- Email: [email protected]
A dot blot assay for sulfated glycosaminoglycans
Journal of Biochemical and Biophysical Methods, 18 (1989) 221-226 Elsevier
221
BBM 00735
A dot blot assay for sulfated glycosaminoglycans J. Scheerer, C. Matzenauer, S. K r e g e r a n d P. M u h n Z M B H Zentrurn )~r Molekulare Biologie Heidelberg Im Neuenheimer Feld 282, D-69 Heidelberg, F.R.G.
(Received 28 December 1988) (Accepted 23 January 1989)
Summary A dot blot assay for detection of low amounts of heparin and sulfated glycosaminoglycans (GAGs) is described. The detection range is between 25 ng/ml and 1000 n g / m l of heparin. The assay is based on the interference of sulfated GAGs with the binding of a synthetic ligand (described in this paper) to defined receptors like collagen type V and histones. Ligand binding to type V collagen was suppressed specifically by heparin, but not by other sulfated GAGs like heparan sulfate and chondroitin sulfate. Ligand binding to histones was suppressed most strongly by heparin, but also by chondroitin sulfate. Hyahironic acid did not interfere. Key words: Glycosaminoglycan; Heparin; Collagen; Histone
Introduction During a search for radioiodinatable analogues of a peptide, head activator [1], it had been noted that conjugates of this peptide with serum albumin were capable of binding to column fractions containing a high molecular weight immunoreactive form of this peptide in human serum and in the conditioned media of the cell line NH15CA2 (to be described elsewhere) [2,3]. From serum-free conditioned media of NH15CA2 cells this binding activity could be enriched by a factor 500 using heparin-Sepharose. The binding activity was located on a product that had been secreted by NH15CA2 cells since it was not present in nonconditioned media. It is reported here that the binding activity of this product is sensitive to collagenase like collagen is and that it bound heparin, but not chondroitin sulfate. Commercially Correspondence address: J. Scheerer, ZMBH Zentrum ftir Molekulare Biologie Heidelberg, Im Neuenheimer Feld 282, D-69 Heidelberg, F.R.G. Abbreviations: PBS, phosphate buffered saline; BSA, bovine serum albumin; MEM, minimal essential medium; FCS, fetal calf serum. 0165-022X/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
222
available batches of collagen type 5 mimicked this behaviour. Beyond collagen type V, a second binding species had been identified and these were the histones. They are distinguished from the endogenous activity in that their binding activity is not sensitive to collagenase and chondroitin sulfate interferes with ligand binding. Although the reason for ligand binding is not understood, the assay is very useful for measuring small amounts of heparin and other sulfated glycosaminoglycans.
Materials
Chemicals were from Merck, Darmstadt, F.R.G. or Roth, Karlsruhe, F.R.G. Lactatbumin, bovine serum albumin, collagens, histories (mixture of ali except H1), glycosaminoglycans, collagenase, purified from Clostridium histolyticum (EC 3.4.24.3), heparinases, neuraminidase and chondroitinase ABC were all from Sigma, Munich. Dexamethasone was from Serva, Heidelberg and Sigma. Media for cell culture and fetal calf serum were from Boehringer, Mannheim. Nitrocellulose (BA85) and the blotting device was from Schleicher & Schuell, Dassel, F.R.G. Peptides were from Bachem, Switzerland. Sep-Pak C18 cartridges were from Waters. Heparin-Sepharose was from Pharmacia, Freiburg, F.R.G.
Methods
Synthesis of lysylpeptide conjugates 0.001 mmol of E-lysyl residue was dissolved in 0.1 ml trifluoroethanol. 0.1 M potassium phosphate, p H 8.5 (4 ml), and a 500 × molar excess of glutaraldehyde was added and the mixture was incubated overnight at 4 ° C with continuous stirring. It was passed over a Sep-Pak column and washed with 10 ml of water and 10 ml of 20% methanol in water. The conjugated peptide was eluted with 5 ml of 80% methanol in water and 5 ml of 100% methanol and the combined eluates were dried in vacuo. The residue was dissolved in 5 ml of 0.1 M potassium phosphate, p H 8.5, and incubated overnight in the cold room with 200 nmoi of bovine serum albumin or, alternatively, lactalbumin. The protein-peptide conjugate was separated from low molecular weight constituents on G-25 Sephadex and was dialyzed against 2 1 of 10 mM Tris-HC1, p H 7.5, with two changes of buffer, each for 12 h.
