Bacteria or melanosomes? A geochemical analysis of micro-bodies on a tadpole from the Oligocene Enspel Formation of Germany
- First Online:
- Cite this article as:
- Barden, H.E., Bergmann, U., Edwards, N.P. et al. Palaeobio Palaeoenv (2015) 95: 33. doi:10.1007/s12549-014-0177-5
- 1.3k Downloads
Many exceptionally preserved fossils have long been thought the product of preservation by bacterial autolithification, based largely upon the presence of, micron-sized, spherical or elongate bodies on their surface. This has recently been challenged by studies of similar fossils which cite morphological and geochemical evidence that these structures could be fossilized melanosomes, melanin-containing organelles. We geochemically analysed a tadpole from the Oligocene Enspel Formation, Germany, which displays such spherical bodies on its surface. Pyrolysis gas chromatography mass spectroscopy (Py-GCMS) and Fourier transform infrared spectrometry (FTIR) indicate that the organic remains of the tadpole are original and are not the result of external contamination, shown by the different chemical compositions of the fossil and its enclosing matrix. Py-GCMS also demonstrates the presence of bacterial and plant biomarkers in the matrix but not the tadpole, suggesting that the spherical bodies are unlikely to be bacterial, and also that such fossils do not develop their dark colour from incorporating plant material, as has been suggested. X-ray absorption spectroscopy (XAS) shows high levels of organically bound Zn(II) in the fossilized soft tissue, a metal known to chelate both eu- and pheomelanin. The zinc in the tadpole shows greater similarity to that bound in pheomelanized extant samples than to that in eumelanized ones. Though further geochemical analysis of both pure pheomelanin and bacterial samples is required to completely exclude a bacterial origin, these results are in line with a pheomelanic origin for the spherical bodies on the tadpole.
KeywordsEnspel Bacterial autolithification Amphibian Melanin Melanosomes Geochemistry
Bacterial autolithification has been a widely accepted taphonomic model for the exceptional preservation of fossil organisms that has been studied across various taxonomic groups and geological formations (Wuttke 1992; Wilby et al. 1996; Briggs et al. 1997, 2005; Gehling 1999). The process is particularly a characteristic of the Oligocene Enspel Formation of Germany and was first described by Wuttke (Wuttke 1983; Cristol et al. 2005) after the observation of coccoid or bacillus cell-like structures within fossilized soft tissues. The phenomenon involves the mineralization of the cells in a bacterial mat that has covered a carcass (Briggs et al. 2005), caused by the precipitation of either siderite, utilising iron from the surrounding sediments (Franzen 1985; Wuttke 1992), or apatite, where the phosphorus from the decaying organism has become trapped by the mat (Briggs and Wilby 1996; Briggs et al. 1997). The typically black colour of the preserved soft tissues is thought to be derived from the subsequent incorporation of organic matter from plants to form a carbonized film (Franzen 1985).
In recent years however, the structures thought to be fossil bacterial cells have in many cases been reclassified as fossilized melanosomes, the membrane-bound organelles in which the chemical pigment melanin is synthesized. This reinterpretation is based on a growing number of studies citing both morphological (Vinther et al. 2008; Clarke et al. 2010; Zhang et al. 2010; Li et al. 2010, 2012; Carney et al. 2012) and geochemical evidence (Barden et al. 2011; Wogelius et al. 2011; Glass et al. 2012; Lindgren et al. 2012, 2014), though the debate is by no means settled (Moyer et al. 2014). Extant melanosomes are known to exist in both spherical and elongate forms depending on the type of melanin they contain (Liu et al. 2005), though this is not always the case (Liu et al. 2005). Thus far, such a combination of geochemical techniques has not been applied to the study of fossilized amphibians, though some fossil tadpoles from the Enspel Formation are known to have dense layers of sub-micrometre-sized spherical bodies within their soft tissues (Toporski et al. 2002). In this study, we performed detailed geochemical analysis of a fossil tadpole from the Enspel Formation to look for biomarkers indicative of the preservation of melanin or bacteria. Such analyses will not only give great insight into the taphonomy of the organism (Gupta et al. 2007a, b) but can also help to improve our understanding of its original biochemistry.
