Cell Tissue Res (1992) 270:485-494
Cell&Tissue Research 9 Springer-Verlag 1992
Distribution of intraventricularly injected horseradish peroxidase in cerebrospinal fluid compartments of the rat spinal cord M. Cifuentes 1, p. Fernfindez-LLebrez ~, J. P~rez 1, J.M. P6rez-Figares z, and E.M. Rodriguez 3 Departamentos de BiologiaAnimaly 2 BiologlaCelular de la Facultad de Cienciasde M/tlaga,Spain 3 Instituto de Histologiay Patologiade la Universidad Austral de Chile,Valdivia,Chile Received March 9, 1992 / AcceptedJuly 14, 1992
Summary. The circulation of the cerebrospinal fluid along the central canal and its access to the parenchyma of the spinal cord of the rat have been analyzed by injection of horseradish peroxidase (HRP) into the lateral ventricle. Peroxidase was found throughout the central canal 13 rain after injection, suggesting a rapid circulation of cerebrospinal fluid along the central canal of the rat spinal cord. It was cleared from the central canal within 2 h, in contrast with the situation in the brain tissue, where it remained in the periventricular areas for 4 h. In the central canal, HRP bound to Reissner's fiber and the luminal surface of the ependymal cells; it penetrated through the intercellular space of the ependymal lining, reached the subependymal neuropil, the basement membrane of local capillaries, and appeared in the lumen of endothelial pinocytotic vesicles. Furthermore, it accumulated in the labyrinths of the basement membrane contacting the basolateral aspect of the ependymal ceils. In ependymocytes, HRP was found in single pinocytotic vesicles. The blood vessels supplying the spinal cord were classified into two types. Type-A vessels penetrated the spinal cord laterally and dorsally and displayed the tracer along their external wall as far as the gray matter. Type-B vessels intruded into the spinal cord from the medial ventral sulcus and occupied the anterior commissure of the gray matter, approaching the central canal. They represented the only vessels marked by HRP along their course through the gray matter. HRP spread from the wall of type-B vessels, labeling the labyrinths, the intercellular space of the ependymal lining, and the lumen of the central canal. This suggests a communication between the central canal and the outer cerebrospinal fluid space, at the level of the medial ventral sulcus, via the intercellular spaces, the perivascular basement membrane and its labyrinthine extensions. Key words: Central canal
Cerebrospinal fluid Circulation - Horseradish peroxidase - Ependyma - Rat (Sprague Dawley)
Correspondence to: P. Fern/mdez-LLebrez,Departamento de Biologia Animal, Facultad de Ciencias, E-29071 Mfilaga,Spain
In 1965, Bradbury and Lathem investigated the flow of cerebrospinal fluid (CSF) along the central canal of lampreys, rats, guinea pigs, rabbits and monkeys by injecting tracer molecules into the lateral ventricle. Among these species, only the rabbit exhibited circulating CSF within the central canal. Since a circulation of CSF along the central canal was not observed in the other species, these authors considered the flow of CSF in the canal compartment of the rabbit as a '"relic of the vertebrate evolutionary story", apparently absent in most recent species. However, Nakayama and Kohno (1974) and Nakayama (1976) reported the presence of india ink in the central canal of guinea pigs, rabbits and rats from 30 rain to 2 h after injection into the lateral ventricle. To our knowledge, there are no other reports dealing with the flow of CSF along the central canal under normal conditions. Eisenberg et al. (1974a, 1974b) have reported a flow in the central canal of cats suffering from kaolin-induced hydrocephalus. The distribution of tracer molecules injected into the cerebral ventricles via the interstitial spaces of the brain has been thoroughly investigated (Rall etal. 1962; Brightman 1965; Wagner et al. 1974; Brightman et al. 1975; Matakas et al. 1978; Krisch et al. 1984; Rennels et al. 1985; Cserr 1988; see also Wood 1983; Davson et al. 1987). The tracers circulate within the CSF and fill the inner (ventricular) and outer (meningeal) spaces of the CSF. From the ventricles of the brain, the tracers penetrate the ependymal layer through the intercellular space and gain access to the interstitial space of the subependymal parenchyma (Matakas etal. 1978). The tracers may then reach the basement membrane of the brain capillaries where they are cleared by specific phagocytotic cells or are transported to the blood stream by transcytosis across the endothelium (Wagner et al. 1974). Many authors have found that, from the meningeal spaces, the tracers penetrate the brain parenchyma by diffusing along the intercellular spaces of the marginal glia (Wagner etal. 1974; Borison etal. 1980; Zervas et al. 1982; Rennels et al. 1985), and have postulated that there is open communication between the subarach-
