Feature: The biggest poisoning in history
24 January 2013. By Penny Bailey
One of the more bitter ironies of human existence is the way the best of our intentions can fall foul of Murphy's Law and wind up as paving stones on the proverbial road to hell. A recent, devastating example was the mass poisoning of 150 million people in South Asia as a result of attempts to supply them with clean drinking water.
For centuries, people living in the region had been drawing their water from the rivers they lived beside, all of which have their beginnings in the Himalayan peaks (also known as 'the water towers of South Asia'). Melted ice from the glaciers and snow lakes flows down the mountains in overground and underground streams, eventually forming the major river systems of Nepal, North India and Bangladesh.
The overground rivers collect vast amounts of environmental and human pollution on the way - including the bacteria that cause cholera and typhoid and other waterborne diseases, meaning that hundreds of thousands of people drinking the water were dying every year. Reasoning that the water in the underground streams and aquifers would be relatively uncontaminated, organisations (including the Bangladeshi government and UNICEF) drilled tube wells to access the underground mountain water. They were right on one count: there were far fewer bacteria in the water, and people stopped dying in such vast numbers.
Ten or 15 years later, however, they began turning up at their doctors' surgeries with some worrying symptoms: lesions on their hands and feet, or cancers of the bladder and lung at a surprisingly young age - classic symptoms of long-term arsenic poisoning. It turned out that many of the underground aquifers accessed by the tube wells had passed through layers of rock containing arsenic, which was leaching into the drinking water, leading to what the WHO has called "the largest mass poisoning in history".
It's not possible to boil arsenic out of drinking water, so the task now is to find out which wells are contaminated and which are safe. With more than ten million tube wells in South Asia, that's a gigantic undertaking. And there are no short cuts: arsenic levels in underground streams just feet apart can vary dramatically, so testing one well won't give any reliable information about nearby wells, even ones in the same field.
Every single well needs to be tested individually - and retested periodically, because the underwater aquifers can shift over a period of months or even weeks. Some may cease to pass through rock containing arsenic and clean themselves up (although they have been labelled as 'contaminated'), whereas others labelled 'safe' will become contaminated.
Current laboratory and field tests are expensive, can be unreliable or hazardous, and require highly trained operators. They simply aren't up to the task. As a result, many of the ten million tube wells in the region haven't been tested once, let alone at regular intervals.
Dr Jim Ajioka at the University of Cambridge and Dr Chris French at Edinburgh University and colleagues have taken up the gauntlet: with a Wellcome Trust Translation Award, they are developing an arsenic testing device that will be more fit for purpose. They aim to pilot it in Nepal's Terai, the humid savannahs at the foothills of the Himalayas, home to rural and farming communities using around 400 000 tube wells. If the device proves successful there, they will roll it out to the rest of South Asia.
On the ground
It's new territory for Ajioka and French - in every sense of the word. For Ajioka, whose expertise is in blue-skies scientific research, the wider logistical and human considerations inherent in creating a product with a practical use, in a very different country, were "a real eye-opener".
"It's one thing having a biosensor that works in the laboratory. It's another to have one that works in the field," says Dr David Nugent, who has come on board as business advisor. A physicist who runs a technology development and investment advisory business in Cambridge (Elucidare), Nugent is advising on the legal and financial complexities at different stages of the project, from patenting new technology, manufacturing the device on a vast scale and distributing it throughout South Asia to marketing and pricing the product so people will buy and use it.
The last point is crucial: to build something that people living in South Asia will actually use to test their drinking water requires understanding their priorities, their competing needs and what resources are available to them. "To do any good, for our product to be any use in Nepal, we had to understand what life is like there - what's happening on the ground," says Ajioka.
That understanding is provided by Professor David Grimshaw, Visiting Professor of Information and Communication Technologies for Development at Royal Holloway, who has joined the team as the development expert. He spent several years working in Nepal as Head of the New Technologies International Programme with Practical Action and had talked to the affected communities about the arsenic problem.
First and foremost, he says, farmers in the Terai need safe drinking water, now, today, for their crops, goats and family. "If you need water, you can't wait till UNICEF comes round and tests the ground water - if they ever even come to your village," he says. "You need to be able to test the water now, yourself, to find out exactly where you should drill the well. Then you need to be able to test it again periodically to check whether it's still safe or whether you need to drill another one."
