Skip to main content
PMC home page
Protein Science : A Publication of the Protein Society logoLink to Protein Science : A Publication of the Protein Society
. 2002 Sep;11(9):2237–2246. doi: 10.1110/ps.0216302

Mapping sequence differences between thimet oligopeptidase and neurolysin implicates key residues in substrate recognition

Kallol Ray 1, Christina S Hines 1, David W Rodgers 1
PMCID: PMC2373592  PMID: 12192079

Abstract

The highly homologous endopeptidases thimet oligopeptidase and neurolysin are both restricted to short peptide substrates and share many of the same cleavage sites on bioactive and synthetic peptides. They sometimes target different sites on the same peptide, however, and defining the determinants of differential recognition will help us to understand how both enzymes specifically target a wide variety of cleavage site sequences. We have mapped the positions of the 224 surface residues that differ in sequence between the two enzymes onto the surface of the neurolysin crystal structure. Although the deep active site channel accounts for about one quarter of the total surface area, only 11% of the residue differences map to this region. Four isolated sequence changes (R470/E469, R491/M490, N496/H495, and T499/R498; neurolysin residues given first) are well positioned to affect recognition of substrate peptides, and differences in cleavage site specificity can be largely rationalized on the basis of these changes. We also mapped the positions of three cysteine residues believed to be responsible for multimerization of thimet oligopeptidase, a process that inactivates the enzyme. These residues are clustered on the outside of one channel wall, where multimerization via disulfide formation is unlikely to block the substrate-binding site. Finally, we mapped the regulatory phosphorylation site in thimet oligopeptidase to a location on the outside of the molecule well away from the active site, which indicates this modification has an indirect effect on activity.

Keywords: Thimet oligopeptidase, neurolysin, specificity, substrate, model

The M3 family of zinc metallopeptidases has nine members that share various degrees of sequence similarity (Rawlings and Barrett 1995). Like several other metallopeptidase families, they have a common active site sequence motif, His-Glu-Xaa-Xaa-His (HEXXH), which forms part of the binding site for the metal cofactor and the catalytic water molecule (Matthews et al. 1974). Two of the family members, neurolysin (EC 3.4.24.16) and thimet oligopeptidase (TOP; EC 3.4.24.15), are mammalian enzymes with ∼60% sequence identity. They are widely distributed in various tissues, including the central nervous system, and appear to be present in different subcellular compartments in different cell types (Vincent et al. 1996; Kato et al. 1997; Vincent et al. 1997a; Crack et al. 1999; Ferro et al. 1999; Garrido et al. 1999; Oliveira et al. 2000). Both peptidases are known to hydrolyze in vitro various bioactive peptides, including neuropeptides (Dahms and Mentlein 1992; Dando et al. 1993; Barrett et al. 1995), and numerous reports have linked the enzymes to the metabolism of these peptides in vivo (Molineaux et al. 1988; Molineaux and Ayala 1990; Dahms and Mentlein 1992; Da Silva et al. 1992; Barelli et al. 1994; Mentlein and Dahms 1994; Vincent et al. 1995, 1997b). They are therefore frequently classified as neuropeptidases, although they undoubtedly also perform other functions. For example, TOP is thought to hydrolyze peptides that are resistant to degradation by other cytosolic peptidases, reducing the pool of peptides available for antigen presentation (Saric et al. 2001).

Both neurolysin and TOP have complex patterns of substrate specificity (Dahms and Mentlein 1992; Dando et al. 1993; Barrett et al. 1995). They cleave at only one or a small number of sites on bioactive peptides, but the recognized cleavage sequences vary widely, with no consistent amino acid preferences at any position relative to the site of hydrolysis. The only common feature of substrates is the prevalence of basic residues, prolines, and glycines. On many peptides, the two enzymes cleave at the same peptide bond, but on a few peptides, they cleave at different positions. The subtle specificity differences between enzymes that metabolize bioactive peptides play a critical role in determining the levels and temporal patterns of these signaling molecules in different tissues and neural circuits. Thimet oligopeptidase and neurolysin are good model systems for investigating the basis for substrate specificity in this class of enzymes.

Recently, we determined the crystal structure of neurolysin at 2.3-Å resolution (Brown et al. 2001). The enzyme (Fig. 1 ▶) has a deep, narrow channel that runs the length of the molecule, and the active site is located about midway along the floor of this channel. Given the high degree of sequence identity with neurolysin, we expect TOP to adopt a nearly identical three-dimensional structure. Neither neurolysin nor TOP is active on peptides >17 residues in length, and the extensive shielding of the active sites provides a basis for this restriction. Modeling of the neuropeptide neurotensin into the channel (Fig. 1 ▶) indicated that the length restriction might be primarily on the sequence N-terminal to the cleavage site and indicated the regions likely to interact with substrate.

