This post is a translation of an article originally published in French in Médecine/Sciences. The Editorial of the same issue (also in French) by Pierre Corvol is entitled Scientific integrity: the need for a systemic approach (open access). You can download a pdf of my article, or, read at the publisher’s website (subscription). For citation, please refer to the original article as follows:
Trois petites (nano) controverses et leurs morales; Raphaël Lévy; Med Sci (Paris), 33 8-9 (2017) 797-800; Publié en ligne : 18 septembre 2017; DOI: https://doi.org/10.1051/medsci/20173308027
« Selon que vous serez puissant ou misérable, les jugements de cour vous rendront blanc ou noir » [1] [Depending on your social height, The law will see your crime as black—or else as white.] Thus concludes the Fable, by Jean de La Fontaine, The Animals sick of the plague : the donkey, guilty of the theft of a few blades of grass, is condemned to death, whilst the Lion and other powerful animals guilty of much more serious crimes are treated to praise and flattery. It is tempting and comforting to think that scientific judgments are of an altogether different nature. Seen in this light, science would reside outside of power struggles and the few mishaps (mistakes, frauds, conflicts of interest) would be rapidly corrected since the reality of the material world would quickly come back to us through experimental results if we were to try to ignore it for too long. The truth is however very different. A large fraction of published scientific results cannot be reproduced. It is not a few mishaps but structural problems which affect the foundations of the scientific enterprise [2, 3]. Peer evaluations seems to encourage the publication of extraordinary stories in high impact factor journals rather than careful and rigorous experimental studies. Contradictory or “negative” data are rarely published: scientific journals are not really interested, and us, scientists, are not particularly motivated by publicly stating our doubts on the work of colleagues who could be in charge of evaluating our next article or grant application. It is particularly urgent to repair our knowledge production system because science is at the center of numerous challenges critical for the future of human beings and the planet. The (real) problems of reproducibility have already been harnessed by lobbies to attack the credibility of scientists [4]. After the election for president of the largest scientific and military power of the world of a man who denies climate change, is very positive about the use of the atomic bomb, and, more broadly wages an open war against science and truth [5], we have a paramount need for science to be open, robust, capable of defending its independence, integrity and universal values. This seems a distant prospect.
The near absence of critical discussion in the scientific literature in many areas of science could make us forget that controversies are an essential aspect of the quest for knowledge, allowing to identify weak points of experiments and theories, thus enabling to consolidate or invalidate them [6]. They are consubstantial to the scientific practice [7]. The analysis of controversies is also a tool to “symmetrically map” the actors to better understand the roles of individuals and social processes [8]. In this piece, I describe three recent controversies in my area of research: gold nanoparticles applied to biology and medicine. This is no “symmetric map”: I am not a neutral observer but a scientist active, to various degrees, in each of those. I am trying nevertheless to draw some lessons and suggestions to improve the ways we work as scientists.
Stripy Nanoparticles
In 2004, Francesco Stellacci’s group at the Massachusetts Institute of Technology (MIT) published in the prestigious journal Nature Materials an article describing gold nanoparticles covered by a mixture of two molecules that self-assemble to form stripes that are observed by scanning probe microscopy [9]. This article and the numerous other ones that will follow in the same journal and in others just as prestigious such as the Journal of the American Chemical Society [10], Science [11] and Proceedings of the National Academy of Sciences (PNAS) [12], suggest that, thanks to their stripes, these nanoparticles have unique properties in terms of wetting, self-organization, interaction with proteins, penetration in cells, with lots of potential applications for biomolecular sensing, or even drug delivery. These articles certainly contributes to the progress of their authors’ careers, but the stripes are an experimental artefact well known by users of scanning probe microscopy. How to explain then that more than 20 “stripy” articles were published between 2004 and 2012? It is obvious that specialists (and even enlightened amateurs) had identified the problem as early as 2004. Yet, the articles and reviews of that period show no sign of it. One now knows that Predrag Djuranovic has been the first to engage into a scientific investigation aiming at testing, and eventually, contesting, the evidence for the the existence of the stripes. In 2005, this rigorous and brave scientist was a student in Francesco Stellacci’s lab. His experimental results and numerical simulation showing how the stripes originate from a poorly adjusted feedback control system were unambiguous but MIT ensured that these results would remain secret [13]. In 2007, I submitted a technical comment responding to the Science article. This first attempt, limited in its scope to the Science article itself, was unsuccessful: Science did ask Francesco Stellacci to respond but then decided not to publish the exchange of views [14]. In 2008, a new article from the MIT group, again in Nature Materials, report that, thanks to their stripes, these nanoparticles can cross the cell membrane and directly access the cytosol [15]. This is accompanied by a commentary entitled “Particles slip cell security” [16]. After discussions with several of my students, we decide to propose a more exhaustive answer. A few months later, the article “Stripy Nanoparticles Revisited” is ready. It includes a new analysis of the stripy images concluding that the stripes are a scanning artefact as well as a critical discussion of the physicochemical and biological properties which, together with experimental results, contradict the claim of direct access to the interior of cells. The article is first submitted to Nature Materials (rejected), then NanoLetters (rejected), and, finally, Small… where it is published after an editorial process that lasted three years [17-19]. The publication of our article, in November 2012, does not end the controversy. Instead, it expands in the scientific literature (a little) and it also takes new forms (in particular on my blog and others [20-23]). Problems with the reuse of images in different publications emerge and eventually lead to two corrections ([12] and [15]). After a number of requests, Philip Moriarty and Julian Stirling (School of Physics and Astronomy, university of Nottingham, UK) are given access to the original data of the 2004 article. They demonstrate, among other things, that the stripes are present in the entire image, i.e. even between the gold nanoparticles [24], a conclusion still rejected by Francesco Stellacci [25].
