Friday, 27 January 2012

What's all the fuss about? Flu isn't so bad, right?
Seasonal flu is an annoyance for most people but, in the young or elderly, or immunocompromised individuals, it can still be fatal. Each year, there are around 4,000 deaths attributed to flu in the UK, despite these people having access to health care and anti-viral medicines.

So why is everyone so worried about bird flu?
The H5N1 bird flu strain has so far caused 578 confirmed infections and 340 deaths, but all of these individuals were in close contact with infected birds. The big fear is if the strain gains the ability to be transmitted between humans, which would give it the potential to cause a global pandemic. But we don't know how likely it is that this will happen, so it's very much a waiting game. Because none of us have encountered this strain before, we have virtually no resistance and the outbreak could be much worse than seasonal flu. It has been estimated that the worldwide death toll could be around 150 million and, because the strains are constantly evolving, we can't make a vaccine until the epidemic begins. The best we can do is stockpile anti-virals and put in place contingency plans for what to do if the worst case scenario should materialise.

What's this about scientists making a transmittable version of the bird flu virus?
Scientists want to understand how the virus might make the leap into being transmittable between humans and, to do this, they have created a version of the virus which can be passed between ferrets. Ferrets are a good model for flu in other mammals so there is a good chance that this virus would also be able to spread between humans (but this isn't known for sure). The two papers in question have not yet been published and there have been calls from the US government for the journals in question, Science and Nature, to censor specific details about how the research was done in case bad guys use it to reverse-engineer a bio-weapon.

So the scientists have played God and created a killer virus just because they can
Let's get this one out of the way first. It is terrifying to see just how many web pages come up if you Google 'H5N1 playing God'. Do people honestly believe scientists are power-crazed maniacs doing really stupid things for the sake of it? Because it simply isn't true. This bird flu research was designed to give us a better understanding of how a potentially deadly virus might emerge naturally and go on to kill millions. Similar research into flu has led to some awesome discoveries that better prepare us to deal with this disease. For more information, here is an article by a scientist behind one of the bird flu papers explaining why he believes this work to be so important.

These papers contain methods that can easily replicated by those that would like to use flu as a bio-terror weapon
A well accepted way to create strains that can be transmitted between animals is to repeatedly grow the virus until it accumulates the mutations that will allow it to be passed on by itself. This is effectively happening in nature right now and many believe it is a matter of time before the bird flu virus does it outside of the lab. We need to understand how this could happen in order to prepare for a situation in which it does. And failing to publish these papers won't stop terrorists from attempting this method if they really want to but it isn't a quick and simple task.

But they could use the mutations detailed in the paper to reverse-engineer a deadly strain
It's been reported that all of the mutations leading to these new strains are already published - that's because the same mutations have already been seen in the wild, albeit separately. A person with enough knowledge of the flu virus would most likely be able to make an educated guess as to how to go about engineering a similar strain but even an expert would find recreating the virus to be a mammoth task. Generally, if you've put all your effort into following the terrorist career path, you probably haven't had the time to become an expert on complicated viral genetic manipulation.

How can scientists even run the risk that someone will use this information for evil purposes?
Because the dissemination of information is a cornerstone of science. Censorship is a slippery slope which could lead us to a point where legitimate scientists cannot access information that could help their work and be of benefit to others. However, it is unlikely to provide a barrier to those intent on doing harm. If we are going to start restricting what information can be published among the scientific community, we need to know that the restrictions will work and that they are being imposed for the right reasons, rather than as a knee-jerk reaction to the slightest whiff of concern. The laboratories at the centre of this controversy have agreed to a 60 day halt in their work so these concerns can be discussed - this is an effort to show they are taking the situation extremely seriously but some have voiced worries that it is simply an empty public relations gesture. But any bad feeling between the scientific community and the public needs to be addressed - it helps no one if scientists adopt a 'we know better than you' attitude and refuse to justify themselves.

