Brain Research. 525 (1990) 92-10tl Elsevier
92 BRES 15765
Functional interactions of neuronal heparan sulphate proteoglycans with laminin R.J. R i o p e l l e I a n d K . E . D o w 2 Departments of 1Medicine (Neurology) and 2pediatrics (Neonatology), Queen's University, Kingston, Ont. (Canada) (Accepted 20 February 1990)
Key words." Neuron; Laminin; Proteoglycan; Neurite growth
Quantitative biosynthetic studies using cellular extracts and neuron conditioned medium demonstrated that heparan sulphate proteoglycans (HSPGs) comprised 20-25% of the sulphated proteoglycans produced by neurons while the remainder consisted of chondroitin sulphate proteoglycans (CSPGs). When chromatographic fractions containing guanidine extracted and partially purified proteoglycans from culture medium conditioned by neurons (NCM) were used to pretreat a laminin substrate, neurite formation by sensory neurons was enhanced. Enhanced neurite promoting activity was not apparent if, during the pretreatment of the laminin substrate with NCM, heparan sulphate glycosaminoglycans (HS) were present. To determine the molecular basis of cell surface HSPG interactions with immobilized laminin, adhesion and neurite growth by dissociated sensory neurons were quantified at 4 h in vitro - - a time at which there was no apparent contribution of released proteoglycans to neurite growth 6. Whereas adhesion was not influenced, neurite growth was partially inhibited in a dose-dependent manner if the sensory neurons were coincubated with HS, and if the cells were pretreated, prior to seeding, with heparitinase. The inhibitory effect produced by coincubation with saturating concentrations of HS was no longer apparent if the cells had been pretreated with heparitinase. These findings distinguish quantitatively between neurite growth on laminin and on laminin-HSPG complexes, and suggest that some neuronal cell surface and released HSPGs are involved in neurite growth by virtue of non-covalent interactions with glycosaminoglycan binding domains of laminin. INTRODUCTION Proteoglycans on the cell surface and in the extracellular matrix have been implicated in cell adhesion and in the regulation of neurite growth. Heparan sulphate proteoglycans (HSPGs) promote neurite growth in vitro that is morphologically distinct from that seen on laminin 12, and on fibronectin a heparan binding domain is involved in process elongation 25. A monoclonal antibody I N O (inhibitor of neurite outgrowth) which recognizes a laminin-heparan sulphate proteoglycan complex, inhibits the neurite growth-promoting activity of PC-12 p h e o c h r o m o c y t o m a cell conditioned medium 3'19 in vitro, and alters cranial neural crest migration in vivo 2. Monoclonal antibodies to a neuronal H S P G deplete neuron conditioned medium of neurite promoting activity is. Embryonic chick neural retinal cells in vitro condition culture medium with molecular species known as adherons 26,27. W h e n adsorbed onto tissue culture plastic adherons promote retinal cell attachment which can be inhibited by heparan sulphate. A n antiserum produced against H S P G labels retinal cell surface H S P G and inhibits the binding of cells to the adherons 26. More recently Cole and Glaser 4 have determined that a
component of adherons is immunologically identical to N - C A M ; these studies have confirmed earlier observations that N - C A M has a heparin/HS binding domain 5, and have provided evidence that this domain is involved in retinal cell-adheron interaction. In recent studies, D o w et al. 6 have demonstrated that embryonic neurons synthesize H S P G s which are both cell-associated and released into the tissue culture medium. The released HSPGs which could contain proteoglycan complexed growth factors have neurite-promoting activity when bound to laminin and the interaction of neurons with the H S P G - l a m i n i n complex appears to be mediated by a domain at or in juxtaposition to the HNK-1 (Leu 7) carbohydrate epitope of a family of cell surface glycoproteins. The present studies were designed to explore the molecular basis of the neurite promoting activity associated with neuronal HSPGs by probing the interactions of these H S P G s with laminin. MATERIALS AND METHODS
Cell culture Dorsal root ganglia (DRG) were removed from 8-day chick embryos and incubated in 0.01% trypsin in 0.05 mg/ml DNAse in Ca2÷,Mg2+-free Gey's balanced salt solution (CMF) for 15 min at
Correspondence." R.J. Riopelle, Apps Medical Research Centre c/o Doran 2, Room 6-200, Kingston General Hospital, Kingston, Ont. Canada K7L 2V7. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
