Progesterone treatment normalizes the levels of cell proliferation and cell death in the dentate gyrus of the hippocampus after traumatic brain injury

Abstract

Traumatic brain injury (TBI) increases cell death in the hippocampus and impairs hippocampus-dependent cognition. The hippocampus is also the site of ongoing neurogenesis throughout the lifespan. Progesterone treatment improves behavioral recovery and reduces inflammation, apoptosis, lesion volume, and edema, when given after TBI. The aim of the present study was to determine whether progesterone altered cell proliferation and short-term survival in the dentate gyrus after TBI. Male Sprague-Dawley rats with bilateral contusions of the frontal cortex or sham operations received progesterone or vehicle at 1 and 6 hours post-surgery and daily through post-surgery Day 7, and a single injection of bromodeoxyuridine (BrdU) 48 hours after injury. Brains were then processed for Ki67 (endogenous marker of cell proliferation), BrdU (short-term cell survival), doublecortin (endogenous marker of immature neurons), and Fluoro-Jade B (marker of degenerating neurons). TBI increased cell proliferation compared to shams and progesterone normalized cell proliferation in injured rats. Progesterone alone increased cell proliferation in intact rats. Interestingly, injury and/or progesterone treatment did not influence short-term cell survival of BrdU-ir cells. All treatments increased the percentage of BrdU-ir cells that were co-labeled with doublecortin (an immature neuronal marker in this case labelling new neurons that survived 5 days), indicating that cell fate is influenced independently by TBI and progesterone treatment. The number of immature neurons that survived 5 days was increased following TBI, but progesterone treatment reduced this effect. Furthermore, injury increased cell death and progesterone treatment reduced cell death to levels seen in intact rats. Together these findings suggest that progesterone treatment after TBI normalizes the levels of cell proliferation and cell death in the dentate gyrus of the hippocampus.

1. INTRODUCTION

Progesterone, a naturally occurring steroid hormone, is present in both male and female brains. Its actions in the central nervous system are not limited to the control of neuroendocrine regulation and reproduction (Baulieu and Schumacher, 2000, Birzniece, et al., 2006); it influences several neural survival functions including neuronal and glial differentiation, hippocampal neurogenesis, and synaptic stability (Ghoumari, et al., 2005, Smith, et al., 1987, Tsutsui, et al., 2004, Zhang, et al., 2010). There are now more than 160 publications showing that progesterone is an effective neuroprotective agent against a number of different insults including blunt and penetrating traumatic brain injury (TBI) (De Nicola, et al., 2009, Garcia-Estrada, et al., 1999, Labombarda, et al., 2010, Schumacher, et al., 2007, Stein, 2007, Stein and Sayeed, 2010, Stein and Wright, 2010), diffuse TBI (O’Connor, et al., 2007), stroke (Ishrat, et al., 2010, Sayeed and Stein, 2009), anoxic brain injury, glutamate toxicity (Ogata, et al., 1993), and spinal cord injury (Labombarda, et al., 2010) in both males and females. Progesterone exerts its beneficial effects in the central nervous system via multiple pathways that are not always dependent on the activation of the classical intranuclear progesterone receptor (Bottino, et al., 2010, Falkenstein, et al., 2000, Losel, et al., 2003). These effects include, among others, protecting the blood-brain barrier, reducing cerebral edema, decreasing the inflammatory cascade, and decreasing cellular necrosis and apoptosis (Djebaili, et al., 2005, Gibson, et al., 2005, He, et al., 2004, Ishrat, et al., 2010, Majewska, 1992, Mani, 2006, Roof, et al., 1997, Smith and Woolley, 2004).

Treatment with progesterone following TBI also improves functional outcome. For example, cognitive deficits, a hallmark of TBI, are attenuated following progesterone treatment (Roof, et al., 1994). Systemic treatment with progesterone facilitates recovery of spatial learning and memory, which involves the hippocampus, in brain-injured male rats compared to placebo-treated male rats (Djebaili, et al., 2004, Roof, et al., 1994, Shear, et al., 2002). The neuroprotective properties of progesterone in response to TBI are also seen in humans. Two recently completed phase IIa single-center clinical trials found a beneficial effect of progesterone as a treatment in moderate to severe TBI in male and female adults, with fewer mortalities and better functional outcome seen in progesterone-treated patients (Wright, et al., 2007, Xiao, et al., 2008). These encouraging findings have led to two phase III double-blind randomized clinical trials now enrolling patients (ProTECT III: NCT00822900; SyNAPSe: NCT01143064).

Hippocampal neurons may be the most vulnerable to TBI-induced degeneration and are among the first population of neurons in the brain to die following experimental TBI (Ariza, et al., 2006, Kotapka, et al., 1991). The dramatic upregulation in apoptotic and necrotic cell death following TBI can contribute to cognitive deficits, and drugs that prevent cell death also reduce cognitive impairment (Sinson, et al., 1997).

