Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

6
THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc Vol. 264, No. 9, Issue of March 25, pp. 5128-5133,1989 Printed in U.S.A. (Received for publication, October 11, 1988) Morey A.Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and Douglas M. TollefsenSII From the $Division of Hematology-Oncology, Departments of Medicine and Biological Chemistry, Washington University School of Medicine, St. Louis, Missouri 631 10 and the TMedical Department A, Aker Hospital, 0514 Oslo 5, Norway Heparin and dermatan sulfate increase the rate of inhibition of thrombin by heparin cofactor I1 (HCII) -1000-fold by providing a catalytic template to which both the inhibitor and the proteinase bind. A variant form of HCII that binds heparin but not dermatan sulfate has been described recently in two heterozy- gous individuals (Andersson, T. R., Larsen, M. L., and Abildgaard, U. (1987) Thromb. Res. 47,243-248). We have now purified the variant HCII (designated HC11o.1~) from the plasma of one of these individuals. HC11o.1, or normal HCII (11 nM) was incubated with thrombin (9 nM) for 1 min in the presence of heparin or dermatan sulfate. Fifty percent inhibition of throm- bin occurred at 26 pg/ml dermatan sulfate with normal HCII and >1600 pg/ml dermatan sulfate withHCIIoslo. In contrast, inhibition of thrombin occurred at a simi- lar concentration of heparin (1.0-1.6 pg/ml) with both inhibitors. To identify the mutation in HCIIodo, DNA fragments encoding the N-terminal 220 amino acid residues of HCII were amplified from leukocyte DNA by the Taq DNA polymerase chain reaction and both alleles were cloned. A point mutation (G + A) resulting in substitution of His for Arg-189 was found in one allele. The same mutation was constructed in the cDNA of native HCII by oligonucleotide-directed mutagenesis and expressed in Escherichia coli. The recombinant HCIIH~..~~~ reacted with thrombin in the presence of heparin but notdermatan sulfate, confirming that this mutation is responsible for the functional abnormality in HCIIoslo. The anticoagulant activities of glycosaminoglycans are me- diated in part by two plasma proteins, antithrombin I11 (ATIII)’ and heparin cofactor I1 (HCII). ATIII andHCII are homologous members of the family of serine proteinase inhib- itors called serpins (1). Proteinase inhibition results from formation of a stable complex between the active center serine hydroxyl group of the target proteinase and aspecific peptide bond (the reactive site) of the inhibitor (2-4). Differences in * This work was supported by National Institutes of Health Grant HL-14147 (Specialized Center of Research in Thrombosis) and by a grant from the Monsanto Company. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact. Fellow of the American Heart Association, Missouri Affiliate. 1) To whom correspondence should be addressed Div. of Hematol- ogy-Oncology, Box 8125, 660 S. Euclid, St. Louis, MO 63110. The abbreviations used are: ATIII, antithrombin III; HCII, hep- arin cofactor 11; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; PEG, polyethylene glycol; TS, Tris-buffered sa- line; PPACK, Phe-Pro-Arg-chloromethyl ketone; serpins, serine pro- teinase inhibitors. reactive site sequence may explain the fact that ATIII inhibits thrombin, coagulation factor Xa, and other proteinases of the intrinsic coagulation pathway, while HCII specificallyinhibits thrombin (5, 6). Heparin increases the rate of proteinase inhibition by ATIII and HCII -1000-fold, apparently by providing a catalytic template to which both the inhibitor and the proteinase bind (7). Although dermatan sulfate increases the rate of the thrombin-HCII reaction by a similar mecha- nism, it has no effect on the rate of proteinase inhibition by ATIII (8). Andersson et al. (9) recently identified two asymptomatic blood donors who had -50% of normal HCII activity deter- mined by athrombin inhibition assay in the presence of dermatan sulfate. The HCII abnormality appeared to be in- herited in the families of these individuals as an autosomal trait. Crossed immunoelectrophoresis suggested the presence of approximately equal amounts of normal HCII and a variant form that bound heparin but not dermatan sulfate. We have now isolated the variant HCII from one of the affected indi- viduals and have demonstrated that it has a marked reduction in apparent affinity for dermatan sulfate while the affinity for heparin is essentially unchanged. Furthermore, we have identified a point mutation in the coding sequence of the HCII gene that accounts for the functional abnormality. Based on the birthplace of the propositus, we have designated the mutation HCIIO,,~. EXPERIMENTAL PROCEDURES Materiak-Bovine lung heparin was purchased from Upjohn. Der- matan sulfate from porcine skin was obtained from Sigma and was treated with nitrous acid to remove contaminating heparin prior to use (10). Chromatographic resins and polyethylene glycol 8000 (PEG) were obtained from Sigma. lZ5I-Labeled human thrombinand affinity purified rabbit anti-HCII antibodies were prepared as described pre- viously (11,12). The concentration of thrombin was determined from the absorbance at 280 nm using an extinction coefficient of 1.83 ml. mg” . cm“ and a molecular weight of 36,600 (13). Taq (Thermus aquuticus) DNA polymerase and other reagents used in the polymerase chain reaction were purchased from Perkin Elmer Cetus. Restriction enzymes and DNA modifying enzymes were pur- chased from New England Biolabs, Amersham Corp., or Promega. The Bluescript KS vector and DNA sequencing reagents were ob- tained from Stratagene. Deoxyadenosine [a-?3]thiotriphosphate ([a- ’%]dATP) was obtained from Du Pont-New England Nuclear. Oli- gonucleotides were synthesized in the Protein Chemistry Facility of Washington University or in the laboratory of Dr. J. Evan Sadler (Washington University). The plasmid expression vector pMON- 5840 was provided by Dr. Peter Olins of the Biological Sciences Division of the Monsanto Company (St. Louis, MO). The strains of Escherichia coli used for transformation included library efficiency DH5a competent cells purchased from Bethesda Research Labora- tories and JMlOl from New England Biolabs. Isolation of HCIZoSb from Plusma-HCII was purified from frozen, citrate-anticoagulated plasma as described by Griffith et al. (14), except that a Sephacryl S-200 column was used for the final step. The concentration of HCII was determined from the absorbance at 5128

Transcript of Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

Page 1: Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 264, No. 9, Issue of March 25, pp. 5128-5133,1989 Printed in U.S.A.

