Development of an Immunoaffinity Process for Factor IX Purification

Original Papers

Vox Sang 1990;58:21-29

Development of an Immunoaffinity Process for Factor IX Purification

John Tharakan, Dudley Strickland, Wilson Burgess, William N . Drohan, David B. Clark American Red Cross, Jerome H. Holland Laboratory for the Biomedical Sciences, Rockville, Md., USA

Abstract, An immunoaffinity process based on monoclonal antibody (MAb) to factor IX (FIX) has been developed. Initially, vitamin-K-dependent proteins from cryoprecipitate-poor plasma are isolated on DEAE-Sephadex. The eluate is applied to an immunoaffinity column that utilizes a divalent metal-ion-dependent MAb directed against FIX. After washing the column with high salt in the presence of magnesium ion, the FIX is eluted using a citrate- or EDTA-containing buffer. Coagulation assays and Western blots show no detectable amounts of any contaminating proteins. Purity of the FIX product is established using reduced and nonreduced Coomassie-stanined SDS-PAGE and HPLC. The N-terminal20 amino acids of the single peak of the HPLC were shown to be identical to those reported for FIX. The process shows no detectable leakage of monoclonal antibodies (MAb), efficient utilization of MAb, and provides yields greater than 95%. The use of solvenddetergent treatment as a potential viral inactivation method is incorporated in the process. Studies with tritiated Triton X-100 indicate that the detergent can be washed out of the MAb column so that less than lppm (total) Triton X-100 coelutes with the FIX.

Introduction

Factor IX (FIX) complex concentrates for replacement therapy in hemophilia B patients are commonly prepared by processes that take advantage of the common specific adsorbability of the vitamin-k-dependent clotting factors on anion-exchange resins [l]. These products contain a number of components including other clotting factors and, typically, only 1 % of the total protein is FIX. Although the clinical effectiveness of these concentrates for treatment of hemophilia B is well established, there are consistent re- ports that the use of FIX complex is associated with the occurrence of adverse thrombotic reactions in some patients [2], including such complications as thrombophlebitis at the site of infusion [3] superficial-vein thrombosis [4], deep-vein thrombosis, pulmonary embolus [5], acute myo- cardial infarction [6], and DIC [7,8]. Other potential prob- lems include the transmission of viral diseases and possibly a reduced immune response 191. The causes of these complications are not well understood, but they are probably

associated with other materials present in the concentrates rather than with FIX [lo]. Various contaminants, such as activated factors (factors Xa, IXa, VIIa) and contact-phase factors [l, 11-17] have been implicated. Thus, it is likely that purer products would be benefical. One recent example is provided by Michalski et al. [18], who produced a virally inactivated FIX concentrate using heparin-affinity chromatography.

The American Red Cross [19] has developed an appar- ently nonthrombogenic FIX concentrate essentially free of FII, FVII, and FX. However, this product is only about 10% FIX by protein, and involves a lengthy purification process. Immunoaffinity purification methods using polyclonal or monoclonal antibodies (MAb) hold the promise of producing essentially 100 % pure products by simpler, more efficient processes. Several investigators have reported both polyclonal antibodies and MAbs to clotting factors [20-281. Many of the reported MAbs to FIX require harsh elution conditions, which may be a reason for the lower yields and specific activities of FIX reported in these stud-

22 TharakanlStricklan~ndlBurgess/Drohan/Clark

ies. The harsh elution conditions may also lower the re- usability of the MAb column, an important considzration in large-scale processing. Smith [a] has recently rt:ported a MAb that binds FIX in the presence of magnesium. Elution of bound FIX is effected simply by using a chelating agent such as EDTA or citrate. Other MAbs that bind FIX in the presence or absence of divalent cations [24,29] or that are specific for calcium-dependent epitopes on the FIX protein [27l have also been reported.

