Colipase residues Glu64 and Arg65 are essential for normal lipase-mediated fat digestion in the presence of bile salt micelles.

Pancreatic triglyceride lipase (PTL) requires colipase for activity. Various constituents in meals and in bile, particularly bile acids, inhibit PTL. Colipase restores activity to lipase in the presence of inhibitory substances like bile acids. Presumably, colipase functions by anchoring and orienting PTL at the oil-water interface. The x-ray structure of the colipase.PTL complex supports this model. In the x-ray structure, colipase has a hydrophobic surface positioned to bind substrate and a hydrophilic surface, lying opposite the hydrophobic surface, with two putative lipase-binding domains, Glu(45)/Asp(89) and Glu(64)/Arg(65). To determine whether the hydrophilic surface interacts with PTL in solution, we introduced mutations into the putative PTL binding domains of human colipase. Each mutant was expressed, purified, and assessed for activity against various substrates. Most of the mutants showed impaired ability to reactivate PTL, with mutations in the Glu(64)/Arg(65) binding site causing the greatest effect. Analysis indicated that the mutations decreased the affinity of the colipase mutants for PTL and prevented the formation of PTL.colipase complexes. The impaired function of the mutants was most apparent when assayed in micellar bile salt solutions. Most mutants stimulated PTL activity normally in monomeric bile salt solutions. We also tested the mutants for their ability to bind substrate and anchor lipase to tributyrin. Even though the ability of the mutants to anchor PTL to an interface decreased in proportion to their activity, each mutant colipase bound to tributyrin to the same extent as wild type colipase. These results demonstrate that the hydrophilic surface of colipase interacts with PTL in solution to form active colipase.PTL complexes, that bile salt micelles influence that binding, and that the proper interaction of colipase with PTL requires the Glu(64)/Arg(65) binding site.

Lipases participate in a variety of physiologic and industrial processes (1,2). They aid in the preparation of foods and pharmaceuticals. They serve to generate second messengers, participate in cytotoxicity reactions, and have become a model system for the study of molecular interactions occurring at lipid-water interfaces. Most importantly, they allow for the digestion and trafficking of dietary triglycerides and phospholipids by converting apolar dietary lipids to more polar products, fatty acids and monoacylglycerols. An understanding of the digestive process has implications for the treatment of human diseases, including malnutrition and obesity.
The digestion of dietary triglyceride begins in the stomach through the process of emulsification and the actions of gastric lipase (3). Digestion continues in the duodenum where pancreatic triglyceride lipase (PTL) 1 releases 50 -70% of dietary fatty acids (3). Although PTL requires bile acids to function efficiently, bile acid micelles inhibit PTL unless another pancreatic protein, colipase, is present (2). Colipase restores PTL activity in the presence of physiologic concentrations of bile acids. Without colipase, PTL could not cleave fatty acids from dietary triglycerides, and fat malabsorption with its consequences would result.
Previous models of colipase function uniformly proposed that colipase acts as a bridge between PTL and the lipid substrate by binding to both PTL and the lipid emulsion surface. The strongest support for this model comes from the crystal structures of the porcine colipase⅐human PTL complex solved under different conditions (4,5). In the first structure, PTL resides in an inactive conformation with the catalytic site sterically blocked by a surface loop, called the lid. The other structure, solved in the presence of mixed micelles, reveals the active conformation of PTL in which the lid moves away from the catalytic site and forms new interactions with colipase. Colipase has similar conformations in both structures (6). Importantly, colipase has two opposing surfaces that can potentially anchor PTL to the substrate. One surface, which contains predominantly hydrophilic residues, forms interactions with the C-terminal domain of PTL. The opposite surface, which faces away from PTL, has four hydrophobic loops positioned to interact with the substrate surface. In this conformation, colipase brings PTL into the proper proximity and orientation for lipid hydrolysis.
