Nonlinear-optical stain-free stereoimaging of astrocytes and gliovascular interfaces

摘要

Methods of nonlinear optics provide a vast arsenal of tools for label-free brain imaging, offering a unique combination of chemical specificity, the ability to detect fine morphological features, and an unprecedentedly high, subdiffraction spatial resolution. While these techniques provide a rapidly growing platform for the microscopy of neurons and fine intraneural structures, optical imaging of astroglia still largely relies on filament-protein-antibody staining, subject to limitations and difficulties especially severe in live-brain studies. Once viewed as an ancillary, inert brain scaffold, astroglia are being promoted, as a part of an ongoing paradigm shift in neurosciences, into the role of a key active agent of intercellular communication and information processing, playing a significant role in brain functioning under normal and pathological conditions. Here, we show that methods of nonlinear optics provide a unique resource to address long-standing challenges in label-free astroglia imaging. We demonstrate that, with a suitable beam-focusing geometry and careful driver-pulse compression, microscopy of second-harmonic generation (SHG) can enable a high-resolution label-free imaging of fibrillar structures of astrocytes, most notably astrocyte processes and their endfeet. SHG microscopy of astrocytes is integrated in our approach with nonlinear-optical imaging of red blood cells based on third-harmonic generation (THG) enhanced by a three-photon resonance with the Soret band of hemoglobin. With astroglia and red blood cells providing two physically distinct imaging contrasts in SHG and THG channels, a parallel detection of the second and third harmonics enables a high-contrast, high-resolution, stain-free stereoimaging of gliovascular interfaces in the central nervous system. Transverse scans of the second and third harmonics are shown to resolve an ultrafine texture of blood-vessel walls and astrocyte-process endfeet on gliovascular interfaces with a spatial resolution within 1 mu m at focusing depths up to 20 mu m inside a brain.