Genetic change creates new functions

Genetic mutations are the raw material of evolution. While some mutations are harmful, the majority are selectively neutral, and a significant number are beneficial—conferring new functions that natural selection can act upon. Decades of research in molecular biology, population genetics, and experimental evolution have documented numerous cases in which mutations produce entirely new genes and functions, from lactase persistence in humans to nylon-degrading enzymes in bacteria.^{2, 3} Gene duplication, whole-genome duplication, and de novo gene birth from noncoding DNA are established mechanisms by which genomes gain new capabilities.²

Types of mutations

A mutation is any heritable change to the nucleotide sequence of DNA. The simplest type is a point mutation (also called a single-nucleotide polymorphism or SNP when it becomes established in a population), in which one base pair is substituted for another. Point mutations are further classified as transitions (purine to purine or pyrimidine to pyrimidine) or transversions (purine to pyrimidine or vice versa).⁴ In protein-coding regions, a point mutation may be synonymous (producing no change in the amino acid sequence due to the redundancy of the genetic code), missense (changing one amino acid to another), or nonsense (creating a premature stop codon that truncates the protein).⁴

Beyond point mutations, several classes of structural variation alter larger stretches of DNA. Insertions add one or more nucleotides into a sequence, while deletions remove them. When insertions or deletions occur within a coding region and involve a number of bases not divisible by three, they shift the reading frame of all downstream codons—a frameshift mutation that typically renders the resulting protein nonfunctional.⁴ Duplications copy a segment of DNA, ranging from a few bases to entire genes or even whole chromosomes. Inversions reverse the orientation of a segment within a chromosome, and translocations move a segment from one chromosomal location to another.^{4, 5}

Crucially, the functional consequence of any mutation depends on its context. The same types of molecular events—substitutions, insertions, duplications—can be harmful, neutral, or beneficial depending on where in the genome they occur, what gene or regulatory element they affect, and what environment the organism inhabits.^{2, 3}

Beneficial mutations in humans

Perhaps the most widely cited example of a beneficial mutation in humans is lactase persistence—the ability to digest lactose, the sugar in milk, into adulthood. In most mammals, the gene encoding the enzyme lactase (LCT) is downregulated after weaning, causing lactose intolerance in adults. In 2002, Enattah and colleagues identified a single-nucleotide variant (C/T–13910) in an enhancer region upstream of LCT that maintains lactase expression throughout life.⁶ This variant is found at high frequency in populations with a long history of dairying, particularly in northern Europe, and subsequent studies identified independent lactase-persistence mutations in East African, West African, and Middle Eastern pastoral populations, a striking example of convergent evolution driven by the same selective pressure—the nutritional advantage of digesting milk.⁷ Genetic evidence indicates that the European variant rose to high frequency within the last 5,000 to 10,000 years, making lactase persistence one of the strongest known signals of recent positive selection in the human genome.⁸

The sickle-cell allele provides a textbook case of a mutation whose effects depend on dosage and environment. A single point mutation in the beta-globin gene (HBB) substitutes valine for glutamic acid at position six, producing hemoglobin S (HbS). Individuals homozygous for HbS develop sickle-cell disease, a serious condition in which red blood cells deform under low-oxygen conditions. However, heterozygous carriers (HbAS)—those with one normal and one sickle-cell allele—enjoy substantial resistance to severe Plasmodium falciparum malaria, a benefit first demonstrated by Anthony Allison in 1954.⁹ This heterozygote advantage explains why the HbS allele is maintained at high frequencies in malaria-endemic regions of sub-Saharan Africa, the Mediterranean, and South Asia, despite being harmful in the homozygous state.¹⁰ The sickle-cell case illustrates that calling a mutation simply "harmful" or "beneficial" is an oversimplification; the same allele can be either, depending on genetic background and ecological context.^{9, 10}

New functions in bacteria

Some of the most compelling evidence that mutations can generate genuinely new biological capabilities comes from microbiology, where large population sizes and short generation times allow evolution to be observed directly. In 1975, Japanese researchers discovered a strain of Flavobacterium (initially classified as Achromobacter guttatus K172) living in wastewater ponds near a nylon factory. The bacterium could metabolize 6-aminohexanoic acid cyclic dimer, a synthetic byproduct of nylon-6 manufacture that did not exist before the industrial era.¹¹ Kinoshita, Negoro, and colleagues characterized three novel enzymes responsible for this ability, none of which bore significant resemblance to any previously known enzymes.¹¹ Subsequent work by Negoro and others showed that the nylonase enzymes arose through a frameshift mutation in a previously noncoding DNA sequence, generating a new open reading frame that encoded a functional protein.¹² Because nylon is an entirely synthetic material first produced in 1935, these enzymes represent new biological functions that evolved in response to a novel, human-created substrate.^{11, 12}

