Seventy-six SD rats (Harlan) were acquired at PND 21 and acclimated to the environment over 2 days. Rats were first weighed on PND 23–25 and daily thereafter. Animals were housed, fed and maintained on a 12-hour reverse light/dark cycle as in Experiment 1. The Institutional Animal Care and Use Committee of Emory University approved the procedures used in this study and the research was conducted in an AAALAC-approved facility (Protocol # 2001801).

Injury was induced on PND 28. Rats were placed in an air-tight induction chamber with oxygen, nitrous oxide, and isoflurane gas (4% induction, 1.5% maintenance, 700 mmHg/min N₂O, and 500 mmHg/min O₂). The rats were then mounted in a Kopf stereotaxic device (model 900) equipped with a Mouse and Neonatal Rat Adaptor (model BJK-030). The animal's head was held in place by ear bars and a bite bar and anesthesia was delivered by nose cone. The rats' heads were shaved and sterilized with 70% ethanol and Betadine™ antiseptic solution. Anesthesia levels were monitored closely throughout surgery and were frequently adjusted between 400–700 mmHg/min, based on heart rate and oxygen saturation. In the adult rat CCI-injury model, a 5.0 to 6.0-mm diameter impactor is commonly used [34], but this size proved fatal in the juvenile rats. Here we found that a 4.0-mm diameter impactor induced a survivable, severe contusion injury in the younger rats.

A SurgiVet™ pulse oximeter (model V3304) was placed on the rear paw to monitor and maintain blood SpO₂ at or above 90%. A heart rate monitor with its sensor attached to the other hind paw was used to observe and maintain a rate ≥300 beats per minute. To prevent hypothermia throughout surgery a homeothermic blanket control unit (Harvard Apparatus, Holliston, MA) was used to monitor and maintain core body temperature (∼37°C). Using aseptic techniques, a midline incision was made along the scalp into the skin and fascia covering the skull, exposing the cranium and its bony landmarks including bregma (β) and lambda (λ).

An electric burr drill was used to produce bilateral craniectomies centered 2.0 mm anterior to β and 2.0 mm lateral to the midline (on each side). After craniectomy, injured animals received an injury centered on the midline. Contusions were made to the mPFC using the Impact One™ Stereotaxic CCI instrument (Leica #39463920) designed to produce pneumatic-CCI injury in small animals. The tip of the impactor (4.0 mm diameter) was moved 2.0 mm anterior to β retracted to set the new dorsal–ventral coordinate 3.0 mm ventral to the dura, and then activated to traumatize the cortical tissue. Dwell time was 500 ms and impactor velocity was 2.25 m/s at a pressure of 1.7 psi. Following lesion or sham surgery, the scalp was closed using 4-0 chromic gut sutures. Sham surgeries (n = 24) included all procedures up to and including craniectomy and the steps for closing the incision. After surgery, all rats were maintained on a heating pad, monitored closely and returned to home cages once they regained consciousness [35], [36]. All animals included in the statistical analysis regained consciousness and motor control no more than 5 min after being removed from the stereotaxic device.

Of the 76 rats at the beginning of the experiment, 5 died under surgery. Following surgery, the remaining 71 rats were assigned to an experimental condition based on their baseline data. CCI+ progesterone groups received a dose of 4 (n = 12), 8 (n = 12), or 16 mg/kg (n = 11). Progesterone (4-pregnene-3, 20-dione, Sigma) was dissolved in 22.5% 2-hydroxyropyl-β-cyclodextrin (HBC) (Sigma). Progesterone doses smaller (4 mg/kg) and larger (16 mg/kg) than 8 mg/kg were included in this dose-response study because previous work found that 8 mg/kg of progesterone was the most effective in adult and rats [33] and humans [16]. Each dose was administered at an equal volume (0.2 ml/100 gm b.wt.) relative to body weight. Vehicle-treated rats with CCI (n = 12) and vehicle-treated rats with sham surgeries (n = 12) received equal volumes of HBC relative to body weight. A total of 9 injections were administered to each animal in all experimental conditions at the following post-injury times: 1, 3, and 24 h, and 2, 3, 4, 5, 6, and 7 days. With the exception of the first dose, which was administered intraperitoneally to ensure more rapid absorption following injury, all injections were given subcutaneously [33]. It is important to note that the 8^th and 9^th injections of progesterone were tapered (one half and one one-quarter of the original dose, respectively) to avoid withdrawal effects [37], [38]. Progesterone was prepared just prior to surgery and then again on the 4^th post-injury night.