Affordability is the other pressing practical need. Around 55 per cent of people in Nepal live on less than 80 pence ($1.29) a day (£24, or $37.50, a month). With seed, fertiliser, tools and food to buy for their farms and families, people can only reasonably be expected to pay for a test for their drinking water if it doesn't take a huge chunk out of the few pence they have at their disposal from day to day. If it's to make a difference, the device will have to be priced in cents.
Arsine and mercury
Current laboratory and field tests don't meet either need. Laboratory tests use atomic absorption spectroscopy, which measures the absorption patterns of light by vaporised arsenic atoms. Water samples from the wells have to be sent to a centralised laboratory to be read by a spectrometer and the results sent back to the community - all of which takes up precious time. "If you're sending a sample from a well in the Terai to a lab in Kathmandu, the fastest you might do that is a day," says Grimshaw. "If it rains it's going to take you several days to get there, because there'll be landslides on the way."
Even if the results do make it back in a day or two, the laboratories' good intentions can once again come face-to-face with Murphy's ineluctable Law. "UNICEF, with the best will in the world, is testing a lot of wells, tens of thousands of them," says Grimshaw. "And sometimes they confuse the results of one well with another - so a contaminated well is given a tick and a clean well marked as unsafe. It's human error; it's bound to happen with that kind of system."
Field tests can give an answer on the spot but can only be carried out by trained operators. Most are based on the Gutzeit method, which converts arsenic in a water sample into arsine gas. The gas reacts with an indicator paper impregnated with mercuric bromide to generate a coloured spot, the intensity of which indicates the concentration of arsenic in the water.
Both arsine gas and the mercury salts in the kit are extremely toxic, posing a hazard to the operator and making disposal a problem. "You can't flush the chemicals down a toilet or pour them into the water stream after you've done the test - you'd be defeating the purpose and poisoning people with something else," says Nugent. "You have to send the chemicals back to the laboratory for safe disposal." Both methods mean villagers in the Terai have to wait for the experts to test their wells (if they ever do), something that's not an option when what people are waiting for is drinking water.
The Trust-funded team are turning to biology rather than chemistry to develop a test that is more fit for purpose, circumventing the hazards of toxic chemicals and arsine gas, as well as the need for specialist equipment and operators. Their approach turns the original irony - of drilling tube wells to access 'bacteria-free' water - around by using bacteria to report the presence of arsenic in those wells.
Some bacteria can sense even minute amounts of arsenic in water and respond by producing enzymes, an efflux pump and other mechanisms to break down and expel the toxin. Fusing a gene that makes fluorescent or coloured proteins with one coding one of these detoxification proteins causes the bacteria to 'light up' when it senses arsenic. It's a subtle effect that can only be detected with expensive laboratory equipment like a luminometer or a fluorometer, making the system unsuitable for field use.
In 2006 a team of young, enthusiastic synthetic biologists from the University of Edinburgh built an arsenic biosensor with a different reporting method. The device was their entry for the 2006 International Genetically Engineered Machine (iGEM) competition, an annual competition in synthetic biology organised by researchers at the Massachusetts Institute of Technology.
"We were trying to think of ways of making an arsenic biosensor that wouldn't need a special laboratory instrument and a technician to read the results," says French, who was an adviser to the team. After mulling over a few ideas, they eventually decided to build a hybrid device that was part biological and part electrical, which would produce a change in pH instead of fluorescing in the presence of arsenic.
They selected Escherichia coli, which naturally ferments sugars into acids, as the sensing bacterium and fused the LacZ alpha gene (responsible for fermenting lactose into acid) to an arsenic-responsive gene. Both genes switched on simultaneously in the presence of arsenic, causing the bacteria to produce more acids and lowering the pH in the water.
The pH change was easily detected by a small glass pH electrode, which converts the pH into a measurable voltage displayed on an electronic meter. Using a pH indicator solution (methyl red, which turns red in acidic solutions, or bromothymol blue, which turns yellow) also gave a very bright and obvious response. "It worked so suspiciously well I was convinced they'd cooked the results till I'd repeated it a few times myself," recalls French. "I think it was the first iGEM project that actually had a practical application. They made a special prize for it, called 'Best Real World Application', which we won."
In the field, however, several other factors - such as the presence of other lactose-fermenting bacteria or buffer ions (which minimise changes in acidity) and changes in temperature - can influence the pH in ground water, possibly confounding the results. So for their arsenic biosensor, Ajioka and French are drawing on another iGEM project by students from the University of Cambridge.