Labelling of conjugates with 12sI The chloramine T method was used for labelling. 0.01 ml of 0.5 M sodium phosphate, pH 8, and 0.1 mCi of Na12SI (Amersham) were added to 0.001 mg of conjugate (1 mg/ml). 0.01 ml of chloramine T (3 m g / m l in phosphate buffer) were added and incubated for 1 rain. The reaction was stopped by addition of 0.01 ml sodium bisulfite (12 m g / m l ) and free iodine was compJexed with 0.1 ml of potassium iodide (1 g/ml). Labelled protein was separated on a G-25 column from free iodine and stored at - 2 0 ° C for maximally 6 weeks.
223
Dot blot assay for sulfated GAGs 0.1-0.45 ml of a sample (maximally 0.002-4 mg protein) was spotted on a dot blot manifold (8 × 12) (Schleicher & Schuell) and applied to nitrocellulose by suction. Washing is not needed. The nitrocellulose was cut in strips (1 × 12) and incubated in s u i t e d containers. The strips were preincubated for 1 - 2 h in 1% B S A / P B S in a volume of 1 - 4 ml. The radiolabelled conjugate was added at a concentration of 100000-150000 c p m / m l . Background on rfitrocellulose was usually 0.5-4% of t h e counts added. The background was generally lower (0.5-2%) when methanoI-precipitated BSA was used in the preincubation mix. Incubations were overnight at room temperature. For competition experiments, the competitor was added in a volume of 0.020-0.2 ml after preincubation of the mtrocellutose strips. After incubation the strips were washed in 0.1% B S A / P B S for 10 min with two changes. Ligand binding was practically not dependent on the length of washing. Washing times can be extended to one hour without affecting the result. The strips were cut i n t o single pieces (12 × 1) and counted in a gammacounter.
Cell culture and chromatography of conditioned media NH15CA2 cells were cultured in 75 cm 2 dishes at a density of 104--105 cells/ml in Dulbecco M E M supplemented with 10% FCS. To obtain conditioned media, cells were harvested by scraping and were cultured on Dulbecco M E M supplemented with insulin (5 rag/l) and transferrin (30 mg/1). Defined media were conditioned for 2 days, harvested and centrifuged for 10 rain at 1000 × g. They were appfied directly onto heparin-Sepharose (10 ml of wet gel per 200 ml of medium) at a flow rate of 10-20 m l / h . The column was washed with PBS and the b o u n d material was eluted with batches of 1.5 column volumes of PBS containing increasing concentrations of sodium chloride, as indicated in the figure. The eluates were dialyzed against PBS for 4 h at 4 ° C and analyzed for conjugate binding.
Resui*s and Discussion When conditioned defined media, obtained from early passages of N H 1 5 C A 2 celis, were selected on heparin-Sepharose it was found that the conjugate binding activity had bound completely to the column and eluted between 1 and 2 M of NaC1 (Fig. 1A). N o activity was found in the flowthrough of the heparin-Sepharose colunm. When heparin wag added to the incubation buffer at 20 m g / 1 it was found that the binding activity was blocked to 100%. Chondroitin sulfate, dextran and hyaluronic acid, all at the same concentration, did not affect binding. The effect was concentration dependent and half-maximal at 50 n g / m l (Fig. 1B). Heparin b o u n d rapidly; a 10 min incubation i n the presence of 20 rag/1 heparin, followed by removal of excess heparin, was sufficient to block the conjugate binding activity completely. U p o n treatment with several enzymes, a decrease in binding activity was observed only with bacterial collagenase but not with neuraminidase, chondroitinase or heparinase. Since bacterial collagenase is a highly specific enzyme it was
224
A
%,,
O J LID
FT
,15
.3
.5
"1
2
3
NaCr [M]
CDIVi 14
B 12
\
10
\°
e~ ® ~
\ Jm
0
~
P 0
~
2
3
4
log heparin [ng/mm] Fig. 1. Conjugate binding activity in conditioned defined media of NH15CA2 cells. (A) Heparin-Sepharose profi!e. Shown are the binding activity of the load (LD) and the flowthrough (FT) per volume of a conditioned defined medium (CDM). The binding activity enriched in the 1-2 M NaC1 fraction of the column. (B) Competition curve of conjugate binding activity with heparin. The fraction that eluted between 1 and 2 M NaC1 was dialyzed against PBS and assayed as described for binding in the presence of increasing amounts of hepari~. The binding activity is blocked by heparin and the effect is half-maximal at 50 n g / m l heparin.
