Eur. J. Biochem. 99, 345-351 (1979)
Studies on the Structure of the Rabbit Kidney Brush Border Mary T. C. KRAMERS and Garth B. ROBINSON Department of Biochemistry, Oxford University (Received March 20, 1979)
The effects of salts and non-ionic detergents on renal brush borders have been studied. 2 M sodium chloride, iodide or thiocyanate dissociated up to 40 o/, of the protein from the brush borders, destroying the core filaments and resulting in the formation of membrane vesicles; EDTA had a similar effect on structure but released little protein. Triton X-100 and Nonidet P-40 extracted up to 60 % of the protein including the major membrane glycoproteins and the enzymes trehalase, maltase and aminopeptidase (microsomal). Triton exhibited a selective effect on lipids removing phosphatidylserine, phosphatidylethanolamine and sphingomyelin but not the bulk of the phosphatidylcholine or cholesterol. The residual structures after Triton extraction comprised the core filaments associated with vesicles of lipid containing alkaline phosphatase and several other proteins. Treatment of these core-vesicle complexes with 2 M sodium chloride dissociated the filaments, releasing the vesicles which could be recovered as a pellicle on centrifugation. It is suggested that the proteins found in the vesicles might serve to interconnect the core filaments with the lipid bilayer. Studies of the topology of renal brush borders (reviewed in [l]) have shown them to be complex organelles. They comprise a lipoprotein membrane containing enzymes, a filamentous core containing actin [2,3], and a terminal web which, at least in the case of intestinal brush borders [4], contains myosin. Here we report attempts to 'dissect' renal brush borders using salts and detergents, correlating the release of brush border components with changes in morphology induced by these agents.
MATERIALS AND METHODS Isolation of Brush Borders Rabbits (Blue Beveren X Chinchilla, male 2.53 kg) were obtained from the Oxford University Farm. The renal brush borders were isolated as described [ 5 ] and were stored overnight at 0 ° C in 0.5 M sucrose prior to treatment. Treatment of the Brush Borders with Disruptive Agents Brush borders ( > 10 mg protein) were suspended (2.8 mg/ml) in disruptive agent dissolved in 0.5 M sucrose/lO mM Tris-HC1 pH 7.4. The mixture was Enzymes. Alkaline phosphatase (EC 3.1.3.1); maltase or a-Dglucoside glucohydrolase (EC 3.2.1.20); trehalase (EC 3.2.1.28); aminopeptidase (microsomal) (EC 3.4.11.2).
kept at 0°C for 15 min and then centrifuged at 100000 x g for 90 min; the soluble fraction was defined as the material not sedimented. In some experiments the brush borders were extracted with 1 % Triton X-100, centrifuged, and the residue extracted with 2 M NaCl; after the second centrifugation a pellicle and a pellet were obtained which were recovered separately. Analysis of Subfractions Enzyme Determination. All enzyme assays were performed at 37°C under conditions of zero order kinetics. Alkaline Phosphatase. The assay mixture contained 40 mM ethanolamine-HC1 buffer pH 10.5, 5 mM sodiump-nitrophenyl phosphate, 5 mM MgC12; the reaction was terminated by the addition of 2.5 vol. 0.1 M NaOH. p-Nitrophenol release was measured at 405 nm; p-nitrophenol was the standard. Maltase and Trehalase. The assays were an adaptation of the method of Dahlquist [6]. The reaction mixture contained 50 mM maleate buffer pH 6.25 and 28 mM maltose or trehalose. Glucose production was measured by adding 2.5 vol. Tris/glucose oxidase reagent and incubating at 37 "C for 60 min. Trisi glucose oxidase reagent contained 100 ml 0.5 M TrisHC1 buffer pH 7.0, 12.5 mg glucose oxidase (type 11, Sigma, London, Chemical Co. Ltd) 4 mg peroxidase type I1 (Boehringer) and 0.5 ml 1 % o-dianisidine in 95 % ethanol.
346
Studies on the Structure of the Rabbit Kidney Brush Border
Aminopeptidase M. The reaction mixture contained 50 mM imidazole buffer pH 7.5 and 1.6 mM leucine p-nitroanilide ; the reaction was terminated by the addition of 0.5 vol. 5 % (w/v) ZnS04. The p-nitroaniline released in the reaction was measured at 405 nm; p-nitroaniline was the standard. Sodium Dodecyl Sulphatel Polyacrylamide Gel Electrophoresis Sodium dodecyl sulphate/polyacrylamide gel electrophoresis was performed by the method of Quirk et al. [7]. Suspensions of brush borders at high ionic strength were dialysed against 0.5 M sucrose for 16 h before analysis. The gels were stained for carbohydrate (glycoprotein) by the periodic acid/Schiff reaction, and for protein with Coomassie blue.
