IntroductionThe advent of high-throughput automated sequencing has given the discipline of genetics the complete genomic sequence of several model organisms, including humans (Lander and others 2001; Venter and others 2001), rice (Goff and others 2002; Yu and others 2002), weeds (Lin and others 1999; Mayer and others 1999; Adam 2000; Salanoubat and others 2000; Tabata and others 2000; Theologis and others 2000), worms (Bargmann 1998; Blaxter 1998; Clarke and Berg 1998; Ruvkun and Hobert 1998), fruit flies (Adams and others 2000; Myers and others 2000), and yeasts (Goffeau and others 1996; Wood and others 2002). This explosion of sequence data has given birth to the new subdiscipline called genomics, which examines organisms from a whole-genome perspective. If our genes are trees and our genomes a forest, then genomics allows geneticists to examine the whole forest at one time instead of spending our time focusing on one or a few trees (Gibson and Muse 2002).
By far the group that hosts the widest range and largest number of organisms whose genomes have been completely sequenced are the prokaryotes, that group we colloquially know as the bacteria. Genomes from both free-living and pathogenic bacteria have been completely sequenced, as have genomes from many eubacteria and several members of the Archaea (Fraser 2002). Because of their small, compact genomes and the relative ease of growing large numbers of them, bacterial organisms make prime candidates for genomic sequencing.
Genomics has revolutionized the way questions are addressed and has provided valuable insight into how genomes evolve (Arber 2002; Doolittle 2002; Knight 2002). Nevertheless, creationists, such as Bryan College's Todd C Wood, are using genomic data to support their contention that living things were independently created only a few thousand years ago. These "recent creationists" claim that completed genomic sequence data from bacteria called mycoplasmas refute evolutionary theory (Wood 2001); this article is an evaluation of Wood's efforts. As we shall see, in order to fit the genomic data into his recent creationist paradigm, Wood has to ignore previous work on mycoplasmal phylogeny and misrepresent contemporary evolutionary thinking with respect to parasitism.
Mycoplasmas — Mighty but MinisculeMycoplasmas are very small, prokaryotic organisms that lack a cell wall. Their small size and flexibility allows them to pass through bacteriological filters — a feature that makes them frequent nuisances in cell cultures. Mycoplasmas also have very small genomes that are one-fourth or less the size of most bacterial genomes — a feature that make them particularly good candidates for genomic studies (Woese and others 1985). Mycoplasmal DNA has higher proportions of the nucleotides adenine and thymine (they are "A-T rich"), and mycoplasmas show unusual nutritional requirements (Weisburg and others 1989). Mycoplasma and mycoplasma-like organisms (MLOs) collectively compose a class of microorganisms called Mollicutes. Mollicutes contains a variety of organisms that show strong symbiotic associations with other living organisms. Some are associated with insects and plants (Entomoplasmas, Mesoplasm, and Spiroplasma), others are oxygen-sensitive and found in the rumens of bovine and ovine mammals (Anaeroplasma and Asteroleplasma), and some associate with plants, insects, and warm-blooded animals, but do not require sterols for growth (Acholesplasma; Tully and others 1993). The largest subgroup within the mollicutes is the mycoplasmas, which consists of organisms from the genera Mycoplasma and Ureaplasma that must associate with humans and other warm-blooded animals to survive; in some cases, they cause human and animal diseases (Razin and others 1998). Not surprisingly, completed genomes of several mycoplasmas are available (Fraser and others 1995; Himmelreich and others 1996; Glass and others 2000; Chambaud and others 2001). The genome of Mycoplasma genitalium (see cover) is the smallest known at 568 070 base pairs (Fraser and others 1995). Many plant MLOs have not been cultured to date, which has delayed their characterization (Lim and Sears 1989).
