Isolation and culture of functional adult human neurons from neurosurgical brain specimens

Abstract The ability to characterize and study primary neurons isolated directly from the adult human brain would greatly advance neuroscience research. However, significant challenges such as accessibility of human brain tissue and the lack of a robust neuronal cell culture protocol have hampered its progress. Here, we describe a simple and reproducible method for the isolation and culture of functional adult human neurons from neurosurgical brain specimens. In vitro, adult human neurons form a dense network and express a plethora of mature neuronal and synaptic markers. Most importantly, for the first time, we demonstrate the re-establishment of mature neurophysiological properties in vitro, such as repetitive fast-spiking action potentials, and spontaneous and evoked synaptic activity. Together, our dissociated and slice culture systems enable studies of adult human neurophysiology and gene expression under normal and pathological conditions and provide a high-throughput platform for drug testing on brain cells directly isolated from the adult human brain.

The neuronal cultures are devoid of any detectable tumour cell contamination. Photomicrographs of neuronal cultures showing a high density of neurite-possessing cells in the absence (A-B) or presence of tumour cell contamination (C-D). Gene expression analysis of cultures devoid of tumour cell contamination using Nanostring® also showed low counts of stem cell and neural progenitor cell marker genes that were not regulated throughout our culture period (E), and the pattern of neuronal marker gene expression was similar in cultured cells to those of tissue (F). Panels G-H shows the relatively low levels of EdU + proliferating cells in the cortical cultures compared to the patient-matched tumour cultures, as well as the near absence of EdU + MAP2 + cells in the cortical cultures. Cells isolated from the adult human brain contain neurons, astrocyte and microglia. Photomicrographs illustrating neuronal cultures (MAP2 + ) that contain both GFAP + astrocytes and CD45 + microglia. Several different combination MAP2 and GFAP antibodies were trialled and all showed MAP2 only and GFAP only cells, but also those co-expressing MAP2 and GFAP (white arrows). The cultures also contained CD45 + microglia that did not colocalise with any of the neuronal markers tested. The interneuron marker, Calretinin, also labelled a large number of neurons. All images were taken from cultures at 28 DIV. Scale: 100 µm.

Supplementary Figure 5.
Cortical slice tissue contained the same neuronal markers expressed in the neuronal cultures. FFPE tissue from the same specimen as were cultured, were probed for the markers detected in neurons in vitro. The presynaptic marker, synapsin I, exhibited punctate staining onto MAP2 + processes (A-C). Another neuronal marker, NeuN and an interneuron marker calretinin were both present in the cortical specimens (D-F). Parvalbumin (G) and GAD65/67 (H) were also present and frequently co-localized onto a single cell (I). IHC studies of neuronal markers in the corresponding cortical tissue was conducted in 15 of the 37 specimens received and the representative case is shown in Figure S3. Scale bar = 100 µm, inset = 10 µm.

Supplementary Figure 6.
A comparison of the electrophysiological properties of dissociated primary human neurons with human brain slice cultures. The recorded dissociated primary neurons were categorised by age (A-D) and brain region (E-H), and assessed for RMP, AP amplitude, afterhyperpolarisation amplitude, and AP half-width, as shown in the 'Assessment Parameters' illustration. These properties were also compared to equivalent neurons recorded from brain slice cultures. Panels A-D compare the neurons recorded from the dissociated cultures and the cortical slice cultures from paediatric and adult cases. Panels E-H compare the neurons cultured from dissociated cerebellar and neocortical tissues, as well as neurons recorded from neocortex-derived slice cultures. For data with equal variance, a One-way ANOVA analysis with Tukey's multiple comparison test was employed, and where variance were different, logarithmically transformed data was analysed using a general linear mixed model. ns = p > 0.05, * = p < 0.05, ** = p < 0.01, *** = p < 0.001, and **** = p < 0.0001.

Supplementary Figure 7.
Full sized blots for the RT-PCR experiments conducted for Figure 2K on whole-brain, pericyte (negative control) and three representative neuronal cultures.

Supplementary Tables
Supplementary Table 1. Neurosurgical specimens used for the neuronal cultures. Different regions of the human cortex were obtained as part of the required neurosurgical procedure for the neurological disorders listed. 51 specimens were obtained from 49 patients with an average age of 42 ± 3.5 years.

Underlying pathology
Cortical region Number of specimens Primary brain tumours -high grade (WHO grades III-IV) GBM, anaplastic astrocytoma III