Homeopathic nanoparticles
The laboratory of Molly Stevens at Imperial College is one of the most prestigious in the field of biomaterials. In 2012, two articles from the group relate the particularly interesting properties of nanoparticles for diagnostic applications. The first one, published in Nature Materials, reports a phenomenon which is entirely extraordinary in which the signal detected increases when the concentration of molecules to detect decreases (“inverse sensitivity”) [26]. Even more incredible, this phenomenon extends to the point where there is less than a molecule of enzyme, on average, in the volume under study. The second article published in Nature Nanotechnology goes further : no need for instruments, the detection of concentration of analytes in the same range is achieved thanks to a colour change visible with the naked eye [27]. Detailed critiques of these articles are available on the website PubPeer [28, 29] as well as in a preprint authored by Boris Barbour; the objections are both simple and profound but the authors have chosen not to respond. One can note that the Avogadro number includes lots of zeros (630 000 000 000 000 000 000 000) and that the detection of a macroscopic change of property due to the presence of a single molecule is therefore an achievement that requires extremely solid proofs. One of the posts on PubPeer indicates that someone contacted the Editor of Nature Nanotechnology in January 2013, but, four years later, no doubts are expressed on the journal website, in the traditional scientific literature nor in the newspapers that had covered this story (e.g. Le Monde and the Daily Mail) when it was initially published [30, 31].
Spherical Nucleic Acids
The laboratory of Chad Mirkin at Northwestern University (USA) is one of the most prestigious in the field of nanosciences applied to biology and medicine. One major theme of their research are the Spherical Nucleic Acids (SNAs), a term introduced by Mirkin to describe gold nanoparticles functionalised with DNA (or RNA) strands. These SNAs are supposed to have properties very different from linear DNA [32]. In particular, they can access the cytosol of live cells, where they could detect and regulate, the presence and quantity of mRNAs. One could ask why this solution did not appear during evolution : to access the cell machinery, viruses and bacteria would have only needed to package themselves within their genetic materials. The first articles (in Science [33], the Journal of the Americal Chemical Society [34], NanoLetters [35], ACS Nano [36]) proposing this surprising theory do no mention the mechanism of the SNAs into cells whatsover. The following one, e.g. [37], propose that the particles enter by endocytosis, but do not explain the mechanism by which the SNAs would escape endosomes. After several dozens of articles on this topic, the proportion of particles reaching the cytosol is still to be measured and reported (in spite of the fact that gold nanoparticles have been used since the 1950s to study intracellular trafficking; such a study would not be difficult). One article from the Mirkin group suggests that SNAs are degraded in the endosomes and that a “small unquantifiable portion escapes […]” [38]. Nevertheless, the particles are now commercially available under the name SmartFlares (Merck Millipore) to detect RNA inside cells. We have studied the entry of nanoparticles in cells and their ability to detect RNA. Given our difficult experience with the publication of Stripy Nanoparticle Revisited, we decided to adopt a different strategy. The project has been open and we have shared our results in quasi real time on our blog. In contradiction with the descriptions made by Mirkin and by Merck Millipore, we have observed that the SmartFlares were degraded in endosomes and were not able to detect mRNA. Mirroring the tale of Predrag Djuranovic and the stripy nanoparticles, we were not the first to have doubts about the technology: Luke Armstrong, who had been in charge of developing the SmartFlares at Merck Millipore in California (before leaving the company) had reached the same conclusion [39]. To ensure speedy publication and transparency, we published our article on the (not so prestigious) ScienceOpen platform where peer review occurs after publication [40]. We invited comments by Mirkin to no avail. Another article by the same group in PNAS describe a new version of the SmartFlares [41]. Our analysis of the raw data (obtained after multiple insistent requests) show that the signal comes from endosomes. Our letter submitted to PNAS has been rejected by the editorial board because it “[did] not contribute significantly to the discussion of this paper” [42].
