Psychiatry Research: Neuroimaging, 40:31-48 Elsevier
31
The Brain Stem Reticular Formation in Schizophrenia Craig N. Karson, Edgar Garcia-Rill, JoAnn Biedermann, Robert E. Mrak, Muhammad M. Husain, and Robert D. Skinner Received August 15, 1990; revised version received January 3, 1991; accepted January 28, 1991. Abstract. Post-mortem brain tissue was obtained from four patients with schizophrenia and five controls to study cell groups in the brain stem reticular formation. Cholinergic neurons in the pedunculopontine nucleus (PPN) and lateral dorsal tegmental nucleus (LDT) were labeled using nicotinamide adenosine dinucleotide phosphate (NADPH)-diaphorase histochemistry, while catecholaminergic neurons of the locus ceruleus (LC) were labeled immunocytochemically using an antibody to tyrosine hydroxylase. In schizophrenic patients, there were increased numbers of neurons in the PPN labeled by NADPH-diaphorase and reduced cell size in the LC. These results implicate the reticular formation as a possible pathophysiological site for at least some patients with schizophrenia. This also suggests that some of the deficits observed may be based on faulty neurodevelopment.
Key Words. Schizophrenia, brain stem reticular formation, pedunculopontine nucleus, locus ceruleus. In schizophrenia, reductions in Stages 3 and 4 of non-rapid eye movement ( N R E M ) sleep (Caldwell and Domino, 1967; Feinberg et al., 1969; Itil et al., 1972) and reductions in rapid eye movement (REM) latency (Jus et al., 1973; Zarcone et al., 1975; T a n d o n and Greden, 1989) have been reported. Peculiar movements and postural aberrations also are c o m m o n in schizophrenia (King, 1974). These and other functions relevant to schizophrenia are modulated by the pedunculopontine nucleus (PPN) and lateral dorsal tegmental nucleus (LDT), the main cholinergic cell groups of the reticular formation (RF) of the brain stem (Armstrong et al., 1983; Mesulam et al., 1984; Satoh and Fibiger, 1985; Isaacson and T a n a k a , 1986; Woolf and Butcher, 1986; Jones and Beaudet, 1987b). For instance, these nuclei generate R E M - o n activity (Steriade and McCarley 1990). Injection of acetylcholine or carbachol into the locus ceruleus (LC) at its junction with the cholinergic cells of the P P N triggers R E M sleep (reduces R E M latency). Atropine suppresses R E M sleep (Shiromani et al., 1988; Vanni-Mercier et al., 1989; Hobson,
Craig N. Karson, M.D., is Professor of Psychiatry, University of Arkansas for Medical Sciences, and Chief, Psychiatry Service, John L. McClellan Memorial Veterans Hospital, Little Rock, AR. Edgar Garcia-Rill, Ph.D., is Professor of Anatomy; JoAnn Biedermann is Research Assistant; Robert E. Mrak, M.D., Ph.D., is Associate Professor of Pathology and Anatomy; Muhammad M. Husain, M.D., is Associate Professor of Pathology; and Robert D. Skinner, Ph.D., is Professor of Anatomy, Universityof Arkansas for Medical Sciences, Little Rock, AR. (Reprint requests to Dr. C.N. Karson, Psychiatry Service, John L. McClellan Memorial Veterans Hospital, 4300 W. 7th St., 116A-NLR, Little Rock, AR 72205, USA.) 0165-1781/91/$03.50 © 1991 ElsevierScientific Publishers Ireland Ltd.
