- Email: [email protected]
Intracranial self-stimulation in the thalamus of the rat
Brain Research Bulktin,
Vol. 8, pp. 353-358, 1982. Printed in the U.S.A.
Intracranial Self-Stimulation in the Thalamus of the Rat RONALD M. CLAVIER’ Depurr~en~ of Biopsych~f~gy, The Clarke Znst~tule of Psy~h~a~~ 250 College St., Toronto, Ontario MST IR8, Canada AND CHARLES
R. GERFEN
Weingart L~~orato~ for Deve~op~eniffl ~euru&iufog~~, The Salk Znstitute P.O. Box 85800, San Diego, CA 92138 Received 21 January
1982
CLAVIER, R. M. AND C. R. GERFEN. Zntrucrunialsetf-st~muiation in the thdamus of the rut. BRAIN RES. BULL. g(4) 353-358, 1%2.-Rats were tested for intracranial set-stim~ation (ICSS) via bipolar ekctrodes situated copout the thalamus. Of 112 animals in the study, 55 met the criteria for ICSS, with scores ranging from 55 to 921 bar-presses in a 15 minute session. A map of both positive and neutral placements is presented. Positive sites for ICSS were found in all aspects of the mediodorsal nucleus, except for the central segment. The ventromediai nuclear compkx was also a positive area of ICSS, with the exception of the submedial nucleus (nucleus gelatinosus). Each of the intmkmimu nuclei (central medial, parafascicular, paracentral, and central lateral) supported ICSS, as did each of the midline nuclei (rhomboid, paratenial and paraventrkukr). No placements were found in the nucleus reuniens, Both “major” relay nuclei, the ventrobasaf and ventrolaterai, supported ICSS; but neither the taterodorsal nor the lateral posterior nuclei had positive sites. Positive sites were found in the anterior and posterior nuclear groups, as we11as in the paraventricular gray area of the caudai thabunus. As a general rule, ICSS scores appeared to be higher as the electrode placements approached the midline. Sites in which no positive placements were seen included the reticular nucleus, as well as the stria medukris, the mammillothakmic tract, and the fasciculus retroflexus. Thakmus
Self-stimulation
BECAUSE of its potential relation to brain reward mechanisms, intracranial self-stimulation (ICSS) has been investigated in great detail. A major focus of this work concerns efforts to identify the anatomical substrates of ICSS. Perhaps the most clear-cut findings of this focus are: (1) that ICSS is obtained from a wide variety of brain regions; and (2) that efforts to identify a common underlying substrate for ICSS are not successful (see Redgrave and Dean [29] for a recent review). Despite this, however, certain consistencies have been found. One of these is the topic of the present study. As a result of studies aimed at describing the ~ato~c~ connections of brain regions that support ICSS, our attention has been drawn repeatedly to the thalamus. For example, Keen and Casey [221 were able to establish a functional relation between stimulation of the medial forebrain bundle via ICSS electrodes and the firing pattern of single units in the ventromedial thalamic region. The presumed anatomical basis for this effect may reside in the GoIgi data of M~ouse [26], which established MFB connections with the midline
thalamic nuclei. Also, Clavier and Routtenberg [7], in describing axonal degeneration after lesions made via brainstem ICSS electrodes, identified 2 ascending pathways: one that ran adjacent to the parafascicular nucleus, and terminated in the ventromedial and dorsomedial thalamic nuclei; and another that terminated in the region of the medial forebrain bundle. Electrolytic lesions in either of these projection regions were shown to attenuate brainstem ICSS [61, indicating a possible functional relation between the medial forebrain bundle and the medial thalamic regions. Two other ICSS areas are related to the thalamus: the sulcal prefrontal cortex; and the substantia nigra. The sulcal cortex is the area on the dorsal bank of the rhinal sulcus in the prefrontal cortex of the rat. This region is the site of ICSS placements [32], and represents one of the projection fields of the “AlO” dopaminergic system [ 12,251, which has been implicated in brain reward mechanisms 110,331. The sulcal cortex also receives input from the thalamus. Specifically, the mediodorsal, ventromedial, and intralaminar (~clu~ng the parafascicular) nuclei all project to this region 113, 23,
‘To whom requests for reprints should be sent.
~opy~g~t
@ 1982 ANKHO
Intemation~
Inc .-0361-9230/82/040353-06$03.00/O
354
241.Moreover, a descending projection exists from the sulcal area to several thalamic nuclei, including the mediodorsal, ventrome~~, and midline groups 1141. Because of these connections, each of these thalamic nuclei received further attention as possible contributors to the observed behavior. Substantia nigra ICSS has been hypothesized to depend on the stimulation of the “A9” dopaminergic cell group 19,101. Yet, ICSS in this area continues after cases of total depletion of dopamine from the ni~ost~at~ system [5]. Thus, we have considered other systems in the vicinity of the electrodes that might support ICSS. Part of this effort deals with the relation of the substantia nigra to the thalamus. As the recipient of nigral inputs, the parafascicular and ventromedial nuclei, and the paralamellar aspect of the mediodorsal nucleus [2, 4, 151, were all considered to be related, potenti~y, to ICSS. One method for assessing the relation of these thalamic nuclei to ICSS is to see whether electrodes in any or all of them are positive sites of ICSS. The results of that approach are the subject of the present report. Prior to presenting our findings, however, a brief summary of previous work on thalamic ICSS is warranted. The most comprehensive ex~nation of thalamic ICSS to date is that of Olds and Olds [ZS]. The authors described 32 electrode sites in 15 nuclear regions. Each site was tested for avoidance, as measured in a bar-pressing paradigm for escape, as well as for approach, as measured by ICSS. This provided the scope to ascribe qualities of pure approach, pure avoidance, or mixed approach-avoid~ce to each site. They used an approach criterion of 200 responses in 8 minutes, which they admitted might have deleted several cases of fairly consistent bar-pressing from the approach category. Had they used less stringent criteria-for example those of Routtenberg and Malsbury [311_and extrapolated their scores to 1.5 minute session equivalents, they might have identified 8 cases of moderate ICSS (2~5~), 14 cases of low ICSS @O-199), and 9 cases of neutrality. In the case of their highest rate of ICSS (860 responses in 8 minutes), the electrode tip was situated in the subthalamic nucleus, and not in the thalamic reticular nucleus as the authors indicated. This discrepancy aside, each of the thalamic nuclear areas examined in the Olds and Olds study is represented by at least one case of ICSS. Yet, according to their own criteria, only 2 of their placements showed ICSS properties, and both of these showed avoiance qualities as well. This led Olds and Glds to conclude that “. . . there is . . . almost no approach in the thalamus,” and that “One is led to wonder about the evolutionary and functional significance of an arrangement which appears to put negative mechanisms only in thalamic systems . . . . ” Olds [27] later revised this view, accepting the thalamus as a valid ICSS site. However, he noted that the presence of clear-cut aversive areas, and their possible overlap with ICSS regions in the thalamus could result in the operational existence of “marginal” areas; these could be moved into or out of brain reward systems based on methodological parameters such as probe size, stimulating parameters, or behaviors paradigm. The aversive potential of thalamic stimulation was ak.0 noted by Gillett and Webster [161, who showed that me&odorsal thalamic stimulation attenuated bar-pressing on a baseline schedule of water reinforcement. The role of the thalamus in ICSS remains unclear today. Rolls [30] has written that “Stimulation at the majority of other brain sites, for example in much of neocortex, thalamus, and cerebellum, is neutral , . . .” Also, Stein 1331
