ERAP-1 (endoplasmic-reticulum aminopeptidase-1) is a multifunctional enzyme with roles in the regulation of blood pressure, angiogenesis and the presentation of antigens to MHC class I molecules. Whereas the enzyme shows restricted specificity toward synthetic substrates, its substrate specificity toward natural peptides is rather broad. Because of the pathophysiological significance of ERAP-1, it is important to elucidate the molecular basis of its enzymatic action. In the present study we used site-directed mutagenesis to identify residues affecting the substrate specificity of human ERAP-1 and identified Gln181 as important for enzymatic activity and substrate specificity. Replacement of Gln181 by aspartic acid resulted in a significant change in substrate specificity, with Q181D ERAP-1 showing a preference for basic amino acids. In addition, Q181D ERAP-1 cleaved natural peptides possessing a basic amino acid at the N-terminal end more efficiently than did the wild-type enzyme, whereas its cleavage of peptides with a non-basic amino acid was significantly reduced. Another mutant enzyme, Q181E, also revealed some preference for peptides with a basic N-terminal amino acid, although it had little hydrolytic activity toward the synthetic peptides tested. Other mutant enzymes, including Q181N and Q181A ERAP-1s, revealed little enzymatic activity toward synthetic or peptide substrates. These results indicate that Gln181 is critical for the enzymatic activity and substrate specificity of ERAP-1.

INTRODUCTION

Aminopeptidases hydrolyse the N-terminal amino acid of proteins or peptide substrates. Among them, the M1 family of zinc aminopeptidases (gluzincins) shares the consensus GAMEN and HEXXH(X)18E motifs essential for the enzymatic activity. This family, which consists of 11 enzymes in humans [13], plays important roles in several pathophysiological processes, such as angiogenesis, cell-cycle regulation, reproduction, memory retention, blood-pressure control and antigen presentation to MHC class I molecules [412].

Previously we cloned a cDNA for P-LAP (placental leucine aminopeptidase)/oxytocinase, a type II membrane-spanning protein belonging to the M1 family [13]. Subsequently, we cloned cDNAs encoding ERAP-1 [endoplasmic-reticulum aminopeptidase-1; as A-LAP (adipocyte-derived leucine aminopeptidase)] and ERAP-2, which is also referred to as L-RAP (leucocyte-derived arginine aminopeptidase) from adipocyte and leucocyte cDNA libraries respectively by searching databases [14,15]. Structural and phylogenetic analyses indicated that these three enzymes are most closely related among the M1 family and are now classified into “the oxytocinase subfamily of M1 aminopeptidases” [3].

ERAP-1 is a multifunctional aminopeptidase with roles in the regulation of blood pressure, angiogenesis and the presentation of antigens to MHC class I molecules [3,4]. Moreover, this enzyme was shown to bind to cytokine receptors such as TNFR1 (tumour-necrosis-factor type I receptor), IL-6α (interleukin 6 α) receptor and IL-1 type II receptor and to promote ectodomain shedding of these receptors [3,16]. Because of its multifunctional properties, ERAP-1 is also designated PILSAP (puromycin-insensitive leucine-specific aminopeptidase) [17], ERAAP (ERAP associated with antigen presentation) [11] and ARTS-1 (aminopeptidase regulator of TNFR1 shedding) [16]. The data suggest that ERAP-1 plays important roles in several pathophysiological processes, such as immune and inflammatory responses, and is a potential therapeutic target. To design a molecule targeting ERAP-1, it is important to clarify its structural features and the molecular basis of its action.

ERAP-1 is a monomeric zinc-metallopeptidase that shows a preference for leucine when using synthetic substrates [18]. On the other hand, it reveals relatively broad substrate specificity toward natural peptide hormones and is shown to cleave several hormones such as Ang II (angiotensin II), kallidin, neurokinin A and neuromedin B [18]. Several antigenic peptides presented to MHC class I molecules are also processed both in vitro and in vivo by the enzyme [11,12]. On the basis of a preference for substrates of a specific length and with a C-terminal hydrophobic amino acid, the ‘molecular ruler’ mechanism was proposed for the processing of antigenic peptides by the enzyme [19]. However, the structural features and enzymatic mechanism of ERAP-1 remain elusive, although several amino acid residues important for the enzymatic activity of the M1 family of enzymes in general were identified by mutational analysis [20,21].

In the present study we searched for an amino acid residue affecting the enzymatic properties of human ERAP-1. We identified Gln181 as a key residue for its enzymatic action and substrate specificity. In particular, replacement of Gln181 by aspartic acid caused a drastic change in the substrate specificity of ERAP-1, which further suggested an evolutionary relationship between ERAP-1 and ERAP-2 [15].

MATERIALS AND METHODS

Molecular modelling of ERAP-1

The published X-ray-crystallographic structure of human LTA4H (leukotriene A4 hydrolase) [22] and of Thermoplasma acidophilum TIFF3 (Tricorn-interacting factor F3) [23] were used as templates for modelling the catalytic site of ERAP-1 with the SWISS-MODEL Internet server (http://www.expasy.org/swissmod/). The structure was displayed using the CueMol program (http://cuemol.sourceforge.jp).

Site-directed mutagenesis

The cDNAs encoding mutant ERAP-1s were generated by two-step PCR. PCRs were carried out in 0.2-ml-volume tubes with a 30 μl reaction volume. Sense oligonucleotide primers were designed to introduce point mutations in ERAP-1 cDNA. First PCRs were carried out for one cycle at 98 °C for 3.5 min, followed by ten cycles at 95 °C for 0.5 min, 55 °C for 1 min, and 72 °C for 2 min using Pyrobest DNA polymerase (TaKaRa, Ohtsu, Japan). Sense primer A (5′-CACCAACCCTAAAAAACCGCCACCATGGTGTTTCTGCCC-3′), containing a CACC sequence for directional cloning and initiation ATG codon, and antisense primers complementary to the desired sequences, were employed for the amplification of upstream fragments. Downstream fragments with the sequence for hexahistidine tag at the 3′-end were amplified using mutagenic sense primers and primer B (5′-GACTGTCGACTTAGTGATGGTGATGGTGATGCATACGTTCAAGCTT-3′).