93 37 °C. The reaction was stopped by removing the incubation solution and adding supplemented Ham's F12 medium 3° with 5% fetal calf serum (FCS). Following trituration the cell suspension was centrifuged (150 g for 8 min at 4 °C) through a concentrated FCS gradient, suspended in supplemented Ham's F12 medium with FCS, plated on a 100 mm Falcon culture dish in 4 ml of medium and incubated for 3 h at 37 °C, 5% CO 2. The non-adherent, neuronenriched population of cells was washed, resuspended in defined medium 1 with 4 pM NGF (final concentration) and used subsequently for bioassay or for Alcian blue labeling. Spinal cord cultures for production of conditioned medium were prepared as described for D R G cultures with the following exceptions: 0.1% trypsin for 30 min was used and cultures were established in defined medium I with 10% FCS for 24 h. Non preplated spinal cord cells were seeded at 105/well in 18 mm poly-D-lysine-coated wells. The medium was then changed to serum-free defined medium with 10/~M cytosine arabinoside. After a total of 7 days in culture, the medium conditioned by the cells (NCM) was removed, filtered through 0.2/~m Millipore filters and kept at 4 °C. For biosynthetic studies, some cultures were labeled for 7 days with Na235SO4 (20/~Ci/ml, New England Nuclear) and [3H]glucosamine (1/zCi/ml, New England Nuclear) in serum-free defined medium with cytosine arabinoside following a 24 h-period of exposure to the same medium without radioisotope.
Enzyme digestions Enzymatic digestions were carried out on dissociated D R G neurons following preplating (37 °C, 5% CO 2, 3 h) but prior to seeding in Terasaki microwells for bioassay. Chondroitinase ABC (Sigma) was used at 100 mU/ml in CMF pH 7.4 for 1 h at 37 °C and heparitinase (ICN) was used at 350 mU/ml in CMF with 5 mM calcium acetate for 1 h at 43 °C. Concentrations of enzymes used produced no proteolytic effect on [3H]leucine labeled proteins synthesized by neurons 6.
Bioassay Sensory neurons were prepared as described and seeded into wells of Terasaki microculture plates at a plating density of 500-1000 cells/well for assays of neurite formation and at 1000-1200 cells/well for assays of adhesion. Defined medium I with 4 pM NGF prepared by the method of Mobley et al. 2° was used in all cases. The wells of the Terasaki plates had been treated overnight (37 °C, 5% CO2) with poly-D-lysine (0.1 mg/ml), washed extensively, then treated overnight with 5 ~g/ml laminin (Bethesda Research Laboratories or a gift from S. Carbonetto) and washed. Subsequent treatment of wells with albumin-containing solutions made no difference to neuron adhesion or neurite growth in control studies and this blocking technique was not used in the biological studies. In some assays plates coated with poly-D-lysine were pretreated with laminin followed by NCM or fresh medium, or with NCM and then laminin or water over the same time period. In some experiments, heat-inactivated (56 °C, 30 min) rabbit anti-laminin and pre-immune serum (a gift from S. Carbonetto), glycosaminoglycans of heparan sulphate (HS) from bovine kidney (Sigma) or chondroitin sulphate (CS) from shark cartilage (Sigma) were added to the wells in a total volume of 10 ~1 at the time neurons were seeded. To assay for the effect of HS and CS on the binding of NCM to laminin, 10/A of spinal cord NCM with either of these glycosaminoglycans (5/~g/ml) was added to Terasaki microwells previously coated sequentially with poly-D-lysine (0.1 mg/ml) and laminin (5 /~g/mi). After an overnight incubation at 4 °C, the medium was removed, the wells washed thoroughly and D R G neurons added in serum-free defined medium with 4 pM NGE Plates were incubated at 4 *C for 30 min in the upright position, then at 4 °C for 30 min in the inverted position followed by incubation in the inverted position at 37 °C in a humidified atmosphere of 5% CO 2 and air for the remainder of the assay (180 min). The inverted well assay was developed 6 to minimize cell-cell interactions that result from settling of cells under 1 g sedimentation conditions in the tapered wells of Terasaki plates. Wells were
examined at 200x magnification under phase optics at 4 h and the number of cells with at least one process greater than 1.5 cell diameters was counted. Typically, 35-40% of seeded cells remained adherent to the wells after 30 min in the upright position and 210 min in the inverted position. Even at 4 h some 20% of these adherent cells had neurites in conditions where the substrate was laminin-NCM. Where neurons were seeded onto a laminin substrate the percent of adherent cells was similar to that seen on the laminin-NCM substrate, but at 4 h neurite extension was 8-10% of adherent cells. In a series of experiments to ascertain that gravity was not a factor in the effects produced by HS, the influence of HS on adhesion of neurons to wells that were not inverted was compared to adhesion to wells that had been handled according to the protocol.