486 n o i d space a n d the interstitial space o f the brain. Such a n o p e n c o m m u n i c a t i o n b e t w e e n the m e n i n g e a l a n d the intercellular spaces o f the b r a i n has been challenged by Krisch et al. (1984) who have p e r f o r m e d a t h o r o u g h ult r a s t r u c t u r a l s t u d y o f the d i s t r i b u t i o n , in the m e n i n g e a l sheaths a n d s u b p i a l b r a i n tissue, o f h o r s e r a d i s h peroxidase ( H R P ) injected i n t o the CSF. T h e d i s t r i b u t i o n o f tracer molecules r e a c h i n g the central c a n a l a n d the m e n ingeal space o f the spinal cord has, a p p a r e n t l y , n o t yet been investigated. O u r p r e s e n t research is m a i n l y c o n c e r n e d with the s t u d y o f the s u b c o m m i s s u r a l o r g a n - R e i s s n e r ' s fiber complex, a b r a i n g l a n d o f u n k n o w n f n n c t i o n ( R o d r i g u e z et al. 1992). O n e o f the w o r k i n g hypotheses p o s t u l a t e s t h a t the s u b c o m m i s s u r a l o r g a n participates in the control o f the c i r c u l a t i o n o f the C S F (see R o d r i g u e z et al. 1992). T h e p r e s e n t i n v e s t i g a t i o n was designed to study, by use o f H R P , the c i r c u l a t i o n o f the C S F a l o n g the central c a n a l o f the rat spinal cord u n d e r n o r m a l conditions.
Materials and methods
Animals Male Sprague-Dawley rats (body weight=250-300 g) were kept under a photoperiod of L:D 12:12 and an environmental temperature of about 25~ C. They were fed ad libitum with rodent food (A04, Panlab, Barcelona, Spain). The animals were anesthetized with ether, placed in a stereotactic instrument and perfused through the right lateral ventricle with one of the working solutions. Two groups of animals were prepared: 1) Control rats: the animals were perfused, by use of a pump, with 30 gl 0.9% NaC1 at a rate of 3 gl/min and sacrificed 20 min after perfusion ended. 2) Experimental rats: these were perfused with 30 gl 3% HRP (Sigma, Type IV, Dorset, England), dissolved in 0.9% NaC1, at a rate of 3 gl/min. The central nervous system of these rats was fixed 13 rain (n = 5), 20 rain (n = 8), I h (n = 5), 2 h (n = 3), and 4 h (n = 3) after starting the HRP perfusion.
ma, Dorset, England) and 0.03% H202 (Merck, Darmstadt, FRG) in TRIS-phosphate-buffered saline (0.05 M, pH 7.8). After being rinsed in the same buffer for 30 rain, the sections were mounted in glycerol or in Aquatex (Merck, Darmstadt, FRG) for light microscopy. Some sections were further processed for electron microscopy.
Electron microscopy After the Vibratome sections had been processed for the demonstration of peroxidase, some were washed in 0.1 M phosphate buffer, pH 7.4, and were fixed in 1% OsO4 in the same buffer for 1 h. After dehydration in a graded series of ethanol and acetone, the sections were embedded in Araldite. Semithin and ultrathin sections were obtained. The former were stained with toluidine blue. The ultrathin sections were stained with 1% lead citrate for I min.
Histochemical and immunocytochemical methods In order to investigate the nature of certain structures apparently related to the capillary wall, normal (untreated) rats were transcardially perfused with Bouin's fluid, and the central nervous system was dissected out and embedded in paraffin. Transverse and horizontal sections (10 gm thick) of the brain and spinal cord were stained with periodic acid-silver methenamine (Burkholder !974). Other sections were processed for immunocytochemistry for the demonstration of type-IV collagen. For this purpose, the sections were hydrated and treated for 12 h with 0.5% papain (Difco, Detroit, USA) in 0.02 M phosphate buffer, pH 4.7, containing 5 mM sodium bisulfite and 0.5 mM EDTA (Junqueira et al. 1980). The sections were then incubated for 18 h at 22~ C in an antiserum against type-IV collagen (Chemicon International, Temecula, Calif., USA) at a dilution of 1:500. The linking antibody (from E.M. Rodriguez, Valdivia, Chile) was used at a dilution 1 : 10 for 30 min at 22~ C and the PAP complex (1 : 75) (Sigma P2030, Dorset, England) for 30 rain at 22~ C. DAB was used as the electron donor. The antisera and the PAP complex were diluted in 0.1 M TRIS buffer, pH 7.8, containing 0.7% lambda carrageenan (Sigma, Dorset, England).
