The role of sialic acid in the determination of distinct properties of the isozymes of rabbit plasminogen.

Abstract Previous studies from our laboratory have resulted in the separation of two major forms of plasminogen from rabbit plasma. These two forms differ in their affinity characteristics for antifibrinolytic amino acids, metabolic survival times in the circulation, charge characteristics of the subforms resolved from each major form, and sialic acid content. In order to investigate the role of sialic acid in determining these distinct properties of the plasminogen forms, we have studied these same properties of the asialo rabbit plasminogen forms. We find that the charge differences between the two forms are essentially abolished upon removal of the sialic acid. The number of subforms resolved from each major form is decreased as an effect of removal of the sialic acid, but the remaining subforms possess greatly increased isoelectric points. Contrarily, the binding of each plasminogen to antifibrinolytic amino acids appears no different from normal. The metabolic survival times of the asialoplasminogen forms are not greatly different from each other but are significantly lower than normal, a phenomena also found for many other circulating proteins. These studies suggest that the charge differences between the two forms of rabbit plasminogen are incidental to their functional differences.

From the Department of Chemistry, Program in Biochemistry and Biophysics, The University of Notre Dame, Notre Dame, Indiana 46556 SUMMARY Previous studies from our laboratory have resulted in the separation of two major forms of plasminogen from rabbit plasma.
These two forms differ in their affinity characteristics for antifibrinolytic amino acids, metabolic survival times in the circulation, charge characteristics of the subforms resolved from each major form, and sialic acid content.
In order to investigate the role of sialic acid in determining these distinct properties of the plasminogen forms, we have studied these same properties of the asialo rabbit plasminogen forms. We find that the charge differences between the two forms are essentially abolished upon removal of the sialic acid. The number of subforms resolved from each major form is decreased as an effect of removal of the sialic acid, but the remaining subforms possess greatly increased isoelectric points.
Contrarily, the binding of each plasminogen to antifibrinolytic amino acids appears no different from normal. The metabolic survival times of the asialoplasminogen forms are not greatly different from each other but are significantly lower than normal, a phenomena also found for many other circulating proteins. These studies suggest that the charge differences between the two forms of rabbit plasminogen are incidental to their functional differences.
We have previously shown that two major forms of rabbit plasminogen can be purified from rabbit plasma by affinity chromatography (1). Each major form can be resolved into five subforms which possess distinct isoelectric points with some staggered overlapping (2). Although the molecular weights (2), amino acid compositions (2), and NHz-terminal amino acid sequences (3) of the two forms are similar, other properties such as the sialic acid content (3), binding to antifibrinolytic amino acids, and circulatory survival times (4) show interesting differences.
Many of the distinct properties * This work was supported by Grants HL-13423 and HL-15747 from the National Heart and Lung Institute, National Institutes of Health and a cooperative grant-in-aid from the Indiana and American Heart Associations.
$ Recipient of Research Career Development Award HL-70717 from the National Heart and Lung Institute, National Institutes of Health.
To whom correspondence should be addressed.
of the two major plasminogen forms could be related in principle to the sialic acid differences. For example, the differential charge characteristics of the two forms could be a direct result of the sialic acid differences especially since the form with the greater content of sialic acid possesses an acid shift in the isoelectric points of the subforms.
With regard to the subforms, it is possible that some of these may also result from sialic acid differences.
It is also interesting that the form with the greater content of sialic acid also survives in the circulation longer than the other form. Ashwell and colleagues (5, 6) have proposed a role for sialic acid in protection of the protein from uptake and consequent degradation by the liver.
Due to the potentially interesting relationships of the sialic acid differences to the observed differences in properties of the two major rabbit plasminogen forms we decided to investigate whether this relationship did in fact exist. Our approach to the problem was to determine whether the differential properties of each form of rabbit plasminogen were conserved upon removal of the sialic acid. This manuscript is a result of these studies.

