Aspartyl aminopeptidase proteases from H. sapiens (Q9ULAO), S. cerevisiae (P38821), Plasmodium falciparum (PFI1570c), Plasmodium chabaudi chabaudi (PC000238.00.0), Plasmodium yoelii yoelii (PY03205), Plasmodium knowlesi (PKH_073050), and Plasmodium vivax (Pv087090) were retrieved from the NCBI and PlasmoDB (version 5.3) database. The sequences were aligned using ClustalW at EMBL-EBI http://www.ebi.ac.uk/Tools/clustalw/index.html.

Plasmodium falciparum was cultured in vitro by the method of Trager and Jensen 21 and genomic DNA extracted by first lysing the erythrocytes and subsequently boiling the parasites to release the DNA 22. The PfM18AAP sequence (nucleotides 7–1689 of gene PFI1570c taken from the PlasmoDB (version 5.3) database) was amplified from the parasite DNA by PCR, using sequence specific primers containing 5' NdeI and 3'BamHI recognition sequences (restriction enzyme recognition sequence is in bold; F: 5' CTGAGGAACTGCATATGAAGAAAGCTAGGGAATACGCC 3'; R: 5' TGTGCTGGATCCTCATAAGACTTGGTTGATGTAGG 3'). The amplified DNA was inserted into the pET15-b vector (Novagen, USA), the insert sequenced and BL21-CodonPlus^® (DE3) RIL competent cells (Stratagene, USA) were transformed with the vector construct. rPfM18AAP, containing an N-terminal hexahistidine-tag, was produced in Overnight Express™ Instant TB Medium (Novagen, USA) and purified from E. coli extracts using HIS-Select™ Magnetic Agarose Beads (Sigma, USA). The rPfM18AAP concentration was determined spectrophotometrically at 595 nm using the Coomassie Plus^® Protein Assay Reagent Kit (Pierce, USA) and the purified protein sample analyzed by sodium dodecylsulphate 10% polyacrylamide gel electrophoresis (SDS-PAGE) 23.

The hexahistidine-tag was removed from rPfM18AAP with the THROMBIN CleanCleave™ KIT (Sigma, USA) and the isoelectric point (pI) of rPfM18AAP determined by two-dimensional gel electrophoresis 24 against 2-D SDS-PAGE Standards (Bio-Rad, USA). The first dimension isoelectric focusing (IEF) gel contained the Bio-Lyte^® 3/10 Ampholyte (Bio-Rad, USA) and the second dimension Laemmli SDS-polyacrylamide gel 23 contained 13% acrylamide. The proteins were visualized by silver staining.

The number of subunits of native rPfM18AAP was determined by Ferguson plots 25. The protein was electrophoresed through non-denaturing agarose-acrylamide tube gels containing 3, 4, 5, 6, 7, 8, 9, and 10% acrylamide and 0.3% agarose 26. The approximate molecular weight of each rPfM18AAP subunit was determined from a standard curve constructed from the relative mobility of bovine serum albumin (BSA) (monomer = 66 kDa; dimer = 132 kDa) and human erythrocyte spectrin (dimer = 460 kDa; tetramer = 920 kDa; hexamer = 1380 kDa). Tube gels were cut in half to facilitate transfer onto Hybond™-C Extra Nitrocellulose membrane (Amersham, UK) and the rPfM18AAP detected with 1:2000 Penta·His™ HRP Conjugate antibody (Qiagen, Germany) using the SuperSignal^® West Pico Chemiluminescent Substrate (Pierce, USA).

A coupled enzyme assay using 0.3 mM Asp-Ala-Pro-beta-Napthylamide (Peptides International, USA) and 0.0025 U dipeptidyl peptidase IV (Sigma, USA) in 50 mM Tris-HCl (pH 7.5) was used to study the enzymatic activity of 2.5 μg rPfM18AAP in a final volume of 50 μl 27. To determine the optimum pH, the enzyme assay was performed with 50 mM Tris-HCl buffer at pH 6.8, 7.5, 8.0, 8.5 and 9.0, or 0.1 M sodium citrate buffer at pH 5.3, 6.0 and 6.5. A temperature study was completed in 50 mM Tris-HCl buffer (pH 7.5) at 25, 30, 33, 37 and 39°C.

Blot overlay assays were used to confirm the interaction between rPfM18AAP and erythrocyte spectrin, as well as to study the protein-protein interactions between rPfM18AAP and other erythrocyte membrane proteins.