Progenitor cells in the dentate gyrus of the hippocampus retain the ability to proliferate into neurons during adulthood in most mammalian species studied including humans (Eriksson, et al., 1998, Gould, et al., 1999, Gould, et al., 1997, Gould, et al., 2001, Huang and Sato, 1998, Lavenex, et al., 2000). Hippocampal neurogenesis consists of at least four processes: cell proliferation, cell migration, cell differentiation, and cell survival. Cell proliferation in the dentate gyrus is the production of new cells from the division of progenitor cells within the subgranular zone. Daughter cells migrate from the subgranular zone to the inner layers of the granule cell layer while undergoing a process of fate determination, differentiating into either neurons or glial cells (Brown, et al., 2003, Cameron, et al., 1993). These new cells continue to mature over a period of weeks to months resulting in new functionally mature neurons (Esposito, et al., 2005, van Praag, et al., 2002). Cell survival is the number of new cells that survive to maturity with the majority of these new cells being neurons (Cameron, et al., 1993).

In addition to its effects on cell death and hippocampus-dependent learning and memory, TBI also affects different aspects of adult hippocampal neurogenesis (Chirumamilla, et al., 2002, Gao, et al., 2009, Lu, et al., 2005, Richardson, et al., 2007). Different experimental models of TBI, including lateral fluid percussion and controlled cortical impact (CCI) injury, induce a 3- to 4-fold number of BrdU-ir cells in the dentate gyrus as early as 2 days postinjury that peaks within the first 7 days following injury (Dash, et al., 2001, Emery, et al., 2005, Kernie, et al., 2001, Lu, et al., 2005, Lu, et al., 2003, Rice, et al., 2003, Sun, et al., 2005, Yoshimura, et al., 2003) and returns to baseline levels by 35 days post-injury (Kleindienst, et al., 2005). However it is important to note that in the majority of these studies cell proliferation was not independently assessed as multiple BrdU injections were given and thus the number of BrdU-ir cells were an index of both proliferating cells and the short-term survival of newly-produced cells. The capacity of TBI to increase cell proliferation and subsequent neurogenesis in the dentate gyrus may not necessarily be beneficial and could actually contribute to injury-induced cognitive deficits. Some authors have suggested that too much or too little neurogenesis may interfere with proper functioning of the hippocampus, as predicted by computer modelling (Butz, et al., 2008). Indeed, prolonged and continuous seizures lead to an increase in neurogenesis which can then directly contribute to cognitive decline seen after the seizures (Jessberger, et al., 2007). Here we offer the hypothesis that the long-term beneficial effects of progesterone on cognitive outcome after TBI are associated with progesterone’s influence on modulating potentially pathological neurogenesis.

The aim of the present study was to determine the effects of progesterone treatment on cell death, cell proliferation and immature neuron survival in the hippocampus after TBI. We tested the idea that TBI would increase cell proliferation, the survival of immature neurons and cell death and that progesterone treatment would reduce this TBI-induced augmentation.

2. MATERIALS AND METHODS

2.1 Animals

A total of twenty-four adult male Sprague-Dawley rats (8 to 12 weeks; Charles River Laboratories, Wilmington, MA) weighing 275–300 g were used according to procedures approved by the Institutional Animal Care and Use Committee, Emory University, Atlanta, GA, USA (protocol 164–2008). The rats were handled for at least 5 days before surgery, and were individually housed. Food and water were provided ad libitum throughout the experiment and the animals lived in a reversed 12-h light/12-h dark cycle controlled environment. Animals were separated into four groups (n = 6): group I: sham-operated vehicle-treated control (S); II: sham-operated and PROG (16 mg/kg) -treated (SP); group III: controlled cortical impact injury (CCI) + vehicle-treated (L); group IV: CCI + PROG (16 mg/kg) -treated (LP). Sham-operated animals underwent surgery similar to that for animals receiving CCI, but they did not receive craniectomies or cerebral contusions (see below for surgical details). Vehicle-treated animals were just given 22.5% 2-hydroxypropyl-β-cyclodextrin (see below for injection details).

2.2 Procedures

2.2.1 Induction of controlled cortical impact model for TBI

Bilateral CCI to the medial frontal cortex (MFC) was induced by a cortical contusion device as previously described (Cutler, et al., 2006). Rats were anesthetized with isoflurane (5%) (Novaplus^™), N₂O (700 cm³/min) and O₂ (300 cm³/min) for 4 min, and mounted in a stereotaxic frame. Under aseptic conditions, a sagittal incision was made in the scalp and the fascia was retracted to expose the cranium. Then, a 6-mm diameter trephan drill was used to open the skull immediately anterior to bregma. Bilateral CCI injuries of the MFC were made with a 5-mm-diameter stainless impactor attached to a computer-controlled piston propelled by compressed air (velocity = 2.25 m/s; depth = 2 mm; duration = 500 ms). After CCI cortical surface haemorrhaging was controlled and the fascia and scalp were sutured. Sham-operated rats were anaesthetized, mounted in the stereotaxic apparatus, and their scalps cut and sutured, but they were not given craniectomy, contusion or neurosteroid treatment. Using a SurgiVet^™ (model V3304) pulse oximeter, blood SpO₂ was monitored and maintained at levels ≥ 90%. Body temperature was maintained at 37°C with a homeothermic heating blanket system (Harvard Apparatus, Holliston, MA).

