Laser capture microdissection-targeted mass spectrometry: a method for multiplexed protein quantification within individual layers of the cerebral cortex

Abstract

The mammalian neocortex is organized into layers distinguished by the size, packing density, and connectivity of their constituent neurons. Many neuropsychiatric illnesses are complex trait disorders with etiologic factors converging on neuronal protein networks. Cortical pathology of neuropsychiatric diseases, such as schizophrenia, is often restricted to, or more pronounced in, certain cortical layers, suggesting that genetic vulnerabilities manifest with laminar specificity. Thus, the ability to investigate cortical layer-specific protein levels in human postmortem brain is highly desirable. Here, we developed and validated a laser capture microdissection-mass spectrometry (LCM-MS) approach to quantify over 200 proteins in cortical layers 3 and 5 of two cohorts of human subjects as well as a monkey model of postmortem interval. LCM-MS was readily implementable and reliably identified protein patterns that differed between cortical layers 3 and 5. Our findings suggest that LCM-MS facilitates the precise quantification of proteins within individual cortical layers from human postmortem brain tissue, providing a powerful tool in the study of neuropsychiatric disease.

Fig. 1

Study overview. a Coronal blocks containing area 46 of the PFC were obtained from 10 male human subjects. Eight DLPFC tissue blocks simulating PMIs of 0, 6, 12, and 24 h were obtained from a single monkey. b Image one of a 12 µm section of human DLPFC after harvesting by LCM. The 4.8E⁷ µm³ of layer 3 and layer 5 (4E⁶ µm² from a 12 µm section) were dissected from each human subject and each monkey PMI cortical block by LCM. Additionally, gray matter homogenates were prepared from the monkey blocks. c For each experiment (human cohort 1, human cohort 2, monkey LCM, and monkey homogenate), a pooled control was created within each experiment to control for variability in sample preparation and MS analysis and a different randomized sample order, keeping layers 3 and 5 from the same section together, was utilized at every experimental step: Addition of the [¹³C₆] lysine-labeled neuronal proteome standard, trypsin digestion, and MS analysis. d All samples were analyzed by targeted MS

Fig. 2

Unsupervised hierarchical clustering heat maps. Unsupervised hierarchical clustering was performed using Cluster 3 to organize subject/layers by peptide expression. a In human cohort 1, layers 3 and 5 segregated into two groups. b In human cohort 2, layers 3 and 5 segregated in all but one subject 1336 (PM 18.4). c In the monkey PMI cohort layers 3 and 5 segregated in all but one slab (PMI 24 h). d CALB1 log2 layer 3–5 difference in human and monkey cohorts. CALB1 levels are significantly elevated in layer 3 of both humans and monkeys, *q < 0.1