A Novel Protein Acts as a Negative Regulator of Prophenoloxidase Activation and Melanization in the Freshwater Crayfish Pacifastacus leniusculus*

Melanization is an important immune component of the innate immune system of invertebrates and is vital for defense as well as for wound healing. In most invertebrates melanin synthesis is achieved by the prophenoloxidase-activating system, a proteolytic cascade similar to vertebrate complement. Even though melanin formation is necessary for host defense in crustaceans and insects, the process needs to be tightly regulated because of the hazard to the animal of unwanted production of quinone intermediates and melanization in places where it is not suitable. In the present study we have identified a new melanization inhibition protein (MIP) from the hemolymph of the crayfish, Pacifastacus leniusculus. Crayfish MIP has a similar function as the insect MIP molecule we recently discovered in the beetle Tenebrio molitor but interestingly has a completely different sequence. Crayfish MIP as well as Tenebrio MIP do not affect phenoloxidase activity in itself but instead interfere with the melanization reaction from quinone compounds to melanin. Importantly, crayfish MIP in contrast to Tenebrio MIP contains a fibrinogen-like domain, most similar to the substrate recognition domain of vertebrate l-ficolins. Surprisingly, an Asp-rich region similar to that found in ficolins that is likely to be involved in Ca2+ binding is present in crayfish MIP. However, crayfish MIP did not show any hemagglutinating activity as is common for the vertebrate ficolins. A mutant form of MIP with a deletion lacking four Asp amino acids from the Asp-rich region lost most of its activity, implicating that this part of the protein is involved in regulating the prophenoloxidase activating cascade. Overall, a new negative regulator of melanization was identified in freshwater crayfish that shows interesting parallels with proteins (i.e. ficolins) involved in vertebrate immune response.

Invertebrate animals do not have any adaptive immune system and have to rely on innate immune systems. Several such innate systems have been described such as the coagulation system (1), melanization synthesis (2), and the production of antimicrobial peptides (3). The melanization reaction is an important component of the innate immune system of invertebrates and is essential for defense as well as for wound healing (4). In arthropods and most other invertebrates melanin synthesis is achieved by the prophenoloxidase (proPO)-activating 3 system, a proteolytic cascade similar to vertebrate complement (4,5). The proPO-activating system is initiated when microbial polysaccharides, such as lipopolysaccharides (LPS), ␤-1,3-glucans or peptidoglycans (PGN) are recognized by pattern recognition proteins, and the complexes formed induce activation of serine proteinase zymogens in the cascade (4,5). The final step in this process is the conversion of proPO into the active enzyme phenoloxidase (PO). To date ϳ40 proPOs have been cloned and characterized, and several other constituent factors of the proPO system have recently been characterized (4,5). Active PO oxidizes o-diphenols into quinones that are toxic to microorganisms and melanin pigments are formed that also restrict the spreading of microorganisms within the host (4,5). Detailed studies of the molecular mechanism by which the activation is achieved has been performed in Tenebrio molitor (6), Holotrichia diomphalia (7), and Manduca sexta (4). In Tenebrio a complex of the peptidoglycan recognition protein Tm PGRP-SA and peptidoglycan fragments lead to the activation of proPO-activating factors, one of which (proPO-activating factor I) in its active form is mediating cleavage of proPO into an active enzyme (6,8). A similar mechanism of proPO activation is implicated also by research done on M. sexta (9). In crustaceans the proPO-activating system is studied in detail at the cellular and molecular level (5). In crayfish the proPO-activating system is produced in hemocytes and stored in secretory granules similar to the clotting system of horseshoe crabs (1). Upon challenge with microorganisms, the system is released into the hemolymph. Crayfish proPO is cleaved by the proPOactivating enzyme, a serine proteinase that is activated by another serine proteinase (4).
Although melanin formation is essential for host defense in crustaceans and insects, the process needs to be tightly regulated because of the danger to the animal of unwanted production of quinone intermediates and melanization in places where it is not appropriate (10,11). Several proteinase inhibitors, such as serpins, have been described as responsible for preventing improper activation of the proPO system (for details see Ref. 4). Pacifastin is a high molecular weight inhibitor consisting of one light chain containing the protease inhibitors and a heavy chain that contains three transferrin lobes (12). Pacifastin is highly efficient in inhibiting the crayfish proPO-activating enzyme (12), as is Drosophila serpin 27 in inhibiting the Drospohila proPO system and its proPO-activating enzyme (13). In addition to inhibitors of the proPO system, activation of a number of endogenous factors acting as competitive inhibitors of PO activity has been described (5). For example the so-called phenoloxidase inhibitors have been found in Musca domestica, and homologous basic lysine-rich peptides were also detected in mosquitoes and M. sexta (14,15).
