A Comparison of Molecular Properties of Hepatic Triglyceride Lipase and Lipoprotein Lipase from Human Post-heparin Plasma*

Hepatic triglyceride lipase was isolated from human post- heparin plasma by the method of Ehnholm et al. using modifications which increased the specific activity l&fold to approximately 3,000 pmol of free fatty acidlhlmg of protein. Lipoprotein lipase with similar specifi’c activity was prepared from the same plasma samples using heparin and concanavalin il affinity chromatography. The molecular weight of hepatic triglyceride lipase (69,000) was slightly greater than that of lipoprotein lipase (67,000) as determined by polyacrylamide electrophoresis in sodium dodecyl sul-fate-containing buffers. These proteins had identical amino acid compositions, terminal amino acid residues, and tryptic peptide maps. However, the differences previously described regarding optima of pH and ionic strength and the require-ment for apolipoprotein CII (only for lipoprotein lipase) were in the highly purified

: the precipitation of one of the enzymes with specific antibodies (6, 71, or separation based on the differences in their affinity for heparin (8,9).

Purification of Lipoprotein Li pase
Step I-The LPL obtai ned as described above i n Step I was used for further purification.
Step II-A concanaval i n

Formatron of Phenylthiohydantoins
Samples of protein (5 to 10 nmol) were treated with phenylisothiocyanate (Beckman), and the phenylthiohydantoin derivatives were formed following the procedure described in the Beckman sequenator manual.
They were then chromatographed on polyamide thin layer plates (23). Samples and standards were inspected under shortwave ultraviolet illumination.

Carboxypeptidase A Digestion
Approximately 10 nmol of protein together with norleucine as an internal standard were treated with carboxypeptidase A (Sigma) using a protein-to-enzyme ratio (by weight) of 15O:l at 2'7" in 0.2 M NaHCO,,.
Aliquots were taken at different time intervals, frozen immediately in liquid nitrogen, and lyophilized. After precipitation of the protein with 10% trichloroacetic acid, the samples were filtered through a capillary pipette filled with glass wool and subjected to amino acid analysis as described above.  (Fig. 3). Both cu-methylglucoside and ol-methylmannoside were partially effective in displacing the enzyme from the column; however, the recovery was best with Lu-methyl-n-glucopyranoside.

Hydrazinolysis
Of the total lipase activity applied, 68% was eluted in 250 ml of the a-methyl-o-glucopyranoside buffer and the specific activity of the pooled eluate was almost 8,000 times that of the post-heparin plasma ( Step I is identical with Step I in Table I except the fraction analyzed was that eluting with 1.5 M NaCl in barbital buffer as described in the text and in Fig. 1 8). In addition, a large amount of material remained near the origin and was poorly resolved. When the two enzymes were digested and the peptides analyzed in parallel, the patterns of staining were indistinguishable (Fig. 8) presence of less than 0.2% (w/w) of these sugars. Thus, there is no evidence for the presence of heparin in either enzyme preparation.
Immunoinactivation -The triglyceride hydrolase activity of H-TGL and LPL were progressively inhibited by increasing volumes of antisera raised against the appropriate enzyme. Normal goat serum and antisera prepared against the alternate enzyme did not inhibit H-TGL or LPL (Fig. 9).

DISCUSSION
The H-TGL has been previously purified to near homogeneity by essentially the same sequence of procedures as reported here. However, important modifications have improved the yields and allowed a 12-fold increase in specific activity.
Increasing the ionic strength of the plasma before loading onto the heparin-Sepharose column was the most important of these changes. This procedure removed the need for a delipidation step and thereby avoided a very large loss of activity for both H-TGL and LPL. The introduction of isoelectric focusing as a preparative procedure added an additional 3-fold purification to the present result with H-TGL. This technique also demonstrated heterogeneity within the H-TGL activity, since at least two and perhaps three isoelectric points were found for enzyme preparations from several different plasma pools. Recently, two forms of H-TGL have been separated by NaCl gradient elution of heparin-Sepharose (40). The nature of this heterogeneity is yet to be defined.
Chromatography on concanavalin A-Sepharose was most successful with both enzymes when the column was eluted with o-methyl-n-glucopyranoside. This is identical with earlier findings with the human hepatic lipase (l), but differs from the chromatography of swine adipose tissue lipoprotein lipase, with which better recovery was obtained using CYmethyl-n-mannoside (37). With LPL, a second heparin-sepharose column was required in part to remove protein contaminants which were introduced during the concanavalin A-Sepharose chromatography step and which may represent degradation products of the concanavalin A itself. These products have been eliminated in the purification of chicken adipose tissue LPL by immuno adsorption with antibodies prepared against concanavalin A (14). Similar material was separated from H-TGL by the isoelectric focusing procedure.
The specific activity of the LPL preparation in its most highly purified form was approximately 3000 wmol/h/mg of protein. This is comparable to values obtained with LPL from swine adipose tissue (37) and an LPL from rat plasma (41). The enzyme obtained from chicken adipose tissue (14) had approximately twice the activity per mg of protein using similar assay techniques. With the purified bovine milk LPL (42), specific activities have been found which are IO-fold greater than those obtained with the human plasma enzyme. This is, in part, explained by the assay of the milk LPL under different conditions, including the use of incubations at 37 versus those at 28" in the present experiments. It is also clear that species differences in kinetic characteristics may exist.
The molecular weight of LPL determined by gel filtration is in agreement with that of the band of protein found on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (67,000). This latter technique has given size estimates of lipoprotein lipase from bovine milk (43), swine adipose tissue (37), chicken adipose tissue (14), rat (44) and swine heart (45), and the hepatic lipase (1) from human plasma that range from 60,000 to 70,000. It also agrees with an earlier report on the human plasma lipoprotein lipase (46). The highly purified bovine milk enzyme has recently been studied by sedimentation equilibrium ultracentrifugation and by gel filtration in the presence of guanidine HCl. By both techniques, the protein was estimated to have molecular weight of approximately 50,000 (42).
It has also been reported recently that two forms of apo-CII activated lipoprotein lipase are released into post-heparin plasma of the rat (47). One of these has a much lower molecular weight (37,500) and a higher affinity for triglyceride-rich lipoproteins than the second species of LPL 04, = 69,000). Both enzymes are inhibited by apolipoprotein CI and apolipoprotein CIII and by high ionic strength or protamine in the assay media. The so-called low affinity lipase is similar in