Mycoplasmas According to WoodWood begins his article by asserting that all disease, pain, and suffering are the result of the sin of Adam and Eve (Wood 2001). He states, "Creationists generally explain the Curse-related imperfections as degenerations of originally beneficial structures." This is a common recent creationist belief, but this statement essentially endorses the evolutionary concept of exaptation, in which characters that serve particular purposes initially are later co-opted for new functions (Futuyma 1998). In the case of pathogenic bacteria, these microorganisms possess virulence factors, which help them to cause disease by adhering to or damaging host tissues or evading the host immune system (Hacker and others 1997; Hacker and Kaper 2000; Henderson and others 1996; Kathariou 2002; Law and Chart 1998; Potempa and others 2000). However, according to Wood, these virulence factors arose from features that were not originally used for that purpose. Consequently, Wood would argue, for example, that exotoxin A from the soil bacterium and opportunistic pathogen Pseudomonas aeruginosa, which inactivates eukaryotic elongation factor-2 and causes cessation of protein synthesis and cell death in vertebrate cells (Beattie and Merrill 1996; Beattie and others 1996; Yates and Merrill 2001), somehow originally had a benign function that degenerated to an inimical function after the Fall.
In summarizing a review article by Christopher Wren (Wren 2000), Wood writes that Wren "discusses three possible origins for bacterial pathogenicity" — lateral gene transfer, antigenic variation, and genomic decay. According to Wood, "Of these three themes, genomic decay is most consistent with the creationist idea of a degenerating creation." The Wren review shows that lateral gene transfer, antigenic variation, and genomic decay are trends observed in the genomic sequence data from pathogens. All three events are well documented in the literature and the completion of genomic sequences from pathogenic bacteria has extended our understanding of them. In other words, these events are not guesses about what makes microorganisms pathogenic, but are events for which we have solid genomic evidence.
Wood mentions these trends in the genomes of pathogenic bacteria because he thinks that the mycoplasmas demonstrate the best-documented case of pathogen-associated genome decay, even though members of the genus Rickettsia show pseudogenes and split genes, which are both signs of continuing genomic decay (Andersson and others 1998; Andersson and Andersson 1999a, 1999b; Ogata and others 2001). However, Wood also seems to accept that lateral gene transfer and antigenic variation are contributors to microbial pathogenesis. How appropriate is it to build a model of acquisition of pathogenesis that selects only one of these mechanisms while ignoring the other two?
Wood regards the mycoplasmas as a bacterial group that shows phylogenetic discontinuity from other bacteria because mycoplasmas lack a cell wall and use an atypical genetic code. Such a statement shows enormous disregard for earlier ribosomal RNA studies of mycoplasmas that clearly links them not only with some of the gram-positive eubacteria with genomes that contain low percentages of guanine and cytosine — which includes the Bacillus/Lactobacillus group (Lim and Sears 1989; Hori and others 1981; Maniloff 1983; Walker 1983; Rogers and others 1985) — but also specifically with a small subgroup of Clostridia represented by Clostridium innocuum and Clostridium ramosum (Woese and others 1985; Rogers and others 1985; Woese and others 1980).
Is the lack of a cell wall an adequate reason to consider the mycoplasmas phylogenetically discontinuous from the other eubacteria? The answer to this has to be no. Other bacteria lack cell walls. For example, the archaebacterium Thermoplasma acidophilum lacks a cell wall, but is completely unrelated to the mycoplasmas (Walker 1983; Woese and others 1980; Woese and Olsen 1986; Sanz and Amils 1988; Gaasterland 1999). Thus the lack of a cell wall by itself is not uniquely derived.
Furthermore, the ability of bacteria to lose their cell walls is well documented. Cell-wall deficient bacteria (CWDB), or "L-forms" as they are sometimes called, can appear under a variety of circumstances, and intensive antibiotic treatment can select for the formation of persistent L-forms that are resistant to antibiotics that attack cell wall synthesis (Domingue and Woody 1997).
