Transcriptome analysis of expressed sequence tags from the venom glands of the fish Thalassophryne nattereri

Biochimie 88 (2006) 693–699 www.elsevier.com/locate/biochi

Transcriptome analysis of expressed sequence tags from the venom glands of the fish Thalassophryne nattereri G.S. Magalhães a,b, I.L.M. Junqueira-de-Azevedo c, M. Lopes-Ferreira a, D.M. Lorenzini d, P.L. Ho c, A.M. Moura-da-Silva a,* a

Laboratório de Imunopatologia, Instituto Butantan, Av. Vital Brasil 1500, 05503-900 São Paulo, SP, Brazil Laboratório de Imunogenética, Instituto Butantan, Av. Vital Brasil 1500, 05503-900 São Paulo, SP, Brazil c Centro de Biotecnologia, Instituto Butantan, Av. Vital Brasil 1500, 05503-900 São Paulo, SP, Brazil d Departamento de Parasitologia, Instituto de Ciências Biomédicas, Universidade São Paulo, São Paulo, Brazil b

Received 19 July 2005; accepted 22 December 2005 Available online 27 January 2006

Abstract Thalassophryne nattereri (niquim) is a venomous fish found on the northern and northeastern coasts of Brazil. Every year, hundreds of humans are affected by the poison, which causes excruciating local pain, edema, and necrosis, and can lead to permanent disabilities. In experimental models, T. nattereri venom induces edema and nociception, which are correlated to human symptoms and dependent on venom kininogenase activity; myotoxicity; impairment of blood flow; platelet lysis and cytotoxicity on endothelial cells. These effects were observed with minute amounts of venom. To characterize the primary structure of T. nattereri venom toxins, a list of transcripts within the venom gland was made using the expressed sequence tag (EST) strategy. Here we report the analysis of 775 ESTs that were obtained from a directional cDNA library of T. nattereri venom gland. Of these ESTs, 527 (68%) were related to sequences previously described. These were categorized into 10 groups according to their biological functions. Sequences involved in gene and protein expression accounted for 14.3% of the ESTs, reflecting the important role of protein synthesis in this gland. Other groups included proteins engaged in the assembly of disulfide bonds (0.5%), chaperones involved in the folding of nascent proteins (1.4%), and sequences related to clusterin (1.5%), as well as transcripts related to calcium binding proteins (1.0%). We detected a large cluster (1.3%) related to cocaine- and amphetamine-regulated transcript (CART), a peptide involved in the regulation of food intake. Surprisingly, several retrotransposon-like sequences (1.0%) were found in the library. It may be that their presence accounts for some of the variation in venom toxins. The toxin category (18.8%) included natterins (18%), which are a new group of kininogenases recently described by our group, and a group of C-type lectins (0.8%). In addition, a considerable number of sequences (32%) was not related to sequences in the databases, which indicates that a great number of new toxins and proteins are still to be discovered from this fish venom gland. © 2006 Elsevier SAS. All rights reserved. Keywords: Fish venom; Thalassophryne nattereri; Kininogenase; Natterin; Lectin; Transcriptome

1. Introduction Venoms of poisonous animals have been extensively studied because they are a potential source of pharmacological agents and physiological tools. However, fish venoms have received little attention. Although the number of reports on fish venoms is low, a number of interesting toxins have been de-

Abbreviations: ESTs, expressed sequence tags. author. Tel.: +55 11 3726 7222; fax: +55 11 3726 1505. E-mail address: [email protected] (A.M. Moura-da-Silva).

* Corresponding

0300-9084/$ - see front matter © 2006 Elsevier SAS. All rights reserved. doi:10.1016/j.biochi.2005.12.008

scribed in [1–4] and showing unknown primary structures to the literature [5,6]. It seems likely that many toxins remain to be discovered from the venoms of these animals, including some that could be useful as research tools or therapeutic agents. Thalassophryne nattereri is a venomous fish that has drawn our attention due to the severity of the accidents it provokes. Commonly known as niquim, it belongs to the Batrachoididae family and is found buried under the mud in shallow water on the northern and northeastern coast of Brazil. Stepping on the fish or grabbing it allows its dorsal spines to protrude and inject the venom produced by the glands located at the base of the spine.