Separation of proteoglycans by ion-exchange chromatography Proteoglycans were extracted from the NCM by addition of guanidine HC1 to a final concentration of 4 M. Extracts were passed over Sephadex G50 columns (bed volume 50 ml) to remove unincorporated radioactivity and to equilibrate the samples in the buffer used for ion-exchange chromatography (8 M urea, 0.05 M sodium acetate, 0.15 M sodium chloride, pH 6.0). Ion-exchange chromatography was performed by the method of Harper et a1.13 on a TSK DEAE-5PW column equilibrated in the above buffer and maintained at a flow rate of 1 ml/min. Samples of between 6 and 10 ml were loaded onto the column via a 2 ml sample loop. Unbound species were eluted by washing the column with buffer for 10 min. Bound species were eluted with a 30-min linear gradient of sodium chloride from 0.15 M to 1 M prepared in buffer and the column was re-equilibrated with buffer for 10 rain before the next sample was loaded. One-milliliter fractions were collected and aliquots were taken for liquid scintillation counting using Beckman ReadySolvEP. Prior to heparitinase digestion, pooled fractions from the ionexchange column were dialyzed against 0.1 M sodium acetate, 0.05 M Tris, 0.02 calcium acetate, pH 7.3. Prior to chondroitinase digestion fractions were dialyzed against the same buffer without calcium acetate. Digestion of the proteoglycans was carried out using 350 mU/ml heparitinase for 4 h at 43 °C and/or 100 mU/ml chondroitinase for 4 h at 37 °C. After incubation the samples were chromatographed on the DEAE-5PW columns as described above. Control samples were dialyzed against the above buffers and subjected to the same temperature conditions as for enzyme digestion. For bioassay, pooled fractions from the ion-exchange column were dialyzed for 3 days against 3 changes of phosphate-buffered saline. The samples were lyophilized, dissolved in defined medium, and dilutions as noted were used to treat microweUs previously coated with poly-D-lysine and laminin. After 24 h at 4 °C, the medium was removed and the wells washed thoroughly prior to seeding of D R G neurons in defined medium with NGF.