(Received for publication, October 11, 1988)

Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and Douglas M. TollefsenSII From the $Division of Hematology-Oncology, Departments of Medicine and Biological Chemistry, Washington University School of Medicine, St. Louis, Missouri 631 10 and the TMedical Department A, Aker Hospital, 0514 Oslo 5, Norway

Heparin and dermatan sulfate increase the rate of inhibition of thrombin by heparin cofactor I1 (HCII) -1000-fold by providing a catalytic template to which both the inhibitor and the proteinase bind. A variant form of HCII that binds heparin but not dermatan sulfate has been described recently in two heterozy- gous individuals (Andersson, T. R., Larsen, M. L., and Abildgaard, U. (1987) Thromb. Res. 47,243-248). We have now purified the variant HCII (designated HC11o.1~) from the plasma of one of these individuals. HC11o.1, or normal HCII (11 nM) was incubated with thrombin (9 nM) for 1 min in the presence of heparin or dermatan sulfate. Fifty percent inhibition of throm- bin occurred at 26 pg/ml dermatan sulfate with normal HCII and >1600 pg/ml dermatan sulfate with HCIIoslo. In contrast, inhibition of thrombin occurred at a simi- lar concentration of heparin (1.0-1.6 pg/ml) with both inhibitors. To identify the mutation in HCIIodo, DNA fragments encoding the N-terminal 220 amino acid residues of HCII were amplified from leukocyte DNA by the Taq DNA polymerase chain reaction and both alleles were cloned. A point mutation (G + A) resulting in substitution of His for Arg-189 was found in one allele. The same mutation was constructed in the cDNA of native HCII by oligonucleotide-directed mutagenesis and expressed in Escherichia coli. The recombinant H C I I H ~ . . ~ ~ ~ reacted with thrombin in the presence of heparin but not dermatan sulfate, confirming that this mutation is responsible for the functional abnormality in HCIIoslo.

The anticoagulant activities of glycosaminoglycans are me- diated in part by two plasma proteins, antithrombin I11 (ATIII)’ and heparin cofactor I1 (HCII). ATIII and HCII are homologous members of the family of serine proteinase inhib- itors called serpins (1). Proteinase inhibition results from formation of a stable complex between the active center serine hydroxyl group of the target proteinase and a specific peptide bond (the reactive site) of the inhibitor (2-4). Differences in

* This work was supported by National Institutes of Health Grant HL-14147 (Specialized Center of Research in Thrombosis) and by a grant from the Monsanto Company. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

Fellow of the American Heart Association, Missouri Affiliate. 1) To whom correspondence should be addressed Div. of Hematol-

ogy-Oncology, Box 8125, 660 S. Euclid, St. Louis, MO 63110. The abbreviations used are: ATIII, antithrombin III; HCII, hep-

arin cofactor 11; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; PEG, polyethylene glycol; TS, Tris-buffered sa- line; PPACK, Phe-Pro-Arg-chloromethyl ketone; serpins, serine pro- teinase inhibitors.

reactive site sequence may explain the fact that ATIII inhibits thrombin, coagulation factor Xa, and other proteinases of the intrinsic coagulation pathway, while HCII specifically inhibits thrombin (5, 6). Heparin increases the rate of proteinase inhibition by ATIII and HCII -1000-fold, apparently by providing a catalytic template to which both the inhibitor and the proteinase bind (7). Although dermatan sulfate increases the rate of the thrombin-HCII reaction by a similar mecha- nism, it has no effect on the rate of proteinase inhibition by ATIII (8).

Andersson et al. (9) recently identified two asymptomatic blood donors who had -50% of normal HCII activity deter- mined by a thrombin inhibition assay in the presence of dermatan sulfate. The HCII abnormality appeared to be in- herited in the families of these individuals as an autosomal trait. Crossed immunoelectrophoresis suggested the presence of approximately equal amounts of normal HCII and a variant form that bound heparin but not dermatan sulfate. We have now isolated the variant HCII from one of the affected indi- viduals and have demonstrated that it has a marked reduction in apparent affinity for dermatan sulfate while the affinity for heparin is essentially unchanged. Furthermore, we have identified a point mutation in the coding sequence of the HCII gene that accounts for the functional abnormality. Based on the birthplace of the propositus, we have designated the mutation HCIIO,,~.

EXPERIMENTAL PROCEDURES

Materiak-Bovine lung heparin was purchased from Upjohn. Der- matan sulfate from porcine skin was obtained from Sigma and was treated with nitrous acid to remove contaminating heparin prior to use (10). Chromatographic resins and polyethylene glycol 8000 (PEG) were obtained from Sigma. lZ5I-Labeled human thrombin and affinity purified rabbit anti-HCII antibodies were prepared as described pre- viously (11,12). The concentration of thrombin was determined from the absorbance at 280 nm using an extinction coefficient of 1.83 ml. mg” . cm“ and a molecular weight of 36,600 (13).

Taq (Thermus aquuticus) DNA polymerase and other reagents used in the polymerase chain reaction were purchased from Perkin Elmer Cetus. Restriction enzymes and DNA modifying enzymes were pur- chased from New England Biolabs, Amersham Corp., or Promega. The Bluescript KS vector and DNA sequencing reagents were ob- tained from Stratagene. Deoxyadenosine [a-?3]thiotriphosphate ( [a - ’%]dATP) was obtained from Du Pont-New England Nuclear. Oli- gonucleotides were synthesized in the Protein Chemistry Facility of Washington University or in the laboratory of Dr. J. Evan Sadler (Washington University). The plasmid expression vector pMON- 5840 was provided by Dr. Peter Olins of the Biological Sciences Division of the Monsanto Company (St. Louis, MO). The strains of Escherichia coli used for transformation included library efficiency DH5a competent cells purchased from Bethesda Research Labora- tories and JMlOl from New England Biolabs.

