We are doomed…

not because of the risks of nanotechnology but because of a broken scientific system.

Last week, I had the privilege of opening, as the first invited speaker, a symposium on ‘Converging technology for nanobio applications’. This was my first slide:

Collage of various images. See links in the paragraph below for reference and credits.

Collage of various images. Credits: top left “Shutterstock”, top right “Cathy Wilcox”, bottom left “Jeremy M. Lange for The New York Times, A scientist at Duke University measures silver nanoparticles”, bottom center “Terminator 3: Rise of the Machines – a vision surely now only decades away. Photograph: Observer”, bottom right “image courtesy of Oregon State University”. See links in the paragraph below for original publication.

I started my talk by asking the audience what these images had in common (I did point out that the one in the top right corner was, in my experience, scientifically accurate).   The answer? These pictures had all been used to illustrate nanoparticle news in the previous week.

Exotic Nanomaterials Claimed Their First Major Workplace Injury said Stephen Leahy, writing for Motherboard on Tuesday about a worker injured by nanonickel while working without protection. The same day, Andrew Maynard, in Slate, published a more reasonable viewpoint on this same event. On Thursday, the Sydney Herald Tribune reported that a ‘Green group [had] called for ban of nano-materials in food’. This has been amplified in various outlets and, Andrew Maynard (again!) attempted to correct the record in the Conversation. On Friday, Deborah Blum, writing on the New York Times blog said that “Silver [is] Too Small to See, but Everywhere You Look”. The same day, Ben Baumont-Thomas informed us in  the Guardian blog that scientists had created bionic particles ‘inspired by Terminator’. Apparently, the latter piece was tongue in cheek, but it is rather difficult to differentiate the satire from the real thing these days (see for example Salon‘s coverage of the same story here).

While these stories are very different, they all originate in peer reviewed scientific research.

The “inspired by Terminator” piece originally comes from a press release by the University of Michigan. The authors of the article and their PR department were seeking this kind of publicity: the original article (very interesting BTW) uses the word “bionic” and the press release starts with “Inspired by fictional cyborgs like Terminator…”. Good to get coverage but it is highly debatable whether this kind of analogies really help improve the general understanding of science.

Deborah Blum article is measured and well researched – as we would expect from this award-winning science journalist – and based on multiple interviews with scientists. Yet in some ways, it also reflects the deep problems we are facing with establishing solid evidence in support of scientific understanding, and, eventually, policy making.

Deborah Blum article quotes Elisabeth Loboa as saying that “There’s evidence that the particles penetrate into plasma membranes, and they can disrupt cell function” [link in the original article, which is excellent practice!]. The idea that nanoparticles can go through the membrane of cells is so often repeated that it must be true, right? Scientists making those sorts of claim should provide a very high level of proof (unfortunately, this does not happen during peer review) because there are at least two fundamental reasons to be highly skeptical of such claims, one related to evolution, and another one related to physical chemistry and thermodynamics.

Nanoparticles are of similar size to viruses. If viruses could so easily penetrate cells, we would not be here discussing the risk of nanoparticles. Thankfully, during evolution, cells have developed very advanced mechanisms to protect themselves from foreign objects. Viruses too have developed very advanced mechanisms to get in there. Quite simply none of the nanomaterials made in the lab today seriously approach the level of sophistication that viruses use to get access to the interior of the cell (see this movie for an example). The linked article by AshaRani et al provides no evidence of particles penetrating through the plasma membrane (apart from the table of content cartoon). The dose used in this particular study is huge: 200 micrograms of nanomaterials per mL of medium (the equivalent of 10 grams of silver for a 50 kg person). In line with many other studies (including our own work), AshaRani et al show nanoparticles overwhelmingly in endosomes, i.e. in bags inside cells where they are isolated from the cell machinery. Endosomal trapping also remains a major limiting factor to siRNA delivery (even using nanoparticles).

Ignoring now the biology, at the simplest level, the membrane of cells is made of a bilayer of lipids. It has an hydrophobic interior and two hydrophilic surfaces. For an object to diffuse through the membrane, it would need to have no significant repulsive or attractive interactions with any of these components (otherwise it would be repelled and not go through, or attracted and then get stuck). It is hard to imagine any nanoparticle that would fulfill such criteria (see also post and comments here for more details). While it is unclear that any nanoparticle can diffuse through the membrane, many small molecules can. We therefore have this strange situation where the supposed capability of nanoparticles to go through the cell membrane is presented as a reason to be particularly worried even though this is unproven, unlikely for nanomaterials, and common for many smaller compounds (e.g. DAPI).

