Journal of Neuroscience Research 29:355-361 (1991)
Glutamine Stimulates Growth in Rat Cerebral Endothelial Cell Culture E. Dux, L. Noble, and P.H. Chan CNS Injury and Edema Research Center, Departments of Neurology (E.D., P.H.C.) and Neurosurgery (L.N., P.H.C.), University of California, San Francisco, School of Medicine, San Francisco, California 1980; Spatz et al., 1980; see also, for review, Joo, 1985; Gerritsen et al., 1988). It is evident from these studies that primary cultures of cerebral endothelial cells exhibit limited growth even if special conditioned media and growth factors were added (Folkman et al., 1979; Hobson and Denekamp, 1984; Gospodarowicz et al., 1986). With this limitation in mind, subcultures of cerebral endothelial cells have been established in order to increase the yield of cells (Meresse et al., 1989). Moreover, many investigators are convinced that the primary cultures have features identical to those present in cerebral endothelial cells (De Bault and Cancilla, 1980; Bowman et al., 1983; Hughes and Lantos, 1986). In the present study, we have avoided the potential problems associated with the use of subcultures of cerebral endothelial cells by focusing on improving the growth potential of primary cultures. For this reason we have examined the effect of glutamine on cell growth and development. Glutamine is produced by astrocytes via amidation of glutamate (Hosli and Hosli, 1976; Norenberg, 1979). Astrocytes not only maintain intimate contact with endothelium in vivo but also contribute to the expression of blood-brain barrier characteristics, includKey words: cerebral endothelial cells, primary cul- ing the induction of y-glutamyl-transpeptidase (De Bault ture, glutamine-y-glutamyl transpeptidase, blood- and CanciIla, 1980). It was therefore conceivable that brain barrier astrocytic products such as glutamine may influence endothelial cell growth and/or functions. INTRODUCTION
Endothelial cells were isolated from rat cerebral cortices using combined enzymatic digestions and Percoll gradient centrifugation. Primary cultures were subsequently grown on collagen-covered dishes in a medium containing 20% fetal calf serum and 0.6 mmol glutamine. The majority of cultures became confluent by day 7 or 8, but some could not reach confluence. The cells were fusiform in shape and exhibited immunoreactivity to factor VIII-related antigen and binding to the lectin Griffonia simplicifolia. Exposure of cultures to media containing 2.6 mmol glutamine resulted in accelerated growth (in cultures were confluent at days 3-4) and change in culture morphology, namely the formation of circular, cell-free areas. However, this treatment did not restore y-glutamyl transpeptidase activity that was lost during cultivation. As for other amino acids, asparagine was less potent, glycine and phenylalanine failed to mimic the glutamine effect. In summary, glutamine stimulates growth of cerebral endothelial cells in vitro and so it may supplement for other growth factors in the culture media.
The absence of glucose and oxygen pools in the central nervous system (CNS) makes the functioning neurons and glial cells dependent on their blood supply, ,which fact underlines the regulatory role of the bloodbrain barrier. Although the glial end-feet system plays an important role in the normal function of the barrier, its most fundamental functions are intrinsic to the microvascular endothelial cells. A number of reports have described the isolation of cerebral microvessels (Joo and Karnushina, 1973; Brendel et al., 1974; Goldstein et al., 1975) and the establishment of cultures of endothelial cells for the purpose of studying the blood-brain barrier in vitro (Bowman et al., 1981; DeBault et al., 1981; DeBault and Cancilla, 0 1991 Wiley-Liss, Inc.
MATERIALS AND METHODS Cerebral endothelial cells were isolated from 2week-old Sprague-Dawley rats of either sex using several modifications of the techniques originally described by Bowman et al. (1981) and Audus and Borchardt (1986). Briefly, for each experiment 20 brains were collected in phosphate-buffered saline (PBS) (without cal-
Received September 21, 1990; revised December 3 , 1990; accepted December 3 , 1990. Address reprint requests to Dr. Pak H. Chan, CNS Injury and Edema Research Center, Department of Neurology, University of California, San Francisco, CA 94143-01 14.
