Science in Christian Perspective

DNA Sequences in Miocene and Oligo-Miocene Fossils: Their Significance to Evolutionary Theory

 Gordon C. Mills*

 Department of Human Biological Chemistry & Genetics
 University of Texas Medical Branch 
Galveston, TX 77555

Three reports of sequencing DNA isolated from ancient fossils have recently been published. Two of these reports describe sequences of DNA from plant chloroplasts in 17-20 million-year-old fossils found in Miocene sediments in Clarkia, Idaho. The third report provides a sequence for a DNA segment from a 25-30 million-year-old termite imbedded in amber. The DNAs of the fossil plants code for a subunit of an important chloroplast enzyme and the DNA of the fossil termite codes for a portion of a ribosomal RNA. In this paper, the author critiques the chemical and biological value of these studies and their significance to Darwinian evolutionary theory, the theory of punctuated equilibria, and his theory of theistic evolution.

 In the early 1990s, some studies on DNA sequences in 17-20 million-year-old fossils received considerable media attention. Fairly long segments of DNA were isolated from chloroplasts of leaves of a magnolia and a bald cypress found in Miocene sediments at Clarkia, Idaho.^1,2 Since then, DNA has been isolated from a fossil termite imbedded in Oligo-Miocene amber from the Dominican Republic.³ The latter is believed to be 25-30 million years old. In this case, DNA isolated from the termite was somewhat shorter. In both cases, the DNA was informational, coding for a chloroplast enzyme in the two fossil tree leaves (magnolia and bald cypress), and coding for ribosomal RNA in the fossil termite.

Since these findings extend the age of previously studied DNA a thousandfold or more, we should carefully evaluate their significance. Preservation of undamaged DNA in ancient fossils is an extremely unusual occurrence.⁴ However, the environmental conditions necessary for preservation appear to have occurred in Clarkia, Idaho sedimentary deposits and in amber trapping of insects. In both cases, differences in DNA sequence of fossil DNA compared to DNA in closely related living species were extremely small. Does this sequencing of fossil DNA provide, as Gould notes "Y a striking illustration of the best kind of evidence that we can produce for the factuality of evolution itself"?⁵ Or are there other possible interpretations for the significance of these fossil DNA sequencing studies?

When Gould uses the word "evolution," he is referring to it in a fully naturalistic sense, i.e., one where only chance events are considered. I recently proposed a theory of theistic evolution, which suggests a possible mode of action for divine agency that accounts for macroevolutionary change in organisms. This theory points to the need for a provision of new genetic information in major processes of change in organisms. I argued strongly that this theory will in no way interfere with scientific investigation and indeed will provide new insight into possible areas of future research.⁶ In proposing this theory of theistic evolution, I do not wish to limit God's sovereignty or governance over all of his creation. God's sovereignty still extends over all of the physical laws of the created world, including those events that I have spoken previously of as being a consequence of chance.

In this paper, I will consider the relationship of fossil DNA research to this theory of theistic evolution as well as to naturalistic theories. I will examine the fossil DNA data carefully, and will make an evaluation of the possible significance of these studies. I will also consider several unanswered questions that should give us cause to avoid unmerited speculation.

Age of Fossils

As a biochemist and molecular biologist, I do not consider myself qualified to assess the accuracy of dating the Miocene Clarkia fossil beds where the magnolia and bald cypress leaves were found or the Oligo-Miocene amber with its imbedded fossil termite. The age of deposits at the Clarkia fossil beds is reported by Golenberg, et al. to be 17-20 million years,⁷ and that of Dominican amber by DeSalle, et al. to be 25-30 million years.⁸ The papers cited herein give some consideration to evidence for the validity of those dates for the fossils, but I will leave it to others to seriously critique the dates they provide. However, I have used the above ranges of dates in some of my calculations, and results of those calculations would be affected if there are serious errors in dating the fossils.

 The Significance of Chloroplasts

A few characteristics of chloroplasts merit careful consideration. In earlier examinations of leaves in the Clarkia fossil beds, scientists found that chloroplasts were the best preserved structural component in the leaf cells with over 90% of cells containing intact chloroplasts. From the standpoint of possible DNA isolation, there are thousands of copies of chloroplast genes in each cell, so this would make these genes a much more favorable target for study than nuclear genes. The chloroplast is the organelle within plant cells that is primarily responsible for photosynthesis. In this process, solar energy is trapped and utilized in fixation of carbon dioxide from air into various organic carbon compounds. Like mitochondria, chloroplasts have their own DNA, although the DNA of the plant cell nucleus codes the majority of chloroplast proteins. Chloroplasts of all plants appear to be quite similar.

