Just the names of nanotechnology imaging probes reveal their variety: cages, diamonds, dots, rods, tubes and even wontons. The range of substances is just as wide: carbon, semiconductors, precious metals and more. But it is not their variety that interests biologists. Compared with existing fluorescent proteins and small-molecule dyes, nanotech probes can offer signals that are severalfold brighter and hundreds of times more stable. “You can just blast them with laser photons and they don't break down like an organic dye,” says David Piwnica-Worms, who directs the Molecular Imaging Center at the Washington University in Saint Louis.

Nanoparticles, particularly ones made of precious metals (Box 1), provide or enhance signals in imaging approaches beyond fluorescence microscopy. Magnetic resonance imaging, photoacoustic imaging and Raman spectroscopy approaches all draw on this branch of materials science. But nanotechnology probes also have drawbacks: they are often large and difficult to introduce into living systems. Once introduced, they may interfere with the very processes they are meant to monitor; they can clump or otherwise behave poorly in a biological milieu; their properties can vary from batch to batch; and the ease of targeting them to specific proteins, subcellular sites or cell types ranges from challenging to impossible.

Nanotechnology bioimaging applications for basic research are both eclipsed and supplemented by the pursuit of nanomaterials as drug-delivery devices and clinical diagnostics, which have an easier time attracting money from investors and translational medicine grants. In fact, scientists developing nanomaterials for bioimaging have plenty of distractions. Carbon nanotubes have been considered for hydrogen storage and as tiny semiconductor devices, for instance. Quantum dots have been explored in solar cells, digital cameras and light-emitting diodes. Nonetheless, the use of nanomaterials in imaging applications continues to advance rapidly, both in the improvement of existing agents and in the development of new ones.

Quantum dots can be designed to fluoresce in many colors. Credit: Brad Kairdolf and Shuming Nie, Emory University

Follow the dots

The optical properties of quantum dots seem made to order. (In fact, quantum dots have long been available for order: the Quantum Dot Corporation, now part of Life Technologies, launched its first product in 2002.) These semiconducting nanocrystals can be excited by a range of light wavelengths, but the light they emit is confined to a narrow spectrum controlled by the dots' size and composition. This property allows researchers to not only home in on their signals but also use several multicolored probes at once. Typically made from cadmium or selenium, the inorganic centers of quantum dots used in imaging are coated to prevent toxicity and allow the attachment of antibodies, peptides or other moieties to direct which cells or biomolecules the quantum dots interact with.

Reading signals from quantum dots would be easier if the dots could fluoresce without requiring an external source of excitation light. Not only does external light damage cells and tissues, but naturally occurring fluorescing molecules also respond to the blue-to-green light typically used to excite quantum dots, producing background noise. Last year, California start-up Zymera began distributing quantum dots that self-illuminate using a strategy known as bioluminescence resonance energy transfer (BRET). Quantum dots are conjugated to a version of the enzyme luciferase. When the enzyme is exposed to its substrate, it produces blue light, and this excites the quantum dots and causes them to emit specified red-to-infrared wavelengths.

Tests on individual nanodiamonds show that even those as small as 4 nanometers can hold color centers stably. Credit: Carlo Bradac, James Rabeau laboratory

Radioactivity can also be used to excite quantum dots. Researchers at Stanford University demonstrated this approach using radioactive iodine to elicit signals1. Independently, researchers at Washington University in Saint Louis reported a technique called Cerenkov radiation energy transfer, which uses probes developed for positron emission tomography (PET). The radioactive elements in PET probes generate Cerenkov radiation, which also causes the characteristic blue glow seen in nuclear reactors. This type of light is quickly absorbed by tissue and thus is inconvenient for imaging. It can, however, excite quantum dots2. Besides doing away with the need for excitation light, Cerenkov radiation energy transfer allows studies of PET probes using less-expensive, more-convenient optical techniques.

A scanning electron micrograph showing large diamond crystallites grown from a nanocrystal. Scale bar, 0.5 μm. Credit: Shaneel Chandra, James Rabeau laboratory

Many researchers are less concerned about how to excite quantum dots than how to maintain their signals. Fluorescent signals from quantum dots tend to turn on and off unpredictably, a quality known as 'blinking'. Blinking can actually be beneficial when a researcher needs to ascertain that a signal is coming from a single quantum dot, but the trait is still considered more a problem than an asset. Nonblinking quantum dots have been reported but have not yet been tested in cells or tissues3.