The crystal structure predicts that van der Waals forces and several polar bonds stabilize the interaction between colipase and PTL. The number of putative polar bonds varies between the inactive and active conformation of PTL. In the active conformation of PTL two colipase residues, Glu 15 and Arg 38 , form polar bonds with residues in the PTL lid. In both conformations of PTL polar interactions occur between five residues of colipase, Arg 44 , Glu 45 , Glu 64 , Arg 65 , and Asn 89 , and mainchain residues in the C-terminal domain of PTL. Four residues form two potential binding domains, one containing residues 45 and 89, which interact with Lys 400 of PTL, and the other containing residues 64 and 65, which interact with Gln 369 of PTL. Arg 44 is not conserved and is replaced by a serine or methionine in most other species. Additionally, Arg 44 does not form a polar bond in the porcine colipase⅐PTL complex. These points argue that Arg 44 may not play an important role in colipase⅐PTL interactions in solution.
A few studies have examined the role of specific amino acids in the formation of the colipase⅐PTL complex. One study examined the interaction between Glu 15 of colipase and the PTL lid by site-specific mutagenesis of Glu 15 (7). This study concluded that the interactions between Glu 15 and the PTL lid stabilized the active confirmation of PTL. Two other studies investigated the interaction between colipase and the C-terminal domain of PTL. Jennens and Lowe (8) introduced substitution and deletion mutations into the C-terminal domain of PTL. Single substitution mutations of multiple PTL residues thought to interact with colipase did not affect lipase activity. Deletion of the C-terminal domain greatly decreased PTL activity, but the truncated PTL still required colipase for activity in the presence of bile acids. These mutants raised questions about the importance of the interactions between colipase and the Cterminal domain of PTL. Another recent study provided support for the predicted interaction between one colipase residue, Glu 45 , and Lys 400 of the PTL C-terminal domain (9). Ayvazian et al. (9) demonstrated that disruption of this interaction impaired lipase activity secondary to an inability of the mutant colipase to bind PTL. None of these studies addressed the relative importance of the other putative lipase binding residues in the hydrophilic surface of colipase.
To determine if the hydrophilic surface of colipase interacts with lipase in solution and to identify participating residues, we introduced mutations into positions 44, 45, 64, 65, and 89 of human colipase by site-specific mutagenesis. Each mutant colipase was characterized for the ability to reactivate PTL in the presence of bile salts and to anchor PTL at an oil-water interface. The results support the orientation of colipase as defined by the crystal structures. That is, the hydrophilic surface of colipase interacts with PTL in solution. Furthermore, the results identify the 64/65 binding domain, specifically position 64, as key to the normal function of colipase, thereby adding additional details to the mechanism of PTL-mediated lipolysis.

MATERIALS AND METHODS
Construction of Mutants-Manipulations of DNA were done by standard methods (10). Mutations were introduced into wild type human colipase cDNA in the pPIC9 vector by polymerase chain reaction using the Stratagene QuikChange site-directed mutagenesis kit per the manufacturer's protocol. Oligonucleotide primers were designed to introduce the desired mutations. Transformations into Epicurian Coli XL-1 Blue Supercompetent cells were performed per the QuikChange protocol. Minipreps were performed using either the Qiagen Plasmid Mini kit or the Qiagen QIAprep Miniprep kit. Presence of the desired mutation was confirmed by dideoxynucleotide sequence analysis using the ABI PRISM Big Dye Terminator Cycle Sequencing Ready Reaction kit.
Protein Methods-Proteins were expressed and purified as previously described (11). Protein concentrations were determined by ultraviolet spectrophotometry at A 280 using an extinction coefficient of 0.47. Equal quantities (3 g) of each mutant protein were run on SDS-PAGE gels and stained with GELCODE Blue stain (Pierce, Rockford, IL) to confirm the homogeneity of the protein product.