layer thickness 0.5 mm (7.5 g silica gel), were prepared and activated at 110 "C for 30 min. 30 pg phospholipid phosphorus in chloroform was applied and the plates were developed with chloroform/methanol/7 M NH3 (37.5/22.5/3.75,v/v/v) which was allowed to migrate to the top of the plate. After drying for 5 min at 110 "C the second dimension was developed with chloroform/methanol/water/acetic acid (50/50/4/1, v/v/v/v), and the solvent was again allowed to migrate to the top of the plate. The plates were air dried, phosphatidylethanolamine and phosphatidylserine were detected with ninhydrin and the Vaskovsky reagent [13] was used to detect the remaining phospholipids. Spots were recovered by scraping, digested in 0.5 or 1.0 ml 70% (w/v) perchloric acid at 220°C for 1 h and assayed for inorganic phosphorus. Electron Microscopy
Chemical Determinations Protein was determined by the method of Lowry et al. [8] after precipitation by trichloroacetic acid at a final concentration of 10 % (w/v) at 0 "C. The precipitates were washed twice with 5 % (w/v) trichloroacetic acid, and lipids were extracted from the residues by washing ( 3 x 1 ml extractions) with methanol/chloroform/diethyl ether (1/1/2, by vol.). The protein was dissolved in 0.5 M NaOH at 90°C prior to assay; bovine serum albumin was the standard. For analysis, lipids were extracted with 20 vol. of chloroform/methanol (2/1, v/v) at room temperature overnight and the pellet re-extracted twice with 5 vol. Combined extracts were partitioned against CaCL by the method of Folch et al. [9] and the lower lipid phase was separated into neutral lipid and phospholipid fractions [lo]. Total cholesterol was determined on the neutral lipid fraction, after alkaline hydrolysis, by the Lieberman-Burchard reaction [l 11. Phospholipid phosphate was determined after digestion with 70 % (w/v) perchloric acid [12]. Thin-layer chromatography of phospholipids was performed as follows. Silica gel H plates (20 x 20 cm),
Pellets were fixed for 2 h in 10% (v/v) p-formaldehyde, 10 % (v/v) glutaraldehyde in 0.1 M sodium phosphate buffer, pH 7 . 3 , containing 0.1 % CaC12 and were then subsequently treated as previously described [14]. Thin sections were stained with uranyl acetate and lead citrate and examined in a Siemens Elmiskop 101 microscope. Negative staining was performed using uranyl acetate (1 % in water). RESULTS Disruption of the Brush Borders with Salt Solutions None of the salt solutions released significant amounts of enzyme at any of the concentrations used, although protein was released, this effect being most marked at high salt concentrations (2 M). Sodium thiocyanate and sodium iodide proved the most effective agents (Table 1). Dodecyl sulphate/polyacrylamide gel electrophoresis of the proteins released showed them to be of low molecular weight in the
Table 1. Disruption of kidney brush borders with salt solutions: release of protein and enzymes into supernatant All disruptive treatments were performed at 0 ° C for 15 min and the figures are the results of separate experiments (duplicate determinations). 10 mg of brush borders were extracted in each experiment. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x lo6 g x min. Percentage distribution is based on the total recovery which was between 90-110% Disruptive agent
Protein
Alkaline phosphatase
Aminopeptidase M
Maltase
0 04 09 19 24
03 03 13 29 39
13 0 0 75 30
Trehalase
%
-
10 mM EDTA pH 7 4 0 15 M NaCl 2 M NaCl 2 M NaI 2 M NaSCN
63 10 3 24 8 41 4 39 8
~~
~-
~~
03 0 0 11 05
M. T. C. Krdmers and G. B. Robinson
347
Fig. 1. Kidney brush borders. One brush border fragment is seen with many radiating microvilli. Each microvillus has a prominent central core. The glycocalyx can be seen at the periphery of the microvillus membrane. Magnification x 100000
Fig. 2. Brush borders treated with 2 M NuSCN. The membranes exhibit a beaded appearance. Extensive vesiculation has occurred and the microvilli appear to be devoid of central core filaments. Desmosomes may also be seen. Magnification x 65000
range 13000-80000 with major bands at 78000, 73000, 65000, 40000 and 20000; none of these proteins was glycosylated as judged by periodic acid/ Schiff staining. Small amounts of phospholipids were released by 2 M sodium thiocyanate and sodium iodide, 5 % and 8 % respectively; no cholesterol was released by any of the reagents. Electron micrographs of the untreated brush borders isolated in 0.5 M sucrose showed well defined brush borders with many microvilli radiating from the terminal web; apical inclusions could be seen basal to the microvilli. A prominent central core was seen when the microvilli were cut normal to the microvillus axis and a fuzzy coating, the glycocalyx, was seen at the outer periphery of the microvillus membrane (Fig. 1). 150 mM sodium chloride did not cause alteration in brush border morphology, whereas 10 mM EDTA produced extensive disruption of the brush borders to form vesicles, although this agent released
little protein (Table 1). Higher ionic strength extractions produced a marked change in brush border morphology. Layers of sheet-like membrane together with membrane vesicles were apparent and the core filaments, prominent in native brush borders, could not be seen. In addition, 2 M sodium thiocyanate and 2 mM sodium iodide produced a marked beading on the outer surface of the membrane bilayer (Fig. 2). The Disruption ojBrush Borders with Detergents
Non-ionic detergents released different amounts of protein from the brush borders. Tween 80 [polyoxyethylene (n = 20) sorbitan fatty acid ester] released < 9 % of the membrane protein (Table 2). Lubrol WX (polyoxyethylene cetyl alcohol) released up to 31 % of the protein. Nonidet P40 and Triton X-100 (isooctyl ethoxylates, n = 9, n = 10, respectively) at concentrations of 1.O % (w/v) (detergent/protein ratio,
Studies on the Structure of the Rabbit Kidney Brush Border
348
Table 2. Disruption of'brush borders tvitli nonionic detergents: re1c.asr nf enzymes into supernatant fraction All disruptive treatments were performed at O ' T for 15 min and the figures are the results of separate experiments (duplicate determinations). 10 mg of brush borders were extracted in each experiment. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x 10' g x min. Percentage distribution is based on the total recovery which was between 90- 110 '%, Disruptive agent
Alkaline phosphatase
~~
Tween 80 (w/v)
1 .O '%, 0.1
Maltase
-
.