Morals
Access to raw data is essential and guaranteed by clear rules adopted by Universities, scientific journals and funding agencies. It is therefore generally possible to access data with some efforts. It is obviously preferable to publish data at the same time as the articles. This is already the norm for some categories of results and it should become generalised. Researchers should also adopt the Manifesto for reproducible research [43]. The tools are in place to improve the practice of science.
Evaluations of science and scientists must imperatively be based on a critical analysis of their work and the robustness or their results, not on the prestige of the institutions or journals. This requires a change of mind and a clear commitment from researchers who are in positions of power, i.e. everyone who features on promotion or recruitment committees. To say that an article is good because it has been published in a prestigious journal is a moral and logical error which needs to be challenged.
Institutions and scientific journals are not motivated by the quest for scientific truth. The decisisons taken by MIT (keeping Predrag Djuranovic’s findings secret), by Nature Materials (not publishing the exchange with Francesco Stellacci [14]), and by PNAS (not publishing [42]) have directly impacted progress of knowledge. These institutions have commendable principles but, in practice, they aim first at defending their reputation and finances [44]. The latter objective only partially aligns with scientific progress which requires rapid and open discussion of results and conclusions. The Worldwide Web, invented for the sharing of science, enables this discussion. Researchers therefore should embrace the following tools: 1) Pubpeer to comment on articles; 2) Preprints to publish rapidly, minimise the influence of editors, and, dissociate publication, i.e. sharing of information, from evaluation, i.e. peer review; 3) Social networks, e.g. Twitter and blogs, which constitute an ongoing scientific conference to discuss experiments, results, methods, analyses, and new publications.
Acknowledgements: I thank Marianne Noel (IFRIS) for her critical reading of this piece, and, Marianne Lévy for comments on the grammar and style [French version] very necessary after 14 years in an English-speaking country…
Conflicts of interest: The author declares that he has no conflict of interest related to this article.
References:
- La Fontaine J (de). Les animaux malades de la peste. Fables, 1678. [full text]
- Ioannidis JPA, Boyack K, Klavans R, et al. How to make more published research true. PLoS Med 2014 ; 11 : e1001747.
- Baker M, Dolgin E. Cancer reproducibility project releases first results. Nature 2017; 541 : 269-270.
- Laframboise D. How many scientific papers just aren’t true? The Spectator 2016.
- Krauss LM. Donald Trump’s war on science. New Yorker 2016.
- Dascal M. The study of controversies and the theory and history of science. Sci Context 1998 ; 11 : 147.
- Latour B. Pasteur et Pouchet : hétérogenèse de l’histoire des sciences (sous la direction de Michel Serres). Éléments d’histoire des sciences 1989 : 423-45.
- Pestre D. L’analyse de controverses dans l’étude des sciences depuis trente ans. Mil neuf cent. Rev d’histoire Intellect 2007 ; 25 : 29-43.
- Jackson AM, Myerson JW, Stellacci F. Spontaneous assembly of subnanometre-ordered domains in the ligand shell of monolayer-protected nanoparticles. Nat Mater 2004 ; 3 : 330-6.
- Jackson AM, Ying Hu Y, Silva PJ, Stellacci F. From homoligand- to mixed-ligand- monolayer-protected metal nanoparticles: a scanning tunneling microscopy investigation. J Am Chem Soc 2006 ; 128 : 11135-47.
- DeVries GA, Brunnbauer M, Hu Y, et al. Divalent metal nanoparticles. Science 2007 ; 315.
- Centrone A, Penzo E, Sharma M, et al. The role of nanostructure in the wetting behavior of mixed monolayer-protected metal nanoparticles. Proc Natl Acad Sci USA 2008 ; 105 : 9886-91.
- Djuranovic P. Seven years of imaging artifacts: what gives? Rapha-Z-Lab 2012. https://raphazlab.wordpress.com/2012/12/11/seven-years-of-imaging-artifacts/
- Levy R. Divalent metal nanoparticles. PubPeer. https://pubpeer.com/publications/4DA88768C5B279E24E469CC0080A47
- Verma A, Uzun O, Hu Y, et al. Surface-structure-regulated cell-membrane penetration by monolayer-protected nanoparticles. Nat Mater 2008 ; 7 : 588-95.
- Xia T, Rome L, Nel A. Nanobiology: particles slip cell security. Nat Mater 2008 ; 7 : 519-20.
- Cesbron Y, Shaw CP, Birchall JP, et al. Stripy nanoparticles revisited. Small 2012 ; 8 : 3714-9.
