Technical report: surgical preparation of human brain tissue for clinical and basic research

For electrophysiological experiments, cold (7–10°C) N-methyl-D-glucamine-based artificial cerebrospinal fluid (aCSF) saturated with carbogen (95% O₂, 5% CO₂) was used as described in the literature [15, 38]. Larger (10–30 mm) samples used for organotypic slice cultures and transcriptomic experiments were transferred to sterile ice-cold carbogenated “preparation medium” (Hibernate-A Medium (Gibco) supplemented with 1 mM GlutaMAX (Gibco), 13 mM Glucose (Sigma-Aldrich), 30 mM NMDG (Sigma-Aldrich), and 1% Anti-Anti). Depending on the distance from the OR to the laboratory facilities, the transport time ranged from 5 to 15 min. For a transport time >5 min, the medium was carbogenated using a nitrile examination glove filled with carbogen (95% O₂, 5% CO₂) and connected to a 3-way valve (Luer/Lock) and perfusor line (B. Braun Melsungen AG) (Supplementary Fig. 1). A single glove (size XL) lasted up to 30 min.

The protocols for the recovery of acute human brain slices, electrophysiological whole-cell patch-clamp recordings, post-hoc visualization of neurons, and immunohistochemistry were performed as previously described [15]. Light microscopic images were acquired using a Leica SP8 laser-scanning confocal microscope equipped with a 40× oil-immersion (NA 1.30; Leica) objective.

Acute 400 μm slices were immersed in EM fixative (4% Paraformaldehyd (Polyscience) (w/v) 2.5% Glutaraldehyd (Serva) (w/v) in 0.1 M phosphate buffer) and kept at room temperature for 2 h, then stored at 4 °C. Next, ~1.5 mm wide samples were trimmed spanning the entire cortical height. Staining and embedding of samples was then performed as described in the literature [15]. Electron micrographs were taken with a Philips CM100 transmission electron microscope equipped with a Gatan Kamera Orius SC600 with a magnification of 6600.

After transport of the samples to the laboratory, they were further divided into blocks ranging from 1 to 40 mm depending on the experimental use case (Table 2). Dissection was performed using sterile tools on a cutting platform consisting of aseptic filter paper (Whatman Merck, UK) moistened with sterile “preparation medium” (see above) and placed on an ice-cold metal block (−20 °C) on a sterile surgical drape. Visibly damaged or pathological tissue and capillaries were removed. For slice culture experiments, tissue blocks (7–20 mm) were cut into 300-μm-thick sections in a disinfected vibratome (Leica VT1200 Wetzlar, Germany) and filled with cold, sterile carbonated preparation medium. Depending on the apparent stiffness of the tissue, the cutting speed ranged from 0.14 to 0.16 mm/s with an amplitude of 1.5 mm. Immunofluorescent labeling was performed, and ex vivo glioblastoma invasion models were generated as described in the literature [7, 23] by injecting primary cultured GBM cells that were virally tagged with zsGreen into human organotypic slice cultures.

Small 1–4 mm large tissue and tumor samples were removed using a 2-mm cutting forceps (8591A; Karl Storz, Tuttlingen, Germany). The samples were then compressed in a custom microscope slide according to the manufacturer’s instructions and imaged in a clinical stimulated Raman scattering microscope (NIO Laser Imaging System, Invenio Imaging, Santa Clara, CA) [21].

Rectangular incisions (~5–7 mm) were made perpendicular to the pial surface using a pointed scalpel (Fig. 1d). Incision should be placed at the top of gyri to avoid entering sulci where larger blood vessels are found to prevent bleeding complications. To avoid compression and shear of the tissue, serrating movements of the blade were applied (see Supplementary Video 1). A narrow dissector was introduced from one side, and the white matter under the sample was carefully dissected while causing minimal mechanical disturbance to the cortical part of the sample (Fig. 1c). It was helpful to angulate (~45°) (Fig. 1a) one side of the rectangular incision to allow the introduction of the dissector and minimize compression of the surrounding tissue during preparation. The tissue block was lifted using the dissector and flipped onto a brain spatula (Fig. 1c). The tissue and the brain spatula were immediately immersed into a non-sterile container filled with transportation media and removed from the operating field and transported to the internal (5 min) and external laboratory facilities (15 min).

