| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Psychology, Bowling Green State University, Bowling Green, Ohio
Submitted 23 February 2005; accepted in final form 7 April 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
"One of the few things that is agreed about the mammillary bodies is that they might or might not be important for memory."—Vann and Aggleton 2004
As this quote illustrates, the functional role of the mammillary bodies has remained elusive despite more than a century of effort. Lesion work suggests the mammillary bodies are critical for memory under certain conditions, perhaps especially conditions requiring spatial memory (Sziklas and Petrides 1998
; Vann and Aggleton 2004). However, it remains unclear how to precisely characterize these conditions.
As reviewed elsewhere (Allen and Hopkins 1989; Vann and Aggleton 2004), the mammillary bodies and their associated structures can be divided into at least two separate systems (Fig. 1A). Both of these are components of a larger loop that travels from the mammillary bodies, to the anterior thalamus, to limbic cortex, and then back.
|
). Each of these areas contains head direction cells like those first documented by Taube et al. (1990), as illustrated in Fig. 1, C and E. Although these direction-signaling cells are found in each of the structures throughout this subloop, and in several additional brain regions as well (e.g., Mizumori and Williams 1993; Wiener 1993), it has been suggested that the signal may originate in a circuit consisting of the reciprocally connected lateral mammillary and dorsal tegmental nuclei (e.g., Sharp et al. 2001). According to this idea, all other brain regions that contain head direction cells receive this directional signal through synaptic relays from the lateral mammillary nucleus.
Thus it could be that the functional contribution of the lateral mammillary nucleus is to provide this directional signal, which may be critical for certain kinds of spatial tasks, and which could possibly also affect memory processing even on tasks that are not obviously spatial in nature (Vann and Aggleton 2004).
The second component of the loop consists of the medial mammillary nucleus, the anteroventral and anteromedial subnuclei of the anterior thalamic nucleus, and the subicular and entorhinal cortices. At the level of the mammillary bodies, this component of the loop has received less study. In fact, to our knowledge, the data reported here constitute the first recordings from medial mammillary body cells in unanesthetized animals. Note, however, that the cortical components of this loop (subiculum and entorhinal cortex) are regions that contain locational signals (e.g., Frank et al. 2000; Fyhn et al. 2004; Quirk et al. 1992; Sharp and Green 1994), similar to the place cells originally discovered by O'Keefe and Dostrovsky (1971). Thus, it might be expected that the medial mammillary nucleus plays a role in coding for spatial location.
To examine this possibility, we here report firing correlates for medial mammillary body cells recorded while rats performed a food-foraging task identical to that used for much of the work on head direction and place cells (Muller et al. 1987).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The subjects were five male, Long–Evans rats, weighing between 250 and 400 g at shipping. The rats were housed individually upon arrival and had a 12-h light–dark schedule. All procedures related to the rats were conducted in accordance with NIH guidelines for animal care and use, and with the approval of the Animal Care and Use Committee at Bowling Green State University.
Recording chamber and behavioral task
All training and recording sessions were conducted in a 76.0-cm-diameter cylindrical recording chamber with 51-cm-high walls. The rats were food deprived to 80% of their ad lib weight and trained to search for 20-mg food pellets (P.J. Noyes, New Brunswick, NJ) that were dispensed to random locations within the chamber at approximately 15-s intervals throughout each session. As a result of the three, daily, 15-min training sessions, the rats developed a pattern of constant locomotion through the cylinder. Thus during subsequent recording sessions, the rats repeatedly traversed each region of the cylinder, using a pattern of diverse, seemingly random trajectories.
Electrode implantation
After training, two drivable microelectrode bundles (one per hemisphere), consisting of six wires each, were chronically implanted just above the medial mammillary nucleus (4.2 mm posterior, 1.72 mm lateral, and 8.2 mm ventral to bregma, at a mediolateral angle of 10 degrees). Each wire was made of insulated stainless steel (FHC, Brunswick, ME) with a 125-µm shank that tapered down to a 1-µm exposed recording tip.
Recording sessions
After recovery from surgery, the rats were given screening-recording sessions during which the activity on the recording electrodes was sampled while the rat performed the pellet-chasing task. Upon isolation of unit activity, recording was begun while the rat continued to chase pellets. Recording sessions were between 30 and 40 min in duration.