Isolation of HCIZoSb from Plusma-HCII was purified from frozen, citrate-anticoagulated plasma as described by Griffith et al. (14), except that a Sephacryl S-200 column was used for the final step. The concentration of HCII was determined from the absorbance at

5128

Page 2: Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

Heparin Cofactor IIoSb 5129 280 nm using an extinction coefficient of 1.17 ml.mg" .cm" (4). The molecular weight of HCII was assumed to be 65,600 (4). The material purified from the individual with HCII deficiency contained a mixture of normal HCII and an abnormal HCII that did not inhibit thrombin in the presence of dermatan sulfate (see "Results"). The normal HCII molecules were removed by first allowing them to react with thrombin in the presence of dermatan sulfate as follows: 850 pg (13 nmol) of HCII was incubated with 357 pg (9.8 nmol) of thrombin and 90 mg of dermatan sulfate in 1000 ml of 0.15 M NaCI, 0.02 M Tris-HCI, pH 7.4,0.1% PEG (TS/PEG) for 60 s at 22 "C. At the concentrations of HCII and thrombin used in this incubation, the rate of glycosami- noglycan-independent complex formation is negligible (4). A t the end of the incubation, 1 ml of 3 mM Phe-Pro-Arg-chloromethyl ketone (PPACK, Calbiochem) was added to inhibit the residual thrombin, and the solution was concentrated to -30 ml with an Amicon YM30 membrane at 4 "C. After dialysis to remove the PPACK, 0.5 ml of the sample was chromatographed on a 10 X 300-mm Superose 12 (Pharmacia LKB Biotechnology Inc.) column equilibrated with 0.15 M NaC1, 0.05 M Tris-HC1, pH 7.5, a t a flow rate of 0.2 ml/min. The peak of unreacted abnormal HCII (HCIIh,.,) was identified by means of a thrombin inhibition assay performed in the presence of heparin.

Inhibition of Thrombin by Normal HCII and HCII&b in the Pres- ence of Glycosaminoglycans-Incubations that included 10 nM HCII, 9.1 nM thrombin, and 0-1.6 mg/ml glycosaminoglycan in 100 pl of TS/PEG were performed in disposable polystyrene cuvettes. The thrombin was added last to initiate the reaction. After 60 s, 500 pl of 78 p~ tosyl-Gly-Pro-Arg-p-nitroanilide (Chromozym TH, Boehringer Mannheim) in TS/PEG was added, and the absorbance at 405 nm was determined continuously for 100 s. The rate of change of absorb- ance was linear with thrombin concentrations from 0 to >13 nM when thrombin alone was incubated with the substrate. No inhibition occurred in control experiments in which thrombin was incubated with HCII in the absence of a glycosaminoglycan, nor did inhibition occur when thrombin was incubated with either heparin or dermatan sulfate alone over the range of concentrations shown in Fig. 3. Effective quenching of the thrombin-HCII reaction after addition of the chromogenic substrate was demonstrated by experiments in which the thrombin was added last to solutions containing HCII, glycosa- minoglycan, and substrate.

Isolation of DNA-Whole blood was collected in the presence of EDTA and stored frozen at -20 "C. After thawing at 37 "C, leukocyte DNA was extracted from the blood by the method of Gustafson et al. (15). The yield of DNA as determined by absorbance at 260 nm was -120 pg/15 ml blood.

Polymerase C h i n Reaction-A portion of the HCII gene was amplified using the polymerase chain reaction technique (16). The two synthetic oligonucleotide primers were (a) a 30-mer identical to nucleotides 86-115 of the coding strand of the HCII cDNA (numbered according to Fig. 2, Ref. 17) and (b) a 21-mer complementary to nucleotides 725-745 of the coding strand. Approximately 2.5 pg of genomic DNA was mixed with 1.0 p~ of each oligonucleotide primer, 200 p~ of each deoxynucleotide triphosphate and 2.5 units of Tag DNA polymerase in 100 p1 of 0.01% gelatin, 50 mM KCI, 1.5 mM MgC12,lO mM Tris-HC1, pH 8.3. Each cycle consisted of the following: (a) denaturation of the DNA for 2.5 min at 94 'C; (b) annealing between the single-stranded DNA and the oligonucleotides for 3 min at 53 "C; and (c) DNA chain polymerization for 3 min at 72 "C. The cycle was repeated 30 times using an automated thermal cycler (Perkin Elmer Cetus).

Cloning and Sequencing of the Amplified DNA-The reaction mix- ture containing the amplified fragment of DNA was subjected to electrophoresis on a 1.4% agarose gel and the 660-base pair band was electroeluted from the gel. After extraction with phenol and chloro- form followed by ethanol precipitation, the 5' termini of the fragment were phosphorylated using T4 polynucleotide kinase and ATP (18). The fragment was ligated into the Bluescript KS vector that had been linearized with Smd and treated with calf intestinal alkaline phosphatase (19). DH5a E. coli cells were transformed as described by the manufacturer. Colonies were selected by ampicillin resistance, and the recombinant plasmids were isolated. Nine plasmids contain- ing the HCII DNA fragment were sequenced by the dideoxy chain termination method of Sanger et al. (20) using [CY-~~SI~ATP.

