Slime protein profiling: a non-invasive tool for species identification in Onychophora (velvet worms)



Onychophorans (velvet worms) use an adhesive, protein-based slime secretion for prey capture and defence. The glue-like slime is ejected via a pair of modified limbs and the sticky threads entangle the victim. In this study, we analysed the protein composition of slime in twelve species of Onychophora from different parts of the world, including two species of Peripatidae from Costa Rica and Brazil and ten species of Peripatopsidae from Australia, using sodium dodecyl sulphate polyacrylamide gel electrophoresis. Our results revealed high intraspecific conservation in protein composition of slime in each species studied. In contrast, the protein profiles differ considerably in both number and position of bands between the species. We observed the highest number of differences (in 20 of 33 considered band positions) between a peripatid and a peripatopsid species, whereas the lowest number of differences (in four band positions) occurs between two closely related egg-laying species. The reconstructed maximum parsimony cladogram based on the electrophoretic characters largely reflects the phylogenetic relationships of the species studied, suggesting that the slime protein profiles contain useful phylogenetic information. Based on our findings, we suggest that the slime protein profiling is a valuable, non-invasive method for identifying the onychophoran species. Moreover, this method might help to discover potentially new species of Onychophora, given that the ~200 described species most likely underrepresent the actual diversity of the group.


Onychophorans or velvet worms comprise a small group of soft-bodied, terrestrial invertebrates, which are confined to microhabitats with high humidity levels, such as within rotted logs and leaf litter, where they mainly feed on different arthropods (Bursell and Ewer 1950; Ghiselin 1984; Read and Hughes 1987; Storch and Ruhberg 1993; Hamer et al. 1997). One of the most peculiar features of onychophorans is their hunting and defence strategy using an adhesive slime secretion, which is ejected via a pair of modified limbs called slime papille (Fig. 1). The slime is mainly composed of proteins (up to 55% of its dry mass), varying from low-molecular-mass proteins (8–25 kD) up to large protein complexes exceeding 600 kD (Mora et al. 1996; Benkendorff et al. 1999; Haritos et al. 2010).

Figure 1.

Photograph of a specimen of Euperipatoides rowelli (Peripatopsidae) caught in the act of slime ejection. Note the jets of the sticky slime, which are ejected via a pair of slime papillae (arrows)

While previous analyses using sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) mainly focused on mid- and low-molecular-mass proteins (Mora et al. 1996; Haritos et al. 2010), high-molecular-mass proteins have been shown to be present in the onychophoran slime by size-exclusion chromatography (Benkendorff et al. 1999). Moreover, a comparative analysis of low-molecular-mass proteins revealed interspecific variation in four onychophoran species from Costa Rica (Mora et al. 1996), suggesting that differences in slime composition might be useful for taxonomic studies of Onychophora. So far, only ~200 onychophoran species have been described worldwide (Oliveira et al. 2012a), but the actual number of species might be much higher, given high cryptic diversity of the group as revealed by molecular, karyologic and morphological studies (Gleeson et al. 1998; Trewick 1998; Daniels et al. 2009; Sampaio-Costa et al. 2009; Oliveira et al. 2011, 2012b). Thus, an additional tool for species identification based on the characterization of slime proteins would provide a quick and non-invasive method for studies of cryptic speciation and species diversity of Onychophora.

To test the applicability of this method for species identification, we analysed the protein composition of slime in twelve onychophoran species from different parts of the world, including representatives of the two major onychophoran subgroups, Peripatidae and Peripatopsidae (Figs 2a,b and 3a–l). To further increase the resolution of this method, we optimized the collection technique of slime and analysed a wide range of protein sizes. In addition, we tested the sensitivity of this method for species identification by including two cryptic Tasmanian species of Peripatopsidae in our analyses, which are indistinguishable morphologically but are clearly two separate species according to their karyotypes (Rowell et al. 1995; Reid 1996).

Figure 2.

