ASABF, a novel cysteine-rich antibacterial peptide isolated from the nematode Ascaris suum. Purification, primary structure, and molecular cloning of cDNA.

Previously, we reported antibacterial activity in the body fluid of the nematode Ascaris suum (Kato, Y. (1995) Zool. Sci. 12, 225-230). The antibacterial activity is due to a heat-stable and trypsin-sensitive molecule that was designated as ASABF (A. suum antibacterial factor). In the present study, the purification, determination of primary structure, and cDNA cloning of ASABF were carried out. The mature peptide of ASABF is a basic peptide consisting of 71 residues and containing four intramolecular disulfide bridges. The amino acid sequence of a precursor for ASABF, deduced from a cDNA clone, indicates that flanking peptides both at the N terminus and at the C terminus are eliminated by processing. ASABF exhibits potent antibacterial activity particularly against Gram-positive bacteria. ASABF has several features that resemble those of insect/arthropod defensins, whereas the statistical significance of the similarity is not observed on comparison of amino acid sequences. A search of data bases revealed ASABF homologues in Caenorhabditis elegans.

Antimicrobial peptides originating from multicellular organisms have been discovered, mainly in arthropods including insects, vertebrates, and plants (1). Interestingly, some antimicrobial peptides isolated from evolutionally distant origins are structurally similar. For example, defensins were originally found in mammalian neutrophil cells (2). Insect/arthropod defensins, isolated from the body fluid of insects and other arthropods, show a certain degree of sequence similarity with mammalian defensins (3). Both mammalian and insect defensins contain six cysteine residues contributing intramolecular disulfide bridges. Cecropins, linear and mostly helical antibacterial peptides without cysteine residues, were first detected in insects (4) and later isolated from porcine small intestine (5). Plant defensins are antifungal peptides with eight cysteine residues (6), and a homologue, drosomycin, was recently demonstrated in the fruit fly Drosophila melanogaster (7).
In addition, most immune proteins of insects, including antimicrobial peptides, are induced by bacterial challenge or wounding. The gene expression of these immune proteins is suggested to be regulated by transcription factors that resemble those controlling the genes for immunoglobulins and acute phase response proteins in vertebrates, e.g. NFB. These results suggest that such regulatory systems are of evolutionally ancient origin, i.e. prior to the divergence of deuterostomes (e.g. vertebrates) from protostomes (e.g. insects) (8).
It is, therefore, possible to argue that some innate immune systems related to antimicrobial peptides may be evolutionally related. However, little has been experimentally studied on the early events in the evolution of the antimicrobial peptide-related defense systems. From this aspect, it is clearly important to explore how antimicrobial peptides and their gene regulation in lower invertebrates diverged during an ancient process of evolution. Although few fossil records are available, nematodes are thought to be of very ancient origin, at least comparable with the divergence time of the lines leading to vertebrates and to arthropods from an ancient group (9). The similarity of the antimicrobial peptide-related defense systems among evolutionally distant organisms, furthermore, encourages the application of model animals for studying the innate immunity. It has already been proposed that D. melanogaster may provide an excellent model for a molecular and genetic approach to innate immune reactions, including organisms other than insects (10). Similarly, the nematode, Caenorhabditis elegans, can also be another candidate for a model.
Parasitic nematodes in animal intestines can survive not only a hostile hydrolytic environment and host immune attacks but also a microbe-rich environment. Hence, the immune defenses against coliform microbes are essential for the parasites. We have already reported antibacterial, bacteriolytic, and agglutinating activities in the body fluid of the intestinal parasitic nematode, Ascaris suum (11). 1 The antibacterial factor ASABF (A. suum antibacterial factor) is a heat-stable and trypsin-sensitive molecule, i.e. peptide/protein. In the present study, the purification, determination of primary structure, and cDNA cloning of ASABF were carried out. The results revealed that ASABF is a novel antibacterial peptide containing four intramolecular disulfide bridges and has several features similar to those of insect/arthropod defensins. ASABF homologues in C. elegans were, moreover, demonstrated by a computer-assisted search of data bases.

