Distribution of major toxins in Rhinella marina parotoid macroglands using Desorption-Electrospray-Ionization mass spectrometry imaging (DESI-MSI).

Amphibian cutaneous glands secrete toxins used in different vital functions including passive defense. Through Desorption Electrospray Ionization-Imaging we analyzed the distribution of the major toxins of the toad Rhinella marina parotoid macroglands. Alkaloids and steroids showed characteristic distribution and intensity within the glands and were also present at lower levels on the skin surface. A comprehensive overview of toxins distribution in toads' skin might help to understand their full biological role within the amphibians.

Main text

Amphibians are known to possess cutaneous glands distributed throughout the body. Basically, two different types of cutaneous glands can be found: (i) mucous glands, generally associated to maintenance of humidity and cutaneous respiration, and (ii) granular (or poison) glands, generally associated with chemical defense against predators and/or microbial infection (Toledo and Jared, 1995)(Jared et al., 2009).

In some species, the cutaneous glands are grouped in large numbers, forming conspicuous protrusions, known as macroglands, among which the parotoids are the most common (Mailho-Fontana et al., 2017). Histologically, the parotoids in toads of genus Rhinella are defined as large accumulations of hypertrophied poison glands arranged side by side and inserted into the dermal connective tissue (Mailho-Fontana et al., 2018).

It is widely accepted that the mucous glands secrete mucins involved in physiological functions (Daly, 1995), while the granular glands would have the functions of producing and storing content rich in a huge diversity of key substances for the defense system against microorganisms and predators (Raaymakers et al., 2017). However, no definitive demonstration of the exclusivity of the granular glands in the production of defensive toxins has yet been performed, in spite of the efforts of several groups in the thorough characterization of the skin secretion composition (Liu et al., 2010, Li et al., 2016, Zhang et al., 2019).

Recently, our group has started to investigate such issues by an integrative approach using biochemical and immunohistochemical methods, together with classic histochemical stains in order to analyze the content and distribution of toxins in the amphibian skin. First we developed a method for retrieving soluble proteins from viscous amphibian skin secretion (Mariano et al., 2018) aiming at posterior proteomic identification (Mariano et al., 2018). In another study we purified and biochemically identified an enzyme in the skin secretion of the leaf frog Phyllomedusa distincta, which was unequivocally localized by immunohistochemistry in all types of skin glands, including the mucous glands (Sciani et al., 2019). In a third study, by the use of immunohistochemistry and mass spectrometry we were able to identify and localize tetrodotoxin in the skin of the newt Taricha granulosa, not only inside the granular glands but also within the mucous glands (Mailho-Fontana et al., 2019). In addition, we have also performed a molecular phenotypic approach in the characterization of different South-American toads, by the analysis of their low molecular mass components in the skin secretion (Sciani et al., 2013). Despite the very informative data obtained, molecules localization within the tissues still remains poorly known (Mailho-Fontana et al., 2018).

In the present work, by performing Desorption-Electrospray-Ionization (DESI) mass spectrometry imaging (MSI), for the first time, we were able to precisely localize toad toxins in the parotoid macroglands.

In order to achieve our goal, we analyzed Rhinella marina parotoids (male) (Suppl. Fig. 1) from Belterra (state of Pará, Brazil) under collection permit provided by the Brazilian Federal Government (SISBIO #48080-2). After euthanasia, the parotoid macroglands were removed, immediately frozen in -80 °C and sectioned (16 μm thick) in transversal and longitudinal axis (Suppl. Fig. 1), using a Leica CM1860 cryostat. All methods were approved by the Ethics Committee on Animal Use of Instituto Butantan (CEUAIB, protocol #1046/13).

The glass slides containing the histological sections were imaged using a 2-D Omni Spray DESI imaging platform (Prosolia Inc, USA) coupled to a Q-Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific, USA), in a spatial resolution set to 200 μm. Analysis were performed under positive ion mode by spraying methanol (MeOH), dimethylformamide (DMF)/acetonitrile (ACN) (1:1) or MeOH/H2O (1:1) (v/v) and ion images were assembled using BioMap (http://www.maldi-msi.org). After DESI-MSI, the sections were stained with hematoxylin-eosin (HE).

Fig. 1 contains the main findings of our analyses: HE-stained tissue section and ion images in a color-coded representation of the whole gland, where red represents abundancy and black represents absence.