The two products of the first PCR were used as templates for the second PCR. Second PCRs were carried out with primers A and B for one cycle at 98 °C for 3.5 min, followed by 25 cycles at 95 °C for 0.5 min, 55 °C for 1 min, and 72 °C for 2 min. The resultant products were inserted into the entry vector, pENTR-D-TOPO, using a TOPO-cloning system (Invitrogen). The sequences of the products were confirmed by automated sequencing on an Applied Biosystems model 377 apparatus.

Expression and purification of recombinant wild-type and mutant ERAP-1s in the baculovirus system

Human wild-type and mutant ERAP-1s with C-terminal hexahistidine tags inserted into pENTR-D-TOPO were transfected to the destination vector pDEST8 via Gateway LR reactions (Gateway LR Clonase Reaction Mix; Invitrogen). The pDEST8 vectors containing ERAP-1 cDNAs were transformed to competent DH10bac Escherichia coli cells harbouring the baculovirus genome (bacmid) and a transposition helper vector (Invitrogen). Subsequently, insect Sf 9 (Spodoptera frugiperda 9) cells were transfected with recombinant bacmids using the Cellfectin® reagent (Invitrogen). After a 3 day incubation period, recombinant baculoviruses were isolated and used to infect Sf 9 cells at a multiplicity of infection of 0.1. At 3 days after infection, the amplified viruses were harvested.

For the production of ERAP-1s, Sf 9 cells were grown at 27 °C in 100 ml of Sf-900III medium (Invitrogen) and 1.5×106 cells/ml were infected at a multiplicity of infection of 1–3. After 3 days, the culture medium was collected after centrifugation at 5000 g for 15 min.

The culture medium was loaded on to a hydroxyapatite (Nacalai Tesque, Kyoto, Japan) column (bed volume 10 ml) pre-equilibrated with 5 mM phosphate buffer, pH 7.5. After extensive washing with the same buffer, ERAP-1 was eluted from the column with 100 mM phosphate buffer, pH 7.5. The eluate was then loaded on to a Co2+-chelating Sepharose (GE Healthcare) column (bed volume 1 ml) pre-equilibrated with 10 mM phosphate buffer, pH 7.5, containing 0.1 M NaCl. After extensive washing with 10 mM phosphate buffer, pH 7.5, containing 0.1 M NaCl and 5 mM imidazole, ERAP-1 was eluted with 10 mM phosphate buffer, pH 7.5, containing 0.1 M NaCl and 100 mM imidazole. The ERAP-1-containing fractions were extensively dialysed against 25 mM Tris/HCl buffer, pH 7.5, containing 0.125 M NaCl, concentrated with an ultrafiltration membrane and stored at −20 °C prior to use.

Expression of wild-type and D198Q ERAP-2s in HEK-293 (human embryonic kidney-293) cells

The full-length version of ERAP-2 or D198Q ERAP-2 with a C-terminal hexahistidine tag was inserted into the pTargeT vector (Promega). For transfection experiments, HEK-293 cells plated in six-well plates were grown to subconfluency in Dulbecco's modified Eagle's medium containing 10% (v/v) horse serum. The cells were then mock-transfected or transfected with either pTargeT/ERAP-2 or pTargeT/D198Q ERAP-2, employing FuGENE 6 Transfection Reagent (Roche) according to the manufacturer's instructions. After 72 h, culture media were collected by centrifugation at 5000 g for 15 min. Aminopeptidase activity in the media was measured as described previously [15].

Measurement of aminopeptidase activity of ERAP-1s

Aminopeptidase activities of wild-type and mutant ERAP-1s were routinely determined by endpoint assay with various fluorogenic amino acid MCAs (amino acid 4-methylcoumaryl-7-amides) as substrates. The reaction mixture containing 100 μM amino acid MCA and 2 μg/ml enzyme in 0.5 ml of 20 mM phosphate buffer, pH 7.5, containing 10 μg/ml BSA was incubated at 37 °C for 15 min. The reaction was terminated by adding 2.5 ml of 0.1 M sodium acetate buffer, pH 4.3, containing 0.1 M sodium monochloroacetate. The amount of 7-amino-4-methylcoumarin released was measured by spectrofluorophotometry (F-2000 instrument; Hitachi) at an excitation wavelength of 360 nm and an emission wavelength of 460 nm and is given in arbitrary units.

To determine the kinetic parameters, we used the synthetic fluorogenic substrates Leu-NA (L-leucine β-naphthylamide) and Lys-NA (L-lysine β-naphthylamide). The reaction mixture containing various concentrations of substrates and either wild-type or mutant ERAP-1 in 500 μl of 20 mM phosphate buffer, pH 7.5, containing 10 μg/ml BSA, was incubated at 37 °C for 5 min. The amount of β-naphthylamine released was then measured by spectrofluorophotometry (F-2000 instrument; Hitachi) at an excitation wavelength of 335 nm and an emission wavelength of 410 nm. The kinetic parameters (Km and Vmax) were calculated using CurveExpert 1.3 (http://curveexpert.webhop.net/) to fit the data directly to the Michaelis–Menten equation.

Cleavage of peptide hormones by recombinant ERAP-1s

Peptide hormones (Peptide Institute, Osaka, Japan; 25 μM each) were incubated with either wild-type or mutant ERAP-1 (1–4 μg/ml) at 37 °C in 20 mM phosphate buffer, pH 7.5, containing 10 μg/ml BSA. The reaction was terminated by addition of 2.5% (v/v) formic acid. The peptides generated were separated by reverse-phase HPLC on a COSMOSIL (4.6 mm internal diameter×250 mm long) column (Nacalai Tesque) using a Shimadzu HPLC system (AT-10) at a flow rate of 0.5 ml/min. Peptides generated from Ang II–Ang IV were eluted isocratically with 19% (v/v) acetonitrile in 0.09% trifluoroacetic acid. Peptides generated from kallidin (Lys-bradykinin), Leu-bradykinin, HIV gag 11 and HIV pol 11 antigen precursors were eluted isocratically with either 22.5 or 17.5% acetonitrile in 0.09% trifluoroacetic acid. The molecular masses of peptides were determined by MALDI-TOF (matrix-assisted laser-desorption-ionization–time-of-flight) MS with a REFLEX mass spectrometer (Bruker-Franzen Analytik), using α-cyano-4-hydroxycinnamic acid as the matrix.