Stability in urea and TDC was determined for colipase and the colipase mutants. Each colipase was incubated for 30 min at room temperature in 8 M urea or 4 mM TDC in 50 mM Tris-HCl, pH 8.0, at a concentration of 0.1 mg/ml. Companion incubations were done in buffer alone. After the incubation, 500 ng of the colipase was assayed with 2.8 g of PTL using tributyrin emulsified with 0.5 mM TDC as described below. The difference between PTL activity in the absence and presence of colipase was defined as the colipase-stimulated PTL activity. Neither colipase nor the colipase mutants lost any function after incubation in buffer for 30 min at room temperature.
Colipase Assays-Assays were performed at 37°C using a VIT 90 Radiometer pH-stat, in assay buffer containing 1 mM Tris-HCl, pH 8, 2 mM CaCl 2 , 150 mM NaCl, and taurodeoxycholate (TDC) concentrations ranging from 0 to 4 mM as described earlier (12). The rate of NaOH titration during the assay was set to maintain a constant pH of 8. Tributyrin (Fluka, Buchs Switzerland and Sigma Chemical Co., St. Louis MO) was used as substrate. Assays were performed with 14.5 ml of assay buffer and 0.5 ml of tributyrin except as noted below. Further analysis of the mutants was performed by measuring lipase activity over a range of colipase concentrations (0 -1372 nM) with a constant concentration of lipase (1400 ng) and excess substrate. To determine the substrate dependence of PTL activity with the various colipase mutants, we measured PTL activity over a range of tributyrin concentrations (0 -43.4 mM), whereas holding the colipase and lipase concentrations constant at a one to one molar ratio (2800 ng of PTL, 560 ng of colipase).
Mutant Colipase Binding Assays-Binding assays were performed as previously described (13). Briefly, the reactions were carried out in binding buffer containing 0.1 M Tris-HCl, pH 8, 150 mM NaCl, 2 mM CaCl 2 , and 4 mM TDC. 2800 ng of wild type lipase and 560 ng of mutant colipase were added to 1 ml of binding buffer in a 1.5-ml microcentrifuge tube that had been precoated with 1 mg/ml BSA in binding buffer. The samples were vortexed for 1 min, then centrifuged at 14,000 rpm for 5 min. 800 l of the upper, aqueous phase was carefully removed and placed in a microcentrifuge tube that had also been precoated with 1 mg/ml BSA in binding buffer. The samples were immediately placed on ice and stored at Ϫ20°C. The amount of lipase remaining in the aqueous phase was determined by activity assay using 200 l of the aqueous phase and 270 ng of wild type procolipase. The amount of lipase bound to the substrate surface was calculated as the difference between the amount of lipase initially added and the amount recovered in the aqueous phase. We developed a radioimmunoassay to determine the amount of colipase remaining in the aqueous phase (see below). The amount of procolipase bound to the substrate was calculated as the difference between the total amount initially added, and the amount remaining in solution.
Radioimmunoassay for Colipase Quantitation-The amount of procolipase bound to the interface was determined by radioimmunoassay. 100-l aliquots of the aqueous phase were tested for each mutant. The wild type control curve samples (0 -16 ng of wild type colipase), which were linear over this range of colipase concentration, were brought up to a total volume of 100 l with binding buffer. All other dilutions were performed with TBST (10 mM Tris-HCl, pH 8.0, 0.15 M NaCl, 0.1% Tween 20) containing 1 mg/ml bovine serum albumin. Wild type procolipase labeled with I 125 was diluted to a final concentration of ϳ120,000 cpm per 200 l. Polyclonal anti-colipase antibody was diluted 1:160. Each diluted procolipase sample was placed in a 1.5-ml microcentrifuge tube that had been precoated with 1 mg/ml BSA in TBST. 200 l of the labeled procolipase dilution was added to each tube, followed by 200 l of the anti-colipase antibody dilution. The contents of each tube were mixed and allowed to sit at room temperature for 1 h. 30 l of a Protein A (Sigma Chemical Co., St Louis, MO) 50% slurry in TBST was added to each tube, mixed, and incubated at room temperature for an additional hour. The precipitate was collected by centrifugation and then washed in 500 l of TBST three times. The pellet was resuspended in 100 l of TBST and counted using a Packard Cobra II Auto Gamma counter.