~~
Trehalase
~-
0
cx,
0.01
-~
Aminopeptidase M
0 0
90%. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x lo6 g x min ~~
Phospholipid
Disruptive agent
Lubrol (w/v)
1.OX 0.1 "/, 0.01 '%,
40.2 (2) 6.0 (2) 3.1 (1)
Triton X-100 (w/v)
1.O% 0.1 "/,
54.4 (1) 25.3 (I)
Nonidet P40 (wiv)
1.O2) 0.1 7; 0.01
65.0 (2) 45.3 (2) 1.3 (1)
Cholesterol
amine, phosphatidylserine and phosphatidylinositol) were preferentially released, whilst phosphatidylcholine was largely retained in the residue (Table 4). After disruption of the brush borders with 1 % (w/v) Triton X-I 00 thin-section microscopy revealed core filaments in the residue which could be seen to radiate from terminal webs, the overlying membrane having apparently been removed ; desmosomes were also observed (Fig. 3). However, negative staining (Fig. 4) revealed vesicles still adhering to the filaments. The vesicles, which were only observed by negative staining, were smooth-surfaced and devoid of the projections (knobs) which were seen on the outer surface of the brush border membrane.
349
M. T. C. Kramers and G. B. Robinson
x
Table 4. Disruption of kidney brush borders with I (wjv) Triton X-100: release of phospholipids into supernatant The values reported are based on three separate determinations. 5 % . Recoveries were Standard deviation from the means was > 85 %. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x lo6 g x min Phospholipid class
Phospholipid
Phosphatidylethanolamine Phosphatidylinositol Phosphatidylserine Sphingomyelin lysophosphatidylcholine Phosphatidylcholine
13 58 86.5 82.4 18.5
x +
200000 and the remaining two had molecular weights of 1 5 000 and 70 000. Microscopy showed the pellet to comprise amorphous material and desmosome-like structures; at least ten proteins were present as judged by electrophoresis. Fig. 3. Brush borders treated with I % Triton X-100. Filaments may be seen radiating from a terminal web. The overlying membrane has been removed. Magnification x 26500
Sequentid Extraction of the Brush Borders Treatment of the brush borders first with 1 % (w/v) Triton X-100 (detergent/protein ratio, 4.4/1) and then with 2 M NaCl yielded a pellicle and a small pellet when the NaCl suspension was centrifuged. The pellicle consisted of smooth-surfaced membrane vesicles, as seen by negative staining, which were enriched in
DISCUSSION
A comparison of the results obtained with concentrated salts and with detergents shows that the two classes of reagents perturb the brush borders in very different ways. Salt extraction removed non-glycosylated, enzymically inactive proteins of low molecular weight, including a protein of molecular weight 40 000, probably actin. These changes were paralleled by a loss of core structures from the brush borders, coupled with vesiculation and fragmentation of the membrane structure. The solubilised proteins, which
Studies on the Structure of the Rabbit Kidney Brush Border
3 50
Table 5. Sequential extraction of brush borders with I ”/, Triton X-100 ( w l v ) and 2 M NaCl Brush borders were extracted for 15 min at 0’C with 1%) Triton-X100 in 0.5 M sucrose/lO mM Tris-HCI pH 7.4 and centrifuged at I00000 x g for 90 min. The residue was extracted for 15 min at 0 ;C with 2 M NaCl in 0.5 M sucrose/l0 mM Tris-HCI pH 7.4 and centrifuged at 100000 x g for 90 min. A pellicle floating above the supernatant, and a pellet at the bottom of the centrifuge tube were obtained. The pellicle and pellet were resuspended in 0.5 M sucrose. The fractions were analysed for phospholipid, protein, cholesterol and alkaline phosphatase. The recoveries were: phospholipid (100‘x); alkaline phosphatase (78 %); cholesterol (100%); protein (1 10 %). Alkaline phosphatase specific activity is expressed as pmolp-nitrophenol x min-’ x mg-’ at 37°C Fraction
Phospholipid
pmol Psim& protein 1%)Triton X-100 supernatant 2 M NaCl supernatant Triton residue pellicle Triton residue pellet Kidney brush borders
0.4 0 2.4 0.6 0.4
Alkaline phosphatase specific activity
‘x 67 0 24 9 100
pol x min-‘ x mg-‘ 0.35 1.59 35.40 5.34 3.1
comprise up to 40% of the brush border, may be regarded as peripheral in the terminology adopted by Singer and Nicholson El51 but they are clearly not peripheral in the sense that they are of little importance. The collapse of the brush border architecture as a consequence of their removal implies that these core proteins are necessary for the physical integrity of the structure. It could be argued that the salts exert chaotropic as well as electrostatic effects [I61 and SO may damage the bilayer structure. However, there was no evidence of phospholipid release under these conditions. Dilute EDTA appeared to destroy the membrane core filaments, leaving the membrane vesiculated, without releasing protein into the supernatant. This result suggests that divalent ions are required to maintain the structure of the core material. It is worthy of note that intestinal brush borders retain their structure in the presence of EDTA [17] and in this respect they differ from renal brush borders. The neutral detergents differed in effectiveness, with Triton X-100 and Nonidet P40 releasing more protein than Tween 80 and Lubrol WX (Table 2). The effects of Triton were studied in some detail since, in common with Nonidet, this detergent effected a selective release of membrane-bound enzymes, leaving the bulk of the alkaline phosphatase activity associated with the core material (Table 2). Removal of lipid from the membranes was also selective, since most of the cholesterol and phosphatidylcholine remained associated with the core material although a major part of the phosphatidylethanolamine, phosphatidylserine and sphingomyelin was released (Tables 3,4). A similar selective effect on the removal of phospholipids from erythrocyte ghosts by Triton has been recorded [IS, 191. It was of interest to note that the lipid remaining with the cores formed vesicles around the cores which could be seen by negative staining
