likes best, or which route it prefers. In many lab animal studies it is useful to study such types of movements, for instance if you let them perform a task like the CognitionWall or an anxiety test. After all, you want to know how quickly a mouse learns. However, it normally takes eons of time. Hence the Pheno- Typer, a cage with associated software developed by information technology company Noldus. Dozens of mice can each go their own way in their own cage, and they are all automatically observed. “The fact that the PhenoTyper is fully automatic and can be configured for all kinds of tests makes it so popular among scientists”, Lucas Noldus, general director, explains. “It is also a commercial success: we have already sold over a thousand of these cages with instrumentation and video tracking software.”

D

2.9588 cm

Feedingzone

Short arrestLong arrestShort movementLong movement

Shelter

Drinkingzone


6362

Why did you opt for mice as lab animals?De Zeeuw: “We share 97 per cent of our genes with mice. This means that a great many genetic defects that are behind our non- or misproduction of proteins cause identical problems in mice. Moreover, a mouse can easily be manipulated genetically, is simple to keep, procreates rapidly and without further ado, and is cheap – hence its suitability as our lab animal.”

Still, there have been cases where medicines worked well in mice, but the eventual clinical trials were disappointing.De Zeeuw: “That happens. We have also seen that happening within NeuroBasic. In practice, it then turns out that even though a process has the same genetic origin in mice and humans, the degree to which the genes come to expression differs. That is why it is so important to stay on top of things as soon as you translate results to a human being: it is not automatically true that it works the same in humans. Sometimes you will be partially dis-appointed. But then it is still interesting to see what exactly is, and what is not, comparable.”Kushner: “A lab animal is not a person: the mouse is explicitly a model, not the real subject. That makes it difficult. We have to carefully ascertain which parts of their physiology are comparable or in fact completely different.”

You could also use another lab animal.Kushner: “Sometimes you might be better off with zebrafish or fruit flies, but for our type of brain research it is essential that the model brain is similar to ours. The brain must be as developed as possible, but at the same time the animal itself should suffer the least. Primates as potential lab animal models come too close: their emotions are very similar to ours.”

De Zeeuw: “There is no better model. Whatever ani-mal you choose, the step from lab animals to humans remains crucial. It is especially because we are very similar to other mammals but we differ from them too that something which looked very promising can turn out to be not, or only partially, successful.”

The fact that lab animals sometimes prove not to be such a good model after all is used by lobbyists as an argument for abolishing lab animal testing altogether. What is your opinion on this issue?De Zeeuw: “I have sympathy for the moral principle that each lab animal is one too many. I believe that you should only sacrifice lab animals if there is a realistic chance that they will help you acquire new knowledge. When I first started out as a researcher and had to kill my first rat, I actually considered a career change. But the reality is that we simply have no alternative. There are no lab animals in the kingdom of mammals more suitable for genetic manipulation than a mouse, and you can only do without any lab animals whatsoever in some cases by testing on cultivated tissue but the tissue still has to come from somewhere. If researchers can do without it, they will. But tissue does not show whether it is depressed or epileptic – only the behaviour of a living organism does.”Kushner: “At the end of the day, no patient would like to test medication that could harm them. If you take lab animals completely out of the equation, you would take an enormous risk. Not everything can be tested on computer models or cultivated cells. However, it certainly is not a matter of scientists loving to work with lab animals. I am still hoping for the day that we don’t need them anymore. If there would be an alternative that was just as good – or even if it would just come close – I would be happy to adopt it, and a great many others too.”

story: René Rector en Rineke Voogtimage: René den Engelsman

‘It is not an automatism; you will be disappointed sometimes’

The principle behind NeuroBasic is to test medication in lab animals in a useful way,

in cases wherein the disease that the medication needs to cure is mostly expressed

in the animals’ behaviour. However, a mouse is still a thirty-gram lab animal with

long whiskers. So, is a mouse really a suitable model when it comes to inducing

human disorders? We give the floor to Chris de Zeeuw, NeuroBasic Programme

Director and Steven Kushner, Schizophrenia Transworkpackage Leader.

Discussion

Viral infections: better understood, better contestedVIRGO

Virgo aims to prevent viral infections by designing new

or improved vaccines, and to improve clinical treatment

with novel diagnostic tools and interventions.

VIRGOVIRGO

6766

of the host response and the regulation of this response is essential in order to develop effective vaccines in a rational way.

PesteringVirgo is applying new techniques to connect this knowledge. By meticulously examining all the steps of the virus-host interaction, Virgo scientists are able to determine exactly when and where to tackle the virus. When a virus enters a host, the immune system is triggered immediately – with a quarter of our more than twenty thousand genes jumping into action. T-helper cells, for instance, put other immune cells to work, B-cells are prompted to make antibodies, and cytotoxic T-cells to kill cells infected by the virus. At the DNA level, all sorts of things happen in these cells; genes are switched on and off in order to clear up the infection and are transcribed in messenger RNA. Protein synthesis and other bio-

chemical intermediary and final products also change under the influence of a viral infection.

The Virgo research projects focused on all these levels. This had led to a range of biomarkers: indica-tors that help towards the diagnosis, treatment and prognosis of an infection. For example, if a bio-marker shows that someone is highly sensitive to a certain viral infection, you are able to predict that the infection will be severe and so give the patient more targeted treatment. The technology has also helped to discover new viruses, such as the MERS virus. In such cases, the strategy is to collect as much data of the unknown infection as possible in order to detect similarities with known viruses and develop a treatment method. And, finally, this approach can result in new or improved vaccines, since a host’s response to a vaccine can now be measured more accurately.

“Naturally, it is impossible to prevent all viral infections”, Andeweg adds. New viruses continue to appear, time and time again. Some viruses that were virtually eliminated, such as measles and polio, still prove capable of raising their head again – for instance due to a low degree of vaccination in regions of crisis. “The pestering is not over yet. We’re still playing a cat-and-mouse game.”

Virgo focuses on four groups of viral infections

which, together, comprise the most common

viruses. The first group is the acute respiratory

viral infections, including influenza, SARS,

MERS and RSV. Viruses causing diarrhoea

(acute enteral viral infections), such as the

Norovirus, are the second group. The third

group includes the chronic liver infections

hepatitis B and C, and the fourth is HIV.

This “hairball” illustrates the network of genes that is known to be involved in the immune system. A line represents “a relationship” and a dot stands for a gene.

“Viruses are smart”, Virgo Coordinator Ab Osterhaus says. Although they cannot reproduce themselves, their parcels of genes suffice for them to be multi-plied by a host and spread from there on. Since they rapidly mutate, viruses can easily take advantage of niches. This makes them quite difficult to control.

The changing world has actually made life easier for viruses. Osterhaus: “When I graduated, the smallpox virus had just been eradicated. At the time, the idea was that we would soon be able to manage all viruses this way. This, however, was far too optimistic. In fact, we are faced with growing numbers of infec-tions and almost all of them stem from the animal world.” Viruses then jump from the animal host to humans, with major consequences sometimes. In a small, isolated human population, a virus has far more difficulty to keep going, simply because there are not enough hosts to survive. Osterhaus: “Due

to intensive cattle breeding as well as the global village effect, viruses are able to spread more quickly than ever before.”

Over-responseIn order to tackle this problem Virgo focuses on various antiviral strategies. “The best thing, of course, would be for us to make a vaccine”, coordinator Arno Andeweg says. Great success was achieved in the past with vaccines of cultured viruses that had been attenuated or killed, but could still lead to an immune response that protected the individual against the real virus. “Many of these vac-cines were discovered by trial and error, but unfor-tunately that approach is not suitable for all viruses. For RSV, a virus that causes respiratory infections in young children, such a vaccine actually triggered an over-response, causing the body itself to further damage the lungs.” Therefore, greater knowledge

Follow the immune response step by step

Eradicating a virus is not as easy as it seems. While viruses used to be isolated

within small host populations, our contemporary lifestyles make it far easier

for them to reach the entire world’s population. Vaccines do not always help.

Research consortium Virgo is seeking a better approach to viruses: when

you know exactly how they work in the body, you can treat infections more

successfully or even prevent them.

story: Rineke Voogt

Introduction

VIRGOVIRGO

68 69

Gene expressionWhen investigating different types of infections, compare the levels of gene expression in infected persons and virus-free persons. If you find different levels of activity, you have identified genes which are involved in the immune system.

Protein interactionStudy databases of measured interactions between human proteins and virus proteins. Chances are that such human proteins (and therefore the genes associated with them) are involved in the immune system.

Genome regulationSome proteins stimulate cells to read DNA. Quite a few of these “transcription factors” are known to play a part in the immune system. If you then determine which part of the DNA they “match”, you will know which gene they activate.

Genome evolution Since virus-host interaction is a matter of “staying one

step ahead of the other party” for both the virus and the host, genes which play a part in the immune system tend to evolve more quickly than other genes. If you compare human genetic material with genetic material obtained from apes and other organisms, you can identify the main differences.

New knowledge on the body’s immune response to viruses

Newly identified

genes

Genes which play a part in the immune system

Discovering gene functions by combining dataScientists derive new knowledge from combining and analysing data sets. But where to get the data sets?

Genetic dataGenetic material

obtained from several sections of the human population can be compared. Immune system- related genes show greater variety than other types of genes since genes in different parts of the world are exposed to different viruses.

Immune system model improved by cleverly combining data Model of immune

system improved by cleverly combining dataWhether or not a person falls ill after contracting a virus partly depends on how

capable their immune system is of fighting off the virus. However, parts of the mecha-

nism used to fight off viruses are still unknown to geneticists. Bioinformaticians of the

Virgo research consortium are investigating which genes play a part in this process.

Both viruses and their hosts profit from being able to recognise each other – viruses because they are unable to reproduce on their own, and hosts because they are unable to fight off intruders if they are unaware the intruders are intrud-ers. They recognise each other through proteins, whose blueprints are stored in the genetic materials of both the virus and the host.

Keeping a tally of genesVirologists are trying to get a good understanding of what happens when a virus and a host interact, as this will help them fight viruses. However, it is hard to determine what actually happens at the genetic level. “If you’re studying the measles virus, you mainly learn to understand what happens during a measles infection. You’ll be measuring the genetic activity going on in the immune system, some of which will be specific to that particular virus. We need to find out what’s happening at a more generic level, because that will help us fight many more viruses”, says Martijn Huynen,

a professor at Nijmegen Radboud Academic Medical Center’s Center for Molecular and Biomolecular Informatics.

Huynen and researcher Robin van der Lee compared countless data sets obtained from molecular and biomedical research so as to iden-tify the key genes in the human immune system. The data sets they used hail from other studies and are made up of various sorts of data (see figure on the right). Each of these separate studies has generated new insights into the human immune system, but many more insights can be obtained from cleverly combin-ing all this data. “When you keep a tally of genes which pop up in all data sets as being somehow ‘involved in the immune system’, you’ll find that there’s not a single gene which is involved in every-thing”, says Huynen. “Instead, you must use your knowledge of virol-ogy and immunology to calculate the likelihood that a gene is involved, and the degree to which it is involved. You need to perform a ‘weighted’ tally.”

After solving many statistical riddles, Van der Lee and Huynen ended up with a collection of a few hundred genes. A new challenge followed, in that they now had to verify that all these newly found genes were actually involved in the human immune system. “This was an immense amount of work”, Van der Lee states. As part of the study, Utrecht University virologists con-ducted a number of experiments. In this way, the bioinformaticians arrived at a model which, accord-ing to their calculations, presented a much more accurate description of the immune system than the current model.

Intervention strategies This sophisticated model offers clues which may help scientists come up with new intervention strategies to fight infections in the future. However, for now the research is of a more fundamental nature: we now know that many more genes are involved in organ-ising our defence against all sorts of pathogens.

story: René Rector image: Robin van der Lee

■

■

■

■

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

VIRGOVIRGO

■ ■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■

7170

The Virgo project has certainly given virology a boost in terms of new knowledge and techniques. New technologies have helped prevent pandemic viral infections, and viral mysteries – like the Trojan horse that delivers the measles virus – have been unravelled. So much more is known now as to how viruses invade the body, which receptors they use as a key to enter a cell and how they reprogramme host cells in order to multiply themselves. Dozens of new viruses have been discovered. Virgo coordi-nator Ab Osterhaus explains: “We are now able to identify all genetic material in, for instance, fluid from the lungs of seriously ill patients through deep sequencing, and then to use bioinformatics tools to determine which virus is causing the disease. The new techniques enable us to quickly recognise a virus like the MERS coronavirus as ‘new’, which means we can also develop diagnostics, medication and a vaccine sooner.”

Top publicationsThe increase in knowledge can more or less be quantified in publications and doctoral theses as well as in numbers of newly discovered viruses, developed vaccines, patents, and spin-offs (see frame). This project has produced dozens of publica-tions in scientific journals, for example, including a great number of articles in top-ranking journals like Science and Nature. Some researchers have obtained

their doctoral degree in this Virgo project and have flown the nest in order to apply their knowledge to new virological issues all around the globe. The Rotterdam Viroscience Lab has grown into a Centre of Excellence.

Self-supportiveThe FES funds were not only intended to promote scientific success, but also to encourage cross- pollination with the world of business. “We have calculated that the FES investment has been paid back ninefold in subsidies from outside the Nether-lands and funds from the industry. In other words, we have injected new money into the Dutch econ-omy”, says Eric Claassen, initially involved in the project as Professor of Immunology and now Pro-fessor of Entrepreneurship in Life Sciences at VU University Amsterdam. “The FES money was the starter engine. However, such funds are only useful if they lead to something sustainable, so we specifically wanted to instate a research group that could pay its own way. We started out with just the six of us. With the ten million grant from the FES funds, we eventually created a hundred jobs in applied biomedical research for the next ten years – a thousand men years of work. This earned us the Valorisation Award of one million euros from the Netherlands Genomics Initiative (NGI).”

Virus killers: a great economic success too

The Virgo consortium has yielded a kaleidoscope of new insights, diagnostic

tools, vaccines and spin-off companies. “It has more than recovered the

investments made by FES, the Economic Structure Enhancing Fund”, says

Ab Osterhaus, Professor of Virology.

story: Joost van der Gevel image: Parkers

Case study

Publications with an impact factor greater than 10

Publications with an impact factor greater than 5

Publications with an impact factor less than 5

PhD theses up till 2015

Another 20 or so in the years thereafter

Presentations at conferences and symposiums

Poster presentations

Patent applications and 2 patents awarded

New spin-offs in 2013

New projects with public partners

New projects with industrial partners

Non-scientific publications

Non-scientific presentations

Contributions to the public debate/discussions

Educational activities and contributions

New viruses discovered

New clinical applications (such as protocols)

New products for the general public (such as informative websites)

Advisory councils and boards with seats held by members of

the consortium

VIRGO

7372

London, Bonn University in Germany, together with the Rotterdam Viroscience Lab, developed a specific test for the identification of the virus. There were just two patients at that time. Based on their experi-ences with SARS and Influenza, the virologists suspected that the virus had “jumped over” from an animal shortly before. But what animal species? They started to investigate.

Coronaviruses – “corona” is Latin for crown or halo – have a crown of spikes: proteins that can bind to receptors on human and animal cells. If such a viral protein binds to a receptor, the virus is able to invade and merge with the cell; the virus genome is then released and the cell starts to make new exemplars of the virus. Humans and animals that have once been infected with the virus have made antibodies against this specific surface protein (and others). If you know what the protein looks like, you can develop a test to identify these antibodies and also use this information to track down the receptor.

Autumn 2012

Bart Haagmans and Stalin Ray identified the human and animal protein DPP4 as the receptor. It is found on the outside of cells in the lungs, kidneys, intestines and on the cells of the immune system. This cellular protein is the key that the virus uses to enter the cell. Smaller animal species like mice, hamsters and ferrets proved not to be able to be infected with the MERS coro-navirus. All animals share this receptor, but due to variation in the amino acid sequence there are differences that make it harder or impossible for the virus to bind.

IMA

GE

: E

LM

OU

BA

SH

ER

AB

U B

AK

ER

AB

D F

AR

AG

story: Joost van der Gevel

Report

Summer 2012

A sixty-year old patient of the Egyptian doctor and virologist Ali Mohamed Zaki died in a Saudi hospital. The patient had pneu-monia and kidney failure. Zaki had managed to cultivate the virus and suspected that the man had an infection with a paramyxovirus, from the same family as measles and mumps. The group led by Ron Fouchier at the Erasmus University Medical Center had recently published a new diagnostic method for such viruses. Zaki contacted Rotterdam and sent his patient’s data and materials.