Maps illustrating the geographical distribution of onychophorans and the localities of the species studied. Abbreviations of the species/taxon names as in Table 1, except for *Tas1 (=‘Tasmania’ sp. 1/locality 1) and **Tas1 (=‘Tasmania’ sp. 1/locality 2). (a) World map showing the distribution of Peripatidae (in red) and Peripatopsidae (in blue) and the localities of representatives of Peripatidae from Costa Rica and Brazil (map modified from: (b) Detailed view of South Eastern Australia and Tasmania (framed region in a) illustrating the localities of representatives of Peripatopsidae studied

Figure 3.

Representatives of Onychophora studied. The size of the photographed specimens ranges from 10 to 85 mm (images not to scale). (a–j) Representatives of Peripatopsidae. (k, l) Representatives of Peripatidae. (a) Euperipatoides rowelli. (b) Phallocephale tallagandensis. (c) Ooperipatus hispidus. (d) Ooperipatus oviparus. (e) Ooperipatellus insignis. (f) Ooperipatellus sp. (g) ‘Tasmania’ sp. 1. (h) ‘Tasmania’ sp. 2. (i) Tasmanipatus anophthalmus. (j) Tasmanipatus barretti. (k) Principapillatus hitoyensis. (l) Epiperipatus sp.

Materials and Methods


Two species of Peripatidae and ten species of Peripatopsidae were studied (Figs 2a,b and 3a–l; Table 1). The specimens were obtained from rotting logs and leaf litter at the corresponding localities and maintained in plastic boxes with perforated lids as described previously (Baer and Mayer 2012; Oliveira et al. 2012b). The animals were collected and exported under the following permits: (1) the Forestry Commission of New South Wales, Australia (permit no. SL100159), (2) the Department of Primary Industries, Parks, Water and Environment of Tasmania, Australia (permit no. TFA 12259), (3) the Department of Sustainability and Environment, Victoria, Australia (permit no. 10004070), (4) the Gerencia Manejo y Uso Sostenible de RR NN–Ministerio del Ambiente y Energia, Costa Rica (permit numbers 123-2005-SINAC and 014950) and (5) Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio), Brazil (permit no. 12414/3909-1 and 113141/2011).

Table 1. Onychophoran species examined in the present study with corresponding locality data
  1. a

    Undescribed species. Preliminary names according to Reid (1996).

PeripatidaePrincipapillatus hitoyensis Oliveira et al., 2012HitReserva Biológica Hitoy Cerere, 09°40′21″N, 83°02′36″W, 300 m, Province of Limón, region of Talamanca, Costa Rica
Epiperipatus sp.EpiPCH Porto das Pedras, 52°32′33″W, 19°28′44″S, 360 m, State of Mato Grosso do Sul, Brazil
PeripatopsidaeEuperipatoides rowelli Reid, 1996RowTallaganda State Forest, 35°26′S, 149°33′E, 954 m, New South Wales, Australia
Ooperipatus hispidus Reid, 1996HisTallaganda State Forest, 35°26′S, 149°33′E, 954 m, sympatric with Euperipatoides rowelli, New South Wales, Australia
Ooperipatellus insignis (Dendy, 1890)InsMount Macedon, 37°23′26″S, 144°34′38″E, 976 m, Victoria, Australia
Ooperipatus oviparus (Dendy, 1895)OviMount Macedon, 37°22′05″S, 144°35′25″E, 899 m, Victoria, Australia
Ooperipatellus sp.OopLyell Highway, 42°20′36″S, 146°28′13″E, 545 m, Tasmania, Australia
Tasmanipatus anophthalmus Ruhberg et al., 1991AnoElephant Pass, 41°38′25″S, 148°14′29″E, 324 m, Tasmania, Australia
Tasmanipatus barretti Ruhberg et al., 1991BarScamander River, 41°26′01″S, 148°11′34″E, 42 m, Tasmania, Australia
‘Tasmania’ sp. 1a (locality 1)Tas1Scotts Peak Road, 43°02′17″S, 146°18′5″E, 308 m, Tasmania Australia
‘Tasmania’ sp. 1a (locality 2)Tas1Scotts Peak Road, 42°50′45″S, 146°22′22″E, 523 m, Tasmania Australia
‘Tasmania’ sp. 2aTas2Lyell Highway, 42°12′42″S, 146°07′15″E, 814 m, Tasmania, Australia
Phallocephale tallagandensis Reid, 1996TalTallaganda State Forest, 35°26′S, 149°33′E, 954 m, sympatric with Euperipatoides rowelli, New South Wales, Australia