EXPERIMENTAL PROCEDURES
Proteins-Each major form of rabbit plasminogen as well as the subforms studied were purified as described in an earlier report from this laboratory (2). Neuraminidase (V'ibrio cholerae) was purchased from General Biochemicals in vials containing 500 units per ml. These were stored as directed and used without further purification.
Lactoperoxidase was purchased from Calbiochem and coupled to Sepharose 4B as described by David (7).
Asialo rabbit plasminogen Forms 1 and 2 were prepared by dissolving the required plasminogen to 5 mg per ml in a buffer consisting of 0. gens-These were obtained by preparing mixtures of the two forms. Approximately 0.6 mg of normal rabbit plasminogen Form 1 and 1.2 mg of normal rabbit plasminogen Form 2 were dissolved in 2 ml of a solution consisting of 0.1 M phosphate, 0.15 M NaCl, pH 8.0. This mixture was passed over a column (1 X 11 cm) of Sepharose 4B-L-lysine, prepared as previously described (2). The column was then washed with 25 ml of 0.3 M phosphate, pH 8.0, and the plasminogen forms were eluted with a linear gradient of e-aminocaproic acid, exactly as described previously (4). The same procedure was used to obtain affinity chromatography profiles of the asialo rabbit plasminogen forms.
Iodination of Plasminogen-This was accomplished with lz61 by the solid state lactoperoxidase procedure (7). Our lactoperoxidase resin was prepared by coupling lactoperoxidase to cyanogen bromide-activated Sepharose 4B at a final concentration of 2.4 mg of lactoperoxidase per ml of resin. The amount actually coupled was not determined.
Iodination was achieved at room temperature in phosphate-buffered saline solution, pH 7.0. The particular plasminogen was dissolved at a concentration of 5 mg per ml in this buffer with addition of a few crystals of L-lysine to enhance solubility.
One milliliter of protein solution was added to 0.05 ml of Sepharose 4Blactoperoxidase.
Following this, 0. Plasminogen-These were determined by injecting 1 mg of 12sI-labeled plasminogen into the exposed jugular vein of the anesthetized rat. Samples of 0.5 ml of blood were withdrawn from the jugular vein over a period of 70 min and added to heparinized tubes. The blood was centrifuged, and 0.25 ml of the plasma was counted according to the techniques which we described previously (4).

Isoelectric Focusing of Each Asialo Rabbit Plasminogen Form-
This was accomplished with a llO-ml LKB column and LKB carrier ampholytes.
Our procedures have been described earlier (2). Isoelectric points were obtained by refocusing the peaks on narrow pH gradients. Electrophoresis-This was performed as described by Weber and Osborn (9). Fig. 1 shows sodium dodecyl sulfate /3-mercaptoethanol gels for each native and iodinated native affinity chromatography form of rabbit plasminogen as well as gels for each asialo and iodinated asialo form of these proteins. Clearly, treatment with neuraminidase as well as the enzyme and reagents used for iodination did not result in any proteolytic cleavages in plasminogen under our conditions. Table I shows the rate of loss of sialic acid from each rabbit plasminogen form upon treatment with neuraminidase as described under "Methods." LOSS of sialic acid was monitored by the thiobarbituric assay (8) on aliquots of the reaction. The content of sialic acid of plasminogen, as determined in this table, is slightly higher by almost 0.3 to 0.4 mole per mole than we reported earlier (3). However, we feel this variation exists from preparation to preparation and the fact that the two plasminogen major forms differ in sialic acid content is not altered. At the conclusion of the neuraminidase reaction, each plasminogen was purified by affinity chromatography, and an aliquot was hydrolyzed by sulfuric FIG. 1. Sodium dodecyl sulfate-mercaptoethanol polyacrylamide gels (7%) of the rabbit plasminogens used in this study. 1, native plasminogen Form 1 (F-l) ; 3, native plasminogen Form 2 (F-2); 8, 1261-labeled native plasminogen F-l ; 4, 1261~labeled native plasminogen F-2; 6, asialoplasminogen F-l ; 6, asialoplasminogen F-2; 7, 1261-labeled asialoplasminogen F-l ; 8, i261-labeled asialoplasminogen F-2. acid to release any residual sialic acid. Assays at this point demonstrated that essentially no sialic acid remained on the proteins. Incubation of each asialo plasminogen with urokinase demonstrated that these plasminogen derivatives could be activated to plasmin. Fig. 2 shows some pH 4.3 gels on mixtures of each native and aisialo rabbit plasminogen major form. It appears clear that the charge differences which exist in the two native forms are greatly reduced and practically abolished upon removal of the sialic acid. This finding is corroborated by the isoelectric focusing profiles shown for each asialo major form of rabbit plasminogen in Fig. 3. In this figure, two to three subforms are resolved from each major form of the desialylated rabbit plasminogen. The isoelectric points of each subform, which were obtained by refocusing each peak on narrow pH gradients, are summarized in Table II. It appears clear that when the two forms are com- 3. Isoelectric focusing profiles of asialo rabbit plasminogen ona pH 3 to 10 gradient at 4". Top, asialo rabbit plasminogen F-l; bottom, asialo rabbit plasminogen F-2. The experimentally determined pH gradient and the absorbance at 280 nm are presented on the graph. 8.97 9.14 9.19 a The number of the peak in' Fig. 3.

