- Email: [email protected]
Identification of proton-translocating adenosine triphosphatases in rat cerebral microvessels
128
Brain Research, 629 (1993) 128-132 © 1993 Elsevier Science Publishers B.V. All rights reserved 0006-8993/93/$06.00
BRES 19450
Identification of proton-translocating adenosine triphosphatases in rat cerebral microvessels A r s h a g D . M o o r a d i a n a,,,* a n d B a h a r B a s t a n i b , , a St. Louis VA Medical Center and the Divisions of ~ Endocrinology and b Nephrology, Department of Internal Medicine, St. Louis University School of Medicine, St. Louis, MO 63104 (USA) (Accepted 6 July 1993)
Key words: Proton pump; H +-ATPase; Blood-brain barrier; Acid-base balance
To determine if proton translocating adenosine triphosphatases (ATPases) can be localized at the blood-brain barrier, isolated rat cerebral microvessels and cerebral synaptosomal preparations were assayed for ATPase activity in the presence of various inhibitors. N-ethylmaleimidesensitivity could be consistently found in both cerebral microvessel and synaptosomal preparations. There was no vanadate sensitive component in the presence of ouabain, oligomycin and EGTA. Immunoblotting of cerebral microvessels and synaptosomes with a monoclonal antibody (Ell) against the 31 kDa subunit of the vacuolar type H-ATPase pump identified a discrete 31 kDa band. Diffuse immunocytochemical staining of cerebral cortical tissue, predominantly in choroid plexus, could be found with E u but not with HKaN2, an H,K-ATPase specific antibody, nor with a non-specific mouse monoclonal antibody (MOPC-21). Immunoblotting with HKaN2 showed an immunoreactive 76 kDa band, not present with the preimmune serum or the antibody preabsorbed with the immunizing synthetic peptide. It is concluded that the vacuolar type H-ATPase and not the gastric H,K-ATPase is present in cerebral tissue including cerebral microvessels and choroid plexus. Non-specific immunoreactivity may account for the 76 kDa band observed in the immunoblots using the HKaN2 antibody although presence of a degradation product of H,K-ATPase can not be ruled out. The functional role of the vacuolar H-ATPase in the blood-brain barrier remains to be determined.
INTRODUCTION
Previous studies on the physiology of acid-base balance in the central nervous system (CNS) have suggested that distribution of hydrogen ion (H ÷) or bicarbonate (HCO 3) between the cerebrospinal fluid (CSF) and blood is a passive process 1-3. However, Severinghaus et al. 4 found that neither H + or HCO 3 in CSF were in electrochemical equilibrium during hypocapnic alkalosis and suggested that these ions may cross the blood-brain barrier (BBB) by an active transport process. Recently it was found that carbonic anhydrase IV (CA_ IV) is expressed on the luminal surface of endothelial cells, of cerebral microvessels and this isozyme along with CA II may have an important role in acid-base homeostasis of the CNS 5. In addition, previous studies have identified the presence of 31 and 56 kDa subunits of the vacuolar H-ATPase in bovine brain 6. Most of the vacuolar proton pump in the brain may be associated with synaptic vesicles7. The presence
of the proton pump at the blood-brain barrier, the major gate keeper of the CNS, has been suspected, but never documented. To test this possibility inhibitor sensitivities of ATPases in isolated rat cerebral microvessels were studied and compared to the findings in synaptosomal membrane preparations. In addition, proton ATPases were identified with immunoblotting and immunocytochemical techniques. MATERIALS AND METHODS Cerebral microvessel and synaptosomal membrane preparations Male Fischer 344 rats were obtained from the National Cancer Institute (Bethesda, MD). New Zealand white rabbit cerebra were purchased from Rockland Farms (Rockland, MD). Human cerebral tissue was obtained from autopsy specimens within 6 h of death. These tissues were immediately frozen on dry ice. The cerebral microvessels were isolated as described previously 8-1°. The purity of the microvessels was documented with more than 20-fold enrichment of gamma glutamyl transpeptidase activity, a biochemical marker of cerebral microvessels (0.22 + 0.013 vs. 5.03 + 0.81 p.mol/mg/h, mean+ S.E.M.). The yield and purity data of rat cerebral microvessels have been r e p o r t e d in previous publicationss-x°.
* With the technical assistance of Liyang Yang and Gary Grabau. * Corresponding author. Address: Division of Endocrinology, St. Louis University Medical Center, 1402 S. Grand Blvd., St. Louis, MO 63104, USA. Fax: (1) (314) 771-0784.
