Neurons and Glial Cells Are Added to the Female Rat Anteroventral Periventricular Nucleus During Puberty

Abstract

The anteroventral periventricular nucleus (AVPV) orchestrates the neuroendocrine-positive feedback response that triggers ovulation in female rodents. The AVPV is larger and more cell-dense in females than in males, and during puberty, only females develop the capacity to show a positive feedback response. We previously reported a potential new mechanism to explain this female-specific gain of function during puberty, namely a female-biased sex difference in the pubertal addition of new cells to the rat AVPV. Here we first asked whether this sex difference is due to greater cell proliferation and/or survival in females. Female and male rats received the cell birthdate marker 5-bromo-2′-deoxyuridine (BrdU; 200 mg/kg, ip) on postnatal day (P) 30; brains were collected at short and long intervals after BrdU administration to assess cell proliferation and survival, respectively. Overall, females had more BrdU-immunoreactive cells in the AVPV than did males, with no sex differences in the rate of cell attrition over time. Thus, the sex difference in pubertal addition of AVPV cells appears to be due to greater cell proliferation in females. Next, to determine the phenotype of pubertally born AVPV cells, daily BrdU injections were given to female rats on P28–56, and tissue was collected on P77 to assess colocalization of BrdU and markers for mature neurons or glia. Of the pubertally born AVPV cells, approximately 15% differentiated into neurons, approximately 19% into astrocytes, and approximately 23% into microglia. Thus, both neuro- and gliogenesis occur in the pubertal female rat AVPV and potentially contribute to maturation of female reproductive function.

The anteroventral periventricular nucleus (AVPV) of the rodent hypothalamus is crucial to regulation of a pivotal event in female reproduction: the preovulatory surge of LH (for review see reference ¹), leading to ovulation. The LH surge is the product of neuroendocrine positive feedback, in which elevated levels of ovarian hormones trigger a surge of GnRH. The AVPV integrates these hormonal signals with circadian signals, and through AVPV kisspeptin-producing neurons, generates a GnRH surge on the afternoon of proestrus (for review see Ref. 2). This positive feedback response is acquired during puberty: only after puberty can female rats generate an LH surge that is temporally and physiologically appropriate (3). Furthermore, male rats are incapable of generating an LH surge at any age, indicating that the male AVPV does not undergo the same pubertal transformation as the female AVPV (4). The mechanisms responsible for this female-specific gain of function during puberty are not well understood, but clues may be found by examining the structural development of the AVPV during puberty in females and males.

The adult AVPV is larger in females than in males, and although sexual differentiation of the AVPV begins during perinatal development (for review, see reference ⁵), the sex difference in AVPV volume does not emerge until puberty (6). We previously discovered that new cells are added to the rat AVPV during puberty, more so in females than in males (7). In that study, we injected rats with bromodeoxyuridine (BrdU) around the onset of puberty (postnatal days [P] 30–32) and 3 weeks later saw approximately twice as many BrdU-immunoreactive (ir) cells in the female AVPV as in the male. The purpose of the present study was to answer two questions related to the AVPV that were not addressed in our previous study: 1) is the sex difference in the number of new cells added to the AVPV during puberty due to greater cell proliferation or greater cell survival in females, and 2) what proportion of pubertally born female AVPV cells differentiate into mature neurons, astrocytes, or microglia? We report here that the sex difference in the number of pubertally born cells added to the AVPV is likely due to greater cell proliferation in females, and not greater cell death in males, the opposite of what occurs during perinatal sexual differentiation of the AVPV. In addition, more than half of the pubertally born female AVPV cells differentiate into mature neurons, astrocytes, or microglia within several weeks of proliferation. These findings point to the addition of neurons and glial cells to the AVPV as a putative new mechanism of maturation of female neuroendocrine positive feedback.

Materials and Methods

Animals

Sprague Dawley rats were ordered from Harlan Sprague Dawley and double housed (experiments 1 and 3) or single housed (experiment 2) in 37.5- × 33- × 17-cm clear polycarbonate cages with ad libitum access to food (Teklad Rodent diet number 8640; Harlan) and water. They were maintained on a 14-hour light, 10-hour dark cycle (lights off at 1:00 pm). Animals were treated in accordance with the National Institutes of Health's Guide for the Care and Use of Laboratory Animals. Michigan State University's Institutional Animal Care and Use Committee approved all protocols.

Design

Experiment 1a

The purpose of this experiment was to determine whether the sex difference in cell addition to the rat AVPV is due to greater cell proliferation or survival in females.

