Targeted Microlesions Reveal Novel Organization of the Hamster Suprachiasmatic Nucleus

Introduction

Virtually all behavioral and physiological processes exhibit circadian cycles with a period of ∼24 hr. The role of the suprachiasmatic nucleus (SCN) in generating these circadian rhythms in physiology and behavior is well established (Moore and Eichler, 1972; Stephan and Zucker, 1972; Lehman et al., 1987; Ralph et al., 1990). Over the last decade, enormous progress has been made in uncovering the mechanisms responsible for circadian oscillations at the cellular and molecular levels (for review, see Okamura et al., 2002; Panda et al., 2002). To understand rhythm generation in multicellular organisms, it is important to investigate how the molecular mechanisms translate to network properties at the tissue level and to the behavior of the whole organism.

The rodent SCN are composed of ∼10,000 cells in each nucleus (van den Pol and Tsujimoto, 1985; Moore et al., 2002). At the tissue level, the SCN of rats and mice have been characterized into “core” (ventrolateral) and “shell” (dorsomedial) subregions based on the pattern of neural inputs and outputs as well as in the organization of phenotypically distinct neuronal cell types (Moore, 1996; Moga and Moore, 1997; Moore et al., 2002). Recently, we have characterized the hamster (Hamada et al., 2001) and mouse (Karatsoreos et al., 2004) SCN as having two functionally distinct regions, one exhibiting light-induced Period (Per) expression and a nonoverlapping region expressing Per rhythmically (see below). In view of this complex topographical and functional organization, it is interesting to note that the SCN has historically been considered to be a functionally homogenous structure. For example, lesion studies suggest that survival of only a small fraction (∼25%) of any portion of the SCN is sufficient to support rhythmic function (Rusak, 1977; Van den Pol and Powley, 1979; Davis and Gorski, 1988; Harrington et al., 1993; Welsh et al., 1995). More recent work on clock gene expression and electrical activity in neurochemically defined SCN subregions suggest that the SCN is not functionally homogenous but instead that different subregions serve different functions (Hamada et al., 2001; Nakamura et al., 2001; Jobst and Allen, 2002).

Syrian hamsters represent an ideal animal model to explore questions of SCN organization, because functional subdivisions have been well-characterized by their chemoarchitecture, providing convenient markers. Our laboratory has shown that lesions bilaterally destroying a small portion of the SCN marked by calbindin-D_28K (CalB), while sparing up to 67% of the remainder of the SCN, abolish rhythms in locomotor behavior (LeSauter and Silver, 1999). Per1 and Per2 are light induced, but not endogenously rhythmic, in this CalB subregion. In contrast, the SCN shell, marked by vasopressin-containing cells, has rhythmic Per1, Per2, and Per3 (Hamada et al., 2001). One hypothesis that may account for these findings proposes that cells in the CalB region are necessary to coordinate independent oscillator cells in the SCN shell to prevent these cells from drifting out of synchrony (Antle et al., 2003). Following from this hypothesis, lesions destroying the CalB portion of the SCN, while sparing other SCN compartments, should abolish all circadian rhythms in physiology and behavior. Alternatively, the CalB subregion may represent a subdivision uniquely important for arousal measured by locomotor rhythm generation or maintenance, and distinct neurochemical subregions of the SCN may be associated with different rhythmic responses. Distinguishing between these competing hypotheses is critical for establishing SCN functional organization.

Materials and Methods

Animals and housing. Eighty-three adult male LVG hamsters (Mesocricetus auratus) were used in the present experiment. Seventy-six animals were lesioned, whereas seven intact animals were used to obtain serum for normal cortisol and melatonin patterns (see below). Animals were purchased from Charles River (Wilmington, MA) at 4-5 weeks of age. All animals were housed in translucent propylene cages (48 × 27 × 20 cm) and provided with ad libitum access to food and water for the duration of the study. Cages were equipped with running wheels (17 cm diameter) connected to a computer (Dataquest; Datasciences International, St. Paul, MN). Animals were maintained in a colony room with a 24 hr light/dark cycle (14/10 hr light/dark) at 23 ± 1°C. At least 1 week before the start of the experiment, animals were transferred to cages equipped with running wheels (see below) and housed in constant darkness (DD). All experimental protocols conformed to the Institutional Animal Care and Use Committee guidelines of Columbia University.

Lesions. After 1-2 weeks in DD, 76 male hamsters (110-120 gm) were anesthetized deeply with 60 mg/kg ketamine and 5 mg/kg xylazine. The head of the animal was shaved and positioned in a stereotaxic apparatus (David Kopf Intruments, Tujanga, CA), and the animal was prepared for aseptic surgery. Lesions were aimed at the following coordinates: 0.8 mm anterior to bregma, 0.1 mm lateral to midline, and 7.9 mm below the dura. Bilateral electrolytic lesions were made using a Grass Instruments (Quincy, MA) LM-5 lesion maker and stainless steel electrodes insulated with Epoxylite (The Epoxylite Corp., Irvine, CA), excluding the tip (0.20 mm). Current was passed for 10 sec at 0.55 mA.