Is the work really of any benefit, though?
Predicting how scientific advances will help us in the future isn't always possible - some of the biggest discoveries in science have emerged when no one was expecting them to. Sometimes in science it is necessary to accumulate a critical level of unusable information before something comes of it and anything that improves our understanding of a disease such as bird flu can only be a good thing. So perhaps this bird flu work will lead to a better vaccine or drug treatment, perhaps it won't. Some people have suggested that it will allow us to better monitor the virus, allowing us to predict when it might start spreading between humans. In theory, this is a great idea but, unfortunately, it relies on quick and reliable reporting of infections and they are currently occurring in parts of the world not well known for their organisational skills when it comes to keeping track of viral infections.

It is inevitable that some of these strains will get released by accident and we'll all die
Safety measures in every category 3 laboratory are extremely tight, training is extensive, and every scientist I've met takes safety very seriously - none of us want to infect ourselves, after all! We all need to consider the risks versus the benefits. Yes, dangerous work does go on in scientific laboratories. No, it is not inevitable that there will be an accidental release of something deadly.

But the work isn't even happening in the highest biosafety level labs
In the country where this work was conducted, bird flu can be used in category 3 laboratories rather than the stricter category 4 facilities (used to study diseases such as ebola). This isn't an indication that the scientists in question are being stupid, it is an indication that they have considered all the possible risks and have gained permission from the relevant agencies for this work to go ahead. But I personally do find it slightly surprising that a cat 3 was used - I know in the UK, our bird flu work takes place in cat 4 labs.

And it has been reported that there are no armed guards protecting the virus!
This is true, universities and research institutes don't tend to station people with machine guns at every entrance. Yes, if armed terrorists stormed the labs, they could potentially steal these strains. But, come on guys, we can't start having armed guards stationed in all the places that everyone's currently panicking about, only to move them on when the next sensationalist newspaper article pops up. Preparing for real threats (such as the natural emergence of bird flu) is surely better than living in fear of something extremely unlikely happening?

If it gets out, this strain will kill 80% of those infected
It is surprising how much this statistic is reported. Yes, nearly 60% of the reported bird flu cases have resulted in death. But the problem here is the word 'reported'. Only those who are very, very sick go to hospital - when was the last time you reported mild cold symptoms to your doctor? So we don't know how many people have been exposed to bird flu but haven't become sick. The death rate is likely to be much lower than 60% - possibly less than 1%, in fact (see this PNAS article discussing the fatality rate, amongst other things).

There is no treatment or vaccine available
Current anti-virals are active against the H5N1 strain and there are vaccines that provide some protection against H5 strains. But it is true that, if there was a global pandemic, it would likely take 4-6 months for a vaccine against the strain to be formulated and distributed.

Nefarious governments are deliberately doing this work to create weapons
There isn't much arguing with conspiracy theorists who believe the world leaders are out to get them. But I can tell you that I've met few scientists who'd be willing to act as minion to some shadowy government organisation intent on killing off half the world's population. That probably isn't much comfort if you've already convinced yourself that the government and all scientists are inherently evil. In truth, very few of us are evil – if we were, we’d have chosen alternative career paths such as banking.

Image: Cybercobra at en.wikipedia

Saturday, 21 January 2012

Forgotten to defrost the chicken overnight in the fridge? That’s fine—you can leave it to thaw at room temperature, right? It will be quicker, after all…

But, within just a few hours, the tiny bacteria hiding in frozen food such as chicken, beef or that left over Chinese takeaway can start to divide. Fail to cook the meal enough and you can end up with Salmonella food poisoning—an infection that affects around around 10,000 people in the UK each year (reported infections, at least) and is responsible for more than 100 deaths.

Salmonella bacteria lurking within food have to run a gauntlet of challenges in their attempt to infect the epithelial cells of our intestines. First, there’s stomach acid to content with, then there’s the normal intestinal flora preventing the attachment and growth of non-commensal species, and our immune system will do what it can to destroy any invading bacteria. For this reason, a large dose of bacteria has to be ingested for the infection to take hold—that’s around 106 cells in a healthy individual. Because Salmonella fails to be killed by freezing and can grow rapidly at 25⁰C, a relatively small number of microbes present in frozen food can multiply to dangerous levels if left at room temperature for long enough.