Here, we describe the use of a divalent-cation-dependent MAb for the affinity purification of FIX. The starting material is FIX complex from an initial capture of'vitamin- K-dependent coagulation factors from cryoprecipitate- poor plasma using DEAE-Sephadex, which is treated with tri-n-butyl phosphate and TXlW for viral inactivation [%I. The MAb is coupled to cyanogen bromide (CNB:r)-activated Sepharose 4B@. The affinity column is loaded in the presence of 40mM magnesium chloride and washed with 1M sodium chloride. Using u)mM citrate, an eluate of extremely pure FIX, as determined by western blots (WB), Laemmli gel electrophoresis and coagulation assays, is obtained with specific activity greater than 200 units FIWmg protein. No measurable amount of HI, FVII, FX, protein C (PC) or leached mouse IgG is present. Also, the zdity of the MAb (defined as the total units of FIX captured per milligram of MAb) increases with decreasing amounts of bound MAb.

Materials and Methods

MAb Production and Purification Male BALBlc mice were immunized with 3 consecutive injections

of purified human FIX as described by Wang et al. "3. lhree days after the final injection, spleen cells (1 X 108) were fused with the mouse myeloma cell line (1 x lo') P3-X63-Ag8.653 [31] by the method of Koh- ler and Milstein [33]. Hybridoma selection was performed Using the hypoxanthine-aminopterin-thymidine technique and limiting dilution.

MAb was purified from ascites by initial precipitation with sat- urated ammonium sulfate followed by dialysis of the centrifuged pre- cipitate against lOmM sodium phosphate, pH 8, and ion-exchange chromatography on a DEAE-Sephacel column. The column was eluted with a 0-30OmM sodium chloride gradient resulting in two peaks, the first containing IgG and the second albumin.

MAb Screening Assay Cell supernatants were assayed by using a solid-phaw enzyme-

linked immunosorbent assay using human FIX as the capturing agent, detected by goat antimouse IgG conjugated to alkaline phcsphatase. All the buffers used in the assay contained lOmM magnesiurr chloride.

Purification of FIX from DUE-Sephadex Eluate Sepharose CNBr 4B (Warmacia) was used as the immobilization

substrate. Freeze-dried resin (2.5g) was suspended and washed in 1mM

HCl for 15 min. MAb solution (31ml,8.86mg/ml protein, >90% IgG by SDS-PAGE) that had been heat treated to inactivate proteases (56°C for 30 min) was then added to the gel and allowed to mix end-over-end for 2h at mom temperature. The gel was then washed with 2OmM sodium phosphate, lOOmM sodium chloride, pH 8 after which any remaining active groups were blocked by washing with u)omM glycine, pH 8 for 2h. Subsequently, the gel was washed in turn with acetate buffer (0.1M sodium acetate, 0.5 M sodium chloride, pH 4) and phosphate buffer ( u ) d sodium phosphate, lOOmM sodium chloride, pH 8) in sequence several times. Total protein in the unbound material as well as all the washes was determined by absorbance at 280nm. Calculations of the amount of protein coupled revealed a MAb coupling efficiency of 93%. This corresponded to 2.9mg protein per milliliter of gel. The product was stored in the cold in phosphate buffer containing 0.2% sodium azide until further use.

The coupled gel was packed in an Amicon GlOXlSO Column. form- ing a bed of volume 7.5cm3. The column was equilibrated with five column volumes of lOmM magnesium chloride, lOOmM sodium chloride, u)mM'Ltis, pH 7.5 buffer, containing 0.2% sodium azide, at a flow rate of 0.4mUmin. The sample to be applied was eluate from a DEAE- Sephadex adsorption of cryo-poor plasma [19]. This sample was equilibrated in 4OmM magnesium chloride in order to overcome the 20mM concentration of citrate ion already present in the sample. The column was loaded at a flow rate of 0.4mUmin. Subsequently, the column was washed with l O m M magnesium chloride, 1 M sodium chloride, 2OmM Tris, pH7.5 buffer, containing0.2% sodium azide, until the absorbance at 28Onm was below 0.09. A this point, the buffer was changed to 2OmM sodium citrate, l l 0 d sodium chloride, pH 6.8. Elution with this buffer yielded one eluate peak. Column effluent fractions were collected during the load, wash, and elution and assayed for protein and FIX clotting activity. The antibody columns were used repeatedly with no decrease in purity of the FIX obtained or in the utility of the MAb.