694

G.S. Magalhães et al. / Biochimie 88 (2006) 693–699

T. nattereri poisoning causes excruciating local pain, edema, and necrosis that may lead to permanent disabilities [7]. In experimental models, T. nattereri venom toxins efficiently induce edema, pain, myotoxicity, impairment of blood flow, platelet lysis and they are cytotoxic to endothelial cells [8,9]. Recently, we demonstrated that the venom has a kininogenase activity, which contributes to the pain and edema induced by the toxin, since these effects were diminished with kininogenase-specific antagonists [10]. Due to the diversity of effects of this venom, the potential pharmacological applications of its toxins, and the lack of information about the venoms of fish in general, we decided to catalogue the toxins. We used an expressed sequence tag (EST) approach because it provides a rapid and reliable method for gene discovery as well as a resource for the large-scale analysis of gene expression of known and unknown genes [11]. To accomplish this task, a cDNA library was generated from the venom glands and partial sequencing of the cDNAs was performed, generating hundreds of ESTs. Using this library, we previously reported the discovery of natterin, the major toxin of T. nattereri venom, and which presents many of the actions of the crude venom. In the current report we show the complete analysis of 775 ESTs from T. nattereri venom glands, revealing the major venom toxins.

ABI prism Big Dye Terminanator kit (PE Applied Biosystems) according to manufacturer’s instructions. To extract the high quality sequence region, the ESTs were subjected to the Phred program [17] with the window length set to 75 and the standard quality to 20. The CrossMatch program was used to remove vector (parameters: minmatch 12, penalty –2, minscore 18); 5′ adapters (minmatch 8, penalty –2, minscore 10); and 3′ primers with poly(A/T) sequences > 15 (minmatch 17, penalty –2, minscore 21). All sequences shorter than 150 bp long were eliminated. ESTs that shared an identity of > 95 out of 100 nucleotides were assembled in contiguous sequences with the CAP3 program [18]. T. nattereri venom gland sequences (clusters and singlets) were searched against public databases (nr/ NCBI, Swissprot+ TREMBL/EMBL) using the BlastX program with the e-value cutoff set to < 10−5 [19] to identify putative functions of the new ESTs. The signal peptide was predicted with the SignalP 3.0 program (http://www.cbs.dtu.dk/ services/SignalP/) [20]. The alignment and shadings were performed with programs available at http://www.ch.embnet.org/ software/TCoffee.html [21] and http://bioweb.pasteur.fr/seqanal/interfaces/boxshade.html#threshold, respectively. In order to evaluate the most expressed ESTs, a less stringent alignment using a shared identity of > 65 out of 100 nucleotides was used.

2. Material and methods

3. Results and discussion

2.1. cDNA library construction

3.1. EST sequencing and clustering

The cDNA library was constructed from mRNA prepared from nine glands extracted from three specimens of T. nattereri collected in the Mundaú lagoon in the state of Alagoas, Brazil. The fish were milked 4 days before RNA extraction, because Paine et al. [13] have suggested that this increases mRNA production in snakes. The integument covering the opercular spines was removed to expose the gland, which appears as a rounded mass at the base of the spine. The venom gland was carefully removed as described by Halstead [14]. The RNA was extracted with Trizol reagent (Invitrogen). The mRNA was isolated with an oligo-dT cellulose column (Amersham-Pharmacia Biotech). The cDNA was prepared with the Superscript Plasmid System for cDNA Synthesis and Cloning (Life Technologies, Inc.), according to the supplier’s instructions. cDNA was linked to Eco RI adapters and subjected to agarose gel electrophoresis. cDNAs between 400 and 800 bp and larger than 800 pb were cloned directionally into pGEM11Zf+ (Promega). Competent Escherichia coli DH5α cells were transformed by calcium precipitation and plated on 2YT agarose plates containing 100 μg/ml of ampicilin [15,16].

The cDNAs were subjected to gel electrophoresis and two sub-libraries were constructed, one with cDNAs ranging in size between 400 and 800 bp, and one with cDNAs > 800 bp. Since most venom toxins are located above 20 kDa on a SDS/PAGE gel [8], we sequenced more cDNAs from the large than from the small library (513 vs. 346). The average readable sequence length was 390 bp. Low-quality sequences and sequences matching with ribosomal RNA were removed. The remaining 775 ESTs were deposited in the dbEST division of the GenBank under accession numbers CO815830 to CO816604. All those ESTs were assembled using stringency parameters of > 95% of identity in a sequence including at least 100 nucleotides in such a way that only ESTs showing this match were considered to be encoded by the same gene. The ESTs assembling resulted in 96 clusters (TNAT0001 to TNAT0096) showing more than one ESTs and 420 singletons (TNAT0100 to TNAT0520).