I’m fairly sure that, were I to be kept in a fridge, it would take me a while to start doing something if returned to room temperature. Salmonella also goes through this same period of adaptation to their new conditions and a recent study from the institute of Food Research in Norwich published in the Journal of Bacteriology looked into what is involved in Salmonella beginning to divide after a period of non-growth. Understanding what is happening in the pre-dividing bacteria could help to find new ways to prevent the growth of harmful bacteria in our food.

Like all bacteria, the growth of Salmonella can be divided into five phases as shown in the image below. The first stage, the lag phase, is a period of adjustment in which the bacteria prepares for active growth—the cells do not divide but, behind the scenes, they are actively preparing for growth. The cells need time to not only repair any damage accumulated during their period of non-growth, but they need to remake and switch on all the cellular machinery that was packed away while it wasn’t required. This is the least understood phase of bacterial growth, however, as there is little data to describe the processes occurring within a lag phase cell.
Growth phases of cultured bacteria. The lag phase involves adaptation to the new environment. This is followed by exponential doubling of the bacteria until they run out of space or nutrients and enter stationary phase. The culture may under a period of death until a long-term stationary phase is reached in which a small number of bacteria can survive for long periods of time in a non-growing state.

Once everything is ready, the cells enter exponential growth in which they double at a rate anywhere in between every 14 days for a slow-growing bacterium such as Mycobacterium leprae, which causes leprosy in humans, to every 20 minutes for Salmonella species. At this rate, one Salmonella bacterium could theoretically divide enough times to form a colony the size and mass of the Earth in less than one day. Of course, this doesn’t occur because the culture runs out of space and nutrients. Once things start to get crowded or food runs low, the culture enters stationary phase in which they hang around doing not very much, waiting for conditions to improve.

The fourth stage is a decline or death stage but, importantly, many bacteria are extremely tough and it can take a very long time for them to die. This means that a growth curve contains a final phase—long-term stationary phase. This particular phase of growth is interesting for research into a number of pathogens as it is thought to best represent the state in which bacteria survive during a number of diseases, such as in the latent stage of tuberculosis.

Move stationary phase bacteria to more favourable conditions and they are able to resume growth after a period of lag, and this is what happens with Salmonella. The problem with understanding exactly how bacteria restarts growth is hampered by the fact that, by definition, few cells are present during lag phase. The Journal of Bacteriology paper described how 750 ml cultures of Salmonella were incubated at 25 ⁰C without shaking (usually used in the lab to increase oxygen availability in the culture) to represent the conditions experienced by Salmonella in room temperature food.

The researchers used RNA isolated from these bacteria to determine which genes are switched on during lag phase. They found that a number of genes are induced within the first four minutes of lag phase and, among them, were a number of genes required for metal uptake and utilisation. Perhaps these metals are required for active growth and represent a limiting factor in whether the cells can regrow or not. A better understanding of the processes occurring during lag phase will result in new ways to prevent growth of bacteria such as Salmonella, leading to new ways to prevent diseases such as food poisoning. For example, perhaps inhibitors could be added to certain foods that prevent the vital metal uptake pathways from functioning.

Graphical representation of the changes in gene expression during lag phase - red represents genes that are switched on and blue are down-regulated.

In addition, the paper addressed an important question—are stationary phase bacterial cells already primed to switch on vital genes as soon as they find themselves in conditions conducive to growth? The rapid induction of genes in Salmonella would suggest that this is the case. However, the researchers determined that this is not true. Instead, those genes switched on within the first four minutes have very strong binding sites for the cell’s transcription machinery that means they can be turned on extremely quickly.