Protein Assays Absorbance at 280nm was used to measure protein content. Pro-

tein mass was estimated using a FIX extinction coefficient of 13.3 1241 for a 1% solution.

Coagurcrwn Assays FII, FVII, FIX andFX activity was measured by standard one-stage

coagulation assays as described by Biggs [34] using F'IX-deficient plasma, MIdeficient plasma (George King), FII- and FVII-, and FX- and FVIIdeficient plasma (Sigma) for FIX, M I , FII and FX assays, respectively. Samples were prediluted in a bovine albumidheen-20- containing buffer 1351 to approximately 1 unit/ml. A pool of more than 10 donors of fresh frozen plasma was used as a standard. The FIX and FVII activities were calculated from a semi-log plot. A log-log plot was used to calculate FII and FX activity.

PC was assayed using a chromogenic substrate [36] and by a coagulation assay [37].

Polyacrylamide Gel Electrophoresis and Western Bloning SDS-PAGE was performed by the method of Laemmli 1381. A 4%

stacking gel preceded a 9% separating gel in a Hoeffer Mighty Small Vertical Slab (Model SE200). Proteins were diluted to the appropriate concentration in Tris-buffered saline, 0.1% SDS with or without 2- mercaptoethanol.

A portion of each gel was stained with Coomassie Brilliant Blue R-250, and the rest was electroblotted onto nitrocellulose overnight at

Development of an Immunoaffinity Process for Factor IX Purification 23

20V. The membranes were then washed with 0.025M Tris, 0.15M NaCI, pH 7.4 (WB buffer) containing 3% BSA and 0.1% liveen and blocked with WB buffer with 1.0% albumin. After further washing in WB buffer with reduced albumin (0.1%), the membranes were cut into strips to be probed with polyclonal rabbit antisera to human FIX (Accurate, lot 014A, 1:2OO dilution in WB buffer with 0.1% BSA and 0.1% Tween). FII (Behring, lot 010508, 1:200 dilution), FX (Diag- nostica, Stago, lot 114, 1:200 dilution) and PC (Diagnostica Stago, lot 6006E15,1:1000 dilution). A second set of blots was performed against inter-a-trypsin inhibitor (rabbit anti-inter-a-trypsin inhibitor, Accurate lot 014A, 1:200 dilution) and murine IgG (goat anti-mouse-IgG, Sigma lot 73F8848,1:200 dilution). The membranes were incubated with these antibodies for 2h, washed extensively with WB buffer containing 0.1% Tween and 0.1% BSA and then incubated with goat-anti-rabbit conjugated to horseradish peroxidase (Biorad, lot 31816, 1:lOOO dilution). After extensive washing, the strips were subsequently developed with color reagent containing H202. Color development was stopped by immersing the blotted strips in water.

HPLC and Amino Terminal Sequence Analysis MAb purified FIX product was analyzed by HPLC with a Vydac C4

Reverse Phase Column with buffer A as 0.1% trifluoracetic acid and buffer B as 0.08% trifluoroacetic acid /acentonitrile. Detection was at 220nm. Approximately 150pmol of the single HPLC peak was subjected to automated Edman degradation, to establish identity. The sample was applied directly from the HPLC fraction to a polybrene- treated filter. Twenty cycles of Edman degradation were performed using an Applied Biosystems 477A protein sequencer using the NOR- MAL-1 program supplied by the manufacturer. Phenylthiohydantoin amino acids were identified using an Applied Biosystems model 120 analyzer.