2.2. DNA sequencing and bioinformatics analysis Single-pass sequencing of the 5′-termini of randomly selected cDNA clones was conducted with standard M13 forward primers on an ABI 3100 automatic DNA sequencer using the

3.2. Cluster identification Since amino acid sequences are more useful to detect homology over long periods [22], the assembled sequences were translated in all six reading frames and compared to the sequences in the NCBI nr and Swissprot protein databases. Sequences that did not match were further compared against the GenBank and dbEST nucleotide databases (Blastn). From 516 sequences, 288 showed similarity (e-value < 10−5) to pre-

G.S. Magalhães et al. / Biochimie 88 (2006) 693–699

viously described sequences from other species, including fish. The remaining 219 clusters were not closely related to known sequences. Sequences closely related to previously identified sequences with known functions were classified in categories. If the previously identified sequence had multiple functions, the classification of the new sequence was based on the main function. The number of ESTs in each category is shown in Fig. 1. Proteins involved in gene and protein expression were abundant (14.0%), which is not surprising in a protein producing and secreting gland. Most of these were ribosomal proteins, but translation initiation and elongation factors were also found. Within the category of cell defense homeostasis (4.4%), several

695

ESTs with homology to genes involved in the immune system, such as the immunoglobulin heavy chain and major histocompatibility complex II, were identified. In addition, ESTs showing homology with genes involved in innate immunity, such as beta-2 microglobulin, ferritin, and transferrin, were found. The cell signaling and cell communication category accounted for 7.4% of the ESTs, and included a homologue of the S100 calcium binding protein. Sequences related to metabolism represented 4.5% of the transcripts. Cell structure and motility accounted for 6.3%. Sequences matching with proteins involved in transport such as Sec-23-like B and apolipoprotein E precursor accounted for 2.9% of the transcripts. One surprising finding in our cDNA library was the presence of sequences that showed homology with retrotransposable elements (1.0%), which included transposases, reverse transcriptase, pol polyprotein and others. The main venom components identified were natterin-related sequences, which accounted for 18% of the transcripts (18%). A group of C-type lectins (0.8%) was also detected and will be described in a future study. It may be that additional toxins are present in the group of ESTs that did not match with presently known proteins. The characterization of these new ESTs will depend on protein characterization from venom samples. 3.3. The most abundant transcripts in the T. nattereri venom gland Since we used a non-normalized library, the frequency of cDNAs in the library is a reflection of the frequency of the mRNA in the venom tissue. We presume that many highly expressed transcripts may be toxins [15]. As many of the small clusters identified were unknown to the database, we re-grouped the clusters using a less stringent criterion for alignment (more than 65% alignment of identity over an extension of 100 nucleotide) to identify the most expressed similar sequences. In this way, highly expressed family of sequences that might represent toxins (paralogous or populational variants) may be revealed. Table 1 shows the 15 most abundant clusters obtained this way, with their respective number of clones. 3.4. Natterins are the most abundant toxins

Fig. 1. Functional classification of the transcripts from T. nattereri venom gland. (A) Relative proportion of the different types of transcripts. Unknown includes ESTs that matched with sequences with unknown function. No hit includes ESTs that did not match with currently known sequences. Others includes ESTs that matched with sequences with known function, but that did not fit in the other functional categories. (B) Table showing the occurrences of the transcripts, their redundancy (ESTs/clusters) and proportions over all the transcripts.

Table 1 shows that natterin-related sequences account for 18% of the library (group 1). The presence of a putative signal peptide and the abundance of these ESTs suggest that these ESTs may encode toxins. In fact, this assumption was confirmed, and Natterin sequences showed to be composed of five related toxins with kininogenase activity [12]. The natterins are the major venom toxins, and we showed that they present most of the effects of the crude venom, including the induction of edema, pain, and kininogenase activity [12].

696

G.S. Magalhães et al. / Biochimie 88 (2006) 693–699

Table 1 List of the most abundant clusters Groups 1

Number of cluster 31

Number of clones 140

2

3

18

3

5

14

4

2

12

5

3

6 7

Clones/ clusters 4.5 6.0

Percent of total 18.0

Putative identification Natterins *

2.3

No hit *

2.8

1.8

Actin

6.0

1.5

Clusterin *

11

3.6

1.4

Cytokeratin

4 1

11 10

2.7 10.0

1.4 1.3

Hsp70 CART *

8

4

8

2.0

1.0

S100-like calcium binding protein

9

3

6

0.7

C-type lectin *

10

1

6

2.0 6.0

0.7

Ribosomal protein L36

11

1

5

5.0

0.6

Tumor necrosis factor receptor *

12

1

4

4.0

0.5

Ribosomal protein P2

13

1

4

4.0

0.5

Annexin I

14 15

1 1

4 4

4.0

0.5 0.5

Proto-oncogene c-fos Apolipoprotein E2 *

4.0

The (*) designates the detection of a putative signal peptide, predicted using SignalP 3.0 server found at http://www.cbs.dtu.dk/services/SignalP/.

