Abstract
Keywords
ABBREVIATIONS
intermediolateral nucleus
Lewy body
large fibers (minimum diameter of at least 6µm)
multiple system atrophy
normal control
orthostatic hypotension
Parkinson disease and DLB with OH
Parkinson disease and DLB without OH
Parkinson disease
small fibers (maximal diameter less than 6µm)
transverse fascicular area
total myelinated fibers
level of the 12th thoracic segment
ventral root
Introduction
The intermediolateral nucleus (IML) of the thoracic spinal cord contains the preganglionic neurons of the sympathetic nervous system. In multiple system atrophy (MSA), a disease spectrum that includes striatonigral degeneration, olivopontocerebellar atrophy, and Shy–Drager syndrome, loss of IML neurons plays a major role in orthostatic hypotension (OH) [1]. OH is also a common clinical presentation in individuals with Parkinson disease (PD) [2]. Histopathologically, Lewy bodies (LBs), related alpha-synucleinopathy with neuronal degeneration in sympathetic [3–10], parasympathetic [8, 11–17], and visceral autonomic neurons [8, 18–22], have been considered to be associated with such autonomic phenomena in PD. In the first reported case of autonomic failure with OH, LBs and severe (90%) loss of IML neurons were observed at the level of the 12th thoracic segment (Th12) [23]. However, few reports have focused on the IML in PD [1, 25], and none have dealt with dementia with LBs(DLB).
The aim of the study is to clarify the neuropathologic alterations of OH in PD and DLB. We analyzed the number of IML neurons at Th12 as well as the number of myelinated fibers in the ventral root (VR). In particular, we focused on small myelinated fibers (SFs), as well as the fine myelinated fibers (FFs) considered to be B fibers (i.e., preganglionic sympathetic nerve fibers) derived from IML neurons.
Materials and methods
Cases
Tissue samples were obtained from autopsy materials collected at the Tokyo Metropolitan Geriatric hospital and Institute of Gerontology [26, 27]. We analyzed 18 PD cases (mean age at death, 83.8 years; 7 women and 11 men) and 15 DLB cases (82.2 years; 7 women and 8 men) in this study. These 33 cases were available for analyzing the IML and VRs at Th12 (Th12-IML and Th12-VRs) (Table 1). Before selecting these cases, we excluded other neurodegenerative diseases (progressive supranuclear palsy, cortico-basal degeneration, amyotrophic lateral sclerosis, fronto-temporal lobar degeneration, Alzheimer disease, argyrophilic grain dementia), cerebral vascular diseases, brain tumors, neuro-inflammatory diseases, and diseases of the spine or spinal cord. DLB was clinically and neuropathologically diagnosed according to the Consensus Guidelines [28–30]. Diagnosis of PD was based on the UK PD Society Brain Bank Clinical Diagnostic Criteria [31].
We also used 18 normal control (NC) cases (mean age at death, 78.6 years; 8 women and 10 men) that satisfied all of the following four conditions: (1) no neurodegenerative diseases, cerebral vascular diseases, brain tumors, neuro-inflammatory diseases, toxic disorders, or diseases of the spine or spinal cord; (2), no phosphorylated alpha synuclein (pSyn#64, monoclonal, 1 : 20000, Wako, Osaka, Japan) immunoreactivity (i.e., no Lewy body–associated pathology) in the brain, spinal cord, skin, esophagus, heart, adrenal gland, or sympathetic ganglia; (3) no phosphorylated TDP-43 (pSer409/410, monoclonal, 1 : 10000, Cosmo Bio, Tokyo, Japan) immunoreactivity in the brain or spinal cord; and (4) Braak neurofibrillary stage of 0 to II and amyloid stage of 0 or A [32, 33]. There were no significant differences in age or gender between NC and PD or DLB cases.
In addition, to compare them with cases having autonomic nervous system abnormality, we analyzed 9 MSA cases (mean age at death, 73.0 years; 6 women and 3 men) diagnosed by the second consensus statement on diagnosing MSA [34].
Three neuropathologists reviewed and diagnosed each case separately and had conferences to determine the final diagnosis.