Nematodes and Collection of Body Fluid
Adult female A. suum were obtained from Tokyo Shibaura Zohki, Tokyo, Japan. The nematodes were kept at 4°C after isolation from pig small intestines, and body fluid was collected within 5 h as described previously (11). The collected body fluid was stored at Ϫ120°C. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence reported in this paper has been submitted to the DDBJ/GenBank TM

Antimicrobial Assay
Inhibition zone assay was performed for the anti-S. aureus assay, as described previously (11). Briefly, LB-agar plates (12) with small wells containing 10 5 colony-forming units/ml (final) logarithmic phase bacteria were prepared. Samples were poured into each well and incubated at 37°C for 18 h. Antibacterial activity was detected as clear zones around the wells after the incubation. The inhibition zone assay was also used for the antifungal assay. Fungi were inoculated on potatodextrose agar plates (Difco). Wells cut at the edge of the developing fungal lawn received samples. The plates were incubated at 25°C for 72 h, and the formation of clear zones was observed.
In order to determine the IC 50 , 2 the microdilution method described by Alvarez-Bravo et al. (13) was used with modification. Each bacterial strain in the logarithmic phase was suspended in 200 l of optimum medium, containing a series of purified ASABF at a 2-fold increase in concentration. The optical density of the bacterial suspension was adjusted to 0.02 at 650 nm. The following media were used: LB medium for E. coli and S. aureus and IFO802 medium (10 g of polypeptone, 2 g of yeast extract, 1 g of MgSO 4 /7H 2 O, 1 liter of distilled water, pH 7.0) for B. subtilis, M. luteus, and P. vulgaris. The bacterial suspension was incubated at 37°C for E. coli and S. aureus and at 30°C for B. subtilis, M. luteus, and P. vulgaris. Twenty-four h after the incubation, the optical density of the bacterial suspension was measured at 650 nm.

Purification of ASABF
Step 1. Gel Permeation HPLC-Defrosted body fluid was centrifuged at 20,000 ϫ g for 10 min to remove debris. The supernatant was applied to a Superdex 75 HR 10/30 column (Pharmacia Biotech Inc.) connected to a Pharmacia fast protein liquid chromatography system. A modified saline solution (11) was used as the mobile phase at a flow rate of 0.5 ml/min. The elution pattern was monitored at 280 nm. A part of each fraction was directly subjected to inhibition zone assay, and the antibacterial activity against S. aureus was assessed. To estimate the molecular mass of ASABF, a series of standards was used: bovine serum albumin (67 kDa), chymotrypsinogen A (25 kDa), ribonuclease A (13.7 kDa), aprotinin (6.5 kDa), and cyanocobalamin (1.36 kDa). We repeated this step to purify a greater amount of ASABF.
Step 2. Reversed-phase HPLC-The fractions exhibiting antibacterial activity, which were derived from 3-6 ml of parent body fluid, were applied to a Purecil C18 column (Millipore Corp.) connected to a Waters HPLC system. Two linear gradient elutions were employed after an elution with ultrapure water for 25 min: 0 -40% acetonitrile over 40 min and 40 -100% acetonitrile over 10 min. Both the ultrapure water and acetonitrile used as mobile phases contained 0.05% trifluoroacetic acid. The flow rate was constant at 1 ml/min at ambient temperature. The elution pattern was monitored at 225 nm. The fractions were vacuum-dried and then dissolved in ultrapure water again to test the antibacterial activity against S. aureus by inhibition zone assay. The purity of ASABF was tested by tricine/SDS-PAGE with 16% acrylamide gels (14). A series of standards were used to estimate the molecular mass of ASABF: myoglobin (16.9 kDa), myoglobin I-II (14.4 kDa), myoglobin I (8.2 kDa), myoglobin II (6.2 kDa), and myoglobin III (2.5 kDa). The proteins in the gels were visualized by a silver staining kit (Wako Pure Chemical Industries Ltd.).