A total of 42 male and 42 female weanling female Sprague Dawley rats arrived on P21. On P30, rats received an ip injection of 5-bromo-2′-deoxyuridine (Sigma-Aldrich) at a dose of 200 mg/kg dissolved in sterile saline (20 mg/mL) at approximately 8:00 am, 4:00 pm, and 12:00 am. Rats were perfused 2 days, 4 days, 7 days, 14 days, 21 days, or 42 days after the final BrdU injection. All perfusions (n = 7/ time point) took place at approximately 1:00 pm (lights out).

Experiment 1b. The purpose of this experiment was to determine whether there is an early sex difference in proliferating cells.

Preliminary analyses of BrdU-ir cells in experiment 1a indicated a sex difference (females > males) as early as 2 days after the BrdU injections. To determine whether a sex difference was present within a few hours of the BrdU injection, when virtually all proliferating cells would be captured, three male and three female P30 rats received a single ip injection of BrdU (200 mg/kg dissolved in sterile saline, 10 mg/mL) at 2:00 pm and were perfused 2 hours later.

Experiment 2

The purpose of this experiment was to determine whether BrdU is incorporated into proliferating, not dying, cells in the AVPV.

Thirty-six female weanling female Sprague Dawley rats arrived on P23. On P30, rats were given a single ip injection of BrdU (Sigma-Aldrich) at a dose of 200 mg/kg, dissolved in sterile saline (10 mg/mL solution) between 12:00 pm and 2:00 pm. Rats were perfused 2 hours, 4 hours, 1 day, 2 days, 4 days, 7 days, 10 days, 14 days, or 21 days after the BrdU injection. With the exception of the 2 hour and 4 hour post-BrdU perfusion groups, all perfusions (n = 4/time point) took place at approximately 3:00 pm (2 h after lights off).

Experiment 3

The purpose of this experiment was to determine whether pubertally born cells in the AVPV differentiate into neurons, astrocytes, or microglia in adulthood.

Twelve weanling female Sprague Dawley rats (P21) were ordered from Harlan. Animals were given a week to acclimate after arrival. During this time, they were handled daily. From P28 to P56, rats received a daily ip injection of BrdU (200 mg/kg in sterile saline, 20 mg/mL) at approximately 12:00 pm. Three weeks after the last injection (P77), the rats were perfused.

Perfusions and sectioning

Animals were killed with an overdose of sodium pentobarbital (90 mg/kg, ip) and transcardially perfused with 0.9% buffered saline rinse (pH 7.4) followed by 4% paraformaldehyde in cold 0.1 M PBS. Brains were removed and postfixed in 4% paraformaldehyde overnight (∼18 h) and then transferred to 30% sucrose. Brains were sectioned on a cryostat at 30 μm (experiments 1 and 3) or 40 μm (experiment 2) and placed into cryoprotectant solution at −20°C (8). One set of sections (1 in 12 series, experiment 1; one in four series, experiment 2) was mounted onto subbed glass slides and Nissl-stained for use as anatomical reference.

Immunohistochemistry/immunofluorescence

General methods

Free-floating sections were rinsed in Tris-buffered saline (TBS; 0.05 M) initially and between all incubations, and all incubations took place at room temperature, unless otherwise noted. Monoclonal rat anti-BrdU (Serotec; MCA2060; 1 μg/mL) served as the primary antibody against BrdU in all immunolabeling experiments. Brain sections used as BrdU-ir positive control tissue (in many cases containing the subgranular zone) were from a 1.5- to 2-month-old rat whose dam received BrdU during the latter end of gestation. Controls excluding primary or secondary antibodies were run using experimental tissue that did not contain the AVPV to conserve the limited number of AVPV sections. The microscopic examination of the control sections revealed little to no nonspecific background staining. For antibody information, see Table 1.

Table 1.

Antibody Table

Peptide/Protein Target	Antigen Sequence (if Known)	Name of Antibody	Manufacturer, Catalog Number, and/or Name of Individual Providing the Antibody	Species Raised (Monoclonal or Polyclonal)	Dilution Used
BrdU		BrdU antibody, clone BU1/75 (ICR1)	AbD Serotec, MCA2060, clone BU1/75 (ICR1)	Rat monoclonal	1:1000
PCNA	Raised against rat PCNA made in the protein A expression vector pR1T2T	PCNA antibody, PC10	Santa Cruz Biotechnology, SC56	Mouse monoclonal	1:250
GFAP		Polyclonal rat anti-GFAP	Dako, Z0334	Rabbit polyclonal	1:5000
NeuN		Anti-NeuN antibody, clone A60	EMD Millipore, MAB377, clone A60	Mouse monoclonal	1:1000
Iba1	C terminus of Iba1	Anti-Iba, rabbit	Wako, 019-1974	Rabbit polyclonal	1:1000

