Substitutions in woolly mammoth hemoglobin confer biochemical properties adaptive for cold tolerance

Journal name:
Nature Genetics
Year published:
Published online

We have genetically retrieved, resurrected and performed detailed structure-function analyses on authentic woolly mammoth hemoglobin to reveal for the first time both the evolutionary origins and the structural underpinnings of a key adaptive physiochemical trait in an extinct species. Hemoglobin binds and carries O2; however, its ability to offload O2 to respiring cells is hampered at low temperatures, as heme deoxygenation is inherently endothermic (that is, hemoglobin-O2 affinity increases as temperature decreases). We identify amino acid substitutions with large phenotypic effect on the chimeric β/δ-globin subunit of mammoth hemoglobin that provide a unique solution to this problem and thereby minimize energetically costly heat loss. This biochemical specialization may have been involved in the exploitation of high-latitude environments by this African-derived elephantid lineage during the Pleistocene period. This powerful new approach to directly analyze the genetic and structural basis of physiological adaptations in an extinct species adds an important new dimension to the study of natural selection.

At a glance


  1. Evolution of the genes encoding the single adult-expressed hemoglobin component of three members of the Elephantidae family.
    Figure 1: Evolution of the genes encoding the single adult-expressed hemoglobin component of three members of the Elephantidae family.

    The position of nucleotide and amino acid changes are shown within the three coding regions (exons) of the HBA-T2 and HBB/HBD globin genes (shown as open horizontal boxes along each branch) superimposed on the Elephantidae phylogeny4. Branch lengths are not proportional to geologic time. Ancestral nucleotide and amino acid residues are shown above, and derived nucleotide and amino acid residues are shown below the exons. The numbers above and the letters to the right of the vertical lines denote the amino acid residue, whereas the numbers below and the letters to the left of each vertical line indicate nucleotide position relative to the ATG initiation codon. Thick vertical lines with bold characters indicate nonsynonymous substitutions, and thin vertical lines represent synonymous substitutions, with red, green and blue characters and bars representing replacements at codon positions 1, 2 and 3, respectively. We employed the α-globin and β-globin chain sequences of other afrotherian mammals (Echinops telfairi (GenBank P24291, P24292), Procavia habessinica (P01957, P02086) and Trichechus inunguis (P07414, P07415)) to deduce the direction of amino acid substitutions.

  2. Surface model of a chimeric Asian elephant (left) and mammoth (right) deoxyhemoglobin molecule bound to 2,3-bisphosphoglycerate (BPG).
    Figure 2: Surface model of a chimeric Asian elephant (left) and mammoth (right) deoxyhemoglobin molecule bound to 2,3-bisphosphoglycerate (BPG).

    The locations of the three mammoth-specific amino acid substitutions are highlighted in blue, and the positions of each heme group are denoted by ball-and-stick diagrams. Regions highlighted in yellow denote positively charged residues (Lys82 of the β/δ-chain (hereafter denoted β/δ82 Lys), β/δ143 His and the amino group of β/δ1 Val) implicated in the binding of BPG to elephant deoxyhemoglobin11, 12, 14. Note that because the polar hydroxyl side chain of β/δ12 Thr of Asian elephant deoxyhemoglobin (and human hemoglobin25) forms a hydrogen bond with the carbonyl group of β/δ8 Lys (green), the negatively charged side chain of β/δ79 Asp (red) is free to project into the BPG binding pocket, where it would tend to repel this anion. Conversely, the methyl side chain of β/δ12 Ala in mammoth hemoglobin cannot bond with β/δ8 Lys, allowing the lysyl side chain to form an ionic interaction with β/δ79 Asp of the E helix and neutralizing its charge (light red). The mammoth-specific β/δ101 Gln residue is spatially distant from this charged cluster and cannot contribute to BPG binding15, 16, 17. However, this central cavity residue alters electrostatic interactions at the sliding interface of the molecule that both destabilizes the low-affinity deoxy-state protein and creates additional proton-linked chloride binding sites in mammoth hemoglobin (see main text for details).