Y, 10.7 17.0 58.0 14.2 100
Cholesterol
pmoli m& 0.15 0 2.30 0.85 0.26
Protein
7(,
m&
XI
43.0 0 36.0 21 .0 100
44.1 15.8 2.4 3.9 60.0
66.1 23.9 3.6 5.9 100
but not in thin sections. Triton X-100 has been used as a ‘stripping’ agent to reveal cytoskeletal structures [20,21] but it has not been generally realised that such preparations may contain remnants of membranes since they are not obvious in thin sections. This partial removal of lipid effected by Triton needs to be borne in mind in interpreting the results of such ‘stripping’ experiments. Further treatment of the core-vesicle complexes with 2 M sodium chloride apparently disrupted the cores, releasing the vesicles, which could be harvested as a pellicle after centrifugation. This material had a high 1ipid:protein ratio and contained a limited array of proteins, one being alkaline phosphatase which exhibited a very high specific activity (Table 5). The association of these lipid vesicles with the core material may be fortuitous, but it is tempting to speculate that it reflects some degree of organisation within the brush borders. Kidney [22] and intestinal [20,23] microvilli are reported to undergo concerted movement. If contraction of the microvilli occurs, mediated by the core filaments, it is reasonable to expect that a mechanical interconnection exists between the core and the membranes. It has been proposed that a-actinin may serve as cross-bridge between the membranes and the actin filaments [2,20] and the a-actinin presumably associates with integral proteins in the membrane. If this interaction is electrostatic then it is unlikely to be disrupted by a non-ionic detergent. The integral protein/@-actinin complex would remain attached to the core together with some lipid, or lipid and detergent, so forming vesicles. Treatment of these complexes with salts might then disrupt the integral protein-actinin interaction releasing the integral protein-lipid vesicles. Thus some of the proteins in the pellicle fraction may serve as ‘anchors’ connecting the membrane to the core. Similar ideas have been proposed for fibroblasts [21], macro-
M. T. C. Kramers and G. B. Robinson
phages [24], and smooth muscle cells [25]. The proteins which must connect the lipid bilayer with the cytoskeleton of the brush border have received little attention. The technique of selective sequential extraction reported here may provide a suitable method for their isolation. M. K. held a scholarship awarded by the M.R.C. for training in research methods during this study. The skilled technical help of Miss Jennifer Byrne is gratefully acknowledged.
REFERENCES 1. Kenny, A. J. & Booth, A. G. (1978) Essays Biochem. 14, 1-44. 2. Booth, A. G. & Kenny, A. J. (1976) Biochem. J. 159, 395-407. 3. Rostgaard, J. & Thuneberg, L. (1972) Z. Zelljorsch. Mikrosk. Anat. 132, 473 - 496. 4. Mooseker, M. S., Pollard, T. D. & Fujiwara, K. (1978) J. Cell Biol. 79, 444-453. 5. Quirk, S. J. & Robinson, G. B. (1972) Biochem. J. 128, 13191328. 6. Dahlquist, A. (1968) Anal. Biochem. 22, 99-107. 7. Quirk, S. J., Byrne, J. & Robinson, G. B. (1973) Biochem. J. 132, 501 - 508. 8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275.
351 9. Folch, J., Lees, M. & Sloane-Stanley, G. H. (1957) J. Biol. Chem. 226, 497 - 509. 10. Dole, V. P. (1956) J. Clin. Invest. 35, 150-154. 11. Stadtman, T. C. (1957) Methods Enzymol. 3, 392-394. 12. Bartlett, G. R. (1959) J . Biol. Chem. 234, 466-471. 13. Vaskovsky, V. E. & Kostelsky, E. Y. (1968) J. Lipid Res. 9, 396 - 403. 14. Sjostrand, F. J. (1967) Electron Microscopy of Cells and Ti.7sues, vol. 1 , pp. 138- 308, Academic Press, London. 15. Singer, S. J. & Nicholson, G. T. (1972) Science (Wash. D.C.) 175, 720-731. 16. Hatefi, Y. & Hanstein, E. G. (1969) Proc. Natl Acad. Sci. U.S.A. 62, 1129-1136. 17. Michell, R. H., Coleman, R. & Lewis, B. A. (1976) Biochem. SOC. Trans. 4,1017- 1020. 18. Yu, J., Fishman, D. A. & Steck, T. L. (1973) J. Supramol. Struct. I , 233-248. 19. Kirkpatrick, F. H., Godesky, S. E. & Marinetti, G. V. (1969) Biochemistry, 9, 50- 57. 20. Mooseker, M. J. (1976) J. Cell Biol. 71, 417-433. 21. Hunt, R. C. & Brown, J. C. (1978) Ann. N . Y . Acad. Sci. 312, 207 - 220. 22. Thuneberg, L. & Rostgaard, J. (1969) J. Ulirastruct. Res. 29, 578. 23. Sandstrom, B. (1971) Cytobiologie, 3, 293-297. 24. Stossel, T. P. & Hartwig, J. H. (1976) J. CellBiol. 68,602-619. 25. Wang, K., Ash, J. F. & Singer, S. J. (1975) Proc. Natl Acad. Sci. U.S.A. 72, 4483-4486.
