*1.5. Amphibian*

Although a witch's recipe benefits from venomous animals, the toe of a frog and the eye of a newt would definitely spice things up. Shakespeare's Macbeth (Act 4, Scene 1) contains a recipe for a witch's brew that goes as follows:

"Fillet of a fenny snake, In the cauldron boil and bake; Eye of newt and toe of frog, Wool of bat and tongue of dog, Adder's fork and blind-worm's sting, Lizard's leg and owlet's wing, For a charm of powerful trouble, Like a hell-broth boil and bubble."

Although most of the above referred ingredients can be traced back to herbs (eye of newt = mustard seed (*Sinapis alba*); toe of frog = buttercup (*Ranunculus acris* L.); wool of bat = holly leaves (*Ilex aquifolium*); tongue of dog = gypsyflower from the genus hound's tounge (*Cynoglossum officinale* L.); adder's fork = least adder's-tongue (*Ophioglossum lusitanicum* L.); blind-worm = slowworm (*Anguis fragilis*)), a really mighty witch might as well as go on literally, seeing the true herpetological powers needed for the spell.

According to Table 1, published papers on amphibian venoms are less common than the triad. The similar figure to scorpion papers is due to two characteristics belonging to the study of the amphibian skin secretion: (i) the discovery of magainin, the first antibiotic peptide by [76], which boosted the literature by making several researchers seek other antibiotic peptides in other species, and (ii) the vast Chinese literature on Chan'Su, the all healing Chinese traditional medicine. These two events have undoubtfully contributed to these numbers. However, in general, amphibian literature on accidents is scarce in comparison to venomous animals.

The amphibian defense strategy against predators/aggressors is the "passive" defense (with the exception of *Rhaebo guttatus*, which is capable of voluntarily compressing its parotoid glands and ejecting its contents [77]), and the chemical nature of their venom is mainly protein/peptide toxins and low-molecular-mass compounds (such as alkaloids, steroids and their respective derivatives).

Some of the authors of the current review have been working with amphibian skin secretion for almost twenty years. As a consequence, they have been able to produce consistent literature on the subject that encompasses the different classes of bioactive molecules commonly found on the amphibian skin secretion. A compilation of these results will be presented below, together with the related literature.

Conceição, et al. [78] have evaluated the skin secretion of the tree frog *Phyllomedusa hypochondrialis* and described that this secretion presents proteins ranging between 68 and 14 kDa, and that proteolytic and phospholipase A2 activities could be detected in vitro. Moreover, authors also report that the injection of 0.6 ug of the venom in mice induced myotoxicity, as evaluated by the increase of creatinine-kinase activity in plasma. The same dose of the skin secretion also elicited vasoconstriction (for 20 min) and leukocyte rolling, as assayed by intravital microscopy. Edema and nociception, in a dose–response manner, could also be observed. Interestingly, the observed vascular permeability alterations displayed a different mechanism, in which the lowest tested concentration caused the most intense effect, in comparison to larger concentrations. This phenomenon is mostly likely due to the presence of different molecules, in distinct relative concentrations, acting on independent biological systems.

As a consequence of the described leukocyte rolling effect, a subsequent study was performed [79] that assessed the mechanisms involved in that effect. Experiments revealed that the toxins could lead to edema formation, within 2 h, which lasted for 24 h. Moreover, authors also described that the numbers of rolling and adherent leukocytes were augmented in post-capillary venules. Cytological analysis showed that macrophages were the main cells present 2 h after the injection, whereas neutrophils were the cells present after 24 h. The cytokine profiles indicted elevated levels of chemokines MCP-1 and KC, and also IL-6 and PGE2.

Mendes et al., 2016 [80], studied the casque-headed tree frog *Corythomantis greening*, a frog bearing a cranial bone adaptation used in phragmosis. The cutaneous secretion of this animal was able to induce inflammation (edema, for 96 h after the injection) and nociception. Moreover, relevant enzymatic activities were detected in the skin secretion, such as fibrinogenolytic, hyaluronidasic and metallopeptidasic. Enzymes presenting such activities have already been described as important toxins for snake venoms [81,82] and were also described in some amphibians from different genus, for example phospholipase in *Pithecopus azureus* [83] and serine peptidases in *Duttaphrynus melanostictus* [84]. Furthermore, Fusco et al. 2020 [85] studied the epidermal secretion of *Argenteohyla siemersi* and described both phospholipasic and hemolytic activities. They also reported that that venom

is cytotoxic and capable of promoting necrosis which is independent of the proteolytic activity, a different activity pattern from *C. greeningi* (included in the same genus).