The researchers in Rotterdam established that it was not a paramyxovirus. The RNA

sequences revealed it to be an unknown coronavirus, family of SARS and various cold-causing viruses. The virus would later be named Middle East Respiratory Syndrome coronavirus (MERS). Dr Zaki notified the Saudi Minis-try of Health, in compliance with the international rules of the World Health Organization (WHO). He also posted the new virus on ProMED mail, the web and email service for outbreaks of infectious diseases. This was on Thursday, 20 September.

September 2012 On Saturday, 22 September, Ab Osterhaus, a renowned virologist of the Viroscience Laboratory of the Erasmus Medical Center, received a phone call from England. A man who had flown with a private jet from Qatar had been admitted to a London hospital. He had been isolated immediately and put on a ven- tilator. The man suffered from pneumonia and kidney failure and was in such bad shape that the British doctors believed he was probably going to die. They asked for the sequencing data of the new virus and probes – small parts of complementary RNA – to help them diagnose the virus.

This man also had MERS. He remained in intensive care for almost a year before he died. Within a week after the call from

It was a death in Saudi Arabia caused by an unknown virus that signalled the start of a test case for the Virgo toolbox of genomics, proteomics and bioinformatics for the discovery of viruses, development of vaccines and prevention of spreading.

The sniffling dromedary and the deadly virus

IMA

GE

: S

TAL

IN R

AJ

VIRGO

7574

December 2014

It worked. The vaccinated drome-daries did not become ill, whereas the unvaccinated animals did. However, this was no guarantee for success. MERS did cause heavy sniffles in the unvaccinated dromedaries, but that was it. Chances were that the Saudis, for this reason, would have little interest in a vaccination pro-gramme. The virologists had anticipated this. Osterhaus: “They also had an entirely different problem: camelpox. We put the MERS protein in a ‘vector’, a cripple pox virus that can no longer do any damage. This MERS vaccine, therefore, also protected these animals against camelpox. That would kill two birds with one stone and would probably make it quite interesting to the Saudis, we thought.”

April 2016

What about MERS today? In April 2016, the WHO counter for MERS worldwide reached 1,698 patients since 2012, including 609 fatalities.

November 2014 Ab Osterhaus and Bart Haagmans tested vaccines on dromedaries in Barcelona. In order to induce antibody production, they used the MERS protein that binds to the receptor. These antibodies then block the binding of the virus to the receptor; they stick to the virus so that it can no longer bind and infect the cells. Haagmans: “You can induce such a block beforehand with a vaccine even without injecting the entire virus. The binding protein may suffice.”

Summer 2015

During a major outbreak in several hospitals in South Korea, 186 people were infected of whom 36 died. Negotiations about the further development of the vaccine, the vaccination of drom-edaries and possible also hospital staff were ongoing. Osterhaus: “The idea is: control at the source. Taking away the source will bring the process to an end. In other words, if you vaccinate dromedar-ies, you also protect people. That is how Western Europe got rid of rabies, for instance. It is a very old principle.”

“There is an entirely different problem: camelpox.”

The snif fling dromedary and the deadly virus

IMA

GE

: D

O W

IDA

GD

OIM

AG

E:

MA

RIO

N K

OO

PM

AN

S

IMA

GE

: C

HA

NTA

L R

EU

SK

EN

May 2014

Half of the dromedaries in the abattoir turned out to test positive for MERS. Koopmans wanted to push back the spread of the virus by, for instance, vaccinating camels from Australia – which were not infected – or by trans-porting them directly from the ship to the abattoir. “Transmis-sion between people in the hospi-tal can be managed by vaccinating hospital staff and good hygiene”, Koopmans explained, “but that will still leave the occasional animal-human infection. We want to block this path.”

April 2014

In response to an outbreak of MERS in Qatar, the WHO asked Koopmans to investigate. Two people at a small farm were ill, and a number of dromedaries had some kind of cold. The patients and the camels were found to have the same virus. The question was: who infected whom? Koopmans focused on how the virus spread. In Qatar she examined different groups of animals: at an abattoir, in the market and at the dromedary races. Koopmans: “This region was booming and still is, thanks to the riches from oil and gas. There is enormous wealth and activity that draws a lot of people and this in turn creates a demand for food. Hence the huge market for dromedary meat. The animals are bred in great numbers. Qatar has one central abattoir. The country also imports a lot of animals from the Horn of Africa and Australia, and these all come together in the same market, where they are kept for months on end before they go to the abattoir. This truly is a ‘pressure cooker’ for the spreading of a virus.”

Spring 2013

Professor of Virology Marion Koopmans and Chantal Reusken of the RIVM National Institute for Public Health and the Envi-ronment in the Netherlands travelled to the Middle East to test large farm animals like cows, goats, sheep and dromedaries for antibodies. Dromedaries were found to be the only animals with antibodies against the MERS coronavirus. This was soon to be confirmed by other researchers. The antibodies were even found in the blood of dromedaries from the 1980s.

IMA

GE

: E

LM

OU

BA

SH

ER

AB

U B

AK

ER

AB

D F

AR

AG

IMA

GE

: M

AR

ION

KO

OP

MA

NS

IMA

GE

: M

AR

ION

KO

OP

MA

NS

VIRGOVIRGO Interview

7776

Such a chronic situation is dangerous. Because the virus divides itself so often it induces a lot of anti-gens, which the immune system sees as a threat. In response to this enormous amount of antigens, the immune system produces a great many virus-spe-cific immune cells, mainly T-cells. Such a strong immune reaction may damage the liver, which is where the hepatitis virus settles.

To prevent serious damage, the body itself pulls the emergency brakes. If the immune system is exposed to a huge amount of antigens, the T-cells become exhausted and lose the ability to divide. This pre-vents further damage to the liver, but it also causes the virus to never be completely eradicated.

“What we have tried is to reactivate the exhausted immune system”, Boonstra explains. “This can be done in two ways. The first is to use exhaustion blockers. An exhausted T-cell is lined on the outside with so-called exhaustion markers, transfer sub-stances that prevent overstimulation of the immune system by reducing the division of T-cells. By block-ing these markers with antibodies, you can relieve the exhaustion and reinvigorate the T-cell response.”

The second method is to apply TLR-stimulation. TLR stands for Toll-Like Receptors. These are molecules on the outside of an immune cell that are able to rec-ognise the structures of pathogens. They serve as the sensors of the immune system. By stimulating these sensors, you can reactivate the immune system.

Both strategies are potentially risky. There is a great risk of overstimulation of the immune system. Boonstra: “The existing therapies enable us to sup-press the hepatitis virus quite effectively, so that few virus particles make it into the blood and liver. How-ever, it is extremely difficult to remove all the virus particles from the body. You want to suppress the hepatitis virus in such a way that there are few virus

particles in the blood, but more in the liver. Our idea is to treat people first until the virus is almost gone and then use this method of reactivation to give the immune system the last shove in the right direction. We will have to increase the dosage extremely slowly in the clinical trials.”

HIV infection, or AIDS, involves a similar problem. Virologist Rob Gruters is trying to stimulate the immune system to enable it to attack HIV. “HIV infects the cells of the immune system”, he explains. “This in itself is quite nasty, but the virus make things even worse by integrating its own DNA into

the cell’s DNA. This way, it becomes part of the host’s DNA and can lurk in the dark. Moreover, the virus can change very rapidly. The result is a race in which the virus always stays one step ahead of the immune system. Eventually, the immune system gets exhausted and gives up.”

“ The great majority of hepatitis infections in children becomes chronic.”

Researcher Rik de Groen is inspecting the blood cells of patients with viral hepatitis under the microscope.

A look into immunologist André Boonstra’s lab.

IMA

GE

: R

OB

GR

UTE

RS

story: Elles Lalieu

Immunologist André Boonstra focuses primarily on hepatitis B viruses. “The hepatitis B virus has a high replication speed over a long period”, he explains. “In other words, there is a continual development of numerous additional virus particles. Ninety per cent

of adults nevertheless get rid of the virus by them-selves. In ten per cent of cases, however, the virus triggers a chronic infection. In children, it is the other way around: the great majority of hepatitis infections in children become chronic.”

Booster for an exhausted immune system

Some viruses activate the immune system to such an extent that it wears itself

out. The result is a chronic infection for which patients will have to take antiviral

medication for the rest of their lives. There might be another way: give the

immune system a shove in the right direction so that it can clean away the virus

under its own steam.

VIRGO Case studyVIRGO

78 79

Peeking through the keyhole:

A cacophony of genetic activity

story: Elles Lalieu

There are thousands of pathogens that could attack us, and we have only one

immune system. This immune system has a limited number of genes to send

into battle, so the first line of defence is their built-in flexibility to respond to

each pathogen accordingly. “In the old days, we were peeking through a key

hole, but know we can open the door and have a view of the entire room.”

So in HIV cases there are two problems at play: the virus breaks down certain immune cells (CD4 T-cells) and eventually outsmarts the remaining immune cells. Problem number one can be tackled by the administration of antiviral drugs. The number of immune cells in the blood can be measured. Healthy people have 500 to 1,500 CD4 T-cells in a 1-microliter blood sample. In people with an untreated HIV infection, this number can fall to less than a hundred CD4 T-cells per microliter. Therapy is started when there are fewer than 350 CD4 T-cells remaining per microliter. Thanks to the antiviral medication, the number of immune cells rises again, but those cells will never be as effective as they were.

The second problem is more difficult to overcome. Gruters: “The immune system does recognise the

HIV virus, but it looks at the wrong parts, namely the parts that the virus can easily modify. We are trying to ‘re-educate’ the immune system through a therapeutic vaccine so that the remaining T-cells can tackle the virus again. We have achieved good results in the lab and in lab animals, and have meanwhile

vaccinated seventeen people. One patient was able to do without treatment for seven years but, gener-ally, the vaccination effect is limited.”

Gruters wants to improve the vaccine by enhancing it with other parts of the virus. “The immune system must be able to recognise the virus as early as possible”, he says. “The virus invades immune cells in order to divide itself. To start the production of additional virus particles, the virus needs the first few small proteins that deregulate such a cell, so these proteins will not change as quickly. If we put these ‘early proteins’ into the vaccine, we’ll get a far better immune response, and then the virus can no longer escape.”

Again, there is a risk of overstimulation of the immune system. An activated immune cell is ready to make all sorts of new products. This makes it an easy prey for the virus, which wants to make new virus particles. Inactive cells are far more difficult for a virus to successfully infect. If the immune system is activated but still does not manage to clear away all of the HIV, you might actually make it easier for the virus to spread.

Gruters compares the overstimulation of the immune system to an allergic reaction. “The immune system’s response to the stimulus that you apply is far too violent – it’s like cracking a nut with a sledgehammer. The sledgehammer does far more damage than required.”

Tackling hepatitis B and HIV is, therefore, quite a tricky challenge. On the one hand, you want to encourage the immune system to disarm the viruses, but, on the other hand, you do not want the immune system to be overstimulated. This calls for a subtle strategy.

“ The immune system does recognise the HIV virus, but it looks at the wrong parts.”

Patrick Boers (senior research analyst) and Cynthia Lungu (postgraduate student) are working on the analysis of viral material.

Booster for an exhausted immune system

IMA

GE

: R

OB

GR

UTE

RS

VIRGOVIRGO

8180

Peeking through the keyhole: A cacophony of genetic activity

important. For some genes there is no question about it, but for thousands of others we do not know for sure”, Van den Ham explains.

Subtle differencesIt is impossible to examine all genes one by one. Andeweg: “A pathogen’s antigen is firstly recognised by one cell in the immune system. But then all the cell reproduction starts, and you end up with a humongous population of immune cells. Thousands of genes activate themselves in these division pro-cesses, all adding to the enormous amount of data.”

The bulk of the data comes from cells that divide, but that is not what Andeweg and his colleagues are looking for. They are trying to detect subtle differ-ences that determine whether someone has a good or bad immune response to a certain pathogen.

In order to discover those subtle differences, the researchers use co-expression network analysis. “Within the entire pattern of genes, there are groups of genes that show a specific type of behaviour”, Van den Ham clarifies. “So you can squeeze the orig-

inal 20,000 measurements into small sets to create a network of genes. There are junctions in that net-work that might correspond to clinical variables.”

In the danger zoneVan den Ham uses an example to demonstrate this. “There will be groups of genes that are active but do not change their activity when you are infected. Take hearing. Genes that are key to hearing do not play a role in infection and so their activity remains low. For dengue, we have put the genes with the same behaviour in the same group. In the end we had over twenty of those groups, approximately half of which can be related to typical dengue symptoms, such as the amount of blood platelets in the blood.”

“If we know which genes are determinant for a certain illness, we can train the body to trigger the correct immune response”, Andeweg adds. For this reason, research into genes is also significant for the development of vaccines. In the case of dengue, such intervention is all the more important since the course of the illness is difficult to predict. “The majority of dengue infections cause no complaints, but fifteen to twenty per cent of patients do become ill. The first symptoms are somewhat similar to the flu. A few days later, it looks like the patient is recovering, but in a few per cent of them the infection then actually enters a critical stage, with leakage of the blood vessels causing haemorrhages all over the body.”

Dengue is spreading because the habitat of the mosquitos that transmit the disease is continuing to expand. In other words, there are more and more people in the danger zone. That is why more knowledge of this unpredictable disease is more than welcome.

Modern techniques show exactly which part of the genome is very (red), partly (turquoise) or not active (black). Researchers can establish where the immune system’s response to viral infections runs out of control by comparing this “heat map” of a standard patient with one of an overreacting patient.

The dengue mosquito (Aedes aegypti) is the host of the dengue virus that causes dengue.

IMA

GE

: M

UH

AM

MA

D M

AH

DI

KA

RIM

This built-in flexibility makes research into the host’s immune response difficult, but not impossi-ble. Researchers within the Virgo research consor-tium went looking for genes that play an important part in the immune response against dengue.

One gene, one role. That was the traditional idea – but it was incorrect. An immune gene can be expressed in one infection, but in fact be suppressed in contact with another pathogen. “This means we have to pull out all the stops to distinguish the star actors from the extras”, virologist Arno Andeweg explains. “On the other hand: measurement is key to knowledge. And now we have the measurement tools we need. Today we can measure the activity of all the 20,000 human genes in one go. In the old days, we were peeking through a key hole, but now we can open the door and have a view of the entire room.”

Gene expressionWe can view the room thanks to DNA sequencing: the sequential reading of all the genetic letters of the DNA molecule (see box). The capacity of sequencing machines has continued to increase in recent years. “In the 1990s, scientists needed hundreds of

machines and then it still took them ten years to map out the complete human genome”, says bio-information scientist Henk Jan van den Ham. “You can now sequence ten people’s genomes in three days with just one machine.”

The unravelling of dengue has been quite a feat: in collaboration with Indonesian doctors, Cox van de Weg and Eric van Gorp took blood samples of dengue infected patients and the expressions of all the 20,000 genes were measured. “You then know which pieces of the entire genome are on or off, but still have a long way to go, because you will also need to find out which of these genes are truly

DNA-sequencingDNA sequencing is a technique for determining the sequential order of all the genetic letters of a DNA molecule.

The DNA, which consists of two strands, is first cut into small pieces. The strands are then separated. One of the

two strands is dipped into a solution with four differently coloured building blocks and glue. The building blocks

which fit then attach themselves to the DNA strand. The reaction continues until the strands have been copied in

full and the colour has betrayed the sequence of the building blocks. Finally, all the pieces of DNA are put together

into one single genome, like a giant jigsaw puzzle.

This technique can also be used to measure the activity of the genome’s separate genes in a piece of tissue or blood

cells, for instance. If you “tick” the gene specific fragments, you can establish a gene’s activity. The more often you

come across a certain fragment, the higher the gene’s activity.

VIRGOVIRGO Case study

8382

measles virus specifically attaches to immune mem-ory cells and those cells are cleared away, patients actually lose their immunological memory.

Memory cellsDe Swart: 'We already knew that infection with measles weakens the immune system and also that this could last for weeks or months. We were also aware that although the number of white blood cells decreases during infection, it returns to normal val-ues within a week. The thing we have discovered now, is that the numbers return, but the composi-tion has changed. The cells that return are immune cells that fight the acute measles infection, but the

number of memory cells that are effective against other infections has dropped.

Memory cells are immune cells that memorise what they have come across and can quickly jump into action at a second encounter with the agent of the disease, the pathogen. After infection with measles, a patient thus temporarily loses part of their immunity against other, primarily bacterial, infections. But temporarily here does not mean a few weeks or months, but more than two years. De Swart draws this conclusion from research in collaboration with American epidemiologists. They compared figures from databases from the United States, the United Kingdom and Denmark about mortality due to mea-sles and other infectious diseases, both before and after the introduction of vaccination against measles.

