The Story Behind The Killer Spinach

escherichia.jpgDon’t eat your spinach.

That’s the word coming today from the FDA: they want everyone to avoid bagged spinach until they can get to the bottom of a nasty outbreak of Escherichia coli O157:H7, a virulent strain that infects an estimated 70,000 people in the United States and kills about 60. A number of people have gotten sick in the new outbreak, apparently from eating contaminated spinach, and there’s been a report of one death in Wisconsin.

There’s a fascinating–albeit gruesome–backstory to this outbreak, which I’ve been researching for my next book, a portrait of Escherichia coli. Escherichia coli is regular inhabitant of the human gut (not to mention the guts of mammals and birds). You carry about a trillion harmless E. coli. E. coli has also become the model par excellence for understanding the nuts and bolts of life. Lots of Nobel Prizes were awarded for research on these fascinating bugs.

Over the twentieth century, scientists began to discover that some strains of Escherichia coli are not so nice. A group of strains called Shigella cause diarrhea, for example, killing over a million people a year. And new virulent strains keep turning up. In March 1982, 25 people in Medford, Oregon, developed cramps and bloody diarrhea. The doctors identified a strain of Escherichia coli in some of the patients. The strain could not be found among the records of the Centers for Disease Control. That had never happened before. Three months later, another outbreak occurred in Traverse City, Michigan. The source of the bacteria proved to be undercooked hamburgers that the victims had eaten at a McDonald’s restaurant. Scientists named the strain O157:H7–a code for its distinctive surface molecules–and began to hunt for it among the bacteria that had been isolated from patients in earlier years. Out of 3,000 strains collected from American patients, a single one proved to be O157:H7, isolated from a woman in California in 1975. Searches in Great Britain and Canada turned up seven more cases, none before 1975.

O157:H7 slipped back into obscurity for a decade. It emerged again in the mid-1990s in a series of outbreaks spread across the world. One outbreak in Washington State, spread in undercooked restaurant hamburgers, sickened 732 people in 1993. Four of them died. More outbreaks flared up in other countries. In Japan radish sprouts tainted with O157:H7 sickened 12,000 people in 1997.

The outbreaks mobilized a lot of scientists to figure out what in the world this pathogen was. It is one of many strains of Escherichia coli that live in cows and sheep, dwelling peacefully in their guts. About 28% of cows in the United States are estimated to carry O157:H7. It can get into people through contaminated meat thanks to bad butchering (a single crumb of undercooked meat carrying 100 bacteria is enough for a potentially fatal infection). The animals also release the bacteria in their manure, from which it can be spread to vegetables and fruits, perhaps by the wind, slugs, or flies. It is also the reason why you really must wash your hands if you visit a petting zoo.

Once it gets in our bodies, E. col O157:H7 starts causing serious trouble. It builds a needle that it can jab into the cells of our guts. Through it they inject a cocktail of molecules. One of the first of these molecules is a receptor, which inserts itself into the wall membrane of the intestinal cells. In other words, Escherichia coli makes our cells part human, part microbe. The receptor can take in more of the molecules Escherichia coli O157:H7 produces, which alter how the cell reads its genes. Over 2000 genes in the cell change their patterns of expression. Some genes make more proteins and others shut down.

As its genes are manipulated, the cell begins to behave oddly. The skeleton-like fibers that support the cell begin sliding over one another to create a new shape. A pedestal-shaped cup rises from the top of the cell, offering Escherichia coli O157:H7 a place to rest. The cell also begins to leak fluids, which rush past the microbes. Safely cradled on their pedestals, the microbes can absorb nutrients in the flow. They also release toxins that dislodge hemoglobin, the oxygen-ferrying molecule, from blood cells. O157:H7 needs iron to grow, and hemoglobin is an iron-rich molecule. As the host bleeds, the microbe pulls fragments of hemoglobin into its interior and pries the iron atoms free.

These changes cause blood diarrhea, but they are not the worst of Escherichia coli O157:H7’s symptoms. Sometimes a few of the bacteria swell with toxins and burst.Their toxins enter our own cells, where they jam up the cellular factories that build proteins. Unable to make new proteins, the cells die and burst open. These toxins can slip into the blood vessels lining the intestines and soon spread to other organs. The kidneys are especially vulnerable to their attacks.

