Biochemical and Toxinological Characterization of Venom from Macrorhynchia philippina (Cnidaria, Hydrozoa)

Macrorhynchia philippina is a colonial benthic hydroid from the Class Hydrozoa (Phylum Cnidaria) distributed in the tropical and subtropical marine waters from Atlantic Ocean, Indo-Pacific, and Mozambique. Its colonies somewhat resemble plants, causing confusion in the bathers who accidentally touch the animal. Acute burning/local pain, edema, erythema, and pruritus were symptoms already described, but its venom composition is unknown, as well as the participation of toxins for the symptom's development. Thus, herein, we show the biochemical composition and toxic effects of M. philippina venom. Colonies were collected and processed for histological analysis; alternatively, they were immersed into methanol containing 0.1% acetic acid for venom attainment, which was analyzed by mass spectrometry and submitted to edema and nociception evaluation in mice, hemolysis and antimicrobial assays in vitro. Before the molecule's extraction, it was possible to see the inoculation structures (hydrocladiums and hydrotheca) containing venom, which was released after the immersion of the animal in the solvents. The venom was composed mainly by low molecular mass compounds, able to cause significant reduction of the paw withdrawal latency from the hot plate test, 30 minutes after the injection. Moreover, significant edema was observed 10 and 30 minutes after the injection, indicating the activity of at least two inflammatory mediators. The venom caused no hemolytic activity but reduced the growth of A. baumannii and K. pneumoniae strains. This study is the first biochemical description of M. philippina venom, with molecules that cause fast inflammatory and painful effects, characteristic of the envenomation.


Introduction
The Phylum Cnidaria is a highly diverse group that includes the Class Anthozoa, Cubozoa, Hydrozoa, Myxozoa, Scypho-zoa, and Staurozoa. All of them have venomous representatives, some with potent toxins for humans, injected in the victim through an inoculation apparatus, named nematocysts [1]. (Figure 1) specimens were collected by SCUBA diving at Ponta do Jarobá, São Sebastião, São Paulo, Brazil (23°49 ′ 44 ″ S; 45°25 ′ 23 ″ W) under the ICMBio license # 68917-1. Upon sampling, colonies were washed with filtered sea water and other animals and plants were manually removed.

Animal's Morphology.
After cleaning, colonies were immediately fixed with Karnovsky's fixative solution (glutaraldehyde 3%, paraformaldehyde 3% in 0.1 M phosphate buffer, pH 7.4) for 48 hours. After fixation, the animals were examined using a light microscope and, in parallel, some colonies fragments were dehydrated in crescent ethanol series (70, 95, and 100%) and submitted to a xylene clarification process using an automatic tissue processor (Lupetec PT 09 TS) and embedded in paraffin (Lupetec DP 2010). Transversal or longitudinal 2 μm sections were obtained using the microtome Lupetec MRPO3. Histological slides were deparaffinized in xylene, hydrated in decreasing concentrations of ethanol (100%, 95%, 80%, and 70%), hydrated in distilled water, and stained with hematoxylin-eosin (HE) for 5 and 3 minutes, respectively, and mounted with a slide and synthetic Canada balsam for general morphological observation under the light microscope (Nikon SE).

Venom Extraction.
Fresh specimens of M. philippina were immersed in methanol containing 0.1% acetic acid for 24 hours at room temperature. Resulting extract was centrifuged at 10,000 × g for 10 minutes, and the supernatant was submitted to a rotary evaporator for solvent removal. The remaining content was then dissolved in the appropriate solution for subsequent analysis, described in the correspondent sections.