We have recently discovered a novel 43-kDa protein from the hemolymph of the beetle T. molitor (Tenebrio MIP) acting as a negative regulator of melanin synthesis (16). This protein, the target of which is presently unknown, is consumed during melanization. Interestingly, no similarity was found between Tenebrio MIP and any other known protein (16).
Here, we now report the presence of a hemolymph protein with an apparently similar function as Tenebrio MIP in the crustacean Pacifastacus leniusculus. Interestingly, this protein has no sequence similarity to Tenebrio MIP. Instead, it is similar to vertebrate ficolin and horseshoe crab Tachylectin 5 (17,18). The crayfish MIP is very efficient in inhibiting activation of the proPO system and thus functions as an important regulatory molecule to prevent unwanted proPO activation.

EXPERIMENTAL PROCEDURES
Animals, Collection of Plasma, and Microbial Organisms-Freshwater crayfish, P. leniusculus were purchased from Torsäng (Lake Vättern, Sweden) and kept in aquaria in aerated tap water at 10°C. Only intermolt animals were used. Hemolymph was collected by bleeding from the abdominal hemocoel through a needle (0.8 mm) into sterile tubes on ice and centrifuged at 800 ϫ g for 10 min at 4°C to remove the hemocytes.
The Gram-negative bacterium Hafnia alvei has earlier been isolated from P. leniusculus hemolymph 4 and was cultured in LB broth. For fresh cultures the bacteria were grown with shaking at 260 rpm overnight at 37°C and then diluted 1:100 and further cultured until the density reached OD 600 of Ϸ0.5. Bacteria H. alvei were injected into the base of the fourth walking leg as earlier described (19).
Hemocyte Lysate Supernatant-Hemocyte lysate supernatant (HLS) was prepared by collecting hemolymph from 8 crayfishes in bleeding buffer (10 mM sodium cacodylate, 250 mM sucrose, pH 7.0). The hemocytes were spun down by centrifugation at 800 ϫ g for 10 min at 4°C and then homogenized in 10 mM sodium cacodylate, 5 mM CaCl 2 , pH 7.0. The homogenate was centrifuged at 25 000 ϫ g for 20 min at 4°C, and the super-natant was adjusted to a protein content of ϳ1 mg/ml, kept on ice, and used as HLS within 1 h.
Induction of proPO Activation and Assay of PO Activity-To confirm the involvement of crayfish MIP in the proPO system, 25 l of HLS (1 mg/ml) was preincubated with 25 l of MIP (wild type or MIP(ϪD 4 ), a mutant protein lacking the tetraaspartic acid stretch, at 0.5-1 g) or buffer for the control for 10 min at 20°C. These mixtures were incubated with 25 g of LPS-PGN (Sigma L3129 from Escherichia coli 0127:B8) and 25 l of 3,4-dihydroxy-L-phenylalanine (L-DOPA, 3 mg/ml) for 5-20 min at 20°C. For analysis of the effect of MIP on PO activity, HLS was preincubated with LPS-PGN for 25 min at 20°C to fully activate proPO prior to the addition of MIP. Phenoloxidase activity was measured as the oxidation of L-DOPA at 490 nm and presented as the means Ϯ standard deviation from four independent experiments. In some experiments the phenoloxidase inhibitor phenylthiourea (PTU) was preincubated with HLS for 5 min at 20°C.
Measurement of Proteinase Activity-To determine whether any activating proteinase was affected by MIP, LPS-PGN-activated amidase activity of HLS was assayed as the hydrolyzing activity toward the chromogenic peptide S-2222 (Suc-Ile-Gly (␥Pip) Gly-Arg-pNA; Chromogenix). Briefly, 25 l of HLS was incubated with 25 g of LPS-PGN and 100 l of 100 mM Tris-HCl, pH 8.0, and 25 l of 2 mM S-2222 at 30°C for 30 min, and then the reaction was terminated by the addition of 50 l of 50% acetic acid, and the absorbance at 405 nM was determined. The effect of MIP was determined by preincubation of MIP (0.5-1 g) or buffer for the controls for 10 min at 20°C prior to the addition of LPS.