Slime collection and sample preparation

Slime samples were first obtained from several individuals of each species to clarify whether there is an intraspecific variation in slime composition (Hebert et al. 1991). To study the interspecific variation, we pooled slime samples from different individuals to obtain a sufficient volume of slime for the electrophoretic analyses using different methods. Specimens were stimulated to eject the slime into 500 μl Eppendorf tubes by carefully touching their anterior ends with a forceps. Collected slime samples were diluted in emulated onychophoran slime gland buffer (41% w/v glycine, 11% w/v glutamic acid and 3% w/v aspartic acid) adjusted to pH 7.0 (Röper 1977) and stored for several days at 4°C. The protein amount of each sample was quantified according to Bradford (1976), and the absorbance at 595 nm was monitored using a Tecan Infinite M200 plate reader and i-control v1.6 software (Tecan, Mainz-Kastel, Germany). Protein samples were prepared by adding sodium dodecyl sulphate (SDS) sample buffer (10% w/v SDS, Laemmli 1970) to each sample up to a final volume of 18 μl followed by denaturing for 10 minutes at 65°C. After extraction of slime, the animals were kept in culture for further experiments. We have refrained from depositing slime samples as vouchers, as we observed variation of the protein composition within a few weeks due to protein degradation. Therefore, samples were investigated as fast as possible after collection.

Protein separation

Slime proteins were separated using denaturing polyacrylamide gel electrophoresis (SDS-PAGE) under non-reducing conditions. The polyacrylamide concentration in gels varied between 4% T and 14% T (total concentration of both acrylamide and the cross-linker bisacrylamide) with 3% C (concentration of the cross-linker) for separating small and large proteins, respectively. Protein samples (10 μg protein per lane in 4% gels and 30 μg protein per lane in 14% gels) were run in a Mini-PROTEAN II electrophoresis cell (Bio-Rad, Hercules, CA, USA). Analyses using PAGE were performed either according to Laemmli (1970), with a Tris-Glycin SDS running buffer (25 mM Tris, 192 mM Glycine, 0.1% SDS, pH 8.3) at 40 mA for approximately 1.5 h (SDS-PAGE) or according to Schägger (2006) using Tris anode buffer and Tris-Tricine cathode buffer at constant 120 V for approximately 3 h (Tricine SDS-PAGE). As protein standards, the prestained protein ladders PageRuler (Fermentas, St. Leon-Rot, Germany; for proteins in the molecular range of 10–170 kD) and Protein Standard HiMark (Invitrogen, Darmstadt, Germany; for larger proteins between 80 and 500 kD) were used.

Staining and imaging

Non-specific staining of SDS gels was performed using Amido Black or Coomassie Blue (0.1% w/v Coomassie Blue R350 or Amido Black, 20% v/v methanol and 10% v/v acetic acid) for 20 min and destained in destaining solution (50% v/v methanol and 10% v/v glacial acetic acid) for 3 h. Low-abundance proteins were detected using the highly sensitive Pierce® Silver Stain Kit (Thermo Scientific, Waltham, MA, USA). Gels were imaged using the documentation system Syngene G-Box supported by the software GeneSnap v7.05 (Synoptics, Cambridge, UK). Final image editing and panel design were performed using adobe (San Jose, CA, USA) Photoshop CS5 and Illustrator CS5.

Comparative analyses of SDS gels

To determine the molecular mass of protein bands for each species, molecular markers with defined weight values were used as a reference. Intermediate molecular mass values were calculated via a standard curve, which was obtained from the logarithmic regression of the values defined by the marker. The distances between the protein bands in each gel were measured and assigned to the respective values with an accuracy of 0.5 mm. Due to the differing readability, only conserved areas of high (350–1300 kD)-, mid (55–110 kD)- and low-molecular-mass protein bands (8–25 kD) were taken into account. For phylogenetic analyses, an electrophoretic character matrix was compiled (Mickevich and Mitter 1981; Richardson et al. 1986; Briscoe and Tait 1995), considering the values between 8 and 1300 kD as characters and the presence or absence of bands at each position as character states of the species studied. Protein bands that occurred in two or more taxa were assumed to be a shared character. Maximum parsimony analyses of the presence–absence matrix (Table S7) were conducted with paup* 4.0b10 (Swofford 2002) using an exact tree search algorithm (branch-and-bound).