S-Pyridylethylation
Twenty to forty g of purified ASABF was dissolved into 200 l of 0.25 M Tris/HCl buffer, pH 8.5, containing 6 M guanidine hydrochloride, 1 M EDTA, and 0.1% 2-mercaptoethanol. The sample was flushed with nitrogen and incubated at 37°C for 2 h. Two l of 4-vinylpyridine was added, and the sample was flushed with nitrogen again and incubated at room temperature for 2 h. S-Pyridylethylated ASABF was separated from the reagents by reversed-phase HPLC using a Sephacil C18 SC2.1/10 microbore column connected to a Pharmacia SMART system. Two linear gradient elutions were employed after an elution with ultrapure water for 3 min: 0 -40% acetonitrile over 40 min and 40 -100% acetonitrile over 5 min. Both the ultrapure water and acetonitrile used as mobile phases contained 0.05% trifluoroacetic acid. The flow rate was constant at 0.1 ml/min at ambient temperature. The elution pattern was monitored at 225 and 280 nm. S-Pyridylethylated ASABF was eluted at 27% acetonitrile.

Protease Digestion
Lysyl Endopeptidase Digestion-Twenty to forty g of purified AS-ABF or S-pyridylethylated ASABF was dissolved in 100 l of 200 mM Tris/HCl buffer, pH 8.5, containing 8 M urea. The dissolved samples were diluted twice with ultrapure water, and 1 g of lysylendopeptidase was added. The digestion was carried out at 37°C for 3 h.
V8 Protease Digestion-Twenty to forty g of S-pyridylethylated ASABF was dissolved in 200 l of 50 mM ammonium bicarbonate buffer (pH 7.0) containing 1 g of S. aureus V8 protease and digested at 37°C for 18 h.
The fragments derived by protease digestion were separated by reversed-phase HPLC as described for S-pyridylethylation.

Determination of Amino Acid Sequence
Purified ASABF, S-pyridylethylated ASABF, and the fragments derived by protease digestion were lyophilized and subjected to automated sequence analysis using an Applied Biosystems Procise TM or a Beckman LF3000.

Mass Spectrometry
The exact molecular mass of intact ASABF was determined by an ion spray ionization mass spectrometer (API 300 triple quadrupole mass spectrometer, Perkin-Elmer). The quadrupole was scanned over 500-2000 Da using a step size of 0.1 Da and a 1.0-ms dwell time/step. A matrix-assisted laser desorption ionization-time of flight mass spectrometry was used (Voyager TM -RP, PerSeptive Biosystems) to determine the mass of the S-pyridylethylated fragment of ASABF, Arg 68 -Gly 89 .

