December 2008
Feature Story

William Check, PhD

It’s remarkable how many words beginning with the letter “q” have to do with seeking answers or looking for information. Think of “question,” “query,” “quiz,” “quest,” and “quandary.” To this list we can now add “quantum dots,” or Qdots, a form of nanotechnology that many people believe holds great promise for providing answers in the world of biomedicine, particularly in pathology.

“From the perspective of the pathologist, nano­technology can be applied in many ways,” says Andrew N. Young, MD, PhD, clinical laboratory director in the Department of Pathology and Laboratory Medicine at Grady Health System/Emory University. “Nanotechnology allows for higher multiplexing, more rapid assays, and smaller assays than we now use,” Dr. Young says. Historically, pathologists began with morphology supplemented with immunohistochemistry, or IHC, using one or two gene products, he says. “We have now moved to larger panels to help with diagnostic decisions and are being pressed to provide more information about prognosis and likely response to specific molecular agents. As a result, in such conditions as breast cancer we are running into the limitations of pure morphology plus information from a handful of gene products by IHC or other current methods of multiplexing.”

Qdots linked to antibodies and applied to tissue sections can help overcome these limitations. Because Qdots can be designed to produce many discrete wavelengths of fluorescent light with minimal spectral overlap, they can achieve a high order of multiplexing while preserving an ability to visualize cell structure and tumor architecture, which is important for heterogeneous tumor tissue. This gives Qdots an advantage over microarrays, RT-PCR, or commercial assays like Oncotype DX, which, Dr. Young says, “sample a large number of gene products but are disruptive to tissue.”

Qdots also address the size squeeze—tissue biopsies are getting smaller while highly multiplexed analysis becomes more imperative and cell morphology remains valuable. “That’s where Qdots may provide enabling technology,” Dr. Young says. “They allow us to evaluate many biomarkers in a single pass.” He is finding one “very exciting” application in subtyping renal tumors.

Lawrence True, MD, professor of pathology and director of urologic pathology at the University of Washington School of Medicine, calls Qdots “a novel promising technology.” And he describes himself as “optimistic” about the quantifiability that measurement of Qdots’ fluorescence could bring to the analysis of biomarkers in tissue. “Standard cancer stains are not really quantitative,” he says, “unless we run many controls.”

Dr. True’s collaborator at the University of Washington, Xiaohu Gao, PhD, is developing Qdot technology for another promising use—traceable drug delivery and targeted therapy. Dr. Gao, an assistant professor of bioengineering, has demonstrated the ability of Qdots to deliver si­RNA into cells in culture to silence designated genes (Qi L, et al. ACS Nano. 2008;5:263–267).

Shuming Nie, PhD, distinguished chair and professor of biomedical engineering, chemistry, and hematology/oncology, and director of the Emory/Georgia Institute of Technology Nanotechnology Center for Personalized and Predictive Oncology, calls Qdots “a radical departure from current RT-PCR or microarray methods.” With Qdots, Dr. Nie says, “we can do quantitative analysis of tissue sections at the cellular-molecular level.” He predicts that the first biomedical applications of nanotechnology will be in pathology. “When we are working in vitro or ex vivo,” Dr. Nie says, “we don’t have to worry about toxicity.” Along with Dr. Young, Dr. Nie is putting a large effort into pathology applications.

One laboratory application of Qdots is in multicolor flow cytometry, which requires analysis of many cells expressing many protein combinations using multiple antibodies. Pratip K. Chattopadhyay, PhD, a staff scientist in the ImmunoTechnology Section of the Vaccine Research Center, National Institute of Allergy and Infectious Diseases, notes that in their work the amount of sample is often limited and antibodies need to be multiplexed. “So we are interested in techniques where we have lots of colors available and can multiplex easily. That’s where Qdots became particularly exciting to our group.” Dr. Chattopadhyay says they “hit the wall” with organic fluorochromes, which allow a maximum of 12 colors because of spectral overlap. “With Qdots we increased the number of fluorescent signals that we can measure simultaneously up to 17 or 18,” he says (Chattopadhyay PK, et al. Nat Med. 2006;12:972–977). “Qdots have fairly narrow emission spectra, so they are amenable to multiplexing.” He has no doubt that this advantage will translate into enhanced laboratory immunophenotyping. “If you have a new marker that you want to incorporate into panels and you have run out of traditional fluorochromes, Qdots give you more options,” he says.

Thomas Grogan, MD, professor of pathology at the University of Arizona, delineates his path to appreciating the virtues of Qdots. “We are going from an era in which we have focused on diagnosis to an era in which we do tests that predict patient response to pharmacologic treatment by finding targets of therapy,” says Dr. Grogan, chief scientific advisor and founder, Ventana Medical Systems Inc. “This will require a matrix of results as opposed to single molecules.” Where he sees the future: “multiplexing in cellular context, with a focus on quantitating one pathway relative to another.” That complex goal implies multiple simultaneous measurements.