3.5. C-type lectin in the fish venom Group 9 in Table 1 includes three clusters (ESTs CO816130–CO816133, CO816461 and CO816545) matching with C-type lectins from the fish Cyprinus carpio, Salmo salar and Carassius auratus, respectively. C-type lectins are proteins of animal origin; they are calcium-dependent lectins that bind carbohydrates. Carbohydrate-binding depends on a highly conserved carbohydrate recognition domain [23]. Animal C-type lectins are involved in extracellular matrix organization, endocytosis, complement activation and also mediate pathogen recognition and cell–cell interactions [24]. Lectins are also present in venoms of various snakes and other poisonous animals [25]. A sialic acid-specific lectin from the horseshoe crab [26] and a galactose-specific lectin from the sea cucumber Cucumaria echinata [27] has direct hemolytic activity. After binding to specific carbohydrate chains on the erythrocyte surface, these lectins damage the cell membrane, leading to cell lysis. These lectins may play an important role against bacterial in-

fections or natural enemies. Echinoidin, a lectin in the venom of the spines of the sea urchin Toxopneustes pileolus, is mitogenic and chemotaxic, and may aid against predators [28]. Fig. 2 shows the alignment of the longest T. nattereri lectin EST (CO816131), echinoidin, and the N-terminal sequence of the myotoxin TmC4-47.2 of Thalassophryne maculosa [29]. T. nattereri lectin and echinoidin have similar putative galactose-specific CRDs (the substitution of Q for N is conservative). The high similarity with the N-terminus of the T. maculosa venom myotoxin [29] suggests that CO816131 encodes a toxic lectin. Indeed, a recently isolated fraction from T. nattereri venom with a Mw around 15 kDa presented a lectin-like activity towards erythrocytes, and this activity was inhibited by galactose (data not shown). 3.6. Chaperones to protect protein from degradation This group that accounted for 1.4% of the sequences contained four clusters that matched with Hsp70 (e-111). A cluster

Fig. 2. Alignment of the partial sequence of a T. nattereri C-type lectin EST (CO816131) with the toxic lectin from T. pileolus (echinoidin) and the N-terminal of T. maculosa myotoxin (TmC-47.2). Top box indicates the putative signal peptide. Bottom box indicates a putative galactose-specific carbohydrate recognition domain. The conserved cysteins residues between T. nattereri lectin and echinoidin are indicated by asterisks. Black and gray areas indicate amino acids that are identical and conserved, respectively. Dashes represent gaps introduced to improve the sequence alignment.

G.S. Magalhães et al. / Biochimie 88 (2006) 693–699

with two ESIs matched closely with Hsp90 (e-102) and two clusters (three ESTs) matched with chaperonins (e-033), especially with chaperone sequences of fish species. Hsp70, Hsp 90 and other chaperonins are a subset of a ubiquitous group of proteins that direct the folding and assembly of cellular proteins [30,31]. Redox-related enzymes included a cluster of two ESTs that matched (2e-013) with protein disulfide isomerase (PDI), an enzyme that breaks and re-forms disulfide bonds between cysteine residues and that affects the folding of many proteins. Such enzymes promote the exchange of paired disulfide bridges in proteins that contain multiple cysteine residues such as the natterins, and allow the proteins to attain the pattern of disulfide bonds that is important to maintain its conformation [32,33]. Another cluster of two ESTs matched with thioredoxin interacting protein (e-104), another enzyme that acts on redox reactions. This ubiquitous protein binds with high affinity to thioredoxin and inhibits its ability to reduce sulfhydryl groups via NADPH oxidation, and may help maintain an oxidizing environment that promotes the formation of disulfide bonds. Group 4 contained two clusters (CO815995–CO816001 and CO816122–CO816126) matching with clusterin (3e-061) from Danio rerio. This protein in its predominant form is a heterodimeric glycoprotein of 75–80 kDa that is secreted in plasma, milk, urine, cerebrospinal fluid and semen [34]. Its function is still not clear, but it is able to bind and form complexes with numerous partners, including immunoglobulins, lipids, heparin, complement components, paraoxonase, beta amyloid, and leptin [34]. A tempting hypothesis is that clusterin, like the small heat shock proteins, is an extracellular chaperone that can stabilize partly unfolded stressed proteins [35]. In addition, clusterin at physiological levels inhibits stress-induced precipitation of proteins in undiluted human serum [36]. We suggest that clusterin-like proteins in the venom may act in a chaperone-like manner to bind to hydrophobic regions of partly unfolded, stressed toxins, and thereby avoids their precipitation. T. nattereri can live in fresh and brackish waters (estuaries), and is highly resistant to periods of up to 12 h out of the water, which allows it to survive in the mud during the tidal cycle. These stressful conditions may affect the solubility of proteins in the cells of the venom gland, and clusterin-like peptides may help reduce this problem.