2.2.2 Progesterone (PROG) administration

PROG (P-0130; Sigma-Aldrich Co., St. Louis, MO) was dissolved in 22.5% 2-hydroxypropyl-β-cyclodextrin. The first injection was administered IP (16 mg/kg) at 1 h post-injury to ensure relatively rapid absorption following injury, and the subsequent doses were administered SC (16 mg/kg) at 6 h post-injury for more gradual absorption and then every 24 h with tapering as shown in Table 1. The tapered dose (16 mg/kg) of PROG was determined from previous research results showing that the selected amount provided the maximal protective effects in TBI model (Cutler, et al., 2006). Vehicle injections were given at the same times.

Table 1.

Post-injury progesterone treatment schedule:

	Days 1–5	Day 6	Day 7
Progesterone	16 mg/kg	8 mg/kg	4 mg/kg
Vehicle	22.5% HBC	22.5% HBC	22.5% HBC

2.2.3 5-Bromo-2-deoxyuridine (BrdU) administration

Bromodeoxyuridine (BrdU) is an exogenous thymidine analogue that is injected and becomes incorporated into cells that are in S-phase (synthesizing DNA) within 2 hours of injection (Packard, et al., 1973). BrdU can be used as a marker of cell proliferation or cell survival depending on the time that has elapsed between injection and perfusion (Barha, et al., 2009). If animals are perfused within 24hrs of injection then BrdU labelling can be used to assess cell proliferation, because it requires approximately 24 h for dividing precursor cells to produce daughter cells in the dentate gyrus (Cameron and McKay, 2001). If animals are perfused days to months after injection then BrdU assesses cell survival. Forty-eight hours after CCI, all the rats received a single IP injection of BrdU (200 mg/kg; Sigma) dissolved in 0.9% saline. On the same day, the dose of PROG was given 2 h prior to the single injection of BrdU. A timeline of the procedure is shown in Figure 1. In the present study, because BrdU was given 6 days prior to perfusion, BrdU-ir cells assessed short-term cell survival.

Figure 1.

Depicts the procedural timeline used in this experiment.

2.2.4 Tissue processing

Six days following BrdU injection (8 days after brain injury) the rats were deeply anaesthetized with an overdose of Nembutal and then perfused with 0.9% saline followed by 4% formaldehyde. Brains were rapidly removed and post fixed in 4% formaldehyde at 4°C for 2 days. For cryoprotection, the intact brain was transferred to 30% sucrose in phosphate buffered saline (PBS) until sectioning.

2.2.5 Immunohistochemistry

Brains were sliced into 40μm coronal sections throughout the entire rostral caudal extent of the hippocampus using a vibratome (Leica, VT1000s). Every 10^th slice was processed for either Ki67, an endogenous protein expressed in all phrases of the cell cycle except G_O and early G₁ and a marker of cell proliferation, or for BrdU, an exogenous marker of DNA synthesis and in the current study a marker of 6-day-old cells (short-term cell survival). In addition, one series of slices were double-stained for BrdU and doublecortin, a microtubule-associated protein expressed in immature neurons during the first few days to weeks following proliferation (Brown, et al., 2003), to allow phenotyping of the stained cells. Another series of slices was stained with Fluoro-Jade B, a polyanionic fluorescein derivative that specifically binds to degenerating (apoptotic and necrotic) neurons (Schmued and Hopkins, 2000). All immunohistochemistry procedures were performed as previously described (Barha, et al., 2010, Barha and Galea, 2009, Barha, et al., 2009, Epp, et al., 2009, Epp, et al., 2010, Sliwowska, et al., 2010).

Briefly, for Ki67 staining (see Figure 2A), sections were pretreated with 0.6% H₂O₂ for 30 min, rinsed three times in 0.1M PBS and then transferred to a primary antibody solution containing a 1: 1000 rabbit anti-Ki67 monoclonal antibody (Novocastra; Newcastle upon Tyne, United Kingdom), 1 % NGS and 0.5% Triton-X in 0.1M PBS for 16 hr. Tissue was rinsed three times again and then incubated in a secondary antibody solution containing 1:1000 goat anti-rabbit (Vector; Burlington, ON, Canada) in 0.1M PBS for 1 h, followed by another rinse (3X) and incubation in an ABC solution (Vector) for 40 mins. The sections were rinsed again, developed with diaminobenzidine (DAB; Vector) for 5 min and then mounted on gel-coated glass slides. All sections were lightly counterstained with cresyl violet, dehydrated and coverslipped with Permount (Fisher Scientific Canada, Ottawa, ON).

Figure 2.

Panel A shows Ki67-ir cells in the granule cell layer. Panel B shows BrdU-ir cells in the granule cell layer. Panel C shows a BrdU (red) and doublecortin (green) double labelled cell in the granule cell layer. Photomicrographs are magnified 1000x. 1 cm scale bar = 10 μm.