Approximately 1 μg spectrin tetramers and dimers 28, 100 ng rPfM18AAP (positive control) and 100 ng BSA (negative control) were applied to a Hybond™-C Extra Nitrocellulose membrane in a Bio-Rad Bio-Dot^® SF chamber (Bio-Rad, USA). To test the interaction of rPfM18AAP with other erythrocyte membrane proteins, 100 ng rPfM18AAP (positive control), 100 ng BSA (negative control) and 20 μg erythrocyte membrane proteins 28 were electrophoresed through a Laemmli SDS 9% polyacrylamide gel 23 or a Fairbanks SDS 3.5–17.5% polyacrylamide gel 2930 and transferred onto a Hybond™-C Extra Nitrocellulose membrane. Membranes were blocked with 5% BSA in TBS (10 mM Tris-HCl, 150 mM NaCl, pH 7.5) and subsequently overlaid for 1 hr at room temperature with 1.25–2.5 μg rPfM18AAP in either 50 mM Tris-HCl (pH 7.5), or 50 mM Tris-HCl (pH 7.5) containing 150 mM NaCl. Membranes were washed four times for 10 minutes in wash buffer (20 mM Tris-HCl, 500 mM NaCl, 0.12% Tween^®-20 (v/v), 0.2% Triton^® X-100 (v/v), pH 7.5) and once for 10 minutes in TBS. The membranes were fixed for 20 minutes with 0.5% (v/v) formaldehyde followed by a 10 minute wash with TBS. The presence of rPfM18AAP was detected with 1:2000 Penta·His™ HRP Conjugate antibody using the SuperSignal^® West Pico Chemiluminescent Substrate.

PfM18AAP is classified as an M18 aminopeptidase in the MEROPS database 31. The 33 amino acid spectrin-binding region of PfM18AAP, which was initially identified by phage display 20, was not found in the H. sapiens or S. cerevisiae sequences (Additional File 1). This region may represent a special evolutionary adaptation of P. falciparum, since it is specific for PfM18AAP and is not found in other Plasmodium species. Despite this species-specific spectrin-binding insert, there is a 64–74% sequence identity between PfM18AAP and the Plasmodium homologues. The sequence probably forms a loop on the surface of the active enzyme 32 and could, therefore, allow PfM18AAP to bind to spectrin, without interfering with the enzymatic function.

PfM18AAP is a metalloprotease and requires cofactors, probably cobalt 32, for enzymatic activity. These cofactors are bound by specific amino acids that are present in all the M18 aminopeptidases (Additional File 1). Conserved amino acids are labeled according to the amino acid numbers of the human aspartyl aminopeptidase 33 and the amino acid numbers of PfM18AAP are given in brackets. The five amino acids, H94, D264, E302, D346 and H440 (H86, D324, E380, D434 and H534) involved in cofactor binding 31, as well as the two amino acids, D96 and E301 (D88 and E379) involved in substrate cleavage 31 are conserved in PfM18AAP. Two additional histidines, H170 (H160), involved in enzymatic activity 33 and H352 (H440), involved in quaternary structure stabilization 33, are also present in all the sequences (Additional File 1).

Only a small proportion of rPfM18AAP was expressed as soluble protein after induction in Overnight Express™ Instant TB Medium and therefore only ~1 μg of soluble protein was obtained from a one liter E. coli culture. The N-terminal hexahistidine-tag present in rPfM18AAP was detected by Western blot analysis (Figure 1) and purification of rPfM18AAP from E. coli extracts yielded a protein sample of more than 85% purity as determined by scanning densitometry. The rPfM18AAP sample generally contained low molecular weight contaminants (Figure 1). SDS-polyacrylamide gels indicated that rPfM18AAP migrated at ~67 kDa (Figure 1), which approximates the calculated molecular weight (66.9 kDa) of the recombinant protein.