Separation of glycosaminoglycans by ion-exchange chromatography Following removal of unincorporated radioactivity as described above, glycosaminoglycans were isolated following the method of Lee et al) 6. Hyaluronic acid and chondroitin sulphate (both 50 /~g/ml of sample) were then added to each sample, which was then boiled for 10 min. After cooling, each sample was mixed with 3 vols. of ethanol (1.3% Na acetate) and left overnight at -20 °C. It was then centrifuged at 10,000 g (15 min) and the pellet was resuspended in 150 mM Tris HCI with 10 mM CaCI 2 and 100 mM glucosamine (pH 8.3). The sample was then incubated with pronase E (1.1 mg/100/A) overnight at 55 °C, boiled for 10 min, separated into subsamples, and precipitated with 3 vols. of ethanol (1.3% Na acetate). After centrifugation (10,000 g for 25 min) each pellet was resuspended in appropriate vehicle. For nitrous acid digestion of glycosaminoglycans, HNO 2 was prepared by adding 0.5 ml H20 and 0.5 ml 1 N n 2 s o 4 to 57 mg Ba(NO2) 2 at 4 °C. After centrifugation at 9000 g for 2 min, the supernatant (HNO2) was removed and added to the sample (4:1 HNO2:sample ) (v:v). After the sample had
94 warmed to ambient temperature, 1 M Na2CO3 was added (0.49 ml/ml of sample). The sample was then washed by precipitation in 3 vols. of ethanol, resuspended in H20, and chromatographed. For chondroitinase AC digestion samples were resuspended in 0.05 M Tris/HCl, 0.05 M Na acetate, 5 mg/ml BSA (pH 7.3), chondroitinase ABC or AC (0.i units/ml) was added, and digestion proceeded at 37 °C for 4 h. The samples were then boiled for 10 min, and precipitated overnight at -20 °C after addition of 3 vols. of ethanol. The GAGs were then resuspended in H20 and chromatographed. Samples were applied to a 5PW DEAE HPLC column and eluted with a NaCI gradient (0-1.0 M NaCI) in 1:9 methanol:H20 (v:v). The following program was used: 0 to 0.33 M NaCI (20 rain), 0.33-0.45 M (20 min), 0.45-0.64 M (20 rain), and 0.64-1.0 M (5 min). The flow rate was 0.6 ml/min, and the fraction volume was 0.6 ml. Aliquots of each fraction were counted for radioactivity.
enhanced neurite growth. To provide further evidence that a complex of laminin and N C M factors could
o x
In previous studies 6"23'z4 we have observed that NCM bound to poly-o-lysine and to laminin in a fashion that
15-
I"
-0.5
/ /
10-
o o
/
01 5O,
35 -
/" -'-'-- ~ = ==-" ¢ : : : : ¢--: : :~. --: 5 10 15 20 25 Elution Volume (rnl)
- - ~ . " : -" : : . 0.0 30 35
B
1.0
30-
? o x
/'~
2025-
i~ 0.5
1510I 5-~
Biosynthesis of proteoglycans
Biological activity associated with proteoglycans in neuron-conditioned medium
o
E
Differences between groups were compared by Student's t-test or analysis of variance and the method of least significant difference29.
In previous qualitative studies extraction of sensory and spinal cord neuronal cells or conditioned m e d i u m for glycosaminoglycans following 35504 labeling revealed the presence of chondroitin sulphate and a family of h e p a r a n sulphates 6 with a p p r o x i m a t e l y similar ratios of the various types of proteoglycans in m e d i u m and cell extracts. In the present experiments intact proteoglycans were isolated from conditioned m e d i u m following 35SO4labeling using ion exchange H P L C techniques (Fig. 1A). Proteoglycans were eluted as a large single peak between 0.5 and 0.6 M NaCI. Pooled fractions which included peaks of activity were subjected to digestion with either chondroitinase A B C or heparitinase and then rechromat o g r a p h e d . A s d e m o n s t r a t e d in Fig. 1A virtually all proteoglycans could be accounted for following digestion with the two enzymes. In a series (n = 5) of similar analyses 79 + 6% of the counts were i n c o r p o r a t e d into chondroitin sulphate proteoglycan while 26 + 1% of counts were in h e p a r a n sulphate proteoglycan. C h r o m a t o g r a p h y of 35504 and [3H]glucosamine-lab e l e d glycosaminoglycans on D E A E - 5 P W revealed two peaks of activity as depicted in Fig. l B . The b r o a d peak eluting before 0.45 M NaC1 was r e m o v e d by nitrous acid digestion identifying it as h e p a r a n sulphate (data not shown). The larger p e a k which accounted for 7 5 - 8 0 % of the glycosaminoglycan eluted at concentrations of NaC1 g r e a t e r than 0.5 M, and was r e m o v e d following chondroitinase A C digestion. Chondroitinase A C digests specifically chondroitin sulphate.
- 1.o
25-
Statistical analysis
RESULTS
A
30-
30
z
-o o'°-o,o, ~r .=- _ O ~ O ~ o ~ .
