Upconversion nanoparticles convert NIR light into visible light, which can then activate optogenetics tools such as ChR2. Credit: Reproduced with permission from Chen et al. (2018).

Minimally invasive methods to stimulate neurons in the rodent brain are available, but they tend to be temporally imprecise or not cell-type specific. In contrast, optogenetic tools can be targeted to specific cell types and can be precisely controlled in time. However, the technology requires the implantation of devices to illuminate brain areas of interest, which can cause damage.

To overcome these problems, Shuo Chen, a postdoc in the lab of Thomas McHugh from the RIKEN Brain Science Institute in Saitama and the University of Tokyo, initiated a collaboration between the McHugh lab and the lab of Xiaogang Liu from the National University of Singapore and A*STAR in Singapore. The teams explored the feasibility of using upconversion nanoparticles (UCNPs) for optogenetic manipulation of neurons in the mouse brain.

UCNPs can emit in the visible range after stimulation with near-infrared (NIR) light. Their emission wavelength can be tuned to match the action spectra of tools such as channelrhodopsin-2 (ChR2) and archaerhodopsin (Arch). UCNPs have been applied in the context of optogenetics in cell culture, as well as in the roundworm and zebrafish. “Some attempts at in vivo work in rodents were reported,” explains McHugh, but they were done in a very invasive way.

The team found that they could inject the UCNPs into the brain area of interest and then illuminate them through the intact skull using 980-nanometer-wavelength NIR light. The UCNPs could be located as deep as 4.5 millimeters and still generate enough short-wavelength light to activate ChR2 or Arch, and the NIR illumination scheme was designed to keep tissue heating at bay. “We included a large amount of data characterizing this and demonstrating it was safe under the conditions we employed,” writes McHugh via email.

The researchers applied their approach first in the ventral tegmental area (VTA), a deeply situated area in the mouse brain. They targeted ChR2 to dopamine neurons in the VTA by using a viral vector in combination with a Cre recombinase line, and they injected blue-emitting UCNPs into this area. Transcranial NIR illumination activated dopamine neurons, as confirmed by higher c-Fos levels and dopamine release. The researchers also targeted ChR2 to parvalbumin-expressing neurons in the medial septum, which is involved in theta rhythmicity, an oscillatory pattern of neural activity in the range of 4–7 Hertz. In this case, the team was able to modulate the frequency of theta oscillations with pulsed NIR illumination. Finally, UCNP-mediated optogenetics can be used to influence the behavior of awake mice, which the authors demonstrated by inducing freezing behavior in response to NIR illumination of ChR2-expressing neurons in the dentate gyrus.

The researchers also showed that UCNPs tuned to emit green light can activate the inhibitory tool Arch. But the strategy is not limited to ChR2 and Arch. “We are working now to expand the use to other opsins,” McHugh writes.

Although the technology works well and is promising, it is not without challenges. The illumination with NIR light leads to temperature increases and therefore needs to be tightly controlled. In addition, although NIR light scatters less than visible light, areas beyond about 4 millimeters in depth cannot be reached because of scattering.

Further developments in the UCNP design could improve the efficiency of upconversion. “We will continue our collaboration with the Liu lab to improve the targeting and efficacy of UCNPs and test their performance in vivo,” writes McHugh. He would also like to transfer the approach to animals with larger brains, such as rats or nonhuman primates.