129 The synaptosomal membrane fractions were prepared by the method of DeRobertis ~ using discontinuous sucrose gradient as previously described 12. The recovered membrane fraction from the sucrose gradient was washed in Tris buffer by centrifugation at 20,000× g for 20 rain twice. The enrichment of Na+-K+-ATPase activity in this fraction compared to total cerebral homogenates was approximately three-fold (0.135±0.013 vs. 0.441_+0.019 /xmol P i / an±n/rag). ATPase enzyme assay The final composition of the assay medium was 150 mM KCI, 30 mM Tris-HC1, pH 7.4, 30 mM 2-(N-morpholine) ethanesulfonic acid (MES), 6 mM MgCI 2, and 3 mM Na2ATP. In different experiments 1 mM N-ethyl male±re±de (NEM), 10/xg/ml oligomycin, 200/xM of dicyclohexyl carbodiimide (DCCD), 10 jxM of 7-chloro-4-nitrobenz2-oxa-l,3-diazole (NBD-CI), 10 /xM of p-chloromercuri phenylsulfonic acid (PCMBS) or 500 # M vanadate (Na2HVO 4) were added to the assay mixture. For assay of NEM-sensitive ATPase activity the ATPase activity in presence of NEM was subtracted from the activity in the absence of NEM. Similar additional experiments were carried out in the presence of 10 # g / m l oligomycin to obviate the contribution of mitochondrial ATPase. The vanadate sensitivity of the ATPase activity was also determined in the absence or presence of 10 /xg/ml oligomycin, 1 mM ouabain, and 0.1 mM EGTA. This assay medium was incubated at 37°C for 20 rain. The inorganic phosphate was determined color±metrically using the method of Daly and Ertingshausen~3. Gel electrophoresis and immunoblotting Primary antibodies. For immunocytochemistry, two different antibodies were used; E11, a mouse monoclonal antibody to the 31 kDa subunit of the vacuolar H-ATPase 14, (gift from Stephen Gluck, M.D., Washington University, St. Louis, MO), and HKaN2-IgG, the IgG fraction ~5 of rabbit antiserum developed to a synthetic peptide based on the N-terminus of hog gastric H,K-ATPase ~6 (gift from Adam Smolka, Ph.D., Medical University of South Carolina, Charleston, SC). For negative control MOPC-21, a non-specific mouse monoclonal antibody, and rabbit preimmune serum replaced E11 and HKaN2, respectively. Immunoblotting. Rat, rabbit, and human cerebral microvessels were prepared as described previously s-l°. Rat brain choroid plexus and synaptosomal membranes were prepared m'u. For comparison, rat kidney cortex and medulla were homogenized separately, and microsomal membranes were prepared as described before 17. We also prepared microsomal protein from rat gastric mucosa using the same protocol as for the kidney 17. Proteins (15 mcg of each sample) were resolved on 11.25% sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE) and transferred to nitrocellulose membrane by electroblotting. Membrane strips were probed with Ell, HKaN2-IgG, preimmune serum-IgG, HKaN2 preabsorbed with the immunizing synthetic peptide, and MOPC-21 using enhanced chemiluminescence (ECL) Western blotting technique (ECL Kit) as described in the manufacturer's protocol (Amersham Co., Arlington Heights, IL) Immunohistochemistry. Brain tissue was fixed in B5 (HgCI 2 0.22 M, sodium acetate 90 raM, formaldehyde 3.7%) overnight, and subsequently embedded in paraffin. Four-micron-thick sections were deparafinized in xylene solution, rehydrated in decreasing concentrations of ethanol, and were serially incubated in Lugol's iodine solution, 5% sodium thiosulfate, followed by phosphate buffered saline (PBS, pH 7.4). To block non-specific binding of the primary antibodies, sections were incubated in a blocking solution (10% calf serum, 10% goat serum, 1% polyethylene glycol, tool. wt, 20,000, in PBS) for 30 rain. Sections were then incubated overnight with 1 : 200 diluted (v:v, in the blocking solution) E~I ascites or HKaN2-1gG. Samples were subsequently rinsed in PBS, reblocked for 15 rain, and incubated in FITC-labeled goat anti-mouse or goat anti-rabbit IgG (each diluted at 1:50, v:v, in blocking solution) for 20 rain. Sections were rinsed in PBS and mounted in a fresh mixture of 2 mg/ml para-phenylene diamine (PPD) in 50% glycerol (v:v) in PBS. The slides were viewed using an inverted Nikon fluorescence microscope
(Nikon Diaphot-TMD, Nikon Inc., Instrument Group, New York, USA). Sections were also stained using the labelled streptavidin-biotin peroxidase-antiperoxidase technique (DAKO Laboratories Kit) as described in the manufacturer's protocol (DAKO Corporation, Carpinteria, CA). RESULTS
The sensitivity of the ATPase to various inhibitors is shown in Table I. In cerebral microvessels, DCCD resulted in almost 77% inhibition of ATPase activity (or 22.9 + 4.2% of control value) and oligomycin resulted in about 38% inhibition (61.8 + 1.8% of control). The NEM sensitivity was relatively modest with 1 mM NEM resulting in inhibition of ATPase activity to 78.9% of control level. Similar results were found for NEM inhibitory effect in the presence of 10 /zg/ml oligomycin. Vanadate sensitivity was similarly modest (78.0% of control) and this modest effect completely disappeared when testing was done in the presence of oligomycin, EGTA and ouabain. A similar profile of inhibitor sensitivity of the ATPase was noted in the synaptosomal membrane preparations except that oligomycin had a greater inhibitory, effect (32.7 + 2.7% of control in synaptosomes vs. 61.8 +_ 1.8% of control in cerebral microvessels). Fig. 1 shows the immunoblots using Ell monoclonal antibody. A prominent 31 kDa band with kidney medullary microsomal protein is shown next to a fainter band of the same size with rat cerebral microvessel protein. A similar band was observed with human and rabbit cerebral microvessels (immunoblots not shown).
TABLE I Inhibitor sensitiL,ities of ATPases of rat cerebral microt,essels (n = 9) and synaptosomal preparations (n = 5) The results are shown as percentages of control in the absence of inhibitor. The control for cerebral microvessels was 0.451_+0.028 /xmol P i / m i n / m g and for synaptosomal membranes was 0.218 ± 0.35 /xmol Pi/min/mg. Mean _+S.E.M.. See text for abbreviations-of inhibitors. Inhibitor Oligomycin (10 ~g/ml) NEM * (1 raM) DCCD (200/zM) NBD-C1 (10/xM) PCMBS (10 ~zM) Vanadate * * (10 txM)
Cerebral microt,essels
Synaptosomal membranes
61.8 +_1.8%
32.7 ± 2.7%
78.9 -+4.3%
82.6 ± 3.7%
22.9 ± 4.2%
25.8 ± 3.5%
90.4 + 1.9%
91.4 ± 3.5%
80.3 + 3.9%
77.5 ± 4.6%
78.0± 4.2%
71.8_+ 3.8%
* Similar results were found for NEM in the presence of oligomycin. ** Vanadate in the presence of ouabain (1 mM) and oligomycin (10 izg/ml) had no effect (see text for details).