Detection of Crayfish MIP in Plasma and Determination of the Amino Acid Sequence-The proteins in plasma were precipitated with acetone and subjected to 12.5% SDS-PAGE or twodimensional gel electrophoresis under reducing conditions. First-dimensional separation was performed according to pI, and second-dimensional separation was done according to molecular weight. The IPG strips (7 cm length, pI range between 3 and 11, nonlinear; GE Healthcare) were rehydrated with rehydration solution including 80 g of protein extracted before, for 12 h or overnight at 20°C. Using the IPGphor system (GE Healthcare), isoelectric focusing was performed at a total of 45.5 kVh at 20°C. The cysteine sulfhydryls were reduced and carbamidomethylated, whereas the proteins were equilibrated in the two-dimensional loading buffer (glycerol, SDS, urea) supplemented with 1% dithiothreitol for 15 min at 20°C, followed by 2.5% iodoacetamide in fresh equilibration buffer for an additional 15 min at 20°C. After equilibration, the IPG strips were applied onto a 7-cm acrylamide gel (12.5%). SDS-PAGE was performed at 30 mA/gel for 50 -60 min at 20°C. All of the electrophoretic procedures were performed at room temperature. One two-dimensional gel was stained with Coomassie Brilliant Blue R-250, and the other gel was transferred to a polyvinylidene difluoride membrane for Western blot. An affinitypurified antibody against Tenebrio MIP (16) was used for Western blot analysis. The proteins were separated by SDS-PAGE and transferred electrophoretically to a polyvinylidene difluoride membrane, blocked by immersion in 5% skimmed milk solution for 1 h. The membrane was then transferred to TBS (10 mM Tris-HCl, pH 7.5, containing 150 mM NaCl) containing the affinity-purified Tenebrio MIP antibody (50 ng/ml) and 1% bovine serum albumin and incubated at 20°C for 1 h. After washing in T-TBS (TBS ϩ 0.1% Tween 20) for 3 ϫ 10 min, the ECL anti-rabbit IgG peroxidase-linked species-specific whole antibody from donkey (GE Healthcare) diluted 1:10000 with TBS ϩ 1% bovine serum albumin was incubated for 1 h and washed with T-TBS for 3 ϫ 10 min. For detection, the ECL Western blotting reagent kit (Amersham Biosciences) was used according to the manufacturer's instructions.
After comparing with the result of Western blot, selected spots from gels stained with Coomassie Brilliant Blue R-250 were excised and cleaved with trypsin by in-gel digestion. The peptides were analyzed by electrospray ionization mass spectrometry on a quadruple time-of-flight mass spectrometer (Waters Ltd.) using Masslynx software.
SDS-PAGE and Western Blot-The proteins in plasma were precipitated with acetone and subjected to 12.5% SDS-PAGE under reducing conditions. To examine what happened to MIP during PO-induced melanization, we induced PO activity as described above. When melanin pigments were generated, the reaction mixture was centrifuged at 28,500 ϫ g at 4°C for 10 min. To confirm whether the disappeared bands on SDS-PAGE were specifically related to melanin synthesis, the inhibitor PTU was added to the same reaction mixture under the same conditions. After incubation, the proteins of the supernatant were precipitated with acetone and analyzed by Western blot after SDS-PAGE under reducing conditions as described above.
cDNA Cloning and Nucleotide Sequencing of 43-kDa Crayfish MIP-Hepatopancreas total RNAs were extracted using GenElute TM mammalian total RNA miniprep kit (Sigma) and followed by RNase-free DNase I (Ambion, Austin, TX) treatment. cDNA was synthesized with ThermoScript (Invitrogen). Several sets of gene-degenerate primers based on the MS sequence results (VVMEDFDANK) were designed for 5Ј-rapid amplification of cDNA ends (RACE) and 3Ј-RACE. 5Ј-RACE was performed according to the manufacturer's protocol (Invitrogen). Oligo(dT) was used to synthesize first strand cDNA. After poly(C) tail were assembled, PCR amplification was performed using the MIP-5R1 (5Ј-TTIGCRTCRAARTCYTCCAT-3Ј) and abridged anchor primer followed by nest PCR amplification with MIP-5R2 (5Ј-TCRAARTCYTCCATIAC-IAC-3Ј) and abridged universal amplification primer (AUAP) using the recommended conditions. For 3Ј RACE, oligo(dT)-adapter was used to synthesize first strand cDNA. An initial amplification by PCR was carried out with primer MIP-3R1 (5Ј-GCGAGTACCAGGA-AGGCTTT-3Ј) and oligo(dT) adapter. The nested PCR was performed with primer MIP-3R2 (5Ј-GCGTTCTCAC-CAGCTGAGAGT-3Ј). Amplified fragments were cloned into PCR 2.1-TOPO TA cloning vector (Invitrogen) and sequenced.