General methodology and caveats. For the purposes of comparison, records from a single animal with an intact SCN have been depicted for all measures. The data for all behavioral and physiological measures were collected sequentially from all animals. As a result, minor differences in the period of each rhythm may be observed within an individual animal. However, these differences are consistent with changes in the circadian period that are commonly observed over time. All rhythms measured simultaneously exhibited equivalent periods.

Locomotor behavior. One week before and at least 3 weeks after lesions, locomotor behavior was measured in all animals using cages equipped with running wheels (17 cm diameter) connected to a computer (Dataquest; Datasciences International). Cumulative counts were recorded every 10 min.

Drinking behavior. After the lesions, drinking behavior was monitored continuously by placing a wire grid on the bottom of the cage connected to the ground of a circuit closure. The positive lead of the sensor was attached to the spout of the water bottle. Every time the hamster drank from the spout, the circuit was closed, and a count was recorded. Cummulative counts were transmitted to a computer (Dataquest) and recorded every 10 min.

Gnawing behavior. After measures of locomotor and drinking behavior, gnawing rhythms were measured as described previously (LeSauter and Silver, 1994). Eighteen animals showing no disruptions in rhythmic locomotor and drinking behavior, suggesting a missed lesion, were eliminated from further study. For all other animals, a wooden stick was attached to the cage top with a clamp holder. A switch mechanism was mounted on the top of the clamp and to the side of the cage. Gnawing on the stick caused the clamp to move vertically, deflecting the wire attached to the clamp into the wire attached to the side of the cage, thereby closing the circuit. Switch closures were recorded continuously, and cumulative counts were sent to a computer (Dataquest) every 10 min for a minimum of 3 weeks.

Monitoring of heart rate and body temperature. After the monitoring of gnawing behavior, physiological transmitters (TATTCTA-F40; Datasciences International) were surgically implanted into animals that were deeply anesthetized with 60 mg/kg ketamine and 5 mg/kg xylazine. The ventral surface of the abdomen and the thorax to the area of the axilla, slightly cranial on the right side, was shaved. The shaved area was swabbed with Betadine. A midline abdominal skin incision 1 cm below the diaphragm and no >2 cm in length was made. The abdomen was opened by making a 2 cm incision along the white line of fascia where the abdominal muscles join on the midline. The body of the transmitter was slipped into the abdominal cavity along the sagittal plane and dorsal to the digestive organs so that both leads were oriented toward the head of the animal. For heart rate measurement, the heart leads were threaded through two small holes in the external oblique muscle of the abdominal wall to the right and left of the incision. For negative lead attachment, using a trochar and sleeve, the lead was passed under the skin to an incision made near the clavicle. Similarly, for positive lead placement, the lead was passed under the skin to an incision made cranial to the last rib. The leads were then secured against the chest muscles by attaching a suture just behind the tabs, and the abdominal opening and the smaller two lead access points were closed. After surgery, the animals were returned to their cages. The animals were allowed 1 week to recover after surgery. To monitor heart rate and body temperature, the animals' cages were placed on an RA-1010 receiver plate (Datasciences International) connected to a computer. The plates were situated so that the hamster was always within range of the receiver plate (even when rearing). Body temperature and heart rate were transmitted and monitored continuously. Mean body temperature and heart rate were recorded every 5 min using Datasciences ART software (Datasciences International).

Gonadal measurements. At the time that transmitters were implanted (8-10 weeks after exposure to DD), the scrotum was dampened with alcohol, and the length and width of the left testis was measured to the nearest 0.1 mm with calipers. The estimated testis volume (ETV) was calculated as: ETV = testis width² × testis length.

Blood sampling. After heart rate and body temperature measurements, animals were implanted with a jugular catheter for future blood sampling. An additional seven control animals not receiving surgery or prior monitoring were also catheterized to obtain standard hormone values. Animals were anesthetized with 60 mg/kg ketamine and 5 mg/kg xylazine, and a catheter made out of SILASTIC tubing (inner diameter, 0.025 inch; outer diameter, 0.047 inch; Dow-Corning, Midland, MI) was inserted into the right jugular vein. The free end of the catheter was passed under the skin and exited near the base of the skull. The catheter was filled with heparinized saline (10 U/ml) and connected to a stainless steel tube (26 gauge) that was sealed with a piece of SILASTIC tubing that was closed at one end. The end of the tube was embedded in dental cement for protection and anchored into the skull with four screws. The animals were allowed 5 d to recover before blood sampling began. For animals that displayed rhythmic behavior, blood sampling began at a random circadian time for each animal (circadian time 1, 4, 7, 10, 13, 16, 19, and 22) to control for effects of prior sampling on later samples. For arrhythmic animals, blood sampling began at a random local time and continued every 3 hr thereafter for 24 hr. To collect blood samples from freely moving animals, 12-15 inches of PE 50 tubing was attached to the tube on the end of the catheter. A syringe was attached to the end of the tubing, and a 0.3-0.4 ml sample of blood was withdrawn. An equal amount of saline was replaced after each sample. Samples were centrifuged at 3000 rpm, and serum was separated. Serum samples were stored at -80°C until assayed.