,o.
92 BRES 15765
Functional interactions of neuronal heparan sulphate proteoglycans with laminin R.J. R i o p e l l e I a n d K . E . D o w 2 Departments of 1Medicine (Neurology) and 2pediatrics (Neonatology), Queen's University, Kingston, Ont. (Canada) (Accepted 20 February 1990)
Key words." Neuron; Laminin; Proteoglycan; Neurite growth
Quantitative biosynthetic studies using cellular extracts and neuron conditioned medium demonstrated that heparan sulphate proteoglycans (HSPGs) comprised 20-25% of the sulphated proteoglycans produced by neurons while the remainder consisted of chondroitin sulphate proteoglycans (CSPGs). When chromatographic fractions containing guanidine extracted and partially purified proteoglycans from culture medium conditioned by neurons (NCM) were used to pretreat a laminin substrate, neurite formation by sensory neurons was enhanced. Enhanced neurite promoting activity was not apparent if, during the pretreatment of the laminin substrate with NCM, heparan sulphate glycosaminoglycans (HS) were present. To determine the molecular basis of cell surface HSPG interactions with immobilized laminin, adhesion and neurite growth by dissociated sensory neurons were quantified at 4 h in vitro - - a time at which there was no apparent contribution of released proteoglycans to neurite growth 6. Whereas adhesion was not influenced, neurite growth was partially inhibited in a dose-dependent manner if the sensory neurons were coincubated with HS, and if the cells were pretreated, prior to seeding, with heparitinase. The inhibitory effect produced by coincubation with saturating concentrations of HS was no longer apparent if the cells had been pretreated with heparitinase. These findings distinguish quantitatively between neurite growth on laminin and on laminin-HSPG complexes, and suggest that some neuronal cell surface and released HSPGs are involved in neurite growth by virtue of non-covalent interactions with glycosaminoglycan binding domains of laminin. INTRODUCTION Proteoglycans on the cell surface and in the extracellular matrix have been implicated in cell adhesion and in the regulation of neurite growth. Heparan sulphate proteoglycans (HSPGs) promote neurite growth in vitro that is morphologically distinct from that seen on laminin 12, and on fibronectin a heparan binding domain is involved in process elongation 25. A monoclonal antibody I N O (inhibitor of neurite outgrowth) which recognizes a laminin-heparan sulphate proteoglycan complex, inhibits the neurite growth-promoting activity of PC-12 p h e o c h r o m o c y t o m a cell conditioned medium 3'19 in vitro, and alters cranial neural crest migration in vivo 2. Monoclonal antibodies to a neuronal H S P G deplete neuron conditioned medium of neurite promoting activity is. Embryonic chick neural retinal cells in vitro condition culture medium with molecular species known as adherons 26,27. W h e n adsorbed onto tissue culture plastic adherons promote retinal cell attachment which can be inhibited by heparan sulphate. A n antiserum produced against H S P G labels retinal cell surface H S P G and inhibits the binding of cells to the adherons 26. More recently Cole and Glaser 4 have determined that a
component of adherons is immunologically identical to N - C A M ; these studies have confirmed earlier observations that N - C A M has a heparin/HS binding domain 5, and have provided evidence that this domain is involved in retinal cell-adheron interaction. In recent studies, D o w et al. 6 have demonstrated that embryonic neurons synthesize H S P G s which are both cell-associated and released into the tissue culture medium. The released HSPGs which could contain proteoglycan complexed growth factors have neurite-promoting activity when bound to laminin and the interaction of neurons with the H S P G - l a m i n i n complex appears to be mediated by a domain at or in juxtaposition to the HNK-1 (Leu 7) carbohydrate epitope of a family of cell surface glycoproteins. The present studies were designed to explore the molecular basis of the neurite promoting activity associated with neuronal HSPGs by probing the interactions of these H S P G s with laminin. MATERIALS AND METHODS
Cell culture Dorsal root ganglia (DRG) were removed from 8-day chick embryos and incubated in 0.01% trypsin in 0.05 mg/ml DNAse in Ca2÷,Mg2+-free Gey's balanced salt solution (CMF) for 15 min at
Correspondence." R.J. Riopelle, Apps Medical Research Centre c/o Doran 2, Room 6-200, Kingston General Hospital, Kingston, Ont. Canada K7L 2V7. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
