- Email: [email protected]
Osmosensitive neurons in the rat's dorsal midbrain
Brain Research, 105 (1976) 105-120 © Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
105
OSMOSENSITIVE NEURONS IN THE RAT'S DORSAL MIDBRAIN
R. B. M A L M O
Allan Memorial Institute, McGill University, 1033 Pine Ave. W., Montreal, P.Q. H3A 1.41 (Canada) (Accepted August 18th, 1975)
SUMMARY
In experiment 1, multiple unit recordingS were taken simultaneously from lateral preoptic and dorsal midbrain areas during a series of intracarotid hypertonic and isotonic NaC1 injections. Subjects were 15 hooded rats (11 males and 4 ovariectomized females) under urethane anesthesia. Results showed that the neuronal reactions to a series of hypertonic NaCI injections (0.30 M, 0.45 M, 0.60 M and 0.75 M) were at least as strong in the dorsal midbrain as in the lateral preoptic area. Strength of neuronal reaction correlated with osmolarity of the NaCI solution injected. Control isotonic NaCI injections were ineffective, and the (monitored) force of injection was found not to affect the results. In experiment 2 with 15 hooded rats (9 males and 6 ovariectomized females), and two male Wistar rats under urethane anesthesia, recordings from dorsal midbrain units were made during intracarotid injections of hypertonic and isotonic NaC1 solutions. In addition, other sensory stimulations, including tail pinches, were presented. Of the 52 units studied, 39 cells (75 700) reacted to injections of hypertonic NaCI, but not to the isotonic (control) solution (Normosol-R). Again, strength of neuronal reaction correlated with osmolarity of the NaCI solution injected, and force of injections was found not to influence results, Eleven cells reacted to hypertonic NaC1 injections but not to tail pinch. This and other evidence indicated that certain dorsal midbrain cells were specifically osmosensitive, and not merely showing general 'arousal' reactions to the injections. These results indicate that, for the rat, the osmosensitive zone extends into the midbrain. The functional significance of these findings is discussed.
I NTRODUCTION
The present status of osmoreceptor theory as related to cellular dehydration thirst has recently been reviewed 9. This review highlighted the demonstrated impor-
106 tance of the lateral preoptic area as delimited in the rat by Blass and Epstein 2, and in the rabbit by Peck and Novin 17, and pointed out the importance of bringing brainrecording experiments to bear on this area of research. In the same article 9 data were presented that clearly showed that the lateral preoptic area of the rat contains cells that respond to challenge with hypertonic NaC1. On the basis of recent data, Blass I stated that the osmosens~tive zone for thirst was probably more extensive than had originally been suggested ~,17. The caudal border was placed on the zona incerta. Earlier, in extending a multiple unit exploration to areas in the midbrain, far posterior to the zona incerta, we were surprised to find strong responses to hypertonic challenge. Although the probable importance of the midbrain for thirst-related behavior is recognized 7,1°-1~, to the best of my knowledge, the literature contains no previous reports of midbrain neurons responding to the intracarotid injection of hypertonic NaCI solutions. Therefore, it seemed important to follow up our preliminary findir, gs with a systematic study. The dorsal locations of recording sites in this study resulted from the practice of moving the electrode tip in small downward steps and testing along the way until a good response to challenge was obtained. The electrode was then held at this point during a series of injections. The purpose of this study was to determine whether osmosensitive neurons could be found as far posterior as the midbrain, rather than to undertake an extensive mapping of the entire midbrain. EXPERIMENT 1:
SIMULTANEOUS MULTIPLE UNIT RECORDING FROM LATERAL PREOPT1C
AREA AND DORSAL MIDBRAIN
METHODS
Full details concerning basic methodology employed in the experiment were previously described by Malmo and MundlL
Subjects Fifteen Charles River Long-Evans hooded rats obtained from the Canadian Breeding Farm and Laboratories were used in this experiment. At time of operation the weights of the 11 males (M4-1, 3, 4, 5, 6, 7, 8, 12, 13, 15, 21) ranged from 441 to 657 g (median 471 g); the weights of the 4 ovariectomized females (M4-22, 23, 26, 27) ranged from 423 to 528 g (median 465.5 g). Procedure Anesthesia was induced with halothane and shifted to urethane 1.2 g/kg i.p. A heating device maintained body temperature (monitored continuously by a Yellow Springs telethermometer) at slightly above 37 °C. Recording electrodes were aimed in the right hemisphere 1.6 mm anterior to bregma, 2.7 mm lateral for the preoptic area placement; and 5.4 mm posterior to bregma, 1.6 mm lateral for the midbrain placement. The plane was that of the Pelle-
107 grino and Cushman atlas la. Exact depth of final recording site in each case depended on the animal's reactions to intracarotid hypertonic saline injections, as the electrodes were lowered step-wise in the two brain areas. The right common carotid artery* was exposed by neck incision, blunt incision and retraction of muscle tissue. A section of bent British 30-gauge stainless steel tubing sharpened at its tip, and fitted at the other end to No. PE 10 polyethylene tubing, was inserted into the artery for cannulation. A Beckton-Dickinson 1 ml tuberculin syringe graduated in 0.01 ml or a No. 725-LT 250 microliter Hamilton syringe (which replaced the tuberculin syringe after the tenth animal) was connected via a 28 gauge hypodermic needle to No. PE 20 polyethylene tubing, which was joined with the No. PE 10 tubing. The basic recording apparatus was the same as that described previously 9. The Diamel-coated platinum iridium electrode wires were 0.0025 in. in diameter. In this study a frequency counter of the kind described by Brown et al. a was added to the recording equipment. This counter supplied an analog voltage whose peak represented the number of single spikes counted over a half second period. An amplitude discriminator at the counter input was used to select the most approp6ate trigger point for each group of multiple units being studied at any particular time. The injection series was the same as that previously described 9, except for doubling the number of isotonic injections by repeating every Normosol-R pH 7.4. This was done in order to ensure that the second isotonic injection in each pair, which was the one examined for reaction, was free from hypertonic residue. The series comprised five 0.1 ml injections of each of the hypertonic NaCI solutions (0.30 M, 0.45 M, 0.60 M and 0.75 M) and, as previously mentioned, 10 injections of Abbott Normosol-R pH 7.4, whose electrolyte concentrations and pH closely correspond to those found in normal plasma and extracellular fluids. The 5-sec 0.1 ml injections were spaced at intervals of at least 90 sec. Beginning with the last 8 rats in this series, and throughout all unit recordings, the practice was adopted of not informing the injector concerning the concentration of the solution he was injecting, and of monitoring the force of each of the injections. The force transducer tbr monitoring injections consisted of a U-shaped brass clip, which was bent when force was applied to the end of one straight arm of the U. The clip was mounted horizontally, with the lower arm attached to a base plate. Two strain gauges, which were cemented to the inner and outer surfaces of the circular part of the U, formed two arms of a bridge, whose output was applied to a differential amplifier, whose output in turn was fed into the driver amplifier of the polygraph. Calibrations with weights showed that deflections were linearly pcoportional to the applied force in grams. Histology
At the end of the recording, electrolytic lesions (50/~A for 10 sec) were made to mark the sites of electrode tips. At the termination of the experiment, the brains were * To the best of my knowledge, investigators in this area of research have used the carotid rather than the vertebral basilar arteries for injection of hypertonic NaC1 in the rat.
":-"
"4
• "--O~
~
"
,-.
o,! s ~ -
~.
.-~
O~
I
on
.'-CO
.on
I
7(;"
I -",~;i~ !~
b
r,o
I'0
I
.on
i, ',V(3.;~
PO r'O
(
0
I
.on
co
,-. ,/'., j~..'
O~
I
-~
o
..