Ohyama, et al. have sequenced the entire chloroplast DNA of the liverwort Marchanti polymorpha.⁹ Liverwort chloroplast DNA contains 121,024 nucleotide base pairs. They detected 128 possible genes in the DNA, including genes for four kinds of ribosomal RNA, 32 species of transfer RNA genes, and 55 possible genes for proteins. Clearly, a chloroplast is still extremely complex. Zurawski, et al. note that most of the genetic components of a chloroplast are extremely resistant to change. They estimate that the average rate of change of chloroplast DNA is about one-hundredth that of mitochondrial DNA.¹⁰ Among the genes in chloroplast DNA is the rbcL gene, which codes for the larger subunit of the enzyme ribulose-bis-phosphate carboxylase. This enzyme catalyzes fixation of carbon dioxide and is therefore a very important component in the photosynthetic pathway of plants. The rbcL gene has been sequenced from a variety of plants. It is approximately 1400 nucleotide base pairs in length and is the gene that was isolated and sequenced from fossil leaves of magnolia and bald cypress.

 Application of New Techniques - The Polymerase Chain Reaction

 The data on the rbcL gene of these two fossil plants are taken from studies of Golenberg et al. for the magnolia,¹¹ and of Soltis, et al. for the bald cypress.¹² These authors deserve great credit for their careful and painstaking work, first in isolating the chloroplast DNA and in the subsequent sequencing work. However, use of the polymerase chain reaction (PCR), which was developed by others, was essential to this type of study. PCR amplifies an extremely tiny amount of fossil DNA into an amount suitable for sequence studies.¹³ The starting position of DNA for amplification is selected by using a synthetic nucleotide primer (usually about 30 nucleotides in length) that has a known sequence identical to a segment from a known rbcL gene. This provides a means of selecting only the desired gene for amplification from all other genes in the fossil chloroplast DNA. Using the PCR procedure, the amount of DNA is doubled (presumably with perfectly sequential copies) at each cycle. When twenty or thirty cycles are carried out automatically over a period of several hours, it produces an amount of DNA a millionfold or more greater than the starting material. This amplified DNA is then used for the actual DNA sequencing. In these studies both strands of the initial fossil DNA were amplified by the PCR procedure. This provides an important check on the validity of these studies because the complementary strand would always have a different base in any position than that in the primary strand. The complementary strand also proceeds in the opposite direction and requires a different primer to initiate amplification.

 Fossil Gene Sequencing. - Are the Results Valid?

One extremely important question regarding the significance of fossil DNA sequence studies is the question of the validity of the studies. Could DNA possibly survive for millions of years without undergoing extensive damage? Initially, this damage might be a consequence of dying tissues, or possibly damage by bacteria or other organisms. Subsequently, damage would be predominantly of a chemical nature and would be dependent upon many environmental conditions such as pH, temperature, presence or lack of water and oxygen, etc. Golenberg, et al. note that the fossil leaves were generally intact and that natural dehydration before abcission may have contributed to preservation. They suggest that the leaves fell into water and sank to colder layers prior to covering with sediment.¹⁴ The most likely types of chemical damage would include cleavages in the DNA chain, removal of individual purines (purine-deoxyribose linkages are particularly sensitive to acid conditions), or modification of adenine, cytosine, or guanine by deamination.

Because I have worked extensively with the various components of DNA, I would be among the first to question whether sequencing of 17-20 million-year-old fossil DNA could possibly give meaningful results. In fact, several papers (e.g., Paabo and Wilson¹⁵ and Lindahl¹⁶) have challenged the validity of the sequencing studies. In both cases, the challenges were based on the known susceptibility of DNA to chemical modification. The authors postulated certain environmental conditions during the millions of years. After extrapolation of the data, they insisted that no DNA could possibly have survived in the fossil leaves for 17-20 million years. However, as noted by Golenberg, empirical confirmation must take precedence over theoretical objections since the actual environmental conditions during the 17-20 million-year period are not known.¹⁷ Neither Paabo and Wilson nor Lindahl critically evaluated the fossil DNA sequence data.

Use of a 30-base DNA primer in the isolation technique assures the selection of the desired chloroplast gene and eliminates the possibility of contamination by bacterial genes.