M. T. C. Kramers, Department of Biophysical Chemistry, Royal Free Hospital School of Medicine. 8 Hunter Street, London, Great Britain, WClN IBP G. B. Robinson, Department of Biochemistry, Oxford University, South Parks Road, Oxford, Great Britain, OX1 3QU
Studies on the Structure of the Rabbit Kidney Brush Border Mary T. C. KRAMERS and Garth B. ROBINSON Department of Biochemistry, Oxford University (Received March 20, 1979)
The effects of salts and non-ionic detergents on renal brush borders have been studied. 2 M sodium chloride, iodide or thiocyanate dissociated up to 40 o/, of the protein from the brush borders, destroying the core filaments and resulting in the formation of membrane vesicles; EDTA had a similar effect on structure but released little protein. Triton X-100 and Nonidet P-40 extracted up to 60 % of the protein including the major membrane glycoproteins and the enzymes trehalase, maltase and aminopeptidase (microsomal). Triton exhibited a selective effect on lipids removing phosphatidylserine, phosphatidylethanolamine and sphingomyelin but not the bulk of the phosphatidylcholine or cholesterol. The residual structures after Triton extraction comprised the core filaments associated with vesicles of lipid containing alkaline phosphatase and several other proteins. Treatment of these core-vesicle complexes with 2 M sodium chloride dissociated the filaments, releasing the vesicles which could be recovered as a pellicle on centrifugation. It is suggested that the proteins found in the vesicles might serve to interconnect the core filaments with the lipid bilayer. Studies of the topology of renal brush borders (reviewed in [l]) have shown them to be complex organelles. They comprise a lipoprotein membrane containing enzymes, a filamentous core containing actin [2,3], and a terminal web which, at least in the case of intestinal brush borders [4], contains myosin. Here we report attempts to 'dissect' renal brush borders using salts and detergents, correlating the release of brush border components with changes in morphology induced by these agents.
MATERIALS AND METHODS Isolation of Brush Borders Rabbits (Blue Beveren X Chinchilla, male 2.53 kg) were obtained from the Oxford University Farm. The renal brush borders were isolated as described [ 5 ] and were stored overnight at 0 ° C in 0.5 M sucrose prior to treatment. Treatment of the Brush Borders with Disruptive Agents Brush borders ( > 10 mg protein) were suspended (2.8 mg/ml) in disruptive agent dissolved in 0.5 M sucrose/lO mM Tris-HC1 pH 7.4. The mixture was Enzymes. Alkaline phosphatase (EC 3.1.3.1); maltase or a-Dglucoside glucohydrolase (EC 3.2.1.20); trehalase (EC 3.2.1.28); aminopeptidase (microsomal) (EC 3.4.11.2).
kept at 0°C for 15 min and then centrifuged at 100000 x g for 90 min; the soluble fraction was defined as the material not sedimented. In some experiments the brush borders were extracted with 1 % Triton X-100, centrifuged, and the residue extracted with 2 M NaCl; after the second centrifugation a pellicle and a pellet were obtained which were recovered separately. Analysis of Subfractions Enzyme Determination. All enzyme assays were performed at 37°C under conditions of zero order kinetics. Alkaline Phosphatase. The assay mixture contained 40 mM ethanolamine-HC1 buffer pH 10.5, 5 mM sodiump-nitrophenyl phosphate, 5 mM MgC12; the reaction was terminated by the addition of 2.5 vol. 0.1 M NaOH. p-Nitrophenol release was measured at 405 nm; p-nitrophenol was the standard. Maltase and Trehalase. The assays were an adaptation of the method of Dahlquist [6]. The reaction mixture contained 50 mM maleate buffer pH 6.25 and 28 mM maltose or trehalose. Glucose production was measured by adding 2.5 vol. Tris/glucose oxidase reagent and incubating at 37 "C for 60 min. Trisi glucose oxidase reagent contained 100 ml 0.5 M TrisHC1 buffer pH 7.0, 12.5 mg glucose oxidase (type 11, Sigma, London, Chemical Co. Ltd) 4 mg peroxidase type I1 (Boehringer) and 0.5 ml 1 % o-dianisidine in 95 % ethanol.
346
Studies on the Structure of the Rabbit Kidney Brush Border
Aminopeptidase M. The reaction mixture contained 50 mM imidazole buffer pH 7.5 and 1.6 mM leucine p-nitroanilide ; the reaction was terminated by the addition of 0.5 vol. 5 % (w/v) ZnS04. The p-nitroaniline released in the reaction was measured at 405 nm; p-nitroaniline was the standard. Sodium Dodecyl Sulphatel Polyacrylamide Gel Electrophoresis Sodium dodecyl sulphate/polyacrylamide gel electrophoresis was performed by the method of Quirk et al. [7]. Suspensions of brush borders at high ionic strength were dialysed against 0.5 M sucrose for 16 h before analysis. The gels were stained for carbohydrate (glycoprotein) by the periodic acid/Schiff reaction, and for protein with Coomassie blue.
