Monitoring of Antibiotic-Induced Alterations in the Human Intestinal Microflora and Detection of Probiotic Strains by Use of Terminal Restriction Fragment Length Polymorphism
Cecilia Jernberg, Åsa Sullivan, Charlotta Edlund, and Janet K. Jansson
Applied and Environmental Microbiology, January 2005, p. 501-506, Vol. 71, No. 1
The Big Picture:
Antibiotics are a great tool for treating bacterial infections, but they are something of an atomic bomb. Broad spectrum antibiotics kill all sorts of things, intended and unintended. Some of the unintended victims are happy symbiotic Lactobacillus and other lactic acid bacteria that love the pH 5 and rich nutrient soup of the small intestine. They help us digest, enrich our lives with vitamins, and so on. People who have taken antibiotics often get various forms of diahrrea, as well as other GI distress, because they have killed their nice symbiotes.
To address this, scientists have recommended investigating "probiotics." These are organisms that help out around the body. The simplest case are some nice lactic acid bacteria to enhance our normal flora, or even replace those wiped out by antibiotics. These lactic acid bacteria are found in sources like kefir, yogurt, sourdough, saurkraut, pickles, etc (only in unpasturized ones!).
Probiotics are not to be confused with prebiotics, which are nutrients meant to encourage the growth of symbiotic bacteria particularly, or biochemical precursors to good things that symbiotes produce for us. Prebiotics include various starches and sugars.
The fields of prebiotics and probiotics have been notoriously sketchy. Sometimes its a quack trying to prove how his "special blend" of bacteria is good for you. Scientists, too, take shortcuts in this field, getting some uncharacterized mess of soil bacteria or food bacteria and trying to demonstrate benefits. Industry hasn't stayed out of the fray, and the natural/bioactive food people are always making unsubstantiated claims about their products. The 'real' dairy industry is happy enough to glom on - not that their products aren't healthful, but just that they haven't necessarily found ideal bacteria for any form of probiotic therapy. If those lactic acid bacteria do you any good at all, it is pretty much dumb luck.
So, recently, there have been some efforts to improve the science behind the probiotics. Some of the big problems: the intestinal flora is hard to sample, hard to observe, and difficult to characterize. Most of the relevant bacteria are not cultured, their properties are unknown, and they change in frequency and growth state rapidly as they move around in the intestines, not to mention in to the field and laboratory. As a result, things like fecal samples are possibly irrelevent, maybe even misleading, when studying the small intestine - and further, even if the samples were good, the means by which they are studied (culturing) hits a relatively small number of the strains, not necessarily the important ones.
Some of these things are changing, particularly with the advent of molecular techniques to characterize strains and communities. Not that these techniques don't have an amazing number of problems... just that they have fewer, potentially, than the culture techniques.
This particular study uses 8 patients. All were given clindamycin, four were given the probiotic in addition. Molecular methods and culture methods were both employed to track the populations in the patients. The molecular method chosen was T-RFLP.
This method basically uses the rRNA as the marker for populations and creates an easy to read fingerprint of the rRNAs to track the populations in a quantifiable fashion.
Samples were taken on three days, first: before antibiotic treatment second: during antibiotic treatment (7 days) third: after all the therapy was over (21 days)
The yogurt was consumed for the first 14 days (both during and after the antibiotic administration).
This is one of the weaknesses with the study... only three sample points per patient. It would have been very interesting to inspect the variance with a little more detail, say, two or three points during treatment, as well as two or three afterwards and two or three before. Also, the yogurt was given for a full week after the antibiotic treatment ended. It would have been very interesting to see a sample taken the day that concluded, rather than merely a week afterward.
Regardless, their statistical analysis appears very impressive, with principal componants and little three element dendrograms. What they are trying to demonstrate with all that is that the probiotic helped the community come back to normal, more or less, after the antibiotic administration ended.
Unfortunately, I am willing to presume that the populations changed - but I don't feel convinced of that by the data. Were I skeptical, all I'd have to say is that they haven't given any sense of the baseline variance in the populations. They would have needed day -14, -7, and 0 or something like that, to give a sense of what their principal componants would do without any perturbation at all.
"The results of the present study demonstrate that T-RFLP is
a useful tool for investigating the human intestinal microflora."
This is the striking result of all this sampling and data analysis. Come on, guys! Can't we at least hear about the probiotics? Well, none of the volunteers had diahrea... so there wasn't much to say about the relative rates of negative side effects from the clindamycin. And big changes in fecal flora due to clindamycin had already been published elsewhere, so that couldn't be their point. And they didn't take the appropriate samples to make a claim about the probiotic, even though they hint at such a claim in the Results section.
Thus, their lack of an impressive conclusion really may spring from the fact that the experiment was not well designed in the above respect. Their entire discussion hinges on the question of whether the molecular methods were better than the culture based ones (they were) and whether they were able to differentiate populations (they were) and whether the relative quantification is good enough to be worth the trouble, given that absolute numbers are obtained via culture methods (yes, relative is useful). So, it becomes, in the end, a methods paper with a hint of medical application.
Frustrating, but still nice to have in the literature. Unfortunate, though, because if they had done even 1 more sample (to get a baseline), they might have been able to demonstrate a statistically significant change in the population, etc.