“That’s what brought us to Qdots,” Dr. Grogan says. “One Qdot’s emission doesn’t interfere with another. And Qdots don’t have fading signals.” Both of these properties facilitate comparing pathways, which demands the ability to quantitate two pathways simultaneously.

Moreover, Qdots can be attached to probes directed to genes, RNA messages, or proteins, Dr. Grogan says. That makes it possible to use Qdots to look simultaneously at all three levels of gene function, to ask whether message is expressed and whether protein follows expression of message and to see derangements in that process. “It is the pathologist’s domain to put biochemical phenotypes into cellular-genetic context,” Dr. Grogan says. He sees this as extending the traditional practice of pathology to deriving an understanding of subcellular chemistry. “We all do chemistry on tissue biopsies one molecule at a time,” Dr. Grogan says. “Imagine being able to do eight or 10 or 12 analytes in a tissue section at once.”

Dr. Grogan looks even further into what Qdots can enable pathologists to do. “Once I can see 10 things per slide in a needle biopsy of a tumor from a treated patient, a biopsy with only a few hundred cells, I can learn what is going on in that tumor. That puts me in a position to monitor therapy. We hear this common complaint that the pathologist is asked to say more and more about less and less,” Dr. Grogan says. “Here is a chance to say more and more about even less.”

Since Qdot analysis is done on tissue, not homogenates, that brings it into the realm of the anatomic pathologist, Dr. Grogan asserts. “Molecular techniques are complex and difficult to perform. Once we can do them with reliable platforms and protocols, pathologists will know best what the data mean in the context of a biopsy. The pathologist is master of the biopsy.” Dr. Grogan suggests that the anatomic pathology laboratory of the future will grow by essentially asking new questions and providing answers that are new to medicine. “As the anatomic pathologist answers questions not only about diagnosis but about optimal therapy and how the patient will respond, you can argue that testing will be 10 times greater,” he says.

Before Qdots find their destined place in pathology practice, however, researchers will have to seek answers to basic questions about this new methodology. Dr. Nie says that nanoparticles, including Qdots, are easy to make and relevant antibodies exist. Yet several steps remain to be optimized, including conjugation of Qdots to probes, staining procedures, and signal analysis. “Optimizing and automating the whole technology will be the key,” Dr. Nie says.

And simple questions about Qdots as reagents still need to be addressed. “One of our concerns is stability,” Dr. Chattopadhyay says. “Can we keep conjugated Qdots long enough to use them across multiple samples? Right now we haven’t used them long enough to know how stable they are in storage.” Qdots also need to be made reliably lot-to-lot. “They are so new that these things haven’t been tested extensively,” Dr. Chattopadhyay says.

Qdots are tiny particles of semiconductor material, such as cadmium selenide, that have quantum properties by virtue of their small size—less than 10 nm in diameter. In a Qdot, changes in the energy levels of electrons excited by incident light occur in discrete quantum steps. Further, the quantum or magnitude of change in the energy level of an excited electron depends on the diameter of the particle, the quantum confinement effect. When an electron relaxes after excitation, it collapses back to its ground state, releasing a photon. The energy of this photon, and thus the color of the emitted light, depends on the magnitude of the initial quantum energy change in the excited electron. Putting all of these factors together, we see that the color of the fluorescence emitted by a Qdot can be “tuned” by changing the size of the particle. A Qdot of 2 nm diameter emits short wavelength blue light, whereas one of 6 to 7 nm emits longer wavelength red light.

Dr. Nie, one of the first researchers to publish on the use of nanoparticles for biomedical application, demonstrated in 1998 several of their advantageous properties (Chan WC, Nie S. Science. 1998;281:2016–2018). One essential set of characteristics is that Qdots can be conjugated with biomarkers and that these conjugates are water-soluble and biocompatible.

Absorption and emission of light by Qdots has several desirable properties. Organic fluorophores have to be excited by multiple light sources, while a variety of Qdots can be excited by one light source and one filter. “The color is defined not by the input stimulus but by the size of the Qdot itself,” Dr. Gao says. Also, light emitted by organic fluorophores has broad and asymmetric peaks, allowing only low resolution, so only a few organic fluorophores can be measured simultaneously. With Qdots, on the other hand, spectral linewidth is much less broad, so many more colors can be used at once.

