All living organisms carry their own evolutionary history in their cells, and this history can be read in analyses of nucleic acid sequences or protein amino acid sequences. Thus, phylogenetic trees constructed from DNA or proteins have helped to clarify evolutionary relationships among species. Although DNA and proteins are also determinants of morphology, the genetic information that morphology provides is indirect and difficult to interpret, since numerous genes and complex interrelations are involved in configuring the structures of a living organism. In addition, convergence or parallel evolution phenomena can lead to similarities between unrelated species in one or more morphological characteristics. Nevertheless, in classical paleontology, species identification and classification are based exclusively on morphological features of the fossil record. As Wilson and Cann remarked, “The fossil record, on the other hand, is infamously spotty because a handful of surviving bones may not represent the majority of organisms that left no trace of themselves. Fossils cannot, in principle, be interpreted objectively: the physical characteristics by which they are classified necessarily reflect the models the paleontologists wish to test”
[1].
2. Short Survival of Fossil DNA, Longer Survival of Fossil Proteins
Although recent decades have seen spectacular developments in molecular paleontology, this branch of science is not as recent as has been suggested. In the 1950s, Abelson first demonstrated the presence of amino acids and peptides in fossils
[2]. In their initial work the detection of amino acids present only in certain proteins (e.g., hydroxyproline in collagen) made it possible to infer which type of protein these amino acids came from, but no additional genetic information could be obtained regarding the species to which the rest of the amino acids belonged. In 1963, Wykoff published electron microscopy images of collagen fibrils in dinosaur bones more than 200 million years old—another example of the early stages of molecular paleontology
[3]. The molecule that has most often been investigated in the tissues of extinct species, ancient bones, or fossil remains is DNA, as the direct carrier of genetic information. The first ancient DNA sequence was obtained by Wilson’s group, who studied a museum specimen of tissue from a quagga—a species from the horse family that became extinct in the late 19th century. To study the DNA remnants in the sample it was necessary to amplify them with a technique first developed in 1984: molecular cloning
[4].
The main drawback that limits the scope of molecular paleontology is that any biomolecules that may survive in fossil remains must necessarily be altered and present at very low concentrations. When an organism dies, most of its biomolecules, as well as the organism itself, quickly disappear. However, under special circumstances in which rapid dehydration or rapid burial in an anaerobic environment occurs, hard (bone, shell, etc.) and even soft tissues (skin, muscle, etc.) can survive, and may thus contain biomolecules—albeit not in an intact form
[5]. Proteins found in fossils are thus usually denatured and fractionated into peptides. In addition, after death, a process of racemization takes place: amino acids with the
L-form spatial configuration are converted to the isomeric
D-form
[6]. DNA, an even more fragile molecule than proteins, is usually fractionated into sequences of only a few hundred base pairs containing abundant lesions such as baseless sites, oxidized pyrimidines, and chain cross-linkings
[7]. Accordingly, Lindahl noted that it would be unlikely that any useful DNA could ever be extracted from very ancient fossils
[8]. In fact, although the entire Neanderthal genome has been sequenced
[9], most studies that focused on DNA found that it is unlikely to survive for more than 100,000 years. Nonetheless, notable exceptions to date are the sequencing of this biomolecule in a 400,000 year old
Homo heidelbergensis fossil
[10], and the genomic data obtained from a 560–780-thousand-year-old equid specimen
[11] and from two mammoth specimens more than 1 million years old
[12]. In addition, Woodward et al. published nine DNA sequences of the gene encoding cytochrome b, which were extracted from an 80-million-year-old dinosaur bone
[13]. A drawback of these results was that the sequences did not show a significant degree of similarity to equivalent sequences from birds and reptiles, i.e., dinosaurs’ closest extant relatives. However, later analyses of the sequences obtained by Woodward and colleagues revealed a greater similarity to human DNA than to that of other animals. This similarity, therefore, ruled out the possibility of dinosaur DNA and showed that the results were probably due to the inadvertent contamination of the sample during processing
[14].