Fig. 1

Transverse histological section of (a) Hematoxylin-eosin staining. (b) Dehydrobufotenin m/z 203.098. (c) Marinobufagin m/z 401.202. (d) Telocinobufagin m/z 403.217. (e) Argininesuberoyol-marinobufagenin m/z 713.383. (f) Hellebrigenin m/z 417.216, as negative control. Poison gland (*). Dermis (d).

Fig. 1A shows a transverse section of a parotoid macrogland stained with HE. Fig. 1B–F shows serial transverse sections obtained by DESI-MSI of four most abundant toxins previously identified in this species (Sciani et al., 2013), namely: dehydrobufotenin (Fig. 1B, m/z 203.098), marinobufagin (Fig. 1C, m/z 401.202), telocinobufagin (Fig. 1D, m/z 403.217), argininesuberoyol-marinobufagenin (Fig. 1E, m/z 713.383), and hellebrigenin (Fig. 1F, m/z 417.216), a molecule absent in R. marina skin secretion (Sciani et al., 2013), used as a negative control. All these structures were confirmed by MS/MS, as shown in the Suppl. Fig. 2, considering their fragmentation profile, in comparison to the already described ones (Sciani et al., 2013). Methanol was the best solvent to solubilize those molecules and get an appropriated spectrum. The mixture MeOH/H2O did not improve the signal, neither solubilized different compounds, whereas DMF/ACN yielded low quality spectra.

Two interesting conclusions can be drawn from our analysis: 1) the low molecular mass toxins (alkaloids and steroids) are unequivocally stored within the glands that compose the parotoid macroglands, and 2) there is a large variation in concentration among these toxins, which is in accordance to the relative concentrations previously assessed by RP-HPLC (Mailho-Fontana et al., 2018) (Sciani et al., 2013). Moreover, the location varies for each of the molecules. While alkaloids (Fig. 1B and E) are distributed throughout the glands, steroids are concentrated in the gland's periphery (Fig. 1C and D). Such differences in location may be related to the compatibility of each molecule stored in the gland to specific cytoplasm locations or to timing and constancy of toxin production.

The analysis of longitudinal sections by DESI-MSI revealed that the dermis is not consistent with any ion-image (Fig. 2), corroborating the results showed in Fig. 1. On the other hand, more superficial sections revealed that the epidermis clearly exhibits the presence of the analyzed molecules (Fig. 2), although in lower intensity when compared to the interior of the glands. Therefore, such molecules seem to be constitutively secreted by the cutaneous glands onto the skin surface, and may act in the chemical defense against microorganisms, as already described for Rhinella marina poison (de Medeiros et al., 2019).

Fig. 2

Longitudinal histological section of (a) Hematoxylin-eosin staining. (b) Dehydrobufotenin m/z 203.098. (c) Marinobufagin m/z 401.202. (d) Telocinobufagin m/z 403.217. (e) Argininesuberoyol-marinobufagenin m/z 713.383. (f) Hellebrigenin m/z 417.216, as a negative control. Dermis (d). Epidermis (arrows).

In conclusion, the present work demonstrates for the first time the actual tissue distribution of different toxins in the skin of a toad. The use of DESI-MSI, mainly if integrated to other morphological approaches, can provide relevant information about amphibian poison sources, contributing to the better understanding of its multifunctional role.

Funding

This work was supported by [grant #001 to PLMF], [grants #01.09.0278.04 and #01.12.0450.03], (CNPq) [406385/2018-1 to DCP] and [grant #2014/15334-4 and #2017/10488-1 to PLMF, #2018/03265-9 to CJ and #2019/04314-6 to AMP]. DCP, CJ and MMA are fellow researchers [grants #303792/2016-7, #305266/2019-5 and #309589/2017-7].

CRediT authorship contribution statement

Pedro L. Mailho-Fontana: Conceptualization, Methodology, Investigation, Writing - original draft. Andreia M. Porcari: Methodology, Investigation, Writing - review & editing. Marcos N. Eberlin: Methodology, Resources. Carlos Jared: Conceptualization, Funding acquisition, Writing - review & editing. Marta Maria Antoniazzi: Conceptualization, Validation, Writing - review & editing. Daniel C. Pimenta: Conceptualization, Funding acquisition, Formal analysis, Writing - original draft. Juliana M. Sciani: Conceptualization, Investigation, Resources, Validation, Writing - original draft, Writing - review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.