More prone to illnessDe Swart: “There is a clear correlation between the occurrence of measles and mortality due to other infectious diseases. This connection lasts up to approximately 28 months after a measles outbreak. This study stands out because it followed the same approach in the various databases and came up with same period for all the countries. After the introduction of vaccination against measles, the number of measles cases naturally decreased, but we can still see a higher mortality caused by other infectious diseases up to two years after an out-

This patient from Nigeria has a red rash all over her back – a well-known symptom of measles. At the moment that the rash occurs, the peak of the infection has already passed. That is why the virus is so contagious: other people are infected before the patients actually show clinical symptoms.

This human dendritic cell was infected with the GFP measles virus. The GFP protein is produced in the cell’s cytoplasm, which makes the entire cell fluorescent.

I M A G E : M I K E B L Y T H

story: Elles Lalieu images: Plos Pathogens/Rik de Swart

Rik de Swart, virologist at the Erasmus Medical Center Rotterdam, has been focusing on measles for many years. He met the Northern-Irish virologist Paul Duprex at a measles conference ten years ago. “Duprex delivered a presentation on a measles virus that makes GFP, a green fluorescent marker pro-tein”, De Swart says. “The only problem was that he had no suitable model to test the fluorescent virus. We did, so ten years ago we decided to join forces. We now have a good model in which infected cells become visible – with a far greater knowledge of measles as a result.”

Trojan horseMeasles was long considered a classic respiratory virus, a virus that infects the epithelial cells of the airways. But in 2000, a receptor for measles was discovered on immune cells. When De Swart and his team released the measles virus in the airways of monkeys, they saw that primarily the immune cells at the bottom of the lungs became infected. This was a strange place to find viruses, De Swart thought,

knowing that the upper airways also contain cells that try to fight infection. “Dendritic cells – immune system cells – regularly stretch their new branches towards the upper airways in order to ‘feel’ if there are any invaders lurking about”, De Swart explains. “We believe that at such moment the measles virus seizes the opportunity and, hitch-hiking on a dendritic cell, is taken along on this Trojan horse to the cells of the immune system.”

It is quite difficult for the virus to enter the gates as a hitchhiker, but once it is inside, it soon starts marauding, infecting various immune cells, includ-ing B-cells and different types of T-cells. The recep-tor to which the measles virus attaches can mainly be found on cells that have already come across parts of other intruders in the past. These cells are affected. At the same time, an effective immune response against the measles virus itself is triggered. The immune system clears away the affected immune memory cells. This in itself is, of course, good news: the patient gets better. But since the

‘Innocent’ measles virus wrecks your immune system

Thanks to vaccination, measles has become a rare disease. But vaccination has

become controversial, so even in this part of the world we have to deal with an

outbreak of measles now and again. This emphasises the need for continued

research into measles if we want to increase our understanding of the infection

VIRGOVIRGO

M F H LN P

N P GFP M F H L

8584

my American colleagues are currently looking into. We still have the problem, though, that you can never know for sure if someone has inhaled the full dosage. In other words, there is room for improve-ment of the measles vaccine, but it is still a long way coming. Until then, we want to use our research findings to reverse the image of measles as a mere ‘innocent’ childhood illness.”

Macaques infected with a GFP measles virus. At the peak of virus replication, fluorescent cells (green dots) are visible in the skin (A), on the gums (B), on the tongue (C), in the lymph nodes (D-F) and in the spleen (G). Some days after this peak, a rash (H) appears, still containing fluorescent cells (I).

‘ Innocent ’ measles virus wrecks your immune system

break. We therefore think that we have found a strong connection.”

The researchers are now trying to find out if other databases reveal the same connection. GP data-bases, for instance, store anonymous data on patients' medical history. “We eventually want to examine the medical history of five hundred to one thousand children for a few years after they’ve had measles”, De Swart continues, “and not focus on mortality, but for instance on the use of antibiotics and admissions to hospital. We would like to com-pare this group with children who have not had measles in order to see whether the first group really is more prone to illness in the two years after infection.”

EradicationMeasles remain a significant health problem, as it is so unbelievably contagious. The virus has difficulties getting in, but is very effective in its way out. That is the trick. At the height of the infection (two to three days before people are actually showing any symp-toms), the immune system is riddled with infected cells. At that stage, the virus also makes contact with the epithelial cells of the airways. This is how a huge number of virus particles end up in the upper airways. Patients can then very efficiently spread these virus particles throughout their environment by coughing. This compensates more than adequately for the difficult entry of the virus.

There is a vaccine, and human beings are the only natural host for the measles virus. Monkeys can be infected by measles, but the groups in which they live are generally too small for the virus to survive very long. Once infected, monkeys cannot be

infected again, so in a small group of monkeys it does not take long before the virus cannot find another “fresh” host. The fact that people are its only host means that theoretically we could eradicate measles by vaccinating everyone against the disease. However, according to De Swart that is easier said than done.

“Vaccinations are certainly not undisputed”, he explains. “In a protected group it does not matter so much if a single member refuses vaccination. The vaccinated others will provide protection. But if the unprotected form a large enough group, the way to an outbreak is wide open – as we saw in 2013 in the Netherland’s orthodox Protestant community.”

Alternative vaccinationApart from the problem of non-acceptance, the vaccine itself is not always practical. It contains a live, weakened virus and must therefore be stored in a cool place. Especially in the tropics this is not always feasible. The consequence is that children are being vaccinated, but with a vaccine that is less effective. Moreover, the vaccine is administered with a syringe, which entails the risk of infection with HIV or hepatitis. De Swart and his team are seeking an alternative way of vaccinating that is better suited to the tropics.

“We are looking for a method that would allow people to inhale the vaccine through a nebulizer”, De Swart clarifies. “In our monkey model, we have already seen a good immune response if the vaccine ends up in the lungs. It is now in the form of a liquid solution, but it would be even better to store it in dehydrated form as the dry vaccine would remain stable for a very long time. That is what

Above: The genome of a measles virus. It contains only six genes (N, P, M, F, H and L), together coding for eight proteins. Below: The genome of a GFP measles virus. In this virus GFP (green fluorescent protein) is expressed. Thanks to this marker protein we can visualize which cells are infected.

VIRGOVIRGO Case study

86 87

Each group included ten to twenty children. This means that the number of blood samples was far lower than the number of genes that were exam-ined, and therefore classic statistics fell short. “That is why we used a method called machine learning”, says statistician Victor Jong. “Machine learning makes it possible to build a model step by step. You start off with the variable (in this case, a gene) that most strongly discriminates between a serious and a mild infection. You then add other variables, one by one, until the model’s predictive value levels off.”

PredictabilityIt works like this: suppose you are at 80 per cent predictability. When you add another variable, the predictability reaches 80.0001 per cent. “Would you then include this variable in your model? Probably not”, Jong says, “for it is of little use. Extra variables would make the model more complex, without contributing very much to its predictive force.”

The RSV model yielded 84 variables that consider-ably improved the predictability: 84 genes that play an important role in the development of a serious infection. The test’s accuracy is good:

between 96 and 97 per cent. So if the outcome of the test is that a child has a 10 per cent chance of developing a serious infection, this prediction is correct in 96 or 97 per cent of cases.

Understanding the outcomesWhether the test also works well in practice has yet to be examined. Fraaij: “Measuring is easy. It usually is far more difficult to understand the out-comes.” Jong compares the test result to the weather forecast. “If there is a 90 per cent chance of rain, you take your raincoat. If a child has a 90 per cent chance of developing a serious form of RSV, it will have to stay in hospital. If there is a 40 per cent chance of rain, however, some people will take a raincoat and others will not. If a child has a 40 per cent chance of becoming seriously ill, the doctor will have to go by observations to decide on the next step.”

“ Sometimes RSV is so serious that the patients need to be ventilated.”

“ As a precaution, there are more children in hospital than necessary.”

“RSV is one of the main viral infections in chil-dren”, paediatrician Pieter Fraaij explains. “Most children are infected for the first time in their first year. It starts with symptoms that are typical of the common cold, but in part of the children the symptoms worsen. One in a hundred children eventually ends up in hospital with respiratory complaints. Sometimes these are so serious that the child needs to be ventilated in intensive care.”

Blood samplesWhich children run the risk of a serious form of infection is, however, difficult to predict. At the moment a child visits a GP or the hospital, it is hard to say whether it actually is RSV and whether the infection will intensify. That is why many chil-dren are referred to the hospital for observation – as a precaution. “This means that there are more children in hospital than necessary”, virologist Arno Andeweg adds. “We have searched the blood for biomarkers that can foretell whether a child will become seriously ill or not as soon as a doctor diagnoses RSV in a child.”

The research team collected blood samples from healthy children, children with a mild RSV infec-tion and children with a serious RSV infection. The human genome boasts 20,000 genes, and the researchers went searching among them in order to find the genes that significantly differ in expres-sion between children with a serious RSV infection and healthy or mildly infected children.

story: Elles Lalieu

Hospital admission or not? Your genes decide

The respiratory syncytial virus (RSV) is an important cause of the common cold. The symp-

toms are usually limited to a runny nose in adults, but RSV infections can actually have very

serious consequences in newborns. In fact, newborns may end up in the intensive care unit

of the hospital and die if they go without treatment of an RVS infection. It would be great if

there were a test that can predict which children will develop a serious form of the infection.

“ The prediction is correct in 96 or 97 per cent of the cases.”


8988

Not all vaccines are available everywhere. What does this mean in terms of virus control? Andeweg: “It certainly is an awful business that not everyone in the world has access to medication or vaccines. In the first place, this has to do with the unfair division of wealth. In the Western world we are quickly inclined to think that a virus no longer exists or is of interest, even if it is still causing major prob-lems elsewhere. Hundreds of millions of people in the developing countries, for instance, fall victim to chronic viral hepatitis or the dengue virus. The availability of vaccines and proper medical care is extremely important – but (often very simple) hygiene measures are also vital.”

Osterhaus: “The distribution of vaccines and antiviral drugs in times of crisis is often closely related to the political situation. If a few countries buy up all the available vaccines when there is danger of a pan-demic, there is nothing left for other countries. Pandemic vaccination should not only be available to the ‘happy few’, but to all who need it. We need to come up with solutions for this problem, for instance by procuring vaccines jointly within Europe. This is not something we scientists at Virgo are looking into, but sometimes we do see some very narrow escapes.”

Still, it is difficult to get everyone to accept vaccina-tion, even in the wealthy Western countries. Take, for instance, the fuss over vaccination against HPV, the virus that causes cervical cancer.Osterhaus: “I do indeed worry about the acceptance of vaccines. People are critical when it comes to science; they no longer take for granted everything their doctor says. On the one hand, there are religious reasons. There is little you can do about that. How-ever, there are also highly educated people organising ‘measles parties’ because they think it would do their kids so much good. There is nothing healthy or good about it, but these parents do not know that. They have never experienced it themselves and are no lon-ger aware of the consequences of the illness and the complications it might cause.” (See also ‘Innocent’ measles virus wrecks your immune system)

Andeweg: “The fact that people do not trust the system unquestioningly is not a problem in itself. Everyone is an articulate citizen these days; the time when only the doctor knew what was good for you fortunately lies in the past. However, the increase of mistrust is worrisome. This scientific research aims to diminish the burden of illness, and this can only be achieved if the generated knowledge can actually be applied.”

How can we take away this mistrust? Andeweg: “Sound information is important. The public’s lack of trust should stimulate us to share our knowledge more effectively. As scientists we have to be able to properly explain the pros and cons of a vaccination. Well-informed people can then make up their own minds, in their own interest and the interest of those around them.”

Osterhaus: “On the one hand, our field has achieved a great deal and is doing good work, but it is a pity that we do too little to promote the acceptance of the fruits of our labour: new vaccines. There are some horrible videos going around showing what suppos-edly happens after vaccination. No one takes a stance against these sceptics. It is up to science to counter their arguments. We should also appeal to social scientists and social media; they can help us learn more about the psychology behind acceptance and the best way to communicate.”

How does this complex problem affect your motivation for this research?Andeweg: “Every researcher knows that their job is a long-term project: fundamental research in view of application in practice. We are used to being in it for the long haul; wars, poverty, and politics will always play a role.”

Osterhaus: “Our research is a dire necessity. We continue to see more and more infections. It is up to us to contain them – it was thanks to our approach to SARS, for instance, that we put a check on a rising pandemic. If we would not do this type of research, we would all suffer the consequences.”

story: Rineke Voogtimage: René den Engelsman

Vaccines do not work when they are left on the shelf

Virus infections can be needlessly problematic. For numerous viruses

there are excellent vaccines, but these are not always available to everyone.

Sometimes they are too expensive, and sometimes people do not want to

be vaccinated because they don’t trust the system. This presents a problem:

vaccines do not work when they are left on the shelf. Virgo coordinators

Ab Osterhaus and Arno Andeweg explain.

Discussion

Towards new cures for diabetes type 1DCTI

The body of a patient with type 1 diabetes no longer produces insulin.

This seriously interferes with the blood sugar control. The Diabetes

Cell Therapy Initiative is searching for new treatment strategies.

DCTIDCTI

9392

subsequently need? Even if they strictly adhere to insulin therapy, there are times that it turns out to be very difficult to control glucose levels.

Some patients may undergo islet transplantation. In this procedure, which could not be performed in the Netherlands until just a few years ago, the Langerhans islets are extracted from a deceased donor’s pancreas and injected into the patient. In virtually everyone who has undergone this transplantation, glucose control is far more stable. Some people can even do without insulin for longer time periods, whereas others will need to keep using insulin or will have to start reusing insulin shortly after the transplantation.

Cell encapsulationTransplantation is not as efficient as we had hoped, and that is for several reasons. Both during the transplanta-tion and directly after injection, many insulin producing cells are lost. Now, in order to improve islet function for longer time periods DCTI has worked on various biomaterials. De Koning: “By encapsulating the donor cells, an environment is created in which cells feel more at home so that the islet loss may remain limited.”

Alternative sourcesApart from the efficiency problem, the short-age in donors also plays an important role. With the current number of donors, 50 to 100 transplantations can be performed each year, whereas we have 1,600 newly diag-nosed patients each year. That is why DCTI focuses on alternative cell sources from which

insulin producing cells can be produced – with one key question in mind: how to culture cells in such a way that we will eventually be able to treat all patients with type 1 diabetes in the Netherlands?

story: Elles Lalieu image: Parkers

Introduction

Type 1 diabetes: on the way towards a cure

Diabetes has the reputation of an invisible killer, since excessively high blood sugar

levels will eventually lead to complications. In patients with type 1 diabetes, the blood

sugar level is sometimes difficult to regulate, as their insulin producing cells are defec-

tive. So far, these patients have nothing else to turn to but insulin injections. Sometimes

a transplantation is considered. The Diabetes Cell Therapy Initiative (DCTI) research

consortium searches for improvements in, and alternatives of transplantation.

The body converts everything that we eat into glu-cose as fuel for the cells. However, the cells cannot simply absorb glucose. They need insulin. It is the pancreas that produces this insulin. This takes place in the so-called islets of Langerhans: clusters of dif-ferent cell types that excrete a wide range of hor-mones. The beta cells in these islets are responsible for the production of insulin.

One millionIn people with type 1 diabetes, the own immune system attacks the islets in the pancreas. These islets thereby suffer so much damage that they cannot produce insulin anymore. Without insulin, glucose can no longer enter the cells and stays in the blood. Chronically elevated glucose in the blood can lead to complications like blindness, kidney failure, heart problems, and amputations. Type 2 diabetes is a completely different story: type 2 diabetes is a metabolic disorder in which the body slowly becomes insensitive to insulin in combination with failure of insulin production –

which may lead to the same problems if it remains untreated.

Approximately one million people in the Nether-lands have diabetes – 900,000 have type 2 and 100,000 have type 1 diabetes. DCTI has, nevertheless, chosen to focus on type 1 diabetes. Project leader Eelco de Koning explains why: “The glucose regulation for type 1 diabetes is generally more complicated than for type 2 diabetes. Patients with type 1 diabetes need to inject insulin immediately after diagnosis, whereas patients with type 2 can usually start off with adjustments to their diet or pills.”

TransplantationAs a result of the constant risk of a too high, or too low, level of glucose patients with type 1 diabetes are aware of their illness all day. Type 1 diabetes has a dramatic impact, as patients continuously need to make calculations. What am I going to do today, what am I going to eat, and how much insulin will I

DCTIDCTI

9594

Instead of transplantations in these times of desperate shortages of donors, and rejection symptoms to boot, the culturing of new insulin producing beta cells would be ideal. Since, unfortunate-ly, these cells are hardly cul-turable, DCTI went searching for alternatives and other methods for culturing.

More recently, the transplanta-tion of islets of Langerhans, where the insulin is made, is also possible. However, the quality of the islets often diminishes rapidly. Clever tricks, such as a protective wrapping, should make these transplants more successful.

Islets of Langerhans

Insulin producing cells

Blood vessel

Infographic


DCTI in a nutshellThe body of a patient with type 1 diabetes

no longer produces insulin. This seriously

interferes with the blood sugar system.

