Illumination Methods For Achieving Large Field-of-view Imaging In Super-resolution Localization Microscopy

Revealing the distribution feature and function of neural circuits over multiple brain regions has a great scientific impact in brain research.Super-resolution localization microscopy(SRLM)is a recent breakthrough in the field of optical microscopy,and provides the required spatial resolution for studying the ultrastructure of neural circuits.However,the imaging field-of-view(FOV)of SRLM is mostly limited to about 30 um x 30 um,which is not sufficient for studying the ultrastructure of neural circuits across a normal mouse brain slice with centimeter size.Such task is possible with the development of SRLM with large FOV,since enhancing the FOV in SRLM is achievable without scarifying either spatial resolution or temporal resolution.This paper aims to solve the technical challenges in developing SRLM with large FOV(hereafter called Large-FOV-SRLM),and consists of three parts.(1)Part I studies the key factors that limit the realization of Large-FOV-SRLM,and proposes a solution of using high-power homogeneous illumination to achieve Large-FOV-SRLM.Through systematic analysis on the imaging principle of SRLM,we find out that the requirement in illumination uniformity and intensity is the key factor responsible for realizing Large-FOV-SRLM.Therefore,we design and make a multi-mode fiber combiner for achieving high-power homogeneous illumination.This fiber combiner provides simultaneously a uniform illumination and a high coupling efficiency(97% for 561 nm and 92% for 639nm).Comparing to other approaches for high-power illumination,our method achieves a higher coupling efficiency and a comparable illumination homogeneity,outputs a square-shape illumination beam without beam clipping to match the square-shape sensor of typical low-light cameras,and decreases the demand for laser power.(2)Part II studies and realizes Large-FOV-SRLM with objective-type illumination.We analyze the advantages and limitations of different illumination methods in SRLM,and propose to use objective-type illumination for realizing Large-FOV-SRLM.For a spatial resolution of 40 nm,the imaging FOV is obtained to be as large as 221 um x 221 um,which is ~50 times bigger than that in popular SRLM,and ~5 times bigger than the current FOV record in SRLM.(3)Part III studies and provides preliminary results in Large-FOV-SRLM with prism-type illumination.We realize that Large-FOV-SRLM with objective-type illumination is not suitable for highly inclined thin illumination(HILO)and total internal reflection illumination(TIRF),which are useful for minimizing the fluorescence background in SRLM.Basing on the new multimode fiber combiner,we design a prism-type illumination setup for Large-FOV-SRLM.Combining this setup with HILO illumination,we realize a spatial resolution of 40 nm with a FOV of 132 μm x 200 μm.This FOV is ~30 times bigger than that in popular SRLM,and ~3 times bigger than the current FOV record in SRLM.Compared with Large-FOV-SRLM using objective-type illumination,this new approach provides a higher signal-to-noise ratio due to the use of HILO illumination.