697

3.7. CART-like peptide in the cDNA library Another surprising finding was a single cluster (CO816062– CO816071) containing 10 clones (Table 1), which matched with cocaine- and amphetamine-regulated transcript (CART) from C. auratus (2e-021). CART was originally described as an mRNA induced in the brain after acute administration of cocaine and amphetamine [37]. CART has a major role in the regulation of feeding in mammals and in fish [38,39]. Fig. 3 shows the alignment CART of T. nattereri, goldfish, humans and rats. The identity between CART of T. nattereri and the other species was 37–40%. The C-terminal portion of the peptide is especially well conserved, which is not surprising because this region is thought to be the biologically active region. This region contains six conserved cysteine residues, which in the recombinant CART peptide are linked by disulfide bridges to form a compact structure [40]. CART is found not only in the mammalian brain, but also in the spinal cord, pituitary, gut, pancreas, and adrenals [41]. However, the effects of CART peptides in the periphery have received less attention. Whether the CART-related peptide in the T. nattereri venom gland affects feeding is not known. This fish uses its venom primarily for defense, and its release depends on a completely involuntary mechanical action, rather than the voluntary expulsion seen in spiders, snakes, and scorpions. The presence of CART suggests that the venom gland has additional functions besides the production of venom. Table 1 shows a group (#13) that matches with the protein c-Fos. Expression of this protein may be related to CART expression, since rats receiving recombinant CART have induced c-Fos expression in several hypothalamic and brainstem structures implicated in the control of food intake [42]. Whether CART induces c-Fos expression in the venom gland remains to be seen. 3.8. Ca2+-binding regulatory proteins Ca2+ plays a pivotal role in regulating cellular responses including cell metabolism, muscle contraction, secretion, cell cycle progression, gene expression, neurotransmission, and intracellular signal transduction [43]. These actions depend on a host of Ca2+-binding proteins [44]. The venom gland library contained a cluster that is related to the S100-like calcium

Fig. 3. Alignment of T. nattereri, goldfish, human and rat CART sequences. The boxes indicate the putative signal peptides (signal peptide for human CART was experimentally determined). Lines between cysteine residues indicate disulfide bonds. Black and gray areas indicate amino acids that are identical and conserved, respectively. Dashes represent gaps introduced to improve the sequence alignment.

698

G.S. Magalhães et al. / Biochimie 88 (2006) 693–699

Fig. 4. Alignment of putative venom toxins. The numbers on the left represent the cluster of each sequence. The box indicates the putative signal peptides. Black and gray areas indicate amino acids that are identical and conserved, respectively. Dashes represent gaps introduced to improve the sequence alignment.

binding protein, and another related to annexin I (Table 1), both of which are Ca2+ binding proteins. In the pancreas and salivary glands, increases in intracellular calcium concentration stimulate secretion [45]. Since S-100 protein has been identified in exocrine glands of several mammalian species, this protein is thought to play a role in secretion [46]. Annexin I can form heterocomplexes with members of the Ca2+-binding S100 protein family [47], and annexin I is known to play a role in insulin secretion [48]. Therefore, these members of the annexin family may be involved in secretion of toxins.

Although there are many sequences from other fishes in the databases, such as Takifugu rubripes, D. rerio, S. salar, Oryzias latipes and Gasterosteus aculeatus, this is the first time a fish venom gland transcriptome is described, which will help to support comparative studies for other fish venom glands. This approach also allowed us to find the natterins, a novel family of kininogenases unknown before our data, a toxic Ctype lectin and also pointed out the presence of putative new toxins. In addition, as many of the sequences are unknown, they will represent new data to the databases.