Statistical Analysis-Descriptive statistics were calculated using Microsoft Excel. Regression analysis was performed with Sigma Plot using a single rectangular hyperbola model. R 2 values Ն 0.94 were obtained for all regression lines used to determine K d , K m , B max , or V max . Data are presented as the mean Ϯ S.D. of the fit.

RESULTS
Colipase Mutants-The crystal structures of the pig colipasehuman PTL complex identified polar interactions between residues Arg 44 , Glu 45 , Glu 64 , Arg 65 , and Asn 89 of porcine colipase and residues in the C-terminal domain of PTL (6). Although the interactions clearly form in crystals, the same bonds may not stabilize the colipase⅐PTL complex in solution. To determine whether or not these interactions contribute to the binding between human colipase and PTL in solution, we created a variety of single-and double-substitution mutants in human colipase (Fig. 1). Initially, we substituted a small neutral amino acid, alanine, in positions Glu 45 , Glu 64 , and Arg 65 , which are conserved in human colipase, and in positions Ser 44 and Asp 89 , which are not conserved. We next substituted amino acids with identical charge but different side-chain length, with opposite charge in the side chain or with a bulky side chain to determine which properties of the side chain influence binding of colipase to PTL. Each of the mutant proteins was expressed in a yeast system and purified by immunoaffinity and gel filtration chromatography. Analysis of the purified proteins by SDS-PAGE revealed a single broad band for each protein indicative of an uncontaminated purified protein suitable for further characterization (Fig. 2).
Activity of Colipase Mutants at Various TDC Concentrations-The effect of bile salts on PTL varies depending on the concentration of the bile salts (2,14). Below their critical micelle concentration (CMC) bile salts stimulate PTL activity. Above their CMC bile salts inhibit PTL activity. Colipase increases PTL activity in both cases, indicating that colipase and PTL interact below and above the CMC of a given bile salt. Although colipase and PTL probably form a complex at all bile salt concentrations, the interactions stabilizing the complex in the presence of bile salt monomers could differ from those in the presence of bile salt micelles. Thus, we tested the ability of the colipase mutants to stimulate PTL activity at various concentrations of TDC, which has a CMC of about 1.1 mM under the assay conditions (Fig. 3). In the absence of colipase, the activity of PTL increased at 0.5 mM TDC and, then, dropped sharply as the TDC concentration increased (Fig. 3, A, OE; B, ࡗ). The addition of native colipase increased PTL activity at all TDC concentrations (Fig. 3, q). The S44A mutant restored full PTL activity at all TDC concentrations tested, suggesting that Ser 44 does not contribute to the binding of human colipase to human PTL (Fig. 3A, E). The other alanine mutants separated into two groups by putative binding site (Fig. 3A). Below the CMC, single alanine substitutions into the Glu 45 /Asp 89 and Glu 64 /Arg 65 binding sites had little effect on PTL activity compared with native colipase. Both double-substitution mutants had decreased function but still increased PTL activity above the no colipase levels. By 4 mM TDC, the E64A and R65A mutants restored less than 5% of the PTL activity compared with native colipase, and the E64A/R65A double mutant restored no activity to PTL. The function of the single-substitution mutants in the other binding site remained relatively preserved even at 4 mM TDC. The E45A restored full activity to PTL, and the D89A restored 45% of activity. Mutating both Glu 45 and Asp 89 decreased the function of the double mutant to about 5% of native colipase. These findings indicate that both sites on colipase contribute to the binding between colipase and PTL on colipase. Of the two sites, Glu 64 /Arg 65 appears to form essential interactions with PTL.