RESULTS
* The asialo form of the first afhnity chromatography peak.
c The asialo form of the second affinity chromatography peak.  4. Affinity chromatography of the asialo rabbit plasminogen forms at 23". 0, neuraminidase-treated rabbit plasminogen F-l and F-2; l , native rabbit plasminogen F-l and F-2 shown for comparison.

IO
The linear gradient of l -aminocaproic acid (e-AHX) used to elute the peaks is shown on the graph. pared, the resolved subforms have nearly identical p1 values. For example, subform 2 of Form 1 has a p1 of 8.91 and subform 2 of Form 2 has a p1 of 8.97 at 22". This result is to be contrasted with our earlier published results on the subforms of each major form of native rabbit plasminogen (2) in which greatly different p1 values for the subforms between the two major forms were obtained. Here, at least five subforms were resolved from each major rabbit plasminogen form. The subforms from the first affinity chromatography form possessed a range of p1 values from 6.2 to 7.78, whereas the subforms from second affinity chromatography form possessed a range of p1 values from 6.95 to 8.74. In this case, Lu S bform 2 of Form 1 has a p1 of 6.56 and'subform 2 of Form 2 has a p1 of 7.18 at 22". Therefore, removal of the sialic acid from each form of native rabbit plasminogen results in a decreased number of subforms, much higher p1 values for the remaining subforms, and near identity in the p1 values of the remaining subforms when the two major forms are compared. Fig. 4 shows the affinity chromatography profiles of each major form of asialo ra.bbit plasminogen compared with each native protein. Near identical behavior is obtained on these columns. This suggests that although the charge differences in the two major forms of rabbit plasminogen can be greatly reduced by removal of sialic acid, their functional difference, i.e. differential binding to antifibrinolytic ammo acids is not affected by this treatment. Further, sedimentation velocity analysis of each native and asialo form of rabbit plasminogen, shown in Table III, indicated that the gross conformational change induced in native plasminogen by the fibrinolytic inhibitor e-aminocaproic acid (I, 10) is still produced in the asialo rabbit plasminogen.
In addition, examination of the native 4O.W values, also shown in this table, leads to the conclusion that removal of the sialic acid from rabbit plasminogen does not markedly alter its gross conformation. Metabolic short term survival times of each major form of native and asialo rabbit plasminogen in the plasma of the rat are plotted according to first order kinetics in Fig. 5. There is a significant decrease in the short term survival times of the asialo forms compared to the native forms. It seems clear that plasminogen follows the general trend of plasma proteins in having greatly decreased circulatory survival times as a consequence of removal of the sialic acid.
Since it appeared possible that some of the subforms of each major rabbit plasminogen form might arise solely from differences in sialic acid, we determined the sialic acid content of each of the subforms.
In this procedure, we hydrolyzed with 0.1 N H&O4 to remove sialic acid. The results of this experiment are shown in Table IV. It is clear that a gradation in sialic acid does exist in the subforms with the most acidic subforms having the greater levels of sialic acid. However, the differences in some of the subforms is rather small and of questionable significance. On the other hand, some of the differences are quite large and are probably significant. DISCUSSION This manuscript deals with the role of sialic acid in determining the distinct characteristics of the many forms of plasminogen found in the animal.
First, two major forms of plasminogen can be resolved based on their differential affinity for antifibrinolytic amino acids of the e-aminocaproic acid class (1). We wish to propose at this time that these two major forms be classified as isozymes. Although we know essentially nothing concerning their site or sites of synthesis we do know that these forms possess certain physical differences such as their charge characteristics (2), sialic acid content (3), and general states of glycosylation (3), and certain functional differences such as their binding constants to antifibrinolytic acids (2). Metabolic studies show differences in their rates of synthesis and degradation and also show that these forms of plasminogen are not interconvertible in the animal (4). We do not mean to infer that the above mentioned structural and functional differences are the only differences found in these forms but any others remain to be established.
It is important to note that removal of the sialic acid from each isozyme of rabbit plasminogen appears to greatly reduce and possibly eliminate the charge differences between these proteins.
On the other hand, the only functional difference that we have so far observed between the two isozymes, i.e. their differential binding to e-aminocaproic acid-like compounds is not affected by this treatment. This is reflected by the identity in affinity chromatography elution profiles of the isozymes of native and asialo rabbit plasminogen.
These observations tend to suggest that the charge differences in the rabbit plasminogen isozymes are incidental to their classification as isozymes.
No definitive statements can be made concerning which, if any, of the subforms resolved from each isozyme can be considered to be isozymes. It appears clear that there are small differences in the sialic acid content of the subforms of each isozyme.
The possibility that these small sialic acid differences