Iron-coated Komodo dragon teeth and the complex dental enamel of carnivorous reptiles

Komodo dragons (Varanus komodoensis) are the largest extant predatory lizards and their ziphodont (serrated, curved and blade-shaped) teeth make them valuable analogues for studying tooth structure, function and comparing with extinct ziphodont taxa, such as theropod dinosaurs. Like other ziphodont reptiles, V. komodoensis teeth possess only a thin coating of enamel that is nevertheless able to cope with the demands of their puncture–pull feeding. Using advanced chemical and structural imaging, we reveal that V. komodoensis teeth possess a unique adaptation for maintaining their cutting edges: orange, iron-enriched coatings on their tooth serrations and tips. Comparisons with other extant varanids and crocodylians revealed that iron sequestration is probably widespread in reptile enamels but it is most striking in V. komodoensis and closely related ziphodont species, suggesting a crucial role in supporting serrated teeth. Unfortunately, fossilization confounds our ability to consistently detect similar iron coatings in fossil teeth, including those of ziphodont dinosaurs. However, unlike V. komodoensis, some theropods possessed specialized enamel along their tooth serrations, resembling the wavy enamel found in herbivorous hadrosaurid dinosaurs. These discoveries illustrate unexpected and disparate specializations for maintaining ziphodont teeth in predatory reptiles.


Page 3
Supplementary Figure 1.Pigmented cutting edges and tooth tips in museum specimens of Varanus komodoensis Supplementary Figure 1.Pigmented cutting edges and tooth tips in museum specimens of Varanus komodoensis.a Functional tooth (AMNH 74606).b Functional tooth (AMNH 37913).c Replacement teeth, from below the gumline and unworn (AMNH 37912).d Functional tooth (AMNH 37909).e Polished thick section of a functional tooth; the pigmented region is still embedded in a thin layer of resin (J94036-4).f Polished thick section along the mesial serrations of a functional tooth; the clear enamel is exposed on the surface of the polished block and the pigmented regions are still embedded in resin (J94036-2).Asterisks indicate orange pigmented regions.Abbreviations: AMNH American Museum of Natural History (New York, New York, USA), de dentine, en enamel.Scale bars in a-d are 1mm, e-f are 0.1mm.

Supplementary Figure 2. Additional synchrotron-based X-Ray MicroFluorescence (S-µXRF) and Scanning Electron Energy-Dispersive x-ray Spectroscopy (SEM-EDS)
elemental maps for Varanus komodoensis teeth.a S-µXRF map (0.5 µm resolution) of iron (red), calcium (green), and zinc (blue) in a horizontal thick section taken through an unerupted tooth crown (MoLS X-263).Map shows iron and zinc sequestration along the outer enamel of a distal serration.b S-µXRF map (0.5 µm resolution) of iron (red), calcium (green), and zinc (blue) in a horizontal thick section taken through the same tooth crown.Map shows iron and zinc sequestration along the outer enamel of another distal serration of MoLS X-263 (same as main text Fig. 2g).c S-µXRF map (0.5 µm resolution) of iron (red), calcium (green), and zinc (blue) in the same horizontal thick section taken through MoLS X-263 as in b.Map shows iron and zinc sequestration along the outer enamel of a mesial serration of MoLS X-263 (same as main text Fig. 2i).d S-µXRF map (0.5 µm resolution) of iron (red), calcium (green), and zinc (blue) in a longitudinal thick section taken parallel to the serrations.Map shows iron and zinc sequestration along the outer enamel of mesial serrations of X-263.e S-µXRF map (0.5 µm resolution) of iron (red), calcium (green), and zinc (blue) in a longitudinal thick section through an erupted, functional tooth crown (J94036-1).Map shows iron and zinc sequestration along the outer enamel of a distal serration.g S-µXRF map of a distal serration of J94036-5 (0.5 µm resolution).h S-µXRF map through the enamel and dentine off-serration (0.5 µm resolution).Note the lack of prominent iron signal.i Scanning Electron Microscope image of mesial serrations of an erupted, functional tooth (J94036-2).j SEM-Energy Dispersive Spectroscopic image of calcium, k iron, and l oxygen along the mesial serrations of J94036-2.Abbreviations: de dentine, edj enamel-dentine junction, en enamel, zn zinc-enriched region of enamel.Asterisks refer to iron-coated regions.Supplementary Figure 3. Elemental maps derived from Laser Ablation Time-of-Flight Inductively-Coupled Mass Spectrometry (LA-TOF-ICP-MS) of Varanus komodoensis tooth serrations.Maps were first normalized to the calcium counts to account for artefacts that arose from differential ablation of enamel vs dentine (see Methods and Supplementary Fig.