Our analyses of slime composition in twelve onychophoran species (Fig. 3a–l; Table 1) revealed that the molecular mass of slime proteins varies between 8 and 1300 kD (Figs 4 and 5a–f; Tables S1–S3). Different-sized proteins appear in the SDS-PAGE gels as distinct bands, which vary in position according to their molecular mass. In total, we considered 33 band positions of low-, mid- and high molecular mass. Thereby, band intensities vary according to the amount of the corresponding protein in the slime (Figs 4 and 5a–f). None of the twelve species studied show a complete set of 33 bands. Moreover, none of the bands occur in all twelve species and characteristic bands only appear in one or two species (Fig. 5a–f, white dots). Instead, the respective position and number of bands clearly differ in 10–15 bands between the species (Fig. 5a–f; Tables S3–S6). We therefore refer to the specific set of bands observed in each species as a ‘band pattern’.

Figure 4.

Slime protein band patterns in the peripatopsid species ‘Tasmania’ sp. 1 and ‘Tasmania’ sp. 2. Non-reducing Tricine SDS-PAGE, silver stain. Positions of molecular mass standards (in kD) are indicated in the left column. Lanes are numbered as follows: 1 and 2, specimens of ‘Tasmania’ sp. 1 from the locality 1; 3 and 4, specimens of ‘Tasmania’ sp. 1 from the locality 2; 5 and 6, specimens of ‘Tasmania’ sp. 2. Note that band patterns are conserved in specimens from the same locality. Also note the intraspecific variation in band patterns between different samples of ‘Tasmania’ sp. 1 from localities 1 and 2 and interspecific variation between ‘Tasmania’ sp. 1 and ‘Tasmania’ sp. 2 (white dots)

Figure 5.

Slime protein band patterns in twelve onychophoran species studied. Sizes of molecular mass standards (in kD) are indicated in the left (a, d) and right columns (a–f). Lanes are numbered as follows: 1 and 2, Peripatidae; 3–12, Peripatopsidae. 1, Principapillatus hitoyensis; 2, Epiperipatus sp.; 3, ‘Tasmania’ sp. 1; 4, ‘Tasmania’ sp. 2; 5, Tasmanipatus anophthalmus; 6, Tasmanipatus barretti; 7, Euperipatoides rowelli; 8, Phallocephale tallagandensis; 9, Ooperipatus hispidus; 10, Ooperipatus oviparus; 11, Ooperipatellus insignis; 12, Ooperipatellus sp. (a, d) 4% non-reducing Tricine SDS-PAGE, bands were visualized by silver stain. (b, e) 14% non-reducing Tricine SDS-PAGE, bands were visualized by Coomassie stain. (c, f) 14% non-reducing Tricine SDS-PAGE, bands were visualized by silver stain. Note the interspecific diversity of band patterns and the accumulation of bands in three major regions (8–25 kD, 55–110 kD, and 350–1300 kD). White dots indicate characteristic bands.

Nearly all species studied show intraspecifically conserved band patterns, that is, our analyses of slime samples obtained from different specimens of the same species retrieved highly congruent results regarding the presence, position, intensity and regions of dominant bands (Fig. 4; Tables S1–S3). The only exception are the observed deviating band patterns in specimens of the undescribed species ‘Tasmania’ sp. 1 from two localities, which differ in a single band position and in the intensities of five bands (Fig. 4, Tables S1–S3).

In contrast to the high intraspecific conservation, the interspecific comparison of band patterns revealed clear differences between the species (Fig. 5a–f, Tables S3–S5). The largest interspecific difference occurs between the peripatid Principapillatus hitoyensis and the peripatopsid Euperipatoides rowelli, in which the band patterns differ in 20 band positions, whereas the lowest interspecific difference is found between the two egg-laying peripatopsid species Ooperipatellus insignis and Ooperipatellus sp., with band patterns deviating only in four positions (Table S6).