BrdU staining (see Figure 2B) was similar to Ki67 staining but using tris-buffered saline (TBS) for rinsing instead of PBS and with the following modifications: after initial incubation in H₂O₂ for 30 min and the subsequent rinses, tissue was incubated in 2N hydrochloric acid at 37 °C, followed by one rinse in 0.1M borate buffer for 10 min and three rinses in PBS. Tissue was then incubated in a solution containing 3% normal horse serum (NHS) and 0.1% Triton-X in PBS for 30 min, before it was transferred to a primary antibody solution containing 1:200 mouse anti-BrdU (Roche; Mississauga, ON, Canada), 3% NHS and 0.1% Triton-X in PBS for 48 hours. The secondary antibody solution contained 1:100 anti mouse IgG and tissue was incubated in this solution for 4 hours. Incubation time for the ABC solution was 1.5 hrs.

To assess the phenotype of the BrdU-ir cells, double labelling was performed using BrdU and doublecortin (see Figure 2C), a marker for immature neurons in the brain. Tissue was incubated for 48 h in 0.1M TBS containing 3% normal donkey serum (NDS), 0.4% Triton-X, 1:1000 goat anti-doublecortin (Santa Cruz Biotechnology, Santa Cruz, CA). The tissue was rinsed three times and then transferred to a 1:500 dilution of donkey anti-goat Alexa 488 in 0.1M TBS for 18 h. Next, the doublecortin signal was protected by fixing the tissue in 4% formaldehyde for 10 min and then rinsed 3 times in TBS. The tissue was then incubated in a BrdU primary antibody solution containing 1:500 rat anti-BrdU (AbD Serotec, Raleigh, NC), 4% NDS and 0.4% Triton-X for 24 h. Sections were then rinsed 3X in 0.1M TBS before being incubated overnight in 0.1M TBS containing 1:500 donkey anti-rat Alexa 594 (Invitrogen, Carlsbad, CA). Sections were then rinsed and mounted on glass slides and coverslipped with PVA-DABCO.

To assess degenerating neurons tissue was stained with Fluoro-Jade B (Schmued and Hopkins, 2000). Specifically, tissue was rinsed three times in PBS, then mounted on slides and left to air-dry for 48 h at room temperature. The slides were soaked in 0.06% potassium permanganate for 20 min, followed by soaking for 2 min in deionized water. The slides were then soaked in 0.0004% Fluoro-Jade B solution (Chemicon, Temecula, CA, USA) in 0.1% glacial acetic acid for 30 min, rinsed three times in de-ionized water, rapidly air dried, cleared in xylene, and cover-slipped with Permount.

2.2.6 Cell counting

Counting was done by an experimenter blind to the animal’s group designation. BrdU- and Ki67-ir cells were counted in every 10th section throughout the entire granule cell layer (GCL), including the subgranular zone (SGZ) using a Nikon microscope at 1000x magnification. Cells in the hilus were counted separately to account for potential changes in the blood-brain barrier permeability. Cells were considered to be BrdU-ir if they were intensely stained and exhibited medium round or oval cell body morphology (Cameron, et al., 1993, Ormerod and Galea, 2001, Ormerod, et al., 2004). Overlapping cells were counted only if a clear outline of each cell was discernable. The total number of BrdU- and Ki67-ir cells was estimated by multiplying the total number counted by 10 as previously described (Eadie, et al., 2005, Kronenberg, et al., 2003). The percentage of BrdU/doublecortin double-labelled cells was obtained by selecting 25 BrdU-ir cells arbitrarily per brain from at least five sections and determining the percentage of these cells that also expressed doublecortin on an Olympus epifluorescent microscope at 600x magnification. This method to determine the proportion of new surviving cells that are immature neurons is well-established (Dalla, et al., 2009, Epp, et al., 2011, Kempermann, et al., 2003, Suzuki, et al., 2007). In order to estimate the total number of immature neurons surviving for 6 days in the granule cell layer of the hippocampus, the total number of BrdU-ir cells was multiplied by the percentage of BrdU/doublecortin cells as has been done previously (Brandt, et al., 2010, Ramirez-Rodriguez, et al., 2009).

2.3 Data analyses

Total number of Ki67-ir cells (cell proliferation), and total number of BrdU-ir cells (cell survival) were analyzed using repeated-measures ANOVAs with group (sham, TBI, TBI + PROG, sham + PROG) as the between-subjects factor and region (GCL + SGZ, hilus) as the within-subjects factor. BrdU/doublecortin co-labelling was analyzed with a one-way ANOVA with group as the between-subjects factor. The total number of immature neurons surviving for 6 days in the granule cell layer of the hippocampus was analyzed using a one-way ANOVA with group as the between-subjects factor. The total number of Fluoro-Jade B labelled cells (cell death) in the granule cell layer and in the CA1 region of the hippocampus were analyzed separately using a one-way ANOVA with group as the between-subjects factor. Post-hoc tests utilized the Newman-Keuls procedure unless otherwise indicated. All statistical procedures were set at α = 0.05.