130
1
2
31Kd rr"# ~ ~
i' " ~
3
.
.
.
.
4
" ~
staining of cerebral parenchymal tissue including cerebral microvessels. T h e greatest intensity of staining, however, was apparent in the epithelial cells of choroid plexus. To exclude the possibility that a non-specific lipophilic property may have resulted in the staining seen with E I~, rat peritoneal fat was fixed and stained using identical techniques. T h e r e was no immunostaining with Ell in the peritoneal fat. H K a N 2 resulted in no staining using the same protocol (figure not shown). DISCUSSION
Fig. 1. Immunoblots using El1 monoclonal antibody specific for the 31 kDa subunit of the vacuolar H-ATPase pump. A 31 kDa band seen in: Lane 1, renal medullary microsomes; Lane 2, cerebral microvessels; Lane 3, choroid plexus; Lane 4, synaptosomal membranes.
An immunoreactive protein of the same size was found in choroid plexus and synaptosomal m e m b r a n e s (Fig. 1). MOPC-21 used as a negative control for Ell did not reveal any immunoreactivity. W h e n immunoblots of cerebral microvessels were treated with H K a N 2 - I g G a prominent band was found at approximately 76 kDa while preimmune s e r u m - I g G or H K a N 2 - I g G preabsorbed with the immunizing synthetic peptide showed no immunoreactivity (Fig. 2). A 76 k D a immunoreactive protein was also found in synaptosomal membranes. This protein was distinctly smaller than the immunoreactive, approximately 94 kDa, protein found in rat gastric mucosa (Fig. 2). Sections of rat cerebral tissue stained with either Eal or MOPC-21 are shown in Fig. 3a and b, respectively. EI~ , but not MOPC-21, resulted in moderate
Mr
2
1
(Kd)
.
.
.
.
3
4
.
205 1 1 6 :
97.6
........
h
....
'
66
Fig. 2. Immunoblots using HKaN2 antibody againsts N-terminus of the gastric H,K-ATPase. Lane 1, synaptosomes with weak immunoreactivity at approximately 76 kDa. Lanes 2 and 3, cerebral microvessels. The immunoreactive band is seen at approximately 76 kDa. Lane 4, gastric mucosa microsomal protein. The major immunoreactive band is seen at approximately94 kDa.
One of the distinguishing features of various proton pumps is their inhibitor sensitivities 18,19. N E M is a specific inhibitor of the vacuolar type H - A T P a s e p u m p TM present in large amounts in some specialized renal tubular epithelial cells and osteoclasts. Oligomycin is a mitochondrial ATPase (F1F 0) inhibitor. D C C D inhibits both mitochondrial and vacuolar ATPases and to a lesser extent the phosphorylated ATPases, while vanadate is an inhibitor of the gastric H , K - A T P a s e (P type ATPase) as well as Na,K-ATPase, and Ca-ATPase. The significant inhibition of ATPase activity with oligomycin in cerebral microvessels and in synaptosomal preparations indicates the presence of large amounts of mitochondrial ATPase activity in both preparations. T h e most prominent inhibition of ATPase activity could be found in the presence of D C C D (Table I). The N E M sensitivity of the ATPase, an index of vacuolar type H - A T P a s e pump, was consistently discernible in cerebral microvessels even in the presence of oligomycin. The presence of this p u m p in cerebral microvessels was also supported by immunologic studies. The presence of H - A T P a s e in cerebral microvessels of two other mammalian species, namely rabbits and humans, was documented with immunoblotting. T h e molecular size of this proton p u m p subunit in cerebral microvessels is identical to that found in the kidney (31 kDa). The intensity of this band found in cerebral microvessels or choroid plexus was not significantly different (experiment repeated 3 times). However, when compared with the kidney the intensity of the 31 kDa band is weaker in immunoblots of cerebral microvessels. This corresponded with the modest NEM-sensitive ATPase activity found in our inhibitor sensitivity assays. Our immunohistologic data indicate that this proton p u m p is by no means restricted to cerebral microvessels. There was diffuse staining of the cerebral tissue which is consistent with the inhibitor sensitivity data of the synaptosomal m e m branes showing NEM-sensitive A T P a s e activity comparable to that of the cerebral microvessels. This is consistent with previously published reports of the pres-
131
,;~
6
Fig. 3. Immunohistochemical staining of rat cerebral cortex treated with (a) E H monoclonal antibody against the 31 kDa subunit of the vacuolar H-ATPase pump, and (b) MOPC-21 a non-specific mouse monoclonal antibody as negative control, Diffuse staining of rat cerebral parenchymal tissue, microvessels and choroid plexus is seen with E11 only.