Laser Ablation Inductively-Coupled Plasma Mass Spectrometry (LA-ICP-MS) data processing
Elemental maps generated from the LA-ICP-MS experiments were sensitive to the mechanical properties of each tissue.For example, for a given transect of ablation along the extant Alligator tooth sample, the laser removed more dentine (which is softer) from the tooth than enamel (which is significantly harder).Consequently, element counts were underrepresented in the enamel relative to the dentine.This bias was especially evident in raw elemental maps for calcium, where the results initially indicated higher calcium counts in the dentine than in the enamel, which was opposite to all XRF and SEM-EDS data.To account for this artifact, we applied a correction factor on the LA-ICP-MS count data for iron, calcium, and zinc.
To implement a correction to the raw element counts, we first estimated the amount of enamel lost during the ablation process and compared it with the amount dentine (Supplementary Fig. 8d, f).We calculated the depths of ablation along the enamel and dentine by generating profile lines from z-stacked microscope images of the tooth surface using the Keyence digital microscope's 3-D imaging function.We calculated an average step height for dentine in Alligator tooth 2 ROI1 of -5.43 µm, whereas we could not reliably detect a step height through the enamel.We therefore had to rescale the elemental maps for iron, calcium, and zinc for this tooth, given that we ablated approximately 5.4 times more dentine than enamel.We did this by manually masking every pixel in the dentine in ImageJ and then dividing this region in each elemental map by the difference in step height (5.433).This resulted in re-scaled maps where the counts between enamel and dentine were comparable with elemental data obtained from the other techniques.
To more accurately depict the relative differences in elemental concentrations, each map needed to be processed differently.For the iron map, the counts for iron along the surface of the enamel were approximately three orders of magnitude higher than for the counts in the rest of the enamel and dentine.We therefore rescaled this map to a log scale (from background to ~3000 counts in the outer enamel).However, zinc and calcium showed much smaller differences along the outer and inner enamel, as well as the dentine.We presented the zinc map from 0 (background) to the 95 th percentile to eliminate outliers and produce a more realistic distribution of zinc through the tooth.Calcium also showed much smaller count differences between enamel and dentine and is therefore also presented on a linear scale from 0 to the 95 th percentile to eliminate outliers.The corrected maps here and in the main text are therefore relative representations of the concentrations of iron, calcium, and zinc, and therefore do not include count scale bars for this reason.We also could not detect any difference in step height between ablated enamel and ablated dentine in the fossil tyrannosaurid teeth and therefore present the LA-ICP-MS elemental maps for these teeth without any correction (Supplementary Figs.11, 13).Supplementary Figure 8.
Step-height correction of elemental maps for Alligator mississippiensis tooth.a Raw iron map from LA-ICP-MS.b raw calcium map.Note the higher counts of calcium in the dentine (lower region) compared with the enamel.This is in direct opposition to all other calcium maps generated from XRF and EDS analyses.c Raw zinc map.Note higher zinc counts in the dentine compared with the enamel.d Profile line drawn through the ablated region of enamel in the tooth sample using the profile function through a z-stacked image of the sample using the Keyence VHX digital microscope.No consistent step height could be detected, suggesting negligible ablation of the enamel surface.e Profile line drawn through the dentine.We measured a step height of 5.43µm between ablated and unablated regions of the dentine.This was used as a correction factor for the derived elemental maps.h corrected map for iron, i calcium, and j zinc.Note reversal of counts for zinc and calcium between enamel and dentine in the corrected maps.Abbreviations: de dentine, en enamel, re resin.Distal view of "Tooth 2".l Lingual view of "Tooth 2" showing lack of any obvious pigmentation along the cutting edges under plain lighting.Orange colouration is most obvious in polished thick sections.m, n Raw LA-ICP-MS maps for iron along the enamel and dentine of "Tooth 2" showing the presence of an iron-enriched outer enamel layer in the same positions as the orange-coloured region in j. o Region of interest along the outer layers of enamel in j where more detailed nanomechanical testing was undertaken to directly compare the mechanical properties of pigmented and non-pigmented enamel.