32 1990). Localized injections of cholinomimetics inhibit the onset of N R E M sleep, whereas cholinergic neurons of the P P N become inactive at the onset of N R E M sleep (Vanni-Mercier et al., 1989; Hobson, 1990). Moreover, injections of cholinomimetics into the pars c o m p a c t a of the P P N produce "hallucinating," erratically behaving animals (Baghdoyan et al., 1984; Vanni-Mercier et al., 1989). Hence, the sleep disruptions and even the hallucinations that occur in schizophrenia could be produced by enhancing the cholinergic tone of the P P N and the LDT. A correlation between sleep disturbances and negative symptoms in schizophrenia (Ganguli et al., 1987), as well as evidence that cholinergic agonists such as physostigmine produce negative symptoms (Janowsky et al., 1979) whereas anticholinergics reduce negative symptoms (Tandon and Greden, 1989), further suggests that the negative symptoms of schizophrenia may relate to enhanced brain cholinergic activity. Finally, locomotion is also modulated by P P N as part of the mesencephalic locomotor region (MLR; Garcia-Rill and Skinner, 1987a, 1987b). The purpose of the current study was to investigate the R F cholinergic and catecholaminergic cell groups in schizophrenia by determining the number and the size of the neurons in the PPN, L D T , and LC in post-mortem brain tissue of schizophrenic patients compared with control subjects.
Methods Subjects. Post-mortem brain tissue was obtained during autopsy from psychiatric patients and control subjects who expired at the John L. McClellan Memorial VA Hospital. A probable lifetime psychiatric diagnosis was established for all subjects by a physician using the Diagnostic Evaluation After Death (DEAD; Salzman et al., 1983). The DEAD uses all available information, including chart review and information obtained from kin. When sufficient information is obtained, the DEAD appears to be a valid and reliable instrument (Karson et al., 1990). The diagnosis of schizophrenia and alcohol abuse relies on the criteria of Feighner et al. (1972), while that of schizoaffective disorder and affective disorder relies on Research Diagnostic Criteria (Spitzer et al., 1977). Kin were interviewed in all but one case in this study including control subjects. Three of the four subjects diagnosed as schizophrenic in the current study had been examined by a neuropsychiatrist (C.N.K.) using DSM-III-R (American Psychiatric Association, 1987), and shortly before death, the DSM-III-R diagnosis was schizophrenia. Handling of Brain Tissue. At autopsy, the brain stem was separated from the remainder of the brain and was then bisected at the midline. Only the right half of the brain stem was used in the neuronal studies described. Immediately after dissection, the right half of the brain stem was placed in a solution of 4% paraformaldehyde with 1 mM MgC12 for 4-6 hours, then transferred to 20% sucrose with 1 mM MgC12 for 3-4 days. Sagittal frozen sections (60/am) were cut to minimize cell counting errors due to the vertical orientation of PPN cells and every fourth section processed for nicotinamide adenosine dinucleotide phosphate (NADPH)diaphorase histochemistry (Vincent et al., 1983). Cholinergic neurons in the midbrain (PPN and LDT) are labeled selectively by this technique, producing a purple-blue cytoplasmic label which extends well into the processes without labeling the nucleus or nucleolus (Vincent et al., 1983; Mesulam et al., 1989). These authors used NADPH-diaphorase histochemistry and choline acetyltransferase (CHAT) immunocytochemistry to confirm colocalization. Every sixth section also was processed for tyrosine hydroxylase (TH) immunocytochemistry. Cell counts were performed using a Biographics Image Analysis System. The outlines of sections were digitized (at 40X) and the locations of each cell type entered (at 100) 18 hours) yield very light or no staining. Such tissues were not quantified. Quantitation was carried out only in brains in which the reported Golgi-like labeling produced by this technique was evident. Another recent study on normal human brain post-mortem demonstrated the presence of double labeling of cholinergic mesopontine neurons using NADPH-diaphorase histochemistry and ChAT immunocytochemistry (Mesulam et al., 1989). Neuronal counts performed by these authors showed that NADPH-diaphorase-positive neurons outnumbered ChAT-positive cells by 10%. This difference cannot account for the 100% difference observed in the present report. The NADPH-diaphorase histochemical technique provides "an excellent first approximation" for cholinergic mesopontine neurons (Mesulam et al., 1989). Our future studies will use both ChAT and