Materials

Asp-MCA (aspartic acid 4-methylcoumaryl-7-amide), Gln-MCA (glutamine 4-methylcoumaryl-7-amide), Bzl-Cys-MCA (benzylcysteine 4-methylcoumaryl-7-amide) and Glu-MCA (glutamic acid 4-methylcoumaryl-7-amide) and Leu- and Lys-NAs were purchased from Bachem AG (Bubendorf, Switzerland). Ala-, Arg-, Leu-, Lys-, Met- and Phe-MCAs and all peptide hormones were obtained from The Peptide Institute (Osaka, Japan). Amastatin was purchased from Sigma. Leu-bradykinin, HIV gag 11 and HIV pol 11 antigen precursors were synthesized by RIKEN Brain Science Institute, Saitama, Japan.

RESULTS

Modelling of the catalytic pocket of ERAP-1

Molecular modelling is a useful way in which to elucidate the structural features of ERAP-1 in the absence of its crystal structure, making it possible to identify residues critical to enzymatic action. The crystal structures for LTA4H and TIFF3 were resolved [22,23], with human LTA4H and Thermoplasma acidophilum TIFF3 showing 11.8 and 22.2% overall similarity to ERAP-1 respectively. Notably, the sequence similarity around the catalytic pockets of ERAP-1 is 25 and 40% respectively, allowing us to model the structure of the catalytic pocket of ERAP-1 using these two enzymes as templates (Figure 1A). Molecular modelling of the catalytic pocket of ERAP-1 suggests that, as with Asp221 of human APA (aminopeptidase A), a residue that is critical to the calcium-induced modulation of APA enzymatic activity [24], Gln181 is located near the catalytic HEXXH(X)18E motif and at a possible S1 site of the enzyme. Alignment of APA and human-oxytocinase-subfamily enzymes indicates that Gln181 of ERAP-1 occupies the corresponding site, Asp221, of APA, further suggesting the importance of this residue for the enzymatic activity (Figure 1B). We therefore examined the role of this residue in the enzymatic activity of ERAP-1.

Molecular modelling of the catalytic pocket of ERAP-1

Figure 1
Molecular modelling of the catalytic pocket of ERAP-1

(A) Modelling of the catalytic site of ERAP-1 using human LTA4H and Thermoplasma acidophilum TIFF3 as templates. Glu320 in the GAMEN motif, His353, Glu354, His357 and Glu376 in the HEXXH(X)18E motif and Tyr438 are shown as residues in the catalytic site of the enzyme. (B) Alignment of the ERAP-1 amino acid sequence with the sequences of APA [30,31] and other oxytocinase-subfamily enzymes P-LAP [13] and ERAP-2 [15]. Gaps are inserted into the sequences for optimal alignment. Residues conserved among the enzymes are shaded. Gln181 of ERAP-1 and the corresponding residue of P-LAP are shown as a white Q on a black background (

graphic
).

Figure 1
Molecular modelling of the catalytic pocket of ERAP-1

(A) Modelling of the catalytic site of ERAP-1 using human LTA4H and Thermoplasma acidophilum TIFF3 as templates. Glu320 in the GAMEN motif, His353, Glu354, His357 and Glu376 in the HEXXH(X)18E motif and Tyr438 are shown as residues in the catalytic site of the enzyme. (B) Alignment of the ERAP-1 amino acid sequence with the sequences of APA [30,31] and other oxytocinase-subfamily enzymes P-LAP [13] and ERAP-2 [15]. Gaps are inserted into the sequences for optimal alignment. Residues conserved among the enzymes are shaded. Gln181 of ERAP-1 and the corresponding residue of P-LAP are shown as a white Q on a black background (

graphic
).

To elucidate the significance of Gln181 to the enzymatic activity of ERAP-1, wild-type and mutant ERAP-1s carrying a single substitution of this residue (i.e. Q181E, Q181N, Q181D and Q181A ERAP-1s) were transiently expressed in a baculovirus system and purified to homogeneity by SDS/PAGE. All the enzymes showed a single band with an apparent molecular mass of ∼105 kDa under reducing conditions (results not shown).

Characterization of the hydrolytic activities of wild-type and mutant ERAP-1s towards synthetic substrates

Figure 2 shows the enzymatic activity of wild-type and mutant ERAP-1s towards various synthetic substrates. As shown previously [14], the wild-type enzyme cleaved Leu-MCA preferentially, followed by Met-MCA, indicating a restricted substrate specificity toward synthetic peptides. On the other hand, the mutant enzymes revealed little activity toward Leu- and Met-MCAs, indicating that glutamine was the most favourable residue at position 181 of ERAP-1 for the hydrolysis of these two substrates.

Enzymatic activities of wild-type and mutant ERAP-1s toward synthetic substrates

Figure 2
Enzymatic activities of wild-type and mutant ERAP-1s toward synthetic substrates

Purified enzymes (2 μg/ml) were incubated with various amino acid MCA substrates (100 μM) at 37 °C for 15 min. Relative activities of the enzymes are shown, taking the hydrolytic activity of wild-type enzyme toward Leu-MCA as 100%. Each bar shows the mean±S.D. (n=3).

Figure 2
Enzymatic activities of wild-type and mutant ERAP-1s toward synthetic substrates

Purified enzymes (2 μg/ml) were incubated with various amino acid MCA substrates (100 μM) at 37 °C for 15 min. Relative activities of the enzymes are shown, taking the hydrolytic activity of wild-type enzyme toward Leu-MCA as 100%. Each bar shows the mean±S.D. (n=3).

When the substrate specificity of Q181D ERAP-1 was examined, an apparent preference for basic amino acids (i.e. Lys- and Arg-MCAs) was observed. It should be noted here that only Q181D ERAP-1 revealed substantial activity toward basic amino acids, and that other mutant enzymes, including Q181E ERAP-1, had little activity towards any synthetic substrates tested (see below). These results indicate that Gln181 is important for the enzymatic activity as well as for the substrate specificity of ERAP-1 towards synthetic substrates.