How can the same species be so gentle and so nasty? It has evolved. Escherichia coli has diversified into different strains. These strains adapted to different ecological niches. In some cases that adaptation made them harmless to us, allowing them to just graze on sugar cast off by other bacteria in our guts. In other cases, that adaptation made Escherichia coli a killer. Many different changes gave rise to this evolution. In some cases, genes spontaneously mutated and provided an extra reproductive edge to a microbe. In other cases, Escherichia acquired genes from other bacteria. In many of those cases, the genes came in bundles, delivered by viruses that pasted themselves into its genome and then became integrated into their host. You can find related versions of much of Escherichia coli O157:H7’s equipment in other strains of E. coli and in other species of bacteria.

The evidence suggests that the ancestors of Escherichia coli O157:H7 and its disease-causing relatives branched off from the best known harmless strain, called K12, a few million years ago. But it acquired some of its most important equipment more recently than that. Here’s a paper that presents some of the latest evidence about how O157:H7 and its relatives began acquiring new things such as toxins starting about 50,000 years ago. But it may not have been evolving into a better human pathogen. Instead, it may have been adapting to the animals that became our livestock about 10,000 years ago. Some research even suggests that O157:H7 helps sheep fight off cancer-causing viruses. For some reason scientists don’t yet really understand, when it gets into us, its behavior causes disease instead of health. In other words, humans created the conditions–from the first herds to the latest giant ranches and slaughterhouses–that fostered the evolution of this scary beast. And in the past 30 years, changes in how we get our food have allowed O157:H7 to emerge as a serious threat to public health.

And this evolution has not stopped. New strains of E. coli that cause much the same disease as O157:H7 have emerged in recent years. They are aquiring viruses bearing genes for toxins and other important equipment for making us sick. Those viruses are floating around in our sewers in vast numbers, with about one percent carrying various version of the toxin genes.

So I for one will be laying off the spinach for the next few days, and spending some more time getting to know these strange creatures and how they came to be.

Update: Naturally, Tara Smith has more. And Mike too.

Update 9/18 1:30 pm: Why antibiotics are the last thing you want to take if you get sick with this bug.

0 thoughts on “The Story Behind The Killer Spinach

  1. Dateline NBC recently covered a similar issue, to use caution with bagged lettuce. People dying… kidney failure, etc. Simple lettuce. Who knew?

    Urban Legends Reference Pages: E. Coli Salad Warning
    For those of slightly less cautious nature, Dateline NBC offered these tips on how to protect yourself from E. coli in lettuce: …

    Unseen danger in bagged salads – Dateline NBC – MSNBC.comEven if the lettuce looks good, you should know E.coli can grow quickly in greens that are … 30, 2006. This report aired on Dateline NBC, April 30, 2006. …

    Google “lettuce dateline nbc e coli”

  2. So, if I have my math right, this bacterium has made .02% of the US population sick, and killed .09% of that .02%? I like those odds.

  3. The most interesting thing about this to me is that the bacterium can modify the shape of the cell so dramatically. How the many molecules responsible for cell structure coordinate to produce the overall shape of a cell is, as far as I know, very poorly understood. It seems like this bacterium/eukaryote interaction would offer a good starting point for further investigation.

  4. I’ve always wondered, with this bacterium and others, if it can occur in so many forms that vary quite radically, how does the definition of the species hold up? Is the range of variation still smaller than differences among bacterial species?

  5. Hi, I’m aware biologists and scientists frequent this blog. I’ve got an off-topic, but pretty serious question about a squirrel I caught on camera earlier. The closest thing I came to on Google was “notoedric mange”, and learned it can kill animals if they lose enough hair, but after looking at the photos, it doesn’t seem to be what was ailing this particular squirrel.

    This squirrel was bare across its shoulders of hair, and had what appeared to be some kind of growth(s) hanging from it’s shoulder.

    I’ve never seen anything like this on any of the squirrels around here, (North Carolina).
    I’ve uploaded some images to Photobucket.
    Can anyone shed some light on this?
    Whatever it is, isn’t pretty.

  6. David Nelson wrote:

    I’ve always wondered, with this bacterium and others, if it can occur in so many forms that vary quite radically, how does the definition of the species hold up? Is the range of variation still smaller than differences among bacterial species?