Experiments 1a and 1b: single-label BrdU immunohistochemistry

A one-in-four series of free-floating sections was immunohistochemically processed for BrdU localization as described previously for experiment 1b (9) and with minor modifications for experiment 1a. Endogenous peroxidase activity was eliminated by incubation in 0.6% H₂O₂ for 30 minutes. Sections were placed in 2 N HCl for 30 minutes at 37°C and then rinsed for 10 minutes in borate buffer (0.1 M; pH 8.5). The tissue was blocked in AffiniPure Fab Fragment donkey antirat IgG (Jackson ImmunoResearch, catalog number 712-007-003, 12 μg/mL) in TBS with 0.1% Triton X-100 and 3% donkey serum for 30 minutes and then incubated overnight at 4°C in rat anti-BrdU. Sections were incubated for 2 hours in Biotin-SP-conjugated AffiniPure donkey antirat secondary antibody (catalog number 712-065-150; Jackson ImmunoResearch; 2 μg/mL) and then for 60 minutes with avidin/biotinylated enzyme complex reagent (ABC Elite kit; Vector Laboratories) and reacted with 3,3′-diaminobenzidine tetrahydrochloride (Sigma-Aldrich, catalog number D5905; 0.25 mg/mL) with 0.012% hydrogen peroxide in 0.05 TBS for 2 minutes. After mounting the tissue onto glass slides, slides were coverslipped using Vectashield mounting medium with 4′,6′-diamino-2-phenylindole (DAPI; Vector H-1200; Vector Laboratories) and stored at 4°C.

Experiment 2: double-label proliferating cell nuclear antigen (PCNA)/BrdU immunofluorescence

To determine whether BrdU was incorporated into proliferating cells and not cells undergoing DNA repair or apoptosis, double-label immunohistochemistry was performed with BrdU and PCNA, which is expressed only during the synthesis phase of the cell cycle (reviewed in reference ¹⁰). Tissue containing an intact AVPV from animals perfused 2 hours, 4 hours, 1 day, 2 days, 4 days, 7 days, 10 days, 14 days, or 21 days after BrdU injection was processed for BrdU/PCNA fluorescent labeling (n = 3 processed/time point). After rinses in 0.05 M TBS, sections underwent antigen retrieval via incubation in 0.01 M citric acid (pH 6.0) for 10 minutes at 95°C and were then cooled for 30 minutes at room temperature. Sections were then incubated in 0.6% H₂O₂ for 30 minutes. After a 2-hour incubation at 65°C in 50% formamide in saline sodium citrate buffer, sections were rinsed in saline sodium citrate for 15 minutes, incubated for 30 minutes in 2 N HCl, and then rinsed in 0.1 M borate buffer (pH 8.5) for 10 minutes. Tissue was blocked in 0.1% Triton X-100 and 3% goat serum (Chemicon; S 26-100) in TBS for 30 minutes and then incubated for 72 hours at 4°C in rat anti-BrdU and monoclonal mouse-anti-PCNA (Santa Cruz Biotechnology; SC 56, 4 μg/mL). Subsequently, sections were incubated for 3 hours in Alexa Fluor 488 goat antirat (Invitrogen; A-11006, 1:250 dilution) and Alexa Fluor 635 goat antimouse (Invitrogen; A-31575, 1:250 dilution) in TBS and 0.2% Triton X-100. After the last rinse, the sections were mounted, coverslipped, and stored as above.

Experiment 3: triple-label BrdU/glial fibrillary acidic protein (GFAP)/neuronal nuclei (NeuN) and double-label BrdU/ionized calcium-binding adapter molecule 1 (Iba1) immunofluorescence

To determine whether pubertally born cells in the female rat AVPV differentiate into astrocytes or neurons, triple-label immunofluorescence for BrdU, the astrocyte marker GFAP, and the mature neuronal marker NeuN was performed (n = 6). The methods were completed as previously described (9), with minor changes. A one-in-three series of free-floating sections was incubated in 0.1% sodium borohydride for 10 minutes to reduce autofluorescence. Sections were incubated in 1 N HCl for 30 minutes at 37°C followed by a 10-minute incubation in 0.1 M borate buffer. Tissue was blocked for 1 hour in TBS-plus (0.3% Triton X-100 and 6% goat serum in TBS), followed by a 48-hour incubation at 4°C in a primary antibody cocktail solution containing rat anti-BrdU, monoclonal mouse anti-NeuN (EMD Millipore; catalog number MAB377, clone A60; 1 μg/mL), and polyclonal rabbit anti-GFAP (Dako; catalog number Z0334; 0.58 μg/mL) in TBS-plus. Sections were incubated for 2 hours in a secondary antibody cocktail solution containing biotin-SP-conjugated AffiniPure goat antirat (Jackson ImmunoResearch; catalog number 112 065 003; 1.3 μg/mL), Cy3-conjugated AffiniPure goat antirabbit (Jackson ImmunoResearch; catalog number 111 165 144; 3 μg/mL) and Alexa Fluor 635 goat antimouse (Life Technologies; catalog number A-31574; 4 μg/mL). Finally, sections were incubated in Cy2-conjugated streptavidin (Jackson ImmunoResearch; catalog number 016 220 084; 1.8 μg/mL) for 1 hour and then mounted and coverslipped using SlowFade gold antifade reagent (Life Technologies; catalog number S36936).