Since the mycoplasmas tend to associate with insect, plant, or warm-blooded-animal cells, the loss of their cell wall is not that difficult to envision. The immune systems of these host organisms constantly search for foreign substances or antigens, and microorganisms that present fewer antigens are less easily recognized by the immune response. Since the cell walls of bacteria contain many potential antigens (Chatterjee 1997; Haslberger and others 2000; Heumann and others 1998; Ryan and others 2001), long-term association of bacteria with specific hosts could select for the generation of CWDB (Paton 1987; Sladek 1986).
Likewise, the origin of the alternative genetic code of some mycoplasmas is not as mysterious as it might initially appear. In some mycoplasmas, the codon UGA, which acts as a translational termination codon in most organisms, encodes the amino acid tryptophan, but besides this exception the genetic code of these organisms is completely standard. Mycoplasmal genomes contain low proportions of guanine (G) and cytosine (C) content, which means that their genomes show the effects of "AT-biased directional mutation pressure", which means that base substitutions in mycoplasmas consistently favor the replacement of C-G base pairs with adenine-thymine (A-T) base pairs. Consequently, C-G-rich codons like CCN, GGN, GCN, or CGN (where N indicates any base) are rare in the coding regions of mycoplasmal genomes, and mycoplasmal proteins have fewer glycine, proline, alanine or arginine residues (see Table 1, p 25). In conserved proteins, mycoplasmas tend to have lysine residues, encoded by AAA and AAG, instead of the arginine residues, encoded by AGG, CGN, and AGA, found in the proteins of other bacteria (Razin and others 1998). Mitochondrial genomes sometimes use an alternative genetic code, and in this case it seems as though selection for a small genome streamlines the total number of tRNAs the genome encodes and favors the use of alternative codons (Knight and others 1999; Saccone and others 2000; Knight and others 2001). Here again the origin of an alternative genetic code is not mysterious (Osawa and others 1992). Therefore the insistence on discontinuity between the mycoplasmas and other eubacteria is almost certainly unwarranted.
Creationist Classification of MycoplasmasSince the discipline of taxonomy attempts to group organisms according to phylogeny, creationist classification schemes often suggest some taxonomic reorganization. Such rearrangements reflect the creationist belief that some organisms were created ex nihilo during the Creation Week and diverged since to produce extant organisms (Sarfati 1999). This emphasis on discontinuity between organisms motivated WJ ReMine to suggest a nomenclature for creationist taxonomy by adapting the term "baramin" coined by Frank Marsh in 1947 to refer to a "created kind". According to the nomenclature formulated by ReMine, a "holobaramin" is a "group containing all and only organisms related by common descent". An "apobaramin" is a "group of holobaramins that are separated from all other organisms by phylogenetic discontinuities". Finally, a "monobaramin" is "a group containing only organisms related by common descent, but not necessarily all of them" and a "polybaramin" is a group of organisms that do not share a common ancestor. ReMine gives the following examples to clarify his nomenclature: mammals are apobaraminic, the placental dogs are holobaraminic, but dogs and wolves are monobaraminic (ReMine 1990). It should be noted that this classification scheme still affirms that biological classification should reflect phylogenetic proximity.
In applying ReMine's nomenclature, Wood proposes that mycoplasmas compose an apobaramin. Since an apobaramin is a group of holobaramins that are separated from other organisms by phylogenetic discontinuities, the mycoplasmas must contain holobaramins. This designation is slightly problematic, since the typical criterion for a holobaramin is the ability to produce fertile offspring; since bacteria lack sexual reproduction, such a standard is unreasonable. Therefore the norm for designating a bacterial group as holobaraminic is somewhat arbitrary. Contemporary bacterial taxonomy often uses the percentage of DNA homology among bacterial genomes to distinguish among bacterial species, and such techniques determine phylogenetic sequences with some accuracy (Martin 2002).
Wood considers Mycoplasma genitalium and Mycoplasma pneumoniae to be members of the "same monobaramin". His reason for this is that the genome of M pneumoniae contains all the genes found in the genome of M genitalium, even in the same gene order. Furthermore, the genetic material unique to M pneumoniae is localized to 6 segments of the genome bordered by repetitive sequences. Since the recombination-inducing protein RecA is encoded by the genome of M pneumoniae, it is entirely conceivable that these M pneumoniae-specific segments were deleted from the genome by RecA-dependent recombination to eventually form a genome that resembles that of M genitalium (Himmelreich and others 1997).