Expression of pMON-HCII and pMON-HCZIHh.lss-The plasmid pKK-HCII previously described (17) was digested with NcoI and HindIII, and the HCII cDNA fragment was isolated by electrophoresis in a 1% low melting point agarose gel. A similar restriction endonu- clease digestion and electrophoresis of pMON-5840 was performed to remove an irrelevant insert, and the vector and HCII cDNA fragment

were ligated to yield pMON-HCII. E. coli JMlOl cells were trans- formed by the CaC12 method and colonies were selected by ampicillin resistance (21). Induction of HCII synthesis by nalidixic acid in cells transformed with pMON-HCII was carried out as described previ- ously (22). The cell pellets were washed and resuspended in 0.15 M NaCI, 0.02 M sodium phosphate, pH 7.4, and they were sonicated as described (17). Cell lysate supernatants were subjected to SDS-PAGE on 7.5% polyacrylamide slab gels in the presence of p-mercaptoetha- no1 as described by Laemmli (23), and immunoblots were prepared (24). HCII was identified with '251-labeled anti-HCII antibodies. The HCII was quantitated by comparison of the intensity of the immu- noblot of the cell lysate to that of serial dilutions of plasma HCII. To assay for HCII activity, cell lysates were incubated with lZ5I-thrombin in the presence of heparin or dermatan sulfate, and 12sI-thrombin- HCII complexes were detected by SDS-PAGE and autoradiography (17).

Oligonucleotide directed mutagenesis of the HCII cDNA was per- formed in M13mp18 phage by the method of Nakamaye and Eckstein (25). The 21-base oligonucleotide corresponding to nucleotides 641- 661 of the coding strand but containing an A instead of G a t position 651 was used to create the point mutation Arg-189 + His. The single base change was confirmed by dideoxy chain termination sequencing as described above. The replicative form of M13 containing the HCII"m.lss cDNA was prepared from E. coli cell pellets (26). A Bsu36I- XhoI fragment (671 bp) spanning the point mutation was isolated and ligated into the corresponding position of pMON-HCII to yield pMON-HCII~m.l~. Expression of the variant protein was performed as described above.

RESULTS

Isolation of HCIIO~~-HCII was isolated from 600 ml of frozen plasma obtained from the individual designated 11-2 (family 2) in the report by Andersson et al. (9). A standard purification procedure which includes chromatography on heparin-Sepharose, QAE-Sephadex, and Sephacryl S-200 was used (14). The purified HCII was homogeneous by SDS- PAGE and had an apparent molecular weight of 72,000 (Fig. 1). Titration of the purified HCII with increasing amounts of thrombin is shown in Fig. 2A. The equivalence point was determined by extrapolation of the linear portion of the titration curve to the horizontal axis. In the presence of heparin, 1.08 mol of thrombin was inhibited per mol of HCII (closed circles). In contrast, 0.54 mol of thrombin was inhib- ited per mol of HCII in the presence of dermatan sulfate (open circles). None of the thrombin was inhibited in similar incubations performed in the absence of a glycosaminoglycan (closed squares). Under similar conditions, HCII purified from

, 200-

116-

97-

66-

43-

kDa

A B FIG. 1. SDS-PAGE of HCII. Lune A, molecular weight standards

(Bio-Rad). Lane B, HCII purified from individual 11-2 (family 2). The gel was stained with Coomassie Blue.

Page 3: Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

5130 Heparin Cofactor I&

A 0.36

0.30

'g 0.24 c

0.18 . 0 d Q 0.12 a

0.06

0.00

I ..A

0 5 10 15 20 25

0 . 0 0 1 0 :" a 4- o/-.

I

0 5 10 15 20 25

Thrombin Added (nM) FIG. 2. Titration of HCII with thrombin. A, HCII (13 nM)

purified from individual 11-2 (family 2) (see Fig. 1, lane B ) was incubated with 0-25 nM thrombin for 60 s in the presence of heparin (0, 67 pglml), dermatan sulfate (0, 100 pglml), or no glycosamino- glycan (m). The remaining thrombin activity was then determined from the rate of hydrolysis of the chromogenic substrate tosyl-Gly- Pro-Arg-p-nitroanilide (AA,N/min). B, HCII (8.4 nM) purified from normal pooled plasma was titrated with thrombin under the same conditions.

normal pooled plasma inhibited 1.11 mol of thrombin/mol HCII in the presence of either heparin or dermatan sulfate (Fig. 2B). These experiments suggest that the plasma of individual 11-2 contains an approximately equal mixture of two forms of HCII, one of which is unable to interact with dermatan sulfate. The results are consistent with the crossed immunoelectrophoretic analysis of this individual's plasma reported previously (9).

To isolate the variant HCII, the mixture was allowed to react with 0.75 mol equivalent of thrombin in the presence of dermatan sulfate as described under "Experimental Proce- dures." The complex of thrombin with normal HCII was then removed from the unreacted variant HCII (designated HCIIhlo) by gel filtration chromatography on a Superose 12 column. We found no evidence for degradation of the abnor- mal HCII after this procedure. An immunoblot of HCII&lo detected a single band with the same molecular weight as normal plasma HCII by SDS-PAGE. In addition, HCIIhlo and normal HCII formed complexes of the same molecular weight when incubated with '251-thrombin and heparin (data not shown).

The relative glycosaminoglycan-binding affinities of HCIIhlo and HCII isolated from normal pooled plasma were estimated indirectly by incubation of HCII (11 nM) with thrombin (9 nM) for 1 min in the presence of various concen- trations of heparin or dermatan sulfate (Fig. 3). The concen- tration of heparin that produced 50% inhibition of thrombin was -1.0 pg/ml with normal HCII and -1.5 pg/ml with HCIIhlo. In contrast, the concentration of dermatan sulfate that produced 50% inhibition of thrombin was -26 pg/ml with normal HCII compared to >1600 pg/ml with H C I I ~ I ~ .

0.1 2 7

0.10 " C .- E 0.08 -. . v) s 0.06

.. -4 0.04

.. a

0.02 "

0.00 J I

0.01 0.1 1 10 100 1000 10000

[Glycosaminoglycan] (pg/ml) FIG. 3. Effect of glycosaminoglycan concentration on inhi-

bition of thrombin by HC110.1.. Incubations were performed with HCII (11 nM), thrombin (9 nM), and heparin (0, 0) or dermatan sulfate (m, 0) at the final concentrations indicated. After 60 s, the remaining thrombin activity was determined with tosyl-Gly-Pro-Arg- p-nitroanilide (AAa/min). Closed symbols, control HCII purified from normal pooled plasma. Open symbols, HCIIelo isolated as de- scribed in the text.