  3. Oxygen equilibrium curves of woolly mammoth (blue) and Asian elephant hemoglobin (red) at 37 [deg]C and pH 7.0.
    Figure 3: Oxygen equilibrium curves of woolly mammoth (blue) and Asian elephant hemoglobin (red) at 37 °C and pH 7.0.

    In the absence of allosteric effectors (solid lines), the mammoth β/δ-chain E101Q substitution destabilizes the tense-state (deoxy) conformation, leading to a protein phenotype with an intrinsic affinity nearly two times higher (curve is shifted to the left). This radical increase in O2 affinity (which would drastically impair tissue O2 offloading) is almost precisely compensated by enhanced H+, Cl and 2,3-BPG binding to mammoth hemoglobin that right-shifts the curve more strongly than in Asian elephant hemoglobin. As a result, the overall O2 affinity of mammoth hemoglobin in the presence of red cell effectors is nearly identical to that of Asian elephants at 37 °C (red dashed line). However, the increased effector binding to mammoth hemoglobin lowers the effect of temperature on O2 affinity, facilitating the release of O2 at cold temperatures in relation to Asian elephant hemoglobin.

  4. Mean enthalpy of oxygenation ([Delta]H; kJ mol-1 O2) values of woolly mammoth (blue columns) and Asian elephant (red columns) hemoglobin in the absence and presence of effector molecules.
    Figure 4: Mean enthalpy of oxygenation (ΔH; kJ mol−1 O2) values of woolly mammoth (blue columns) and Asian elephant (red columns) hemoglobin in the absence and presence of effector molecules.

    Error bars for each treatment ('stripped', 0.1 M Cl, and 0.1 M Cl plus saturating levels of 2,3-bisphosphoglycerate (Cl + BPG)) are ± s.e.m. of four calculated ΔH values: one from O2 equilibria measured at 10 °C and 25 °C at pH 7.0; one from measurements at 25 °C and 37 °C at pH 7.0; one from measurements at 10 °C and 25 °C at pH 7.4; and one from measurements at 25 °C and 37 °C at pH 7.4. The temperature dependence of the oxygenation process is governed by the associated overall ΔH of this reaction19, where numerically low ΔH values correspond to small effects of temperature on hemoglobin-O2 affinity. Student's unpaired t-tests (α = 0.05, n = 4) illustrate that the intrinsic thermal sensitivity of mammoth hemoglobin is not different from that of Asian elephant hemoglobin (P = 0.9174). Conversely, as denoted by asterisks, the endothermic dissociation of Cl (P = 0.0198) and BPG (P = 0.0168) each independently lower the oxygenation enthalpy of mammoth hemoglobin to significantly greater degrees than for Asian elephant hemoglobin, as predicted by the E101Q and T12A substitutions on the mammoth β/δ-globin chain, respectively (see text for details). The ΔH of mammoth hemoglobin was independent of pH under all conditions employed here, illustrating that the binding of Bohr protons does not directly contribute to lowering the ΔH value.