I am not blaming Deborah Blum nor  Elisabeth Loboa for this confusion. Such statements have become the norm. Although a more detailed investigation would be necessary (and I am not qualified to do it though I’d be happy to collaborate), my hypothesis is that the confusion results from a combination of nano hype (both in terms of risks and potential applications – see the Terminator for one striking example), bias towards the publication of positive findings, absence of post-publication peer review and debate, and poor scientific standards in an interdisciplinary area where editors and referees often lack some of the skills to properly evaluate the work (e.g. material scientists with very little understanding of biology, etc).

The situation is however now extremely serious since it has reached the point where it affects understanding of science for both basic scientists and the general public. It is our responsibility to try to fix the system.


  1. I remember nano-world being magic foofoo powder back in 1995. Doesn’t seem like much has changed and we are 15 years later.

    When does the thing become passé? These things go in cycles (look at the solar cells and fuel cells getting sexy every 20 years or so).

    I suggest to call the field, colloid science. The antithesis of hype. Make it humble.

    Carbon black is a nanomaterial. But it’s a common substance and no big deal and used in industry and blablabla. And even with it being a soup of stuff and made from soot, I wonder if the controls are better than these people with their little metal balls.



  2. Funny post, but unfortunately it is rather misinforming.

    Zwitterionic quantum dots and bigger mesoporous silica nanoparticles passively internalize in red blood cells, and gold nanoparticles that get internalized are not toxic if passivated so as not to release toxic ions.



    1. I have commented on the first DOI (Zwitterionic QDs) at Pubpeer

      I’ll comment on the other two when I have time. Regarding the third, I am impressed by this new euphemism for stripy nanoparticles:
      “gold nanoparticles [.] passivated so as not to release toxic ions”,
      I am even more impressed by the fact that an article based entirely on the existence of stripy nanoparticles does not have a single STM image to back up the existence of those stripes (for that specific composition, there is only one image of one nanoparticle that has ever been published to my knowledge) – 35 supporting information figures but not one STM…


      1. Regarding the zwitterionic QDs study (this has been posted in PubPeer):

        You can’t make any believable statement about a homogeneous distribution of isolated spots when there are only very few (3 per cell at most).

        It is highly implausible that water-soluble zwitterionic nanoparticles would aggregate in the cell. Please show evidence of zwitterionic aggregates in the literature to back up your statement.

        The spot outside of the cells in Fig. 1b (not 2b!) is much smaller than the brighter spots in the cells, and therefore should not affect the integrated fluorescence intensity.

        About the image stacks in Fig. 2 (not Fig. 3!): the xy stack is taken at a z value roughly at the middle of the cell (its interior); for the xz and yz cuts, the bulk of the bright area is inside the cell. This would be hard to explain if the particle were adsorbed to or stuck in the membrane.

        This seems to me a pretty solid study.


    2. The second DOI (10.1021/nn103077k) does not report passive diffusion of silica nanoparticles at all. Key relevant sentence from this abstract:

      “adsorption of large SBA-15-type MSNs (∼600 nm) to RBCs induced a
      strong local membrane deformation leading to spiculation of RBCs, internalization of the
      particles, and eventual hemolysis.”

      I may be tempted to write a blog post about the third paper… but in the mean time, you may want to provide other more convincing examples?


      1. There is strong evidence of passive internalization (or diffusion, yet not in the strict definition of the word as it is understood in fluid mechanics) in the paper (for example Fig. 5), despite your general dismissive claim.


  3. Reply to Author | June 12, 2014 at 9:50 pm
    also posted @PubPeer

    You are right about one thing: I got the figure number wrong! The reason is that the ACS figure viewer which allows to browse through the figures show the wrong figure numbers because it counts schemes as figures…

    Apart from that, according to the authors themselves:
    “Moreover, the fluorescent spots inside the cells (Figure 1b,d), which are much brighter than individual DPA-QDs, reveal their ability to migrate and cluster inside the RBCs.”


    1. The zwitterionic QDs “showed no aggregation over the physiologically relevant pH range of 5-9”.

      In any case, the evidence is that the particles have undoubtely diffused across the membrane.


      1. Author,
        That isn’t quite fair, you offered the original paper(10.1021/nn203892h, 2012) and are now referring to another publication (10.1021/nn900600w, 2009 with one shared author). The passage Raphael cites is from the more recent one, and you’ve resorted to quoting an older one. You don’t see the obvious point that needs to be reconciled, how is “aggregation” different from “clustering”? The two papers seem to be contradicting eachother!

        I would hope you would at least consider elaborating on this rather than just listing a citation.


