#MicroTwJC : The Creation of a Superbug

The year was 2004. The patient was a 6 month old baby girl. She was about to enter thoracic surgery, when the doctors found that she was harbouring methicillin resistant Staphylococcus aureus. Now, in most western hospitals, the origin of this bacterium would not be a mystery. But this was a hospital based in the Netherlands. The Dutch have a "search and destroy" mentality when it comes to dealing with superbugs, and have been very successful at keeping their hospitals free of MRSA. They wanted it to stay that way. They had to find the source of this MRSA, and put a stop to it. The hospital equipment was scrutinised for any traces of the bacterium. None could be found.
They eliminated the MRSA from the baby, and then sent her home with her parents. But when they followed up, the baby was once again colonised with MRSA. They went through the same process again and again, until they realised that the baby was continuously being re-infected with the bacterium from an unknown source. The doctors found that the babies parents were also carriers of MRSA. but where did they get the disease from ? If it wasn't coming from the hospital, then where was it coming from ?
It turned out that the family lived on a farm raising pigs. The pigs were tested. They were the source of the MRSA.
Other pigs on different farms in that area also carried this strain of MRSA. A number of other cases of farmers and vets catching MRSA off their pigs. They concluded that farmers were 760x more likely to get an MRSA infection than any other Dutch people. Further research revealed that 39% of pigs entering a slaughterhouse carried MRSA. Hospitals in close proximity to pig farms tended to see more patients with MRSA than hospitals that were far from pig farms. This MRSA appears to be different from the hospital associated MRSA's we are more familiar with. It is primarily carried by pigs, and was a leading cause of MRSA infection in the Netherlands.
 So now we know that pigs can carry MRSA, it is time to ask an important question. How did they get MRSA  ?  How did this particular strain evolve ? These are the questions that this weeks #MicroTwJC paper aims to answer.

So how would you go about tracing the heritage of a bacterium ? It's not like we can use census records or old photo albums to track down their ancestors. But there are records that we can use to study the genealogy of a bacterium. Every single living thing on this planet keeps a record of their genealogy within their genetic code. All you need are the right tools and talent to read this code, and the ancestry of the bacterium should unfold before you.

Just like us, bacteria inherit their genetic code from their forebears. When a bacterium reproduces, it replicates it;s DNA , which is inherited by its two daughter cells. As the DNA is replicated, errors occasionally occur in the code, changing it from its ancestors. These errors are known as mutations, and can change the functions of genes. Most importantly, once these mutations creep into the code, they stay there. If the bacteria carrying a particular mutation survives to reproduce, it's descendants will still carry the mutation.
So if we see a group of bacteria with the same mutation, we can trace that quirk of the genetic code back to a common ancestor. This can tell us a lot about how bacteria are related.

This brings us to an important point. You can't look at one bacterium and divine the history of its genetic code from nothing. You always need to compare it with other bacteria. This is crucial. In order to build a family tree for this bacteria, we need to get a good look at its family. To this end, researchers sequenced the genomes of 63 livestock associated MRSA. The researchers needed to go further, so they also looked at other closely related bacteria, such as MSSA (methicillin sensitive Staphylococcus aureus. Also known as Staphylococcus aureus)  and hospital MRSA's, which are the two most likely candidates from which this bacterium evolved. They looked at 40 and 43 genomes from each of these groups respectively. They also made sure to take samples from all over the world. This allowed them to peg the evolution of each of these strains geographically.
There are a number of possibilities that could explain how this particular livestock associated MRSA evolved.

1. MRSA Tourism: MRSA is common in hospitals, and most of the strains we know of evolved in the clinic. It could be that a human carrier of MRSA passed it on to a pig. The pigs would have been fed growth promoting antibiotics, which would select for MRSA, allowing it to be sustained within this population. In this case, the strain would show clear indicators that it had evolved from an MRSA strain, without showing any characteristics unique to MSSA strains.
2. The Origin of MRSA: Perhaps an MSSA strain jumped into livestock, and became resistant to agricultural antibiotics to become MRSA. It then leapt over into man where it eventually formed the basis for the whole MRSA population that currently infest our hospitals. In this case,Livestock MRSA would show mutations that are similar to the MSSA strains that also distinguish it from the MRSA strains.
3. MRSA Independant evolution: Perhaps Livestock associated MRSA is completely unrelated to the MRSA's we find in humans. It is like the above scenario, only this MRSA never made the jump into hospitals.  Maybe In this case it would show no similarity to any of the MRSA strains tested, but definite similarity with the MSSA strains.
4. Convergent Evolution in Multiple Regions: Perhaps we are not just looking at one single pandemic strain of livestock associated MRSA. Perhaps MRSA evolved in livestock multiple times in different regions. In this case, we would see that different regions have different livestock associated MRSA strains, with different origins. This would be a situation in which one or more of the above pathways of evolution could have occurred in order to reach the same end point.