Table 1 shows the kinetic parameters of the wild-type and mutant ERAP-1s using Leu- and Lys-NA as substrates. We could not obtain parameters by using conventional amino acid MCAs, because higher concentrations of amino acid MCAs tended to inhibit the enzymatic activity of ERAP-1. All mutant enzymes revealed reduced hydrolytic activity toward Leu-NA as compared with the wild-type enzyme. An increase in Km together with a decrease in kcat caused a significant reduction in the enzymatic activity of mutant enzymes toward Leu-NA. When Lys-NA was employed as a substrate, only Q181D revealed considerable activity. The affinity of Leu-NA for wild-type ERAP-1 was 8.4-fold that for the Q181D mutant and comparable with that of Lys-NA for the mutant, suggesting a conformational change of the substrate pocket caused by the replacement of Gln181 by aspartic acid.

Table 1
Kinetic parameters of wild-type and mutant ERAP-1s toward synthetic substrates

Kinetic parameters were determined from the Michaelis–Menten equation. Reactions took place at 37 °C for 5 min. Values are means±S.D. (n=3). Relative kcat/Km values are shown taking Leu-NA hydrolytic activity of the wild-type enzyme as 100%.

   kcat/Km 
Substrate/enzyme Km (μM) kcat (s−1(mM−1·s−1(%) 
Leu-NA     
 Wild-type 250±41 19±1.6 75±5.6 100 
 Q181E 1800±580 6.8±1.6 3.8±0.27 5.1 
 Q181D 2100±790 1.9±0.65 0.93±0.042 1.2 
 Q181N 2400±140 6.2±1.5 2.6±0.59 3.5 
 Q181A 3100±780 0.52±0.044 0.18±0.062 0.20 
Lys-NA     
 Wild-type 650±98 3.6±0.22 5.6±0.68 7.5 
 Q181E 320±50 1.9±0.090 5.9±0.72 7.9 
 Q181D 320±42 8.2±0.26 26±2.5 35 
 Q181N 870±93 0.64±0.046 0.74±0.030 1.0 
 Q181A 1100±480 0.10±0.023 0.097±0.023 0.13 
   kcat/Km 
Substrate/enzyme Km (μM) kcat (s−1(mM−1·s−1(%) 
Leu-NA     
 Wild-type 250±41 19±1.6 75±5.6 100 
 Q181E 1800±580 6.8±1.6 3.8±0.27 5.1 
 Q181D 2100±790 1.9±0.65 0.93±0.042 1.2 
 Q181N 2400±140 6.2±1.5 2.6±0.59 3.5 
 Q181A 3100±780 0.52±0.044 0.18±0.062 0.20 
Lys-NA     
 Wild-type 650±98 3.6±0.22 5.6±0.68 7.5 
 Q181E 320±50 1.9±0.090 5.9±0.72 7.9 
 Q181D 320±42 8.2±0.26 26±2.5 35 
 Q181N 870±93 0.64±0.046 0.74±0.030 1.0 
 Q181A 1100±480 0.10±0.023 0.097±0.023 0.13 

Characterization of the hydrolytic activities of wild-type and mutant ERAP-1s toward peptide substrates

We next compared the hydrolytic activities of the wild-type and mutant ERAP-1s toward several natural peptides. Figure 3 shows the cleavage of Ang II–Ang IV by the wild-type, Q181D and Q181E ERAP-1s. The wild-type enzyme cleaved the N-terminal amino acid of both Ang II and Ang IV sequentially, and degradation products were clearly detected within 30 min (Figures 3A and 3C). By contrast, the Q181D and Q181E ERAP-1s showed little activity toward these two hormones. Other mutant enzymes (i.e. Q181N and Q181A) also had little activity (results not shown), suggesting that Gln181 of ERAP-1 was required for the maximal enzymatic activity toward these two hormones.

Cleavage of Ang II, Ang III and Ang IV by wild-type, Q181D and Q181E ERAP-1s

Figure 3
Cleavage of Ang II, Ang III and Ang IV by wild-type, Q181D and Q181E ERAP-1s

Ang II (25 μM) (A), Ang III (25 μM) (B) or Ang IV (25 μM) (C) was incubated with purified enzymes (4 μg/ml) at 37 °C for 60 min. The peptides generated were loaded on to an HPLC column and separated, and their amino acid sequences were determined. Abbreviation: mAU, milli-absorbance unit.

Figure 3
Cleavage of Ang II, Ang III and Ang IV by wild-type, Q181D and Q181E ERAP-1s

Ang II (25 μM) (A), Ang III (25 μM) (B) or Ang IV (25 μM) (C) was incubated with purified enzymes (4 μg/ml) at 37 °C for 60 min. The peptides generated were loaded on to an HPLC column and separated, and their amino acid sequences were determined. Abbreviation: mAU, milli-absorbance unit.

When Ang III was employed as a substrate (Figure 3B), sequential release of the N-terminal amino acid from Ang III to des-[Val-Tyr]-Ang IV by the wild-type enzyme was observed. The conversion of Ang III into Ang IV by Q181D ERAP-1 was comparable with that by the wild-type enzyme, and consistent with its preference for basic amino acids. Further degradation to des-[Val]-Ang IV was barely detectable. In addition, Q181E ERAP-1 also revealed substantial activity to convert Ang III into Ang IV. This result indicates that, although Q181E ERAP-1 showed little hydrolytic activity toward synthetic substrates, it retained enzymatic activity toward peptide substrates with an N-terminal basic amino acid.

