globulin region of SSPE low-pH extracts were directed against measles virus, the extracts were suitably absorbed with M P A and the patterns of unabsorbed and absorbed preparations were compared by electrophoresis. The MPA-absorbed preparations lacked most of the oligoclonal bands and contained a diffuse band in the gamma globulin region. Similar results were obtained when the electrophoretic patterns of unabsorbed and MPA-absorbed SSPE brain extracts of neutral p H were compared [GI. IMMUNOLOGICALSTUDIES. Immunological comparison by Ouchterlony test of the low-pH extracts of SSPE brains with IgG isolated from neutral p H extracts using rabbit antiserum specific for SSPE IgG gave a reaction of identity.
Discussion These data confirm earlier studies of Weil et al [lo] and extend them by further characterizing the antibodies eluted at low pH. Our studies show that the low-pH extracts from 3 different SSPE brains not only possessed significant levels of antibodies to measles virus, but also had high measles-specific I g G content and showed oligoclonal bands on electrophoresis. Since these results are similar to those seen in I g G extracted at neutral p H [6,9], one can conclude that the IgG eluted at low and neutral p H is indistinguishable. A number of reports [ 1,7] have demonstrated the existence of immune complexes in serum, CSF, and kidney from SSPE patients. O u r data support the presence of immune complexes with measles virus in SSPE brains, since 40 to 60% of the bound IgG is measles specific. The low-pH extracts of 1 of the MS brains (Patient 4) had a low level of bound antibody and showed oligoclonal bands on electrophoresis. Since the antibodies bound to brain tissue had low but detectable titers against measles virus, the results indicate that a measles virus immune complex was present. Studies to amplify this finding are in progress. References
6. Mehta PD, Kane A, Thocmar H: Quantitation of measles virus specific immunoglobulinsin serum, CSF and brain extract from patients with subacute sclerosing panencephalitis. J Immunol
118:22%-2261, 1977 7. Perrin LH, Oldstone MBA: The formation and fate of virus antigen-antibody complexes. J Immunol 118:316-322,1977 8. Vandvik B, Norrby E: Oligoclonal IgG antibody response in the central nervous system to difTerent measles virus antigens in subacute sclerosing panencephalitis. Proc Natl Acad Sci USA 70:1060-1063, 1973 et aL Oligoclonal measles 9. Vandvik B, Norrby E, Nordal virus-specific IgG antibodies isolated from cerebrospinal fluids, brain extracts and sera from patients with subacute sclerosing panencephalitis and multiple sclerosis. &and J Immunol 5979-992, 1976 10. Weil ML, Leiva WA, Heiner DC, et al: Release of bound immunoglobulin from SSPE brain by acid elution. J Immunol
w,
115:1603-1606, 1975
Cytoplasmic Filaments in Intracerebral Cortical Vessels Sukria Nag, MD, David M. Robertson, MD, and Henry B. Dinsdale, MD
Actin filaments measuring 5 nm in diameter ate present in the endothelium of systemic vessels and presumably have a contractile function that may be related to vascular permeability. Little attention has been directed to the presence of similar cytoplasmic filaments in the endothelium of cerebral vessels. This study was undertaken to determine whether filaments are present in the endothelial cells of intracerebral cortical vessels of rats and humans. Ultrastructural studies revealed two populations of endothelial cytoplasmic filaments 5 nm and 10 nm in diameter, respectively. The 5-nm filaments were abundant in the endothelium of penetrating cortical arterioles, being grouped near the cytoplasmic margins in proximity to cell junctions and along the abluminal side of the endothelial cells. The filaments were scant in capillaries and venules.
Nag S, Robertson DM, Dinsdale HB: Cytoplasmic filaments in intracerebral cortical vessels. Ann Neurol 3:555-559, 1978
1. Dayan AD, Stokes MI: Immune complexes and visceral depos-
its of measles antigens in subacute sclerosing panencephalitis. Br Med J 2~374-376,1972 2. Fahey JL, McKelvey EM: Quantitative determination of serum immunoglobulins in antibody-agar plates. J Immunol 94:8490, I965 3. Johansson BG: Agarose gel electrophoresis. Scand J Clin Lab Invest 29:7-19, 1972 4. Link H,Panelius M, Salmi AA: Immunoglobulins and measles antibodies in subacute sclerosing panencephditis. Arch Neurol
28:23-30, 1973 5 . Mehta PD, Kane A, Thormar H: Relationship between homogeneous I g G fractions and measles virus antibody activities in subacute sclerosing panencephalitis brain. J Immunol 117:2053-2060, 1976
Several reports [ 5 , 11, 201 describe actin filaments in the endothelium of systemic vessels other than cerebral vessels. These filaments are thought to produce endothelial cell contraction, thereby widening interendothelial spaces and allowing direct passage of From the Departments of Pathology and Medicine, Queen's University and Kingston General Hospital, Kingston, Ont, Canada. Accepted for publication Jan 25, 1978. Address reprint requests to D r Robernon, Department of Pathology, Queen's University, Kingston, Ont, Canada K7L 3N6.