We extended these observations with studies on additional colipase mutants. These mutants were created to alter the physical properties of the side chains in specific ways. We changed the side-chain charge, length, and bulk in Glu 45 , Asp 89 , Glu 64 , and Arg 65 . Several of these mutants, E45D, E45Q, and D89R, did not express well and could not be analyzed. Of the remaining mutants, only the substitution of lysine for arginine at position 65 was tolerated. PTL had full activity with the R65K colipase mutant (Fig. 3B, E). The other mutants had relatively preserved function below the CMC of TDC compared with their greatly decreased function above the CMC. At position 45, the substitution of a bulky side group abolished the ability to restore PTL activity above the CMC of TDC (Fig. 3B,  f). Likewise, the substitution of a negatively charged amino acid, glutamic acid, for a positively charged amino acid, arginine, at position 65 eliminated function above the CMC (Fig.  3B, Ⅺ). Position 64 was sensitive to both chain length and charge with both the E64D and E64Q mutant colipases showing decreased function (Fig. 3B, and ƒ). The results confirm the importance of the targeted sites in the interaction between colipase and PTL.
Stability of the Mutants-Decreased stability of the mutants in 4 mM TDC could also explain the above results. To test the stability of the mutants, we incubated them in 8 M urea or 4 mM TDC for 30 min at room temperature. After the incubation we determined the ability of colipase and each colipase mutant to activate PTL. Under both conditions native colipase retained full function, as did each of the colipase mutants. These results make it unlikely that any of the mutations altered the stability of colipase to a significant degree. Concentration Dependence of the Alanine Mutants-The assays at various TDC concentrations suggested that bile salt micelles might affect formation of productive complexes between colipase and PTL. To better characterize any differences in the interaction between colipase and PTL in the presence and absence of bile salt micelles, we measured the ability of the colipase mutants to reactivate PTL over a range of colipase concentrations in the presence of excess substrate and a constant amount of PTL (Fig. 4). From these data, we determined the concentration of the colipase mutants that restored halfmaximal activity to PTL (K d ) and the concentration that restored maximal PTL activity (B max ) in 0.5 and 4 mM TDC. The apparent K d and B max were determined by nonlinear regres- sion with a rectangular hyperbola function. Given the large excess of substrate in the assays, the apparent K d and B max most likely reflect the properties of interactions between colipase and PTL rather than between colipase and substrate. The experimental curves for the activity measured in 0.5 mM TDC are shown in Fig. 4 (A and B). All of the tested mutants had an apparent K d and B max equal to that of wild type colipase with the exception of the R65E mutant, which had an apparent K d value ϳ4-fold weaker than wild type ( Table I). The preserved function of the colipase mutations under these conditions indicated that the mutations did not grossly disrupt the structure of the colipase mutants.
At 4 mM TDC, we see differences in the curves (Fig. 4, C-F). The curves shift to the right for most of the mutants indicating a change in affinity of the colipase mutants for PTL. The apparent B max for all mutants equals that of native colipase, whereas the apparent K d increased for all of the mutants (Table II). Mutations of the Glu 64 /Arg 65 binding site affected the apparent K d to a greater degree than mutations of the Glu 45 /Asp 89 binding site. The greatest effect was seen with the R65E mutant, which had an apparent K d for lipase that was ϳ260-fold weaker than that of native colipase, a further 60-fold reduction in apparent K d from that measured in 0.5 mM TDC. The decrease in apparent K d of the colipase mutants agrees with the hypothesis that the targeted residues participate in the formation of colipase⅐PTL complexes through direct interactions with PTL. The large effect of TDC micelles on the apparent K d suggests that micelles may influence the binding of colipase and PTL either directly or indirectly.