rPfM18AAP minus the hexahistidine-tag (~65 kDa) separated as three entities with pI 6.6, 6.7 and 6.9 during isoelectric focusing (data not shown). Electrophoresis through non-denaturing gels (Figure 2) and Ferguson plot analysis (Additional file 2) showed that the rPfM18AAP occurred mainly as a monomer, tetramer and higher oligomeric forms. The rPfM18AAP dimer was present in very small amounts. Western blot analysis confirmed that the lower protein band was the rPfM18AAP monomer and that the upper smear included the dimer and tetramer (Figure 2). The smear extended past the tetramer to the top of the gels indicating that there were higher oligomeric forms (e.g. octamer and dodecamer) of rPfM18AAP present. The lower bands seen on the non-denaturing gels were not detected during Western blot analysis with the PentaHis™ HRP Conjugate antibody, indicating that these proteins were contaminants isolated during the purification procedure. Teuscher et al showed that recombinant PfM18AAP occurred as an octamer in its active native state by assaying HPLC fractions for enzymatic activity 32. This is the same number of subunits as the human aspartyl aminopeptidase 27. However, the unpublished crystal structures of Clostridium acetobutylicum, Thermotoga maritima, and Pseudomonas aeruginosa M18 proteases (Protein Data Bank: http://www.rcsb.org/pdb/home/home.do), showed that these enzymes occur as dodecamers. To resolve the issue of the subunit composition of the native PfM18AAP, crystallization of the enzyme will be required.

Enzyme activity assays showed that rPfM18AAP cleaves the N-terminal aspartate from the tripeptide Asp-Ala-Pro-beta-Napthylamide. Teuscher et al showed that rPfM18AAP also cleaves dipeptides and that the enzyme is more active on N-terminal glutamates in comparison to N-terminal aspartates 32. Both these amino acids are acidic indicating that PfM18AAP cleaves peptides and proteins that have an acidic N-terminal amino acid.

A pH study showed that rPfM18AAP is active from pH 6.8 to 8.5 (Figure 3), with the highest activity at the physiological pH of 7.5 and no great difference in enzyme activity between pH 7.3–7.7. The enzyme had no activity at pH 5.3, which is close to the pH of the food vacuole (5.0–5.2 34). The parasite uses the food vacuole for the digestion of hemoglobin into 2–10 amino acid peptides 3536, which are subsequently exported into the parasite cytosol where they are converted into amino acids by PfM18AAP and other aminopeptidases 3236. The parasite cytosol has a pH of 7.3–7.4 37, allowing PfM18AAP to be active in this parasite compartment. Even though PfM18AAP has no export signals, it has been localized to the parasitophorous vacuole by immunoblot analysis 32 and the erythrocyte membrane by mass spectroscopy 38. The enzyme could, therefore, also be active in the erythrocyte cytosol which has a neutral pH of 7.2–7.3 39.

A temperature study revealed that rPfM18AAP is functional from 33–39°C with maximum enzymatic activity at 37°C (Figure 3), which indicates that the enzyme is active when the parasite resides in the human host. rPfM18AAP has 90% enzymatic activity at 39°C (Figure 3), indicating that the enzyme is still functional when the human host experiences fever (maximum temperature 42°C), which occurs after the erythrocytes rupture 40. PfM18AAP could therefore be active during invasion and growth in new erythrocytes.

Blot overlay assays with rPfM18AAP against spectrin (Figure 4) and erythrocyte membrane proteins (Figure 5) confirmed our phage display results, which showed an interaction between spectrin and a 33 amino acid peptide of PfM18AAP 20. The anti-His6 antibody does not interact with any of the erythrocyte membrane proteins as shown in Figure 1 lane 1 and Additional file 3. Beta spectrin showed much stronger binding to rPfM18AAP than alpha spectrin (Figure 5 and Fairbanks SDS-PAGE, data not shown), which explains why there is no signal evident for alpha spectrin in figure 5A. rPfM18AAP also reacted with other erythrocyte membrane proteins, including protein 4.1, protein 4.2, actin and glyceraldehyde 3-phosphate dehydrogenase (G3PD) (Figure 5). In addition, performing the overlay under isotonic conditions (Figure 5B) increased binding to all the target erythrocyte membrane proteins.

Actin and alpha spectrin could be prime targets for PfM18AAP, because they have three aspartates (actin) and one glutamate (alpha-spectrin) after the N-terminal methionine, which would be cleaved from newly synthesized proteins by the human methionine aminopeptidases 41. Beta spectrin, protein 4.1 and protein 4.2 contain either an aspartate or a glutamate 32 close to their N-termini. These acidic residues could potentially become available after cleavage of the preceding amino acids by endogenous proteases. Alternatively, another scenario may also be possible. The 33 amino acid binding region of PfM18AAP is on the surface of the enzyme and is not implicated in the catalytic site, which raises the interesting possibility that the enzyme could bind to a protein, but not necessarily cleave it. This could be the case with proteins that do not have an appropriate acidic residue at the N terminus. These proteins could be used as anchors and the enzyme could cleave proteins situated adjacent to it on the membrane, which consists of a network of interacting proteins.