Qdots are extremely photostable, which is important for applications in pathology. Organic fluorophores photobleach rapidly under the light microscope, whereas the color of a Qdot remains stable for hours, which allows quantitative study. And Qdots are much brighter than organic fluorophores, 20 to 40 times brighter on a particle-to-molecule basis. This is important to achieve a high signal-to-noise ratio, Dr. Gao notes, since biological tissue has high background autofluorescence.

Several advances in microscopy bear on the value of Qdots. Dr. True says Aperio, Ariel, and Olympus are developing instruments for fluoroscopic image scanning, like those for transmission light scanning. They will be useful in exploiting the photostability of Qdots. In addition, Drs. True and Gao have demonstrated that analysis of Qdot emission by hyperspectral imaging (also called multispectral imaging, or MSI) can extend the practical range of Qdots to 10 colors. They wrote, “[H]ighly multiplexed analysis of cancer biomarkers is limited if hyperspectral imaging or quantum dots are used separately. The combination of the nanoprobe and spectral imaging enhances each other’s capability” (True LD, Gao X. J Mol Diagn. 2006;9:7–11). MSI is particularly helpful for in situ quantitation with Qdots of tumor markers in prostate and breast cancer because it is difficult to separate stromal and malignant cells. Dr. Nie’s group has reported efficient removal of background autofluorescence with MSI during in vivo imaging with Qdots (Xing Y, et al. Nat Protoc. 2007;2:1152–1165).

“MSI allows pathologists to take advantage of new opportunities for multiplexing,” says Richard M. Levenson, MD, director of research for biomedical systems at CRI Inc., Woburn, Mass. “It combines molecular specificity with exquisite imaging precision.” In addition to separating out tissue autofluorescence from otherwise hard-to-distinguish molecular signals, MSI solves a problem that Dr. Levenson says “has kept fluorescence imaging from being optimally used in pathology,” namely, the need to go beyond simple color imaging (including visual examination), which can’t discriminate, for example, between a fluorescent source that emits light in the true yellow range from combinations of dyes—for instance, red and green—that when spatially commingled will appear yellow. “A multispectral camera can separate those two sources of yellow,” Dr. Levenson says, “allowing us to use red, yellow, and green Qdots—and many others—and have no trouble unmixing and quantitating them.”

One feature of MSI that is especially important to the pathology community, Dr. Levenson says, is that it can reformat fluorescent signals so they look like conventional brightfield microscopy. “Most pathologists are unfamiliar with fluorescent images,” Dr. Levenson says. “That makes it hard to know what we are looking at when the background is black and nuclei are blue. I can foresee a time,” he says, “when a pathologist can’t tell what kind of image he or she is looking at. This would go a long way toward moving fluorescence into the pathologist’s world.” And toward realizing the potential of Qdots. CRI markets a CCD-camera-based system for MSI with a liquid crystal tunable filter that has no moving parts. Basic software enables the system to create and capture images at each Qdot wavelength. More sophisticated software performs image analysis, for instance, identifying where cancer is and restricting analysis to those areas.

Use of MSI, as well as more conventional methods of separating blended colors due to overlap of different-colored Qdots, requires sophisticated computational analysis. May Wang, PhD, assistant professor of biomedical engineering and hematology/oncology at Georgia Tech and Emory University, is responsible for this aspect of the Atlanta group’s research. Computer algorithms make it possible, she says, to separate two Qdots as close as 1 nm apart. Working with Dr. Young, Dr. Wang performs computational analysis to help select which possible biomarkers on microarrays are statistically associated with prognosis, therapeutic response, or subtype, so that Qdot-antibody conjugates can be made against them.

Dr. Wang is a member of the Cancer Biomedical Informatics Grid, or CaBIG, which is working to make computation tools freely available and usable on the National Cancer Institute Web site. She says the programs are “very interactive” and “very intuitive.”

Movement of Qdots into the clinical arena is “coming along,” Dr. True says. He and Dr. Gao have experiments underway to demonstrate that Qdot staining can be quantified and can detect changes in expression levels in prostate cancer cell lines. They have begun multiplexed Qdot analysis of prostate cancer cells, with the goal of analyzing expression levels of multiple genes in prostate needle biopsy tissue using just one section. “There are at least 20 gene markers that are prognostic or predictive for targeted therapy,” Dr. True says. He believes that Qdots will improve the ability to measure biomarkers, profile tumors, and predict response in prostate cancer. In another project, they will use Qdots to analyze expression of several androgen-regulated genes that are altered in prostate cancer, such as androgen receptor, PSA, IGF1R, and the transmembrane protease TMPRSS2.