This iGEM team genetically modified E. coli to generate a range of coloured pigments by optimising the expression of genes such as the violacein operon, which gives Chromobacterium violaceum its eponymous violet hue. "I was astounded that they were able to make an E. coli strain that either turned dark green or a deep purple by reconfiguring genes from another bacteria," says Ajioka, who was the adviser on the project. The production of a range of genetic devices generating a rainbow of colours, including more conventional pigments such as melanin and carotene, won them the iGEM 2009 Grand Prize.
Ajioka and French are now working on linking the violacein operon to the arsenic detoxification system - in a different bacterium - so that it will turn purple or green depending on whether arsenic is present or not.
E. coli was the natural choice for both iGEM projects because people have been working with it since the 1930s, it's extremely well characterised and there are hundreds of thousands of strains available. But for their current purpose, it has a couple of drawbacks: the main one is the fact that it doesn't produce spores or any kind of stable resting state. This would make it hard to keep the sensing E. coli cells alive and viable during transit, particularly en route to remote villages off the beaten track. Freeze-drying would make the cells much more stable but requires expensive electrical equipment.
French and Ajioka are instead using a less well-characterised bacterium called Bacillus subtilis, which forms spores ('sporulates') when it detects that conditions are becoming unfavourable for growth. The spores are extremely resistant to heat, drying and other stresses and can survive for decades - even centuries - remaining perfectly viable. Once a spore detects that conditions are favourable again, it germinates rapidly, regenerating an active, growing cell in around 30 minutes.
The process can be uneven and slow because the spore wall has to be broken down before the bacteria can germinate, but French has found a handy low-tech solution that kills two birds with one stone: "You can speed the process up by boiling the spores for two minutes. That's enough to break down the spore wall without killing the actual bacteria." It also kills off other bacteria in the water sample that might interfere with the test.
"The whole bed is going to be a plastic device with wells containing the spores," says Ajioka. "You add your water sample and boil it for a couple of minutes, and the spores will germinate. If the water's safe, the bacteria will turn green. If its unsafe, they will turn purple. We can make the bacteria produce different degrees of colour at different arsenic concentrations so you can tell instantly what level of contamination you have."
The other advantage B. subtilis has over E. coli is safety. It's an innocuous and ubiquitous bacteria that is used in probiotic yoghurts and laundry detergents, and it competes poorly with other bacteria, so the biosensor can be thrown away after use without any risk to the environment. Just to be on the safe side - and with another ironic twist, in view of widespread concerns about the safety of genetically modified organisms - the team will be giving B. subtilis another genetic tweak to ensure it can't survive or function out in the real world.
The biosensor looks set to tick all the boxes for ease of use, safety and reliability, and to meet the affordability criteria the team aim to keep the price at around 31 pence (50 cents) apiece - less than half the estimated daily wage of a farmer living in the Terai.
But however good the fit between product and price, need and income, if the consumer ultimately doesn't 'buy' it - in every sense of the word - all the work and investment has been in vain. It's a depressingly familiar story. "When we talked to other NGOs who've tried to roll out other healthcare products in low-income countries, we started to realise that a lot of the time the local populations don't want to use them," says Ajioka. "You can put a product out there, but if no one uses it it's useless."
If it's safe, easy to use and affordable, why wouldn't a farmer in the Terai buy an arsenic test every now and then to safeguard his health and that of his family and livestock? The reasons are manifold. For a start, 'affordable' is a relative term, and 50 cents is still a lot of money if it's over a third of your daily income and you need to buy food and clothes for your family now.
Then the health effects of arsenic poisoning don't become apparent for many years, and because not all wells are contaminated there's no guarantee you will get poisoned if you don't test your well. Confronted with the need for drinking water now, and the greater, more immediate threat of bacterial disease and death from river water, it's understandable why people might take a long-term gamble and use a well without testing it first.
There are also wider concerns about the social impact of knowing the arsenic status of someone's well. "When the UN test the wells and find arsenic, they put red paint on the hand pump. That tells everybody that you've got arsenic in your well," says Grimshaw. "Suppose I'm a poor farmer and I've got three or five daughters I need to get married. If my well is advertising that it's got arsenic in it, no one's going to want my daughters because their long-term health prospects aren't looking good, so I'm going to get rid of the red paint. People are forever inventive within whatever boundaries and systems they live in."