layer thickness 0.5 mm (7.5 g silica gel), were prepared and activated at 110 "C for 30 min. 30 pg phospholipid phosphorus in chloroform was applied and the plates were developed with chloroform/methanol/7 M NH3 (37.5/22.5/3.75,v/v/v) which was allowed to migrate to the top of the plate. After drying for 5 min at 110 "C the second dimension was developed with chloroform/methanol/water/acetic acid (50/50/4/1, v/v/v/v), and the solvent was again allowed to migrate to the top of the plate. The plates were air dried, phosphatidylethanolamine and phosphatidylserine were detected with ninhydrin and the Vaskovsky reagent [13] was used to detect the remaining phospholipids. Spots were recovered by scraping, digested in 0.5 or 1.0 ml 70% (w/v) perchloric acid at 220°C for 1 h and assayed for inorganic phosphorus. Electron Microscopy
Chemical Determinations Protein was determined by the method of Lowry et al. [8] after precipitation by trichloroacetic acid at a final concentration of 10 % (w/v) at 0 "C. The precipitates were washed twice with 5 % (w/v) trichloroacetic acid, and lipids were extracted from the residues by washing ( 3 x 1 ml extractions) with methanol/chloroform/diethyl ether (1/1/2, by vol.). The protein was dissolved in 0.5 M NaOH at 90°C prior to assay; bovine serum albumin was the standard. For analysis, lipids were extracted with 20 vol. of chloroform/methanol (2/1, v/v) at room temperature overnight and the pellet re-extracted twice with 5 vol. Combined extracts were partitioned against CaCL by the method of Folch et al. [9] and the lower lipid phase was separated into neutral lipid and phospholipid fractions [lo]. Total cholesterol was determined on the neutral lipid fraction, after alkaline hydrolysis, by the Lieberman-Burchard reaction [l 11. Phospholipid phosphate was determined after digestion with 70 % (w/v) perchloric acid [12]. Thin-layer chromatography of phospholipids was performed as follows. Silica gel H plates (20 x 20 cm),
Pellets were fixed for 2 h in 10% (v/v) p-formaldehyde, 10 % (v/v) glutaraldehyde in 0.1 M sodium phosphate buffer, pH 7 . 3 , containing 0.1 % CaC12 and were then subsequently treated as previously described [14]. Thin sections were stained with uranyl acetate and lead citrate and examined in a Siemens Elmiskop 101 microscope. Negative staining was performed using uranyl acetate (1 % in water). RESULTS Disruption of the Brush Borders with Salt Solutions None of the salt solutions released significant amounts of enzyme at any of the concentrations used, although protein was released, this effect being most marked at high salt concentrations (2 M). Sodium thiocyanate and sodium iodide proved the most effective agents (Table 1). Dodecyl sulphate/polyacrylamide gel electrophoresis of the proteins released showed them to be of low molecular weight in the
Table 1. Disruption of kidney brush borders with salt solutions: release of protein and enzymes into supernatant All disruptive treatments were performed at 0 ° C for 15 min and the figures are the results of separate experiments (duplicate determinations). 10 mg of brush borders were extracted in each experiment. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x lo6 g x min. Percentage distribution is based on the total recovery which was between 90-110% Disruptive agent
Protein
Alkaline phosphatase
Aminopeptidase M
Maltase
0 04 09 19 24
03 03 13 29 39
13 0 0 75 30
Trehalase
%
-
10 mM EDTA pH 7 4 0 15 M NaCl 2 M NaCl 2 M NaI 2 M NaSCN
63 10 3 24 8 41 4 39 8
~~
~-
~~
03 0 0 11 05
M. T. C. Krdmers and G. B. Robinson
347
Fig. 1. Kidney brush borders. One brush border fragment is seen with many radiating microvilli. Each microvillus has a prominent central core. The glycocalyx can be seen at the periphery of the microvillus membrane. Magnification x 100000
Fig. 2. Brush borders treated with 2 M NuSCN. The membranes exhibit a beaded appearance. Extensive vesiculation has occurred and the microvilli appear to be devoid of central core filaments. Desmosomes may also be seen. Magnification x 65000
range 13000-80000 with major bands at 78000, 73000, 65000, 40000 and 20000; none of these proteins was glycosylated as judged by periodic acid/ Schiff staining. Small amounts of phospholipids were released by 2 M sodium thiocyanate and sodium iodide, 5 % and 8 % respectively; no cholesterol was released by any of the reagents. Electron micrographs of the untreated brush borders isolated in 0.5 M sucrose showed well defined brush borders with many microvilli radiating from the terminal web; apical inclusions could be seen basal to the microvilli. A prominent central core was seen when the microvilli were cut normal to the microvillus axis and a fuzzy coating, the glycocalyx, was seen at the outer periphery of the microvillus membrane (Fig. 1). 150 mM sodium chloride did not cause alteration in brush border morphology, whereas 10 mM EDTA produced extensive disruption of the brush borders to form vesicles, although this agent released
little protein (Table 1). Higher ionic strength extractions produced a marked change in brush border morphology. Layers of sheet-like membrane together with membrane vesicles were apparent and the core filaments, prominent in native brush borders, could not be seen. In addition, 2 M sodium thiocyanate and 2 mM sodium iodide produced a marked beading on the outer surface of the membrane bilayer (Fig. 2). The Disruption ojBrush Borders with Detergents
Non-ionic detergents released different amounts of protein from the brush borders. Tween 80 [polyoxyethylene (n = 20) sorbitan fatty acid ester] released < 9 % of the membrane protein (Table 2). Lubrol WX (polyoxyethylene cetyl alcohol) released up to 31 % of the protein. Nonidet P40 and Triton X-100 (isooctyl ethoxylates, n = 9, n = 10, respectively) at concentrations of 1.O % (w/v) (detergent/protein ratio,
Studies on the Structure of the Rabbit Kidney Brush Border
348
Table 2. Disruption of'brush borders tvitli nonionic detergents: re1c.asr nf enzymes into supernatant fraction All disruptive treatments were performed at O ' T for 15 min and the figures are the results of separate experiments (duplicate determinations). 10 mg of brush borders were extracted in each experiment. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x 10' g x min. Percentage distribution is based on the total recovery which was between 90- 110 '%, Disruptive agent
Alkaline phosphatase
~~
Tween 80 (w/v)
1 .O '%, 0.1
Maltase
-
.