"~
I
~
109 Rot M 4 - 2 7 A
B
~r~2~J~:~'~'~:`~'~'~``'~'t,It~tttIIl~I¢r~I~t~tttt~IIgt~'~¢~t~t¢~'~b:~I~:;'`~t~b~ttItIt~z'~I~]'f'`~tr~'~t'eI~`'I~j'~tg~`~;~ÈI.'~¢~;~`';~I~'I:t~zt`'~.``'tÈ~t~/`~t~..~eg/I' 13
F
H
1
d I ,5,,e.4
iic~
i~O
Fig. 2. Correlations between magnitudes of multiple unit reactions (as indicated by frequency counters) and the concentrations of NaCI solution injected. Top trace in each pair (A, C; E, G, I) recording from lateral preoptic area. Bottom trace (B, D, F, H, J), from dorsal midbrain. N: Normosol-R. Amount of force applied during injection indicated by deflection beneath each pair of frequency traces (see 100 g calibration at bottom of chart).
p e r f u s e d with isotonic saline a n d followed by 10 ~ f o r m o l - s a l i n e . Brains were then r e m o v e d a n d s t o r e d in 1 0 ~ f o r m a l i n for at least 7 days. The b r a i n s were b l o c k e d at the angles used b y Pellegrino a n d C u s h m a n is. Serial frozen sections were s t a i n e d with L u x o l fast blue a n d c o u n t e r s t a i n e d with P y r o n i n Y. RESULTS Fig. 1 shows the pairs o f r e c o r d i n g sites in the f o r e b r a i n a n d m i d b r a i n for the 15 animals. Fig. 1. Representations of loci of two recording sites in each of 15 rats, indicated on frontal diagrammatic sections from the Pellegrino and Cushman atlas TM.
110 4.0
Averaged Ratings of Multiple Unit Responses N=I5 //
3.5
///
7"
3.0 Dorsal Midbrain 2.5
_~
/i
20
J //'/
1.0
0.5
o~o~
o4~ M
o~o~
oA
Concentration of NaCI Solution
Fig. 3. Graphs showing average ratings of multiple unit reactions in 15 rats to intracarotid injections of NaCI solutions ranging from isotonic through 4 degrees of hypertonicity.
Fig. 2, which was taken from the records of Rat M4-27, illustrates the correlation between degree of reaction and the concentration of NaCI injected. Note, in the tracings from the frequency counter, the progressive increase in magnitude of reaction through 0.30 M, 0.45 M, 0.60 M and 0.75 M NaC1; and also note the absence of reaction to the control solution, Normosol-R pH 7.4 (see N in Fig. 2). Fig. 2 represents a typical finding. In order to combine all the data, a rating-scale method, previously described 9, was used. The results are shown in Fig. 3 and in Table I. Multiple unit recording from the dorsal midbrain recording sites showed degrees of reaction to each of the 4 hypertonic NaCI concentrations that, on the average, were at least as great as those shown by the forebrain recording sites in the lateral preoptic area. The comparisons in Table I indicate that with the exception of the data for the 0.60 M NaC1 concentration, there were no statistically significant differences between average forebrain and midbrain reactions to challenge. From careful analysis of the force transducer data it was concluded that the data in Table I and in Fig. 3 were not associated with any systematic trends in forces used during injections. DISCUSSION
These results from multiple unit recording show a remarkable similarity between
111 TABLE
I
AVERAGED MULTIPLE UNIT RESPONSES TO HYPERTONIC SALINE INJECTIONS ( N = 15)
NaCI concentrations
Mean Rating S.D. N higher Ties Statistical significance
0.30 M
0.45 M
0.60 M
0.75 M
Forebrain
Midbrain
Forebrain
Midbrain
Forebrain
Midbrain
Forebrain
Midbrain
1.00 0.76 3
0.97 0.55 2
1.37 1.11 3
1.60 1.06 5
2.60 1.88 1
3.07 1.98 6
3.73 1.53 5
3.77 1.82 5
10 N.S.
7 N.S.
8 P < 0.05*
5 N.S.
* Based on t-test for related measures.
the reactions of forebrain and midbrain neurons to challenge with hypertonic NaC1 solutions. It is interesting to note that the only significant difference between forebrain and midbrain was observed with the 0.60 M solution and not with the more concentrated 0.75 M solution. It would appear that if the reactions of the midbrain neurons were merely 'arousal' elicited at a certain intensity threshold, that the divergence would have been even greater at 0.75 M than it was at 0.60 M; whereas in fact the curves converge at 0.75 M (see Fig. 3). The problem of 'arousal' will be considered again later. The purpose of Experiment 2 was to extend the study of osmosensitive neurons in the dorsal midbrain by means of unit recordings. The advantages of a combined approach, employing both unit and multiple unit techniques were previously discussed 9. EXPERIMENT
2:
U N I T R E C O R D I N G S FROM CELLS I N T H E DORSAL M I D B R A I N
METHODS
Following a modification of Hubel's s method, electrodes were made of 0.005 in. 25 m m length tungsten wire, electrolytically etched to a tip diameter of 1 # m or less and insulated with four coats of Epoxylite vacnish (The Epoxylite Corp., E1 Monte, California). The recording apparatus was previously described 9. All injections were monitored with the force transducer described above and the injector was not informed concerning the concentration o f the NaC1 solution being injected. Subjects
Fifteen Charles River Long-Evans hooded rats and two rats of the Wistar strain, all obtained from the Canadian Breeding F a r m and Laboratories were used in this experiment. At time of operation the weights of the 9 male hooded rats (M4-U2, 3, 4, 6, 7, 9, 10, 12, 13) ranged from 475 to 610 g (median 553 g); the weights of the
I .m
I
I
~
•
0
r,o
3ilii--Q ~
I
I
i
oo
I
I
i
po
I
113 6 female ovariectomized hooded rats (M4-U14, 15, 16, 17, 18, 19) ranged from 282 to 325 g (median 312.5 g). The weights of the two Wistar males (M4-U20, 21) were 630 g and 485 g.
Procedure When a cell was found, it was challenged with a hypertonic NaCI solution (in the range between 0.30 M and 0.75 M), followed by an isotonic (Normosol-R) control solution. With osmosensitive cells, the full range of NaC1 concentrations, previously listed for multiple unit recordings, was tried in order to observe correlations between NaC1 concentrations and varying strengths of reactions to challenges. A number of control isotonic injections (Normosol-R) were regularly given. Usually, the series of injections included the following: Normosol-R (N), NaCI concentrations: 0.75 M, 0.30 M, 0.45 M, 0.60 M, 0.75 M, 0.60 M, 0.45 M, 0.30 M; N, N. Cells were regularly tested tor reactions to visual, auditory, and cutaneous stimuli. Special care was taken in challenging cells that were responsive to any of these stimuli, in order to avoid confounding the reaction to challenge with reactions to other forms of stimulation. A flashlight was used for visual stimulation. Where there was a reaction to light, further tests of the effects of light distribution were made. For auditory stimulation toy clickers were chiefly used. Tail pinch was effected by firm pressure with forceps applied near the end of the tail. From the work of Hayward and Vincent ~ it is clearly important to determine whether there are cells that respond to hypertonic intracarotid injections and not to other sensory stimuli. Even though the animals were deeply anesthetized, careful attention was given to the question of 'arousal', because Hayward and Vincent's 6 'non-specific' osmosensitive cells responded to arousing stimuli. Therefore, the animals were carefully observed for any sign of 'arousal' following every stimulation.
Histology Histological procedure was as described for Experiment 1. Marking lesions were either 50 #A for 10 sec or 20/~A for 5 sec. Lesions were placed at critical points in each brain in order to locate each recording site. RESULTS Fig. 4 presents the histological findings. In this paper, a particular cell is identified by the rat number and the cell number for that particular rat. Each locus plotted in Fig. 4 refers to one of the 52 cells studied. Cell loci for a particular rat are connected by a vertical line, which represents the electrode track. The number at the top of each Fig. 4. Histological findings for unit recordings in Experiment 2. Rat numbers are the large ones, identifying the electrode tracks along which cells (identified by smaller numbers) are located. N = 52 cells. All cells reacted to hypertonic NaCI except the following: (animal number is given first; cell numbers follow in parentheses) 2 (1, 2, 3); 3 (5, 7); 4 (1, 3, 5, 6); 13 (1); 19 (1, 2); 20 (2). These 13 cells did not react to hypertonic NaC1.
114
B !