Furthermore, the majority of dominant bands in all species studied are located in three regions of the gels ranging from 8 to 25, 55 to 110 and 350 to 1300 kD (Fig. 5a–f). In these regions, most similarities between the species are found in areas of large-sixed (~780 and 350 kD) and mid-sized (~95 kD) proteins, showing a variation of only a few kilodaltons between the bands. In contrast, the highest diversity of band patterns between the species occurs in the low-molecular-mass range of 8–25 kD (Fig. 5c, f).

Our phylogenetic analyses of the electrophoretic character matrix using maximum parsimony retrieved a single best-scoring tree (tree length = 66), in which all species studied fall into five major clades (Fig. 6). One clade consists of the two peripatid species Principapillatus hitoyensis and Epiperipatus sp. and is separated by the longest branch from the remaining four clades consisting of the Peripatopsidae species. The first Peripatopsidae clade consists of the New South Wales sympatric species Euperipatoides rowelli and Phallocephale tallagandensis, which forms the sister group to a second clade comprising the two undescribed species ‘Tasmania’ sp. 1 and ‘Tasmania’ sp. 2, whereas the third clade, which includes Tasmanipatus anophthalmus and Tasmanipatus barretti, clusters with a fourth major clade formed by the egg-laying species Ooperipatus hispidus, Ooperipatus oviparus, Ooperipatellus insignis and Ooperipatellus sp. (Fig. 6).

Figure 6.

Unrooted maximum parsimony tree using the electrophoretic data. Phylogenetic analysis was based on the character matrix provided in Tables S3–S5. The branch length values represent the number of evolutionary changes reflected by the deviating protein bands. Number of taxa: 12; number of characters: 33; consistency index (CI): 0.500; rescaled consistency index (RC): 0.548.


The slime of onychophorans is composed of proteins that are able to aggregate to sticky threads that adhere to various surfaces and are used to entangle their prey and potential predators (Heatley 1936; Read and Hughes 1987; Reid 1996; Haritos et al. 2010). Since the efficiency of this mechanism depends on particular properties of slime proteins, one would expect that the composition of slime has remained unchanged in different onychophoran lineages. Hence, the observed band patterns of proteins from different species should be similar, as the migration rates of proteins in SDS-polyacrylamide gels mainly depend on the number and identity of amino acids, which determine its molecular mass (Stryer et al. 2003). Typically, proteins with a higher molecular mass migrate slower than those with a lower molecular mass in the gel, and, thus, quantitative and/or compositional differences in the amino acid sequences are reflected in distinct bands in SDS-PAGE analyses (Stryer et al. 2003).

However, a previous study of four peripatids from Costa Rica revealed that the protein composition of slime varies between species (Mora et al. 1996), suggesting that a comparative analysis of slime composition might be useful for the identification of onychophoran species. Our large range separation of proteins via SDS-PAGE, applied to the slime of twelve additional onychophoran species, including representatives of Peripatidae and Peripatopsidae, revealed a high diversity of band patterns, thus confirming the previous finding of Mora et al. (1996). According to our data, the band patterns are highly conserved intraspecifically, while their number and composition vary considerably between the species, as none of the 33 identified band positions are shared by all species studied. Among the three major molecular mass classes, in which most slime proteins are accumulated, we observed that the highest band diversity occurs in the lowest molecular mass range, that is, between 8 and 25 kD. Whether this variation is due to evolutionary changes in amino acid sequences or to the occurrence of different protein isoforms (e.g. Talmadge and Roy 1993; Stryer et al. 2003), or both, remains unclear.

In addition to compositional changes in the amino acid sequences, differing migration behaviour of proteins in electrophoretic analyses might be caused by post-translational modifications, such as glycosylation or phosphorylation (Claverol et al. 2003). However, since such post-translational modifications must be also encoded in the genome, the corresponding evolutionary changes will be reflected in deviating band patterns (Uy and Wold 1977). Irrespective of the mechanisms involved, our data clearly show that the band patterns differ between the onychophoran species studied, whereas they are conserved within each species. This suggests that SDS-PAGE analysis of slime proteins (=slime protein profiling) is a useful, non-invasive method that might serve as a rapid guide for the identification of onychophoran species. Hence, it can be used in conjunction with other methods, including DNA sequencing, allozyme analyses, karyology, light microscopy and scanning electron microscopy for studying the diversity of Onychophora (Hebert et al. 1991; Briscoe and Tait 1995; Trewick 1998; Mayer 2007; Oliveira et al. 2012a,b).