Semiquantitative studies Fixation Under ether anesthesia, the animals were transcardially perfused with 100 ml 0.1 M phosphate buffer, pH 7.4, for 10 rain. Then, 300 ml fixative (1% glutaraldehyde, 1% paraformaldehyde in 0.1 M phosphate buffer) were perfused for 30 rain. After perfusion, the brain and the entire spinal cord were dissected out. The spinal cord was divided into five fragments: 1) cervical, including cervical segments C1 to C6; 2)proximal thoracic, including thoracic segments TI to T4; 3) medial thoracic (T5-T8); 4) distal thoracic (T9T12); 5)lumbo-sacral,'includingsegments L1 to the filum terminale. The brain and all pieces of the spinal cord were left immersed in the same fixative overnight, at 4~ C.
Light microscopy Vibratome sections, approximately 40 gm thick, were obtained from all tissue blocks. The brain was cut transversally; each one of the five fragments into which the spinal cord was cut was, in turn, subdivided into two halves: the distal half was sectioned transversally and the proximal half was sectioned horizontally. Peroxidase was visualized according to the method of Graham and Karnovsky (1966), using 0.1% 3,3' diaminobenzidine (DAB, Sig-
An image analyzer IBAS-2000 (IPS) (Kontron, FRG) with a program IPS (image processing system) was used. By means of a video camera connected to a light microscope, the image of a frame containing the central canal was digitalized as a matrix of 100 x 100 points. A value, representing the relative optic density (ROD), as a gray scale from 0=white to 255=black, was assigned to each point. This value was automatically corrected with respect to the background density of an area of the section apparently devoid of peroxidase reaction product. At least three transverse sections from each of the five fragments of the spinal cord (see above) were used for the analysis. The value given to a section of the central canal represented the mean of the values corresponding to the 10000 points, with the exception of those points having a value lower than the background unstained tissue. The value for a given fragment of the spinal cord was the mean of the values recorded for the three sections; the value ascribed to an experimental group represented the mean of values estimated for corresponding segments in animals of such a group.
Statistical analysis The data of ROD obtained for the different experimental groups and segments of the spinal cord were submitted to the Kolmogor-
487 ov-Smirnov test (P
Cell&Tissue Research 9 Springer-Verlag 1992
Distribution of intraventricularly injected horseradish peroxidase in cerebrospinal fluid compartments of the rat spinal cord M. Cifuentes 1, p. Fernfindez-LLebrez ~, J. P~rez 1, J.M. P6rez-Figares z, and E.M. Rodriguez 3 Departamentos de BiologiaAnimaly 2 BiologlaCelular de la Facultad de Cienciasde M/tlaga,Spain 3 Instituto de Histologiay Patologiade la Universidad Austral de Chile,Valdivia,Chile Received March 9, 1992 / AcceptedJuly 14, 1992
Summary. The circulation of the cerebrospinal fluid along the central canal and its access to the parenchyma of the spinal cord of the rat have been analyzed by injection of horseradish peroxidase (HRP) into the lateral ventricle. Peroxidase was found throughout the central canal 13 rain after injection, suggesting a rapid circulation of cerebrospinal fluid along the central canal of the rat spinal cord. It was cleared from the central canal within 2 h, in contrast with the situation in the brain tissue, where it remained in the periventricular areas for 4 h. In the central canal, HRP bound to Reissner's fiber and the luminal surface of the ependymal cells; it penetrated through the intercellular space of the ependymal lining, reached the subependymal neuropil, the basement membrane of local capillaries, and appeared in the lumen of endothelial pinocytotic vesicles. Furthermore, it accumulated in the labyrinths of the basement membrane contacting the basolateral aspect of the ependymal ceils. In ependymocytes, HRP was found in single pinocytotic vesicles. The blood vessels supplying the spinal cord were classified into two types. Type-A vessels penetrated the spinal cord laterally and dorsally and displayed the tracer along their external wall as far as the gray matter. Type-B vessels intruded into the spinal cord from the medial ventral sulcus and occupied the anterior commissure of the gray matter, approaching the central canal. They represented the only vessels marked by HRP along their course through the gray matter. HRP spread from the wall of type-B vessels, labeling the labyrinths, the intercellular space of the ependymal lining, and the lumen of the central canal. This suggests a communication between the central canal and the outer cerebrospinal fluid space, at the level of the medial ventral sulcus, via the intercellular spaces, the perivascular basement membrane and its labyrinthine extensions. Key words: Central canal