Materials and methods
Samples and handling
The tadpole appears to be broken in half with the end of the tail curling round to meet the head (Fig. 1a, b). Preservation is too poor to allow species identification with any confidence and also makes assignment of growth stage difficult. In addition, though no legs are apparent, the tadpole lays on the edge of the block leaving the possibility that either it became disassociated from the main organism or was separated from the main body during collection. As measured from the part (FS003a), the tadpole is 67 mm long including soft tissues. On comparison with other fossil tadpoles (Roček 2003; McNamara et al. 2010), it is likely to be between Gosner stages 36 and 37 (Gosner 1960) due to its size and apparent ossification of the skull.
Samples of redbreast feathers from an extant black-headed grosbeak (Pheucticus melanocephalus) were obtained from partially intact specimens curated at The Manchester Museum, and blonde and black human hair samples were donated by members of staff. These acted as modern pheomelanin and eumelanin standards, respectively. Zinc standards used for XAS edge calibration and for comparison to biological samples were zinc metal foil, zinc acetate (Zn(O2CCH3)2) and zinc sulfate (ZnSO4·7H2O). Zinc sulfate and acetate were purchased from Sigma-Aldrich, as was a sample of Sepia officinalis eumelanin (product number M2649). Once in the laboratory, all samples were kept in individual aluminium foil packets within sealed plastic bags. Gloves were worn at all times when handling the fossils.
Scanning electron microscopy (SEM) and energy-dispersive x-ray spectroscopy (EDS)
Samples were analysed using both variable pressure field emission gun scanning electron microscopy (VP-FEG-SEM) and conventional SEM. Extant grosbeak feathers were prepared for imaging according to Zhang et al. (2010), gold coated and mounted on an aluminium stub; fossil samples were left uncoated. All images collected were secondary electron images. A ZEISS Supra 40VP SEM was used to collect VP-SEM images of the tadpole and matrix (FS003b) at 15.0 keV accelerating voltage and 14.5 mm working distance and high vacuum SEM images of the same sample at 1.0 keV and 3.5 mm. An FEI XL30 VP-SEM was also used to collect VP-SEM images of smaller matrix pieces (FS003b) at 15.0 keV and 12.4 mm and SEM images of extant grosbeak feathers at 15.0 keV and 12.4 mm. Standardless EDS spectra of tadpole and matrix were taken using an EDAX energy-dispersive x-ray spectrometer with the ZEISS Supra 40VP SEM at 15.0 keV and 14.5 mm for 100 s. Errors are estimated to be approximately 30 % of the reported value. Image features were analysed using ImageJ (Schneider et al. 2012).
Pyrolysis gas chromatography mass spectrometry (Py-GCMS)
Matrix samples (FS003a) and 2.4 mg of powdered tadpole were analysed using normal flash Py-GCMS using a Chemical Data Systems (CDS) 5200 series pyroprobe pyrolysis unit attached to an Agilent 7890A gas chromatograph fitted with an HP fused column (J&W Scientific 5 % diphenyl-dimethylpolysiloxane; 30 m, 0.32 mm i.d.; 0.25 μm film thickness) and an Agilent 5975C MSD single quadrupole mass spectrometer operated in electron ionization (EI) mode (EI source temperature 230 °C and MS quadrupole held at 150 °C) with helium as a carrier gas. Samples were pyrolysed at 600 °C for 20 s, the pyrolysis transfer line and injected using a split ratio of 2:1; the injector port temperature was maintained at 350 °C and the interface at 280 °C. The oven was programmed to heat from 40 °C for 3 min to 320 °C at 4 °C min−1 held for 5 min, and there was a solvent delay of 4 min. Mass spectrometric detector (MSD) mass range scanned from was m/z 50–650 at one scan s−1. Thermochemolysis of samples was carried out by reaction with 10 μl of tetramethylammonium hydroxide (TMAH) solution and then analysed using the same equipment with the oven programmed to heat from 40 °C for 3 min to 300 °C at 4 °C min−1 held for 15 min, with a solvent delay of 10 min and an MSD mass range of m/z 60–700 at one scan s−1. Compounds were identified using the National Institute of Standards and Technology (NIST) database and by comparison with spectra from the literature.