 With sequencing of the second fossil DNA from a bald cypress,¹⁸ it appears that most of the critics have been silenced. In this case, the fossil bald cypress gene differed by only 11 bases out of 1320 from the corresponding gene of a closely related modern species. None of these 11 bases would produce a change in amino acids of the corresponding rbcL protein. It is very difficult to postulate that results such as these could be due to an artifact. Artifactual changes would be equally likely to affect each of the three codon positions in DNA. In the bald cypress, all differences occurred in the third position of codons. It is probably true that much of the fossil chloroplast DNA has been damaged or modified over the 17-20 million year period since the leaves were formed, but some undamaged DNA molecules have apparently survived in fossil bald cypress and magnolia leaves. Use of modern isolation and amplification techniques has permitted selection of undamaged molecules for sequencing studies. Use of a 30-base DNA primer in the isolation technique assures the selection of the desired chloroplast gene and eliminates the possibility of contamination by bacterial genes. To be sure, there is still a question whether these techniques may be applied at other paleontological sites, since preservation conditions were unique at the Clarkia, Idaho site. Soltis, et al.¹⁹ and Golenberg²⁰ indicate, however, that rbcL genes from several other plants have been isolated for study from Clarkia fossil beds.

One type of artifactual change would be deamination of individual bases of fossil DNA since one pyrimidine (cytosine) and both purines (adenine and guanine) have amino groups. However, deamination events at any particular site on a DNA strand would only occur at that site on one of the two strands. For example, if cytosine at a given position was deaminated to form uracil in one strand, the base paired guanine at that position on the other strand would likely be intact. Sequencing of the other strand would show guanine in the base-paired position of that strand. Since the strands go in opposite directions, the base-paired nucleotides in each of the two DNA strands would have different numbers in the linear sequence. One may estimate that up to 25% deamination at a given position on a DNA chain would probably be required before a deamination artifact would be evident. Even then there would be an indication on the sequencing gels of a possible artifact.

A careful study of the data has convinced me that sequenced fossil chloroplast DNA does actually represent the DNA sequence of the fossil leaves when they were first formed.

 It is more difficult to predict the consequences of a depurination artifact (loss of adenine or guanine) on the fossil DNA chain, but it seems likely that the effect might be similar to that suggested for deamination. Depurination at any site on one of the chains would not necessarily cause a break in that DNA chain, since the backbone on DNA is made up of alternating phosphate and deoxyribose groups, with the purines and pyrimidines extending inward in the double stranded DNA helix. Depurination appears to make phosphate-deoxyribose linkages more susceptible to breakage, however. The procedures used in isolating DNA for study remove most DNA that is extensively broken. Since purines are always base paired with pyrimidines, the complementary strand of DNA would always have a correct base at any given position. It seems likely that depurination would have to be fairly extensive before errors, such as finding two different bases at a single position on the amplified DNA, would be evident. Golenberg has provided a careful evaluation of criticisms of fossil DNA sequencing studies and has outlined various controls utilized to assure that results are valid.²¹ A careful study of the data has convinced me that sequenced fossil chloroplast DNA does actually represent the DNA sequence of the fossil leaves when they were first formed.

 Results of Magnolia and Bald Cypress DNA Sequence Studies

Golenberg, et al. sequenced the 759 base segment of the rbcL gene from chloroplasts of a fossil magnolia (Magnolia latahensis).²² This is slightly over 50% of the coding sequence of the larger subunit of the gene for ribulose-bis-phosphate carboxylase (rbcL gene). The above authors compared this DNA sequence with that of the corresponding gene from the most closely related modern species (Magnolia macrophylla) and several other related species (Persea americana and Liriodendron tulipifera). A comparison of the fossil magnolia gene (M. latahensis) with the corresponding gene from M. macrophylla indicated 17 base differences in the 759 base DNA segment (2.2% difference). Of the 17 differences, 11 were transitions (purine-purine or pyrimidine-pyrimidine) and six were transversions (purine-pyrimidine or pyrimidine-purine). There were no insertions or deletions. Thirteen of the base differences were in the third position of codons and would cause no change in amino acid sequence of the expressed protein. They, therefore, would be called synonymous differences. The other four differences were in the first or second position of codons and would cause amino acid changes in the expressed protein (i.e., they were nonsynonymous differences). Of the four amino acid differences, only one at codon 91, alanine-proline, might be classed as a radical difference. The other differences at codons 95, 97, and 255 respectively were: asparagine-serine, tyrosine-phenylalanine, and valine-isoleucine with the amino acid for the fossil protein listed first. Three of the nonsynonymous differences were grouped closely (codons 91, 95, and 97). They appear to be in a region of the gene that is seen to be quite variable when the same gene is examined in other plant species.^23,24 Therefore, there is a possibility that these nonsynonymous differences might represent a single gene conversion event rather than three separate mutational events.