SolventlDetergent Treatment The eluate from the DEAE-Sephadex was treated with solvent

(tri-n-butyl phosphate) and detergent (TX100) as follows: TXlOO (Sig- ma) was added to the DEAE eluate at 1% v/v and tri-n-butyl phosphate (Sigma) was added at 0.3% v/v and the mixture was allowed to mix end-over-end for 2 h at room temperature. This was then applied to the immunoaffinity column. Subsequently, the column was washed extensively and then the FIX was eluted. Experiments with tritiated TXlOO were also performed. In this case, the column was loaded and then washed while fractions were collected to be read in a scintillation counter. After washing, the FIX was eluted in fractions. The eluate fractions were pooled after measurement of radioactivity and the radioactivity of the pool was measured to determine the amount of TXlOO in the final pooled FIX preparation.

MAb Utility MAb utility was compared for three different densities of MAb on

Sepharose-4B resin. MAb was coupled to CNBr-activated Sepha- rose-4B as previously described. The concentration of MAb in the coupling mixture was changed twice yielding resins with MAb coupled at densities of 1.6 and 9.6mg MAb/ml resin, respectively. These resins were poured into two columns as described, and used in experiments identical to the FIX purification runs done with the resin coupled at a density of 2.9mg MAb/ml resin. The amount of FIX activity bound was determined by subtracting the unadsorbed activity from the total loaded. A constant ratio of FIX loaded per milligram MAb coupled was maintained for all three loading densities. Four separate purification runs were conducted with all three densities.

Results and Discussion

Monoclonal Antibody The purified antibody exhibited a single band of approx-

imately 150 kDa in SDS-PAGE under nonreducing conditions. ’ SDS-PAGE analysis under reducing conditions revealed that the molecule is comprised of two subunits, with approximate molecular weights of SO and 2.5 kDa, which is characteristic of the heavy and light chain of IgG. The purified antibody was subtyped by the Ouchterlony technique and identified as belonging to an IgG subclass with a k light chain.

Purification of FIX f rom DEAE-Sephadex Eluate The elution of adsorbed protein from the MAb column

yielded a single protein peak as shown in figure 1. The FIX specific activity and percent recovery for several different experiments are shown in table 1. The specific activity was calculated using the activity determined by the coagulation assays and the protein concentration determined by absorbance at 280nm. It is clear that the bulk of the FIX in the load sample is eluted from the column by the addition of a chelating agent. Figure 2 shows SDS-PAGE, both reduced and nonreduced, of the eluates from four different runs on one column. In all cases a single major band is present. Other faint bands are detectable, which have been shown by western blots to be degradation products.

Factor Activity Assays The results of the assays for FII, FVII, FX and PC are

shown in table 2. The data show the average of duplicate clot times measured for buffer and purified FIX product. The clot times for FII, FVII and FX for the buffer are all less than the clot time for the MAb purified product suggesting the absence of any of these coagulation factors in the product. The buffer clot time for PC, an inhibitor of coagulation, is greater than the clot time for the monoclonal purified product, suggesting the absence of any PC in the product. Thus, the data indicate quite clearly that clotting factors 11, VII, X and PC are not present as measured by coagulation assays.

SDS-PAGE and WB Analysis Figure 3 shows an SDS-PAGE and WB against FII, FIX,

FX and PC, both in the starting material, DEAE-Sephadex eluate and in the final product. As the blots clearly indicate, the starting material contains FII, FIX and FX as well as many others. However, none of these are present in the MAb purified product. The FIX antibody reveals several

24 Tharakan/Strickland/Burgess/Drohan/Clark

Fig. 1. Chromatogram from MAb purification of FIX. 0 = Absorbance at 280nm; A = FIX activity.