Rats were tested before surgery on rotarod, grip strength and open field performance to obtain baseline data. All post-surgery testing was delayed for 1 week, during which rats received daily vehicle or progesterone injections as described above. Animal IDs were coded to keep experimenters blind to group identity throughout behavioral testing and histological analysis.

At 4 weeks post-injury, all animals were deeply anesthetized and their brains perfused with saline, then with 10% formalin, and then extracted for histological analysis. After post-fixation in 10% formalin for 24–48 h, brains were immersed successively for 2–3 days each in 15%, 20%, and 30% sucrose solution in dH₂O, then removed and stored at −80°C until sectioned. The forebrain was cut into 20-µm sections on a cryostat and stored at −80°C on 1% gelatin coated (subbed) slides. Slides were stained in 0.1% cresyl violet solution (0.1 g cresyl violet acetate and 0.3 ml of glacial acetic acid dissolved in 100 ml dH₂O) for 10 min at 45°C and then rinsed in distilled water. Slides were then washed in 95% alcohol for 5–10 min, in 100% alcohol (2×5 min), and then in xylene (3×5 min). Sections were mounted with xylene-based cytoseal. Lesion surface area was identified and analyzed from Nissl-stained sections between 4.5 and −0.5 mm from β (see section 2.4.2. for details) using the Image J™ System (Media Cybernetics, Silver Spring, MD), a free computer-assisted image analysis program.

Following immersion in 30% sucrose, brains were stored at −80°C. Each rat brain was then removed from the −80°C freezer and placed in the cryostat at −20°C for 20 minutes before being sliced into 20-micron thick sections. Six evenly spaced anterior-to-posterior sections (4.5, 3.5, 2.5, 1.5, 0.5 mm anterior to β, and −0.5 mm posterior to β were then stained with hematoxylin and eosin (H&E), examined for evidence of CCI-induced tissue damage across groups [41], and selected for lesion cavity measurements. Stained slides were scanned using Silverfast Pathscan software (PathScan Enabler IV, Meyer Instruments, Houston TX). ImageJ™ was used to analyze the scanned images. The percentage of injured tissue from a single section was calculated by tracing the perimeter of the injury and determining its surface area, dividing this by an estimate of the total surface area of the section (taken by tracing both the remaining tissue and the estimated perimeter of the necrotic cavity), and multiplying by 100. The estimated percentage of damaged tissue from H&E-stained sections for each rat was determined by averaging the percentage of damaged tissue across six sections, which spanned 4 mm (i.e., +4.5, +3.5, +2.5, +1.5, −0.5 mm to β). In the sham group, rats were excluded because of slight but noticeable drill-induced injury during the craniectomy. We wanted to reduce any potential variance due to this factor.

A mixed factorial ANOVA for repeated measures was performed on the behavioral results and lesion reconstruction data, which were expressed as the mean ±SEM. Mean comparisons were used for post-hoc analyses of repeated ANOVAs. Independent paired t-tests were also used to compare the differences between baseline (pre-injury) and post-injury data when data was normally distributed. Statistical significance was established at p≤0.05. Based on a delta-value of 1.5 (using latency to find the hidden platform during Run 1 on the MWM task) we calculated the sample sizes and power needed to reject the null hypothesis (of no differences among the treatment groups relative to untreated CCI-injured controls) to achieve >80% power to detect a 30% difference (medium effect). The number of rats per group at these criteria was determined to be at least 9.