- Yu M, Stellacci F. Response to “Stripy nanoparticles revisited”. Small 2012 ; 8 : 3720-6.
- Levy R. Stripy timeline. 2012. Rapha-z-lab https://raphazlab.wordpress.com/2012/12/20/stripytimeline/
- Dove A. Do these stripes make my nanoparticles look weird? 2012. http://alandove.com/static/2012/12/do-these-stripes-make-my-nanoparticles-look-weird/
- Fernig DG. Ferniglab Blog. https://ferniglab.wordpress.com/?s=stripy
- Neuroskeptic. Science is interpretation. Discov Mag Blogs 2014. http://blogs.discovermagazine.com/neuroskeptic/2014/01/04/reanalysis-science/
- Natelson D. A nano-controversy. Nanoscale views 2012. http://nanoscale.blogspot.co.uk/2012/12/a-nano-controversy.html
- Stirling J, Lekkas I, Sweetman A, et al. Critical assessment of the evidence for striped
nanoparticles. PLoS One 2014 ; 9 : e108482.
- Ong QK, Stellacci F, Jeschke G, et al. Response to “Critical Assessment of the Evidence for Striped Nanoparticles.” PLoS One 2015 ; 10 : e0135594.
- Rodríguez-Lorenzo L, la Rica R de, Álvarez-Puebla RA, et al. Plasmonic nanosensors with inverse sensitivity by means of enzyme-guided crystal growth. Nat Mater 2012 ; 11 : 604-7.
- la Rica R de, Stevens MM. Plasmonic ELISA for the ultrasensitive detection of disease biomarkers with the naked eye. Nat Nanotechnol 2012 ; 7 : 821-4.
- PubPeer “Plasmonic nanosensors with inverse sensitivity by means of enzyme-guided crystal growth”. https://pubpeer.com/publications/3E8208F0654769A44C22D4E78DA2B8
- PubPeer “Plasmonic ELISA for the ultrasensitive detection of disease
biomarkers with the naked eye”. https://pubpeer.com/publications/54AECF24E96162E3A563AED08BE0B3
- Des nanoparticules d’or pour dépister VIH ou cancer à l’œil nu. Le Monde
2012.
- Bates C. Colour-coded blood test that turns blue if you have HIV is 10 times
more sensitive than current methods. The Daily Mail 2012.
- Cutler JI, Auyeung E, Mirkin CA. Spherical nucleic acids. J Am Chem Soc 2012 ;
134 : 1376-91.
- Rosi NL, Giljohann DA, Thaxton CS, et al. Oligonucleotide-modified gold
nanoparticles for intracellular gene regulation. Science 2006 ; 312 :
1027-30.
- Seferos DS, Giljohann DA, Hill HD, et al. Nano-flares: probes for transfection
and mRNA detection in living cells. J Am Chem Soc 2007 ; 129 : 15477-9.
- Zheng D, Seferos DS, Giljohann DA, et al. Aptamer nano-flares for molecular
detection in living cells. Nano Lett 2009 ; 9 : 3258-61.
- Prigodich AE, Seferos DS, Massich MD, et al. Nano-flares for mRNA
Regulation and Detection. ACS Nano 2009 ; 3 : 2147-52.
- Choi CHJ, Hao L, Narayan SP, et al. Mechanism for the endocytosis of
spherical nucleic acid nanoparticle conjugates. Proc Natl Acad Sci USA
2013 ; 110 : 7625-30.
- Wu XA, Choi CHJ, Zhang C, et al. Intracellular fate of spherical nucleic acid
nanoparticle conjugates. J Am Chem Soc 2014 ; 136 : 7726-33.
- Schneider L. Do nanoparticles deliver? Merck’s Smart Flares and other
controversies, 2015. https://forbetterscience.com/2015/11/20/do-nanoparticles-deliver-mercks-smart-flares-and-other-controversies/
- Mason D, Carolan G, Held M, et al. The spherical nucleic acids mRNA
detection paradox. ScienceOpen Res 2016 ; DOI: 10.14293/S2199-
1006.1.SOR-CHEM.AZ1MJU.v1.
- Briley WE, Bondy MH, Randeria PS, et al. Quantification and real-time
tracking of RNA in live cells using Sticky-flares. Proc Natl Acad Sci USA
2015 ; 112 : 9591-5.
- Mason D, Levy R. Sticky-flares: real-time tracking of mRNAs…
or of endosomes? bioRxiv 2015. http://biorxiv.org/content/early/2015/10/19/029447
- Munafò MR, Nosek BA, Bishop DVM, et al. A manifesto for reproducible
science. Nat Hum Behav 2017 ; 1 : 21.
- Glanz J, Armendariz A. Years of ethics charges, but star cancer researcher
gets a pass. New York Times 2017
” 
Rapha-z-lab 