Accession codes

Referenced accessions

NCBI Reference Sequence


  1. Benner, S.A., Caraco, M.D., Thomson, M. & Gaucher, E.A. Planetary biology— paleontological, geological, and molecular histories of life. Science 296, 864868 (2002).
  2. Jermann, T.M., Optiz, J.G., Stackhouse, J. & Benner, S.A. Reconstructing the evolutionary history of the artiodactyl ribonuclease superfamily. Nature 374, 5759 (1995).
  3. Beintema, J.J., Schuller, C., Irie, M. & Carsana, A. Molecular evolution of the ribonuclease superfamily. Prog. Biophys. Mol. Biol. 51, 165192 (1988).
  4. Rohland, N. et al. Proboscidean mitogenomics: chronology and mode of elephant evolution using mastodon as outgroup. PLoS Biol. 5, e207 (2007).
  5. Lister, A.M., Sher, A.V., van Essen, H. & Wei, G. The pattern and process of mammoth evolution in Eurasia. Quat. Int. 126–128, 4964 (2005).
  6. Lister, A.M. The impact of Quaternary Ice Ages on mammalian evolution. Phil. Trans. R. Soc. Lond. B 359, 221241 (2004).
  7. Lister, A. & Bahn, P.G. Mammoths: Giants of the Ice Age (University of California Press, Berkeley, California, USA, 2007).
  8. Repin, V.E., Taranov, O.S., Ryabchikova, E.I., Tikhonov, A.N. & Pugachev, V.G. Sebaceous glands of the woolly mammoth, Mammuthus primigenius Blum: histological evidence. Dokl. Biol. Sci. 398, 382384 (2004).
  9. Römpler, H. et al. Nuclear gene indicates coat-color polymorphism in mammoths. Science 313, 62 (2006).
  10. Miller, W. et al. Sequencing the nuclear genome of the extinct woolly mammoth. Nature 456, 387390 (2008).
  11. Braunitzer, G., Jelkmann, W., Stangl, A., Schrank, B. & Krombach, C. Hemoglobins, XLVIII: the primary structure of hemoglobin of the Indian elephant (Elephas maximus, Proboscidea): beta2=Asn. Hoppe-Seyler's Z. Physiol. Chem. 363, 683691 (1982).
  12. Braunitzer, G., Stangl, A., Schrank, B., Krombach, C. & Wiesner, H. Phosphate-haemoglobin interaction. The primary structure of the haemoglobin of the African elephant (Loxodonta africana, Proboscidea): asparagine in position 2 of the beta-chain. Hoppe-Seyler's Z. Physiol. Chem. 365, 743749 (1984).
  13. Opazo, J.C., Sloan, A.M., Campbell, K.L. & Storz, J.F. Origin and ascendency of a chimeric fusion gene: the beta/delta-globin gene of paenungulate mammals. Mol. Biol. Evol. 26, 14691478 (2009).
  14. Perutz, M.F. Species adaptation in a protein molecule. Mol. Biol. Evol. 1, 128 (1983).
  15. Shih, D.T., Jones, R.T., Imai, K. & Tyuma, I. Involvement of Glu G3(101)beta in the function of hemoglobin. Comparative O2 equilibrium studies of human mutant hemoglobins. J. Biol. Chem. 260, 59195924 (1985).
  16. Baudin, V. et al. Functional consequences of mutations at the allosteric interface in hetero- and homo-hemoglobin tetramers. Protein Sci. 2, 13201330 (1993).
  17. Jones, R.T. & Shih, T.B. Hemoglobin variants with altered oxygen affinity. Hemoglobin 4, 243261 (1980).
  18. Nakatsuji, T., Shimizu, K. & Huisman, T.H. Hb F-La Grange or α2γ2101(G3)Glu→Lys; 75Ile; 136Gly: A high oxygen affinity fetal haemoglobin variant observed in a causcasian newborn. Biochim. Biophys. Acta 789, 224228 (2004).
  19. Clementi, M.E., Condò, S.G., Castagnola, M. & Giardina, B. Hemoglobin function under extreme life conditions. Eur. J. Biochem. 223, 309317 (1994).
  20. De Rosa, M.C., Castagnola, M., Bertonati, C., Galtier, A. & Giardina, B. From the Arctic to fetal life: physiological importance and structural basis of an 'additional' chloride-binding site in haemoglobin. Biochem. J. 380, 889896 (2004).
  21. Irving, L. & Krog, J. Temperature of skin in the arctic as a regulator of heat. J. Appl. Physiol. 7, 355364 (1955).
  22. Fronticelli, C. et al. Allosteric modulation by tertiary structure in mammalian hemoglobins. Introduction of the functional characteristics of bovine hemoglobin into human hemoglobin by five amino acid substitutions. J. Biol. Chem. 270, 3058830592 (1995).
  23. Shen, T.J. et al. Production of unmodified human adult hemoglobin in Escherichia coli . Proc. Natl. Acad. Sci. USA 90, 81088112 (1993).
  24. Weber, R.E. Use of ionic and zwitterionic (Tris/BisTris and HEPES) buffers in studies on hemoglobin function. J. Appl. Physiol. 72, 16111615 (1992).
  25. Park, S.-Y., Yokoyama, T., Shibayama, N., Shiro, Y. & Tame, J.R.H. 1.25 Å resolution crystal structures of human haemoglobin in the oxy, deoxy and carbonmonoxy forms. J. Mol. Biol. 360, 690701 (2006).
  26. Abraham, D.J., Peascoe, R.A., Randad, R.S. & Panikker, J. X-ray diffraction study of di and tetra-ligated T-state hemoglobin from high salt crystals. J. Mol. Biol. 227, 480492 (1992).
  27. Turner, G.J. et al. Mutagenic dissection of hemoglobin cooperativity: Effects of amino acid alteration on subunit assembly of oxy and deoxy tetramers. Prot. Str. Funct. Gen. 14, 333350 (1992).
  28. Römpler, H. et al. Multiplex amplification of ancient DNA. Nat. Protoc. 1, 720728 (2006).
  29. Barnes, I. et al. Genetic structure and extinction of the woolly mammoth, Mammuthus primigenius . Curr. Biol. 17, 10721075 (2007).
  30. Oostenbrink, C., Villa, A., Mark, A.E. & van Gunsteren, W.F. A biomolecular force field based on the free enthalpy of hydration and solvation: the GROMOS force-field parameter sets 53A5 and 53A6. J. Comput. Chem. 25, 16561676 (2004).