cDNA Cloning
A cDNA for ASABF was cloned using three-step PCR amplification.
Step 1. Reverse Transcriptase-PCR-The poly(A) ϩ RNA isolated from the body walls of adult female A. suum, as described by Kuramochi et al. (15), was kindly given by Prof. Kiyoshi Kita, Tokyo University. Singlestranded cDNAs were synthesized from 0.3 g of the poly(A) ϩ RNA and oligo(dT) 20 -M4 adaptor primer, 5Ј-GTTTTCCCAGTCACGAC(T) 20 -3Ј, using avian myeloblastosis virus reverse transcriptase. The cDNA coding Thr 34 -Arg 43 was amplified by PCR using a set of degenerate primers: the sense primer (29-mer) whose sequence is deduced from Cys 27  All reagents used in this step were obtained from an RNA LA PCR kit (AMV) (Takara). Denaturation was carried out at 95°C for 6 min (first cycle) or 1 min (second and following cycles), annealing at 35°C (initial 10 cycles) or 45°C (following 30 cycles) for 1 min, and polymerization at 72°C for 1 min. The total number of cycles was 40. Only the product of expected size was found. This product was subcloned into pGEM-T vector (Promega) and sequenced by a dye terminator system (PRISM TM , Applied Biosystems) with an automated DNA sequencer (373A, Applied Biosystems).
Step 2. Amplification of 5Ј-End Using SL1 Primer-Most of the mRNAs in Ascaris lumbricoides are trans-spliced and acquire a common 22-nt SL1 sequence at the 5Ј-end (16). It is thus highly possible that cDNAs for ASABF contain the SL1 sequence. PCR was performed using a set of primers: the sense primer (22-mer) whose sequence is identical to the SL1 sequence, 5Ј-GGTTAAATTACCCAAGTTTGAG-3Ј; the antisense primer (23-mer) whose sequence is identical to that for Thr 34 -Arg 43 revealed in "Step 1," 5Ј-CGACCTCCACGTTTCTCA-CAGTG-3Ј. All reagents used in this step were obtained from an LA PCR kit Ver.2 (Takara). Denaturation was carried out at 95°C for 6 min (first cycle) or 1 min (second and following cycles), annealing at 55°C for 1 min, and polymerization at 72°C for 1 min. Taq DNA polymerase was added during the first denaturation, i.e. a hot start mode. The cycle was repeated 30 times. The major product was shown to be of 0.2 kbp and was subcloned. The nucleotide sequence of the product was se-quenced as described above. The nucleotide sequence deduced the putative signal sequence and the N-terminal region of the mature ASABF.
Step 3: 3Ј Rapid Amplification of cDNA Ends-To determine the sequence of a full-length cDNA for ASABF, 3Ј rapid amplification of cDNA ends was carried out using a set of primers: the sense primer (35-mer) whose sequence is identical to the 5Ј untranslated region revealed in "Step 2," 5Ј-GATATTCAGCAAAAAAGACAAAACTACT-GTCGACC-3Ј; and M13M4 primer (17-mer) as an antisense primer, 5Ј-GTTTTCCCAGTCACGAC-3Ј. PCR conditions were identical to those described under "Step 2." Major products were found to be of 0.6 and 0.25 kbp. Their sequences revealed that the product of 0.6 kbp was the full-length cDNA for ASABF, except for the SL1 sequence.

Computer-assisted Sequence Analysis
Standard sequence analyses were performed using Genetyx-Mac Ver. 7.3 (Software Development, Tokyo, Japan). The MPsrch TM (Smith-Waterman algorithm, University of Edinburgh, U. K.) was used for searching the nucleic acid data bases at DDBJ, GenBank, and EBI Data Bank and the protein data bases at Swiss-Prot, Protein Information Resource, GenPept, and Protein Data Bank via the on-line E-mail server of the DNA Information and Stock Center, Tsukuba, Japan. Furthermore, the cDNA catalogue of C. elegans, including unpublished data, was searched using the BLAST algorithm (17) through the kindness of Prof. Yuji Kohara (National Institute of Genetics, Mishima, Japan). The statistical significance of sequence similarity was estimated by a jumbling test (18) using the program employed by Nagata et al. (19). The criteria described by Doolittle (20) were used to evaluate the score of the jumbling test.