To determine whether pubertally born cells in the female rat AVPV differentiate into microglial cells in adulthood, double-label immunofluorescence for BrdU and the microglial marker, Iba1, was performed (n = 6). A one-in-three series of free-floating sections was incubated in 0.1% trypsin/0.1% calcium chloride (CaCl₂) in 0.1 M TBS for 5 minutes. Tissue was incubated at 37°C in 2 N HCl for 30 minutes, followed by a 10-minute incubation in 0.1 M borate buffer (pH 8.5). After blocking in 10% BSA (Sigma; catalog number A3733)/6% normal goat serum (Pel-Freez, code number 32130-5)/0.4% Tween 20 in TBS for 1 hour, tissue was incubated overnight in a primary antibody cocktail solution containing rat anti-BrdU and rabbit anti-Iba1 (Wako; catalog number 019-19741; 1 μg/mL). Tissue then underwent a 2-hour incubation in a secondary antibody cocktail solution containing biotin-SP-conjugated AffiniPure goat antirat (Jackson ImmunoResearch; catalog number 112 065 003; 1.3 μg/mL), and Cy3-conjugated AffiniPure goat antirabbit (Jackson ImmunoResearch; catalog number 111 165 144; 1.5 μg/mL). Finally, sections were incubated in Cy2-conjugated streptavidin (Jackson ImmunoResearch; catalog number 016 220 084; 1.8 μg/mL) for 1 hour and then mounted and coverslipped using the SlowFade gold antifade reagent (Life Technologies; catalog number S36936).

Microscopic analyses and quantification

Criterion for analyses

Harsh conditions are required for BrdU immunolabeling, damaging some of the sections. Only sections that had the entire AVPV visible as well as the presence of the optic tract were included in microscopic analyses, and, for colocalization experiments, only animals showing robust single label for each marker were included. For these reasons, some animals were entirely excluded. The final sample sizes for each experiment are provided in the Statistical analyses section below. For all microscopic analyses, observers were blinded as to the age or sex of the animals.

Single-label BrdU, experiments 1a and 1b

To analyze the total number of BrdU-ir cells in the AVPV of female and male rats, four anatomically matched sections of the AVPV were selected. Animals that did not have all four sections were excluded from the analyses. Analyses were performed on an Olympus BX51 microscope with bright-field illumination as well as epifluorescence illumination (mercury arc lamp with an UV filter) using Neurolucida software (MBF Bioscience). The AVPV was located and traced bilaterally using a DAPI counterstain with a 10× objective. A 1-in-12 series of Nissl-stained sections also served as a guide for tracing AVPV. All BrdU-ir cells within the AVPV sections were counted at ×400 magnification using an UPlanSApo ×40 (0.9 NA) objective.

BrdU/PCNA, experiment 2

Z stacks containing BrdU-ir cells in the AVPV were taken with a Leica DMR microscope under epifluorescence illumination using the MBF Bioscience Image Stack Module for Neurolucida and a Plan Fluotar ×40 (NA 0.70) objective. The AVPV was first traced at ×10 using a UV filter and the DAPI counterstain to identify landmarks. Next, the objective was switched to ×40, and a meander scan was initiated that allowed stacks to be captured at each field of view throughout the AVPV, switching between fluorescein isothiocyanate and tetramethylrhodamine isothiocyanate filters to later visualize BrdU- and PCNA-positive cells, respectively. Each stack was taken at 1-μm steps spanning 20 μm without regard to whether the immunolabel was present in the field of view. Image stacks were later opened, cells were counted, and colocalization was determined using the Neurolucida 3D Visualization feature.

BrdU/GFAP/NeuN, experiment 3

Confocal Z stacks of every BrdU-ir cell in the AVPV were taken with an Olympus FluoView FV1000 confocal laser-scanning microscope equipped with FV1000 ASW software. Stacks were taken at 0.5-μm steps spanning the Z-axis of the cell using an UPLFLN ×40 oil objective (1.3 NA). The multiline argon laser (488 nm), green HeNe laser (543 nm), and red HeNe laser (633 nm) were used to image BrdU-, GFAP- and NeuN-positive cells, respectively. All BrdU-ir cells were imaged in three to five sections per animal, yielding 29–52 images of BrdU-positive cells per animal for phenotypic analysis. Image stacks were imported into MBF Bioscience's Image Stack Module for Neurolucida, in which the colocalization was determined using the 3D Visualization feature. The Neurolucida software was used for quantification of cells.

