GFAP image analysis protocol. Coronal rat brain sections were analyzed with a computer-based system for their relative neocortical (left, experimental, compared to right, control) area of GFAP staining. Staining was measured on a linear gray scale (illustrated by two upper rows of blue color-coded images) from a darkest value threshold, where GFAP visualization began (first row, first image) to a lightest value threshold, where GFAP visualization reached saturation in the lighter-stained neocortex (second row, third image). This gray scale range was divided into six equally sized increments (shown as an accumulative increase in blue area above and with corresponding individual incremental areas in the two lower rows). To enhance visualization, six incremental areas of GFAP staining were color coded (two lower rows). Beginning with darkest shading (third row, first image; shaded with dark brown and reflecting the most intense GFAP staining), each progressively less intense incremental area was shaded a different color with the lightest intensity of GFAP staining shaded orange (fourth row, third image). Most of the GFAP area visualized occurred within the first three lighter zones. These individual areas were given a relative weighting factor to reflect their differences in staining intensity: The darkest area was multiplied by 6, and each progressively lighter increment was multiplied by either 5, 4, 3, 2, or 1, respectively. Then weighted areas were added together, and the resultant sum was normalized by dividing this value by the total area measured for a given cortical side. Finally, experimental-side values were divided by the corresponding control-side values, and these ratios were averaged for the nine sections measured for each animal. Thus, a single number was generated for each animal, which gave a weighted, normalized, and relative estimate of the area of GFAP staining in the experimental neocortex as compared to the control cortex.

Hematoxylin- and eosin-stained brain sections. KCl induced coagulation necrosis, as indicated by wedge-shaped area of reduced staining (upper left). Most neurons within this area showed shrunken eosinophilic cytoplasm and pyknotic nuclei (center left, arrow). Nuclear fragmentation was evident, and nuclear debris was scattered in this zone. NaCl, on the other hand, never produced coagulation necrosis (upper right). Instead, only “dark” neuronal changes are evident (center right, arrow). This change is consistent with reversible neuronal injury (see Duchen, 1984). KCl induced recurrent SD in the left hemisphere, and no evidence of reversible or irreversible neuronal injury is seen; section from left frontal cortex is shown (lower left). For comparative purposes, a right frontal section is also shown (lower right). No evidence of neuronal injury is evident. Both images show normal histological features of frontal neocortex. Scale bars: 500 μm, upper two micrographs; 50 μm for remaining micrographs.

Effect of a combined increase in O₂ and CO₂ on SD- and KCl-induced increase in GFAP staining. The dramatic reduction of the increase in GFAP staining from SDs caused by exposure to elevated O₂ and CO₂ is further emphasized by a nearly similar reduction in GFAP staining around the necrotic zone where KCl was applied. The full nine brain sections (three upper rows) from the animal whose results are shown in Figure 3 are presented. In this experiment, 25 SDs were seen in the left neocortex, and none in the right. Notice that GFAP staining is more intense in all left neocortices (darkest color indicates darkest staining). In addition, the three images of the third row show changes from application of KCl to the left cortex. Notice that tissue is missing from the uppermost aspect of the left parietal cortex in the first and second images of the third row, and GFAP staining is less intense in a focal distribution in the same area (second and third images of third row). These latter changes reflect necrotic brain tissue induced by application of KCl. The three bottom rows show all nine sections from the results presented in Figure 4. Here, the animal was pretreated with increased inhaled O₂ and CO₂ before KCl was applied to the left, and NaCl to the right, parietal cortex for 3 hr. Notice that, though the corresponding KCl-application zone (all images of sixth row) is less destroyed in this example, a nonsignificant (p > 0.10), 10% increase in GFAP staining is evident on the left compared to the right cortex when only the last two brain sections (n = 6) are analyzed. Thus, increased O₂ and CO₂ provided some protection against any increased GFAP staining from KCl-induced brain necrosis.

Bright-field photomicrographs of GFAP staining. Image from a frontal coronal section showing GFAP staining in experimental (left) and control (right) neocortices at low power (top). KCl was applied to the left parietal cortex and NaCl to the right parietal cortex for 3 br while DC signals were simultaneously monitored from both frontal cortices. No SD occurred in the right, while in this example, 37 SDs occurred in the left neocortex. Animals were allowed 10 recover for 48 hr and then were processed for GFAP visualization by a peroxidase-anti-peroxidase staining technique. Coronal section from normal animal is shown for comparison (bottom). These images were used to produce color-coded images in Figure 3.

Color-coded GFAP staining profile and evidence for unilateral recurrent SDs from a single animal. The top color-coded image shows relative GFAP staining intensities (color bar is described in Fig. 2; darkest color indicates darkest staining). Left neocortex is shown on left. Associated DC signals are shown in the middle traces. Twenty-five SDs occurred in the left neocortex, while no SDs occurred in the right neocortex. Hence, recurrent SD produces a significant (p < 10⁻⁴), 43% increase in GFAP staining compared to control contralateral cortices, which experienced no SDs (n = 8). For comparison, the bottom image shows color-coded GFAP staining from a normal animal.

The effect of a combined increase in inhaled O₂ and CO₂ on GFAP staining and SD. After electrodes were implanted into frontal cortices to record DC signals, inhaled gasses were first altered to 100% O₂ and then, 5–10 min later, to include a progressively increasing amount of inhaled CO₂. CO₂ was increased and then adjusted to keep arterial CO₂ tension between 100 and 120 torr. Inspired halothane remained constant at 1.0–2.0% of inhaled gasses. KCl was then applied to the left and NaCl to the right parietal cortex for 3 hr. No SDs were seen in this example (bottom). Color-coded GFAP staining is shown in the lop (darkest color indicates darkest staining). Notice that in this example there is less intense GFAP staining in the experimental neocortex (left neocortcx is on left) exposed to KCl than in the right neocortex exposed to NaCl. However, in six animals exposed to this modification of inhaled gasses, a nonsignificant (p > 0.20), 7% increase in GFAP staining was evident in left as compared to right neocortices.

Toluidine blue- and GFAP-stained thin sections showing astrocytic hypertrophy after SD. The three upper images show thickened and more extensive GFAP processes of hypertrophied astrocytes from cortex that underwent 25 SDs. The three lower images show astrocytes from contralateral neocortex. The latter cells are smaller and have less shorter and thinner processes than those from the SD hemisphere. Scale bar, 50 μm.