Báo cáo khoa học: Role of the structural domains in the functional properties of pancreatic lipase-related protein 2

Role of the structural domains in the functional properties of pancreatic lipase-related protein 2 Ame´ lie Berton, Corinne Sebban-Kreuzer and Isabelle Crenon

UMR, INSERM 476, INRA 1260, Universite´ de la Me´ diterrane´ e, Nutrition Humaine et Lipides, Faculte´ de Me´ decine de la Timone, Marseille, France

Keywords chimera; colipase; domain; pancreatic lipase; PLRP2

Correspondence I. Crenon, UMR, 476 INSERM ⁄ 1260 INRA, Faculte´ de Me´ decine, 27 Boulevard Jean-Moulin, 13385 Marseille Cedex 5, France Fax: +33 4 91 78 21 01 Tel: +33 4 91 29 41 10 E-mail: Isabelle.Crenon@medecine. univ-mrs.fr

(Received 7 August 2007, revised 10 September 2007, accepted 1 October 2007)

doi:10.1111/j.1742-4658.2007.06123.x

Although structurally similar, classic pancreatic lipase (PL) and pancreatic lipase-related protein (PLRP)2, expressed in the pancreas of several species, differ in substrate specificity, sensitivity to bile salts and colipase depen- dence. In order to investigate the role of the two domains of PLRP2 in the function of the protein, two chimeric proteins were designed by swapping the N and C structural domains between the horse PL (Nc and Cc domains) and the horse PLRP2 (N2 and C2 domains). NcC2 and N2Cc proteins were expressed in insect cells, purified by one-step chromatogra- phy, and characterized. NcC2 displays the same specific activity as PL, whereas N2Cc has the same as that PLRP2. In contrast to N2Cc, NcC2 is highly sensitive to interfacial denaturation. The lipolytic activity of both chimeric proteins is inhibited by bile salts and is not restored by colipase. Only N2Cc is found to be a strong inhibitor of PL activity, due to compe- tition for colipase binding. Active site-directed inhibition experiments dem- onstrate that activation of N2Cc occurs in the presence of bile salt and does not require colipase, as does PLRP2. The inability of PLRP2 to form a high-affinity complex with colipase is only due to the C-terminal domain. Indeed, the N-terminal domain can interact with the colipase. PLRP2 properties such as substrate selectivity, specific activity, bile salt-dependent activation and interfacial stability depend on the nature of the N-terminal domain.

human pancreatic lipase-related juice from different species and also in other secretions [8,11–15].

Abbreviations E600, diethyl p-nitrophenyl phosphate; ho, horse; NaTDC, sodium taurodeoxycholate; PL, classic pancreatic lipase; PLRP, pancreatic lipase-related protein.

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6011

In 1992, Giller et al. isolated mRNA coding for two novel proteins (PLRPs) showing a high level of identity with the human classic pancreatic lipase [1]. On the basis of amino acid sequence comparisons, Giller et al. pro- posed the classification of pancreatic lipases in three subgroups: classic pancreatic lipase (PL), PLRP1 and PLRP2. Numerous PLRP sequences have been identi- fied in several species by isolating mRNA [2–11]. Furthermore, by using classic protein purification procedures, the presence of PLRP1 and ⁄ or PLRP2 has been demonstrated in the pancreas or in the pancreatic PLRP and PL differ in enzymatic properties such as substrate specificity, sensitivity to inhibition by bile salts and colipase dependence [16]. Pancreatic lipases are highly active and selective for triglyceride sub- strates. Under physiological conditions, the PL activity is dependent on the presence of colipase, which able to overcome the inhibitory effect of bile salts [17,18]. Despite extensive studies on a large variety of sub- strates, only very low lipolytic activity against triglyce- rides has been reported with PLRP1 [1,4,14,15]. The

A. Berton et al.

PLRP2 and colipase ) no interaction

the structure of

lipases by PLRP2s are distinguishable from classical their substrate specificity, because, besides triglycerides, they are able to hydrolyze phospholipids, galactolipids and esters of vitamins [3,8,9,19,20]. Moreover, the activities of PL and PLRP2 seem to be different according to the vehicles in which the substrate is solu- bilized [21].

indicate that, as predicted by high primary structure homology, the three-dimensional PLRP members can be superimposed on that of PL, which cannot explain the particular features of the PLRPs. Indeed, they possess an N-terminal domain with the same catalytic triad, a C-terminal domain in which the residues implicated in colipase binding are conserved, and a lid domain (except for guinea pig PLRP2, which has a naturally truncated lid [3]), which must move during the activation process. Previous data indicate that the motion of the PLRP2 lid is dependent on the presence of bile salts and does not require the presence of colipase [32].

Concerning the effect of bile salts and colipase on the activity of PLRP2, there is no clear conclusion, because the results appear to change according to the different species and seem to depend on the triglyceride substrate and the bile salt [9,16]. Indeed, some of them, such as horse PLRP2 and human PLRP2, are inhibited on tributyrin substrate by the presence of bile salts, and this inhibition is only slightly overcome or not overcome by the presence of an excess of colipase [8,13,22]. Another PLRP2 group including guinea pig and coypu PLRP2s is affected neither by the bile salt concentration nor by the addition of colipase on tri- butyrin substrate, but is strongly inhibited on trioctan- oin substrate [3]. Concerning the rat PLRP2, because of contradictory results, no conclusions can be drawn [4,23]. the various PLRP2s imply that

The two-domain structural organization of the pan- creatic lipases allowed the development of the domain- exchange strategy to provide further insights into the structure–function relationships of pancreatic lipases [14,33–36]. These studies show that the lid domain alone is responsible neither for the substrate selectivity nor for the activation process. They did not show whether the PLRP2 C-terminal domain could or could not bind colipase. The differences in kinetic properties of these proteins should not be grouped together and that it is impor- tant to obtain new information about the properties of PLRP2 family members.

contribution of respective

In the present study, we produced, purified and characterized chimeric proteins designed by N-terminal and C-terminal domain exchange between horse PLRP2 (hoPLRP2) and horse PL (hoPL) in order to investigate the role of the two domains in the function of hoPLRP2. The influence of bile salt and colipase on the lipase activity of the different chimeras was investi- gated using tributyrin as substrate. Experiments were performed to investigate active site-directed inhibition and competition for colipase binding. The properties of the chimeras were compared to those of the original proteins bearing the modifications induced by the con- structions in the chimeric proteins and compared with the properties of the native hoPL and hoPLRP2 pro- teins. This work provided new information about the ability of the hoPLRP2 C-terminal domain to bind each colipase and the PLRP2 domain to the activation process, the substrate specificity and the interfacial stability of this protein.