An even more detailed case comes from Richard Lenski's Long-Term Evolution Experiment (LTEE), which has tracked twelve populations of Escherichia coli in a glucose-limited medium since 1988. Under aerobic conditions, E. coli cannot use citrate as a carbon source—this inability is in fact one of the defining characteristics used to identify the species in clinical microbiology.¹³ Around generation 31,500 (roughly 15 years into the experiment), one of the twelve populations suddenly evolved the ability to grow aerobically on the citrate present in the medium, dramatically increasing its population density. Blount, Borland, and Lenski reported this result in a 2008 paper in PNAS, showing through "replay" experiments with frozen ancestral populations that the innovation required a specific historical sequence of prior mutations—a phenomenon they called historical contingency.¹³ Further analysis revealed that the key genetic change was a tandem duplication that placed a previously silent citrate transporter gene (citT) under the control of a promoter active under aerobic conditions, effectively rewiring an existing gene's regulation to produce a new metabolic capability.¹⁴

Mutation rates and population sizes enabling observed beneficial mutations^{3, 13}

Gene duplication and new genes

One of the most important mechanisms for generating new genetic material is gene duplication. In his influential 1970 book Evolution by Gene Duplication, Susumu Ohno argued that duplication of genes (and sometimes entire genomes) is the principal source of evolutionary novelty.¹⁵ When a gene is duplicated, the organism possesses two copies where previously there was one. Because one copy continues to perform the original function, the second copy is freed from the constraints of purifying selection and can accumulate mutations that would otherwise be eliminated. Over time, this redundant copy may acquire a new function (neofunctionalization), subdivide the ancestral function between the two copies (subfunctionalization), or degrade into a nonfunctional pseudogene.^{15, 16}

The vertebrate globin gene superfamily provides one of the best-studied examples of this process. Hemoglobin (which transports oxygen in red blood cells) and myoglobin (which stores oxygen in muscle tissue) are encoded by separate genes that arose from the duplication and divergence of a single ancestral globin gene hundreds of millions of years ago. The alpha-globin and beta-globin gene clusters themselves contain multiple paralogous genes produced by additional rounds of duplication, including the fetal hemoglobin genes that enable efficient oxygen transfer across the placenta—a function not present in the ancestral single-globin system.¹⁷ The entire trajectory from one gene to a family of functionally distinct genes is precisely the kind of increase in genetic information that the "mutations only destroy" claim denies is possible.^{15, 17}

Whole-genome duplication (WGD) represents an even more dramatic expansion of genetic material. Two rounds of WGD occurred early in vertebrate evolution, providing the raw material for the diversification of developmental regulatory genes, including the four clusters of Hox genes that pattern the vertebrate body plan (invertebrates typically possess a single cluster).¹⁸ Additional WGD events have been documented in the ancestry of teleost fishes, flowering plants, and yeasts, each followed by extensive gene loss and neofunctionalization of the retained duplicates.¹⁸

Beyond duplication of existing genes, entirely new genes can arise de novo from previously noncoding DNA sequences. A striking example was described in codfishes, where antifreeze glycoprotein (AFGP) genes evolved from noncoding genomic regions approximately 13 to 18 million years ago, coinciding with the cooling of the Northern Hemisphere. Researchers identified a nine-nucleotide element in the noncoding founder region that supplied the codons for one Thr-Ala-Ala unit, which was then amplified through tandem duplication to build the repetitive AFGP-coding sequence.¹⁹ This case demonstrates that mutations can convert nonfunctional DNA into an essential survival gene—a clear instance of new genetic information arising where none existed before.¹⁹

Endogenous retroviruses and shared ancestry

Endogenous retroviruses (ERVs) provide a particularly powerful form of genetic evidence for common descent. When a retrovirus infects a germ-line cell and integrates its DNA into the host genome, that viral insertion is passed to all descendants of that cell. Because retroviruses integrate at essentially random positions among billions of base pairs, the probability that two species would independently acquire an insertion at the same genomic location is vanishingly small.²⁰ When the same ERV is found at the same chromosomal position in two species, the most parsimonious explanation is that the insertion occurred once in their common ancestor and was inherited by both lineages.²⁰

Humans and chimpanzees share thousands of ERV insertions at identical genomic locations.²¹ When phylogenetic trees are constructed based on the pattern of shared and lineage-specific ERV insertions across primates, the resulting branching order matches the tree derived from morphological data and other molecular markers: humans share more ERVs with chimpanzees than either shares with gorillas, and all three share more with orangutans than with Old World monkeys.^{20, 21} Some ERV families tell an even more detailed story. The PTERV1 element, for instance, is found in chimpanzees, gorillas, and Old World monkeys but is absent from the human genome, consistent with either a deletion in the human lineage or an insertion after the human-chimpanzee divergence.²⁰ These patterns of shared insertions are exactly what common descent predicts and are extremely difficult to explain under independent creation, since there is no functional reason for a designer to place identical viral sequences at the same positions in the genomes of related species.^{20, 21}