Sample environment
Step size (microns) 8 and text therein for explanation).The calcium map is therefore not shown.a White light image of ablated region of distal serrations of J94036-1 (dashed lines).b Map of magnesium, showing higher counts in the dentine.c Iron map showing its restriction to the outer layer of enamel.d Map of iron and magnesium.e Zinc map, showing its restriction to the outer enamel layer.f Map of zinc and magnesium.Asterisks indicate position of pigmented enamel.Abbreviations: de dentine, en enamel.Supplementary Figure 4. Serration and tooth tip colouration in museum specimens of Varanus.All images taken in lingual or labial views.Asterisks indicate orange pigmentation.All scale bars except the one in d are 1mm.Abbreviations: AMNH American Museum of Natural History, SAMA South Australian Museum.Supplementary Figure 6.Comparisons of tooth crown colouration, iron and zinc sequestration along cutting edges in extant crocodylian teeth.a White light (WL) and b Laser Stimulated Fluorescence images of two teeth from a Varanus komodoensis (Zoological Society of London) showing differential fluorescence of the serrations (asterisks) for comparisons with crocodylian samples.c Anterior tooth of Tomistoma schlegelii in distal view.d LSF image of distal carina, showing different fluorescence patterns between carina (asterisks) and the rest of the tooth crown, similar to V. komodoensis.e Posterior tooth of T. schlegelii under white light (WL) and f Laser Stimulated Fluorescence (LSF) showing differential fluorescence of carina (asterisks).g Mesial view of a posterior tooth crown of Osteolaemus tetraspis.h Closeup of tooth tip showing pigmented cutting edges.i Longitudinal section taken through mesial and distal cutting edges of tooth in g. j Synchrotron X-Ray Microfluorescence (S-µXRF) map of iron (red), calcium (green), and zinc (blue) taken from tip of tooth section in h.Iron and zinc are restricted to the outermost enamel layers along the tooth tip and cutting edges.k Mesial view of a tooth crown of Crocodylus porosus.l Closeup of tooth tip in j, showing a lack of obvious pigmentation along the carina under white light.m Longitudinal section taken through the mesial and distal carinae for S-µXRF analysis.n S-µXRF map of iron (red), calcium (green), and zinc (blue), showing iron and zinc sequestration along the tip and cutting edge enamel in the same tooth as in l.Abbreviations: ca carina, de dentine, en enamel Supplementary Figure 7. Comparisons of elemental compositions of extant and fossil crocodylian teeth.a Posterior tooth of Osteolaemus tetraspis before (left) and after sectioning along mesiodistal axis (right).b S-µXRF map of iron (red) and calcium (green) along the tooth tip.Iron is located only in the outermost enamel layers.c Posterior tooth of a fossil crocodylian from Dinosaur Provincial Park (UALVP 60546) with similar morphology to O. tetraspis before (left) and after sectioning along mesiodistal axis (right).d S-µXRF map of iron (red) and calcium (green) showing the abundance of iron within the dentine and enamel.e Anterior tooth of Crocodylus porosus before (left) and after sectioning along the mesiodistal axis (right).f S-µXRF map of iron (red) and calcium (green) along the tooth tip and g along the carina.h Anterior tooth of a fossil crocodylian from Dinosaur Provincial Park (UALVP 60550) with similar morphology to C. porosus before (left) and after sectioning along mesiodistal axis (right).i S-µXRF map of iron (red) and calcium (green) along the tooth tip showing the abundance of iron within the dentine and enamel.j S-µXRF map of iron (red) and calcium (green) along a carina showing the abundance of iron within the dentine and enamel.Abbreviations: de dentine, en enamel.Asterisks indicate positions of iron-enriched enamel.