RESULTS
Purification of ASABF-Because only a limited amount of A. suum body fluid was available, we selected a short step HPLCbased procedure to minimize loss. The body fluid was centrifuged to remove debris and directly subjected to gel permeation HPLC (Fig. 1A). The peak of antibacterial activity against S. aureus was detected at 6 kDa, estimated with a standard curve of molecular mass. Further purification was achieved by reversed-phase HPLC (Fig. 1B). The antibacterial activity against S. aureus was detected as a single peak. This peak was separated by tricine/SDS-PAGE, and a single band was detected at 8 kDa under non-reducing conditions or reducing conditions with ␤-mercaptoethanol (data not shown).
Primary Structure-The N-terminal sequence of intact, Ala 19 -Gly 89 , and S-pyridylethylated ASABF was determined using an automated gas-phase sequenator (Fig. 2). S-pyridylethylated ASABF was digested by lysyl endopeptidase, and the fragments Ala 19 -Lys 35 , Val 36 -Lys 46 , Phe 47 -Lys 57 , and Arg 58 -Lys 83 were separated, whereas Gly 84 -Gly 89 was not found. Furthermore, the fragments Ala 19 -Glu 56 , Lys 57 -Asp 67 , and Arg 68 -Gly 89 were derived by the S. aureus V8 protease digestion of S-pyridylethylated ASABF. The sequence of these fragments was determined. The sequence from overlapping fragments was compared, and the entire amino acid sequence of mature ASABF was determined.
We next submitted the S-pyridylethylated fragment, Arg 68 -Gly 89 , to a matrix-assisted laser desorption ionization-time of flight mass spectrometry and obtained a molecular mass of 2419.9 Da. This is in good agreement with the mass calculated for this fragment, 2418.7 Da. Thus, the C-terminal amino acid residue is confirmed to be Gly 89 .
The mature peptide of ASABF contains eight cysteine residues. The molecular mass of intact ASABF was determined to be 7412.3 Da by ion spray mass spectrometry. Comparison with the molecular mass assessed by gel permeation HPLC (6 kDa) and tricine/SDS-PAGE (8 kDa) suggests that no oligomerization occurred, i.e. no intermolecular disulfide bridge is indicated. On the other hand, the mass calculated for the intact ASABF is 7420.4 Da, i.e. in excess of 8 Da relative to the experimental mass. This difference of 8 Da is well explained if all of the eight cysteine residues contributed to the intramolecular disulfide bridges. This elucidation is also supported by the following experiment. Intact ASABF was digested with lysyl endopeptidase and subjected to reversed-phase HPLC. Only a single peak was detected in this case. However, after S-pyridylethylation of this peak, the same profile with four major peaks was observed as for that of S-pyridylethylated ASABF digested with lysyl endopeptidase. These results suggest that the four regions, Ala 19 -Lys 35 , Val 36 -Lys 46 , Phe 47 -Lys 57 , and Arg 58 -Lys 83 , bind to each other with intramolecular disulfide bridges.
No modification, e.g. glycosylation or phosphorylation, was indicated based on the data of the HPLC profiles of the sequenator and the mass spectrometry. The pI calculated from the entire sequence of mature ASABF is 8.7, i.e. ASABF is a cationic molecule.
cDNA Cloning and Deduced Precursor-A cDNA for ASABF was cloned using a three-step PCR-based approach with poly(A) ϩ RNA from the body walls of A. suum as a template ( Fig. 2; and see "Materials and Methods"). A precursor peptide for ASABF is deduced from the nucleotide sequence of the cDNA. From the amino acid sequence of the precursor, mature ASABF was indicated to be flanked by a hydrophobic putative signal peptide, Met 1 -Ala 18 , at the N terminus. In addition, a four-residue peptide, Arg 90 -Ser 93 , was found at the C terminus as a flanking peptide that should be eliminated by processing. Some pre-mRNAs in nematodes are processed by trans-splicing and acquire a common 22-nt SL sequence from a small SL RNA (16). In the cloning procedure, the PCR amplification of the 5Ј-end was achieved using the SL1 sequence as a sense primer. This result suggests that the mRNA for ASABF contains the SL1 sequence at the 5Ј-end (Fig. 2).
Antimicrobial Activity-Purified ASABF was tested for antibacterial activity (Table I). Gram-positive bacteria, S. aureus, M. luteus, and B. subtilis were very sensitive to ASABF. Their IC 50 s were estimated to be 0.6 -5 g/ml, i.e. 0.08 -0.7 M. Gram-negative bacteria E. coli and P. vulgaris were less sensitive than Gram-positive bacteria. S-pyridylethylated ASABF exhibited no antibacterial activity against S. aureus. No antifungal activity was detected against the tested fungi, A. brassicicola, S. tritci, and T. viride.