3. RESULTS

3.1 Brain injury increases cell proliferation compared to sham operations. Progesterone decreases cell proliferation in TBI rats but increases cell proliferation in intact rats

The total number of Ki67-ir cells in the granule cell layer and the hilus for all groups are shown in Figure 3A. A repeated-measures ANOVA conducted on the total number of Ki67-ir cells in the granule cell layer and the hilus showed a significant interaction between group and region [F(3, 16) = 8.05, p < 0.002]. Post-hoc analysis revealed that the sham group had significantly fewer total Ki67-ir cells in the granule cell layer compared to the sham + progesterone group (p < 0.004) and the TBI group (p < 0.03), but had significantly more Ki67-ir cells compared to the TBI + progesterone group (p < 0.002). Furthermore, the TBI + progesterone group had significantly fewer Ki67-ir cells compared to the TBI group (p < 0.0002) and the sham + progesterone group (p < 0.0002). The sham + progesterone group and the TBI group did not significantly differ from each other (p = 0.15). Groups did not differ in the total number of Ki67-ir cells in the hilus (all p’s > 0.55). Significant main effects of group [F(3, 16) = 13.13, p < 0.0002) and region [F(1, 16) = 249.20, p < 0.00001] were also found.

Figure 3.

A) Mean (+ SEM) total number of Ki67-ir cells in the granule cell layer (GCL) and hilus of the dentate gyrus. The sham group had significantly fewer Ki67-ir cells compared to the sham + progesterone group and the traumatic brain injury (TBI) group and had significantly more Ki67-ir cells compared to the TBI + progesterone group. The TBI + progesterone group had significantly fewer Ki67-ir cells than the TBI group and the sham + progesterone group. There were no significant differences between groups in the hilus. B) Mean (+ SEM) total number of BrdU-ir cells in the GCL and hilus of the dentate gyrus. Groups did not differ in the total number of BrdU-ir cells in the GCL or the hilus. C) Mean (+ SEM) total number of immature neurons surviving for 6 days. The TBI group had significantly more immature neurons compared to the sham group. All other groups did not differ from the sham group.

a denotes significantly different from the sham group (p < 0.05), b denotes significantly different from the TBI group and the sham + progesterone group (p < 0.05).

3.2 All treatments increase the percentage of immature neurons in the granule cell layer

An ANOVA conducted on the percentage of BrdU-ir cells co-expressing doublecortin showed a significant main effect of group [F(3,15) = 5.74, p < 0.009; see Table 2], with the sham group having the lowest percentage of BrdU-ir cells co-expressing doublecortin compared to all other groups (all p’s< 0.02).

Table 2.

Mean (+ SEM) percentage of BrdU/Doublecortin double-labelled cells in the granule cell layer with the sham group having the lowest percentage of double-labelled cells (doublecortin and BrdU).

Group	% of BrdU/DCX
Sham	73.11% ± 4.74
Sham + Progesterone	88.00% ± 2.53*
TBI	89.00% ± 3.42*
TBI + Progesterone	89.60% ± 2.04*

^*indicates significantly different from sham group.

3.3 Brain injury and progesterone treatment do not influence short-term cell survival of BrdU-ir cells but did alter the estimated number of immature neurons

The total number of BrdU-ir cells in the granule cell layer and the hilus for all groups is shown in Figure 3B. A repeated-measures ANOVA conducted on the total number of BrdU-ir cells in the granule cell layer and the hilus did not find a significant interaction between group and region [F(3, 22) = 0.11, p = 0.95], or significant main effect of group [F(3, 22) = 0.75, p = 0.54]. As expected, a significant main effect of region was seen [F(1, 22) = 343.48, p = 0.00001], with a higher number of BrdU-ir cells seen in the granule cell layer compared to the hilus.

We multiplied the total number of BrdU-ir in the granule cell layer by the percentage of BrdU-ir cells that were co-labelled with doublecortin to obtain a measure of immature neurons that survived 6 days (Brandt, et al., 2010, Ramirez-Rodriguez, et al., 2009). An ANOVA conducted on the total number of immature neurons was not significant [main effect of group: F(3, 15) = 2.34, p = 0.11; see Figure 3C]. However, a priori we expected that TBI and progesterone treatment would influence the total number of immature neurons surviving. A priori tests indicate that the TBI group had more doublecortin cells surviving than the sham group (p < 0.018). The TBI + progesterone group and the sham + progesterone group did not differ from the sham group (p < 0.20), indicating that progesterone treatment reduced the genesis of new neurons in the damaged brain.

3.4 Brain injury increases cell death in the granule cell layer of the dentate gyrus and progesterone treatment normalized cell death back to intact levels

The total number of Fluoro-Jade-ir cells in the granule cell layer and the CA1 region of the hippocampus for all groups are shown in Figure 4. An ANOVA conducted on the total number of Fluoro-Jade-ir cells in the granule cell layer showed a significant main effect of group [F(3, 22) = 8.12, p < 0.001], with the TBI group containing more Fluoro-Jade-ir cells than the sham group (p < 0.003), the sham + progesterone group (p < 0.002), and the TBI + progesterone group (p < 0.04). An ANOVA conducted on the total number of Fluoro-Jade-ir cells in the CA1 region showed a tendency for a main effect of group [F(3, 23) = 2.75, p < 0.07]. A priori we expected TBI to increase cell death and A priori tests indicate that the TBI group has more Fluoro-Jade-ir cells than the sham group (p < 0.03) and the sham + progesterone group (p < 0.02).

