Lin et al. 10.1073/pnas.0701106104.

Supporting Information

Files in this Data Supplement:

SI Figure 6
SI Figure 7
SI Figure 8
SI Figure 9
SI Movie 1
SI Movie 2
SI Movie 3
SI Methods
SI Figure 10




Fig. 6. The waveforms and basic firing properties of the three nest-responsive cells. The interspike-interval histogram of each nest cell is shown above the image. The waveform is included as an inset on the top right. The autocorrelogram of the same cell is shown below each image. (The horizontal scale bar is 200 ms, and the vertical bar is 50 mV.)





Fig. 7. Robust firing changes of the other five nest-responsive cells (Cells #4-8). The snapshots of the recorded video are presented at the top. The corresponding firing rate histograms are listed below. Please note that Cell #6 and Cell #7 were simultaneously recorded from independent electrodes of the left hippocampus in mouse F but exhibited opposite patterns. The gray lines show the animal's movement paths, and the red segments indicate the duration in which the mouse was approaching the nest (for Cells #4 and #5) or moved inside the nest (for Cells #6, #7, and #8).





Fig. 8. The nest-responsive cells did not respond to other episodic stimuli and events. (A) Although Cell #1 responded robustly to nests, it was unresponsive when the mouse encountered sudden air-blows. (B) Cell #1 also did not respond to earthquake-like shakes, which are also known to elicit responses in some hippocampal cells. The effectiveness of those stimuli to trigger responses of some hippocampal cells is evident from other simultaneously recorded units from the same animal. One representative air-blow cell is shown on the left side, and one representative shake cell is shown on the right side. (C) Cell #1 also did not respond to other objects such as toys. The numbers listed in the snapshot indicate the position of encounters. A total of seven repetitions were performed for perievent spike raster and perievent spike histogram analyses. The time bin of the perievent spike histograms is 250 ms.





Fig. 9. Invariant responses of other types of nest cells to changes in the physical appearance, shape, and construction materials of nests. (A) Cell #2 of mouse B exhibited persistent-on-type firing once the mouse entered nests that were made in different shapes or different styles, including a square nest (Left), a triangular nest (Center), and a circular plastic nest (Right). (B) Cell #3 of mouse C exhibited a persistent-off-type response once the animal moved into nests made in the circular shape (Left), square shape painted with stripes (Center), or constructed from a metal cap (Right). The firing rate histograms are shown beneath the video snapshots. The gray lines show the animal's movement paths, and the red segments indicate the duration in which the mouse was inside the nests.





Fig. 10. Examples of a variety of nests that were used in these experiments. The top two nests (from the left) were used as home nests with water cups glued next to the nests. The mice have been housed for 2 or 3 d before the surgery in the plastic bucket (50 cm in diameter) in which home nests, water, and food pellets were provided. All other nests were newly constructed and kept away from mice so that mice had no prior experiences with the objects. The function of these objects as nest was further confirmed by measuring sleep occupancy. (Red scale bars, 5 cm.)





SI Movie 1

Movie 1. "Transient-on" type of nest cell responses and its functionality-based encoding of conceptual knowledge of nests. The first movie segment shows that Cell #1 of mouse-A exhibited "transient-on"-type responses to the home nest but not to another similarly shaped, smaller circular object (water cup). The second movie segment shows that Cell #1 did not respond to the plastic nest that was placed in an inverted manner (so that it would function as a small stage). However, once the plastic nest reverted back to its normal nest position, the cell exhibited robust firing up to 40 Hz. Because it is the same object and placed at the same location, this inversion experiment demonstrates that the nest cell encodes the functionality of the nest rather than merely physical appearances, materials, or spatial location, etc.





SI Movie 2

Movie 2. "Persistent-on" type of nest cell responses. The first movie segment shows that Cell #2 exhibited persistent-on-type responses to the home nest. Mouse B entered the home nest twice from two different directions, and the cell fired persistently as long as the mouse was inside of the nest. The second movie segment shows that Cell #2 responded significantly to a nest buried into the ground (the underground nest). This also illustrates that the functionality of nests is important in determining the firing property of the nest cell.





SI Movie 3

Movie 3. Persistent-off-type responses of nest cell and its functionality-based encoding of conceptual knowledge of nests. The first movie segment shows that Cell #3 of mouse C exhibited persistent-off-type responses to the nest. This cell would turn off its firing as soon as the mouse entered the nest. The second movie segment shows that the nest Cell #3 did not cease its firing when the animal crossed over the inverted plastic bottle cap (which functions as a small stage). However, once the plastic cap was flipped back into a nest position, the cell ceased to fire when the mouse moved into the plastic cap. Because it is the same object and placed at the same location, this inversion experiment again demonstrates that the nest cell is tuned toward the functional features that define nests.





SI Methods

Construction of Recording Microdrives and Animal Surgery.

The 96-channel electrodes, consisting of two independently movable bundles of 48 stereotrodes or 24 tetrodes (48-channel on each side of the hippocampi), were constructed as described (1, 2). Each stereotrode or tetrode was constructed by twisting a folded piece of two or four wires (STABLOHM 675, H-FORMVAR, 25 mm for stereotrode and 13 mm for tetrode, California Fine Wire), securing the two strands together with a low-intensity heat source, and removing the insulation from the tips of the free ends over an open flame. After all electrodes had been inserted into separate polyimide tubes, the twisted ends of the wires were cut to a length that extended 3-4 mm beyond the end of the polyimide bundle, and the wires were then secured to the polyimide tubing with glue. A reference wire (magnet wire, 0.01 sq mm, Belden electronic division) was soldered to the four pins on the ends of each connector array. In addition, the tips of the tetrode were plated with gold (Cyanida Gold solution, SIFCO Selective plating) to a final impedance of 500-800 kW.

