Brain Research 881 2000 103–111 www.elsevier.com locate bres
Research report
The substrate for brain-stimulation reward in the lateral preoptic area I. Anatomical mapping of its boundaries
a b ,
b
Tamara Bushnik , Catherine Bielajew , Anne T.M. Konkle
a
TBI SCI Grants Office , San Jose, CA 95128, USA
b
School of Psychology , University of Ottawa, Ottawa, Ontario, K1N 6N5, Canada
Received 7 February 2000; accepted 30 May 2000
Abstract
Given the putative role of the lateral preoptic area as a primary contributor of the cell bodies of origin of the descending pathway linking a subset of lateral hypothalamic and ventral tegmental area reward neurons, the distribution of self-stimulation sites in this
structure was mapped in 22 animals using moveable electrodes and threshold procedures. Ninety-seven electrode sites were evaluated with placements ranging from just rostral to the midline convergence of the anterior commissure back to the transition zone between the
lateral preoptic and lateral hypothalamic areas; of these, roughly 2 3 supported self-stimulation which was widely observed throughout the lateral preoptic area and medial forebrain bundle. In general, self-stimulation thresholds obtained from lateral sites were most stable,
and progressively so approaching more caudal regions. Examination of the slopes of the period current trade-off functions revealed a tendency for higher values in lateral and caudal sites; in contrast, dorsoventral excursions did not influence these estimates. Taken
together, these data provide support for the notion that the substrate for brain-stimulation reward in the lateral preoptic area has a relatively homogeneous distribution that is more diffusely organized than that found in reward sites activated further caudally in the
medial forebrain bundle.
2000 Elsevier Science B.V. All rights reserved.
Theme : Neural basis of behavior
Topic : Motivation and emotion
Keywords : Brain-stimulation reward; Mapping; Lateral preoptic area; Trade-off functions; Moveable electrode; Medial forebrain bundle
1. Introduction have shown that neurons in the LPO and the area between
the preoptic and anterior hypothalamic nuclei respond The lateral preoptic area LPO, situated rostral to the
antidromically to stimulation of the medial forebrain lateral hypothalamus LH, has reciprocal connections with
bundle MFB [32,36]. Cells responsive to activation of both the ventral tegmental area VTA and LH [33,40], and
sites in the MFB that support self-stimulation are present is linked to the latter by a short transition zone, frequently
in the lateral and medial preoptic areas, the ventral referred to as the bed nucleus of the medial forebrain
pallidum, and the parastrial nucleus, with a much larger bundle [29]; rostral and caudal parcellations correspond to
region of antidromically driven neurons existing in and the LPO and LH, respectively [15,16]. In view of its
around the septal complex [38]. particular anatomical arrangement, the LPO is considered
Metabolic markers have been used to further define an important candidate for the cell bodies of origin of the
regions selectively activated by self-stimulation electrodes descending pathway that links a subset of LH and VTA
in the
MFB. Autoradiography
with 2-deoxyglucose
reward neurons [6,21,25,37,44]. However compelling this [13,34,35,39] has shown activation in the vertical limb of
hypothesis, the LPO has not been systematically mapped the
diagonal band
of Broca
included within
the for self-stimulation. For example, single cell recordings
boundaries of the LPO in some studies [28], the bed nucleus of the stria terminalis, the medial preoptic area
with suppression of activity in the lateral habenula, as a
Corresponding author. Tel.: 11-613-562-5800; fax: 11-613-562-
consequence of self-stimulation in the anterior VTA [13].
5147. E-mail address
: catchuottawa.ca C. Bielajew.
