Premium
Computer‐assisted estimation in the CNS of 3D multimarker ‘Overlap’ or ‘Touch’ at the level of individual nerve endings: A confocal laser scanning microscope application
Author(s) -
Wouterlood Floris G.,
Boekel Amber J.,
Meijer Gerrit A.,
Beliën Jeroen A.M.
Publication year - 2007
Publication title -
journal of neuroscience research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.72
H-Index - 160
eISSN - 1097-4547
pISSN - 0360-4012
DOI - 10.1002/jnr.21244
Subject(s) - confocal , neuroscience , colocalization , postsynaptic potential , calbindin , parvalbumin , synaptophysin , biology , computer science , pattern recognition (psychology) , artificial intelligence , immunohistochemistry , physics , optics , biochemistry , receptor , immunology
Abstract Presynaptic boutons and associated postsynaptic structures in the CNS express markers that are highly synapse type‐specific. In multilabel immunofluorescence imaging, coexpression of such markers appears as overlap of signals in the same structures whereas closely related yet segregated markers, e.g., located pre‐and postsynaptically, translate into signals that touch. ‘Overlap’ and ‘touch’ occur in three dimensions (3D). The instrument of choice to study overlap vs. touch of small objects in tissue volumes is the confocal laser scanning microscope (CSLM). To quantify overlap and touch we used two paradigms. Overlap was studied in rat brain sections triple‐immunostained with antibodies against markers predominantly located presynaptically: glutamic acid decarboxylase, vesicular glutamate transporter 2, and calretinin. Touch was studied in rat temporal cortex where afferent, tracer‐labeledentorhinohippocampal fibers in the subiculum were imaged together with possible postsynaptic target neurons immunostained with an antibody against the calcium binding protein, parvalbumin. Z‐series of CLSM images were obtained in multiple channels. After post‐acquisition deconvolution we further processed the images via software written in the C/C++ SCIL_Image computer programming environment. The software receives parameters via scripts, after which it first identifies 3D objects through establishing isodensity envelopes around pixels representing small biologic structures (in our case: boutons) and then compares associated Z‐series in which it determines whether there is overlap or touch between recognized 3D objects. Finally, graphic and numeric output is produced. With this script‐commanded software we feel equipped to accurately and objectively quantify overlap and touch. © 2007 Wiley‐Liss, Inc.