- Email: [email protected]
Postnatal development of γ-GT activity in rat brain microvessels corresponds to capillary growth and differentiation
III/. J. Ikvl.
V<,I. Nc,rr~r,.\i~ic,rrce.
J. No. 0. pp. 5O.%SI I,
IWh.
0736-S7JXIXh $03.cnl+o.o0 Pcrgamon Journals Ltd. @ 19X6 ISDN
Pnnrcd in Great Britain.
POSTNATAL DEVELOPMENT OF y-GT ACTIVITY IN RAT BRAIN MICROVESSELS CORRESPONDS TO CAPILLARY GROWTH AND DIFFERENTIATION A. Max-Planck-lnstitut fiir Systemphysiologie, Anatomie und Entwicklungsneurobiologie,
W.
BUDI
SANTOSO
and TH.
BAR*
Rheinlanddamm 201, D-4600 Dortmund Georg-August Universittit. Kreuzberger (Accepfed
I. F.R.G.; *Abteilung fiir Klinische Ring 36, D-3400 GGttingen, F.R.G.
I4 July 1986)
Abstract-It has been demonstrated histochemically that endothelial cells of the cerebral capillaries show a high activity in y-glutamyl transpeptidase (y-CT), an enzyme which takes part in the transfer of large neutral amino acids across the blood-brain barrier. Reports of the disappearance of enzyme activity in endothelial cells (EC) grown in culture” suggest that the presence of astroglial cells (AC) is required for the expression of y-GT in these cells. The present study deals with the developmental changes in ?I-GT activity in capillaries of rat cerebral cortex during ontogenesis (i.e. on days 2,7. I I, 14, 21 and 60 after birth). y-GT activity is determined by measuring the enzyme kinetics of the histochemical reaction on isolated brain capillaries using a flying spot microscope densitometer. Enzyme activity is expressed as an increase in relative optical density (at SO0 nm) in arbitrary units/min/pm” during the 2 min immediately after initiating incubation and corresponds to the y-CT active area (in pm”) of the capillary segment. During early postnatal development, a biphasic change of y-CT activity in capillaries of rat cerebral cortex is observed. The first phase (i.e. the postnatal period between the 2nd and 12th day) is characterized by a significant decrease in y-GT activity, which coincides with the onset of the rapid mitotic proliferation of cortical endothelial cells. During the second phase (i.e. the postnatal period between the 12th and 2lst day), a fast increase in the y-CT activity can be measured. Then enzyme activity reaches the adult level between the 2lst and 60th postnatal day. With regard to changes in capillary length density and in astroglial proliferation the results correspond to the differentiation and maturation of the blood-brain barrier and confirm evidence stressing the important role of astroglial cells for the expression of y-CT activity. Key words: Isolated microvessels, Microdensitometric measurements,
Rat brain cortex, y-GT activity, Ontogenesis, Endothelial cells.
Semiquantitative
histochemistry,
y-Glutamyl transpeptidase (y-GT) plays a key role in the y-glutamyl cycle; it also catalyzes transfer of the y-glutamyl (-y-GT) residue of intracellular glutathione, S-substituted glutathion derivatives, and other y-glutamyl compounds to a number of acceptors forming y-glutamyl peptides. “‘J y-GT is a membrane-bound heterodimeric glycoprotein. “J It has also been considered to metabolize leucotrien C, to D_,.Zh This enzyme is widely distributed in the various tissues of mammals involved in secretory and transport mechanisms, such as in the brush border of the kidney proximal tubules,y,32,“7 in the lactating mammary glands,‘,“” in biliary duct epithelial cells,2” in cornea1 endothelial cells,” in small intestinal mucosa,‘” in colon epithelial cells,3” and in lymphocytes.“” The enzyme’s activity varies within the pre- and postnatal development stages and in different organs.2.“,‘“.“‘,‘h At the dedifferentiation stage, y-GT activity is at the level usually found in early embryonic development of the particular organ. Thus, the enzyme has been reported to be a useful marker enzyme for preneoplastic and neoplastic foci in the liver after carcinogen application, ‘. ‘0-34 and during carcinogenesis of the respiratory tract epithelium.“’ Orlowski et ~1.~’ found, using histochemical methods, that y-GT is associated with the endothelial cells of isolated brain capillaries, and suggested y-GT functions to transfer large neutral amino acids across the blood-brain barrier. -y-GT is localized predominantly in the capillaries of the brain,14 and is immunocytochemically detectable in high concentration on the luminal surface of the cell membrane of cerebellar endothelial cells. Ix As suggested by DeBault and Cancilla,‘2 the presence of adjacent processes of astroglia (AG) is required for continuous induction of y-GT. The enzyme usually disappears when the endothelial cells migrate or proliferate from the isolated microvessel. It has been recognized that coculturing mouse endothelial cells with Ch rat glioma cells can induce the re-expression of y-GT activity in endothelial cells. “.I2 -y-GT activity, in cultured brain capillary endothelium, increases
A. W. Rudi Santoso and Th. Biir
so3
significantly after in vitro transformation of these cells by an avian retrovirus. Activity declines to a low level by passage 80 coinciding with the disappearance of F VIII and angiotensin converting enzyme (differentiation markers for brain endothelial cells).’ The association between the differentiation of endothelial cells and the presence of y-GT activity prompted us to examine the developmental changes of y-GT activity on isolated brain capillaries at different ontogenetic stages. The postnatal period was selected since, according to and ultrastructural changes occur in neocortithe detailed study of Bar,4 obvious morphometric cal capillaries during that period. It was shown that a phase of vigorous proliferation of new capillaries, resulting in the most dramatic increase in capillary length, was followed by the functional maturation of the capillary bed. Difficulties in detecting low levels of y-GT activity with common biochemical methods, required that a micromethod for semiquantitative histochemical measurement of y-GT activity can be used together with a microscope densitometer. This method offers some advantages, because the capillary fragments, which are expected to exhibit y-GT activity, can be selectively measured for their enzyme kinetic properties. By means of the newly developed method, we demonstrated a significant decrease in y-GT activity in brain capillaries during the period of rapid mitotic proliferation of endothelial cells in the first postnatal week. After that period, y-GT activity increases to adult levels at day 60 postnatal. Possible implications of the results are discussed with regard to changes in the capillary length density and in astroglial proliferation. EXPERIMENTAL
PROCEDURES
Isolation of the rat brain microvessels
The cortical microvessel preparations performed were slightly modified according to the method described by Diglio et a1.13 using Sprague-Dawley rats (supplier: Zentrales Institut fur Versuchstierkunde, Hannover, F.R.G.). For the ontogenic study, groups of animals at different postnatalages,i.e.2days(n=6).7days(n=7),11days(n=10),14days(n=10),16days(n=19), 21 days (n = 6), and 60 days (II = 6) were taken. Under pentobarbital anesthesia (Ncmbutal’“‘i.p., 40 mg/kg body wt). the animals were perfused transcardially with ice-cold perfusion medium (5 mM MgCl:. 20 mM HEPES. pH 7.2, in 500 ml IO times diluted Hank’s balanced salt solution). The perfusion was carried out by using a 20 ml syringe under gentle pressure. After quick removal of the brain from the skull, the cerebral cortex was freed from the meninges. the brain stem, and the white matter. Pieces of the cerebral cortex were collected in a beaker containing ice-cold homogenization medium (5 mM MgCl*, 20 mM HEPES, 20 mM Tris, 2 mM EDTA, pH 7.2. in 500 ml Hank’s balanced salt solution). The material was chopped into small pieces and divided in two equal parts. Homogenization was performed in a Teflon-glass homogenizer (X IO. 350 rpm) with 20 ml homogenization medium. After adding 40 ml of the homogenization medium, homogenization was repeated. Centrifugation was performed for IO min (1000 g). The sediment was resuspended in a medium (40 ml) containing 5 mM MgC12, 5% fetal calf serum (FCS), 20 mM HEPES and 15% dextran in x IO diluted Hank’s balanced salt solution. To obtain a crude pellet of the brain microvessels, the suspension was centrifuged again (4OOOg, 20 min). The crude pellet was resuspended in 40 ml x IO diluted Hank’s balanced salt solution containing 5 mM MgC17, 20 mM HEPES and 5% fetal calf serum (washing medium). The suspension was washed thoroughly through a glass column containing glass beads (diameter 0.25%0.3 mm) with approximately 200-300 ml washing medium. Thereafter, the microvessel-rich medium, above the glass beads column, was collected by gentle pipetting. To enrich the microvessels, a final centrifugation was performed (5 min, 1000 g). The microvessel fractions were spread onto glass slides (precoated with 3% gelatin which was stabilized in 2% Na2S03 for 1 hr before use), air-dried and fixed and defatted in 100% aceton (-4O”C, 12 hr), 100% aceton (4”C, I hr), 2:l (v/v) aceton/chloroform (4°C. I hr). 2:l (v/v) chloroform/aceton (4”C, 1 hr) and finally in 100% chloroform (4°C. 1 hr) according to Chang and Hori.” The specimens were stored at -40°C prior to examination. Histochemical
demonstration
of‘ y-GT
activity
Demonstration of y-GT activity was performed by using an azo dye coupling method first described by Rutenburg et al.” The y-glutamyl residue of the y-L-glutamyl-4-methoxy-2-naphthyla-
y&T
activity in rat brain microvessels
505
mide (as a substrate) is transferred by y-GT to the acceptor glycylglycine. In a simultaneous azo coupling reaction with Fastblue BB. a red azo dye is developed. To obtain a micromethod for the quantitative measurement of the enzyme kinetic, using a flying spot microscope densitometer, the method was modified as demonstrated
in Fig.