Results

Expression and purification of chimeric proteins

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6012

Chimeric proteins designed by domain exchange between hoPL and hoPLRP2 were constructed and expressed in insect cells. The strategy that we followed The three-dimensional structure resolution of pancre- atic lipases provides important information concerning the structure–function relationship of the PL [24–28]. In agreement with biochemical data, these structures demonstrate the functional organization of the PL into two structural domains: a large N-terminal domain, which contains the active site with the catalytic triad formed by Ser152, Asp176 and His263, and a smaller C-terminal domain, which is important for colipase binding. In the inactivated state, the PL catalytic site is inaccessible to substrate, being covered by a surface loop called the lid domain (residues 237–261). In partic- ular conditions, the lid must move to accommodate a lipid substrate. The closed PL conformation converts into the open form upon interaction with lipid [26]. The principal elements that undergo space reorganization during the activation of the enzyme are the lid (resi- dues 238–262) and loop b5 of the N-terminal domain (residues 77–86). The functional consequences of the structural reorganization are as follows: (a) the active site is accessible to the substrate; (b) the oxyanion hole is created; (c) an important hydrophobic surface is formed; and (d) a new binding site is generated between the colipase and the open lid. Some studies have ques- tioned whether lipase activation is even interfacial in the presence of bile salt and colipase, on the basis of attaining an activated ternary complex of PL, colipase and a small micelle in the absence of any interface [29]. Three-dimensional structures of canine PLRP1 and rat PLRP2 have also been reported [30,31]. These data

A. Berton et al.

PLRP2 and colipase ) no interaction

than the lane 5, and NcC2,

to construct plasmids expressing the two chimeric cDNAs was to exchange the cDNA fragment encoding the C-terminal domain of the two lipases between plas- mids pVLhoPL and pAcGP67hoPLRP2, which carry the cDNA of hoPL and hoPLRP2, respectively. This exchange could be done because an Eag1 site was engi- neered in each plasmid at the junction between the N-terminal and C-terminal domain sequences of each protein. The chimeric protein composed of the N-ter- minal domain of hoPL and of the C-terminal domain of hoPLRP2 was named NcC2. Conversely, the other chimeric protein, bearing the N-terminal domain of hoPLRP2 and the C-terminal domain of hoPL, was named N2Cc. This construction procedure induced substitutions in the amino acid sequence of each C-ter- minal domain, as shown in Fig. 1. To ensure that these the substitutions did not influence the behavior of C-terminal domain as compared to the wild-type proteins, we expressed NcCc and N2C2 as controls.

In the absence of dithiothreitol in the sample buffer, native hoPL and hoPLRP2 ran as a single band of about 50 kDa (Fig. 2A, lanes 1 and 2). In the presence of dithiothreitol in the sample buffer, in contrast to lane 3), hoPLRP2 ran as two frag- hoPL (Fig. 2A, ments of 27.5 and 22.5 kDa (Fig. 2A, lane 4), in agree- ment with previous results demonstrating the high sensitivity of the hoPLRP2 Ser244–Thr245 bond to proteolytic cleavage [32]. The four chimeric proteins ran as a major single band with a molecular mass of about 50 kDa (Fig. 2A, lanes 5–8). Nevertheless, the two chimeras bearing the N2 domain (N2C2, lane 6, lane 8) had a slightly higher molecular and N2Cc, chimera bearing the Nc domain mass (NcCc, lane 7) according to the theoretical value, as indicated in Table 1. Micro- sequencing of purified NcCc and NcC2 yielded the N-terminal sequence NEVCY, corresponding to the N-terminal sequence of the mature hoPL. The N-ter- minal sequence of N2C2 and N2Cc was ADLKE, corresponding to the three terminal amino acid exten- sion resulting from the construction strategy, followed by the N-terminal sequence of the mature hoPLRP2. These results indicated that the cleavage of the signal sequence by the insect signal peptidase was correct.

Fig. 1. Functional maps of the plasmids expressing natural and chimeric isoforms of hoPLRP2 and hoPL. A novel EagI site was engineered (see Experimental procedures). Above each plasmid map, the nucleotide and amino acid sequences of the region at the junction between the two protein domains are reported. The EagI site is underlined.

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6013

The chimeric proteins were expressed in insect cells using the Baculovirus Expression System. The four proteins were secreted into the culture medium with yields reaching 10–40 mg of recombinant proteinÆL)1. After 5 days of culture, the secreted proteins were purified from the dialyzed supernatant by a one-step anionic exchange chromatography procedure, with a recovery yield of 50%. The four purified recombinant proteins were analyzed and compared to native hoPL and hoPLRP2 by SDS ⁄ PAGE followed by Coomassie blue staining (Fig. 2A) or western blot (Fig. 2B,C). Despite crossreactions due to the strong homology between the two proteins, hoPL antibodies recognized hoPL and NcCc better than hoPLRP2 and N2C2, and conversely, hoPLRP2 antibodies reacted better with hoPLRP2 and N2C2 than they did with hoPL and

A. Berton et al.

PLRP2 and colipase ) no interaction

A

of N2C2 and N2Cc to proteolytic cleavage generating one fragment detected by hoPLRP2 antibodies and corresponding to the larger proteolytic fragment of native hoPLRP2 (Fig. 2C, lanes 6 and 8). Using differ- ent preparations of N2C2 and N2Cc, we checked that this proteolysis did not have an effect on either the activity or the behavior of the proteins.

Kinetic properties of chimeric proteins ) effects of bile salts and colipase

B

C

the surface of

Fig. 2. Analysis of purified protein by SDS ⁄ PAGE 12% Coomassie blue staining (A) and western blots using hoPL antibodies (B) and hoPLRP2 antibodies (C). Lanes 1 and 2: protein migration without dithiothreitol. Lanes 3–8: protein migration with dithiothreitol. Lanes 1 and 3: hoPL. Lanes 2 and 4: hoPLRP2. Lane 5: NcCc. Lane 6: N2C2. Lane 7: NcC2. Lane 8: N2Cc.

the modifications introduced at

Table 1. Theorical biochemical properties of the chimeric proteins.

The lipolytic activity of the different chimeric proteins was investigated by titrimetry using emulsified tributy- rin as substrate. In a first experiment, the assays were performed in the absence of bile salts [sodium tauro- deoxycholate (NaTDC)], in the absence or in the pres- ence of colipase. As shown in Fig. 3, the kinetic rate for NcCc (3000 UÆmg)1) rapidly decreased in the absence of colipase and bile salts. NcCc was probably irreversibly inactivated at tributyrin droplets. Prior addition of colipase enhanced the kinetic rate (7200 UÆmg)1) and prevented this inactiva- tion. This well-known phenomenon, named interfacial inactivation, has been extensively described with sev- eral pancreatic classic lipases, and in particular with hoPL [37]. The kinetic rate for N2C2 in the absence of bile salt was constant (560 UÆmg)1), as it was in the absence and in the presence of colipase (Fig. 3). Simi- lar data were observed with hoPLRP2 [8]. These results indicated that NcCc and N2C2 behaved like respectively [8,35,37]. native hoPL and hoPLRP2, Thus, the junction between the N-terminal and C-terminal domains as compared to the native proteins influence neither their stability nor their activity.