CLAVIER
AND GERFEN
noted that “Most regions of the thalamus do not support self-st~ulation . . . .” Nevertheless, Stein did report finding 5 probes in the region of the paratenial thalamic nucleus that supported “respectable” rates of ICSS. Thalamic sites of ICSS were also reported by Cooper and Taylor [8]. They showed positive sites in the midline and intralaminar nuclear groups, as well as mixed ICSS-pain placements in the periventricular and central gray areas of the thalamus. In the case of human brain stimulation, Heath [ 171cited an instance of centromedian ICSS, which carried the subjective sensation of frustration and anger related to partial loss of memory. Thus, while the thalamus appears to be related to ICSS. it is often unrelated to pure reward properties, and the anatomical details of its involvement are not clear. METHOD One hundred twenty-eight adult male albino rats (Holtzman) weighing 275-350 grams prior to surgery were used in this study. Each rat had a single bipolar nichrome wire electrode (Plastic Products Co., MS 303.081 -0.3120.005 in.) implanted while under pentobarbital anesthesia. Electrodes were implanted ~~endic~~ to the skull surface using stereotaxic coordinates that ranged from 5.0 mm to 6.5 mm rostra& 0.1 mm to 3.5 mm lateral, and 2.2 mm to 4.2 mm dorsal to stereotaxic zero, and with the incisor bar set at 4.2 mm below horizontal. At least 10 days were allowed for recovery prior to the start of behavioral testing. Testing for ICSS was conducted in operant conditioning chambers, in which each depression of a lever resulted in a 0.5 set pulse of 60 Hz sine wave current. Each rat underwent a minimum of 21 days of pre-test trials, each lasting at least 2 hours, unless ICSS criteria were met prior to this. During the pre-test trials, the rats were given Experimenter-delivered stimulations for 1-2 minutes at 30 minute intervals. The current level for these stimulations was in the order of 55 PA @MS). Current was raised gradually in the case of neutral reactions on the part of the rats, or else lowered if an aversive or motor response was elicited. Between bouts of Experimenter-delivered stimulations, the animals were left in the chambers and the number of bar-presses was recorded. If the number of daily bar-presses by a rat did not exceed 100 on any of the 21 days of such trials, the animal was killed by pentob~bit~ overdose followed by intracardiac perfusion; its electrode site was recorded and considered neutral for ICSS. If a rat passed the initial criterion for ICSS, it was run daily in 15 minute sessions with the current set at the level at which it responded with a steady rate. We have, moreover, set a criterion for stabilized ICSS. This was said to have occurred when there were any 5 consecutive daily test sessions in which the score in 15 minutes did not fall by more than 1 standard deviation from the mean of those 5 days. If the score of a rat fell below this criterion, the current was raised until the response rate was in the range of the criterion scores. In the event that an ICSS score fell in our assessment due to motor or aversive qualities of the current, the current was lowered until such qualities were no longer apparently interfering with the ICSS. In addition to the criterion for stabilized ICSS, we have used an animal’s highest daily 15 minute score to identify a given placement. This is in conformity with the procedure of Olds and Olds [28]. At the end of testing, all animals were killed by pentobarbital overdose and intracardiac perfusion with 0.9%
THALAMIC
SELF-STIMULATION
FIG. 1. Schematic representation of positive (numbers) and negative (filled circles) ICSS placements. Numbers indicate the rahge (x 100)in which a given animal’s 15 minute ICSS score fell on its highest performance day (see text). List of abbreviations: AD Anterodorsal N. Hm medial Habenula Paraventricular N. Pv Am Anteromedial N. Internal capsule IC Pvgt Periventrictdar 8ray (thalami) Av Anteroventral N. Ld Laterodorsal N. Reuniens N. Re Central segment c Lgd dorsal Iateral Geniculate Rh Rhomboid N. Central medial N. Ce Lateral posterior N . Reticular N. Rt LP Central lateral N. Cl Md Mediodorsal N. SM Stria Medulatis FR Fasciculus Retroflexus PC Paracentral N. Vb Ventrobasal N. N . Gelatinosus Ge Pf Parafascicular N . Vbp Posterior ven~ob~~ N. lateral Habenula HI PO Posterior N. Ventrolateral N. VI Pt Paratenial N. Approx~te levels of the sections in this figure correspond to the appropriate number taken from K&rig, F.R. and R. A. Khppei, The Ret Brain: A Stereoraxic Atlas of the Forebrain and Lower Parts of ihe Brain Stem. Baltimore: Williams and WiIkins, 1963. A-_5@jO,C-5340, E-4890, G-4380, I-41 10, K-3290. saline f&owed by 10% Forma&n. Frozen sections (40 pm) in the coronaI plane were stained with cresyl violet, and examined under lightfield acumination using a Leitz Dialux microscope. RESULTS Of the 128 animals run in this study, 16 electrode sites were either lost during histological preparation, or else proved to be in the lateral ventricle. Of the remaining 112 animals, 57 failed to meet our criteria for ICSS, and were sacrificed after the pre-test trials. Their electrode sites are represented by the filled circles in Fig. 1. The remaining 55 rats responded with highest ICSS scores rauging from 55 to
921 in 15 minutes. In addition, 21 of these rats met the criterion for stabilized ICSS (described above), while 34 rats did not. A schematic summary of these results is presented in Fig. 1. Drawings in this figure are not taken from a published atlas, but are based upon our assessment of the histological boundaries in our material, and on the nuclear parce~ation schemes of Altmau and Bayer [i], Dvnogkue et al. ill], Jones and Leavitt 1211and Herkenham [191. The numbers on the right side of each section are placed at the sites of electrode tips of rats that passed the pre-test criterion for ICSS. Each number indicates the range in which the highest daily score for that animal fell. Thus, the number 1 indicates that
CLAVIER
.^
_--
-
.
AND GERFEN
--
FIG. 2. ~oto~cro~phs of certain positive placements for ICSS. Arrows indicate the tip of the electrode in each case; associated number indicates the highest daily score achieved at this site, and the letter indicates whether the ICSS at the site stabilized to a constant rate (C), or else was non-stabilized (N). Placements were considered to be in: (A) Md; (B) Vm; (C) pf, (D) I?; (E) Pv; (F) Vb.