    This is a really good question, and I am not sure that there is yet an answer that everyone finds satisfactory. However, as I understand it, there exists a conserved “backbone” which in essence defines the “species.”

    However, even the idea of “species” is difficult to apply to bacteria as most reproduce asexually, and therefore a species cannot be defined by a reproductive barrier. (Yet there are some bacteria which reproduce by sexual means, and some only by sexual means, and those which reproduce only by sexual means have linear chromosomes with telomeres rather than the standard circular chromosomes found in most bacteria.)

    Beyond this, it is sometimes argued that what promotes the stability of a given bacterial species is essentially that similar populations fill similar ecological niches.

  7. Carl,

    You briefly mention the phages. I personally find them quite fascinating. In the human body alone, we have 4 to 5 trillion eukaryotic cells, but 100 trillion prokaryotic cells, and of course there are plenty of bacteria which reside outside of multicellular organisms. At the same time, while there exist 10^30 bacteria in the biosphere, there exists ten phages for every prokaryote, and they may very well play an important role (for example) in the carbon cycle. Many damage their hosts upon replication, but others bear presents to pay for their keep.

    Pathogenicity islands (which are composed of genes for virulence) are the accreted from such interactions, but then so are metabolic islands and symbiosis islands – and this sort of exchange has been going on for about as long as bacteria themselves have existed. In the process, phages have undoubtedly driven much of bacterial evolution, but likewise have been shaped by it, giving rise to the highly modular structure of phages themselves. Likewise, they are a major component in and beneficiary of the small-world lateral gene transfer network which exists in the bacterial world. Their relationship with bacteria is often largely symbiotic.

  8. Tim–The phages in O157:H7 are particularly weird, since they carry the toxin genes. That’s why you can’t give antibiotics for this strain. The phages sense the stress and break out, producing toxins.

  9. The transfer of proteins by E. coli into human cells is a fascinating subject. This process occurs through the Type III secretion pathway, whose apparatus forms the bridge between the bacterial and human cell. Elements similar to the Type III pathway are implicated in the formation of the bacterial flagellum, of “irreducible complexity” fame.

    I believe the genes used by E. coli to form the Type III pathway have been shown to have originated from Yersinia pestis (bubonic plague), at least their sequence similarity is high enough to support this hypothesis. Thus, E. coli acquired these genes ‘across the species barrier’.

  10. Carl (and Tim),
    The phages are interesting, but not particularly weird, as it turns out. Phage genomes are frequently found to contain virulence genes (cholera toxin, for example, is in the CTX phage in V. cholerae; Staph enterotoxin A is also on a phage, as are several other (read, many) staph toxins). This is analogous to why naturally occuring plasmids and transposons frequently contain antibiotic resistance elements, and forms the basis (in some sense) for “selfish gene” evolutionary ideas.

    Carl, this is a beautiful essay, that describes O157:H7 very nicely. The issue of species in bacteria is a tough one; Shigella is considered by classical microbiologists as a separate species, but is very, very close to E. coli metabolically, physiologically, and genomically. O157:H7 and K12 are at least as different as either one and Shigella, at least in some ways. We tend to use old metabolic keystones to type bacterial strains and species, and old habits die hard…

  11. I hadn’t read Mike’s post (well, I did, a long time ago, but I’d forgotten it). There’s no doubt E. coli is Shigella phylogenetically.

  12. Carl wrote:

    Tim–The phages in O157:H7 are particularly weird, since they carry the toxin genes. That’s why you can’t give antibiotics for this strain. The phages sense the stress and break out, producing toxins.

    I find it interesting that the phages become lytic (destroying the bacteria, and thereby spreading their virons to other bacteria) when it would seem that they are normally temperate, particularly as there still exists in many circles the view that viruses (and therefore phages) aren’t really alive since they have no innate metabolism. With respect to the gene for iron uptake, this is of course part of a pathogenicity island, but within a different ecological niche (outside of the prokaryote’s current host or within a different prokaryote) would simply be a metabolic island.