225
A
4000I ~ [] ~
3000~
•
• .._...
HA
,[] CS
II
--
0L
T
0
,-
~"
,
1000
Hs
2000
3000
ng/ml
B 10000 r
~
l
i
.
.
.
.
.
.
•
HA
~ooo~\
cs
2000
0
~
.
i
! 1000
,
I 2000
, 3000
ng/ml
Fig. 2. Effectof heparin and sulfatedGAGs on the conjugatebinding activityassociatedwith collagen type V and histones.H, heparin; HS, heparan sulfate; CS, chondroitinsulfate; HA, hyaluronicacid. (A) CollagentypeV; (B) histones.
226 conjectured that the binding activity could be a collagen. When commercially available collagens were tested for a heparin-suppressible binding activity it had been :found indeed that some preparations (obtained from Sigma) of interstitia] collagen type V mimicked the endogenous activity. Binding was sensitive to bacterial collagenase and suppressible by heparin, but not by chondroitin sulfate or hyaluronic acid (Fig. 2a). Furthermore, binding was not affected by a commercially available heparan sulfate (Fig. 2a). The fibrillar collagens type I, II, Iii were different in that they did not bind the HA-conjugate (not shown). A suited control experiment was provided by the histones which had been accidentally identified as a conjugate binding activity. When histones were incubated with bacterial collagenase the binding activity remained stable which demonstrated the absence of proteolytic activities other than collagenase in the lots used. The binding activity of histories was suppressed by heparin in a similar dose response curve as it was for collagen type V. However, in contrast to the collagen type V binding activity, the histone binding activity was suppressed by heparan sulfate and chondroitin sulfate (Fig. 2b). These largely empirical data suggest that this method can be used for the determination of very small quantities of sulfated glycosaminoglycans, tt appears that sulfation is necessary because nonsulfaCed glycosaminoglycans, like hyaluronic acid and dextran did not interfere with ligand binding. Since the ligand is not itself a sugar molecule it should also be possible to include enzymatic digests such as chondroitmase and heparinase digests into the incubation mixture. This would simplify a qualitative assessment of the respective glycosaminoglycan. Simplified description of Che method and its application The ligand described in this paper is capable of binding to receptors (collagen type V, histones) that had been immobilised on nitrocellulose. Sulfated glycosaminoglycans suppress ligand binding to these receptors in small concentrations. The method is simple and suited for large sample numbers. Its application consists in the determination of very small quantities of sulfated glycosaminoglycans. Furthermore, the presence of glycolytic enzymes wilt not effect the determination since the ligand is a protein. Acknewledgements The members of the laboratory of H.C. Schaller are thanked for vivid discussion. This work has been supported by the Bundesministerium fibr Forschung and Technologie (BCT 365/1). References
1 Schaller,H.C. and Bodenmf~lter,H. (1985) Bid. Chem. Hoppe Scyler 366, 1003-1007. 2 Heumann, R., Ocalan, M., Kachel, V. and Hamprecht, B. (1979) Proc. Natl. Acad. Sci. USA 76, 4574-4577. 3 Roberge, M. (1985) Doctoral thesis, University of Heidelberg,Heidelberg, F.R.G,
221
BBM 00735
A dot blot assay for sulfated glycosaminoglycans J. Scheerer, C. Matzenauer, S. K r e g e r a n d P. M u h n Z M B H Zentrurn )~r Molekulare Biologie Heidelberg Im Neuenheimer Feld 282, D-69 Heidelberg, F.R.G.