Figure 4.

Mean (+ SEM) total number of Fluoro-Jade B labelled cells in the granule cell layer (GCL) of the dentate gyrus and the CA1 region of the hippocampus. The traumatic brain injury (TBI) group had significantly more Fluoro-Jade B labelled cells in the GCL compared to the sham group, the sham + progesterone group, and the TBI + progesterone group. The TBI group tended to have more Fluoro-Jade B labelled cells in the CA1 region compared to the sham group and the sham + progesterone group. a denotes significantly different from the TBI group (p < 0.05).

4. DISCUSSION

The results from the present study demonstrate that TBI increased cell proliferation and cell death in the dentate gyrus of the adult male rat, while progesterone treatment normalized the injury-induced levels of both cell proliferation and cell death to sham levels. Progesterone’s ability to influence cell proliferation was dependent on whether it was administered to a healthy or injured brain. In the present study neither injury nor progesterone treatment influenced short-term cell survival of 6-day-old BrdU-ir cells. However, all treatments increased the percentage of new cells with an immature neuronal phenotype. Furthermore, only injury increased the number of immature neurons relative to sham-treated rats, and progesterone after injury reduced the number of immature neurons to sham control levels. We also found that TBI increased neuronal death in the dentate gyrus and CA1 region of the hippocampus and that progesterone restored cell death to control levels. To our knowledge this is the first demonstration of how progesterone treatment following TBI can influence adult neurogenesis in the dentate gyrus of male rats.

Progesterone normalizes the injury-induced increase in cell proliferation and progesterone alone increases cell proliferation

Bilateral controlled cortical impact to the medial frontal cortex increased the number of cells proliferating in the dentate gyrus of adult male rats 8 days after injury compared to sham-operated controls. Exogenous progesterone treatment augmented the number of new cells proliferating in sham-operated controls but decreased the number of new cells proliferating in brain-injured rats. This observation indicates a differential neurogenic response to progesterone in the healthy versus the injured brain. Although perhaps counterintuitive, we think it is possible that proliferating cells may actually interfere with hippocampal function until they mature and make appropriate connections. In fact many studies have found that high rates of cell proliferation are not necessarily beneficial for hippocampus-dependent learning (Epp and Galea, 2009) and can directly contribute to impairments in learning (Jessberger, et al., 2007). This may be because new cells that are not yet connected to the existing circuitry could “add noise” and interfere with the proper functioning of the dentate gyrus. Furthermore, the findings in the present study are consistent with progesterone’s protective effects via neurogenesis in a focal ischemia model of injury. Zhang et al. (2010) found that progesterone treatment reduced cell proliferation induced by ischemia but increased 28-day survival of neurons. Recently, Chrousos (2010) pointed out that the idea that steroid hormones have to have uniform actions in various tissues is overly simplistic. He suggests that the response to steroid exposure depends on the environmental context of the cells and the state of the individual cells (i.e. healthy or injured), among many other factors that can determine cellular response to steroid hormones.

Progesterone reduces the injury-induced increase in cell death in the dentate gyrus

We found that bilateral injury to the medial frontal cortex increased the number of degenerating neurons in the granule cell layer of the dentate gyrus and to a lesser extent, in the CA1 region of the hippocampus. Importantly, progesterone treatment did not influence cell death in sham-operated rats, but after cortical contusion injury, it decreased the number of dying cells in the dentate gyrus, indicating that progesterone’s neuroprotective effects are more obvious in the injured brain. It is important to note that Fluoro-Jade B stains degenerating apoptotic and necrotic mature neurons. Therefore the cells that are stained with Fluoro-Jade B in this experiment are of a different, much older population than Ki67-ir cells and BrdU-ir cells (which are very immature neurons) and thus in our study, most of the Fluoro-Jade B-ir cells existed well before the bilateral controlled cortical impact.

Cell death is a prominent feature seen following TBI in the cortex, cerebellum, thalamus, and hippocampus. The hippocampus is particularly vulnerable to TBI. Bilateral loss of hippocampal neurons is seen in 85% of fatal head injuries in humans (Kotapka, et al., 1992). Within the hippocampus, the CA1, CA3, and dentate gyrus subregions show differing degrees of neuronal death after controlled cortical impact injury. A higher number of degenerating neurons are seen in the dentate gyrus, and fewer in the CA1 region. This is consistent with our other findings (Anderson, et al., 2005, Ariza, et al., 2006, Gao, et al., 2008, Pullela, et al., 2006, Saatman, et al., 2006). The results of our experiment support the regional distinction in TBI-induced pathology as cellular death was significantly increased by TBI in the dentate gyrus, but less so in the CA1 region. Newborn immature neurons in the dentate gyrus of the hippocampus are the most vulnerable to TBI (Gao, et al., 2008), as seen in the present study, where the cortical contusions increased cell proliferation but did not affect 6-day-old BrdU cells in the dentate gyrus. The effects of a cortical contusion on hippocampal neurogenesis could be due to the inhibitory effects of diaschisis, and may involve many different secondary pathways elicited by the traumatic injury, like increases in inflammation (Rola, et al., 2006), which lead to disruptions in the neurogenic microenvironment of the dentate gyrus. The cortical impact, penetrating about 2mm into the brain also produces both proximal and distal effects on the brain vasculature leading to ischemic injury in underlying tissue such as the dorsal hippocampus. This can produce a cascade of inflammation, excitotoxicity, secondary neurodegeneration and gliosis—all of which can damage or kills vulnerable hippocampal neurons.