2
Fig. 5. Film from oscilloscope camera showing reaction of cell 2 in rat M4-UI7 to injections of 0.60 M and 0.75 M NaC1. Note the 'burster' pattern of background firing, which is interrupted by injections. Note also the long inhibitory phase in the 'biphasic' reaction to 0.75 M NaC1.
line refers to the rat number. The smaller numbers along the line below the rat number refer to the cells that were studied in that particular rat. In classifying the reaction types of the 52 cells, the terminology of Hayward and Vincent 6 will be used. In reacting to hypertonic NaC1 injections, 22 of the 52 cells consistently showed clear 'monophasic' activation, and 8 others, though less consistent could be classified as mainly 'monophasic' acceleratory. One cell was probably 'monophasic' inhibitory (wilh perhaps a small rebound). Seven cells were 'biphasic', 6 with the acceleratory phase appearing first, and one with the inhibitory phase first. Finally, there was one cell that showed reversal from inhibition to acceleration with diminution of the NaCI concentration injected. (Further details concerning this cell will be given later in connection with Fig. 7.) Thirteen cells of 52 showed no reaction to hypertonic NaC1. Percentage of cells responding to hypertonic NaCI was thus 75 %.
115 Rot M4-UI7
A liliill
DJillil i l i ' ......... i IJllli I~ i l i l i " ~ " II t l i i l i i ~---:~-~-ii m i l i ! ~ l l i l i l l i J i
lJl~l jiilm IIPlII
B
IilililliUilil,l#l:li
i,t ;~l I i/~iill i I l~li~nii I i i i i l t l i i ~ i l i i i i / 1 1
iii,illlilliliiilliliiliiiilll
c llliilllllilit
III/1111 liiillll
tl fill/if
l i ! i i l lill/i I i i i l i i i / l l i / l i
LIIi l l li.lll t[iill I l l l i l i l l
I
Fig. 6. Oscillograph tracings from pulse stretcher and integrator showing reactions of burster-tyl~ cell 2 in rat M4-U17 to injections of 0.30 M and 0.45 M NaC1 solutions, and the absence of reaction to the control isotonic solution: Normosol-R (N).
Forty cells (77~) were studied 15 min or longer. Of the 23 cells (44%) that were studied 30 min or longer, 12 cells were studied approximately 1 h or longer. Absence of any bias associated with force of injection was clear from the measurements of deflections produced by the force transducer, which monitored every injection. Fig. 5 presents a photograph from the oscilloscope camera showing reaction of cell 2 in rat M4-U17 to injections of 0.75 M and 0.60 M NaCI. The cell obviously had a 'burster' pattern of background firing, which resembles 'burster-type' cells observed by Hayward and Jennings5 in the supraoptic and internuclear zone of the unanesthetized monkey, and in the lateral preoptic area of the rat by Malmo and Mundl 9. Although it would be unwise to make too much of it at the present time, the resemblance is interesting nevertheless. In comparing the reactions to 0.75 M and to 0.60 M NaC1 (shown in Fig. 5), two points may be noted. First, the acceleratory phase of the former is stronger; and second, there is an obvious inhibitory phase in the reaction to 0.75 M, which is absent (or nearly so) in the reaction to 0.60 M. In both instances the 'burster' pattern is interrupted for a period. A related observation was noted in the recordings of cell 4 in rat M4-U19. This cell when first encountered showed a rhythmically varying background firing pattern, which persisted during and after an isotonic injection (Normosol-R) but disappeared following injection with 0.30 M NaC1. Thereafter, reactions to challenge occurred in a background devoid of rhythmical bursts. Fig. 6 presents oscillographic records, again from rat M4-U17, showing reactions to 0.45 M and 0.30 M NaCI and the absence of reaction to Normosol-R. It is
116
Rat M4-UI7 A
/
0.75 M
;'
B '11,,
--'f
0.60M
]
0,45M
]
050M
, ..............
,
/.
C
D
E
/.............. / N
L.
~
.5_sec
~
I lOOg
Fig. 7. Oscillographic tracings from pulse stretcher and integrator for cell 1 in rat M4-U17. Note in A, B and C the inhibitory effects of injecting 0.75 M, 0.60 M and 0.45 M NaC[. Observe in D the reversal of the effect of injection (with 0.30 M NaC]) from inhibitory to excitatory; and note the absence of any reaction to injection of the isotonic solution (N). This cell did not respond to light, sound, or to tail pinch. important to make the point that in this study, neuronal reactions were regularly obtained with such low concentrations as 0.30 M NaC1. Lawful relations between degree o f neuronal reaction and strength o f injected NaC1 solution, like those seen in Figs. 5 and 6, were regularly observed. Careful ratings o f all reactions o f the 22 cells classified as 'clearly m o n o p h a s i c ' yielded the following mean values, and standard deviations (in parentheses): N o r m o s o l - R , 0; 0.30 M, 1.53 (2.20); 0.45 M, 3.28 (3.02); 0.60 M, 4.36 (3.09); 0.75 M, 5.95 (3.08). The rating m e t h o d employed was similar to the one described for multiple unit recordings 9.
117 Fifteen of the 22 'monophasic' cells had response patterns that allowed latency measurements to be made with confidence. Taking, for each cell, the modal latency reaction to 0.60 M and 0.75 M NaC1, the mean for the 15 cells was 2.90 sec with a standard deviation of 0.63. There was considerable variation in the background firing rates of these cells, from 25 Hz or more to cells showing large intervals of no background firing, and with prompt strong reactions to relatively weak hypertonic challenges. An example of a cell with complex reactions to a graded series of hypertonic injections is presented in Fig. 7. The 0.75 M NaC1 injection had a strong inhibitory effect on the cell, as did the 0.60 M NaC1 injection, although the latency of the latter was longer than that of the former. The inhibitory response of this cell to the injection of 0.45 M NaC1 was significantly weaker than it was to 0.60 M and 0.75 M NaCl. With the injection of 0.30 M NaC1 there was a reversal from inhibition to activation (see D in Fig. 7). As usual, there was no response to the injection of the isotonic Normosol-R. Another cell (no. 1 in rat M4-UI5) reacted to injection of 0.75 M NaC1 in a manner similar to the reaction to 0.75 M NaCl shown in Fig. 7. However, (in cell l, rat M4-U15) the 0.60 M injection produced an initial acceleration followed by inhibition (i.e. the reaction became 'biphasic'). This biphasic reaction continued to be evident in going from 0.60 M to 0.45 M to 0.30 M, with the acceleratory phase growing larger and the inhibitory phase becoming smaller. Eleven cells that reacted to injections of hypertonic NaCl did not react to tail pinch. Five of these cells (including the one whose recordings are shown in Fig. 7) also failed to react to either auditory or visual stimuli. Because of losing the cell, tail pinch could not be tried with 10 cells that responded positively to hypertonic NaC1. In all, 38 ~ of the cells that responded to hypertonic NaCI failed to respond to tail pinch. O f the 39 cells with reactions to hypertonic NaC1, definite response to light was observed in 11 cells, and to sound in 8 cells, with none of these 19 cells responding to both light and sound. Although the number of observations was small and the brain areas explored rather limited, there appeared to be a definite tendency for responses to light to be found in sites more dorsally placed than those for responses to sound; which is in line with the findings in the extensive study by Groves et al. 4. The cells that were responsive to hypertonic NaCl were distributed over the brain area studied; and with the data at hand (see Fig. 4) there appeared to be no significant relation between cell locus and sensitivity of the cell to hypertonic NaC1. DISCUSSION
The results from the multiple unit and unit recordings have clearly demonstrated that the dorsal midbrain contains cells that react to intracarotid injections of hypertonic NaC1. Furthermore, their reactions closely resemble those of neurons in the lateral preoptic area, which is regarded as a probable site of osmoreceptors ~,17. To the best of my knowledge, this is the first study in which simultaneous multiple unit