ence of vacuolar proton pump in the brain, probably within the synaptic vesicles 7. The intensity of staining was somewhat more prominent in the epithelial cells of the choroid plexus suggesting that this H + translocating ATPase may have a more active role in the regulation of cerebrospinal fluid pH. Our immunoblotting studies also indicate that a peptide reactive with HKaN2, the H,K-ATPase specific antibody is present in cerebral microvessels. However, the molecular size of this protein (approximately 76 kDa) is significantly lower than the 94 kDa gastric H,K-ATPase a-subunit. Immunostaining with H K a N 2
antibody did not reveal reactivity in any regions of cerebral cortex. In addition, the vanadate sensitive ATPase activity could not be demonstrated in the presence of oligomycin and ouabain, indicating the non-specific nature of vanadate sensitivity of the enzyme in this tissue. It appears, therefore, that vanadate sensitive H,K-ATPase, found in the gastric mucosa is not present in significant quantities in central nervous system. An immunologically related protein which could represent a degradation product, however, with a molecular weight of 76 kDa is detectable in isolated cerebral microvessels, and to a lesser extent in synaptoso-
132 mal m e m b r a n e s , using the highly sensitive E C L techn i q u e of i m m u n o b l o t t i n g . T h e biological significance of this m o l e c u l e is n o t known. T h e a n t i b o d y against H,KA T P a s e was raised a g a i n s t a synthetic p e p t i d e b a s e d o n the N - t e r m i n u s of hog gastric H , K - A T P a s e . T h u s it is n o t u n e x p e c t e d that it m a y recognize a series of p r o t e i n b a n d s which may have a c o m m o n s e q u e n c e b u t are otherwise u n r e l a t e d . This study provides p h a r m a c o l o g i c a n d i m m u n o l o g i c evidence for the p r e s e n c e of v a c u o l a r type H ÷ - A T P a s e p u m p in the cerebral microvessels a n d s y n a p t o s o m a l m e m b r a n e s . However, the precise s y n a p t o s o m a l localization w o u l d r e q u i r e e l e c t r o n microscopic studies. T h e i m m u n o r e a c t i v i t y a n d i n h i b i t o r sensitivity data o n isolated c e r e b r a l microvessels a n d s y n a p t o s o m a l m e m b r a n e s indicate that the activity of this p u m p in the CNS is relatively weak c o m p a r e d to t h a t in the kidney. T h e p r e s e n c e of a f u n c t i o n a l H , K - A T P a s e in the CNS can n o t be d e m o n s t r a t e d with the p r e s e n t techniques. Acknowledgements. This work is supported by the Medical Research of the Department of Veterans Affairs (A.D.M.), and an extramural grant from Baxter Healthcare Corporation (B.B.).
REFERENCES 1 Pavlin, E.G. and Hornbein, T.F., Distribution of H ÷ and HCO 3 between CSF and blood during respiratory acidosis in dogs, Am. J. Physiol., 228 (1975) 1145-1148. 2 Hornbein, T.F. and Pavlin, E.G., Distribution of H ÷ and HCO 3 between CSF and blood during respiratory alkalosis in dogs, Am. Z Physiol., 228 (1975) 1149-1154. 3 Pavlin, E.G. and Hornbein, T.F., Distribution of H + and HCO 3 between CSF and blood during metabolic acidosis in dogs, Am. J. Physiol., 228 (1975) 1134-1140. 4 Severinghaus, J.W., Mitchell, R.A., Richardson, B.W. and Singer, M.M., Respiratory control at high altitude suggesting active transport regulations of CSF pH, J. Appl. Physiol., 18 (1963) 1155-1166.
5 Ghandour, M.S., Langley, O.K., Zhu, X.L., Waheed, A., and Sly, W.S., Carbonic anhydrase IV in brain capillary endothelial cells: a marker associated with the blood-brain barrier, Proc. Natl. Acad. Sci. USA, 89 (1992) 6823-6827. 6 Nelson, R.D., Guo, X.L., Masood, 1(i, Brown, D., Kalkbrenner, M. and Gluck, S., Selectively amplified expression of an isoform of the vacuolar H÷-ATPase 56-kilodalton subunit in renal intercalated cells, Proc. Natl. Acad. Sci. USA, 89 (1992) 3541-3545. 7 Sndhof, T.C. and Jahn, R, Proteins of synaptic vesicles involved in exocytosis and membrane recycling, Neuron, 6 (1991) 665-677. 8 Mooradian, A.D. and Smith, T.L., The effect of age on lipid composition and order of rat cerebral microvessels, Neurochem. Res., 17 (1992) 233-237. 9 Mooradian, A.D., Morin, A., Clip, L. and Haspel, H., Glucose transport is reduced in the blood-brain barrier of aged rats, Brain Res., 551 (1991) 145-149. 10 Mooradian, A.D. and Scarpace, P.J., /3-Adrenergic receptor activity of cerebral microvessels in experimental diabetes mellitus, Brain Res., 583 (1992) 155-160. 11 DeRobertis, E., Ultrastructure and cytochemistry of the synaptic region, Sc&nce, 156 (1967) 907-914. 12 Mooradian, A.D., Dickerson, F. and Smith, T.L., Lipid order and composition of synaptic membranes in experimental diabetes mellitus, Neurochem. Res., 15 (1990) 981-985. 13 Daly, J.A. and Ertingshausen, G., Direct method for determining inorganic phosphate in serum with the "Centrifichem", Clin. Chem., 18 (1972) 263-265. 14 Hemken, P.H., Xiao-Li, G., Zhi-Qiang, W., Kun, Z. and Gluck, S., Immunologic evidence that vacuolar H + ATPase with heterogeneous forms of Mr = 31,000 subunit have different membrane distributions in mammalian kidney, J. Biol. Chem., 267 (1992) 9948-9957. 15 Ey, P.L., Prowse, S.J. and Jenkin, C.R., Isolation of pure IgG1, IgG2, IgG2a, and IgG2b immunoglobulins from mouse serum using protein A-Sepharose, Biochemistry, 15 (1978) 429-436. 16 Smolka, A. and Swiger, K.M., Site-directed antibodies as topographical probes of the gastric H,K-ATPase a-subunit, Biochim. Biophys. Acta, 1108 (1992) 75-85. 17 Bastani, B., Purcell, H., Hemken, P.H., Trigg, D. and Gluck, S., Expression and distribution of renal vacuolar proton-translocating adenosine triphosphatase in response to chronic acid and alkali loads in the rat, J. Clin. Invest., 88 (1991) 126-136. 18 Forgac, M., Structure and function of vacuolar class of ATPdriven proton pumps, Physiol. Rev., 69 (1989) 765-796. 19 Stone, D.K., Crider, B.P. and Xie, X.-S., Structure of vacuolar proton pumps, Sere. Nephrol., 10 (1990) 159-165, 1990.