Table 1. Purification of FIX from DEAE eluate Table 2. Coagulation assay data

Total FIX Recovery % Utility Product- U loaded/ U FIWmg MAb specific U adsorbed/ activity U eluated U FIX/mg

protein

1,050/909/906 99.8 41.8 248 1,798/1,1120/1,069 95.5 51.0 239 1,914/1,147/1,115 97.0 52.2 272 1,264/775/662 85.0 32.1 170

The starting material was eluate from an adsorption of cryo-super- natant plasma on DEAE-Sephadex, which had a specific activity of 1.5 units FIWmg. The activity of FIX was determined by one-!itage coagulation assays. The data are presented from four separate runs of the same column loaded with 2.9 mg MAbfml gel.

Component Buffer clot time FIX product S clot time

S

FII 180.5 190.8 194.4

FVII 9.3 9.5 9.3

FX 52.7 54.7 54.2

PC 39.0 34.5

Comparison of clot times for buffer and MAb-purified FIX product. Clot times shown are averages of duplicates.

bands in the WB suggesting degradation of FIX. 'This degradation is evident in the starting material as well and suggests that the MAb recognizes degraded as we1 as intact FIX. It is likely that, since the degradation is present in the starting material, it is not a direct result of the immunoaffinity process. In addition to FII, FIX and FX and PC, the MAb purified FIX product was also blotted against inter-a- trypsin inhibitor. From the WB shown in figure 4, no inter- a-trypsin inhibitor was detected in the product, although it was present in the starting material.

A major concern when using immunoaffinity purification techniques is the leaching of affinity ligand from the column, causing a contamination of the product with the antibody or antibody fragments. A WB against mouse IgG, as shown in figure 4, suggests that there is no leaching of antibody during this process. The immunoaffinity-purified FIX product was also assayed for murine IgG using an ELISA. The ELISA was sensitive to 2.8ng of IgG/ml. No murine IgG was detected in the undiluted product at this sensitivity. The product samples used in this assay were


Fig. 2. SDS-PAGE of four runs on the MAb column. The lanes are: lane 1: molecular weight standards reduced (200,116, 97, 66 and 43kDa); lane 2: empty; lanes 3-6, nonreduced FIX from the four different runs on the immunoaffinity column; lane 7: empty; lane 8-11: reduced FIX from the same runs, above; lane 12: empty; lane 13: molecular weight standards.

Fig. 3. Characterization of MAb purified FIX. SDS-PAGE and WB against FII, FIX, FX and PC for the starting material and the MAb-purified product. SDS-PAGE is also shown. Lane 1: HMW standards, Coomassie Blue stain; lane 2: DEAE eluate starting material, Coomassie Blue stain; lane 3: MAb FIX product, Coomassie Blue stain; lanes 4-7: starting material, antiserum to FIX, FII, FX and PC, respectively; lanes 8-11: MAb FIX product, antiserum to FIX, FII, FX and PC, respectively.

I

from the first preparatory runs on this column. It is prob- able that more ligand leakage occurs with the initial runs on a column. The absence of any detectable murine IgG suggests that the MAb-coupling method utilized provides a stable attachment of the affinity ligand to the substrate matrix. Also, the ionic strength, pH and compositions of all the buffers utilized in the process are quite mild and appar- ently do not disrupt the binding between the substrate matrix and the affinity ligand.

HPLC and Amino Terminal Sequence Analysis A fraction of approximately 150pmol of the single

HPLC peak of FIX was subjected to automated Edman degradation. The initial yield for the first cycle was approximately 75%. The yields of the 20 cycles are shown in table 3. The failure to detect phenylthiohydantoin amino acids in cycles 7,8,15,17,18 and 20 is consistent with the presence of y-carboxy glutamic acid or cysteine at these positions. The 20 N-terminal amino acids correspond to published

26 TharakanlStricklandfBurgesslDrohanfClark

Fig. 4. Characterization of MAb-purified FIX. Lane 1: reduced DEAE eluate starting material; lanes 2,3: reduced MAb FIX product; lane 4: empty; lanes 5,6: nonreduced MAb FIX product. Lanes 2 and 5 are from the same run and lanes 3 and 6 are from the same run. The antibody against mouse IgG was shown to be reactive in a separate ELISA. a Coomassie-stained gel. b Antiserum to FIX. c Antiserum to inter-a-trypsin inhibitor. d Anti-mouse IgG.