NADPH-diaphorase labeling to confirm our initial observations. Moreover, analysis of both sides of the brain stem will be carried out to determine if there is a lateralization of the effects observed (only the right brain stems were quantified in the present studies). A final issue which should be addressed is that, by selectively labeling a population of cells with these techniques, statements can be made regarding that subpopulation of neurons. There are a large number of noncholinergic PPN neurons which would need to be selectively labeled or counted using Nissl-stained sections. The problem with this approach is that the boundaries of the PPN are a matter of controversy, in addition to the fact that the nucleus has a concentrated pars compacta and a diffuse pars dissipata, each with varying percentages of noncholinergic neurons. The direct approach of counting only selectively labeled cells allows us to address differences in NADPH-diaphorase-positive neurons by not regarding density differences or noncholinergic cell differences. Cell area measures were obtained at 250× magnification by randomly measuring the soma perimeter of at least 200 cells in each nucleus (LC, LDT, and PPN) per half brain stem. Only cells with an identifiable nucleus were measured. Cells that were measured were selected on the basis of the completeness of perikaryal shape and profile of a nucleus and dendrites. Random
34 sections were measured by two individuals with no significant difference found across observers. Overlapping neurons were excluded. Statistical analyses were performed using SAS software, Version 6.0 (SAS Institute Inc., 1987). Nonparametric comparisons between groups used the Wilcoxon rank-sums test, and correlations between variables were performed using the Spearman correlation.
Results Subjects. Four subjects with a diagnosis of schizophrenia were studied (Table 1) and compared to five age-matched control subjects (2 normal and 3 psychiatric controls). The mean age of the schizophrenic group was 63 (SD = 10) years, which was not statistically different from that of the control group at 60 (SD ----9) years. Table 1 summarizes the cause of death and medications at the time of death in both groups. Two of the psychiatric control subjects suffered from ongoing alcohol dependence. The only female subject was a schizophrenic (HB4).
Cell Number. The NADPH-diaphorase histochemical procedure yielded a dark blue-purple reaction product that helped visualize the cell body as well as a considerable portion of the dendrites of the cell. The nucleus and nucleolus were not labeled. The distribution of NADPH-diaphorase-positive mesopontine cells was similar to that previously described for the normal human brain (Kowall and Mueller, 1988; Nakamura et al., 1988; Mesulam et al., 1989). Both N A D P H diaphorase histochemistry and ChAT immunocytochemistry have shown these neurons to correspond with cholinergic cells of the PPN and LDT. Table 2 shows that the mean cell number in the PPN of 8,500 (SD = 1,621) for control brain stems was significantly increased to 18,275 (SD = 8,220) for schizophrenic subjects. Cell number ranged from 6,435 to 9,994 for control subjects and 9,936 to 27,623 for schizophrenic patients, with cell number exceeding 22,000 in two schizophrenic subjects. Cell number was increased to a similar extent in LDT--6,268 (SD = 3,526) for control subjects vs. 12,266 (SD = 7,774) for schizophrenic patients--but this increase was not statistically significant due to the larger relative standard deviation. The T H immunocytochemically labeled neurons in the mesopontine area were found to correspond with those of the LC and substantia nigra (SN). The label produced a brown reaction product filling the cell body and some of the proximal dendrites. In sections processed only for NADPH-diaphorase histochemistry, LC and SN neurons were seen as pigmented, melanin-containing cells with varying degrees of pigmentation. The number of cells in the LC was virtually identical in the two groups (mean + SD = 11,353 ___4,791 vs. 11,496 ___2,956). It should be noted that the LC is located between the more lateral PPN and the more medial LDT. In fact, both cholinergic and noradrenergic neurons are found intermingled at the transition zones of these cell groups. Fig. I is a photomicrograph of a section from the transition zone between LC and PPN showing the blue-purple labeling of NADPH-diaphorase-positive cells and the brown pigment of melanincontaining cells. Such an overlap has been reported previously for other species (Jones and Beaudet, 1987a; Reiner and Vincent, 1987), in addition to humans (Mesulam et al., 1989; Garcia-Rill and Skinner, 1991).