The institutional review board approved on this particular research theme.
Clinical information
All clinical information, including the presence or absence of autonomic symptoms as well as diabetes mellitus, was obtained from medical charts and reviewed by two board-certified neurologists. The presence of OH was defined as follows: a fall in systolic blood pressure of at least 20 mm Hg and/or a fall in diastolic blood pressure of at least 10 mm Hg within 3 minutes of standing [37] and/or postural syncope.
Neuropathologic analysis
Sampling
Spinal cords with VRs were fixed in 20% buffered formalin (WAKO, Osaka, Japan) for 7 to 13 days, dehydrated in a graded-alcohol series, cleared in xylene, and embedded in paraffin.
To identify the level of Th12, we identified the Th1 segment and counted the number of posterior roots.
Quantitative analysis of Th12–IML
We quantified the IML neurons by using a modified version of previously reported methodology [25, 36]. Three serial transverse slices, 5 mm in thickness, were cut at Th12 of the formalin-fixed paraffin-embedded spinal cord. The IML was defined as a triangular area of gray matter [1, 24]. Four serial 6- μm-thick sections were obtained from each block and subjected to Klüver–Barrera staining, and 12 slices (24 IMLs; bilateral) were analyzed for each case. To count IML neurons, neurons were identified by the presence of Nissl substance and prominent nucleoli. We only counted neurons with nucleoli.
The degree (density score) of Lewy-related α-synucleinopathy in the IML was semiquantitatively assessed in a low-power field (×10) on the basis of the scoring system of the third report of the DLB consortium [30]. In brief, 0, absent; 1, only neurites or dots or cytoplasmic appearance in a glia. 2, 1–3 LBs or neuronal cytoplasmic bodies. 3, more than 4 LBs or neuronal cytoplasmic bodies. 4, numerous LBs filled with severe immunoreactivity in the neuropil.
Sampling of Th12-VRs
In the present study, we employed two methods for processing Th12-VRs. In 5, 5, 3, and 6 cases of NC, PD, DLB, and MSA, respectively, Th12-VRs were obtained at the time of autopsy, and fixed in 2% glutaraldehyde in 0.1 M phosphate buffer (pH, 7.4) for 6 h. After fixation, the tissue was post-fixed in 2% osmium tetroxide for 2 h at room temperature. In contrast, in 13, 13, 12, and 6 other cases of NC, PD, DLB, and MSA, Th12-VRs were obtained after fixation in 20% buffered formalin. In 3 MSA cases, one side of the Th12-VR was fixed in 2% glutaraldehyde and the other side in 20% buffered formalin. After obtaining Th12-VRs from the formalin-fixed spinal cord, they were fixed in 2% glutaraldehyde in 0.1 M phosphate buffer (pH, 7.4) for 6 h followed by post-fixation in 2% osmium tetroxide for 2 h at room temperature. To avoid pooling the effects of two different fixatives, we analyzed Th12-VRs separately for each fixation method.
Quantitative analysis of myelinated fibers in Th12-VRs
After dehydration, the Th12-VRs were embedded in epoxy resin and 1- μm thick transverse sections stained with toluidine blue were examined. The total transverse fascicular area (TFA) of stained slides was digitized, by using a high-resolution automated slide scanner (ScanScope CS Scanner, Aperio, Vista, California). To measure the number and transverse profile diameter of myelinated fibers for each fascicle of a Th12-VR, as well as the TFA, the contour of the outer edge of the endoneurium was traced by using analysis software (Image-Pro PLUS ver. 5.1, Media Cybernetics, Silver Spring, Maryland). The fiber size distribution in each root was plotted on a linear scale.
Myelinated fibers were divided into two groups: large fibers (LFs), defined as fibers with a minimal diameter of at least 6 μm; and SFs, defined as fibers with a diameter of less than 6 μm. SFs were further divided in two categories: fine myelinated fibers (FFs) with a diameter of less than 3 μm, and non-FFs. FFs were considered to be preganglionic sympathetic nerve fibers derived from IML neurons [37]. Therefore, SFs included FFs in this study. The densities of total myelinated fibers (TFs), SFs, LFs, and FFs were calculated. In addition, the areal ratios of each fiber group to the TFA (density per 1 mm2), of LFs to SFs, and of FFs to TFs were calculated.
