Inhibition of lipoprotein binding to cell surface receptors of fibroblasts following selective modification of arginyl residues in arginine-rich and B apoproteins.

Treatment of human low density lipoproteins (LDL) with 0.1 M 1,2-cyclohexanedione in borate buffer selectively modified half of the arginyl residues of the apolipoproteins and almost totally abolished the binding of the LDL to the high affinity cell surface receptors of human fibroblasts. Except for the modification of the arginyl residues, there were no apparent alterations in the other amino acid residues, lipid composition, size and morphologic appearance, or apoprotein pattern. Removal of cyclohexanedione by incubation of the modified LDL with 0.5 M hydroxylamine for 7 h restored more than 80% of the original activity as determined by competitive binding, internalization, and degradation studies with Y-LDL. Likewise, selective modification with cyclohexanedione of the arginyl residues of certain canine lipoproteins (LDL, HDL,, and HDL,) prevented their interaction with the cell surface receptors. In particular, the binding activity of the cholesterol-induced HDL, which contain the arginine-rich apoprotein as the only detectable protein was abolished by cyclohexanedione. The only detectable alteration in the canine lipoproteins was a modification of the arginyl residues. Several conclusions are suggested by these studies: (a) the recognition site on the lipoprotein which determines specificity for binding to the cell surface receptors resides with the apoprotein; (b) both the B and arginine-rich apoproteins can react with the receptor; (cl a structural sequence or similarly charged (stereospecific) region may be common to both apoproteins; and (d) a limited number of arginyl residues are functionally significant in or near the recognition site on the lipoproteins.

The binding and degradation assays were performed at 37" by the methods of Goldstein and Brown (20) with minor modifications as previously described (4). The binding studies at 4" were as described (5) except that 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid buffer (25 mM; pH 7.41 was used and the plates were agitated during incubation.

Cyclohexanedione
Modificatmn of Lipoproteins-The most effective conditions as determined by the preliminary experiments were as follows.
Lipoprotein protein (2 to 5 mg) in 1 ml of 0.15 M NaCl and 0.01% EDTA was mixed with 2 ml of0.15 M 1,2cyclohexanedione in 0.2 M sodium borate buffer (pH 8.1) and incubated at 35" for 2 h. The sample was then dialyzed for 20 to 40 h against 0.15 M NaCl and 0.01% EDTA at 4". The modified lipoproteins were filtered through Millipore filters and used immediately in the binding and degradation assays or prepared for amino acid analysis. This procedure was based on previously described methods (7, 9).

Regeneration of Lzpoproteins by Removal of Cyclohexanedione -
An aliquot of the cyclohexanedione-modified lipoprotein was mixed with an equal volume of 1 M hydroxylamine and 0.3 M mannitol at pH 7. The solution was incubated at 35" for 7 to 16 h and then dialyzed for 13 to 24 h against 0.15 M NaCl and 0.01% EDTA, pH 7, at 4". The procedure was based on previously described methods (7, 9).