Figure 9 .
Comparisons of Iron X-ray AbsorptionNear Edge Structure (Fe-XANES) spectra for the iron layers in extant beaver, crocodile, and Komodo dragon.a Comparisons of XANES spectra for magnetite, haematite, and ferrihydrite standards with the spectra derived from the iron coatings in a V. komodoensis tooth.The V. komodoensis spectra most closely resembled that of ferrihydrite.b Closeup of iron coatings in a polished thick section of a V. komodoensis tooth (J94036-4).c Closeup of iron layer within the outer enamel of a A. mississippiensis tooth ("Tooth 2").d Closeup of iron layer within the outer enamel of a Castor canadensis tooth (UALVP 56017-3).e Comparisons of Fe-XANES spectra of the iron layers in V. komodoensis, C. porosus, and C. canadensis.Though consistent with ferrihydrite, the iron layers in V. komodoensis differ from those of the iron layers in the other two species.Abbreviations: de dentine, en enamel.Asterisks indicate positions of pigmented enamel layers.Supplementary Figure 10.Laser-Stimulated Fluorescence (LSF) imaging of cutting edges in selection of fossil theropod teeth from the NHMUK collections.Note that none of the samples show differential colouration along the cutting edges (worn edges appear darker due to the exposure of underlying dentine).a Distal serrations of a tooth of the tyrannosaurid Albertosaurus showing no differences in fluorescence pattern between serrations and the rest the crown.b Distal serrations of another Albertosaurus tooth showing similar fluorescence patterns between serrations and rest of crown.Blue regions are areas covered in adhesives.c Distal serrations of a tooth of Tyrannosaurus rex showing no differential fluorescence patterns along the crown.Blue colour is the result of the fluorescence of adhesives.d Distal serrations of a partial crown of the megalosaurid Megalosaurus bucklandii.Serrations show no differential fluorescence compared with the remainder of the crown.e Distal serrations along a tooth of the theropod "Megalosaurus" insignis showing no difference between the serrations and remainder of the crown.f Distal serrations of a tooth of "Megalosaurus" dunkeri showing no difference between serrations and remainder of crown.g Distal carina of a spinosaurid tooth showing no difference in fluorescence between cutting edge and remainder of crown.Dark patches along the carina result from breakage of the enamel and exposure of the underlying dentine.Supplementary Figure 11.LA-ICP-MS elemental maps for two tyrannosaurid teeth.a Longitudinal section through distal serrations of UALVP 60555 with elemental map for barium, b Calcium, c Iron, d Zinc, e Magnesium, f Yttrium, g Strontium.h White light image of polished thick section through a distal serration of UALVP 60554.i elemental map of calcium, j Magnesium, k Yttrium, l Barium, m Iron, n Strontium, o Zinc.None of these elemental distributions match those seen from elemental analyses of extant Varanus komodoensis or crocodylian teeth.Supplementary Figure 12.Additional synchrotron-based X-Ray MicroFluorescence (S-µXRF) elemental maps for two tyrannosaurid teeth.a Distal view of a tyrannosaurid premaxillary tooth (UALVP 60553) used for S-µXRF analyses.b Overview image of horizontal section taken through UALVP 60553, showing position of S-µXRF elemental maps in c-f.c S-µXRF map of horizontal section through a premaxillary tooth serration, showing distribution of iron, d Calcium, e Zinc, and f Composite of all three elements.Note the lack of iron and zinc sequestration along the serration enamel towards the bottom left of the image.Instead, iron counts are highest in the dentine and along cracks in the tooth, suggesting iron concentrations are primarily driven by fossilization artifacts.g Closeup of a longitudinal section through a mesial serration of another tyrannosaurid tooth (UALVP 53472).h Lower magnification image showing position of mapped serration.i S-µXRF elemental map for iron, j Calcium, k Zinc, l Tungsten, and m Barium.Abbreviations: UALVP University of Alberta Laboratory of Vertebrate Paleontology.Supplementary Figure 13.LA-ICP-MS elemental maps for a dromaeosaurid dinosaur tooth (UALVP 61165).a Longitudinal section through distal serrations of UALVP 61165 with elemental map for barium, b