Expression and Purification of Recombinant
Crayfish MIP-The construction of the recombinant baculovirus vector and the expression of the recombinant MIP (rMIP) were performed according to manufacturer's instructions (Invitrogen). The cDNA encoding the mature MIP was subcloned into pFastBac1 vector (Invitrogen) using BamHI and SalI enzyme sites. The recombinant virus for MIP expression was generated according to the manufacturer's instruction manual (Bacto-Bac Baculovirus expression system; Invitrogen). The recombinant virus was amplified using Spodoptera frugiperda 9 (Sf9; Invitrogen) cells in SF-900II serum-free medium (Invitrogen) at 27°C. To produce the protein, Sf9 cells were grown in Sf-900II serum-free medium (Invitrogen) in a 75-cm 2 tissue culture flask. The cells were infected with a cell density of 2 ϫ 10 6 cells/ml at a multiplicity of infection of 10 and were incubated for 3-4 days. The supernatant was collected by centrifuging at 500 ϫ g for 10 min. After discarding the pellets, the supernatant containing MIP protein was desalted by using a PD-10 desalting column (code 17-0851-01; Amersham Biosciences) and applied to a 1-ml HiTrap Q HP column (0.7 ϫ 2.5 cm; Amersham Biosciences, Uppsala, Sweden) column, pre-equilibrated with 20 mM Tris-HCl buffer, pH 8.0, and then this column was washed with the same buffer until no material appeared in the effluent and then gradient eluted with 20 mM Tris-HCl, 1 M NaCl, pH 8.0 (100 ml, 0 -50%) at a flow rate of 60 ml/h. Fractions of 1 ml were collected. The purified protein was subjected to 12.5% SDS-PAGE and stained with Coomassie Brilliant Blue R-250 or transferred to polyvinylidene difluoride membrane to determine the purity by Western blot.
Site-directed Mutagenesis-The cDNA encoding the mature crayfish MIP was subcloned into pFastBac1 vector (Invitrogen), and this plasmid was used for mutagenesis. Site-directed mutagenesis was done using the QuikChange site-directed mutagenesis kit (Stratagene). The mutants were obtained by deleting the four aspartic acids in the Asp-rich region in the cDNA clone: MIP(ϪD 4 ) Ϫ858ϩ: 5Ј-GTTTTCTACCTACGA-CAAGAACAAAGATGGTAACTGCTC-3Ј; and MIP(ϪD 4 ) Ϫ858-: 5Ј-GAGCAGTTACATCTTTGTTCTTGTCGTAG-GTAGAAAAC-3Ј. The PCR was done as follows: 95°C for 30 s, and 18 cycles of 95°C for 30 s, 55°C for 1 min, final extension at 68°C, 6 min. The nicked vector DNA containing the desired mutations was then transformed into XL1-Blue supercompetent cells, and then mutant MIP(ϪD 4 ) plasmids were purified and sequenced to verify the mutated sequence. The generation of the recombinant virus for crayfish MIP(ϪD 4 ) expression and the purification of recombinant crayfish MIP(ϪD 4 ) are similar to the methods used for crayfish MIP.
Hemagglutinating Activity Assay-Hemagglutinating activity toward human erythrocytes type A, B or O of rMIP was determined in assay buffer containing 5 mM CaCl 2 as described in Ref. 20.