Data acquisition
The signals from each electrode wire were amplified at a gain of 20,000 to 50,000 and filtered at 600 Hz high pass and 6,000 Hz low pass and then sent to a computer for automatic data collection (Neuralynx, Tucson, AZ). Superthreshold events were analyzed, using waveform-specific characteristics, to determine the time of occurrence for the spikes from each individual cell.
The rat's moment-to-moment position within the cylinder was also sampled continuously, at 30 Hz, throughout the sessions, using a video tracker system (Neuralynx), which detected the position of a red and a green diode mounted on either side of the rat's head.
Data analysis
SPATIAL AND MOVEMENT CORRELATES. The positions of the red and green headlights for each sample throughout the session were used to calculate, for each intersample interval, the rat's momentary head direction, location, running speed, and angular velocity.
These values were then used to calculate each cell's average firing rate for each value of head direction (divided into 60 bins, from 0 to 360 degrees), location within the cylinder (divided into 3.16 x 3.16-cm pixels), angular head velocity (divided into 20 bins over a range of –200 to +200 deg/s), and running speed (divided into 40 bins, each covering 1 cm/s). For angular velocity, negative values indicated counterclockwise turns (with larger negative values corresponding to faster counterclockwise turning speeds), whereas positive values indicated clockwise turning speed.
The plots of average firing rate as a function of running speed, as well as angular head velocity, were often linear (see Fig. 3 and 4). Because of this, the strength and sign of these correlations could be assessed using Pearson's correlation coefficient (referred to below as the R value).
|
|
EXAMINATION OF SPATIAL (LOCATION-RELATED) CORRELATES.
Because the mammillary bodies are strongly interconnected with hippocampal regions known to contain place cells (see Fig. 1), it was of interest to examine the present data set for any possible place cell-like properties. For this, spatial firing rate maps (as in Fig. 1D) of the cylinder floor were constructed using methods identical to those used for cells in the hippocampal formation (e.g., Sharp and Green 1994). First, the cylinder floor was divided into a set of 2.9-cm2 pixels. Then, the average firing rate for each pixel throughout the session was calculated. Next, the mean and SD of these pixel averages were calculated. Then the relative rates for each pixel were displayed using a gray scale in which pixels that were 2SDs above the mean were represented using the darkest shade, pixels 1SD above the mean were shown using the second darkest shade, pixels within 1SD of the mean were shown using the next darkest shade, and so forth.
TEMPORAL ANALYSIS OF SPIKE TRAINS. Autocorrelation histograms were constructed by summing, across all spikes, the number of spikes in each of 300 subsequent 1-ms bins after the occurrence of that spike.
Reconstruction of electrode location
After recording, rats were perfused transcardially under deep anesthesia with a formyl saline solution. Beforehand, a small marking current (30 µA x 5 s) was passed through at least one wire of each electrode bundle. The brains were removed and then sectioned at 40-µm intervals, so that the location of each recorded cell could be determined.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Data from the medial mammillary nucleus were collected from a total of 51 cells in five rats. The average firing rate for this sample was 4.8 Hz.
Figure 2 shows the right and left hemisphere electrode tracks from a representative brain as they traverse through the level of the medial mammillary nucleus. Note that the coronal plane of section for this brain was slightly off from the plane at which the electrodes descended, and this is why the tracks can be seen to descend from a level just above and at the top of the nucleus in Fig. 2B, on to the bottom of the nucleus in the more posterior section shown in Fig. 2D.
|
The electrode track on the right, in contrast, only grazed the right-most edge of the medial mammillary nucleus. This suggests that, although some of the wires in this electrode bundle appear to have passed through the nucleus, it is likely that some did not. For this reason, none of the cells recorded from this electrode was included here.
The medial mammillary nucleus has been divided into five subnuclei (Allen and Hopkins 1988): pars lateralis, pars basalis, pars medialis, pars medianus, and pars posterior. Of these, all but pars medianus are strongly interconnected with the hippocampal formation and also with the VTN, as illustrated in Fig. 1A (Allen and Hopkins 1989; Shibata 1989). Of the 51 cells recorded here, 41 were in pars medialis, eight were in pars posterior, and two were in pars lateralis.
The firing rate for many cells showed a significant relationship to angular head velocity
The most salient behavioral correlate observed for the medial mammillary cells was angular head velocity. Seventeen of the 51 recorded cells (33.33%) showed a statistically significant (P < 0.01, two-tailed) correlation (R) with angular motion of the head. For eight of these 17 cells, the correlation was positive, suggesting that the cell increased its rate during clockwise turns, while decreasing for counterclockwise turns. For these eight cells, the average R value was +0.70. For the remaining nine of 17 cells, the correlation with angular velocity was negative, with an average R value of –0.64.