118 recordings were taken from the lateral preoptic area and from the dorsal midbrain during intracarotid injections of hypertonic NaC1 solutions. Interpretation of these results, however, must be carefully considered. First of all, were all of the reactions of the midbrain cells merely general 'arousal' reactions to the injection of 0.1 ml of hypertonic NaCI? (The effect of the force of injection was ruled out by control injections of isotonic saline, and by the data from the force transducer.) There appear to be valid reasons for concluding that the general 'arousal' explanation is improbable. First, the anesthetized animals, who were carefully observed during injections, showed no signs of 'arousal'. Second, eleven ceils that were reactive to hypertonic NaCl failed to react to firm tail pinch, which certainly was more likely to produce an 'arousal' reaction than the slow injection of 0.1 ml of weakly hypertonic saline. The strengths of hypertonic NaCI solutions that were effective in activating neuronal responses in the present study were of the same order as those used by Hayward and Vincent 6 in their investigation of osmosensitive neurons in the supraoptic nucleus of unanesthetized monkeys. This observation is important in relation to Mogenson's 11 point that physiologically meaningful osmosensitivity is reflected in a low threshold to osmotic stimulation. As previously stated, 38 ~ of the cells that responded to hypertonic NaC1 tailed to respond to tail pinch, which in the awake animal is a highly noxious stimulus, and in a lightly anesthetized animal causes movement. Finally, it seems impossible to account for the reactions of a cell like the one illustrated in Fig. 7 in terms of general 'arousal'. In the first place, this cell appeared 'specific' in the terminology of Hayward and Vincent 6. It responded to hypertonic NaCI but did not respond to tail pinch, auditory or visual stimuli. Secondly, although there was a lawful relation between osmolarity of injected NaCI and reaction, the relation was certainly not a simple progression of increasing activation in going from weaker to stronger concentrations, as might be expected if the reactions were merely due to 'arousal' produced by intracarotid NaCI injections. Hayward and Vincent's 6 criteria for defining a cell as 'osmosensitive' are as follows: 'a prompt response to intracarotid hypertonic solutions during or immediately after injection with a return to base line firing within 30-60 sec'. The cells studied in the present experiments conform to these criteria, and it therefore seems valid to apply the term 'osmosensitive' to these cells, including those in the midbrain. To those midbrain cells that responded to hypertonic NaCI, but not to other forms of sensory stimulation, the term 'specific' osmosensitive cell appears applicable. It should be pointed out, however, that extension of this research in the future should include a systematic study of other compounds (especially dextrose or sucrose). A more difficult problem concerns the functional significance of osmosensitive neurons in the midbrain. According to current theories as summarized recently by Mogenson 10,11 and by Mogenson and Huang 12, the most probable interpretation of midbrain osmosensitive cells is that they are being synaptically driven by neuronal activity from elsewhere in the brain. The effects of electrical stimulation of the lateral preoptic area, lateral hypothalamus, hippocampus, amygdala and olfactory bulb on unit activity of the lateral habenular nucleus in the rat were demonstrated by Mok
119 and Mogenson14,15; and they 15 suggested, on the basis of their findings, that the lateral habenular nucleus serves as an integrating area for limbic forebrain inputs. Furthermore, M o k and Mogenson 16 found that midbrain neurons (some in dorsal midbrain) were in turn influenced by stimulation of the lateral habenular nucleus. As Mogenson 10 points out: 'the critical output from the hypothalamic integrative sites appears to be to the midbrain (p. 123)'. On the other hand, the present finding of osmosensitive neurons in the midbrain does raise the question of whether there may be osmoreceptors in this dorsal midbrain area. While, from the point of view of classical neurophysiology, this seems improbable, this question nonetheless seems to deserve consideration in further research. Recent evidence led Blass to suggest that the caudal border of the osmosensitive zone for cellular thirst as originally conceived2,17 was too far anterior, and should be moved back to the zona incerta. Present data suggest the possibility that this border should be placed even more posteriorly, in the midbrain. To the best of my knowledge, the kind of thorough application of the intracranial injection technique, exemplified in the recent work of Blass 1 on the basal forebrain, has not as yet been extended to a study of the dorsal midbrain. This technique should be useful in determining whether midbrain osmosensitive neurons appear to be osmoreceptive. There are some recent findings on the possible role of the midbrain in extracellular thirst. According to Sharpe and Swanson 19, the mesencephalic central gray (but not the midbrain reticular formation) in the monkey appears to be sensitive to angiotensin II. These investigators inferred sensitivity to angiotensin II on the basis of the amount of water the monkey drank following injection into a specific brain site. ACKNOWLEDGEMENTS The author acknowledges the substantial technical assistance of W. J. Mundl, and also the help throughout the project and the critical reading of the manuscript by H. P. Malmo. This research was supported in part by grants from the National Institute of Mental Health, United States Public Health Service, and from the National Research Council of Canada. I wish to express my thanks to Abbott Laboratories, Limited, Montreal, P.Q., for supplying Normosol-R p H 7.4.
REFERENCES 1 BLASS,E. M., Evidence for basal forebrain thirst osmoreceptors in rat, Brain Research, 82 (1974) 69-76. 2 BLASS,E. M., AND EPSTEIN,A. N., A lateral preoptic osmosensitive zone for thirst in the rat, J. comp. physioL PsychoL, 76 (1971) 378-394. 3 BROWN, K. A., WEBER,D. S., AND BUCHWALD,J. S., Chronic recording and quantification of subcortical single and multiple units. In M. I. PmLLIPS(Ed.), Brain Unit Activity During Behavior, Thomas, Springfield, I11., 1973, pp. 41-52. 4 GROVES, P. M., MILLER, S.W., PARKER, M.V., AND REBEC, G. V., Organization by sensory modality in the reticular formation of the rat, Brain Research, 54 (1973) 207-224.
120 5 HAYWARD, J. N., AND JENNINGS, D. P., Activity of magnoceUular neuroendocrine cells in the hypothalamus of unanaesthetized monkeys. I. Functional cell types and their anatomical distribution in the supraoptic nucleus and the internuclear zone, J. Physiol. (Lond.), 232 (1973) 515-543. 6 HAYWARD, J. N., AND VINCENT, J. D., Osmosensitive single neurones in the hypothalamus of unanaesthetized monkeys, J. PhysioL (Lond.), 210 (1970) 947-972. 7 HUANG, Y. H., AND MOGENSON, G. J., Neural pathways mediating drinking and feeding in rats, Exp. Neurol., 37 (1972) 269-286. 8 I'-IUBEE, D. H., Tungsten microelectrode for recording from single units, Science, 125 (1957) 549-550. 9 MALMO, R. B., AND MUNDL, W. J., Osmosensitive neurons in the rat's preoptic area: mediallateral comparison, J. comp. physiol. Psychol., 88 (1975) 161-175. l0 MOt3ENSON, G. J., Hypothalamic limbic mechanisms in the control of water intake. In A. N. EPSTEIN, H. R. KISSILEFF AND E. STELLAR (Eds.), The Neuropsychology of Thirst: New Findings and Advances in Concepts, Winston, Washington, D.C., 1973, pp. 119-142. 11 MOGENSON,G. J., Electrophysiological studies of the mechanisms that initiate ingestive behaviours with special emphasis on water intake. In G. J. MOGENSON AND F. R. CALARESU (Eds.), Neural Integration of Physiological Mechanisms and Behaviour: J. A. F. Stevenson Memorial Volume University of Toronto Press, Toronto, 1975, pp. 248-266. 12 MOGENSON,G. J., AND HUANG, Y. H., The neurobiology of motivated behavior. In G. A. KERKUT AND J. W. PHILLIS (Eds.), Progress in Neurobiology. Vol. 1, Pergamon, Oxford, 1973, pp. 53-83. 13 MOGENSON, G. J., AND HUANG, Y. H., Facilitation and suppression of centrally elicited feeding and drinking by electrical stimulation of the midbrain, Canad. J. Psychol., 28 (1974) 252-259. 14 MoK, A. C. S., AND MOGENSON, G. J., Effect of electrical stimulation of the septum and the lateral preoptic area on unit activity of the lateral habenular nucleus in the rat, Brain Research, 43 (1972) 361-372. 15 MOK, A. C. S., AND MOGENSON, G. J., Effects of electrical stimulation of the lateral hypothalamus, hippocampus, amygdala and olfactory bulb on unit activity of the lateral habenular nucleus in the rat, Brain Research, 77 (1974) 417-429. 16 MOK, A. C. S., AND MOGENSON, G. J., Effects of electrical stimulation of the lateral habenular nucleus and lateral hypothalamus on unit activity in the upper brain stem, Brain Research, 78 (1974) 425435. 17 PECK,J. W., AND NOVIN, D., Evidence that osmoreceptors mediating drinking in rabbits are in the lateral preoptic area, J. comp. physiol. Psychok, 74 (1971) 134-147. 18 PELLEGRINO, L. J., AND CUSHMAN, A. J., ,4 Stereotaxic Atlas of the Rat Brain, Appleton-CenturyCrofts, New York, 1967. 19 SHARPE,L. G., AND SWANSON,L. W., Drinking induced by injections of angiotensin into forebrain and mid-brain sites of the monkey, J. Physiol. (Lond.), 239 (1974) 595-622.