Proteins

N-terminal sequence

Molecular mass (Da)

Amino acids

Isoelectric point

hoPL hoPLRP2 NcCc N2C2 NcC2 N2Cc

NEVCY ADLKE NEVCY ADLKE NEVCY ADLKE

49 710.47 50 362.16 49 696.45 50 387.23 49 559.32 50 524.36

449 455 449 455 451 453

5.19 5.50 5.19 5.62 5.80 5.10

(6000 UÆmg)1)

the protein at

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6014

NcCc (Fig. 2B,C). Concerning the chimera, NcC2 was recognized better by hoPL antibodies than by hoPLRP2 antibodies, and conversely, N2Cc was recog- nized better by hoPLRP2 antibodies than by hoPL antibodies. This suggests that both antibodies were preferentially raised against the N-terminal domain of the respective proteins. We observed slight sensitivity In the absence of bile salts and colipase, the kinetic and N2Cc rates of both NcC2 (650 UÆmg)1) decreased, and this decrease was even more rapid for NcC2 (Fig. 3). These results indicated that both the N-terminal and C-terminal domains of hoPLRP2 contributed to the stability of the protein in the presence of the water–lipid interface. Also, both the N-terminal and C-terminal domains of hoPL were involved in the inactivation of the water–triglyceride interface. The inactivation of N2Cc in the absence of bile salts was prevented by prior addition of colipase, suggesting that N2Cc was able to bind the colipase. In contrast, the inactivation of NcC2 was not prevented by prior addition of colipase, suggesting that NcC2 was not able to bind colipase. These results indicated that only the proteins possess- ing the PL C-terminal domain are able to bind coli- pase.

A. Berton et al.

PLRP2 and colipase ) no interaction

M), NcC2 M) or

M), N2C2 (1.75 · 10)9

M) without (in black) and

M).

Fig. 3. Kinetics of hydrolysis of tributyrin by chimeric proteins without bile salt and in presence or absence of colipase. Lipolytic activity was measured titrimetrically at pH 7.5 with NcCc (0.99 · 10)9 (0.89 · 10)9 N2Cc (0.5 · 10)9 with (in gray) colipase (5 · 10)9

to those obtained for

at (300 UÆmg)1

Fig. 4. Bile salt and colipase dependence of the chimeric protein activity. The assays were done using 10)9 M each lipases in the pH-stat at various concentrations of NaTDC and in the absence (in black) or presence (in gray) of a 5 M excess of colipase.

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6015

In a second experiment, the activities of the chime- ras were tested on emulsified tributyrin in the presence of increasing concentrations of bile salts (0–6 mm), in the absence or in the presence of colipase. As seen in Fig. 4, increasing the concentration of NaTDC inhib- ited the activity of NcCc (1500 UÆmg)1 at 6 mm NaTDC versus 3200 UÆmg)1 at 0 mm NaTDC) and 6 mm NaTDC versus N2C2 560 UÆmg)1 at 0 mm NaTDC). In the presence of colipase, only NcCc activity was increased, even in the presence of bile salt (8000 UÆmg)1 at 0.1 mm NaTDC and 5600 UÆmg)1 at 6 mm NaTDC). These results were similar the native proteins [8,35,37]. The activity of NcC2 was slightly increased at a very low NaTDC concentration (8000 UÆmg)1 at 0.1 mm NaTDC) and inhibited when the NaTDC concentration increased. The inhibitory effect was complete for NaTDC concentrations above 2 mm. The colipase was not able to restore the NcC2 activity. For N2Cc, a slight activator effect was observed at a very

A. Berton et al.

PLRP2 and colipase ) no interaction

a plateau value corresponding

low NaTDC concentration (700 UÆmg)1 at 0.1 mm NaTDC). An inhibitory effect then appeared, increased up to the NaTDC critical micellar concentration, and stabilized at to 400 UÆmg)1. Interestingly, the colipase failed to restore the maximal activity for N2Cc. The colipase effect on the lipase activity in the presence of bile salt depended not only on the presence of the classic C-terminal domain, but also on the nature of the N-terminal domain.

Fig. 5. Competition for colipase between PL and inactive or chime- ric proteins. Colipase (10)9 M) was incubated with increasing concentrations of inhibitor protein in the presence of a tributyrin emulsion and bile salts at a final concentration of 4 mM. After 5 min, PL (10)9 M) was added. The activity was determined and expressed as a percentage compared to the lipase activity mea- sured in the absence of inhibitor protein. (A) E600-hoPL (d) and E600-hoPLRP2 (.). (B) N2Cc (d) and NcC2 (.).

Inhibition of the PL by the chimeras

effect of E600-hoPL and N2Cc is probably due to competition for colipase binding. These data suggested that the C-terminal domain of PL is able to bind coli- pase whatever the nature of the N-terminal domain. Moreover, the C-terminal domain of hoPLRP2 was not able to bind colipase even in the presence of the PL N-terminal domain.

In the presence of a supramicellar concentration of NaTDC (4 mm), PL needs the colipase to develop its full activity. In the same conditions, the N2Cc and NcC2 activities were inhibited and not restored in the presence of colipase (see above). The influence of increasing concentrations of NcC2, N2Cc, hoPL inac- tive forms [diethyl p-nitrophenyl phosphate (E600)- hoPL] and hoPLRP2 inactive forms (E600-hoPLRP2) on the native PL activity was investigated. In these experiments, the concentrations of lipase (10)9 m) and colipase (10)9 m) were constant. Inactive forms of hoPL and hoPLRP2 were prepared as previously described [32] using high concentrations of E600, which covalently binds to the active site serine. As shown in Fig. 5A, E600-hoPL was found to be an excellent inhibitor of the lipase activity, as 50% inhibi- tion was obtained with an [E600-hoPL] ⁄ [PL] molar ratio of 0.5. Only 18% of residual activity remained when E600-hoPL was used at a molar excess of 2. Interestingly, no inhibition of the lipase test activity was observed when E600-hoPLRP2 was added, even at a molar excess of 1800. As shown in Fig. 5B, the inhibitory effect with N2Cc was similar to that of E600-hoPL. Fifty per cent inhibition was obtained with an [N2Cc] ⁄ [PL] molar ratio of about 0.5, and complete inhibition was observed when N2Cc was used at a molar excess of 10. The inhibitory effect of E600- hoPL and N2Cc was abolished when an excess of coli- pase was added during the assay, and was observed only in the presence of NaTDC (data not shown). In the case of NcC2, no inhibition of the lipase activity was observed, as 100% of the lipase activity still remained even at a molar excess of 45. The effect of NcC2 was similar to that of E600-hoPLRP2. The same results were obtained using human or porcine lipase and colipase.

Influence of NaTDC and colipase on chimera inhibition by E600

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6016

The activation of the pancreatic lipase is a mechanism allowing accessibility of the active site to the substrate The proteins bearing the C-terminal domain of PL were efficient inhibitors of the lipase activity, whereas the proteins bearing the C-terminal domain of PLRP2 had no effect on the lipase activity. The inhibitory

A. Berton et al.

PLRP2 and colipase ) no interaction

M, were incubated in M). T50% is the time needed

Table 2. Influence of bile salt and colipase on the chimeric protein inhibition by E600. Chimeric proteins, at 2 · 10)6 the presence of E600 in the absence or in the presence of bile salt (NaTDC 0.5 mM or 4 mM) or colipase (10)5 to reach 50% inhibition. ND, not determined.