    Here is one older technical article which deals with this sort of thing that I ran across within the past few weeks (for those who are interested):

    A Genomic Island, Termed High-Pathogenicity Island, Is Present in Certain Non-O157 Shiga Toxin-Producing Escherichia coli Clonal Lineages
    H. Karch, et al.
    Infection and Immunity, November 1999, p. 5994-6001, Vol. 67, No. 11

    “The genomic islands may contribute to the fitness (fitness islands) or metabolic flexibility (metabolic islands) of the organisms, or they may increase their pathogenic potential (PAIs). The particular function of an island will thus depend strongly on the genetic background of the individual strains. Further experiments are necessary in order to define the exact role of the HPI element in the life cycle of STEC strains.”

  13. Here are a couple of other articles which have a somewhat broader view:

    Ecological fitness, genomic islands and bacterial pathogenicity
    A Darwinian view of the evolution of microbes
    J�rg Hackera and Elisabeth Carniel
    EMBO reports 2, 5, 376-381 (2001)

    From the article:

    “Fitness islands can be subdivided into different subsets, depending on the life-style of the microbe (its niche) (Figure 3), rather than on the intrinsic composition of the islands. Fitness islands that help microorganisms to live in the environment or to persist as saprophytes in a host may be considered ‘ecological islands’ and ‘saprophytic islands’, respectively. Other bacteria reside temporarily or permanently in a host (another microorganism, a plant or an animal), where they either provide some benefits to the host-organism (symbiont) or cause damage to it (pathogen). Accordingly, a ‘symbiosis island’ is a specific type of fitness island that helps bacteria to positively interact with their hosts, while a fitness island that participates directly or indirectly in the induction of lesions is a true pathogenicity island.”

    Evolving Insights: Symbiosis Islands and Horizontal Gene Transfer
    Turlough M. Finan
    J Bacteriol. 2002 June; 184(11): 2855-2856.

    Then here are two more reviews, the first of which is a real classic on the modularity of phages…

    Evolutionary relationships among diverse bacteriophages and prophages: All the world’s a phage
    Roger W. Hendrix, et al.
    PNAS | March 2, 1999 | Vol. 96, no. 5 | 2192-2197

    Phages and the Evolution of Bacterial Pathogens: from Genomic Rearrangements to Lysogenic Conversion
    Harald Br�ssow, Carlos Canchaya, and Wolf-Dietrich Hardt
    Microbiology and Molecular Biology Reviews, September 2004, p. 560-602, Vol. 68, No. 3

  14. … and given that I know how Carl likes the subject of the origin of life, here are a couple of real classics which became available on the web not too long ago – also dealing with phages:

    An Extracellular Darwinian Experiment with a Self-Duplicating Nucleic Acid Molecule
    D. R. Mills, R. L. Peterson, and S. Spiegelman
    Communicated May 18, 1967
    PNAS 1967; 58: 217-224.

    Evidence for De Novo Production of Self-Replicating and Environmentally
    Adapted RNA Structures by Bacteriophage Q3 Replicase
    Manfred Sumper and Rudiger Luce
    Proc. Nat. Acad. Sci. USA
    Vol. 72, No. 1, pp. 162-166, January 1975
    Communicated by Manfred Eigen, October 11, 1974


    PS I sent in links some other more directly relevant articles, but the blog mistook them for spam and is holding them for the moment.

  15. My apologies to Mr. Zimmer for posting off topic earlier in the thread, I know, I know. But I hoped to get an answer from a scientist.

    I tried the talk origins group to identify what was wrong with the squirrel, and “Klaus” responded, “It looks like a grey squirrel that survived a serious injury… massive scar tissue on its back.” in, Susan responded about a pet fox which was attacked, “She lost all fur in that spot forever and it had the same appearance of your squirrels problem. Those nodules remind me of when I cut my wrist on an aquarium. The wound healed but a nasty, painful nodule formed due to the trauma… a friend with open heart surgery had the same nodules running up and down the incision area. Could the squirrel have been attacked (cat, hawk, whatever)..? Not a clue but that would be the most likely area a predator would grab.”

    It’s probably an old nasty injury as they both suggested.

    This presents a question for Creationists,
    Lk:12:24: Consider the ravens: for they neither sow nor reap; which neither have storehouse nor barn; and God feedeth them: how much more are ye better than the fowls?
    Lk:12:7: But even the very hairs of your head are all numbered. Fear not therefore: ye are of more value than many sparrows.