Although the ideal outcome is to read genetic information directly from the DNA nucleotide sequence, proteins also provide useful information, albeit indirectly, about amino acid sequences. In contrast to DNA, which appears to survive for only a thousand years, some proteins, under certain conditions, can persist in fossils for millions of years. Proteins bind to the mineral phase (hydroxyapatite) of bone, and this binding provides considerable protection from degradation by exogenous agents. Moreover, the amount of hydroxyapatite crystals increases after death, and this may favor protein encapsulation
[15][16]. Compared to DNA, however, proteins present a technical obstacle in that they are not amplifiable, so their concentration cannot be increased—as can be attempted for DNA with polymerase chain reaction techniques. Although initial studies conducted between the 1950s and 1970s identified amino acids and peptides in fossils up to millions of years old, they did not provide information on the species specificity of these biomolecules, that is, on their ascription to or kinship with other species
[2][3].
3. Detection of Fossil Proteins with Immunological Methods: Applications in Paleontological Controversies
Proteins undergo profound changes over time; nonetheless, these molecules, although fragmented or altered, can in some cases retain intact amino acid sequences. The protein fragments may be detectable with antibodies, which identify sequences comprising between 4 and 12 amino acids (epitopes)
[17]. Mass spectrometry (MS) is also able to detect amino acid sequences in peptides
[18]; however, this technology had not yet been implemented for fossil proteins in the twentieth century. In this period, most studies that aimed to identify fossil proteins were carried out with antibodies. Jerold M. Lowenstein was the first to identify genetic information contained in fossil proteins by applying radioimmunoassay (RIA)
[19], an immunological technique able to specifically detect proteins in quantities as low as 10
−13 M. Lowenstein and colleagues found human collagen, the most abundant protein in bone, in fossil samples of 20,000-year-old
Homo sapiens, 50,000-year-old
Homo neanderthalensis, 0.5-million-year-old
Homo erectus, and 1.9-million-year-old
Australopithecus robustus [20][21]. Collagen was also detected with dot-blotting in a 10-million-year-old fossil bone
[22]. Osteocalcin, another abundant protein in bone, was identified by Ulrich et al. with antibodies in 13-million-year-old fossil bovine bones and 30-million-year-old rodent teeth. These researchers observed that osteocalcin in bovine fossil material still retained its functional ability to bind calcium
[23]. Osteocalcin was also detected in a sample of 75-million-year-old dinosaur bone
[24].
Particularly interesting is the detection of proteins in
Ramapithecus fossils. In the 1960s, some paleoanthropologists considered this species, which lived 8 to 20 million years ago, to be a hominid, and it was thus suggested that the human lineage had diverged from that of apes about 20 million years ago. Molecular data, however, contradicted this hypothesis. Sarich and Cronin used immunological techniques to study modern chimpanzee, gorilla and human albumin, and concluded that these three species diverged from a common ancestor only 5 million years ago
[25]. If this hypothesis is correct, it rules out hominin ancestry for
Ramapithecus. Lowenstein produced antibodies by injecting an extract prepared from this fossil into a rabbit, and found that these antibodies reacted more strongly with gorilla, orangutan and gibbon sera than with chimpanzee or human sera. According to these results,
Ramapithecus was genetically as closely related to Asian monkeys as to African monkeys, and more distantly related to humans
[26]. Currently, paleontologists do not include
Ramapithecus in the human lineage and consider it more closely related to orangutans.
At the turn of the twentieth century, a skull of modern human appearance was discovered in Sussex, England, which appeared in association with a jaw displaying ape-like morphological characteristics. Because the morphology of these bones was consistent with then-current theories of human evolution, the so-called Piltdown Man (
Eoanthropus dawsoni) was accepted in 1912 by English anthropological authorities as the missing link between apes and humans, and was considered the first English human. It was not until 1953 when it became evident, based on an analysis of fluoride content, that the purported fossil was a fraud: a 500-year-old human skull to which the artificially aged jaw of a monkey had been added and the teeth modified to make them look human
[27]. It remained to be determined whether the jaw was from a chimpanzee or orangutan. Lowenstein et al. studied a sample from the jaw and observed that antibodies to orangutan collagen reacted more strongly with an extract from this sample than did antibodies to human or chimpanzee collagen
[28].