Traumatic brain injury and progesterone did not influence short-term survival of new cells

TBI and/or progesterone treatment did not increase short-term (6-day) survival of BrdU-ir cells. However, TBI and/or progesterone treatment did increase the percentage of new cells that expressed doublecortin compared to sham control rats. This finding is consistent with previous work showing that TBI increases long-term (> 21 days) survival of new neurons in the dentate gyrus (Dash, et al., 2001, Sun, et al., 2010, Sun, et al., 2005). In the current study we found that TBI increased the number of immature neurons surviving in the granule cell layer compared to shams. We also observed that the number of immature neurons in rats given progesterone treatment after TBI was not different from sham controls, indicating that progesterone after TBI normalized the number of immature neurons. Thus far we have examined only the survival of 6-day-old cells. It is possible that the effects of TBI and/or progesterone on survival of new neurons can be seen at longer time points, as reported in a study conducted by Zhang et al. (2010) in which 28-day survival of neurons was assessed.

Seven days of progesterone treatment increases cell proliferation and neuronal phenotype in male rats

We found that 7 days of progesterone treatment increased both cell proliferation and the percentage of new cells expressing a neuronal phenotype, but not short-term survival of immature neurons in male rats. In agreement with the current study, in vitro work found that progesterone enhanced cell proliferation (Liu, et al., 2009, Wang, et al., 2005).

In contrast, a recent study by Zhang et al. (2010) did not see an increase in a mixture of cell proliferation and 2 day cell survival following progesterone treatment in healthy adult male mice, but we found that 8 progesterone injections (tapered dose: 16 mg/kg to 4 mg/kg) over 7 days led to an increase in cell proliferation in the dentate gyrus. This regimen of progesterone has previously been shown to provide maximal protective effects in a TBI model (see Table 1 and Figure 1; Cutler, et al., 2006). Thus the Zhang et al. study differs from ours in the dose of progesterone given (4 mg/kg vs tapered schedule), number of progesterone injections (3 vs 8), the method used to detect cell proliferation (BrdU vs Ki67), and the species tested (mouse vs rat). Importantly, Zhang et al. gave multiple injections of BrdU over 2 days, which would have resulted in labelling a heterogeneous population of cells of both proliferating and short-term survival (see Taupin, 2007), and would not have been a pure measure of cell proliferation unlike in our study.

In the present study we saw a TBI effect on cell proliferation, using Ki67, 8 days after contusion, but we did not observe a TBI effect on BrdU-ir cells surviving 8 days after TBI that were labelled 2 days after the injury. It is important to note that in the present study, Ki67-ir and BrdU-ir cells were labeling very different populations of cells. Ki67 labelled a population proliferating at the time of death which was 8 days after TBI and BrdU was labeling cells that proliferated 2 days after TBI and survived a further 6 days. A single pulse of BrdU is incorporated into approximately 50% of the Ki67-ir cell population in the dentate gyrus (Kee, et al., 2002, Leuner, et al., 2009). This difference between the total number of cells labelled by BrdU and Ki67 is due to the fact that Ki67 is expressed throughout the cell cycle, except G0, whereas BrdU is only incorporated during the S-phase.

Progesterone receptors and potential role in hippocampal neurogenesis

The reduction in the injury-induced increase in cell proliferation and cell death by progesterone treatment may occur through many possible mechanisms. Progesterone exerts its influence on the central nervous system through classical progesterone receptor (PR) mediated pathways and alternative non-genomic mechanisms. Two major classical PR exist: PRB and PRA (Boonyaratanakornkit, et al., 2008). Several splice variants of each of the classical PRs have also been identified in several species (Zhu, et al., 2003, Zhu, et al., 2003). Additionally two main putative membrane-bound PRs have been identified in recent years, the seven transmembrane domain G protein-coupled receptor 7TMPR (Zhu, et al., 2003, Zhu, et al., 2003) and the single transmembrane domain 25-Dx PR membrane components, the PGRMCs 1 and 2 (Guennoun, et al., 2008, Meffre, et al., 2007, Meffre, et al., 2007). The membrane-associated progesterone-binding protein 25-Dx is expressed in brain regions involved in water homeostasis and is up-regulated after traumatic brain injury (Meffre, et al., 2005).