I,
Preparation of the stock solutions 8.27 mM y-glutamyl-4-methoxy-2-naphthylamide (y-Glu-MNA) 5 mg yGlu-MNA dissolved in 0.1 ml dimethylsulfoxide (DMSO), 0.1 ml 1 N NaOH and 1.8 ml bidest . HzO. 12.0 mM Fastblue BB Salt 25 mg Fastblue BB Salt dissolved in 10 ml bidest. HZO. 300 mM Glycylglycine (Gly-Gly) 100 mg Gly-Gly dissolved in 2 ml 0.1 M Tris-HCI buffer, pH 7.4. 145 mM NaCl 0.85 g dissolved in 100 ml bidest. H20. 0.4 M Tris-HCI buffer (diluted 4 times before use to obtain 0.1 M buffer solution) 25 ml 0.4 M Tris (5.61 g/100 ml bidest. H,O)+33.5 ml bidest. Hz0 adjusted with 42.5 ml 0.2 N HCI to pH 7.4. 0.1 M CuSO, 2.5 g/l00 bidest. H20. ISOLATED
0.200ml lL5mM
NaCl
MICROVESSELS
p-7
A-1
u
BRAIN
O.OZOmt
SPACER
WIRE
i@=200@i1
300 mM Gly -Gly
O.lMTris-HCI
0 ZOOmI
DH8.0 8.27mM
Y-GiuMNA
12 OOmM
Fastblue.
MlCRODENSlTOMETRlC
BB
MEASUREMENT
I~UBATION SOLUTION
Fig. I. Schematic diagram showing the preparation of the incubation medium and the isolation of microvessels. 0.2 ml of the incubation medium is dropped on a sample of isolated microvessels before starting the enzyme kinetic measurements.
Microdensitometric measurement of the y-GT activity To obtain quantitative data on enzyme kinetics of y-GT we used a flying spot microdensitometer (Type M 85 A, Vickers Instr. Ltd., London, U.K.) and a computer-assisted system (Videoplan, Kontron Ltd., Munich, F.R.G.), linked by a BCD interface and compatible special measuring software (for details see Ref. 22). Figure 1 schematically shows the general procedure including the preparation of the incubation solution and the isolated brain microvessel specimens demonstrated prior to microdensitometric measurement. The incubation temperature was kept constant (37°C) during the measurements, by perfusing a small chamber of the microdensitometric stage between condenser lens and microscopic slide with a constant flow of water at 37°C from a thermostat (Type KARD, Mel3gertite Werk-Lauda, F.R.G.). Before starting the incubation, an appropriate capillary segment was seIected using differential contrast illumination. Microdensitometric measurements (i.e. registration of the relative optical density caused by development of the reaction product) is started
SO6
A. W.
470
Budi
Santoso
and Th
510 490 WAVELENGTH
530 Inml
Biir
550
570
Fig. 2. Curve showing the absorption maximum of the red azo dye developed as reaction product during the y-GT histochemical reaction at the wavelength getween 470 and 570 nm. The enzyme-kinetic measurements are done at the optimal absorption maximum (SO0 nm).
immediately after applying 0.2 ml of incubation solution. The measurements were performed at a wavelength of 500 nm (i.e. at the absorption maximum of the reaction product, Fig. 2). The changes of the relative optical density (expressed in arbitrary units; AOD,,,,,) were measured at 3 set intervals controlled by the computer. During the 10 min after the start of incubation the values of the optical density (AODsoo) were stored automatically in the computer. Finally, the area of the capillary segment displaying the reaction product (monitored by means of a video planimetrically using a digitizing tablet camera Sanyo, Model No. VC x 1600) was measured (Fig. 3a). The stored densitometric values obtained during the histochemical reactions were plotted against time. To evaluate the y-GT activity, the slope of the regression lines within the first 2 min of the reaction was determined (i.e. within the initial linear part of the kinetic curve); the value represents AOD,,,,,/min. To obtain the relative specific activity of the y-GT, this value was related to the morphomerically measured reactive area and expressed in AODs,,,,lp,m2/min. The following instrument settings were used for this study: objective x40. spot size 2. guting mask A4. scanning time 2 set, band width 50. wavelength 500 nm. For the histochemical localization of y-GT activity the microvessels were incubated at 37°C in a moist chamber for 30 min. The reaction was stopped by addition of 0. I ml 0.1 M CuSO,. the slides were washed several times in phosphate-buffered saline solution and embedded in glycerol gelatin (Sigma Chem., Munich, F.R.G.), and examined immediately under the light microscope. Control incubations are done using, (1) substrate free incubation, (2) incubation with the incubation medium containing 0.2 ml 200 mM serine (dissolved in 0.01 M Tris-HCI buffer, pH 6.9)/0.2 ml 0.05 M borate buffer solution (pH 7.6) added to 1 ml incubation medium.