Fourier transform infrared spectroscopy (FTIR)
Spectra of the tadpole and matrix (FS003b) and a eumelanin standard were taken in Attenuated Total Reflectance (ATR) mode with a germanium crystal using a Spotlight 400 PerkinElmer FTIR imaging system (wave number range 4000–650 cm−1). Scans were taken with an aperture of 20 μm2 and a 4 cm−1 resolution; each spectrum is an average of ten scans and is background subtracted. The ATR crystal was cleaned using isopropanol after each scan to avoid cross contamination. Peaks were identified using the Bio-Rad KnowItAll Informatics System 8.2 Multi-Technique database. Inorganic bands were identified by comparison with spectra from the RRUFF mineral database (Downs 2006).
Synchrotron rapid scanning x-ray fluorescence (SRS-XRF)
SRS-XRF analysis of tadpole and matrix (FS003b) was carried out at wiggler beamline 6-2 of the Stanford Synchrotron Radiation Lightsource (SSRL). Scans were undertaken in ambient conditions with a beam energy of 13.5 keV and a flux of 1.5 × 1011 photons s−1. A pinhole aperture was used to generate a beam spot size of 50 μm, and samples were mounted on an X-Y motorized stage at a fixed 45° incident angle to the beam. Fluoresced x-rays were detected using a single element drifted (Vortex) detector fixed at a 90° scattering angle to the beam and 70 mm from the sample. Zinc, copper and calcium were mapped with energy windows 350 eV wide to capture their respective characteristic x-ray emission line intensities. Point analyses were taken by locating and driving the mounted sample to a point of interest and collecting a full energy-dispersive spectrum for 100 s. Spectra were calibrated by comparison with a point analysis from a Durango apatite standard of known elemental composition. Errors are conservatively estimated to be approximately ±40 %. Spectra were fit using PyMCA software from fundamental parameters (Solé et al. 2007). Further details of methods are published in Bergmann et al. (2012), Anné et al. (2014) and Edwards et al. (2014).
X-ray absorbance spectroscopy (XANES and EXAFS)
X-ray absorbance near-edge structure spectroscopy (XANES) and extended x-ray absorbance fine structure (EXAFS) spectra were collected for FS003b at the microfocus beamline I18 of the Diamond Light Source synchrotron. The fossil material and extant human hair were analysed under fluorescence mode, the zinc standards in transmission mode. The emitted intensity of the zinc Kα line was recorded as a function of incident beam energy. A linear combination analysis was carried out in Athena version 071 (Ravel and Newville 2005).
Zinc K edge EXAFS were obtained using Kirkpatrick-Baez mirrors to produce a spot size of approximately 5 μm, a double crystal Si (111) monochromator to scan incident beam energy, and a 4-element Vortex silicon drift detector. Flux was estimated to be between 1011 and 1012 photons s−1. A zinc metal foil was used to calibrate edge energy. Data were analysed with SIXpack (Webb 2005) fit with FEFF6 using reference scattering paths.
(where v is the degrees of freedom in the fit). All shells presented here passed this significance test.
SEM and EDS
The soft tissues of the tadpole are composed of a thin film which has become cracked into a number of interconnected polygonal shapes (Fig. 1g). The surface of the film is covered in a dense layer of spherical bodies (Fig. 1c, d) 0.76 μm in diameter ±18 nm (average of ten measurements), as well as amorphous filamentous material occasionally entangling diatom skeletons (Fig. 1d). This filamentous material also appears on the matrix (Fig. 1e), which is a diatomaceous sediment consisting of a dense mat of siliceous diatom skeletons (Fig. 1f). Also present on both tadpole (Fig. 1g) and matrix (Fig. 1f) are chains of up to 20 unidentified spherical bodies 1.6 μm in diameter ±18 nm (average of ten measurements); these could be bacterial bodies or pollen grains. Pheomelanosomes examined for comparison were found within the fractured barbules of an extant grosbeak feather (Fig. 1h). They are spherical, 0.3 μm in diameter ±18 nm (average of ten measurements), and entangled within the filamentous material within the centre of the barbules.