3.9. Unknown sequences that may encode venom toxins

Acknowledgements

Group 2 (Table 1) shows clusters that are not related to known sequences and contain a putative signal peptide, suggesting that they are secreted. This group is composed of three clusters, TNAT0012 (CO815883–CO815885), TNAT0074 (CO816108–CO816119) and TNAT0077 (CO816127– CO816129) containing three, 12 and three ESTs, respectively (2.3% of total transcripts). Fig. 4 shows the alignment of the translated clusters, and shows a highly conserved signal peptide and high overall identity (≈ 50%). Clusters TNAT0074 and TNAT0077 have nucleotide sequences coding for proteins with predicted molecular mass of 12.0–12.7 kDa and an isoelectric point around 3.7. The stop codon of cluster TNAT0012 was not identified, suggesting that the protein it encodes is longer than that encoded by the other two clusters. This incomplete sequence apparently also encodes an acidic protein. The acidic properties of these sequences are due to their high content (≈ 35%) of glutamic (E) and aspartic (D) acids. The abundance of these ESTs, their failure to match known sequences, and the presence of a signal peptide, suggest that they may code for novel toxins, as already observed for Natterin sequences.

This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Brazilian Research Council (CNPq) and Fundação Butantan.

4. Conclusion In this study we describe a transcriptome analysis of 775 ESTs which represents a general panorama of the transcripts responsible for the physiological events taking place in the venom gland, since those data were generated from a non-normalized cDNA library.

Annex A. Supplementary data Supplementary data associated with this article can be found in the online version at doi: 10.1016/j.biochi.2005.12.008. References [1] S. Karmakar, D.C. Muhuri, S.C. Dasgupta, A.K. Nagchaudhuri, A. Gomes, Isolation of a haemorrhagic protein toxin (SA-HT) from the Indian venomous butterfish (Scatophagus argus, Linn) sting extract, Indian J. Exp. Biol. 42 (2004) 452–460. [2] H.C. Ng, S. Ranganathan, K.L. Chua, H.E. Khoo, Cloning and molecular characterization of the first aquatic hyaluronidase, SFHYA1, from the venom of stonefish (Synanceja horrida), Gene 346 (2005) 71–81. [3] G. Ouanounou, M. Malo, J. Stinnakre, A.S. Kreger, J.T. Molgo, A neurosecretory protein isolated from stonefish (Synanceia trachynis) venom, forms nonselective pores in the membrane of NG108-15 cells, J. Biol. Chem. 277 (2002) 39119–39127. [4] S.T. Hahn, J.M. O’Connor, An investigation of the biological activity of bullrout (Notesthes robusta) venom, Toxicon 38 (2000) 79–89. [5] P. Garnier, F. Ducancel, T. Ogawa, J.C. Boulain, F. Goudey-Perrière, C. Perrière, A. Ménez, Complete amino-acid sequence of the β-subunit of VTX from venom of the stonefish (Synanceia verrucosa) as identified from cDNA cloning experiments, Biochim. Biophys. Acta 1337 (1997) 1–5. [6] F.G. Ghadessy, D. Chen, R.M. Kini, M.C.M. Chung, H.E. Khoo, K. Jeyaseelan, R. Yuen, Stonustoxin is a novel lethal factor from stonefish (Synanceja horrida) venom cDNA cloning and characterization, J. Biol. Chem. 271 (1996) 25575–25581.

G.S. Magalhães et al. / Biochimie 88 (2006) 693–699 [7]

[8]

[9]

[10]

[11]

[12]

[13]

[14] [15]

[16]

[17]

[18] [19]

[20] [21]

[22] [23] [24] [25]

[26]

[27]

