A group led by scientists from Lawrence Berkeley National Laboratory has coupled gene-expression analysis with high-content cellular imaging to provide what is believed to be one of the most comprehensive examinations to date of the toxicity of semiconductor nanocrystals, also known quantum dots.
The research is the first step toward developing a standardized system that could associate different types of QDs or other nanomaterials with specific molecular biomarkers, toxicity responses, and imaging characteristics, and could provide guidance for further improving their toxicity.
The study also demonstrates that QDs coated with polyethylene glycol and silica have an extremely low cellular toxicity profile, and thus may be promising for long-term live-cell and in vivo imaging studies.
As a result, the study may have gained the attention of Invitrogen, which is interested in improving the toxicity profile of its own quantum dot reagents — which are based on core technology licensed a few years ago from a co-author of the LBNL study.
"People have done incubation with QDs, and tracked cells for up to 30 days. But our study is the first one to look at this at the molecular level, using a very comprehensive panel of all the genes in a cell to look at the perturbation caused by QDs."
Combining data from gene-expression studies and high-content imaging is not new. For instance, many researchers have already combined the methods to provide a more accurate assessment of small-molecule toxicity. However, this is the first time such an approach has been used to study the toxicity of QDs.
"People have done incubation with QDs, and tracked cells for up to 30 days," said Fanqing Chen, an LBNL scientist and co-author of the study. "But our study is the first one to look at this at the molecular level, using a very comprehensive panel of all the genes in a cell to look at the perturbation caused by QDs."
The group's findings were published in the April 12 issue of Nano Letters. Besides LBNL, the paper also featured contributions from researchers at the University of California, Berkeley; Lawrence Livermore National Laboratory; and Affymetrix.
One notable collaborator was Paul Alivisatos, a professor of chemistry and materials science at Berkeley and director of the materials sciences division at LBNL, and one of the earliest developers of QD technology. Alivisatos is also co-editor of Nano Letters.
As detailed in the paper, the researchers examined the impact of two dosages of PEG-silane-QDs on human lung and epithelial cells. They used a dosage previously reported to be non-toxic to human breast cells, and a 10-fold higher concentration.
After incubating the dosed cells for two days, they used a Cellomics KineticScan high-content imaging system and various labeling reagents to measure phenotypic responses such as cell proliferation, apoptosis, necrosis, and cell-cycle distribution.
"High-content screening is very advantageous here because the cost is pretty low," Chen told CBA News. "Drug companies are using it for drug screening, but we used it to try to establish all these morphological changes, cell-cycle distribution changes, indices for necrosis and apoptosis, and then within the cell, how the cytoskeleton proteins redistribute. All these can be detected with HCS very quickly."
The researchers then conducted gene expression-profiling experiments on the cells using an Affymetrix GeneChip system. The chip contained approximately 22,000 probes, of which 18,400 were known genes or probes sets.
"High-content screening is very advantageous here because the cost is pretty low. Drug companies are using it for drug screening, but we used it to try to establish all these morphological changes, cell-cycle distribution changes, indices for necrosis and apoptosis, and then within the cell, how the cytoskeleton proteins redistribute. All these can be detected with HCS very quickly."
The PEG-silane-QDs did not induce any statistically significant cell-cycle changes, and the researchers observed minimal apoptosis and necrosis in lung cells, regardless of the treatment dosage. They did, however, observe a slight increase in apoptosis and necrosis in the skin cells at both dosages.
In the gene expression-analysis portions of the research, only about 50 genes, or about 0.2 percent, had a greater-than-two-fold change in expression level. Upregulated genes involved in carbohydrate binding, intracellular vesicle formation, and cellular response to stress; while downregulated genes included those involved in controlling the M-phase progression of mitosis, spindle formation, and cytokinesis.
More importantly, the researchers noted, the QDs did not activate genes typically involved in a strong immune or inflammatory response, or related to heavy-metal toxicity. The latter point is crucial since QDs at their core are composed of extremely toxic heavy metals such as cadmium and selenide.
"One major finding here is that these QDs are very safe to use," Chen said. "We don't see a lot of perturbation of genes. There are less than 0.2 percent of the total genes being changed, even at a very high dose."
Invitrogen Skeptical, But Interested
Since they were invented a little over a decade ago, QDs have been heralded as the next great imaging reagent because of their small size, brightness, and easily tunable emission spectra.
However, QDs have suffered from two major limitations: highly finicky surface chemistry and the perception that their heavy-metal composition could be highly toxic to cells. Scientists have been able to tailor the surface chemistry to bind to molecules of interest but not bind to others, but considering the dots' potential as in vivo imaging agents, researchers have yet to convincingly demonstrate that they are not toxic.