1 2 3

FIG. 4. Amplification of a 660-base pair (bp) fragment of the HCII gene. Conditions for the Tq DNA polymerase chain reaction are described under "Experimental Procedures." The reac- tion mixtures were subjected to electrophoresis on a 1.4% agarose gel which was stained with ethidium bromide. Each lane contained 10% of the total incubation. Ten picograms of a plasmid containing the HCII cDNA was amplified with primers to the HCII gene ( l o n e I ) . Approximately 2.5 pg of leukocyte DNA from individual 11-2 (family 2) was amplified with X phage primers as a control (lane 2 ) or with primers to the HCII gene (lane 3).

Thus, the affinity of HCIIhlo for dermatan sulfate appears to be reduced at least 60-fold, while the affinity for heparin is only minimally affected.

Amplification, Cloning, and Sequence Analysis of HCIIW DNA-Based on a comparison of the amino acid sequences of HCII and ATIII, we have proposed that the glycosamino- glycan-binding site of HCII resides in the N-terminal half of the protein and is comprised of basic amino acid residues in the vicinity of Arg-103 and/or Lys-185 (17). We and others have found that amino acid residues -19 to 277 of HCII are encoded by a single exon (exon 11) of the HCII gene (27).* To determine the molecular defect in HCIIO.~,, we chose to am- plify and sequence a portion of this exon which corresponds to nucleotides 86-745 of the cDNA and encodes amino acid residues 1-220 of HCII.

The Taq DNA polymerase chain reaction yielded a single band of amplified DNA when leukocyte DNA from the indi- vidual with HCIIalo was used as the template (Fig. 4, lune 3) . This band was similar in size to DNA amplified with the same primers from a plasmid containing the HCII cDNA (lune 1 ). No amplification occurred when oligonucleotide primers for

* J. C. Marasa and D. M. Tollefsen, unpublished observations.

Page 4: Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

Heparin Cofactor IIoSb

X phage DNA were used ( l a n e 2). The amplified DNA was eluted from the agarose gel and ligated into the Bluescript KS plasmid for sequencing. Nine clones containing DNA inserts of the appropriate size were isolated. Three of the clones were found to have a base change (G + A) in the coding strand corresponding to position 651 of the cDNA (Fig. 5). Expres- sion of this allele would result in an amino acid substitution of His for Arg-189 of HCII. The remainder of the clones had the native sequence at this site. The opposite strand was examined through this region in all nine clones to confirm the nucleotide sequences. Single clones with and without this point mutation were sequenced in both directions in their entirety, and no other differences from the cDNA previously sequenced in our laboratory (17) were observed.

We also amplified DNA from a second, apparently unre- lated individual (11-3 of family 1, Ref. 9) with HCII deficiency. The plasma from this individual had an abnormal crossed immunoelectrophoresis pattern in the presence of dermatan sulfate that was identical to the pattern observed in the subject from whom HCIIh1, was isolated (9). Three clones of amplified DNA were sequenced in their entirety on one strand. Two of the clones had the same G + A mutation described above, while the third clone had the native sequence. This finding strongly suggests that the mutation in the codon for Arg-189 represents a naturally occurring allele of the HCII gene rather than an artifact of the amplification or cloning procedures.

Expression of Recombinant HCII and H C I I H ~ . , ~ i n E. coli- To determine whether the Arg-189 + His mutation caused the functional defect in HCIIO.~,, we expressed the native and mutant forms of HCII in E. coli and compared their activities. The expression vector pMON-HCII shown in Fig. 6 was constructed as described in “Experimental Procedures.” This vector contained the entire coding sequence of the native HCII cDNA except that the codons for the signal peptide (19 residues) and the N-terminal18 amino acid residues of plasma HCII were replaced by the sequence encoding Met-Ala-. We previously demonstrated that this slightly truncated form of HCII was active when expressed in E. coli by the vector pKK- HCII (17). E. coli transformed with pMON-HCII expressed -10-20 pg of HCII/ml cell lysate, representing a 40-80-fold increase in the level of expression over that obtained with pKK-HCII. The point mutation causing Arg-189 to be re- placed by His was constructed in pMON-HCII as described under “Experimental Procedures” to yield the vector pMON- HCIIH~..~~. Similar amounts of recombinant HCII and

HCll HCll Oslo

Thr18, ’:I C G

C G Hisle8 A T T A HiSlae

T A A T

C G G C

C G G C

C G G C A T Leulg0

3’ C G 1; Phelgl

Argleg G C -& 4 T A His lag

LeulgO T A

Phelgl T :r A

A C G T A C G T 3’

FIG. 5. Partial nucleotide sequences of two alleles of the HCII gene. Left, the sequence surrounding the codon for Arg-189 of the normal allele, Right, the corresponding sequence of the mutant allele (HC110.1~). The sequencing gel of the non-codingstrand is shown in each case. The arrows indicate that the normal codon for Arg-189 (CGC) has undergone a point mutation to CAC resulting in His-189 in HCII~,,.

pMON-HCII 4.7kb

Amp‘

5131

Ncol Bsu36 I Xhol Hindlll

‘ ~ G C T GCA ‘CGC‘ GG TAC CGA CGT GCG

- 100 bp

Met A l a A l a l g A r g l n 9

FIG. 6. Schematic diagram of the expression vector pMON- HCII. The NcoI-Hind111 fragment of pKK-HCII (17) was cloned into pMON-5840 as described under “Experimental Procedures.” The RecA promoter ( P d ) and GlOL translation control element, derived from the T7 gene, are indicated. The black rectangle in the NcoI- Hind111 insert represents the HCII coding sequence. The open rectan- gle denotes the 3”untranslated sequence from the HCII cDNA and adjacent nucleotides from the pKK-HCII polylinker. The Bsu36I- XhoI segment indicates the region inserted from the M13 replicative form containing the His-189 mutation. The codons and amino acids at the translation initiation site and at Arg-189 are shown. bp, base pair(s).