Statistical analysis
Data were expressed as mean ± 1 S.D. The Mann–Whitney U test was used to analyze differences in age, number of neurons, or density and ratio of myelinated fibers as a function of disease. Fisher’s exact test was also used to analyze differences in gender or presence of clinical symptoms as a function of disease. Statistical analysis was performed by using SPSS 15.0J software (SPSS, Chicago, IL, USA). The criterion for statistical significance was set at p < 0.05.
Results
Clinical and pathological characteristics
Seven of the 18 PD cases and five of the 15 DLB cases had OH (OH+ cases) (Table 1). None of the 18 NC cases had OH. Six of the 18 NC cases had diabetes mellitus (DM), but only one DLB case out of all other cases (PD, DLB, and MSA) had this condition. Nine of the 15 DLB cases had parkinsonism, and 13 of the 18 PD cases had dementia.
α-synucleinopathy and number of neurons in the IML
In the IML of all PD, DLB and MSA cases except only one DLB case, α-synuclein immunopositive structures were observed (Table 1).
The total IML neurons obtained from 24 IML regions are shown Fig. 1. The mean number of Th12-IML neurons in NC cases was 64.1 ± 12.6. Those in PD and DLB cases were 33.9 ± 10.5 (PD / NC = 53%) and 30.0 ± 11.2 (DLB/NC = 47%), respectively. These values were significantly lower than that of NC (PD, p < 0.001; DLB, p < 0.001). The number of IML neurons in MSA cases was 17.7 ± 7.8 (MSA/NC = 28%), which was significantly less than that of PD, DLB, or NC (p < 0.001). There was no significant difference between numbers of IML neurons in NC with DM and without DM.
Myelinated fibers in the VR
In NCs, LFs and SFs were unevenly present (Fig. 2A). In PD and DLB, SFs were apparently reduced relative to NC (Fig. 2 B, C). In MSA, SFs were more severely reduced relative to PD and DLB (Fig. 2D).
Fixation via glutaraldehyde
In NC, the mean number and density of TFs in Th12-VR were 4455 ± 2962 /fascicle and 10992 ± 2344 /mm2, respectively (Table 1). The histogram showed 2 peaks (a higher SF peak at 3 μm and a lower LF peak at 7 or 9 μm) and a valley bottom between them (at 6 or 8 μm) (Fig. 3 A, C). The SF peak was twice as high as the LF peak. Although the histogram showed 2 peaks even in PD, DLB, and MSA, SFs were fewer and LF peak heights were lower than in NC.
The mean SF/LF ratio in DLB was 1.78 ± 1.15, which was lower than that in NC (3.17 ± 2.01). The ratio in PD was 1.22 ± 0.29, which was significantly lower than that in NC (p = 0.016).
One of the 5 NC cases had DM. The number and density of TFs and SFs, and SF/LF ratio of this case was lower the mean numbers of 4 NC cases without DM.
Fixation by formalin and glutaraldehyde
In NC, the mean number and density of TFs in one Th12-VR fascicle were 4244 ± 1845 /fascicle and 11168 ± 1486 /mm2, respectively (Table 1). The histogram showed 2 peaks (a higher SF peak at 3 μm and lower LF peak at 9 μm) and a valley bottom between them (at 6 μm) (Fig. 3B, D). The SF peak was twice as high as the LF peak. Although the histogram showed 2 peaks even in PD and DLB, the SF peaks were lower and nearer to the LF peaks than in NC. This tendency was stronger in MSA, and the histogram showed two peaks of roughly the same height.
The mean number of FFs in PD was 703 ± 455/fascicle, respectively, and was significantly lower than that in NC (1239 ± 470, p = 0.004). This was also the case for the corresponding value in DLB (519 ± 308, p < 0.001). The mean number of FFs in MSA was 291 ± 223 /fascicle, respectively, was significantly lower than that in PD (p = 0.044). The densities of FF relative to NC (3348 ± 682/mm2) were significantly reduced in PD (2233 ± 1001/mm2, 67%, p = 0.003) and DLB (2388 ± 1288 /mm2, 71%, p = 0.022) but less so than in MSA (1419 ± 980/mm2, 42%).