Two or 3 d before surgery, wild-type B6BCA/J mice were removed from the standard mouse cage and placed in the customized recording home environment (plastic bucket, 50 cm in diameter) in which a cardboard circular nest, a water cup, and food pellets were provided. Mice were handled for 10-15 min each day before surgery to minimize the potential stress of human interaction. On the day of surgery, the mouse was anesthetized with an i.p. injection of 60 mg/kg ketamine (Bedford Laboratories, Bedford, OH) and 4 mg/kg Domitor (Pfizer Animal Health, New York, NY). The mouse head was immobilized in a stereotaxic frame and the positions for the two bundles (2.0 mm lateral to the bregma and 2.3 posterior to the bregma on the both right and left sides) were measured and marked. The stereotaxic apparatus was then used to lower the electrode bundles into the mouse's cortex. The gap surrounding the electrodes was filled with softened paraffin, and the headstage was stabilized with dental cement. The reference wire attached to the two posterior head screws was then soldered to the reference wire affixed to the connector pin arrays of the headstage, and copper mesh was wrapped around the entire headstage to protect the wires from potential damage. The mouse was then aroused with an injection of 2.5 mg/kg Antisedan and returned to the customized recording home environments. The mouse was allowed to recover for at least 2-3 d before advancing the electrodes. The electrode bundles were advanced slowly toward the hippocampal CA1 region, in daily increments of ≈0.07 mm until the tips of the electrodes had reached the CA1, as deduced from an assessment of high-frequency ripples, field potential, and neuronal activity patterns (1, 2).

On-Line Search for Nest-Responsive Cells and Off-Line Spike Sorting.

Overall off-line spike sorting was carried out by using the MClust 3.0 and KlustaKwik 1.5 programs (3), and the details are the same as described in our recent publications (1, 2). To search for nest-responsive cells during the mouse's encounters with the nest, we first used Sort Client of the Plexon MAP system to obtain well isolated units by adjusting the MAP operating parameters and setting the specific sorting parameters for each channel. We used the Boxes sorting method of Sort Client to achieve initial on-line sorting and classification. The spike activities during various nest encounters were recorded and then off-line sorted again by using the MClust 3.0 and KlustaKwik 1.5 programs as described (1-4). Only stable units (for at least 6 h) with clear boundaries and <0.5% of spike intervals within a 1-ms refractory period are included in the analysis.

If a well-isolated unit from on-line sorting was visually identified to exhibit nest encounter-related firing changes, we subjected the mouse to the first set of experiments in which the nest was relocated to different positions in the home environment and to new environments. The mouse was allowed to freely explore the nest and environments. If the nest-responsive cell was still stable after those experiments, we then conducted the second set of experiments in which new nests with different shapes, colors, styles, or construction materials were introduced to the animals. The number of different types of nests examined often depended on how actively and how often the mouse was willing to explore them during the stable recordings. In several cases (e.g., Cell #1 or Cell #3), we were able to record over a course of 7 or 4 days, respectively. Experiments were continued until we lost the cell from the on-line sort. The mouse was then anesthetized and a small amount of current was applied to four channels in the headstage to mark the positioning of the electrode bundle. Histological staining, with (1% cresyl echt violet) was used to confirm the electrode positions. A total of 255 single units from seven mice were recorded and carefully examined for nest responsiveness. Of them, eight single units were identified to respond significantly to nests. The examples of waveforms and basic firing characteristics of the units are shown in SI Figs. 6 and 7). For plotting the perievent histogram and perievent rasters of short one-type cells, we marked the time point at which the mouse's nose tip was 1 cm away from the edge of the nest before crossing as time "zero".

Construction of Nests.

A variety of nests in different geometric shapes and sizes were made from cardboard coffee cups or other cardboard materials. Plastic caps from Maxwell house instant coffee bottles or sucrose bottles (Sigma) and metal caps from phenol containers (Sigma) or Altoids candy tins were also used as nests. The shapes, diameters, and heights of each nest are described in either the text or figure legends. The examples of various nests used in the study are shown in SI Fig. 10.

All animals received normal light-dark cycle and were not subjected to food and water deprivation or restriction. Behavioral exploration and interactions with nests were completely voluntary and natural. All nests, other than the home nests as shown in the text (the top two left panels of SI Fig. 10), were new to mice and had not been presented to mice until the recording experiments. The functional categorization of these new objects as a nest was also demonstrated by confirming that mice indeed slept inside of these new objects once introduced to the mouse cages (data not shown).

1. Lin L, Osan R, Shoham S, Jin W, Zuo W, Tsien JZ (2005) Proc Natl Acad Sci USA 102:6125-6130.

2. Lin L, Chen G, Xie K, Zaia K, Zhang S, Tsien JZ (2006) J Neurosci Methods 155:28-38.

3. Harris KD, Henze DA, Csicsvari J, Hirase H, Buzsaki G (2000) J Neurophysiol 84:401-414.

4. Buzsaki G (2004) Nat Neurosci 7:446-451.