Amyloid β Precursor Protein Regulates Male Sexual Behavior

Animals.

Male B6D2F1 hybrid mice (Mus musculus) were produced in the Jordan Hall Vivarium at the University of Virginia by crossing C57BL/6J females with DBA/2J males. Three different cohorts of male hybrid mice were used for the experiments. One cohort (n = 35) was used for the microarray study, another cohort (n = 40) for the quantitative reverse transcription (qRT)-PCR, and a third for the Western blots (n = 48). Hybrid B6D2F1 male mice were weaned at 20–21 d of age, single-sex group housed until the beginning of each of the experiments (between 50 and 80 d of age), and individually housed afterward for the rest of the experiments. All animals were maintained on a 12 h light/dark cycle (lights off at 12:00 P.M. EST). All of the mice received food (Harlan Diet 8604) and water ad libitum in the University of Virginia Animal Care Facility.

C57BL/6J females were used as stimulus females for MSB testing. Females were ovariectomized and injected subcutaneously with 0.5 μg of estradiol (dissolved in sesame oil) 48 h before testing. Three to 5 h before testing, stimulus females were injected subcutaneously with 0.83 μg of progesterone. The females were group-housed in the same colony room as the experimental males.

APP transgenic mice (B6.129S2-Tg(APP)8.9Btla/J; Jackson Laboratory; stock #5301; n = 10) were developed by transfecting a 650 kb YAC transgene containing entire human APP gene (and ∼350 kb flanking sequence) into 129S2/SvPas-derived D3 embryonic stem cells (Lamb et al., 1993). Founder animals were backcrossed to C57BL/6J for 11 generations. These mice express all mRNA and protein isoforms of the wild-type human amyloid β (A4) precursor protein, APP, and the expression pattern of the various protein isoforms of human APP mimics endogenous mouse gene expression patterns. There is a ∼70% increase of total APP levels in brain extracts of the APP transgenic mice when compared with controls (Lamb et al., 1993). C57BL/6J males from our colony (n = 10) were used as controls. All animal procedures were conducted in accordance with our animal protocol, approved by the University of Virginia Committee on Animal Care and Use.

Behavioral testing.

MSB was tested under dim red lights during the dark phase of a light/dark cycle (Wersinger et al., 1997). All males were habituated 1 h before the introduction of the stimulus female in an 18 × 40 × 11 cm clear Plexiglas testing cage containing the male's home cage bedding. Tests began with the introduction of a hormone-treated stimulus female. Once the male mounted, the test continued to a criterion of a successful ejaculatory reflex or for 120 min, whichever occurred first. If the stimulus female became unreceptive during testing she was replaced with a receptive female.

All tests were videotaped and scored by an observer blind to the classification of the individuals. During each behavioral test, the behavioral components recorded were as follows: mount latency (ML; time from the introduction of a receptive female to the first mount), intromission latency (IL; time from the introduction of a receptive female to the first intromission), and ejaculation latency (EL; interval between the first intromission and ejaculation).

All males received sexual experience before castration. Males that continued to copulate after castration were considered to be “maters” if they demonstrated the ejaculation reflex on at least three of the last four behavioral tests, including the last test. Males used in the gene expression study were tested between weeks 22 and 28 postcastration, for the qRT-PCR experiment testing was performed between weeks 14 and 20 postcastration and males used for protein studies were tested on weeks 16–22 postcastration. APP transgenic mice and controls had 4 weekly MSB tests before castration and were retested for 12 weeks after castration.

Gene expression microarray.

Brains were rapidly dissected and frozen in dry ice and stored at −80°C. Brains were cut into 120-μm-thick coronal sections with a Bright cryostat. Based on the mouse brain atlas (Franklin and Paxinos, 1997), the brain areas containing the mPOA and BNST were dissected bilaterally from ∼8 single sections. Samples were homogenized at room temperature in QIAzol Lysis Reagent (Qiagen) and stored at −80°C until ready to be processed for RNA isolation. Qiagen kits were used to collect total RNA from brain areas containing both the mPOA and BNST from maters (n = 5) and non-maters (n = 5). Sufficiently high quality RNA was generated from each individual animal; thus providing five biological replicates for both groups.

The University of Virginia Biomolecular Facility (BMF) hybridized the mRNA samples to Affymetrix GeneChip Mouse Genome 430 2.0 Arrays that contains probe sets that monitor >39,000 transcripts. The BMF performed a number of quality control checks on the samples including the purity and concentration of the RNA. A number of quality control analyses were performed on the array data including intensity histogram plots, calculation of background levels, and fraction of probes sets detecting RNA. Normalized unscaled SE (NUSE) and relative log expression (RLE) plots, which are sensitive indicators of hybridization quality (Gentleman et al., 2005) were generated. The data were further checked for reproducibility by generating MvA plots which are pairwise scatter plots of M = difference of log intensities (y-axis) versus A = average log intensities (x-axis) for all replicates. All replicates compared well to each other; therefore all 10 arrays were included in the analysis. A principle components analysis also revealed that the maters and non-maters were well separated from each other, supporting the fact that the data were reproducible within each group and there was a significant change in global gene expression patterns when comparing maters and non-maters (supplemental Fig. S1, available at www.jneurosci.org as supplemental material).