Unfamiliarity with the product is another disincentive to buy it. "Often a new technology comes along, out of the laboratories, where scientists are making these things with the best will in the world, but in isolation. Then the whole thing fails when it gets out in the field because for local people it's like something has been dumped on them from outer space. Nobody's been involved, they don't know anything about the product, and there's been no ownership."
The key, says Grimshaw, is to involve the product's consumers from the very start of the design process, something he began with a workshop in Nepal in 2009. The meeting brought together representatives from the government, regulatory authorities, NGOs, Nepalese scientists and village communities to discuss how to apply new technologies to tackle the arsenic problem.
Opening up a dialogue from the outset can provide important feedback at an early enough stage to adapt the final product before it goes into large-scale production. "You can ask what are the good points? What are the bad points? What can be changed - and what should be changed?"
Knowing what to change requires an understanding of the reasons underlying particular preferences. "At the 2009 workshop, the people from the local communities were saying they would prefer a biosensor with a digital read-out rather than a colour range. A mobile phone app could convert a colour to a number, if it turns out to be an important issue, but we need to get to the bottom of why people have that preference. Is it because a proportion of them are colour blind? Or because high sunlight levels make it difficult to discriminate between colours? Or because people have more trust in something that looks more technical?"
In addition to providing feedback, the process engenders that crucial sense of ownership, which is so important if people with very little money to live on are going to be willing to buy a new product. "We all know that when we adopt a piece of technology to improve our lives - an iPhone, for example - we want to be able to personalise it, however slightly, to make it our own. There's lots of research and evidence on that."
The sense of ownership can be enhanced by engaging all the stakeholders in Nepal's arsenic problem, including local governments and businesses. "It's important to tap into the set-up people already know and trust, particularly local governance structures at the village and district level," says Grimshaw. If community leaders endorse the biosensor, people are more likely to feel comfortable buying and using it. Involving local Nepalese manufacturers and distributors (e.g. allowing those that package or transport the device to put their logos on it) will also help to give the device a sense of familiarity.
Ownership and familiarity are both fundamental to ensuring the project has a sustainable impact. "If you just employ an NGO to buy arsenic testing kits and go out and distribute them, it will tend to be a one-off project. Those kinds of initiatives aren't sustainable because they don't become embedded in people's daily lives," explains Grimshaw.
Ideally, the biosensor will become a familiar tool that people will use at regular intervals as a matter of course, either to test their existing well or when they want to drill a new well. If it does, its users are also more likely to contribute to another important aim of the Trust-funded project: to create a detailed map of every well in the entire region, showing exactly where the arsenic contamination is concentrated, where levels of arsenic in the water are increasing or decreasing and what factors are influencing those fluctuations.
"We've got to be thinking about using this on a massive scale - something that hasn't happened until now," says Nugent, pointing out that the tools to do so are already available. "The sensors on modern mobile phones are now better than those you'd find in some laboratory instruments. Some mobile phone manufacturers have adopted backside illumination, which can capture images in low light. So someone could take a picture of the biosensor and coordinate it with the GPS coordinates on the phone to capture the location of the site, along with the test result. Then we could load all the data from hundreds of thousands - potentially millions - of wells, across the entire South-east Asian basin, into a centralised database."
Crucial to that part of the project will be the involvement of the other key stakeholder group: local scientists. "They've already done quite a lot of work testing wells, so they know what testing technology works and what doesn't work," says Grimshaw. "And they've mapped the area; they know the geography. They know about any seasonality effects - whether flooding affects the levels of arsenic, for example." In turn, working alongside scientists from Cambridge will give Nepalese scientists the opportunity to develop their scientific knowledge and experience and ultimately assimilate it into the country's education system, further strengthening the long-term sustainability of the project.
Grimshaw has coined a phrase for technology developed by involving stakeholders from the start of the design process: FairTechnology, influenced by 'fair trade'. His vision of FairTechnology is one of "diverse stakeholders working together towards a world where science-led new technologies deliver products which fulfil human needs, rather than consumer wants." If the arsenic biosensor manages to do that, it could be a big step towards mitigating the colossal problem brought to South Asia by the best of intentions.
This article was amended to remove an inaccurate mention of the British Geological Survey in the third paragraph.
Top image: The Himalayas. Credit: Carole Reeves, Wellcome Images.