RESULTS Histochemicul
localization
of y-glutamyl
transpeptiduse
The activity of y-GT detected in isolated brain cortical microvessels was restricted exclusively to the capillary segments (Fig. 3b). Pre- and postcapillary vessels were not stained. Observation under higher magnification revealed that the reaction produced was mostly localized at the luminal side of the endothelial cells (Fig. 4). The brain capillary fragments from 2-day-old rats showed large rounded nuclei lying close together (Fig. 5a). After incubation with y-GT, the capillary endothelial cells were stained intensively. The nuclei, however, showed only a weak staining (Fig. 4). In the capillary fragments from 7-day-old animals, a distinct reduction of y-GT activity was observed. The nuclei were densely packed. They remained large, and round in shape (Fig. 5b). On day 15, the maximal inhomogeneity of staining among single capillary segments was found. From 16 days on and in the older age groups an increased intensity of y-GT staining could be observed in the cortical capillary
y-GT
activity in rat brain microvessels
Fig. 3. (a) Capillary segment as seen on the monitor at the end of the measurement, 10 min after the start of the enzyme reaction for y-GT activity. The area on which the development of the reaction product can be detected, is circumscribed for the morphometric measurement. The area of the endothelial cells encircled by the gating mask is used to estimate for calculation of the specific activity of y-GT. (b) y-GT activity is demonstrated predominantly in the endothelial cells of capillaries within the microvessel fraction isolated from adult rat cortex (linear magnification: X40@.
507
508
A. W. Budi Santoso
and Th. Btiir
Fig. 4. y-GT activity in brain capillary fragment at higher magnification. intraluminal localization of the reaction product (Interference-Contrast magnification: X 1600).
Arrows demonstrate the after Nomarski; linear
Fig. 5. (ax) Demonstration of -r-GT activity from rat brain cortical capillaries isolated at different postnatal ages. The specimens were incubated with the y-CT incubation medium at 37°C for 30 min. Microphotography was performed by identical exposure time and at the same magnification (linear magnification: X250).
y-GT
activity
in rat
brain microvcsscls
50~
segments. Capillaries isolated 60 days after birth showed the most intensive staining of y-GT activity (Fig. 5c, d, e). When 0.2 ml 200 mM serine (in 0.01 M Tris-HCI, pH 6.9)/0.2 ml 0.05 M borate buffer solution (pH 7.6) were added to 1 ml incubation medium, no activity (or staining) was detected in any of the specimens. Additional examination of rat brain cryosections revealed that the staining was also restricted to the capillary segments (result not shown). Quantitative
analysis
of the histochemical
y-CT
activity
On day 2, the relative specific activity of y-GT in capillary segments of the brain cortex was higher than in similar fragments isolated from the brain of 7- and IO-day-old animals. In the second postnatal week y-GT activity was obviously decreased. The lowest activity was found on postnatal day 11. If the mean value of y-GT activity in capillary segments obtained from 60-day-old rat brains was regarded as lOO%, the relative specific activity of y-GT at postnatal days 7, 11 and 16 varied between 33 and 48% (Fig. 6). In early postnatal development, a biphasic change of y-GT activity in capillaries of rat cerebral cortex was apparent. The first phase (i.e. at the postnatal period between the 2nd and 12th day) was characterized by a significant decrease of y-GT activity, whereas during the following second phase (i.e. the postnatal period between the 12th and 21st day) a fast increase of y-GT activity could be measured. Later on, only a slight increase in y-CT activity was observed before the curve of activity reached its plateau approximately on day 60 (Fig. 6).
AGE
[DAYS
AFTER
BIRTH1
Fig. 6. Relative specific activity of y-GT in isolated brain cortical capillaries as measured by the microscope densitometer (each symbol represents the mean values r?rS.E.M.; n = 10 measurements). *P
DISCUSSION Our results demonstrate that yGT activity of capillary endotheliai cells is closely related to the developmental changes of the brain vascular system. The semiquantitative microdensitometric approach offers the possibility of comparing differences in the enzyme kinetics of y-GT in isolated brain capillaries at various ages after birth. Our method seems a suitable tool for studies of other enzymes, especially in isolated or cultured cells, even if only a small amount of cells is available for investigation. The vascularization of the rat brain cortex during the postnatal period represents a suitable model for studies of vascular growth and differentiation, because most major changes occur predominantly after birth. Additional knowledge is availabie concerning timing and intensity of endothelial and astroglial proliferation as well as the postnatal growth of the astroglial processes and lamellae covering the intracerebral vasculature.“~2’.2x
510
A. W. Budi Santoso and Th. BLir
High y-GT activity in the brain cortex is exhibited exclusively at the capillary level. In isolated microvessels the enzyme activity is 35 times higher than in the crude brain hom0genate.l’ Shortly after birth, y-CT activity decreases significantly. The diminished activity coincides with the onset of the rapid mitotic proliferation and results in a dramatic increase in capillary length density.’ During the rapid increase in capillary length, the highest numerical density of capillary sprouts is found on day 8 after birth.28 At this stage, the ?I-GT activity in the brain capillaries remains low. We suggest a state of low differentiation of the membranes of endothelial cells during the rapid mitotic proliferation. Szamei and Resch’3 demonstrated a modulation of y-GT activity in lymphocytic plasma membranes by incorporation of long-chain unsaturated fatty acids (linoleic acid or arachidonic acid). It has been reported that the activity of membrane-bound enzymes is modified by increasing the content of polyunsaturated fatty acids in the plasma membrane. This change can be induced by stimulating lymphocytes with the mitogen Concanavalin A. A mechanism similar to “stimulation of endothelial growth by an angiogenesis factor” may also exist during rat brain vascularization which may cause the observed decrease of ?I-GT activity. BBr and Wolff’ demonstrated that the fraction of the capillary surface in direct contact with the astroglial sheath increased from 66 to 84% during the first postnatal week. At the same time, the postnatal proliferation of the astroglial cells reaches maximum values.” Thus, both the mitotic proliferation of the endothelial cells and of the astroglial cells is correlated with a decrease of y-GT activity in brain capillaries. The surprisingly high level of y-GT activity at day 2 after birth (69% when compared to the adult level) can be explained by enzyme induction in the resting endothelial cells which are already partially covered by astroglial lamellae during that period. Glio-vascular contact seems to be necessary for induction of y-GT in endothelial cells. as already shown by DeBault and Cancilla.‘2 The final increase of r-GT activity in the newly formed capillary segments, observed during the third postnatal week, coincides in time with the phase of telescope-like elongation of the brain capillaries (with the consecutive slow increase in the capillary length density, see Fig. 6). and with the complete expansion of the astroglial lamellae as described by BPr and Wolff.’ Interestingly, the increase of the y-GT activity was initiated on day 14. In this period, inhomogeneity of staining among the capillary segments increased showing that the onset of differentiation did not occur simultaneously in every capillary segment at that time. After the third postnatal week, the y-GT activity of the endothelial cells became fully expressed, possibly by induction by the complete astroglial sheath. Wapnir et ~1.~” described similar ontogenctic changes of y-G?‘ activity in rat brain homogenates by using a biochemical assay. They found a decrease of y-GT activity between the fetal (2 days prenatal) and the neonatal period, where the minimal enzyme activity is found. The difference in the timing of the minimum of y-GT activity may be due to the different sampling procedures used (whole brain homogenate versus capilIary fragments from cerebral cortex). Biochemical measurement of y-GT activity in isolated mouse brain microvessels resulted in 24 This finding correlates with the differentiasimilar changes during ontogenetic development. tion and maturation of capillary wall structures and confirms the evidence that the presence of astroglial cells is important for the expression of -y-GT activity. This enzyme is suggested to play a role in functional maturation of the blood-brain barrier. Ack,iowtedgemetzr.s-We
wish to thank Dr I. Karnushina for helpful advices and introduction in the tecbniquc of the microvessel preparation. We are also grateful to Prof. A. Miodonski for stimulating discussions. to Prof. R. Kinnc for his constructive criticism, and to Miss R. Schebaum for typing the manuscript.