Fig. 1.

Fig. 1.

Open in a new tab

Molecular surface views of neurolysin (Brown et al. 2001). (A) Two views of neurolysin, rotated by 90° relative to each other, showing the long narrow substrate-binding channel. The active site of the enzyme is located near the floor of the channel. (B) Cutaway view of the channel floor with a model of the bound substrate neurotensin, a 13-residue neuropeptide. This figure and all molecular surface representations were made with the program GRASP (Kraulis 1991).

We have taken advantage of the probable structural similarity between neurolysin and TOP to help identify sequence differences that may contribute to the differential cleavage site recognition in the two enzymes. By mapping residue differences between the two enzymes onto the surface of neurolysin, we have been able to visualize their positions. Because relatively few sequence differences map to the surface near the active site region, we can assess the significance of these changes for substrate recognition. We also use mapping to the neurolysin surface to visualize the locations of residues in TOP thought to be involved in possible regulation of the activity of the enzyme. These regulatory sites are well removed from the active site and substrate-binding regions, indicating indirect effects of modification.

Results

Mapping residue sequence changes

The aligned sequences of TOP and neurolysin, which differ in length by six residues, are shown in Figure 2 ▶. There are a total of 259 differences between the two rat enzymes over the 680 aligned residues. The differences in primary sequence are well distributed over the length of the proteins, with no completely conserved continuous sequences longer than 12 residues.

Fig. 2.

Fig. 2.

Open in a new tab

Amino acid sequence alignment of thimet oligopeptidase (TOP) and neurolysin. Sequence differences are highlighted in green. Cysteine residues in TOP thought to be involved in disulfide mediated multimerization of the enzyme are in red, and the site of protein kinase A phosphorylation of TOP is in blue.

Using the crystal structure of neurolysin (Brown et al. 2001), we have determined which residues at the molecular surface differ from the corresponding residues in TOP and mapped their positions. Of the 259 sequence differences, 86.5% contribute to the surface of the neurolysin structure. The most apparent feature of the mapped sequence differences is their asymmetric distribution on the molecular surface, which contrasts with the uniform distribution of sequence changes in the primary sequence (Fig. 3 ▶). Only 11% of the changes map to the walls and floor of the substrate-binding channel, even though they account for >23% of the total surface of the molecule.

Fig. 3.

Fig. 3.

Open in a new tab

Surface view of neurolysin showing (blue) positions of surface amino acid differences with thimet oligopeptidase (TOP). (A) View showing the exterior surface of the molecule. (B) Cutaway view of the substrate-binding channel floor with sequence differences labeled. Neurolysin residue numbers are given together with the one-letter codes for neurolysin (before the residue number) and TOP (after the residue number) amino acids. Subtract one from the neurolysin residue numbers for the corresponding residue numbers in TOP. Red labels indicate positions at which the TOP amino acid is conserved in all known TOP orthologs and the neurolysin amino acid is conserved in all neurolysin orthologs. The active site zinc is colored orange. (C) The two walls of the channel with sequence differences labeled.

On the floor of the channel (Fig. 3b ▶), four isolated changes—R470/E469, R491/M490, N496/H495, and T499/R498 (the neurolysin residue is listed first)—occur in residues close to the cleavage site that potentially could interact with bound substrate. In two of these positions, R470/E469 and T499/R498, the neurolysin and TOP residues are conserved across all known sequences of the enzymes from different species. More distant from the active site are two patches of multiple differences, one expected only to interact with substrates having more than about six or seven residues N-terminal to the cleavage position, and the other to interact with residues about the same distance C-terminal to the cleavage position.

The sides of the channel have even fewer changes that map to the surface. On the wall opposite the active site residues (Fig. 3c ▶, left panel), there are only a total of six changes in three different patches. One patch relatively near the active site contains three altered positions, with one, L338/M337, conserved across species. The other wall has five changes in three patches (Fig. 3c ▶, right panel). The patch from positions E355/N354 and S454/T453 and the patch from I77/T76 and T81/N80 might interact with a substrate peptide five or six residues (C- and N-terminal, respectively) from the cleavage site. The isolated change N424/G423 is quite high on the wall over the active site, and it seems unlikely to play a role in contacting substrate bound in a catalytically competent position.