To characterize further the N-terminal preference of the wild-type, Q181D and Q181E ERAP-1s, we compared the rate of cleavage of the N-terminal residue from kallidin (Lys-bradykinin) and Leu-bradykinin. Because of proline at the second last N-terminal position, we could examine the role of the N-terminal amino acid in the hydrolytic activity of the enzyme directly. As shown in Figures 4(A) and 4(B), whereas the wild-type enzyme cleaved Leu-bradykinin much more quickly than did the Q181D and Q181E ERAP-1s, its cleavage rate for kallidin was rather low when compared with that of the mutant enzymes. The wild-type enzyme converted 85% of Leu-bradykinin into bradykinin within 5 min in our assay system, but only 33% of kallidin to bradykinin, again supporting its preference for non-basic amino acids. By contrast, the Q181D and Q181E mutants cleaved kallidin (65 and 45% conversion respectively) more quickly than Leu-bradykinin (40 and 35% conversion respectively). Therefore, although the data do not precisely reflect the substrate specificity of the enzymes toward synthetic substrates, it is still possible to conclude that, although the wild-type enzyme ‘prefers’ non-basic amino acids, the Q181D and Q181E ERAP-1s show a preference for basic amino acids. In addition, the results also suggest that dependent only on the N-terminal amino acid of the substrates, the amino acid at position 181 mediates the hydrolytic efficiency, and thus determines the substrate specificity, of ERAP-1.

Cleavage of various substrate peptides by wild-type, Q181D and Q181E ERAP-1s

Figure 4
Cleavage of various substrate peptides by wild-type, Q181D and Q181E ERAP-1s

Leu-bradykinin (25 μM) (A), kallidin (Lys-bradykinin) (25 μM) (B) or the precursor of HIV gag antigen (25 μM) (C) was incubated with purified ERAP-1s (2 μg/ml) at 37 °C for 5 min. The precursor of HIV pol antigen (25 μM) (D) was incubated with purified enzymes (2 μg/ml) at 37 °C for 60 min. Peptides in the reaction mixture were then separated by HPLC and their amino acid sequences were determined. Abbreviation: mAU, milli-absorbance unit.

Figure 4
Cleavage of various substrate peptides by wild-type, Q181D and Q181E ERAP-1s

Leu-bradykinin (25 μM) (A), kallidin (Lys-bradykinin) (25 μM) (B) or the precursor of HIV gag antigen (25 μM) (C) was incubated with purified ERAP-1s (2 μg/ml) at 37 °C for 5 min. The precursor of HIV pol antigen (25 μM) (D) was incubated with purified enzymes (2 μg/ml) at 37 °C for 60 min. Peptides in the reaction mixture were then separated by HPLC and their amino acid sequences were determined. Abbreviation: mAU, milli-absorbance unit.

It is well established that ERAP-1 is the final processing enzyme of antigenic peptides presented to MHC class I molecules in the ER [3,11,12]. Therefore, we next examined the cleavage of N-extended precursors of HIV gag 11 and HIV pol 11 antigens to compare further the hydrolytic activity and substrate specificity of the wild-type and mutant ERAP-1s. As shown in Figure 4(C), whereas ERAP-1 removed 49% of the N-terminal leucine from the HIV gag antigen precursor within 5 min, only 7 and 11% was released by Q181D and Q181E ERAP-1s respectively, indicating that, as in the case of synthetic substrates, the wild-type enzyme cleaved N-terminal leucine much more efficiently than did the mutant enzymes. By contrast, when the HIV pol antigen precursor, which has arginine at its N-terminal end, was employed, both mutants cleaved the arginine residue more efficiently than did the wild-type enzyme (Figure 4D). These results again indicate that the Q181D and Q181E ERAP-1s cleave N-terminal basic amino acid residues more efficiently than the wild-type enzyme and that replacement of Gln181 by aspartic acid or glutamic acid causes a change of substrate specificity to a preference for basic amino acids. Taken together, these results suggest that Gln181 of ERAP-1 is crucial in the processing of peptide hormones and antigenic peptides by determining preferential substrates of the enzyme and thus mediating various pathophysiological functions of the enzyme.

Effects of aminopeptidase inhibitors on the enzymatic activities of wild-type and Q181D ERAP-1s

We then compared the effects of Zn2+ and amastatin, both well-known inhibitors of M1 aminopeptidases, on the enzymatic activity of the wild-type and Q181D ERAP-1s. As shown in Figure 5(A), Zn2+ inhibited the enzymatic activities of both enzymes equally, suggesting that Zn2+ interfered with hydrolytic processes shared by the M1 family members. On the other hand, amastatin was about 50-fold less effective towards Q181D ERAP-1 than towards the wild-type enzyme (Figure 5B). The Ki values of amastatin for the wild-type and Q181D ERAP-1s were 41.8±10.3 and 168±46 μM respectively. Considering that amastatin [(2S,3R)-3-amino-2-hydroxy-5-methylhexanoyl-Val-Val-Asp] is a small peptide having an amino-acid-like residue at its N-terminus, these results are consistent with the notion that wild-type ERAP-1 binds to non-basic amino acids with higher affinity than does the Q181D mutant enzyme.

Effects of Zn2+ and amastatin on the enzymatic activity of wild-type and Q181D ERAP-1s

Figure 5
Effects of Zn2+ and amastatin on the enzymatic activity of wild-type and Q181D ERAP-1s

Purified wild-type ERAP-1 and Q181D ERAP-1 (2 μg/ml) were incubated with various concentrations of ZnCl2 (A) or amastatin (B) on ice for 5 min. Aminopeptidase activities were then measured by using Leu-MCA (100 μM) for the wild-type and Lys-MCA (100 μM) for Q181D ERAP-1.

Figure 5
Effects of Zn2+ and amastatin on the enzymatic activity of wild-type and Q181D ERAP-1s

Purified wild-type ERAP-1 and Q181D ERAP-1 (2 μg/ml) were incubated with various concentrations of ZnCl2 (A) or amastatin (B) on ice for 5 min. Aminopeptidase activities were then measured by using Leu-MCA (100 μM) for the wild-type and Lys-MCA (100 μM) for Q181D ERAP-1.

