Detailed Colocalization Analysis of A- and B-type Nuclear Lamins: a Workflow Using Super-Resolution STED Microscopy and Deconvolution

Abstract

The inner nuclear membrane is covered by a filamentous network, the nuclear lamina, consisting of A- and B-type lamins as its major components. The A- and B-type lamins form independent but interacting and partially overlapping networks, as demonstrated by previous super-resolution studies. The nuclear lamina in fibroblast cultures derived from laminopathy patients shows an increased segregation of the A- and B-type lamin networks, which can be quantitatively expressed by the Pearson's Correlation Coefficient (PCC). Blurring and noise (convolution), however, significantly affect the quality of microscopy images, which led us to optimize the deconvolution approach for Confocal Scanning Laser Microscopy (CSLM) and Stimulated Emission Depletion (STED) microscopy images. For that purpose, the differences in using a theoretical, experimental, or semi-experimental Point Spread Function (PSF), an important parameter for deconvolution, was evaluated for its use in deconvolution of CSLM and STED microscopy images of double immunolabeled healthy and laminopathy patient fibroblasts. The semi-experimental is a new PSF introduced in this study, which combines the theoretical and experimental PSF to solve issues that arise from noisy PSF recordings due to very small and thereby low intensity fluorescent beads. From these deconvoluted images, the colocalization of the lamin networks could not only be quantified at the level of the nucleus as a whole, but also at a subnuclear level. The latter was achieved by dividing the nucleus into multiple equal rectangles using a custom-made ImageJ macro in Fiji. In this detailed analysis, we found heterogeneity in the colocalization of lamins A/C and B1 within and between nuclei in both healthy and laminopathy dermal fibroblasts, which cannot be detected in one single analysis for the entire nucleus.