Table 3. Amino-terminal analysis ~

Cycle No. Amino acid Content pmol

obtained expected

1 2 3 4 5 6 7 8 9

10 11 12 13 14 15 16 17 18 19 20

Phe Val Gln

Asn Leu

GlY

- Met -

5 r Asn Ser GlY LYS Leu Y Y Phe Val Gln

Asn Leu Y

Y

Met Y

GlY

Arg

CYS

111 86 58 91 57

104

- 64 58 64 70 65 60

31 -

- -

24 -

Comparison of experimentally obtained and expected N-terminal sequences. The actual N-terminal amino acid sequence was obtained from published y refers to y-carboxy glutamic acid.

sequence data [39], suggesting that the identity of the affinity-purified product is FIX.

SolventlDetergent Treatment of DEAE Eluate Figure 5 compares the purification of FIX from solvent/

detergent-treated and untreated DEAE eluate starting material. The loading, washing and elution profiles for the FIX activity and the protein are similar for both cases. Although most of the TX 100 is washed out of the column in the washing step, a small amount of TXlOO coelutes with FIX. The total radioactivity measured in the pool of the eluate fractions shows that less than 1 ppm (0.03 ng TX100/unit of FIX) of TXlOO was present in the total eluate pool.

MAb Utility and Eficiency of FIX Capture 6y Immobilized MAb The utility of the antibody is defined as the amount of

FIX activity bound per milligram of MAb. This was calculated based on the number of units adsorbed (determined from the difference between the amount loaded and the drop-through) and the amount of MAb present. From table 1, the average utility of the MAb is 44.3 units FIWrng MAb which corresponds to 34% of the MAb being active, assum- ing a theoretical limit of two molecules of FIX for every molecule of MAb.


Fig. 5. Chromatograms for solvenudetergent-treated and untreated FIX affinity adsorption. 0,. = Ac- tivity; A,A = protein absorbance at 280nm. Closed symbols are for the solvent/detergent-treated run, in which case, the DEAE eluate was treated with 1% TXlOO and 0.3% tri-n-butyl phosphate, and open symbols are for the untreated run. Elution for the untreated DEAE starting material commenced at 125ml and for the solvent/detergent-treated material at 105ml. Frac- tion collection stopped at 136 and 160ml for the treated and untreated materials, respectively.

Fig. 6. Comparison of MAb utility for various antibody loading on CNBr-activated Sepharose-4B. All the loading levels utilized the same coupling protocol. Each data point represents an average of four experiments.

In optimizing an immunoaffinity process, it is necessary to determine the efficiency of antigen capture by the immobilized antibody. Diffusional resistances afforded by heavi- ly loaded columns are not uncommon. Intuitively, there- fore, it would seem that the lower the loading of antibody on a column, the higher might be the efficiency of antigen capture. This will be balanced by the theoretical limit of two antigen molecules or less per antibody molecule. Fig- ure 6 compares the utility of the MAb, as defined earlier, for three separate loadings (mg MAb/ml gel) of antibody

and suggests that the utility of the MAb (units FiWmg MAb) increases as the loading of the gel decreases. For a loading of 1.6mg MAb/ml gel, the average utility was 80.7 unitdm1 gel or 61% of MAb active. As the total amount of protein bound to the gel is decreased, the resistance to transport of the FIX decreases, making the antibody more accessible to FIX for binding, as suggested by the data. It is clear that an optimization of the amount of MAb immobilized per milliliter of resin is important in the development of an immunoaffinity process.