kDa

86 -I

I 2 3 4 5 6 7 8 9 1 0 ~~- pMON- pMON- pMON- 5840 HCll HCII,,b,m

FIG. 7. Complex formation between ‘a61-thrombin and re- combinant HCII and HCIIAL~~B. Thirty-microliter samples of 10- fold concentrated E. coli cell lysates were incubated with -10,OOO cpm of 12SI-thrombin for 2 min a t 22 ‘C in the presence or absence of a glycosaminoglycan. The samples were reduced, subjected to SDS- PAGE on a 7.5% gel, and autoradiographed. Lane 1 contains ’%I- thrombin alone. The other lanes contain lysates of cells transformed with pMON-5840 (lanes 2 4 ) , pMON-HCII (lanes 5-7), or pMON- HCIIHh.lm (lanes 8-10). Incubations were performed with no glycos- aminoglycan (lanes 2,5 , and 8), with 50 pg/ml heparin (lanes 3 ,6 , and 9 ) , or with 100 pg/ml dermatan sulfate (lanes 4, 7, and 10). The 86,000-dalton band represents the 1Z61-thrombin-HCII complex. The radioactivity at the bottom of the gel represents uncomplexed lZ6I- thrombin.

HCII~is-189 were present in lysates of cells transformed with the two vectors as assessed by either immunoblot intensity or by a thrombin inhibition assay in the presence of heparin (not shown).

The activities of recombinant HCII and HCIIH~-IW were determined by incubation of cell lysates with 1251-thrombin in

Page 5: Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

5132 Heparin Cofactor IIosh

the presence of heparin or dermatan sulfate. Each incubation contained -5 ng of recombinant HCII as estimated from the intensity of immunoblots standardized with plasma HCII. Formation of covalent complexes with '251-thrombin was de- tected by SDS-PAGE and autoradiography. Cell lysates con- taining recombinant native HCII formed an 86-kDa complex in the presence of either heparin or dermatan sulfate as shown in Fig. 7. Recombinant HCIIHi,.lss formed a similar amount of the complex with '251-thrombin in the presence of heparin. In contrast, only a trace amount of the complex was detected in the presence of dermatan sulfate. No complexes were observed in the absence of a glycosaminoglycan or when lysates of E. coli transformed with the parent vector pMON- 5840 were used. These experiments indicate that the single base mutation leading to substitution of His for Arg-189 causes the functional abnormality in HCIIo,~,.

DISCUSSION

The ability of glycosaminoglycans to increase the rate of inhibition of proteinases by HCII and ATIII -1000-fold dis- tinguishes these inhibitors from other members of the serpin family (28). Although the precise mechanism for this catalytic reaction is unknown, binding of the glycosaminoglycan to the inhibitor is required for activity. Binding is thought to occur by ionic interactions between sulfate groups on the glycosa- minoglycan and a domain of the protein rich in basic amino acid residues. This hypothesis is supported by the observation that binding of HCII and ATIII to heparin-Sepharose is disrupted at high ionic strength (3,4). Furthermore, chemical modification of basic residues in ATIII and HCII leads to a marked decrease in heparin binding and catalysis without affecting the uncatalyzed rate of proteinase inhibition (3, 29, 30).

A specific pentasaccharide sequence in heparin, which in- cludes a unique 3-0-sulfated glucosamine residue, is required for high affinity binding to ATIII (31, 32). Low affinity heparin molecules that lack this structure have <5% of the activity of high affinity heparin in a thrombin inhibition assay with ATIII (34). Similarly, dermatan sulfate and other gly- cosaminoglycans that lack this structure do not catalyze pro- teinase inhibition by ATIII. In contrast, the specific penta- saccharide sequence is not required for binding of heparin to HCII or for catalysis of the thrombin-HCII reaction (33, 34). Dermatan sulfate and a variety of other natural and synthetic polyanions (e.g. chondroitin sulfate E (35), pentosan polysul- fate (36), polyaspartic acid (37), and dermatan sulfate (38)) all increase the rate of inhibition of thrombin by HCII to various degrees. These findings indicate that the glycosami- noglycan-binding site of HCII has a much broader ligand specificity than that of ATIII.

Two domains of the ATIII molecule have been implicated in heparin binding (Fig. 8). One domain includes Arg-47, where naturally occurring mutations to Cys, His, and Ser are associated with ATIII molecules that fail to bind heparin (39-

41). A mutation of the nearby Pro-41 to Leu and chemical modification of Trp-49 are also associated with a decrease in affinity for heparin (42, 43). The second domain of ATIII implicated in heparin binding consists of a cluster of basic amino acid residues that includes Lys-114, Lys-125, Arg-129, Arg-132, and Lys-133. Although no mutations have been described for these residues, there is evidence that Lys-114 and Lys-125 are important. These two lysine residues are protected from chemical modification in the presence of hep- arin, implying that they are directly involved in heparin binding (44, 45), and selective chemical modification of Lys- 125 greatly reduces the affinity of ATIII for heparin (44). Furthermore, a proteolytic fragment of ATIII containing amino acid residues 114-156 has been reported to bind hepa- rin (46). The two non-contiguous domains may in fact be brought together by protein folding to form a single binding site for heparin. Alignment of the sequence of ATIII with the x-ray crystallographic structure of al-antitrypsin cleaved at the reactive site predicts a group of positively charged residues that are spaced close to one another on the surface of the molecule (47). These residues include Arg-47 and the cluster of four basic amino acids beginning with Lys-125. Confirma- tion of this structure will require crystallization of native ATIII.