Brain Research, 629 (1993) 128-132 © 1993 Elsevier Science Publishers B.V. All rights reserved 0006-8993/93/$06.00
BRES 19450
Identification of proton-translocating adenosine triphosphatases in rat cerebral microvessels A r s h a g D . M o o r a d i a n a,,,* a n d B a h a r B a s t a n i b , , a St. Louis VA Medical Center and the Divisions of ~ Endocrinology and b Nephrology, Department of Internal Medicine, St. Louis University School of Medicine, St. Louis, MO 63104 (USA) (Accepted 6 July 1993)
Key words: Proton pump; H +-ATPase; Blood-brain barrier; Acid-base balance
To determine if proton translocating adenosine triphosphatases (ATPases) can be localized at the blood-brain barrier, isolated rat cerebral microvessels and cerebral synaptosomal preparations were assayed for ATPase activity in the presence of various inhibitors. N-ethylmaleimidesensitivity could be consistently found in both cerebral microvessel and synaptosomal preparations. There was no vanadate sensitive component in the presence of ouabain, oligomycin and EGTA. Immunoblotting of cerebral microvessels and synaptosomes with a monoclonal antibody (Ell) against the 31 kDa subunit of the vacuolar type H-ATPase pump identified a discrete 31 kDa band. Diffuse immunocytochemical staining of cerebral cortical tissue, predominantly in choroid plexus, could be found with E u but not with HKaN2, an H,K-ATPase specific antibody, nor with a non-specific mouse monoclonal antibody (MOPC-21). Immunoblotting with HKaN2 showed an immunoreactive 76 kDa band, not present with the preimmune serum or the antibody preabsorbed with the immunizing synthetic peptide. It is concluded that the vacuolar type H-ATPase and not the gastric H,K-ATPase is present in cerebral tissue including cerebral microvessels and choroid plexus. Non-specific immunoreactivity may account for the 76 kDa band observed in the immunoblots using the HKaN2 antibody although presence of a degradation product of H,K-ATPase can not be ruled out. The functional role of the vacuolar H-ATPase in the blood-brain barrier remains to be determined.
INTRODUCTION
Previous studies on the physiology of acid-base balance in the central nervous system (CNS) have suggested that distribution of hydrogen ion (H ÷) or bicarbonate (HCO 3) between the cerebrospinal fluid (CSF) and blood is a passive process 1-3. However, Severinghaus et al. 4 found that neither H + or HCO 3 in CSF were in electrochemical equilibrium during hypocapnic alkalosis and suggested that these ions may cross the blood-brain barrier (BBB) by an active transport process. Recently it was found that carbonic anhydrase IV (CA_ IV) is expressed on the luminal surface of endothelial cells, of cerebral microvessels and this isozyme along with CA II may have an important role in acid-base homeostasis of the CNS 5. In addition, previous studies have identified the presence of 31 and 56 kDa subunits of the vacuolar H-ATPase in bovine brain 6. Most of the vacuolar proton pump in the brain may be associated with synaptic vesicles7. The presence
of the proton pump at the blood-brain barrier, the major gate keeper of the CNS, has been suspected, but never documented. To test this possibility inhibitor sensitivities of ATPases in isolated rat cerebral microvessels were studied and compared to the findings in synaptosomal membrane preparations. In addition, proton ATPases were identified with immunoblotting and immunocytochemical techniques. MATERIALS AND METHODS Cerebral microvessel and synaptosomal membrane preparations Male Fischer 344 rats were obtained from the National Cancer Institute (Bethesda, MD). New Zealand white rabbit cerebra were purchased from Rockland Farms (Rockland, MD). Human cerebral tissue was obtained from autopsy specimens within 6 h of death. These tissues were immediately frozen on dry ice. The cerebral microvessels were isolated as described previously 8-1°. The purity of the microvessels was documented with more than 20-fold enrichment of gamma glutamyl transpeptidase activity, a biochemical marker of cerebral microvessels (0.22 + 0.013 vs. 5.03 + 0.81 p.mol/mg/h, mean+ S.E.M.). The yield and purity data of rat cerebral microvessels have been r e p o r t e d in previous publicationss-x°.
* With the technical assistance of Liyang Yang and Gary Grabau. * Corresponding author. Address: Division of Endocrinology, St. Louis University Medical Center, 1402 S. Grand Blvd., St. Louis, MO 63104, USA. Fax: (1) (314) 771-0784.