The Diabetes Cell Therapy Initiative (DCTI)

is hunting for new treatment methods.

If the pancreas is defective, transplanta-tion is a (scarce) option. In order to keep the pancreas in good condition during the transplantation and to increase the success of the operation, a special pump has been developed.

The body regulates the blood sugar level with insulin. This insulin is produced by beta cells in the pancreas.

Pancreas

Stomach

Small intestine

DCTI

links URL

DCTI

9796

enough new ones to transplant. Culturing beta cells is not as easy as it sounds. They do not want to multiply, nor in the body or in a petri dish in the lab. Moreover, the cells that you try to culture change their identity. They start off as insulin pro-ducing beta cells, but spontaneously transform into another type of cell that no longer produces insulin. “The good thing is that we can multiply those cells,” Roefs says. “However, now we have to find a way to make them change back again into beta cells.” The idea is that the cultured cells must be suitable for insulin production because that is what they originally did.

Growth factorsResearch into turning this process around is ongoing. By allowing the cells to group and adding var-ious growth factors, the researchers have managed to make them somewhat return to their original function. What they have not discovered as yet is the exact trigger that makes them transform back into insulin producing cells.

Roefs collaborates with the biotech company Galapa-gos to find this trigger. The company has the capacity to test the effects of various medications on tissue on a major scale. For this study, it comes in handy: the effects of substances that may be growth factors can be tested on cultured beta cells within a short space of time. This quick first screening of substances helps the quest for factors that qualify as effective in helping the cells return to their beta origin.

If the quest is successful, it would be the break-through that Roefs is hoping for. The cell could be placed in a scaffold (see also Transplanting islets in an artificial pancreas) and a patient with diabetes could receive new islets without any chance of shortages of donor material. Of the 250 pancreases that are available in the Netherlands each year, some are transplanted in their entirety into one single patient. Other pancreases are used to extract islets, approxi-mately one per cent of the whole organ, which are then transplanted. The rest of the pancreas is often considered to be of no value. This is a waste, the

Islets (in red) can be isolated from the pancreas and separated from the other tissues. The diameter of each islet is approximately one tenth of a millimetre.

story: Rineke Voogt images: Françoise Carlotti

Case study

Wanted: cells for insulin production

There is still no cure for people with type 1 diabetes (whose pancreas no

longer produces insulin). Transplantation of the pancreas is a possibility, but

it is a drastic operation and there is a desperate shortage of donors. That is

why DCTI sought a small-scale solution: transplanting the insulin producing

cells only. These are scarce too, but the Leiden researchers reasoned that

you can culture them.

In type 1 diabetes patients, the beta cells that are responsible for the insulin production are either defective or have been destroyed. To make up for the lack and to keep their blood sugar at an adequate level, these patients take insulin by injection. That boils down to the treatment of symptoms: it is not a solution to the underlying problems. Once beta cells are wrecked, they cannot be repaired.

However, you can replace them. A patient can receive a donor pancreas, including islets of Langer-hans with beta cells that do function. The problem is, that a complete pancreas transplantation is a major intervention and there are very few donors. (see also Pump keeps donor pancreas in shape) Some 250 pancreases become available each year – far too few for the 100,000 type 1 patients in the Netherlands. Moreover, the organ can still be rejected after the transplantation.

Since 2008, the islets are sometimes transplanted separately. They are injected into the liver’s portal

vein. The pancreas itself remains untouched: the pancreas is sensible to inflammation and the liver has proved to be just as fine an accommodation. There is, however, still a great shortage of donors and the patients must continue with their medica-tion to fight rejection symptoms.

Culturing beta cells Leiden researcher Françoise Carlotti and her col-leagues are, therefore, seeking another way of making insulin producing cells without having to rely on donors. Their goal: cultured islets containing beta cells, ready for transplantation. There may be several ways to get there, says Carlotti, Assistant Professor and member of the Diabetes Group of the Department of Nephrology at LUMC Leiden Univer-sity Medical Center. The easiest is to extract beta cells from a healthy donor, multiply them in the lab and then implant them in the type 1 diabetes patient. Maaike Roefs, PhD student, concentrates on this possibility: further culturing of a small amount of beta cells from a donor until you have

DCTIDCTI

9998

Wanted: cells for insulin production

BioreactorsA third option for the culture of insulin producing cells is to start from alpha cells, which can also be found in islets, but produce insulin’s counterpart: glucagon. There are approximately four times fewer alpha cells than beta cells. The research group at LUMC discovered that beta cells sometimes change into alpha cells spontaneously as soon as you disinte-grate an islet and allow the cells to regroup. The idea is that if this spontaneous transformation can be reversed so that an alpha cell becomes a beta cell, then this opens up a new perspective for an alterna-tive source for insulin producing cells.

The work by Carlotti and her colleagues is special in that the researchers focus on human cells. “Many research groups that work with animal cells publish great results – but these usually say little about the way it works in people.” Working with human cells

is essential when the findings are to be translated to the clinic. However, it does have a negative side: it limits the speed of the study. This is not only due to the stricter rules, but also because there is so little material to work with. Approximately eighty of all the pancreases donated each year can be used for research – provided that the surviving relatives give their permission.

No wonder that Carlotti and her colleagues use the available material as efficiently as possible. In order to upscale the cell culture, they collaborate with biotech company Xpand, which is specialised in culturing cells in bioreactors (see also Assignment: culture millions of beta cells). Carlotti: “If we want to translate our work to the clinic, we need a great many cells. Our collaboration with Xpand makes the culturing less labour intensive: in our own lab we culture cells on a two-dimensional surface and it requires quite some work to feed them. With the 3D cell culture developed by Xpand we can upscale the culturing far more rapidly.”

Beta cells (in green) can be cultured in a culture flask. Meanwhile, they change identity. A char-acteristic of this “new” identity is marked in red. Cell nuclei are marked in blue.

researchers believe. The rest of the material might actually come in handy.

Other candidatesOne of the chief elements of the pancreas is its extensive network of corridors, which collects all the digestive fluids that are produced by the organ and leads them to the bowels. Those corridors are formed by so-called duct cells. It is these duct cells that are important in the formation of beta cells: when the pancreas develops in the womb, the Langerhans islets emerge from embryonic duct cells. “This embryonic development might be mimicked in a petri dish,” says Tim Dielen, member of the LUMC Diabetes Group, who collaborates in this project with PhD student Jeetindra Balak.

Embryonic developmentWe already know how to isolate pancreatic duct

cells. These adult cells now have to be enticed to behave like embryonic duct cells so that they can subsequently specialise in the direction of a beta cell. Dielen: “It is quite a puzzle. We use all of our knowl-edge of embryonic development. You have to iden-tify the genes that are necessary for the development of beta cells and chart the chain reactions that take place as a baby grows. This information can then be used to give the cell a push in the right direction.”

The researchers have experimented with different growth factors. It is now possible to culture duct cells for the long term in a special type of gel, in which cells seem to create corridors all by them-selves – as in the embryonic pancreas. The next step is these cells’ development into beta cells. The researchers have managed to show proof of concept: a small percentage of the cultured duct cells are now producing insulin.

An islet in the pancreas under the microscope. Cell nuclei are marked in blue, insulin produc-ing cells (beta cells) in red, and glucagon producing cells (alpha cells) in green.

DCTI

101100

Spring 2014

A breakthrough for the stripped beta cells: a new batch of cells, provided with a fluorescent label, proved to attach properly. Microcarrier CellBind glows with fluorescence as soon as Tra prepared it. The success was short-lived. Betas do want to attach to CellBind now, but they never let go again.

Winter 2014

Growing duct cells seemed to be a good alternative. However, they proved to be quite difficult cus-tomers. The population of cul-tured cells was not pure; they were difficult to sort. Moreover, they needed quite some medium. After a few months of experi-menting, Tra pulled the plug. Duct cells are no longer a priority. Xpand will focus on beta cells from now on.

Spring 2015

You need to know your cells before you can grow them. It had already become clear that betas are not the easiest of cells. They were not impressed either by the restriction enzymes that can usually seperate cells from their microcarriers pretty easily. Even after taking a bath for an hour in a mixture with the enzyme, by no means all the cells had let go. Do they have to soak a bit longer? If truth be told, we had better not, for long soaks also destroy cells.

Summer 2015

The cells of two donors were growing in flat culture flasks. They seemed to be doing well: the correct microcarrier, the correct growth factors, and fifty per cent of the cultured cells could be har-vested. The yields are not great, however. Das and Tra were not prepared to give up on half their culture. There certainly was room for improvement in the harvest-ing process.

20??

The first rich harvest was a fact! On to the next step: culturing the cells in the bioreactor. It can hold an equivalent of a hundred flat culture flasks of cells simultane-ously. Two hundred million cells, if they are lucky. That would be enough beta cells for tens of patients with type 1 diabetes. That is to say, if they manage to change stripped betas back into insulin producing cells. That work is cur-rently in progress.

IMA

GE

: X

PA

ND

IMA

GE

: M

AA

IKE

RO

EF

S

IM

AGE : WI K IM

ED I A

story: Rineke Voogt

Process report

The transplantation of islets of Langerhans that contain good-quality, insulin producing beta cells requires an enormous amount of them, but the culturing of these cells has proved to be a perilous undertaking – even for specialists.

Summer 2012

The whole idea behind DCTI is to transplant healthy, well-func-tioning beta cells into patients who no longer have any of their own. Naturally, this requires a sufficient number of beta cells. There is a shortage of organ donors, so the researchers are searching for ways to create new insulin producing cells (see also Wanted: cells for insulin produc-tion). This is not the only chal-lenge, however, for the culture should not only be successful in a test tube, but also on a large scale. A healthy pancreas has around a million of islets of Langerhans, containing for sixty per cent beta cells. In other words: there is plenty of culture work to do. That is where Xpand comes in.

Xpand is a biotechnology com-pany specialised in the culturing of mesenchymal stem cells (that can later specialise in, for exam-ple, fat cells, bone tissue or mus-cle tissue). Ruud Das and Wendy Tra at Xpand want to use the same technique to culture beta cells, using a special bioreactor that they developed themselves and where a great many cells can be cultured in a plastic bag in a con-trolled environment requiring far less need of medium. The same technique should work for beta cells, they figured.

Autumn 2013

The first batch of cells from Leiden arrived at Xpand. Since adult beta cells do not divide spontaneously, the Leiden researchers had changed them: the “stripped” cells no longer produced insulin, but they could divide. Both the stripped beta cells and the duct cells must be prepared for cultur-ing. It was now paramount to find out as soon as possible under which circumstances the cells thrived best. Unfortunately, dis-appointment came soon after. In order to culture the cells in the bioreactor (and stop them from being flushed away with the medium), they needed to attach to a so-called microcarrier: a marble with a diameter of two hundred micrometres covered with pro-teins that would serve as a good base for cell growth. When Tra added cells to the medium with microcarriers in a culture dish and removed the medium after a while, there were hardly any clus-ters of cells: alas, they did not want to attach.

IMA

GE

: X

PA

ND

Assignment: culture million s of beta cells

DCTIDCTI

102 103

Report

Transplanting islets in an artificial pancreas

A fifty eurocent coinIn the lab of bioengineer Aart van Apeldoorn, researcher Don Hertsig is pouring some type of elastic biomaterial in a tiny, especially designed mould. After a while, the material has hardened enough to be removed from the mould, and ready for the controlled puncturing of tiny holes in the material. The result is an elastic, thin layer of biomaterial that is not much larger than a fifty eurocent coin.

This wafer thin “coin” is meant to be a carrier for islets of Langerhans during transplantation. “Pressing or pouring the biomaterial creates a kind of micro egg container with tiny cups”, Van Apeldoorn explains. “You can collect one islet in each cup. The holes make it easy for blood vessels to grow in to give islets access to oxygen and nutritional substances. When we cover up the graft with yet another layer of material, the islets have their own protective environment.”

In the meanwhile, the new location for the trans-plantation material has been successfully tested in small lab animals. The next step is to test the material in pigs.

story: Elles Lalieu images: Aart van Apeldoorn

DCTIDCTI

104 105

Transplanting islets in an artif icial pancreas

Eleven artificial pancreasesImmunologist Paul de Vos is researching the effec-tiveness of an artificial pancreas in rats and mice. “First, the empty graft is placed underneath the skin of a mouse or rat”, he explains. “The material is foreign to their body, so a rejection reaction is initiated. Once this reaction is done and over with and the graft contains sufficient blood vessels the islets are administered intravenously. If we do it sooner, all the islets will die immediately.”

The first results of the artificial pancreas are very promising. In the large picture, the back of a rat has been cut open in order to assess the blood cir-culation in the graft. The circulation is fine; the islets are surviving and producing a normal amount of insulin. In the small picture, the isolated islets are visible under a microscope.

Eleven lab animals have been supplied with an arti-ficial pancreas, and some animals have been doing well for four months already. In the next few years, the graft will be gradually scaled up from small lab animals via bigger lab animals to humans.

images: Paul de Vos

DCTIDCTI

107106

Interview

story: Rineke Voogt images: Marjolein Leemkuil

Patients with type 1 diabetes usually manage to keep their glucose level under reasonable control by injecting insulin when needed. For some patients, however, it proves to be very difficult: their blood sugar varies enormously and they often already suf-

fer the con-

sequences at an early age. A continuously high blood sugar level may lead to kidney failure, for instance, or damage to the retinas or nerves; an excessively low glucose level may make you lose consciousness or even induce a coma.

These patients are better off with a pancreas trans-plant. Leemkuil: “This is a far-reaching operation with a high risk of complications. Some of the patients are not physically fit enough to undergo such a major operation. Sometimes a patient has to stay in the hospital for months. The results after a successful transplantation are good, by the way; many patients are without insuline after the operation.”

In the Netherlands, around thirty pancreases are transplanted each year. Such an operation demands extensive preparation and great precision. A pan-creas has a soft structure and is far less solid than a kidney or a liver – which means that it is easily

A human pancreas is prepared for connection to the pump.

In some type 1 diabetes patients, the glucose level in the blood is so

poorly controllable despite insulin treatment that a transplantation of

a new pancreas is their only option. This transplantation is a difficult

procedure with a great risk of damage to the organ. Researchers

Henri Leuvenink and Marjolein Leemkuil at the University Medical

Center Groningen are working on a solution so as to improve the

condition of donor pancreases.

Pump keeps donor pancreas in shape

Transplanting islets in an artif icial pancreas

Since several years, patients with severe forms of type 1 diabetes are eligible for an islets of Langerhans transplantation. These islets are extracted from a donor pancreas and injected into the patient’s liver. The transplantation in itself is success-ful, but since the islets have to be infused into the liver, many of them are lost during transplantation and in the following period. Bioengineer Aart van Apeldoorn and immunologist Paul de Vos are trying to find a better transplantation location with the help of biomaterials.

The fact that the islets are transplanted into the liver is mainly for practical reasons. De Vos explains: “Doctors do not want to touch the pancreas, as the organ also produces digestive juices. If these start leaking, the problems are incalculable. Islet transplantation is a good alternative option but more then fifty per cent of the islets die after the transplantation. We want to prevent this loss by transplanting these very delicate islets into an artificial environment that mimics the pancreas.”

This artificial environment must meet certain criteria. “The material may not be degradable and should be easily removable from the body”, Van Apeldoorn explains. “Furthermore, the material may not be toxic to cells and the islets may not attach to it, for this could make the islets change shape and lose their function.”

Creating an artificial pancreas using islets should at least be as efficient as the current method. Otherwise, it is useless to con-tinue. If and whether a method is adequate enough, however, is not an open-and-shut case. De Vos: “We receive islets from deceased donors, so we have to take care. They are not lying on a shelf waiting for us to be used whenever we feel like it. That is also the reason why we want to monitor as much as possible.”

“ Islet transplantation is a good alternative option but more then fifty per cent of the islets die.”

DCTIDCTI

109108

Pump keeps donor pancreas in shape

it. Leuvenink: “If you buy a used car, of course you want to make sure that the motor runs properly. But when you receive a donor organ in a polystyrene box, you are completely in the dark. ‘Test driving’ the organ first is, therefore, very relevant. If it is a pancreas, the first thing you want to find out is: does it produce insulin?”

Similar pumps are already being used for kidneys, livers and lungs. Machine perfusion for the pancreas, on the other hand, is still uncharted territory. Within DCTI, Leemkuil worked on a prototype for the pump, using the existing devices for kidneys as a basis. “We have found that the kidney pump as such is not suitable for a pancreas and that we will have to make quite some adjustments, but we wanted to find

out first whether this type of installation would be useful for this organ in the first place. If it looked safe, we could develop a pump specifically for the pancreas.”