105
OSMOSENSITIVE NEURONS IN THE RAT'S DORSAL MIDBRAIN
R. B. M A L M O
Allan Memorial Institute, McGill University, 1033 Pine Ave. W., Montreal, P.Q. H3A 1.41 (Canada) (Accepted August 18th, 1975)
SUMMARY
In experiment 1, multiple unit recordingS were taken simultaneously from lateral preoptic and dorsal midbrain areas during a series of intracarotid hypertonic and isotonic NaC1 injections. Subjects were 15 hooded rats (11 males and 4 ovariectomized females) under urethane anesthesia. Results showed that the neuronal reactions to a series of hypertonic NaCI injections (0.30 M, 0.45 M, 0.60 M and 0.75 M) were at least as strong in the dorsal midbrain as in the lateral preoptic area. Strength of neuronal reaction correlated with osmolarity of the NaCI solution injected. Control isotonic NaCI injections were ineffective, and the (monitored) force of injection was found not to affect the results. In experiment 2 with 15 hooded rats (9 males and 6 ovariectomized females), and two male Wistar rats under urethane anesthesia, recordings from dorsal midbrain units were made during intracarotid injections of hypertonic and isotonic NaC1 solutions. In addition, other sensory stimulations, including tail pinches, were presented. Of the 52 units studied, 39 cells (75 700) reacted to injections of hypertonic NaCI, but not to the isotonic (control) solution (Normosol-R). Again, strength of neuronal reaction correlated with osmolarity of the NaCI solution injected, and force of injections was found not to influence results, Eleven cells reacted to hypertonic NaC1 injections but not to tail pinch. This and other evidence indicated that certain dorsal midbrain cells were specifically osmosensitive, and not merely showing general 'arousal' reactions to the injections. These results indicate that, for the rat, the osmosensitive zone extends into the midbrain. The functional significance of these findings is discussed.
I NTRODUCTION
The present status of osmoreceptor theory as related to cellular dehydration thirst has recently been reviewed 9. This review highlighted the demonstrated impor-
106 tance of the lateral preoptic area as delimited in the rat by Blass and Epstein 2, and in the rabbit by Peck and Novin 17, and pointed out the importance of bringing brainrecording experiments to bear on this area of research. In the same article 9 data were presented that clearly showed that the lateral preoptic area of the rat contains cells that respond to challenge with hypertonic NaC1. On the basis of recent data, Blass I stated that the osmosens~tive zone for thirst was probably more extensive than had originally been suggested ~,17. The caudal border was placed on the zona incerta. Earlier, in extending a multiple unit exploration to areas in the midbrain, far posterior to the zona incerta, we were surprised to find strong responses to hypertonic challenge. Although the probable importance of the midbrain for thirst-related behavior is recognized 7,1°-1~, to the best of my knowledge, the literature contains no previous reports of midbrain neurons responding to the intracarotid injection of hypertonic NaCI solutions. Therefore, it seemed important to follow up our preliminary findir, gs with a systematic study. The dorsal locations of recording sites in this study resulted from the practice of moving the electrode tip in small downward steps and testing along the way until a good response to challenge was obtained. The electrode was then held at this point during a series of injections. The purpose of this study was to determine whether osmosensitive neurons could be found as far posterior as the midbrain, rather than to undertake an extensive mapping of the entire midbrain. EXPERIMENT 1:
SIMULTANEOUS MULTIPLE UNIT RECORDING FROM LATERAL PREOPT1C
AREA AND DORSAL MIDBRAIN
METHODS
Full details concerning basic methodology employed in the experiment were previously described by Malmo and MundlL
Subjects Fifteen Charles River Long-Evans hooded rats obtained from the Canadian Breeding Farm and Laboratories were used in this experiment. At time of operation the weights of the 11 males (M4-1, 3, 4, 5, 6, 7, 8, 12, 13, 15, 21) ranged from 441 to 657 g (median 471 g); the weights of the 4 ovariectomized females (M4-22, 23, 26, 27) ranged from 423 to 528 g (median 465.5 g). Procedure Anesthesia was induced with halothane and shifted to urethane 1.2 g/kg i.p. A heating device maintained body temperature (monitored continuously by a Yellow Springs telethermometer) at slightly above 37 °C. Recording electrodes were aimed in the right hemisphere 1.6 mm anterior to bregma, 2.7 mm lateral for the preoptic area placement; and 5.4 mm posterior to bregma, 1.6 mm lateral for the midbrain placement. The plane was that of the Pelle-
107 grino and Cushman atlas la. Exact depth of final recording site in each case depended on the animal's reactions to intracarotid hypertonic saline injections, as the electrodes were lowered step-wise in the two brain areas. The right common carotid artery* was exposed by neck incision, blunt incision and retraction of muscle tissue. A section of bent British 30-gauge stainless steel tubing sharpened at its tip, and fitted at the other end to No. PE 10 polyethylene tubing, was inserted into the artery for cannulation. A Beckton-Dickinson 1 ml tuberculin syringe graduated in 0.01 ml or a No. 725-LT 250 microliter Hamilton syringe (which replaced the tuberculin syringe after the tenth animal) was connected via a 28 gauge hypodermic needle to No. PE 20 polyethylene tubing, which was joined with the No. PE 10 tubing. The basic recording apparatus was the same as that described previously 9. The Diamel-coated platinum iridium electrode wires were 0.0025 in. in diameter. In this study a frequency counter of the kind described by Brown et al. a was added to the recording equipment. This counter supplied an analog voltage whose peak represented the number of single spikes counted over a half second period. An amplitude discriminator at the counter input was used to select the most approp6ate trigger point for each group of multiple units being studied at any particular time. The injection series was the same as that previously described 9, except for doubling the number of isotonic injections by repeating every Normosol-R pH 7.4. This was done in order to ensure that the second isotonic injection in each pair, which was the one examined for reaction, was free from hypertonic residue. The series comprised five 0.1 ml injections of each of the hypertonic NaCI solutions (0.30 M, 0.45 M, 0.60 M and 0.75 M) and, as previously mentioned, 10 injections of Abbott Normosol-R pH 7.4, whose electrolyte concentrations and pH closely correspond to those found in normal plasma and extracellular fluids. The 5-sec 0.1 ml injections were spaced at intervals of at least 90 sec. Beginning with the last 8 rats in this series, and throughout all unit recordings, the practice was adopted of not informing the injector concerning the concentration of the solution he was injecting, and of monitoring the force of each of the injections. The force transducer tbr monitoring injections consisted of a U-shaped brass clip, which was bent when force was applied to the end of one straight arm of the U. The clip was mounted horizontally, with the lower arm attached to a base plate. Two strain gauges, which were cemented to the inner and outer surfaces of the circular part of the U, formed two arms of a bridge, whose output was applied to a differential amplifier, whose output in turn was fed into the driver amplifier of the polygraph. Calibrations with weights showed that deflections were linearly pcoportional to the applied force in grams. Histology
At the end of the recording, electrolytic lesions (50/~A for 10 sec) were made to mark the sites of electrode tips. At the termination of the experiment, the brains were * To the best of my knowledge, investigators in this area of research have used the carotid rather than the vertebral basilar arteries for injection of hypertonic NaC1 in the rat.
":-"
"4
• "--O~
~
"
,-.
o,! s ~ -
~.
.-~
O~
I
on
.'-CO
.on
I
7(;"
I -",~;i~ !~
b
r,o
I'0
I
.on
i, ',V(3.;~
PO r'O
(
0
I
.on
co
,-. ,/'., j~..'
O~
I
-~
o
..
