**3. Results**

## *3.1. Seizure Score*

Twenty-four rats were used for PTZ kindling, from which 18 ended up reaching a kindled state and, due to this, were each consecutively tested with a challenge dose of PTZ. The remaining six animals (*n* = 5 in the kindled and *n* = 1 in the CBD-treated group) died before completing the study, so the behavioral assessment was performed on *n* = 7 kindled and *n* = 11 CBD-treated animals. The mortality rate tended to decrease following chronic CBD treatment (5/12 vs. 1/12, Chi square = 3.429, *p* = 0.064, Figure 3a), especially in the early phase of the kindling.

Analyzing the parameters characterizing the development of the kindling process, it was observed that the total number of PTZ injections required to achieve a kindled state did not show a significant difference between CBD-treated and control groups (median 15.5 vs. 16, *p* = 0.833, Figure 3b). The seizure scores registered during kindling were analyzed using the mixed-effect model with Geisser–Greenhouse correction, and it was noted that the time factor was significant in the development of generalized seizures (Racine 4 and 5 seizures) for both groups (F (1.667, 16.39) = 53.63, *p* < 0.001, Figure 3c). However, treatment and the time x treatment interaction were not significant (F (1, 15) = 2.397, *p* = 0.142 and F (6, 59) = 1.174, *p* = 0.333, respectively). Furthermore, there was no difference in the duration of generalized seizures between the CBD-treated and control groups (74 ± 25.7 vs. 89 ± 40 s, *p* = 0.425). Conversely, the mean (± SEM) latency to first generalized seizure was significantly longer in the CBD-treated group (925.3 ± 120.0 vs. 550.1 ± 69.62 s, Figure 3d), and the mixed-effects analysis confirmed that treatment had a significant influence on this parameter (F (1, 15) = 6.3872, *p* = 0.023).

**Figure 3.** Effects of chronic CBD treatment on the kindling scores induced by pentylenetetrazole in rats. The development of kindling was characterized by the following parameters: (**a**) mortality rate due to generalized seizures, (**b**) the total number of PTZ injections to reach the kindled state, (**c**) the number of generalized seizures, and (**d**) the latency to the first generalized seizure. Data are expressed as mean ± SEM in the bar graph and as mean (solid line) with the range in the floating bar graphs; \* *p* < 0.05 vs. control.

The challenge dose of PTZ induced the same seizure pattern in both groups, and all animals had Racine 5 seizures that did not differ in duration between groups (127.4 ± 63.58 vs. 131.2 ± 31.53 s, *p* = 0.312). However, the latency to maximal seizure in the CBD-treated group showed high variability among animals (777.8 ± 228.4 vs. 651.1 ± 220.0 s, Figure 4a) with no significant differences between groups (t(15) = 0.397, *p* = 0.697).

#### *3.2. CBD Plasma and Brain Concentrations*

To verify whether the in-house-prepared CBD solution administered by the oral route would achieve the desired concentrations at the site of action, serum and brain CBD concentrations were assessed at 1 h and 24 h after administration, corresponding to Cmax and Cmin, respectively. The mean peak concentration was 1976.1 ± 1151.41 ng/mL in serum and 5260 ± 3284 ng/g in brain, while the minimum concentrations were 24.2 ± 3.25 ng/mL and 91.4 ± 20 ng/g, respectively (Table 1). Brain-to-plasma ratios calculated for each animal showed higher values at 24 h compared to 1 h after administration, suggesting a slower brain CBD-elimination rate.

**Figure 4.** Lack of long-term effect of CBD (60 mg/kg) in the PTZ-kindling model. (**a**) Latency to maximal seizure observed after PTZ injection. (**b**) Duration of the generalized seizures (clonic convulsion with loss of righting reflex and/or bouncing, two or more clonic convulsions, tonic convulsion or status epilepticus) induced by a challenge dose of PTZ. Data are expressed as mean ± SEM (*n* = 7–11) for each parameter.

**Table 1.** Cannabidiol concentrations in plasma and brain at Cmax and Cmin after the oral administration of 60 mg/kg body weight.


\* As a result of inadequate tissue perfusion, the brain tissue sample was excluded from analysis. # Data points are labeled by rat number.

#### *3.3. Open Field Test*

Chronic CBD treatment had no effect on exploratory and locomotor parameters in the open field test. The control animals apparently entered more times in the center zone of the arena (7.4 ± 15.1 vs. 3.43 ± 5.62, *p* = 0.77, Figure 5a), but they spent the same amount of time exploring it (17.69 ± 4.48 vs. 18.15 ± 2.49, *p* = 0.92, Figure 5b). The distance travelled and the vertical exploration expressed as supported rearings did not vary between groups (*p* > 0.05) (Figure 5c,d).

**Figure 5.** Assessment of locomotor activity in the open field. Rats were submitted to a PTZ-kindling protocol with or without chronic cannabidiol (60 mg/kg body weight) administration (mean ± SEM, *n* = 18). (**a**) The number of entries in the center zone; (**b**) time spent in the central zone of the arena, defined as a 30 × 30 cm square out of an 60 × 60 cm total surface of the arena; (**c**) the vertical exploration of the animals was expressed as the time spent leaning on the walls of the arena; (**d**) total distance moved during the 5 min testing session. CBD, cannabidiol. Data are expressed as mean ± SEM (*n* = 7–11) for each parameter.