129 The synaptosomal membrane fractions were prepared by the method of DeRobertis ~ using discontinuous sucrose gradient as previously described 12. The recovered membrane fraction from the sucrose gradient was washed in Tris buffer by centrifugation at 20,000× g for 20 rain twice. The enrichment of Na+-K+-ATPase activity in this fraction compared to total cerebral homogenates was approximately three-fold (0.135±0.013 vs. 0.441_+0.019 /xmol P i / an±n/rag). ATPase enzyme assay The final composition of the assay medium was 150 mM KCI, 30 mM Tris-HC1, pH 7.4, 30 mM 2-(N-morpholine) ethanesulfonic acid (MES), 6 mM MgCI 2, and 3 mM Na2ATP. In different experiments 1 mM N-ethyl male±re±de (NEM), 10/xg/ml oligomycin, 200/xM of dicyclohexyl carbodiimide (DCCD), 10 jxM of 7-chloro-4-nitrobenz2-oxa-l,3-diazole (NBD-CI), 10 /xM of p-chloromercuri phenylsulfonic acid (PCMBS) or 500 # M vanadate (Na2HVO 4) were added to the assay mixture. For assay of NEM-sensitive ATPase activity the ATPase activity in presence of NEM was subtracted from the activity in the absence of NEM. Similar additional experiments were carried out in the presence of 10 # g / m l oligomycin to obviate the contribution of mitochondrial ATPase. The vanadate sensitivity of the ATPase activity was also determined in the absence or presence of 10 /xg/ml oligomycin, 1 mM ouabain, and 0.1 mM EGTA. This assay medium was incubated at 37°C for 20 rain. The inorganic phosphate was determined color±metrically using the method of Daly and Ertingshausen~3. Gel electrophoresis and immunoblotting Primary antibodies. For immunocytochemistry, two different antibodies were used; E11, a mouse monoclonal antibody to the 31 kDa subunit of the vacuolar H-ATPase 14, (gift from Stephen Gluck, M.D., Washington University, St. Louis, MO), and HKaN2-IgG, the IgG fraction ~5 of rabbit antiserum developed to a synthetic peptide based on the N-terminus of hog gastric H,K-ATPase ~6 (gift from Adam Smolka, Ph.D., Medical University of South Carolina, Charleston, SC). For negative control MOPC-21, a non-specific mouse monoclonal antibody, and rabbit preimmune serum replaced E11 and HKaN2, respectively. Immunoblotting. Rat, rabbit, and human cerebral microvessels were prepared as described previously s-l°. Rat brain choroid plexus and synaptosomal membranes were prepared m'u. For comparison, rat kidney cortex and medulla were homogenized separately, and microsomal membranes were prepared as described before 17. We also prepared microsomal protein from rat gastric mucosa using the same protocol as for the kidney 17. Proteins (15 mcg of each sample) were resolved on 11.25% sodium dodecyl sulfate-polyacrylamide gels (SDS-PAGE) and transferred to nitrocellulose membrane by electroblotting. Membrane strips were probed with Ell, HKaN2-IgG, preimmune serum-IgG, HKaN2 preabsorbed with the immunizing synthetic peptide, and MOPC-21 using enhanced chemiluminescence (ECL) Western blotting technique (ECL Kit) as described in the manufacturer's protocol (Amersham Co., Arlington Heights, IL) Immunohistochemistry. Brain tissue was fixed in B5 (HgCI 2 0.22 M, sodium acetate 90 raM, formaldehyde 3.7%) overnight, and subsequently embedded in paraffin. Four-micron-thick sections were deparafinized in xylene solution, rehydrated in decreasing concentrations of ethanol, and were serially incubated in Lugol's iodine solution, 5% sodium thiosulfate, followed by phosphate buffered saline (PBS, pH 7.4). To block non-specific binding of the primary antibodies, sections were incubated in a blocking solution (10% calf serum, 10% goat serum, 1% polyethylene glycol, tool. wt, 20,000, in PBS) for 30 rain. Sections were then incubated overnight with 1 : 200 diluted (v:v, in the blocking solution) E~I ascites or HKaN2-1gG. Samples were subsequently rinsed in PBS, reblocked for 15 rain, and incubated in FITC-labeled goat anti-mouse or goat anti-rabbit IgG (each diluted at 1:50, v:v, in blocking solution) for 20 rain. Sections were rinsed in PBS and mounted in a fresh mixture of 2 mg/ml para-phenylene diamine (PPD) in 50% glycerol (v:v) in PBS. The slides were viewed using an inverted Nikon fluorescence microscope
(Nikon Diaphot-TMD, Nikon Inc., Instrument Group, New York, USA). Sections were also stained using the labelled streptavidin-biotin peroxidase-antiperoxidase technique (DAKO Laboratories Kit) as described in the manufacturer's protocol (DAKO Corporation, Carpinteria, CA). RESULTS
The sensitivity of the ATPase to various inhibitors is shown in Table I. In cerebral microvessels, DCCD resulted in almost 77% inhibition of ATPase activity (or 22.9 + 4.2% of control value) and oligomycin resulted in about 38% inhibition (61.8 + 1.8% of control). The NEM sensitivity was relatively modest with 1 mM NEM resulting in inhibition of ATPase activity to 78.9% of control level. Similar results were found for NEM inhibitory effect in the presence of 10 /zg/ml oligomycin. Vanadate sensitivity was similarly modest (78.0% of control) and this modest effect completely disappeared when testing was done in the presence of oligomycin, EGTA and ouabain. A similar profile of inhibitor sensitivity of the ATPase was noted in the synaptosomal membrane preparations except that oligomycin had a greater inhibitory, effect (32.7 + 2.7% of control in synaptosomes vs. 61.8 +_ 1.8% of control in cerebral microvessels). Fig. 1 shows the immunoblots using Ell monoclonal antibody. A prominent 31 kDa band with kidney medullary microsomal protein is shown next to a fainter band of the same size with rat cerebral microvessel protein. A similar band was observed with human and rabbit cerebral microvessels (immunoblots not shown).
TABLE I Inhibitor sensitiL,ities of ATPases of rat cerebral microt,essels (n = 9) and synaptosomal preparations (n = 5) The results are shown as percentages of control in the absence of inhibitor. The control for cerebral microvessels was 0.451_+0.028 /xmol P i / m i n / m g and for synaptosomal membranes was 0.218 ± 0.35 /xmol Pi/min/mg. Mean _+S.E.M.. See text for abbreviations-of inhibitors. Inhibitor Oligomycin (10 ~g/ml) NEM * (1 raM) DCCD (200/zM) NBD-C1 (10/xM) PCMBS (10 ~zM) Vanadate * * (10 txM)
Cerebral microt,essels
Synaptosomal membranes
61.8 +_1.8%
32.7 ± 2.7%
78.9 -+4.3%
82.6 ± 3.7%
22.9 ± 4.2%
25.8 ± 3.5%
90.4 + 1.9%
91.4 ± 3.5%
80.3 + 3.9%
77.5 ± 4.6%
78.0± 4.2%
71.8_+ 3.8%
* Similar results were found for NEM in the presence of oligomycin. ** Vanadate in the presence of ouabain (1 mM) and oligomycin (10 izg/ml) had no effect (see text for details).