0364-5134/78/0003-0617$01.00 @ 1978 by Sukriti Nag
555
substances from the vascular compartment into the subendothelium. The present study was undertaken to determine whether normal intracerebral cortical vessels of rats and humans have similar filaments.
Materials and Methods Twenty female Wistar-Furth rats (200 to 250 gm) were perfusion fixed via a wide-bore needle in the ascending aorta. The fixative solution [lo] was infused at apressure of 120 mm H g for elght minutes while fixative was allowed to escape through a slit in the right atrium. Brains were removed and immersed in fixaave solution for a further two hours. Frontal lobe biopsies of 2 normotensive paaents were immersion fixed in the same fixative. One biopsy was from a 44-year-old man with a progressive dementia of unknown MCUR.The other biopsy was obtained from a 22-year-old patient suspected of having vasculitis. Neither biopsy showed pathological changes on light or electron microsCOPY. Tissues were posdixed for ninety minutes in 1% osmium tetroxide, dehydrated through graded alcohols, impregnated with propylene oxide and Epon mixtures, and embedded in Epon 812. Sections approximately 300 nm thick were cut (Sorvd Porter-Blum MT-ZB) and stained with 1% taluidine blue for light microscopy. Ultrathin sections (60 to 100 nm) of selected blocks mounted on copper grids were stained with uranyl acetate and lead ciuate and exarnined with a Hitachi H5OO electron microscope at 100 kv. Thirty penetrating arterioles measuring 18 to 34 p m in diameter of rat cerebral cortex and 6 penetraang arterioles of human frontal cortex were studied. The ultrastructural features of numerous capillaries and venules were also observed.
Results The ultrastructural features of endothelial cells of both rat and human arterioles were as described in the literature [19]. In addition, there were two populations of cytoplasmic filaments. Thin filaments were numerous in the endothelial cells of both human and rat cortical arterioles. They were arranged in discrete bundles along the entire circumference of the vessel wall and were particularly prominent along the abluminal side of the endothelial cells. The filaments measured 5.2 0.4 nm in diameter in rats and 5.8 +- 0.3 nm in humans. The morphological resemblance of these filaments to myofilaments in the subjacent smooth muscle cells was striking (Fig 1). The bundles of thin filaments were often arrayed along the endothelial cell cytoplasms bordering the intercellular spaces between endothelial cells (Figs 2, 3). Capillaries and venules of both rat and human cortex showed a few small, inconspicuous bundles of similar thin filaments near the intercellular spaces between endothelial cells. The endothelium of rat cortical arterioles showed small numbers of thzckfilaments, each measuring 10.2
*
556 Annals of Neurology Vol 3 No 6 June 1978
Pig 1 rRat cortex.) Segment ofpenetrating corticalurteriole showing the endothelium and smooth muscle kayer. B u n d h of thinfilaments (F) are present along thc abluminalside of the enahtbelialceII.Note their resemblance to the myofkaments in the adjacent .smooth muscle cell. A group of thickkfilaments is present in cross-section (arrow). (x56,400 before 5 % reduction.) I
k 0.4 nm in diameter. These filaments were also arranged in groups along the longitudinal axis of the vessel and were often observed around cytoplasmic organelles, particularly mitochondria (see Fig 2). They were sometimes randomly distributed in the cytoplasm of the endothelial cells and not associated with other organelles. These filaments showed no periodicity and o n cross-section appeared to be tubular with hollow cores (Fig 4). The thick filaments were more numerous in the endothelial cells of arterioles in human cortex. They measured 9.7 f 1.0 nm in diameter. Their arrangement was similar to that of the thick filaments of rat endothelial cells. Occasional groups of thick filaments were observed in the endothelium of capillaries and venules.
Discussion The present study demonstrates two populations of filaments in the cytoplasm of endothelial cells of normal cerebral cortical vessels of rats and humans. Little attention has been directed to the presence of endothelial filaments in normal cerebral vessels. Hirano [8] briefly mentioned that filamentous components are relatively inconspicuous in normal cerebral capillaries while in many pathological conditions, especially neoplasia, the endothelial cells become rich in fine 40- to 60-A filaments. Filaments were not observed in the endothelium of normal cerebral and
Fig 2. (Rat cortex.) Bundles ofthinfilaments (F)arepresent adjacent to the intercelldar space between two endothelial cells. Cross-sections of clnsters of thin (€9und thickfilaments (arrow) are seen in the endothelialcell nn the right side. (x95.200 before 10% reduction.)
F i g 3 . (Human cortex.)A Rroup of thinfilamentJ (F) is present a t the basal portion of the endothelial cell on the &fi side. f ~ 5 6 , 4 0 before 0 5% ! reduction.)
retinal arterioles of rats by Giacomelli et a1 [G, 71. However, these authors found filaments measuring 60 to 70 A in diameter in the arteriolar endothelium of rats with chronic renal hypertension. The thin filaments measured about 5 nm in diameter, while the thick filaments were 10 nm in diameter. Thin and thick filaments of similar dimensions have been observed in endothelial cells of myometrial arterioles [ZO], coronary arteries 1251, and pulmonary lymphatics [ll]. Numerous other studies describe
o n l y thin filaments in vascular endothelial cells [ 3 , 51 and in a variety of other cell types [4, 9, 18, 221.