While it is certainly reasonable to suggest that M genitalium and M pneumoniae are directly related by common descent, why should we exclude other mycoplasmas, since 16S rRNA analyses link other mycoplasmas, like M muris, with M pneumoniae (Weisburg and others 1989)? Also, these same studies definitively link the mycoplasmas to the gram-positive bacteria with low percentages of G-C base pairs, even though the mycoplasmas do show some diversity as a group (Woese and others 1985). These data suggest that the mycoplasmas are related to low G-C gram-positive bacteria and form a coherent, though diverse, phylogenetic unit. Such a close affinity with another bacterial group and the somewhat downsized nature of mycoplasmas is hardly coincidental. Certainly the best inference to draw from these data is that the mycoplasmas evolved from a common ancestor (Weisburg and others 1989; Maniloff 1983; Woese and others 1980). This makes their designation as "apobaraminic" highly questionable.
In discussing the sequenced Mycoplasma genomes, Wood uses outdated information. At its initial publication, workers thought that the genome of M genitalium contained 468 genes (Fraser and others 1995) and this is the number used by Wood. Since that time, however, further work and annotation have definitively shown that this was an underestimate. A global mutagenesis study published in 1999 has shown that the genome of M genitalium contains 480 protein-coding sequences and 517 total genes (Hutchison and others 1999). In addition, the genome of M pneumoniae does not encode the 677 genes that Wood quotes from the original reference (Himmelreich and others 1996). Instead, further annotation has shown that the genome of M pneumoniae encodes 688 proteins and 42 RNAs, for a grand total of 730 genes (Dandekar and others 2000). All of these studies were published before Wood's paper, but none is cited or discussed by him.
Mycoplasmas - Made to be Small or Got Small After Getting Made?Because of their greatly reduced genomes, mycoplasmas lack a variety of biosynthetic genes, and Wood thinks that this is an important feature of their genomes. This leads to a potentially interesting question:
But how do we know whether the created ancestors of M genitalium or M pneumoniae had the ability to synthesize amino acids? Could the lack of amino acid synthesis genes be a design feature of this baramin? (Wood 2001: iii)First of all, a lack of biosynthetic capacity is a common feature in many pathogenic bacteria, and genomic reduction is a hallmark of strict parasites (Andersson and others 1998; Ogata and others 2001; Fraser and others 1997; Fraser and others 1998; Kalman and others 1999; Stephens and others 1998; Read and others 2000). Therefore there is nothing unusual about the lack of biosynthetic machinery in the mycoplasmas.
Second, Wood never really answers the questions he posed above, even though he makes it clear that he thinks that M genitalium arose from M pneumoniae or an M pneumoniae-like organism. Therefore, we will answer them. According to contemporary evolutionary thinking, since bacteria arose before warm-blooded animals, all microorganisms that live on or inside animals had to evolve from free-living bacteria that eventually formed symbiotic relationships with warm-blooded animals. All organisms must have some kind of biosynthetic capacity in order to survive unless they are parasites and acquire all their nutrients from the host. Thus it makes sense to postulate that the ancestors of contemporary mycoplasmas almost certainly had some kind of biosynthetic capacity.
From the creationist perspective, if mycoplasmas were originally created to inhabit the bodies of animals, then they might have already had reduced biosynthetic capacities, in the same way that organisms living in milk are unable to synthesize amino acids found in milk. Alternatively, mycoplasmas could have been created as free-living organisms that eventually became animal commensals and parasites.
Which of these hypotheses fits the evidence? According to Wood, the decay of the genomes of mycoplasmas fits the Creation/Fall model, since the Fall is the event that begins the cycle of degradation. However, genomic reduction as an adaptation to a parasitic lifestyle also fits the theory of evolution, and many obligate intracellular parasites show extensive genome reduction (Andersson and others 1998; Ogata and others 2001; Kalman and others 1999; Stephens and others 1998; Read and others 2000). Furthermore, the kinship the mycoplasma share with the Clostridium group is not a surprise for the evolutionary model, but it does pose some problems for the Creation/Fall model.