the highest score for that animal was between 100 and 199; 2 indicates that the highest score was between 200 and 299, and so on. A zero indicates that the rat pressed between 50 and 99 times on its best day. Representative histological sections are shown in Fig. 2. MD Nucleus
A total of 25 p~cements fell within the limits of the mediodorsal (MD) nucleus. Of these, 13 met the pretest criterion for ICSS, but only 5 showed stabilized performance. Though no lirm trend could be established with so few cases, we have noted that the instance of stabilized ICSS also showed the highest daily rates, ranging from 200 to 500, while nonstabilized rates ranged from below 100 to the 300’s. Also, scores appeared to depend on proximity to the midline, with higher scores in the more medial placements and lower score more laterally (see for example, Fig. 11 and Fig. 2A). Interestingly, 7 of the MD placements were in the region of the central segment of that nucleus. All 7 of these placements were neutral with respect to ICSS (Fig. lC, G, H. I). Intra~a~~nar Nuclei
For the purposes of this paper, the following nuclei are considered to be in the intralaminar group: central medial; parafascicular; paracentral; and central lateral. In the central medial nucleus, 4 of the 5 placements were positive for ICSS; 2 of these showed stabilized scores. In the parafascicular nucleus, 3 of the 5 placements showed ICSS (Fig. ZC); 1 of these was stable. It may be noted here that none of the 6 placements in the fasciculus retroflexus supported ICSS. In the paracentral nucleus, 4 of the 8 placements supported ICSS; 2 of these were stable. In the central lateral
nucleus, 4 of the 5 placements supported ICSS, but none of these were stable. In summa~, 1.5 of the 23 intralaminar placements supported ICSS. Ventromedial
Nucleus
Of the 4 electrode placements in the ventromedial thalamic nucleus, all were positive for ICSS, and 2 showed stabilized rates (Fig. 2B). Another electrode tip was situated just medial to the ven~omedi~ nucleus in the nucleus gelatinosus, which is also known as the submedial nucleus. This site was neutral for ICSS (Fig. 1D). Relay Nuclei
This classiikation includes the ventrobasal, ventrolateral, lateral posterior, and laterodorsal nuclei (A&man and Bayer [l]). In all, 12 electrode sites were obtained in the ventrobasal nucleus. Of these, 6 showed ICSS, but only 1 with stable rates (Fig. 2F). 12 sites in the ventrolateral nucleus were also found, with 5 supporting ICSS, and 2 stabilized rates. Neither of the 2 lateral posterior sites, nor the single laterodorsal site supported ICSS. Midline Nuclei
This category includes the rhomboid, reuniens, paratenial, and paraventricular nuclei. In both the rhomboid and paraventricular nuclei, 2 placements were found, and in both cases, only 1 of these supported ICSS. However, each of those positive sites were related to stabilized scores in the 400’s (Fig. IB, D; Fig. 2E). The paratenial nucleus had 2 placements, both of which supported ICSS, and one of these showed stabilized rates (Fig. 2D). No placements were seen in the reuniens nucleus.
THALAMIC
357
SELF-STIMULATION
Posterior Nucleus
A total of 7 electrode sites were obtained in the posterior thalamic nucleus, and 3 of these supported ICSS. However, 2 of these sites were rather dorsally situated, and in close relation to the ventrobasal nucleus (Fig. lI), while the 3rd site was medially located, and adjacent to the parafascicular nucleus (Fig. 1L). Each of the 4 neutral placements in the posterior nucleus were more centrally located within its limits (Fig. 1H). Anterior Nuclei
These nuclei are represented by the anterodorsal, anteromedial, and anteroventral cell groups. The only placement in the anteroventral nucleus (Fig. 1A) supported ICSS that reached into the 500’s, but never stabilized. There were 3 sites in the anteromedial nucleus, but only 1 of these supported ICSS. However, though in the anteromedial nucleus, the 2 neutral placements were closely related to the stria medularis and the mammillothalamic tract, respectively (Fig. lA, C). Reticular Nucleus
None of the 3 placement ported ICSS. Paraventricular
in the reticular
nucleus
sup-
Gray of the Thalamus
In all, 6 placements were found in this region, and 3 of these supported ICSS (Fig. 1L). All 3 positive placements were ventral to the level of the 3 neutral sites, and in closer association with the parafascicular nucleus than were the latter sites. Of the 3 positive sites, 1 showed stabilized rates.
DISCUSSION
Positive ICSS placements have been found in each thalamic region whose potential relation to brain reward mechanisms was suggested by its anatomical connections with either the sulcal prefrontal cortex, or the substantia nigra. However, fewer than 40% of the positive sites showed stabilized performance rates. Nor were positive sites restricted to areas related to the 2 regions mentioned above; indeed, every thalamic region, including classical relay cell groups, had positive sites within their limits. Furthermore, the average current level at which ICSS occurred (approximately 45 PA, RMS) was higher than that observed in many ICSS reports. Though the degree of current spread from the tip of the electrode cannot be stated, Steiner et al. [34] present data indicating that ICSS corresponds to neuronanatomical entities closely associated with the bipolar electrode tips. (The reader may consult this paper for a critical discussion of relative current levels and current spread in ICSS.) Still, the possibility must be considered that certain of the placements that were considered positive in this study may have been so due to the spread of current beyond the anatomical limits of the nucleus in which the tip was situated. Two factors argue against this, however. First, more than half of the animals in this study failed to reach the initial ICSS criterion. If the current were able to spread from neutral electrode sites to positive ICSS regions, one might have expected to find a higher proportion of positive ICSS placements in this report. Second, superimposed upon the factor of current spread is that of possible aversive properties of the stimulation in both
the positive and the neutral placements. Though such properties were not assessed in this study, it is possible that elevating currents in this study would serve to drive responding rates downward rather than upward. Prefrontal Sulcal Cortex Relation
The sulcal cortex has been considered to be a major projection field of the MD thalamic nucleus 1241. More detailed analysis of this relation [13,23] has revealed a differential thalamic input to the sulcal cortex, based on the subdivision of the latter area into dorsal and ventral regions. Thalamic input to the dorsal region is from the caudal ventromedial aspect of the MD nucleus, and the area extending caudally from there to a point where it is continuous with the rostral aspect of the parafascicular nucleus. The latter nucleus, as well as the other intralaminar nuclei also supply inputs to the dorsal region, as does the ventromedial nucleus. Each of these nuclei supported ICSS in the present study. By contrast, thalamic inputs to the ventral region of the sulcal cortex originate from the central segment of the MD nucleus and from the submedial aspect of the ventromedial complex (nucleus gelatinosus). As noted, none of the 7 placements in the central segment of the MD nucleus supported ICSS in this study; nor did the single placement in the nucleus gelatinosus. However, the number of placements is too small to warrant any conclusions about the connections of the thalamus to the ventral region of the sulcal cortex, in regard to ICSS. Efferents from the sulcal cortex to the thalamus project to the MD, ventromedial, and midline nuclei [3, 14, 241. The first two of these thalamic regions have been discussed above. As for the midline nuclei, only 6 placements were found in this study; moreover, neither of the 2 paratenial placements were, strictly speaking, in the midline. On the other hand, 2 of the central medial, and 3 of the MD placements were in the midline. In all, then, 11 of the electrode tips in our study were situated within a vertical band that extended l/2 mm on either side of the midline. Of these, only 2 failed to support ICSS. The average high score for the remaining 9 animals was approximately 450 responses in 15 minutes, compared with an average high score of 120 responses in I5 minutes for positive sites that fell outside of this arbitrary boundary. Interestingly, there was no significant difference in the current level at which these animals bar-pressed for ICSS, compared with the animals whose placements fell outside this boundary. Nevertheless, the tendency toward higher scores as the placements approach the midline remains an interesting aspect of our findings. One may speculate that this focus on the midline thalamus relates to the anatomical relation between that region and the medial forebrain bundle detailed by Millhouse [26]. Cooper and Taylor [8] also described medially situated thalamic ICSS probes. Specifically, ICSS without pain-related behavior was seen with electrodes in the centralis medialis, centralis lateralis, rhomboideus, reuniens, and parafascicular nuclei. They noted that these nuclei “correspond to nuclei making up thalamic reticular system” of Jasper [20], and suggested that “specific motivational processes” could be added to the list of hypothesized functions for that system. In view of our present finding of ICSS outside of that system, and of recent revisions of the parcellation of thalamic nuclei 11,191, this conclusion may not be salient. More important, however, was their focus on midline thalamic nuclei in terms ofICSS.Cooper and Taylor did not see the midline nuclei as
358
CLAVIER
the exclusive mediators of thalamic ICSS; placements in the periventricular and central grey regions of the thalamus also supported ICSS, but with pain-related behavioral components. Substantia
Nigra
Relation
A recent description of nigrothalamic connections confirmed the existence of inputs to the ventromedial, parafascicular, and paralamellar MD nuclei [141. This study also showed that nigral inputs to the ventromedial nucleus is bilateral, and that the relationship between the substantia nigra and the parafascicular nucleus is reciprocal. In addition, some evidence indicated the existence of an ipsilateral nigral link with the laterodorsal nucleus; in the present study, the latter nucleus had only one placement, and it did
AND GERFEN
not support ICSS. However, as noted above, each of the other nigrothalamic projection areas had electrode sites related to ICSS. In summary then, the finding of ICSS sites in diverse aspects of the thalamus should serve as an impetus for further anatomical, physiological, pharmacological, and behavioral investigations of this complex region. This work will undoubtedly shed new light on the concept of brain stimulation-induced reward in general, and on the role of the thalamus in this phenomenon.