Actin and protein 4.1 are located in the junctional complexes and protein 4.2 is located in the band 3 complex of the erythrocyte membrane 3. The N-termini of beta spectrin molecules are linked to actin and protein 4.1 in the junctional complex and the N-termini of alpha spectrin are located at the self-association sites of spectrin tetramers 3. Cleavage of any of these proteins could therefore destabilize and disrupt the junctional complexes, the band 3 complexes and the spectrin tetramers.

The presumed primary function of PfM18AAP in P. falciparum is to cleave aspartates or glutamates from the oligopeptides that are exported from the food vacuole into the parasite cytosol after hemoglobin digestion 3236. Amino acids are released from the cell and this regulates the colloid-osmotic pressure within the infected erythrocyte to prevent premature cell lysis and to establish a concentration gradient that facilitates the entry of rare amino acids into the infected erythrocyte 42. PfM18AAP may therefore be indirectly involved in regulating the volume of the host cell to accommodate the increasing size of the growing parasite.

Our studies, as well as PfM18AAP mRNA and protein data from other laboratories, indicate that the enzyme could also perform additional functions in the parasitophorous vacuole, the erythrocyte cytosol and particularly at the infected erythrocyte membrane skeleton. Microarray data showed that PfM18AAP mRNA is expressed throughout the erythrocytic stages, with the highest expression levels in early and late trophozoites and in gametocytes 43, whilst Northern blot analysis revealed predominant expression in rings 32. Protein data localized PfM18AAP in merozoites, trophozoites, schizonts 44 and at the infected erythrocyte membrane in trophozoite/schizont stage parasites 38. Western blot analysis, utilizing anti-PfM18AAP antiserum, also revealed the enzyme in rings, the parasite cytosol and the parasitophorous vacuole 32. PfM18AAP mRNA is therefore transcribed into the active enzyme at several stages in the erythrocytic life cycle and the PfM18AAP protein is located in several compartments within the parasite-infected erythrocyte. These data imply that PfM18AAP could play a role in parasite invasion, growth and exit from the host cell, since parasite proteins interact with the erythrocyte membrane and the underlying erythrocyte membrane skeleton during intracellular development.

Merozoites are the invasive form of the parasite in erythrocytic schizogony and utilize proteases to gain entry into the new host cell. Inhibition studies with 1,10-phenanthroline have shown that metalloproteases are involved in parasite invasion 45 and hence PfM18AAP could be involved in this process. During invasion and early growth of the parasite, the host experiences febrile paroxysms triggered by the release of toxins during parasitized erythrocyte rupture 40. Since PfM18AAP is functional at elevated temperatures, we speculate that it may be responsible for regulating, processing and activating other parasite proteins within the ring stage by removing their N-termini.

Several parasite proteases, such as plasmepsin-II 18 and falcipain-2 19, which are primarily involved in hemoglobin digestion inside the food vacuole, also facilitate the release of the parasite from its erythrocytic host by cleaving membrane skeleton proteins such as spectrin. Studies with inhibitors have proven that parasite proteases first weaken the parasitophorous vacuole membrane and then the erythrocyte membrane prior to parasite release 46, which is caused by an osmotic pressure build-up within the erythrocyte 47. Weakening of the erythrocyte membrane would involve the disruption of the spectrin-junctional complex network below the plasma membrane. Given that PfM18AAP is located in the parasitophorous vacuole 32 and at the erythrocyte membrane 38, and that the enzyme binds to several erythrocyte membrane proteins and functions at neutral pH, it is, therefore, plausible that PfM18AAP could also play a role in the release of the parasite from its erythrocyte host.

PfM18AAP may represent a novel drug target, since knockdown experiments utilizing a plasmid containing an antisense copy of the PFI1570c gene, resulted in a lethal phenotype 32. A knockout experiment utilizing a single-crossover strategy 36 produced a truncated PfDAP (PfM18AAP), which retained ~10% enzymatic activity when compared to wild-type parasites. This did not have deleterious effects on parasite replication, possibly because the small amount of active PfM18AAP was sufficient to perform the appropriate enzymatic reactions in the parasite. Additional evidence is thus required to validate PfM18AAP as a drug target. If the enzyme is essential for parasite survival, the development of PfM18AAP inhibitors could lead to new drugs that can be employed to fight malaria infections.