T50% (min)

Without colipase

With colipase

– NaTDC

0.5 mM NaTDC

4 mM NaTDC

– NaTDC

4 mM NaTDC

Proteins

E600 (mM)

> 1440 16 > 1440 30

> 1440 75 > 1440 90 > 1440 > 1440

ND ND

> 1440 70 > 1440 64 > 1440 > 1440

> 1440 7 > 1440 6 70 79

> 1440 6 > 1440 3 > 1440 75

NcCc N2C2 NcC2 N2Cc NcCc NcC2

0.05 0.05 0.05 0.05 2.5 2.5

increased versus (T50% ¼ 64 min inhibition was T50% ¼ 90 min), in contrast to N2C2.

Blank experiments performed in the absence of E600 showed that, in any case, proteins retained at least 85% of activity after 24 h, indicating that the enzymes were stable under the conditions used for the study.

and resulting in the unmasking of the catalytic triad of the enzyme induced by the motion of the flap. The accessibility of the active site can be tested using the ability of an organophosphate, E600, to react with the active site serine only when the enzyme adopts an opened flap conformation. E600 inhibition experiments were carried out to investigate the influence of NaTDC and colipase on the activation of the chimeric proteins. Table 2 shows T50%, corresponding to the time needed to reach 50% inhibition.

With 0.05 mm E600, no inhibition was observed for NcCc and NcC2. At 2.5 mm E600, inhibition of NcCc activity was observed after incubation in the presence of bile salt and colipase (T50% ¼ 70 min). In contrast, inhibition of NcC2 activity was observed in the pres- ence of bile salt alone (T50% ¼ 75 min), and the coli- pase addition had no effect on the rate of inhibition.

These results indicated that the NcCc active site was accessible to high E600 concentrations only in the presence of colipase and bile salt, whereas the accessi- bility of the NcC2 active site depended only on the presence of bile salt. The accessibility of the N2C2 and N2Cc active sites to E600 was possible even in the absence of colipase and bile salt, and was considerably increased by the presence of bile salt. Therefore, the concentration of E600 needed to obtain clear inhibi- tion of NcC2 and NcCc was 50 times higher than that used for N2C2 and N2Cc. In conclusion, the accessi- bility of the active site was better in the protein bear- ing the N2 domain than in the protein bearing the Nc domain. Thus, the accessibility of the active site in the N2 proteins was independent of the nature of the C-terminal domain, in contrast to the situation with Nc proteins. Indeed, the C2 domain induced sensitivity of the Nc active site to E600 inhibition in the presence of bile salt.

Discussion

In the absence of bile salt, noticeable inhibition of N2C2 by 0.05 mm E600 was observed (T50% ¼ 75 min), and the addition of colipase had no signifi- cant effect (T50% ¼ 70 min). In contrast, the rate of inhibition was significantly increased in the presence of NaTDC monomers (T50% ¼ 16 min). At NaTDC con- centrations beyond the critical micellar concentration, inhibition of N2C2 increased (T50% ¼ the rate of 6 min). The addition of colipase still had no significant influence (T50% ¼ 7 min). These results were in agree- ment with previous experiments on inhibition by E600 performed on native hoPLRP2 [32]. Significant

considerable variability

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6017

inhibition by E600 was observed for N2Cc in the absence of colipase and bile salt (T50% ¼ 90 min). The rate of inhibition was increased in the presence of NaTDC, concentrations beyond the critical micellar concentration having a higher efficiency than concentration (T50% ¼ 3 min versus the monomer T50% ¼ 30 min). The addition of colipase alone had a slight influence on N2Cc inhibition, as the rate of Despite their structural similarities, the PLRP2s form a subfamily that is clearly distinct from the classic lipase subfamily, notably concerning their functional properties. Moreover, is observed among the members of the PLRP2 subfamily. The aim of our study was to investigate the contribu- tion of the N-terminal and C-terminal domains to the particular behavior of hoPLRP2. The structural orga- nization of the pancreatic lipases is completely suitable

A. Berton et al.

PLRP2 and colipase ) no interaction

interfacial activation [3,9], and preferentially hydro- lyzes triglycerides with short chains.

for the domain-exchange strategy, which has previ- ously been used successfully in the study of the struc- ture–function relationships of different lipases [35,38]. Chimeric proteins, named NcC2 and N2Cc, were designed by N and C structural domain exchange between hoPL (Nc and Cc domains) and hoPLRP2 (N2 and C2 domains). NcC2 and N2Cc were produced as secreted proteins and purified. Their properties were compared to those of NcCc and N2C2, corresponding to the original proteins PL and PLRP2, respectively, bearing the modifications induced by the construction in the chimeric proteins. These modifications have no effects on the behavior of the proteins [8,35,37]. The kinetic characterizations of proteins

hoPLRP2 is characterized by a specific activity on TC4 of about 600–700 UÆmg)1 in the absence of bile salts. The specific activity of hoPL is 8000 UÆmg)1 (in the presence of colipase). In the absence of bile salts, the proteins containing the same N-terminal domain show a similar specific activity on tributyrin (in the presence of colipase) (500–700 UÆmg)1 for N2 proteins and 6000–8000 UÆmg)1 for Nc proteins). In the pres- ence of bile salts, we observed that the behavior of chimeric proteins is also related to the nature of the N-terminal domain. Indeed, the specific activities of NcCc and NcC2 are very strongly decreased, whereas those of N2C2 and N2Cc are much less sensitive to the inhibitory effect of bile salts. This indicates that in the absence or in the presence of bile salts, the specific activity of hoPL and hoPLRP2 depends on the nature of their N-terminal domain.