Biochemical Characterization.
In order to verify the presence of proteins, M. philippina venom (50 μg/mL) was submitted to an electrophoresis in a 12% polyacrylamide gel containing SDS (SDS-PAGE) performed under reducing conditions. Gel was later stained by silver.
Alternatively, the M. philippina venom (50 μg/mL diluted in 100 mM Tris-HCl, pH 8.5) was added to 8 M urea or 15 minutes at 80°C, followed by the addition of dithiothreitol 100 mM for 30 minutes at 60°C and 300 mM iodoacetamide for 30 minutes at room temperature, protected from light. The solution was diluted with 100 mM Tris-HCl pH 8.5 to reach urea concentration to 2 mM and digested with 10 ng/μL of trypsin (Proteomics Grade, from porcine pancreas, Sigma-Aldrich) overnight, at 37°C. The reaction was stopped by the addition of 5% formic acid, and the material was lyophilized to be analyzed in the mass spectrometry.
The raw data was analyzed by PEAKS®7.0 software (Bioinformatics Solution Inc., Canada) for protein identification using the following parameters: MS and MS/MS error mass 10 ppm and 0.01 Da; methionine oxidation and carbamidomethylation as variable and fixed modification, respectively; trypsin as cleavage enzyme; 3 maximum missed cleavages and 3 maximum variable PTMs per peptide; and the false discovery rate ≤ 0:5%. All data were analyzed in a Cnidaria Database, built by retrieving all UniProt entries associated with this term.
In parallel, the venom was loaded into a C18 column (ACE C18, 5 μm, 100 Å, 2:0 mm × 50 mm) coupled to a reversed-phase ultraperformance liquid chromatography (Acquity, Waters co., USA) to verify the complexity of the sample and the molecular mass distribution.
The elution was performed in a two Equipment control and data acquisition were conducted using the MassLynx 4.2. Raw files were manually verified and processed using PEAKS®7.0 (Bioinformatics200 Solution Inc., Canada) for peptide de novo sequence and Progenesis QI Software (Waters Co., USA) for small molecules, which were identified by exact molecular mass, m/z, and spectra similarity using the GNPS and Vaniya/Fiehn Natural Products databases. For both, 10 ppm was used as precursor and fragment tolerance. Moreover, ions previously described for M. philippina were manually searched in the peak list, in order to find known molecules in our extract. Animals were placed into microisolators, in a room set at a 12/12 h light/dark cycle, with a 70% air humidity, and a  Results were expressed as the difference (%) of thickness between M. philippina-injected paw and control paws.  The inoculum was added in Mueller Hinton Agar plates, and discs of filter-paper were placed in the plates. Ten microliters of M. philippina venom (10 mg/mL) was pipetted in the discs. The results were interpreted by visual inspection, by observing the inhibition halo formed around the discs.
2.6. Statistical Analysis. Data were evaluated using GraphPad Prism to verify statistical differences between groups. Oneway ANOVA test, followed by Tukey posttest, was applied, and differences were considered significant when p < 0:05.

Macrorhynchia philippina Morphology.
To study M. philippina morphology and understand its possible source of toxins, its colonies were firstly observed by light microscopy after fixation, without staining. Figure 2 shows details of the hydrocaulus, axial structure, and its ramifications, the hydrocladiums. Figure 2(a) depicts the longitudinal view of the basal portion of the hydrocaulus in the first plane, with the hydrocladiums behind. A detail of the hydrocaulus is shown in Figure 2(b), where it is possible to note the secretion inside the structure. In a transversal view (Figure 2(c)), it is possible to see that the hydrocaulus is composed of several canals, from the tip to the base of the structure (data not shown).
The hydrocladium magnification is shown in Figure 2(d), where more ramifications are present in the tip of the structure, named hydrotheca. In this structure, the secretion is stored (black spots in the tip). The detail of hydrotheca, in a lateral view, is shown in Figures 2(e) and 2(f), where the content is ready to be released. When the structures were stained with HE (Figures 3(a)  and 3(b)), we could confirm the presence of canals, as observed before (Figure 2(c)). Apparently, the hydrocaulus is composed of structural proteins, such as keratin, that sus-tains all the structure (as a stalk). Nevertheless, it is possible to see a secretion inside the structure, where probably the secretory cells are present as well (Figures 3(c) and 3(d)).
To confirm the release of secretion after immersing the animal in methanol/acetic acid, we analyzed the morphology before and after the procedure. The secretion before the extraction was present in the hydrocladiums and hydrothecas (Figure 4(a)), where it is possible to see black spots. However, after 24 h immersion in methanol and acetic acid, no more secretions were visible in these structures, suggesting that its contents were completely released (Figure 4(b)). This content was analyzed by its composition and biological activity.