Differentially expressed genes were identified by first quantile normalizing all arrays together (Bolstad et al., 2003) and then estimating relative log2 RNA abundance using the GCRMA (Wu et al., 2004) package in the Bioconductor software suite (Gentleman et al., 2005). Modified t tests on the log2 expression estimates using the limma package were performed (Smyth, 2004). The false discovery rate (FDR) (Benjamini and Hochberg, 1995; Benjamini and Yekutieli, 2001; Smyth, 2004) for each gene was estimated (i.e., the estimated rate of false-positives divided by the total positives derived from the t test p-values) to correct for the fact that ∼40,000 statistical tests were performed. A list of differentially expressed genes was derived by applying an FDR cutoff of 5%.

Quantitative RT-PCR.

Total RNA was isolated from mouse brain tissues using RNeasy Lipid Tissue Mini kit (Qiagen) as described by the manufacturer's protocol. The quantity (A ₂₆₀) and quality (A ₂₆₀/A ₂₈₀) of RNA were determined (Bio-Rad SmartSpec Plus spectrophotometer). The cDNA templates were prepared using SuperScript Reverse Transcriptase (Invitrogen). The reverse transcription reaction consisting of 1 μg of total RNA, 500 ng of oligo (dT)_12–18, 500 μm each dNTP, 10 mm DTT, 40 U of RNaseOUT RNase inhibitor, and 200 U of SuperScript RT was incubated at 37°C for 1 h and then heat-inactivated at 70°C for 15 min.

Real-time PCR was performed using ABI Prism 7300 Real-Time PCR System with Sequence Detection Software version 1.2.3 (Applied Biosystems). Separate β-actin endogenous control reactions were used to normalize RNA input. Oligonucleotide primers were designed using Primer Express version 2.0 and synthesized by Invitrogen (supplemental Table S3, available at www.jneurosci.org as supplemental material). The real-time PCR conditions were 95°C for 3 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. After the last PCR cycle, a dissociation melting curve stage was run according to software protocol. Target and endogenous control genes were measured in triplicate for each cDNA sample during each real-time run to avoid intersample variance. For each RNA sample a no-reverse transcriptase reaction was run in parallel to cDNA synthesis, and measured by qRT-PCR to control for contamination and genomic amplification. Each qRT-PCR was verified for a single PCR product of expected size with the disassociation melting curve stage, and some samples were checked with gel electrophoresis.

Normalization and quantifications of the genes of interest, and β-actin mRNA were performed with the comparative cycle thresholds (C _T) method as described in the ABIPRISM7700 sequence detection system user bulletin (#2). Validation experiments were conducted to test for equally efficient target and endogenous control gene amplification as described in the user bulletin. All of the primers were at between 90 and 110% efficient for all amplifications.

Western blot analysis.

Maters (n = 10), non-maters (n = 6) and sexually experienced gonad intact (n = 6) hybrid mice were killed, and the mPOA and BNST were collected from each. Males were tested biweekly for sexual behavior between 10 and 20 weeks after castration; those that ejaculated on at least 5 of the 6 tests, including the last test, were considered maters and those that did not ejaculate on any of the tests were non-maters.

For protein extraction, brain tissues were thawed and homogenized in cold RIPA buffer (0.05 m Tris, 0.9% NaCl, 5 mm EDTA, pH = 7.4) with a protease inhibitor (Sigma). Tissue was homogenized, centrifuged and total protein concentrations were determined by BCA (bicinchoninic acid) Protein Assays (Pierce Chemical Co.). Samples were subjected to electrophoresis on 14% polyacrylamide-SDS gels and transferred to nitrocellulose. Membranes were blocked for 1 h then rinsed and incubated with APP antibodies (Millipore) overnight at 4°C. After rinsing, blots were incubated for 1 h in a horseradish peroxidase (HRP)-conjugated sheep anti-mouse IgG secondary antibody (1:10,000; Jackson Laboratory), followed by detection on x-ray film (Kodak X-OMAT) with SuperSignal West Pico Chemiluminescent Substrate (Pierce Chemical Co.). Later, blots were reprobed with antibody against β-actin (1:50,000; Sigma-Aldrich Corp.), and after rinsing, the blots were incubated for 1 h in an HRP-conjugated sheep anti-mouse IgG secondary antibody (1:10,000; Jackson Laboratory) followed by chemiluminescent detection. The intensities of each of the candidate proteins and β-actin were measured and analyzed by densitometry with ImageQuant (Molecular Dynamics). Levels of each of the candidates were normalized to those of β-actin in each sample and a standard which was run on every gel. The standard used to normalize between blots was the protein concentration of the frontal cortex of a randomly selected hybrid mouse.

Statistical analysis.

Chi-square tests were used to compare differences in the proportion of males displaying copulatory behavior between groups. Repeated ANOVAs were used to analyze the number of mounts and intromissions, as well as mount, intromission and ejaculation latencies of all mice. One-way ANOVAs were used to analyze mRNA levels and protein levels between groups. Post hoc comparisons were conducted using the Fisher Protected Least Significant Difference test where appropriate. Survival analyses were performed using the Mantel-Cox log-rank test to generate Kaplan-Meier curves which showed the percentage of mice displaying MSB at a specific time point during a behavioral test. Any test in which an animal never mounted was a censored data point of 7200 s indicating that the animal did not mount for at least the first 7200 s of the test. Any test in which an animal did not mount was dropped from the dataset for latency to intromit and ejaculate. The latency to intromit in any test in which an animal mounted but did not intromit was a censored data point calculated as 7200 s minus the latency to mount, and the test was dropped from the dataset for latency to ejaculate only. The latency to ejaculate in any test in which an animal intromitted but did not ejaculate, was a censored data point calculated as 7200 s minus latency to intromit starting from when the female was introduced. Observed differences were considered significant if p < 0.05. Statistical tests were run using the Statview program (Statview 5; SAS Institute).