 Soltis, et al. sequenced the 1320 base segment of the rbcL gene from chloroplasts of a fossil bald cypress and from the same species of a living bald cypress (Taxodium distichum).²⁵ This sequence constitutes about 92% of the entire coding sequence for this gene. For comparison, they also sequenced the same gene from another related modern species (Metasequoia glyptostroboides). In this case, the DNA sequence of the fossil rbcL gene from the bald cypress differed from that of the modern gene of the same species in only 11 positions (0.83%). There were no insertions or deletions, and the 11 differences were all transitions in third positions of codons (synonymous differences). Therefore, the proteins expressed by rbcL genes of living and fossil bald cypress would have identical amino acid sequences. For comparison, the fossil bald cypress gene and that of the modern bald cypress both differ from the corresponding M. glyptostroboides gene in 38 positions (2.9%). Seven of these base differences would cause amino acid differences in the expressed protein.

The proteins expressed by rbcL genes of living and fossil bald cypress would have identical amino acid sequences.

If we assume C for the purpose of making calculations C that the fossil bald cypress, T. distichum, has an ancestral relationship to the corresponding living species, we may calculate a value for the rate of change in nucleotide sequence of the rbcL gene. Likewise, the fossil magnolia species, M. latahensis, and the closely related sister species, M. macrophylla, may be considered to share a common ancestor, and be separated by a minimum of 17-20 million years. Consequently, we may calculate a maximum value for the rate of change in nucleotide sequence for the rbcL gene of this species. For the bald cypress, we obtain an average change rate for the rbcL gene of 0.4-0.5 synonymous changes per nucleotide site per billion years. For the magnolia, the corresponding maximal value for synonymous changes would be 0.8-1.0 per nucleotide site per billion years. For nonsynonymous changes, the change rate for the bald cypress rbcL gene would be zero, and for the magnolia, a maximal value of 0.3 changes per nucleotide site per billion years.

The data of Zurawski, et al. provide some figures for comparison.²⁶ They reported 47 synonymous differences and 25 nonsynonymous differences in rbcL genes (1279 bases) of barley and maize. They estimate a divergence date for these two plant species of 50-65 million years. If we assume that half of the synonymous base changes were in the line from the common plant ancestor to maize and the other half were in the line from the common plant ancestor to barley, we obtain an average change rate for the rbcL gene of 0.3-0.4 synonymous nucleotide differences per site per billion years. If we make similar calculations for nonsynonymous changes, the value obtained is 0.15-0.2. None of the above calculated values are corrected for multiple substitutions. It is interesting that the rate of DNA sequence changes based on comparisons of the fossil DNAs with living plant chloroplast DNAs are of the same order of magnitude as the estimates of average change rates calculated from comparisons of DNA sequences in the rbcL genes of maize and barley.

Neither Golenberg, et al.²⁷ nor Soltis, et al.²⁸ studied the 5'-noncoding region of the rbcL gene. Most of the short nucleotide sequences that control the expression of genes lie in adjacent noncoding regions. In studies of 5'-noncoding regions of rbcL genes in chloroplasts of barley and maize, Zurawski, et al. noted that insertion/deletion differences were often seen in these gene regions when making comparisons. They noted that other types of changes (single base differences, for example) were less evident in 5'-noncoding regions of these genes than in the third position of codons in coding regions.²⁹ If we were to study these same 5'-noncoding regions of rbcL genes of fossil cypress and magnolia leaves and their living counterparts, we might learn whether the control regions of these genes were as stable for 17-20 million years as their coding regions.

 Studies with Fossil Termite DNA

 DeSalle, et al. isolated some DNA from genes that code for ribosomal RNA from both cell nuclei and mitochondria of a 25-30 million-year-old fossil termite (Mastotermes electrodominicus) imbedded in amber.³⁰ This would be informational DNA since these nuclear and mitochondrial DNAs code for nucleotide sequences in RNA used in ribosomal particles of cells. However, the actual function of this ribosomal RNA during translation is not known. After amplification of the isolated fossil termite DNA using the PCR procedure, the amplified DNA was sequenced. The authors provide nucleotide sequences for two fragments of the nuclear gene, designated 18SA (116 bases long) and 18SC (121 bases long), and one fragment of a mitochondrial gene designated 16S (94 bases long). The length of these segments contrasts sharply with the 759 and 1320 bases found for fossil chloroplast genes in leaves of magnolia and bald cypress. DeSalle and coworkers compared the nuclear DNA sequences with that of a closely related living termite (Mastotermes darwiniensis) and also with comparable sequences of several other species of termites and related insects (cockroach, mantid, grasshopper, stone fly and fruit fly). The primary thrust of their paper was in providing molecular evidence for phylogenetic relationships of these organisms.