These last authors showed the actin nature of these filaments by the ultrastructural demonstration of characteristic complexes (arrowhead structures) that these filaments form with heavy meromyosin. This technique was also used to prove the actin nature of thin filaments in the endothelial cells of renal arteries [Z 11 and pulmonary lymphatics [ll]. Gabbiani e t a1 [ 5 ] used immunofluorescent techniques employing antiactin antisera to demonstrate the actin nature of the microfilaments they observed in aortic endothelial cells. Immunological techniques applied to cerebral vessels have produced conflicting results. Becker and Shustak [ 11, using the immunoperoxidase technique on frozen sections, found that normal cerebral capillaries contain less contractile protein than capillaries of skeletal and heart muscle. Owman et al[16], using immunofluorescent techniques, observed actin and myosin in pericytes and endothelial cells of vessels which they interpreted as being brain capillaries. Although the limited resolution obtained with light microscopical immunofluorescent studies limits precise localization of the site of the filaments, these results appear to be consistent with our observations. We would emphasize, however, that thin filaments are numerous in arteriolar endothelium but are much fewer in capillary endothelium. The functional role of the filaments remains uncertain but is no doubt related to the contractile properties of endothelial cells [S, 14, 201. It has been suggested that in systemic vessels, endothelial cell contraction would widen the interendothelial gaps between the discontinuous tight junctions, thus enhancing vascular permeability [14]. However, in cerebral vessels the continuous bandlike interendothelial tight junctions might be more resistant to the effects of endothelial contraction. In many experimental models in which cere-
Brief Communication: Nag, Robemon, and Dinsdale: Cerebral Endothelial Filaments
557
brovascular permeability is increased, transport of tracer substances is at the capillary and venule level and is associated with markedly enhanced pinocytotic activity in the absence of demonstrable disruption of the interendothelial clefts [2, 12, 171. The paucity of endothelial filaments in these small vessels sheds further doubt o n their having a major role in the enhancement of permeability. In experimental hypertension, however, increased permeability of penetrating arterioles is the principal mechanism of horseradish peroxidase extravasation, and relatively minor changes are found in capillaries and venules [15, 241. Arteriolar permeability is also enhanced during experimental seizures [13, 17, 231 and after portacaval anastomoses [23]. In these situations the endothelium again shows increased pinocytotic transport. Although interendothelial passage has been suggested as occurring in experimental hypertension, it has not been convincingly shown. Thus, although normal cerebral arteriolar endothelium ordinarily contains abundant cytoplasmic filaments, their role, if any, in alterations of vascular permeability remains to be determined.
F i g 4 . iRut cortex.) High magnification showing cmss-sections ofbundlcs of thin andthickfilaments. (x113,OOO before 10% redHfl*on.)
5.
6.
7.
8.
9.
10.
11.
References 1. Becker CG. Shustak S R Contractile protein in endothelial cells: comparison of cerebral capillaries with those in heart and skeletal muscle and with liver sinusoids. Circulation 45, 46:Suppl 2:87, 1972 2. BriBhtman M W , Matzo I, Olsson Y,et al: The blood-brain barrier to proteins under normal and pathological conditions. J Neurol Sci 10:215-239. 1070 3. De Bruyn PPH, Cho Y :Contractile structures in endothelial cells of splenic sinusoids. J Ultrastruct Res 4924-33, 1974 4. French SW. Davies PL Ultrastructural localization of actin-
558
Annals of Neurology Vol 3
No 6 June 1978
12.
13.
14.
15.