Human chromosome 2 fusion

Humans possess 23 pairs of chromosomes, while chimpanzees, gorillas, and orangutans each have 24 pairs. If humans and the other great apes share a common ancestor, then either one chromosome fission occurred in the ape lineage or one fusion occurred in the human lineage. In 1991, IJdo and colleagues at Yale University discovered direct molecular evidence for a fusion event. They identified a site on the long arm of human chromosome 2 (at band 2q13) containing two arrays of the vertebrate telomeric repeat sequence (TTAGGG)_n arranged in a head-to-head orientation—exactly the pattern expected if two chromosome ends had fused together.²²

Further analysis revealed a vestigial (inactivated) centromere in the region corresponding to the position of the centromere on one of the two ancestral chromosomes, providing additional confirmation that human chromosome 2 was formed by the fusion of two separate chromosomes.^{22, 23} The two ancestral chromosomes correspond precisely to chimpanzee chromosomes 2A and 2B (chromosomes 12 and 13 in the older nomenclature), as demonstrated by comparative genomic mapping and confirmed by the 2005 sequencing of the chimpanzee genome.^{21, 23} The fusion site, the vestigial centromere, and the gene-by-gene synteny (matching order of genes) between human chromosome 2 and chimpanzee chromosomes 2A and 2B together constitute strong evidence that humans and chimpanzees descended from a common ancestor with 48 chromosomes.^{22, 23}

Neutral mutations and the molecular clock

The claim that mutations are "always harmful" overlooks the largest category of mutations entirely: those that are selectively neutral. In 1968, Motoo Kimura proposed the neutral theory of molecular evolution, arguing that the vast majority of mutations that become fixed (reach 100% frequency) in a population are not driven by natural selection but instead spread through the random process of genetic drift.² Kimura showed mathematically that for neutral mutations, the rate at which substitutions accumulate in a lineage equals the mutation rate per individual per generation—a result independent of population size.²

The neutral theory provided a theoretical foundation for the molecular clock, a concept first proposed by Emile Zuckerkandl and Linus Pauling in 1962, who observed that the number of amino acid differences between homologous proteins in different species is roughly proportional to the time since those species diverged.²⁴ If most molecular changes are neutral and accumulate at a constant rate, then the degree of sequence divergence between two species can be used to estimate the time of their last common ancestor. Molecular clocks have since been calibrated using fossil dates and applied across the tree of life, from estimating the human-chimpanzee divergence at 6 to 8 million years ago to dating the origin of animal phyla in the Precambrian.^{24, 25}

Empirical studies have confirmed that neutral and nearly neutral mutations dominate the molecular landscape. The 2005 chimpanzee genome project found that the patterns of evolution in human and chimpanzee protein-coding genes are "highly correlated and dominated by the fixation of neutral and slightly deleterious alleles," with only a small fraction of fixed differences attributable to positive selection.²¹ This finding is consistent across mammals and indeed across all domains of life: most DNA sequence changes over evolutionary time are neither beneficial nor harmful but simply neutral passengers, drifting to fixation by chance.^{2, 21}

The question of "genetic information"

At the heart of the "mutations cannot create new information" claim lies an ambiguity about what "information" means. Intelligent-design proponents frequently invoke information theory to argue that natural processes cannot increase the information content of a genome, but they typically employ a colloquial definition of "information" that conflates it with functional complexity, rather than using the precise technical definition from Claude Shannon's mathematical theory of communication.^{1, 26}

In Shannon's framework, the information content of a message is a function of its entropy—the degree of uncertainty or surprise it conveys. By this measure, a random sequence of nucleotides has higher information content than a highly repetitive one, because each position is less predictable. A point mutation that changes one base to another in a sequence increases the entropy of that position across a population, thereby increasing information in the Shannon sense.²⁶ This is the opposite of what the "mutations destroy information" argument claims.

Christoph Adami, applying information theory to biological systems, has argued that the biologically relevant measure is not Shannon entropy per se but rather the information a genome encodes about its environment—the degree to which the genome's sequence is correlated with conditions that affect fitness.²⁶ Natural selection increases this environmental information by preferentially retaining sequences that improve an organism's fit to its ecological niche. Computational experiments using digital organisms have demonstrated that populations evolving under selection accumulate genomic information over time, with more complex environments producing organisms with higher genomic information content.²⁶ There is, in short, no theoretical barrier in information theory to the generation of biological complexity through mutation and selection.²⁶

The empirical examples discussed in this article—gene duplication producing new globin genes with new functions, noncoding DNA giving rise to antifreeze proteins, a tandem duplication enabling citrate metabolism in E. coli, nylonase enzymes arising from frameshift mutations in noncoding sequence—are all concrete instances of new functional genetic information arising through documented mutational mechanisms.^{11, 13, 17, 19} The claim that mutations "cannot create new information" is not supported by information theory, molecular biology, or the observed history of life on Earth.^{1, 26}