"~
I
~
109 Rot M 4 - 2 7 A
B
~r~2~J~:~'~'~:`~'~'~``'~'t,It~tttIIl~I¢r~I~t~tttt~IIgt~'~¢~t~t¢~'~b:~I~:;'`~t~b~ttItIt~z'~I~]'f'`~tr~'~t'eI~`'I~j'~tg~`~;~ÈI.'~¢~;~`';~I~'I:t~zt`'~.``'tÈ~t~/`~t~..~eg/I' 13
F
H
1
d I ,5,,e.4
iic~
i~O
Fig. 2. Correlations between magnitudes of multiple unit reactions (as indicated by frequency counters) and the concentrations of NaCI solution injected. Top trace in each pair (A, C; E, G, I) recording from lateral preoptic area. Bottom trace (B, D, F, H, J), from dorsal midbrain. N: Normosol-R. Amount of force applied during injection indicated by deflection beneath each pair of frequency traces (see 100 g calibration at bottom of chart).
p e r f u s e d with isotonic saline a n d followed by 10 ~ f o r m o l - s a l i n e . Brains were then r e m o v e d a n d s t o r e d in 1 0 ~ f o r m a l i n for at least 7 days. The b r a i n s were b l o c k e d at the angles used b y Pellegrino a n d C u s h m a n is. Serial frozen sections were s t a i n e d with L u x o l fast blue a n d c o u n t e r s t a i n e d with P y r o n i n Y. RESULTS Fig. 1 shows the pairs o f r e c o r d i n g sites in the f o r e b r a i n a n d m i d b r a i n for the 15 animals. Fig. 1. Representations of loci of two recording sites in each of 15 rats, indicated on frontal diagrammatic sections from the Pellegrino and Cushman atlas TM.
110 4.0
Averaged Ratings of Multiple Unit Responses N=I5 //
3.5
///
7"
3.0 Dorsal Midbrain 2.5
_~
/i
20
J //'/
1.0
0.5
o~o~
o4~ M
o~o~
oA
Concentration of NaCI Solution
Fig. 3. Graphs showing average ratings of multiple unit reactions in 15 rats to intracarotid injections of NaCI solutions ranging from isotonic through 4 degrees of hypertonicity.
Fig. 2, which was taken from the records of Rat M4-27, illustrates the correlation between degree of reaction and the concentration of NaCI injected. Note, in the tracings from the frequency counter, the progressive increase in magnitude of reaction through 0.30 M, 0.45 M, 0.60 M and 0.75 M NaC1; and also note the absence of reaction to the control solution, Normosol-R pH 7.4 (see N in Fig. 2). Fig. 2 represents a typical finding. In order to combine all the data, a rating-scale method, previously described 9, was used. The results are shown in Fig. 3 and in Table I. Multiple unit recording from the dorsal midbrain recording sites showed degrees of reaction to each of the 4 hypertonic NaCI concentrations that, on the average, were at least as great as those shown by the forebrain recording sites in the lateral preoptic area. The comparisons in Table I indicate that with the exception of the data for the 0.60 M NaC1 concentration, there were no statistically significant differences between average forebrain and midbrain reactions to challenge. From careful analysis of the force transducer data it was concluded that the data in Table I and in Fig. 3 were not associated with any systematic trends in forces used during injections. DISCUSSION
These results from multiple unit recording show a remarkable similarity between
111 TABLE
I
AVERAGED MULTIPLE UNIT RESPONSES TO HYPERTONIC SALINE INJECTIONS ( N = 15)
NaCI concentrations
Mean Rating S.D. N higher Ties Statistical significance
0.30 M
0.45 M
0.60 M
0.75 M
Forebrain
Midbrain
Forebrain
Midbrain
Forebrain
Midbrain
Forebrain
Midbrain
1.00 0.76 3
0.97 0.55 2
1.37 1.11 3
1.60 1.06 5
2.60 1.88 1
3.07 1.98 6
3.73 1.53 5
3.77 1.82 5
10 N.S.
7 N.S.
8 P < 0.05*
5 N.S.
* Based on t-test for related measures.
the reactions of forebrain and midbrain neurons to challenge with hypertonic NaC1 solutions. It is interesting to note that the only significant difference between forebrain and midbrain was observed with the 0.60 M solution and not with the more concentrated 0.75 M solution. It would appear that if the reactions of the midbrain neurons were merely 'arousal' elicited at a certain intensity threshold, that the divergence would have been even greater at 0.75 M than it was at 0.60 M; whereas in fact the curves converge at 0.75 M (see Fig. 3). The problem of 'arousal' will be considered again later. The purpose of Experiment 2 was to extend the study of osmosensitive neurons in the dorsal midbrain by means of unit recordings. The advantages of a combined approach, employing both unit and multiple unit techniques were previously discussed 9. EXPERIMENT
2:
U N I T R E C O R D I N G S FROM CELLS I N T H E DORSAL M I D B R A I N
METHODS
Following a modification of Hubel's s method, electrodes were made of 0.005 in. 25 m m length tungsten wire, electrolytically etched to a tip diameter of 1 # m or less and insulated with four coats of Epoxylite vacnish (The Epoxylite Corp., E1 Monte, California). The recording apparatus was previously described 9. All injections were monitored with the force transducer described above and the injector was not informed concerning the concentration o f the NaC1 solution being injected. Subjects
Fifteen Charles River Long-Evans hooded rats and two rats of the Wistar strain, all obtained from the Canadian Breeding F a r m and Laboratories were used in this experiment. At time of operation the weights of the 9 male hooded rats (M4-U2, 3, 4, 6, 7, 9, 10, 12, 13) ranged from 475 to 610 g (median 553 g); the weights of the
I .m
I
I
~
•
0
r,o
3ilii--Q ~
I
I
i
oo
I
I
i
po
I
113 6 female ovariectomized hooded rats (M4-U14, 15, 16, 17, 18, 19) ranged from 282 to 325 g (median 312.5 g). The weights of the two Wistar males (M4-U20, 21) were 630 g and 485 g.
Procedure When a cell was found, it was challenged with a hypertonic NaCI solution (in the range between 0.30 M and 0.75 M), followed by an isotonic (Normosol-R) control solution. With osmosensitive cells, the full range of NaC1 concentrations, previously listed for multiple unit recordings, was tried in order to observe correlations between NaC1 concentrations and varying strengths of reactions to challenges. A number of control isotonic injections (Normosol-R) were regularly given. Usually, the series of injections included the following: Normosol-R (N), NaCI concentrations: 0.75 M, 0.30 M, 0.45 M, 0.60 M, 0.75 M, 0.60 M, 0.45 M, 0.30 M; N, N. Cells were regularly tested tor reactions to visual, auditory, and cutaneous stimuli. Special care was taken in challenging cells that were responsive to any of these stimuli, in order to avoid confounding the reaction to challenge with reactions to other forms of stimulation. A flashlight was used for visual stimulation. Where there was a reaction to light, further tests of the effects of light distribution were made. For auditory stimulation toy clickers were chiefly used. Tail pinch was effected by firm pressure with forceps applied near the end of the tail. From the work of Hayward and Vincent ~ it is clearly important to determine whether there are cells that respond to hypertonic intracarotid injections and not to other sensory stimuli. Even though the animals were deeply anesthetized, careful attention was given to the question of 'arousal', because Hayward and Vincent's 6 'non-specific' osmosensitive cells responded to arousing stimuli. Therefore, the animals were carefully observed for any sign of 'arousal' following every stimulation.
Histology Histological procedure was as described for Experiment 1. Marking lesions were either 50 #A for 10 sec or 20/~A for 5 sec. Lesions were placed at critical points in each brain in order to locate each recording site. RESULTS Fig. 4 presents the histological findings. In this paper, a particular cell is identified by the rat number and the cell number for that particular rat. Each locus plotted in Fig. 4 refers to one of the 52 cells studied. Cell loci for a particular rat are connected by a vertical line, which represents the electrode track. The number at the top of each Fig. 4. Histological findings for unit recordings in Experiment 2. Rat numbers are the large ones, identifying the electrode tracks along which cells (identified by smaller numbers) are located. N = 52 cells. All cells reacted to hypertonic NaCI except the following: (animal number is given first; cell numbers follow in parentheses) 2 (1, 2, 3); 3 (5, 7); 4 (1, 3, 5, 6); 13 (1); 19 (1, 2); 20 (2). These 13 cells did not react to hypertonic NaC1.
114
B !