The first tests with pig pancreases were successful. Leemkuil also tested the pump’s effects on the tissue of rejected human donor pancreases and compared these with the tissue of pancreases that were stored in the conventional way. The pump system did not harm the tissue, and the vitality of the pancreases was greater: they contained more ATP, the carriers of energy in cells. Today, the pump’s effects on the quality of isolated Langerhans islets are being stud-ied – in a collaborative partnership with the Leiden University Medical Center.

“The principle works. That was the first step”, Leemkuil concludes. “But we cannot yet recom-mend the method to everyone in the field. We first have to develop a specific pump for the pancreas that can be set according to the organ’s needs and desires.” There are still some catches when it comes to logistics too: as soon as a surgeon lifts a pancreas from a donor’s body, someone must immediately connect the organ to the pump. The pump is then to travel along with the organ to its recipient – but what if the organ has to be transported by airplane? Is the pump also allowed on board? If so, who travels with it? “We will have to solve these kinds of issues first”, Leuvenink says.

Once the pancreas pump is finally there, it might also be used in order to optimize conditions for islet isolation and improve islet transplantation . The expectations are that the islets in a donor pancreas that is connected to the pump will stay in a better condition. This would be an added bonus, since separated islets do not function as well as an entire pancreas. That is the reason why a transplantation now still requires the islets of two or three donors. Leuvenink: “We want to try and use the pump to improve the condition of the pancreas and thereby the quality of the islets, so that in the future one single pancreas will do.”

A pancreas that has been connected to a pump. This pump supplies it with oxygen and nutrients so that the organ stays in a better condition.

“ Sometimes a patient has to stay in the hospital for months.”

damaged. Such damage has major consequences: only two per cent of a pancreas is made up of islets of Langerhans (containing insulin producing cells); the greater part is comprised of cells that produce digestive enzymes. If these enzymes start to leak into the recipient’s abdomen, they might also digest their own organs. Furthermore, despite the doctors’ carefulness, complications are anything but rare.

Leuvenink and Leemkuil want to lower the risk of damage to the donor pancreas as much as possi-ble, so that the number of suitable donor organs increases. “Damage to organs occurs in different stages of the donation procedure”, Leuvenink explains. Even at the moment of the donor’s death there is the risk of damage. Next, the organ is cooled in the operating room, taken from the body, and transported on ice in a polystyrene box to the hospital of the receiving patient where it is con-nected to the recipient’s blood vessels. This is the

moment when, finally, warm blood flows through the organ again. There must be a different way, the Groningen researchers believed.

An alternative for the polystyrene box is the so- called machine perfusion, a method that has been copied from the heart-lung machine. If you can connect the organ that is to be transplanted to this type of machine as soon as you have taken it out of the donor’s body, you limit the chances of damage. “When an organ cools down, its metabolism falls to approximately ten per cent. So the organ still needs oxygen and nutrients. The pump allows you to con-tinually flush the organ with the specific fluid of choice and, moreover, it helps keep it at the desired temperature. An oxygenator provides the oxygen. You could even administer medication or repair the organ”, Leuvenink says.

If you connect the organ to a pump straightaway, you can also perform tests on it before you transplant

DCTIDCTI

111110

story: Elles Lalieu image: LUMC

After a transplantation, the Langerhans islets do not give their best performance yet. Why are these transplantations performed anyway?“Because the treatment is often effective. The first islet transplantation took place in the 1980s. This procedure has been performed worldwide 2,000 to 3,000 times so far. The treatment improves difficult glucose control in the patient, albeit that there are side effects, such as infections or a higher risk of certain types of skin cancer. At the moment, we want to improve the treatment through the use of carrier materials.”

As for those carrier materials, can you simply try them on humans?“No, we cannot. They will have to be tried in a research setting, with the consent of the hospital’s Medical Ethics Board. We have tested the carrier materials in small lab animals within the DCTI project. We will now have to prove their effective-ness and safety in larger lab animals, such as pigs. If, and only if, those tests are successful, the treatment may first be applied in patients.”

What would, or could, this first application involve?“We could transplant most of the islets in their carrier material into the liver in the usual way, apart from a small amount of islets that we could implant in another part of the body, for instance underneath the skin. In this way, the first patients will receive the regular treatment, while we can also see how the islets in their carrier materials behave in the human body.”

Due to the shortage in donors, there is talk of a treatment with stem cells. Would that involve the same procedure?“Each carrier material has already been approved by the authorities for use in humans. That is also one of the reasons why we chose these materials at the start of our research. If you add stem cells, you will enter a completely different area.”

What is the big difference?“We now isolate the donor’s pancreatic islets and try to keep them alive until they are transplanted a few days later. We do nothing special to the cells them-selves. Stem cells, on the other hand, will first have to multiply so that you end up with a nice amount. You will then have to let them grow to maturity with the help of growth factors. It takes longer to culture stem cells than donor cells, and during their culturing they may undergo all kinds of changes, including genetic ones.”

What consequences does this have for tests on humans?“It means that you will first have to soundly test the cell product. You will, for instance, have to prove that your culturing procedure does not leave you with cancer cells. It will take some time before the treatment with stem cells wil be available.”

If we ever get there, for cancer cells make it all sound pretty risky.“There are not enough donors to treat everyone with type 1 diabetes, so if we want to do more than treat symptoms with insulin injections alone and come up with a definite solution, we will have to find it in other cell sources. There is no other way. It is impor-tant, though, that patients will always have to give their consent if they are to participate in research. They will have to be very well informed about the potential risks beforehand and the safety of the cell products will have to be tested extremely thoroughly first.”

What is the current state of affairs?“One in six hospital beds in England are now occupied by a patient with diabetes who is suffering from com-plications from treatment, so a true solution would be more than welcome. However, as with all medical research, it is not a matter of weeks or months, but years and years. Patients with type 1 diabetes will have to be patient before they can grasp this last straw.”

Discussion

First tests versus the last straw

To patients whose glucose level cannot be controlled no matter what

they do, any new form of therapy feels like the last straw. How to deal

with this as a doctor as well as a scientist? When is the application of

a new therapy safe enough to involve the patient? Eelco de Koning,

Professor of Diabetology at the Leiden University Medical Center

explains.

Help the body to repair itselfNIRM

If something is broken, it needs to be fixed.

This also goes for our body. When the body itself

cannot manage, regenerative medicine tries to help.

NIRMNIRM Infographic

115114

Sometimes, tissues and organs are damaged to the extent that they can no longer repair themselves. When the tissue or organ is essen-tial for a reasonable quality of life, patients usually have to rely on treatment of their symptoms or a transplant. However, donor material is on short supply. Furthermore, after transplanta-tion patients have to take medica-

tion for the rest of their lives to prevent rejection. Regenerative medicine works differently. Instead of relying on a donor, autologous cell material is cultivated outside the body and then implanted back into the patient.This is a complicated process. Not only does the treating physician have to be able to cultivate suffi-

cient tissue from the patient’s cells outside their body, this tis-sue also has to develop in exactly the right way so that it will func-tion as intended. As an aid, tissue engineers often start off by mak-ing a scaffold from biodegradable material outside the body. By making a scaffold from “smart” material and then injecting autologous cells onto it, tissue can be grown in the lab to have the right shape and function. “Smart” means that instructions are built into the material so that the cells will know what to do. Stem cells work best for this job because they still have to develop.

3 Stem cells are injected into the scaffold. The scaffold “directs” the stem cells to develop in a specific way.

4 The new construct is surgically implanted in the right spot. The body then takes over the repair work. Because it has been cultivated from the patient's own cells, the implant will not be rejected.


When stem cell research and tissue engineering meet

Building entire organs or tissue structures outside the body and using

them to help the body repair itself is what tissue engineering is all about.

However, this is trickier than it sounds, because how do you make sure

that only the part of the tissue you need grows?

1 Doctor extracts stem cells from the patient’s body.

2 A scaffold is built in the lab. This can be done in sever-al ways. It can be printed using a 3D printer, or the biodegrad-able material can be moulded into the right shape.

NIRMNIRM Interview

117116

for joint repair

A horse is placed on a special treadmill after treatment. Using cameras and measurement equipment, researchers monitor the exact joint pressure and whether it is the same in all four legs.

Why is research on horse cartilage so important?“Most domesticated animals are important for their products. Horses, however, are important for their movement. They have been used by armies, in agriculture and for transport, and today are used primarily in sports. Their locomotor sys-tem is essential. What I do is comparable to sports medicine for horses. Most of the health problems we come across in horses affect their musculo- skeletal system, with damaged cartilage in the joints being one of the main causes of poor performance.”

Can’t such damage simply be repaired?“Cartilage is complicated stuff. Unlike almost all other types of tissue, in adults it doesn’t repair itself after an injury. Cartilage cells need an extremely long time to produce new collagen: it would take around three hundred years for joint cartilage to completely regrow. We’ve known that cartilage is tough to repair since scientists first wrote about it in 1743, but we have yet to find a solution. Cur-rently, a hole in someone's cartilage is usually repaired by drilling small openings in the underly-ing bone so that bone marrow cells can fill the hole

with new tissue. However, the quality of the result-ing scar tissue is much worse than the original car-tilage. Another option is to replace the entire joint with an artificial one, but artificial joints are not ideal and have a limited lifespan. We are now look-ing for a solution using stem cells so that cartilage can finally repair itself.”

How would that work?“Though cartilage cannot be reproduced, you can implant a biomaterial (called a scaffold) in which cartilage-producing cells can survive. We are col-laborating with the University Medical Center Utrecht (UMCU), where they are working on a technique to produce such biomaterial using 3D printers. There are still quite a few variables in their research. For example, you could use carti-lage cells, cartilage stem cells or more general IPS cells. The researchers are in NIRM also working on a biomaterial with ‘vesicles’. These are pouches filled with substances such as RNA that regulate

Veterinarian and Utrecht University Professor René van Weeren’s research

focuses on bioprinting cartilage for horses. Damage to cartilage is something

equestrians would love to be able to remedy. But his work offers prospects for

treating cartilage injuries in human patients too, says Van Weeren.

“ It takes around three hundred years for joint cartilage to completely regrow.”

Horses for courses – printed cartilage

story: Rineke Voogt

NIRM Case studyNIRM

119118

There is an urgent need for implantable heart valves that more closely resemble the real thing. The artifi-cial valves currently in use have a lot of drawbacks. Carlijn Bouten, Professor of Biomedical Engineering at Eindhoven University of Technology, explains, “A mechanical prosthesis is not alive. It is made of syn-thetic material and metal, and you can hear it tick. Even worse, they knock blood cells to pieces. This means you have to take medication to prevent thrombosis for the rest of your life. But these blood thinners also make bleeding difficult to control. In other words, though the patient’s quality of life improves, the cost is another disorder.” A biological valve from a donor (human or animal) is also far from ideal, because it has to be replaced after fifteen years. This means that young patients – every year six thousand European children need a new heart valve – face a series of risky operations because arti-ficial heart valves cannot be tailor-made. Moreover, donor valves are scarce.

Design your own heart valve

People with a leaking or constricted heart valve can have an artificial replacement

implanted. But there are drawbacks. A mechanical artificial heart valve requires

lifelong use of medication, and a valve prosthesis made of animal material wears

out quickly. Moreover, neither type of valve grows with a patient. Carlijn Bouten is

working on an alternative: an implant that attracts autologous cells as soon as it

is implanted and uses them to build a new heart valve.

1 The heart pumps blood around the body continuously.

2 A malfunctioning heart valve makes the heart work too hard. Ultimately, this leads to complications.

3 The tissue-engineered artificial heart valve is implanted to replace the patient’s own defective heart valve.

story: Rineke Voogt images: Hartstichting

Horses for courses – printed cartilage for joint repair

communication between cells. If these vesicles can prompt your own cartilage cells to produce collagen again, you don’t even need stem cells anymore.”

IPS cellsInduced pluripotent stem cells (IPSCs or IPS cells)

are stem cells made from adult skin cells. These

stem cells can develop into a wide range of

specialised cell types. The advantage of using

these cells is that they can be harvested from

patients themselves, thus preventing problems

with the immune system.

Have these techniques been applied in practice yet?“You can test cell growth in a biomaterial in vitro, but we’re conducting the tests in horses. Fortu-nately, it is far easier to take various measurements in horses than in classic lab animals like mice or rats. For our tests, we inflict a small injury to the horse’s cartilage and then administer the intervention with the printed biomaterial. We thoroughly examine the horse before, during and after the experiment. We place the horse on a treadmill, for instance, to mea-sure the exact force exerted on the ground, the angle of the joints and which leg has gone lame. We also

analyse all its movements using cameras, and a kind of keyhole operation even lets us visually monitor if and how the cartilage is repairing itself. Biomarkers in the joint fluid can tell us if any inflammation has occurred and whether any additional collagen has been produced or broken down.”

And?“We still have a long way to go – generating proper cartilage tissue would be a feat worthy of the Nobel Prize. To begin with, the material is still not ideal. In some tests, the cartilage quickly broke down again. The procedure itself also has to be improved. The downside of using horses is that the implanted material is put to the test straight away, since the animal puts pressure on its leg immediately after the operation, when it gets on its feet. In this respect, working with humans is easier. You can simply tell them to stay off their feet for a while.”

Painless wear and tearDamage to joint cartilage is not noticeable at first as

there are no nerves in cartilage tissue. But, once inju-

red, the rest of the cartilage also wears out faster. This

is when it becomes painful, because the underlying

bone – which does have a lot of nerves and is there-

fore very sensitive – is eventually exposed.

This research also opens up prospects for people with joint problems. How big is the leap from horse to human?“At the start of our study we examined the thick-ness, composition and other characteristics of car-tilage in various animals. We compared some 120 animals across 58 mammal species – all of them zoo animals sent here for analysis after they died. We found that the cartilage in many animals that weigh over one kilogram is very similar. In horses and humans it’s even extremely similar. Therefore, if we manage to repair it in horses, it’s almost a foregone conclusion that we can do it in humans. That’s also why we are able to work so closely with researchers on the human side of things at the UMCU: the tar-get group for this research is not only horses, but humans as well.”

“ Generating proper cartilage tissue would be a feat worthy of the Nobel Prize.”

NIRMNIRM

121120

4 The artificial valve is made of biodegradable material.

5 A close-up of the material (a type of plastic) shows that all kinds of substances can be attached to the scaffold.

6 The scaffold is supplied with molecules that attract monocytes – a type of white blood cell – to nestle within.

7 Attracted by the bait substances, white blood cells colonise the scaffold.

8 Due to the presence of the white blood cells, the scaffold is gradually reinforced with collagen.

9 While the scaffold is broken down, its task is taken over by autologous tissue.

10 The body has replaced the old, defective valve with new material.

Design your own heart valve

In the “in vivo situation” mimicked at Bouten’s lab, she can also simulate the conditions of a sick or old body. Because wounds heal more slowly in diabetics, for example, than in healthy people, this would have to be taken in to account when building heart valves for diabetic patients. Similarly, cell characteristics differ between old people and young people.

The scaffold method offers solutions for these more difficult conditions too, by means of adding auxiliary substances. Bouten and her colleagues want to find out which substances work for several key patient groups. “Naturally, the heart valve can be injured again, so we have to determine whether this technique is also suitable for diabetics, for instance.”

SheepThe next stop for this treatment is lab animal research. Although the human model system is actually more complicated than lab animal testing in some ways – after all, you can use it to simulate a sick patient – the technique has to be proven effective in lab animals before it can be used on humans. The researchers have already selected the ideal candidate: sheep, which were previously used to test the lab-grown valves. Sheep also represent the worst case scenario because of the pace at which their heart valves can calcify. If the researchers manage to grow a functioning, healthy heart valve in sheep, it is very likely that it can also be done in humans.

That final stage – implanting scaffolds in people with poorly functioning heart valves – is still some way off. However, once the valve has proved effective in sheep for a full two years, the road to the first clinical trial is open.

Bouten and her colleagues believed there had to be an alternative. They had already worked on a new type of tissue-engineered artificial heart valve that they cultivated in the lab. “Heart valves open and close a hundred thousand times a day. We know that live valves perform much better. They do not wear out and are self-healing. They also adapt to new circumstances, such as increased blood pressure. Tissue engineering is already being used to create live prostheses, and we want to do the same for heart valves.”

LifelongA tissue-engineered valve made of autologous cells would have none of the drawbacks of standard artificial heart valves – no risk of thrombosis or rejection symptoms. Furthermore, live tissue- engineered valves can grow with patients and last a lifetime.

However, a lab-grown heart valve has drawbacks of its own. For example, it is not clear who actually owns the valve – the researchers, the patient, the doctor? There is also a high risk of infection. More-over, cultivation takes more than eight weeks and is very expensive. The Eindhoven researchers have therefore opted to take a different approach: the implantation of a lifeless framework – known as a “scaffold” – which develops on its own into a fully functioning heart valve inside the body.