like filamenu in rat hepatocytes. Gastroenterology 68:765774, 1975 Gabbiani G , Badonnel MD, Rona G: Cytoplasmic contractile apparatus in aortic endothelial cells of hypertensive rats. Lab Invest 32:227-234, 1975 Giacomelli F, Wiener J, Spim D: Cellular pathology of experimental hypertension: V. Increased permeability of cerebral arterial vessels. Am J Pathol 59:133-159, 1970 Giacomelli F, Juechter KB, Wiener J: The cellular pathology of experimental hypertension: VI. Alterations in retinal vasculature. Am J Pathol 68:81-96, 1972 Hirano A: Further observations of the fine structure of pathological reaction in cerebral blood vessels. in Cervos-Navarro J, Betz E, Matakas F. et al (eds): Cerebral Vessel Wall. New York, Raven Press, 1976, pp 41-49 Ishikawa H , Bischoff R, Holtzer H: Formation of arrowhead complexes with heavy meromyosin in a variety of cell types. J Cell Biol 43:312-328, 1969 Karnovsky MJ: T h e ultrastructural basis of capillary permeability studied with peroxidase as a tracer. J Cell Biol 35:213-236, 1967 Lauweryns JM, Baert J, De Loecker W: Intracytoplasmic filamenu in pulmonary lymphatic endothelial cells. Cell Tissue Res 163:111-124, 1975 Lee JC: Evolution in the concept of the blood-brain barrier phenomenon. Prog Neuropathol 1:84-145, 107 1 Lorenm AV, Hedley-Whyte ET, Eisenberg HM, et al: Increased penetration of horseradish peroxidase across the blood-brain barrier induced by Metrazol seizures. Brain Res 88:136-140, 1975 Majno G, Shea SM, Leventhal M: Endothelial contraction induced by histamine-type mediators. A n electron microscopic study. J Cell Biol 42:647-672, 1969 Nag S, Robertson DM. Dinsdale HB: Cerebral cortical
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
changes in acute experimental hypertension. An ultrasrructural study. Lab Invest 3 6 1 5 C 1 6 1 . 1977 Owman CH, Edvinsson L, Hardebo JE, et al: Immunohistochemical demonstration of actin and myosin in brain capillaries, in Ingvar DH, Lassen N A (eds): Cerebral Function, Metabolism and Circulation. Copenhagen, Munksgaard, 1977, pp 384-385 Petito CK, Schaefer JA, Plum F The blood-brain barrier in experimental seizures, in Pappius HM, Feindel W (eds): Dynamics of Brain Edema. Heidelberg, New York, Springer, 1976, pp 38-42 Pollard TD, Shelron E, Weihing RR, et al: Ultrastructural characterization of F-actin isolated from Aranthamorba ratelLaii and identificarion of cytoplasmic filaments as F-actin by reaction with rabbit heavy meromyosin. J Mol Biol50:91-97, 1970 Roggendorf W, Cervos-Navarro J, Matakas F The ultrastructural criteria of intracerebrd arterioles, in Cerv6s-Navarro J, Betz E, Matakas F, et al (eds): The Cerebral Vessel Wall. New York, Raven Press,pp 23-31, 1976 Rohlich P, Olah I: Cross-striated fibrils in the endothelium of the rat myomeuial arterioles. J Ultrastruct Res 18:667-676, 1967 Rostgaard J, Kristensen BJ, Nielson LE: Myofilarnents in nonmuscular cells, in Takeuchi T, Ogawa K, Fujita S (eds): Histochemisrry and Cytochemistry. Kyoto, Nakanishi Printing Company, 1972, p 377 Wessells NK, Spooner BS. Ash JF, et al: Microfilaments in cellular and developmental processes. Science 171:135-143. 1971 Westergaard E: The blood-brain barrier to horseradish peroxidase under normal and experimental conditions. Acta Neuropathol (Berl) 39:181-187, 1977 Westergaard E, van Deurs B, Brdndsred HE: Increased vesicular transfer of horseradish peroxidase across cerebral endothelium evoked by acute hypertension. Acta Neuropathol (Berl) 37:141-152, 1977 Yohro T, Burnstock G Filament bundles and contractility of endotheiial cells in coronary arteries. Z Zellforsch 138:8595, 1973
The .-- Malignancy of Uernentias Rodney C. P. Go, PhD, Alexandre B. Todorov, MD, Robert C. Elston, PhD, and Jean Constantinidis, MD
Survival times were determined for 982 patients admitted over a ten-year period to the Geneva Psychiatric Clinic. Patients with dementia had one-third the life expectancy of controls. Among patients classed as having Alzheimer disease, the longest survival times were in those with neurofibrillary tangles involving the neocortex, while those lacking this anatomical abnormality had the shortest survival times. Except €or women with Alzheimer disease, patients with dementia had less than 10% life expectancy from the time of their admission to the clinic compared to the life expectancy of Geneva’s population.
Go RCP, Todorov AB, Elston RC, et al: The malignancy of dementia. Ann Nearol3559-561, 1 9 7 8
Several studies 12-7, 91 describe a markedly decreased life span in the dementias. We report an analysis of the life expectancy of 982 persons treated at the Geneva Psychiatric Clinic with available postmortem examination. The simple was from a welldefined geographic and socioeconomic area, collected over a ten-year period. Survival time after onset of the condition and after first admission to the hospital were both investigated. The statistical method consisted of adjusting the survival time for each patient to a common age of onset or hospitalization of 65 years and comparing this adjusted survival time (AST)with the expected survival time (EST). Control life expectancy figures were obtained from life tables for the population of Geneva. This study confirms a decreased AST in the demented population, especially for the time after admission to a psychiatric hospital. It was also possible to demonstrate a significantly longer AST in patients with widespread distribution of neurofibrillary tangles compared to those without. Subjects A detailed description of the nosology has been reported 181. Six anatomical classes of patients were considered: type From the Department of Neurology, University of California, Davis, CA, the Department of Biostatistics and the Genetics Curriculum, University of North Carolina, Chapel Hill, NC, and the University Psychiatric Clinic, Geneva, Switzerland.