  4. Author | June 12, 2014 at 8:51 pm:
    “There is strong evidence of passive internalization (or diffusion, yet not in the strict definition of the word as it is understood in fluid mechanics) in the paper (for example Fig. 5), despite your general dismissive claim.”

    Good: we agree that those particles are not going through cell membrane through molecular diffusion. They do not magically slip through the membrane because they are nano. The process here involves considerable interactions with the membranes, followed by bending, and, eventually membrane fusion. It is often disruptive. I don’t know of any paper which uses the word “diffusion” to describe such a phenomenon. The paper cited does not.

    There is indication of “passive internalization” for large particles from the EM but it is likely (though not discussed in the paper) that those particles are still surrounded by a piece of membrane. At least this would be in agreement with the mechanism proposed (a mechanism which, by the way, works for large particles over 100 nm and not for smaller particles because membrane bending becomes energetically too costly).

    Interestingly, the fluorescence images in Fig 7 show particles localized at the membrane. No particles are seen inside. This seems to somewhat contradict the EM or at least suggest that penetration is a rare event.


    1. It’s obviously not molecular diffusion; anyone who reads this forum would have understood this. In fact, I used the word internalization.

      Yes, the particles are likely to be wrapped by a piece of membrane. This is discussed in the paper (p. 1368).

      Note that in Fig 7 the concentration of particles is half that in Fig 5, and as the authors explain in the paper speculation strongly depends on particle concentration. Rather than internalization, Fig 7 was meant to show the effect of particle-membrane interactions by comparing various polymer-funcationlized particles.

      Would appreciate it if you could read carefully the paper before making claims of apparent contradiction.


  5. Raphaël Lévy on PubPeer (https://pubpeer.com/publications/A8ED352E38B377276C9CB818637F7C#fb10760):
    ( June 14th, 2014 12:47pm UTC )

    There would be lots of interesting things to discuss about zwitterions, protein aggregation, evolution (while most peptide sequences are prone to aggregation, evolution has selected against those; e.g. http://www.ncbi.nlm.nih.gov/pubmed/24472658, http://www.ncbi.nlm.nih.gov/pubmed/17668004, etc), but: 1) this is pretty far from the topic of this conversation; 2) it is simply a diversion by a commentator who is clearly not interesting in an honest scientific discussion.

    I appreciate that 2) is quite strongly worded but consider this:
    ** On June 12th, 2014 9:49pm UTC, Unreg2 argues that isolated spots are not an argument against a homogeneous distribution because those spots correspond to single particles and therefore the concept of an homogeneous distribution does not make sense.
    ** This is a legitimate argument, but it turns out, that the authors themselves make it clear that those spots correspond to many particles.
    ** If Unreg2 was arguing in good faith, he/she would either agree with me and the authors that the distribution is indeed inhomogeneous and that what he/she called highly implausible is in fact what is happening.
    ** Instead however, Unreg2 shifts goal posts, argues that those QDs don’t aggregate, that zwitterions in general don’t aggregate (pretty close to complete nonsense), while at the same time affirming that all of this does not matter any way since “in any case, the evidence is that the particles have undoubtely diffused across the membrane.”

    In addition Unreg2 is arrogant. “We would appreciate it if non-experts would refrain from discussing.” The best questions are often asked by non experts, but good discussion need good faith.

    We are wasting time. None of the points I have made in my first comment has been addressed in any way by Unreg2. I would very much welcome any interesting comments by others but won’t engage further with Unreg2.


  6. That is quite an amazing twisting of facts, Mr Levy.

    First, you haven’t shown me any evidence of the aggregation of zwitterionic particles in physiological conditions. In fact, the DPA-QDs the authors made are well described in the publication I referred to (10.1021/nn900600w), where they show no aggregation in physiological conditions.
    If the QDs cluster inside, it may be caused by protein-induced clustering after the QDs have diffused across the membrane.

    It’s unfortunate that Mr Levy accuses me of shifting goal posts and writing ‘complete nonsense’, when he has failed to address the criticisms I noted on his remarks about aggregation of zwitterionic particles, integrated fluorescence intensity, and the image stacks in Fig 2.

    After perusing Mr Levy’s blog site and his comments on PubPeer, I note that he has been engaging in various debates in the literature, and that some have considered his approach to be rather tasteless and his manners to be aggressive and harassing (http://mag.digitalpc.co.uk/Olive/ODE/physicsworld/LandingPage/LandingPage.aspx?href=UEhZU1dvZGUvMjAxNC8wNi8wMQ..&pageno=OA..&entity=QXIwMDgwMA..&view=ZW50aXR5)


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.