2
Fig. 5. Film from oscilloscope camera showing reaction of cell 2 in rat M4-UI7 to injections of 0.60 M and 0.75 M NaC1. Note the 'burster' pattern of background firing, which is interrupted by injections. Note also the long inhibitory phase in the 'biphasic' reaction to 0.75 M NaC1.
line refers to the rat number. The smaller numbers along the line below the rat number refer to the cells that were studied in that particular rat. In classifying the reaction types of the 52 cells, the terminology of Hayward and Vincent 6 will be used. In reacting to hypertonic NaC1 injections, 22 of the 52 cells consistently showed clear 'monophasic' activation, and 8 others, though less consistent could be classified as mainly 'monophasic' acceleratory. One cell was probably 'monophasic' inhibitory (wilh perhaps a small rebound). Seven cells were 'biphasic', 6 with the acceleratory phase appearing first, and one with the inhibitory phase first. Finally, there was one cell that showed reversal from inhibition to acceleration with diminution of the NaCI concentration injected. (Further details concerning this cell will be given later in connection with Fig. 7.) Thirteen cells of 52 showed no reaction to hypertonic NaC1. Percentage of cells responding to hypertonic NaCI was thus 75 %.
115 Rot M4-UI7
A liliill
DJillil i l i ' ......... i IJllli I~ i l i l i " ~ " II t l i i l i i ~---:~-~-ii m i l i ! ~ l l i l i l l i J i
lJl~l jiilm IIPlII
B
IilililliUilil,l#l:li
i,t ;~l I i/~iill i I l~li~nii I i i i i l t l i i ~ i l i i i i / 1 1
iii,illlilliliiilliliiliiiilll
c llliilllllilit
III/1111 liiillll
tl fill/if
l i ! i i l lill/i I i i i l i i i / l l i / l i
LIIi l l li.lll t[iill I l l l i l i l l
I
Fig. 6. Oscillograph tracings from pulse stretcher and integrator showing reactions of burster-tyl~ cell 2 in rat M4-U17 to injections of 0.30 M and 0.45 M NaC1 solutions, and the absence of reaction to the control isotonic solution: Normosol-R (N).
Forty cells (77~) were studied 15 min or longer. Of the 23 cells (44%) that were studied 30 min or longer, 12 cells were studied approximately 1 h or longer. Absence of any bias associated with force of injection was clear from the measurements of deflections produced by the force transducer, which monitored every injection. Fig. 5 presents a photograph from the oscilloscope camera showing reaction of cell 2 in rat M4-U17 to injections of 0.75 M and 0.60 M NaCI. The cell obviously had a 'burster' pattern of background firing, which resembles 'burster-type' cells observed by Hayward and Jennings5 in the supraoptic and internuclear zone of the unanesthetized monkey, and in the lateral preoptic area of the rat by Malmo and Mundl 9. Although it would be unwise to make too much of it at the present time, the resemblance is interesting nevertheless. In comparing the reactions to 0.75 M and to 0.60 M NaC1 (shown in Fig. 5), two points may be noted. First, the acceleratory phase of the former is stronger; and second, there is an obvious inhibitory phase in the reaction to 0.75 M, which is absent (or nearly so) in the reaction to 0.60 M. In both instances the 'burster' pattern is interrupted for a period. A related observation was noted in the recordings of cell 4 in rat M4-U19. This cell when first encountered showed a rhythmically varying background firing pattern, which persisted during and after an isotonic injection (Normosol-R) but disappeared following injection with 0.30 M NaC1. Thereafter, reactions to challenge occurred in a background devoid of rhythmical bursts. Fig. 6 presents oscillographic records, again from rat M4-U17, showing reactions to 0.45 M and 0.30 M NaCI and the absence of reaction to Normosol-R. It is
116
Rat M4-UI7 A
/
0.75 M
;'
B '11,,
--'f
0.60M
]
0,45M
]
050M
, ..............
,
/.
C
D
E
/.............. / N
L.
~
.5_sec
~
I lOOg
Fig. 7. Oscillographic tracings from pulse stretcher and integrator for cell 1 in rat M4-U17. Note in A, B and C the inhibitory effects of injecting 0.75 M, 0.60 M and 0.45 M NaC[. Observe in D the reversal of the effect of injection (with 0.30 M NaC]) from inhibitory to excitatory; and note the absence of any reaction to injection of the isotonic solution (N). This cell did not respond to light, sound, or to tail pinch. important to make the point that in this study, neuronal reactions were regularly obtained with such low concentrations as 0.30 M NaC1. Lawful relations between degree o f neuronal reaction and strength o f injected NaC1 solution, like those seen in Figs. 5 and 6, were regularly observed. Careful ratings o f all reactions o f the 22 cells classified as 'clearly m o n o p h a s i c ' yielded the following mean values, and standard deviations (in parentheses): N o r m o s o l - R , 0; 0.30 M, 1.53 (2.20); 0.45 M, 3.28 (3.02); 0.60 M, 4.36 (3.09); 0.75 M, 5.95 (3.08). The rating m e t h o d employed was similar to the one described for multiple unit recordings 9.
117 Fifteen of the 22 'monophasic' cells had response patterns that allowed latency measurements to be made with confidence. Taking, for each cell, the modal latency reaction to 0.60 M and 0.75 M NaC1, the mean for the 15 cells was 2.90 sec with a standard deviation of 0.63. There was considerable variation in the background firing rates of these cells, from 25 Hz or more to cells showing large intervals of no background firing, and with prompt strong reactions to relatively weak hypertonic challenges. An example of a cell with complex reactions to a graded series of hypertonic injections is presented in Fig. 7. The 0.75 M NaC1 injection had a strong inhibitory effect on the cell, as did the 0.60 M NaC1 injection, although the latency of the latter was longer than that of the former. The inhibitory response of this cell to the injection of 0.45 M NaC1 was significantly weaker than it was to 0.60 M and 0.75 M NaCl. With the injection of 0.30 M NaC1 there was a reversal from inhibition to activation (see D in Fig. 7). As usual, there was no response to the injection of the isotonic Normosol-R. Another cell (no. 1 in rat M4-UI5) reacted to injection of 0.75 M NaC1 in a manner similar to the reaction to 0.75 M NaCl shown in Fig. 7. However, (in cell l, rat M4-U15) the 0.60 M injection produced an initial acceleration followed by inhibition (i.e. the reaction became 'biphasic'). This biphasic reaction continued to be evident in going from 0.60 M to 0.45 M to 0.30 M, with the acceleratory phase growing larger and the inhibitory phase becoming smaller. Eleven cells that reacted to injections of hypertonic NaCl did not react to tail pinch. Five of these cells (including the one whose recordings are shown in Fig. 7) also failed to react to either auditory or visual stimuli. Because of losing the cell, tail pinch could not be tried with 10 cells that responded positively to hypertonic NaC1. In all, 38 ~ of the cells that responded to hypertonic NaCI failed to respond to tail pinch. O f the 39 cells with reactions to hypertonic NaC1, definite response to light was observed in 11 cells, and to sound in 8 cells, with none of these 19 cells responding to both light and sound. Although the number of observations was small and the brain areas explored rather limited, there appeared to be a definite tendency for responses to light to be found in sites more dorsally placed than those for responses to sound; which is in line with the findings in the extensive study by Groves et al. 4. The cells that were responsive to hypertonic NaCl were distributed over the brain area studied; and with the data at hand (see Fig. 4) there appeared to be no significant relation between cell locus and sensitivity of the cell to hypertonic NaC1. DISCUSSION
The results from the multiple unit and unit recordings have clearly demonstrated that the dorsal midbrain contains cells that react to intracarotid injections of hypertonic NaC1. Furthermore, their reactions closely resemble those of neurons in the lateral preoptic area, which is regarded as a probable site of osmoreceptors ~,17. To the best of my knowledge, this is the first study in which simultaneous multiple unit