Comparison of substrate specificity between the wild-type and D198Q ERAP-2s

As shown in Figure 1(B), Asp198 of ERAP-2 corresponds to Gln181 of ERAP-1. To obtain further insight into the role of these residues in the enzymatic properties, we compared the substrate specificity of the wild-type and D198Q ERAP-2s toward synthetic substrates. The enzymes were transiently overexpressed in HEK-293 cells and secreted into culture medium. Western-blot analysis revealed that the secreted enzymes had molecular masses of ∼120 kDa (Figure 6A). As shown in Figure 6(B), whereas wild-type ERAP-2 preferentially cleaved basic amino acids, as reported previously [15], D198Q ERAP-2 cleaved Leu-MCA most efficiently, followed by Met-MCA. In addition, the mutant enzyme still retained a substantial activity toward (S)-Bzl-Cys-, Arg-, and Lys-MCAs. These results indicate that Asp198 of ERAP-2 also affects the enzymatic properties of ERAP-2 and that the substrate specificity of D198Q ERAP-2 towards synthetic substrates is broad and similar to that of P-LAP rather than ERAP-1 [14,25].

Substrate specificity of wild-type and D198Q ERAP-2s towards synthetic peptides

Figure 6
Substrate specificity of wild-type and D198Q ERAP-2s towards synthetic peptides

HEK-293 cells transiently mock-transfected and transfected with either wild-type or mutant ERAP-2 were incubated at 37 °C for 72 h. The culture medium was then collected and the cells were incubated further with various amino acid MCA substrates (100 μM) to assess aminopeptidase activity. (A) Western-blot analysis of wild-type and D198Q ERAP-2s expressed in HEK-293 cells. Enzymes secreted into medium were detected by using rabbit polyclonal anti-ERAP-2 antibody. (B) Protein levels of ERAP-2s for the aminopeptidase assay was adjusted by Western-blot analysis. Differences in activity between either wild-type or mutant ERAP-2 and the mock-transfected HEK-293 cell culture medium are shown as relative values, taking the wild-type enzyme activity toward Arg-MCA as 100%. Each bar shows the mean±S.D. (n=3).

Figure 6
Substrate specificity of wild-type and D198Q ERAP-2s towards synthetic peptides

HEK-293 cells transiently mock-transfected and transfected with either wild-type or mutant ERAP-2 were incubated at 37 °C for 72 h. The culture medium was then collected and the cells were incubated further with various amino acid MCA substrates (100 μM) to assess aminopeptidase activity. (A) Western-blot analysis of wild-type and D198Q ERAP-2s expressed in HEK-293 cells. Enzymes secreted into medium were detected by using rabbit polyclonal anti-ERAP-2 antibody. (B) Protein levels of ERAP-2s for the aminopeptidase assay was adjusted by Western-blot analysis. Differences in activity between either wild-type or mutant ERAP-2 and the mock-transfected HEK-293 cell culture medium are shown as relative values, taking the wild-type enzyme activity toward Arg-MCA as 100%. Each bar shows the mean±S.D. (n=3).

DISCUSSION

Because of their pathophysiological significance, it is important to elucidate the structural and enzymatic features of the M1 family of aminopeptidases, including ERAP-1. In the present study we identified Gln181 as a residue critical to the enzymatic activity and substrate specificity of ERAP-1. Although Gln181 of ERAP-1 was required for maximal enzymatic activity toward peptide substrates having an N-terminal non-basic amino acid, replacement of this residue with aspartic acid increased the enzyme's preference for basic amino acids. This replacement might cause rather restricted substrate specificity through electrostatic interaction. Cleavage of peptides with an N-terminal basic amino acid by Q181E ERAP-1 also supported a role for electrostatic interaction between Glu181 and the N-terminal basic amino acid of the substrates. Because of the length of the side chain carrying a negative charge at position 181, Q181E ERAP-1 might be less active towards the N-terminal basic amino acid of the substrates than Q181D ERAP-1. Thus it is plausible that the length, as well as uncharged nature, of the side chain of Gln181 is critical for the enzymatic properties of ERAP-1. Considering that ERAP-1 is the final processing enzyme of antigenic peptides presented to MHC class I molecules, we speculate that Gln181 has been selected during evolution to maintain a rather broad specificity of the enzyme toward peptide substrates.

Figure 7 shows a possible mechanism for the enzymatic action of ERAP-1. It is apparent that Gln181 is involved in the cleavage of the N-terminal amino acid of the substrates. However, because it is unlikely that Gln181 interacts with this amino acid (such as leucine) directly, we speculate that it is involved in the maintenance of the catalytic pocket's structure by bridging with another unidentified residue and thus contributes to the formation of the S1 site. Replacement of Gln181 by the acidic amino acids aspartic acid or glutamic acid might cause a local conformational change of the substrate pocket, resulting in a rather restricted substrate specificity through electrostatic interaction between the residue and N-terminal basic amino acid of the substrates. To elucidate the molecular basis of the role of this residue in the enzymatic activity of ERAP-1 definitively, it is essential to determine the crystal structure of the enzyme.

Schematic representation of the catalytic pockets of wild-type and Q181D ERAP-1s

Figure 7
Schematic representation of the catalytic pockets of wild-type and Q181D ERAP-1s

Position 181 of either the wild-type or mutant enzyme presumably acts as the S1 site. Gln181 of the wild-type enzyme contributes to the formation of the catalytic pocket of the enzyme by bridging with another residue, which causes the broad substrate specificity. Replacement of this residue by aspartic acid may induce a conformational change in the pocket. Through electrostatic interaction, the mutant enzyme shows a preference for a basic N-terminal amino acid.

Figure 7
Schematic representation of the catalytic pockets of wild-type and Q181D ERAP-1s

Position 181 of either the wild-type or mutant enzyme presumably acts as the S1 site. Gln181 of the wild-type enzyme contributes to the formation of the catalytic pocket of the enzyme by bridging with another residue, which causes the broad substrate specificity. Replacement of this residue by aspartic acid may induce a conformational change in the pocket. Through electrostatic interaction, the mutant enzyme shows a preference for a basic N-terminal amino acid.

Phylogenetic analysis suggests an evolutionary relationship and the recent divergence of ERAP-1 and ERAP-2 from P-LAP [3,15]. In fact, the human genes for the oxytocinase subfamily are located contiguously around chromosome 5q15, and the gene structure of each member is similar, with exon–intron junctions well conserved [15,2628].