130
1
2
31Kd rr"# ~ ~
i' " ~
3
.
.
.
.
4
" ~
staining of cerebral parenchymal tissue including cerebral microvessels. T h e greatest intensity of staining, however, was apparent in the epithelial cells of choroid plexus. To exclude the possibility that a non-specific lipophilic property may have resulted in the staining seen with E I~, rat peritoneal fat was fixed and stained using identical techniques. T h e r e was no immunostaining with Ell in the peritoneal fat. H K a N 2 resulted in no staining using the same protocol (figure not shown). DISCUSSION
Fig. 1. Immunoblots using El1 monoclonal antibody specific for the 31 kDa subunit of the vacuolar H-ATPase pump. A 31 kDa band seen in: Lane 1, renal medullary microsomes; Lane 2, cerebral microvessels; Lane 3, choroid plexus; Lane 4, synaptosomal membranes.
An immunoreactive protein of the same size was found in choroid plexus and synaptosomal m e m b r a n e s (Fig. 1). MOPC-21 used as a negative control for Ell did not reveal any immunoreactivity. W h e n immunoblots of cerebral microvessels were treated with H K a N 2 - I g G a prominent band was found at approximately 76 kDa while preimmune s e r u m - I g G or H K a N 2 - I g G preabsorbed with the immunizing synthetic peptide showed no immunoreactivity (Fig. 2). A 76 k D a immunoreactive protein was also found in synaptosomal membranes. This protein was distinctly smaller than the immunoreactive, approximately 94 kDa, protein found in rat gastric mucosa (Fig. 2). Sections of rat cerebral tissue stained with either Eal or MOPC-21 are shown in Fig. 3a and b, respectively. EI~ , but not MOPC-21, resulted in moderate
Mr
2
1
(Kd)
.
.
.
.
3
4
.
205 1 1 6 :
97.6
........
h
....
'
66
Fig. 2. Immunoblots using HKaN2 antibody againsts N-terminus of the gastric H,K-ATPase. Lane 1, synaptosomes with weak immunoreactivity at approximately 76 kDa. Lanes 2 and 3, cerebral microvessels. The immunoreactive band is seen at approximately 76 kDa. Lane 4, gastric mucosa microsomal protein. The major immunoreactive band is seen at approximately94 kDa.
One of the distinguishing features of various proton pumps is their inhibitor sensitivities 18,19. N E M is a specific inhibitor of the vacuolar type H - A T P a s e p u m p TM present in large amounts in some specialized renal tubular epithelial cells and osteoclasts. Oligomycin is a mitochondrial ATPase (F1F 0) inhibitor. D C C D inhibits both mitochondrial and vacuolar ATPases and to a lesser extent the phosphorylated ATPases, while vanadate is an inhibitor of the gastric H , K - A T P a s e (P type ATPase) as well as Na,K-ATPase, and Ca-ATPase. The significant inhibition of ATPase activity with oligomycin in cerebral microvessels and in synaptosomal preparations indicates the presence of large amounts of mitochondrial ATPase activity in both preparations. T h e most prominent inhibition of ATPase activity could be found in the presence of D C C D (Table I). The N E M sensitivity of the ATPase, an index of vacuolar type H - A T P a s e pump, was consistently discernible in cerebral microvessels even in the presence of oligomycin. The presence of this p u m p in cerebral microvessels was also supported by immunologic studies. The presence of H - A T P a s e in cerebral microvessels of two other mammalian species, namely rabbits and humans, was documented with immunoblotting. T h e molecular size of this proton p u m p subunit in cerebral microvessels is identical to that found in the kidney (31 kDa). The intensity of this band found in cerebral microvessels or choroid plexus was not significantly different (experiment repeated 3 times). However, when compared with the kidney the intensity of the 31 kDa band is weaker in immunoblots of cerebral microvessels. This corresponded with the modest NEM-sensitive ATPase activity found in our inhibitor sensitivity assays. Our immunohistologic data indicate that this proton p u m p is by no means restricted to cerebral microvessels. There was diffuse staining of the cerebral tissue which is consistent with the inhibitor sensitivity data of the synaptosomal m e m branes showing NEM-sensitive A T P a s e activity comparable to that of the cerebral microvessels. This is consistent with previously published reports of the pres-
131
,;~
6
Fig. 3. Immunohistochemical staining of rat cerebral cortex treated with (a) E H monoclonal antibody against the 31 kDa subunit of the vacuolar H-ATPase pump, and (b) MOPC-21 a non-specific mouse monoclonal antibody as negative control, Diffuse staining of rat cerebral parenchymal tissue, microvessels and choroid plexus is seen with E11 only.