Auxiliary substances The principle is simple: take a solid but porous plastic-like material that is degradable in the body, implant it, and allow the body to populate the scaffold with its own cells. According to Bouten, “The body is great at making its own tissue. Take the way it repairs wounds. Cells are attracted to sites of inflammation. The only thing you need to do is steer the process in the right direction.” Bouten’s group directs this growth process, for example, by injecting certain active molecules into the scaffold

that promote healing by attracting monocytes (a type of white blood cell) in the blood. Inside the scaffold, the cells gradually form a strong tissue. The Eindhoven scientists have already successfully used this method to replace blood vessels.

The scaffoldThe implant – the scaffold that forms the basis for the

heart valve – consists of wires of biologically degradable

polymers a thousandth of a millimetre in diameter.

These scaffolds are built by Xeltis, a biotechnology com-

pany. “It’s like making candy floss”, Martijn Cox at Xeltis

explains. Under an electric current, a long, thin thread is

spun around a mould for a heart valve to make a 3D

network that autologous cells can latch onto.

Special groupsThe mechanics of the valve seem to work well. The challenge now is to have the valve grow precisely the right tissue within the body. A heart valve has to be strong, but also flexible, and it has to have the right shape. Moreover, the scaffold has to break down at a certain pace, neither too slowly nor too quickly. “We still have to fine-tune this process”, Bouten says. That work, however, takes place out-side the body, at the Eindhoven lab. The researchers place the scaffolds in a bioreactor where they are supplied with an artificial blood flow to reproduce conditions in the body as closely as possible. A “human model system”, they call it. “We still have a lot of testing to do”, continues Bouten. “We want to find out how quickly the tissue grows, and how the breakdown of the scaffold material can be switched on and off. A scaffold has certain characteristics, but they change as the tissue grows. Compare it to passing the IKEA test: you first have to make sure the valve can ultimately do what it’s meant to do and will carry on doing it.”

NIRMNIRM Interview/case study

123122

Stem cells are cells that can still

become anything they like. But how

do these cells decide what they will

be in the end? We are getting better

and better at answering that ques-

tion. Different groups within NIRM

are trying to make stem cells develop

into bone or cartilage tissue – and

they are well on their way.

story Elles Lalieu images UMC Utrecht

A shove in the direction of bone or cartilage

NIRMNIRM

125124


face and the cell’s behaviour, Van Blitterswijk and his team developed the Topochip, a chip with over two thousand different surface structures.

Stem cells are seeded onto the chip and subse-quently attach to one of the available structures. Cells seem to “read” the structure on which they grow, just like the blind read braille. Stem cells spe-cialise in a certain direction, depending on the sur-face structure. On one surface they evolve into muscle cells, for instance, whereas they grow into bone cells on the other.

Building blocksThe structures on the chip are combinations of three basic forms: a circle, a rectangle and a trian-gle. The Topochip comprises over 2,000 structures, but a total of 158 million combinations are possible. All these combinations are stored in a library. For example, if researchers want stem cells to turn into bone cells and know that the circular structures are the most suitable for that purpose, they can make a new chip with all kinds of related circular struc-tures to find out which surface is the best stimulant for the desired development.

While the Topochip focused on individual cells, Van Blitterswijk has meanwhile also developed a system to grow mini organs on different structures. “We use quite a few 96-well plates in the lab”, he explains. “These allow us to test 96 conditions in one go. What we have done is cover the bottom of each of these wells with 580 tiny micro cultivation containers – so that we can now grow 50,000 micro-organs on one plate.”

On top of the micro cultivation containers is a layer of gel where the stem cells are seeded. Van Blitterswijk: “The cells cannot attach to the gel, so they sink through the gel into the containers. That is where the cells coagulate and eventually form a small ball of approximately one tenth of a millimetre. We can harvest these balls and bring them together into a greater mould, thus ending up with a piece of a puzzle of approximately one millimetre. Such pieces can be used as building blocks towards tissue replacement.”

The collection of different types of cells in one con-tainer makes it possible to create complex tissues, such as a bit of bone or cartilage with blood vessels.

The bioprinter in the biofabrication facility. The printer is not only used for orthopaedic research, but can be widely applied.

Cell biologist Clemens van Blitterswijk focused his doctoral studies in the 1980s on the development of ceramic ossicles. He is now professor of Regenera-tive Medicine at Maastricht University. Over the past thirty years he has conducted a great deal of research in the area of tissue regeneration: creating natural tissue for the purpose of replacing defective body parts.

“My research group has now created the smallest ossicle, the stapes, from cartilage”, Van Blitterswijk says enthusiastically. “In the ear this ossicle is made of bone; unfortunately, that is still a step too far. But I do see it as a proof of concept. We can easily

make a great number of ossicles, at very low cost. This is key, since affordability is one of the chal-lenges for regenerative medicine: we can make all sorts of things, but if the costs are prohibitive, it is no use to anyone.”

TopochipThe stapes made of cartilage is a practical example, and a result of extensive research. Stem cells grow on different surfaces, but not in the same way on each and every surface. The form and structure of the surface determine the stem cell’s behaviour as well as the type of cell it eventually grows into. In order to examine the relationship between the sur-

Previous pages: A small part of printed material in a petri dish.

Cartilage cells in a micro cultivation container. The cells coagulate and eventually form the building blocks for micro-organs.

NIRMNIRM

127126

implants to undergo their first test. Malda and his team have tried to replace the entire shoulder of a rabbit by printed tissue. Whether it worked, remains the exciting question for now: the results will be evaluated over the next few months. There is funding for this research for the next five years, so the story of these experiments with joint replac-ing tissues will definitely not end here.

While improving their ability to create different tissues, do Malda and his group now also under-stand exactly why a stem cell becomes a muscle cell in one instance and a bone cell in the next? Malda: “Our insight is increasing, but we haven’t yet reached the stage where a stem cell forms bone whenever we like it to.” According to Van Blitter-

swijk we are only now starting to understand how it works. “We can now see how cells respond. If we grow organoids in the shape of a triangle, for instance, we see high concentrations of blood vessels in the vertices. Why? Because of the high concentration of the VEGF growth factor there, which is needed for the formation of blood vessels. But what causes the high concentration of growth factor in these exact spots? We hope to unravel these types of mechanisms in the next five years. We will not be able to truly predict a stem cell’s behaviour until then.”


In Van Blitterswijk’s lab, the stirrup bone or stapes, one of the ossi-cles, was made out of cartilage. Normally, the stapes is made of bone, so this cultivated speci-men is not functional, but it does show that it is relatively simple to create a great many ossicles.

The NIRM researchers implanted one such culti-vated building block in mice, where it shaped into bone in a spot where normally no bone is found.

Fleece sweaterPrinting entire tissue with living cell blocks is a step further. Bioprocess technologist Jos Malda saw an opportunity in bioprinting a few years ago, and he jumped on it – with success. The University Medical Center Utrecht is now the proud owner of a special biofabrication facility, which is entirely focused on 3D printing.

What makes the bioprinter so suitable for repairing cartilage defects? Natural cartilage has a layered structure which is ideal for reproduction using a 3D printer. The “ink” that Malda uses is a hydrogel filled with stem cells, cartilage-forming cells or a combination of both, plus growth factors. As the cells in the gel have to stay alive during and after the printing process, the hydrogel has been made to provide them with an ideal living environment.“The hydrogel is a weak substance, a bit like gela-

tine”, Malda explains. “It is, of course, not some-thing you can walk on. That is why, as part of the NIRM project, we have tried to do clever things with it, so that the material gained more strength. We first tried to print the gel on a mesh of thick fibres, like fortified concrete.” Although the result was certainly solid, it needed relatively extensive fortification making the implant rather inflexible. The fibres had to be thinner. Malda: “We managed to make thin fibres, comparable to the fibres of a fleece sweater. These fibres are in themselves not robust, but this changes when you print them together with the hydrogel. The forces exerted on these fibres keep the entire construct together. One plus one does not equal two in this case, but fifty.”

Unravelling mechanismsMalda is now making implants with a combination of thick and thin fibres. The thick fibres are more suitable for creating bone and the thin ones for cartilage. Printing the tissue in combination with supportive materials has also enabled him to make different shapes, such as slices or tubes. The progress that he made soon allowed the printed

The bioprinter in action. The material is deposited in layers.

NIRMNIRM Case study

129128

Sometimes, it is not clear how medicines will affect patients. The effects of a medication to treat intestinal cancer, for instance, depends on the genetic characteristics of the tumour, which can-not be seen from the outside. What’s needed, reali-sed Hans Clevers and his colleagues at the Hubrecht Institute (part of the KNAW Netherlands Academy of Sciences), is to be able to test the medication on the patient’s cells first, without the patient suffe-ring any consequences. That way, you can better predict the response. If the test shows that aggres-sive chemotherapy has no chance of success, you can spare the patient a pointless treatment.

Patient test kits can be built by taking a small amount of tissue from the patient’s body, cultiva-ting it into miniature organs and then testing the medication on them. Researchers at Clevers’ lab figured out a way to do this: using growth factors, the patient’s stem cells can be stimulated to multiply and specialise (see also “Repairing liver damage using patient cells”). The result is a so- called organoid that is not only similar in shape and function, but also has the same genetic code as the original organ.

“You can then expose these petri dish organoids to medications to see what the effect will be”, Rob Vries explains. He is the managing director of the HUB Foundation, founded by the Hubrecht Institute to test medications on a large scale and gain a better understanding of why patients respond to them so differently. Their work is still in the research phase. When a doctor prescribes a treatment for a patient, Vries and his colleagues try to make a prediction on the basis of the patient’s cell material. “If we manage to consistently make accurate predictions, we will be able to make real recommendations one day.”

Testing medicines on organoids has also turned out to be useful for disorders like cystic fibrosis. Hubrecht researchers have even managed to cure the cultivated mini bowels of a cystic fibrosis

patient by replacing a defective gene with a healthy one. This offers hope for a potential treatment for the disease.

Testing existing medications on cultivated patient cells allows for tailor-made medications. Taking this a step further, brand-new medicines could also be tested in petri dishes. This is exactly what the biotechnology company Pluriomics does, focusing on the heart. Cardiac muscle cells, culti-vated from what are known as induced pluripotent stem cells, make it possible to study the effect of a medication on the heart. The cells behave almost identically to how they do inside the body. “You can see them contract, or ‘beat’”, says Marijn Vlaming, a researcher at Pluriomics. “Using a type of ECG test, we can then see if a substance causes arrhythmia, for instance.”

The test organoids have various applications. They are particularly useful for the development of medications, says Vlaming. “Many medicines, cancer drugs for instance, pose a risk to the heart. Currently, their safety is often tested in animal models, but something that is dangerous to a dog’s heart is not necessarily dangerous to a human heart too, or vice versa. Our test seeks to provide greater certainty. This technology is already replacing lab animal testing, but it will still be a while before lab animals become completely unnecessary.”

With mini test organs offering so many uses, Pluriomics is currently developing a “toolkit” with which researchers can perform their own tests on cardiac muscle cells they cultivate in their labs. “This is convenient”, Vlaming says, “because academics often want to do things themselves. Plus, the more the method is used, the sooner and more widely the technology will be accepted.”

How effective medicines are in treating

specific disorders differs considerably

from one patient to another. By ex-

tracting and cultivating stem cells from

a patient’s body, the effectiveness or

risk of medications can be tested

beforehand – making it possible to

identify the most effective medication

for that patient.

Build a tailored test kit

Cardiac muscle cells grown from human stem cells shown in a “multi-electrode array” dish. Electrodes are used to measure the effect of medication on the contraction of the cardiac muscle cells.

Microscopic image of the structure of cardiac muscle cells grown from human stem cells. This sarcomere structure, which helps cardiac muscle cells contract, consists of the overlapping proteins myosin (red) and actin (green).

Contracting heart muscle cells grown from human stem cells in a petri dish.

story: Rineke Voogt images: Marijn Vlaming

NIRMNIRM

131130

?

Stem cells surviving on synthetic hydrogelIt looks like a landscape in a science fiction movie, but this is an image of a natural extracellular matrix. Chemist Patricia Dankers tried to syntheti-cally copy this matrix in order to serve as a syn-thetic base for the growth of stem cells. Stem cells in the lab now grow on a natural hydrogel called matrigel. It may never be clinically applied, however, since tumour issue is requiered fot the preparation of the matrigel. Dankers: “Matrigel contains around ten thousand different compo-nents. The trick is to identify the relevant ones and reproduce them synthetically. Five years ago, we thought that it would be a straightforward job, but nothing could be further from the truth.”

The idea was to use small building blocks to create a larger structure, held together by hydro-gen bridges. Dankers and the research team devel-oped building blocks based on the ureidopyrimidi-none (UPy) group. If you couple other molecules to these building blocks, you end up with different UPy-blocks with widely varying functions. It is the combination of specific building blocks that will get you a specific bioactive hydrogel.

Over the past few years, the researchers have cre-ated a structure to which three or four UPy-blocks can be added. Dankers: “The stem cells survive and that in itself is quite an achievement, as there have been many synthetic gels made over the years in which they simply died. However, it will take years before we create a synthetic gel that also provides us full control over stem cell behaviour.”

story: Elles Lalieu

Report

NIRM

Testmedium A

Testmedium B

NIRM Infographic

133132

3a After differentiation, the healthy cells are injected into the patient again, where they can multiply further. The great advantage: the cells are autologous, so the patient will not suffer any rejection symptoms.

3b The cultivated tissue serves as a test medium in order to see how the patient responds to certain medication. This is useful when patients differ strongly in their respon-ses to the specific medicine or when there is a chance of it being harmful.

If an organ is defective, you can try to recreate the entire organ in a cultivation flask or with a printer. This is not always necessary however for each and every tissue. Sometimes it will suffice to grow something new at the place of the defect. How do you do this?This type of research basically starts out in the same way as tissue engineers do, by extracting stem cells from somewhere in the body. The researchers then cultivate those stem cells outside the body, after which they are differentiated into the right kind of cells. These cells are subsequently injected into the body again, where they can take on the task that the body itself is no longer able to perform.This sounds simpler than it is. In reality, each step in the process is a challenge. First of all, there is the difficulty of finding the right stem cells. Next, they

have to be grown in a petri dish. Probably the most complicated challenge is to make certain that it is safe to return them into the body.The latter is not always necessary. For various disorders drugs are already available, but they provoke widely different responses in patients. One cardiac or cancer patient may greatly benefit from a certain drug, whereas in another it may be totally ineffective or actually harmful. In such cases, what doctor would not like to be able to predict whether their prescriptions are indeed going to help? We are improving our ability to do just that, as researchers are now able to cultivate patients’ autol-ogous tissue outside the body with the help of stem cells. The cultivated material then serves as a “guinea pig” to predict the effects of the medication.

Reparation with stem cellsThe cultivation of body material outside the body so that you can tinker

with it appeals to the imagination. However, you can do far more with

stem cells. You do not always need to create entire grafts and even

without placing them back, stem cells are of great value to patients.


1 A doctor extracts stem cells from the body.

2 The stem cells are cultivated outside the body, until their number suffices.

NIRMNIRM

135134

at hand – there is still no remedy for people with dry mouth syndrome. Coppes and his team are looking for a way to repair the salivary glands using stem cell therapy. “The method is actually similar to a bone marrow transplant”, Coppes explains. “We take some of the patient’s healthy cells, cultivate them, and then place them back, returning healthy mate-rial that came from the patient’s own body.”

The stem cells can be directly extracted from the patient’s salivary glands, which are still intact prior to the radiation treatment. A small cut under the ear suffices for a biopsy. Coppes: “Of course, you cannot take away the entire gland – there is a limit to the number of cells that you can harvest. But we have found a way to grow six thousand stem cells outside the body from a single stem cell within seven weeks.”

There was no doubt about the possibility of cultivat-ing salivary gland stem cells in itself. “After all, if you chew chewing gum for two weeks, your salivary glands grow as well.” But it proved to be more diffi-

cult to mimic this process in the lab. The question was: what made a stem cell behave like a stem cell? And how could the researchers get them to multiply? Those were the tough nuts that Coppes’ team had to crack. Growth factors (chemical substances natu-rally present in the body that promote cell growth) proved to play an important role, but finding the right growth factors was far from easy.

The research conducted by Coppes’ team is the first investigation into salivary gland stem cells. There is one significant difference with research in other organs: other organs are often already damaged by disease at the start of treatment, and that makes it difficult to harvest healthy stem cells. That is not a problem here: the doctor can determine whether the patient is at risk of developing dry mouth syndrome before the radiation therapy actually starts. With the help of a CT scan, the radiotherapist can predict exactly which areas will receive how much radia-tion. If it looks like there is a chance the salivary glands will be damaged, the doctor can decide to perform a biopsy on the salivary gland to extract stem cells. These can then be cultivated while the patient is receiving radiotherapy, so that they can be placed back and do their work once the therapy has finished.