Figure 3 shows activity from four representative cells, each from a different rat. Each of these showed a clear, dramatic change in firing rate that was coincident with the head-turning motion. For example, Cell A showed a decrease in rate (middle trace) during clockwise turns, with a time course that appeared to mirror the angular head velocity (top trace) itself. In contrast, this cell showed no detectable change in rate during counterclockwise turns. Note that the plot of average rate in relation to momentary angular velocity (bottom trace) also reflects this cell's relative preference for counterclockwise turns.
The pattern shown by Cell C consists of an increase during clockwise turns and a decrease during counterclockwise turns. This results in the strong positive correlation between momentary firing rate and angular velocity, as shown in the bottom trace for Cell C.
Cells B and D each showed somewhat more complex temporal patterns. Thus, Cell B showed a decrease in rate that was coincident with counterclockwise turns, but showed an increase that was slightly delayed in relation to clockwise turns. Cell D showed a decrease, then increase during clockwise turns, while showing no obvious relationship to counterclockwise turns. Note, the bottom trace for Cell D indicates that the overall correlation between rate and momentary angular velocity was not significant. (Thus, this cell was not counted, above, among the 17 cells that showed a significant R value.) This probably results from the fact that the timing of the changes during clockwise turns was such that the cell was in transition from low to high rates at the peak of any fast clockwise turning motion. Also, as just mentioned, this cell seemed to show no relation to counterclockwise turns.
Many cells showed a linear relationship to running speed
Figure 4 (left) shows average firing rate as a function of running speed (translational motion) for a set of four representative cells, each recorded from a different rat. Overall, 29/51 (57.9%) of the cells showed a significant correlation between firing rate and running speed. For 26 of these cells the correlation was positive, indicating that the cell fired at higher rates during faster running. The average R value of these positive correlations was +0.74. For the remaining three cells the correlation was negative, with an average R of –0.82. Note, the top cell in Fig. 4 provides an example of a cell with no detectable relationship to running speed. In contrast, the bottom three cells all show a significant, positive correlation with this variable.
In contrast to the angular velocity correlates described above, the running speed correlates were not typically tightly tied temporally to instances of fast running. Rather, firing rate for most cells was coarsely correlated with running speed, so that there were bouts of relatively high frequency firing that were accompanied by bouts of relatively fast running. These alternated with bouts of relatively slow running and lower firing rates (data not shown). For most cells, momentary bursts of particularly fast or slow speed were not accompanied by noticeable changes in rate.
No medial mammillary cells showed either head direction or place correlates
Figure 4 (middle) shows average firing rate as a function of directional heading for each of the representative cells. For each of these, there was no indication of a single preferred direction like that observed for head direction cells (see Fig. 1E). Indeed, no cell showed evidence of any form of consistent relationship to directional heading.
Figure 4 (right) shows the locational firing rate map for each cell. None of these cells showed evidence of preferred spatial locations like that typically observed for cells in the hippocampal, subicular, and entorhinal cortices (see Fig. 1D). Indeed, none of the cells recorded here showed any sign of consistent location-related signaling.
Almost all medial mammillary cells showed a strong modulation of firing rate at theta frequency
Figure 5 shows autocorrelation histograms for four representative medial mammillary cells. Each of these showed a strong temporal modulation of rate, so that there were peaks in firing probability at approximately 140-ms intervals. This corresponds to a frequency of about 7 Hz, which is within the range of the theta EEG rhythm exhibited in the hippocampus. Only four cells (from three rats) failed to show this temporal pattern.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
One third of the cells recorded here showed a significant, closely timed relationship to angular head motion. Each cell had its own idiosyncratic pattern of rate changes over the course of rapid clockwise and/or counterclockwise head turns, as illustrated in Fig. 3. Importantly, each of these cells showed a differential response for clockwise versus counterclockwise turns. This means that the cell population rate vector provides an unambiguous indication of whether the movement trajectory is to the left or to the right.
These findings can be contrasted to those from the lateral mammillary nucleus, where only 3/41 cells showed this type of angular velocity activity (Blair et al. 1998). [Note, Stackman and Taube (1998) reported angular velocity correlates for 44% of their lateral mammillary body cells, but for these cells the response (either an increase or a decrease) was identical for clockwise and counterclockwise turns.]