Cecilia Jernberg, Åsa Sullivan, Charlotta Edlund, and Janet K. Jansson
Applied and Environmental Microbiology, January 2005, p. 501-506, Vol. 71, No. 1
The Big Picture:
Antibiotics are a great tool for treating bacterial infections, but they are something of an atomic bomb. Broad spectrum antibiotics kill all sorts of things, intended and unintended. Some of the unintended victims are happy symbiotic Lactobacillus and other lactic acid bacteria that love the pH 5 and rich nutrient soup of the small intestine. They help us digest, enrich our lives with vitamins, and so on. People who have taken antibiotics often get various forms of diahrrea, as well as other GI distress, because they have killed their nice symbiotes.
To address this, scientists have recommended investigating "probiotics." These are organisms that help out around the body. The simplest case are some nice lactic acid bacteria to enhance our normal flora, or even replace those wiped out by antibiotics. These lactic acid bacteria are found in sources like kefir, yogurt, sourdough, saurkraut, pickles, etc (only in unpasturized ones!).
Probiotics are not to be confused with prebiotics, which are nutrients meant to encourage the growth of symbiotic bacteria particularly, or biochemical precursors to good things that symbiotes produce for us. Prebiotics include various starches and sugars.
The fields of prebiotics and probiotics have been notoriously sketchy. Sometimes its a quack trying to prove how his "special blend" of bacteria is good for you. Scientists, too, take shortcuts in this field, getting some uncharacterized mess of soil bacteria or food bacteria and trying to demonstrate benefits. Industry hasn't stayed out of the fray, and the natural/bioactive food people are always making unsubstantiated claims about their products. The 'real' dairy industry is happy enough to glom on - not that their products aren't healthful, but just that they haven't necessarily found ideal bacteria for any form of probiotic therapy. If those lactic acid bacteria do you any good at all, it is pretty much dumb luck.
So, recently, there have been some efforts to improve the science behind the probiotics. Some of the big problems: the intestinal flora is hard to sample, hard to observe, and difficult to characterize. Most of the relevant bacteria are not cultured, their properties are unknown, and they change in frequency and growth state rapidly as they move around in the intestines, not to mention in to the field and laboratory. As a result, things like fecal samples are possibly irrelevent, maybe even misleading, when studying the small intestine - and further, even if the samples were good, the means by which they are studied (culturing) hits a relatively small number of the strains, not necessarily the important ones.
Some of these things are changing, particularly with the advent of molecular techniques to characterize strains and communities. Not that these techniques don't have an amazing number of problems... just that they have fewer, potentially, than the culture techniques.
This particular study uses 8 patients. All were given clindamycin, four were given the probiotic in addition. Molecular methods and culture methods were both employed to track the populations in the patients. The molecular method chosen was T-RFLP.
This method basically uses the rRNA as the marker for populations and creates an easy to read fingerprint of the rRNAs to track the populations in a quantifiable fashion.
Samples were taken on three days, first: before antibiotic treatment second: during antibiotic treatment (7 days) third: after all the therapy was over (21 days)
The yogurt was consumed for the first 14 days (both during and after the antibiotic administration).
This is one of the weaknesses with the study... only three sample points per patient. It would have been very interesting to inspect the variance with a little more detail, say, two or three points during treatment, as well as two or three afterwards and two or three before. Also, the yogurt was given for a full week after the antibiotic treatment ended. It would have been very interesting to see a sample taken the day that concluded, rather than merely a week afterward.
Regardless, their statistical analysis appears very impressive, with principal componants and little three element dendrograms. What they are trying to demonstrate with all that is that the probiotic helped the community come back to normal, more or less, after the antibiotic administration ended.
Unfortunately, I am willing to presume that the populations changed - but I don't feel convinced of that by the data. Were I skeptical, all I'd have to say is that they haven't given any sense of the baseline variance in the populations. They would have needed day -14, -7, and 0 or something like that, to give a sense of what their principal componants would do without any perturbation at all.
"The results of the present study demonstrate that T-RFLP is
a useful tool for investigating the human intestinal microflora."
This is the striking result of all this sampling and data analysis. Come on, guys! Can't we at least hear about the probiotics? Well, none of the volunteers had diahrea... so there wasn't much to say about the relative rates of negative side effects from the clindamycin. And big changes in fecal flora due to clindamycin had already been published elsewhere, so that couldn't be their point. And they didn't take the appropriate samples to make a claim about the probiotic, even though they hint at such a claim in the Results section.
Thus, their lack of an impressive conclusion really may spring from the fact that the experiment was not well designed in the above respect. Their entire discussion hinges on the question of whether the molecular methods were better than the culture based ones (they were) and whether they were able to differentiate populations (they were) and whether the relative quantification is good enough to be worth the trouble, given that absolute numbers are obtained via culture methods (yes, relative is useful). So, it becomes, in the end, a methods paper with a hint of medical application.
Frustrating, but still nice to have in the literature. Unfortunate, though, because if they had done even 1 more sample (to get a baseline), they might have been able to demonstrate a statistically significant change in the population, etc.
posted by BenK at 12:58 PM
0 Comments:
Post a Comment
<< Home