involvement of

the lipid–water

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6018

In contrast to NcCc, NcC2 was not protected from the interfacial denaturation by colipase and not reacti- vated by colipase in the presence of bile salt, suggest- ing that NcC2 is not able to form a stable complex with colipase. Competition experiments on colipase binding reveal that NcC2, like hoPLRP2, is a very bad competitor. This observation indicates that NcC2 and N2C2 do not bind well to colipase, probably due to the C2 domain. However, as shown in Fig. 6, the resi- dues of the C-terminal domain involved in the primary interaction of PL with colipase are preserved in the C-terminal domain of hoPLRP2 (Asn366, Gln369, Lys400). It is possible that these residues are not in an ideal conformation to allow either binding to colipase in particular for or correct orientation of colipase, stabilizing the lid. Although N2Cc activity was not restored by colipase in the presence of bile salt, there are substantial arguments in favor of N2Cc–colipase complex formation. N2Cc is protected from interfacial denaturation by colipase and behaves as an excellent inhibitor of colipase binding. Indeed, the [N2Cc] ⁄ [lipase] molar ratio needed to obtain 50% inhibition is the same as that found with E600-hoPL competitor or with other inactive forms of PL used as competitors by Miled et al. [41]. This result indicates that N2Cc binds to colipase as well as hoPL. Experiments previ- ously carried out with the C-terminal domain of PL as inhibitor showed that a [C-terminal domain] ⁄ [lipase] molar ratio of 1000 was needed to give 50% inhibition [42]. It was assumed that the whole lipase is a better inhibitor than the C-terminal domain alone, because new interactions, which stabilize the lipase–colipase complex, were created between colipase and the lid lipase in the opened conformation. The results of in the absence of bile salts and colipase show that, in con- trast to N2C2, the NcCc, NcC2 and N2Cc proteins undergo irreversible inactivation, which is thought to result from denaturation of these enzymes in the lipid– water interface [39]. These observations underline the fact that the proteins possessing at least one of the two domains of hoPL are more sensitive to interfacial the C-terminal denaturation. The domain in the interfacial denaturation of the PL was already proposed by Carrie` re et al. [35]. Indeed, these authors showed that, in the absence of bile salts, a chimeric protein composed of the N-terminal domain of guinea pig PLRP2 and of the C-terminal domain of human PL (gpN2 ⁄ huCc) was inactivated at the inter- face. Moreover, it was reported that the C-terminal domain of PL bound efficiently to a triglyceride–water interface and was an absolute requirement for possible interfacial binding of PL [40]. In the case of the gpN2 ⁄ huCc chimera, the rate of denaturation was higher, indicating that the C-terminal domain of hoPL is less sensitive to the interface than that of huPL. In the present work, the similarities between NcCc and NcC2 with regard to the rate of inactivation show for the first time the dominant role of the N-terminal domain of PL in the phenomenon of interfacial dena- turation. The N-terminal domain of PLRP2 does not possess this feature. PLRP2 is not sensitive to interfa- cial denaturation; either the two domains confer high stability on the lipid–water interface, or PLRP2 has no affinity for interface. We recently showed that PL and PLRP2 hydrolyzed retinyl esters. Moreover, PL preferentially hydrolyzed the substrate when it was included in droplets, and PLRP2 was more efficient when it was included in micelles of smal- ler size [21]. Even if PL and PLRP2 hydrolyze triglyce- rides, it is probable that the physical property of the substrate is specific for each enzyme: droplet for PL, and a water-soluble structure for PLRP2. It has also been previously reported that PLRP2 does not display

A. Berton et al.

PLRP2 and colipase ) no interaction

A

B

Fig. 6. (A) Amino acid sequence comparison between hoPLRP2 and hoPL. The residues of the catalytic triad are in red, the lid sequence is in blue, and the amino acids of the C-terminal domain involved in colipase binding are in green. (B) Superimposition of hoPLRP2 (1W52) and hoPL (1HPL) Ca traces are displayed in red and blue, respectively.

et al. [21]. Nevertheless,

active site. This observation supports the idea that PLRP2 preferentially hydrolyzes substrates that are sol- uble or included in micelles, as was proposed by Re- the mechanism of boul activation of PLRP2, which involves the C2 domain, is different from that of PL, which involves the Cc domain and colipase. The C-terminal domain alone is involved in the affinity of the PLRP2 for micellar sub- strates, and probably allows the interaction of the enzyme with the substrate structure. Whether the motion of the lid promotes the recognition of the sub- strate structure, or the recognition of this structure pro- motes the displacement of the lid, is still questionable.

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6019

obtained with N2Cc as a competitor mean that the C-terminal domain in this context is able to bind colipase, but especially that the complex formed would probably be stabilized by the open lid of the N2 domain. The movement of the lid making it possible to adopt an open conformation is a crucial stage in the mecha- nism of action of lipase. The motion of the flap makes the active site accessible to the substrate, simulta- neously forming a functional oxyanion hole and gener- ating lipase interfacial binding. It has been previously proposed from active site-directed inhibition experi- ments with an organophosphate that the accessibility of the active site of pancreatic lipase, in the absence of interface, could be obtained in the presence of colipase and bile salts, probably through the formation of a ter- nary lipase–colipase–micelle complex of biliary com- pounds [29]. The same type of experiment indicates that the lid of PLRP2 is already more mobile than that of PL, and especially that it moves and uncovers the active site in the presence only of the bile compounds [32]. The accessibility of E600 to the N2 active site is considerably increased in the presence of bile salts, which masks the slight influence of colipase observed with N2Cc in the absence of bile salt. The unmasking of the active site of the Nc domain absolutely requires colipase and bile salts in micellar concentrations in the presence of the Cc domain and bile salt only in the presence of the C2 domain. The inhibition of the active site serine by E600 needs both the motion of the flap and the recognition of the vehicle in which E600 was solubilized. It was clearly established that soluble E600 can be included in bile salt micelles, and that this inclu- sion is a prerequisite for inhibition of PL [43]. Our work indicates that E600 included in bile salt micelles is a better inhibitor of the N2 active site than of the Nc Three structures of PLRP2 are now available in the Protein Data Bank: rat PLRP2 (Protein Data Bank code 1BU8 [31], human PLRP2 (Protein Data Bank code 2OXE, to be published), and hoPLRP2 (Protein Data Bank code 1W52 [44]). These three PLRP2 struc- tures are comparable to the hoPL structure in the closed conformation, or to the porcine PL structure in the opened conformation [25,26]. With respect to the conformation of loop b5 of the N-terminal domain, hoPLRP2 and human PLRP2 are probably in the opened conformation, in contrast to rat PLRP2, which is in the closed conformation. For the human and rat proteins, it is not possible to draw conclusions about the exact position of the lid, because a sequence of approximately 20 amino acids is missing. In the case of hoPLRP2, the lid is partially opened (Fig. 6). A comparison of the exposed surface between PLRP2 and PL in the opened conformation would explain the difference in behavior between PL and PLRP2 with respect to the interface. The resolution of the structure of N2Cc would be very useful to determine the posi- tion of the opened lid, whether it can be stabilized by

A. Berton et al.

PLRP2 and colipase ) no interaction

junction between the N-terminal and C-terminal domain sequences (named Nc and Cc, respectively) that induced the substitution A337G [14]. The resulting vector pVLNcCc encoded the protein named NcCc.

shows

The N-terminal and C-terminal domain sequences of hoP- LRP2 (named N2 and C2, respectively) were amplified by PCR using pAcGP67hoPLRP2, previously described [8], as template. This plasmid resulted in the insertion of the mature hoPLRP2 sequence into the BamH1 ⁄ EcoR1 restriction site of the pAcGP67 transfer vector, downstream of the signal sequence of the baculovirus glycoprotein GP67. The two oligonucleotides 5¢-N2 (5¢-GGAATTCAGATCTCAAAGA GGTTTGCTATACCCC-3¢) and 3¢-N2 (5¢-CCCGGCCG TAGTCACCACTTTCTCC-3¢) were used as 5¢ and 3¢ pri- mer, respectively, to amplify the N2 sequence. The sequences in bold correspond to the Bgl2 restriction site for primer 5¢-N2 and Eag1 restriction site for primer 3¢-N2. The under- lined sequences in the primers correspond to the sequences encoding the first and the last residues of N2, respectively.

colipase, and what the nature is of the amino acids that correspond to the exposed surface. The superposi- tion of hoPL and hoPLRP2 (Fig. 6) that loop b5¢ of the C-terminal domain (residues 405–414) is oriented differently. This observation is very interest- ing, because this loop was shown to play an important role in lipase function and could influence the binding of colipase [45]. No conclusion is possible about the orientation of this b5¢ loop in the other PLRP2s, as this fragment was found to have no interpretable elec- tron density.