BrdU/Iba1, experiment 3

All BrdU-ir cells were imaged in four sections per animal, with 87–147 BrdU-ir cells per animal for phenotypic analysis. Z stacks containing BrdU-ir cells in the AVPV were taken with an Olympus BX51 microscope under epiillumination using MBF Bioscience's Image Stack Module for Neurolucida. Stacks were taken at 0.5-μm steps spanning the Z-axis of the cell using an UPlanSApo ×40 (0.9 NA) objective and 12-bit color camera. Fluorescein isothiocyanate and tetramethylrhodamine isothiocyanate filters were used to visualize BrdU-ir and Iba1-ir cells, respectively. Red fluorescence was pseudocolored magenta. Colocalization was determined using the 3D Visualization feature, and Neurolucida software was used for quantification of cells.

Statistical analyses

All analyses were initially carried out with hemisphere as a variable. Because no differences in cell number were observed between the right and left hemispheres, the two sides were combined for statistical analyses.

Experiments 1a and 1b

For experiment 1a, an ANOVA with Bonferroni post hoc tests were performed to determine whether the total number of BrdU-ir cells in four anatomically matched sections of the AVPV differed between sexes or as a function of cell age (ie, days after BrdU injection). Thus, the dependent variable was the total number of BrdU-ir cells, and the independent variables were sex (female, male) and cell age (2, 4, 7, 10, 14, 21, and 42 d). Sample sizes for males and females, respectively were: 2 days (6, 5); 4 days (3, 3); 7 days (5, 5); 14 days (5, 7); 21 days (5, 4); and 42 days (5, 5). Hedges' g, a measure of effect size that accounts for small sample size, was calculated at each time point to determine the effect of sex on the total number of BrdU-ir cells in the AVPV. Linear regression analyses were performed on the time course data after first determining that the goodness of fit was best for linear (r² ∼ 0.7) vs quadratic or exponential curves (r² values <0.3). For experiment 1b, Hedges' g was calculated to determine whether there was an effect of sex on the BrdU-ir cells 2 hours after the BrdU injection. The sample sizes were three per sex.

Experiment 2

After analysis, the final sample sizes used to determine BrdU and PCNA colocalization were as follows: 2 hours (2), 4 hours (2), 1 day (2), 2 days (1), 4 days (2), 7 days (2), 10 days (2), 14 days (3), or 21 days (3). To determine whether the percentage of BrdU-ir cells that colocalize with PCNA-ir in the AVPV differs with time after BrdU in female rats, a one-way ANOVA was performed with cell age (time after BrdU) as the independent variable and percentage of BrdU-ir cells that colocalized with PCNA-ir as the dependent variable.

Results

Qualitative results

In single-label studies using diaminobenzidine as the chromagen, BrdU-ir appeared as a dense nuclear dark brown reaction product (Figure 1, A–C). We note that BrdU-ir cells were not confined to the AVPV and were observed in surrounding gray and white matter. Within a given section, the BrdU-ir cells in female and male rats were evenly distributed throughout the AVPV and were not confined to a particular region of the AVPV (eg, the BrdU-ir cells were not clustered along the lining of the third ventricle, Figure 1A). At the 2-hour post-BrdU time point, many of the BrdU-ir cells were paired, forming visible doublets (Figure 1C), indicative of in situ proliferation within the parenchyma (11); doublets remained at all other post-BrdU time points. The intensity of immunolabel within BrdU-labeled cells did not appear to decrease with increasing cell age (ie, time after BrdU); however, the optical density of BrdU-ir was not quantified. In phenotyping studies, the newborn neurons, astrocytes, and microglial cells were also evenly distributed throughout the AVPV.

Figure 1.

The sex difference in pubertally born cells in the AVPV is due to greater proliferation in females compared with males. A and B, Representative bright-field images of BrdU-labeled cells in the AVPV at low- (×10 objective, A) and high-power (×100 objective, B, C). Boxed area in panel A indicates location of panel B. C, A pair of BrdU-ir cells in the AVPV. Image is a minimum intensity projection of 30 slices (15 μm thick, 0.5 μm step) taken using a ×100 oil objective. Scale bars, 10 μm. D, Total number of BrdU-ir cells in the AVPV of four anatomically matched sections. ANOVA revealed a main effect of days after BrdU (F [6, 53] = 3.518, P = .005), in which Bonferroni post hoc tests indicated that the number of BrdU-ir cells was significantly higher 7 days after BrdU compared with 42 days after BrdU treatment (P = .004). Lines indicate line of best fit from regression analyses (n = 3–7 per sex per time point). E, ANOVA revealed a main effect of sex, in which females have more BrdU-ir cells in the AVPV compared with males (F [1, 53] = 5.065, P = .029; n = 34 and 33 for males and females, respectively). F, Two hours after the BrdU injection, there is a medium effect size, with more BrdU-ir cells in the AVPV of females compared with males. Graphs represent means ± SEMs.