There are at least two possible sources for the medial mammillary angular head turning information. One is from the vestibular and prepositus hypoglossis nuclei, which contain angular velocity signals from the vestibular system and project, by the ventral tegmental nucleus (see Fig. 1A), to the medial mammillary body (Irle et al. 1984). The other possible source is from the habenula, which also projects to the ventral tegmental nucleus (Herkenham and Nauta 1979; Irle et al. 1984). The habenula, in turn, receives its major input from the internal segment of the globus pallidus (Herkenham and Nauta 1977), which, presumably, provides motor-related information. A recent study (PE Sharp, S Turner-Williams, and S. Tuttle, unpublished observations) has shown that the habenula does, in fact, contain angular velocity cells similar to those observed here.
In addition to these angular head velocity correlates, many (nearly 60%) of the cells were correlated with running speed. The majority of these showed a positive correlation, so that faster running was accompanied by higher firing rates. Observation of these firing patterns during episodes of fast running (data not shown) indicated that the timing for the fast firing was not tightly correlated with brief bursts of high speed. Rather, these cells showed generally higher rates during periods of relatively high locomotor activity, and vice versa.
These running speed correlates could also, possibly, be projected from the habenula because that region also contains cells that are coarsely correlated with running speed (Sharp et al., unpublished observations).
Indeed, one possible explanation for both the angular velocity and running speed correlates reported here could be that they are both driven by motor command–related activity from the entopeduncular nucleus (by the habenula–ventral tegmental pathway). Thus the angular velocity signal could be driven by motor commands for neck muscle contraction. Similarly, the running speed correlates could possibly involve signals to control other individual muscle groups, such as forearm or back muscles. If the cells are related to this type of specific movement, this might explain why the running speed correlations are not closely temporally tied to episodes of fast running. For example, any cell that is tied to movement of the right forelimb would be expected to be generally more active during fast rather than slow running, but would probably not coincide exactly with brief bursts of high speed.
Yet another possible source of running-related input is the superior central nucleus, which also provides a major projection to the medial mammillary nucleus (Hayakawa and Zyo 1991). Cells in this nucleus, as studied in freely moving cats, fire in relation to behavioral state, so that they are optimally active during active wakefulness, and then systematically decrease their rate as the animal descends into quiet wakefulness, slow-wave sleep, and, finally, rapid eye movement sleep (Rasmussen et al. 1984). Interestingly, Rasmussen et al. (1984) reported that the cells were not tightly tied to phasic increases in EMG activity. Rather, they were tonically active at high rates during periods of high motor activity. Thus, it is possible that if these centralis superior cells were recorded during the pellet-chasing task used here, periods of relatively fast running would be coarsely correlated with relatively higher tonic firing rates, just as were the medial mammillary body cells recorded here.
Medial mammillary cells do not show location or head direction correlates
There was no indication that any medial mammillary cell activity was correlated with spatial location, as is the case for many cells in the hippocampal formation (see Fig. 1D). This is surprising given that the medial mammillary nuclei receive a strong input from both the dorsal and ventral subiculum (Allen and Hopkins 1989; Shibata 1989), and the majority of dorsal subicular cells show strong locational correlates when recorded during this task (Sharp and Green 1994). Thus, either the subicular projection to the mammillary bodies arises from a select population of nonspatial cells, or the locational signal from the subiculum is somehow averaged out, or otherwise cancelled, upon arrival in the medial mammillary nucleus.
One possibility is that there are place signals in portions of the medial mammillary bodies that were not sampled here. As described in the RESULTS section, cells in this study were recorded from only three of the five subnuclei of which the medial mammillary nucleus is composed. Thus, it is possible that place cells would have been discovered if each subnucleus had been thoroughly sampled. However, anatomical work has demonstrated that, with the expection of pars medianus of the medial mammillary nucleus (from which no cells were recorded here), all the subnuclei share very similar connectivity (Allen and Hopkins 1989; Shibata 1989). Specifically, all are strongly connected with the subicular cortex and the ventral tegmental nucleus, as diagrammed in Fig. 1A. This makes it seem unlikely that these four subnuclei would differ strongly in their behavioral correlates.
Of course, given the limited sample (51 cells) recorded here, it cannot be entirely ruled out that there could be a small population of place cells within the medial mammillary nuclei. However, if such cells do exist, they would constitute a much smaller percentage of the overall population than is the case for other areas from which place cells have been recorded.