Because substrate peptide is expected to extend largely along the channel floor and the lower portion of one wall (see Fig. 1b ▶), the extensive sequence conservation on both channel walls is puzzling. One possibility is that the walls are involved in guiding substrate to the floor of the deep, narrow channel. We therefore computed the surface potential for the walls and found that both surfaces are negatively charged (Fig. 4 ▶), with the wall containing active site residues carrying the stronger potential. Because substrate peptides tend to be basic, it is possible that some of the sequence conservation on the channel walls results from their functioning to steer the peptide into the channel.

Fig. 4.

Fig. 4.

Open in a new tab

Electrostatic potential at the channel wall surfaces. The wall opposite the active site is on the left and the wall containing the active site on the right. The surface potential was calculated in GRASP (Kraulis 1991).

Differences in substrate specificity

The mapped sequence differences between TOP and neurolysin help to account for known differences in specificity between the two enzymes. As noted, the two peptidases cleave most tested bioactive peptides and other peptide sequences at the same sites (Dahms and Mentlein 1992; Dando et al. 1993; Barrett et al. 1995). There are, however, three peptides in which they have been shown to cleave at distinct single sites: neurotensin and two peptides related to portions of the bioactive peptides dynorphin A and kininogen (Fig. 5a ▶).

Fig. 5.

Fig. 5.

Open in a new tab

Substrate targeting and sequence differences in thimet oligopeptidase (TOP) and neurolysin. (A) Alignment of three substrates cleaved at different sites by TOP and neurolysin relative to sequence changes in the substrate-binding site. The three substrates are neurotensin (NT) and fluorescently labeled peptides based on portions of the dynorphin A (QF370) and kininogen (Kin) sequences. (B) Amino acids giving the lowest Km values (Oliveira et al. 2001) at each position in the parent sequence shown aligned relative to sequence differences between TOP and neurolysin in the substrate-binding site.

Based on the model of neurotensin binding (Brown et al. 2001), some of the neurolysin residues altered in TOP appear well positioned to interact with the peptide. In particular, Arg491 could interact with Glu at the P7 position, Asn496 with the Asn at P6 or Lys at P5, and Thr499 with either Arg at P2 or Arg at P3. On the C-terminal side of the scissile bond, Arg470 may be in position to interact with the C terminus of the peptide. In contrast, substitution of positively charged residues, His495 and Arg498, in TOP for the corresponding polar residues in neurolysin could create an electrostatically less favorable binding surface for the positively charged residues in neurotensin at positions P2, P3 and P5 (neurolysin registration). Also the substitution of Met490 removes the potential Arg491 interaction with Glu at P7, and substitution of Glu469 for Arg470 eliminates the possible interaction of the enzyme with the C-terminal carboxylate of the peptide. Clearly then, sequence changes could reduce the affinity of TOP for neurotensin in the registration found in neurolysin.

Altering the registration to that found in TOP (Fig. 5a ▶) shifts some of the positive electrostatic potential on the peptide away from His495 and Arg498. Indeed, this registration places the acidic glutamate, now at P5, in a position to interact with His495. Also, the leucine at P7 might contact Met490, and the arginine at P1` could potentially contact Glu469. Based on this analysis, changes in the residues likely to contact bound neurotensin are consistent with the differences in the peptide bonds targeted by TOP and neurolysin.

A similar analysis can be used to rationalize the different cleavage positions in two synthetic peptides (Fig. 5a ▶). In both cases, the hydrolyzed peptide bond shifts one position toward the N terminus of the substrate in TOP relative to neurolysin. This change moves an arginine away from Arg498 of TOP and places positively charged residues in what appear to be favorable positions (see below) at P1` or P3` (or both positions). Here they might interact with Glu469. Changes between the two enzymes at residues 491/490 and 496/495 are unlikely to favor one cleavage site over the other, and these residues probably do not play a large role in altering specificity with these peptides.

Juliano and coworkers (Oliveira et al. 2001) have determined neurolysin and TOP kinetic parameters for every possible amino acid substitution (except glycine) at each individual position in the fluorogenic substrate Abz-Gly-Phe-Ser-Pro-⇓-Phe-Arg-Gln-EDDnp. Differences in Km values, the parameter most easily related to possible changes in binding interactions, can also be at least partially rationalized by sequence variations in the active site channel (Fig. 5b ▶). On the C-terminal side of the scissile bond, the clear preference of neurolysin for aspartate at P2` is consistent with an interaction with Arg470. In TOP, on the other hand, there is no strong preference at this position. Instead, positively charged residues are preferred at P1` and P3` and are among the most favored residues at P2`, possibly reflecting interactions with Glu469. Contrary to expectations, arginine is clearly preferred by neurolysin at P1`, despite the close proximity of Arg470. In this case, other interactions must determine the preference for a positively charged residue.