Targeting antibiotic peptides—a consequence of Zasloff's study—Conceição et al., 2006 [86], screened the skin secretion of *P. hypochondrialis* for antimicrobial peptides against Gram-positive and -negative bacteria and successfully described Phylloseptin-7 and Dermaseptin (DPh-1). These peptides were active over common pathogens, such as *Staphylococcus aureus*, *Escherichia coli*, *Pseudomonas aeruginosa* and *Micrococcus luteus*. In a complementary study, Huang and collaborators identified a new Dermaseptin from the same *P. hypochondrialis* (Dermaseptin-PH), which was active against Gram-positive/-negative bacteria and inhibited biofilm formation. This peptide was also effective against *Candida albicans.*

Other authors also reported complementary phylloseptins. For example, Wu et al., 2017 [87], isolated PNS-PC from *P. camba* the PNS-PC. This peptide displays inhibitory action against Methicillin-resistant *Staphylococcus aureus*. They also isolated PBa1–3 from *P. Burmeister*, a peptide with antibacterial and antifungal activities [88]. A recent study by Liu et al., 2020 [89], reported the antibacterial activity of PV-1, a Phylloseptin from *P. vaillantii* in vitro and in vivo. In spite of observed hemolysis (in vitro), this peptide was not toxic to hepatic and renal tissues in vivo, indicating the possible therapeutical potential of this peptide for bacterial infection.

Zhang et al., 2010 [90], isolated Phylloseptin-1 (PSN-1) from *P. sauvagei*. This peptide was active against *Staphylococcus aureus* in vitro, including bacterial biofilm formation inhibition. A few years later, Raja et al., 2013 [91], described five more Phylloseptins displaying antimicrobial activity from this species. Their work proved that the structural differences among those peptides were responsible for the different observed bactericidal potency, suggesting that the alpha-helix amphipathic conformation leads to microbial membrane disruption.

Using Zasloff's classic strategy, Conlon et al. 2007 [92] stimulated *Hylomantis lemur* skin secretion with norepinephrine and successfully purified Dermaseptin-L1 and Phylloseptin-L1, which were active against Gram-negative bacteria and *Batrachochytrium dendrobatidis*, a fungus that infect frogs.

In 2009, an unexpected antimicrobial peptide was described by Sousa et al. [93]. Leptoglycin, a peptide comprised basically by Leu and Gly (with an import Pro at the center of the sequence) was isolated from the skin secretion of *Leptodactylus pentadactylus* and was active against Gram-negative bacteria.

Bradykinin-potentiating peptides are protagonists of the most important example of drug discovery from animal venoms. Rocha e Silva's discovery of bradykinin [94] ultimately led to the discovery of the bradykinin-potentiating peptides (BBPs) from snake venoms. Such a peptide, on the other hand, led to the development of Captopril, the first drug belonging to angiotensin-converting enzyme inhibitor (ACEi) class, widely used around the world to treat arterial hypertension. In another unexpected study, Conceição et al., 2007 [78], described the first canonical BPP isolated from another source than snake venoms. Phypo-Xa, a decapeptide isolated from *P. hypochondrialis,* inhibited ACE and potentiated bradykinin both in vivo and in vitro. A few years later, those authors [95] also isolated three bradykinin-related peptides from *P. nordestina* skin secretion: two were vasodilators (Pnor3 and Pnor7) and one was a vasoconstrictor (Pnor5).

Some amphibians, particularly toads, can be considered major biological sources of low-molecular-mass compounds, such as alkaloids and steroids. Tempone et al. [96], through biomonitored assays, have isolated two bufadienolidc steroids displaying antiparasitic activity from the skin secretion of *Rhinalla jimi*. Telecinobufagin and hellebrigenin were not new molecules at that time; however, the activity against *Leishmania* sp. promastigotes and amastigotes in macrophage culture (without NO production modulation) and the anti-*Trypanossoma cruzi* trypomastigotes activity were the novelties they reported in their paper. The mechanism of action of these molecules seems to be related to the disturbance of cellular membrane and mitochondrial function. Neither steroid presented hemolytic or cytotoxic activities in the tested conditions.

That same group of authors [97] later assayed the skin secretion of *P. nordestina* on antiparasitic models. They were able to demonstrate that four antimicrobial peptides (Dermaseptins 1 and 4, and Phylloseptins 7 and 8) were able to decrease the in vitro viability of *T. cruzi*, with a high theoretical therapeutic index. The proposed mechanism of action of the peptides is cell death induction, through cellular membrane permeabilization. Phylloseptin-7 was also effective against *Leishmania* sp.