3
Glutamine Effects on Endothelial Culture
cium and magnesium, pH 7.4). After the meninges and white matter were removed, the cortex was minced and incubated in MI99 medium with Earle’s salt containing 0.37% collagenase (type CLS2, Worthington, NJ, USA), 0.1% bovine serum albumin (BSA), gentamycin (50 pg/ml), and fungizone (2.5 pg/ml) at 37°C for 2 h. After this first enzymatic digestion, 25% BSA in PBS was added to the suspension and centrifuged at 1,OOOg for 20 min. The washed pellet was resuspended and further incubated in the presence of 0. l % collagenase-dispase (Boehringer Mannheim, Germany) in MI99 medium with antibiotics for additional 2 h. After centrifugation at 150g for 5 min, the pellet was suspended in culture medium and layered on a 35% continuous Percoll gradient. After a 10-min centrifugation at 1,OOOg in a swing-out rotor (HSA 13.94, Hereaus, Germany), the endothelial cells were carefully collected from the gradient above the visible layer, which contained red blood cells. The cells were then thoroughly washed three times in M199 tissue culture medium (containing 0.6 mmol glutamine) and subsequently seeded onto 20 collagen-coated plastic 35-mm dishes (200 pg/ dish) (type I from lathyritic rat skin, Boehringer Mannheim, Germany). The culture medium was supplemented with 20% fetal calf serum (representing not more than 20 pmol glutamine, the analysis was performed by HyClone Lab Inc., Logan, UT) and 20 mmol HEPES . The cells were cultivated in 5% CO, atm at 37°C. In one group of dishes, an additional 2 mmol glutamine was given to the medium at the initial plating and at each feeding. In other dishes, asparagine and ammonium chloride were given, respectively, in 2-mmol as well as glycine or phenylalanine in 1 mmol concentrations, from the very beginning. The medium was routinely changed every 3 days, beginning at the third day of cultivation. The morphology of cultivated cells was examined at both light and electron microscopic levels. In addition immunostaining for factor VIII-related antigen (antibody was purchased from Dakopatts, Santa Barbara, CA) and binding of the lectin Grzffonia simplicifolia (GSL) (Dako, Carpinteria, CA) with the use of the avidin-biotin kit (Vector Lab, Burlingame, CA) was evaluated. Fig. 1 . Seven-day-old control (0.6 mmol glutamine in the medium) endothelial cell culture from rat cerebral cortex. a. Phase-contrast micrograph. Bar: SO pm. b. Factor VIII immunoreactivity in the endothelial cytoplasm. Bar: 25 pm. c. Grqfonia simplicifolia lectin (GSL) binding to the cellular membrane. Bar: 25 pm. d. Electron histochemical visualization of GSL binding. M , mitochondria. Arrows point toward membrane zones without GSL reactivity. (These sites may represent membrane specializations between adjacent cells). Bar: 0.5 Pm.
357
TABLE I. Specific Activity of y-Glutamyl Transpeptidase Enzymet Uimg Protein Cortical homogenate Capillary endothelial cells Endothelial cells in culture (7-day old) Control (0.6 mmol glutamine) 2.6 mmol glutamine
5.1 2 1.2 128.5 ? 9.7* 14.7 ? 12.8 10.9 2 12.2
*
‘Mean SD; n = 4. *Significant difference from the cortical homogenate (P c0.05).
During the endothelial cell preparation as well as on the 7th day of cultivation, the specific activity of y-glutamyl transpeptidase was determined using the Sigma diagnostic kit (No. 545). The activity of y-glutamyl transpeptidase in the isolated capillaries and in cultivated endothelial cells was also visualized histochemically using the methods of DeBault and Cancilla ( 1980). Protein determination was carried out according to Lowry et al. (1951). To assess the effects of glutamine, asparagine, glycine, phenylalanine, and ammonium chloride, morphometric evaluation was carried out in the cultures after 12 and 48 h, i.e., the numbers of colonies, cells, and the bridges formed between the adjacent colonies were counted in 1-mm2 area in 20 randomly selected places of 10 dishes from each group. Statistical evaluation was made by Student’s t-test.