Model showing the major processes occurring during lag phase, exponential phase, and stationary phase. Each symbol represents groups of functionally related proteins and is colored according to the level of expression of the appropriate genes under each growth condition. Blue shows that the relevant genes are expressed at low levels, yellow shows genes expressed at medium levels, and red shows genes expressed at high levels in each growth phase.

The authors proposed a model to explain the processes are occurring in the various phases of Salmonella growth, shown above. They conclude the paper by saying that ‘It seems fitting that as Salmonella was the first bacterium in which lag phase was studied, it is now the first Gram-negative bacterium to be understood at the level of global gene expression during lag phase.’

Sunday, 15 January 2012

A Salmonella infection has some very obvious and unpleasant effects on the unfortunate person who failed to wash or cook their food correctly. But it also messes with the bacteria that live in our guts, and that’s not a good thing when it comes to the emergence of new pathogens that can make us extremely sick.

Our guts are home to a huge community of bacteria. In fact, bacteria in our intestines outnumber our own cells ten to one, weighing in at around 2 kg in total. Without our intestinal microflora, our digestive and immune systems would not work very well at all. There's a huge industry centred around taking care of these little guys (and now is probably not the time to dwell on whether those probiotic drinks do us any good at all—I will undoubted come back to that another day). But one thing we can all agree on is that we need our intestinal microbes as much as they need us.

Normally, we co-exist peacefully with the approximately 1012 bacteria per gram that live in our intestines. It’s a thriving community with over a hundred different species, the complement of which is as unique between individuals as our fingerprints. The majority belong to oxygen-hating species that are fairly greedy when it comes to space and nutrients—so much so that they keep less common species including the Enteriobacteria under control.

Enteriobacteria such as E. coli keep to themselves most of the time, but they are capable of making us ill under the right conditions. Luckily for us, their low numbers in our guts stop them from getting together and carrying out a kind of bacterial arms race in which different species trade genes with each other and acquire new weapons such as antibiotic resistance. But this cold war turns hot when a bacterial infection such as Salmonella is thrown into the mix.

A recent paper in PNAS from a lab at ETH Zurich showed that mice infected with Salmonella undergo blooms of native E. coli in their guts due to inflammation caused by the infection. These blooms can result in up to 80% of the bacteria in the gut being E. coli—that puts them at high enough levels to undergo the bacterial equivalent of holding hands and then they start swapping DNA. 

In ten E. coli isolates investigated in this study, four were found to contain a plasmid derived from the infecting Salmonella. And when the researchers co-infected sterile mice lacking any gut bacteria with both E. coli and Salmonella, they found that nearly all of the E. coli acquired a plasmid from the Salmonella strain—that’s a level of transfer which no one could have expected.

It was the swapping of DNA between different strains that led to a huge outbreak of E. coli O104:H4 that infected 3,785 people and killed 45 in Germany just last year. In this case, a strain of E. coli acquired the genes required to make a toxin that turned it from being able to cause mild food poisoning to an infection that could kill.

But this new study highlighted another important point. Maintaining a plasmid uses up valuable resources that the cell could use for more important processes. So a population of bacteria in the lab will only maintain a plasmid if it confers some kind of advantage, such as resistance to an antibiotic in the culture. Remove the antibiotic, and the cells eventually ditch the plasmid.

In the gut things are slightly more complicated. In this study, the original Salmonella plasmid (P2) transferred to the E. coli contained a gene called cib. This makes a toxic protein which can kill other bacteria lacking the P2 plasmid. It’s a bacterial mechanism for ensuring their own survival at the expense of any other species that could use up all the nutrients and space. So the high levels of plasmid uptake by the E. coli was down to it needing this cib-containing plasmid in order to survive? Not exactly.

When the researchers did the same experiment with a Salmonella strain containing a modified version of P2 lacking cib, the levels of transfer were still huge. Part of the problem is that transfer between cells is so great that cells don’t get a chance to lose plasmids that give them no advantage. This is important when considering how bacteria acquire resistance to antibiotics. The results of this study suggest that an antibiotic resistance plasmid could be maintained in a population of bacteria even in the absence of the antibiotic.