RNA Interference Experiments-Oligonucleotide primers including T7 promoter sequences (italics) at the 5Ј end were designed to amplify a 776-bp region of the P. leniusculus MIP gene: 217ϩ, 5Ј-TAATACGACTCACTATAGGGCCACTCACCTCA-GCCGACAC-3Ј; 993Ϫ, 5Ј-TAATACGACTCACTATAGGGCC-AGGCCGCCATACTCGTTA-3Ј. Control 657-bp templates were generated by PCR using primers specific for the green fluorescent protein gene from the pd2EGFP-1 vector (Clontech, Palo Alto, CA) as follows: 63ϩ, 5Ј-TAATACGACTCACTATAGGG-CGACGTAAACGGCCACAAGT-3Ј; 719Ϫ, 5Ј-TAATACGA-CTCACTTAGGGTTCTTGTACAGCTCGTCCATGC-3Ј. To generate dsRNA, the PCR products purified by gel extraction (Qiagen) were used as templates for in vitro transcription using the MegaScript kit (Ambion, Austin, TX), and dsRNA was purified with the TRIzol LS reagent (Invitrogen) method. Small intermolt crayfish (20 Ϯ 2 g, fresh weight) were used for in vivo RNA interference experiments. Briefly, 150 g of MIP or green fluorescent protein control dsRNA dissolved in crayfish saline (0.2 M NaCl, 5.4 mM KCl, 1 mM CaCl 2 , 2.6 mM MgCl 2 , 2 mM NaHCO 3 , pH 6.8) (200 l) was injected via the base of the fourth walking leg. The injection was repeated 24 h after the first dsRNA injection. The presence of MIP protein in the hemolymph of dsRNA-treated crayfish was analyzed by Western blot as described above.
Homology Modeling-A homology model of Pacifastacus MIP protein was built using the structure of lectin called L-ficolin (pdbid 2J61) (18). An alignment of these two proteins was carried out using the PROBCONS web service (21) and manually edited. The MODELLER program version 9.2 (22) was used to build three-dimensional models of crayfish MIP. The quality of the final model was checked using the ProCheck (23) and WhatCheck programs (24).

Tenebrio MIP Antibody Recognizes a Protein in Crayfish
Plasma-We have earlier reported about the existence of MIP, a 43-kDa protein with no significant similarity to other reported proteins, acting as a negative regulator of melanization in the mealworm larvae. One characteristic of Tm MIP is the Asp-rich region including 11 adjacent Asp residues in its central part. To determine whether the freshwater crayfish P. leniusculus expresses a similar MIP protein, we used antibodies raised against Tm MIP in a Western blot experiment of crayfish hemocytes and plasma. As shown in Fig. 1A, a clear band was detected by the antibodies at a molecular mass of ϳ43 kDa. To characterize further the protein that reacted with the MIP antibody, we performed a two-dimensional analysis of crayfish plasma, followed by Western blot. Two clear spots (and another two very weak) appeared at ϳ43 kDa after immunoblotting of the two-dimensional gel (Fig. 1, B and C), and these spots were all subjected to amino acid analysis by matrix-assisted laser desorption ionization time-of-flight. The only peptide sequence that could be identified from these spots by tandem MS was VVMEDFDANK, showing no match with Tenebrio MIP or any other known protein. Thus we assumed that this crayfish protein might represent a new protein of unknown function.
Cloning and Characterization of the cDNA for This Crayfish Protein-To determine the amino acid sequence of this crayfish protein, we designed degenerate primers to the peptide VVMED-FDANK and performed 5Ј-RACE and 3Ј-RACE using cDNA synthesized from hepatopancreas RNA. When we used hemocyte cDNA no transcript could be detected. However, we obtained a cDNA of 1764 base pairs from hepatopancreas, and the deduced amino acid sequence of the open reading frame of this cDNA is shown in Fig. 2A. This cDNA encodes a protein consisting of 326 amino acid residues with a signal sequence of 26 residues. The deduced protein sequence was confirmed to have identical theoretical tandem MS spectra to the protein spot shown in Fig. 1B. The crayfish protein contains three putative N-linked glycosylation sites in the deduced amino acid sequence (open circles in Fig.  2A), and the first at position 30 (NPSI) is likely to have carbohydrates bound according to the tandem MS result.