This preparation technique is optimized for the preservation of superficial layer cortical neurons. Deep layers of neurons, as well as the underlying white matter, are at risk of mechanical damage. An optimally prepared single block of 5–7 mm size can immediately be processed in the vibratome (Supplementary Fig. 2), minimizing delay and mechanical damage. Without the need for further trimming, the slices can be cut at slightly higher temperatures (7–11 °C), reducing the effect of low temperatures on ultrastructure and electrophysiological properties [11]. Blood clots on the sample may lead to detachment during processing in the vibratome and are not easily removed once they are attached to the sample. Therefore, constant irrigation of the sample during surgery proved to be beneficial (Table 1).

Using this technique, 36 brain tissue samples, of which 8 were reported previously [15], were obtained from frontal (n = 14), parietal (n = 2), temporal (n = 16), and occipital lobe (n = 4). Twenty-nine samples were from tumors and 7 from epilepsy surgeries. Eight to ten brain slices can result from one single block of tissue. Under ideal conditions, up to 20–70 cells could be recorded from a single sample.

The same surgical steps as described above were applied for the microdissection of larger (10–30 mm) brain tissue samples. In larger samples, subcortical structures remained intact (Fig. 2). After transfer to the laboratory in preparation medium, larger samples were divided into blocks of various sizes according to the intended experimental use case (Table 2). Dissection was best performed in an ice-cold preparation medium, taking care to discard macroscopically infiltrated and tumorous tissue (Fig. 2a). For the generation of tumor invasion models [7], a certain portion (3–5 mm) of the white matter was preserved (Fig. 5). As previously shown in the literature [23], a single block of the size of 7–20 mm resulted in 18–20 organotypic slice models that could be cultivated for up to 12 days (n = 25).

Stimulated Raman histology (SRH) [8, 21, 22, 35] allows the intraoperative label-free histological evaluation of tumorous samples. These samples were the least sensitive to mechanical damage, as the preparation itself consisted of a squash preparation. However, cauterization and the use of ultrasonic aspirators (CUSA) should be avoided. We found an endoscopic biopsy forceps optimal for targeted sampling of tumor and infiltrated brain tissue [21] (Fig. 6b). Samples sized 1–4 mm were sufficient for SRH imaging (Fig. 6b, c). The ease of use and the standardized imaging conditions make intraoperative SRH also an ideal tool to assess the quality and tumor burden of unstained neocortical access tissue slices (Fig. 6e–g).

Case 1: Figure 3 shows a whole-cell patch-clamp recording of a superficial layer 2/3 pyramidal neuron from a 73-year-old male patient undergoing surgery for the resection of an IDH wild-type glioblastoma in the left frontal lobe (Figs. 1d and 3a, Supplementary Video 1). The action potential firing patterns in response to a current injection as well as spontaneous excitatory synaptic currents demonstrate the viability of the tissue (Fig. 3b). Two-layer 2/3 pyramidal neurons from the same sample were recorded and visualized with a streptavidin-coupled fluorescent labeling (Fig. 3c). The recordings were performed in a laboratory situated 10 min away from the neurosurgical operating room [15].

Case 2: Figure 4a shows the case of a 19-year-old female patient who underwent the resection of a diffuse astrocytoma CNS-WHO grade 2 (IDH mutated) in the right supramarginal gyrus during awake surgery. Electron micrographs generated from an acute slice of a small block of the tissue showed a high quality of ultrastructural preservation that allows the analysis of synapses as well as intracellular organelles like the spine apparatus organelle (Fig. 4b) [29].

Case 3: Figure 5 shows an ex vivo glioblastoma invasion model based on organotypic slice cultures generated from the access tissue of a 59-year-old male patient who underwent surgery for the removal of a glioblastoma IDH wild-type in the right frontal lobe. The slice cultures were subcortically injected with fluorescently labeled GBM cells, and the growth was monitored for 10 days, highlighting the interaction with astrocytes of the slices.