Once they had all of the MRSA and MSSA genomes sequenced, they had to analyse them. They tried MLST, a technique that I talked a fair bit about in a previous #MicroTwJC. The key thing about MLST is that it looks at seven genes and using their deletion/ alteration to measure the relatedness of strains. Conceptually it is sort of similar to the example of mutations I talked about above. But there is one quirk of bacterial evolution that I didn't mention, which is the main reason why MLST is necessary.

Bacteria can steal genes off eachother, and occasionally completely lose genes as well. If you are looking at the mutations in just one gene, you have to realise that the bacterium carrying that gene may not have evolved it itself, it may have stolen it off of another bacterium. Or it may have completely lost the gene, thus giving you no information at all. MLST avoids this problem by looking at multiple genes. These genes are spaced out along the bacteria's chromosome. If a bacteria decides to get rid of one section of its genome, it means that only one of the genes you are looking at is lost and you still have the other genes to study. In fact, this gene loss works in your favour. You can now tell how closely other bacteria are related by looking at the MLST pattern they have. If a group have similar gene patterns, except for one missing gene, then they are likely to be related. If one bacteria steals a gene off another, it creates a new lineage with a distinct gene pattern. These groupings tend to be names "eBURST" groups, after the kind of statistical analysis used to discover them.

Whilst MLST is good, it only tells you about the lives of seven genes. If you want to go into even more detail, you will need to look at the whole genome sequence. Instead of looking at just seven genes on the chromosome, we can look at all of them. But this comes with some degree of problems. Because there are some portions of the genome that are constantly mutating and changing at such a fast rate that there is a high probability they will generate the same sequences due to chance. You need to focus on relatively stable genes. But if you know what you're doing, this can be really useful. If there are specific genes you are interested in, like the ones which encode antibiotic resistance, or virulence, all of that information is stored in the genome sequence.

Whole genome sequencing also allows for an unparalleled look at the exchange of genes between different strains of bacteria. Bacteria exchange genes through a variety of factors, but one of the most important are through viruses called bacteriophages (phages for short). These are parasites that insert themselves within the genome of their hosts, tricking the host into using the genetic information they encode into making new viruses. When they splice their DNA out of the host genome, they can take a souvenir. In some cases, this can include a virulence gene, or an antibiotic resistance gene. These viruses can then become inactive for generations while their host goes forth and multiplies. On occasion, these viruses can become reactivated and spread their genes to a new host.

Over the years, a number of phages which benefit bacteria have evolved, and insinuated themselves in the genomes of bacteria. Whole genome sequencing allows for these phages to be identified, and we can go some way to tracking their journey across different strains of bacteria.

They used all of this information to build up a family tree of the different strains of livestock associated MRSA, which is shown below :

The left of this diagram shows the family, and on the right we see a table giving more details of each strain. Each block of colour indicates a specific group. The grey and orange blocks on the bottom are groups of MSSA that have been found to infect humans. The green colour indicates the live-stock associated MRSA. Each successive colour indicates a further subgroup for each lineage.
Those lines of numbers at the end of each tree are the names given to each strain. These strains are the family members that were used to work this tree out. We can learn further details about them if we look to the right side of this diagram. The first column indicates each strain's country of origin with a two letter code, Shown below.

AT, Austria; BE, Belgium; CA, Canada; CH, Switzerland; CN, China; DE, Germany; DK, Denmark; ES, Spain; FI, Finland; FR, France; GF, French Guiana; HU, Hungary; IT, Italy; NL, The Netherlands; PE, Peru; PL, Poland; PT, Portugal; SI, Slovenia; US, United States

The next column notes the host of these bacteria. H is for human, T is for Turkey, P is for Pig and B is for Cow.
The columns to the right of this are a bit more involved. The next column tells us about a gene called spa. spa encodes Staphylococcal Protein A, which is one of the proteins that has been used a lot in the past to classify different strains of staphylococci. It is usually produced in high amounts, making it easy to detect. It binds to antibodies, preventing it's host from using them against it*.
The next column indicates whether these strains carry a specific bacteriophage φSa3. This phage carries a few genes that help Staphylococcus aureus survive in humans. There are other types, but we don't really need to get into them for this figure. I could talk about φAvβ phage which allows Staphylococcus to colonise birds.
The next three columns indicate what types of antibiotic resistances each of these strains have. The Tet column indicates whether the bacteria are resistant to tetracycline or not. The MET column indicates whether the bacteria are resistant to methicillin or not. The final column SCCmec tells us about the kind of genes the Staphylococcus aureus carries in order to encode these resistances.

