How do we know what causes an infectious disease? Part 2November 18, 2011
So in Part 1 I discussed the set of guidelines, or postulates, designed by Robert Koch (and his colleagues) that helps scientists get the evidence they need to establish that a specific infectious agent causes a specific infectious disease.
1) Establish association of the organism with the disease.
2) Isolate the organism and grow in pure culture.
3) Put the organism into a healthy host and show the host gets diseased.
4) Reisolate the organism from the now healthy host and show it is the same now as when you put it in.
I also discussed the main limitations of the guidelines:
Asymptomatic carriers – not every host infected will show signs of diseases.
Organisms that are difficult to culture – these include organisms for which we just don’t know the right culture conditions as well as viruses which need to be in a cell in order to multiply and so by definition cannot be grown in the “pure culture” that the postulates require.
Another limitation which I didn’t discuss in my previous post: the type of healthy host that you need to use for 3). Say chocolatitis is only found in humans and that Chocolobacter has no harmful effect on other species – what would I do then? It wouldn’t be ethical to use a human host for 3).
These limitations mean that sometimes Koch’s postulates are an inappropriate set of guidelines to try to use. If a microbe fulfills all of Koch’s postulates it is most likely the cause of the disease you’re looking at. If a microbe doesn’t fulfill all of them it might still be the cause but it might instead be there coincidentally.
But is there anything else scientists can do to provide more evidence for whether or not microbes that don’t fulfill the postulates are the cause of the disease?
Over the years since Koch first published his postulates scientists have adapted them in various ways to help them identify disease-causing microbes. Now that we have the ability to isolate, amplify and study nucleic acid sequences (those that make up DNA or RNA) a new set of guidelines has been proposed(1).
And so we return once more to chocolatitis, and the agent I suspect is causing it – Chocolobacter. It turns out that Chocolobacter is very difficult to isolate alive and culture but I have managed to sequence at least some of its genome. So although this means that I can’t easily use Koch’s postulates to help me establish Chocolobacter‘s guilt, I can use these ones (I’ve quoted the guidelines as they are written in the paper in bold and then explained it in the non-bold font):
1) “A nucleic acid sequence belonging to a putative pathogen should be present in most cases of an infectious disease.” To fulfil this I need to find my Chocolobacter genome sequence in the diseased host, preferably in the brain (which I’m assuming is the organ most severely affected in this infection) and I need to find it in most cases of the disease.
2) “Fewer, or no, copy numbers of pathogen-associated nucleic acid sequences should occur in hosts or tissues without disease.” So the Chocolobacter sequence must not be found (or be only rarely found) in healthy hosts and preferably, even in diseased hosts it shouldn’t be found in organs unaffected by the disease.
3) “With resolution of disease (for example with clinically effective treatment), the copy number of pathogen-associated nucleic acid sequences should decrease or become undetectable.” So if my diseased host gets better, I should no longer find any Chocolobacter sequences (or at least there should be fewer) in their brain. (I know my host has now apparently had several brain biopsies – this is totally fine in my imaginary world!) If my host then gets ill again the sequences should also return.
4) “When sequence detection predates disease, or sequence copy number correlates with severity of disease or pathology, the sequence-disease association is more likely to be a causal relationship.” If either I could detect the Chocolobacter sequences before the host started to show any symptoms, or if the host getting sicker corresponded to an increase in the number of sequences detected, this would strongly suggest that Chocolobacter was the cause.
5) “The nature of the microorganism inferred from the available sequence should be consistent with the known biological characteristics of that group of organisms.” What this is saying is that first, I need to identify other organisms that are related to Chocolobacter (I can use the genome sequence to help me with this). Let’s say it’s related to Spinachobacter (infection with this causes people to eat lots of spinach – I’m so imaginative…) and Coffeobacter (guess what this does…). To fulfil this guideline the behaviour of Chocolobacter and the characteristics of the disease it causes should be similar to its relatives Spinachobacter and Coffeobacter.
6) “Tissue-sequence correlates should be sought at the cellular level.” Ideally I should be able to make a nucleic acid sequence that will bind to the Chocolobacter sequence, and I should label my sequence with a fluorescent dye. Using tissue samples (for example on a slide) my sequence should bind to regions with the Chocolobacter sequence and should fluoresce. This can then be looked at for example by using a microscope. (This is a type of in situ hybridisation.) The areas of fluorescence should correspond to either visible Chocolobacters or to areas which I presume to be affected by Chocolobacter.
7) “These sequence-based forms of evidence for microbial causation should be reproducible.” I can’t just fulfil the above six guidelines once. I have to do it lots of times to be sure.
No doubt as the technology we have available improves the guidelines will once again be adapted and adjusted as we find new ways of detecting microbes but in a way this is irrelevant.
What really matters is that, regardless of the set of guidelines we as scientists use to prove the cause of a disease, the process should be logical. I can’t just conclude that Chocolobacter causes chocolatitis just because it happened to be present in a few chocolatitis cases and I’m strongly suspicious of it. But if I follow a logical set of steps, whether the original ones published by Koch or a newly designed set of my own, I should be able to generate enough evidence to prove whether or not Chocolobacter causes chocolatitis.
References/where to find more info
(1) Fredericks DN, & Relman DA (1996). Sequence-based identification of microbial pathogens: a reconsideration of Koch’s postulates. Clinical microbiology reviews, 9 (1), 18-33 PMID: 8665474
(2) Inglis, T. (2007). Principia aetiologica: taking causality beyond Koch’s postulates Journal of Medical Microbiology, 56 (11), 1419-1422 DOI: 10.1099/jmm.0.47179-0