Case 4: Figure 6a–d shows the case of a 57-year-old female patient who underwent surgery for a recurrent CNS-WHO grade 4 astrocytoma (IDH mutated) in the left occipital lobe; 7 months prior to surgery, radiation therapy was performed. A 2–3 mm sample of the tumor was removed using biopsy forceps and stimulated Raman histology was performed in the OR.

Case 5: Figure 6e–g shows the neocortical access tissue of a 60-year-old female who underwent resection of a left temporal metastasis of an adenocarcinoma. Stimulated Raman histology was used to assess tissue quality and to rule out tumor infiltration in a slice of access tissue.

To ensure a consistent and ethically sensitive approach, in the absence of standard guidelines on this emerging field of translational research, the pre-, peri-, and postoperative procedures should be formalized in a research protocol and ethics approval for this protocol should be sought by the local research ethics committee. To ensure consistency in the ethical standards for such procedures, an international and participatory consensus process should be initiated, e.g., organized by professional neurosurgery associations.

The technique outlined here allows to maximize the scientific gain of every cortical and subcortical tissue donation, leading to fewer sample numbers required to answer a range of questions in translational neuroscience. The combination of several techniques maximized the use of each individual sample. We encourage to set up an experimental protocol and biobanking infrastructure that allows the exchange of tissue and data among research institutions. With emerging technologies, the benefit of analyzing human tissue with a multitude of technologies is paramount.

Human cortical access tissue has to be considered pathologically altered. The alterations arise from sample removal, sample processing [11], and the pathology (tumor, epilepsy) leading to the operation. The effects of the underlying pathology on cortical access tissue may range from tissue edema and histopathologically diagnosable tumor infiltration to subtle changes in the ultrastructure [5].

The macroscopic pathological effect can be controlled by measuring the distance of the cortical access tissue to the pathologies (contrast-enhancing border or FLAIR signal) [23] and microscopically using histological methods (e.g., immuno/H&E stain, stimulated Raman histology [22, 26, 35]) to assess tumor infiltration (Fig. 6e–f). Nonetheless, subtle changes will most likely only be discernible by exploration of the electrophysiological [9] and ultrastructural properties [16] across different samples, under the condition that the artifacts introduced by tissue microdissection and processing are reduced to a minimum.

The resection of neocortical tissue blocks and subsequent slice preparations involve a severing of long range and local axons as well as dendrites, thus preserving only parts of the neocortical circuitry [18]. Subcortical neuromodulatory inputs, which are essential for physiological function in vivo, will not be preserved ex vivo, a fact that has to be taken into consideration in the experimental design. In animal models, a solution to overcome the limitation of acute slice preparations includes a combination of in vivo recordings and post-hoc morphological reconstructions across entire brains [20, 28].

A standardized sampling technique that reliably produces high-quality human brain tissue specimens forms the basis for the investigation of unique neurophysiological features in individual patients. Such characteristics can, for instance, include comorbidities such as neurodevelopmental and neurodegenerative disorders that can lead to or arise from alterations of neuronal wiring and plasticity [39, 43], or the intake of pharmacological agents. Furthermore, it will be possible to study the effect of interventions on the human brain, such as invasive and non-invasive brain stimulation techniques (e.g., TMS [4]), radiation [27], or tumor-treating fields [36]. A standardized sampling technique is therefore of utmost importance for current and future basic and clinical research and treatment development.

Improved tissue quality will also have an impact on pathological diagnosis. Although neuropathological diagnostic methods are moving toward molecular diagnostics that are also performed with lower-quality tissue [3], the interpretation of tumor infiltration in intact neuropil might greatly benefit from optimally preserved tissue. Further novel intraoperative imaging and spectroscopic techniques (SRH, FTIR) [8, 21, 22, 26, 34] that permit close to real-time histological examination of brain tissue in the operating room (Fig. 6) may influence surgical decision-making in the future [35]. Thus, optimally preserved tissue is essential for the validation of anatomical, molecular, and physiological findings of methods for intraoperative tissue analysis.