93 37 °C. The reaction was stopped by removing the incubation solution and adding supplemented Ham's F12 medium 3° with 5% fetal calf serum (FCS). Following trituration the cell suspension was centrifuged (150 g for 8 min at 4 °C) through a concentrated FCS gradient, suspended in supplemented Ham's F12 medium with FCS, plated on a 100 mm Falcon culture dish in 4 ml of medium and incubated for 3 h at 37 °C, 5% CO 2. The non-adherent, neuronenriched population of cells was washed, resuspended in defined medium 1 with 4 pM NGF (final concentration) and used subsequently for bioassay or for Alcian blue labeling. Spinal cord cultures for production of conditioned medium were prepared as described for D R G cultures with the following exceptions: 0.1% trypsin for 30 min was used and cultures were established in defined medium I with 10% FCS for 24 h. Non preplated spinal cord cells were seeded at 105/well in 18 mm poly-D-lysine-coated wells. The medium was then changed to serum-free defined medium with 10/~M cytosine arabinoside. After a total of 7 days in culture, the medium conditioned by the cells (NCM) was removed, filtered through 0.2/~m Millipore filters and kept at 4 °C. For biosynthetic studies, some cultures were labeled for 7 days with Na235SO4 (20/~Ci/ml, New England Nuclear) and [3H]glucosamine (1/zCi/ml, New England Nuclear) in serum-free defined medium with cytosine arabinoside following a 24 h-period of exposure to the same medium without radioisotope.
Enzyme digestions Enzymatic digestions were carried out on dissociated D R G neurons following preplating (37 °C, 5% CO 2, 3 h) but prior to seeding in Terasaki microwells for bioassay. Chondroitinase ABC (Sigma) was used at 100 mU/ml in CMF pH 7.4 for 1 h at 37 °C and heparitinase (ICN) was used at 350 mU/ml in CMF with 5 mM calcium acetate for 1 h at 43 °C. Concentrations of enzymes used produced no proteolytic effect on [3H]leucine labeled proteins synthesized by neurons 6.
Bioassay Sensory neurons were prepared as described and seeded into wells of Terasaki microculture plates at a plating density of 500-1000 cells/well for assays of neurite formation and at 1000-1200 cells/well for assays of adhesion. Defined medium I with 4 pM NGF prepared by the method of Mobley et al. 2° was used in all cases. The wells of the Terasaki plates had been treated overnight (37 °C, 5% CO2) with poly-D-lysine (0.1 mg/ml), washed extensively, then treated overnight with 5 ~g/ml laminin (Bethesda Research Laboratories or a gift from S. Carbonetto) and washed. Subsequent treatment of wells with albumin-containing solutions made no difference to neuron adhesion or neurite growth in control studies and this blocking technique was not used in the biological studies. In some assays plates coated with poly-D-lysine were pretreated with laminin followed by NCM or fresh medium, or with NCM and then laminin or water over the same time period. In some experiments, heat-inactivated (56 °C, 30 min) rabbit anti-laminin and pre-immune serum (a gift from S. Carbonetto), glycosaminoglycans of heparan sulphate (HS) from bovine kidney (Sigma) or chondroitin sulphate (CS) from shark cartilage (Sigma) were added to the wells in a total volume of 10 ~1 at the time neurons were seeded. To assay for the effect of HS and CS on the binding of NCM to laminin, 10/A of spinal cord NCM with either of these glycosaminoglycans (5/~g/ml) was added to Terasaki microwells previously coated sequentially with poly-D-lysine (0.1 mg/ml) and laminin (5 /~g/mi). After an overnight incubation at 4 °C, the medium was removed, the wells washed thoroughly and D R G neurons added in serum-free defined medium with 4 pM NGE Plates were incubated at 4 *C for 30 min in the upright position, then at 4 °C for 30 min in the inverted position followed by incubation in the inverted position at 37 °C in a humidified atmosphere of 5% CO 2 and air for the remainder of the assay (180 min). The inverted well assay was developed 6 to minimize cell-cell interactions that result from settling of cells under 1 g sedimentation conditions in the tapered wells of Terasaki plates. Wells were
examined at 200x magnification under phase optics at 4 h and the number of cells with at least one process greater than 1.5 cell diameters was counted. Typically, 35-40% of seeded cells remained adherent to the wells after 30 min in the upright position and 210 min in the inverted position. Even at 4 h some 20% of these adherent cells had neurites in conditions where the substrate was laminin-NCM. Where neurons were seeded onto a laminin substrate the percent of adherent cells was similar to that seen on the laminin-NCM substrate, but at 4 h neurite extension was 8-10% of adherent cells. In a series of experiments to ascertain that gravity was not a factor in the effects produced by HS, the influence of HS on adhesion of neurons to wells that were not inverted was compared to adhesion to wells that had been handled according to the protocol.
