Skip to main content

Analytical Chemistry

Google’s Nanoparticle Diagnostic Ideas

Google’s “Google X” division, the part that works on odd high-risk high-reward projects, is apparently interested in diagnostic nanoparticles. That Wired article is pretty short on specifics, but the company’s Andrew Conrad revealed a few details in a talk yesterday. The idea, apparently, is to use magnetic-cored nanoparticles to interrogate various body functions, then to reconcentrate them in some superficial vein for a readout. I had thought initially that there would be a blood draw at that point, which seems like less of a leap, but apparently the idea is for some sort of across-the-skin readout.
That’s still not a crazy idea, although it has a ways to go (and an awful lot of work in animals). And I’m not sure how the noninvasive readout thing is supposed to work – I can imagine a lot more being done if you take some of these things back out. The nanoparticles could be tagged in various ways for sorting after removal, and (in theory) you could get quite a bit of information that way. The tricky part, either way, will be targeting the things that you can’t get from just circulating – otherwise you’d just take a blood sample as usual and add your nanoparticle brew to it ex vivo (which is not such a bad idea, either, and has been worked on by many others). Perhaps that’s why Google’s team is going the extra step.
So they’re presumably checking up on solid tissues somewhere, not the soluble blood factors, and that brings up a lot of pharmacokinetic issues. Many things that interact well enough to be diagnostic might well stick to the tissue instead of circulating back around, for example. And the ultimate fate of all these particles will be key – what effects will they have themselves, how well are they cleared, by what routes and at what rate, and so on. But I’ll reserve judgment until we know more about this. Google is saying that they’re not planning on developing this all the way themselves, but are trying to get other life sciences companies interested. How interested anyone gets might be a measure to watch.

37 comments on “Google’s Nanoparticle Diagnostic Ideas”

  1. Wondering says:

    Could the Google nanoparticles be small enough to pull through tissue with a capture magnet (like with immunoprecipitation beads) so that they can undergo “standard” biomarker analysis?

  2. steve says:

    Seems to me that the easiest way would simply be to do imaging – if enough particles accumulate where there’s a tumor you should be able to image them. Not sure that’s cool enough for Google though. One problem with the magnetic particles they’re thinking about is that clinicians like to use MRI but then they’d become nano-bullets tearing through tissues in response to the magnet.

  3. Anonymous says:

    I wonder how they’ll incorporate Google Ads (AdSense) into this.

  4. Anonymous says:

    What happens when billions of people start excreting things like non-degradable quantum dots down the toilet after diagnostic imaging? We already have problems with pharmaceuticals in our water that may be affecting marine life.

  5. lt says:

    My first thought was that they’ll cover the nanoparticles with some sort of a fancy fluorescent probe, so the optical signal will be quenched/enhanced by binding to their target. Taking frequent blood samples is unpleasant after all.

  6. Anon says:

    Yes, fluorescence would be good with particles at the surface. They would not really be pulled to the surface though, rather trapped as they flow past a magnet. Magnetic nanoparticles are used in MRI and do not fly around the body in the magnetic field.

  7. Wavefunction says:

    #3: Onto every nanoparticle will be imprinted a tiny nanoad, preferably one about Google’s new nanoparticles.

  8. Brian says:

    More here:
    It seems they think they will not have to test in animals…
    There are so many problems with this, I am looking forward to a good critique.
    But one meme we’ve been hearing lately from Google and others is the idea of defining the attributes of a healthy population so they can detect disease as outliers. My guess is they will find it more difficult to define health than to define various disease states.

  9. Prairie Boy says:

    One of the founders of Google Life Sciences, a google[x] division, is an NMR expert.

  10. SM says:

    Sounds like it could also be used like TumorPaint? For surgical tumor removal.

  11. bacillus says:

    I used to inject magnetic microparticles into mice. They were rapidly phagocyosed by various tissue macrophages in the liver and spleen. Thereafter, I only had to remove the appropriate organ, make single cell preps, and use a magnet to capture the tagged macs for additional analysis. I’m not sure whether nanoparticles would suffer the same fate, or be forced into e.g. the space of Diss in the liver before hitting any intended target.

  12. alf says:

    Can anyone tell me why they would go “public” at this stage?

  13. johnnyboy says:

    I find the basic concept and motivation (early detection, constant monitoring) to be admirable, and absolutely the way to go, especially for cancer. However having read that interview it’s pretty clear that they don’t really have a clue what goes on inside an organism that isn’t an in vitro machine model – and have no idea of the regulatory steps to market either. Conrad says it’s a matter of years, not decades, before this becomes available. I say bring on the flying cars.

  14. Anonymous says:

    In my (admittedly very limited, graduate student here) experience, they do seem to make a beeline for the liver and the kidneys. It’s a big issue, especially since they do this no matter how many targeting ligands you stick on them, so perhaps that’s one of the things for the Google team to figure out. Best of luck to them.
    That said, a colleague had an interesting experience with her batch: Whenever she injected her particles into healthy mice, they all headed out for the liver and kidneys, but when she did the same with tumor-bearing mice, their organs were clear and the tumor was chock-full of NPs. Even a concentration readout would be useful for a nanoparticle of this bent (though said colleague was less than pleased with the results, what with the toxicity issues).

  15. John Wayne says:

    @13 Is anybody else really disappointed that we don’t have flying cars? It’s *just* an engineering problem 😉

  16. Anonymous says:

    Perhaps these nanoparticles will reassemble in the brain to form the image of an advert?