Alignment of the amino acid sequences of HCII and ATIII reveals a strong similarity in the vicinity of the proposed heparin-binding site described above (Fig. 8) (17). Specifi- cally, Arg-103 of HCII corresponds to Arg-47 of ATIII. Begin- ning with Lys-173 of HCII (corresponding to Lys-114 of ATIII), 9 of 21 residues are identical, including Lys-185, Arg- 189, and Arg-192. In addition, Arg-193 of HCII corresponds to a Lys residue in ATIII. Thus, five of the six basic amino acid residues between positions 173 and 193 of HCII align perfectly with basic residues in ATIII. It is interesting to note that these basic residues are poorly conserved in other serpins that do not react with glycosaminoglycans (see Ref. 48 for alignments). For example, human al-antitrypsin has only one basic residue in this region, a*-antiplasmin has two, and al- antichymotrypsin has three, none of which align with basic residues in HCII or ATIII. Plasminogen activator inhibitor type 1 has two basic residues in this region, one of which corresponds to Arg-189 of HCII.

The structure of HCIIo,l, suggests that Arg-189 is directly involved in binding to dermatan sulfate but that this residue is not required for binding to heparin. The mutation Arg-189 + His would result in loss of a positive charge at pH 7.4, which could disrupt an important ionic bond between HCII and a sulfate or carboxylate group on dermatan sulfate. On the other hand, it is clear that the presence of Arg-189 is insufficient for dermatan sulfate binding, since ATIII and plasminogen activator inhibitor 1 contain an Arg residue at this position but do not interact with dermatan sulfate. Per- haps the presence of the positive charge at Arg-184 of HCII (corresponding to Ala-124 of ATIII) is required for binding to

8 8 8 8

4 7 1 1 4 1 2 5 1 2 9 132 1 3 3

FIG. 8. Alignment of the putative . . . . . . . . . . . .

glycosaminoglycan-binding sites of $ Q 8 8 8 8 @

numbering of amino acid residues are 103 1 7 3 1 8 4 185 1 8 9 1 9 2 1 9 3

described in Ref. 17. . . . . . . . . . . . . . . . . . . _ _ . . . . _ . . . . . . _

HCII and ATIII. The alignment and Hell ~ r g / / L y s - T y r - G 1 u - I l e - T h r - T h r - I l e - H i s - A s n - L e u - P h e - A ~ g - L y ~ - L ~ u - T h ~ - H i s - A ~ ~ - L e u - P h e - A ~ g - A ~ g

. . . . . . . .

8 0 8 0 $ 8 HCI loslo Arg / / L y s - T y r - G l u - I l e - T h r - T h r - I l e - H i s - A s n - L e u - P h e - A ~ g - L y ~ - L e u - T h ~ - H i ~ - H i s - L ~ u - P h e - A ~ g - A ~ g 1 0 3 1 7 3 184 185 1 8 9 1 9 2 1 9 3

Page 6: Morey A. Blinder$$, Terje R. Anderssonq, Ulrich Abildgaardll, and ...

Heparin Cofactor IIoSb 5133

dermatan sulfate. Mutagenesis of Arg-184 and other residues in this region of HCII may delineate the requirements for heparin and dermatan sulfate binding more precisely.

The occurrence of HC11o.1, can be explained by a C .--, T mutation in a CpG dinucleotide of the non-coding strand of the HCII gene. Coulondre et al. (49) have postulated that methyl cytosine residues are mutational hotspots. Methyl cytosines may spontaneously deaminate to thymine causing a C + T transition during the next round of replication. Since most of the methyl cytosines in eukaryotic DNA occur in the setting of a CpG dinucleotide, C --., T mutations in this dinucleotide would be expected. In ATIII, six point mutations have been identified which result in abnormal heparin binding (47, 50). All except ATIIIR~~~,,,, (a C + A mutation in the codon for Arg-47) could be explained by a C .--, T mutation in a CpG dinucleotide.

Although two kindreds have been described in whom an -50% decrease in plasma HCII was associated with throm- bosis (51,52), a definitive role for HCII in hemostasis has yet to be firmly established. We previously observed that cultured fibroblasts and smooth muscle cells, but not vascular endo- thelial cells, stimulate the thrombin-HCII reaction (53). Moreover, the effect of fibroblasts is mediated by dermatan sulfate proteoglycans on their surface. These results have led us to propose that the physiological function of HCII may be to inhibit thrombin in extravascular tissues rather than in the bloodstream. The individual from whom HCIIO,~, was isolated is a heterozygote who has -50% of normal plasma HCII activity in the presence of dermatan sulfate (9). The fact that she remains healthy suggests either that binding of HCII to dermatan sulfate is unimportant for its physiological function or, more likely, that half the normal activity of HCII is sufficient to prevent disease. Identification of a homozygous individual with HCII deficiency may eventually suggest the in vivo function of this inhibitor.

Acknowledgments-We wish to thank Drs. Evan Sadler and Stan Korsmeyer for their helpful suggestions and Dr. Peter Olins for providing the expression vector.

REFERENCES 1. Carrell, R. W., and Boswell, D. R. (1986) in Proteinase Inhibitors

(Barrett, A. J., and Salveson, G., eds) pp. 403-420, Elsevier, Amsterdam

2. Laskowski, M., and Kato, I. (1980) Annu. Rev. Biochem. 49,593- 626

3. Rosenberg, R. D., and Damus, P. S. (1973) J. Biol. Chem. 248, 6490-6505

4. Tollefsen, D. M., Majerus, D. W., and Blank, M. K. (1982) J. Biol. Chem. 257,2162-2169

5. Rosenberg, R. D. (1977) Semin. Hematol. 14 , 427-440 6. Parker, K. A., and Tollefsen, D. M. (1985) J. Biol. Chem. 2 6 0 ,

7. Griffith, M. J. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 5460-

8. Tollefsen, D. M., Pestka, C. A., and Monafo, W. J. (1983) J. Biol.

9. Andersson, T. R., Larsen, M. L., and Abildgaard, U. (1987)

10. Teien, A. N., Abildgaard, U., and Hook, M. (1976) Thromb. Res.

11. Tollefsen, D. M., and Blank, M. K. (1981) J. Clin. Invest. 68,

12. Jaffe, E. A., Armellino, D., and Tollefsen, D. M. (1985) Biochem. Biophys. Res. Commun. 132,368-374

13. Fenton, J. W., 11, Fasco, M. J., Stackrow, A. B., Aronson, D. L., Young, A. M., and Finlayson, J. S. (1977) J. Biol. Chem. 252, 3587-3598

14. Griffith, M. J., Noyes, C. M., and Church, F. C. (1985) J. Biol. Chem. 260,2218-2225

15. Gustafson, S., Proper, J. A., Bowie, E. J. W., and Sommer, S. S.

3501-3505

5464

Chem. 268,6713-6716

Thromb. Res. 47,243-248

8,859-867

589-596

16.