35
II
o7o
~o~o :~_~ E
g II
Z o
~ oE O
Sg
- -< -< o~o
0 0 0 0
II
m. 12.
o o
31
The Brain Stem Reticular Formation in Schizophrenia Craig N. Karson, Edgar Garcia-Rill, JoAnn Biedermann, Robert E. Mrak, Muhammad M. Husain, and Robert D. Skinner Received August 15, 1990; revised version received January 3, 1991; accepted January 28, 1991. Abstract. Post-mortem brain tissue was obtained from four patients with schizophrenia and five controls to study cell groups in the brain stem reticular formation. Cholinergic neurons in the pedunculopontine nucleus (PPN) and lateral dorsal tegmental nucleus (LDT) were labeled using nicotinamide adenosine dinucleotide phosphate (NADPH)-diaphorase histochemistry, while catecholaminergic neurons of the locus ceruleus (LC) were labeled immunocytochemically using an antibody to tyrosine hydroxylase. In schizophrenic patients, there were increased numbers of neurons in the PPN labeled by NADPH-diaphorase and reduced cell size in the LC. These results implicate the reticular formation as a possible pathophysiological site for at least some patients with schizophrenia. This also suggests that some of the deficits observed may be based on faulty neurodevelopment.
Key Words. Schizophrenia, brain stem reticular formation, pedunculopontine nucleus, locus ceruleus. In schizophrenia, reductions in Stages 3 and 4 of non-rapid eye movement ( N R E M ) sleep (Caldwell and Domino, 1967; Feinberg et al., 1969; Itil et al., 1972) and reductions in rapid eye movement (REM) latency (Jus et al., 1973; Zarcone et al., 1975; T a n d o n and Greden, 1989) have been reported. Peculiar movements and postural aberrations also are c o m m o n in schizophrenia (King, 1974). These and other functions relevant to schizophrenia are modulated by the pedunculopontine nucleus (PPN) and lateral dorsal tegmental nucleus (LDT), the main cholinergic cell groups of the reticular formation (RF) of the brain stem (Armstrong et al., 1983; Mesulam et al., 1984; Satoh and Fibiger, 1985; Isaacson and T a n a k a , 1986; Woolf and Butcher, 1986; Jones and Beaudet, 1987b). For instance, these nuclei generate R E M - o n activity (Steriade and McCarley 1990). Injection of acetylcholine or carbachol into the locus ceruleus (LC) at its junction with the cholinergic cells of the P P N triggers R E M sleep (reduces R E M latency). Atropine suppresses R E M sleep (Shiromani et al., 1988; Vanni-Mercier et al., 1989; Hobson,
Craig N. Karson, M.D., is Professor of Psychiatry, University of Arkansas for Medical Sciences, and Chief, Psychiatry Service, John L. McClellan Memorial Veterans Hospital, Little Rock, AR. Edgar Garcia-Rill, Ph.D., is Professor of Anatomy; JoAnn Biedermann is Research Assistant; Robert E. Mrak, M.D., Ph.D., is Associate Professor of Pathology and Anatomy; Muhammad M. Husain, M.D., is Associate Professor of Pathology; and Robert D. Skinner, Ph.D., is Professor of Anatomy, Universityof Arkansas for Medical Sciences, Little Rock, AR. (Reprint requests to Dr. C.N. Karson, Psychiatry Service, John L. McClellan Memorial Veterans Hospital, 4300 W. 7th St., 116A-NLR, Little Rock, AR 72205, USA.) 0165-1781/91/$03.50 © 1991 ElsevierScientific Publishers Ireland Ltd.