Such results (selective membrane permeation) convinced Sciani et al. [98] to investigate the possible antitumor activities of the skin secretion of some Brazilian toads. MCF-7 and MDA-MB-231 lineages (breast tumor) displayed reduced proliferation and apoptosis induction when treated with eight different amphibian skin secretions. Among them, the most promising results came from *R. guttatus*, *R. margaritifera* and *P. hypochondrialis*. Moreover, *R. guttatus* and *R. marina* displayed selective antitumor activity over HL-60 (leukemia lineage), without toxicity to human leukocytes. It is believed that the observed antiproliferative effect is due to the known presence of bufadienolides in this toad secretion.

Schemda-Hirschmann in 2014 [99] related the presence of argininyl bufadienolides in *R. schneideri* dermic secretions, which were active on different tumor lineages AGS, SK-MES-1, J82 and HL-60 (gastric adenocarcinoma, lung carcinoma, bladder carcinoma and leukemia, respectively). Later, the same group showed similar activity in the Peruvian *R. marina* venom, and the mechanism of action seems related to ROS production and cell cycle arrest, for breast cancer lineages [100]. Antitumor properties were also described for the Paraguayan *Rhinella* sp. Such skin secretion is traditionally used by locals in folk medicine to treat skin lesions and tumors [101].

The crude extract of *Physalaemus nattereri* is cytotoxic for the B16F10 melanoma cell line. Carvalho et al. [102] observed that the secretion was able to induce conformational changes in cells, exposure of phosphatidylserine on cell membrane, reduction of mitochondrial membrane potential and arrest of cell cycle in S phase, indicating that apoptosis is the probable mechanism of action that explains the antitumor activity. RP-HPLC fractionated *P. nattereri* extract points out that this biological action is due to peptides

Skin venom from the Malaysian toad *B. asper* was active against HCT 116 colorectal tumor line by apoptosis induction, via caspase 3/7 activation and mitochondrial membrane potential disruption [103]. Bufadienolides also possess the ability to inhibit Na+/K+ ATPase and trigger caspase-induced apoptosis, being more selective to cancer cells than normal cells [104]. The venoms of two Turkish Salamandrine amphibians were tested against cancer cell lineages. The venoms, which presented proteins in their biochemical content, were active against cervix, alveolar, colon colorectal, pancreas, prostate, astrocytoma and breast carcinoma lines. However, these secretions were also toxic to human fibroblasts (HEK 293) [105].

Marinobufagin is a molecule present on *R. marina* venom displaying activity against leukemic cells without being toxic to normal blood cells. According to Machado et al. [106], this steroid induces toxicity via apoptosis, antimitotic action and cycle cell arrest at interphase in leukemia cells, without any genotoxicity.

The bufadienolides, bufotoxins, alkaloids and arginiyl derivatives from *R. jimi* cytotoxicity effects on cancer cell lineages were studied by Filho et al. [107], whereas Spinelli et al. [108] revaluated the antitumor action of 11 different Argentine amphibians: 6 *Hylidae*/*Microhylidae* and 5 *Leptodactylidae*. These venoms induced apoptosis and autophagy. Interestingly, *Leptodactylidae* skin secretion induced aggregation on cancer cells.

Finally, we present bufotenine: a tryptamine alkaloid found in many species and genera across nature (animals and plants), particularly in *R. crucifer*, *R. granulosa*, *R. schneideri*, *R. icteria* and *R. jimi* [109]. This molecule was selected in biomonitored assays and has the capacity to inhibit the penetration of rabies virus in mammalian cells, through an apparent competitive mechanism [110]. Complementary studies conducted by those authors [111] showed that this molecule was active in vivo, by increasing the survival rate of intracerebrally virus-infected mice from 15 to 40%. The safety of bufotenine was then evaluated [112] and no significant effects on mice could be detected at the effective

antiviral dose. Interestingly, bufotenine acts synergically with ocellatin-F1—an antimicrobial peptide obtained from the frog *Leptodactylus labyrinthicus* skin secretion—in the rabies virus model [113]. Finally, recent in vitro assays showed that bufotenine has no antiviral action against canine coronavirus (CCoV), canine adenovirus type 2 (CAV-2) or herpesvirus type 1 (HSV-1), indicating some specificity against distinct types of viruses [114]. The mechanism of action of this alkaloid remains unclear (although the evaluation of its effects in the immune system is being assayed by these authors), but bufotenine is the perfect example of the potential of bioactive molecules isolated from a neglected venom, serving as biotechnological tool for a neglected disease drug development study.