Chen and colleagues may have killed two birds with one stone by developing a method for assessing QD toxicity, which they hope to eventually patent (see sidebar), and by showing a lack of toxicity in a particular type of QD.
The PEG-silane-QDs that were the subject of the Nano Letters study were developed in Alivisatos' lab. Alivisatos, along with MIT scientist Moungi Bawendi, also developed the core QD technology — that is, the CdSe/ZnS nanocrystal — which was licensed and commercialized by the firm Quantum Dot, now a division of Invitrogen.
The QDs commercially available through Invitrogen, known by the trademarked name Qdot, are similar to the PEG-silane-QDs, but are instead coated with a proprietary polymer developed by Quantum Dot.
There are also QDs with other surface modifications that are either commercially available from companies such as Evident Tech or being developed in academic labs. However, according to Chen, none have successfully solved the toxicity problem.
"The [PEG-silane-QDs] are probably the only ones that are non-poisonous," Chen said. "The coating around other QDs is thinner, and much leakier, and doesn't have all this cross-linking between the molecules, which is why cadmium is still coming out. You can do short-term studies — maybe a couple of hours — but longer than that, you run into secondary effects from the QDs. The goal of your imaging application is that you don't want to perturb the system with the thing you are introducing."
Chen said he and his colleagues have compared the PEG-silane-QDs with Invitrogen's Qdots and found that the toxicity profile of the silanized dots "is much better. We think this coating is demonstrated as having a minimal impact on the cell, and hopefully we can eventually use the dots for in vivo imaging studies."
Vicki Singer, director of the labeling and detection technology business segment at Invitrogen, told CBA News that the company hasn't conducted comprehensive toxicology studies on its Qdots, but is currently ramping up to do so. However, she said that there are many examples of researchers who have used the dots for long-term live cell or even animal imaging without toxic effects, and that the researchers' claim that existing QDs are toxic due to cadmium leakage must be taken with a grain of salt.
"The materials that we sell, which are polymer coated and also have PEG on the surface, also have never been shown to be toxic to cells," Singer said. "No one has ever shown that there is cadmium ion release from Qdot nanocrystals, [although] people have seen that with homebrewed ones."
The difference, she said, is that Quantum Dot successfully industrialized the process of making the CdSe cores, as well as the ZnS shells, and that the integrity of the shell will actually protect the cadmium — not necessarily the outer silanized or polymer coatings.
Singer also pointed out that a direct comparison of the technologies may be misleading, again due to the fact that Qdots are industrially manufactured. The QD's discussed in Nano Letters were synthesized in the Alivisatos lab, and although Singer concedes that the expertise in that lab is likely unparalleled in academia, all non-industrial QD preparations lack the kind of reproducibility of those manufactured in Quantum Dot's labs.
"Every single quantum dot is going to be different if it's from a different manufacturer," Singer said. "So every single batch that this lab makes is going to be different from every other batch they make. Unless you have an industrial process, you do not have reproducibility. It's difficult to make conclusions on toxicity or lack thereof based on materials that are not reproducible. That's one of the difficulties in the interpretation of these results."
Singer also said that Invitrogen would be interested in collaborating with Chen and colleagues to use similar methods to test the toxicity of Qdots and industrially produced silanized dots. If silanized QDs turn out to be less toxic, she said, Invitrogen would "absolutely" be interested in acquiring the technology.
"We're very excited about all the interesting coatings out there, and it could be that the silanized is better than the polymer coating we put on," she added. "We don't know the answer, and that's why we're going to do the studies to see if there is a problem with our coatings."
— Ben Butkus ([email protected])
On to the USPTO
In their study, LBNL researcher Fanqing Chen and colleagues believe they may have unearthed a method that can be used to create a reference of sorts for evaluating the toxicity of existing and new versions of QDs. Chen said that the group is filing a patent on the combined HCS-gene expression method for quantifying the toxic effects of QDs on cells and in living organisms.
"We are actually filing a patent for that, because we believe this could serve as a first step for establishing an evaluation matrix for nanoparticles," Chen said. "There are probably thousands of types of nanoparticles out there right now, but each has a different composition and different physical and chemical properties. So to evaluate them one by one is a very daunting task — almost impossible."
Chen said that a long-term goal of his group is to establish a standardized system, or matrix, of different nanoparticles and the biomarkers associated with them; or, certain imaging characteristics generated from high-content imaging studies.
"That way, this system can be applied to all nanoparticles, or a few key ones that can be used as a standard, and then you can use that to predict other ones," Chen said.
The next step, Chen said, is to conduct longer-term but similar toxicity studies on live cells in culture, and eventually, in live animals.
"We're also going to move to different organs or tissues, because this one was done on a lung cell line and skin cell line," he said. "We want to extend that to liver, kidney, brain, et cetera, and look at everything, and see what the difference is between them."