DISCUSSION
This paper describes the purification, primary structure, and cDNA cloning of the novel antibacterial peptide ASABF discovered in the body fluid of the nematode A. suum. ASABF has been confirmed as a peptide, and it is thus strongly suggested that antibacterial peptides contribute to the immune defense of the ancient animal nematodes. Mature ASABF is a basic 71residue peptide containing eight cysteines engaged in intramolecular disulfide bridges. To the best of our knowledge, this is the first report on the structure of an antibacterial protein in a nematode.
In some trials, a protein exhibiting weak antibacterial activ-ity was eluted at higher acetonitrile concentration than that of ASABF, by reversed-phase HPLC in step 2 of the purification. The IC 50 of this minor antibacterial protein against S.aureus was estimated to be 200 g/ml, which is much higher than that of ASABF. The N-terminal 19-amino acid sequence of this peptide was completely identical to that of ABA-1 (21). ABA-1 is known to be the most abundant protein in the body fluid of A. suum and binds fatty acids at high affinities. We separated ABA-1 by gel permeation HPLC using a Superdex 75 HR 10/30 column. This partially purified ABA-1 was subjected to reversed-phase HPLC. Neither the partially purified ABA-1 nor any fractions separated by reversed-phase HPLC exhibited antibacterial activity against S. aureus and M. luteus. We speculate that the minor antibacterial activity might be attributed to a protein similar to ABA-1 or a degraded ABA-1. However, further analyses could not be carried out because this minor antibacterial activity was not always observed. There is no evidence to indicate other antibacterial factors in the body fluid. Furthermore, the antibacterial spectrum of ASABF is in good agreement with that of the body fluid (11). 3 It is, therefore, suggested that ASABF is the major antibacterial molecule in the body fluid of A. suum.
The cDNA cloning studies reveal that mature ASABF is processed from a 93-residue precursor. A 4-residue peptide, Arg 90 -Ser 93 , is thought to be removed in addition to the elimination of a putative signal peptide by the processing. The exact mechanism of the processing remains to be elucidated. It is noteworthy that the 4-residue peptide includes a dibasic cleavage site Arg 90 -Arg 91 (22,23).
Several types of antimicrobial peptides containing cysteine residues have been reported. Insect/arthropod defensins are antibacterial peptides containing six cysteine residues engaged in intramolecular disulfide bridges (24). We found similarity between ASABF and insect/arthropod defensins in several features. Both of them are cationic peptides and more effective against Gram-positive bacteria than Gram-negative bacteria. Insect/arthropod defensins have a consensus sequence, Cys1- (25). This consensus sequence is highly conservative among insect/arthropod defensins, except for the Gly between Cys3 and Cys4 of sapecin B (26). We can find the consensus sequence in the sequence of mature ASABF ignoring Cys 50 and Cys 69 . The intramolecular disulfide bridges are essential for the antibacterial activity of the insect defensin, sapecin (27). The antibacterial activity of ASABF was also lost by S-pyridylethylation. In contrast to these similar features, some other data do not support the belief that ASABF is a member of the insect/arthropod defensin family. A computerassisted multiple alignment suggests that tenecin (28) is the insect/arthropod defensin most similar to ASABF (Fig. 3). The homology between tenecin and ASABF is 25% identity and 52% similarity in the optimum region corresponding to Ser 24 -Cys 66 of ASABF. The statistical significance of the sequence similar-3 Y. Kato, unpublished data.