#### *3.4. Novel Object Recognition Test*

There was no effect of CBD treatment on the total distance traveled and on the exploration time of the arena or objects. Conversely, the mean discrimination index was decreased by CBD treatment (0.57 ± 0.15 vs. −0.01 ± 0.17, *p* = 0.0367, Figure 6). The results indicate that CBD-treated animals were impaired in distinguishing between the novel and familiar objects.

**Figure 6.** The effects of chronic CBD (60 mg/kg body weight) treatment on cognitive performance of rats subjected to a PTZ-kindling protocol. (**a**) Discrimination index (DI), which shows the discrimination between the novel and familiar objects, i.e., the difference in exploration time for a familiar object, but then dividing this value by the total amount of exploration of the novel and familiar objects [DI = (TN − TF)/(TN + TF)] showed a significant decrease in the CBD-treated group. + sign shows the arithmetic mean, whereas the line represents the median; (**b**) the total exploration time of the arena (mean ± SEM) did not show statistically significant difference between groups; (**c**) the total exploration time of the objects (mean ± SEM) was also similar for both groups; (**d**) the total distance moved by both groups (mean ± SEM) during the 5 min testing session was almost equal, and the results were similar to those found in the OF test. CBD, cannabidiol; \* *p* < 0.05 vs. control. Data are expressed as mean ± SEM (*n* = 7–11) for each parameter.

## *3.5. Fluorescent Immunohistochemistry*

#### 3.5.1. Validation of Manual and Algorithmic Cell Counter Strategies

Neurons, astrocytes, and microglia were triple-immunostained in the hippocampus, where the CA1 and CA3 regions were analyzed (Figure 7). Sham, i.e., healthy control (*n* = 6), PTZ-kindled control (*n* = 5), and CBD-treated (*n* = 8) groups showed different pattern of immunostaining for astrocytes labeled with anti-GFAP antibody (cyan), for neurons labeled with anti-NeuN antibody (green), and for microglia labeled with anti-IBA1 antibody (red).

**Figure 7.** Neuron, astocyte, and microglia triade. Representative confocal microscopy image of the tripple immunostaining of the (**a**) neuron-NeuN, (**b**) astrocyte-GFAP, and (**c**) microglia-IBA1 in the CA3 subfield of the hippocampus.

We used a custom cell-counter algorithm to quantify the cells of interest. The manual counting method was used as the reference method. The mean of the two observers was compared with the values provided by the algorithm. For the statistical analysis the bias, precision and limits of agreemen<sup>t</sup> were used as proposed by Bland and Altman to compare two methods of measurement. The methods were considered to be interchangeable if the bias was smaller than ±10% for each cell type and if there was no tendency of the difference to increase with the mean (Supplementary Figure S4).

#### 3.5.2. Effects of CBD on the Neuron–Astrocyte–Microglia Triad in the Hippocampus

The thickness of the stratum pyramidale showed differences between groups, a twoway ANOVA analysis showing a significant impact of CBD treatment (F (2, 119) = 3.963, *p* = 0.02). The PTZ-kindling procedure decreased the stratum pyramidale thickness compared to sham animals, and CBD treatment induced a further significant thinning in the CA3. A similar tendency could be observed in the pyramidal layer of the CA1 region as well, but it did not reach significance (Figure 8a, Supplementary Figure S5).

The cell density of GFAP-positive astrocytes was decreased by CBD treatment (F (2, 119) = 6.546, *p* = 0.002,) both in the CA1 and CA3 regions of the hippocampus (Figure 8b). The post hoc analysis confirmed that the CBD-treated group had significantly decreased astrocyte density in the CA1 region (*p* = 0.045). Conversely, neither the PTZ-kindling nor the CBD treatment had any effect on the total number of IBA1-positive cells (F (2, 119) = 0.2522, *p* = 0.78, Figure 8c).

Neuronal debris, i.e., fragments of NeuN positive cells closely attached to the branches of astrocytes in the stratum radiatum, was observed in both the CA1 and the CA3 regions. Despite the fact that PTZ-kindling did not influence the density of neuronal debris when compared to sham animals, the CBD-treated group had significantly decreased neuronal debris in the CA3 region of the hippocampus when compared to PTZ-kindled controls (Figure 8d, *p* = 0.0359).

**Figure 8.** Effects of chronic CBD (60 mg/kg body weight) treatment on the hippocampal reorganization induced by PTZ-kindling in rats. (**a**) Mean thickness of the pyramidal cell layer; (**b**) the density of GFAP-positive astrocytes; (**c**) the density of IBA1-positive microglia; (**d**) the density of neuronal debris, defined as NeuN-positive fragments surrounded by astrocytes. Data are expressed as mean ± SEM (*n* = 5–8).; CBD, cannabidiol; \* *p* < 0.05; \*\* *p* < 0.01.