Experiments 1a and 1b: the sex difference in pubertally born cells in the AVPV is likely due to greater proliferation in females compared with males

Cells were added to both the female and male AVPV during early puberty (P30), and approximately half of these cells survived at least 6 weeks in both sexes (Figure 1D). An ANOVA was performed to determine whether the number of BrdU-ir cells in four anatomically matched sections of the AVPV differed between females and males as a function of days post-BrdU injection. Main effects of sex and days post-BrdU were found, with females having approximately 1.25 times more BrdU-ir cells in the AVPV compared with males (F [1, 53] = 5.065, P = .029; Figure 1E) and with approximately half the number of BrdU-ir cells present 6 weeks after BrdU compared with 2 days after BrdU (F [6, 53] = 3.518, P = .005; Table 2). At 2 days and 21 days post-BrdU survival times, there was a large effect size such that females had more BrdU-ir cells in the AVPV compared with males, and at 4, 7, and 42 days, after the BrdU injection, there was a medium effect of sex on the number of BrdU-ir cells (Table 2). Bonferroni post hoc tests to compare the total number of BrdU-ir cells at different days after BrdU injection revealed that there were significantly more BrdU-ir cells at 7 days compared with 42 days after BrdU treatment (P = .004). When collapsed across time points and AVPV sections, the mean area analyzed was not different in females and males (0.339 ± 0.02 mm² and 0.334 ± 0.03; Hedges' g = 0.2, no difference), whereas the density of BrdU cells was larger in females than in males (total BrdU/mm²; 94.4 ± 6.9 vs 83.6 ± 7.5; Hedges' g = 1.5, large effect size). Within 2 hours of the BrdU injection, there is a medium effect of sex, with females having twice as many BrdU-ir cells in the AVPV compared with males (Figure 1F). Although the sex difference at 2 hours after BrdU was not statistically significant, the medium effect size suggests that a significant sex difference may be detected if a larger sample size were studied. The absence of an interaction between sex and cell age suggested that the rate of attrition of newly born AVPV cells did not differ between males and females. To substantiate this, regression analyses were performed separately by sex. Lines of best fit are superimposed onto the graph in Figure 1D. The slope of the line of best fit was similar in females and males (−0.42 and −0.52, respectively; P = .77), further indicating that the rate of attrition of newly born cells in the AVPV was similar between females and males.

Table 2.

Effect of Sex on Number of Newborn AVPV Cells

Cell Age (Days Post-BrdU)	BrdU-ir Cells in AVPV		Effect Size (Hedges' g)
	Males	Females
2	28 ± 2	38 ± 7	0.8a
4	28 ± 8	37 ± 9	0.5b
7	33 ± 6	44 ± 10	0.5b
10	26 ± 5	29 ± 4	0.2
14	27 ± 5	25 ± 2	−0.1
21	17 ± 1	29 ± 9	0.9a
42	14 ± 3	20 ± 4	0.6b

Mean total number of BrdU-ir cells in AVPV ± SEM.

^aA large effect of sex (Hedges' g > 0.8).

^bA medium effect of sex (Hedges' g > 0.5).

Experiment 2: BrdU is incorporated into proliferating cells in the AVPV

Two hours after the BrdU injection, 94.4% ± 5.6% of BrdU-ir cells colocalized PCNA-ir, indicating that BrdU was incorporated into proliferating cells and not in cells undergoing DNA repair in the AVPV (Figure 2). An ANOVA revealed a significant main effect of cell age/time after BrdU (F [7, 11] = 3.071, P = .0472). By 24 hours after BrdU injection, virtually no BrdU-ir cells expressed PCNA. At 7, 10, and 14 days, we did observe a small number of BrdU-ir cells that also colocalized PCNA-ir, suggesting that these cells were proliferating on the day of perfusion, possibly the daughters or granddaughters of a cell that was born on P30 (the day of a single BrdU injection). However, the absolute number of BrdU-ir cells was low (ie, an average of five BrdU-ir cells per animal per time point), making the percentages less meaningful for these later time points post-BrdU.

Figure 2.

