Effects of mental illness and aging in two thalamic nuclei

Abstract

We previously reported a schizophrenia associated reduction of neuronal and oligodendrocyte number in the anterior principal thalamic nucleus (APN) in a cohort of severely impaired elderly subjects with schizophrenia (SZ) relative to age matched nonpsychiatric controls (NCs). The present study was undertaken to determine 1) if those findings could be replicated in an independent sample of less chronically impaired subjects with SZ and NCs stratified across a broader age range; 2) if the findings are specific to SZ or are also seen in unipolar major depressive (MDD) or bipolar disorder (BPD); and 3) if the findings are specific to the APN or also seen in another thalamic nucleus. Computer assisted stereological methods were employed to determine the number of neurons and oligodendrocytes in the APN and centromedian nucleus (CMN) of the Nissl-stained thalamic sections maintained by the Stanley Foundation Brain Bank. This collection includes specimens from NCs and age matched subjects with diagnoses of SZ, MDD, or BPD who died between the ages of 25 and 68. Data were analyzed by mixed-effects linear regressions adjusting for demographic variables and known history of exposure to psychotropic medications. Oligodendrocyte number was decreased in both nuclei relative to NCs in subjects with SZ and in that subset of subjects with BPD who had experienced psychotic episodes. Compared to NCs both of these patient groups also exhibited an attenuation of an age-related increase in the number of oligodendrocytes. Contrary to our previous report, we did not detect a SZ-associated deficit in neuronal number in the APN. A history of exposure to neuroleptics, however, was associated with a decrease in neuronal number in both nuclei, but this decrease did not vary in relation to cumulative lifetime neuroleptic exposure in fluphenazine equivalents. Among subjects with psychiatric diagnoses, exposure to lithium was associated with an increase in the number of oligodendrocytes. No effects were detected for exposure to anticonvulsants or for abuse of alcohol or other substances.

Figure 1. Delineation of APN

Coronal 60 micron thick Nissl stained sections through the APN arranged rostrocaudally from panels a) to c). The APN comprises both the anteroventral (AV) and anteromedial (AM) nuclei. Because the boundary between these two subdivisions is indistinct, they were not assessed separately in this study. At most levels, the APN is clearly delineated by the ventricular space medially, the internal medullary lamina medially, laterally and superiorly (arrowheads), and clusters of deeply staining neurons (arrows) of the internal medullary lamina ventrally and ventrolaterally. a) At this most anterior level both the AV and AM are present and a wide lamina separates them medially from the ventral anterior nucleus (VA). b) More caudally, AM has disappeared and only AV remains. Here AV is well circumscribed by the lamina. Ventrally, CL, a prominent member of the intralaminar nuclei, intervenes between AV and the mediodorsal nucleus (not shown in this panel). c) In its most caudal extent, AV is separated from the ventricle by LD whose neurons stain less intensely than those of AV. The boundary between AV and LD is indicated by the dotted line. At higher power, or with dark field optics (not shown), fascicles can be seen separating these two nuclei. The white arrow indicates the superior border of CL which intervenes between AV and the MDN. The scale bar corresponds to 5 mm. Abbreviations: AM anterior medial nucleus, APN anterior principal nucleus, AV anteroventral nucleus, CL central lateral nucleus, LD lateral dorsal nucleus, VA ventral anterior nucleus, VLp ventral lateral posterior nucleus.