In conclusion, the studies on the functional properties the two structural N-terminal and C-terminal of domains of hoPLRP2 show that the enzyme stability in the presence of the lipid–water interface, the motion of the lid and the substrate specificity are properties that are mainly related to the nature of the N-terminal domain. On the other hand, PLRP2 is not able to form a stable complex with colipase, and its C-terminal domain is responsible for this feature. Structural analy- sis of this domain will provide new information to enable a better understanding of the role of the C-termi- nal domain in the function of PLRP2, mainly with regard to the orientation of the residues essential for col- ipase binding and the behavior of PLRP2 towards the lipid–water interface or water-soluble micelles. These structural data will be very important to determine the real contribution of PLRP2 to intestinal lipolysis.

Experimental procedures

To amplify the C2 sequence, the two oligonucleotides (5¢-C2) 5¢-CCCGGCCGTTGGAGATATAGAGTATC-3¢ and (3¢-C2) 5¢-GGTTCTTGCCGGGTCCCCAGG-3¢ were used. The sequence in bold corresponds to the Eag1 restriction site. The sequence in italic corresponds to the sequence encoding the first residue of C2. The 3¢-C2 primer corresponds to the end of the multiple cloning site of the pAcGP67 vector. The underlined sequence corresponds to the substitutions intro- duced in the C2 domain as compared to the wild-type PLRP2. The PCR reactions were carried out under standard conditions, with 0.5 min at 95 (cid:2)C, 1 min at 50 (cid:2)C and 1 min at 72 (cid:2)C for 25 cycles. After the PCR reaction, the N2 and C2 PCR fragments were purified and digested by Bgl2–Eag1 and by Eag1, respectively, and introduced into the Bam- H1 ⁄ Eag1 restriction site of the pAcGP67 transfer vector. The resulting construction pAcN2C2 encoded the protein named N2C2. NcCc and N2C2 were used as controls.

respectively. The

resulting

The BaculoGold Starter Package pVL1393 and pAcGP67 transfer vectors were purchased from Pharmingen (San Diego, CA). X-Press medium and fetal bovine serum were supplied by BioWhittaker (Walkersville, MD, USA). Anti- biotics were obtained from Invitrogen (Carlsbad, CA, USA). Alkaline phosphatase-labeled goat anti-(rabbit IgG), E600, tributyrin and NaTDC were purchased from Sigma-Aldrich (St Louis, MO, USA). Taq polymerase, restriction enzymes and T4 DNA ligase were purchased from New England Bio- labs (Ipswich, MA, USA) or Eurogentec (Seraing, Belgium).

The pVLNcCc and pAcN2C2 vectors were digested by Eag1 and subjected to electrophoresis on polyacrylamide gel, and the Eag1 fragments were electroeluted. The small Eag1 fragments corresponding to the Cc and C2 domains were interchanged and cloned in the Eag1 pAcN2 and pVLNc fragments, constructions pVLNcC2 and pAcN2Cc encoded the chimeric proteins NcC2 and N2Cc. All the constructions were propagated in the JM101 Escherichia coli strain and checked by DNA sequencing carried out by Genome Express (Grenoble, France).

Reagents

Construction of chimeric proteins

After purification using the Qiagen (Hilden, Germany) plas- mid purification protocol, the different constructions were used with linearized genomic DNA from Autographa

The constructions encoding the chimeric proteins composed of a PL domain and a PLRP2 domain are described in Fig. 1. First, pVLhoPL, previously described, resulted in the integration of the nucleotide sequence encoding hoPL, including the peptide signal, at the EcoR1 restriction site of pVL1393 transfer vector. Thereafter, an Eag1 restriction the site was introduced by site-directed mutagenesis at

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6020

Expression of chimeric proteins using the Baculovirus Expression System

A. Berton et al.

PLRP2 and colipase ) no interaction

1 : 5000 dilution, and the reacting antibodies were detected with goat anti-rabbit IgG conjugated with alkaline phos- phatase at a 1 : 5000 dilution.

The lipase activity was measured titrimetrically at 25 (cid:2)C using emulsified 0.11 m tributyrin in 1 mm Tris ⁄ HCl buffer (pH 7.5) containing 0.1 m NaCl and 5 mm CaCl2. The assays were performed either in the absence or in the pres- ence of bile salt (NaTDC) 0.1–4 mm and ⁄ or a five-fold molar excess of colipase. One unit corresponds to the release of 1 lmol fatty acidÆmin)1.

Protein concentrations were determined with the bicinch- oninic acid protein assay reagent (Pierce, Rockford, IL, USA).

Activity measurements and protein assays

californica virus (BaculoGold DNA from the BaculoGold transfection kit) for cotransfection into Sf21 insect cells as described in the Baculovirus Expression Vector System Manual (Pharmingen). The SF21 cells were grown in a monolayer at 27 (cid:2)C in tissue culture flasks using X-press medium containing 5% fetal bovine serum, 50 UIÆmL)1 penicillin and 50 mgÆmL)1 streptomycin. Recombinant viruses were purified by plaque assay and amplified by an additional SF21 cell infection cycle. For the production of the chimera, six 162 cm2 tissue culture flasks were seeded with 6 · 107 cells per flask in complete X-press medium. When the cells were attached, the complete medium was removed and replaced with 20 mL of serum-free X-press medium. The high-titer stock solutions of recombinant baculoviruses were added to the cells at a multiplicity of infection close to 2. In all cases, the chimeric proteins were secreted into the culture media, as observed by electrophore- sis on SDS ⁄ PAGE. After 5 days of culture at 27 (cid:2)C, the cells were pelleted by centrifugation at 900 g for 5 min 4 (cid:2)C, and the supernatants were kept at 4 (cid:2)C for further purification.

Chimeric protein inhibition by E600

The inhibition experiments were performed in 50 mm sodium acetate buffer (pH 6.0) containing 0.1 m NaCl. Pro- teins (2 · 10)6 m) were treated with E600 (0.05 or 2.5 mm as indicated), either in the absence or in the presence of bile salt and ⁄ or colipase (five-fold molar excess). The mixture was incubated at 25 (cid:2)C, and aliquots were withdrawn at various time intervals and used to determine the remaining lipase activity as described above. Control experiments were also performed without E600 to check protein stability.