It is worth noting that this absence of locational signals in the medial mammillary nucleus is in sharp contrast to parallel findings in the lateral mammillary nucleus. As illustrated in Fig. 1A, the lateral mammillary nucleus is connected to other limbic regions that contain head direction cells, and the lateral mammillary nucleus itself also contains these directional cells (Blair et al. 1998; Stackman and Taube 1998). In contrast, although the medial mammillary nucleus is connected with limbic areas that signal spatial location, this nucleus does not appear to show any location-related (place cell) activity itself.
These medial mammillary nucleus cells also showed no evidence of head direction–like properties. This is not surprising, given that this nucleus is not connected to regions that contain head direction cells.
Most medial mammillary cells showed a strong modulation of rate at theta frequency
Almost all of the cells recorded here showed rhythmic modulation of their firing rate at a frequency of about 7 Hz. This frequency is within the range of the theta EEG rhythm as recorded in the hippocampal formation during locomotion. This suggests that these medial mammillary cells may fire in a fixed relation to the hippocampal theta rhythm, when present, in the awake animal, as is the case for mammillary body cells in anesthetized animals (Kocsis and Vertes 1994). Indeed, as mentioned above, the subicular region of the hippocampal formation projects strongly to all portions of the medial mammillary nucleus (Allen and Hopkins 1989; Shibata 1989), and spatial cells that are strongly modulated at theta frequency have been recorded in this area in freely moving rats as they performed the pellet-chasing task used here (Sharp and Green 1994).
The large proportion of theta-modulated cells found here can be contrasted with recordings from the lateral mammillary nucleus, where much smaller percentages have been reported (Blair et al. 1998; Stackman and Taube 1998).
The medial and lateral mammillary bodies together may provide necessary inputs for the calculation of the place cell signals in the hippocampal formation
Empirical work on place cells in the hippocampal formation has demonstrated that this location-specific firing is controlled, in part, by a path-integration process. In this process, the animal's own movement through space is used to constantly update the hippocampal place cell activity vector. Specifically, theoretical models of the place cell system (e.g., McNaughton et al. 1996) require three types of information for this path-integration process: 1) current position (spatial location), 2) current directional heading, and 3) current movement state. Logically, it can be seen that these are the necessary and sufficient inputs for translational path integration because, for any given start position and directional heading, a particular movement will always result in a predictable new locational position.
The necessary information for current position is assumed to arise from the place cells themselves, which are assumed to provide feedback projections for the path-integration process. The directional information is assumed to come from the limbic system head direction signal, which, as mentioned above, is postulated to arise from the lateral mammillary nucleus. The data presented here suggest that the necessary movement trajectory information may be provided, at least in part, by the medial mammillary nucleus.
Thus it could be that the lateral and medial mammillary nuclei together provide necessary building blocks for construction of the limbic system place cell activity. This could, in turn, explain why mammillary body damage seems to preferentially affect memory on spatial tasks.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: P. E. Sharp, Department of Psychology, Bowling Green State University, Bowling Green, OH 43403 (E-mail: psharp{at}bgnet.bgsu.edu).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Allen GV and Hopkins DA. Mammillary body in the rat: topography and synaptology of projections from the subicular complex, prefrontal cortex and midbrain tegmentum. J Comp Neurol 286: 311–336, 1989.[CrossRef][ISI][Medline]
Blair HT, Cho J, and Sharp PE. Role of the lateral mammillary nucleus in the rat head direction circuit: a combined single unit recording and lesion study. Neuron 21: 1387–1397, 1998.[CrossRef][ISI][Medline]
Frank LM, Brown EN, and Wilson M. Trajectory encoding in the hippocampus and entorhinal cortex. Neuron 27: 169–178, 2000.[CrossRef][ISI][Medline]
Fyhn M, Molden S, Witter MP, Moser EI, and Moser M-B. Spatial representation in the entorhinal cortex. Science 305: 1258–1264, 2004.
