Quantitative 3-D morphometric analysis of individual dendritic spines

Abstract

The observation and analysis of dendritic spines morphological changes poses a major challenge in neuroscience studies. The alterations of their density and/or morphology are indicators of the cellular processes involved in neural plasticity underlying learning and memory, and are symptomatic in neuropsychiatric disorders. Despite ongoing intense investigations in imaging approaches, the relationship between changes in spine morphology and synaptic function is still unknown. The existing quantitative analyses are difficult to perform and require extensive user intervention. Here, we propose a new method for (1) the three-dimensional (3-D) segmentation of dendritic spines using a multi-scale opening approach and (2) define 3-D morphological attributes of individual spines for the effective assessment of their structural plasticity. The method was validated using confocal light microscopy images of dendritic spines from dissociated hippocampal cultures and brain slices (1) to evaluate accuracy relative to manually labeled ground-truth annotations and relative to the state-of-the-art Imaris tool, (2) to analyze reproducibility of user-independence of the segmentation method, and (3) to quantitatively analyze morphological changes in individual spines before and after chemically induced long-term potentiation. The method was monitored and used to precisely describe the morphology of individual spines in real-time using consecutive images of the same dendritic fragment.

Figure 1

Confocal light microscopy image of hippocampal dendrite covered with dendritic spines. (A) Maximum intensity projection (MIP) of the z-stack with an outlined region-of-interest (ROI). (B) Cropped and enlarged MIP image of the selected ROI. (C) 3-D rendering of the selected ROI from the confocal z-stack with enhanced morphological details of individual spines. (D) 3-D segmentation result of the selected ROI with the extracted spines marked in red.

Figure 2

Intensity distribution between the dendrite and spine segments in a sample confocal light microscopy image. (Top) Color-coded regions in a dendritic segment. (Bottom) Intensity histogram of the dendritic segment that represents the overlap of the pure dendrite region and shared space between the dendrite and spines.

Figure 3

Illustration of a segmented spine structure with automatic quantitative assessment of different spine attributes.

Figure 4

Examples of the segmented spines of different categories that were obtained by the currently developed method. (A) stubby, (B) filopodia, (C) mushroom, and (D) spine-head protrusions, (E) branched spines.

Figure 5

Accuracy analysis of the developed method relative to the manually segmented ground-truth annotations. (A) Two standard features (i.e., volume and length) were considered for the analysis of agreement of the estimated feature values over the sample spine population. (B) Cropped image (MIP and 3-D segmentation result) of dendritic segment with numbered spines corresponding to the selected ROI at the graph. (C) Difference plots between the estimated by current method and ground-truth values for the spine volume (left) and spine length (right) features (Bland-Altman plots).

Figure 6

Comparative analysis of the 3-D segmentation results relative to the state-of-the-art Imaris tool on three sample dendritic segments taken from three different cell images at baseline condition. (A) Sample 2-D MIP image with a highlighted ROI segment, (B) 2-D MIP image of the selected dendritic segment; (C) 3-D segmentation result of image b using the Imaris tool (D) 3-D segmentation result of image b using the developed segmentation methodology.

Figure 7

Reproducibility analysis of the currently developed method over a sample spine population with three blinded, independent experts. (A) Quantitative analysis of dendritic spine Volume, Length and Head Width performed by three independent users together with (B) corresponding 3-D reconstruction results.

Figure 8

Qualitative assessment of spine plasticity relative to the 3-D segmentation results for a dendritic segment at baseline, 10 minutes and 40 minutes after cLTP induction. (A) Comparisons of 3-D rendition results relative to the MIP images, (B) 3-D segmented results and morphological changes are observed on sample spines before and after cLTP induction.

Figure 9

Quantitative analysis of dendritic spine morphology by 3-D segmentation method. (A) Average relative changes (percentage) in spine morphology (volume) estimated by the currently developed method at two time points in primary neuronal culture were found to be: 0.94 ± 2.64 (DMSO), 43.25 ± 7.02 (cLTP) at ten minutes stimulation and 37.29 ± 2.43 (DMSO), 42 ± 5.34 (cLTP) at 40 minutes stimulation. (B) Changes in dendritic spine morphology (neuronal culture). Results of automatic spine classification where S, F, and M represent spine categories: Stubby, Filopodia and Mushroom respectively. The numbers represent relative transition (in % of total population) between different spine categories before and after 10 minutes and 40 minutes of cLTP induction. (C) Average volume of dendritic spines from brain slices before 1.13 ± 0.09 µm³ and after stimulation 1.53 ± 0.12 µm³. The results are expressed as mean ± SEM.

Figure 10

Segmentation of dendritic spines from complex dendritic tree in brain slice. (A) 2-D MIP image with a highlighted ROI for 3-D analysis, (B) spine segmentation result using the proposed methodology on the selected ROI.