Better living through pee-sniffing, or What can urine tell us? Volume 2.

Previously, I told you about how rodents can avoid predators by detecting specific metabolites in carnivore urine, but today I'd like to tell you about some new research being done on human urine in an effort to diagnose certain diseases.

Last week I introduced the topic of metabolites, which are the by-products of the breakdown of things your body consumes, like food, drugs, and vitamins. These by-products are waste materials and usually leave the body via the urine or feces. The metabolites that are present in your urine depend on a lot of factors: not just what comes into your body but also what's happening inside your body. If you are currently hosting a pathogen, the metabolites present in your urine will likely change. This can be directly due to the pathogen itself producing the metabolites (hey, a bacterium has to eat and make waste too) or indirectly by the pathogen influencing metabolic activity (i.e., energy production pathways) in the cells of the infected tissues.

As it turns out, it may be possible to isolate specific urinary metabolite profiles common to people infected with certain pathogens. This could potentially be as simple as scanning the metabolites present in someone's urine and saying, "Okay, you have elevated levels of A and B and lower levels of C and D in your urine, so we have strong evidence that you're hosting pathogen Y." A research group at the International Center for Genetic Engineering and Biotechnology in New Delhi recently approached the question as to whether or not tuberculosis (TB), a disease caused by the bacterium Mycobacterium tuberculosis, can be detected in such a manner by identifying the volatile¹ organic compounds present in the urine of healthy and TB-infected individuals.

After screening the urine of healthy and infected individuals with mass spectrometry², the researchers came up with a few compounds that stood out as being different between the two groups. Two compounds (isopropyl acetate and o-xylene) increased at least two-fold in people infected with TB, whereas three other compounds (cymol, 2,6-dimethystyrene, and 3-pentanol) decreased by about half. To determine the predictive power of these results, the researchers generated a new pool of healthy and infected individuals, and they were able to accurately determine which individuals were healthy and which were TB-infected on the basis of these five metabolites.

It was beyond the scope of this study to determine the physiological origin of these five metabolites, but some of them are related to glycolysis and lipid metabolism, which are two different methods of providing cells with energy via the breakdown of carbohydrates or fat, respectively. It is also unknown if they are originating from the infected body tissues or directly from the tuberculosis bacteria. Clearly more research is needed on the physiology of these metabolites, but for now it seems that they are good potential biomarkers for diagnosing tuberculosis cases through urinary analysis.

But why bother trying to diagnose with urine at all? Traditional culture tests for TB involve taking a sputum (that's the mucus you cough up) sample and seeing if M. tuberculosis bacteria grow from it, but this procedure can take weeks. There's also the common skin prick test, which involves injecting a small amount of bacterial protein under the skin (there is no risk of infection because the proteins are not the actual bacterium, just something the bacterium produces). If a person has been exposed to TB, their immune system will recognize and attack the TB proteins, resulting in a hard, raised bump on the skin. This procedure is more invasive, subject to complications from existing conditions, and can still take 2-3 days for results. Urine samples have the advantage of being noninvasive (you're going to be peeing every 2-5 hours anyway), and the person doing the testing is never directly exposed to the bacterium, as can be the case with sputum samples.

While mass spectrometry was used to identify urine metabolites in this study, future point-of-care diagnostics may make use of electronic nose systems, which are currently being explored as a method of identifying volatile organic compounds and could potentially be developed as automated sensors for the specific volatile metabolites present in the urine of TB-infected individuals. This would streamline the procedure, make it more cost-effective, and provide a time advantage over other diagnostic methods currently being used. This would make a big difference in developing nations where skilled manpower, money, and resources are scant and where shortening the delay to diagnosis and treatment are key for reducing TB-related deaths.

¹ In chemistry, a volatile compound is one that readily evaporates into the air. I was being facetious when I mentioned pee-sniffing in the title, but volatile compounds do make it to the nose more quickly... I'm just saying...

² Would anyone be interested in a primer on how mass spectrometry works? It gets thrown around on TV procedurals a lot (CSI, Bones, etc.) and is depicted as a machine where you pump in a substance and it spits out all the molecules present in the sample. It ain't nearly that easy by a long shot.

Originally posted at Scientific American on August 8, 2011, where it was a Research Blogging Editor's Selection.

---

Banday, K., Pasikanti, K., Chan, E., Singla, R., Rao, K., Chauhan, V., & Nanda, R. (2011). Use of Urine Volatile Organic Compounds To Discriminate Tuberculosis Patients from Healthy Subjects Analytical Chemistry, 83 (14), 5526-5534 DOI: 10.1021/ac200265g