L.A. Fonseca, M. Lopes-Ferreira, Clinical and experimental studies regarding poisoning caused by a fish Thalassophryne nattereri (niquim), An. Bras. Dermatol. 75 (2000) 435–443. M. Lopes-Ferreira, K.C. Barbaro, D.F. Cardoso, A.M. Moura-da-Silva, I. Mota, Thalassophryne nattereri fish venom: biological and biochemical characterization and serum neutralization of its toxic activities, Toxicon 36 (1998) 405–410. M. Lopes-Ferreira, A.M. Moura-da-Silva, A.A. Piran-Soares, Y. Ângulo, B. Lomonte, J.M. Gutierrez, S.H.P. Farsky, Hemostatic effects induced by Thalassophryne nattereri fish venom: a model of endotheliummediated blood flow impairment, Toxicon 40 (2002) 1141–1147. M. Lopes-Ferreira, J.A. Emim, V. Oliveira, L. Puzer, M.H. Cezari, S. Mda, L. Araujo, A.J. Juliano, C. Lapa, A.M. Souccar, Moura-da-Silva, Kininogenase activity of Thalassophryne nattereri fish venom, Biochem. Pharmacol. 68 (2004) 2151–2157. M.D. Adams, J.M. Kelley, J.D. Gocayne, M. Dubnick, M. H. Polymeropoulos, H. Xiao, C.R. Merril, A. Wu, B. Olde, R. F. Moreno, Complementary DNA sequencing: expressed sequence tags and human genome project, Science 252 (1991) 1651–1656. G.S. Magalhães, M. Lopes-Ferreira, I.M.J. Junqueira de Azevedo, P. J. Spencer, M.S. Araújo, F.C.V. Portaro, L. Ma, R.H. Valente, L. Juliano, J.W. Fox, P.L. Ho, A.M. Moura-da-Silva, Natterins, a new class of proteins with kininogenase activity characterized from Thalassophryne nattereri fish venom, Biochimie 87 (2005) 687–699. M.J. Paine, H.P. Desmond, R.D. Theakston, J.M. Crampton, Gene expression in Echis carinatus (carpet viper) venom glands following milking, Toxicon 30 (1992) 379–386. B.W. Halstead, Poisonous and Venomous Marine Animals of the World, Ed. US Government Printing Office, Washington, 1970 (v.3). I.M.J. Junqueira de Azevedo, P.L. Ho, A survey of gene expression and diversity in the venom glands of the pitviper snake Bothrops insularis through the generation of expressed sequence tags (ESTs), Gene 299 (2002) 279–291. J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: a Laboratory Manual, second ed, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1989. G.R. Lazo, J. Tong, R. Miller, C. Hsia, C. Rausch, Y. Kang, O. D. Anderson, Software scripts for quality checking of high throughput nucleic acid sequencers, Biotechniques 30 (2001) 1300–1305. X. Huang, A. Madan, CAP3: a DNA sequence assembly program, Genome Res. 9 (1999) 868–877. S.F. Altschul, T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller, D.J. Lipman, Gapped BLAST and PSI-BLAST: a new generation of protein database search programs, Nucleic Acids Res. 25 (1997) 3389– 3402. J.D. Bendtsen, H. Nielsen, G. von Heijne, S. Brunak, Improved prediction of signal peptides: SignalP 3.0, J. Mol. Biol. 340 (2004) 783–795. C. Notredame, D.G. Higgins, J. Heringa, T-coffee: a novel method for fast and accurate multiple sequence alignment, J. Mol. Biol. 302 (2000) 205–217. W.R. Pearson, Identifying distantly related protein sequences, Comput. Appl. Biosci. 13 (1997) 325–332. K. Drickamer, Two distinct classes of carbohydrate-recognition domains in animal lectins, J. Biol. Chem. 263 (1988) 9557–9560. W.I. Weis, M.E. Taylor, K. Drickamer, The C-type lectin superfamily in the immune system, Immunol. Rev. 163 (1998) 19–34. T. Ogawa, T. Chijiwa, N. Oda-Ueda, M. Ohno, Molecular diversity and accelerated evolution of C-type lectin-like proteins from snake venom, Toxicon 45 (2005) 1–14. P.B. Armstrong, S. Swarnakar, S. Srimal, S. Misquith, E.A. Hahn, R. T. Aimes, J.P. Quigley, A cytolytic function for a sialic acid-binding lectin that is a member of the pentraxin family of proteins, J. Biol. Chem. 271 (1996) 14717–14721. T. Uchida, T. Yamasaki, S. Eto, H. Sugawara, G. Kurisu, A. Nakagawa, M. Kusunoki, T. Hatakeyama, Crystal structure of the hemolytic lectin CEL-III isolated from the marine invertebrate Cucumaria echinata: implications of domain structure for its membrane pore-formation mechanism, J. Biol. Chem. 279 (2004) 37133–37141.