Another piece of data that fits the theory of evolution is that mycoplasmal genomes show signs of gene duplication as well as genomic degradation. The genome M pneumoniae shows duplication of the lipoprotein genes (Himmelreich and others 1997) and in the genome of Ureoplasma urealyticum there are 6 closely related iron transporter genes that apparently arose by means of gene duplication (Glass and others 2000). In addition, Mycoplasma pulmonis is capable of phase variation whereby it alters its outer membrane protein composition to escape detection by the immune system. The genome of M pulmonis contains several variable surface antigen or vsa genes, and the number of vsa genes varies between strains, thus demonstrating the occurrence of gene duplications within one species of Mycoplasma (Chambaud and others 2001).
Because mycoplasmas show reduced genomes, any gene duplications are probably indications of adaptations to a parasitic or commensal lifestyle. Other examples of obligate intracellular parasites with genomes that host both examples of gene decay and adaptive duplications are the Rickettsia (Ogata and others 2001). Gene duplications are examples of organisms' increasing the "information content" of their genomes, and they conflict with the creationist dictum that "mutations never add information but only reduce it" (Grigg 2000). Thus the evidence suggests that the mycoplasma not only downsized their genomes, but also reinforced other genes to make themselves better pathogens. This favors the evolutionary explanation for the origin of mycoplasmas, since the gene decay found in mycoplasmas does not occur alone, but in combination with gene duplications.
Parasites — Creation or Evolution?
Finally, Wood wishes to construct a framework for how mycoplasmas became human parasites after the Fall. To do so, he compares his ideas with mainstream thoughts on the evolution of parasitism. Wood writes: In the evolutionary model, pathogenicity and parasitism is thought to progress from very virulent (aggressive) forms to harmless or even mutually beneficial relationships. Advocates claim that natural selection will favor hosts that are resistant to the parasite and parasites that are not rapid killers of their own host environments. Thus as time progresses, the parasites evolve to less virulent forms and the hosts become tolerant of the more benign forms of the parasites (Wood 2001: iii).Instead, Wood argues, God created the mycoplasmas as mutualists or commensals that became parasites after the Fall. The adaptation to parasitism included degradation of the mycoplasmal genome. Thus, the evolutionary scenario is challenged by the Creation/Fall model, which predicts just the opposite.
Unfortunately, the evolutionary story Wood tells is oversimplified. His reference for the concept of natural selection's decreasing virulence is a 13-year-old textbook (Pianka 1988). In the 1988 edition, Pianka qualifies this general expectation, writing: "In other situations, such as when a parasite finds itself engaged in a race against its host's immune response, selection may actually favor increased virulence" (Pianka 1988: 296). One must ask why Wood did not consult a more recent edition of Pianka, in which he would have found this revised discussion of the action of natural selection on virulence in parasitic organisms:
To the extent that natural selection favors evolution of reduced parasite virulence (see also subsequent discussion), parasite interactions may evolve gradually toward commensalisms and ultimately even become mutualistic interactions. Of course, selection could also proceed in the opposite direction (reverse arrows). Such changes may also occur during ecological time, as during the ontogeny of parasites (Pianka 1999: 323-5).Pianka then gives examples of natural selection's decreasing virulence in the case of the myxoma and influenza viruses and increasing virulence in malarial parasites (Pianka 1999). Therefore the result of natural selection on the virulence of parasites is not a simple equation that applies to every case. Pianka closes this discussion with the statement "natural selection should favor levels of virulence for parasites with different types of transmission between hosts" (Pianka 1999). Thus Wood has contrasted his own recent creationist view with an inaccurate rendition of contemporary evolutionary thinking regarding parasitism, which amounts to the construction of a straw man.