BrdU is incorporated into proliferating cells in the AVPV. A, Z stack images of BrdU-ir cells colocalized with PCNA-ir in the AVPV. BrdU, PCNA, DAPI, and merged images are maximum-intensity projections of 20 images (20 μm thick, 1 μm step). White scale bar represents 10 μm. Orthogonal view through the middle of the cell confirms colocalization; the red arrow signifies the x plane, the green arrow the y plane, and the blue arrow the z plane. B, Percentage of BrdU-ir cells that express PCNA-ir. Black circles represent data points from individual rats. Graph represents means ± SEMs.

Experiment 3: some pubertally born cells in the female rat AVPV differentiate into neurons, astrocytes, and microglial cells

BrdU was injected every day for 4 weeks during puberty (P28-P56), and brain tissue was collected on P77. Thus, BrdU-ir cells examined in this experiment were between 3 and 7 weeks old. Throughout anatomically matched sections of the AVPV, Z stacks containing BrdU-ir cells were taken and examined for colocalization with the mature neuronal marker, NeuN, the astrocyte marker, GFAP, or the microglia marker, Iba1 (Figure 3A). By P77, more than half of the BrdU-ir cells (57%) had differentiated into mature neurons (15% ± 2% colocalized with NeuN), astrocytes (19% ± 2% colocalized with GFAP), or microglial cells (23% ± 3% colocalized with Iba1); 43% ± 2% of BrdU-ir cells did not colocalize with any of the three markers (Figure 3B).

Figure 3.

Pubertally born cells in the female rat AVPV differentiate into neurons, astrocytes, and microglia. A, Z stack images of BrdU-ir cells colocalized with NeuN-, GFAP-, and Iba1-ir in the AVPV. BrdU, NeuN, GFAP, Iba1, and merge images are maximum intensity projections of 30 images (15 μm thick, 0.5 μm step). White scale bars represent 10 μm. Orthogonal view through the middle of the cell confirms colocalization; the red arrow signifies the x plane, the green arrow the y plane, and the blue arrow the z plane. B, Pie chart represents the mean proportion of BrdU-ir cells that express BrdU only or that also express Iba1-ir, GFAP-ir, or NeuN-ir (n = 6; means ± SEM).

Discussion

Cell proliferation, not cell survival, likely accounts for the female-biased sex difference in the number of cells added to the AVPV during early puberty

When we examined the number of BrdU-ir cells in the AVPV at time points ranging from 2 hours to 6 weeks after BrdU administration on P30, we found that, overall, there were more cells in females than in males. As time since BrdU injection grew longer, the number of BrdU-ir cells grew smaller, reflecting the loss of approximately half of the cells by 6 weeks after cell proliferation on P30. The decrease in the number of BrdU-ir cells with time could reflect cell death or migration of newborn cells out of the AVPV. We detected no difference in the rate of attrition of P30-born cells in females and males and conclude that a female bias in cell proliferation, not cell survival or retention, underlies the sex difference in number of cells added to the AVPV at around the onset of puberty. One caveat to this conclusion is that at the shortest time point studied (2 h after BrdU), although females had twice as many BrdU cells as males, this difference was not statistically significant. In addition, it is possible that there is a sex difference in cell death between 2 and 48 hours after BrdU that was undetected in the present study. However, the most parsimonious explanation of our results is that at least on P30, more cells proliferate in females than in males, leading to a sex difference in P30-born cells 6 weeks later. The sex difference in AVPV volume emerges during puberty; specifically, AVPV volume increases over this period of development in females but not in males (6). We propose that the addition of more pubertally born cells to the AVPV of females than males contributes to the adult sexual dimorphism in AVPV volume.

Our finding that more AVPV cells proliferate in females than in males during puberty is surprising in light of investigations on perinatal sexual differentiation of the AVPV. The earlier studies showed that approximately equal numbers of cells proliferate in the embryonic AVPV in males and females only during days 14–18 of gestation (12) and that exposure of the developing male brain to T promotes AVPV cell death during early postnatal life (13–16). Our previous work shows that between the ages of P20 and P40, more cells are added to the female than to the male AVPV and that this sex difference is driven by ovarian hormones (7). Thus, it appears that two different mechanisms are at play in the perinatal and pubertal periods of sexual differentiation of the AVPV: T-induced cell death preferentially in males during the perinatal period and ovarian hormone-induced cell proliferation preferentially in females during the pubertal period.

AVPV cells born during puberty are born locally

BrdU-labeled cells were present in the female AVPV 2 hours after BrdU injection, and 94% of these cells expressed the cellular proliferation marker PCNA, validating the use of BrdU as a cell proliferation marker and indicating that these are indeed newly born cells. Two hours is not enough time for a new cell to move from the classical neurogenic zones (ie, the subventricular zone of the lateral ventricle and the subgranular zone of the hippocampus) into the AVPV. The subventricular zone of the third ventricle has been recently established as a neurogenic zone (for reviews see references 17 and 18), as have circumventricular organs, including the nearby organum vasculosum of the lamina terminalis (19), and these neurogenic niches may give rise to the newly born AVPV cells. Alternatively, about 15% of newly born cells in the rat AVPV appeared as pairs in the parenchyma, possibly indicative of in situ production (11). Together these data provide evidence that cell proliferation occurs very close to or within the AVPV during early puberty in the female rat.