Researchers would also like quantum dots that are better suited for use inside cells. Including their coatings, the quantum dots most frequently used for imaging are 10–20 nanometers (nm) in diameter. Particles this large generally do not readily enter cells and, once inside, tend to get trapped in vesicles, away from cellular events. “There is a push to develop quantum dots that will be much smaller,” says Shuming Nie, a nanotechnologist at Emory University and the Georgia Institute of Technology. Smaller quantum dots would allow their use as tracers to observe molecular events in real time, an application that is generating considerable excitement4.

Although quantum dots have been used to image dynamic events on the cell membrane, researchers are eager for facile applications inside cells. The goal is to have widely available, functionalized dots with a diameter smaller than 5 nm, which is comparable to that of many genetically encoded fluorescent tags, Nie says. “If the quantum dot can be minimized to be the same size as green fluorescent protein, it could open a new revolution in cellular imaging.”

Nanotechnologists have found ways to produce quantum dots that are considerably smaller than commercially available ones, says Shimon Weiss at the University of California, Los Angeles. There are two main approaches to decreasing the size of quantum dots. One is to make the coating around the dots extremely thin. Right now, the coating on commercial dots can be 5 nm thick or more, says Nie, who has developed a polydentate coating that is only 1 nm thick but still provides a stable organic-inorganic interface and prevents toxicity.

The other approach Nie and others are taking is to shrink quantum dots without interfering with their optical properties. As the semiconducting nanocrystals get smaller, they emit shorter wavelengths that are less efficient at penetrating tissue and reaching a microscope's detector, thus counteracting the advantages of a smaller size. Much work has gone into creating new alloys that emit longer, even infrared wavelengths; these include exotic alloys such as zinc sulfide buried in cadmium telluride, or copper combined with indium and selenium, but these quantum dots have not yet been widely adopted5.

In unpublished work, researchers in Weiss's lab have used a combination of these two approaches—a peptide coating and alloys of indium and arsenic—to create fully functionalized quantum dots with diameters of 5–7 nm. Because these and similar materials are still under development, biologists may not be able to access the technology easily. “Currently it's mostly in the hands of only a few labs that have been developing quantum dots,” he says.

Nano goes organic

Carbon, one of the most abundant elements, is an appealing substance for nanotechnology imaging probes. Unlike quantum dots, however, the fluorescent properties of carbon-based probes are harder to control. Single-walled carbon nanotubes are the most explored of potential imaging tools. They are about 1 nm in diameter and about 100 nm long, dimensions that provide a good surface-to-volume ratio for coupling biomolecules and drugs. They also naturally photoluminesce in the tissue-penetrating near-infrared region. However, the fluorescence of carbon nanotubes is quenched when they aggregate.

Nanodiamonds can be fed to Caenorhabditis elegans and imaged over several days. Credit: Huan-Cheng Chang laboratory

Carbon nanotubes are typically coated with a layer of phospholipids to make them biocompatible, but this technique damages the nanotubes and can prevent fluorescence. However, Hongjie Dai and colleagues at Stanford University reported an alternative, gentler technique. When injected into mice, nanotubes manufactured using this technique gave a stronger, clearer signal than a 15-fold higher dose of nanotubes produced by standard techniques. And the nanotubes will be useful for other imaging modalities, predicts Dai6. “A single injection of well-functionalized SWNTs [single-walled nanotubes] could be a multimodal imaging probe for fluorescence, Raman and photoacoustic imaging.”

Nanodiamonds are taken up by intestinal cells of Caenorhabditis elegans. Credit: Huan-Cheng Chang laboratory

Dai says that a 6-month study in mice has not revealed signs of toxicity from these nanotubes and has shown that the nanotubes are excreted over several months. However, toxicity is still a concern. Like quantum dots, nanotubes can accumulate in tissues. In addition, researchers worry that their needle-like shape could damage cells, though Dai says he has not seen evidence of this.