N-terminal to the cleavage site, the preference of neurolysin for positively charged residues at P4 and its ability to accommodate arginine at P1 might reflect the relative absence of positive charge on the binding surface. TOP does not prefer positively charged residues at any position, consistent with the change to positively charged residues at positions 495 and 498 in the enzyme. The strong preference of TOP for acidic residues at P3 also may reflect the presence of the two positively charged enzyme residues, which are close enough to interact directly with substrate residue P3. Other preference differences, particularly at positions P1 and P2, are more difficult to rationalize, and visualization by structure determination will likely be necessary to understand them. Overall, though, many of the preference differences at individual positions can be rationalized by considering the relatively small number of sequence differences discussed here.

Modification sites

TOP is activated by low concentrations of reducing agents such as dithiothreitol or mercaptoethanol, and thiol-modifying reagents including iodoacetate and N-ethylmaleimide inhibit both TOP and neurolysin (Tisljar 1993; Serizawa et al. 1995; Shrimpton et al. 1997). Recently, activation was shown to result from reductant-induced dissociation of disulfide-linked inactive multimers of the enzyme (Shrimpton et al. 1997). The multimers lose affinity for a competitive inhibitor of TOP and therefore appear unable to bind substrate. Shrimpton et al. (1997) propose that multimerization blocks access to the TOP active site and indicate that this is a physiological mechanism by which the redox characteristics of the environment regulate the activity of the enzyme.

To assess this model of inactivation by multimerization, we mapped (Fig. 6a ▶) the positions of the three cysteine residues in TOP thought to mediate intermolecular disulfide bond formation (Shrimpton et al. 1997) onto the surface of neurolysin. These residues are all located within ∼10 Å of each other on the outer surface of one channel wall. Because no other surface cysteines appear to be involved in multimerization (Shrimpton et al. 1997), disulfide bond formation must place equivalent patches of cysteines from monomers together. The positions of the patches on the outer surface of one channel wall, however, make it unlikely that linked molecules would directly restrict access to their substrate-binding channels. It seems, therefore, that the loss of substrate affinity must result not from simple steric restriction but rather from either an induced conformational change or the inability of TOP multimers to accommodate a conformational change associated with substrate binding.

Fig. 6.

Fig. 6.

Open in a new tab

Positions of regulatory sites in thimet oligopeptidase (TOP) mapped to the neurolysin structure. (A) Two 90°-rotated views showing (blue) the positions of the cysteine residues thought to be involved in multimerization of the enzyme. Residue numbers are for neurolysin, with the identity of the neurolysin residue given before the number. Substract one for TOP residue numbers. (B) View of the neurolysin active site showing the nearby cysteine (Cys428) that may be responsible for the sensitivity of both neurolysin and TOP (Cys427) to thiol modifying reagents. The active site zinc (orange sphere) and side-chains of coordinating residues (His474, His478, and Glu503) are shown. This panel was produced with the program RIBBONS (Carson 1987). (C) Position of the TOP serine (residue 644) phosphorylated by protein kinase A.

In contrast to activation by reducing agents, inactivation by thiol modifying reagents may occur through a surface-accessible cysteine residue near the active site (Fig. 6b ▶) that is present in both TOP (Cys 427) and neurolysin (Cys 428). It seems likely that modification of this residue by a bulky reagent would inactivate the enzyme by interfering with substrate binding. In the structural model for substrate binding to neurolysin (Brown et al. 2001), the modeled peptide passes within 7 Å of the Cys 428 Sγ, indicating that any sizable addition to its side-chain would probably clash with bound substrate.

The activity of rat TOP is also modulated by protein kinase A phosphorylation at Ser 644 (Tullai et al. 2000). This modification affects activity on at least one substrate peptide, increasing both Km and kcat by sevenfold, and phosphorylation at this site was shown to occur in rat PC12 and mouse AtT-20 cells. The equivalent residue in neurolysin, Leu 645 (Fig. 2 ▶), occurs in a helix on the surface opposite the floor of the active site channel (Fig. 6c ▶). In this position, it is ∼20 Å from the active site zinc, with no obvious connection to any possible substrate-binding surfaces.