 Although the means of preservation of the termite in amber is entirely different from that of the fossil leaves, DeSalle and coworkers note that trapping and encapsulation of an insect in the resinous sap of Hymenaea would result in fairly rapid dehydration as isoprene components of the sap polymerize. They also note that bactericidal action of terpenes gives amber its natural embalming characteristics. The authors selected genes coding for ribosomal RNA from both the cell nucleus and mitochondria for study because of the high copy numbers of these genes in each cell. Other workers had previously studied these genes and their work had indicated a high degree of likelihood that useful information could be obtained by sequencing these particular types of DNA.

Clearly, the DNA of the particular nuclear ribosomal gene segment studied Y is very highly conserved.

 In their paper, DeSalle and coworkers do not consider the types of artifactual chemical modification of DNA that might have occurred in amber fossils. However, they do very carefully consider the possibility of contamination by bacteria during processing of fossil DNA and in the subsequent amplification of fossil DNA by the PCR procedure.

They also indicate that M. darwiniensis appears to be one of the most primitive of 2000 or more described species of termites, and is the only living species of the family Mastotermitidae. Other species of the genus Mastotermes are extinct and the best preserved fossils are found in amber (e.g., M. electromexicus). When the two nuclear gene fragments (18SA and 18SC) of the fossil termite (M. electrodominicus) are compared with corresponding fragments of the closely related living termite (M. darwiniensis), the DNA sequence differs in only three positions out of a total of 237 (1.3% difference). Using the same type of comparisons, nuclear DNA of the fossil termite differed from two other termites of other genera (Nasutotermes costalis and Zootermopsis nevadensis) by 3.4% and 4.2%; from a cockroach (Blaberus sp.) by 3.8%; from a mantid (Mantis religiosa) by 3.8%; from a stone fly (Pteronarcys sp.) by 5.5%; from a grasshopper (Warramaba picta) by 6.3%; and from a fruit fly (Drosophila melanogaster) by 14.3%.

Since the fossil termite, M. electrodominicus, and the living termite, M. darwiniensis, are closely related sister groups, they may be considered to share a common ancestor. Consequently, the three nucleotide differences in nuclear ribosomal DNA segments may represent the change in 237 bases for this type of DNA for a minimum period of 25-30 million years. Since the genus Mastotermes occurs as early as 40 million years ago, the time separating these two species could be even longer. Consequently, the maximum value for the average change per nucleotide site would be 0.4-0.5 per billion years. The value for the termite is not too different from the average nucleotide changes for the rbcL gene of the fossil leaves: a maximal value of 1.1-1.3 per site per billion years for the magnolia, and a value 0.4-0.5 per site per billion years for the bald cypress. Clearly the DNA of the particular nuclear ribosomal gene segment studied by DeSalle and coworkers is very highly conserved.

 The relatively small sequence differences in nuclear ribosomal DNA of other closely related insects (based on morphological comparisons) also show that this segment of nuclear ribosomal DNA is highly conserved. In contrast, the mitochondrial ribosomal DNA segment studied in four of the organisms showed much greater differences in DNA sequence. The larger number of differences in mitochondrial ribosomal DNA between the fossil and living termite makes it difficult to provide an assessment of their evolutionary significance. This is due partially to the short length of the DNA sequence studied and partially because the DNA sequence could not be related to any specific ribosomal function. Why should ribosomal DNA that is repeatedly replicated in mitochondria change much more rapidly than that which is replicated in the nucleus? Is it related in some manner to functions of RNA molecules transcribed from mitochondrial DNA genes? Or is it related in some manner to repair mechanisms that prevent modification of nucleotide sequences in genes? The answers to these questions are not evident at the present time.

Significance of Fossil DNA Studies to Evolutionary Theory

DNA sequence studies with these two tree leaf fossils provide very good evidence of the ancestral descent of living bald cypress (T. distichum) from the fossil bald cypress (T. distichum) that lived 17-20 million years ago. The studies also provide reasonable evidence that the modern magnolia (M. macrophylla) and the fossil magnolia (M. latahensis) are closely related sister groups that may be considered to share a common ancestor that lived at least 17-20 million years ago. Likewise, the living termite (M. darwiniensis) and the fossil termite (M. electrodominicus) are closely related sister groups that share a common ancestor that lived at least 25-30 million years ago.

