 |
Loss of Sex Discrimination and Male-Male Aggression in Mice Deficient for TRP2
Lisa Stowers, Timothy E. Holy, Markus Meister, Catherine Dulac, Georgy Koentges
|
Supplementary Material
Supplemental Figure 1. Construction of TRP2-/- mice. (A) Targeting strategy for inactivation of the TRP2 gene. Map of the 32 kb region surrounding the
mTrp2 genomic locus, with exons indicated by open boxes. Sequences encoding the putative transmembrane domain 4, TM4, through the pore region of a 13.4 kb genomic clone were replaced with a loxP-PGKneo-loxP cassette. This modified construct was used to create the deletion in ES cells by homologous recombination.
(B) Confirmation of the
Trp-/- deletion by Southern analysis. Modified alleles result in a 4 kb BamH1 fragment detected by the probe represented in
(A).
(C) Western blot analysis performed on VNO extracts with a polyclonal antibody directed against the C-terminal end of TRP2
(1) confirms the absence of TRP2 protein expression in the VNO of the homozygous TRP2-/- mutant.
Medium version | Full size version
Supplemental Figure 2. VNO receptor survival in TRP2+/- and TRP2-/- mice. In situ hybridization of VNO neuroepithelium from 2 month old mice with antisense probes of pheromone receptor mixes and G-protein probes. The number of V1R-V3R neurons in the TRP2-/- mice is reduced by ~50% and the number of V2R neurons is reduced by ~75% in comparison with TRP2+/- mice. Trp+/-: V1R-V3R=55 (n=5), V2R=158 (n=5) positive neurons/section. TRP2-/-: V1R-V3R=30, V2R=52 (n=5) positive neurons/section. The dashed line indicates the external limit of the neurepithelium bordering the VNO nerve layer. Scale bar, 80
m.
Medium version | Full size version
To view these movies, download a QuickTime viewer.
- Movie 1
- Movie 2
- Movie 3
(A) Male odor cues are necessary for male-male aggression. Wild-type resident (right at start of sequence) with castrated intruder (left at start of sequence). Resident investigates the environment but does not behave aggressively in the presence of castrated intruder. (B) Same mouse pair as shown in (A), but the castrated intruder is now swabbed on back and urogenital region with male urine. The wild-type mouse is at the right at the start of the sequence. Aggressive behaviors include tail rattling and attacks include biting and wrestling. (C) TRP2-/- male (left at start of sequence) does not behave aggressively in presence of male odor cues. Castrated intruder swabbed with male urine is at the right at the start of the sequence. TRP2-/- mutant displays sexual behavior of mounting and thrusting towards male intruder. These videos were recorded on the same day using the same pool of male mouse urine to swab indicated intruders. The 20s sample video clips all occur approximately 300s into a 600s trial.
Supplementary Materials and Methods:
Construction of TRP2-/- mice.
A rat Trp2 cDNA fragment (1) was used to retrieve a full-length (3368bp) mouse Trp2 cDNA clone from a C57Bl/6J male VNO HybriZAP cDNA library. The sequence of mouse TRP2 is as reported (2). From a mouse 129/SV genomic library (Stratagene) a 19kb clone (41 containing 5' regions and exons 1-3 of the locus) and a 13.4kb clone (27, containing exons 3-12) were mapped. A deletion of exons 7-10 was generated by ligating a 5.9kb arm spanning exon 3 to 6 and a 3kb arm spanning exons 11 and 12 on both sides of a neomycin resistance cassette. A loxpPGKneoloxp cassette was introduced with a BamHI site at its 3' end for diagnostic purposes. For detection of the Trp2-/- mutant allele, a probe external to the 3' end of the targeting construct was identified and isolated from a 129 genomic library (Stratagene). Linearized targeting construct was electroporated into W9.5 129/SV ES cells that were cultured as described (3). Two different clones carrying a targeted allele were injected into C57Bl/6J blastocysts to produce three male germline chimeras, which were mated with C57Bl/6 females.
Western Blot Analysis
Protein was extracted by dounce homogenization from the VNOs of two animals of each genotype. Equivalent protein amounts were resolved through 9.5% SDS-PAGE and transferred to nylon. The blot was blocked in TBS, 0.1% Tween, and 6% non-fat dried milk. Anti-TRP2 polyclonal antibody (1) was used 1:1000, the secondary anti-rabbit alkaline phosphatase was used 1:10,000. A band of approximately 110 kD was visualized using Super Signal (Pierce). To control for total protein loaded, the blot was washed in TBS, blocked and then re-probed with anti-
-Tubulin, 1:1000 that detects a band at 34 kD.
