General Background: Quantum dots, first introduced for biological labeling in 1998 (Bruchez et al., 1998; Chan and Nie, 1998) have proved to be extraordinarily useful fluorescence reagents that have significant advantages over other types of fluorescent dyes. They combine very high brightness, due both to high absorbency and high quantum yields, with unprecedented resistance to photobleaching. Emission wavelengths, governed primarily by composition and secondarily by size, range from the near-ultraviolet to the infrared. For a given quantum dot composition, emission bandwidths depend on the size range of the quantum dots and can be very narrow; <35 nm width is now routine for production quantum dots. Core-shell quantum dots can have quantum yields greater than 90%, and quantum dots may be excited in a broad spectral window shorter than their emission wavelength, with absorbency rising as wavelength decreases. For example, molar absorbencies of 6 × 106 at 450 nm and quantum yields in excess of 60% are found for the 655 nm polyethylene glycol (PEG)-coated cadmium selenide-zinc sulfide (CdSe-ZnS) core-shell quantum dots used in our laboratory. Because all quantum dots show rising absorbency at wavelengths shorter than their emission wavelengths, many quantum dots can be excited using one illumination window. (Readers are referred to the Quantum Dot Corporation Website, http://www.qdots.com, for more details and for several tutorials on the properties of quantum dots.). The combination of high brightness, photostability, and narrow emission bandwidths with the ability to excite many colors naturally leads to the possibility of using multicolor combinations of quantum dots to label ("bar-code") large numbers of different objects (e.g., different cell types in a mixed population) (Chan and Nie, 1998; Gao and Nie, 2003, 2004; Han et al., 2001; Jaiswal et al., 2003; Lagerholm et al., 2004; Voura et al., 2004). Quantum dots may be used for single-molecule imaging in living cells (Dahan et al., 2003). Finally, quantum dots are well suited for two-photon microscopy (Larson et al., 2003). There are potential drawbacks to the use of quantum dots: their large size and high molecular weights may limit applications that require measurement of molecular mobility, and attached quantum dots might interfere with molecular interactions. Finally, since the current generation of quantum dots is composed of toxic heavy metals (CdSe and cadmium telluride [CdTe] cores, with ZnS shells), toxicity might be anticipated if the quantum dots degrade during use. Several recent reviews have summarized progress in biological applications of quantum dots (Jaiswal and Simon, 2004; Parak et al., 2003; Smith et al., 2004). In this chapter we briefly review recent advances in the use of quantum dots for biological imaging, then summarize our work on the effects of chemically varying quantum dot surface properties to improve cellular uptake and imaging in vivo. Applications of nucleic acid conjugates are not discussed. Stabilizing the Surface: Core-shell quantum dots are stable and highly fluorescent in nonpolar organic solvents but are not very fluorescent in aqueous solution. Much current work involves making surface coatings that preserve high fluorescence, confer stability and solubility in aqueous solution, and allow ready conjugation of biological molecules. During the usual core-shell manufacturing process, freshly prepared quantum dots are coated with trioctylphosphine oxide (TOPO) (Dabbousi et al., 1997; Hines and Guyot-Sionnest, 1996). Thus the quantum dots start with a hydrophobic surface. Several approaches have been reported to work reasonably well for creating a stable primary coat around the core-shell that preserves fluorescence in aqueous solution:1Coating the quantum dot with an amphiphilic (amp) polymer (Gao et al., 2004; Watson et al., 2003; Wu et al., 2003)2Replacing the TOPO using organosulfur or other organophosphorous ligands that compete for binding sites on the quantum dot shell. Examples include mercaptoacetic acid (Chan and Nie, 1998), mercaptopropionic acid (Mitchell et al., 1999), dihydrolipoic acid (DHLA) (Mattoussi et al., 2000), dl-cysteine (Sukhanova et al., 2004), and an organic phosphene oligomer (Kim and Bawendi, 2003)3Forming a micelle around quantum dots using a mixture of phosphatidylcholine and PEG-substituted phosphatidylethanolamine (Dubertret et al., 2002), or by coating with a cone-shaped amphiphile that self-assembles around the quantum dot (Osaki et al., 2004)4Creating a silica layer around the dot (Bruchez et al., 1998; Chen and Gerion, 2004; Gerion et al., 2001)5Adsorbing albumin directly to quantum dots (Hanaki et al., 2003).These methods may be combined; for example, a primary cysteine coat was overcoated using