One of the most characteristic features of ERAP-2 is a preference for basic amino acids [15], implying that replacement of Gln181 by aspartic acid in ERAP-1 resulted in a substrate specificity similar to that of ERAP-2. It is possible that ERAP-2 acquired its preference for basic amino acids during evolution and thus expanded the repertoire of antigenic peptides processed in the ER. Complementary and/or co-operative actions of two enzymes were reported in the processing of antigenic peptides presented to MHC class I molecules in the ER [29]. In this context, the substrate specificity of D198Q ERAP-2 is rather broad and similar to that of P-LAP. Since glutamine occupies the corresponding site of both P-LAP and ERAP-1, which cleave Leu-MCA most efficiently of the synthetic substrates tested, it is most plausible that glutamine at this site is required for the efficient cleavage of the substrate. However, considering the difference in substrate specificity between P-LAP and ERAP-1 [14,25], other residue(s) may be involved. Taken together, our data suggest that Gln181 of ERAP-1 and the corresponding sites of ERAP-2 and P-LAP are crucial to the enzymatic activity and substrate specificity, and further confirm the evolutionarily close relationship between the members of the oxytocinase subfamily.

In the present study we have shown that Gln181 is important for the enzymatic activity and substrate specificity of ERAP-1. Both Q181D ERAP-1 and ERAP-2 with an aspartic acid residue at this site show a preference for basic amino acids, which might reflect the evolutionary relationship between ERAP-1 and ERAP-2. Our data should be useful in clarifying the enzymatic mechanism of ERAP-1 and ERAP-2 and in developing therapeutic reagents targeting these enzymes.

This work was supported in part by grants-in-aid from the Ministry of Education, Science, Sports and Culture of Japan and by a Chemical Biology Research Program grant from RIKEN.