Accepted for publication Dec 30, 1977. Address reprint requests to D r Todorov, Department of Neurology, University of California. Davis, 4301 X St, Sacramento, CA 95817.
0364-5134/7810003-0618$01.00 @ 1978 by Rodney C. P. Go 559
Discussion These data confirm earlier studies of Weil et al [lo] and extend them by further characterizing the antibodies eluted at low pH. Our studies show that the low-pH extracts from 3 different SSPE brains not only possessed significant levels of antibodies to measles virus, but also had high measles-specific I g G content and showed oligoclonal bands on electrophoresis. Since these results are similar to those seen in I g G extracted at neutral p H [6,9], one can conclude that the IgG eluted at low and neutral p H is indistinguishable. A number of reports [ 1,7] have demonstrated the existence of immune complexes in serum, CSF, and kidney from SSPE patients. O u r data support the presence of immune complexes with measles virus in SSPE brains, since 40 to 60% of the bound IgG is measles specific. The low-pH extracts of 1 of the MS brains (Patient 4) had a low level of bound antibody and showed oligoclonal bands on electrophoresis. Since the antibodies bound to brain tissue had low but detectable titers against measles virus, the results indicate that a measles virus immune complex was present. Studies to amplify this finding are in progress. References
6. Mehta PD, Kane A, Thocmar H: Quantitation of measles virus specific immunoglobulinsin serum, CSF and brain extract from patients with subacute sclerosing panencephalitis. J Immunol
118:22%-2261, 1977 7. Perrin LH, Oldstone MBA: The formation and fate of virus antigen-antibody complexes. J Immunol 118:316-322,1977 8. Vandvik B, Norrby E: Oligoclonal IgG antibody response in the central nervous system to difTerent measles virus antigens in subacute sclerosing panencephalitis. Proc Natl Acad Sci USA 70:1060-1063, 1973 et aL Oligoclonal measles 9. Vandvik B, Norrby E, Nordal virus-specific IgG antibodies isolated from cerebrospinal fluids, brain extracts and sera from patients with subacute sclerosing panencephalitis and multiple sclerosis. &and J Immunol 5979-992, 1976 10. Weil ML, Leiva WA, Heiner DC, et al: Release of bound immunoglobulin from SSPE brain by acid elution. J Immunol
w,
115:1603-1606, 1975
Cytoplasmic Filaments in Intracerebral Cortical Vessels Sukria Nag, MD, David M. Robertson, MD, and Henry B. Dinsdale, MD
Actin filaments measuring 5 nm in diameter ate present in the endothelium of systemic vessels and presumably have a contractile function that may be related to vascular permeability. Little attention has been directed to the presence of similar cytoplasmic filaments in the endothelium of cerebral vessels. This study was undertaken to determine whether filaments are present in the endothelial cells of intracerebral cortical vessels of rats and humans. Ultrastructural studies revealed two populations of endothelial cytoplasmic filaments 5 nm and 10 nm in diameter, respectively. The 5-nm filaments were abundant in the endothelium of penetrating cortical arterioles, being grouped near the cytoplasmic margins in proximity to cell junctions and along the abluminal side of the endothelial cells. The filaments were scant in capillaries and venules.
Nag S, Robertson DM, Dinsdale HB: Cytoplasmic filaments in intracerebral cortical vessels. Ann Neurol 3:555-559, 1978
1. Dayan AD, Stokes MI: Immune complexes and visceral depos-
its of measles antigens in subacute sclerosing panencephalitis. Br Med J 2~374-376,1972 2. Fahey JL, McKelvey EM: Quantitative determination of serum immunoglobulins in antibody-agar plates. J Immunol 94:8490, I965 3. Johansson BG: Agarose gel electrophoresis. Scand J Clin Lab Invest 29:7-19, 1972 4. Link H,Panelius M, Salmi AA: Immunoglobulins and measles antibodies in subacute sclerosing panencephditis. Arch Neurol
28:23-30, 1973 5 . Mehta PD, Kane A, Thormar H: Relationship between homogeneous I g G fractions and measles virus antibody activities in subacute sclerosing panencephalitis brain. J Immunol 117:2053-2060, 1976
Several reports [ 5 , 11, 201 describe actin filaments in the endothelium of systemic vessels other than cerebral vessels. These filaments are thought to produce endothelial cell contraction, thereby widening interendothelial spaces and allowing direct passage of From the Departments of Pathology and Medicine, Queen's University and Kingston General Hospital, Kingston, Ont, Canada. Accepted for publication Jan 25, 1978. Address reprint requests to D r Robernon, Department of Pathology, Queen's University, Kingston, Ont, Canada K7L 3N6.
0364-5134/78/0003-0617$01.00 @ 1978 by Sukriti Nag
555
substances from the vascular compartment into the subendothelium. The present study was undertaken to determine whether normal intracerebral cortical vessels of rats and humans have similar filaments.