ence of vacuolar proton pump in the brain, probably within the synaptic vesicles 7. The intensity of staining was somewhat more prominent in the epithelial cells of the choroid plexus suggesting that this H + translocating ATPase may have a more active role in the regulation of cerebrospinal fluid pH. Our immunoblotting studies also indicate that a peptide reactive with HKaN2, the H,K-ATPase specific antibody is present in cerebral microvessels. However, the molecular size of this protein (approximately 76 kDa) is significantly lower than the 94 kDa gastric H,K-ATPase a-subunit. Immunostaining with H K a N 2
antibody did not reveal reactivity in any regions of cerebral cortex. In addition, the vanadate sensitive ATPase activity could not be demonstrated in the presence of oligomycin and ouabain, indicating the non-specific nature of vanadate sensitivity of the enzyme in this tissue. It appears, therefore, that vanadate sensitive H,K-ATPase, found in the gastric mucosa is not present in significant quantities in central nervous system. An immunologically related protein which could represent a degradation product, however, with a molecular weight of 76 kDa is detectable in isolated cerebral microvessels, and to a lesser extent in synaptoso-
132 mal m e m b r a n e s , using the highly sensitive E C L techn i q u e of i m m u n o b l o t t i n g . T h e biological significance of this m o l e c u l e is n o t known. T h e a n t i b o d y against H,KA T P a s e was raised a g a i n s t a synthetic p e p t i d e b a s e d o n the N - t e r m i n u s of hog gastric H , K - A T P a s e . T h u s it is n o t u n e x p e c t e d that it m a y recognize a series of p r o t e i n b a n d s which may have a c o m m o n s e q u e n c e b u t are otherwise u n r e l a t e d . This study provides p h a r m a c o l o g i c a n d i m m u n o l o g i c evidence for the p r e s e n c e of v a c u o l a r type H ÷ - A T P a s e p u m p in the cerebral microvessels a n d s y n a p t o s o m a l m e m b r a n e s . However, the precise s y n a p t o s o m a l localization w o u l d r e q u i r e e l e c t r o n microscopic studies. T h e i m m u n o r e a c t i v i t y a n d i n h i b i t o r sensitivity data o n isolated c e r e b r a l microvessels a n d s y n a p t o s o m a l m e m b r a n e s indicate that the activity of this p u m p in the CNS is relatively weak c o m p a r e d to t h a t in the kidney. T h e p r e s e n c e of a f u n c t i o n a l H , K - A T P a s e in the CNS can n o t be d e m o n s t r a t e d with the p r e s e n t techniques. Acknowledgements. This work is supported by the Medical Research of the Department of Veterans Affairs (A.D.M.), and an extramural grant from Baxter Healthcare Corporation (B.B.).
REFERENCES 1 Pavlin, E.G. and Hornbein, T.F., Distribution of H ÷ and HCO 3 between CSF and blood during respiratory acidosis in dogs, Am. J. Physiol., 228 (1975) 1145-1148. 2 Hornbein, T.F. and Pavlin, E.G., Distribution of H ÷ and HCO 3 between CSF and blood during respiratory alkalosis in dogs, Am. Z Physiol., 228 (1975) 1149-1154. 3 Pavlin, E.G. and Hornbein, T.F., Distribution of H + and HCO 3 between CSF and blood during metabolic acidosis in dogs, Am. J. Physiol., 228 (1975) 1134-1140. 4 Severinghaus, J.W., Mitchell, R.A., Richardson, B.W. and Singer, M.M., Respiratory control at high altitude suggesting active transport regulations of CSF pH, J. Appl. Physiol., 18 (1963) 1155-1166.
5 Ghandour, M.S., Langley, O.K., Zhu, X.L., Waheed, A., and Sly, W.S., Carbonic anhydrase IV in brain capillary endothelial cells: a marker associated with the blood-brain barrier, Proc. Natl. Acad. Sci. USA, 89 (1992) 6823-6827. 6 Nelson, R.D., Guo, X.L., Masood, 1(i, Brown, D., Kalkbrenner, M. and Gluck, S., Selectively amplified expression of an isoform of the vacuolar H÷-ATPase 56-kilodalton subunit in renal intercalated cells, Proc. Natl. Acad. Sci. USA, 89 (1992) 3541-3545. 7 Sndhof, T.C. and Jahn, R, Proteins of synaptic vesicles involved in exocytosis and membrane recycling, Neuron, 6 (1991) 665-677. 8 Mooradian, A.D. and Smith, T.L., The effect of age on lipid composition and order of rat cerebral microvessels, Neurochem. Res., 17 (1992) 233-237. 9 Mooradian, A.D., Morin, A., Clip, L. and Haspel, H., Glucose transport is reduced in the blood-brain barrier of aged rats, Brain Res., 551 (1991) 145-149. 10 Mooradian, A.D. and Scarpace, P.J., /3-Adrenergic receptor activity of cerebral microvessels in experimental diabetes mellitus, Brain Res., 583 (1992) 155-160. 11 DeRobertis, E., Ultrastructure and cytochemistry of the synaptic region, Sc&nce, 156 (1967) 907-914. 12 Mooradian, A.D., Dickerson, F. and Smith, T.L., Lipid order and composition of synaptic membranes in experimental diabetes mellitus, Neurochem. Res., 15 (1990) 981-985. 13 Daly, J.A. and Ertingshausen, G., Direct method for determining inorganic phosphate in serum with the "Centrifichem", Clin. Chem., 18 (1972) 263-265. 14 Hemken, P.H., Xiao-Li, G., Zhi-Qiang, W., Kun, Z. and Gluck, S., Immunologic evidence that vacuolar H + ATPase with heterogeneous forms of Mr = 31,000 subunit have different membrane distributions in mammalian kidney, J. Biol. Chem., 267 (1992) 9948-9957. 15 Ey, P.L., Prowse, S.J. and Jenkin, C.R., Isolation of pure IgG1, IgG2, IgG2a, and IgG2b immunoglobulins from mouse serum using protein A-Sepharose, Biochemistry, 15 (1978) 429-436. 16 Smolka, A. and Swiger, K.M., Site-directed antibodies as topographical probes of the gastric H,K-ATPase a-subunit, Biochim. Biophys. Acta, 1108 (1992) 75-85. 17 Bastani, B., Purcell, H., Hemken, P.H., Trigg, D. and Gluck, S., Expression and distribution of renal vacuolar proton-translocating adenosine triphosphatase in response to chronic acid and alkali loads in the rat, J. Clin. Invest., 88 (1991) 126-136. 18 Forgac, M., Structure and function of vacuolar class of ATPdriven proton pumps, Physiol. Rev., 69 (1989) 765-796. 19 Stone, D.K., Crider, B.P. and Xie, X.-S., Structure of vacuolar proton pumps, Sere. Nephrol., 10 (1990) 159-165, 1990.