Neurons, astrocytes, and microglia are added to the female AVPV during puberty

We observed a population of pubertally born neurons (15% of pubertally born cells) in the female rat AVPV. Although the specific neurochemical phenotype of these newborn neurons is presently unknown, at least some of them could be kisspeptin-producing neurons, based on previous findings that the female AVPV contains more kisspeptin neurons than the male AVPV (20–22) and that AVPV kisspeptin neuron number in the female increases from P20 to adulthood (23). Dopaminergic, tyrosine hydroxylase (TH)-ir neurons comprise another sexually dimorphic neuronal population in the AVPV, being more numerous in females compared with males (1), and some newborn neurons in the present study could be dopaminergic. TH and kisspeptin neurons in the adult female AVPV number in the hundreds, not thousands, so the pubertal addition of even just several dozen TH or kisspeptin neurons, or another as-yet-unknown cohort, could be consequential for AVPV maturation and function.

In our cell phenotyping study, more than 40% of the BrdU cells were either microglia (23%) or astrocytes (19%), indicating that glial cells comprise a significant proportion of pubertally born AVPV cells. Recent evidence suggests a role for microglia in cell proliferation and survival, programmed cell death, synaptogenesis, synaptic pruning, and sexual differentiation (for reviews, see references 24–26). Microglia acting in a classical manner, eg, activated in response to neuroinflammation, die within a day of a neuroimmune challenge (27). The BrdU-ir/Iba1-ir cells observed in the current study were presumably derived from resident microglia (28) and were between 3 and 7 weeks old, suggesting that they are involved in regulatory, not inflammatory, processes. Astrocytes within the AVPV are crucial to the generation of the preovulatory LH surge. Specifically, astrocytes synthesize neuroprogesterone, which synergizes with estradiol to provide the neuroendocrine positive feedback required to induce the surge mode of GnRH secretion (29–32). Interestingly, astrocytes from adult female mice respond to estradiol by synthesizing neuroprogesterone (33), but astrocytes from either prepubertal female or adult male mice fail to show this response (32). We conjecture that some of the pubertally born AVPV astrocytes seen in the current study may be estrogen sensitive and contribute to the pubertal gain of positive feedback.

The remaining 40%–45% of BrdU-ir cells that did not colocalize with any of the phenotypic markers examined in this study could be a combination of multiple different cell types, including undifferentiated precursor cells or immature neurons not yet expressing NeuN. Undifferentiated cells or immature neurons could be destined for eventual differentiation or cell death as internal signals or experience-dependent activity might dictate.

Significance and functional implications

We find evidence that the sex difference in pubertal addition of AVPV cells is a consequence of greater cell proliferation, not survival, in females than in males. Although our approach of focusing on cells born on a single day was necessary to precisely map the time course of proliferation and survival of cells, it has a distinct limitation. The half-life of BrdU is about 2 hours (34), and therefore, BrdU administration on a single day results in a small absolute number of BrdU-labeled cells. This not only poses challenges for statistical analysis but also undervalues the full scope of pubertal addition of AVPV cells and its consequences. For example, if BrdU is continuously infused into the lateral ventricle of female rats throughout puberty (P25-P50), well more than 1000 pubertally born cells survive in the AVPV into young adulthood (∼P90 [35]). Thus, extrapolation of our data provides a more comprehensive developmental picture and leads to two important propositions. First, the sex difference in the pubertal addition of new cells to AVPV is of sufficient magnitude to explain in part the pubertal emergence of the sex difference in AVPV volume. Second, the number of pubertally added neurons and glial cells to the female AVPV is substantial, confers plasticity to the AVPV, and is a potential mechanism for the pubertal acquisition of neuroendocrine-positive feedback and mature female reproductive function.

The female AVPV coordinates estrogen-positive feedback stimulation of the GnRH/LH surge, a function that is acquired at puberty and occurs only in females. The current study, in conjunction with prior work (7), suggests a novel mechanism for this sex-specific gain of function, namely the pubertal addition of new cells to the AVPV preferentially in females. Alternatively, because cell death is known to occur during development, and may continue long after, by inference, given the results presented here, perhaps continuous addition of new cells is an important component of the fail-safe mechanism that ensures redundancy in the AVPV, a cell group that is vital to reproduction.