Materials and Methods Twenty female Wistar-Furth rats (200 to 250 gm) were perfusion fixed via a wide-bore needle in the ascending aorta. The fixative solution [lo] was infused at apressure of 120 mm H g for elght minutes while fixative was allowed to escape through a slit in the right atrium. Brains were removed and immersed in fixaave solution for a further two hours. Frontal lobe biopsies of 2 normotensive paaents were immersion fixed in the same fixative. One biopsy was from a 44-year-old man with a progressive dementia of unknown MCUR.The other biopsy was obtained from a 22-year-old patient suspected of having vasculitis. Neither biopsy showed pathological changes on light or electron microsCOPY. Tissues were posdixed for ninety minutes in 1% osmium tetroxide, dehydrated through graded alcohols, impregnated with propylene oxide and Epon mixtures, and embedded in Epon 812. Sections approximately 300 nm thick were cut (Sorvd Porter-Blum MT-ZB) and stained with 1% taluidine blue for light microscopy. Ultrathin sections (60 to 100 nm) of selected blocks mounted on copper grids were stained with uranyl acetate and lead ciuate and exarnined with a Hitachi H5OO electron microscope at 100 kv. Thirty penetrating arterioles measuring 18 to 34 p m in diameter of rat cerebral cortex and 6 penetraang arterioles of human frontal cortex were studied. The ultrastructural features of numerous capillaries and venules were also observed.
Results The ultrastructural features of endothelial cells of both rat and human arterioles were as described in the literature [19]. In addition, there were two populations of cytoplasmic filaments. Thin filaments were numerous in the endothelial cells of both human and rat cortical arterioles. They were arranged in discrete bundles along the entire circumference of the vessel wall and were particularly prominent along the abluminal side of the endothelial cells. The filaments measured 5.2 0.4 nm in diameter in rats and 5.8 +- 0.3 nm in humans. The morphological resemblance of these filaments to myofilaments in the subjacent smooth muscle cells was striking (Fig 1). The bundles of thin filaments were often arrayed along the endothelial cell cytoplasms bordering the intercellular spaces between endothelial cells (Figs 2, 3). Capillaries and venules of both rat and human cortex showed a few small, inconspicuous bundles of similar thin filaments near the intercellular spaces between endothelial cells. The endothelium of rat cortical arterioles showed small numbers of thzckfilaments, each measuring 10.2
*
556 Annals of Neurology Vol 3 No 6 June 1978
Pig 1 rRat cortex.) Segment ofpenetrating corticalurteriole showing the endothelium and smooth muscle kayer. B u n d h of thinfilaments (F) are present along thc abluminalside of the enahtbelialceII.Note their resemblance to the myofkaments in the adjacent .smooth muscle cell. A group of thickkfilaments is present in cross-section (arrow). (x56,400 before 5 % reduction.) I
k 0.4 nm in diameter. These filaments were also arranged in groups along the longitudinal axis of the vessel and were often observed around cytoplasmic organelles, particularly mitochondria (see Fig 2). They were sometimes randomly distributed in the cytoplasm of the endothelial cells and not associated with other organelles. These filaments showed no periodicity and o n cross-section appeared to be tubular with hollow cores (Fig 4). The thick filaments were more numerous in the endothelial cells of arterioles in human cortex. They measured 9.7 f 1.0 nm in diameter. Their arrangement was similar to that of the thick filaments of rat endothelial cells. Occasional groups of thick filaments were observed in the endothelium of capillaries and venules.
Discussion The present study demonstrates two populations of filaments in the cytoplasm of endothelial cells of normal cerebral cortical vessels of rats and humans. Little attention has been directed to the presence of endothelial filaments in normal cerebral vessels. Hirano [8] briefly mentioned that filamentous components are relatively inconspicuous in normal cerebral capillaries while in many pathological conditions, especially neoplasia, the endothelial cells become rich in fine 40- to 60-A filaments. Filaments were not observed in the endothelium of normal cerebral and
Fig 2. (Rat cortex.) Bundles ofthinfilaments (F)arepresent adjacent to the intercelldar space between two endothelial cells. Cross-sections of clnsters of thin (€9und thickfilaments (arrow) are seen in the endothelialcell nn the right side. (x95.200 before 10% reduction.)
F i g 3 . (Human cortex.)A Rroup of thinfilamentJ (F) is present a t the basal portion of the endothelial cell on the &fi side. f ~ 5 6 , 4 0 before 0 5% ! reduction.)
retinal arterioles of rats by Giacomelli et a1 [G, 71. However, these authors found filaments measuring 60 to 70 A in diameter in the arteriolar endothelium of rats with chronic renal hypertension. The thin filaments measured about 5 nm in diameter, while the thick filaments were 10 nm in diameter. Thin and thick filaments of similar dimensions have been observed in endothelial cells of myometrial arterioles [ZO], coronary arteries 1251, and pulmonary lymphatics [ll]. Numerous other studies describe
o n l y thin filaments in vascular endothelial cells [ 3 , 51 and in a variety of other cell types [4, 9, 18, 221.