ACKNOWLEDGEMENTS
This work was supported by Grant MH 33987 to RMC. The authors wish to thank Mrs. Christine Stewart and Mr. Donald Henkelman for their excellent technical assistance.
REFERENCES 1. Altman, J. and S. A. Bayer. Development of the diencephalon in
the rat. V. Thymidine radiographic observations on internuclear and intranucleargradients. .I. camp. Neural. 188: 473-500, 1979. 2. Beckstead, R. M. An autoradiographic examination of corticocortical and subcortical projections of the mediodorsal projection (prefrontal) cortex in the rat. J. camp. Neurol. 106: 43-62, 1976. 3. Beckstead, R. M., V. Domesick and W. J. H. Nauta. Efferent connections of the substantia nigra and ventral tegmental area of the rat. Brain Res. 175: 191-217, 1979. 4. Clavier, R. M., S. Atmadja and H. C. Fibiger. Nigrothalamic projections as demonstrated by orthograde and retrograde tracing methods. Brain Res. Bull. 1: 379-384, 1976. 5. Clavier, R. M. and H. C. Fibiger. On the role of ascending catecholaminergic projections in intracranial self-stimulation of the substantia nigra. Brain Res. 131: 271-281, 1976. 6. Clavier, R. M. and A. Routtenberg. Brain stem self-stimulation attenuated by lesions of medial forebrain bundle but not by lesions of locus coeruleus or the caudal ventral norepinephrine bundle. Brain Res. 101: 251-271, 1976. 7. Clavier, R. M. and A. Routtenberg. Fibers associated with brain stem self-stimulation: Fink-Heimer study. Brain Res. 105: 325332, 1976. 8. Cooper, R. M. and L. H. Taylor. Thalamic reticular system and central gray: Self-stimulation. Science 156: 102-103, 1%7. 9. Corbett, D. and R. A. Wise. Intracranial self-stimulation in relation to ascending dopaminergic systems of the midbrain: A moveable electrode mapping study. Bruin Res. 185: l-16, 1980. 10. Crow, T. J. A map of the rat mesencephalon for electrical selfstimulation. Brain Res. 37: 265-273, 1972. 11. Donoghue, J. P., K. L. Kerman and F. F. Ebner. Evidence for two organizational plans within the somatic sensory cortex of the rat. .I. come. Neural. 183: 647-664. 1979. 12. Failon, J. H. Ad R. Y. Moore. Catecholamine innervation of the basal forebrain. IV. Topography of dopamine projections to the basal forebrain and neostriatum. J. camp. Neurol. 180: 545-580, 1978. 13. Gerfen, C. R. and R. M. Clavier. Neural inputs to the prefrontal agranular insular cortex in the rat: Horseradish peroxidase study. Brain Res. Bull. 4: 347-353, 1979. 14. Gerfen, C. R. and R. M. Clavier. Intracranial self-stimulation from the sulcal prefrontal cortex in the rat: The effect of 6-hvdroxydopamine or kainic acid lesions at the site of stimulation. Brain Rhs. 224: 291-304, 1981. 15. Gexfen. C. R.. W. A. Staines. G. W. Arbuthnott and H. C. Fibigerl Crossed connections of the substantia nigra of the rat. Submitted, 1981. 16. Gillett, G. R. and D. M. Webster. Mediodorsal nucleus and behavioral regulation in the rat. Physiol. Behav. 14: 883-885. 1975.
17. Heath, R. G. Electrical self-stimulation of the brain in man. Am. J. Psychiat. 120: 571-577, 1%3. 18. Herkenham, M. The afferent and efferent connections of the ventromedial thalamic nucleus in the rat. J. camp. Neural. 183: 487-518, 1979. 19. Herkenham, M. Laminar organization of thalamic projections to rat neocortex. Science 207: 532-535, 1980. 20. Jasper, H. H. In: Handbook of Physiology, vol. 2, section I: Neuroohvsioloev. edited bv J. Field, H. W. Maaoun and V. E. Hall. W&hi&&, DC: American Physiological-Society, 1960, p. 1307. 21. Jones, E. G. and R. Y. Leavitt. Retrograde axonal transport and the demonstration of non-specific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei in the rat. .I. camp. Neural. 154: 349-378, 1974. 22. Keene, J. J. and K. L. Casey. Rewarding and aversive brain stimulation: Opposite effects on medial thalamic units. Physiol. Behav. 10: 283-287, 1973. 23. Krettek, J. E. and J. L. Price. The cortical projections of the mediodorsal nucleus and adjacent thalamic nuclei in the rat. J. camp. Neural. 171: 157-192, 1977. 24. Leonard, C. M. The prefrontal cortex of the rat. I. Cortical projections of the mediodorsal nucleus. II. Efferent connections Brain Res. 12: 321-343, 1969. 25. Lindvall, O., A. Bjorklund and I. Divac. Organization of the catecholamine neurons projecting to the frontal cortex in the rat. Brain Res. 142: l-24, 1978. 26. Millhouse, 0. E. A golgi study of the descending medial forebrain bundle. Brain Res. 15: 341-363, 1%9. 27. Olds, J. Drives and Reinforcements: Behavioral Studies of Hyoothalamic Functions. New York: Raven, 1977. analysis of rat 28. ’Olds, M. E. and J. Olds. Approach-avoidance diencephalon. J. romp. Neural. 120: 259-295, 1963. 29. Redgrave, P. and P. Dean. Intracranial self-stimulation. Br. med. Bull. 37: 141-146, 1981. 30. Rolls, E. T. The neurophysiological basis of brain-stimulation reward. In: Brain Stimulation Reward, edited by A. Wauquier and E. T. Rolls. Amsterdam: North Holland, 1976, pp. 65-88. 31. Routtenberg, A. and C. W. Malsbury. Brainstem pathways of reward. J. camp. physiol. Psychol. 68: 22-30, 1969. 32. Routtenberg, A. and M. Sloan. Self-stimulation in the frontal cortex of Rattus norvegicus. Behav. Biol. 7: 567-572, 1972. 33. Stein. L. The chemistrv of reward. In: Biology of Reinforcement: Facets of Brain-Stimulation Reward, e&ed-by A.-Routtenberg. New York: Academic Press, 1980, pp. 109-130. 34. Steiner, S. S., R. J. Bodnar, W. T. Nelson, Jr., R. F. Ackermann and S. J. Ellman. Intracranial self-stimulation site specificity: The myth of current spread. Brain Res. Bull. 3: 349-356, 1978.