  17. paranoid says:

    @3 and @7
    The idea is to have the particle outputs readable through the skin, no? Add a unique identifier and the readout could be designed to proximity trigger targeted advertising on a phone, tablet, billboard, etc. Also tracking or interrogation apps.
    “Do no evil” – just enable it instead.

  18. Hap says:

    @15: I’ve had people create their own lanes near me twice in two days; if someone does that with a flying car, very bad things will happen.
    I’m not sure why anyone who has driven would want anyone else [well other than me, of course ;)] to have a flying car. Maybe that’s why they haven’t happened the way people hoped.

  19. Sven says:

    I suppose we shouldn’t be surprised that they are trying to hype up their projects and be vague on the science.
    It seems that all along their plan is to provide proof-of-concept and then monetize on it by licensing to ‘partners’.
    @8: I agree, I doubt telling them ‘but the concept has been applied by others and it’s fine there’ will fly with the FDA.

  20. MikeC says:

    I think the main reason not to just take a blood sample is to avoid all of the I.P. covering diagnostics dependent on taking a sample.

  21. Wile E. Coyote, Genius says:

    They don’t need FDA approval. They can bypass that with the appropriate political campaign support $$$$. Truly is a “do no evil” type of company.

  22. molecular architect says:

    In the Backchannel article, Andrew Conrad is quoted as talking about detecting “molecules of sodium”. I have to assume he knows more chemistry that that but come on reporter and editors, that’s a high school level mistake.

  23. Cato the Elder says:

    @15 I remember at least two covers of Popular Science magazine titled: “your flying car is here!” Got me every time. 🙁

  24. Anonymous says:

    Orally ingested iron oxide nanoparticles will not be able to remain as particles for long as they are very easily digested in acid.

  25. johnnyboy says:

    @22: He is also quoted as discussing “oncologic cells”, which for a company emphasizing its cancer applications is not a good sign.
    @24: And what about those fancy protein markers they want to “decorate” the particles with ? How long will those last in the stomach ? And why would the particles be absorbed by the intestine anyway ?

  26. hypnos says:

    Why does this kind of research come from Google and not from one of the large pharma / diagnostics companies (who should in principle have more experience / expertise in this area)? Why do we not have a division for “high-risk high reward” projects in our companies. (And by that, I mean not “working on a target for which no phase II PoC is available from a competitor).

  27. Anonymous says:

    @26: Because pharma is full of the kind of nay-sayers that wrote comments 1 through 25. I.e., people who haven’t been able to innovate, so they spend their time knocking other ideas down and explaining why they can’t work.

  28. LiqC says:

    There’s an emerging imaging method called magnetic particle imaging (MPI), based on iron oxide nanoparticles but much more sensitive than MRI. Perhaps they’re betting on that?

  29. samadamsthedog says:

    @26 You have to be rolling in dough to take on extremely high-risk, high-payoff projects. Google is rolling in dough. Big pharma ain’t… regardless of management competence. (And besides, it take a different kind of competence to manage an effort that is rolling in dough than it takes to manage an effort that ain’t.)
    According to Conrad in his interview (, “preventative maintenance has been proven without a doubt to be the better model.”
    This conclusion is of limited ubiquity in the realm of medicine. For example, despite the ubiquity of mammography, breast cancer rates have held steady. A contributing cause seems to be that when cancers are detected early, many operations are performed that weren’t needed, because those cancers would never have developed to be life-threatening.

  30. cancer_man says:

    Google is going to leave pharmaceuticals in the dust. But note it is working with one…

  31. Wage_Slave says:

    Combine the responses from comment 2 “nano-bullets” and from 14 “tumor was chock-full of NPs” and you could just get a novel way to attack tumors mechanically. You would of course just get them to oscilate over a couple of cell widths rather than a big MRI type pull.
    Now that would be worth investing in.

  32. Observer 1 says:

    @29, Pharma has enough dough to be making multi-billion dollar stock buybacks, so I don’t think I can agree with your thought.

  33. Paul says:

    This idea is reminding me of recent news about the artificial spleen, which uses magnetic nanoparticles to remove pathogens from the blood.

  34. Hap says:

    This sounds obvious, but if you were talking about operating through the skin, wouldn’t glucose quantitation be an obvious target? When Exubera became the Exxon Valdez of inhaled insulins, some commenters complained that insulin administration was not nearly as onerous as the repeated needle sticks for glucose tests. Something that could eliminate those would probably make a big impact.

  35. scandium says:

    @25 I guess the idea is to inject the particles initially to avoid GI degradation.
    @16 Yes, the particles will also bind to certain brain regions during REM sleep to trigger dreams relating to Google Ads.

  36. anon says:

    @35: He does repeatedly mention in the interview that all that needs to be done is ingest a pill and then it will be in the serum. pharma does its all the time! I am just hoping he is bad at dumbing down his science, but he sounds very naive throughout the interview. not an encouraging sign.

  37. Anonymous says:

    @27: Indeed, pharma is full of people with actual scientific training and decades of experience in what works and what doesn’t – ie. realists, but what you would probably call ‘cynics’. Whereas tech is full of young dreamy-eyed earnest types who talk about ‘changing the world’ every other sentence, but whose main real-world achievements have been to make porn available free for all and putting games onto telephones. Just because you have big dreams and your pockets are full of free money doesn’t mean you can actually realize anything you dream up.

Comments are closed.