(Received 28 December 1988) (Accepted 23 January 1989)
Summary A dot blot assay for detection of low amounts of heparin and sulfated glycosaminoglycans (GAGs) is described. The detection range is between 25 ng/ml and 1000 n g / m l of heparin. The assay is based on the interference of sulfated GAGs with the binding of a synthetic ligand (described in this paper) to defined receptors like collagen type V and histones. Ligand binding to type V collagen was suppressed specifically by heparin, but not by other sulfated GAGs like heparan sulfate and chondroitin sulfate. Ligand binding to histones was suppressed most strongly by heparin, but also by chondroitin sulfate. Hyahironic acid did not interfere. Key words: Glycosaminoglycan; Heparin; Collagen; Histone
Introduction During a search for radioiodinatable analogues of a peptide, head activator [1], it had been noted that conjugates of this peptide with serum albumin were capable of binding to column fractions containing a high molecular weight immunoreactive form of this peptide in human serum and in the conditioned media of the cell line NH15CA2 (to be described elsewhere) [2,3]. From serum-free conditioned media of NH15CA2 cells this binding activity could be enriched by a factor 500 using heparin-Sepharose. The binding activity was located on a product that had been secreted by NH15CA2 cells since it was not present in nonconditioned media. It is reported here that the binding activity of this product is sensitive to collagenase like collagen is and that it bound heparin, but not chondroitin sulfate. Commercially Correspondence address: J. Scheerer, ZMBH Zentrum ftir Molekulare Biologie Heidelberg, Im Neuenheimer Feld 282, D-69 Heidelberg, F.R.G. Abbreviations: PBS, phosphate buffered saline; BSA, bovine serum albumin; MEM, minimal essential medium; FCS, fetal calf serum. 0165-022X/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
222
available batches of collagen type 5 mimicked this behaviour. Beyond collagen type V, a second binding species had been identified and these were the histones. They are distinguished from the endogenous activity in that their binding activity is not sensitive to collagenase and chondroitin sulfate interferes with ligand binding. Although the reason for ligand binding is not understood, the assay is very useful for measuring small amounts of heparin and other sulfated glycosaminoglycans.
Materials
Chemicals were from Merck, Darmstadt, F.R.G. or Roth, Karlsruhe, F.R.G. Lactatbumin, bovine serum albumin, collagens, histories (mixture of ali except H1), glycosaminoglycans, collagenase, purified from Clostridium histolyticum (EC 3.4.24.3), heparinases, neuraminidase and chondroitinase ABC were all from Sigma, Munich. Dexamethasone was from Serva, Heidelberg and Sigma. Media for cell culture and fetal calf serum were from Boehringer, Mannheim. Nitrocellulose (BA85) and the blotting device was from Schleicher & Schuell, Dassel, F.R.G. Peptides were from Bachem, Switzerland. Sep-Pak C18 cartridges were from Waters. Heparin-Sepharose was from Pharmacia, Freiburg, F.R.G.
Methods
Synthesis of lysylpeptide conjugates 0.001 mmol of E-lysyl residue was dissolved in 0.1 ml trifluoroethanol. 0.1 M potassium phosphate, p H 8.5 (4 ml), and a 500 × molar excess of glutaraldehyde was added and the mixture was incubated overnight at 4 ° C with continuous stirring. It was passed over a Sep-Pak column and washed with 10 ml of water and 10 ml of 20% methanol in water. The conjugated peptide was eluted with 5 ml of 80% methanol in water and 5 ml of 100% methanol and the combined eluates were dried in vacuo. The residue was dissolved in 5 ml of 0.1 M potassium phosphate, p H 8.5, and incubated overnight in the cold room with 200 nmoi of bovine serum albumin or, alternatively, lactalbumin. The protein-peptide conjugate was separated from low molecular weight constituents on G-25 Sephadex and was dialyzed against 2 1 of 10 mM Tris-HC1, p H 7.5, with two changes of buffer, each for 12 h.