The ancestral descent that we are dealing with in these cases is clearly of a very limited nature. In the bald cypress, it is ancestral descent within a particular species. In the other two instances (magnolia and termite), it is ancestral descent within a particular genus. After reviewing the data on the bald cypress, Gould notes "Y in this case we may be looking at an unbroken and unbranched evolutionary sequence - a true continuity over 20 million years - and the smaller percentage of changes, with no alterations at all in amino acids may record the actual architecture of evolutionary stability.³¹ I would agree perfectly with Gould's conclusion to this point.³² One might even use this case to illustrate the stasis of Eldredge and Gould's punctuated equilibria theory of evolution.³³ Gould then goes on to say, however, that the data represents the " best kind of evidence that we can produce for the factuality of evolution itself.^"34 Can Gould really illustrate the dramatic change aspect of punctuated equilibrium by citing cases of stasis? I believe not.

The fossil DNA studies provide the first direct evidence of average rates of change for some types of DNA within very closely related organisms.

If these studies cannot be interpreted as providing real support for either the Darwinian concept of evolution or the change aspect of the punctuated equilibria concept of Eldredge and Gould, what is their real significance? It is simply that the fossil DNA studies provide the first direct evidence of average rates of change for some types of DNA within very closely related organisms. The studies are important because they provide a new and independent approach. Previous values of average rates of change were dependent upon estimations of organismal relationships and on divergence dates of related species, both of which were often uncertain. It is important to realize that all of the nucleotide differences noted in this study between DNA of fossil organisms and DNA of living species may be explained by point mutations. There is no indication that any of these differences could account for any change in function, particularly when considering the changes in the gene coding for the chloroplast enzyme. At this point in our knowledge of evolutionary biology, the possibility of even minor changes at the species level is far more likely to be a consequence of more extensive changes in an organismal genome (gene crossovers, gene conversions, gene duplications, etc.), than they are to be due to point mutations.

 In a recent review of the molecular clock hypothesis,³⁵ I provided evidence that there is no constant rate of change in the incorporation of mutations into specific proteins of organisms. In that paper, I discussed the types of changes in organismal genomes that may occur because of chance events. Of particular interest are those events that may involve a transfer of genetic information, both within an organism (intraspecies transfer) and from one species to another (interspecies transfer). Within a particular cell, the intraspecies transfer might be a consequence of gene conversions, gene crossovers, gene duplications, etc. In interspecies transfer, movement of plasmid DNA between bacterial cells appears to be fairly common. This transfer process is termed conjugation. In contrast, transfer of genetic information between different species of higher organisms is rare, but there is increasing evidence that it does occur. Usually a viral vector is involved as a carrier of the transferred gene. Also, a number of organisms at various taxonomic levels use symbiont organisms (usually bacteria) in their metabolic processes. These organisms most commonly function extracellularly but sometimes they do act intracellularly. There seems to be no proof that these symbiont organisms transfer their DNA into the host cell genome, however. The transfer of genetic information, either intra- or interspecies, does not produce new genetic information. The types of genetic changes described above are all very likely a consequence of chance events. Indeed they may have a major role in speciation events (microevolution), but it seems very unlikely that they could account for any macroevolutionary events.

   Recently I proposed the following theory of theistic evolution to account for macroevolutionary changes: "that in the history of the origin and development of living organisms, at various levels of organization, there has been a continuing provision of new genetic information by an intelligent cause.³⁶ What is meant by new genetic information, how the theory would be related to theories of common ancestry, punctuated equilibria, etc., are considered in some detail in that paper.

It would appear that developmental genes would have to be involved in macroevolutionary changes that might account for large morphologic change in an organism. These developmental genes control the migration and positioning of cells in the formation of various morphologic structures, and the length of time that they act during embryogenesis is under tight control. Some of these genes have a broad specificity and act in a wide variety of organisms while others appear to have very limited specificity. Although some developmental genes might be retained throughout evolutionary history, it appears likely that the provision of new genetic information in these developmental genes would be essential to account for major macroevolutionary change.

Possibilities for Future Research Involving Fossil Genes

Although developmental genes would appear to have the highest potential for involvement in evolutionary change, it seems very unlikely that these particular genes could ever be isolated from fossil leaves or fossilized insects. Only if these developmental genes were present in organelles such as chloroplasts or ribosomes, where many copies of the gene are present per cell, would their isolation from fossils seem possible.