Elemental composition of FS003a and FS003b, from ED and SRS-XRF analysis
Combinatorial analysis comparing the XANES of the fossil tadpole (FS003b) to standards containing zinc and a modern blonde human hair. The R-factor represents the sum of the squares of the residuals
Blonde human hair
Black human hair
Zinc Kα EXAFS analysis of standards and the Enspel tadpole (radial distances [r] in ångstroms). Values in parentheses are theoretical values for the standards
The cracked nature of the carbonaceous film of the tadpole, presence of small spherical bodies on its surface, and abundance of carbon and sulfur as seen in EDS are all found in other studies of such samples (Toporski et al. 2002; McNamara et al. 2009). As they have observed, the spherical bodies are fully comparable in both size and form to coccoid bacteria; however they are also of a similar form to pheomelanosomes in extant tissues (Fig. 1h). The presence of the pigment pheomelanin, contained within pheomelanosomes, has recently been demonstrated in amphibian skin (Wolnicka-Glubisz et al. 2012); however this does not appear to be common. The reliance on the geometry of fossilized melanosomes to denote original colour has also been shown to be unreliable due to the effects of diagenesis. Recent work (McNamara et al. 2013) has demonstrated that melanosome shape can change dramatically when exposed to high temperatures and pressures and therefore should not be taken as a definitive indicator of the type of melanin present. Indeed, though eumelanin in extant organisms is generally found in elongate melanosomes, a study identifying eumelanin in a Jurassic cephalopod found its ink sacs were covered in spherical bodies (Glass et al. 2012).
The presence of sulfur in other fossil tadpoles has been cited as evidence of organosulfur compounds which potentially aid the preservation of the organism by the process of sulfurization (McNamara et al. 2006). Sulfur may also be indicative of the presence of pheomelanin, as the sulfur-containing amino acid cysteine is one of the key triggers for its formation (Robins 1991; Prota 1992); however, we find no evidence of this in our samples. There is also no evidence of eu- or pheomelanin breakdown products in the Py-GCMS chromatograms (Dworzanski and Debowski 1985; Latocha et al. 2000), which show the presence of generic organic matter, including series of benzene and naphthalene compounds.
Though the distribution pattern of n-alkane/n-alkene doublets is slightly different in the tadpole and its enclosing matrix, it is not different enough to rule out the possibility of some organic matter transfer. However, the fact that the distribution patterns of fatty acid methyl esters are distinctly different makes this scenario very improbable, indicating that the organic material of the tadpole is most likely endogenous to the original organism. This is strengthened by the detection of phytol, indicative of plant material, in the matrix and not the tadpole, though it is only present in trace amounts. This runs counter to previous studies where this incorporation of molecular plant material has been suggested as the mechanism for the darkening of the carbonaceous film in such samples (Franzen 1985). The fact that a series of bacterial hopane compounds was detected in the matrix but not in the tadpole (Fig. 2) is suggestive but not conclusive of the absence of bacteria in the tadpole; it shows only the absence of aerobic bacteria. Indeed, there is potential evidence of bacterial chain structures on both matrix and tadpole (Fig. 1f, g) though this identification is not certain. The FTIR data suggests the presence of amino acids (C = O, NH and CN groups) within the tadpole and matrix. Given that the organic material of the tadpole is most likely to be endogenous, this suggests the amino acids are from the original tadpole tissues, whereas in the matrix, the signal is most likely from bacteria. However, the functional groups do not allow the identification of specific amino acids, and therefore, this distinction is again inconclusive. One area of the tadpole does however seem to show both C = O ketone and carboxylic acid peaks seen in eumelanin (Fig. 4); the C = O carboxylic acid group at 1350 cm−1 in particular is only seen in the tadpole. This is suggestive but not conclusive of the presence of melanin, due to the likely presence of co-occurring inorganic mineral peaks. Additionally, despite the fact that FTIR is useful in identifying melanin, it has been shown to be less so in distinguishing between eumelanin and pheomelanin (Liu et al. 2005).