Labelling of conjugates with 12sI The chloramine T method was used for labelling. 0.01 ml of 0.5 M sodium phosphate, pH 8, and 0.1 mCi of Na12SI (Amersham) were added to 0.001 mg of conjugate (1 mg/ml). 0.01 ml of chloramine T (3 m g / m l in phosphate buffer) were added and incubated for 1 rain. The reaction was stopped by addition of 0.01 ml sodium bisulfite (12 m g / m l ) and free iodine was compJexed with 0.1 ml of potassium iodide (1 g/ml). Labelled protein was separated on a G-25 column from free iodine and stored at - 2 0 ° C for maximally 6 weeks.
223
Dot blot assay for sulfated GAGs 0.1-0.45 ml of a sample (maximally 0.002-4 mg protein) was spotted on a dot blot manifold (8 × 12) (Schleicher & Schuell) and applied to nitrocellulose by suction. Washing is not needed. The nitrocellulose was cut in strips (1 × 12) and incubated in s u i t e d containers. The strips were preincubated for 1 - 2 h in 1% B S A / P B S in a volume of 1 - 4 ml. The radiolabelled conjugate was added at a concentration of 100000-150000 c p m / m l . Background on rfitrocellulose was usually 0.5-4% of t h e counts added. The background was generally lower (0.5-2%) when methanoI-precipitated BSA was used in the preincubation mix. Incubations were overnight at room temperature. For competition experiments, the competitor was added in a volume of 0.020-0.2 ml after preincubation of the mtrocellutose strips. After incubation the strips were washed in 0.1% B S A / P B S for 10 min with two changes. Ligand binding was practically not dependent on the length of washing. Washing times can be extended to one hour without affecting the result. The strips were cut i n t o single pieces (12 × 1) and counted in a gammacounter.
Cell culture and chromatography of conditioned media NH15CA2 cells were cultured in 75 cm 2 dishes at a density of 104--105 cells/ml in Dulbecco M E M supplemented with 10% FCS. To obtain conditioned media, cells were harvested by scraping and were cultured on Dulbecco M E M supplemented with insulin (5 rag/l) and transferrin (30 mg/1). Defined media were conditioned for 2 days, harvested and centrifuged for 10 rain at 1000 × g. They were appfied directly onto heparin-Sepharose (10 ml of wet gel per 200 ml of medium) at a flow rate of 10-20 m l / h . The column was washed with PBS and the b o u n d material was eluted with batches of 1.5 column volumes of PBS containing increasing concentrations of sodium chloride, as indicated in the figure. The eluates were dialyzed against PBS for 4 h at 4 ° C and analyzed for conjugate binding.
Resui*s and Discussion When conditioned defined media, obtained from early passages of N H 1 5 C A 2 celis, were selected on heparin-Sepharose it was found that the conjugate binding activity had bound completely to the column and eluted between 1 and 2 M of NaC1 (Fig. 1A). N o activity was found in the flowthrough of the heparin-Sepharose colunm. When heparin wag added to the incubation buffer at 20 m g / 1 it was found that the binding activity was blocked to 100%. Chondroitin sulfate, dextran and hyaluronic acid, all at the same concentration, did not affect binding. The effect was concentration dependent and half-maximal at 50 n g / m l (Fig. 1B). Heparin b o u n d rapidly; a 10 min incubation i n the presence of 20 rag/1 heparin, followed by removal of excess heparin, was sufficient to block the conjugate binding activity completely. U p o n treatment with several enzymes, a decrease in binding activity was observed only with bacterial collagenase but not with neuraminidase, chondroitinase or heparinase. Since bacterial collagenase is a highly specific enzyme it was
224
A
%,,
O J LID
FT
,15
.3
.5
"1
2
3
NaCr [M]
CDIVi 14
B 12
\
10
\°
e~ ® ~
\ Jm
0
~
P 0
~
2
3
4
log heparin [ng/mm] Fig. 1. Conjugate binding activity in conditioned defined media of NH15CA2 cells. (A) Heparin-Sepharose profi!e. Shown are the binding activity of the load (LD) and the flowthrough (FT) per volume of a conditioned defined medium (CDM). The binding activity enriched in the 1-2 M NaC1 fraction of the column. (B) Competition curve of conjugate binding activity with heparin. The fraction that eluted between 1 and 2 M NaC1 was dialyzed against PBS and assayed as described for binding in the presence of increasing amounts of hepari~. The binding activity is blocked by heparin and the effect is half-maximal at 50 n g / m l heparin.