Briefly, the six sections from each rat brain were photographed with an Epson scanner (and ImagePro™ software). Group means of percent of damaged tissue averaged across six anterior-posterior sections were used to determine overall lesion size. Sham rats with a visible drill-induced lesion (n = 4), were excluded from behavioral statistical analysis. Tissue sections from the remaining sham rats had no observable damage. Results from the reconstruction analysis of sham and CCI-injured lesions treated with vehicle or progesterone are shown in Figure 1. There was a significant main effect of Lesion/Treatment on estimated lesion size (F(5,45) = 30.31, p<0.05). Post-hoc comparisons further revealed significant differences between sham and CCI (p's<0.05). Finally, while significant differences were observed between progesterone and vehicle-treated CCI rats (p<0.05), no differences were detected between 4-, 8- and 16-mg/kg doses of progesterone (p>0.05).

Immediately after surgery, we documented the extent of hemorrhage, hematoma, and herniation as a result of craniectomies and/or CCI. We found that the group differences in these parameters did not seem to predict injury severity or progesterone effectiveness on recovery of functions regardless of the doses used. We did not see any sustained (>3 days) effects of injury or treatment on body weight in our animals. With respect to metabolic activity such as weight gain and activity, the young rats in our experiment appeared to be much more resilient to CCI than can be expected in more mature animals with the same type and extent of damage.

The midline CCI injury to the mPFC in one-month-old rats produced a cavity highly consistent with a moderate contusion seen in adult rodents [35], [43]. In the Hoffman, et al. study [35], CCI injury to the mPFC in the adult male SD rat produced lesion cavities that, at 18 days post-injury, were largest (≥4 mm²) at the most anterior (+4.7, +3.7 and +2.7 mm to β) sections we examined for neuronal lesion volume.

We found that progesterone at 4, 8, and 16 mg/kg reduced infarct cavity size (by ∼25%), despite variable functional effects. Our vehicle-treated CCI PND 28 rats had >20% damage, while the progesterone-treated CCI groups had <15% damage. Roof and colleagues [44] were the first to demonstrate that in adult male and female SD rats, seven injections of 4 mg/kg progesterone could significantly reduce edema and increase neuronal cell counts in the lateral and medial dorsal thalamic nuclei [44]. Goss and colleagues [33] later demonstrated that post-injury treatment with 8 and 16 mg/kg, as opposed to 32 mg/kg, doses of progesterone in adult male SD rats were optimal for functional recovery after injury to the mPFC [33]. In Roof's study, despite improvement in MWM performance, rats given 4 mg/kg progesterone or vehicle had injuries that represented 46.07% and 40.89% of the total coronal section, respectively [44]. In Goss's study, despite the ability of progesterone to reduce bilateral somatosensory neglect and improve MWM and EPM performance, at PID 25, 20% of the tissue section was damaged regardless of treatment (vehicle or progesterone). Progesterone did not reduce lesion size in these experiments.

It is tempting to conjecture that progesterone's effect on lesion cavity size is age-dependent, but it should be noted that almost half (4 out of 9) of the vehicle-treated CCI rats had relatively smaller cystic infarcts (2–3 mm in diameter). Importantly, the behavioral deficits in these animals did not differ from those of the other rats in the same group. Future studies will need to determine whether the reduction in lesion cavity size can be explained by age or gender/hormonal dependency on neuronal sparing or aberrant sprouting in response to pediatric brain injury. Notably, in non-injured rats, the 16-mg/kg dose of progesterone suppressed the distance travelled in the Open Field test and decreased swim speed in MWM learning. This is not surprising, since relatively high doses of progesterone can be used as an anesthetic and anxiolytic in human and animal subjects 29 [45], [46].