Introduction

In addition to complex morphological, physiological and behavioral adaptations, millipedes evolved chemical weapons: they were probably the first land animals with chemical defenses against predators (Rodriguez et al. 2018). Millipedes possess paired exocrine defense glands arranged segmentally, that open laterally or dorsally on the body. The glands produce a variety of chemicals, including alkaloids, quinones, phenols and cyanogenic compounds (Makarov 2015; Shear 2015). Depending on the morphology of the glands and the presence of specific defense substances, four types of defense glands can be distinguished: the glomeridan type (order Glomerida) produces quinazolinone alkaloids in a sticky proteinaceous matrix; the julidan type (orders Julida, Spirobolida, Spirostreptida, Callipodida and probably Stemmiulida) produces phenols and quinones; the colobognathan type (orders Polyzoniida, Platydesmida, Siphonophorida and Siphonocryptida) produces heterocyclic nitrogen-containing alkaloids; and the polydesmidan type (order Polydesmida) relies mainly on cyanogenic compounds (Gruner et al. 1993; Makarov 2015; Shear 2015). Diplopods release the contents of defense glands during defensive behavior, but other functions are also assumed, such as functions in alarm behavior and pheromonal communication, protection against parasites as well as antimicrobial activity (e.g., Makarov 2015; Ilić et al. 2018, 2019).

With respect to the total number of described millipede species (currently about 13 000), defense secretions have so far only been analyzed in about 3% of species. Recent studies have shown that millipede defense secretions represent an untapped pool of novel components, with their potential functions and applicability still being unknown (Bodner et al. 2016, 2017, 2018; Kunert et al. 2023). Within the Diplopoda, representatives of the order Glomerida are poorly studied: According to literature, chemoprofiles of only five species from three genera, Glomeris, Latreille, 1802, (Glomeris marginata (Villers, 1789), G. klugii Brandt, 1833, G. hexasticha Brandt, 1833), Loboglomeris Verhoeff, 1906 (Loboglomeris rugifera (Verhoeff, 1906)) and Onomeris Cook, 1896 (Onomeris sinuata (Loomis, 1943)) – all of which belong to the family Glomeridae Leach, 1916 – have been analyzed (Schildknecht et al. 1967; Carrel and Eisner 1984; Makarov 2015; Shear et al. 2011; Shear 2015). The secretions of all species were predominated by unique quinazolinone alkaloids, trivially called “glomerins”. Currently, only two glomerins are known, namely glomerin (1,2-dimethylquinazolin-4(1H)-one) and homoglomerin (2-ethyl-1-methylquinazolin-4(1H)-one) (Makarov 2015; Shear 2015).

In this study, the chemoprofiles of two members of the presumably basal family Glomeridellidae Cook, 1896, were analyzed to confirm a possible consistency in the composition of alkaloid glomeridan chemical defenses.

Material and methods

Specimens analyzed in this study: Five specimens of Typhloglomeris coeca Verhoeff, 1898 (Figs. 1a, b, 2), were collected by Dragan Antić on April 27, 2018 in Zaćirska Cave (42.34, 18.99; 435 m elevation), village of Gornji Ceklin, near Cetinje, Montenegro; 10 specimens of T. varunae Makarov et al. 2003 (Figs. 1c, d, 2), were collected by Dragan Antić on April 30, 2023 at type-locality Mlečnik Cave (41.27, 20.65; 1033 m elevation), village of Tašmaruništa, near Struga, North Macedonia. The material is deposited at the Institute of Zoology, University of Belgrade – Faculty of Biology. The photos of the living animals in situ were taken with a Canon PowerShot SX530 HS (Fig. 1a, b) and Olympus Stylus Tough TG-6 (Fig. 1c, d) digital cameras.

Fig. 1
figure 1

Individuals of Typhloglomeris coeca Verhoeff, 1898 (a, b) and Typhloglomeris varunae Makarov et al. 2003 (c, d), in situ. a Male and females on bat guano in Golubova Cave, Montenegro. b Male in Golubova Cave, Montenegro. c Male in Mlečnik Cave, North Macedonia. d Female and male in defense position in Mlečnik Cave, North Macedonia. Photos by Dragan Antić

Fig. 2
figure 2

Localities of the analyzed species. Blue circle: Typhloglomeris varunae Makarov et al. 2003, Mlečnik Cave, village Tašmaruništa, near Struga, North Macedonia. Yellow circle: Typhloglomeris coeca Verhoeff, 1898, Zaćirska Cave, village of Gonji Ceklin, near Cetinje, Montenegro

