Experiences and observations presented are based on experiments performed using ddY (10–12 weeks, 21–26 g, Kyudo Co Ltd., Japan) as well as C57BL/6N (10–12 weeks, 21–26 g, Kyudo Co Ltd., Japan) mice.

Standard microsurgical instruments including a bipolar forceps were used. The operating microscope had a 7- to 45-fold magnification (Arms Systems Co. Ltd., Japan). Mice were anesthetized with isoflurane (Escain, Mylan Co. Ltd., Japan): 5 % induction, 1.5 % maintenance, using a facemask. A thermostatically regulated, feedback-controlled heating pad (NS-TC10, Neuroscience Inc., Japan) was used to maintain body temperature at 37.5 °C. Intracranial pressure was measured using a fiber-optic micro pressure transducer (Samba Sensors AB, Goteborg, Sweden). In prone position, the tip of the sensor was placed on the left scull base.

The mice were placed in a supine position. A midline incision was made in the neck. The underlying subcutaneous tissue was separated in a blunt fashion and the submaxillary glands were gently pushed aside. A triangle formed by the sternohyoid/omohyoid, sternocleidomastoid, and posterior belly of the digastric muscle was visualized (Fig. 2a). By dissecting the triangle, the neurovascular bundle consisting of the common carotid artery (CCA), vagal nerve, and internal jugular vein was visualized in the depth (Fig. 2b). The CCA was isolated, marked, and gently lifted with a holding thread. The CCA was traced cranially until the bifurcation. At the bifurcation, the occipital artery (OA)—the first branch of the external carotid artery (ECA)—could be seen beside the internal carotid artery (ICA). The ECA was traced in a cranial direction after the digastric muscle was pushed laterally (Fig. 2b). The ECA was traced until its further branching, preserving the ECA segment as long as possible. After a ligation, the ECA was coagulated and cut. Then, the OA was sacrificed. Afterwards, the ICA, which runs in an inferior-medial direction (point of view), was traced. The removal of the carotid body might be necessary because it might hamper the view. Next, the pterygopalatine artery (PPA) arising from the extracranial section of the ICA was identified (Fig. 2b). The tiny branches from the glossopharyngeal nerve lying on the ICA/PPA bifurcation were gently pushed cranially to gain an adequate view. Once the paths of the ICA and PPA were visible, the PPA was sacrificed (Fig. 2c). If the PPA was not sacrificed, the filament tended to take a via falsa toward the PPA instead of the ICA. A ligation was prepared on the most proximal part of the ECA stump. Mini clips (Roboz Surgical Instrument Co., Inc., USA) were placed at the ICA and CCA. An arteriotomy was performed on the distal part of the ECA stump (Fig. 2c). A 5-0 nylon filament (0.1-mm diameter, Ethilon, Ethicon Inc., USA) was pushed through the arteriotomy toward the ICA (Fig. 2d) until the tip reached the occluded part by the distal clip. One might consider rotating the ECA stump in the inferior direction before inserting the filament, so that the path for the filament would be straight. However, the tip of the filament can get stuck at the bifurcation while it is being advanced. Therefore, we inserted the filament at a more “physiological” angle ≤90° (Fig. 2d). Once the tip overcame the bifurcation, the stump with the filament was rotated inferiorly (Fig. 2e). The clips were removed and the filament was pushed further cranially. After removal of the clips, retrograde bleeding might occur. Gentle tightening of the prepared ligation avoids this. This inconvenience might be further avoided if the proximal clip was left in place and removed at the end of the procedure, i.e., after the perforation and complete withdrawal of the filament and ligation. However, accumulated experience revealed that the latter maneuver decreases the degree of success and volume of SAH, possibly because of rapid activation of the coagulation cascade. Therefore, we recommend the procedure as shown in Fig. 2g. The filament was gently pushed forward (~5 mm) until some resistance was felt. The filament was pushed an additional ~1 mm further to perforate the vessel. The filament was withdrawn quickly and the prepared ligature tightened (Fig. 2f). After ensuring hemostasis, the wound was adapted and sutured.

To evaluate the degree of hemorrhage, animals were killed at day 0, 1, and 2 after SAH. The brains were quickly removed and photographed under the operating microscope. For histological examination, euthanasia was performed by transcardial perfusion–fixation at day 2 after SAH. The brains were removed, postfixed and paraffin embedded. The blocks were sliced into 7-μm sections using a microtome (Leitz 1512 Microtome, Wetzlar Germany). Hematoxylin and eosin (HE) staining was performed for overview and CVS identification. For the assessment of neuronal injury, Fluoro-Jade C (FJC) staining (Merck Millipore, USA) was performed as previously described [21]. To evaluate the occurrence of apoptotic neurons, terminal deoxynucleotidyl transferase dUTP nick labeling (TUNEL) staining was performed (Roche Diagnostics, Germany). The numbers of FJC- and TUNEL-positive cells were determined in randomly selected regions of interest in the hippocampal region and cerebral cortex. Because microthrombosis has been discussed more recently as an additional explanation for neuronal injury after SAH, we investigated the occurrence of microthrombosis by immunohistochemistry (sheep anti-fibrinogen antibody; LifeSpan, BioSciences, USA).