Originally discovered from broken bits of carbon nanotubes, carbon dots are a relative newcomer on the imaging scene. Ya-Ping Sun at Clemson University has been able to obtain carbon dots (irregular clumps of carbon with many functional groups coming off them) that are individually as bright as quantum dots, though doing so requires separating the brightest carbon dots from dimmer ones. Carbon dots are easy to functionalize, and preliminary work shows that they are internalized into cells much more easily than quantum dots are, says Sun, but their fluorescence is in the green range. Sun believes that the low toxicity of carbon dots will spur their development for clinical applications and that their use for basic research imaging will follow, particularly if their fluorescence in the near-infrared range could be boosted.

Whereas atoms in nanotubes and carbon dots are arranged in a soft, graphene-like state, carbon's harder form is generating considerable excitement. Though they can also be produced in laboratories by chemical vapor deposition, nanodiamonds are commonly purified from what is essentially an industrial waste product collected after explosions (the process is called shock synthesis). Due to imperfections called nitrogen-vacancy centers or color centers, nanodiamonds absorb light around 560 nm and emit light at 700 nm, creating signals distinct from other fluorescent materials in cells.

Nanodiamonds also offer advantages not yet demonstrated in other probes; they do not blink like quantum dots or bleach like proteins. Their “perfect photostability” makes nanodiamonds useful for long-term tracking inside cells, such as following molecular motors, says François Treussart of Ecole Normale Supérieure de Cachan. The photostability also means nanodiamonds could be used to monitor cell processes at the nanometer scale. In fact, nanodiamonds' use in super-resolution microscopy has already been demonstrated, albeit outside biological systems7.

Nanodiamonds have also been imaged in vivo. Researchers led by Huan-Cheng Chang at the Institute of Atomic and Molecular Sciences of Academia Sinica recently fed nanodiamonds coated with biomolecules such as dextran and bovine serum albumin to the flatworm Caenorhabditis elegans. The nanodiamonds were taken up by intestinal cells, allowing the researchers to image the development of the digestive system for several days8. “It demonstrates that fluorescent nanodiamonds are nontoxic at both cellular and organismic levels,” says Chang.

Concerns that smaller diamonds would lose crucial imaging properties were recently assuaged by work led by James Rabeau, a physicist at Macquarie University, who was able to examine the optical properties of individual nanodiamonds as small as 4 nm across and show that these could still hold a stable color center, which opens up the possibility of creating smaller and brighter diamonds9. Quantum dots still have an advantage over nanodiamonds in terms of brightness and the ability to emit a range of colors. “The light collected from a nanodiamond containing a single nitrogen-vacancy center has an intensity three times lower than that of a quantum dot,” but this can be compensated for if a diamond has more color centers, Treussart says. Chang, Treussart and others are working to find ways to put more color centers in small diamonds, as well as to create versions that emit other colors.

Not all of nanodiamonds' exceptional properties are optical. They have interesting magnetic and electronic properties as well, and their lattice-like structure gives them promise as a drug-delivery device. Researchers Thomas Meade and Dean Ho of Northwestern University recently explored the possibility of using nanodiamonds to increase contrast in magnetic resonance imaging10. The nanodiamonds were functionalized with gadolinium, a contrast agent used to improve signals in magnetic resonance imaging studies. Because of their low toxicity, Meade wanted to use the nanodiamonds to monitor drug delivery over a course of several days—any boost to signal was a bonus. When the boost was measured to be an order of magnitude larger than that from other contrast agents, Meade initially thought the result was an artifact. “I told my researcher to go back to the lab and repeat the experiments,” he recalls. Additionally, the particular conjugation of gadolinium to nanodiamond used in the study is just one of many possibilities to form this link. It is possible that further work will produce more powerful signals and introduce more versatility, he says.

So far, though, much work with nanodiamonds is focused on proving them as imaging tools rather than using them as imaging tools. “If you start searching through the biological journals for diamonds, you won't find much,” says Rabeau. Perhaps that is because biologists are not yet ready to adopt them over existing, tried-and-tested fluorophores. Nonetheless, there is a good possibility that for some applications nanodiamonds may be even bigger—and smaller—than quantum dots.

Table 1 Suppliers guide: Companies ofering imag ing equipment, supplies and nanomaterials