The mean FF/TF ratios in PD (22.3 ± 8.6%) and DLB (21.0 ± 8.6%) were significantly lower than the ratio in NC (29.9 ± 3.8%; p = 0.013 and 0.019, respectively).
Five of the 13 NC cases had DM. There were no significant difference between the all items (mean number, density and ratio) in NC cases with DM and without DM.
Orthostatic hypotension
Twelve of the 33 PD and DLB cases had OH (OH+ cases) (Table 1). There were no significant differences in age or gender between PD and DLB cases without OH (OH–cases) and OH+ cases. The mean number of Th12-IML neurons in OH–cases was 35.1 ± 10.6 (55% of that in NC; Fig. 3). The corresponding number in OH+ cases was 27.0 ± 9.6 (42% of that in NC and 74% of that in OH–), which was significantly lower than that in OH–cases (p = 0.039).
Fixation by glutaraldehyde
The histogram of OH–cases showed 2 peaks; because the SF peak was lower than that in NC, the two peaks were about the same height (Fig. 3E, G). In OH+ cases, these tendencies were stronger, and the SF peak was difficult to identify. The mean number of FFs in the Th12-VR of OH–cases differed significantly from that of OH+ cases (565 ± 222 vs. 261 ± 194/fascicle, respectively; p = 0.025) but not from that of NC cases (Table 1).
Fixation by formalin and glutaraldehyde
The histogram of OH–cases showed 2 peaks, and the SF peak was lower than that of NC cases (Fig. 3F, H). In OH+ cases, the latter tendency was more apparent. The mean ratio of small to large myelinated fibers in OH+ (1.18), but not that in OH–(1.58), was significantly lower than that in NC (3.17). The mean number and density of FFs in the Th12-VR of OH–cases were 728 ± 438/fascicle and 2489 ± 1047/mm2, respectively; the corresponding values for OH+ cases were 412 ± 189/fascicle and 1985 ± 1250/mm2 (Table 1). The values for both OH–and OH+ cases were significantly lower than those for NC (OH–: p = 0.005, p = 0.007; OH+: p < 0.001, p = 0.012). The values for MSA were significantly lower than those for OH–(p = 0.018, p = 0.039) but not significantly different from those for OH+. The mean FF/TF ratios in OH–and OH+ cases were 23.8 ± 6.4% and 17.9 ± 10.6%, respectively, and were significantly lower than those in NC (p = 0.014, p = 0.018).
Discussion
Our present results show the following: (1) In PD and DLB, IML neurons and preganglionic nerve fibers (B fibers) decrease, albeit to a lesser degree than in MSA. (2) IML neurons and B fibers in PD and DLB with OH (OH+ cases) were fewer than those in PD and DLB without OH (OH–cases). The latter finding suggests that loss of IML neurons and B fibers are one of the causes of OH.
Rajput and Rozdilsky [6] suggested that OH in PD is related to neuronal loss with LBs in the sympathetic ganglia. LBs were also found in the paravertebral sympathetic ganglia in 23 of 25 patients with PD [25]. However, there are few convincing data for neuronal loss in the IML. IML neuron counts in PD were reported to be lower than those in controls [1, 24]. Another study [18] reported that IML neuron counts in 25 PD cases were lower than those in normal controls at Th2 and Th9. However, IML neuron counts did not differ between PD cases with OH and those without OH. Our study is the first to show that OH+ cases show significantly greater loss of IML neurons than do OH–cases and that the degree of this loss is less than in MSA. In addition, PD and DLB cases showed fewer IML neurons than did normal controls. Therefore, we suggest that IML neurons always decrease in PD and DLB by up to 50% of NC. Degeneration of preganglionic IML neurons in the thoracic spinal segments was suggested as a possible cause of OH in PD and DLB. In fact, axonal loss was predominantly observed in the thin myelinated fibers of autonomic preganglionic axons by degeneration of preganglionic IML neurons [38].