32 1990). Localized injections of cholinomimetics inhibit the onset of N R E M sleep, whereas cholinergic neurons of the P P N become inactive at the onset of N R E M sleep (Vanni-Mercier et al., 1989; Hobson, 1990). Moreover, injections of cholinomimetics into the pars c o m p a c t a of the P P N produce "hallucinating," erratically behaving animals (Baghdoyan et al., 1984; Vanni-Mercier et al., 1989). Hence, the sleep disruptions and even the hallucinations that occur in schizophrenia could be produced by enhancing the cholinergic tone of the P P N and the LDT. A correlation between sleep disturbances and negative symptoms in schizophrenia (Ganguli et al., 1987), as well as evidence that cholinergic agonists such as physostigmine produce negative symptoms (Janowsky et al., 1979) whereas anticholinergics reduce negative symptoms (Tandon and Greden, 1989), further suggests that the negative symptoms of schizophrenia may relate to enhanced brain cholinergic activity. Finally, locomotion is also modulated by P P N as part of the mesencephalic locomotor region (MLR; Garcia-Rill and Skinner, 1987a, 1987b). The purpose of the current study was to investigate the R F cholinergic and catecholaminergic cell groups in schizophrenia by determining the number and the size of the neurons in the PPN, L D T , and LC in post-mortem brain tissue of schizophrenic patients compared with control subjects.
Methods Subjects. Post-mortem brain tissue was obtained during autopsy from psychiatric patients and control subjects who expired at the John L. McClellan Memorial VA Hospital. A probable lifetime psychiatric diagnosis was established for all subjects by a physician using the Diagnostic Evaluation After Death (DEAD; Salzman et al., 1983). The DEAD uses all available information, including chart review and information obtained from kin. When sufficient information is obtained, the DEAD appears to be a valid and reliable instrument (Karson et al., 1990). The diagnosis of schizophrenia and alcohol abuse relies on the criteria of Feighner et al. (1972), while that of schizoaffective disorder and affective disorder relies on Research Diagnostic Criteria (Spitzer et al., 1977). Kin were interviewed in all but one case in this study including control subjects. Three of the four subjects diagnosed as schizophrenic in the current study had been examined by a neuropsychiatrist (C.N.K.) using DSM-III-R (American Psychiatric Association, 1987), and shortly before death, the DSM-III-R diagnosis was schizophrenia. Handling of Brain Tissue. At autopsy, the brain stem was separated from the remainder of the brain and was then bisected at the midline. Only the right half of the brain stem was used in the neuronal studies described. Immediately after dissection, the right half of the brain stem was placed in a solution of 4% paraformaldehyde with 1 mM MgC12 for 4-6 hours, then transferred to 20% sucrose with 1 mM MgC12 for 3-4 days. Sagittal frozen sections (60/am) were cut to minimize cell counting errors due to the vertical orientation of PPN cells and every fourth section processed for nicotinamide adenosine dinucleotide phosphate (NADPH)diaphorase histochemistry (Vincent et al., 1983). Cholinergic neurons in the midbrain (PPN and LDT) are labeled selectively by this technique, producing a purple-blue cytoplasmic label which extends well into the processes without labeling the nucleus or nucleolus (Vincent et al., 1983; Mesulam et al., 1989). These authors used NADPH-diaphorase histochemistry and choline acetyltransferase (CHAT) immunocytochemistry to confirm colocalization. Every sixth section also was processed for tyrosine hydroxylase (TH) immunocytochemistry. Cell counts were performed using a Biographics Image Analysis System. The outlines of sections were digitized (at 40X) and the locations of each cell type entered (at 100) 18 hours) yield very light or no staining. Such tissues were not quantified. Quantitation was carried out only in brains in which the reported Golgi-like labeling produced by this technique was evident. Another recent study on normal human brain post-mortem demonstrated the presence of double labeling of cholinergic mesopontine neurons using NADPH-diaphorase histochemistry and ChAT immunocytochemistry (Mesulam et al., 1989). Neuronal counts performed by these authors showed that NADPH-diaphorase-positive neurons outnumbered ChAT-positive cells by 10%. This difference cannot account for the 100% difference observed in the present report. The NADPH-diaphorase histochemical technique provides "an excellent first approximation" for cholinergic mesopontine neurons (Mesulam et al., 1989). Our future studies will use both ChAT and NADPH-diaphorase labeling to confirm our initial observations. Moreover, analysis of both sides of the brain stem will be carried out to determine if there is a lateralization of the effects observed (only the right brain stems were quantified in the present studies). A final issue which should be addressed is that, by selectively labeling a population of cells with these techniques, statements can be made regarding that subpopulation of neurons. There are a large number of noncholinergic PPN neurons which would need to be selectively labeled or counted using Nissl-stained sections. The problem with this approach is that the boundaries of the PPN are a matter of controversy, in addition to the fact that the nucleus has a concentrated pars compacta and a diffuse pars dissipata, each with varying percentages of noncholinergic neurons. The direct approach of counting only selectively labeled cells allows us to address differences in NADPH-diaphorase-positive neurons by not regarding density differences or noncholinergic cell differences. Cell area measures were obtained at 250× magnification by randomly measuring the soma perimeter of at least 200 cells in each nucleus (LC, LDT, and PPN) per half brain stem. Only cells with an identifiable nucleus were measured. Cells that were measured were selected on the basis of the completeness of perikaryal shape and profile of a nucleus and dendrites. Random
34 sections were measured by two individuals with no significant difference found across observers. Overlapping neurons were excluded. Statistical analyses were performed using SAS software, Version 6.0 (SAS Institute Inc., 1987). Nonparametric comparisons between groups used the Wilcoxon rank-sums test, and correlations between variables were performed using the Spearman correlation.