REFERENCES I. Albert Z. (196s) Gamma-glutamyl transpeptidase in cancers of different human organs. Nurtrre 205, 407. 2. Albert 2.. Rzucidlo Z. and Starzyk H. (1970) Biochemical and histochemical investigations of the gamma glutamyl transpeptidase in embryonal and adult organs of man. Acta Histochem. 37, 74-79. 3. Anderson E. 1. and Wright D. D. (1982) The roles of glutathione reductase and y-glutamyl transpeptidase in cornea1 transendothelial fluid transport mediated by oxidized glutathione and glucose. E.rpl Eye Res. 35, 1I-19. 4. BPr Th. (1983) Pattern of vascularization in the developing cerebral cortex. Ciha FOUNT. 108, 20-36. 5. B& Th. and Wolff J. R. (1977) Morphomet~ of interendothelial and glio-vascular contacts of rat brain capillaries during postnatal development. Riblrhca Annr. IS, 514-517.
y-GT 6. Baumrucker Diary
activity in rat brain
C. R. (1979) y-Glutamyltranspeptidase
511
microvessels
of bovine milk membranes:
Distribution
and characterization.
J.
Sri. 62, 253-258.
7. Caspers
M. L. and Diglio C. A. (1984) Expression of y-glutamyl transpeptidase in transformed rat cerebral endothelial cell line. Biochim. biophys. Acta 803, l-6. 8. Chang J. P. and Hori S. H. (1961) The freeze substitution technique: I. Method. J. Histochem. Cytochem. 9, 292-
300. 9. Dass P. D.. Misra R. P. and Welbourne T. C. (1982) Renal y-glutamyltranspeptidase: In situ antiluminal localization. J. Histochem. Cytochem. 30, 148-1.52. IO. Daoust R. (1982) The histochemical demonstration of gamma-glutamyltranspeptidase like activity in different populations of the rat liver during azo dye carcinogenesis. J. Histochem. Cytochem. 30, 312-316.
Il.
DeBault L. E. (1981) y-Glutamyltranspeptidase induction mediated by glial foot process to endothelium contact in co-culture. Brain Res. 220, 432-435. 12. DeBault L. E. and Cancilla P. A. (1980) y-Glutamyltranspeptidase in isolated brain endothelial cells: Induction by glial cells in vitro. Science 207, 6.53-655. 13. Diglio C. A., Grammas P., Giacomelli F. and Wiener J. (1982) Primary culture of rat cerebral microvascular endothelial cells. Lab. Invest. 46, 554-563. 14. Djuricic B. M. and Mrsulja B. B. (1977) Enzyme activity of the brain: microvessels vs. total forebrain homogenate. Brain Res. 138, 561-564.
15. Fleming N., Groscurth P. and Kistler G. S. (1977) The activity and distribution of gamma-glutamyl transpeptidase (y-GT) in human foetal organs. Histochemistry 51. 209-218. 16. Fukiwara K., Katyal S. L-and Lombardi B..(1982) Influence of age, sex and cancer on the activities of gammaglutamyl transpeptidase and of dipeptidyl aminopeptidase IV in rat tissues. Enzyme 27, 114-l IX. 17. Gardell S. J. and Tate S. S. (1982) Modification of the acceptor binding site of y-glutamyl transpeptidase by the diazonium derivatives of p-aminohippurate p-aminobenzoate. Archs Biophys. 216, 719-726. IX. Ghandour M. S., Langley 0. K. and Varga V. (1980) lmmunohistological localization of y-glutamyltranspeptidase in cerebellum at light and electron microscope levels. Neurosci. Len. 20, 125-129. 19. Glenner G. G., Folk J. E. and McMillan P. J. (1962) Histochemical demonstration of a gamma-glutamyl transpeptidase-like activity. J. Histochem. Cyrochem. 10, 481489. 20. Grafstrom R.. Stead A. H. and Orrenius S. (1980) Metabolism of extracellular glutathione in rat small-intestinal mucosa. Eur. J. Biochem. 106, Y-576. 21. lchikawa M., Shiga T. and Hirata Y. (1983) Spatial and temporal pattern of postnatal proliferation of glial cells in the parietal cortex of the rat. Devl Brain Res. 19, 181-198. 22. Kugler P. (1982) Aminopeptidase A is angiotensinase A. I. Quantitative histochemical studies in the kidney glomerulus. Histochemistry 174, 220-245. 23. Lidinsky W. A. and Drewes L. R. (1983) Characterization of the blood-brain barrier: Protein composition of the capillary endothelial cell membrane. J. Neurochem. 41, 1341-1348. 24. Lisy V., Stasny F. and Lodin J. (1983) Gamma-glutamyl transpeptidase activity in isolated cells and in various regions of the mouse brain during early development and aging. Physiologia Bohemoslov. 32, l-9. 2.5. Meister A.. Tate S. S. and Griffith 0. W. (1981) Gamma-glutamyl transpeptidase. Meth. Etuymol. 77, 237-253. 26. Grning L. and Hammarstrom S. (1982) Kinetics of the conversion of leucotriene c by y-glutamyl transpeptidase. Biochem. 27. Orlowski 28. 29. 30.
31. 32.
hiophys.
Res. Commun.
106,
1304-1309.
M., Sessa G. and Green J. P. (1974) Gamma-glutamyl transpeptidase in brain capillaries: Possible site of a blood-brain barrier for amino acids. Science 184, 66-f%. Rowan R. A. and Maxwell D. S. (1981) Pattern of vascular sprouting in the postnatal development of the cerebral cortex of the rat. Am. J. Anat. 160, 247-255. Rutenburg A. M., Kim H.. Fischbein J. W., Hanker J. S.. Wasserkrug H. L. and Seligman A. M. (1969) Gammaglutamyl transpeptidase activity. J. Histochem. Cytochem. 17, 517-526. Sepulveda F. V. and Burton K. A. (1982) Gamma-glutamyl transferase activity in the pig proximal colon during early postnatal development. FEBS Lett. 139, 171-173. Shiba M. and Klein-Szanto A. J. P. (1983) Variation of gamma glutamyl transpeptidase activity and lectin binding in the course of carcinogenesis of the respiratory tract epithelium. Carcinogenesis 4, 687491. Spater H. W.. Pornchynsky M. S., Quintana N., lnoue M. and Novikoff A. B. (1982) lmmunocytochemical localization of y-glutamyltransferase in rat kidney with protein-A-horseradish peroxidase. Proc. natn. Acad. Sci. U.S.A. 79, 3547-3550.
33.
34. 35. 36. 37. 38.
Szamel M. and Resch K. (1981) Modulation of enzyme activities in isolated lymphocyte plasma membranes by enzymatic modification of phospholipid fatty acids. J. biol. Chem. 256, ll618-I 1623. Vanderlaan M.. Cutter C. and Dolbeare F. (1979) Flow microfluorometric identification of liver cells with elevated gamma-glutamyl transpeptidase activity after carcinogen exposure. J. Histochem. Cytochem. 27, 114-l 19. Viria J.. Puertes I. M., Estrela J. M.. Vida J. R. and Galbis J. L. (1981) Involvement of y-glutamyltransferase in amino-acid uptake by the lactating mammary gland of the rat. Biochem. J. 194, 99-102. Wapnir R. A.. Mancusi V. J. and Goldstein L. A. (1982) Comparative ontogenesis of gamma-glutamyl transpeptidase in rat tissue. E.vperierrtia 38, 647448. Welbourne T. C. and Dass P. D. (1983) Minireview: Function of renal y-glutamyltransferase: Significance of glutathione and glutamine interactions. Life Sci. 30, 793-801. Yamashita K.. Tachibana Y., Shichi H. and Kobata A. (1983) Carbohydrate structures of bovine kidney yglutamyltranspeptidase. J. Biochem. 93, 135-147.