Biochemical
Composition of the Venom. The general biochemical profile of the M. philippina venom is shown in Figure 5(a), which shows peaks distributed along the acetonitrile gradient, being more abundant in the beginning, in a more hydrophilic elution. Moreover, according to the ion intensity map (Figure 5(b)), several ions between 100 and 1200 m/z were detected, being abundant in the 100-600 m/ z region and between 0 and 15 minutes of gradient. This data indicates the abundance of low molecular mass compounds in the secretion.
Overall, 19 peptides were de novo sequenced (Table 1). When they were compared to the database (Blast analysis), no similarity with toxins from animal venoms were identified using BLAST search, indicating undescribed sequences. In addition, no proteins were visualized in the SDS-PAGE, even after stained with silver (data not shown) and only 5 proteins were identified by mass spectrometry, being 4 of them uncharacterized and 1 dynein heavy chain 12 from Hydra vulgaris (accession number tr|T2MAX4|T2MAX4_HYDVU). On the other hand, 210 small molecules were identified, using the natural products and toxins database (Supplementary material (available here)). This data corroborated with the abundance of ions found around 500 m/z, shown in Figure 5(b). Ions and predicted molecular formula (obtained by the isotopic distribution) were compared to molecules previously described for hydroids, and then, we could find five known molecules, shown in Table 2. 3.3. Biological Activities. The M. philippina venom caused no hemolytic/cytotoxic activity in the tested doses, the same used for in vivo studies. However, in terms of its nociceptive activity, Figure 6 shows the time-course evaluation of paw withdrawal latency, in which it is possible to see that after 30 minutes of the secretion injection, a significant reduction of latency was observed, indicating a painful stimulus of the M. philippina venom.
Moreover, the venom induced a mild paw edema with two significant peaks-5/10 minutes and 30 minutes after injection. The edema was no longer visible after 45 minutes (Figure 7).
The 100 μg M. philippina venom caused growth inhibition of A. baumannii and K. pneumoniae, observed as a halo around the discs. Although it is a high concentration, it was enough to identify an antibacterial effect, very important considering that these strains are resistant to antibiotics.