Electrophysiology
Electrode array recordings were performed as described previously (14). Wire bundles were fabricated from nichrome wire, 12m in diameter with 2m polyimide insulation (Kanthal). A bundle of 31 nichrome wires and one silver wire reference, threaded through a 30-gauge Teflon tubing sheath, was cut flush so that the recording tips covered a region approximately 150m in diameter. The tips were platinized to reduce the impedance, typically to ~500 kOhm. The entire bundle was positioned with a single micromanipulator. Adult VNO neuroepithelium was mechanically detached from the blood vessel and connective tissue in 4°C Ringer's solution. For electrode array recordings, the basal lamina was also mechanically removed and the tissue was mounted on the array (4) and maintained under superfusion at 34°C for at least one hour before recording commenced. For wire bundle recordings tissue was held flat with a slice holder under superfusion at room temperature. The wire bundle was advanced into the tissue until a significant fraction of electrodes detected spontaneous spiking activity. Recording commenced after a 10 min recovery period. At the completion of the stimulus cycle (see below) the wire bundle was either advanced 50-100m into the tissue or retracted and moved horizontally to a new region of the epithelium and the entire sequence repeated at the new location. Ringer's (in mM) NaCl 115, KCl 5, CaCl2 2, MgCl2 2, NaHCO3 25, HEPES 5, glucose 10, equilibrated with 95% O2/5% CO2 to pH 7.4. This solution was used for superfusion as a negative-control stimulus (see below), and for diluting urine stimuli. High-potassium solutions were made by substituting equimolar quantities of KCl for NaCl. Stimuli were delivered from the tip of a 13-to-1 micromanifold (DAD-12, ALA Scientific) positioned a few hundred micrometers from the tissue. Flow from this tip was continuous, alternating between stimuli and a Ringer's flush solution under the control of 13 computer-actuated valves. All stimuli were presented once in sequence, and the entire cycle repeated four to nine times. Electrical signals were amplified and filtered, and then digitized at 10kHz with a 12-bit 64-channel A/D card (National Instruments) and written to disk for later analysis. Urine used in electrophysiological recordings was collected in customized cages so that voided urine fell directly into a bucket of liquid nitrogen. After several hours, balls of frozen urine were pooled from eight animals of the same age and sex. Urine was centrifuged (2000 g for 90 s) and frozen in aliquots and stored at -80°C.
Electrophysiology Data Analysis
On each electrode, threshold-crossing events were identified in the recorded waveform. The threshold was set to six times the average absolute deviation from the mean. Firing rate changes upon stimulation were computed as 
r=r1-r2, where r2is the firing rate (number of spikes/unit time) in the ith time bin. r2 was computed from the 10 s before valve opening, while r1 was computed from a period after valve opening equal to twice the valve open time. Urine stimuli and Ringer's controls were delivered for 5 s, while 50mM KCl solutions were delivered for 1 s. A few sites showed responses to a Ringer's solution (negative control) with r> 2.5Hz (array recordings, TRP2+/-: 1 out of 131 recording sites; array recordings, TRP2-/-: 0/117; wire bundle recordings, TRP2+/-: 8/236; wire bundle recordings, TRP2-/-: 14/436), and these were excluded from further analysis. Of the remaining recording sites, only those responding to 50mM KCl (positive control) with r> 5Hz were included in the analysis; these will be called "active sites." To make a systematic comparison between animals of different genotypes, we performed many of the recordings blind with respect to genotype. These recordings were subjected to a first-pass analysis which identified "active" electrodes. Only these active sites were used for further analysis. In array recordings, these criteria were satisfied by 90 sites from 3 preparations of TRP2+/- animals and 105 sites from 5 preparations of TRP2-/- animals. In wire bundle recordings, we obtained 70 sites from 3 preparations of TRP2+/- animals, and 157 sites from 4 preparations of TRP2-/-. In array recordings of 3 preparations of TRP2+/- containing 90 active sites, we found that 53 sites had elevations in firing rate of 5Hz or higher to female mouse pheromones, and 20 sites showed response to male mouse pheromones. By contrast, recordings from TRP2-/- mice never showed a significant change in firing rate in response to pheromonal stimuli. In 5 experiments comprising 105 active recording sites, we consistently failed to detect any elevation in firing rate higher than 5Hz in response to male or female pheromones. Similar results were obtained when we recorded with a bundle of 31 wire electrodes inserted from the apical side of the tissue. With both array and wire bundle recordings, pairs of channels would occasionally record from the same unit, as determined by a significant degree of synchrony over time scales of a few hundred microseconds. We did not correct for this infrequent duplication, as these channels usually recorded from additional units not common to the pair.