Separation of proteoglycans by ion-exchange chromatography Proteoglycans were extracted from the NCM by addition of guanidine HC1 to a final concentration of 4 M. Extracts were passed over Sephadex G50 columns (bed volume 50 ml) to remove unincorporated radioactivity and to equilibrate the samples in the buffer used for ion-exchange chromatography (8 M urea, 0.05 M sodium acetate, 0.15 M sodium chloride, pH 6.0). Ion-exchange chromatography was performed by the method of Harper et a1.13 on a TSK DEAE-5PW column equilibrated in the above buffer and maintained at a flow rate of 1 ml/min. Samples of between 6 and 10 ml were loaded onto the column via a 2 ml sample loop. Unbound species were eluted by washing the column with buffer for 10 min. Bound species were eluted with a 30-min linear gradient of sodium chloride from 0.15 M to 1 M prepared in buffer and the column was re-equilibrated with buffer for 10 rain before the next sample was loaded. One-milliliter fractions were collected and aliquots were taken for liquid scintillation counting using Beckman ReadySolvEP. Prior to heparitinase digestion, pooled fractions from the ionexchange column were dialyzed against 0.1 M sodium acetate, 0.05 M Tris, 0.02 calcium acetate, pH 7.3. Prior to chondroitinase digestion fractions were dialyzed against the same buffer without calcium acetate. Digestion of the proteoglycans was carried out using 350 mU/ml heparitinase for 4 h at 43 °C and/or 100 mU/ml chondroitinase for 4 h at 37 °C. After incubation the samples were chromatographed on the DEAE-5PW columns as described above. Control samples were dialyzed against the above buffers and subjected to the same temperature conditions as for enzyme digestion. For bioassay, pooled fractions from the ion-exchange column were dialyzed for 3 days against 3 changes of phosphate-buffered saline. The samples were lyophilized, dissolved in defined medium, and dilutions as noted were used to treat microweUs previously coated with poly-D-lysine and laminin. After 24 h at 4 °C, the medium was removed and the wells washed thoroughly prior to seeding of D R G neurons in defined medium with NGF.