Whilst levels of copper are higher in the fossil than the matrix, levels of zinc are much higher. This suggests that there is no or very little eumelanin present (Wogelius et al. 2011). The presence of zinc in fossilized tadpoles has been noted before (McNamara et al. 2009), but this is the first time it has been shown to be organometallic Zn(II) and to occur throughout the soft tissues. Zn(II), Cu(II), Mg(II), Fe(III) and Ca(II) are the most common metals present within both eu- and pheomelanin (Liu et al. 2005). In solutions of mild acidic or neutral pH, Cu(II) and Zn(II) have a higher binding affinity to eumelanin than Mg(II) and Ca(II) (Liu et al. 2004). Melanin can act like a metal ion sink during diagenesis, leading to the build-up of these metals where melanin was present and allowing them to be used as a melanin biomarker (Wogelius et al. 2011). Zinc is known to naturally accumulate in red human hair, and levels of zinc commonly exceed those of copper in eu-and pheomelanin extracted from human hair (Latocha et al. 2000). Another reason for higher zinc levels in the tadpole may be the presence of matrix metalloproteinases. These enzymes are interstitial collagenases (Brinckerhoff and Matrisian 2002) and members of the metzincin group of proteins, so named for, among other things, their dependence on the Zn(II) ion (Page-McCaw et al. 2007). These enzymes were first described after their study in tadpoles (Gross and Lapiere 1962) and are an essential part of the process of metamorphosis they undergo to develop an adult body. The tadpoles’ dependence on this protein may have increased the amount of zinc present within the tissues before death which could have subsequently become chelated to any melanin present and thereby retained in the soft tissues.
Whilst XANES and EXAFS data attest to the organic nature of the zinc and its similarity to that in pheomelanic samples, a comparison to such analyses of a pure pheomelanin sample as well as to bacterially bound zinc would be necessary to confirm or refute the presence of pheomelanin in the tadpole. The zinc in the human hair is also not a pure sample; there are likely to be contributions from both eu- and pheomelanin-bound zinc as well as that in proteins. The zinc XANES spectrum of the tadpole is also somewhat similar to that of zinc bound to bacterial enzymes (Meyer-Klaucke et al. 1999; Outten et al. 2001).
The fossil tadpole is covered in small spherical bodies consistent in size and form with both coccoid bacteria and pheomelanosomes.
The tadpole is preserved as a thin, carbon-rich, organic film, with no evidence of secondarily precipitated minerals on its surface.
The organic material in the tadpole is different from that in the encasing matrix and is therefore most likely original to the organism and not a result of transfer from the matrix. This rules out the possibility that the carbonaceous layer of the tadpole being a result of the incorporation of plant material.
Py-GCMS indicates the presence of bacterial biomarkers in the matrix but not the tadpole, though this does not completely rule out the possibility of there being bacteria on the tadpole
FTIR shows possible melanin groups in the tadpole but not the matrix; SRS-XRF analyses show high levels of zinc are associated with the soft tissues of the tadpole, and XANES and EXAFS indicate the zinc to be organic Zn(II) comparable to that seen in blonde human hair. Though further analysis of zinc in both pure pheomelanin and bacterial samples is required to completely exclude a bacterial origin, these results are in line with a pheomelanic origin.
Samples were kindly provided by M. Wuttke of the General Department of Cultural Heritage Rhineland Palatinate, Germany. FTIR analysis facilities were provided by K. Palmer of PerkinElmer and ESEM facilities by V. Vishnyakov of the DRIAM Analytical Service funded by the ESPRC. HEB was supported by NERC studentship grant. The authors would like to acknowledge the valuable contributions of two anonymous reviewers.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.