PRs are found throughout the brain (for review see Brinton, et al., 2008) and PRB and PRA are expressed throughout the hippocampus of both pre- and postsynaptic neurons (Foy, et al., 2010) throughout the hippocampus. Although PRs have been located in the subgranular zone of the dentate gyrus (Waters, et al., 2008), colocalization of PRs and progenitor cells that also reside in the subgranular zone has yet to be determined. However, progesterone-induced increase in cell proliferation is dependent on the ERK/MAPK signalling pathway and involves activation of the membrane PR, PGMRC1, as depleting PGRMC1 levels via siRNA reduces the progesterone-induced increase in cell proliferation (Liu, et al., 2009).

Progesterone’s effects on hippocampal morphology and function

The effects of progesterone on learning and memory are complex and depend on timing of administration. Progesterone treatment prior to training impairs spatial working memory and footshock avoidance (Bimonte-Nelson, et al., 2004, Farr, et al., 1995, Orr, et al., 2009) while post-training treatment with progesterone can improve spatial reference memory, object recognition and Y-maze inhibitory avoidance (Harburger, et al., 2008, Johansson, et al., 2002). Furthermore progesterone rapidly decreases the estradiol-induced increase in neuronal spine and synapse density and synaptic proteins such as syntaxin, synaptophysin, and spinophilin in the CA1 region of the hippocampus (Choi, et al., 2003, Gould, et al., 1990, Woolley and McEwen, 1993). Microtubule-associated protein 2 (MAP2) is involved in the assembly and stabilization of microtubules and is involved in the growth of the dendritic spines of neurons (Leclerc, et al., 1996). Interestingly both estradiol and progesterone increase protein levels of MAP2 in the hippocampus (Reyna-Neyra, et al., 2002). Thus, taken together with the results of this experiment showing that progesterone affects some aspects of neurogenesis in the hippocampus, we believe that progesterone likely works through a number of mechanisms to alter hippocampal morphology and function.

The role of neurosteriods on TBI

Progesterone can work directly and indirectly to exerts its effects on neuroprotection. Importantly, treatment with progesterone or its metabolite, allopregnanolone, decreases infarct size, improves functional recovery, reduces inflammatory response, reduces edema, and protects blood brain barrier integrity following brain injury (Djebaili, et al., 2005, Gibson, et al., 2009, Ishrat, et al., 2010, Ishrat, et al., 2009, Sayeed, et al., 2006, Wang, et al., 2009, Wright, et al., 2001). Depending on the outcome variable of interest, allopregnanolone may be a more potent inducer of neuroprotection than progesterone (Djebaili, et al., 2005, Sayeed and Stein, 2009). Interestingly, allopregnanolone also increases progenitor cell proliferation (Brinton and Wang, 2006, Wang, et al., 2008). Importantly these two neurosteroids do not work through the same receptor. Unlike progesterone, allopregnanolone does not bind to the intracellular membrane-bound progesterone receptor, the sigma-1 receptor, and the putative progesterone dx-25 receptor (Monnet, et al., 1995, Rupprecht, et al., 1993). Allopregnanolone does act on the GABA-A receptor and it may be through this action that it is exerting its effects (Ghoumari, et al., 2003).

Translocator protein (TSPO) is an 18 kDA protein with a high affinity for binding cholesterol and is found mainly in the outer mitochondrial membrane. Tissues rich in steroidgenesis throughout the body, including the brain, express TSPO (Papadopoulos, et al., 1997) and its expression is increased in the brain after various insults including TBI (for review see Chen and Guilarte, 2008). TSPO is involved in multiple functions, including regulating apoptosis, cell proliferation, and mitochondrial function (for review see Batarseh and Papadopoulos, 2010, Chen and Guilarte, 2008, Rupprecht, et al., 2010). Furthermore, TSPO is a key component in steroidogenesis, as it binds cholesterol via steroidogenic acute regulatory protein (StAR) (Lacapere and Papadopoulos, 2003) and then facilitates the transport of cholesterol from the outer to the inner mitochondrial membrane (Papadopoulos, et al., 1997). Within the inner mitochondrial membrane, the side chain cleavage cytochrome P-450 enzyme converts the cholesterol to pregnenolone, which is further converted to many neurosteroids including allopregnanolone (Brown and Papadopoulos, 2001). It may be through this pathway in neurosteroid synthesis that TSPO and TSPO selective ligands influence neuroprotection.

5. CONCLUSIONS

Severe TBI is often very debilitating and leads to alterations in hippocampal function and structure. Progesterone is an effective neuroprotective agent that improves functional outcome following TBI. The results of our study extend the evidence for the beneficial effects of progesterone following TBI, and show that progesterone administration following contusion reduces cell proliferation, number of immature neurons, and cell death in the dentate gyrus of male rats. These findings are particularly interesting in light of recent studies reporting that high levels of cell proliferation in the dentate gyrus correlate with impairments in hippocampus-dependent learning and memory, further indicating that the beneficial effects seen with progesterone treatment on cognitive functioning following TBI may be associated with the ability of progesterone to reinstate neurogenic homeostasis in the hippocampus.

Research Highlights.

• Progesterone improves outcome following traumatic brain injury in humans and rodents

• Traumatic brain injury increased cell proliferation and cell death in hippocampus

• Progesterone normalized these levels in hippocampal neurogenesis

• Improved cognitive outcome could be related to progesterone effects on neurogenesis