28 Tharakan/Strickland/Burgess/Drohan/Clark

Conclusion

There may be advantages to a pure FIX product for use in replacement therapy, including possible decreased fre- quency of thrombotic episodes and decreased risk of adverse effects on the immune system of the recipient of this product. The immunoaffinity process presented here provides several advantages. The FIX preparation is free of contaminating proteins. The process is relatively straight- forward and simple to implement. The use of a metal-ion- dependent antibody and mild buffers suggests that the affinity adsorbent may be re-used several times. Recovery of FIX is very high, and there is no detectable leakage of murine IgG from the affinity column. The process is also amenable to solvenddetergent treatment and thus may be useful in providing a virally safe product.

Acknowledgements

The authors would like to express their gratitude to Dr. Shirley Miekka and Ms. Craigenne Williams for performing the WB and SDS- PAGE experiments. Also, we would like to thank Mr. Evan Behre for assistance in the tritiated TX 100 experiments.

References

1

2

3

4

5

6

7

8

9

10

11

White GC, Lundblad RL, Kingdon HS: Prothrombin complex concentrates: Preparation, properties and clinical uses. Curr Top He- matol 1979;2:203. Kingdon HS, Lundblad R, Veltkamp JJ, Aronson DL: Potentially thrombogenic materials in factor IX concentrates. Thromb Diath Haemorrh 1975;33:617. Loeliger EA, Hensen A, Mattern MJ, Veltkamp JJ, Bruning PF, Hemker HC: Treatment of haemophilia B with purified factor IX (PPSB). Folia Med Neerl1967;10:112. Marchesi SL, Burney R: Prothrombin-complex concentrates and thromboses. N Engl J Med 1974;290:403. Kasper CK: Postoperative thromboses in hemophilia B. N Engl J Med 1973;289:160. Steinberg MH, Dreiling BJ: Vascular lesions in hemophilia B. N Engl J Med 1973;289:592. Edson JR: Prothrombin-complex concentrates and thromboses. N Engl J Med 1974;290:403. Schimpf K, Zimmermann K, Kompf B: DIC and postoperative wound bleeding under factor IX substitution therapy in a case of hemophilia B; successful treatment with heparin. Thromb Res 1976;8:65. Brettler D, Forseberg A, Levine P, Petillo J, Lamon K, Sullivan J: Factor VIII: C purified from plasma via monoclonal antibodies: Human studies. Thromb Haemost 1985;58:307. Magner A, Aronson D: Toxicity of factor IX concentrates in mice. Dev Biol Stand 1979;44: 185. Aronson DL: Inhibition of thrombotic effect of factor IX. N Engl J Med 1974;290:861.