RESULTS During the procedure of capillary endothelial cell preparation, the specific activity of y-glutamyl transpeptidase increased more than twentyfold in the endothelial cells compared with the initial cortical homogenate. By contrast, after 7 days of cultivation, the endothelial cells lost their enzymatic activity measured either biochemically (Table I) or histochemically (not shown). The morphologic appearance of cultivated endothelial cells was characteristic, i.e., continuous monolayer of homogeneous, nonoverlapping cells. Their cytoplasm showed reactivity to factor VIII-related antigen and GSL labelled mostly the cell membranes (Fig. 1). Under normal conditions, the endothelial cells started growing on the first day, then formed colonies of characteristic cobblestone appearance and ultimately reaching confluence not earlier than on the 7th-8th day, similar to those treated with 2 mmol asparagine. Some cultures (about 20-30% of controls) remained at subconfluent state even after 10 days. When additional 2 mmol glutamine was added to the medium, cell growth was significantly stimulated. The cells started growing within 8 h, and as early as 12 h numerous colonies of many cells
Fig. 2 .
Glutamine Effects on Endothelial Culture
359
On the 7th day, the endothelial cells were harvested by trypsin-EDTA in PBS (without calcium and magneCell Bridge sium) and resuspended in culture medium containing trypsin inhibitor (Boehringer Mannheim, Germany). The 64.6 ? 27 2.7 t 2.1 secondary cultures of glutamine treated cells did not show any inhibition in growth or development compared 264.7 f 76.7* 27.0 ? 9.5% with the controls. However, in this study, we only report 157.3 ? 17.9* 6.2 ? 4.4 the results using primary endothelial cultures. 37.9 2 22.0 1.3 ? 1.7
TABLE 11. Number of Colonies, Cells, and Bridges in 1 mm2 of Rat Cerebral Capillary Endothelial Cells After 12 ht Colony
Control (0.6 mmol Gln) plus 2 mmol Gln 2 mmol Asn 1 mmol Gly 1 mmol Phe 2 mmol NH,Cl
20.5 t 7.8 42.2 37.7 16.8 18.7 21.3
?
11.3%
2 16.5
t 10.7 t 12.3 t 10.2
31.8 t 12.4 75.2 2 16.3
*
1.4 1.2 3.3 t 1.9
‘Gln, glutamine; Asn, asparagine; Gly , glycine; Phe, phenylalanine; NH,Cl, ammonium chloride. Mean t SD of 20 area of 10 dishes. *Significant difference from controls (P
Glutamine Stimulates Growth in Rat Cerebral Endothelial Cell Culture E. Dux, L. Noble, and P.H. Chan CNS Injury and Edema Research Center, Departments of Neurology (E.D., P.H.C.) and Neurosurgery (L.N., P.H.C.), University of California, San Francisco, School of Medicine, San Francisco, California 1980; Spatz et al., 1980; see also, for review, Joo, 1985; Gerritsen et al., 1988). It is evident from these studies that primary cultures of cerebral endothelial cells exhibit limited growth even if special conditioned media and growth factors were added (Folkman et al., 1979; Hobson and Denekamp, 1984; Gospodarowicz et al., 1986). With this limitation in mind, subcultures of cerebral endothelial cells have been established in order to increase the yield of cells (Meresse et al., 1989). Moreover, many investigators are convinced that the primary cultures have features identical to those present in cerebral endothelial cells (De Bault and Cancilla, 1980; Bowman et al., 1983; Hughes and Lantos, 1986). In the present study, we have avoided the potential problems associated with the use of subcultures of cerebral endothelial cells by focusing on improving the growth potential of primary cultures. For this reason we have examined the effect of glutamine on cell growth and development. Glutamine is produced by astrocytes via amidation of glutamate (Hosli and Hosli, 1976; Norenberg, 1979). Astrocytes not only maintain intimate contact with endothelium in vivo but also contribute to the expression of blood-brain barrier characteristics, includKey words: cerebral endothelial cells, primary cul- ing the induction of y-glutamyl-transpeptidase (De Bault ture, glutamine-y-glutamyl transpeptidase, blood- and CanciIla, 1980). It was therefore conceivable that brain barrier astrocytic products such as glutamine may influence endothelial cell growth and/or functions. INTRODUCTION
Endothelial cells were isolated from rat cerebral cortices using combined enzymatic digestions and Percoll gradient centrifugation. Primary cultures were subsequently grown on collagen-covered dishes in a medium containing 20% fetal calf serum and 0.6 mmol glutamine. The majority of cultures became confluent by day 7 or 8, but some could not reach confluence. The cells were fusiform in shape and exhibited immunoreactivity to factor VIII-related antigen and binding to the lectin Griffonia simplicifolia. Exposure of cultures to media containing 2.6 mmol glutamine resulted in accelerated growth (in cultures were confluent at days 3-4) and change in culture morphology, namely the formation of circular, cell-free areas. However, this treatment did not restore y-glutamyl transpeptidase activity that was lost during cultivation. As for other amino acids, asparagine was less potent, glycine and phenylalanine failed to mimic the glutamine effect. In summary, glutamine stimulates growth of cerebral endothelial cells in vitro and so it may supplement for other growth factors in the culture media.