118 recordings were taken from the lateral preoptic area and from the dorsal midbrain during intracarotid injections of hypertonic NaC1 solutions. Interpretation of these results, however, must be carefully considered. First of all, were all of the reactions of the midbrain cells merely general 'arousal' reactions to the injection of 0.1 ml of hypertonic NaCI? (The effect of the force of injection was ruled out by control injections of isotonic saline, and by the data from the force transducer.) There appear to be valid reasons for concluding that the general 'arousal' explanation is improbable. First, the anesthetized animals, who were carefully observed during injections, showed no signs of 'arousal'. Second, eleven ceils that were reactive to hypertonic NaCl failed to react to firm tail pinch, which certainly was more likely to produce an 'arousal' reaction than the slow injection of 0.1 ml of weakly hypertonic saline. The strengths of hypertonic NaCI solutions that were effective in activating neuronal responses in the present study were of the same order as those used by Hayward and Vincent 6 in their investigation of osmosensitive neurons in the supraoptic nucleus of unanesthetized monkeys. This observation is important in relation to Mogenson's 11 point that physiologically meaningful osmosensitivity is reflected in a low threshold to osmotic stimulation. As previously stated, 38 ~ of the cells that responded to hypertonic NaC1 tailed to respond to tail pinch, which in the awake animal is a highly noxious stimulus, and in a lightly anesthetized animal causes movement. Finally, it seems impossible to account for the reactions of a cell like the one illustrated in Fig. 7 in terms of general 'arousal'. In the first place, this cell appeared 'specific' in the terminology of Hayward and Vincent 6. It responded to hypertonic NaCI but did not respond to tail pinch, auditory or visual stimuli. Secondly, although there was a lawful relation between osmolarity of injected NaCI and reaction, the relation was certainly not a simple progression of increasing activation in going from weaker to stronger concentrations, as might be expected if the reactions were merely due to 'arousal' produced by intracarotid NaCI injections. Hayward and Vincent's 6 criteria for defining a cell as 'osmosensitive' are as follows: 'a prompt response to intracarotid hypertonic solutions during or immediately after injection with a return to base line firing within 30-60 sec'. The cells studied in the present experiments conform to these criteria, and it therefore seems valid to apply the term 'osmosensitive' to these cells, including those in the midbrain. To those midbrain cells that responded to hypertonic NaCI, but not to other forms of sensory stimulation, the term 'specific' osmosensitive cell appears applicable. It should be pointed out, however, that extension of this research in the future should include a systematic study of other compounds (especially dextrose or sucrose). A more difficult problem concerns the functional significance of osmosensitive neurons in the midbrain. According to current theories as summarized recently by Mogenson 10,11 and by Mogenson and Huang 12, the most probable interpretation of midbrain osmosensitive cells is that they are being synaptically driven by neuronal activity from elsewhere in the brain. The effects of electrical stimulation of the lateral preoptic area, lateral hypothalamus, hippocampus, amygdala and olfactory bulb on unit activity of the lateral habenular nucleus in the rat were demonstrated by Mok
119 and Mogenson14,15; and they 15 suggested, on the basis of their findings, that the lateral habenular nucleus serves as an integrating area for limbic forebrain inputs. Furthermore, M o k and Mogenson 16 found that midbrain neurons (some in dorsal midbrain) were in turn influenced by stimulation of the lateral habenular nucleus. As Mogenson 10 points out: 'the critical output from the hypothalamic integrative sites appears to be to the midbrain (p. 123)'. On the other hand, the present finding of osmosensitive neurons in the midbrain does raise the question of whether there may be osmoreceptors in this dorsal midbrain area. While, from the point of view of classical neurophysiology, this seems improbable, this question nonetheless seems to deserve consideration in further research. Recent evidence led Blass to suggest that the caudal border of the osmosensitive zone for cellular thirst as originally conceived2,17 was too far anterior, and should be moved back to the zona incerta. Present data suggest the possibility that this border should be placed even more posteriorly, in the midbrain. To the best of my knowledge, the kind of thorough application of the intracranial injection technique, exemplified in the recent work of Blass 1 on the basal forebrain, has not as yet been extended to a study of the dorsal midbrain. This technique should be useful in determining whether midbrain osmosensitive neurons appear to be osmoreceptive. There are some recent findings on the possible role of the midbrain in extracellular thirst. According to Sharpe and Swanson 19, the mesencephalic central gray (but not the midbrain reticular formation) in the monkey appears to be sensitive to angiotensin II. These investigators inferred sensitivity to angiotensin II on the basis of the amount of water the monkey drank following injection into a specific brain site. ACKNOWLEDGEMENTS The author acknowledges the substantial technical assistance of W. J. Mundl, and also the help throughout the project and the critical reading of the manuscript by H. P. Malmo. This research was supported in part by grants from the National Institute of Mental Health, United States Public Health Service, and from the National Research Council of Canada. I wish to express my thanks to Abbott Laboratories, Limited, Montreal, P.Q., for supplying Normosol-R p H 7.4.
REFERENCES 1 BLASS,E. M., Evidence for basal forebrain thirst osmoreceptors in rat, Brain Research, 82 (1974) 69-76. 2 BLASS,E. M., AND EPSTEIN,A. N., A lateral preoptic osmosensitive zone for thirst in the rat, J. comp. physioL PsychoL, 76 (1971) 378-394. 3 BROWN, K. A., WEBER,D. S., AND BUCHWALD,J. S., Chronic recording and quantification of subcortical single and multiple units. In M. I. PmLLIPS(Ed.), Brain Unit Activity During Behavior, Thomas, Springfield, I11., 1973, pp. 41-52. 4 GROVES, P. M., MILLER, S.W., PARKER, M.V., AND REBEC, G. V., Organization by sensory modality in the reticular formation of the rat, Brain Research, 54 (1973) 207-224.
120 5 HAYWARD, J. N., AND JENNINGS, D. P., Activity of magnoceUular neuroendocrine cells in the hypothalamus of unanaesthetized monkeys. I. Functional cell types and their anatomical distribution in the supraoptic nucleus and the internuclear zone, J. Physiol. (Lond.), 232 (1973) 515-543. 6 HAYWARD, J. N., AND VINCENT, J. D., Osmosensitive single neurones in the hypothalamus of unanaesthetized monkeys, J. PhysioL (Lond.), 210 (1970) 947-972. 7 HUANG, Y. H., AND MOGENSON, G. J., Neural pathways mediating drinking and feeding in rats, Exp. Neurol., 37 (1972) 269-286. 8 I'-IUBEE, D. H., Tungsten microelectrode for recording from single units, Science, 125 (1957) 549-550. 9 MALMO, R. B., AND MUNDL, W. J., Osmosensitive neurons in the rat's preoptic area: mediallateral comparison, J. comp. physiol. Psychol., 88 (1975) 161-175. l0 MOt3ENSON, G. J., Hypothalamic limbic mechanisms in the control of water intake. In A. N. EPSTEIN, H. R. KISSILEFF AND E. STELLAR (Eds.), The Neuropsychology of Thirst: New Findings and Advances in Concepts, Winston, Washington, D.C., 1973, pp. 119-142. 11 MOGENSON,G. J., Electrophysiological studies of the mechanisms that initiate ingestive behaviours with special emphasis on water intake. In G. J. MOGENSON AND F. R. CALARESU (Eds.), Neural Integration of Physiological Mechanisms and Behaviour: J. A. F. Stevenson Memorial Volume University of Toronto Press, Toronto, 1975, pp. 248-266. 12 MOGENSON,G. J., AND HUANG, Y. H., The neurobiology of motivated behavior. In G. A. KERKUT AND J. W. PHILLIS (Eds.), Progress in Neurobiology. Vol. 1, Pergamon, Oxford, 1973, pp. 53-83. 13 MOGENSON, G. J., AND HUANG, Y. H., Facilitation and suppression of centrally elicited feeding and drinking by electrical stimulation of the midbrain, Canad. J. Psychol., 28 (1974) 252-259. 14 MoK, A. C. S., AND MOGENSON, G. J., Effect of electrical stimulation of the septum and the lateral preoptic area on unit activity of the lateral habenular nucleus in the rat, Brain Research, 43 (1972) 361-372. 15 MOK, A. C. S., AND MOGENSON, G. J., Effects of electrical stimulation of the lateral hypothalamus, hippocampus, amygdala and olfactory bulb on unit activity of the lateral habenular nucleus in the rat, Brain Research, 77 (1974) 417-429. 16 MOK, A. C. S., AND MOGENSON, G. J., Effects of electrical stimulation of the lateral habenular nucleus and lateral hypothalamus on unit activity in the upper brain stem, Brain Research, 78 (1974) 425435. 17 PECK,J. W., AND NOVIN, D., Evidence that osmoreceptors mediating drinking in rabbits are in the lateral preoptic area, J. comp. physiol. Psychok, 74 (1971) 134-147. 18 PELLEGRINO, L. J., AND CUSHMAN, A. J., ,4 Stereotaxic Atlas of the Rat Brain, Appleton-CenturyCrofts, New York, 1967. 19 SHARPE,L. G., AND SWANSON,L. W., Drinking induced by injections of angiotensin into forebrain and mid-brain sites of the monkey, J. Physiol. (Lond.), 239 (1974) 595-622.