Separation of glycosaminoglycans by ion-exchange chromatography Following removal of unincorporated radioactivity as described above, glycosaminoglycans were isolated following the method of Lee et al) 6. Hyaluronic acid and chondroitin sulphate (both 50 /~g/ml of sample) were then added to each sample, which was then boiled for 10 min. After cooling, each sample was mixed with 3 vols. of ethanol (1.3% Na acetate) and left overnight at -20 °C. It was then centrifuged at 10,000 g (15 min) and the pellet was resuspended in 150 mM Tris HCI with 10 mM CaCI 2 and 100 mM glucosamine (pH 8.3). The sample was then incubated with pronase E (1.1 mg/100/A) overnight at 55 °C, boiled for 10 min, separated into subsamples, and precipitated with 3 vols. of ethanol (1.3% Na acetate). After centrifugation (10,000 g for 25 min) each pellet was resuspended in appropriate vehicle. For nitrous acid digestion of glycosaminoglycans, HNO 2 was prepared by adding 0.5 ml H20 and 0.5 ml 1 N n 2 s o 4 to 57 mg Ba(NO2) 2 at 4 °C. After centrifugation at 9000 g for 2 min, the supernatant (HNO2) was removed and added to the sample (4:1 HNO2:sample ) (v:v). After the sample had
94 warmed to ambient temperature, 1 M Na2CO3 was added (0.49 ml/ml of sample). The sample was then washed by precipitation in 3 vols. of ethanol, resuspended in H20, and chromatographed. For chondroitinase AC digestion samples were resuspended in 0.05 M Tris/HCl, 0.05 M Na acetate, 5 mg/ml BSA (pH 7.3), chondroitinase ABC or AC (0.i units/ml) was added, and digestion proceeded at 37 °C for 4 h. The samples were then boiled for 10 min, and precipitated overnight at -20 °C after addition of 3 vols. of ethanol. The GAGs were then resuspended in H20 and chromatographed. Samples were applied to a 5PW DEAE HPLC column and eluted with a NaCI gradient (0-1.0 M NaCI) in 1:9 methanol:H20 (v:v). The following program was used: 0 to 0.33 M NaCI (20 rain), 0.33-0.45 M (20 min), 0.45-0.64 M (20 rain), and 0.64-1.0 M (5 min). The flow rate was 0.6 ml/min, and the fraction volume was 0.6 ml. Aliquots of each fraction were counted for radioactivity.
enhanced neurite growth. To provide further evidence that a complex of laminin and N C M factors could
o x
In previous studies 6"23'z4 we have observed that NCM bound to poly-o-lysine and to laminin in a fashion that
15-
I"
-0.5
/ /
10-
o o
/
01 5O,
35 -
/" -'-'-- ~ = ==-" ¢ : : : : ¢--: : :~. --: 5 10 15 20 25 Elution Volume (rnl)
- - ~ . " : -" : : . 0.0 30 35
B
1.0
30-
? o x
/'~
2025-
i~ 0.5
1510I 5-~
Biosynthesis of proteoglycans
Biological activity associated with proteoglycans in neuron-conditioned medium
o
E
Differences between groups were compared by Student's t-test or analysis of variance and the method of least significant difference29.
In previous qualitative studies extraction of sensory and spinal cord neuronal cells or conditioned m e d i u m for glycosaminoglycans following 35504 labeling revealed the presence of chondroitin sulphate and a family of h e p a r a n sulphates 6 with a p p r o x i m a t e l y similar ratios of the various types of proteoglycans in m e d i u m and cell extracts. In the present experiments intact proteoglycans were isolated from conditioned m e d i u m following 35SO4labeling using ion exchange H P L C techniques (Fig. 1A). Proteoglycans were eluted as a large single peak between 0.5 and 0.6 M NaCI. Pooled fractions which included peaks of activity were subjected to digestion with either chondroitinase A B C or heparitinase and then rechromat o g r a p h e d . A s d e m o n s t r a t e d in Fig. 1A virtually all proteoglycans could be accounted for following digestion with the two enzymes. In a series (n = 5) of similar analyses 79 + 6% of the counts were i n c o r p o r a t e d into chondroitin sulphate proteoglycan while 26 + 1% of counts were in h e p a r a n sulphate proteoglycan. C h r o m a t o g r a p h y of 35504 and [3H]glucosamine-lab e l e d glycosaminoglycans on D E A E - 5 P W revealed two peaks of activity as depicted in Fig. l B . The b r o a d peak eluting before 0.45 M NaC1 was r e m o v e d by nitrous acid digestion identifying it as h e p a r a n sulphate (data not shown). The larger p e a k which accounted for 7 5 - 8 0 % of the glycosaminoglycan eluted at concentrations of NaC1 g r e a t e r than 0.5 M, and was r e m o v e d following chondroitinase A C digestion. Chondroitinase A C digests specifically chondroitin sulphate.
- 1.o
25-
Statistical analysis
RESULTS
A
30-
30
z
-o o'°-o,o, ~r .=- _ O ~ O ~ o ~ .
,o.