For extraction of defensive secretions, live specimens were placed in extraction vials with 500 µL methylene chloride (DCM) for 5 min. Since the small species (6–8 mm) are adapted to cave environment and very sensitive to transport, they were extracted directly after collection in the field. In addition, to increase the chemical yield of possible compounds in the solvent, pooled extracts of all collected individuals of each species were prepared. Extracts of both species were analyzed in the laboratory of the Institute of Biology, University of Graz, Austria, using an Agilent 5977B GC/ MSD (Vienna, Austria). After the first run with the GC–MS the 500 µL extracts were reduced under nitrogen to around 50 µL, and 1.5 µL of the concentrated extracts were re-injected into the GC–MS system. Solvent reduction was necessary due to the small amounts of compounds in the extracts and the poor GC response of quinazoline alkaloids in general.

The GC was equipped with two connected HP-5MS ultra inert capillary columns (15 m × 0.25 mm id., 0.25 µm film thickness; Agilent, Austria). Injection was splitless with helium as carrier gas at a flow rate of 1.0 ml min−1 for the first, and 1.2 ml min−1 for the second column. The temperature program of the GC started with an initial temperature of 40 °C, held for 1 min, followed by a temperature ramp of 10 °C min−1 to 300 °C. The final temperature was held for 5 min. The ion source of the MS and the transfer line were kept at 200 °C and 310 °C, respectively. Electron impact (EI) spectra were recorded at 70 eV. GC–MS data were processed with MassHunter Workstation 10.0 (Agilent, Vienna, Austria). Retention indices (RI) of compounds were calculated according to Van den Dool and Kratz (1963), using a standard mixture of n-alkanes (C9–C36) (Sigma Aldrich, Vienna, Austria).

Results and discussion

In the extracts of both species, Typhloglomeris varunae and T. coeca, a single peak only, showing an RI of 2376 could be detected (Fig. 3). The mass spectrum (Fig. 4) exhibited a base peak at m/z 188 (M+), along with characteristic fragments at m/z 104 and 105, fully matching MS-data for the already known glomerid defensive alkaloid homoglomerin (Shear et al. 2011). Thus, our results confirm the presence of quinazolinone alkaloids as defensive compounds in Glomeridellidae, supporting so far the consistent chemical nature of defensive secretions in this order. Based on these findings, we infer the following considerations on the evolutionary origin of alkaloids in the chemical defense of Glomerida.

Fig. 3
figure 3

Chemical profile of the defensive secretion of Typhloglomeris varunae Makarov et al. 2003 and T. coeca Verhoeff, 1898. Due to the small number of T. coeca individuals in the pooled extract, the amount of the compound homoglomerin was much lower

Fig. 4
figure 4

Characteristic EI mass spectrum of the peak at RI 2376 in the extracts of both Typhloglomeris species, identified as homoglomerin (M.+ at m/z 188)

The millipede order Glomerida comprises about 320 species in 36 genera with Holarctic distribution (Wesener 2022). Mauriès (2005) distinguishes three families of Glomerida: Glomeridellidae and Protoglomeridae Brölemann, 1913 with a small number of species and Glomeridae with the majority of described species. The family Glomeridellidae contains three genera: Glomeridella Brolemann, 1913 with four species, Typhloglomeris Verhoeff, 1898 with 14 species, and Tonkinomeris Nguyen, Sierwald & Marek, 2019 with two species (Liu and Golovatch 2020). The genus Typhloglomeris includes well-pigmented epigean species as well as pallid troglo-and geomorphic species, from the Balkan Peninsula, the Caucasus and the Near East (Liu and Golovatch 2020). Both species investigated in this study were found in caves of the Balkan Peninsula and are undoubtedly troglobiotic. Like other glomeridans Typhloglomeris species are capable of complete enrollment as the first line of defense (Fig. 1d).