V<,I. Nc,rr~r,.\i~ic,rrce.
J. No. 0. pp. 5O.%SI I,
IWh.
0736-S7JXIXh $03.cnl+o.o0 Pcrgamon Journals Ltd. @ 19X6 ISDN
Pnnrcd in Great Britain.
POSTNATAL DEVELOPMENT OF y-GT ACTIVITY IN RAT BRAIN MICROVESSELS CORRESPONDS TO CAPILLARY GROWTH AND DIFFERENTIATION A. Max-Planck-lnstitut fiir Systemphysiologie, Anatomie und Entwicklungsneurobiologie,
W.
BUDI
SANTOSO
and TH.
BAR*
Rheinlanddamm 201, D-4600 Dortmund Georg-August Universittit. Kreuzberger (Accepfed
I. F.R.G.; *Abteilung fiir Klinische Ring 36, D-3400 GGttingen, F.R.G.
I4 July 1986)
Abstract-It has been demonstrated histochemically that endothelial cells of the cerebral capillaries show a high activity in y-glutamyl transpeptidase (y-CT), an enzyme which takes part in the transfer of large neutral amino acids across the blood-brain barrier. Reports of the disappearance of enzyme activity in endothelial cells (EC) grown in culture” suggest that the presence of astroglial cells (AC) is required for the expression of y-GT in these cells. The present study deals with the developmental changes in ?I-GT activity in capillaries of rat cerebral cortex during ontogenesis (i.e. on days 2,7. I I, 14, 21 and 60 after birth). y-GT activity is determined by measuring the enzyme kinetics of the histochemical reaction on isolated brain capillaries using a flying spot microscope densitometer. Enzyme activity is expressed as an increase in relative optical density (at SO0 nm) in arbitrary units/min/pm” during the 2 min immediately after initiating incubation and corresponds to the y-CT active area (in pm”) of the capillary segment. During early postnatal development, a biphasic change of y-CT activity in capillaries of rat cerebral cortex is observed. The first phase (i.e. the postnatal period between the 2nd and 12th day) is characterized by a significant decrease in y-GT activity, which coincides with the onset of the rapid mitotic proliferation of cortical endothelial cells. During the second phase (i.e. the postnatal period between the 12th and 2lst day), a fast increase in the y-CT activity can be measured. Then enzyme activity reaches the adult level between the 2lst and 60th postnatal day. With regard to changes in capillary length density and in astroglial proliferation the results correspond to the differentiation and maturation of the blood-brain barrier and confirm evidence stressing the important role of astroglial cells for the expression of y-CT activity. Key words: Isolated microvessels, Microdensitometric measurements,
Rat brain cortex, y-GT activity, Ontogenesis, Endothelial cells.
Semiquantitative
histochemistry,
y-Glutamyl transpeptidase (y-GT) plays a key role in the y-glutamyl cycle; it also catalyzes transfer of the y-glutamyl (-y-GT) residue of intracellular glutathione, S-substituted glutathion derivatives, and other y-glutamyl compounds to a number of acceptors forming y-glutamyl peptides. “‘J y-GT is a membrane-bound heterodimeric glycoprotein. “J It has also been considered to metabolize leucotrien C, to D_,.Zh This enzyme is widely distributed in the various tissues of mammals involved in secretory and transport mechanisms, such as in the brush border of the kidney proximal tubules,y,32,“7 in the lactating mammary glands,‘,“” in biliary duct epithelial cells,2” in cornea1 endothelial cells,” in small intestinal mucosa,‘” in colon epithelial cells,3” and in lymphocytes.“” The enzyme’s activity varies within the pre- and postnatal development stages and in different organs.2.“,‘“.“‘,‘h At the dedifferentiation stage, y-GT activity is at the level usually found in early embryonic development of the particular organ. Thus, the enzyme has been reported to be a useful marker enzyme for preneoplastic and neoplastic foci in the liver after carcinogen application, ‘. ‘0-34 and during carcinogenesis of the respiratory tract epithelium.“’ Orlowski et ~1.~’ found, using histochemical methods, that y-GT is associated with the endothelial cells of isolated brain capillaries, and suggested y-GT functions to transfer large neutral amino acids across the blood-brain barrier. -y-GT is localized predominantly in the capillaries of the brain,14 and is immunocytochemically detectable in high concentration on the luminal surface of the cell membrane of cerebellar endothelial cells. Ix As suggested by DeBault and Cancilla,‘2 the presence of adjacent processes of astroglia (AG) is required for continuous induction of y-GT. The enzyme usually disappears when the endothelial cells migrate or proliferate from the isolated microvessel. It has been recognized that coculturing mouse endothelial cells with Ch rat glioma cells can induce the re-expression of y-GT activity in endothelial cells. “.I2 -y-GT activity, in cultured brain capillary endothelium, increases
A. W. Rudi Santoso and Th. Biir
so3
significantly after in vitro transformation of these cells by an avian retrovirus. Activity declines to a low level by passage 80 coinciding with the disappearance of F VIII and angiotensin converting enzyme (differentiation markers for brain endothelial cells).’ The association between the differentiation of endothelial cells and the presence of y-GT activity prompted us to examine the developmental changes of y-GT activity on isolated brain capillaries at different ontogenetic stages. The postnatal period was selected since, according to and ultrastructural changes occur in neocortithe detailed study of Bar,4 obvious morphometric cal capillaries during that period. It was shown that a phase of vigorous proliferation of new capillaries, resulting in the most dramatic increase in capillary length, was followed by the functional maturation of the capillary bed. Difficulties in detecting low levels of y-GT activity with common biochemical methods, required that a micromethod for semiquantitative histochemical measurement of y-GT activity can be used together with a microscope densitometer. This method offers some advantages, because the capillary fragments, which are expected to exhibit y-GT activity, can be selectively measured for their enzyme kinetic properties. By means of the newly developed method, we demonstrated a significant decrease in y-GT activity in brain capillaries during the period of rapid mitotic proliferation of endothelial cells in the first postnatal week. After that period, y-GT activity increases to adult levels at day 60 postnatal. Possible implications of the results are discussed with regard to changes in the capillary length density and in astroglial proliferation. EXPERIMENTAL
PROCEDURES
Isolation of the rat brain microvessels
The cortical microvessel preparations performed were slightly modified according to the method described by Diglio et a1.13 using Sprague-Dawley rats (supplier: Zentrales Institut fur Versuchstierkunde, Hannover, F.R.G.). For the ontogenic study, groups of animals at different postnatalages,i.e.2days(n=6).7days(n=7),11days(n=10),14days(n=10),16days(n=19), 21 days (n = 6), and 60 days (II = 6) were taken. Under pentobarbital anesthesia (Ncmbutal’“‘i.p., 40 mg/kg body wt). the animals were perfused transcardially with ice-cold perfusion medium (5 mM MgCl:. 20 mM HEPES. pH 7.2, in 500 ml IO times diluted Hank’s balanced salt solution). The perfusion was carried out by using a 20 ml syringe under gentle pressure. After quick removal of the brain from the skull, the cerebral cortex was freed from the meninges. the brain stem, and the white matter. Pieces of the cerebral cortex were collected in a beaker containing ice-cold homogenization medium (5 mM MgCl*, 20 mM HEPES, 20 mM Tris, 2 mM EDTA, pH 7.2. in 500 ml Hank’s balanced salt solution). The material was chopped into small pieces and divided in two equal parts. Homogenization was performed in a Teflon-glass homogenizer (X IO. 350 rpm) with 20 ml homogenization medium. After adding 40 ml of the homogenization medium, homogenization was repeated. Centrifugation was performed for IO min (1000 g). The sediment was resuspended in a medium (40 ml) containing 5 mM MgC12, 5% fetal calf serum (FCS), 20 mM HEPES and 15% dextran in x IO diluted Hank’s balanced salt solution. To obtain a crude pellet of the brain microvessels, the suspension was centrifuged again (4OOOg, 20 min). The crude pellet was resuspended in 40 ml x IO diluted Hank’s balanced salt solution containing 5 mM MgC17, 20 mM HEPES and 5% fetal calf serum (washing medium). The suspension was washed thoroughly through a glass column containing glass beads (diameter 0.25%0.3 mm) with approximately 200-300 ml washing medium. Thereafter, the microvessel-rich medium, above the glass beads column, was collected by gentle pipetting. To enrich the microvessels, a final centrifugation was performed (5 min, 1000 g). The microvessel fractions were spread onto glass slides (precoated with 3% gelatin which was stabilized in 2% Na2S03 for 1 hr before use), air-dried and fixed and defatted in 100% aceton (-4O”C, 12 hr), 100% aceton (4”C, I hr), 2:l (v/v) aceton/chloroform (4°C. I hr). 2:l (v/v) chloroform/aceton (4”C, 1 hr) and finally in 100% chloroform (4°C. 1 hr) according to Chang and Hori.” The specimens were stored at -40°C prior to examination. Histochemical
demonstration
of‘ y-GT
activity
Demonstration of y-GT activity was performed by using an azo dye coupling method first described by Rutenburg et al.” The y-glutamyl residue of the y-L-glutamyl-4-methoxy-2-naphthyla-
y&T
activity in rat brain microvessels
505
mide (as a substrate) is transferred by y-GT to the acceptor glycylglycine. In a simultaneous azo coupling reaction with Fastblue BB. a red azo dye is developed. To obtain a micromethod for the quantitative measurement of the enzyme kinetic, using a flying spot microscope densitometer, the method was modified as demonstrated
in Fig.