Vol. 8, pp. 353-358, 1982. Printed in the U.S.A.
Intracranial Self-Stimulation in the Thalamus of the Rat RONALD M. CLAVIER’ Depurr~en~ of Biopsych~f~gy, The Clarke Znst~tule of Psy~h~a~~ 250 College St., Toronto, Ontario MST IR8, Canada AND CHARLES
R. GERFEN
Weingart L~~orato~ for Deve~op~eniffl ~euru&iufog~~, The Salk Znstitute P.O. Box 85800, San Diego, CA 92138 Received 21 January
1982
CLAVIER, R. M. AND C. R. GERFEN. Zntrucrunialsetf-st~muiation in the thdamus of the rut. BRAIN RES. BULL. g(4) 353-358, 1%2.-Rats were tested for intracranial set-stim~ation (ICSS) via bipolar ekctrodes situated copout the thalamus. Of 112 animals in the study, 55 met the criteria for ICSS, with scores ranging from 55 to 921 bar-presses in a 15 minute session. A map of both positive and neutral placements is presented. Positive sites for ICSS were found in all aspects of the mediodorsal nucleus, except for the central segment. The ventromediai nuclear compkx was also a positive area of ICSS, with the exception of the submedial nucleus (nucleus gelatinosus). Each of the intmkmimu nuclei (central medial, parafascicular, paracentral, and central lateral) supported ICSS, as did each of the midline nuclei (rhomboid, paratenial and paraventrkukr). No placements were found in the nucleus reuniens, Both “major” relay nuclei, the ventrobasaf and ventrolaterai, supported ICSS; but neither the taterodorsal nor the lateral posterior nuclei had positive sites. Positive sites were found in the anterior and posterior nuclear groups, as we11as in the paraventricular gray area of the caudai thabunus. As a general rule, ICSS scores appeared to be higher as the electrode placements approached the midline. Sites in which no positive placements were seen included the reticular nucleus, as well as the stria medukris, the mammillothakmic tract, and the fasciculus retroflexus. Thakmus
Self-stimulation
BECAUSE of its potential relation to brain reward mechanisms, intracranial self-stimulation (ICSS) has been investigated in great detail. A major focus of this work concerns efforts to identify the anatomical substrates of ICSS. Perhaps the most clear-cut findings of this focus are: (1) that ICSS is obtained from a wide variety of brain regions; and (2) that efforts to identify a common underlying substrate for ICSS are not successful (see Redgrave and Dean [29] for a recent review). Despite this, however, certain consistencies have been found. One of these is the topic of the present study. As a result of studies aimed at describing the ~ato~c~ connections of brain regions that support ICSS, our attention has been drawn repeatedly to the thalamus. For example, Keen and Casey [221 were able to establish a functional relation between stimulation of the medial forebrain bundle via ICSS electrodes and the firing pattern of single units in the ventromedial thalamic region. The presumed anatomical basis for this effect may reside in the GoIgi data of M~ouse [26], which established MFB connections with the midline
thalamic nuclei. Also, Clavier and Routtenberg [7], in describing axonal degeneration after lesions made via brainstem ICSS electrodes, identified 2 ascending pathways: one that ran adjacent to the parafascicular nucleus, and terminated in the ventromedial and dorsomedial thalamic nuclei; and another that terminated in the region of the medial forebrain bundle. Electrolytic lesions in either of these projection regions were shown to attenuate brainstem ICSS [61, indicating a possible functional relation between the medial forebrain bundle and the medial thalamic regions. Two other ICSS areas are related to the thalamus: the sulcal prefrontal cortex; and the substantia nigra. The sulcal cortex is the area on the dorsal bank of the rhinal sulcus in the prefrontal cortex of the rat. This region is the site of ICSS placements [32], and represents one of the projection fields of the “AlO” dopaminergic system [ 12,251, which has been implicated in brain reward mechanisms 110,331. The sulcal cortex also receives input from the thalamus. Specifically, the mediodorsal, ventromedial, and intralaminar (~clu~ng the parafascicular) nuclei all project to this region 113, 23,
‘To whom requests for reprints should be sent.
~opy~g~t
@ 1982 ANKHO
Intemation~
Inc .-0361-9230/82/040353-06$03.00/O
354
241.Moreover, a descending projection exists from the sulcal area to several thalamic nuclei, including the mediodorsal, ventrome~~, and midline groups 1141. Because of these connections, each of these thalamic nuclei received further attention as possible contributors to the observed behavior. Substantia nigra ICSS has been hypothesized to depend on the stimulation of the “A9” dopaminergic cell group 19,101. Yet, ICSS in this area continues after cases of total depletion of dopamine from the ni~ost~at~ system [5]. Thus, we have considered other systems in the vicinity of the electrodes that might support ICSS. Part of this effort deals with the relation of the substantia nigra to the thalamus. As the recipient of nigral inputs, the parafascicular and ventromedial nuclei, and the paralamellar aspect of the mediodorsal nucleus [2, 4, 151, were all considered to be related, potenti~y, to ICSS. One method for assessing the relation of these thalamic nuclei to ICSS is to see whether electrodes in any or all of them are positive sites of ICSS. The results of that approach are the subject of the present report. Prior to presenting our findings, however, a brief summary of previous work on thalamic ICSS is warranted. The most comprehensive ex~nation of thalamic ICSS to date is that of Olds and Olds [ZS]. The authors described 32 electrode sites in 15 nuclear regions. Each site was tested for avoidance, as measured in a bar-pressing paradigm for escape, as well as for approach, as measured by ICSS. This provided the scope to ascribe qualities of pure approach, pure avoidance, or mixed approach-avoid~ce to each site. They used an approach criterion of 200 responses in 8 minutes, which they admitted might have deleted several cases of fairly consistent bar-pressing from the approach category. Had they used less stringent criteria-for example those of Routtenberg and Malsbury [311_and extrapolated their scores to 1.5 minute session equivalents, they might have identified 8 cases of moderate ICSS (2~5~), 14 cases of low ICSS @O-199), and 9 cases of neutrality. In the case of their highest rate of ICSS (860 responses in 8 minutes), the electrode tip was situated in the subthalamic nucleus, and not in the thalamic reticular nucleus as the authors indicated. This discrepancy aside, each of the thalamic nuclear areas examined in the Olds and Olds study is represented by at least one case of ICSS. Yet, according to their own criteria, only 2 of their placements showed ICSS properties, and both of these showed avoiance qualities as well. This led Olds and Glds to conclude that “. . . there is . . . almost no approach in the thalamus,” and that “One is led to wonder about the evolutionary and functional significance of an arrangement which appears to put negative mechanisms only in thalamic systems . . . . ” Olds [27] later revised this view, accepting the thalamus as a valid ICSS site. However, he noted that the presence of clear-cut aversive areas, and their possible overlap with ICSS regions in the thalamus could result in the operational existence of “marginal” areas; these could be moved into or out of brain reward systems based on methodological parameters such as probe size, stimulating parameters, or behaviors paradigm. The aversive potential of thalamic stimulation was ak.0 noted by Gillett and Webster [161, who showed that me&odorsal thalamic stimulation attenuated bar-pressing on a baseline schedule of water reinforcement. The role of the thalamus in ICSS remains unclear today. Rolls [30] has written that “Stimulation at the majority of other brain sites, for example in much of neocortex, thalamus, and cerebellum, is neutral , . . .” Also, Stein 1331