Amino
Acid Analysis-Solutions of lipoproteins for amino acid analysis were lyophilized and lipid was extracted with chloroform:methanol:ether (2:l:l). Hydrolysis of samples of 0.5 mg was performed in 1 ml of 6 N HCl in the presence of 20 ~1 of mercaptoacetic acid at 110" for 24 h in sealed tubes flushed with nitrogen and evacuated.
To establish the stability of the cyclohexanedione derivative of arginine, N7,N"-(1,2-dihydroxycyclohex-1,2-ylen)-~-arginine hydrochloride was prepared (7) and subjected to hydrolysis conditions. In the presence of mercaptoacetic acid, the cyclohexanedione derivative of arginine was converted to a neutral product without regeneration of arginine as described by Patthy and Smith (7). Amino acids were quantitated on an automated analyzer equipped with sample applicator.
The data were arbitrarily normalized assuming 250 amino acid residues/m01 of LDL protein and 290 amino acid residues/m01 of HDL, protein. in the reaction mixture, human lZ51-LDL were incubated for 2 h at 35" with varying concentrations of cyclohexanedione from 0.05 to 0.1 M. The binding and degradation of cyclohexanedione-treated Y-LDL were decreased with increasing concentrations of cyclohexanedione (Fig. 4). At a concentration of 0.1 M cyclohexanedione, only 1% of the modified 'Y-LDL was bound and internalized by the tibroblasts as compared to the untreated 1251-LDL (see inset, Fig. 4). Since 0.05 and 0.075 M cyclohexanedione were less effective, subsequent experiments were performed with 0.1 M cyclohexanedione.
To determine the optimal time for removal of cyclohexanedione from the arginyl residues of LDL, the modified '251-LDL were incubated with hydroxylamine for 2 to 8 h (Fig. 5). Seventy per cent of the binding capacity could be restored by Human lzSI-LDL protein (240 rg) was incubated for 2 h at 35" with the indicated concentrations of 1,2+yclohexanedione in 0.166 M borate buffer (pH 8.1). The lzSI-LDL was incubated with the cells at a lipoprotein protein concentration of 5 pg/ml with a specific activity of 340 cpm/ng. In the inset, the 100% bound represents the untreated (native) Y-LDL bound and internalized (total bound) minus the nonspecific total binding (19.9 ng/mg) which was arrived at by incubating the native lZ51-LDL with 300 pglml of unlabeled LDL. The radioactivity still bound to the cells in the presence of excess unlabeled LDL has been shown to be due to nonspecific binding. The nonspecific total binding was subtracted from all values and the per cent bound calculated from these adjusted values. The tissue culture conditions were as described in Fig. 1.   FIG. 5. Recovery of the ability of the cyclohexanedione-modified Y-LDL (0) to be bound and degraded by fibroblasts after removal of cyclohexanedione by incubation with hydroxylamine for the indicated times. Human Y-LDL (535 gg) was treated with 0.1 M 1,2+yclohexanedione in 0.166 M borate buffer (pH 8.1) for 2 h at 35". After 2 h, half of the modified IV-LDL was added to an equal volume of 1 M hydroxylamine, 0.3 M mannitol (pH 7.01, and aliquots were removed at the indicated times. The regenerated lS51-LDL was added at a protein concentration of 10 pg/ml with a specific activity of 279 cpmlng. The value for binding and degradation of the untreated Y-LDL (@) at a concentration of 10 pg/ml is indicated. Other conditions were as described in Fig. 1. a 7-to 8-h incubation.
In other experiments to be shown later, it was possible to regenerate more than 80% of the original reactivity. Incubation for 16 h did not restore any additional activity to the LDL. Therefore, an incubation time of 7 to 16 h was used to release the cyclohexanedione from the lipoproteins in subsequent experiments.
Reaction of human LDL with cyclohexanedione and the subsequent release of the cyclohexanedione under the conditions described above did not alter the physical or chemical properties of the lipoproteins except for their electrophoretic mobility. On paper electrophoresis (Fig. 61, the cyclohexanedione-treated LDL were relatively more negative and migrated further toward the anode than the untreated or regenerated LDL. Both the untreated and regenerated LDL had typical p mobility. By negative-staining electron microscopy, the untreated LDL, cyclohexanedione-treated LDL, and regenerated LDL (cyclohexanedione treatment followed by cyclohexanedione removal) were of the same size and had the same morphologic appearance (Fig. 7). The chemical composition (Table I) and apoprotein pattern on SDS-polyacrylamide gel electrophoresis were likewise identical for the untreated, treated, and regenerated LDL. These results in combination with the competitive binding and degradation studies shown in Fig. 1 indicated that the reaction was mild, that it did not alter the physical or chemical properties, and that it was almost completely reversible.
To determine the extent of the modification of the apoproteins by the cyclohexanedione treatment, the amino acid compositions of the lipoprotein proteins were compared before and after reaction with cyclohexanedione and after the release of cyclohexanedione with hydroxylamine treatment. Previously, it was reported by Patthy and Smith (7) that the product of the reaction of 1,2-cyclohexanedione with arginine was stable under acid hydrolysis conditions in the presence of mercaptoacetic acid. We have confirmed these results. Since under acidic conditions cyclohexanedione-modified arginine did not undergo conversion to the unmodified amino acid, it was possible to determine by subtractive amino acid analysis the number of arginine residues which reacted with cyclohexanedione. Amino acid analyses of the LDL revealed that the only amino acid modified was arginine (Table II). Untreated LDL contained 9 residues of arginine per mol, assuming 250 amino acid residues per mol, as compared to 5 residues following the reaction with cyclohexanedione (i.e. 4 of 9 FIG. 6. Paper electrophoretograms of untreated (native) human LDL (top pattern), LDL treated with cyclohexanedione (LDL + 1,2-CHD), and LDL from which the cyclohexanedione was removed (LDL -1 &CHD).
Conditions for the modification and regeneration were as described in Fig. 1.  residues were modified). After the release of cyclohexanedione with hydroxylamine, 8 residues of arginine were detected by amino acid analysis (Table II). An aliquot of the same untreated, treated, and regenerated LDL subjected to amino acid analyses was utilized in the competitive binding and degradation studies. The results were essentially identical with those depicted in Fig. 1 and revealed that the moditication of 4 arginine residues virtually abolished the capability of the LDL to compete for binding and degradation.
Eighty per cent of the original activity was recovered after the removal of the cyclohexanedione by hydroxylamine incubation.
There was a direct correlation between the number of arginyl residues modified and the ability of the LDL to competitively inhibit the binding, internalization, and degradation of Y-LDL.
When 0.05 M cyclohexanedione was allowed to react with LDL for 15, 30, 60, and 120 min, the number of modified arginyl residues increased with time (Table III). Competitive binding and degradation studies revealed a pro-0 Amino acid residues calculated per mol assuming 250 residues per mol. Value in parentheses is rounded off to nearest whole number.
Conditions for the modification and regeneration were as described in Fig. 1. gressive decrease in the capability of the modified LDL to compete with the iz51-LDL for both binding and degradation. As shown in Fig. 8, modification of 2 arginine residues resulted in a 50% decrease in the capacity of the modified LDL to inhibit degradation of the Y-LDL as compared to the untreated LDL at a concentration of 20 pglml of LDL protein. The modification of approximately half of the arginyl residues (4 or 5 out of a total of 9 residues) resulted in the abolishment of greater than 85% of the activity of the LDL (Fig. 8).   (Fig. 9). In addition, the modified LDL did not competitively inhibit degradation of the 'Y-LDL.
With the release of cyclohexanedione from the modified LDL, it was possible to restore most of the activity to the LDL (Fig. 9)