Discussion

The conservation of surface residues in the substrate-binding channels of TOP and neurolysin seems to reflect the similarities in their targeted substrate sequences rather than a requirement for conserving structure in the vicinity of the active site. Saccharolysin (Achstetter et al. 1984, 1985) is another member of the M3 family of metalloproteases expected to have the same overall and active site structure as TOP and neurolysin. Yet, we find many sequence changes that map to the channel region when we compare it with neurolysin (data not shown), indicating that more numerous changes can be tolerated in this region without compromising enzymatic activity. Our analysis here indicates that substrate targeting differences can be rationalized in part by considering the locations of four sequence changes in the channel relative to residue positions in a modeled bound substrate (Brown et al. 2001). Efforts to convert specificity from one enzyme to the other by making the appropriate changes at these positions are underway.

A somewhat surprising result is the high degree of sequence conservation on both walls of the channel, even in regions far away from the active site. These surfaces carry overall negative electrostatic potential in neurolysin, and it is reasonable to suppose that substrate access to the narrow deep channel is assisted by electrostatic steering (Adam and Delbrück 1968; Chou and Zhou 1982). It is also possible that the substrate may not always adopt the extended conformation assumed in the model of neurotensin binding. Somatostatin (Brazeau et al. 1973), for example, is a 14-residue peptide cleaved by both enzymes at multiple sites (Dahms and Mentlein 1992), despite the presence of an internal disulfide bond that prevents an extended conformation. Jacchieri and coworkers (1998) have also suggested that in some peptides, TOP hydrolysis sites may correspond to positions of β turns. Peptides that bind with a folded conformation would necessarily interact with residues higher on the channel walls than would an extended peptide bound at the floor of the channel. Conservation of channel wall residues then may maintain activity on those peptide substrates that adopt a hairpin or folded conformation.

TOP and neurolysin share the ability to cleave specifically at a variety of sequences that have no identifiable discerning features. The channel of neurolysin is lined with extended loop segments that may alter conformation to accommodate a range of substrate sequences (Brown et al. 2001). One loop (residues 600–612) adjacent to the active site is known from the crystal structure to be disordered in the absence of bound substrate peptide. In unpublished work (C.S. Hines, K. Ray, and D.W. Rodgers), we have established that altering a tyrosine residue at the tip of this loop in neurolysin differentially affects recognition of target sequences, showing that this loop is important in recognizing at least some substrates. The loop contains five glycine residues in neurolysin, but one glycine (Gly608) is changed to alanine (Ala607) in TOP. This change to alanine may alter the conformational flexibility of the loop and thereby affect target site recognition. Thus, sequence changes may have more effects than the direct interactions with substrate considered here, particularly given the potential importance of plasticity in the substrate-binding site.

The locations of the three clustered cysteine residues in TOP thought to be involved in multimerization (Shrimpton et al. 1997) indicate that some effect other than simply blocking the substrate-binding channel is involved in inactivation. The residues map to the external face of one channel wall, well removed from any possible direct interactions with bound substrate. Contacts with another molecule in this area might, however, affect the positions of two large helices (α9 and α10 in neurolysin). These helices in turn interact with the open coil region starting at residue 584 in neurolysin, which includes the disordered loop (residues 600–612) discussed above. Residues in this open coil region are positioned to interact with substrate peptides, indicating a possible mechanism for transmitting multimerization effects into the substrate-binding site. It is interesting to note that there is an intermolecular contact near this region in crystals of neurolysin, and it has not been possible to soak in substrate analogs without damaging the crystals (C.S. Hines, K. Ray, and D.W. Rodgers, unpubl.). It may be then that contacts here inhibit a conformational change necessary for substrate binding. One puzzling feature of multimerization is the apparent formation of trimers, tetramers, and even higher order aggregates (Shrimpton et al. 1997). The tight clustering of the three implicated cysteines makes it difficult to imagine how more than two molecules could link via these residues.

Finally, the location of the shown phosphorylation site (Tullai et al. 2000) is also somewhat puzzling, because it is positioned on the outside surface of the enzyme, well removed from the active site region. The presence of a negatively charged group below the substrate-binding site might increase the electrostatic interaction with the frequently basic peptide substrates, but the measurements actually show a substantial increase in Km for the only substrate strongly affected (Tullai et al. 2000). That phosphorylation of the enzyme induced only small variations in Km and kcat for the three other substrates tested indicates that the effect of phosphorylation may be specific for certain substrates. It may be that particular conformations of the enzyme required for binding these peptides are inhibited by phosphorylation.