“ It works in principle, but there are still some catches.”

Life without properly functioning salivary glands is pretty uncom-

fortable, and repairing them has turned out to be difficult. At the

University Medical Center Groningen, Rob Coppes is working on a

solution: the cultivation of salivary gland stem cells.

Constantly having a dry mouth, not being able to talk or swallow properly, damaged teeth, sleep deprivation: not having enough saliva is the source of some unpleasant complaints, which are a daily reality for people who have to undergo radiation treatment due to a tumour in their throat or head. The radiation that is meant to combat the tumour damages their healthy salivary glands. These glands, which can be found under the ear and tongue, among other places, are highly sensitive to radiation. “It is a nasty side effect of radiation”, Rob Coppes, Professor of Radiotherapy at the UMCG explains. Saliva is essential to oral health as well as speaking and swallowing. “Try eating two pieces of dry toast one after another – that is how it feels if you have to eat without saliva.”

Approximately forty per cent of the people who undergo radiation therapy in the head-and-neck area experience problems with their salivary glands, Coppes estimates. That boils down to some four

hundred patients each year. They end up with xero-stomia: dry mouth syndrome. Marijke Baks is one of them: ever since she underwent radiation treatment in 2007, her saliva production has virtually stopped. “It is something that you really need to learn to live with: for instance, you have to walk with your mouth closed, because otherwise it gets too dry. Singing is no longer possible, and it is especially difficult at night, as you often unknowingly sleep with your mouth open.” And that is not the end of it. Sports are out of the question without artificial saliva, intimacy is more problematic, and hot weather is also a bother: drinking too much water actually gives you a dry mouth. “You have to adapt the way you live”, Baks says. “Meals are difficult; they always have to be moist. I often eat mashed potatoes and vegetables, and whenever I eat a sandwich, I have to take a sip with each bite.”

Although there are makeshift solutions available – artificial saliva, or always having a bottle of water

Never eating toast again

story: Rineke Voogt image: René den Engelsman

Interview

NIRM Case studyNIRM

136 137

Some diseases are difficult to investigate simply because there is no research material. To solve this problem, Joost Gribnau, develop-mental biologist at the Erasmus University Medical Center, uses a clever trick. He creates stem cells (IPS cells) from patients’ skin cells and lets them mature into adult cells to make his own research material.

Gribnau conducted research into disorders where X-chromosome inactivation plays an important role, such as Rett syndrome. In a female foetus, one ran-dom X-chromosome out of the two in each cell is switched off during the embryonic phase. Therefore, if there is a mutation in X, the foetus will have both sick and healthy cells. It is the ratio between the numbers of sick and healthy cells that determines the extent to which the disorder becomes manifest.

Without growing his own research material, Gribnau would not have been able to do his research. “We wish to better understand and predict the selection process in X-chromosome inactivation”, Gribnau explains. Matured cells of patients were of not much use to them. “The process of inactivation has already come to an end in matured cells. So to go back to the beginning of life we decided to make new cells ourselves.”

A special facility was set up for the purpose of making IPS cells from skin cells. The researchers can then do all sorts of things with the stem cells they have created. They can change them into brain cells, for instance. This means that it should eventually be possible to study and explain the behaviour of the sick cells of each and every patient.

Rare“Unfortunately, the stem cells did not behave as expected, so recently we have mainly focused on pushing both X-chromosomes into action”, Gribnau says. They have succeeded, so they can now finally use these cells to research Rett syndrome and other X-chromosome related disorders.

Gribnau's fundamental research is of great impor-tance to patients. “We can now study disorders for which there used to be no research material available, such as brain diseases or very rare disorders. By cultivating patients’ sick cells we gain not only a better understanding of the process behind a disor-der, but also an opportunity to test new medicines or treatments.”

Make your own research material

The coat of a calico is a fine example of X-chromosome inactivation. The genes for coat color, red or black, in cats are coupled to the X-chromosome. In female cat embryos one of the two X-chromosomes is switched off. Sometimes the color is black switched off and sometimes red. This results in the familiar patch pattern.

story: Elles Lalieu image: Riosafari

Microscope image of a salivary gland organoid from Rob Coppes’ lab (UMCG). The colours represent different types of cells, with the nuclei shown in blue.

IMA

GE

: M

AR

TTI

MA

IME

TS

“We have already booked successes in mice”, says Coppes. Mice whose salivary glands were damaged by radiation had cultivated cells injected into their salivary glands. Around one month later, the cells had distributed themselves over the entire gland. “You do not even need a microscope to see whether the experiment succeeded. Mice clean themselves with saliva, so a shiny fur clearly shows whether or not their salivary glands are working properly.”

The cells do have to be reinjected in exactly the right place. Salivary glands are shaped like a bunch of grapes, Coppes explains: the stem cells are in the ‘twigs’, and the ‘grapes’ contain the active cells that produce saliva. So the stem cells must get into the twigs and spread from there, specialising into sali-va-producing cells. In mice, it takes quite some fiddling, but in humans you could easily check this with ultrasound equipment.

There is still a long way to go. The fact that the prin-ciple works in mice is a major but first step. Human salivary gland stem cells that were transplanted into mice with a suppressed immune system also did their work. However, cell therapy is not freely allowed, and the rules are strict. Coppes expects it to take at least another eighteen months before the first patient can be treated with this method. “The growth medium has to be approved first, and you have to be absolutely certain that you do not inad-

vertently inject tumour cells into the patient. It works in principle, but there are still some catches.” One bottleneck, for instance, is the medium used for the cultivation of the cells. The current medium is appropriate for animal cells, but has not been approved for human cells yet; finding an approved replacement has turned out to be difficult.

Apart from cultivating the stem cells, Coppes also hopes to find a way to grow the entire gland. He is looking for the correct growth factors and environ-ment which allows salivary gland cells to grow into an organ. He has succeeded in cultivating small, two-millimetre salivary glands; in the future, he would like to try to grow larger organs using each individual patient’s own cells, which could then be placed back into their body in their entirety and might even be able to function straight away.

“ Whenever I eat a sandwich, I have to take a sip with each bite.”

Never eating toast again

NIRM

139138

2010

To create a synthetic heart, we first had to be able to produce all the cell types of the heart in the lab. This sounds easy, but it was an enormous challenge. Sixty per cent of a foetal heart is made up of beating heart cells, but in adults this is only thirty per cent. More-over, beating heart cells come in different types. The cells of the atriums, for instance, look slightly different from the cells of the ventricles, not to mention the blood vessel cells that supply the heart with oxygen, connective tissue cells that ensure its solidity and pacemaker cells that regulate its rhythm.

Autumn 2009

We started afresh with a new goal: to make a synthetic piece of heart. Of course, we did not do this just for the sake of it: the synthetic heart could be used as a test model for the development of drugs. We are not only able to make heart cells from healthy people, but also from patients with heart disorders. We simply isolate kidney cells from their urine and treat them in the lab to make them turn into stem cells. And who knows, we might be able to use these synthetic heart pieces in the future to repair damaged hearts.

2011

Together with engineers, we went looking for the perfect matrix for our stem cells. In the body, heart cells grow on a soft substrate. This is a good environment for the cells, so we needed to reproduce it in the lab. We let the cells grow on synthetic polymer, which is somewhat similar to silicon. This polymer is placed on a chip with spiral-shaped electrodes inside and a vacuum underneath. The spiral-shaped electrodes form a kind of stretch system that makes the cells come into action. The heart rate can be adapted by changing the vacuum. In this way we can test the diseased heart cells at rest, but also look at what happens if a patient does sports, for instance.

I M A G E : B E R E N D V A N D E R M E E R ( L U M C )

IMA

GE

: B

ER

EN

D V

AN

DE

R M

EE

R (

LU

MC

)

IMA

GE

: R

EN

É D

EN

EN

GE

LS

MA

N

story: Elles Lalieu

Process report

2003

The project started with a simple idea: take stem cells, turn them into healthy heart cells, and inject these directly into the heart mus-cle. The objective: to prevent heart failure in patients who just had a heart attack, for instance, and whose heart muscle is par-tially defective.

2004

Our first idea proved to be diffi-cult to put into practice. There was no suitable lab animal at hand. The mouse is not a good model for cardiac disorders. The heart of a mouse beats over five hundred times per minute, a human heart only sixty times. Moreover, the heart rate of a mouse does not rise in stressful situations, but a human’s does. The heart of a pig would have been a better model, but it is rather difficult to suppress their immune system; the stem cells would quickly be rejected.

2007 Apart from the problem with the lab animal, we now also had a practical problem. The heart is one great lump of muscle – how do we get healthy heart cells into the right places? Stem cells do not travel and then settle exactly where you want them to. I always compare it to a piece of chewing gum. If you inject red dye into it, only a small part turns red, not the entire piece. It works similarly in hearts. Injecting separate cells into the heart muscle was not a success. A setback, but that is also part of doing research. In any case, the experiments taught us quite a bit about the human heart.

To repair damaged organs: that is the great challenge in regenerative medicine. It was also the challenge that Professor Christine Mummery faced when she started to conduct research on the heart. Repairing the heart using stem cells proved more difficult than expected – so Mummery took a different tack.

IMA

GE

: R

EN

É D

EN

EN

GE

LS

MA

N

IMA

GE

: G

UID

O G

ER

DIN

G

The heart still has some secrets to reveal

NIRM

141140

20??

At the moment we only have a test model, but I do think there will be applications for patients in the future. We could use a piece of synthetic heart as a patch or plas-ter on a damaged heart. Of course, this will bring its own set of prob-lems. A heart attack often leaves damage throughout the heart, whereas a patch would only be placed on the outside. Would that work, or should we attach the patch somewhere else? What about attaching it just below the outer layer of the heart? That would be comparable to putting something just underneath the peel of an orange, without peeling the rest of the orange. Repairing a heart is not easy, but that does not mean that you should not try.

IMA

GE

: D

EN

KE

R

Autumn 2015

Our model was successful. We also received a number of inter-esting follow-up assignments. Even the pharmaceutics devel-oper GSK showed an interest. However, the system is still not good enough. In the development of drugs, it is important to be able to examine a great many condi-tions in one comprehensive test. This is done using plates with a lot of compartments (wells). As it is, the chip does not fit into these wells. This means that we are only able to examine one condition at a time. Downsizing the chips so that they fit is a matter of engi-neering. To get there, we are now collaborating with Delft Univer-sity of Technology.

IMA

GE

: JE

FF

RE

Y M

. V

INO

CU

R

Summer 2015

Time for a celebration: after more than five years, we managed to make eight different types of heart cells and cultivate them in the lab. We knew how heart cells develop in an embryo, but figur-ing out which signals are involved in this process was quite a job. Zebra fish and mouse embryos came in useful here: in these ani-mals you can switch off one “sig-nal” at a time and see the effects on the development of the heart. If you switch one off and see that pacemaker cells are no longer made, for instance, you can add this switched-off signal to the stem cells in the hope that this will stimulate them to actually make pacemaker cells.

I M A G E : F O N S V E R B E E K

The heart still has some secrets to reveal

Mummery’s approach is successful as a test model, but it cannot be directly applied to patients yet. Cell biologist Marie José Goumans adopted another tactic: she isolates stem cells from the heart itself and subse-quently tries to shape them into functional heart cells. Since the cells originate from the organ to which they will eventually be returned, this method is closer to patients.

“We retrieve the cells from the right atrium of the heart”, Goumans explains. “There is an area there that we refer to as the ‘heart’s appendix’. We do not really know what its function is, but it is always removed in open heart surgery. Instead of discarding this bit of ‘waste’, the surgeons now bring it to our lab.”

The removed piece of cardiac tissue contains stem cells, albeit only a few. “We isolate the existing stem cells, cultivate them until we have many more, and then use growth factors to try and grow them into heart muscle or blood vessel cells”, Goumans explains. “Together with Carlijn Bouten’s research group (see also Design your own heart valve, ed.) we have also looked into their willingness to differentiate if we stretch them a little, like they do inside a beating heart.” However, stretching alone has proved not to be enough for stem cells to develop into heart cells.

The cultivated heart cells were implanted in mice that recently had a heart attack, but this brought another problem. Human heart cells do not like to attach to a mouse heart, as this forces them to beat over five hun-dred times per minute, causing them to blow them-selves up. Without good lab animal results, however, it is not possible to test the cells in patients.

According to Goumans, the solution lies in combining her cells and Mummery’s test system. “We could make a synthetic piece of heart and use this to grow the cultivated stem cells.”

X-ray of an appendix (the worm-shaped pouch below the villi). There is a similar piece of tissue

IMA

GE

: H

EL

LE

RH

OF

F

Cultivating cells from the heart’s “appendix”

NIRMNIRM Interview

143142

?

may even grow back again. A sick liver has lost this ability but would, in theory, only need a few trans-planted healthy stem cells to regain its function. This makes the liver – unlike, for instance, the heart – very suitable for this type of research.

In January 2015, Clevers published a so-called proof-of-concept study, based on precisely this idea, in the scientific journal Cell. The researchers extracted stem cells from the hepatic ducts of a donor liver, and started cultivating them. “You will first have to make billions of cells with the help of the correct growth factors”, Clevers explains. “If you have many stem cells and precursors, they can specialise into the two most important parts of the liver: hepato-cytes (liver cells) and hepatic duct cells.” The liver organoids made from those cells consist of tens of thousands of cells and are only two to three millime-tres in size, but they do have a real 3D structure.

The researchers placed these mini livers back into mice with liver damage and a suppressed immune system. The mice clearly recuperated: the healthy cells replaced parts of their damaged liver.

“Although the liver damage in mice would have been repaired in the long run anyway, we did see that the mice treated this way recuperated more quickly”, Clevers says.

In this stem cell therapy, the sick liver acts like a scaffold, as regenerative specialists would call it. The organoids need the infrastructure of the damaged liver to attach themselves. This is relatively simple in livers: “You can inject the organoids in the portal vein, which transports blood directly into the liver. The portal vein branches off into very tiny capillar-ies, where the organoids simply get stuck and easily colonise the liver.” The portal vein in mice proved to be too small, so they received the mini livers in their spleen. Even from there, the organoids could find their way into the liver and nest there.

It is a big step from lab animal research to humans in this type of research. However, clinical experiments with mature liver cells are already taking place, in which billions of separate cells are administered through the portal vein. This would be far more effective with stem cells. But there are some catches.

Repairing liver damage with the patient’s own cells

Hans Clevers: “It would be ideal if we could already start treating patients with organoids while they’re on the waiting list for a transplant.”

I M A G E : S A N D E R H E E Z E N

Using stem cells to repair body parts: the idea is pretty obvious, since stem cells are also responsible for small repairs in organs in healthy people. In fact, it is the damage that triggers stem cells into action. Stem cell research has been going on for several decades, Clevers explains. “The most important pro-teins that play a role in embryonic development – the WNT proteins that pass on signals between cells – were discovered in the 1980s. We found out that they are crucial in the continued activity of mature stem cells. If they are overactive, this will lead to cancer.”

Clevers started his stem cell research in the bowel, which has the most active stem cells of all organs, renewing its cells every four days. In 2009, Clevers’ group succeeded in cultivating stem cells that had only been discovered two years before. Last year, the researchers at Clevers’ lab and their Japanese col-leagues proved that it is possible to reorganise these cells into mini bowels and return those so-called organoids into the damaged bowels of mice – thereby taking another few steps towards the ulti-mate goal of stem cell therapy. Clevers: “If you can use stem cells to repair a sick or damaged organ, the

shortage of donors would be far less of a problem. The patient’s own body would serve as a ‘donor’ to help repair its own tissue.”

The treatment method is not exclusively for the bowel, although its great regenerative capability makes it ideal. Clevers and his colleagues are also working on the use of the same method for the liver. Millions of people worldwide suffer from chronic liver failure, caused by a genetic defect or by viruses, alcohol or other toxic substances. A liver transplant can sometimes be helpful, but there are far too few donor organs available. With the method that Clev-ers’ group has developed within the NIRM consor-tium, a simple injection would suffice to repair the entire liver.

The liver is special, as it is usually able to repair itself quite effectively. If you take away part of the liver, it

When a damaged liver is no longer able to repair itself, a transplant is the only option.

However, to the tens of millions of people with liver damage in this world, there are far

too few donors available. Hans Clevers, professor of Molecular Genetics at the Hubrecht

Institute of the KNAW Royal Netherlands Academy of Arts and Sciences, is developing a

technique to grow mini livers from stem cells that can make the organ as good as new.