When thinking about how different pathogens evolve, it is over-simplifying the situation to consider the species as individuals. The evolution of commensal bacteria in our guts are intertwined with the pathogenic species that occasionally infect us. So not only does a bout of Salmonella make us extremely ill, but it can also turn our harmless commensal species into dangerous pathogens.

Image courtesy of A. Canossa, M. Sommer and G. Dantas

Tuesday, 10 January 2012

What does an Egyptian mummy have in common with one-third of the world’s population? The answer is tuberculosis (TB)—a disease which has been affecting mankind since prehistoric times. But, I hear you muttering, didn’t we already cure TB? Um, not really. Around 1.7 million people die from TB every year and the HIV epidemic is only set to make things. So what went wrong?

The big problem with the bacterium responsible for TB, Mycobacterium tuberculosis, is that it is very difficult to kill. The antibiotics required to sterilise an infection are no fun to take, meaning plenty of people fail to finish the arduous nine month course of up to four side-effect-laden drugs. And this leads to the emergence of drug resistant strains—recently we’ve seen reports of totally drug resistant strains that cannot be treated except by removing big chunks of people’s lungs. Replacing those drugs lost through resistance is only delaying the inevitable. Which is, without a new and improved way to treat TB, we are at risk of returning to the pre-antibiotic dark ages. So understanding how one of mankind’s oldest foes continues to best us is a huge area of research.

A current area of interest is in the development of drugs and vaccines which work with the host’s immune system. But this is a disease which doesn’t induce any protective immunity of its own (unlike infections such as measles or chicken pox). This is an issue for TB researchers—if our own immune systems are too rubbish to know how to fight TB, how can we know what sort of immune response we need to combat the infection? One difficulty is that TB both suppresses and activates branches of the immune system for its own benefit.

So any old immune response won’t do—it has to be the right kind of response. An illustration of this can be seen by looking back at the first vaccine for TB, developed by Robert Koch in the 19th century. Koch’s vaccine caused a very strong immune response in the 2000 patients it was administered to. Good thing? Not really. Unfortunately, because these patients already had TB, the vaccine ended up making their infections much worse and most of them died as a result.

So how can we work out what kind of immune response is required to fight TB? A recent paper published in Proceedings of the National Academy of Science sought to investigate which host genetic factors have a role in the development of protective immunity against TB.

Outcomes of TB infection - adapted from Stefan H.E. Kaufmann, Nature Reviews Immunology 1, 20-30

One of the big unanswered questions in TB research is what governs susceptibility to TB. For all those people who develop active disease (productive cough, weight loss, fever, night sweats, and various other unpleasant symptoms), nine times as many fail to get TB. A small number of these people are able to completely clear the infection, whereas the majority develop a latent infection in which their immune system keeps the bacteria in check for many decades. Understanding the differences between these patients could allow us to better understand how we can nudge our immune systems in the right direction when it comes to fighting TB.

The authors of the PNAS paper investigated the response of a type of immune cell known as a dendritic cell to infection with M. tuberculosis. The method they used was to compare which genes are induced and repressed in dendritic cells when they are infected with M. tuberculosis. They then looked at differences in host genetics that were associated with specific gene inductions. These differences are likely to be those responsible for individual variations in the immune response to TB.

Genes up- and down-regulated in M. tuberculosis infected dendritic cells. Only those genes whose induction or repression was over-represented in TB-infected cells are shown. 