Abbreviations

     
  • A-LAP

    adipocyte-derived leucine aminopeptidase

  •  
  • Ang

    angiotensin

  •  
  • APA

    aminopeptidase A

  •  
  • ER

    endoplasmic reticulum

  •  
  • ERAP

    ER aminopeptidase

  •  
  • HEK-293

    human embryonic kidney-293

  •  
  • IL

    interleukin

  •  
  • LTA4H

    leukotriene A4 hydrolase

  •  
  • Leu-NA

    L-leucine β-naphthylamide

  •  
  • Lys-NA

    L-lysine β-naphthylamide

  •  
  • L-RAP

    leucocyte-derived arginine aminopeptidase

  •  
  • MCA

    4-methylcoumaryl-7-amide

  •  
  • P-LAP

    placental leucine aminopeptidase

  •  
  • TNFR1

    tumour-necrosis-factor type I receptor

References

References
1
Taylor
A.
Aminopeptidases: structure and function
FASEB J.
1993
, vol. 
7
 (pg. 
290
-
298
)
2
Hooper
N. M.
Families of zinc metalloproteases
FEBS Lett.
1994
, vol. 
354
 (pg. 
1
-
6
)
3
Tsujimoto
M.
Hattori
A.
The oxytocinase subfamily of M1 aminopeptidases
Biochim. Biophys. Acta
2005
, vol. 
1751
 (pg. 
9
-
18
)
4
Sato
Y.
Role of aminopeptidase in angiogenesis
Biol. Pharm. Bull.
2004
, vol. 
27
 (pg. 
772
-
776
)
5
Constam
D. B.
Tobler
A. R.
Rensing-Ehl
A.
Kemler
I.
Hersh
L. B.
Fontana
A.
Puromycin-sensitive aminopeptidase. Sequence analysis, expression and functional characterization
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
26931
-
26939
)
6
Osada
T.
Watanabe
G.
Kondo
S.
Toyoda
M.
Sakaki
Y.
Takeuchi
T.
Male reproductive defects caused by puromycin-sensitive aminopeptidase deficiency in mice
Mol. Endocrinol.
2001
, vol. 
15
 (pg. 
960
-
971
)
7
Albiston
A. L.
McDowall
S. G.
Matsacos
D.
Sim
P.
Clune
E.
Mustafa
T.
Lee
J.
Mendelsohn
F. A. O.
Simpson
R. J.
Connolly
L. M.
Chai
S. Y.
Evidence that the angiotensin IV (AT4) receptor is the enzyme insulin-regulated aminopeptidase
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
48623
-
48626
)
8
Goto
Y.
Hattori
A.
Ishii
Y.
Tsujimoto
M.
Reduced activity of the hypertension-associated Lys528Arg mutant of human adipocyte-derived leucine aminopeptidase (A-LAP)/ER-aminopeptidase-1
FEBS Lett.
2006
, vol. 
580
 (pg. 
1833
-
1838
)
9
Wright
J. W.
Mizutani
S.
Murray
C. E.
Amir
H. Z.
Harding
J. W.
Aminopeptidase-induced elevations and reductions in blood pressure in the spontaneously hypertensive rat
J. Hypertens.
1990
, vol. 
8
 (pg. 
969
-
974
)
10
Reaux
A.
Fournie-Zaluski
M. C.
David
C.
Zini
S.
Roques
B.
Corvol
P.
Llorens-Cortes
C.
Aminopeptidase A inhibitors as potential central antihypertensive agents
Proc. Natl. Acad. Sci. U.S.A.
1999
, vol. 
96
 (pg. 
13415
-
13420
)
11
Serwold
T.
Gonzalez
F.
Kim
J.
Jacob
R.
Shastri
N.
ERAAP customizes peptides for MHC class I molecules in the endoplasmic reticulum
Nature
2002
, vol. 
419
 (pg. 
480
-
483
)
12
Saric
T.
Chang
S-C.
Hattori
A.
York
I. A.
Markant
S.
Rock
K. L.
Tsujimoto
M.
Goldberg
A. L.
An IFN-γ-induced aminopeptidase in the ER, ERAPI, trims precursors to MHC class I-presented peptides
Nat. Immunol.
2002
, vol. 
3
 (pg. 
1169
-
1176
)
13
Rogi
T.
Tsujimoto
M.
Nakazato
H.
Mizutani
S.
Tomoda
Y.
Human placental leucine aminopeptidase: a new member of type II membrane-spanning zinc metallopeptidase family
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
56
-
61
)
14
Hattori
A.
Matsumoto
H.
Mizutani
S.
Tsujimoto
M.
Molecular cloning of adipocyte-derived leucine aminopeptidase highly related to placental leucine aminopeptidase/oxytocinase
J. Biochem. (Tokyo)
1999
, vol. 
125
 (pg. 
931
-
938
)
15
Tanioka
T.
Hattori
A.
Masuda
S.
Nomura
Y.
Nakayama
H.
Mizutani
S.
Tsujimoto
M.
Human leukocyte-derived arginine aminopeptidase: the third member of the oxytocinase subfamily of aminopeptidases
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
32275
-
32283
)
16
Cui
X.
Hawari
F.
Alsaary
S.
Lawrence
M.
Combs
C. A.
Geng
W.
Rouhani
F. N.
Miskinis
D.
Levine
S. J.
Identification of ARTS-1 as a novel TNFR1-binding protein that promotes TNFR1 ectodomain shedding
J. Clin. Invest.
2002
, vol. 
11
 (pg. 
515
-
526
)
17
Schomburg
L.
Kollmus
H.
Friedrichsen
S.
Bauer
K.
Molecular characterization of a puromycin-insensitive leucyl-specific aminopeptidase, PILS-AP
Eur. J. Biochem.
2000
, vol. 
267
 (pg. 
3198
-
3207
)
18
Hattori
A.
Kitatani
K.
Matsumoto
H.
Miyazawa
S.
Rogi
T.
Tsuruoka
N.
Mizutani
S.
Natori
Y.
Tsujimoto
M.
Characterization of recombinant human adipocyte-derived leucine aminopeptidase expressed in Chinese hamster ovary cells
J. Biochem.
2000
, vol. 
128
 (pg. 
755
-
762
(Tokyo)
19
Chang
S. C.
Momburg
F.
Bhutani
N.
Goldberg
A. L.
The ER aminopeptidase, ERAP1, trims precursors to lengths of MHC class I peptides by a “molecular ruler” mechanism
Proc. Natl. Acad. Sci. U.S.A.
2005
, vol. 
102
 (pg. 
17107
-
17112
)
20
Reaux
A.
Iturrioz
X.
Vazeux
G.
Fournie-Zaluski
M. C.
David
C.
Roques
B. P.
Corvol
P.
Llorens-Cortes
C.
Aminopeptidase A, which generates one of the main effector peptides of the brain renin-angiotensin system, angiotensin III, has a key role in central control of arterial blood pressure
Biochem. Soc. Trans.
2000
, vol. 
28
 (pg. 
435
-
440
)
21
Tsujimoto
M.
Goto
Y.
Maruyama
M.
Hattori
A.
Biochemical and enzymatic properties of the M1 family of aminopeptidases involved in the regulation of blood pressure
Heart Fail. Rev.
2008
, vol. 
13
 (pg. 
285
-
291
)
22
Thunnissen
M. M. G. M.
Nordlund
P.
Haeggstrom
J. Z.
Crystal structure of human leukotriene A4 hydrolase, a bifunctional enzyme in inflammation
Nat. Struct. Biol.
2001
, vol. 
8
 (pg. 
131
-
135
)
23
Kyrieleis
O. J. P.
Goettig
P.
Kiefersauer
R.
Huber
R.
Brandstetter
H.
Crystal structure of the tricorn interacting factor F3 from Thermoplasma acidophilum, a zinc aminopeptidase in three conformations
J. Mol. Biol.
2005
, vol. 
349
 (pg. 
787
-
800
)
24
Goto
Y.
Hattori
A.
Mizutani
S.
Tsujimoto
M.
Aspartic acid 221 is critical in the calcium-induced modulation of the enzymatic activity of human aminopeptidase A
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
37074
-
37081
)
25
Matsumoto
H.
Rogi
T.
Yamashiro
K.
Kodama
S.
Tsuruoka
N.
Hattori
A.
Takio
K.
Mizutani
S.
Tsujimoto
M.
Characterization of a recombinant soluble form of human placental leucine aminopeptidase/oxytocinase expressed in Chinese hamster ovary cells
Eur. J. Biochem.
2000
, vol. 
267
 (pg. 
46
-
52
)
26
Rasmussen
T. E.
Pedraza-Diaz
S.
Hardre
R.
Lausten
P. G.
Carrion
A. G.
Kristensen
T.
Structure of the human oxytocinase/insulin-regulated aminopeptidase gene and localization to chromosome 5q21
Eur. J. Biochem.
2000
, vol. 
267
 (pg. 
2297
-
2306
)
27
Tanioka
T.
Hattori
A.
Mizutani
S.
Tsujimoto
M.
Regulation of the human leukocyte-derived arginine aminopeptidase/endoplasmic reticulum-aminopeptidase 2 gene by interferon-γ
FEBS J.
2003
, vol. 
272
 (pg. 
916
-
928
)
28
Hattori
A.
Matsumoto
K.
Mizutani
S.
Tsujimoto
M.
Genomic organization of the human adipocyte-derived leucine aminopeptidase gene and its relationship to the placental leucine aminopeptidase/oxytocinase gene
J. Biochem. (Tokyo)
2001
, vol. 
130
 (pg. 
235
-
241
)
29
Saveanu
L.
Carroll
O.
Lindo
V.
Del. Val
M.
Lopez
D.
Lepellertier
Y.
Greer
F.
Schomburg
L.
Fruci
D.
Niedermann
G.
van Endert
P. M.
Concerted peptide trimming by human ERAP1 and ERAP2 aminopeptidase complexes in the endoplasmic reticulum
Nat. Immunol.
2005
, vol. 
7
 (pg. 
689
-
697
)
30
Migaud
M.
Durieux
C.
Viereck
J.
Soroca-Lucas
E.
Fournie-Zaluski
M. C.
Roques
B. P.
The in vivo metabolism of cholecystokinin (CCK-8) is essentially ensured by aminopeptidase A
Peptides
1996
, vol. 
17
 (pg. 
601
-
607
)
31
Li
L.
Wang
J.
Cooper
M. D.
cDNA cloning and expression of human glutamyl aminopeptidase (aminopeptidase A)
Genomics
1993
, vol. 
17
 (pg. 
657
-
664
)