FIG. 2.
Nucleotide sequence of a cDNA clone for ASABF. The deduced amino acid sequence of a precursor is represented below the nucleotide sequence. The region corresponding to mature ASABF is underlined. The primary structure of mature ASABF was determined using an automated gas-phase sequenator and was completely identical to this underlined sequence. The 22-nt trans-spliced leader sequence SL1 is shown as ([SL]). The termination codon is marked with an asterisk. The polyadenylation consensus signal is double underlined. ity was evaluated by a jumbling test. The normalized alignment score was estimated to be 3.82 of the standard deviation, and its evaluation is "marginal" (20). All of the normalized alignment scores between ASABF and other insect/arthropod defensins are Ͼ3.0 of the standard deviation, i.e. "improbably significant," except for sapecin B (3.01 of the standard deviation). 4 In conclusion, the significant sequence similarity is not verified, whereas some similar features are observed between ASABF and insect/arthropod defensins. Moreover, ASABF contains eight cysteine residues, whereas the number of cysteine residues is six, without exception, in insect/arthropod defensins (24). We thus propose to classify ASABF into a novel group of antibacterial proteins in the present situation. However, we are not rejecting the proposal that ASABF and insect/arthropod defensins are possibly related by common ancestry. Are they evolutionally related? This question is curious from the aspect of a search for the origin of cysteine-rich antibacterial peptides. Insect/arthropod defensins show a certain degree of sequence similarity with mammalian defensins as mentioned in the Introduction. It is unclear whether these antibacterial peptides diverged from a common ancestor molecule. The discovery of ASABF, however, suggests that the cysteine-rich antibacterial peptides could be very ancient in origin. Further studies on antibacterial proteins in lower invertebrates should elucidate the evolutional relationship among ASABF, insect/ arthropod defensins, and mammalian defensins. In addition, antifungal peptides containing eight cysteine residues have been also reported, i.e. plant defensins and drosomycin (see Introduction). ASABF exhibits no potent antifungal activity and no significant sequence similarity to these antifungal peptides. However, allowing for several gaps, the array consisting of eight cysteine residues seems to be arranged in a similar pattern between ASABF and these antifungal peptides. Interestingly, it has been suggested that structural and functional properties of plant defensins resemble those of insect and mammalian defensins (6). Revealing the relationship between AS-ABF and these antifungal peptides might be another key to studying the evolutional relationship among cysteine-rich antimicrobial peptides. One of our goals is to introduce the nematode C. elegans as a model animal for investigation on innate immunity, as mentioned in the Introduction. From this aspect, it is very curious regarding whether ASABF homologues exist in C. elegans. A cDNA catalogue by Prof. Yuji Kohara was searched. BLAST data base searches revealed significant sequence identity with a deduced protein from the cDNA sequence, yk150c7 (Fig. 4). In the optimum region corresponding to Leu 33 -Gly 72 of ASABF, yk150c7 exhibits 42% identity and 57% similarity with a normalized alignment score of 5.61 of the standard deviation, i.e. "probably significant." Furthermore, nucleic acid data bases and protein data bases were searched. The protein deduced from the putative gene, T22H6.5, was found to be a protein most similar to both ASABF and yk150c7 by the MPsrch TM data base search (Fig. 4). In the optimum region corresponding to Leu 40 -Cys 69 of ASABF, T22H6.5 exhibits 39.4% identity and 54.5% similarity with a normalized alignment score of 4.61 of the standard deviation. T22H6.5 is also similar to yk150c7 with 48.6% identity and 67.6% similarity in the optimum region corresponding to Phe 38 -Cys 74 of T22H6.5. T22H6.5 contains nine cysteine residues, and one of the cysteines is found in the highly hydrophobic putative signal sequence at the N-terminal region. The array consisting of eight other cysteine residues is similar to that of ASABF. It is noteworthy that the highly similar region among ASABF, yk150c7, and T22H6.5 is almost identical to the region overlapping insect/arthropod defensins (Figs. 3 and 4). The function of the deduced proteins from yk150c7 and T22H6.5 has been unknown and is not predicted. Further experimental analyses are necessary to confirm the function of these ASABF homologues in C. elegans.
To date, a number of antimicrobial proteins were isolated from multicellular animals. Most of them are, however, derived from higher animals, e.g. vertebrates and arthropods. The higher animals seem to develop characteristic defense systems, e.g. B-and T-cell-based adaptive immunity in vertebrates and prophenoloxidase cascades in arthropods, overlaying primitive immunity as described previously (11). Studying the immune defense of lower invertebrates, such as nematodes, could be a way to isolate the primitive systems from these additional systems. The present work was carried out as the initial step of this project. Center) for searching for information on T22H6.5, Prof. Malcolm Kennedy (University of Glasgow) for permission to cite his unpublished data as a personal communication, Drs. Hitoshi Saito and Jun Ishibashi (National Institute of Sericultural and Entomological Science) for helpful discussions, Prof. Kohichiro Fujita and Dr. Nobuaki Akao (Tokyo Medical and Dental University) for providing general information on Ascaris, and Prof. Yasumi Ohsima and his laboratory members (Kyushu University) for providing general information on C.elegans. We are grateful to Dr. Keiji Kurata (National Institute of Sericultural and Entomological Science) for continuous help in all aspects.