Cerebrospinal fluid Circulation - Horseradish peroxidase - Ependyma - Rat (Sprague Dawley)
Correspondence to: P. Fern/mdez-LLebrez,Departamento de Biologia Animal, Facultad de Ciencias, E-29071 Mfilaga,Spain
In 1965, Bradbury and Lathem investigated the flow of cerebrospinal fluid (CSF) along the central canal of lampreys, rats, guinea pigs, rabbits and monkeys by injecting tracer molecules into the lateral ventricle. Among these species, only the rabbit exhibited circulating CSF within the central canal. Since a circulation of CSF along the central canal was not observed in the other species, these authors considered the flow of CSF in the canal compartment of the rabbit as a '"relic of the vertebrate evolutionary story", apparently absent in most recent species. However, Nakayama and Kohno (1974) and Nakayama (1976) reported the presence of india ink in the central canal of guinea pigs, rabbits and rats from 30 rain to 2 h after injection into the lateral ventricle. To our knowledge, there are no other reports dealing with the flow of CSF along the central canal under normal conditions. Eisenberg et al. (1974a, 1974b) have reported a flow in the central canal of cats suffering from kaolin-induced hydrocephalus. The distribution of tracer molecules injected into the cerebral ventricles via the interstitial spaces of the brain has been thoroughly investigated (Rall etal. 1962; Brightman 1965; Wagner et al. 1974; Brightman et al. 1975; Matakas et al. 1978; Krisch et al. 1984; Rennels et al. 1985; Cserr 1988; see also Wood 1983; Davson et al. 1987). The tracers circulate within the CSF and fill the inner (ventricular) and outer (meningeal) spaces of the CSF. From the ventricles of the brain, the tracers penetrate the ependymal layer through the intercellular space and gain access to the interstitial space of the subependymal parenchyma (Matakas etal. 1978). The tracers may then reach the basement membrane of the brain capillaries where they are cleared by specific phagocytotic cells or are transported to the blood stream by transcytosis across the endothelium (Wagner et al. 1974). Many authors have found that, from the meningeal spaces, the tracers penetrate the brain parenchyma by diffusing along the intercellular spaces of the marginal glia (Wagner etal. 1974; Borison etal. 1980; Zervas et al. 1982; Rennels et al. 1985), and have postulated that there is open communication between the subarach-
486 n o i d space a n d the interstitial space o f the brain. Such a n o p e n c o m m u n i c a t i o n b e t w e e n the m e n i n g e a l a n d the intercellular spaces o f the b r a i n has been challenged by Krisch et al. (1984) who have p e r f o r m e d a t h o r o u g h ult r a s t r u c t u r a l s t u d y o f the d i s t r i b u t i o n , in the m e n i n g e a l sheaths a n d s u b p i a l b r a i n tissue, o f h o r s e r a d i s h peroxidase ( H R P ) injected i n t o the CSF. T h e d i s t r i b u t i o n o f tracer molecules r e a c h i n g the central c a n a l a n d the m e n ingeal space o f the spinal cord has, a p p a r e n t l y , n o t yet been investigated. O u r p r e s e n t research is m a i n l y c o n c e r n e d with the s t u d y o f the s u b c o m m i s s u r a l o r g a n - R e i s s n e r ' s fiber complex, a b r a i n g l a n d o f u n k n o w n f n n c t i o n ( R o d r i g u e z et al. 1992). O n e o f the w o r k i n g hypotheses p o s t u l a t e s t h a t the s u b c o m m i s s u r a l o r g a n participates in the control o f the c i r c u l a t i o n o f the C S F (see R o d r i g u e z et al. 1992). T h e p r e s e n t i n v e s t i g a t i o n was designed to study, by use o f H R P , the c i r c u l a t i o n o f the C S F a l o n g the central c a n a l o f the rat spinal cord u n d e r n o r m a l conditions.
Materials and methods
Animals Male Sprague-Dawley rats (body weight=250-300 g) were kept under a photoperiod of L:D 12:12 and an environmental temperature of about 25~ C. They were fed ad libitum with rodent food (A04, Panlab, Barcelona, Spain). The animals were anesthetized with ether, placed in a stereotactic instrument and perfused through the right lateral ventricle with one of the working solutions. Two groups of animals were prepared: 1) Control rats: the animals were perfused, by use of a pump, with 30 gl 0.9% NaC1 at a rate of 3 gl/min and sacrificed 20 min after perfusion ended. 2) Experimental rats: these were perfused with 30 gl 3% HRP (Sigma, Type IV, Dorset, England), dissolved in 0.9% NaC1, at a rate of 3 gl/min. The central nervous system of these rats was fixed 13 rain (n = 5), 20 rain (n = 8), I h (n = 5), 2 h (n = 3), and 4 h (n = 3) after starting the HRP perfusion.
ma, Dorset, England) and 0.03% H202 (Merck, Darmstadt, FRG) in TRIS-phosphate-buffered saline (0.05 M, pH 7.8). After being rinsed in the same buffer for 30 rain, the sections were mounted in glycerol or in Aquatex (Merck, Darmstadt, FRG) for light microscopy. Some sections were further processed for electron microscopy.
