Heme compounds in dinosaur trabecular bone

Abstract

Six independent lines of evidence point to the existence of heme-containing compounds and/or hemoglobin breakdown products in extracts of trabecular tissues of the large theropod dinosaur Tyrannosaurus rex. These include signatures from nuclear magnetic resonance and electron spin resonance that indicate the presence of a paramagnetic compound consistent with heme. In addition, UV/visible spectroscopy and high performance liquid chromatography data are consistent with the Soret absorbance characteristic of this molecule. Resonance Raman profiles are also consistent with a modified heme structure. Finally, when dinosaurian tissues were extracted for protein fragments and were used to immunize rats, the resulting antisera reacted positively with purified avian and mammalian hemoglobins. The most parsimonious explanation of this evidence is the presence of blood-derived hemoglobin compounds preserved in the dinosaurian tissues.

Figure 1

(A) Reverse phase HPLC profiles of extracted tissues and controls; 20 μl of concentrated extract was injected onto a phenyl analytical column and monitored at 410 nm. From top to bottom, samples are: T. rex tissues in extraction buffers, sandstone matrix, and extraction buffers alone. IP, point of injection with time running to the right. (B) UV/visible absorbance spectrum of dinosaurian tissue extract, precipitated in HCl/acetone/ether as described. The large peak at 405 nm is within the range of variation seen for heme compounds and is characteristic of these compounds. The smaller peaks at the longer wavelengths (quasi-allowed, or Q-, bands) also are characteristic of these compounds. The Inset represents an expansion of the longer wavelengths of the tracing for visualization and identification of individual peaks in this region. Plant and sandstone samples, similarly extracted, showed no specific absorbance at this wavelength.

Figure 2

High resolution, solution phase proton NMR profile obtained from extracts of dinosaur tissues precipitated in HCl/acetone/ether as described. For this spectrum, samples were filtered through a 5000 MW cut-off filter, and the higher molecular weight fraction was again filtered through a 30,000 molecular weight filter for better resolution before being subjected to NMR. The arrows indicate resonance in regions that are consistent with the presence of a paramagnetic compound such as heme.

Figure 3

A, ESR profile of the HCl/acetone/ether precipitate and B, the same extract, vacuum degassed, flushed with nitrogen, and injected with 40 μl of anaerobic 0.1 M sodium nitrite followed by 40 μl of anaerobic 0.1 M sodium dithionite. Scan range, 1000 g; field set, 3240 g; modulation amplitude, 10 g; microwave power, 3.4 mW; microwave frequency, 9.132 GHz; receiver gain, 2.0 × 10⁵; temperature, 77 K.

Figure 4

Raman profile of dinosaur extracts, precipitated in HCl/acetone/ether, as described. Human whole blood and commercially prepared pigeon met–hemoglobins are shown for comparison. Roman numerals indicate peaks that are mentioned in the literature as marker bands for heme compounds.

Figure 5

Rats immunized with T. rex tissue extracts produce antibodies that recognize hemoglobin. Antisera drawn from rat IN9 before and after the immunization procedure were compared at increasing dilutions for reactivity against purified turkey hemoglobin by ELISA. The intensity of absorbance measured at 405 nm is directly proportional to the quantity of anti-hemoglobin antibodies present in each serum. No reactivity of the antisera with components of sandstone and plant extracts was noted.

Figure 6

(A) Electrophoretic separation of commercially prepared hemoglobins stained with Coomassie blue. (B) Western blot of the hemoglobins in A exposed to rat preimmune sera. No reaction to any of the hemoglobins is visualized. (C) Positive and specific binding between dinosaur antisera and purified hemoglobins on Western blot. Background was high because of necessary low dilutions of the antisera. The reaction to turkey hemoglobin was measureable but could not be reproduced photographically. There was no reaction to the snake hemoglobin, either by immunoblot or ELISA, nor was there reactivity with plant or sandstone extracts. Protein-specific reactivity, which is variable in intensity, supports a true immune response rather than artifact.