699

[28] Y. Giga, A. Ikai, K. Takahashi, The complete amino acid sequence of echinoidin, a lectin from the coelomic fluid of the sea urchin Anthocidaris crassispina. Homologies with mammalian and insect lectins, J. Biol. Chem. 262 (1987) 6197–6203. [29] J.I. Sosa-Rosales, G. D’Suze, V. Salazar, J. Fox, C. Sevcik, Purification of a myotoxin from the toadfish Thalassophryne maculosa (Gunter) venom, Toxicon 45 (2005) 147–153. [30] M.J. Gething, J. Sambrook, Protein folding in the cell, Nature 355 (1992) 33–45. [31] W.J. Welch, The role of heat-shock proteins as molecular chaperones, Curr. Opin. Cell Biol. 3 (1991) 1033–1038. [32] R.B. Freedman, T.R. Hirst, M.F. Tuite, Protein disulphide isomerase: building bridges in protein folding, Trends Biochem. Sci. 19 (1994) 331–336. [33] R.B. Freedman, P.J. Gane, H.C. Hawkins, R. Hlodan, S.H. McLaughlin, J.W. Parry, Experimental and theoretical analyses of the domain architecture of mammalian protein disulphide-isomerase, Biol. Chem. 379 (1998) 321–328. [34] S.E. Jones, C. Jomary, Clusterin, Int. J. Biochem. Cell Biol. 34 (2002) 427–431. [35] D.T. Humphreys, J.A. Carver, S.B. Easterbrook-Smith, M.R. Wilson, Clusterin has chaperone-like activity similar to that of small heat shock proteins, J. Biol. Chem. 274 (1999) 6875–6881. [36] S. Poon, S.B. Easterbrook-Smith, M.S. Rybchyn, J.A. Carver, M. R. Wilson, Clusterin is an ATP-independent chaperone with very broad substrate specificity that stabilizes stressed proteins in a folding-competent state, Biochemistry 39 (2000) 15953–15960. [37] J. Douglass, A.A. McKinzie, P. Couceyro, PCR differential display identifies a rat brain mRNA that is transcriptionally regulated by cocaine and amphetamine, Neurosci. 15 (1995) 2471–2481. [38] H. Volkoff, R.E. Peter, Effects of CART peptides on food consumption, feeding and associated behaviors in the goldfish, Carassius auratus: actions on neuropeptide Y- and orexin A-induced feeding, Brain Res. 887 (2000) 125–133. [39] P. Kristensen, M.E. Judge, L. Thim, U. Ribel, K.N. Christjansen, B. S. Wulff, J.T. Clausen, P.B. Jensen, O.D. Madsen, N. Vrang, P.J. Larsen, S. Hastrup, Hypothalamic CART is a new anorectic peptide regulated by leptin, Nature 393 (1998) 72–76. [40] L. Thim, P.F. Nielsen, M.E. Judge, A.S. Andersen, I. Diers, M. Egel-Mitani, S. Hastrup, Purification and characterization of a new hypothalamic satiety peptide, cocaine and amphetamine regulated transcript (CART), produced in yeast, FEBS Lett. 428 (1998) 263–268. [41] P.R. Couceyro, E.O. Koylu, M.J. Kuhar, Further studies on the anatomical distribution of CART by in situ hybridization, J. Chem. Neuroanat. 12 (1997) 229–241. [42] N. Vrang, M. Tang-Christensen, P.J. Larsen, P. Kristensen, Recombinant CART peptide induces c-Fos expression in central areas involved in control of feeding behaviour, Brain Res. 818 (1999) 499–509. [43] M.K. Bennett, Ca2+ and the regulation of neurotransmitter secretion, Curr. Opin. Neurobiol. 7 (1997) 316–322. [44] I. Niki, H. Yokokura, T. Sudo, M. Kato, H. Hidaka, Ca2+ signaling and intracellular Ca2+ binding proteins, J. Biochem. (Tokyo) (1996) 685–698. [45] R.M. Case, T.A. Ansah, S. Dho, A. Miziniak, L. Wilson, Calcium homeostasis in exocrine secretory cells, in: C.H. Gerday, R. Gilles, L. Bolis (Eds.), Calcium and calcium binding proteins. Molecular and functional aspects, Springer, Berlin Heidelberg New York, 1988, pp. 211–219. [46] H. Haimoto, S. Hosoda, K. Kato, Differential distribution of immunoreactive S-100a and S-100b proteins in normal nonnervous human tissues, Lab. Invest. 57 (1987) 489–498. [47] W.S. Mailliard, H.T. Haigler, D.D. Schlaepfer, Calcium-dependent binding of S100C to the N-terminal domain of annexin I, J. Biol. Chem. 271 (1996) 719–725. [48] M. Ohnishi, M. Tokuda, T. Masaki, T. Fujimura, Y. Tai, T. Itano, H. Matsui, T. Ishida, R. Konishi, J. Takahara, Involvement of annexin-I in glucose-induced insulin secretion in rat pancreatic islets, Endocrinology 136 (1995) 2421–2426.