Electron microscopy After the Vibratome sections had been processed for the demonstration of peroxidase, some were washed in 0.1 M phosphate buffer, pH 7.4, and were fixed in 1% OsO4 in the same buffer for 1 h. After dehydration in a graded series of ethanol and acetone, the sections were embedded in Araldite. Semithin and ultrathin sections were obtained. The former were stained with toluidine blue. The ultrathin sections were stained with 1% lead citrate for I min.
Histochemical and immunocytochemical methods In order to investigate the nature of certain structures apparently related to the capillary wall, normal (untreated) rats were transcardially perfused with Bouin's fluid, and the central nervous system was dissected out and embedded in paraffin. Transverse and horizontal sections (10 gm thick) of the brain and spinal cord were stained with periodic acid-silver methenamine (Burkholder !974). Other sections were processed for immunocytochemistry for the demonstration of type-IV collagen. For this purpose, the sections were hydrated and treated for 12 h with 0.5% papain (Difco, Detroit, USA) in 0.02 M phosphate buffer, pH 4.7, containing 5 mM sodium bisulfite and 0.5 mM EDTA (Junqueira et al. 1980). The sections were then incubated for 18 h at 22~ C in an antiserum against type-IV collagen (Chemicon International, Temecula, Calif., USA) at a dilution of 1:500. The linking antibody (from E.M. Rodriguez, Valdivia, Chile) was used at a dilution 1 : 10 for 30 min at 22~ C and the PAP complex (1 : 75) (Sigma P2030, Dorset, England) for 30 rain at 22~ C. DAB was used as the electron donor. The antisera and the PAP complex were diluted in 0.1 M TRIS buffer, pH 7.8, containing 0.7% lambda carrageenan (Sigma, Dorset, England).
Semiquantitative studies Fixation Under ether anesthesia, the animals were transcardially perfused with 100 ml 0.1 M phosphate buffer, pH 7.4, for 10 rain. Then, 300 ml fixative (1% glutaraldehyde, 1% paraformaldehyde in 0.1 M phosphate buffer) were perfused for 30 rain. After perfusion, the brain and the entire spinal cord were dissected out. The spinal cord was divided into five fragments: 1) cervical, including cervical segments C1 to C6; 2)proximal thoracic, including thoracic segments TI to T4; 3) medial thoracic (T5-T8); 4) distal thoracic (T9T12); 5)lumbo-sacral,'includingsegments L1 to the filum terminale. The brain and all pieces of the spinal cord were left immersed in the same fixative overnight, at 4~ C.
Light microscopy Vibratome sections, approximately 40 gm thick, were obtained from all tissue blocks. The brain was cut transversally; each one of the five fragments into which the spinal cord was cut was, in turn, subdivided into two halves: the distal half was sectioned transversally and the proximal half was sectioned horizontally. Peroxidase was visualized according to the method of Graham and Karnovsky (1966), using 0.1% 3,3' diaminobenzidine (DAB, Sig-
An image analyzer IBAS-2000 (IPS) (Kontron, FRG) with a program IPS (image processing system) was used. By means of a video camera connected to a light microscope, the image of a frame containing the central canal was digitalized as a matrix of 100 x 100 points. A value, representing the relative optic density (ROD), as a gray scale from 0=white to 255=black, was assigned to each point. This value was automatically corrected with respect to the background density of an area of the section apparently devoid of peroxidase reaction product. At least three transverse sections from each of the five fragments of the spinal cord (see above) were used for the analysis. The value given to a section of the central canal represented the mean of the values corresponding to the 10000 points, with the exception of those points having a value lower than the background unstained tissue. The value for a given fragment of the spinal cord was the mean of the values recorded for the three sections; the value ascribed to an experimental group represented the mean of values estimated for corresponding segments in animals of such a group.
Statistical analysis The data of ROD obtained for the different experimental groups and segments of the spinal cord were submitted to the Kolmogor-
487 ov-Smirnov test (P