Genes found to have a role in susceptibility to TB included DUSP14, which makes a regulator of proinflammatory cytokines - signalling molecules involved in switching on the immune response. This gene is associated with levels of TNF-α and IFN-γ, two of the most important cyctokines involved in protection against TB. Other strong candidates for susceptibility genes included RIPK, which is involved in host signalling pathways in response to TB infection, and ATP6V0A2, encoding a component of the proton pump involved in acidifying the phagosome. The phagosome is the compartment where TB bacteria reside after being engulfed by cells such as macrophages. M. tuberculosis is capable of excluding the proton pump from the phagosome to allow its survival in a compartment which, if the immune cell had its own way, would be able to kill the bacteria. Levels of ATP6V0A2 could therefore impact on the ability of TB to survive inside cells.

In the long run, studies such as this could lead to a better understanding of why some people develop TB while others do not, and could help in the discovery of better treatments for this disease.

Saturday, 7 January 2012

Researchers at UC San Diego have created living 'neon signs' from glowing bacteria which flash on and off in unison. As described in this week's Nature, the researchers used this technology to create a bacterial sensor for detecting low levels of arsenic.
I like to imagine that the scientists working in Jeff Hasty’s lab at the University of California, San Diego were the sorts of kids who liked to taking their toys apart and reassemble the pieces into something new and better. Because this is the basic idea behind synthetic biology, a relatively new area of science that creates artificial biological systems using building blocks taken from the natural world.

Engineering living organisms as if they are machines is as awesome as it sounds. Current developments put us at the point of having created many of the individual components needed to make entire circuits and systems—individual modules such as switches, counters and clocks. Teaching a cell to count as described by Friedland et al, for example, might sound a little like it has no use in the real world. But what if we could use this to engineer safeguards into genetically-engineered microbes, ensuring that after a certain number of divisions the cell’s self-destruct mechanisms are activated? Putting the pieces together is where the future challenges lie.

One of the more well-publicised examples of synthetic biology is the creation of artificial life in the lab. Rather than wanting the play God at the risk of wiping out the human race, the true goal of this work is to understand existing life by learning what building blocks are required and testing our existing models for how cellular machinery works. But the future of synthetic biology goes beyond blue skies research. Potential uses include creating new plants or microbes that can synthesise foodstuffs or biofuels, pest-resistant crops, or living sensors such as those described in the recent paper published in Nature.

Living sensors make sense—bacteria encounter numerous chemicals, some harmful and some useful, and thus already possess methods for detecting substances in their natural environments. And bacteria are better at doing this than our current chemical detection methods, which often provide single rather than continuous readings and can require calibration each time they are used. The trick is in harnessing these natural systems to create low cost sensors capable of informing us of the presence of poisons, pollutants, or disease-causing organisms.

The paper from Arthur Prindle and Phillip Samayoa et al, describes the development of a sensor capable of detecting arsenic. But this has been a long time coming—Jess Hasty’s lab have been publishing their work in Nature for years, each time building on their earlier discoveries. And it all started with the creation of a simple biological clock in which cells were synchronised to produce blinking light in unison.

It’s been known for years that bacteria are able to talk to their neighbours using a mechanism known as quorum sensing. This involves the release of diffusible chemicals that effectively tell neighbouring cells that they are not alone—only when this signal reaches a critical level do the cells switch on the various genes controlled by this system. One of the prettiest examples of quorum sensing in action is the bacterium Vibrio fisheri’s symbiosis with the Hawaiian bobtail squid. When highly concentrated in the squid, the bacteria are triggered to produce a bioluminescent chemical that would be wasteful to make were the cells free-living in the ocean. The bioluminescence allows the squid to camoflague itself from its prey; the squid in turn provides the bacteria with nutrients.

Bobtail squid from East Timor. The blue glow is a result of Vibrio fischeri luminescence. Photo by Nick Hobgood

This ability of cells to talk to each other was the basis of the biological clock, shown in the diagram below. It involves three genes all under the control of the same promoter—a region of DNA that controls whether a gene is switched on or off. This promoter is controlled by a signalling molecule, AHL, synthesised by the product of the luxI gene. In this way, luxI drives its own expression by creating more AHL which, in turn, induces more luxI expression. As levels of AHL increase, so does the induction of the gfp gene—this makes green fluorescent protein (GFP) which makes the cells glow under UV light. The third gene, aiiA, however, makes a protein which breaks down AHL. Once levels of this inhibitor reach a critical level, AHL levels decrease, all three genes are switched off, and the process starts all over again. This creates oscillations of GFP production with a specific frequency. While the overall amount of GFP can vary depending on growth conditions, this oscillation frequency remains the same because relative amounts of gene expression are kept constant.
Oscillator setup. Figure adapted from paper. AHL concentration drives expression of all three genes. The combination of LuxL making AHL and AiiA breaking down AHL sets up oscillations in AHL concentration.