Cytochrome oxidase [2] and glycogen phosphorylase
0006-8993 00 – see front matter
2000 Elsevier Science B.V. All rights reserved. P I I : S 0 0 0 6 - 8 9 9 3 0 0 0 2 5 6 4 - 6
104 T
[17,22] have been similarly exploited to examine sites ranged from 0.2 to 0.92 mm posterior to bregma, 0.6 to 2.0
activated by stimulating electrodes in LH reward sites. mm lateral to the mid-sagittal suture, and 7.6 to 8.2 mm
Cytochrome oxidase activity is evident in the lateral septal ventral to the skull surface reading at bregma. The
nucleus and nucleus accumbens when short pulse durations anesthetic consisted of a combination of sodium pen-
are used to activate reward loci while a more widespread tobarbital Somnotol, 65 mg kg i.p. and xylazine Rom-
activation is observed from the use of longer pulse pun, 0.05 ml i.m.; atropine sulfate 0.05 ml s.c. was
durations [2]. In the case of glycogen phosphorylase, the administered just prior to surgery in order to minimize
diagonal band of Broca and the caudate nucleus show respiratory distress.
greater activation of reaction product on the stimulated side, and the habenula in the contralateral hemisphere in
2.2. Apparatus one study [22], and weak activation in the ipsilateral lateral
septum in another study [17]. More recently, c-fos im- Electrical stimulation was delivered via an integrated
munohistochemistry has been similarly exploited and circuit pulse generator, manufactured in house, and a
generally supports the implication of basal forebrain constant-current amplifier [24]. The current was continu-
structures such as the LPO as responsive to rewarding ously monitored on an oscilloscope by reading the voltage
MFB stimulation [1,10,20,28]. drop across a 1 kV precision resistor in series with the rat.
The purpose of this mapping study was to obtain In order to prevent polarization at the electrode tip in the
information about the distribution of reward sites in the absence of a pulse, the output was shorted to ground via a
region of the LPO and their spatial organization; the 1 kV resistor.
location of self-stimulation sites within and around the LPO was therefore documented. At sites that reliably
2.3. Screening and training supported this behavior, the trade-off relationship between
current and period was established in order to permit the Three days after surgery, the animals were trained to
characterization of the neural density and the boundaries of lever-press on a continuous reinforcement schedule for 0.5
the behaviorally relevant bundle. Estimates of the relative s trains of square wave monophasic pulses, 0.1 ms in
neural density of elements within the effective stimulation duration, using conventional operant shaping techniques. If
field, at individual sites, can be gleaned from period after two screening sessions no self-stimulation behavior
thresholds [23,45]. Due to the trade-off between current was evident, then the electrode was moved a ventral
and period the reciprocal of frequency, any change in distance of 0.08 to 0.24 mm, aided by a calibrated driver;
current must be accommodated by roughly the same the new site was evaluated the following day. This
change in period in order to maintain equivalent levels of procedure continued until either bar-pressing was estab-
behavior. The function that describes this relationship, lished or the full ventral travel of the electrode was used,
when plotted in logarithmic space, ascends to level off at based on the estimated location of the electrode tip from
high currents where the relationship between these two the Paxinos and Watson atlas [31].
parameters breaks down. Thus differences in slope across During the training phase of the study, the animals were
sites can be used to infer the density of the relevant introduced to 60 s trials preceded by five priming stimula-
substrate. The assumptions underlying this interpretation tions, separated by 1 s, set at the same values as the trial
are explained in detail in Gallistel et al. [14]. This strategy parameters; the current was 200 mA. A descending se-
has been profitably employed by others to map more quence of periods was presented, starting with a value at
posterior reward sites from the stria medullaris [8,42] and which no response occurred. Each successive period was
ventral pallidum regions [30] to the posterior mesence- 0.1 log 10 unit below the previous one. The series
phalon [9,11], as well as rostrally to study the connections continued until the bar pressing rate exceeded the criterion
between the caudate–putamen and the medial prefrontal of 35 responses min. A rate period curve was plotted for
cortex in our own laboratory [7,41]. each current and the period threshold, i.e. the period
required to support the criterion response rate was de- termined. The testing phase, during which complete trade-
2. Materials and methods off functions were generated, was initiated when stable