    The young lions roar after their prey, and seek their meat from the Lord…[What if that “meat” is a Christian in a Roman arena? ? Skip Church]…Oh Lord, how manifold are thy works! In wisdom hast thou made them all…both small and great beasts… These all wait upon thee; that thou may give them their meat in due season.
    – Psalm 104

    Speaking of how well “the Lord” “satisfies the desire of every living thing,” let’s take “the young ravens which cry” as a prime example. A recent study showed that one-third of adult birds and four-fifths of their offspring die of starvation every year (David Lack, “Of Birds and Men,” New Scientist, Jan., 1996). Not surprising, since birds have to eat from one-quarter to one-half their body weight daily, so starvation is a common killer of birds.

    This creator has time to count hairs, and “feed the ravens”, part of the time. But no time to heal their wounds. No, the creator is too busy “intelligently designing” bigger and better tooth and claw and the arms race goes on … “better to see you with, better to hear you with, better to EAT YOU with.”

    Carl, my apologies again for my tactless interuption

  16. Sharon,

    The squirrel was a bit off topic, although not necessarily that far off, but the passages from the Bible is much more of a reach. What you are bringing up is simply the traditional problem of evil – which more a problem in theology than biology. With respect to the creationists, it is not their belief in Christianity which is a problem, but rather their extremist views which they would seek to impose upon the rest of society, eventually as part of a theocracy. For those who have any doubts in this regard, I would recommend looking up the Reconstructionists/Dominionists – and remind them that Howard Ahmanson (probably the largest benefactor of the Discovery Institute) was a board member of the Chalcedon Foundation – a leading organization within this movement.

    For more on the Reconstructionists, there was a fairly recent post (a few weeks back) on ScienceBlogs a few weeks back which could serve as an introduction:

    Dispatches from the Culture Wars: A Weak Defense of Reconstructionism
    Category: Theocracy
    Posted on: August 23, 2006 9:24 AM, by Ed Brayton

    … but for the moment at least, I would prefer to focus on something I find far more cheery, like phages, e coli, or for that matter, the bubonic plague.

  17. The squirrel was a bit off topic,

    Oh, I know, and my sincerest apologies.
    … but for the moment at least, I would prefer to focus on something I find far more cheery, like phages, e coli, or for that matter, the bubonic plague.

    That’s why I posted. I feared something like that had crept up in my front yard, and might have a scientist identify the condition, as many do frequent this blog. I was unsure who to contact, that could answer… As for your outside reference, I had not even considered, does have numerous science experts working with them, perhaps there is one available that can identify diseases or ailments in animals, in the future.

    Again, the squirrel took me by surprise, and my apologies for panicking.

  18. … but for the moment at least, I would prefer to focus on something I find far more cheery, like phages, e coli, or for that matter, the bubonic plague.

    I spoke too soon. I just checked the talk origins thread, and “Jerry Sparks” has identified the filthy cause of the squirrel’s condition.

    Again, my apologies to Carl for cluttering his blog.

  19. The phage lysis response is classic temperate phage behavior. The lysis/lysogeny switch (best studied in lambda) is triggered by DNA damage or increased activity of DNA repair enzymes. Antibiotic action (bactericidal) is indirect; the damage to the cell’s metabolism results in autolytic responses, but beforehand, the SOS response is activated, including lots of DNA repair mechanisms. This triggers phage lysis. It’s very elegant, actually, from the phage point of view-as long as the cell is succeeding (and the toxins and virulence factors help that happen) the phage gets a free ride, replicating at a pretty high rate (E. coli optimally every 20 m. When there’s trouble, the phage turns on the lytic cycle, and releases infective particles to find a new host that isn’t in trouble. From the bacterial point of view, it works pretty well-when times are good, the bacterium has some extra useful genes for acquiring goodies. When times get bad, well, you’re probably dead anyway, so what difference does it make what kills you?

  20. Paul Orwin wrote:

    The phage lysis response is classic temperate phage behavior. The lysis/lysogeny switch (best studied in lambda) is triggered by DNA damage or increased activity of DNA repair enzymes.

    Exotoxins of Staphylococcus aureus

  21. Is O157:H7 more resistant to antibiotics than the harmless form, and could it have been encouraged by the feeding of antibiotics to livestock?