Conclusion
So now we can understand this graph, let us take a closer look at what it tells us.

• It shows that all of these livestock associated MRSA's share a common ancestor
• That ancestor came from the same family as the MSSA's studied here. This knocks out "MRSA tourism" theory of evolution. At some point, a human MSSA crossed over to a farm animal, which could have been a pig or a cow. 
• In the course of making this transition, these MSSA's lost the φSa3, which helps it infect humans better.
• These bacteria also acquired a resistance to tetracycline. This shouldn't be too surprising, because tetracyclines were amongst the first antibiotics that were discovered to be growth promoters, and they are still used by many farms. Any bacteria attempting to survive in livestock would have to develop resistance to these antibiotics.
• The evidence suggests that the crossover between humans and their animals occurred in the US. After this crossover, the Staphylococci branched into two lineages.
• In one lineage, bacteria acquired the φAvβ phage, which allowed them to infect turkeys. This lineage remained methicillin sensitive, and has not been known to colonise humans. This strains descendants have not been found outside the US
• The other lineage acquired Methicillin resistance through different SCCmec genes on multiple occasions. Furthermore, this strain had legs. It spawned descendants that would invade Canada and Europe.
• The Strains that invaded Europe evolved and diversified, until you get to the top group in pink, which began to spread to humans in Denmark.

From this information, we can see that livestock associated MRSA evolved in the US, and then made the jump into Europe, where it was eventually discovered in the Netherlands.

Because of the common ancestor, we can rule out the "Convergent evolution in multiple regions" hypothesis.

So this leaves two hypotheses standing. The "Origins of MRSA" hypothesis would require that these livestock MRSA's spread to a hospital and cause a raft of human infections. This requires is a cluster of MRSA's within the green area of this graph that cause human infections. What we actually see is an occasional human infection occurring across a broad range of lineages that infect livestock. This suggests that the livestock MRSA's have not completely made the jump back into human infections.

This leaves one hypothesis left for the evolution of livestock associated MRSA- the one that is completely independent of the MRSA's we usually get in hospitals.

So why is it that livestock associated MRSA's were first discovered in the Netherlands, and not the US ?  Have the Danish MRSA strains become better at infecting humans ? Or are livestock MRSA most commonly observed there because they conduct the most thorough searches for it ? Should we even be concerned about these strains of MRSA, especially seeing as they don't tend to infect humans ? Is the intensive agriculture required to raise cheap meat worth the potential danger of creating new types of antibiotic resistant bacteria, as demonstrated here ?
 These are the questions that this paper raises, and how you answer it affects us all whether we like it or not. Perhaps you think that this is reason enough to ban all antibiotics in agriculture, and food budgets across the world tighten to cope with the rising cost of meat. Or you may decide that the threat from these bacteria in livestock is overblown, and stopping the creation of antibiotic resistant bacteria is not worth the extra costs.
But always remember that as consumers and active participants in democratic society**, you have the power to choose.
But remember, that the choice you make could determine whether a dangerous bacterium in the Netherlands attacks 6 month old baby.

References

Price L.B., Stegger M., Hasman H., Aziz M., Larsen J., Andersen P.S., Pearson T., Waters A.E., Foster J.T. & Schupp J.  (2011). Staphylococcus aureus CC398: Host Adaptation and Emergence of Methicillin Resistance in Livestock, mBio, 3 (1) e00305-11-e00305-11. DOI: 10.1128/mBio.00305-11

van Rijen M.M.L., Van Keulen P.H. & Kluytmans J.A. Increase in a Dutch hospital of methicillin-resistant Staphylococcus aureus related to animal farming., Clinical infectious diseases : an official publication of the Infectious Diseases Society of America, PMID: 18171259

Voss A., Loeffen F., Bakker J., Klaassen C. & Wulf M. Methicillin-resistant Staphylococcus aureus in pig farming., Emerging infectious diseases, PMID: 16485492

I've been writing a five part series of blog posts about our relationship with Antibiotics and Animals. If you liked this post, you may want to check them out. You can check out parts 1, 2 and 3 . Parts four and five will be published on the next two days that begin with a T.

 *If you ever try to perform a western blot, or any other process reliant on antibodies, on Staphylococcus aureus without accounting for this protein, you will run the risk of getting the following reaction.
** Unless you are one of my readers from China. However, as most of my Chinese friends assure me, democracy is over-rated, and the leaders only ever do what's best for the health of their nation.