Substrate Dependence of the Colipase Mutants-To assure that mutations of the hydrophilic surface of colipase did not also affect the interaction of the colipase⅐PTL complex to substrate, we measured the activity of the mutants over a range of substrate concentrations (Fig. 5). Enzymatic lipolysis does not lend itself to standard steady-state analysis due to the interfacial mechanism, constantly changing substrate, and the requirement of a cofactor (14). As a result, true Michaelis-Menten parameters cannot be easily measured. Still, parameters determined under constant, predetermined conditions provide a measure of the relative function among the colipase mutants. The assays contained 0.5 or 4.0 mM TDC and equimolar amounts of PTL and the colipase mutants. In 0.5 mM TDC, wild type colipase and all of the colipase mutants increased PTL activity to the same level. In 4.0 mM TDC, two mutants, E45A and R65K, had curves identical to that of native colipase (Fig.  5, A, E; B, E). With the remaining mutants, PTL had decreased maximal velocities. These impressions were confirmed by calculation of the apparent K m and maximal V max of the colipase⅐PTL complex for tributyrin (Table III). The mutants and wild type colipase had similar values for K m and V max in 0.5 mM TDC. The complex produced with all tested mutants in 4.0 mM TDC had an apparent K m equal to that of the complex with wild type colipase. E45A and R65K⅐PTL complexes had normal apparent maximal velocities. Complexes with the remaining mutants had decreased apparent V max . PTL had low activity with E64A, and the data did not provide a good fit. Consequently, the apparent constants were not reported, but the shape of the curve suggested that the E64A mutation primarily affected the maximal velocity. The decreased apparent V max indicated that some of the mutants form fewer productive complexes with PTL in 4.0 mM TDC than does native colipase, a finding consistent with decreased affinity of the colipase mutants for PTL. Once a complex forms between the various mutant colipases and PTL, the complex has a normal affinity for the substrate. That is, the mutations in the hydrophilic surface of colipase do not affect the binding of the complex to the substrate interface once the complex has formed.
Anchoring of PTL to Substrate by the Colipase Mutants-To address the possibility that the mutant proteins anchor PTL to substrate, but in an inactive confirmation, we directly measured the ability of each colipase mutant to bind substrate and to anchor lipase on the substrate surface. We measured binding with a centrifugation assay that rapidly separates the organic lipid phase from the aqueous phase. Each assay contained an equimolar ratio of colipase and PTL in an emulsion of tributyrin and 4.0 mM TDC. After separation of the phases the amount of colipase and PTL remaining in the aqueous phase was determined as described under "Materials and Methods." In this assay, all of the colipase mutants partitioned with the substrate phase to a degree similar to native colipase, 80% compared with 67% to 93% for the various mutants, which indicates that binding of the colipase mutants to tributyrin was minimally impaired (Fig. 6, black bars). The amount of lipase anchored to the substrate surface varied from 85% for E45A to 8% for E45W, as compared with 90% for wild type colipase (Fig.  6, light gray bars). The amount of lipase anchored to the substrate was proportional to the activity of each mutant (Fig. 6,  dark gray bars). Even though the colipase mutants bind to substrate normally, they cannot anchor PTL efficiently, which is predicted by the hypothesis that the targeted residues in colipase mediate complex formation between colipase and PTL. DISCUSSION In the crystal structure of porcine colipase bound to PTL, polar interactions form between colipase residues Glu 45 , Asn 89 (Asp 89 in human colipase), Glu 64 , and Arg 65 and residues in the C-terminal domain of PTL (6). We mutated each putative lipase-binding residue of colipase to determine whether these amino acids participate in complex formation when the proteins are in solution. Characterization of the colipase mutants showed that substitutions in each targeted site decreased the ability of colipase to bind PTL in the presence of bile salt   , forms a hydrogen bond with Asn 366 , and contributes to the water bridge noted above. Asp 89 has a hydrogen bond with Lys 400 of PTL and contributes to a water bridge with Gly 66 of colipase and Leu 444 of PTL. Additionally, site-specific mutagenesis of Glu 45 and Lys 400 of PTL suggested that the ion pairing between these two residues was essential for lipolysis (9). Our results question this conclusion. The preserved function of the E45A colipase mutant suggests that the salt bridge between Glu 45 of colipase and Lys 400 of PTL is not essential or, alternatively, may not form in solution.