I,
Preparation of the stock solutions 8.27 mM y-glutamyl-4-methoxy-2-naphthylamide (y-Glu-MNA) 5 mg yGlu-MNA dissolved in 0.1 ml dimethylsulfoxide (DMSO), 0.1 ml 1 N NaOH and 1.8 ml bidest . HzO. 12.0 mM Fastblue BB Salt 25 mg Fastblue BB Salt dissolved in 10 ml bidest. HZO. 300 mM Glycylglycine (Gly-Gly) 100 mg Gly-Gly dissolved in 2 ml 0.1 M Tris-HCI buffer, pH 7.4. 145 mM NaCl 0.85 g dissolved in 100 ml bidest. H20. 0.4 M Tris-HCI buffer (diluted 4 times before use to obtain 0.1 M buffer solution) 25 ml 0.4 M Tris (5.61 g/100 ml bidest. H,O)+33.5 ml bidest. Hz0 adjusted with 42.5 ml 0.2 N HCI to pH 7.4. 0.1 M CuSO, 2.5 g/l00 bidest. H20. ISOLATED
0.200ml lL5mM
NaCl
MICROVESSELS
p-7
A-1
u
BRAIN
O.OZOmt
SPACER
WIRE
i@=200@i1
300 mM Gly -Gly
O.lMTris-HCI
0 ZOOmI
DH8.0 8.27mM
Y-GiuMNA
12 OOmM
Fastblue.
MlCRODENSlTOMETRlC
BB
MEASUREMENT
I~UBATION SOLUTION
Fig. I. Schematic diagram showing the preparation of the incubation medium and the isolation of microvessels. 0.2 ml of the incubation medium is dropped on a sample of isolated microvessels before starting the enzyme kinetic measurements.
Microdensitometric measurement of the y-GT activity To obtain quantitative data on enzyme kinetics of y-GT we used a flying spot microdensitometer (Type M 85 A, Vickers Instr. Ltd., London, U.K.) and a computer-assisted system (Videoplan, Kontron Ltd., Munich, F.R.G.), linked by a BCD interface and compatible special measuring software (for details see Ref. 22). Figure 1 schematically shows the general procedure including the preparation of the incubation solution and the isolated brain microvessel specimens demonstrated prior to microdensitometric measurement. The incubation temperature was kept constant (37°C) during the measurements, by perfusing a small chamber of the microdensitometric stage between condenser lens and microscopic slide with a constant flow of water at 37°C from a thermostat (Type KARD, Mel3gertite Werk-Lauda, F.R.G.). Before starting the incubation, an appropriate capillary segment was seIected using differential contrast illumination. Microdensitometric measurements (i.e. registration of the relative optical density caused by development of the reaction product) is started
SO6
A. W.
470
Budi
Santoso
and Th
510 490 WAVELENGTH
530 Inml
Biir
550
570
Fig. 2. Curve showing the absorption maximum of the red azo dye developed as reaction product during the y-GT histochemical reaction at the wavelength getween 470 and 570 nm. The enzyme-kinetic measurements are done at the optimal absorption maximum (SO0 nm).
immediately after applying 0.2 ml of incubation solution. The measurements were performed at a wavelength of 500 nm (i.e. at the absorption maximum of the reaction product, Fig. 2). The changes of the relative optical density (expressed in arbitrary units; AOD,,,,,) were measured at 3 set intervals controlled by the computer. During the 10 min after the start of incubation the values of the optical density (AODsoo) were stored automatically in the computer. Finally, the area of the capillary segment displaying the reaction product (monitored by means of a video planimetrically using a digitizing tablet camera Sanyo, Model No. VC x 1600) was measured (Fig. 3a). The stored densitometric values obtained during the histochemical reactions were plotted against time. To evaluate the y-GT activity, the slope of the regression lines within the first 2 min of the reaction was determined (i.e. within the initial linear part of the kinetic curve); the value represents AOD,,,,,/min. To obtain the relative specific activity of the y-GT, this value was related to the morphomerically measured reactive area and expressed in AODs,,,,lp,m2/min. The following instrument settings were used for this study: objective x40. spot size 2. guting mask A4. scanning time 2 set, band width 50. wavelength 500 nm. For the histochemical localization of y-GT activity the microvessels were incubated at 37°C in a moist chamber for 30 min. The reaction was stopped by addition of 0. I ml 0.1 M CuSO,. the slides were washed several times in phosphate-buffered saline solution and embedded in glycerol gelatin (Sigma Chem., Munich, F.R.G.), and examined immediately under the light microscope. Control incubations are done using, (1) substrate free incubation, (2) incubation with the incubation medium containing 0.2 ml 200 mM serine (dissolved in 0.01 M Tris-HCI buffer, pH 6.9)/0.2 ml 0.05 M borate buffer solution (pH 7.6) added to 1 ml incubation medium.
RESULTS Histochemicul
localization
of y-glutamyl
transpeptiduse
The activity of y-GT detected in isolated brain cortical microvessels was restricted exclusively to the capillary segments (Fig. 3b). Pre- and postcapillary vessels were not stained. Observation under higher magnification revealed that the reaction produced was mostly localized at the luminal side of the endothelial cells (Fig. 4). The brain capillary fragments from 2-day-old rats showed large rounded nuclei lying close together (Fig. 5a). After incubation with y-GT, the capillary endothelial cells were stained intensively. The nuclei, however, showed only a weak staining (Fig. 4). In the capillary fragments from 7-day-old animals, a distinct reduction of y-GT activity was observed. The nuclei were densely packed. They remained large, and round in shape (Fig. 5b). On day 15, the maximal inhomogeneity of staining among single capillary segments was found. From 16 days on and in the older age groups an increased intensity of y-GT staining could be observed in the cortical capillary
y-GT
activity in rat brain microvessels
Fig. 3. (a) Capillary segment as seen on the monitor at the end of the measurement, 10 min after the start of the enzyme reaction for y-GT activity. The area on which the development of the reaction product can be detected, is circumscribed for the morphometric measurement. The area of the endothelial cells encircled by the gating mask is used to estimate for calculation of the specific activity of y-GT. (b) y-GT activity is demonstrated predominantly in the endothelial cells of capillaries within the microvessel fraction isolated from adult rat cortex (linear magnification: X40@.