The absence of glucose and oxygen pools in the central nervous system (CNS) makes the functioning neurons and glial cells dependent on their blood supply, ,which fact underlines the regulatory role of the bloodbrain barrier. Although the glial end-feet system plays an important role in the normal function of the barrier, its most fundamental functions are intrinsic to the microvascular endothelial cells. A number of reports have described the isolation of cerebral microvessels (Joo and Karnushina, 1973; Brendel et al., 1974; Goldstein et al., 1975) and the establishment of cultures of endothelial cells for the purpose of studying the blood-brain barrier in vitro (Bowman et al., 1981; DeBault et al., 1981; DeBault and Cancilla, 0 1991 Wiley-Liss, Inc.
MATERIALS AND METHODS Cerebral endothelial cells were isolated from 2week-old Sprague-Dawley rats of either sex using several modifications of the techniques originally described by Bowman et al. (1981) and Audus and Borchardt (1986). Briefly, for each experiment 20 brains were collected in phosphate-buffered saline (PBS) (without cal-
Received September 21, 1990; revised December 3 , 1990; accepted December 3 , 1990. Address reprint requests to Dr. Pak H. Chan, CNS Injury and Edema Research Center, Department of Neurology, University of California, San Francisco, CA 94143-01 14.
3
Glutamine Effects on Endothelial Culture
cium and magnesium, pH 7.4). After the meninges and white matter were removed, the cortex was minced and incubated in MI99 medium with Earle’s salt containing 0.37% collagenase (type CLS2, Worthington, NJ, USA), 0.1% bovine serum albumin (BSA), gentamycin (50 pg/ml), and fungizone (2.5 pg/ml) at 37°C for 2 h. After this first enzymatic digestion, 25% BSA in PBS was added to the suspension and centrifuged at 1,OOOg for 20 min. The washed pellet was resuspended and further incubated in the presence of 0. l % collagenase-dispase (Boehringer Mannheim, Germany) in MI99 medium with antibiotics for additional 2 h. After centrifugation at 150g for 5 min, the pellet was suspended in culture medium and layered on a 35% continuous Percoll gradient. After a 10-min centrifugation at 1,OOOg in a swing-out rotor (HSA 13.94, Hereaus, Germany), the endothelial cells were carefully collected from the gradient above the visible layer, which contained red blood cells. The cells were then thoroughly washed three times in M199 tissue culture medium (containing 0.6 mmol glutamine) and subsequently seeded onto 20 collagen-coated plastic 35-mm dishes (200 pg/ dish) (type I from lathyritic rat skin, Boehringer Mannheim, Germany). The culture medium was supplemented with 20% fetal calf serum (representing not more than 20 pmol glutamine, the analysis was performed by HyClone Lab Inc., Logan, UT) and 20 mmol HEPES . The cells were cultivated in 5% CO, atm at 37°C. In one group of dishes, an additional 2 mmol glutamine was given to the medium at the initial plating and at each feeding. In other dishes, asparagine and ammonium chloride were given, respectively, in 2-mmol as well as glycine or phenylalanine in 1 mmol concentrations, from the very beginning. The medium was routinely changed every 3 days, beginning at the third day of cultivation. The morphology of cultivated cells was examined at both light and electron microscopic levels. In addition immunostaining for factor VIII-related antigen (antibody was purchased from Dakopatts, Santa Barbara, CA) and binding of the lectin Grzffonia simplicifolia (GSL) (Dako, Carpinteria, CA) with the use of the avidin-biotin kit (Vector Lab, Burlingame, CA) was evaluated. Fig. 1 . Seven-day-old control (0.6 mmol glutamine in the medium) endothelial cell culture from rat cerebral cortex. a. Phase-contrast micrograph. Bar: SO pm. b. Factor VIII immunoreactivity in the endothelial cytoplasm. Bar: 25 pm. c. Grqfonia simplicifolia lectin (GSL) binding to the cellular membrane. Bar: 25 pm. d. Electron histochemical visualization of GSL binding. M , mitochondria. Arrows point toward membrane zones without GSL reactivity. (These sites may represent membrane specializations between adjacent cells). Bar: 0.5 Pm.