Hayakawa T and Zyo K. Quantitative and ultrastructural study of ascending projections to the medial mammillary nucleus in the rat. Anat Embryol 184: 611–622, 1991.[CrossRef][Medline]
Herkenham M and Nauta WJH. Afferent connections of the habenular nuclei in the rat. A horseradish peroxidase study, with a note on the fiber-of-passage problem. J Comp Neurol 173: 123–146, 1977.[CrossRef][ISI][Medline]
Herkenham M and Nauta WJH. Efferent connections of the habenular nuclei in the rat. J Comp Neurol 187: 19–48, 1979.[CrossRef][ISI][Medline]
Irle E, Sarter M, Guldin WO, and Markowitsch HJ. Afferents to the ventral tegmental nucleus of Gudden in the mouse, rat and cat. J Comp Neurol 228: 509–541, 1984.[CrossRef][ISI][Medline]
Kocsis B and Vertes RP. Characterization of neurons of the supramammillary nucleus and mammillary body that discharge rhythmically with the hippocampal theta rhythm in the rat. J Neurosci 14: 7040–7052, 1994.[Abstract]
McNaughton BL, Barnes CA, Gerrard J, Gothard K, Jung MW, Knierim JJ, Kudrimoti H, Qin Y, Skaggs WE, Suster M, and Weaver KL. Deciphering the hippocampal polyglot: the hippocampus as a path integration system. J Exp Biol 199: 173–185, 1996.[Abstract]
Mizumori SJY and Williams JD. Directionally selective mnemonic properties of neurons in the lateral dorsal nucleus of the thalamus of rats. J Neurosci 13: 4015–4028, 1993.[Abstract]
Muller RU, Kubie JL and Ranck JB Jr. Spatial firing patterns of hippocampal complex spike cells in a fixed environment. J Neurosci 7: 1935–1950, 1987.[Abstract]
O'Keefe J and Dostrovsky J. The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat. Brain Res 34: 171–175, 1971.[CrossRef][ISI][Medline]
Paxinos G and Watson C. The Rat Brain in Stereotaxic Coordinates (compact 3rd ed.). San Diego, CA: Academic Press, 1997.
Quirk GJ, Muller RU, Kubie JL, and Ranck JB Jr. The positional firing properties of medial entorhinal neurons: description and comparison with hippocampal place cells. J Neurosci 12: 1945–1963, 1992.[Abstract]
Rasmussen K, Heym J, and Jacobs BL. Activity of serotonin-containing neurons in nucleus centralis superior of freely moving cats. Exp Neurol 83: 302–317, 1984.[ISI][Medline]
Sharp PE, Blair HT, and Cho. The anatomical and computational basis of the rat head-direction cell signal. Trends Neurosci 24: 289–294, 2001.[CrossRef][ISI][Medline]
Sharp PE and Green C. Spatial correlates of firing patterns of single cells in the subiculum of the freely-moving rat. J Neurosci 14: 2339–2356, 1994.[Abstract]
Shibata H. Descending projections to the mammillary nuclei in the rat, as studied by retrograde and anterograde transport of wheat germ agglutinin-horseradish peroxidase. J Comp Neurol 285: 436–452, 1989.[CrossRef][ISI][Medline]
Stackman RW and Taube JS. Firing properties of rat lateral mammillary single units: head direction, head pitch, and angular head velocity. J Neurosci 18: 9020–9037, 1998.
Sziklas V and Petrides M. Memory and the region of the mammillary bodies. Prog Neurobiol 54: 55–70, 1998.[CrossRef][ISI][Medline]
Taube JS, Muller RU, and Ranck JB Jr. Head-direction cells recorded from the postsubiculum in freely moving rats. I. Description and quantitative analysis. J Neurosci 10: 420–435, 1990.[Abstract]
Vann SD and Aggleton JP. The mammillary bodies: two memory systems in one. Nat Rev Neurosci 5: 35–44, 2004.[CrossRef][ISI][Medline]
Wiener SI. Spatial and behavioral correlates of striatal neurons in rats performing a self-initiated navigation task. J Neurosci 13: 3802–3817, 1993.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  M. E. Hasselmo Arc length coding by interference of theta frequency oscillations may underlie context-dependent hippocampal unit data and episodic memory function Learn. Mem., November 15, 2007; 14(11): 782 - 794. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. Bassett, M. L. Tullman, and J. S. Taube Lesions of the Tegmentomammillary Circuit in the Head Direction System Disrupt the Head Direction Signal in the Anterior Thalamus J. Neurosci., July 11, 2007; 27(28): 7564 - 7577. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kepecs, N. Uchida, and Z. F. Mainen The Sniff as a Unit of Olfactory Processing Chem Senses, February 1, 2006; 31(2): 167 - 179. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