polyallylamine (Sukhanova et al., 2004). All these primary coats allow conjugation of biomolecules, usually by amide or thioether formation. Minimizing Uptake In Vitro and In Vivo: Minimizing nonspecific aggregation and binding requires further modification of quantum dot surfaces. Most authors have used PEG conjugation to minimize nonspecific binding, as PEG derivatives for conjugation are readily available and work effectively in many systems (see reviews by Chapman, 2002; Greenwald, 2001; Harris and Chess, 2003; Harrington et al., 2002; Molineux, 2002). A potential drawback to PEG conjugation is that neither tissue-cultured cells nor live animals metabolize PEG (Ibid.). Sugar or polysaccharide derivatives offer another potential way to avoid nonspecific binding or to cause binding to appropriate receptors (Osaki et al., 2004). Conjugates that Confer Specificity: Conjugation of biomolecules to the primary coat is usually performed by conventional methods, using active esters, carbodiimides, or maleimides. An interesting approach is the use of cationic proteins or polyhistidine tagged chimeric proteins for self-assembly onto DHLA-coated quantum dots. Binding is mediated by electrostatic interactions between the negatively charged surface of the quantum dot and the cationic protein or tag (Goldman et al., 2002a,b; Mattoussi et al., 2000; Voura et al., 2004). This primary coat allows fluorescence resonance energy transfer (FRET) between adsorbed surface components and quantum dots (Clapp et al., 2004; Medintz et al., 2003a,b, 2004). Binding and Uptake Using Cultured Cells In Vitro: Binding to cell surfaces nonspecifically can promote uptake (Derfus et al., 2004a). Quantum dots layered on a surface can be taken up by cells that traverse the surface (Parak et al., 2002). Binding to, and in some cases uptake by, specific cell surface receptors has been demonstrated in many cases (transferrin receptor [Chan and Nie, 1998], serotonin receptors [Rosenthal et al., 2002], glycine receptor [Dahan et al., 2003], and epidermal growth factor receptor [Lidke et al., 2004]). Membrane labeling and monitoring of membrane integrity was performed using concanavalin A-biotin bound to streptavidin quantum dots (Minet et al., 2004). Cell labeling by quantum dots after microinjection (Dubertret et al., 2002), electroporation (Chen and Gerion, 2004), and cationic lipid-mediated cell entry (Derfus et al., 2004a; Voura et al., 2004) have all been demonstrated. Although internalized quantum dots frequently localize in endosomes, organelle-specific localization to the cell nucleus and to mitochondria has been shown using appropriate peptide conjugates to DHLA-coated quantum dots (Derfus et al., 2004a) and to the nucleus using peptide conjugates of silica-coated quantum dots (Chen and Gerion, 2004); thus once quantum dots are transported across the cell membrane, tagging of intracellular structures is possible. Derfus et al. (2004a) compared four methods of internalizing quantum dots: a membrane transport peptide, cationic lipids, electroporation, and microinjection. All methods except microinjection caused aggregation of the mercaptoacetic acid-coated quantum dots used in their experiments; cationic lipids provided the highest delivery. We have successfully incorporated large numbers of polymer-coated quantum dots by using polyarginine conjugates. Results from our laboratory (Lagerholm et al., 2004) are discussed in the following text. Cell Tracking In Vivo: Two groups have used quantum dot-labeled cells to follow circulation and extravasation of labeled tumor cell in living mice. Hoshino et al. (2004) used quantum dots coated by cross-linked sheep albumin to label mouse EL-4 lymphoma cells. Although high concentrations of these quantum dot conjugates proved toxic to cells, stable nontoxic labeling could be obtained at lower input levels. After injection into mice, persistence of labeled cells in circulation could be followed using a fluorescence-activated cell sorter (FACS) for at least 5 days, and uptake in various tissues (mainly spleen and lungs) could be monitored by microscopy. Voura et al. (2004) used cationic lipid-mediated internalization to label mouse B16F10 melanoma cells with DHLA-capped quantum dots. Cells were co-labeled with organic dyes to determine whether any loss of quantum dots occurred, as would be shown by a loss of coincidence of the two labels; on the time scale of these experiments (5 hours), no such loss was detectable. Again, by comparison with organic dye labels, quantum dots had no effects on the ability of the cells to survive in circulation, extravasate (emerge from the vascular compartment), or form metastatic tumors. Both standard fluorescence microscopy and two-photon microscopy could be used to detect the labeled cells.