Prognostic and predictive ability is also a goal in Dr. Young’s Qdot work. He is attempting to correlate extensive clinical, pathological, and gene expression data on a tissue bank of prostate cancer specimens with published data to identify potential biomarkers. A “lethal phenotype” in prostate cancer has been described, which carries an increased probability of mortality and bone metastases. Using antibodies to a handful of biomarkers conjugated to Qdots, John Petros, MD, a urologist and senior member of the Atlanta team, will look for varying expression of the markers on one slide to see what correlates with the lethal phenotype.

Working with Walter Curran, MD, chair of the Department of Radiation Oncology at Emory, the Atlanta team is also taking part in a prospective multicenter study on prostate cancer being conducted by the Radiation Therapy Oncology Group. It will be the job of the Atlanta group to see whether any Qdot-measured biomarkers correlate with response to radiation therapy. Ultimately, it may be possible to guide treatment decisions by doing multiplexed biomarker analysis on small needle core biopsies.

In breast cancer, a group led by Brian Leyland-Jones, MD, PhD, director of the Winship Cancer Institute at Emory, is evaluating several known biomarkers—ER, PR, HER2—in a Qdot-based IHC-like assay and correlating the results with treatment outcomes. Using Qdots to do profiling for prognostic categories will prove what the technology can do.

Using gene expression microarrays, Drs. Wang and Young have identified different profiles in different classes of renal cell cancer (Young AN, et al. Am J Pathol. 2001;158:1639– 1651; Schuetz AN, et al. J Mol Diagn. 2005;7: 206– 218; Young A, et al. Conf Proc IEEE Eng Med Biol Soc. 2005;1:723–726). These potential markers are being evaluated now for their correlation with prognosis in IHC assays. Analysis with Qdots could enhance this approach, since many more cases of renal cell tumors are being diagnosed earlier through abdominal imaging, leading to treatments that don’t require the entire kidney to be removed. The ability of Qdots to enable microscopic analysis of many biomarkers on a single small slice of tissue would be helpful in these cases.

In clinical flow cytometry, most laboratories routinely use three to four colors, while some research labs reach eight to 10. “Our group is one of the leaders in multicolor flow cytometry,” says Dr. Chattopadhyay. “We have special instruments that help us detect more, so we routinely measure 14 or 15 colors, and sometimes as many as 18.” To

do this, their instruments employ four lasers (compared with one or two in typical clinical settings) and the range is extended through the combination of Qdots with standard organic fluorophores.

“We have gone past the development stage and are applying Qdots to biology,” Dr. Chattopadhyay says. In a recent flow cytometry study, he and his colleagues measured 15 parameters simultaneously, using Qdots to measure three. The study looked at cryopreserved immune cells collected over 15 years from 470 HIV-positive individuals. Large amounts of reagents were used and work was carried out over three to four months. “Despite those challenges, Qdots worked at the same concentration,” Dr. Chattopadhyay says. Their color remained stable and consistent. “Qdots are definitely usable,” he concludes.

Dr. Grogan notes that some pharmaceutical companies are beginning to recognize there is an emerging science of predictive medicine and to “put a value on this cutting edge of pathology.” Early this year Roche purchased Ventana for more than $3 billion. “Roche was putting a high value on predictive pathology, betting that advanced diagnostics and predictive testing would make pharma delivery more efficient,” Dr. Grogan says. In the past, predictive testing was not a major field because there were few drugs directed at specific molecular targets. Even if there had been more drugs directed at molecular targets, the appropriate technology didn’t exist. “Now with things like quantum dots we have the means of taking on these questions,” Dr. Grogan says.

To realize the promise of personalized and predictive medicine, the NCI has funded eight national Centers of Cancer Nanotechnology Excellence, or CCNE. Dr. Nie, principal investigator of the CCNE at Emory and Georgia Tech, says the use of multicolor Qdot probes in pathology is one of the most important and clinically relevant applications. “We have gathered a team of clinicians, engineers, and biocomputing experts for clinical trials involving breast and prostate cancers in the next two years,” he says.

“We still have some way to go,” Dr. Young says, “to make all the tools related to quantum dots—from the detection and quantification of signal to the interpretation of data in a clinical context—automated and intuitive for pathologists and other clinicians.” Still, he believes, Qdots could move out of research into medical applications in the next several years.

Dr. Grogan thinks it is reasonable to expect to see a Qdot-based assay enter clinical use in the next couple of years. “What we are seeing now is an enormous amount of feasibility activity showing that in principle this technology works,” he says. “The next step will be specific applications. We need to come up with a test that affects patient management. That is what you will see emerging soon, a test for finding a target of therapy or predicting response to therapy.”

Related Links

• Breast cancer cells
  (PDF. 824 KB)
• Hyperspectral imaging of human prostate cancer tissue
  (PDF. 16 KB)
• Neonatal brain tissue specimens
  (PDF. 7.3 MB)