225
A
4000I ~ [] ~
3000~
•
• .._...
HA
,[] CS
II
--
0L
T
0
,-
~"
,
1000
Hs
2000
3000
ng/ml
B 10000 r
~
l
i
.
.
.
.
.
.
•
HA
~ooo~\
cs
2000
0
~
.
i
! 1000
,
I 2000
, 3000
ng/ml
Fig. 2. Effectof heparin and sulfatedGAGs on the conjugatebinding activityassociatedwith collagen type V and histones.H, heparin; HS, heparan sulfate; CS, chondroitinsulfate; HA, hyaluronicacid. (A) CollagentypeV; (B) histones.
226 conjectured that the binding activity could be a collagen. When commercially available collagens were tested for a heparin-suppressible binding activity it had been :found indeed that some preparations (obtained from Sigma) of interstitia] collagen type V mimicked the endogenous activity. Binding was sensitive to bacterial collagenase and suppressible by heparin, but not by chondroitin sulfate or hyaluronic acid (Fig. 2a). Furthermore, binding was not affected by a commercially available heparan sulfate (Fig. 2a). The fibrillar collagens type I, II, Iii were different in that they did not bind the HA-conjugate (not shown). A suited control experiment was provided by the histones which had been accidentally identified as a conjugate binding activity. When histones were incubated with bacterial collagenase the binding activity remained stable which demonstrated the absence of proteolytic activities other than collagenase in the lots used. The binding activity of histories was suppressed by heparin in a similar dose response curve as it was for collagen type V. However, in contrast to the collagen type V binding activity, the histone binding activity was suppressed by heparan sulfate and chondroitin sulfate (Fig. 2b). These largely empirical data suggest that this method can be used for the determination of very small quantities of sulfated glycosaminoglycans, tt appears that sulfation is necessary because nonsulfaCed glycosaminoglycans, like hyaluronic acid and dextran did not interfere with ligand binding. Since the ligand is not itself a sugar molecule it should also be possible to include enzymatic digests such as chondroitmase and heparinase digests into the incubation mixture. This would simplify a qualitative assessment of the respective glycosaminoglycan. Simplified description of Che method and its application The ligand described in this paper is capable of binding to receptors (collagen type V, histones) that had been immobilised on nitrocellulose. Sulfated glycosaminoglycans suppress ligand binding to these receptors in small concentrations. The method is simple and suited for large sample numbers. Its application consists in the determination of very small quantities of sulfated glycosaminoglycans. Furthermore, the presence of glycolytic enzymes wilt not effect the determination since the ligand is a protein. Acknewledgements The members of the laboratory of H.C. Schaller are thanked for vivid discussion. This work has been supported by the Bundesministerium fibr Forschung and Technologie (BCT 365/1). References
1 Schaller,H.C. and Bodenmf~lter,H. (1985) Bid. Chem. Hoppe Scyler 366, 1003-1007. 2 Heumann, R., Ocalan, M., Kachel, V. and Hamprecht, B. (1979) Proc. Natl. Acad. Sci. USA 76, 4574-4577. 3 Roberge, M. (1985) Doctoral thesis, University of Heidelberg,Heidelberg, F.R.G,