  22. According to what both Paul Orwin and Carl Zimmer wrote, antibiotics don’t really improve the situtation. It isn’t so much that O157:H7 itself is resistant to antibiotics as much as that the phage responsible for its virulence will enter lysis (where it reproduces then bursts out of the bacterium) in response to the stress caused by the antibiotics (or judging from what Paul Orwin wrote, in response to the bacterium’s own stress response), releasing virions which will then infect other e coli. As Carl Zimmer pointed out, we have about a trillion in our intestines which are harmless members of intestinal society – but if they become infected, they will be transformed into the dangerous kind. So in this case, it isn’t so much that the bacterial strain is resistant to antibiotics, but that the bacteria’s response to the antibiotics will result in the release of virions which infect the otherwise harmless strains.

    Nevertheless, obviously we are having problems with the misuse of antibiotics which is resulting in strains of bacteria which are resistant to antibiotics and the genes for such resistance are being transmitted by other phages. This is why putting antibiotics into cattle carcasses is a bad idea and why putting it into feed is probably a bad idea as well. In essence, bacteria have established a kind of small-world network reminiscent of the “Six Degrees to Kevin Bacon,” in which it takes relatively few steps to get from any one part of the network to any other – given the structure of the network and use of centralized hubs. In some case, the phages which infect different species of bacteria are even able to reach across biome boundaries – from the ocean into the soil.

    It would appear that one of those hubs may be in the soil itself. Of course, this sort of network isn’t something which has sprung up only since the introduction of antibiotics by humans. It has been there for a very long time, providing phages with hosts to infect and bacteria with genes with which to adapt to a changing world. Interestingly enough, we are also sometimes finding that bacteria are resistant to some antibiotics before they ever reach the market. It appears that some strains of bacteria will produce antibiotics for use against other strains of bacteria which will respond by developing resistance. Thus when we introduce an antibiotic which is sufficiently similar to antibiotics already in use in nature, some bacteria may already be resistant.

    Incidentally, understanding this provides us with a certain edge. Some companies creating new antibiotics will test them against the strains of bacteria found in soil to see whether they already have a degree of resistance. If they do, the company will know to hold off on mass-production until they come up with something else.

    One more point which may be of some interest: many of the genes and pathways of pathogenicity which the bacteria use against their multicellular hosts also seem to have been around for a very long time. In some cases, they are using essentially the same mechanisms against us as they were using against our single-celled ancestors and continue to use against single-celled eukaryotes today.

  23. Dan,

    The article you have cited is well worth reading, and I will be adding it to my collection. My primary interest is in how things evolve – and this is definitely down my alley.

    Incidentally, I have also posted “links” on Tara Smith’s blog since she is also writing pieces on this subject.

    Please see:

    I myself will be keeping an eye on both blogs as there is bound to be some interesting material of significance which goes well beyond the current outbreak.

  24. carl–you write: “One of the first of these molecules is a receptor, which inserts itself into the wall of the intestinal cell. In other words, Escherichia coli makes our cells part human, part microbe. ” I thought that the second sentence was a bit ‘breathlessly dramatic’- which i’ve enjoyed your writing style for NOT having. But , more to the point- since when do eukaryotic intenstinal cells have WALLS?
    granted this post isn’t as heavy as evolution and phages etc, but c’mon carl, don’t dumbdown the basic science for us layfolks.

  25. My name is Gevea Youmans. I live in Fayetteville North Carolina and attend Seventy First Classical Middle School. I am 13 years old and my Social Studies teachers name is Mrs. Parker and i think that this Spinach is not right and im glad im not one of the people who died.

  26. Just a nit that you don’t need to publish. I noticed that the photo on your article and RSS feed is the same photo you used a few days before for a similar article.

    I don’t know if this affects other readers, but I unconsciously index stories visually and seeing the same photo on a later story confused my mental timeline quite a bit.

    It’s surprising to me, but I was completely disoriented when I saw that familiar photo of cells. I didn’t know what day it was or where in your string of RSS articles I was suddenly transported, but I knew that photo was familiar and the fact that it was familiar was what was disorienting.

    Just FYI, another of those annoying “user interface” issues to think about. I wouldn’t think to annoy you with this, except that I absolutely love your blog and figure that you care about such things.

Leave a Reply

Your email address will not be published. Required fields are marked *