507
508
A. W. Budi Santoso
and Th. Btiir
Fig. 4. y-GT activity in brain capillary fragment at higher magnification. intraluminal localization of the reaction product (Interference-Contrast magnification: X 1600).
Arrows demonstrate the after Nomarski; linear
Fig. 5. (ax) Demonstration of -r-GT activity from rat brain cortical capillaries isolated at different postnatal ages. The specimens were incubated with the y-CT incubation medium at 37°C for 30 min. Microphotography was performed by identical exposure time and at the same magnification (linear magnification: X250).
y-GT
activity
in rat
brain microvcsscls
50~
segments. Capillaries isolated 60 days after birth showed the most intensive staining of y-GT activity (Fig. 5c, d, e). When 0.2 ml 200 mM serine (in 0.01 M Tris-HCI, pH 6.9)/0.2 ml 0.05 M borate buffer solution (pH 7.6) were added to 1 ml incubation medium, no activity (or staining) was detected in any of the specimens. Additional examination of rat brain cryosections revealed that the staining was also restricted to the capillary segments (result not shown). Quantitative
analysis
of the histochemical
y-CT
activity
On day 2, the relative specific activity of y-GT in capillary segments of the brain cortex was higher than in similar fragments isolated from the brain of 7- and IO-day-old animals. In the second postnatal week y-GT activity was obviously decreased. The lowest activity was found on postnatal day 11. If the mean value of y-GT activity in capillary segments obtained from 60-day-old rat brains was regarded as lOO%, the relative specific activity of y-GT at postnatal days 7, 11 and 16 varied between 33 and 48% (Fig. 6). In early postnatal development, a biphasic change of y-GT activity in capillaries of rat cerebral cortex was apparent. The first phase (i.e. at the postnatal period between the 2nd and 12th day) was characterized by a significant decrease of y-GT activity, whereas during the following second phase (i.e. the postnatal period between the 12th and 21st day) a fast increase of y-GT activity could be measured. Later on, only a slight increase in y-CT activity was observed before the curve of activity reached its plateau approximately on day 60 (Fig. 6).
AGE
[DAYS
AFTER
BIRTH1
Fig. 6. Relative specific activity of y-GT in isolated brain cortical capillaries as measured by the microscope densitometer (each symbol represents the mean values r?rS.E.M.; n = 10 measurements). *P
DISCUSSION Our results demonstrate that yGT activity of capillary endotheliai cells is closely related to the developmental changes of the brain vascular system. The semiquantitative microdensitometric approach offers the possibility of comparing differences in the enzyme kinetics of y-GT in isolated brain capillaries at various ages after birth. Our method seems a suitable tool for studies of other enzymes, especially in isolated or cultured cells, even if only a small amount of cells is available for investigation. The vascularization of the rat brain cortex during the postnatal period represents a suitable model for studies of vascular growth and differentiation, because most major changes occur predominantly after birth. Additional knowledge is availabie concerning timing and intensity of endothelial and astroglial proliferation as well as the postnatal growth of the astroglial processes and lamellae covering the intracerebral vasculature.“~2’.2x
510
A. W. Budi Santoso and Th. BLir
High y-GT activity in the brain cortex is exhibited exclusively at the capillary level. In isolated microvessels the enzyme activity is 35 times higher than in the crude brain hom0genate.l’ Shortly after birth, y-CT activity decreases significantly. The diminished activity coincides with the onset of the rapid mitotic proliferation and results in a dramatic increase in capillary length density.’ During the rapid increase in capillary length, the highest numerical density of capillary sprouts is found on day 8 after birth.28 At this stage, the ?I-GT activity in the brain capillaries remains low. We suggest a state of low differentiation of the membranes of endothelial cells during the rapid mitotic proliferation. Szamei and Resch’3 demonstrated a modulation of y-GT activity in lymphocytic plasma membranes by incorporation of long-chain unsaturated fatty acids (linoleic acid or arachidonic acid). It has been reported that the activity of membrane-bound enzymes is modified by increasing the content of polyunsaturated fatty acids in the plasma membrane. This change can be induced by stimulating lymphocytes with the mitogen Concanavalin A. A mechanism similar to “stimulation of endothelial growth by an angiogenesis factor” may also exist during rat brain vascularization which may cause the observed decrease of ?I-GT activity. BBr and Wolff’ demonstrated that the fraction of the capillary surface in direct contact with the astroglial sheath increased from 66 to 84% during the first postnatal week. At the same time, the postnatal proliferation of the astroglial cells reaches maximum values.” Thus, both the mitotic proliferation of the endothelial cells and of the astroglial cells is correlated with a decrease of y-GT activity in brain capillaries. The surprisingly high level of y-GT activity at day 2 after birth (69% when compared to the adult level) can be explained by enzyme induction in the resting endothelial cells which are already partially covered by astroglial lamellae during that period. Glio-vascular contact seems to be necessary for induction of y-GT in endothelial cells. as already shown by DeBault and Cancilla.‘2 The final increase of r-GT activity in the newly formed capillary segments, observed during the third postnatal week, coincides in time with the phase of telescope-like elongation of the brain capillaries (with the consecutive slow increase in the capillary length density, see Fig. 6). and with the complete expansion of the astroglial lamellae as described by BPr and Wolff.’ Interestingly, the increase of the y-GT activity was initiated on day 14. In this period, inhomogeneity of staining among the capillary segments increased showing that the onset of differentiation did not occur simultaneously in every capillary segment at that time. After the third postnatal week, the y-GT activity of the endothelial cells became fully expressed, possibly by induction by the complete astroglial sheath. Wapnir et ~1.~” described similar ontogenctic changes of y-G?‘ activity in rat brain homogenates by using a biochemical assay. They found a decrease of y-GT activity between the fetal (2 days prenatal) and the neonatal period, where the minimal enzyme activity is found. The difference in the timing of the minimum of y-GT activity may be due to the different sampling procedures used (whole brain homogenate versus capilIary fragments from cerebral cortex). Biochemical measurement of y-GT activity in isolated mouse brain microvessels resulted in 24 This finding correlates with the differentiasimilar changes during ontogenetic development. tion and maturation of capillary wall structures and confirms the evidence that the presence of astroglial cells is important for the expression of -y-GT activity. This enzyme is suggested to play a role in functional maturation of the blood-brain barrier. Ack,iowtedgemetzr.s-We
wish to thank Dr I. Karnushina for helpful advices and introduction in the tecbniquc of the microvessel preparation. We are also grateful to Prof. A. Miodonski for stimulating discussions. to Prof. R. Kinnc for his constructive criticism, and to Miss R. Schebaum for typing the manuscript.