One type of fossil gene study of evolutionary significance that would appear to be possible is to search for genes or gene segments added or lost to the chloroplast genome

 One type of fossil gene study of evolutionary significance that would appear to be possible is to search for genes or gene segments added or lost to the chloroplast genome in the millions of years separating the fossil DNA from its present day counterpart. Noting an absent gene or gene segment in chloroplast DNA of living organisms would not be too difficult to establish, particularly if DNA regions adjacent to the gene or gene segment were sequenced. It would be more difficult to establish whether the absent gene may have been transferred to another position on the chloroplast DNA. Similarly, an added gene or gene segment could be established in the same manner, i.e., by sequencing DNA regions at both the 5'- and 3'-ends of a particular coding region. These searches for gain or loss of DNA would provide very important information regarding the extent of transfer of DNA into and out of particular DNA regions of the chloroplast genome during the millions of years separating the fossil leaves from living organisms. This type of study would have more evolutionary significance than a study of possible point mutations in coding regions of genes. Either of the proposed studies involving addition or loss of genes would require the isolation and sequencing of fairly large segments of chloroplast DNA, both the fossil DNA and corresponding DNA of similar living plants. Although the isolation and sequencing of very large fossil chloroplast DNA segments is unlikely, sequencing of overlapping smaller segments could provide the same information regarding the genes.

It would be much more difficult to carry out the types of studies that I have suggested for fossil chloroplast DNA on fossil ribosomal DNA of insects. The reasons for the increased difficulty are: (1) there is a distinct advantage to studying DNA that codes for a functional protein because effects of changes in triplet codes can be more readily assessed; (2) at the present time, fossil nuclear ribosomal DNA segments that have been isolated are much shorter than the isolated fossil chloroplast DNA segments; and (3) possible bacterial contamination of the insect fossils imbedded in amber is a much greater problem since both bacteria and insects have ribosomal DNA.

Theological Implications

Brooke, in his book Science and Religion: Some Historical Perspectives, has reviewed attempts by scientists such as T. H. Huxley and Ernst Haeckel to interpret Darwinian evolution in such a manner that there was no longer any need for God. Brooke also reviews carefully views of those who held strongly to a Christian belief and who tried to modify evolutionary theory to make it compatible with their beliefs. He notes that others, both scientists and theologians, rejected Darwinian evolution entirely, particularly after hearing the extreme interpretations of Huxley and Haeckel.³⁷ A. H. Dupree and F. Gregory, in their chapters in God and Nature: Historical Essays on the Encounter between Christianity and Science, also provide insight into views of scientists and theologians of the nineteenth century. They deal particularly with issues such as design, purpose, order, natural law, etc.^38,39

In this paper, I note that there is no problem in interpreting the fossil DNA studies as being consistent with my proposed theory of theistic evolution. The fossil DNA data provide little support, however, for the gradualism inherent in views of traditional Darwinian evolution, but are consistent with the stasis aspect of Eldredge and Gould's theory of punctuated equilibria. Continuation of studies of fossil DNA of the types described earlier in this paper, in conjunction with studies of corresponding DNA from related living organisms, could provide much valuable information. This information might be consistent with the theory that I have proposed or it might suggest the need for modifications. There are many other possible studies that might indicate whether this theory of theistic evolution will continue to be consistent with experimental evidence, or whether the theory might need to be modified or even rejected.

The fossil DNA studies call attention again to questions regarding resistance of DNA in an organism to change with time. Why do genes for some types of proteins (e.g., fibrinopeptides) appear to change so rapidly, and others, such as chloroplast genes, somatic genes for cytochrome c,⁴⁰ and especially genes for histones change so slowly? To what extent do functional constraints of specific proteins have a role in preventing change in the DNA coding for that particular protein? Or do DNA repair mechanisms built into the design of various cells serve as the primary constraint against change in DNA? Does not the information encoded in all genes of living organisms suggest the need for an intelligent designer? Previously I presented arguments that God is involved in some manner as an intelligent cause behind all of life.^41,42 For many of the questions posed above, we have no clear answers at present, but the questions surely merit careful consideration.

©1996

Notes

¹D. E. Giannasi, M. T. Clegg, C. J. Smiley, M. Durbin, D. Henderson, & G. Zurawski, "Chloroplast DNA Sequence from a Miocene Magnolia Species," Nature 344 (1990): 656-658.

²P. S. Soltis, D. E. Soltis, & C. J. Smiley, "An rbcL Sequence from a Miocene Taxodium (Bald Cypress)," Proc. Natl. Acad. Sci. USA 89 (1992): 449-451.