Surprisingly the sequence of this Pacifastacus 43-kDa protein is totally different from the Tm MIP, and instead the crayfish 43-kDa protein shows a significant sequence similarity in its C terminus to vertebrate fibrinogens. A homology search revealed also a certain similarity with the fibrinogen-related domains (FReD) of vertebrate ficolins, but the collagenous domain found in these proteins is missing in the corresponding region of the crayfish 43-kDa protein (supplemental Fig. S1). Pairwise comparisons of the FReDs showed that the crayfish 43-kDa protein showed higher sequence similarities with vertebrate ficolins as compared with horseshoe crab tachylectin (TL) 5A and 5B, although the molecular size and missing collagenous domain is shared with the tachylectins. TL-5A and TL-5B (TLs-5) contain six and seven cysteine residues, respectively, and two/three of these are supposed to be involved in interchain disulfide linkages. However, the crayfish 43-kDa protein is apparently present in plasma as monomers because a Western blot under nonreducing conditions did only show one band at ϳ43-45 kDa (Fig. 1A) and contains solely four cysteines corresponding to the cysteine residues that in tachylectins and ficolins are involved in intrachain disulfide linkages (Fig. 2B). The TLs-5s have efficient hemagglutinating activity and are easily purified by binding to N-acetyl group-immobilized resins as are L-ficolins (21). In contrast, the crayfish 43-kDa protein does not bind to this resin and has no hemagglutinating activity toward A, B, or O type human erythrocytes (a concentration range of 0.005-25 g/ml was tested).
When comparing the amino acid sequence of the crayfish 43-kDa protein with that of Tm MIP, the similarities were few. To explain the similar antigenicity of these two proteins, we compared their antigenic index plot using the MacVector 7.0 software. The highest index was obtained for the Asp-rich regions that are common to both proteins. As shown in Fig. 2C, Tm MIP contains a region in its central part with 11 Asp residues, whereas the crayfish 43-kDa protein has an Asp-rich region containing five Asp residues with four Asp in one row. This region is probably recognized by the Tm MIP antibody.
Expression of the Crayfish 43-kDa Protein-We cloned the first transcript using RNA extracts from hepatopancreas, whereas no mRNA encoding this protein was found in hemocytes. After obtaining the full sequence, we then analyzed the expression pattern in different tissues by reverse transcription-PCR. As shown in Fig. 3, the transcript for the crayfish 43-kDa protein was detected at fairly low level in hepatopancreas and eyestalk, whereas high expression occurred in nerve tissue, heart, and intestine. Hemocytes and hematopoietic tissue cells did not express this transcript.
Recombinant Pacifastacus MIP Inhibits Melanization in Vitro-Because of the similar antigenicity of the crayfish 43-kDa protein and Tm MIP, we decided to explore the relationship between the 43-kDa protein and the proPO-activating system in crayfish. Accordingly, we produced the recombinant protein and purified it to homogeneity (supplemental Fig. S2). From these results (supplemental Fig. S2), it is also evident that this recombinant protein was detected by the Tm MIP antibody. Then we added the recombinant protein to a fresh preparation of HLS containing an inactive proPO system. As shown in Fig. 4A, this recombinant protein could inhibit LPS-PGN or ␤-1,3-glucan induced PO activity in a dose-dependent manner assayed with L-DOPA as substrate.
When proPO was activated prior to incubation with the recombinant protein, no such inhibition was achieved (Fig. 4B).
To investigate whether the antigenic Asp-rich region is important for the function of the crayfish 43-kDa protein, site-directed mutagenesis was performed by deleting the four-Asp amino acids in the recombinant protein. As anticipated, this mutated protein showed significantly decreased inhibitory activity compared with when the aspartic acid motif was intact (Fig. 4C).
Furthermore, even if the enzyme catalyzed oxidation of L-DOPA by the active PO was unaffected by the recombinant 43-kDa protein, the following nonenzymatic autocatalytic oxidation leading to melanin formation was completely blocked in its presence (Fig. 5A). Because we found that the crayfish 43-kDa pro-  tein also appears to be involved in regulating the proPO system and melanization, we decided to name it P. leniusculus (Pl) MIP). Our results suggest that Pl MIP functions by two different mechanisms. One is to inhibit proPO activation, and the other is to block or delay melanin formation, once PO is activated. We therefore decided to investigate the influence of recombinant Pl MIP on LPS-PGN-induced proteolytic activity, using the commercial substrate S-2222 that we have shown is a good substrate for Pacifastacus prophenoloxidase-activating enzyme (22). The concentration of rMIP chosen in these in vitro inhibitory experiment was in physiologically relevant concentrations. As shown in Fig. 5B, LPS-PGN could induce proteolytic activity in an inactive HLS as compared with the control without LPS-PGN. If r-Pl MIP was preincubated with the inactive HLS prior to the addition of LPS-PGN, a delay in the appearance of proteolytic activity was evident (Fig. 5B). Because Ca 2ϩ is known to be a prerequisite for proPO activation, we also performed experiments where Ca 2ϩ was added to the mixture to test whether the inhibitory effect of r-Pl MIP was a result of Ca 2ϩ entrapment. However, no effect on the inhibitory activity was achieved by the addition of Ca 2ϩ . This indicates that rMIP does affect the activating mechanism by some other mechanism.