Discussion
Until now, the Phylum Cnidaria comprises more than 11000 species, according to the World Register of Marine Species (WorMS). They have one feature in common: the presence of nematocysts, a structure that releases a mixture of toxins, such as proteins and peptides, for prey capture and digestion, defense/predator repelling, locomotion, and intra-/ interspecies spatial competition [17]. Although many cnidarians have this specialized structure, they may differ in shape and size according to the class. As they are full of venom, they can also cause human envenomation [15].
In general, human accidents caused by cnidarians are characterized by local repercussions and systemic, depending on the specie. Local symptoms are diverse, ranging from mild pain and itching to excruciating pain with necrosis and scarring, according to the amount of toxins injected and the venom-containing substances, cnidarian species, and its geographic distribution [18]. Cutaneous eruption can become persistent or generalized [17].
Accidents involving benthic hydroids are neglected, as such animals usually do not cause severe reactions and the venom action is apparently restricted to the site of the puncture [2]. According to Rifkin et al. [3], accidents with M. philippina are characterized by burning local pain, edema, erythema, and pruritus. Although these are mild symptoms, this specie has been recognized as health risk of population by envenomation in the Mediterranean region [4].
Similar patterns of envenomation have been related for other hydroids, for example Nemalecium lighti (Haleciidae family) that causes burning pain sensation, erythema, edema, and pruritus once in contact with the skin [19]. The Christmas-tree-hydroid Pennaria disticha also causes intense pain, burning sensation, erythema, and edema [20].
Although there are reports of envenomation, the biochemical nature of hydroids' venom is poorly explored, as     [3] showed that the administration of vinegar triggered the discharge of the nematocysts (cells that harbor the venom) from Lytocarpus philippinus. Similarly, [21] used ethanol to induce nematocyst discharge to assess the venom content of the box jellyfish Chironex fleckeri.
The extraction showed here was confirmed by microscopy that enabled the visualization of empty structures and later corroborated by the biochemical and biological analysis. Moreover, the histology of colonies of M. philippina revealed a semirigid structure entirely filled with secretion, which is released in the tip (hydrotheca) [2].
Data from our group showed that methanol containing 0.1% acetic acid was enough to induce the extraction of molecules from Chiropsalmus quadrumanus and Olindias sambaquiensis [22,23]. Indeed, methanol and acetic acid were enough to extract several small molecules from M. philippina with mild polarity that would not be dissolved in the sea water. The presence of higher masses, such as 1000 to 1500 m/z may correspond to peptides, which need to be investigated. Nevertheless, the abundance of molecules below 600 m/z is evident.
The biochemical composition of the most studied cnidarians has a protein nature. In general, hydrozoans have a predominance of phospholipases, cytotoxins, and hemolysins [24], whereas its jellyfishes have metallopeptidases, neurotoxins, and serine peptidases [9]. Also, the venom of Tubastraea coccinea was determined to harbor several neurotoxins, cytolysins, and dyshomeostasis [25]. On the other hand, the venom of M. philippina does not contain protein toxins, and the toxic effects were elicited by low molecular mass compounds. This data explains the lack of cytotoxic/ hemolytic activity by M. philippina in opposite of similar cnidarians.
In terms of peptides, although 19 were identified herein, none of them is described in the literature. Peptides have been reported in the peptidome of the sea anemone Phymanthus crucifer, with several biological activities associated, which revealed 504 natural peptides secreted [26]. In Hydra and other cnidarians, peptides play an important role in the development as hormones and neurotransmitters, regulating reproduction, metamorphosis, muscle contraction, sensorial activity, among other crucial biological activities [27]. On the other hand, in the same phylum exists an important group of nonpeptidic toxins, well described for anthozoans [28].
Among the nonpeptidic toxins, we can list sarcophine-a, a cembranolide from Sarcophyton glaucum. This diterpene toxin was isolated from a hexane fraction and displayed muscarinic activities [29] and inhibits phosphofructokinase [30]. Regarding neurotransmission, lophotoxin is a furanocembranolide isolated by chloroform extraction from Leptogorgia sp. and has blocking activity on nicotinic acetylcholine receptors [31]. However, the most famous nonpeptidic marine toxin is the potent palytoxin, isolated by water/ethanol extraction from Palythoa sp. [32]. This toxin belongs to a family of polyhydroxylated (polyol) toxins that range between 2659 and 2680 Da, and blocks Na + /K + -ATPase channels as a toxic mechanism [33].
In Millepora complanata (Hydrozoa), a fraction of an aqueous extract had the presence of at least four nonpeptidic thermostable toxins that induced a potent vasoconstriction, hemolysis, and death in mice. Spectrometric analysis revealed that these toxins are polyoxygenated alkylbenzenes [34].
Here, we could find only five molecules already described from M. philippina among the 210 identified: isololiolide, Macrophilones and lytophilippines.
An isololiolide described by [35] had antitrypanosomal activity, without inducing cytotoxicity for mammalian cells. Cytotoxicity was observed by an isololiolide described by Vizetto-Duarte et al. [36], with apoptosis in tumor cell line. Zlotkowski et al. [37] showed that Macrophilone A increased intracellular reactive oxygen species levels and caused cytotoxicity in lung adenocarcinoma cells. Further studies showed that other Macrophilones caused cytotoxicity as well [38]. Lytophilippines were able to induce antibacterial, antiviral, and antitumor activities, besides displaying lethality in shrimps [39].
Antibacterial activity was identified here, against A. baumannii and K. pneumoniae strains, both gram-negative bacteria. K. pneumoniae is resistant to a broad spectrum of antibiotics and then causes serious health infection problems [40]. A. baumannii is another opportunistic bacterium, also resistant to antibiotics and frequently associated to respiratory infections [41]. The action of M. philippina venom on those strains indicates a different mechanism of bacterial death or growth inhibition, important as a chemical defense for the animal and, consequently, the maintenance of the specie in a hostile environment. Moreover, the discovery of new antibacterial molecules is extremely relevant for the clinics and biomedical field, considering the high incidence of antibiotic-resistant infections observed for several strains.
Considering that M. philippina accident causes pain and edema, we used animal models of mammals to evaluate these effects. We could observe a reduction of latency to pain behavior, indicating a hyperalgesia of the extract. The heat induces the activation of heat capsaicin-gated channel TRPV-1 by the release of inflammatory mediators, such as bradykinin, prostaglandins, chemokines, and ATP [42].
Besides activation of nociceptors, bradykinin is an inflammatory mediator that causes increase of vascular permeability, which increases the plasma flow to the extravascular compartment and causes edema. The action of such mediator is fast [43], explaining the effect after only 10 minutes of M. philippina venom injection.
Overall, the edema had two peaks-10 minutes and 30 minutes after the injection. This profile indicates the participation of more than one inflammatory mediator. Thus, besides bradykinin, other molecules might be acting to cause the inflammatory reaction.
Local inflammation is developed by the action of prostaglandins (PGs), histamine, serotonin, bradykinin, leukotrienes, and nitric oxide [44]. In the mouse edema examined herein, the fast inflammatory action (up to 30 minutes) indicates the release of stored mediators, such as histamine and serotonin. However, it is important to mention that the fast inflammatory effect summed to hyperalgesia is compatible with the envenomation profile of venom from M. philippina.

8
BioMed Research International

Conclusion
In conclusion, the M. philippina venom contains peptides and several low molecular mass compounds, able to cause fast inflammatory and nociceptive reactions, characteristic of envenomation, and antimicrobial activity, important for animal's defense.

Data Availability
Data is available in supplementary information files.

Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this article.