Hormone assays. Enzyme immunoassay (EIA) kits were used for measuring serum levels of cortisol and melatonin (Alpco Diagnostics, Windham, NH). Melatonin was measured after extraction by reverse-phase column chromatography. The cortisol EIA did not require an extraction procedure. The samples were assayed in duplicate according to the manufacturer's instructions. The sensitivity of the cortisol assay was 8.4 nmol/ml, whereas the sensitivity for melatonin was 0.34 pg/ml. The intra-assay variability for the melatonin and cortisol assays was 5.6 and 3.9%, respectively. Both assays were validated in the present study for use in hamsters by assessing parallelism with the standard curve in serially diluted pooled hamster serum. All samples were run in single assay (one for cortisol and one for melatonin) in duplicate.

Perfusion and histology. One day after blood sample collection, hamsters were anesthetized deeply with sodium pentobarbital (200 mg/kg) and perfused intracardially with 150 ml of 0.9% saline, followed by 300-400 ml of 4% paraformaldehyde in 0.1 m PBS, pH 7.3. Brains were postfixed for 2-3 hr at 4°C and cryoprotected in 20% sucrose in 0.1 m PBS overnight. Coronal sections were cut on a cryostat and processed as free-floating sections. Every fourth section was stained for vasopressin or vasoactive intestinal polypeptide (VIP), whereas every other section was stained for CalB. Polyclonal antibodies raised against vasopressin and VIP (Incstar, Stillwater, MN) were diluted at 1:10,000. The monoclonal CalB antibody (Sigma, St. Louis, MO) was used at 1:20,000. Antisera were detected using either biotinylated goat anti-rabbit (vasopressin and VIP) or biotinylated horse anti-mouse secondaries (Vector Laboratories, Burlingame, CA) and an avidin-biotin-HRP conjugate (Vector Laboratories) with DAB (Polysciences, Warrington, PA) as the chromogen. Sections were mounted onto glass slides, dehydrated in a graded series of ethanols, and cleared in xylenes, and coverslips were applied.

Microscopy and analysis of the SCN. To assess the extent of the lesions, we used vasopressin and VIP as markers for distinct SCN subregions. The presence of vasopressin-immunoreactive (-ir) and VIP-ir neurons in adjacent sections of the SCN was taken as evidence of surviving SCN tissue. Whether or not CalB-ir cells remained after the lesion was assessed as described previously (LeSauter and Silver, 1999). Images of brain sections were captured using a CCD video camera (XC77; Sony, Tokyo, Japan) attached to an Olympus BH-2 microscope (Olympus, Melville, NY). The extent of SCN tissue surviving ablation was assessed using NIH Image and measuring the remaining SCN tissue in sections stained with peptidergic markers. Total SCN size averaged over two intact hamsters of the same age and body mass was used as a point of comparison to estimate the percentage of SCN tissue remaining. Restated, surviving SCN tissue was measured in lesioned animals, and the percentage of SCN tissue remaining was calculated as: percentage of surviving tissue = (two-dimensional area of surviving tissue/two-dimensional area of intact SCN) × 100.

Delineation of SCN subregions. Various terms have been used to demarcate distinct SCN subregions. Here, we have defined subregions based on chemoarchitecture. For example, we used the descriptive terms “CalB subregion” and “vasopressin subregion” to delineate nonoverlapping regions of the SCN, recognizing that these regions contain other cell phenotypes. Because putative functional characteristics of the hamster SCN map onto neurochemical organization (Hamada et al., 2001, 2003), we felt that this type of analysis was most appropriate for the present investigation. Additional analysis will be required to determine how hamster SCN regions characterized by their clock gene expression correspond to those previously classified as “core” and “shell” in studies of rats and mice (Moore, 1996; Abrahamson and Moore, 2001; Moore et al., 2002)

Analyses and statistics. The power of all rhythms was assessed using a Fourier analysis (Dataquest or Clocklab programs). An animal was considered rhythmic when the highest peak occurred at ∼1 cycle/d, with an absolute power of at least 0.005 mV/Hz. For hormonal data, rhythms from individual animals were considered rhythmic if two consecutive sample values were greater than 2 SDs above the mean nadir values of unlesioned control animals.

For animals that maintained rhythmic function after the lesions, a Pearson Product Moment correlation was performed between the power of the rhythm and a qualitative measure of vasopressin, VIP, and CalB surviving ablation. The extent of sparing of cells and fibers was estimated as follows: as noted above, every fourth section was immunostained for vasopressin and VIP, whereas alternate sections were stained for CalB. Each of these sections were examined by an investigator blind to the experimental condition of the animal and scored on a scale of 1-5, in which 1 represents <25% of the intact nucleus, 5 represents an intact SCN, and a unilateral lesion is scored as 3.

Results