REFERENCES I. Albert Z. (196s) Gamma-glutamyl transpeptidase in cancers of different human organs. Nurtrre 205, 407. 2. Albert 2.. Rzucidlo Z. and Starzyk H. (1970) Biochemical and histochemical investigations of the gamma glutamyl transpeptidase in embryonal and adult organs of man. Acta Histochem. 37, 74-79. 3. Anderson E. 1. and Wright D. D. (1982) The roles of glutathione reductase and y-glutamyl transpeptidase in cornea1 transendothelial fluid transport mediated by oxidized glutathione and glucose. E.rpl Eye Res. 35, 1I-19. 4. BPr Th. (1983) Pattern of vascularization in the developing cerebral cortex. Ciha FOUNT. 108, 20-36. 5. B& Th. and Wolff J. R. (1977) Morphomet~ of interendothelial and glio-vascular contacts of rat brain capillaries during postnatal development. Riblrhca Annr. IS, 514-517.
y-GT 6. Baumrucker Diary
activity in rat brain
C. R. (1979) y-Glutamyltranspeptidase
511
microvessels
of bovine milk membranes:
Distribution
and characterization.
J.
Sri. 62, 253-258.
7. Caspers
M. L. and Diglio C. A. (1984) Expression of y-glutamyl transpeptidase in transformed rat cerebral endothelial cell line. Biochim. biophys. Acta 803, l-6. 8. Chang J. P. and Hori S. H. (1961) The freeze substitution technique: I. Method. J. Histochem. Cytochem. 9, 292-
300. 9. Dass P. D.. Misra R. P. and Welbourne T. C. (1982) Renal y-glutamyltranspeptidase: In situ antiluminal localization. J. Histochem. Cytochem. 30, 148-1.52. IO. Daoust R. (1982) The histochemical demonstration of gamma-glutamyltranspeptidase like activity in different populations of the rat liver during azo dye carcinogenesis. J. Histochem. Cytochem. 30, 312-316.
Il.
DeBault L. E. (1981) y-Glutamyltranspeptidase induction mediated by glial foot process to endothelium contact in co-culture. Brain Res. 220, 432-435. 12. DeBault L. E. and Cancilla P. A. (1980) y-Glutamyltranspeptidase in isolated brain endothelial cells: Induction by glial cells in vitro. Science 207, 6.53-655. 13. Diglio C. A., Grammas P., Giacomelli F. and Wiener J. (1982) Primary culture of rat cerebral microvascular endothelial cells. Lab. Invest. 46, 554-563. 14. Djuricic B. M. and Mrsulja B. B. (1977) Enzyme activity of the brain: microvessels vs. total forebrain homogenate. Brain Res. 138, 561-564.
15. Fleming N., Groscurth P. and Kistler G. S. (1977) The activity and distribution of gamma-glutamyl transpeptidase (y-GT) in human foetal organs. Histochemistry 51. 209-218. 16. Fukiwara K., Katyal S. L-and Lombardi B..(1982) Influence of age, sex and cancer on the activities of gammaglutamyl transpeptidase and of dipeptidyl aminopeptidase IV in rat tissues. Enzyme 27, 114-l IX. 17. Gardell S. J. and Tate S. S. (1982) Modification of the acceptor binding site of y-glutamyl transpeptidase by the diazonium derivatives of p-aminohippurate p-aminobenzoate. Archs Biophys. 216, 719-726. IX. Ghandour M. S., Langley 0. K. and Varga V. (1980) lmmunohistological localization of y-glutamyltranspeptidase in cerebellum at light and electron microscope levels. Neurosci. Len. 20, 125-129. 19. Glenner G. G., Folk J. E. and McMillan P. J. (1962) Histochemical demonstration of a gamma-glutamyl transpeptidase-like activity. J. Histochem. Cyrochem. 10, 481489. 20. Grafstrom R.. Stead A. H. and Orrenius S. (1980) Metabolism of extracellular glutathione in rat small-intestinal mucosa. Eur. J. Biochem. 106, Y-576. 21. lchikawa M., Shiga T. and Hirata Y. (1983) Spatial and temporal pattern of postnatal proliferation of glial cells in the parietal cortex of the rat. Devl Brain Res. 19, 181-198. 22. Kugler P. (1982) Aminopeptidase A is angiotensinase A. I. Quantitative histochemical studies in the kidney glomerulus. Histochemistry 174, 220-245. 23. Lidinsky W. A. and Drewes L. R. (1983) Characterization of the blood-brain barrier: Protein composition of the capillary endothelial cell membrane. J. Neurochem. 41, 1341-1348. 24. Lisy V., Stasny F. and Lodin J. (1983) Gamma-glutamyl transpeptidase activity in isolated cells and in various regions of the mouse brain during early development and aging. Physiologia Bohemoslov. 32, l-9. 2.5. Meister A.. Tate S. S. and Griffith 0. W. (1981) Gamma-glutamyl transpeptidase. Meth. Etuymol. 77, 237-253. 26. Grning L. and Hammarstrom S. (1982) Kinetics of the conversion of leucotriene c by y-glutamyl transpeptidase. Biochem. 27. Orlowski 28. 29. 30.
31. 32.
hiophys.
Res. Commun.
106,
1304-1309.
M., Sessa G. and Green J. P. (1974) Gamma-glutamyl transpeptidase in brain capillaries: Possible site of a blood-brain barrier for amino acids. Science 184, 66-f%. Rowan R. A. and Maxwell D. S. (1981) Pattern of vascular sprouting in the postnatal development of the cerebral cortex of the rat. Am. J. Anat. 160, 247-255. Rutenburg A. M., Kim H.. Fischbein J. W., Hanker J. S.. Wasserkrug H. L. and Seligman A. M. (1969) Gammaglutamyl transpeptidase activity. J. Histochem. Cytochem. 17, 517-526. Sepulveda F. V. and Burton K. A. (1982) Gamma-glutamyl transferase activity in the pig proximal colon during early postnatal development. FEBS Lett. 139, 171-173. Shiba M. and Klein-Szanto A. J. P. (1983) Variation of gamma glutamyl transpeptidase activity and lectin binding in the course of carcinogenesis of the respiratory tract epithelium. Carcinogenesis 4, 687491. Spater H. W.. Pornchynsky M. S., Quintana N., lnoue M. and Novikoff A. B. (1982) lmmunocytochemical localization of y-glutamyltransferase in rat kidney with protein-A-horseradish peroxidase. Proc. natn. Acad. Sci. U.S.A. 79, 3547-3550.
33.
34. 35. 36. 37. 38.
Szamel M. and Resch K. (1981) Modulation of enzyme activities in isolated lymphocyte plasma membranes by enzymatic modification of phospholipid fatty acids. J. biol. Chem. 256, ll618-I 1623. Vanderlaan M.. Cutter C. and Dolbeare F. (1979) Flow microfluorometric identification of liver cells with elevated gamma-glutamyl transpeptidase activity after carcinogen exposure. J. Histochem. Cytochem. 27, 114-l 19. Viria J.. Puertes I. M., Estrela J. M.. Vida J. R. and Galbis J. L. (1981) Involvement of y-glutamyltransferase in amino-acid uptake by the lactating mammary gland of the rat. Biochem. J. 194, 99-102. Wapnir R. A.. Mancusi V. J. and Goldstein L. A. (1982) Comparative ontogenesis of gamma-glutamyl transpeptidase in rat tissue. E.vperierrtia 38, 647448. Welbourne T. C. and Dass P. D. (1983) Minireview: Function of renal y-glutamyltransferase: Significance of glutathione and glutamine interactions. Life Sci. 30, 793-801. Yamashita K.. Tachibana Y., Shichi H. and Kobata A. (1983) Carbohydrate structures of bovine kidney yglutamyltranspeptidase. J. Biochem. 93, 135-147.