Further analysis of substrate recognition in TOP and neurolysin will be greatly assisted by structures of the enzymes with bound peptide analogs of different sequences. Attempts to determine structures of these complexes are ongoing.

Materials and methods

The rat neurolysin and TOP sequences were aligned using CLUSTAL W (Thompson et al. 1994) through the Biology Workbench interface (Subramaniam 1998). Residues that differ in the two enzymes and contribute to the molecular surface of the neurolysin crystal structure (Protein Data Bank code 1I1I) were identified with the program GRASP (Kraulis 1991) by virtue of their solvent accessibility, which was determined with a probe of small radius (0.01 Å). The locations of these residues were visualized by coloring patches from those atoms in them that contribute to the molecular surface. Residues inside the active site channel important for differential specificity were identified with both GRASP and the program O (Jones et al. 1991), as were the positions of residues thought to be involved in covalent modification. Electrostatic potential maps of channel wall surfaces were also calculated with GRASP.

To assess our assumption that the backbone structure of TOP will be nearly identical to that of neurolysin, we submitted the TOP amino acid sequence to the SWISS-MODEL comparative protein-modeling server (Peitsch 1996), which produced a TOP model based on the neurolysin structure. After energy minimization the root mean square deviation on all backbone atoms is only 0.1 Å, with no changes in secondary structure assignments. We then used two protein structure verification methods that characterize the environments of residues to assess the TOP model relative to the known neurolysin structure. Both the PROFILE3D (Luethy et al. 1992) and ERRAT (Colovos and Yeates 1993) programs give comparable scoring for the TOP model and neurolysin, with only 2% or 3% of the residues respectively rejected at the 95% confidence level in ERRAT, for example. Thus, using the neurolysin backbone does not result in TOP residues being placed in incompatible environments. Given the high degrees of sequence identity and similarity, the assumption that the backbone structures are nearly identical is valid for the purposes of the analysis presented here.

Acknowledgments

We thank Louis Hersh for helpful discussions. This work is funded by National Science Foundation grant MCB-9904886 and National Institutes of Health grant NS38041 to D.W.R. C.S.H. is a predoctoral fellow of the National Institute on Drug Abuse (DA14596).

The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.0216302.