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

White GC, Roberts HR, Kindgon HS, Lundblad RL: Prothrom- bin complex concentrates: Potentially thrombogenic materials and clues to the mechanism of thrombosis in vivo. Blood 1977; 49:159. Elodi S, Varadi K: Activation of clotting factors in prothrombin complex concentrates as demonstrated by clotting assays for factors IXa and Xa. Thromb Res 1978;12:797. Chandra S, Wickerhauser M: Contact factors are responsible for the thrombogenicity of prothrombin complex. Thromb Res 1979;14:189. Hultin MB: Studies of factor IX concentrate therapy in hemophilia. Blood 1983;62:677. Seligsohn U, Kasper CK, Osterud B, Rapaport SI: Activated factor VII: Presence in factor IX concentrates and persistence in the circulation after infusion. Blood 1979;53:828. Giles AR, Nesheim ME, Hoogendoorn H, Tracy PB, Mann KG: The coagulant-active phospholipid content is a major determinant of in vivo thrombogenicity of prothrombin complex (factor IX) concentrates in rabbits. Blood 1982;59:401. Michalski C, Bal F, Burnouf T, Goudemand M: Large scale production rind properties of a solveddetergent treated factor IX concentrate from human plasma. Vox Sang 1988;55:202. Menache D, Behre HE, Orthner CL, Nunez H, Andersen HD, Triantaphyllopoulos DC, Kosow DP: Coagulation factor IX concentrate: Method of preparation and assessment of potential in vivo thrombogenicity in animal models. Blood 1984;64: 1220. Smith K:Immunoaffinity purification of factor IX from commercial concentrates and infusion studies in animals. Blood 1988;72: 1269. Natelson EA, DeLallo LJ, Coltman CA: Isolation of human factor IX by immunoadsorption. Proceedings of the 44th Annu Meet, 1971. Lab Clin Med;78:846. Jenny R, Church W, Odegaard B, Litweller R, Mann K: Puri- fication of six human vitamin K-dependent proteins in a single chromatographic step using immunoaffinity columns. Prep Bio- chem 1986;16:227. Croissant MP, Pol H vd, Lee HH, Allain J-P: Characterization of four monoclonal antibodies to factor VIII coagulant protein and their use in immunopurification of factor VII. Thromb Haemostasis 1986;56:271. Liebman HA, Limentani SA, Furie BC, and Furie B: Immunoaffin- ity purification of factor IX (Christmas factor) by using conforma- tionspecific antibodies directed against the factor IX-metal complex. Proc Natl Acad Sci 1985;82:3879. Bessos H, Micklem LR, McCann M, James K, Brian D, McClel- land LL, Prowse CV: The Characterization of a panel of monoclonal antibodies to human coagulation factor IX. Thromb Res 1985;40:863. Smith KJ, Ono K: Monoclonal antibodies to factor IX: Character- ization and use in immunoassays for factor IX. Thromb Res 1984;33:211. Thompson AR: Monoclonal antibody to an epitope on the heavy chain of factor IX missing in three hemophilia-B patients. Blood 1983;62:1027. Bessos H, Prowse CV: Immunopurification of human coagulation factor IX using monoclonal antibodies. Thromb Haemostasis 1986;56:86.

29 Reisner HM, de Serres M, Monroe D , Roberts H: A monoclonal antibody to human FIX whose binding is inhibited by divalent cations. Blood 1985;66 (suppl 1).

30 Horowitz B, Wiebe ME, Lippin A, Stryker MH, Richardson L,


Van den Ende MC: Inactivation of viruses in labile blood derivatives. I. Disruption of lipid-enveloped viruses by tri(n)butyl phosphate detergent combinations. Transfusion 1984;25:516.

31 Wang J, Steiner J, Battey F, Strickland D: An immunoaffinity method for the purification of human factor IX. Fed Proc 1987;46:2119.

32 Kearny JF, Radbruch A, Liesang B, Rajewsky K: A new mouse myeloma cell line that has lost immunoglobulin expression but permits the construction of antibody-secreting hybrid cell lines. J Immunol 1979;123:1548.

33 Kohler G, Milstein C: Continuous culture of fused cells secreting antibody of predefined specificity. Nature 1975;256:495.

34 Biggs R: Human Blood Coagulation Haemostasis and Thrombosis, ed I. Oxford, Blackwell Scientific, 1972, p 614.

35 Miekka, SI: Use of albumin and Tween as stabilizers to prevent activity loss during clotting assays of coagulation factor IX and X concentrates. Thromb Haemostasis 1987;58;349.

36 Odegaard OR, Try K, Anderson TR: Protein C: An automated activity assay. Haemostasis 1987;17: 109.

37 Francis RB, Seyfert U: Rapid amidolytic assay of Protein in whole plasma using an activator from the venom of Agkistrodon contorfix. Am J Clin Pathol 1987;87:619.

38 Laemmli U: Cleavage of structural proteins during the assembly of the head of Bacteriophage T4. Nature 1970;227:680.

39 Thompson AR: Structure, function and molecular defects of factor IX. Blood 1986;67:565.

Received: April 4, 1989 Revised manuscript received: June 19,1989 Accepted: June 22,1989

John Tharakan, PhD American Red Cross Plasma Derivatives Laboratory 15601 Crabbs Branch Way Rockville, MD 20855 (USA)

Development of an Immunoaffinity Process for Factor IX Purification

Documents