~~
Trehalase
~-
0
cx,
0.01
-~
Aminopeptidase M
0 0
90%. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x lo6 g x min ~~
Phospholipid
Disruptive agent
Lubrol (w/v)
1.OX 0.1 "/, 0.01 '%,
40.2 (2) 6.0 (2) 3.1 (1)
Triton X-100 (w/v)
1.O% 0.1 "/,
54.4 (1) 25.3 (I)
Nonidet P40 (wiv)
1.O2) 0.1 7; 0.01
65.0 (2) 45.3 (2) 1.3 (1)
Cholesterol
amine, phosphatidylserine and phosphatidylinositol) were preferentially released, whilst phosphatidylcholine was largely retained in the residue (Table 4). After disruption of the brush borders with 1 % (w/v) Triton X-I 00 thin-section microscopy revealed core filaments in the residue which could be seen to radiate from terminal webs, the overlying membrane having apparently been removed ; desmosomes were also observed (Fig. 3). However, negative staining (Fig. 4) revealed vesicles still adhering to the filaments. The vesicles, which were only observed by negative staining, were smooth-surfaced and devoid of the projections (knobs) which were seen on the outer surface of the brush border membrane.
349
M. T. C. Kramers and G. B. Robinson
x
Table 4. Disruption of kidney brush borders with I (wjv) Triton X-100: release of phospholipids into supernatant The values reported are based on three separate determinations. 5 % . Recoveries were Standard deviation from the means was > 85 %. The supernatant is defined as that fraction which was not sedimented by a centrifugal force of 9 x lo6 g x min Phospholipid class
Phospholipid
Phosphatidylethanolamine Phosphatidylinositol Phosphatidylserine Sphingomyelin lysophosphatidylcholine Phosphatidylcholine
13 58 86.5 82.4 18.5
x +
200000 and the remaining two had molecular weights of 1 5 000 and 70 000. Microscopy showed the pellet to comprise amorphous material and desmosome-like structures; at least ten proteins were present as judged by electrophoresis. Fig. 3. Brush borders treated with I % Triton X-100. Filaments may be seen radiating from a terminal web. The overlying membrane has been removed. Magnification x 26500
Sequentid Extraction of the Brush Borders Treatment of the brush borders first with 1 % (w/v) Triton X-100 (detergent/protein ratio, 4.4/1) and then with 2 M NaCl yielded a pellicle and a small pellet when the NaCl suspension was centrifuged. The pellicle consisted of smooth-surfaced membrane vesicles, as seen by negative staining, which were enriched in
DISCUSSION
A comparison of the results obtained with concentrated salts and with detergents shows that the two classes of reagents perturb the brush borders in very different ways. Salt extraction removed non-glycosylated, enzymically inactive proteins of low molecular weight, including a protein of molecular weight 40 000, probably actin. These changes were paralleled by a loss of core structures from the brush borders, coupled with vesiculation and fragmentation of the membrane structure. The solubilised proteins, which
Studies on the Structure of the Rabbit Kidney Brush Border
3 50
Table 5. Sequential extraction of brush borders with I ”/, Triton X-100 ( w l v ) and 2 M NaCl Brush borders were extracted for 15 min at 0’C with 1%) Triton-X100 in 0.5 M sucrose/lO mM Tris-HCI pH 7.4 and centrifuged at I00000 x g for 90 min. The residue was extracted for 15 min at 0 ;C with 2 M NaCl in 0.5 M sucrose/l0 mM Tris-HCI pH 7.4 and centrifuged at 100000 x g for 90 min. A pellicle floating above the supernatant, and a pellet at the bottom of the centrifuge tube were obtained. The pellicle and pellet were resuspended in 0.5 M sucrose. The fractions were analysed for phospholipid, protein, cholesterol and alkaline phosphatase. The recoveries were: phospholipid (100‘x); alkaline phosphatase (78 %); cholesterol (100%); protein (1 10 %). Alkaline phosphatase specific activity is expressed as pmolp-nitrophenol x min-’ x mg-’ at 37°C Fraction
Phospholipid
pmol Psim& protein 1%)Triton X-100 supernatant 2 M NaCl supernatant Triton residue pellicle Triton residue pellet Kidney brush borders
0.4 0 2.4 0.6 0.4
Alkaline phosphatase specific activity
‘x 67 0 24 9 100
pol x min-‘ x mg-‘ 0.35 1.59 35.40 5.34 3.1
comprise up to 40% of the brush border, may be regarded as peripheral in the terminology adopted by Singer and Nicholson El51 but they are clearly not peripheral in the sense that they are of little importance. The collapse of the brush border architecture as a consequence of their removal implies that these core proteins are necessary for the physical integrity of the structure. It could be argued that the salts exert chaotropic as well as electrostatic effects [I61 and SO may damage the bilayer structure. However, there was no evidence of phospholipid release under these conditions. Dilute EDTA appeared to destroy the membrane core filaments, leaving the membrane vesiculated, without releasing protein into the supernatant. This result suggests that divalent ions are required to maintain the structure of the core material. It is worthy of note that intestinal brush borders retain their structure in the presence of EDTA [17] and in this respect they differ from renal brush borders. The neutral detergents differed in effectiveness, with Triton X-100 and Nonidet P40 releasing more protein than Tween 80 and Lubrol WX (Table 2). The effects of Triton were studied in some detail since, in common with Nonidet, this detergent effected a selective release of membrane-bound enzymes, leaving the bulk of the alkaline phosphatase activity associated with the core material (Table 2). Removal of lipid from the membranes was also selective, since most of the cholesterol and phosphatidylcholine remained associated with the core material although a major part of the phosphatidylethanolamine, phosphatidylserine and sphingomyelin was released (Tables 3,4). A similar selective effect on the removal of phospholipids from erythrocyte ghosts by Triton has been recorded [IS, 191. It was of interest to note that the lipid remaining with the cores formed vesicles around the cores which could be seen by negative staining