CLAVIER
AND GERFEN
noted that “Most regions of the thalamus do not support self-st~ulation . . . .” Nevertheless, Stein did report finding 5 probes in the region of the paratenial thalamic nucleus that supported “respectable” rates of ICSS. Thalamic sites of ICSS were also reported by Cooper and Taylor [8]. They showed positive sites in the midline and intralaminar nuclear groups, as well as mixed ICSS-pain placements in the periventricular and central gray areas of the thalamus. In the case of human brain stimulation, Heath [ 171cited an instance of centromedian ICSS, which carried the subjective sensation of frustration and anger related to partial loss of memory. Thus, while the thalamus appears to be related to ICSS. it is often unrelated to pure reward properties, and the anatomical details of its involvement are not clear. METHOD One hundred twenty-eight adult male albino rats (Holtzman) weighing 275-350 grams prior to surgery were used in this study. Each rat had a single bipolar nichrome wire electrode (Plastic Products Co., MS 303.081 -0.3120.005 in.) implanted while under pentobarbital anesthesia. Electrodes were implanted ~~endic~~ to the skull surface using stereotaxic coordinates that ranged from 5.0 mm to 6.5 mm rostra& 0.1 mm to 3.5 mm lateral, and 2.2 mm to 4.2 mm dorsal to stereotaxic zero, and with the incisor bar set at 4.2 mm below horizontal. At least 10 days were allowed for recovery prior to the start of behavioral testing. Testing for ICSS was conducted in operant conditioning chambers, in which each depression of a lever resulted in a 0.5 set pulse of 60 Hz sine wave current. Each rat underwent a minimum of 21 days of pre-test trials, each lasting at least 2 hours, unless ICSS criteria were met prior to this. During the pre-test trials, the rats were given Experimenter-delivered stimulations for 1-2 minutes at 30 minute intervals. The current level for these stimulations was in the order of 55 PA @MS). Current was raised gradually in the case of neutral reactions on the part of the rats, or else lowered if an aversive or motor response was elicited. Between bouts of Experimenter-delivered stimulations, the animals were left in the chambers and the number of bar-presses was recorded. If the number of daily bar-presses by a rat did not exceed 100 on any of the 21 days of such trials, the animal was killed by pentob~bit~ overdose followed by intracardiac perfusion; its electrode site was recorded and considered neutral for ICSS. If a rat passed the initial criterion for ICSS, it was run daily in 15 minute sessions with the current set at the level at which it responded with a steady rate. We have, moreover, set a criterion for stabilized ICSS. This was said to have occurred when there were any 5 consecutive daily test sessions in which the score in 15 minutes did not fall by more than 1 standard deviation from the mean of those 5 days. If the score of a rat fell below this criterion, the current was raised until the response rate was in the range of the criterion scores. In the event that an ICSS score fell in our assessment due to motor or aversive qualities of the current, the current was lowered until such qualities were no longer apparently interfering with the ICSS. In addition to the criterion for stabilized ICSS, we have used an animal’s highest daily 15 minute score to identify a given placement. This is in conformity with the procedure of Olds and Olds [28]. At the end of testing, all animals were killed by pentobarbital overdose and intracardiac perfusion with 0.9%
THALAMIC
SELF-STIMULATION
FIG. 1. Schematic representation of positive (numbers) and negative (filled circles) ICSS placements. Numbers indicate the rahge (x 100)in which a given animal’s 15 minute ICSS score fell on its highest performance day (see text). List of abbreviations: AD Anterodorsal N. Hm medial Habenula Paraventricular N. Pv Am Anteromedial N. Internal capsule IC Pvgt Periventrictdar 8ray (thalami) Av Anteroventral N. Ld Laterodorsal N. Reuniens N. Re Central segment c Lgd dorsal Iateral Geniculate Rh Rhomboid N. Central medial N. Ce Lateral posterior N . Reticular N. Rt LP Central lateral N. Cl Md Mediodorsal N. SM Stria Medulatis FR Fasciculus Retroflexus PC Paracentral N. Vb Ventrobasal N. N . Gelatinosus Ge Pf Parafascicular N . Vbp Posterior ven~ob~~ N. lateral Habenula HI PO Posterior N. Ventrolateral N. VI Pt Paratenial N. Approx~te levels of the sections in this figure correspond to the appropriate number taken from K&rig, F.R. and R. A. Khppei, The Ret Brain: A Stereoraxic Atlas of the Forebrain and Lower Parts of ihe Brain Stem. Baltimore: Williams and WiIkins, 1963. A-_5@jO,C-5340, E-4890, G-4380, I-41 10, K-3290. saline f&owed by 10% Forma&n. Frozen sections (40 pm) in the coronaI plane were stained with cresyl violet, and examined under lightfield acumination using a Leitz Dialux microscope. RESULTS Of the 128 animals run in this study, 16 electrode sites were either lost during histological preparation, or else proved to be in the lateral ventricle. Of the remaining 112 animals, 57 failed to meet our criteria for ICSS, and were sacrificed after the pre-test trials. Their electrode sites are represented by the filled circles in Fig. 1. The remaining 55 rats responded with highest ICSS scores rauging from 55 to
921 in 15 minutes. In addition, 21 of these rats met the criterion for stabilized ICSS (described above), while 34 rats did not. A schematic summary of these results is presented in Fig. 1. Drawings in this figure are not taken from a published atlas, but are based upon our assessment of the histological boundaries in our material, and on the nuclear parce~ation schemes of Altmau and Bayer [i], Dvnogkue et al. ill], Jones and Leavitt 1211and Herkenham [191. The numbers on the right side of each section are placed at the sites of electrode tips of rats that passed the pre-test criterion for ICSS. Each number indicates the range in which the highest daily score for that animal fell. Thus, the number 1 indicates that
CLAVIER
.^
_--
-
.
AND GERFEN
--
FIG. 2. ~oto~cro~phs of certain positive placements for ICSS. Arrows indicate the tip of the electrode in each case; associated number indicates the highest daily score achieved at this site, and the letter indicates whether the ICSS at the site stabilized to a constant rate (C), or else was non-stabilized (N). Placements were considered to be in: (A) Md; (B) Vm; (C) pf, (D) I?; (E) Pv; (F) Vb.