Calcium, c Iron, d Magnesium, e Strontium, f Yttrium, g Zinc.h Transverse section through a distal serration in UALVP 61165 showing the elemental map for barium, i Calcium, j Iron, k Magnesium, l Strontium, m Yttrium, n Zinc.None of these distributions match those of extant Varanus komodoensis or crocodylian teeth.Abbreviations: UALVP University of Alberta Laboratory of Vertebrate Paleontology.Supplementary Figure 14.Representative XRF spectra for extant reptile and tyrannosaurid teeth examined in this study.a XRF spectra taken from the iron-enriched region, enamel, and dentine of a tooth of Varanus komodoensis (Beamline ID-21, European Synchrotron Radiation Facility, Grenoble, France).b XRF spectra from iron-enriched region, enamel, and dentine of a posterior tooth of Osteolaemus tetraspsis (Beamline BM-28, European Synchrotron Radiation Facility, Grenoble, France).c XRF spectra from ironenriched region, enamel, and dentine of an anterior tooth of Crocodylus porosus (Beamline BM-28, European Synchrotron Radiation Facility, Grenoble, France).d XRF spectra from analogous positions of iron-enriched regions in extant reptiles, enamel, and dentine taken along the serration of a tyrannosaurid tooth (UALVP 53472) (Beamline B-16, Diamond Light Source, Oxfordshire, UK).Note the differences in intensities of iron, calcium, and zinc signals between the three extant reptile teeth and that of the tyrannosaurid.Red boxes in inset elemental map images correspond to regions where "iron hotspot" spectra were taken in each tooth.Grey boxes indicate positions where "dentine" spectra were taken.Black boxes indicate positions where "enamel" spectra were taken.Supplementary Figure 15.Scanning Electron Microscope (SEM) imaging of enamel microstructure across tyrannosaurid tooth crowns.a Lateral view of the partial tyrannosaurid tooth UALVP 60556.b Low-magnification SEM image of horizontal section taken through a distal serration of UALVP 60556.c High-magnification SEM image of the wavy enamel along the serration of UALVP 60556.d High-magnification SEM image of the columnar enamel found elsewhere on the same tooth crown.e Labiolingual view of partial tyrannosaurid tooth (UALVP 60554).f Low-magnification SEM image of a horizontal section through one of the distal serrations of UALVP 60554.g High-magnification SEM image of the wavy enamel found along the distal serration of UALVP 60554.h High-magnification SEM image of columnar enamel along the rest of the same tooth crown.i Distal view of a tyrannosaurid premaxillary tooth (UALVP 60553).j Low-magnification SEM image of a horizontal section taken through UALVP 60553, showing one of the distal serrations.k High-Closeup of distal serrations.h Wholeview of horizontal section taken through Dromaeosaurus sp.tooth for SEM.i SEM image of serration enamel showing mostly parallel crystallite enamel, with slight divergences of crystallites along mid-axis of serration.j SEM image of off-serration, parallel crystallite enamel.k Complete Dromaeosauridae indet.tooth crown sectioned for SEM.l Closeup of distal serrations of dromaeosaurid tooth.m Wholeview of longitudinal section used for SEM.n SEM image of serration enamel, showing microunit and possible wavy enamel (crystallite bundles are not parallel to neighbouring bundles).o SEM image of offserration enamel, showing simpler, parallel crystallites.Abbreviations: en enamel, de dentine, re resin.Supplementary Figure 17.Histological comparisons of wavy enamel along tyrannosaurid serrations and hadrosaurid teeth.a Lingual view of tyrannosaurid tooth UALVP 60556.b Polished thick section through worn mesial serrations of a tyrannosaurid tooth (UALVP 60555).Black arrowheads indicate directions of wear, based on surface striations prior to sectioning (see Suppl.Fig. 15f, g).c Longitudinal thin section of serrations in a tyrannosaurid tooth (UALVP 60398) under cross-polarized light.d Higher-magnification image of a serration in c, showing wavy enamel effect under cross-polarized light.Arrowheads indicate presumed direction of wear.e Isolated hadrosaurid tooth (ROM 58630), showing position of horizontal section in g-i.f Longitudinal section through three functional maxillary teeth (image flipped for comparisons) of a hadrosaurid dental battery (ROM 