357
TABLE I. Specific Activity of y-Glutamyl Transpeptidase Enzymet Uimg Protein Cortical homogenate Capillary endothelial cells Endothelial cells in culture (7-day old) Control (0.6 mmol glutamine) 2.6 mmol glutamine
5.1 2 1.2 128.5 ? 9.7* 14.7 ? 12.8 10.9 2 12.2
*
‘Mean SD; n = 4. *Significant difference from the cortical homogenate (P c0.05).
During the endothelial cell preparation as well as on the 7th day of cultivation, the specific activity of y-glutamyl transpeptidase was determined using the Sigma diagnostic kit (No. 545). The activity of y-glutamyl transpeptidase in the isolated capillaries and in cultivated endothelial cells was also visualized histochemically using the methods of DeBault and Cancilla ( 1980). Protein determination was carried out according to Lowry et al. (1951). To assess the effects of glutamine, asparagine, glycine, phenylalanine, and ammonium chloride, morphometric evaluation was carried out in the cultures after 12 and 48 h, i.e., the numbers of colonies, cells, and the bridges formed between the adjacent colonies were counted in 1-mm2 area in 20 randomly selected places of 10 dishes from each group. Statistical evaluation was made by Student’s t-test.
RESULTS During the procedure of capillary endothelial cell preparation, the specific activity of y-glutamyl transpeptidase increased more than twentyfold in the endothelial cells compared with the initial cortical homogenate. By contrast, after 7 days of cultivation, the endothelial cells lost their enzymatic activity measured either biochemically (Table I) or histochemically (not shown). The morphologic appearance of cultivated endothelial cells was characteristic, i.e., continuous monolayer of homogeneous, nonoverlapping cells. Their cytoplasm showed reactivity to factor VIII-related antigen and GSL labelled mostly the cell membranes (Fig. 1). Under normal conditions, the endothelial cells started growing on the first day, then formed colonies of characteristic cobblestone appearance and ultimately reaching confluence not earlier than on the 7th-8th day, similar to those treated with 2 mmol asparagine. Some cultures (about 20-30% of controls) remained at subconfluent state even after 10 days. When additional 2 mmol glutamine was added to the medium, cell growth was significantly stimulated. The cells started growing within 8 h, and as early as 12 h numerous colonies of many cells
Fig. 2 .
Glutamine Effects on Endothelial Culture
359
On the 7th day, the endothelial cells were harvested by trypsin-EDTA in PBS (without calcium and magneCell Bridge sium) and resuspended in culture medium containing trypsin inhibitor (Boehringer Mannheim, Germany). The 64.6 ? 27 2.7 t 2.1 secondary cultures of glutamine treated cells did not show any inhibition in growth or development compared 264.7 f 76.7* 27.0 ? 9.5% with the controls. However, in this study, we only report 157.3 ? 17.9* 6.2 ? 4.4 the results using primary endothelial cultures. 37.9 2 22.0 1.3 ? 1.7
TABLE 11. Number of Colonies, Cells, and Bridges in 1 mm2 of Rat Cerebral Capillary Endothelial Cells After 12 ht Colony
Control (0.6 mmol Gln) plus 2 mmol Gln 2 mmol Asn 1 mmol Gly 1 mmol Phe 2 mmol NH,Cl
20.5 t 7.8 42.2 37.7 16.8 18.7 21.3
?
11.3%
2 16.5
t 10.7 t 12.3 t 10.2
31.8 t 12.4 75.2 2 16.3
*
1.4 1.2 3.3 t 1.9
‘Gln, glutamine; Asn, asparagine; Gly , glycine; Phe, phenylalanine; NH,Cl, ammonium chloride. Mean t SD of 20 area of 10 dishes. *Significant difference from controls (P