Whereas TFs were fewer in PD and DLB than in NC, LFs differed little among PD, DLB, and NC. Therefore, the difference in TFs was attributed to a difference in SFs. In normal individuals, a histogram of myelinated fibers in the thoracic VR shows two peaks: a higher SF peak and a lower LF peak [38, 39]. In the present study, the NC histogram also showed 2 peaks and a valley. The SF peak was twice as high as the LF peak. In this study, the SF peaks in PD, DLB, and MSA were lower than in NC. The SF peak in Shy–Drager syndrome was reported as lower than in NC [38]. However, SFs in PD and DLB have never been reported. The results above suggest that OH in PD and DLB is associated with pathologic changes in SFs.
Erlanger and Gasser divided mammalian nerve fibers into groups A, B, and C. B fibers have preganglionic autonomic function, and their diameter is usually less than 3 μm [37]. Therefore, we considered that FFs (<3 μm) were B fibers derived from IML neurons at Th12 in the present study. We found an apparent decrease of FF density and the FF/TF ratio in the Th12VR of PD and DLB cases. In addition, the results were independent of the fixative used.
Impaired autonomic function may occur in peripheral neuropathies. There were six cases of diabetes mellitus (DM) in NC and one in DLB. There is no significant difference of results between NC cases with DM and without DM. No other peripheral nervous system diseases were identified in the present study. Because DM patients were predominantly observed in NC group, we believe that DM did not affect our results.
In the present study, neuronal counting method in the IML was not done using an unbiased morphometric method, and neuronal and/or nuclear size and shape were not measured. Therefore, it is possible that at least some of the apparent decrease in IML neuronal number was due to shrinkage of these neurons, and/or their nuclei, in diseased cases relative to controls. Also, only neurons with a nucleolus were counted, and therefore if nucleolar numbers per neuron differed between diseased cases and controls, this would have influenced the neuronal counts. In addition, we could not directly identify the B fibers derived from IML neurons. We evaluated the number of IML neurons as well as the numbers of SFs and FFs. Because SFs consist of FFs and non-FFs, we assumed that a reduction in FFs reflects a decrease of B fibers and may be associated with OH. Therefore, our determination of B fibers was indirect. In addition, because clinical data were obtained retrospectively from medical records, detailed information on laboratory tests of the autonomic nervous system were not readily available.
In conclusion, we present novel neuropathological evidence that FFs in VR are decreased in PD and DLB. In particular, FFs and IML neurons are significantly reduced in OH+ cases relative to those in NC and OH–cases (PD and DLB). In the future, comparisons of physiological and neuropathologic studies will be important to support the present results.
Sponsorship
This study was supported in part by a Grant-in-Aid for Young Scientists (B) (Kakenhi), Japan (to HH), Study on propagation of Lewy body- associated synucleinopathy, Grant-in-Aid for Scientific Research (B) (24300133), the Ministry of Education, Sports, Science and Technology, Japan (to SM), the innovative development in the treatment of protein propagation scheme, the Grants in Aid from the Japan Agency for Medical Research and Development, Japan (to SM), the epidemiological neuropathology of neurodegenerative disorders, the Research on Policy Planning and Evaluation for Rare and Intractable Diseases, the Health and Labour Sciences Research Grants from the Ministry of Health, Labour and Welfare, Japan (to SM), and Establishment of Japanese Brain Bank Network for Neuroscience Research, Grants-in-Aid for Scientific Research on Innovative Areas (Comprehensive Brain Science Network, 221S0003), the Ministry of Education, Sports, Science and Technology, Japan (to SM).
The authors have no interest of conflict.
Footnotes
ACKNOWLEDGMENTS
We thank Mr. Naoo Aikyo, Mr. Fumio Hasegawa, Ms. Mieko Harada, Ms. Nobuko Naoi, and Ms. Yuki Kimura for preparing sections and Dr. Kinuko Suzuki for helpful discussions. The authors also thank the 2 anonymous neurologists for preparing the clinical dementia rating used in this study.