696) showing complexity of the grinding surface in a hadrosaurid dinosaur.g Horizontal thin section through an isolated hadrosaurid tooth (UALVP 55127), showing general histological features and position of higher-magnification images in subsequent panels.h Higher magnification image of g under cross-polarized light, showing similar wavy enamel optical effect as that seen in the tyrannosaurid serrations.i Higher magnification image of wavy enamel under cross-polarized light.Abbreviations: ce cementum, de dentine, en enamel, idf interdental fold.Supplementary Figure 18.Schematic representation of the machine learning based pipeline used to cluster orientation data by similarity to facilitate parameter extraction and 2D fitting of the 002-diffraction peak(s).a 1D orientation data, vertically offset for clarity, demonstrates greater variability compared with conventional diffraction data.Data is grouped through the application of b principal components analysis and c k-means clustering.Fitting parameters, including peak position and width, are extracted from the orientation data d and are subsequently used for e 2D pseudo-Voigt fitting of diffraction images truncated about the 002 peak(s) to obtain preferred orientation and c axis parameters for constituent crystallite populations.Supplementary Figure 19.Synchrotron-based X-Ray Micro-diffraction (S-µXRD) maps of two serrations in longitudinal section (UALVP 53472).The crystallographic c axis lattice and texture parameters of the three constituent crystallite populations within the tooth enamel are shown arranged by columns for population one (a and d), two (b and e) and three (c and f).Lines within each pixel indicate the average preferred apatite crystal orientation and hotter colours correspond to more highly textured regions (lower full-width half maxima).Orientation direction and FWHM in d, e, and f were used to calculate average values illustrated in main text Figure 4p.Supplementary Figure 20.Synchrotron-based X-Ray Micro-diffraction (S-µXRD) map of a horizontal section through a serration and the surrounding enamel (UALVP 60554) with statistical comparisons of Full Width Half Maxima (FWHM) of enamel on-and offserration.Heat map (left) is derived from the same region as in main text Fig. 4q.Two-tailed t-tests were conducted to compare enamel and dentine regions on either side of the serration.Laterally symmetrical regions (e.g., a and b) were grouped together as single samples and compared with other regions.Mean Full Width Half Maxima (FWHM) values for grouped regions and associated standard deviations are summarised in the table (right).Statistical comparisons between each region were all statistically significant (p<0.001),indicating that each region contained apatite crystallite populations that significantly differed in terms of their Full Width Half Maxima (FWHM), which is a measure of the degree of variation around the principal orientations (small lines in each pixel) derived from S-µXRD analyses.See Extended Data 5 for full statistical analysis outputs and raw data for each grouping.Abbreviations: FWHM Full Width Half Maximum.Heat map colours indicate magnitude of FWHM, with hotter colours indicating lower FWHM values and therefore more highly ordered crystallites around a preferred (principal) orientation.Note that the lowest FWHM values are concentrated towards the serration tip (top right corner of map), corresponding to the wavy enamel identified under SEM.Supplementary Figure 21.Nanoindentation analysis of Varanus komodoensis J94036-2.a Enamel and dentine indentation hardness plotted against the indentation elastic modulus.b Coaxial light image of the polished longitudinal thick section of J94036-2 used for the nanoindentation tests.c Region of interest imaged in the nanoindenter prior to the experiment.d Heat map of Hardness (GPa) in the region indented in c. e Heat map of Reduced Elastic Modulus (GPa) in the region indented in c. f Region of interest imaged in the nanoindenter prior to the experiment.g Heat map of Hardness (GPa) in the region indented in f. h Heat map of Reduced Elastic Modulus (GPa) in the region indented in f.Abbreviations: de dentine, en enamel, re resin.See Methods for experimental parameters for nanoindentation tests.