³R. DeSalle, J. Gatesy, W. Wheeler, & D. Grimaldi, "DNA Sequences from a Fossil Termite in Oligo-Miocene Amber and their Phylogenetic Implications," Science 257 (1992): 1933-1936.

⁴S. J. Gould, "Magnolias from Moscow," Natural History 101, no. 9 (1992): 10-18.

⁵Ibid., 16.

⁶G. C. Mills, "A Theory of Theistic Evolution as an Alternative to the Naturalistic Theory," Perspec. Sci. Christian Faith 47, no. 2 (1995): 112-122.

⁷E. M. Goldenberg, et al., see note 1.

⁸R. DeSalle, et al., see note 3.

⁹K. Ohyama, H. Fukuzawa, T. Kohchi, H. Shirai, T. Sano, S. Sano, K. Umesono, Y. Shiki, M. Takeuchi, Z. Chang, S. Aota, H. Inokuchi, & H. Ozeki, "Chloroplast Gene Organization Deduced from Complete Sequence of Liverwort Marchantia polymorpha Chloroplast DNA," Nature 322 (1986): 572-574.

¹⁰G. Zurawski, M. T. Clegg, & A. H. D. Brown, "The Nature of Nucleotide Sequence Divergence between Barley and Maize Chloroplast DNA," Genetics 106 (1984): 735-749.

¹¹E. M. Golenberg, et al., see note 1.

¹²P. M. Soltis, et al., see note 2.

¹³N. Arnhelm, & C. H. Levenson, "Polymerase Chain Reaction," Chem. Eng. News 68, no. 40 (1990): 36-47.

¹⁴E. M. Golenberg, et al., see note 1.

¹⁵S. Paabo, & A. C. Wilson, "Miocene DNA Sequences C a Dream Come True?" Curr. Biol. 1 (1991): 45-46.

¹⁶T. Lindahl, "Instability and Decay of the Primary Structure of DNA," Nature 362 (1993): 709-715.

¹⁷E. M. Golenberg, "Amplification and Analysis of Miocene Plant Fossil DNA," Phil. Trans. R. Soc. Lond. B 333 (1991): 419-427.

¹⁸P. S. Soltis, et al., see note 2.

¹⁹P. S. Soltis, et al., see note 2.

²⁰E. M. Golenberg, see note 17.

²¹E. M. Golenberg, see note 17.

²²E. M. Golenberg, et al., see note 1.

²³P. S. Soltis, et al., see note 2.

²⁴G. Zurawski, et al., see note 10.

²⁵P. S. Soltis, et al., see note 2.

²⁶G. Zurawski, et al., see note 10.

²⁷E. M. Golenberg, et al., see note 1.

²⁸P. S. Soltis, et al., see note 2.

²⁹G. Zurawski, et al., see note 10.

³⁰R. DeSalle, et al., see note 3.

³¹S. J. Gould, see note 4, p. 16.

³²The author agrees with most of Gould's evaluation of the DNA sequence studies in the fossil leaves. It is only when Gould discusses the significance of the studies to the theory of evolution that this author finds himself disagreeing. >

³³N. Eldredge & S. J. Gould, "Punctuated Equilibria: An Alternative to Phyletic Gradualism," in T. J. M. Schopf, ed., Models in Paleobiology (San Francisco: Freeman, Cooper & Co., 1973), 82-115.

³⁴S. J. Gould, see note 4, p. 16.

³⁵G. C. Mills, "The Molecular Evolutionary Clock: A Critique," Perspec. Sci. Christian Faith 46, no. 3 (1994): 159-168.

³⁶G. C. Mills, see note 6.

³⁷J. H. Brooke, Science and Religion: Some Historical Perspectives (Cambridge Univ. Press, 1991), 275-320.

³⁸A. H. Dupree, in D. C. Lindberg, & R. L. Numbers, eds., God and Nature: Historical Essays on the Encounter between Christianity and Science (Univ. Calif. Press, 1986), 351-368.

³⁹F. Gregory, in God and Nature (1986): 369-390.

⁴⁰G. C. Mills, "Cytochrome c: Gene Structure, Homology and Ancestral Relationships," J. Theor. Biol. 152 (1991): 177-190.

⁴¹G. C. Mills, "Presuppositions of Science as Related to Origins," Perspec. Sci. Christian Faith 42 (1990): 155-161.

⁴²G. C. Mills, "Structure of Cytochrome c and c-like Genes: Significance for the Modification and Origin of Genes," Perspec. Sci. Christian Faith 44 (1992): 236-245.