“ It would be ideal if we could already start treating patients while they’re on the waiting list.”

story: Rineke Voogt

Case study NIRM Case studyNIRM

144 145

Cell biologist Gerald de Haan at the University Medi-cal Center Groningen found a solution. He equipped cells with a “barcode” so that they can always be identified. “We take stem cells from the bone marrow of mice, to which we introduce a small piece of new DNA”, he explains. “We then transplant the stem cells back into the bone marrow. If one of these stem cells

starts to divide, all of its offspring share the same code in their DNA. After transplantation we extract some blood, isolate the DNA and then determine how many times the barcode appears.”

The first results of barcoding, as the researchers call it, are promising. In mice (which live to an average age of eighteen months to two years), one or two stem cells are responsible for 95 per cent of all the blood cells formed during their lives. The same experiments are now being repeated with human stem cells from the umbilical cord.

Fewer lab animalsBarcoding not only provides insight into the forma-tion of blood cells, but it also teaches us more about blood cancer – leukaemia. “Not all leukaemia cells are identical”, De Haan explains, “which means that there is not just one single type of cancer in the body, but a number of subtypes. Because we are now able to recognise the stem cells, we can count how many different stem cells are involved in a disease like leukaemia and whether they respond differently to treatment.”

An added advantage of barcoding is that research now requires far fewer lab animals. Say, you have one hundred stem cells and you want to know what each of them does. De Haan: “Normally you would transplant those one hundred stem cells into one hundred different mice. Now you can equip the stem cells with a barcode and transplant them into a single mouse. This makes the study both cheaper and more efficient.”

A piece of DNA (green) is placed in a viral vector – a kind of transport vehicle that can deliver DNA into the cells. Once it is inside the cells, the newly added DNA integrates into the existing DNA, which leaves all the cells with their own unique barcode (line patterns on the right).

Bone marrow contains stem cells that can make new blood cells. Once these new blood cells have entered the bloodstream, they are all identical and you can no longer recognise which blood cell originated from which stem cell. That is why we do not exactly know how many stem cells are involved in the formation of blood cells, nor if their number changes when we age.

Barcoded blood cells

story Elles Lalieu image Evgenia Verovskaya

RetroviralVector

Technically, according to Clevers, researchers are making good progress. They have managed to upscale the cell cultivation in order to grow billions of cells from just a few stem cells. Legislation is the main problem. The biggest concern is the risk of cancer in the tissue due to mutations in the organ-oids. This risk seems to be nil, says Clevers: “We don’t see that happening in cell cultivation. If you maintain the optimal environment for the cells – add the correct substances and ensure that the cells can grow properly – you will avoid cancer. The cells are genetically very stable.”

“It would be ideal if we could already start treating patients with liver organoids while they’re on the waiting list for a transplant”, Clever believes. “But in order to do so, we still have to clear some hurdles.” The researchers first want to know for certain, for instance, how many organoids they

will have to administer for an optimal effect and whether these should be stem cells or mature cells.

Clevers is confident that the principle also works in humans. “There have been patients with a liver dis-ease whose genetic defect was corrected in a single cell, purely by chance. Little by little, the liver repaired itself from that one cell, and the patients regained their health. If a single healthy cell can be enough to heal someone, treatment with mini livers opens up a new perspective for millions of patients.”

A microscope image of a liver organoid.

“ It would be like using the patient’s body as a donor.”

Repairing liver damage with the patient ’s own cells

NIRMNIRM

147146

story: Rineke Voogt image: René den Engelsman

Regenerative medicine is not being used in practice to its full potential. Why is that?Clevers: “It is difficult to get approval for treatments like stem cell therapy to repair a defect in the body. If we succeed in repairing the liver of a mouse with a mini liver, this is not a carte blanche to try out the same approach in humans in a clinical experiment. In a way this is understandable. After all, we are working with living cells which, potentially, may behave badly – you have to make certain that the cell therapy that you want to use does not cause cancer in the patient.”

So you have to provide evidence that the treatment is safe. How do you do that for a new therapy?Clevers: “This is precisely the problem. When is it safe to apply a new technique? The cells that we grow, to make liver organoids for example, are very stable and do not mutate. Cancer emerges from mutations. This is why the cultivated cells are, in all probability, quite safe. However, even with the large numbers of patients with a defective liver on the waiting list, a transplant with liver organoids cannot take place as yet. You will first have to pro-vide numerous examples to show that it is safe, but it is in fact a catch-22: you will need clinical trials to provide such examples.”

Why is it so difficult to test a treatment in a clinical trial?Clevers: “There is great fear of repeating the thalido-mide scandal (when a great many pregnant women were administered this drug against morning sick-ness and it was found to cause serious deformities in their foetuses, ed.). Moreover, the pharmaceutical industry is not very keen on cell therapy. It is easier to produce and sell a treatment wrapped up in a pill; living material is logistically complicated. This resis-tance is an additional challenge in our work.”

Bouten: “As regenerative medicine is still a rather new field, many rules have not been fixed as yet. For example, for how many lab animals should you provide evidence that your treatment works before you are allowed to start a clinical trial? The therapy is new, so there are no clear guidelines for each case.

As a researcher, therefore, you partly make your own rules. For instance, in collaboration with doctors we have decided that the heart valve for our study (see “Design your own heart valve”, ed.) should work in a lab animal for at least two years before we can take the next step to humans. It is possible to accelerate the same processes in the lab, but then the question is whether all the safety aspects are covered and whether the regulators would accept this.”

How do we get regenerative medicine from the lab into patients?Bouten: “The process of successfully developing a new treatment from idea to patient usually takes approximately fifteen years. This might be even longer for regenerative medicine, which is quite difficult to fit in with the daily clinical practice. It is a great step from the lab to the market, as many treatments are simply quite expensive. And once a new method has finally reached the clinical phase, the question is which patient group should be tack-led first. Doctors tend to prefer the group that runs the least risk, but we have noticed that patients take a completely different view. That is an ethical issue that needs further debate.”

What should change in order to make regenerative medicine more common?Bouten: “To find a solution, we should look to the new generation. Not everyone wants to go along with a new treatment, even if it is already techni-cally possible. We should introduce young doctors to this new type of medicine and technological progress at a very early stage of their training. This will encourage them to start exploring the options and adopt a more informed position. If you want to apply regenerative medicine, you will need the sup-port of the practitioner.”

Clevers: “The main challenge is to take away the fear of a new treatment being administered without hav-ing been properly thought out and researched – the idea that it is bound to go wrong. This is precisely the reason why we should also look beyond the medical world: we need to keep society well-informed and enter into public debate about the issue.”

Discussion

Are the regulations ready for regenerative medicine?

Many of the projects within NIRM were “proof-of-concept” studies:

studies to find out if a proposed concept actually works. Technical

glitches have mostly been overcome, but the application in clinical

practice is a great step further. Are the regulations ready for it?

What must be done in order to make regenerative medicine available

to patients? Hans Clevers (Hubrecht Institute) and Carlijn Bouten

(Eindhoven University of Technology) explain.

149

Hier het Health Holland-logo bij: logo HealthHolland.epsLife Sciences & Health (LSH) is one of nine sectors of excellence (“top sectors”) in the Netherlands. The sec-tors of excellence were designated by the Dutch Minis-try of Economic Affairs and were selected on their ability to contribute substantially to global societal challenges. The Life Sciences & Health executive office is housed at Health~Holland.Life Sciences & Health entails a broad scope of discipli-nes, from pharmaceuticals to medical technology, and from health care infrastructure to vaccination. In order to realise its mission – vital citizens in a healthy eco-nomy – this sector of excellence builds on the strengths of the former Dutch Life Sciences & Health sector to address the biggest societal challenges in prevention, cure and care: improving the quality of life (vitality) while restraining the costs of health care.On December 18th, 2009 the LSH-FES research propo-sal was awarded a FES contribution of � 81 million by the Dutch government matching an investment of � 95.4 million by companies, universities and others. The pro-posal was submitted by six research consortia: tEPIS, Cyttron II, NeuroBasic PharmaPhenomics, Virgo, the Dutch Stem Cell Initiative (DCTI) and the Netherlands Institute of Regenerative Medicine (NIRM).

Contact: www.health-holland.com, 070-3495404

René Rector

+31 70

Contact: www.cyttron.org

151150

Partners

Academic Medical Center Amsterdam www.amc.nl

Delft University of Technology www.tudelft.nl

Erasmus University Medical Center Rotterdam www.erasmusmc.nl

FEI Electron optics bv www.fei.com

Leiden University www.leidenuniv.nl

Leiden University Medical Center www.lumc.nl

Maastricht University www.maastrichtuniversity.nl

Naturalis Biodiversity Center www.naturalis.nl

Nikon Instruments Europe bv www.nikoninstruments.eu

Pepscan Therapeutics bv www.pepscan.com

Philips Electronics Nederland bv www.philips.nl

Science and Technology Facilities Council www.stfc.ac.uk

Utrecht University www.uu.nl

Virtual Proteins bv (until December 2013)

Understanding how molecular alterations affect the healthy state of a human being is essential in order to develop new diagnostics and treatments. Yet answering this question requires an all embracing picture – from molecules to cells to organs to organisms. Current bioimaging tools can only reveal the important aspects on one single level. In Cyttron II, fourteen academic and industrial partners collaborate to overcome this problem. We develop tools and integrate the quantitative data of complementary bioimaging technologies to understand the whole picture. This way we contribute to new powerful diagnostic tools.

Conventional pathology involves specimens (microscopy slides), and either physical gathering of (clinical) researchers and pathologists is required to examine the slides under the microscope and discuss their findings, or the slides need to be mailed back and forth between locations. This is an inefficient and time consuming way to operate and it puts the fragile and sometimes irreplaceable pathology samples at risk during transport. The tEPIS project aims to move this process into the online world, by setting up an online platform for storage and exchange of digitalized pathology slides. Such an environment allows researchers to examine slides from their respective workplaces and to perform data analysis on slides from other studies.

Partners

Academic Medical Center Amsterdam www.amc.nl


Radboudumc Nijmegen www.radboudumc.nl

University Medical Center Groningen www.umcg.nl

University Medical Center Utrecht www.umcutrecht.nl

VU University Medical Center Amsterdam www.vumc.nl

Philips Electronics Nederland bv www.philips.nl

Contact: Jan-Willem Boiten, [email protected], +31 6-18639236 Nikolas Stathonikos, [email protected], +31 88-7556563

AIMMTherapeutics

PharmaPhenomics

153152

VIRGO aspires to gain a better understanding of the most common acute and chronic viral diseases, such as influenza, hepatitis, and AIDS. To this end, we study the interaction between the viruses causing these diseases and their hosts on both a molecular and a physiological level.

Partners

AIMM Therapeutics www.aimmtherapeutics.com

Academic Medical Centre Amsterdam www.amc.nl

Bristol-Myers Squibb www.bms.com


F. Hoffman-La Roche Ltd. www.roche.com

Gilead Sciences Nederland bv www.gilead.com

GlaxoSmithKline www.gsk.com

Leiden University www.leidenuniv.nl

Leiden University Medical Centre www.lumc.nl

Netherlands Proteomics Centre www.netherlandsproteomicscentre.nl

Novavax AB www.novavax.com

Psynova Neurotech psynova.com

Radboudumc Nijmegen www.radboudumc.nl

Universiteit Utrecht www.uu.nl


Partners


VU University Amsterdam www.vu.nl


Wageningen University www.wageningenur.nl

Netherlands Institute for Neuroscience www.herseninstituut.knaw.nl

Delta Phenomics bv

Griffin Discoveries bv www.griffindiscoveries.com

Neurasmus bv www.neurasmus.com

Noldus Information Technology bv www.noldus.com

PsyNova Ltd. www.psynova.com

Synaptologics bv www.sylics.com

NeuroBasic PharmaPhenomics is a Dutch public-private consortium that is working on the next generation of drugs for brain disorders. Central to their approach is the eventual large-scale and rapid design and testing of potentially effective compounds that can serve as a basis for medication against brain disorders. The increased understanding of complex signal transduction routes has led to the identification of a great number of candidate genes for major brain disorders as well as the development of specific mouse models. These new models are pivotal in making pre-clinical research into neurological and psychiatric conditions more efficient and ultimately in developing new treatments for human patients.

Contact: Arjen Brussaard, [email protected], +31 20-5987003

155154

Partners

Dutch Diabetes Research Foundation www.diabetesfonds.nl

Galapagos www.glpg.com


Polyganics www.polyganics.nl


Xpand Biotechnology bv www.xpand-biotech.com

University of Twente www.utwente.nl

DCTI is a research initiative to improve beta-cell replacement therapy and to make this therapy accessible to patients with failing beta cells. These patients have developed diabetes because beta cells in the islets of Langerhans no longer produce insulin. The main research goals are to increase the availability of high quality islets for transplantation, identify factors for long-term islet function and survival, and to exploit alternative transplantation sites with the use of biomaterials.

Contact: Eelco de Koning, [email protected], +31 71-5263964 Aart van Apeldoorn, [email protected], +31 53-4892153Contact: www.virgo.nl, [email protected], +31 10-7044770

National Institute for Public Health and the Environment www.rivm.nl


ViroClinics Biosciences www.viroclinics.eu

Vironovative www.vironovative.com

Thermo Fisher / Dionex www.dionex.com

PMS 130 / PMS 369 / black

full color on a light background

full color on a dark background

in black & white

157156

Life Tec Group lifetecgroup.com

Landsteiner Foundation for Blood Transfusion Research www.lsbr.nl

Lead Pharma Holding bv www.leadpharma.com

Materiomics bv www.materiomics.com

Netherlands Kidney Foundation www.nierstichting.nl

Percuros bv www.percuros.com

Pluriomics bv www.pluriomics.com

PolyVation bv www.polyvation.com

Xeltis www.xeltis.com

Rijk Zwaan Zaadteelt en Zaadhandel B.V. www.rijkzwaan.com

ServiceXS bv www.servicexs.com

KWF Dutch Cancer Foundation www.kwf.nl

SupraPolix bv www.suprapolix.com

Synvolux therapeutics bv www.synvolux.com

Dutch Arthritis Foundation www.reumafonds.nl

Xpand Biotechnology bv www.xpand-biotech.com


KNAW Royal Netherlands Academy of Sciences (Hubrecht Institute) www.hubrecht.eu

When something breaks down, you can try to repair or replace it – likewise in our body. But how do you actually repair a defective heart valve, an injured knee or severely burned skin? And, even more importantly, how do you repair it for the long term and without any side effects? These challenging questions are the focus of the Netherlands Institute of Regenerative Medicine (NIRM). Its research is divided into five clusters: heart and blood vessels, muscles and bones, blood, nerve system and internal medicine.

Partners

Absea Biotechnology Ltd www.absea-antibody.com

ACTA Amsterdam Academic Center for Dentistry www.acta.nl

Araim Pharmaceuticals, Inc. www.araimpharma.com

Arcarios bv www.arcarios.com

ASkin Nederland www.a-skin.nl

Beckman Coulter Nederland bv www.beckmancoulter.com

Becton Dickinson Biosciences www.bd.com

Bender Analytical Holding bv www.gatt-tech.com

BiOrion Technologies bv www.biorion.com

CellCoTec bv www.cellcotec.com

Encapson bv www.encapson.com

Feyecon Development & Implementation bv www.feyecon.com

FlexGen bv piculet-biosciences.com

Harbour Antibodies bv harbourantibodies.com

Hubrecht Organoid Technology hub4organoids.eu

158

ColophonEditors

Joost van der Gevel, Elles Lalieu, Rineke Voogt

Chief editor

René Rector, Sciencestories.nl

Graphic design

Rick Verhoog and Sara Kolster, Parkers

Infographics

Parkers, Marjolein Fennis and Sara Kolster

Project leader

Giovanni Stijnen, Science Center NEMO

Coordination

Giovanni Stijnen and Sanne Deurloo, hoofdredactie Kennislink

ISBN 978-90-807981-5-1

This publication was realised with the financial support of the LSH-FES subsidy programme and in collaboration with Science Center Nemo (publishers of Kennislink.nl) and Jan-Willem Boiten (tEPIS), Christa Recourt (Cyttron II), Annemieke Steenbergen (NeuroBasic PharmaPhomics, Paul Huber (Virgo), Margot Beukers (DCTI), and Marja Miedema and Miriam Boersema (NIRM)

For current information about the topic in this publication,

please visit: www.kennislink.nl


Radboud University Medical Center Nijmegen www.radboudumc.nl

University of Groningen www.rug.nl

Netherlands Cancer Institute www.nki.nl

Eindhoven University of Technology www.tue.nl



University of Twente www.utwente.nl

Utrecht University www.uu.nl

VU University Medical Center Amsterdam www.vumc.nl

Wageningen University www.wageningenur.nl

Contact: Ruud Bank, [email protected], +31 50-3618043 Rini de Crom, [email protected], +31 10-7043063

Documents

Diagnosis, medicines, recovery