These last authors showed the actin nature of these filaments by the ultrastructural demonstration of characteristic complexes (arrowhead structures) that these filaments form with heavy meromyosin. This technique was also used to prove the actin nature of thin filaments in the endothelial cells of renal arteries [Z 11 and pulmonary lymphatics [ll]. Gabbiani e t a1 [ 5 ] used immunofluorescent techniques employing antiactin antisera to demonstrate the actin nature of the microfilaments they observed in aortic endothelial cells. Immunological techniques applied to cerebral vessels have produced conflicting results. Becker and Shustak [ 11, using the immunoperoxidase technique on frozen sections, found that normal cerebral capillaries contain less contractile protein than capillaries of skeletal and heart muscle. Owman et al[16], using immunofluorescent techniques, observed actin and myosin in pericytes and endothelial cells of vessels which they interpreted as being brain capillaries. Although the limited resolution obtained with light microscopical immunofluorescent studies limits precise localization of the site of the filaments, these results appear to be consistent with our observations. We would emphasize, however, that thin filaments are numerous in arteriolar endothelium but are much fewer in capillary endothelium. The functional role of the filaments remains uncertain but is no doubt related to the contractile properties of endothelial cells [S, 14, 201. It has been suggested that in systemic vessels, endothelial cell contraction would widen the interendothelial gaps between the discontinuous tight junctions, thus enhancing vascular permeability [14]. However, in cerebral vessels the continuous bandlike interendothelial tight junctions might be more resistant to the effects of endothelial contraction. In many experimental models in which cere-
Brief Communication: Nag, Robemon, and Dinsdale: Cerebral Endothelial Filaments
557
brovascular permeability is increased, transport of tracer substances is at the capillary and venule level and is associated with markedly enhanced pinocytotic activity in the absence of demonstrable disruption of the interendothelial clefts [2, 12, 171. The paucity of endothelial filaments in these small vessels sheds further doubt o n their having a major role in the enhancement of permeability. In experimental hypertension, however, increased permeability of penetrating arterioles is the principal mechanism of horseradish peroxidase extravasation, and relatively minor changes are found in capillaries and venules [15, 241. Arteriolar permeability is also enhanced during experimental seizures [13, 17, 231 and after portacaval anastomoses [23]. In these situations the endothelium again shows increased pinocytotic transport. Although interendothelial passage has been suggested as occurring in experimental hypertension, it has not been convincingly shown. Thus, although normal cerebral arteriolar endothelium ordinarily contains abundant cytoplasmic filaments, their role, if any, in alterations of vascular permeability remains to be determined.
F i g 4 . iRut cortex.) High magnification showing cmss-sections ofbundlcs of thin andthickfilaments. (x113,OOO before 10% redHfl*on.)
5.
6.
7.
8.
9.
10.
11.
References 1. Becker CG. Shustak S R Contractile protein in endothelial cells: comparison of cerebral capillaries with those in heart and skeletal muscle and with liver sinusoids. Circulation 45, 46:Suppl 2:87, 1972 2. BriBhtman M W , Matzo I, Olsson Y,et al: The blood-brain barrier to proteins under normal and pathological conditions. J Neurol Sci 10:215-239. 1070 3. De Bruyn PPH, Cho Y :Contractile structures in endothelial cells of splenic sinusoids. J Ultrastruct Res 4924-33, 1974 4. French SW. Davies PL Ultrastructural localization of actin-
558
Annals of Neurology Vol 3
No 6 June 1978
12.
13.
14.
15.