Tenebrio-MIP was detected because of its disappearance after activation of the proPO system during melanization of the hemolymph (10), indicating that the protein was degraded during melanin formation. We decided to examine whether Pl MIP disappeared in a similar way during melanin synthesis. Therefore, we induced activation of proPO by LPS-PGN in the presence or absence of the PO inhibitor PTU after preincubation with r-Pl MIP. As we expected, incubation in the presence of PTU did not result in oxidation of L-DOPA, and r-Pl MIP was unaffected (Fig. 5C). When PTU was absent from the reaction mixture, oxidation of L-DOPA proceeded, and then r-Pl MIP disappeared completely from the reaction mixture (Fig. 5C).
Crayfish MIP Is Affecting Melanin Synthesis in Vitro as well as in Vivo-To get information about the in vivo function of Pl MIP, we performed RNA interference experiments in live crayfish using a method we have successfully used to silence proPO as well as the proPO-activating enzyme inhibitor pacifastin (13). However, we were unable to silence MIP expression in live crayfish, probably because Pl MIP is transcribed at a high level in many different tissues. Therefore, we carried out further experiments to reveal information about the in vivo function of Pl MIP.
First we separated granular cells and incubated these cells in L15 culture medium together with crayfish plasma containing Pl MIP overnight at 16°C. After this treatment hardly detectable small melanized areas were found around the cells (Fig. 6A). If the granular cells instead were incubated in medium containing plasma that had been immunodepleted with antibodies against Pl MIP, the medium was heavily melanized, and melanin particles were spread all over the culture dish (Fig. 6B). These experiments clearly indicate that Pl MIP is involved in regulating melanin formation in crayfish. Furthermore, when recombinant MIP protein was added to cells treated with anti-MIP-immunodepleted plasma, melanization was clearly inhibited (Fig.  6C). To further establish whether Pl MIP is related to melanin synthesis of the hemolymph in vivo, we induced melanization in crayfish by injecting the Gram-negative bacterium Hafinia alvei, known to induce melanin synthesis in crayfish. 5 In Fig. 6D it is shown that injection of H. alvei results in a heavy melanization of the hemolymph, and simultaneously a decrease in hemolymph Pl MIP concentration was clearly visible (Fig. 6E). On the other hand injection with the highly pathogenic bacterium Aeromonas hydrophila neither caused a change in Pl MIP content of the hemolymph nor induced melanization (Fig. 6, D and E).
Structure of the Protein and Calcium Binding-The sequence similarity between Pacifastacus MIP and the recognition domain of human L-ficolin was used to build a homology model 5 I. Sö derhäll, unpublished results.  of the Pacifastacus MIP three-dimensional structure (Fig. 7). The model suggests that Pacifastacus MIP is an alpha/beta protein stabilized by two cysteine bridges. The binding site for Ca 2ϩ ions as described in the L-ficolin structure and similar to invertebrate lectin tachylectin 5A (13) is rather well conserved in Pl MIP. Our experiments using a mutant lacking this putative Ca 2ϩ -binding site showed limited inhibitory activity (Fig. 4C), suggesting a role for this site in the function of Pl MIP. Crucially, two Asp residues that form specific contacts to Ca 2ϩ ions through their side chains carboxyl groups in the ficolin structure appear in the model, too. Interestingly, the putative Ca 2ϩbinding site of Pl MIP is longer by one amino acid (Lys) compared with ficolin. Nevertheless, this should not prevent Ca 2ϩ ions from binding because the remaining interactions between Ca 2ϩ ions and ficolin are mediated through interactions with the carboxyl groups of the main chain. Therefore, the insertion of an amino acid into the site will not influence the availability of a carboxyl main chain group for ion coordination. The amino acids described as defining the specificity of substrate binding in L-ficolin structure are not conserved in the suggested model of Pl MIP; therefore different ligands might be expected.