To turn this simple system into a sensor for arsenic, Jess Hasty’s lab took a promoter capable of detecting levels of arsenite and used it to modulate levels of AHL. They included a second copy of the luxI gene required to make AHL, this one regulated by the concentration of arsenite. When arsenite is present, there is less induction of this second luxI gene. This alters the ratio of luxI:aaiA induction and results in an increase in the period between the GFP oscillations which is correlated with the concentration of arsenic. And we have a sensor!
Increase in oscillation period with the addition of arsenite.

A flashing bacterial colony is pretty, but it’s a long way from a usable sensor. One problem is that one colony of bacteria isn’t going to generate enough GFP to allow detection without expensive and high-tech equipment, plus background noise can also be high. What was needed was many colonies all working together. But AHL diffuses too slowly to synchronise bacteria that aren’t in close proximity to each other.

To solve this problem, the scientists used a microfluidics setup (an example of which I talked about the other day) and a second signalling molecule, H2O2. On a chip approximately the size of a paperclip, 13,000 separate colonies (termed as ‘biopixels’ by the researchers) are induced to oscillate in a synchronised manner.
A Microfludics setup showing comminucation between colonies by H2O2. B Diagram showing the synchronisation of the oscillations by H2O2, which switches on the AHL-controlled promoters to start each oscillation.

The AHL-dependant system was engineered to also be switched on by the presence of H2O2, shown in the figure above. As H2O2 is a vapour, it can travel between colonies of bacteria very quickly. The individual colonies set up their own AHL-dependant oscillations, which produces small bursts of H2O2 that kick starts the synthesis of AHL in neighbouring colonies. In this way, H2O2 acts as a sort of pacemaker, synchronising the production of GFP as shown in the video below. Because everyone likes science when it is pretty, the researchers also used their flashing biopixels to create neon signs. It might sound more like art than science, but this technology has the potential to be turned into a simple hand-held detector for substances such as arsenic within a few years.

1. Danino, T., Mondragon-Palomino, O., Tsimring, L. & Hasty, J. A synchronized quorum of genetic clocks. Nature 463, 326–330 (2010)
2. Arthur Prindle, Phillip Samayoa, Ivan Razinkov, Tal Danino, Lev S. Tsimring & Jeff Hasty. A sensing array of radically coupled genetic ‘biopixels’. Nature 481,39–44(2012)
3. Ari E. Friedland1, Timothy K. Lu, Xiao Wang, David Shi, George Church,and James J. Collins. Synthetic Gene Networks That Count. Science 324:5931, 1199-1202 (2009)
4. Daniel G. Gibson, John I. Glass, Carole Lartigue, Vladimir N. Noskov, Ray-Yuan Chuang, Mikkel A. Algire, Gwynedd A. Benders, Michael G. Montague, Li Ma, Monzia M. Moodie, Chuck Merryman, Sanjay Vashee, Radha Krishnakumar, Nacyra Assad-Garcia, Cynthia Andrews-Pfannkoch, Evgeniya A. Denisova, Lei Young, Zhi-Qing Qi, Thomas H. Segall-Shapiro, Christopher H. Calvey, Prashanth P. Parmar, Clyde A. Hutchison III, Hamilton O. Smith and J. Craig Venter. Creation of a Bacterial Cell Controlled by a Chemically Synthesized Genome. Science 329(5987), 52-56  (2010)