Regardless of which binding site was mutated, bile salts altered the behavior of most single substitution mutants. The magnitude of the effect depended on the bile salt concentration.  At 0.5 mM TDC, when monomers of bile salts spread at the oil-water interface, the affinity of the colipase mutants for PTL was unchanged except for R65E. At higher TDC concentrations, above the CMC, the affinity of the colipase mutants decreased appreciably for all but the E45A and S44A mutants. At least two models can be offered to explain this observation. The first model proposes that colipase and PTL must form a complex with bile salt micelles in aqueous solution before binding to the interface. There is support for this model in the literature. Multiple studies have demonstrated that bile salt micelles bind to colipase and to PTL (15)(16)(17)(18). A recent study with small angle neutron scattering found a detergent micelle associated with the PTL⅐colipase complex indicating that a complex can form between PTL, colipase, and micelles (19).
The second model states that colipase and PTL have poor affinity for each other in aqueous solution. In the absence of substrate PTL preferentially binds to bile salt micelles. When a substrate is added, colipase binds to the substrate first. After formation of a complex with substrate, the bound colipase has a higher affinity for PTL than that of the bile salt micelles. PTL then binds to colipase on the substrate surface. Several studies support this model. The concept of a sequential mechanism where colipase binds to the substrate first followed by PTL has been previously proposed and supported by various experiments (2). Other studies using microcalorimetry, two-phase partitioning, and affinity chromatography with immobilized colipase demonstrated that bile salt micelles bind PTL and weaken the binding of colipase to PTL, which may favor the binding of colipase to the substrate surface rather than to complexes of PTL and bile salt micelles (20 -22). In another study, monomeric bile salt solutions allowed the absorption of PTL to an interface, but a micellar bile salt solution completely inhibited PTL binding to an interface (23). Two kinetic studies suggested that bile salt micelles or multiple bile salts bind to PTL and inhibit absorption of PTL to the substrate-water interface even in the presence of colipase (24,25).
Both models could account for our data. Mutations in the hydrophilic surface of colipase could destabilize the complex of PTL, bile salts and colipase predicted to form by in the first model. This disruption could occur because the colipase mutants have decreased affinity for the PTL⅐bile salt complex or because the mutants bind to PTL in an incorrect conformation producing an inactive complex. The fact that the colipase mutants have lowered affinity for PTL and can bind to the substrate but not anchor PTL even as an inactive complex argues against the last possibility. The weakened binding affinity of the colipase mutants for PTL is also consistent with the second model. In this instance the decreased affinity of the colipase mutants for PTL allows the bile salt micelles to compete for PTL binding, and thereby inhibit the formation of an active colipase⅐PTL complex.
Although the role of bile salt micelles in the lipolytic mechanism remains unclear, many molecular details of lipolysis have been described in the past decade. X-ray crystallography and site-specific mutagenesis identified the catalytic triad of PTL. The crystal structures of the colipase⅐PTL complex and the characterization of naturally occurring and designed lipases with mutations in the lid have demonstrated a role for the lid in substrate selectivity, in substrate binding, and interactions with colipase. Monolayer experiments, kinetic analysis, and site-specific mutagenesis have broadened the role of colipase in the lipolytic mechanism. It stabilizes the lid in the open, active conformation and may direct PTL to substrate patches on the surface of dietary emulsions, which contain high concentrations of phospholipids and cholesterol esters. This work extends these observations to better define the crucial interactions between colipase and PTL. An increased understanding of the molecular details of this important physiologic process provides the information needed to rationally modulate PTL activity for therapeutic purposes.
FIG. 6. Binding of colipase mutants to substrate and colipasemediated anchoring of PTL to substrate. Colipase binding to tributyrin and its ability to anchor PTL to tributyrin was determined as described under "Materials and Methods." The binding and anchoring data are expressed as the percent bound of the total colipase or PTL added. The ability of each mutant to activate PTL in the presence of 4.0 mM TDC is graphed as a percentage of PTL activity with native colipase. The activity was determined with a 1:2 molar ratio of colipase to PTL. A, alanine mutants; B, nonalanine mutants. Black bars, colipase binding; light gray bars, anchoring of PTL; dark gray bars, PTL activity.