DISCUSSION
In the present study we have identified a novel protein named Pacifastacus MIP from crayfish hemolymph that interferes with the melanization reaction in this animal. We recently described for the first time a protein with a similar function but with a different structure in an insect, T. molitor (10). This is therefore the second protein found to act as a negative regulator of melanin formation in invertebrates. A number of nonproteinaceous factors acting as inhibitors of PO activity have earlier been described (3,23), and phenoloxidase inhibitors peptides are described in several insect species (24,25), but all are shown to inhibit the PO activity. Other proteins affecting melanization are known as proteinase inhibitors that will inhibit one or several of the proteinases that are components of the proPO system, and therefore they will inhibit proPO activation. Among these latter inhibitors is a large molecular weight inhibitor, pacifastin, from P. leniusculus acting as a potent inhibitor of the proPO-activating protease, and Drosophila Spn27, a small serpin supposed to inhibit the proPO-activating enzyme (14,26). In contrast, Pl MIP and Tm MIP do not affect PO activity in itself but instead interfere with the melanization reaction from quinone compounds to melanin. However, in contrast to Tm MIP, the crayfish protein was also shown to delay the induced protease activity that is responsible for proPO activation, but it did not inhibit an already LPS-PGN-activated protease, suggesting that proPO activation is affected by this protein in addition to its effect of inhibiting the following melanin formation from L-DOPA. Pl MIP was found to be expressed in many different tissues, but not in the hemocytes or in the hematopoietic tissue, and the highest expression was detected in heart, nerve tissue, and intestine. Unfortunately, we were unable to silence the expression of Pl MIP, and therefore no detailed in vivo functional studies of this protein could be performed. However, we earlier identified a bacterium from crayfish hemolymph as H. alvei (previously isolated from melanized hemolymph) that is inducing melanization of the hemolymph after injection. When plasma was analyzed for the presence of Pl MIP after injection of this bacterium, melanization was induced, and the MIP protein was found to be absent (Fig. 6E), indicating that Pl MIP has a role as a negative regulator of melanin formation in vivo as well.
Interestingly, we have been able to isolate a protein from crayfish hemolymph with an apparently similar function as Tm MIP by using heterologous antibodies, although these two proteins are completely different when their amino acid sequence are compared. However, the most probable common antigenic surface of Tenebrio and Pacifastacus MIP is the long Asp-rich region in Tm MIP and the shorter Asp stretch found in Pl MIP. These Asp-rich regions are likely to be involved in Ca 2ϩ binding of these proteins. Interestingly, the Asp-rich region in Pl MIP is surrounded by cleavage sites for trypsin-like proteinases, and the sequence is highly similar to the acidic tetra-aspartate sequence of the activation peptide of human trypsinogen (27). The conserved tetra-Asp in trypsinogen is required for enteropeptidase recognition and cleavage and the activation peptide functions as an effective inhibitor of trypsin autoactivation (28). The proPO-activating cascade involves the activation of several proteinases with trypsin-like activity (3). The ability of MIP in inhibiting the activating PO activating cascade was significantly decreased, after its Asp-rich region was deleted (Fig. 4C). This result proves that the tetra-Asp stretch in crayfish MIP is involved in regulating the activating cascade.
Although Tm MIP so far does not show any significant similarity with any known protein, Pl MIP in contrast contains a FReD, most similar to the recognition sites of vertebrate ficolins (14). Lack of the collagenous domain and the molecular size (43 kDa) of Pl MIP is similar to that of TL5s in horseshoe crabs (21), but Pl MIP did not show any hemagglutinating activity and is not likely to have lectin activity as the TL5s. Human L-ficolin was then used as a template for building a homology model of the Pl MIP three-dimensional structure, and the binding site for Ca 2ϩ was found to be highly conserved. The putative Ca 2ϩbinding site of Pl MIP is longer than that of ficolins, and the first two Asp-residues of this site are flanked by another two Asp forming a tetra-Asp stretch that is not found in ficolins or other fibrinogen-related domains. These sequence differences may indicate a different ligand in Pl MIP.
In summary we have isolated a protein, crayfish MIP, with a similar function as Tm MIP, but with a totally different molecular structure. The Pl MIP is most likely an important regulator of the proPO system and will keep the proPO system in a nonactive form until specific inducers such as pathogen-associated molecular patterns or microorganisms are present. Then Pl MIP as well as Tm MIP (10) are degraded, which then will allow activation of the proPO system and melanization. The structural similarities of Pl MIP with ficolins known as activators of vertebrate complement is also interesting and indicates parallels in the regulation between proteolytic cascades involved in defense in vertebrates and invertebrates.