the highest score for that animal was between 100 and 199; 2 indicates that the highest score was between 200 and 299, and so on. A zero indicates that the rat pressed between 50 and 99 times on its best day. Representative histological sections are shown in Fig. 2. MD Nucleus
A total of 25 p~cements fell within the limits of the mediodorsal (MD) nucleus. Of these, 13 met the pretest criterion for ICSS, but only 5 showed stabilized performance. Though no lirm trend could be established with so few cases, we have noted that the instance of stabilized ICSS also showed the highest daily rates, ranging from 200 to 500, while nonstabilized rates ranged from below 100 to the 300’s. Also, scores appeared to depend on proximity to the midline, with higher scores in the more medial placements and lower score more laterally (see for example, Fig. 11 and Fig. 2A). Interestingly, 7 of the MD placements were in the region of the central segment of that nucleus. All 7 of these placements were neutral with respect to ICSS (Fig. lC, G, H. I). Intra~a~~nar Nuclei
For the purposes of this paper, the following nuclei are considered to be in the intralaminar group: central medial; parafascicular; paracentral; and central lateral. In the central medial nucleus, 4 of the 5 placements were positive for ICSS; 2 of these showed stabilized scores. In the parafascicular nucleus, 3 of the 5 placements showed ICSS (Fig. ZC); 1 of these was stable. It may be noted here that none of the 6 placements in the fasciculus retroflexus supported ICSS. In the paracentral nucleus, 4 of the 8 placements supported ICSS; 2 of these were stable. In the central lateral
nucleus, 4 of the 5 placements supported ICSS, but none of these were stable. In summa~, 1.5 of the 23 intralaminar placements supported ICSS. Ventromedial
Nucleus
Of the 4 electrode placements in the ventromedial thalamic nucleus, all were positive for ICSS, and 2 showed stabilized rates (Fig. 2B). Another electrode tip was situated just medial to the ven~omedi~ nucleus in the nucleus gelatinosus, which is also known as the submedial nucleus. This site was neutral for ICSS (Fig. 1D). Relay Nuclei
This classiikation includes the ventrobasal, ventrolateral, lateral posterior, and laterodorsal nuclei (A&man and Bayer [l]). In all, 12 electrode sites were obtained in the ventrobasal nucleus. Of these, 6 showed ICSS, but only 1 with stable rates (Fig. 2F). 12 sites in the ventrolateral nucleus were also found, with 5 supporting ICSS, and 2 stabilized rates. Neither of the 2 lateral posterior sites, nor the single laterodorsal site supported ICSS. Midline Nuclei
This category includes the rhomboid, reuniens, paratenial, and paraventricular nuclei. In both the rhomboid and paraventricular nuclei, 2 placements were found, and in both cases, only 1 of these supported ICSS. However, each of those positive sites were related to stabilized scores in the 400’s (Fig. IB, D; Fig. 2E). The paratenial nucleus had 2 placements, both of which supported ICSS, and one of these showed stabilized rates (Fig. 2D). No placements were seen in the reuniens nucleus.
THALAMIC
357
SELF-STIMULATION
Posterior Nucleus
A total of 7 electrode sites were obtained in the posterior thalamic nucleus, and 3 of these supported ICSS. However, 2 of these sites were rather dorsally situated, and in close relation to the ventrobasal nucleus (Fig. lI), while the 3rd site was medially located, and adjacent to the parafascicular nucleus (Fig. 1L). Each of the 4 neutral placements in the posterior nucleus were more centrally located within its limits (Fig. 1H). Anterior Nuclei
These nuclei are represented by the anterodorsal, anteromedial, and anteroventral cell groups. The only placement in the anteroventral nucleus (Fig. 1A) supported ICSS that reached into the 500’s, but never stabilized. There were 3 sites in the anteromedial nucleus, but only 1 of these supported ICSS. However, though in the anteromedial nucleus, the 2 neutral placements were closely related to the stria medularis and the mammillothalamic tract, respectively (Fig. lA, C). Reticular Nucleus
None of the 3 placement ported ICSS. Paraventricular
in the reticular
nucleus
sup-
Gray of the Thalamus
In all, 6 placements were found in this region, and 3 of these supported ICSS (Fig. 1L). All 3 positive placements were ventral to the level of the 3 neutral sites, and in closer association with the parafascicular nucleus than were the latter sites. Of the 3 positive sites, 1 showed stabilized rates.
DISCUSSION
Positive ICSS placements have been found in each thalamic region whose potential relation to brain reward mechanisms was suggested by its anatomical connections with either the sulcal prefrontal cortex, or the substantia nigra. However, fewer than 40% of the positive sites showed stabilized performance rates. Nor were positive sites restricted to areas related to the 2 regions mentioned above; indeed, every thalamic region, including classical relay cell groups, had positive sites within their limits. Furthermore, the average current level at which ICSS occurred (approximately 45 PA, RMS) was higher than that observed in many ICSS reports. Though the degree of current spread from the tip of the electrode cannot be stated, Steiner et al. [34] present data indicating that ICSS corresponds to neuronanatomical entities closely associated with the bipolar electrode tips. (The reader may consult this paper for a critical discussion of relative current levels and current spread in ICSS.) Still, the possibility must be considered that certain of the placements that were considered positive in this study may have been so due to the spread of current beyond the anatomical limits of the nucleus in which the tip was situated. Two factors argue against this, however. First, more than half of the animals in this study failed to reach the initial ICSS criterion. If the current were able to spread from neutral electrode sites to positive ICSS regions, one might have expected to find a higher proportion of positive ICSS placements in this report. Second, superimposed upon the factor of current spread is that of possible aversive properties of the stimulation in both
the positive and the neutral placements. Though such properties were not assessed in this study, it is possible that elevating currents in this study would serve to drive responding rates downward rather than upward. Prefrontal Sulcal Cortex Relation
The sulcal cortex has been considered to be a major projection field of the MD thalamic nucleus 1241. More detailed analysis of this relation [13,23] has revealed a differential thalamic input to the sulcal cortex, based on the subdivision of the latter area into dorsal and ventral regions. Thalamic input to the dorsal region is from the caudal ventromedial aspect of the MD nucleus, and the area extending caudally from there to a point where it is continuous with the rostral aspect of the parafascicular nucleus. The latter nucleus, as well as the other intralaminar nuclei also supply inputs to the dorsal region, as does the ventromedial nucleus. Each of these nuclei supported ICSS in the present study. By contrast, thalamic inputs to the ventral region of the sulcal cortex originate from the central segment of the MD nucleus and from the submedial aspect of the ventromedial complex (nucleus gelatinosus). As noted, none of the 7 placements in the central segment of the MD nucleus supported ICSS in this study; nor did the single placement in the nucleus gelatinosus. However, the number of placements is too small to warrant any conclusions about the connections of the thalamus to the ventral region of the sulcal cortex, in regard to ICSS. Efferents from the sulcal cortex to the thalamus project to the MD, ventromedial, and midline nuclei [3, 14, 241. The first two of these thalamic regions have been discussed above. As for the midline nuclei, only 6 placements were found in this study; moreover, neither of the 2 paratenial placements were, strictly speaking, in the midline. On the other hand, 2 of the central medial, and 3 of the MD placements were in the midline. In all, then, 11 of the electrode tips in our study were situated within a vertical band that extended l/2 mm on either side of the midline. Of these, only 2 failed to support ICSS. The average high score for the remaining 9 animals was approximately 450 responses in 15 minutes, compared with an average high score of 120 responses in I5 minutes for positive sites that fell outside of this arbitrary boundary. Interestingly, there was no significant difference in the current level at which these animals bar-pressed for ICSS, compared with the animals whose placements fell outside this boundary. Nevertheless, the tendency toward higher scores as the placements approach the midline remains an interesting aspect of our findings. One may speculate that this focus on the midline thalamus relates to the anatomical relation between that region and the medial forebrain bundle detailed by Millhouse [26]. Cooper and Taylor [8] also described medially situated thalamic ICSS probes. Specifically, ICSS without pain-related behavior was seen with electrodes in the centralis medialis, centralis lateralis, rhomboideus, reuniens, and parafascicular nuclei. They noted that these nuclei “correspond to nuclei making up thalamic reticular system” of Jasper [20], and suggested that “specific motivational processes” could be added to the list of hypothesized functions for that system. In view of our present finding of ICSS outside of that system, and of recent revisions of the parcellation of thalamic nuclei 11,191, this conclusion may not be salient. More important, however, was their focus on midline thalamic nuclei in terms ofICSS.Cooper and Taylor did not see the midline nuclei as
358
CLAVIER
the exclusive mediators of thalamic ICSS; placements in the periventricular and central grey regions of the thalamus also supported ICSS, but with pain-related behavioral components. Substantia
Nigra
Relation
A recent description of nigrothalamic connections confirmed the existence of inputs to the ventromedial, parafascicular, and paralamellar MD nuclei [141. This study also showed that nigral inputs to the ventromedial nucleus is bilateral, and that the relationship between the substantia nigra and the parafascicular nucleus is reciprocal. In addition, some evidence indicated the existence of an ipsilateral nigral link with the laterodorsal nucleus; in the present study, the latter nucleus had only one placement, and it did
AND GERFEN
not support ICSS. However, as noted above, each of the other nigrothalamic projection areas had electrode sites related to ICSS. In summary then, the finding of ICSS sites in diverse aspects of the thalamus should serve as an impetus for further anatomical, physiological, pharmacological, and behavioral investigations of this complex region. This work will undoubtedly shed new light on the concept of brain stimulation-induced reward in general, and on the role of the thalamus in this phenomenon.