like filamenu in rat hepatocytes. Gastroenterology 68:765774, 1975 Gabbiani G , Badonnel MD, Rona G: Cytoplasmic contractile apparatus in aortic endothelial cells of hypertensive rats. Lab Invest 32:227-234, 1975 Giacomelli F, Wiener J, Spim D: Cellular pathology of experimental hypertension: V. Increased permeability of cerebral arterial vessels. Am J Pathol 59:133-159, 1970 Giacomelli F, Juechter KB, Wiener J: The cellular pathology of experimental hypertension: VI. Alterations in retinal vasculature. Am J Pathol 68:81-96, 1972 Hirano A: Further observations of the fine structure of pathological reaction in cerebral blood vessels. in Cervos-Navarro J, Betz E, Matakas F. et al (eds): Cerebral Vessel Wall. New York, Raven Press, 1976, pp 41-49 Ishikawa H , Bischoff R, Holtzer H: Formation of arrowhead complexes with heavy meromyosin in a variety of cell types. J Cell Biol 43:312-328, 1969 Karnovsky MJ: T h e ultrastructural basis of capillary permeability studied with peroxidase as a tracer. J Cell Biol 35:213-236, 1967 Lauweryns JM, Baert J, De Loecker W: Intracytoplasmic filamenu in pulmonary lymphatic endothelial cells. Cell Tissue Res 163:111-124, 1975 Lee JC: Evolution in the concept of the blood-brain barrier phenomenon. Prog Neuropathol 1:84-145, 107 1 Lorenm AV, Hedley-Whyte ET, Eisenberg HM, et al: Increased penetration of horseradish peroxidase across the blood-brain barrier induced by Metrazol seizures. Brain Res 88:136-140, 1975 Majno G, Shea SM, Leventhal M: Endothelial contraction induced by histamine-type mediators. A n electron microscopic study. J Cell Biol 42:647-672, 1969 Nag S, Robertson DM. Dinsdale HB: Cerebral cortical
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
changes in acute experimental hypertension. An ultrasrructural study. Lab Invest 3 6 1 5 C 1 6 1 . 1977 Owman CH, Edvinsson L, Hardebo JE, et al: Immunohistochemical demonstration of actin and myosin in brain capillaries, in Ingvar DH, Lassen N A (eds): Cerebral Function, Metabolism and Circulation. Copenhagen, Munksgaard, 1977, pp 384-385 Petito CK, Schaefer JA, Plum F The blood-brain barrier in experimental seizures, in Pappius HM, Feindel W (eds): Dynamics of Brain Edema. Heidelberg, New York, Springer, 1976, pp 38-42 Pollard TD, Shelron E, Weihing RR, et al: Ultrastructural characterization of F-actin isolated from Aranthamorba ratelLaii and identificarion of cytoplasmic filaments as F-actin by reaction with rabbit heavy meromyosin. J Mol Biol50:91-97, 1970 Roggendorf W, Cervos-Navarro J, Matakas F The ultrastructural criteria of intracerebrd arterioles, in Cerv6s-Navarro J, Betz E, Matakas F, et al (eds): The Cerebral Vessel Wall. New York, Raven Press,pp 23-31, 1976 Rohlich P, Olah I: Cross-striated fibrils in the endothelium of the rat myomeuial arterioles. J Ultrastruct Res 18:667-676, 1967 Rostgaard J, Kristensen BJ, Nielson LE: Myofilarnents in nonmuscular cells, in Takeuchi T, Ogawa K, Fujita S (eds): Histochemisrry and Cytochemistry. Kyoto, Nakanishi Printing Company, 1972, p 377 Wessells NK, Spooner BS. Ash JF, et al: Microfilaments in cellular and developmental processes. Science 171:135-143. 1971 Westergaard E: The blood-brain barrier to horseradish peroxidase under normal and experimental conditions. Acta Neuropathol (Berl) 39:181-187, 1977 Westergaard E, van Deurs B, Brdndsred HE: Increased vesicular transfer of horseradish peroxidase across cerebral endothelium evoked by acute hypertension. Acta Neuropathol (Berl) 37:141-152, 1977 Yohro T, Burnstock G Filament bundles and contractility of endotheiial cells in coronary arteries. Z Zellforsch 138:8595, 1973
The .-- Malignancy of Uernentias Rodney C. P. Go, PhD, Alexandre B. Todorov, MD, Robert C. Elston, PhD, and Jean Constantinidis, MD
Survival times were determined for 982 patients admitted over a ten-year period to the Geneva Psychiatric Clinic. Patients with dementia had one-third the life expectancy of controls. Among patients classed as having Alzheimer disease, the longest survival times were in those with neurofibrillary tangles involving the neocortex, while those lacking this anatomical abnormality had the shortest survival times. Except €or women with Alzheimer disease, patients with dementia had less than 10% life expectancy from the time of their admission to the clinic compared to the life expectancy of Geneva’s population.
Go RCP, Todorov AB, Elston RC, et al: The malignancy of dementia. Ann Nearol3559-561, 1 9 7 8
Several studies 12-7, 91 describe a markedly decreased life span in the dementias. We report an analysis of the life expectancy of 982 persons treated at the Geneva Psychiatric Clinic with available postmortem examination. The simple was from a welldefined geographic and socioeconomic area, collected over a ten-year period. Survival time after onset of the condition and after first admission to the hospital were both investigated. The statistical method consisted of adjusting the survival time for each patient to a common age of onset or hospitalization of 65 years and comparing this adjusted survival time (AST)with the expected survival time (EST). Control life expectancy figures were obtained from life tables for the population of Geneva. This study confirms a decreased AST in the demented population, especially for the time after admission to a psychiatric hospital. It was also possible to demonstrate a significantly longer AST in patients with widespread distribution of neurofibrillary tangles compared to those without. Subjects A detailed description of the nosology has been reported 181. Six anatomical classes of patients were considered: type From the Department of Neurology, University of California, Davis, CA, the Department of Biostatistics and the Genetics Curriculum, University of North Carolina, Chapel Hill, NC, and the University Psychiatric Clinic, Geneva, Switzerland.
Accepted for publication Dec 30, 1977. Address reprint requests to D r Todorov, Department of Neurology, University of California. Davis, 4301 X St, Sacramento, CA 95817.
0364-5134/7810003-0618$01.00 @ 1978 by Rodney C. P. Go 559