Purification of chimeric protein

All the chimeric proteins were purified following the one- step procedure reported previously for the purification of horse recombinant PLRP2 expressed in insect cells [8]. The supernatants were dialyzed overnight at 4 (cid:2)C culture against 20 mm Tris ⁄ HCl buffer (pH 8) containing 1 mm benzamidine and loaded onto a Q-sepharose Fast Flow col- umn equilibrated in the same buffer. Elution was performed (from 0 to using a linear NaCl concentration gradient 200 mm NaCl). The fractions were analyzed by measuring the lipase activity and by SDS ⁄ PAGE. The fractions con- taining the protein of interest were pooled, dialyzed over- night at 4 (cid:2)C against distilled water, lyophilized or not, and kept at ) 20 (cid:2)C. Native hoPL and native hoPLRP2 were purified as previously described [8].

Competition experiments

The lipase activity was measured as described above in the presence of 4 mm NaTDC, colipase and increasing concen- trations of chimeric proteins. The lipase and colipase con- centrations were 10)9 m. The colipase and the chimeric proteins were added at the beginning of the test, and the lipase was added during the test. Under these conditions, the lipase activity was determined and expressed as a percentage of the lipase activity measured in the absence of chimeric protein. Native hoPLRP2 and hoPL previously inactivated by E600 as described in [32] were used as controls.

N-terminal sequence analysis

Acknowledgements

The purified chimeric proteins were submitted to N-termi- nal microsequencing. Stepwise Edman degradation was per- formed using an automatic sequencer, model Procise 494, from Applied Biosystems (Foster City, CA, USA).

electrophoresis, proteins were

Gel electrophoresis and western blotting

Electrophoresis on 12% polyacrylamide gels was carried out in the presence of SDS as described by Laemmli [46]. Western blots were performed according to Burnette [47]. After transferred to a polyvinylidene difluoride membrane. Membranes were incubated with rabbit polyclonal antibodies raised against either hoPL or hoPLRP2. The rabbit sera were used at a

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6021

This research was supported by grants from the Insti- tut National de la Sante´ et de la Recherche Me´ dicale and from the Institut National de la Recherche Agron- omique. The PhD work of Miss Ame´ lie Berton was supported by a grant from Institut National de la Recherche Agronomique and ARILAIT RECHER- CHE Industry. We thank Re´ gine Lebrun and Danielle Moinier for performing the sequence analyses, and Mouhcine Louaste for helpful technical assistance. We

A. Berton et al.

PLRP2 and colipase ) no interaction

thank Dr Catherine Chapus for helpful advice and fruitful discussions. We are very grateful to Dr Franc¸ - oise Guerlesquin for critical reading of the manuscript.

References

1 Giller T, Buchwald P, Blum-Kaelin D & Hunziker W

13 De Caro J, Sias B, Grandval P, Ferrato F, Halimi H, Carriere F & De Caro A (2004) Characterization of pancreatic lipase-related protein 2 isolated from human pancreatic juice. Biochim Biophys Acta 1701, 89–99. 14 Crenon I, Foglizzo E, Kerfelec B, Verine A, Pignol D, Hermoso J, Bonicel J & Chapus C (1998) Pancreatic lipase-related protein type I: a specialized lipase or an inactive enzyme. Protein Eng 11, 135–142.

(1992) Two novel human pancreatic lipase related pro- teins, hPLRP1 and hPLRP2. Differences in colipase dependence and in lipase activity. J Biol Chem 267, 16509–16516.

15 De Caro J, Carriere F, Barboni P, Giller T, Verger R & De Caro A (1998) Pancreatic lipase-related protein 1 (PLRP1) is present in the pancreatic juice of several species. Biochim Biophys Acta 1387, 331–341.

16 Lowe ME (2002) The triglyceride lipases of the pan-

creas. J Lipid Res 43, 2007–2016.

17 Borgstrom B (1975) On the interactions between pancre-

2 Kerfelec B, LaForge KS, Puigserver A & Scheele G (1986) Primary structures of canine pancreatic lipase and phospholipase A2 messenger RNAs. Pancreas 1, 430–437.

3 Hjorth A, Carriere F, Cudrey C, Woldike H, Boel E, Lawson DM, Ferrato F, Cambillau C, Dodson GG, Thim L et al. (1993) A structural domain (the lid) found in pancreatic lipases is absent in the guinea pig (phos- pho) lipase. Biochemistry 32, 4702–4707.

atic lipase and colipase and the substrate, and the importance of bile salts. J Lipid Res 16, 411–417. 18 Brockman H (2000) Kinetic behavior of the pancreatic lipase–colipase–lipid system. Biochimie 82, 987–995. 19 Andersson L, Carriere F, Lowe ME, Nilsson A & Ver- ger R (1996) Pancreatic lipase-related protein 2 but not classical pancreatic lipase hydrolyzes galactolipids. Biochim Biophys Acta 1302, 236–240.

4 Payne RM, Sims HF, Jennens ML & Lowe ME (1994) Rat pancreatic lipase and two related proteins: enzy- matic properties and mRNA expression during develop- ment. Am J Physiol 266, G914–G921.

20 Sias B, Ferrato F, Grandval P, Lafont D, Boullanger P, De Caro A, Leboeuf B, Verger R & Carriere F (2004) Human pancreatic lipase-related protein 2 is a galacto- lipase. Biochemistry 43, 10138–10148.

5 Wicker-Planquart C & Puigserver A (1992) Primary structure of rat pancreatic lipase mRNA. FEBS Lett 296, 61–66.

6 Remington SG, Lima PH & Nelson JD (1999) Pancre-

21 Reboul E, Berton A, Moussa M, Kreuzer C, Crenon I & Borel P (2006) Pancreatic lipase and pancreatic lipase-related protein 2, but not pancreatic lipase-related protein 1, hydrolyze retinyl palmitate in physiological conditions. Biochim Biophys Acta 1761, 4–10.

22 Sebban-Kreuzer C, Deprez-Beauclair P, Berton A &

atic lipase-related protein 1 mRNA in female mouse lac- rimal gland. Invest Ophthalmol Vis Sci 40, 1081–1090. 7 Wishart MJ, Andrews PC, Nichols R, Blevins GT Jr, Logsdon CD & Williams JA (1993) Identification and cloning of GP-3 from rat pancreatic acinar zymogen granules as a glycosylated membrane-associated lipase. J Biol Chem 268, 10303–10311.

Crenon I (2006) High-level expression of nonglycosylat- ed human pancreatic lipase-related protein 2 in Pichia pastoris. Protein Expr Purif 49, 284–291.

8 Jayne S, Kerfelec B, Foglizzo E, Chapus C & Crenon I (2002) High expression in adult horse of PLRP2 dis- playing a low phospholipase activity. Biochim Biophys Acta 1594, 255–265.

23 Jennens ML & Lowe ME (1995) Rat GP-3 is a pancre- atic lipase with kinetic properties that differ from coli- pase-dependent pancreatic lipase. J Lipid Res 36, 2374– 2382.