Y, 10.7 17.0 58.0 14.2 100
Cholesterol
pmoli m& 0.15 0 2.30 0.85 0.26
Protein
7(,
m&
XI
43.0 0 36.0 21 .0 100
44.1 15.8 2.4 3.9 60.0
66.1 23.9 3.6 5.9 100
but not in thin sections. Triton X-100 has been used as a ‘stripping’ agent to reveal cytoskeletal structures [20,21] but it has not been generally realised that such preparations may contain remnants of membranes since they are not obvious in thin sections. This partial removal of lipid effected by Triton needs to be borne in mind in interpreting the results of such ‘stripping’ experiments. Further treatment of the core-vesicle complexes with 2 M sodium chloride apparently disrupted the cores, releasing the vesicles, which could be harvested as a pellicle after centrifugation. This material had a high 1ipid:protein ratio and contained a limited array of proteins, one being alkaline phosphatase which exhibited a very high specific activity (Table 5). The association of these lipid vesicles with the core material may be fortuitous, but it is tempting to speculate that it reflects some degree of organisation within the brush borders. Kidney [22] and intestinal [20,23] microvilli are reported to undergo concerted movement. If contraction of the microvilli occurs, mediated by the core filaments, it is reasonable to expect that a mechanical interconnection exists between the core and the membranes. It has been proposed that a-actinin may serve as cross-bridge between the membranes and the actin filaments [2,20] and the a-actinin presumably associates with integral proteins in the membrane. If this interaction is electrostatic then it is unlikely to be disrupted by a non-ionic detergent. The integral protein/@-actinin complex would remain attached to the core together with some lipid, or lipid and detergent, so forming vesicles. Treatment of these complexes with salts might then disrupt the integral protein-actinin interaction releasing the integral protein-lipid vesicles. Thus some of the proteins in the pellicle fraction may serve as ‘anchors’ connecting the membrane to the core. Similar ideas have been proposed for fibroblasts [21], macro-
M. T. C. Kramers and G. B. Robinson
phages [24], and smooth muscle cells [25]. The proteins which must connect the lipid bilayer with the cytoskeleton of the brush border have received little attention. The technique of selective sequential extraction reported here may provide a suitable method for their isolation. M. K. held a scholarship awarded by the M.R.C. for training in research methods during this study. The skilled technical help of Miss Jennifer Byrne is gratefully acknowledged.
REFERENCES 1. Kenny, A. J. & Booth, A. G. (1978) Essays Biochem. 14, 1-44. 2. Booth, A. G. & Kenny, A. J. (1976) Biochem. J. 159, 395-407. 3. Rostgaard, J. & Thuneberg, L. (1972) Z. Zelljorsch. Mikrosk. Anat. 132, 473 - 496. 4. Mooseker, M. S., Pollard, T. D. & Fujiwara, K. (1978) J. Cell Biol. 79, 444-453. 5. Quirk, S. J. & Robinson, G. B. (1972) Biochem. J. 128, 13191328. 6. Dahlquist, A. (1968) Anal. Biochem. 22, 99-107. 7. Quirk, S. J., Byrne, J. & Robinson, G. B. (1973) Biochem. J. 132, 501 - 508. 8. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275.
351 9. Folch, J., Lees, M. & Sloane-Stanley, G. H. (1957) J. Biol. Chem. 226, 497 - 509. 10. Dole, V. P. (1956) J. Clin. Invest. 35, 150-154. 11. Stadtman, T. C. (1957) Methods Enzymol. 3, 392-394. 12. Bartlett, G. R. (1959) J . Biol. Chem. 234, 466-471. 13. Vaskovsky, V. E. & Kostelsky, E. Y. (1968) J. Lipid Res. 9, 396 - 403. 14. Sjostrand, F. J. (1967) Electron Microscopy of Cells and Ti.7sues, vol. 1 , pp. 138- 308, Academic Press, London. 15. Singer, S. J. & Nicholson, G. T. (1972) Science (Wash. D.C.) 175, 720-731. 16. Hatefi, Y. & Hanstein, E. G. (1969) Proc. Natl Acad. Sci. U.S.A. 62, 1129-1136. 17. Michell, R. H., Coleman, R. & Lewis, B. A. (1976) Biochem. SOC. Trans. 4,1017- 1020. 18. Yu, J., Fishman, D. A. & Steck, T. L. (1973) J. Supramol. Struct. I , 233-248. 19. Kirkpatrick, F. H., Godesky, S. E. & Marinetti, G. V. (1969) Biochemistry, 9, 50- 57. 20. Mooseker, M. J. (1976) J. Cell Biol. 71, 417-433. 21. Hunt, R. C. & Brown, J. C. (1978) Ann. N . Y . Acad. Sci. 312, 207 - 220. 22. Thuneberg, L. & Rostgaard, J. (1969) J. Ulirastruct. Res. 29, 578. 23. Sandstrom, B. (1971) Cytobiologie, 3, 293-297. 24. Stossel, T. P. & Hartwig, J. H. (1976) J. CellBiol. 68,602-619. 25. Wang, K., Ash, J. F. & Singer, S. J. (1975) Proc. Natl Acad. Sci. U.S.A. 72, 4483-4486.
M. T. C. Kramers, Department of Biophysical Chemistry, Royal Free Hospital School of Medicine. 8 Hunter Street, London, Great Britain, WClN IBP G. B. Robinson, Department of Biochemistry, Oxford University, South Parks Road, Oxford, Great Britain, OX1 3QU