In most glomeridan species studied, glomerin and homoglomerin occur together (except for L. rugifera, which contains only glomerin) (Schildknecht et al. 1967). By contrast, the defensive chemistry of representatives of the other two families of Glomerida was unknown to date. The presence of homoglomerin in the two typhloglomerids of our study confirms the chemical consistency of quinazolinone alkaloids in the secretions of glomerids. The absence of glomerin in Typhloglomeris might represent a phylogenetic signal, either indicating homoglomerin (the larger of the two glomerins) as the original compound or as the derivative glomerid quinazolinone. This may be argued as follows: According to Golovatch (1989) and Makarov et al. (2003), the family Glomeridellidae includes the most primitive glomerids. However, the opposite view has been expressed as well, stating that Protoglomeridae and Glomeridellidae represent a relatively advanced group of Glomerida (Liu and Golovatch 2020). Based on incomplete phylogenetic analyses of Glomerida and according to comparative morphological studies (Golovatch 1989; Makarov et al. 2003), we consider that the glomeridellids are one of the groups with the primitive state of some characters. With this understanding, we present a possible evolutionary scenario in which the presence of homoglomerin is plesiomorphic for glomeridans, while glomerin may have evolved later. Another scenario involves the possibility that the troglobiotic mode of life in Typhloglomeris species studied led to a reduction of glomerin as a defense substance. Reduction of pigment and eyes, cuticular thinning in arthropods, elongation of appendages and increases in extra-optic sensory structures are characteristic of animals in subterranean environments (Culver and Pipan 2019). Research of the cave-dwelling diplopods (polydesmidans and julidans), coleopterans and opilionids posed the hypothesis that, due to decreased selective pressures by predators in subterranean habitats, chemical defenses would also be simplified compared to epigean taxa (Shear et al 2010; Makarov et al. 2012, 2017; Vesović et al. 2015). However, the results of these studies did not support such a reduction. The retention of diversified chemical arsenal among the arthropods in subterranean habitats might be considered an example of chemical conservatism, implying that factors that drive the evolution of subterranean animals do not easily affect the chemistry of defensive secretions (Makarov et al. 2012, 2017; Vesović et al. 2015). Nevertheless, quinazolinones appear to be potent repellents, and effects of these compounds on vertebrates such as mice, frogs and birds were shown to be sedative, paralytic, emetic and hypnotic (Schildknecht et al. 1967; Carrel and Eisner 1984). Beside vertebrates, different groups of invertebrates are millipede predators, including ants, beetle adults and larvae, predaceous bugs, spiders, slugs (Herbert 2000; Shear 2015). A study on the interaction of wolf spiders and Glomeris marginata documented the sedative and antifeedant effect of alkaloids on the spider, with homoglomerin being the more potent as an antifeedant than glomerin (Carrel and Eisner 1984). Also, the adhesive nature of the proteinaceous matrix of glomeridan secretions is an effective mode of defense against ants, inhibiting their movements after contact with the millipedes’ secretions (Eisner and Meinwald 1966). Based on the knowledge on the cave fauna, we presume that the dominant predators of the two glomeridellid species in their habitats are invertebrates, and they might represent the selective pressure that shaped the evolution of chemical defence in T. coeca and T. varunae.

Regarding the evolutionary history of millipede secretions Rodriguez et al. (2018) investigated the chemical pathways in millipede defense systems. They found that the key chemical compound in the evolution of the millipede defense system was probably phenol. The origin of phenol production could be traced between 315 and 275 million years ago (Ma). It may be presumed that this chemical is the first one produced by millipede ancestors, and possibly even the oldest defense compound in terrestrial animals. However, Rodriguez et al. (2018) also noted that phenols are not directly involved in the production of alkaloids, suggesting that alkaloids may have evolved independently.

Furthermore, Shelley and Golovatch (2011) discussed the origin and early evolution of millipede orders, appointing the origin of the order Glomerida between 480 and 448 Ma. Paleontological data support the early split of Pentazonia and Helminthomorpha (Miyazawa et al. 2014; Fernández et al. 2018; Wesener 2019). Within the Helminthomorpha branch, Colobognatha first separated from the eugnathan groups Juliformia and Merocheta. As Rodriguez et al. (2018) noted, the phenolic biosynthetic pathway is present in Helminthomorpha except for the alkaloid-producing Colobognatha. Therefore, we assume that the alkaloid biosynthetic pathway evolved earlier or at about the same time as phenolics evolved. Thus, the first diversification of chemical defense in millipedes may have led to the alkaloid (or rather anthranilic acid) biosynthetic pathway in glomeridans (Schildknecht and Wenneis 1967) and a pathway in helminthomorphs, which again split into the pathways of the alkaloid Colobognatha and the phenolic helminthomorphs (Rodriguez et al. 2018). In these terms, alkaloids in glomeridans and colobognathans may have different phyletic origins, which is also reflected in the highly different chemistry of alkaloids involved, namely quinazolinone alkaloids in Glomerida and unusual, spirocyclic terpene alkaloids in Colobognatha.

Regardless of currently proposed time frames and hypotheses for the origin of alkaloid compounds, the evolution and divergence of millipede alkaloids remains vague and unclear. It is important to analyze as many diplopod taxa as possible that belong to chemically distinct groups to obtain a robust picture of richness and heterogeneity, distribution and evolution of alkaloid chemical defense systems in extant Diplopoda. This is the first study on the defensive chemistry of representatives of the family Glomeridellidae and our data strongly support the view that the members of the order Glomerida generally rely on alkaloid defensive chemicals.