References

  1. Achstetter, T., Ehmann, C., Osaki, A., and Wolf, D.H. 1984. Proteolysis in eukaryotic cells: Proteinase yscE, a new yeast peptidase. J. Biol. Chem. 259 13344–13348. [PubMed] [Google Scholar]
  2. Achstetter, T., Ehmann, C., and Wolf, D.H. 1985. Proteinase yscD: Purification and characterization of a new yeast peptidase. J. Biol. Chem. 260 4585–4590. [PubMed] [Google Scholar]
  3. Adam, G., and Delbrück, M. 1968. Reduction of dimensionality in biological diffusion processes. In Structural chemistry and molecular biology. (eds. A. Rich and N. Davidson) pp. 198–207. Freeman, San Francisco.
  4. Barelli, H., Fox-Threlkeld, J.E.T., Dive, V., Daniel, E.E., Vincent, J.P., and Checler, F. 1994. Role of endopeptidase 3.4.24.16 in the catabolism of neurotensin, in vivo, in the vascularly perfused dog ileum. Br. J. Pharmacol. 112 127–132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barrett, A.J., Brown, M.A., Dando, P.M., Knight, C.G., McKie, N., Rawlings, N.D., and Serizawa, A. 1995. Thimet oligopeptidase and oligopeptidase M or neurolysin. Meth. Enzymol. 248 529–556. [DOI] [PubMed] [Google Scholar]
  6. Brazeau, P., Vale, W., Burgus, R., Ling, N., Butcher, M., Rivier, J., and Guillemin, R. 1973. Hypothalamic polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone. Science 179 77–79. [DOI] [PubMed] [Google Scholar]
  7. Brown, C.K., Madauss, K., Lian, W., Beck, M.R., Tolbert, W.D., and Rodgers, D.W. 2001. Structure of neurolysin reveals a deep channel that limits substrate access. Proc. Natl. Acad. Sci. 98 3127–3132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Carson, M. 1987. Ribbon models of macromolecules. J. Mol. Graphics 5 103–106. [Google Scholar]
  9. Chou, K. and Zhou, G. 1982. Role of protein outside the active site on the diffusion-controlled reaction of an enzyme. J Am. Chem. Soc. 104 1409–1414. [Google Scholar]
  10. Colovos, C. and Yeates, T.O. 1993. Verification of protein structures: patterns of nonbonded atomic interactions. Protein Sci. 2 1511–1519. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Crack, P.J., Wu, T.J., Cummins, P.M., Ferro, E.S., Tullai, J.W., Glucksman, M.J., and Roberts, J.L. 1999. The association of metalloendopeptidase EC 3.4.24.15 at the extracellular surface of the AtT-20 cell plasma membrane. Brain Res. 835 113–124. [DOI] [PubMed] [Google Scholar]
  12. Dahms, P. and Mentlein, R. 1992. Purification of the main somatostatin-degrading proteases from rat and pig brains, their action on other neuropeptides, and their identification as endopeptidases 24.15 and 24.16. Eur. J. Biochem. 208 145–154. [DOI] [PubMed] [Google Scholar]
  13. Dando, P.M., Brown, M.A., and Barrett, A.J. 1993. Human thimet oligopeptidase. Biochem. J. 294 451–457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Da Silva, A., Dhuy, J., Waeldele, F., Bertrand, C., and Landry, Y. 1992. Endopeptidase 24.15 modulates bradykinin-induced contraction in guinea pig trachae. Eur. J. Pharmacol. 212 97–99. [DOI] [PubMed] [Google Scholar]
  15. Ferro, E.S., Tullai, J.W., Glucksman, M.J., and Roberts, J.L. 1999. Secretion of metalloendopeptidase 24.15 (EC 3.4.24.15). DNA Cell Biol. 18 781–789. [DOI] [PubMed] [Google Scholar]
  16. Garrido, P.A., Vandenbulcke, F., Ramjaun, A.R., Vincent, B., Checler, F., Ferro, E., and Beaudet, A. 1999. Confocal microscopy reveals thimet oligopeptidase (EC 3.4.24.15) and neurolysin (EC 3.4.24.16) in the classical secretory pathway. DNA Cell Biol. 18 323–331. [DOI] [PubMed] [Google Scholar]
  17. Jacchieri, S.G., Gomes, M.D., Juliano, L., and Carmago, A.C.M. 1998. A comparative conformational analysis of thimet oligopeptidase (EC 3.4.24.15) substrates. J. Pept. Res. 51 452–459. [DOI] [PubMed] [Google Scholar]
  18. Jones, T.A., Zou, J.Y., Cowan, S.W., and Kjeldgaard, M. 1991. Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr. A 47 110–119. [DOI] [PubMed] [Google Scholar]
  19. Kato, A., Sugiura, N., Saruta, Y., Hosoiri, T., Yasue, H., and Hirose, S. 1997. Targeting of endopeptidase 24.16 to different subcellular compartments by alternative promoter usage. J. Biol. Chem. 272 15313–15322. [DOI] [PubMed] [Google Scholar]
  20. Kraulis, P. 1991. MOLSCRIPT: A program to produce both detailed and schematic plots of protein structure. J. Appl. Crystallogr. 24 946–950. [Google Scholar]
  21. Luethy, R., Bowie, J.U., and Eisenberg, D. 1992. Assessment of protein models with three-dimensional profiles. Nature 356 83–85. [DOI] [PubMed] [Google Scholar]
  22. Matthews, B.W., Weaver, L.H., and Kester, W.R. 1974. The conformation of thermolysin. J. Biol. Chem. 249 8030–8044. [PubMed] [Google Scholar]