Download references

Author information

  1. These authors contributed equally to this work.

    • Roy E Weber &
    • Alan Cooper


  1. Department of Biological Sciences, University of Manitoba, Winnipeg, Canada.

    • Kevin L Campbell,
    • Jason E E Roberts,
    • Angela M Sloan,
    • Anthony V Signore &
    • Jesse W Howatt
  2. Australian Centre for Ancient DNA, School of Earth and Environmental Sciences, University of Adelaide, Adelaide, Australia.

    • Laura N Watson,
    • Jeremy J Austin &
    • Alan Cooper
  3. Department of Chemistry, University of Manitoba, Winnipeg, Canada.

    • Jörg Stetefeld
  4. Protein Design Laboratory, Yokohama City University, Yokohama, Japan.

    • Jeremy R H Tame
  5. Junior Research Group Molecular Ecology, Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany.

    • Nadin Rohland &
    • Michael Hofreiter
  6. Department of Biological Sciences, Carnegie Mellon University, Pittsburgh, Pennsylvania, USA.

    • Tong-Jian Shen &
    • Chien Ho
  7. Zoophysiology, Institute of Biological Sciences, University of Aarhus, Aarhus, Denmark.

    • Roy E Weber
  8. Present address: Department of Genetics, Harvard Medical School, Boston, Massachusetts, USA.

    • Nadin Rohland
  9. Present address: Department of Biology, University of York, York, UK.

    • Michael Hofreiter


K.L.C. conceived the research. K.L.C., J.S., M.H., J.J.A., T.-J.S., C.H., R.E.W. and A.C. designed the experiments. K.L.C., J.E.E.R., L.N.W., A.M.S., A.V.S., J.W.H., N.R., T.-J.S., R.E.W. and J.J.A. conducted the experiments. K.L.C., J.S., A.V.S., J.R.H.T., R.E.W. and A.C. analyzed the data. K.L.C. and A.C. drafted the manuscript, and K.L.C., M.H., J.R.H.T., C.H., R.E.W. and A.C. contributed to the final manuscript writing and its revisions.

Competing financial interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to:

Author details

Supplementary information

PDF files

  1. Supplementary Text and Figures (2M)

    Supplementary Note, Supplementary Figures 1–12 and Supplementary Tables 1–4

Additional data