17.

18.

19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

(1987) Anal. Biochem. 165,294-299 Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi,

R.. Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988) Science 239,487-491

Blinder. M. A.. Marasa. J. C.. Revnolds. C. H.. Deaven, L. L., andTollefsen, D. M. (1988) bio&mistry 27,752-759

Maniatis, T., Fritsch, E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, .~

Cold Spring HarboriNY Berger. S. L.. and Kimmel, A. R. (1987) Guide to Molecular

C&ng Techniques, Academic Press, Orlando, FL

Acad. Sci. U. S. A. 74,5463-5467 Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl.

Cohen, S., Chang, A. C. Y., and Hsu, L. (1972) Proc. Natl. Acad.

Li, E., Locke, B., Yang, N. C., Ong, D. E., and Gordon, J. I. (1987)

Laemmli, U. K. (1970) Nature 227,680-685 Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad.

Nakamaye, K. L., and Eckstein, F. (1986) Nucleic Acids Res. 14,

Davis, L. G., Dibner, M. D., and Battey, J. F. (1986) Basic Methods in Molecular Biology, pp. 102-105, Elsevier, New York

Ragg, H., and Preibisch, G. (1988) J. Biol. Chem. 263 , 12129- 12134

Bjork, I., and Danielsson, A. (1986) in Proteinase Inhibitors (Barrett, A. J., and Salveson, G., eds) pp. 489-513, Elsevier,

Church, F. C., and Griffith, M. J. (1984) Biochem. Biophys. Res. Amsterdam

Commun. 124, 745-751 Church, F. C., Villanueva, G. B., and Griffith, M. J. (1986) Arch.

Biochem. Biophys. 246 , 175-184 Lindahl, U., Thunberg, L., Backstrom, G., Riesenfeld, J., Nor-

dling, K., and Bjork, I. (1984) J. Biol. Chem. 259,12368-12376 Atha, D. H., Lormeau, J . C., Petitou, M., Rosenberg, R. D., and

Choay, J. (1985) Biochemistry 24,6723-6729 Hurst, R. E., Poon, M.-C., and Griffith, M. J. (1983) J. Clin.

Invest. 72,1042-1045 Maimone, M. M., and Tollefsen, D. M. (1988) Biochem. Biophys.

Res. Commun. 152,1056-1061 Scully, M. F., Ellis, V., Seno, N., and Kakkar, V. V. (1986)

Biochem. Biophys. Res. Commun. 137,15-22 Scully, M. F., Ellis, V., and Kakkar, V. V. (1986) Thromb. Res.

Church, F. C., Treanor, R. E., Sherrill, G. B., and Whinna, H. C.

Yamagishi, R., Niwa, M., Kondo, S., Sakuragawa, N., and Koide,

Koide, T., Odani, S., Takahashi, K., Ono, T., and Sakuragawa,

Owen, M. C., Borg, J. Y., Soria, C., Soria, J., Caen, J., and Carrell,

Duchange, N., Chasse, J. F., Cohen, G. N., and Zakin, M. M.

Chang, J.-Y., and Tran, T. H. (1986) J. Biol. Chem. 2 6 1 , 1174-

Blackburn, M. N., Smith, R. L., Carson, J., and Sibley, C. C.

Peterson, C. B., Noyes, C. M., Pecon, J. M., Church, F. C., and

Lui, C.-S., and Chang, J.-Y. (1987) J. Biol. Chem. 2 6 2 , 17356-

Smith. J. W.. and Knauer. D. J. (1987) J. Biol. Chem. 262.

Sci. U. S. A. 69,2110-2114

J. Biol. Chem. 262,13773-13779

Sci. U. S. A. 76, 4350-4354

9679-9698

4 1,489-499

(1987) Biochem. Biophys. Res. Commun. 148 , 362-368

T. (1984) Thromb. Res. 36,633-642

N. (1984) Proc. Natl. Acad. Sci. U. S. A. 81, 289-293

R. W. (1987) Blood 6 9 , 1275-1279

(1987) Thromb. Res. 4 5 , 115-121

1176

(1984) J. Biol. Chem. 259,939-941

Blackburn, M. N. (1987) J. Biol. Chem. 262,8061-8065

17361

11964-11972

R. W. (1988) J. Clin. Invest. 81, 1292-1296

. ,

Borg, J. Y., Owen, M. C., Soria, C., Soria, J., Caen, J., and Carrell,

Ye, R. D., Wun, T.-C., and Sadler, J. E. (1987) J. Biol. Chem. 262,3718-3725

Coulondre, C., Miller, J. H., Farabaugh, P. J., and Gilbert, W.

Hultin, M. B., McKay, J., and Abildgaard, U. (1988) Thromb.

Tran, T. H., Marbet, G. A., and Duckert, F. (1985) Lancet 2,

Si6, P., Dupouy, D., Pichon, J., and Boneu, B. (1985) Lancet 2,

McGuire, E. A., and Tollefsen, D. M. (1987) J. Biol. Chem. 262 ,

(1978) Nature 274 , 775-780

Huemost. 59,468-473

413-414

414-416

169-175