Organic dyes for deep bioimaging

Small-molecule organic dyes that fluoresce in the short-wave infrared region of the spectrum could improve the resolution of in vivo bioimaging methods. Such dyes have now been made by adapting those that fluoresce visible light.

Fluorescence-based imaging methods have transformed the way that scientists visualize and interpret biological events. An enduring goal is to make fluorescence imaging broadly useful in complex physiological settings, especially for clinical applications. One strategy is to use fluorescence wavelengths between 1,000 and 2,000 nanometres — the short-wave infrared (SWIR) region of the electromagnetic spectrum. But organic molecules that absorb and emit these wavelengths are needed to realize the full potential of SWIR imaging¹. Writing in Angewandte Chemie, Cosco et al.² report the brightest known organic small-molecule dyes that absorb and emit light at wavelengths greater than 1,000 nm.

Red light is preferentially transmitted through tissue — which is why your hand lights up red when you hold a torch behind it. To take advantage of this property, there have been substantial efforts to develop probes and in vivo imaging techniques that use far-red and near-IR wavelengths (650–1,000 nm).

However, imaging in the SWIR range offers important benefits for in vivo applications, compared with that in the far-red and near-IR, for two main reasons. First, SWIR produces less autofluorescence — light emitted naturally by biological structures that have previously absorbed light — so that images are produced against a nearly black background. Second, light scattering caused by particulate matter generally decreases with longer wavelengths. A theoretical study³ has predicted a 100- to 1,000-fold improvement in image resolution at centimetre depths of tissue when imaging in the SWIR range (1,350 nm), compared with imaging at a near-IR wavelength (850 nm).

Several challenges must be overcome to translate theory into experimental reality. One technological issue is that conventional light detectors (such as the sensors used in most digital cameras) lose effectiveness for wavelengths beyond 1,000 nm. The use of detectors that function in the SWIR range has been restricted, in part by national-defence concerns, because they are used in night-vision devices and in other military applications. However, sensitive cameras equipped with SWIR sensors based on the semiconductor indium gallium arsenide are now sold for biomedical use, and the availability and cost of these cameras are likely to improve.

The central chemical challenge in this field is to identify probes that emit in the SWIR range. A seminal study⁴ published in 2009 revealed that surface-modified carbon nanotubes absorb light in the near-IR range and emit at SWIR wavelengths. Imaging experiments with these compounds revealed dramatic benefits in resolution compared with conventional fluorescence imaging at near-IR wavelengths — greatly improving the resolution of through-skull imaging of the brain vasculature in mice⁵, for example. And earlier this year, a class of quantum dots (fluorescent nanometre-scale particles that contain heavy metals) was reported⁶ that enable a range of in vivo applications for SWIR imaging. These two types of nanomaterial finally brought the benefits of SWIR imaging to fruition.

But clinical protocols for optical bioimaging almost exclusively use fluorescent small molecules (known as fluorophores) as imaging agents, rather than nanomaterials⁷. This is because fluorophores generally have low toxicity, are cleared from the body by well-defined mechanisms, and can be straightforwardly targeted to biomolecules. The same is probably true for molecules that fluoresce in the SWIR region, but few such molecules are known, and even fewer that have useful properties for biological applications⁸.

The new dyes have several potentially useful properties for bioimaging applications.

Making bright SWIR molecules is a challenging problem for organic chemists. Solving it requires advanced synthetic methods combined with design insights from theoretical chemistry. Enter Cosco and colleagues, whose strategy was to make rational modifications to the molecular scaffold of fluorescent dyes already used for optical imaging using visible and near-IR wavelengths. Perhaps best known as the Cy dye series⁹, these compounds contain two heterocycles (rings that contain atoms other than carbon) connected by 3, 5 or 7 methine units (=CH– groups; Fig. 1a). They are exceptionally good light absorbers, a key property that contributes to these molecules being excellent emitters in response to light irradiation. Cosco and co-workers gave Cy dyes a chemical makeover by replacing the conventionally used heterocycles (indolenines) with dimethyl-flavylium (Flav) heterocycles (Fig. 1b), devising a concise synthetic route by which to prepare the new compounds.

Figure 1: Increasing the wavelength of light emitted and absorbed by dyes.

a, The Cy dye series of compounds is used for optical fluorescence imaging. The dyes consist of indolenine heterocycles (red) connected by polymethine linkers (orange). I⁻, iodide ion. b, Cosco et al.² modified the structures of these dyes, replacing the indolenines with dimethyl-flavylium heterocycles (blue), and dubbing the resulting compounds Flav dyes. Me, methyl group; Ph, phenyl group; ClO₄⁻, perchlorate ion. c, Each Flav dye emits and absorbs light at a longer wavelength than the analogous Cy dye. Notably, Flav7 emits and absorbs light in the short-wave infrared region (SWIR) of the electromagnetic spectrum. The authors used Flav7 as a probe for SWIR fluorescence bioimaging, which offers higher resolution and visualization of deep tissue than can be achieved using visible-light fluorescence imaging. Wavelengths are shown in nanometres. (Adapted from Fig. 1 of ref. 2.)

The resulting dyes have several potentially useful properties for bioimaging applications. The wavelengths of light absorbed and emitted by the new dyes are shifted roughly 200 nm towards longer wavelengths — that is, towards the infrared region — compared with the corresponding Cy dyes (Fig. 1c). Notably, the compound that contains 7 methine units in its linker emits light in the 1,100-nm range, and is the brightest small-molecule SWIR dye reported so far. The authors used a formulation of this dye to visualize deep vasculature in a mouse.

A major goal for clinical optical imaging is to enable surgeons to readily visualize, and then remove, tumour tissue¹⁰. The improved tissue penetration and optical resolution of SWIR imaging should help surgeons to identify metastases — secondary tumours formed at sites away from the primary tumour — that would be hard to find using existing methods.

Additional advances are required on multiple fronts before SWIR imaging can be broadly used for clinical applications. For example, fluorophores that have high solubility in water must be identified — this might be possible by making modifications to the molecules reported by Cosco and colleagues. Other key objectives include finding ways to target fluorophores to specific biomolecules, and developing fluorescent probes that can be 'switched on' in response to biological stimuli. These chemical advances will have to be paired with emerging developments from engineering and biomedical research. For example, a portable, ideally hand-held, SWIR imaging apparatus is needed that can be readily used in a surgical setting. Nevertheless, Cosco and colleagues' study provides a key breakthrough required to move SWIR imaging towards general clinical use.¹