9 Thirstrup K, Verger R & Carriere F (1994) Evidence

24 Winkler FK, D’Arcy A & Hunziker W (1990) Structure

of human pancreatic lipase. Nature 343, 771–774.

for a pancreatic lipase subfamily with new kinetic prop- erties. Biochemistry 33, 2748–2756.

25 Bourne Y, Martinez C, Kerfelec B, Lombardo D, Cha- pus C & Cambillau C (1994) Horse pancreatic lipase. The crystal structure refined at 2.3 A resolution. J Mol Biol 238, 709–732.

26 van Tilbeurgh H, Egloff MP, Martinez C, Rugani N,

Verger R & Cambillau C (1993) Interfacial activation of the lipase–procolipase complex by mixed micelles revealed by X-ray crystallography. Nature 362, 814–820.

10 Grusby MJ, Nabavi N, Wong H, Dick RF, Bluestone JA, Schotz MC & Glimcher LH (1990) Cloning of an interleukin-4 inducible gene from cytotoxic T lympho- cytes and its identification as a lipase. Cell 60, 451–459. 11 Sias B, Ferrato F, Pellicer-Rubio MT, Forgerit Y, Guil- louet P, Leboeuf B & Carriere F (2005) Cloning and seasonal secretion of the pancreatic lipase-related pro- tein 2 present in goat seminal plasma. Biochim Biophys Acta 1686, 169–180.

12 Fauvel J, Bonnefis MJ, Sarda L, Chap H, Thouvenot

27 van Tilbeurgh H, Sarda L, Verger R & Cambillau C (1992) Structure of the pancreatic lipase–procolipase complex. Nature 359, 159–162.

28 Hermoso J, Pignol D, Kerfelec B, Crenon I, Chapus C & Fontecilla-Camps JC (1996) Lipase activation by

JP & Douste-Blazy L (1981) Purification of two lipases with high phospholipase A1 activity from guinea-pig pancreas. Biochim Biophys Acta 663, 446–456.

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6022

A. Berton et al.

PLRP2 and colipase ) no interaction

nonionic detergents. The crystal structure of the porcine lipase–colipase–tetraethylene glycol monooctyl ether complex. J Biol Chem 271, 18007–18016.

38 Wong H, Davis RC, Nikazy J, Seebart KE & Schotz MC (1991) Domain exchange: characterization of a chimeric lipase of hepatic lipase and lipoprotein lipase. Proc Natl Acad Sci USA 88, 11290–11294.

39 Rietsch J, Pattus F, Desnuelle P & Verger R (1977)

29 Pignol D, Hermoso J, Kerfelec B, Crenon I, Chapus C & Fontecilla-Camps JC (1998) The lipase ⁄ colipase com- plex is activated by a micelle: neutron crystallographic evidence. Chem Phys Lipids 93, 123–129.

Further studies of mode of action of lipolytic enzymes. J Biol Chem 252, 4313–4318.

40 Chahinian H, Bezzine S, Ferrato F, Ivanova MG, Perez B, Lowe ME & Carriere F (2002) The beta 5¢ loop of the pancreatic lipase C2-like domain plays a critical role in the lipase–lipid interactions. Biochemistry 41, 13725– 13735.

30 Roussel A, de Caro J, Bezzine S, Gastinel L, de Caro A, Carriere F, Leydier S, Verger R & Cambillau C (1998) Reactivation of the totally inactive pancreatic lipase RP1 by structure-predicted point mutations. Proteins 32, 523–531.

41 Miled N, Berti-Dupuis L, Riviere M, Carriere F &

31 Roussel A, Yang Y, Ferrato F, Verger R, Cambillau C & Lowe M (1998) Structure and activity of rat pancre- atic lipase-related protein 2. J Biol Chem 273, 32121– 32128.

Verger R (2003) In vitro lipolysis by human pancreatic lipase is specifically abolished by its inactive forms. Biochim Biophys Acta 1645, 241–246.

32 Jayne S, Kerfelec B, Foglizzo E, Granon S, Hermoso J,

42 Ayvazian L, Kerfelec B, Granon S, Foglizzo E, Crenon

Chapus C & Crenon I (2002) Activation of horse PLRP2 by bile salts does not require colipase. Biochemistry 41, 8422–8428.

I, Dubois C & Chapus C (2001) The lipase C-terminal domain. A novel unusual inhibitor of pancreatic lipase activity. J Biol Chem 276, 14014–14018.

33 Crenon I, Jayne S, Kerfelec B, Hermoso J, Pignol D & Chapus C (1998) Pancreatic lipase-related protein type 1: a double mutation restores a significant lipase activ- ity. Biochem Biophys Res Commun 246, 513–517.

43 Rouard M, Sari H, Nurit S, Entressangles B & Desnu- elle P (1978) Inhibition of pancreatic lipase by mixed micelles of diethyl p-nitrophenyl phosphate and bile salts. Biochim Biophys Acta 530, 227–235.

34 Bezzine S, Roussel A, de Caro J, Gastinel L, de Caro A, Carriere F, Leydier S, Verger R & Cambillau C (1998) An inactive pancreatic lipase-related protein is activated into a triglyceride-lipase by mutagenesis based on the 3-D structure. Chem Phys Lipids 93, 103–114.

35 Carriere F, Thirstrup K, Hjorth S, Ferrato F, Nielsen

44 Mancheno JM, Jayne S, Kerfelec B, Chapus C, Crenon I & Hermoso JA (2004) Crystallization of a proteolyzed form of the horse pancreatic lipase-related protein 2: structural basis for the specific detergent requirement. Acta Crystallogr D Biol Crystallogr 60, 2107–2109. 45 Freie AB, Ferrato F, Carriere F & Lowe ME (2006) Val-407 and Ile-408 in the beta5¢-loop of pancreatic lipase mediate lipase–colipase interactions in the pres- ence of bile salt micelles. J Biol Chem 281, 7793–7800.

46 Laemmli UK (1970) Cleavage of structural proteins

PF, Withers-Martinez C, Cambillau C, Boel E, Thim L & Verger R (1997) Pancreatic lipase structure–function relationships by domain exchange. Biochemistry 36, 239–248.

36 Yang Y & Lowe ME (2000) The open lid mediates pan-

during the assembly of the head of bacteriophage T4. Nature 227, 680–685.

creatic lipase function. J Lipid Res 41, 48–57.

47 Burnette WN (1981) ‘Western blotting’: electrophoretic transfer of proteins from sodium dodecyl sulfate–poly- acrylamide gels to unmodified nitrocellulose and radio- graphic detection with antibody and radioiodinated protein A. Anal Biochem 112, 195–203.

37 Gargouri Y, Bensalah A, Douchet I & Verger R (1995) Kinetic behaviour of pancreatic lipase in five species using emulsions and monomolecular films of synthetic glycerides. Biochim Biophys Acta 1257, 223–229.

FEBS Journal 274 (2007) 6011–6023 Journal compilation ª 2007 FEBS. No claim to original French government works

6023