The scientific genre of “animals detecting human diseases by smell” just got larger. Well, by about 1mm. And as long as you count nematode chemotaxis as odor perception.
A Japanese research team has reported that C. elegans can detect chemicals in the urine of patients with early stage pancreatic cancer. This is not the first time C. elegans has been used to detect disease, but it’s the first time for early stage pancreatic cancer.
How does one get nematodes to make this judgment? Well, take 50 to 100 of them (similar ages, please!), rinse them a few times, and squirt them onto the middle of a culture plate. At one end of the plate you squirt clinically-verified cancer patient urine or urine from healthy controls. Squirt some sodium azide in both ends of the plate; this compound immobilizes the worms that enter either end zone (score!) and makes it easier to count them 30 minutes later. Remember to keep the culture plate at 23 ± 1 ºC: these rascals are fussy.
If I’m reading the methods section correctly, these are not sensory discrimination tests, i.e., the nematodes are not choosing between patient urine and control urine. The test merely records how much the worms prefer a particular kind of urine over . . . nothing.
Here’s the actual calculation:
Chemotaxis index = (A − B)/(A + B), where A is the number of nematodes on the urine-spotted side of the plate and B is the number of nematodes on the opposite side.
In one test, done at x10 dilution, the median index value for healthy control urine was -0.015 and that from patients with pancreatic ductal adenocarcinoma was -0.038. Note that both indexes are negative. In other words, the nematodes avoided both types of urine; it’s just that they avoided the patients’ urine more than control urine.
How do you get to the rather small chemotaxis index of -0.015? Well, if you’re using 100 worms per trial, that means that across trials 50.75 worms avoided the cancer urine and 49.25 worms approached it. Not exactly a landslide result. Nevertheless, the authors say the difference is statistically significant at p = 0.034 (they don’t specify mention what statistical test they used). The authors also argue that clinical sensitivity of the nematode index compares favorably with assays based on tumor marker antigens.
If you’ve been following the ADHDBS genre, you know that there are two ways to play. There is the proof-of-principle route and the organism-as-sensor route. The former says “we’ve demonstrated (with dogs, say, or African giant pouched rats) the existence of a volatile diagnostic marker, and now we will develop a specialized sensor for use in a test kit.” The latter says “we intend to trains dogs (or giant rats) to perform the actual medical screening themselves.”
The absurdities of the organism-as-sensor route are obvious. (Oh, were you getting excited about becoming a worm-wrangler? Sorry!) What about the PoP route? Did the Japanese research team come up with any promising tell-tale molecules? After a bit of hand waving and citing of other studies they mention a couple. There is 4-methyl-2-heptanone (you know, the molecule found in the glandular secretions of certain leaf-cutting ants). And also d-limonene, an abundant terpene in some of the more popular sativa-style cannabis cultivars.
So, like . . . what if . . . we used nematodes . . . to tell us when to harvest the Durban Poison?
[That’s idiotic.—Ed.]
No, you’re idiotic.
[You’re a towel!—Ed.]
No, you’re a towel!
The study discussed here is “Scent test using Caenorhabditis elegans to screen for early-stage pancreatic cancer,” by Ayumu Asai, Masamitsu Konno, Miyuki Ozaki, Koichi Kawamoto, Ryota Chijimatsu, Nobuaki Kondo, Takaaki Hirotsu and Hideshi Ishii, published in Oncotarget 12:1687-1696, 2021.