Results Subjects. Four subjects with a diagnosis of schizophrenia were studied (Table 1) and compared to five age-matched control subjects (2 normal and 3 psychiatric controls). The mean age of the schizophrenic group was 63 (SD = 10) years, which was not statistically different from that of the control group at 60 (SD ----9) years. Table 1 summarizes the cause of death and medications at the time of death in both groups. Two of the psychiatric control subjects suffered from ongoing alcohol dependence. The only female subject was a schizophrenic (HB4).
Cell Number. The NADPH-diaphorase histochemical procedure yielded a dark blue-purple reaction product that helped visualize the cell body as well as a considerable portion of the dendrites of the cell. The nucleus and nucleolus were not labeled. The distribution of NADPH-diaphorase-positive mesopontine cells was similar to that previously described for the normal human brain (Kowall and Mueller, 1988; Nakamura et al., 1988; Mesulam et al., 1989). Both N A D P H diaphorase histochemistry and ChAT immunocytochemistry have shown these neurons to correspond with cholinergic cells of the PPN and LDT. Table 2 shows that the mean cell number in the PPN of 8,500 (SD = 1,621) for control brain stems was significantly increased to 18,275 (SD = 8,220) for schizophrenic subjects. Cell number ranged from 6,435 to 9,994 for control subjects and 9,936 to 27,623 for schizophrenic patients, with cell number exceeding 22,000 in two schizophrenic subjects. Cell number was increased to a similar extent in LDT--6,268 (SD = 3,526) for control subjects vs. 12,266 (SD = 7,774) for schizophrenic patients--but this increase was not statistically significant due to the larger relative standard deviation. The T H immunocytochemically labeled neurons in the mesopontine area were found to correspond with those of the LC and substantia nigra (SN). The label produced a brown reaction product filling the cell body and some of the proximal dendrites. In sections processed only for NADPH-diaphorase histochemistry, LC and SN neurons were seen as pigmented, melanin-containing cells with varying degrees of pigmentation. The number of cells in the LC was virtually identical in the two groups (mean + SD = 11,353 ___4,791 vs. 11,496 ___2,956). It should be noted that the LC is located between the more lateral PPN and the more medial LDT. In fact, both cholinergic and noradrenergic neurons are found intermingled at the transition zones of these cell groups. Fig. I is a photomicrograph of a section from the transition zone between LC and PPN showing the blue-purple labeling of NADPH-diaphorase-positive cells and the brown pigment of melanincontaining cells. Such an overlap has been reported previously for other species (Jones and Beaudet, 1987a; Reiner and Vincent, 1987), in addition to humans (Mesulam et al., 1989; Garcia-Rill and Skinner, 1991).
35
II
o7o
~o~o :~_~ E
g II
Z o
~ oE O
Sg
- -< -< o~o
0 0 0 0
II
m. 12.
o o