Supplementary Figure 23 .
Nanoindentation analysis of a tyrannosaurid tooth (UALVP 60555).a Enamel and dentine hardness plotted against the reduced elastic modulus.Note the lack of separation between dentine and enamel measurements and the higher magnitude of both metrics compared with data from extant reptile teeth.b Wholeview image of mesial serrations in polished thick section under coaxial light, showing region of interest in nanoindentation experiments.c Wholeview image of distal serrations under coaxial light, showing region of interest for nanoindentation experiments.d Region of interest along a mesial serration imaged in the nanoindenter prior to the experiment.e Heat map of hardness measured from first region of interest along the middle of the mesial serration.Note the similarity between the enamel and underlying dentine.f Heat map of reduced elastic modulus in same region.g Second region of interest along mesial serration imaged in the nanoindenter prior to the experiment.h Heat map of hardness measured from first region of interest along the middle of the mesial serration.Note the similarity between the enamel and underlying dentine.i Heat map of reduced elastic modulus in same region.j Third region of interest along distal serration imaged in the nanoindenter prior to the experiment.k Heat map of hardness of enamel and dentine measured in third region of interest.l Heat map of reduced elastic modulus in same region.m Fourth region of interest along a distal serration imaged in the nanoindenter prior to the experiment.n Heat map of hardness of enamel and dentine measured in the fourth region of interest.o Heat map of the reduced elastic modulus in the same region.Abbreviations: de dentine, en enamel, re resin.Supplementary Figure 24.Comparisons of indentation hardness and reduced elastic moduli of extant Varanus komodoensis.Alligator mississippiensis, and tyrannosaurid teeth.a Combined plot of nanomechanical properties of tyrannosaurid (red), Varanus komodoensis (black), and Alligator mississippiensis enamel and dentine.Tyrannosaurid tooth indents yielded higher hardness and elastic moduli compared with equivalent regions in the two extant reptiles.b Two hardness heat maps (Supplementary Fig.18k, n) superimposed on polished thick section of the tyrannosaurid tooth Note the more subtle differences in hardness between the dentine and enamel in the fossil tooth, due to chemical and structural alterations to the two tissues.c Comparisons with a relative hardness map of Varanus komodoensis (Supplementary Fig.17g).Note the stark contrast between the enamel and dentine.These comparisons demonstrate the impact of fossilization on the mechanical properties of tyrannosaurid enamel and dentine.tyrannosaurid tooth (UALVP 60553) under reflected light, highlighting many (post-mortem) cracks (arrows) through the columnar enamel of the crown, and the lack of these cracks within the wavy enamel of the serrations.Abbreviations: de dentine, en enamel.Unless otherwise indicated, arrows indicate worn surfaces of teeth.Supplementary Figure 26.Scanning Electron Microscope (SEM) imaging of acid-etched serrations in Varanus komodoensis, showing acid-resistance of the outer iron-rich coating.a Mesial serrations of a V. komodoensis tooth under SEM following a 30-second immersion in 1M HCl.The resistance of the outer iron-rich coating (asterisks) lead to the formation of an overhang (arrow), created by the dissolution of the underlying enamel (J94036-4).b Similar feature after 30 seconds of etching in 1M HCl.The underlying enamel has nearly completely dissolved away, leaving an unsupported outer shell of the iron-rich material (asterisks), which collapsed under its own weight (arrow).c Serration in J94036-2 showing nearly complete dissolution of enamel after 30 seconds of 1M HCl etching.d Higher magnification image showing dissolution of enamel and preservation of the outer iron-rich coating as an unsupported shell over the dentine (arrow).Abbreviations: de dentine, en enamel, re resin.

Survey of tooth pigmentation in reptiles.
Supplementary Table1.

Table 4 . Raw data and t-tests comparing enamel hardness and elastic modulus along unpigmented and pigmented regions via nanoindentation in an Alligator tooth.
See Supplementary Figure22for locations of indents on tooth specimen.