In situ and histological analysis
Mice were perfused with 4% paraformaldehyde, VNOs were removed, allowed to fix an additional two hours and cryoprotected (30% sucrose). Tissue was frozen in OCT (Miles), and 15-30m coronal sections were cut on a cryostat. Sections were treated according to standard histochemistry protocols. Whole-mount X-Gal staining was performed as described (3).
Mating Behavior
Testing occurred in three trials over a one-week period during the first three hours of the subject's dark phase. In the home cage with the feeding tray removed, the socially naive resident male accommodated to the test environment for 15 min prior to the start of each assay. An estrus female, which status was determined after vaginal lavage, was introduced into the resident's cage and the behavior was recorded through infrared light of the digital video recorder (Sony). The male's behavior including mounting, intromission and ejaculation was subsequently analyzed using Observer Video Pro (Noldus).
Male-Male Aggression
Territorial aggression assays were based on the resident intruder paradigm (5). C57Bl/6J intruder animals were surgically castrated at 6-7 weeks by Charles River. Testing occurred during the last three hours of the subject's light phase. Fresh urine was obtained in a separate room from either 129/SV-C57Bl/6J TRP2+/+ males or C57Bl/6J males, pooled from several individuals and kept at room temperature until utilized. Testing occurred in the resident's home cage. Digital video recording began when the castrated male intruder was introduced into the resident's cage. The intruder was removed after the 10 min trial. In a subsequent trial, a castrated male intruder was introduced that had been swabbed by pipette tip with 50ul of pure male urine (40ul on the back and 10ul on the urogenital region). A different intruder animal was used for each resident. Upon application of urine the intruder was not re-used for at least four days to allow urine ligands to dissipate. 7-12 resident animals were each assayed in 1-3 trials. Gender choice was conducted similar to the resident/intruder assay except both a urine swabbed castrated male and an estrus female (determined by vaginal lavage) were simultaneously introduced into the resident male's cage for 10 min. Digital video data was transferred onto a recordable CD using Broadway Pro (Data Translation). Observer Video Pro (Noldus) facilitated the analysis of each assay at 1/5-1/2 speeds. Non-aggressive behavior of the resident included periscoping, self-grooming, and digging. Non-aggressive social behavior included licking and smelling the intruder and sexual mounting. Aggressive behavior included tail rattling, biting, chasing, tumbling/wrestling, and cornering the intruder. An attack is a subset of aggressive behavior that contains contact between the resident and the intruder such as biting or wrestling.
Testosterone RIA analysis
Animals were anaesthetized by an IP injection of 2.5% avertin. 100ul of blood was removed by intraorbital bleeding with a heparinized capillary tube and stored on. After centrifugation 50ul of serum was removed and stored at -20(C prior to analysis. Testosterone concentration was measured following the exact protocol of Testosterone 125I RIA Kit (ICN).
Ultrasonic vocalization analysis
The resident cage with feeding tray removed was placed inside a wood recording box that had one plastic side for observation. The lid contained two ports; through one a microphone was used to gather high frequency vocalizations. The second port connected to a fan to allow circulation of fresh air. Animals accommodated to the test environment for 10 min immediately prior to testing. Video and acoustic recording began 1 min prior to the introduction of intruder animals to test for baseline vocalizations. Recording continued 10 min after intruder introduction. Sounds over the frequency range 20 Hz-110 kHz were recorded with a microphone and amplifier (Bruel & Kjaer), digitized at 250 kHz, and saved to disk. The sonogram was computed using 256 frequencies and averaging 64 successive non-overlapping blocks, resulting in time bins 130 ms in width. Power from ultrasonic vocalizations was confined to the frequency band from 55-110kHz. In trials with a significant vocalization response, these vocalizations were the major source of power in this band. However, there were other events (e.g., attacks and mounting) that introduced power into this band observed, albeit with a different spectral profile. To reject such spurious events, time bins containing vocalizations were identified by requiring that the maximum power in the band 55-110 kHz exceeded the maximum power in a "control" frequency band 25-35 kHz by a factor of 3. The average 55-110 kHz power in such time bins during the 10 min presence of the intruder is reported as the vocalization power.
References and Notes:
- E. R. Liman, D. P. Corey, C. Dulac, Proc. Natl. Acad. Sci. USA 96, 5791 (1999).
- T. Hofmann, M. Schaefer, G. Schultz, T. Gudermann, Biochem J. 351, 115 (2000).
- L. Belluscio, G. Koentges, R. Axel, C. Dulac, Cell 97, 209 (1999).
- T. E. Holy, D. Dulac, M. Meister, Science 289, 1569 (2000).
- J. A. Maruniak, C. J. Wysocki, J. A. Taylor, Physiol. & Behav. 37, 655 (1986).