According to our data, slime protein profiling is a sensitive method, as it revealed differences in a single band position (marked with white dots in lanes 1 and 2 in Fig. 4) as well as in band intensities between populations of ‘Tasmania’ sp. 1 from two localities situated ~22 km apart from each other. Whether these differences are due to intraspecific variation caused by a recent geographical isolation of the two populations, or whether it indicates the existence of a third, hitherto unknown cryptic species within the ‘Tasmania’ species complex, has to be clarified by molecular and karyotypic analyses, which have proven useful for studies of cryptic speciation in Onychophora (Gleeson et al. 1998; Trewick 1998; Daniels et al. 2009; Oliveira et al. 2011, 2012a,b). The sensitivity of slime protein profiling is also evident by comparing our results from ‘Tasmania’ sp. 1 and ‘Tasmania’ sp. 2, as these species show distinct band patterns that differ in seven to eight band positions, depending on which of the two localities of ‘Tasmania’ sp. 1 is considered. Notably, ‘Tasmania’ sp. 1 and ‘Tasmania’ sp. 2 are morphologically indistinguishable, that is, they are cryptic species, although karyotype analyses clearly show they are separate species (Rowell et al. 1995; Reid 1996).

Interestingly, the species-specific band patterns of slime proteins resemble the diversity of allozymes (=variant forms of enzymes encoded by different alleles) among the onychophorans species, which have been commonly used as phylogenetic markers in the past (Briscoe and Tait 1995; Trewick 1998). Similar to these studies, our phylogenetic analysis based on an electrophoretic character matrix, including protein bands as characters, reflects to some extent the phylogenetic relationships of the species studied (Reid 1996; Allwood et al. 2010; Murienne et al. 2014). For example, representatives of Peripatidae and Peripatopsidae form separate clades and closely related species cluster together, such as ‘Tasmania’ sp. 1 and ‘Tasmania’ sp. 2, Tasmanipatus anophthalmus and Tasmanipatus barretti, Euperipatoides rowelli and Phallocephale tallagandensis, whereas the egg-laying species Ooperipatus sp., Ooperipatus oviparous, Ooperipatus hispidus, Ooperipatellus insignis and Ooperipatellus sp. also group together, although their relationship has not been clarified yet with certainty (Ruhberg 1985; Reid 1996; Allwood et al. 2010). This suggests that the evolution of slime proteins at least to some extent mirrors the speciation events in different onychophoran lineages. The position of Ooperipatus hispidus in a different clade from the two other sympatric species, Euperipatoides rowelli and Phallocephale tallagandensis, further suggests that the slime composition is unlikely to be influenced by environmental factors alone, as these three species occur in the same habitat.


We are thankful to Susann Kauschke for technical support and maintaining the animals, to David Rowell, Noel Tait, Robert Mesibov, Franziska Anni Franke, Michael Gerth and Sandra Treffkorn for their help with specimen collection, to Lars Hering for his assistance with phylogenetic analyses and to Stefan Hannig and Leonhard Kätzel for their help with macrophotography (Fig. 1). The staffs of the Forestry Commission of New South Wales (NSW, Australia), the Department of Sustainability and Environment (Victoria, Australia), the Department of Primary Industries, Parks, Water and Environment (Tasmania, Australia), the Instituto Nacional de Biodiversidad (INBio, Costa Rica), the National System of Conservation Areas (SINAC, MINAE, Costa Rica) and the Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio, Brazil) are gratefully acknowledged for providing permits.

Funding: University of Leipzig (Doktorandenförderplatz, DFPL U00022) (to AB); Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq: 290029/2010-4) (to ISO); German Research Foundation (DFG: TRR67, A4) (to AGBS); Emmy Noether Programme of the German Research Foundation (DFG: Ma 4147/3-1) (to GM).