ACKNOWLEDGEMENTS
This work was supported by Grant MH 33987 to RMC. The authors wish to thank Mrs. Christine Stewart and Mr. Donald Henkelman for their excellent technical assistance.
REFERENCES 1. Altman, J. and S. A. Bayer. Development of the diencephalon in
the rat. V. Thymidine radiographic observations on internuclear and intranucleargradients. .I. camp. Neural. 188: 473-500, 1979. 2. Beckstead, R. M. An autoradiographic examination of corticocortical and subcortical projections of the mediodorsal projection (prefrontal) cortex in the rat. J. camp. Neurol. 106: 43-62, 1976. 3. Beckstead, R. M., V. Domesick and W. J. H. Nauta. Efferent connections of the substantia nigra and ventral tegmental area of the rat. Brain Res. 175: 191-217, 1979. 4. Clavier, R. M., S. Atmadja and H. C. Fibiger. Nigrothalamic projections as demonstrated by orthograde and retrograde tracing methods. Brain Res. Bull. 1: 379-384, 1976. 5. Clavier, R. M. and H. C. Fibiger. On the role of ascending catecholaminergic projections in intracranial self-stimulation of the substantia nigra. Brain Res. 131: 271-281, 1976. 6. Clavier, R. M. and A. Routtenberg. Brain stem self-stimulation attenuated by lesions of medial forebrain bundle but not by lesions of locus coeruleus or the caudal ventral norepinephrine bundle. Brain Res. 101: 251-271, 1976. 7. Clavier, R. M. and A. Routtenberg. Fibers associated with brain stem self-stimulation: Fink-Heimer study. Brain Res. 105: 325332, 1976. 8. Cooper, R. M. and L. H. Taylor. Thalamic reticular system and central gray: Self-stimulation. Science 156: 102-103, 1%7. 9. Corbett, D. and R. A. Wise. Intracranial self-stimulation in relation to ascending dopaminergic systems of the midbrain: A moveable electrode mapping study. Bruin Res. 185: l-16, 1980. 10. Crow, T. J. A map of the rat mesencephalon for electrical selfstimulation. Brain Res. 37: 265-273, 1972. 11. Donoghue, J. P., K. L. Kerman and F. F. Ebner. Evidence for two organizational plans within the somatic sensory cortex of the rat. .I. come. Neural. 183: 647-664. 1979. 12. Failon, J. H. Ad R. Y. Moore. Catecholamine innervation of the basal forebrain. IV. Topography of dopamine projections to the basal forebrain and neostriatum. J. camp. Neurol. 180: 545-580, 1978. 13. Gerfen, C. R. and R. M. Clavier. Neural inputs to the prefrontal agranular insular cortex in the rat: Horseradish peroxidase study. Brain Res. Bull. 4: 347-353, 1979. 14. Gerfen, C. R. and R. M. Clavier. Intracranial self-stimulation from the sulcal prefrontal cortex in the rat: The effect of 6-hvdroxydopamine or kainic acid lesions at the site of stimulation. Brain Rhs. 224: 291-304, 1981. 15. Gexfen. C. R.. W. A. Staines. G. W. Arbuthnott and H. C. Fibigerl Crossed connections of the substantia nigra of the rat. Submitted, 1981. 16. Gillett, G. R. and D. M. Webster. Mediodorsal nucleus and behavioral regulation in the rat. Physiol. Behav. 14: 883-885. 1975.
17. Heath, R. G. Electrical self-stimulation of the brain in man. Am. J. Psychiat. 120: 571-577, 1%3. 18. Herkenham, M. The afferent and efferent connections of the ventromedial thalamic nucleus in the rat. J. camp. Neural. 183: 487-518, 1979. 19. Herkenham, M. Laminar organization of thalamic projections to rat neocortex. Science 207: 532-535, 1980. 20. Jasper, H. H. In: Handbook of Physiology, vol. 2, section I: Neuroohvsioloev. edited bv J. Field, H. W. Maaoun and V. E. Hall. W&hi&&, DC: American Physiological-Society, 1960, p. 1307. 21. Jones, E. G. and R. Y. Leavitt. Retrograde axonal transport and the demonstration of non-specific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei in the rat. .I. camp. Neural. 154: 349-378, 1974. 22. Keene, J. J. and K. L. Casey. Rewarding and aversive brain stimulation: Opposite effects on medial thalamic units. Physiol. Behav. 10: 283-287, 1973. 23. Krettek, J. E. and J. L. Price. The cortical projections of the mediodorsal nucleus and adjacent thalamic nuclei in the rat. J. camp. Neural. 171: 157-192, 1977. 24. Leonard, C. M. The prefrontal cortex of the rat. I. Cortical projections of the mediodorsal nucleus. II. Efferent connections Brain Res. 12: 321-343, 1969. 25. Lindvall, O., A. Bjorklund and I. Divac. Organization of the catecholamine neurons projecting to the frontal cortex in the rat. Brain Res. 142: l-24, 1978. 26. Millhouse, 0. E. A golgi study of the descending medial forebrain bundle. Brain Res. 15: 341-363, 1%9. 27. Olds, J. Drives and Reinforcements: Behavioral Studies of Hyoothalamic Functions. New York: Raven, 1977. analysis of rat 28. ’Olds, M. E. and J. Olds. Approach-avoidance diencephalon. J. romp. Neural. 120: 259-295, 1963. 29. Redgrave, P. and P. Dean. Intracranial self-stimulation. Br. med. Bull. 37: 141-146, 1981. 30. Rolls, E. T. The neurophysiological basis of brain-stimulation reward. In: Brain Stimulation Reward, edited by A. Wauquier and E. T. Rolls. Amsterdam: North Holland, 1976, pp. 65-88. 31. Routtenberg, A. and C. W. Malsbury. Brainstem pathways of reward. J. camp. physiol. Psychol. 68: 22-30, 1969. 32. Routtenberg, A. and M. Sloan. Self-stimulation in the frontal cortex of Rattus norvegicus. Behav. Biol. 7: 567-572, 1972. 33. Stein. L. The chemistrv of reward. In: Biology of Reinforcement: Facets of Brain-Stimulation Reward, e&ed-by A.-Routtenberg. New York: Academic Press, 1980, pp. 109-130. 34. Steiner, S. S., R. J. Bodnar, W. T. Nelson, Jr., R. F. Ackermann and S. J. Ellman. Intracranial self-stimulation site specificity: The myth of current spread. Brain Res. Bull. 3: 349-356, 1978.