Figure 2. Delineation of CMN

Coronal 60 micron thick Nissl stained sections through the CMN. a) In the coronal plane, CMN emerges anteriorly just as LD replaces the APN more dorsally. The CMN is bounded at this level by the MDN medially, Pf ventromedially and VPM laterally. At most levels the CMN is clearly circumscribed by the internal medullary lamina which can be seen in the higher power views of panels b) and c). b) Higher power view of same section depicted in a): Here the internal medullary lamina (arrowheads) can be seen surrounding the CMN medially and laterally. Ventromedially, the border with Pf is less clear. At times fascicles of the lamina (arrowheads) may be seen along this border and in other places deeply staining clusters (arrows) of neurons associated with the lamina may be seen. Care was taken to minimize the extent to which these clusters were included within the perimeter drawn for the CMN. c) In the posterior extent of CMN, PA replaces VPM medially. CMN is well circumscribed by the lamina (black arrowheads) and clusters of intralaminar neurons (arrows). A prominent grouping of neurons, sometimes designated as the magnocellular division of the CMN (Dewulf 1971) is conspicuous at this level (white arrow). No attempt was made to subdivide the CMN in the present study. The scale bar corresponds to 5 mm. Abbreviations: CMN centromedian nucleus, LD lateral dorsal nucleus, MDN mediodorsal nucleus, PA anterior pulvinar nucleus, Pf parafascicular nucleus, VLp ventral lateral posterior nucleus, VPL ventral posterior lateral nucleus, VPM ventral posterior medial nucleus.

Figure 3. Characteristics Cells Counted

Neurons were only counted if in addition to cytoplasm, a clear nucleus containing a nucleolus (white arrows) was visible. Small, nuclei which were almost perfectly round, uniformly darkly stained, and lacking visible cytoplasm were counted as presumed oligodendrocytes (arrowheads). a) A single neuron with conspicuous nucleolus (white arrow) is seen. Several oligodendrocytes are seen, one in close association with the neuron (probably a satellite), one with no apparent association with neurons or other oligodendrocytes, and three closely spaced and aligned as if associated with a single axon. b) Although 2 neurons are seen, only on has a distinct nucleolus (white arrow) in focus here. A satellite oligodendrocyte (arrowhead) is seen closely opposed to one of the neurons. The other darkly stained profiles are not in focus and, therefore, not classified here. c) A single neuron with nucleolus (white arrow) is seen and 2 nuclei meeting our criteria for presumptive oligodendrocytes. d) Only two nuclei meeting criteria for oligodendrocytes are seen (black arrowheads). Scale bar corresponds to 10 microns.

Figure 4. Oligodendrocyte Number and Density Across Diagnostic Categories

The main effect of SZ was significant for both number (z = −2.47, p = .01) and density (z = −2.89; p = .004) of oligodendrocytes. Main effects for other diagnoses were nonsignificant. There were no statistically significant nucleus by diagnostic group interactions. In the absence of such interactions, post hoc tests are not warranted and it cannot be concluded that any diagnosis differentially affected either nucleus. Means and SEs for each group are given in Table 2. The numbers in parentheses refer to the numbers of subjects in each group.

Figure 5. Age by Diagnosis Interaction for NC vs SZ

The diagnosis by age interaction was significant in the comparison of NCs with SZs for number of oligodendrocytes (z = −2.40, p = .02). As shown here, the regression of number on age has a greater positive slope in NCs than subjects with SZ. This interaction approached significance for density of oligodendrocytes (z = −1.99, p = .05) (data not shown). When all subjects were stratified by history vs no history of psychotic episodes, the psychosis by age interaction approached significance for number of oligodendrocytes (z = −2.00, p = .05) with the age response in nonpsychotics being similar to that of the NCs and the response of the psychotics similar to that of the SZs (data not shown).

Figure 6. Neuronal Number in Patients With and Without History of Neuroleptic Exposure

Mean neuronal number was lower in both nuclei in patients with (+) compared to those without (−) a history of neuroleptic exposure giving rise to a significant main effect of exposure (z=−2.35, p=.018). The numbers in parentheses refer to the numbers of subjects in each group for which the comparisons were made.

Figure 7. Oligodendrocyte Number in Patients With and Without History of Lithium Exposure

Mean oligodendrocyte number was higher in both nuclei in patients with (+) compared to those without (−) a history of lithium exposure giving rise to a significant main effect of exposure (z=3.57, p=.0004). The main effect was also significant for oligodendrocyte density (z=4.41,p=.00001) (Data not shown). The numbers in parentheses refer to the numbers of subjects in each group.