  23. Mentlein, R. and Dahms, P. 1994. Endopeptidases 24.16 and 24.15 are responsible for the degradation of somatostatin, neurotensin, and other neuropeptides by cultivated rat cortical astrocytes. J. Neurochem. 64 27–37. [DOI] [PubMed] [Google Scholar]
  24. Molineaux, C.J. and Ayala, J.M. 1990. An inhibitor of endopeptidase 24.15 blocks the degradation of intraventricularly administered dynorphins. J. Neurochem. 55 611–618. [DOI] [PubMed] [Google Scholar]
  25. Molineaux, C.J., Lasdun, A., Michaud, C., and Orlowski, M. 1988. Endopeptidase 24.15 is the primary enzyme that degrades luteinizing hormone releasing hormone. J. Neurochem. 51 624–633. [DOI] [PubMed] [Google Scholar]
  26. Oliveira, V., Ferro, E.S., Gomes, M.D., Oshiro, M.E., Almeida, P.C., Juliano, M.A., and Juliano, L. 2000. Characterization of thiol-, aspartyl-, and thiol-metallo-peptidase activities in Madin-Darby canine kidney cells. J. Cell Biochem. 76 478–488. [PubMed] [Google Scholar]
  27. Oliveira, V., Campos, M., Melo, R.L., Ferro, E.S., Carmago, A.C.M., Juliano, M.A., and Juliano, L. 2001. Substrate specificity characterization of recombinant metallo oligo-peptidases thimet oligopeptidase and neurolysin. Biochemistry 40 4417–4425. [DOI] [PubMed] [Google Scholar]
  28. Peitsch, M.C. 1996. PROMOD and SWISS-MODEL: Internet-based tools for automated comparative protein modeling. Biochem. Soc. Trans. 24 274–279. [DOI] [PubMed] [Google Scholar]
  29. Rawlings, N.D. and Barrett, A.J. 1995. Evolutionary families of metallopeptidases. Meth. Enzymol. 248 183–228. [DOI] [PubMed] [Google Scholar]
  30. Saric, T., Beninga, J., Graef, C.I., Akopian, T.N., Rock, K.L., and Goldberg, A.L. 2001. Major histocompatibility complex class I–presented antigenic peptides are degraded in cytosolic extracts primarily by thimet oligopeptidase. J. Biol. Chem. 276 36474–36481. [DOI] [PubMed] [Google Scholar]
  31. Serizawa, A., Dando, P.M., and Barrett, A.J. 1995. Characterization of a mitochondrial metallopeptidase reveals neurolysin as a homolog of thimet oligopeptidase. J. Biol. Chem. 270 2092–2098. [DOI] [PubMed] [Google Scholar]
  32. Shrimpton, C.N., Glucksman, M.J., Lew, R.A., Tullai, J.W., Margulies, E.H., Roberts, J.L., and Smith, A.I. 1997. Thiol activation of endopeptidase EC 3.4.24.15: A novel mechanism for the regulation of catalytic activity. J. Biol. Chem. 272 17395–17399. [DOI] [PubMed] [Google Scholar]
  33. Subramaniam, S. 1998. The Biology Workbench: A seamless database and analysis environment for the biologist. Proteins 32 1–2. [PubMed] [Google Scholar]
  34. Thompson, J.D., Higgins, D.G., and Gibson, T.J. 1994. CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22 4673–4680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Tisljar, U. 1993. Thimet oligopeptidase: A review of a thiol dependent metallo-endopeptidase also known as Pz-peptidase endopeptidase 24.15 and endo-oligopeptidase. Biol. Chem. Hoppe Seyler 374 91–100. [PubMed] [Google Scholar]
  36. Tullai, J.W., Cummins, P.M., Pabon, A., Roberts, J.L., Lopingco, M.C., Shrimpton, C.N., Smith, A.I., Martignetti, J.A., Ferro, E.S., and Glucksman, M.J. 2000. The neuropeptide processing enzyme EC 3.4.24.15 is modulated by protein kinase A phosphorylation. J. Biol. Chem. 275 36514–36522. [DOI] [PubMed] [Google Scholar]
  37. Vincent, B., Dive, V., Yiotakis, A., Smadja, C., Maldonado, R., Vincent, J.-P., and Checler, F. 1995. Phosphorous-containing peptides as mixed inhibitors of endopeptidase 3.4.24.15 and 3.4.24.16: Effect on neurotensin degradation in vitro and in vivo. Br. J. Pharmacol. 115 1053–1063. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Vincent, B., Beaudet, A., Dauch, P., Vincent, J.-P., and Checler, F. 1996. Distinct properties of neuronal and astrocytic endopeptidase 3.4.24.16: A study on differentiation, subcellular distribution, and secretion processes. J. Neuroscience 16 5049–5059. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Vincent, B., Dauch, P., Vincent, J.P., and Checler, F. 1997a. Stably transfected human cells overexpressing rat brain endopeptidase 3.4.24.16: Biochemical characterization of the activity and expression of soluble membrane-associated counterparts. J. Neurochem. 68 837–845. [DOI] [PubMed] [Google Scholar]
  40. Vincent, B., Jiracek, J., Nobel, F., Loog, M., Roques, B., Dive, V., Vincent, J.-P., and Checler, F. 1997b. Effect of a novel selective and potent phosphinic peptide inhibitor of endopeptidase 3.4.24.16 on neurotensin-induced analgesia and neuronal inactivation. Br. J. Pharmacol. 121 705–710. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Protein Science : A Publication of the Protein Society are provided here courtesy of The Protein Society

RESOURCES