Also Available in:

Nylon-degrading bacteria: update

Nylonase does not support microbes-to-mankind evolution


Theistic evolutionists connected with Biologos,1 continue to assert that random mutations have created a ‘new gene’ in bacteria that degrades nylon. This assertion comes from a misunderstanding that was popularized by atheist professor of biology William Thwaites in 1985, who claimed that the enzymatic activity arose from a frameshift mutation—thus from a randomness.2 This was in turn based on the speculation of a Japanese geneticist, Susumu Ohno. Such an occurrence would indeed be fortuitous. I have been following this subject, but after reading a helpful post by Ann Gauger3 of the Discovery Institute I realized it was time to publish an update.

Figure 1: A cartoon (ribbon) representation (green) of a nylon-degrading (mutant) enzyme composed of 392 amino acids.
Source: ebi.ac.uk

Research published since my 2003 paper4 has vindicated my argument that the main nylon-degrading enzyme did not arise ‘presto’ from a frame-shift mutation. Recent reviews acknowledge that the gene, nylB, that codes for the main nylon-degrading enzyme under discussion (nylB, figure 1) came from an existing gene (nylB’) that codes for a protein that had some existing enzymatic activity for degrading nylon compounds. The enzyme was a carboxylesterase that had a particular β-lactamase fold that could grab hold of and degrade nylon naturally. Because nylon is a man-made fibre, it was thought that no natural enzyme would be able to attack it. However, the basic amide bond of nylon is common in living things (figure 3), so it is not surprising that an existing enzyme can degrade nylon to some extent.

The popular science media tried to find a way to put an evolutionary spin on this, but the initial explanations, based on incomplete knowledge and evolutionary assumptions, were completely wrong. Even the anti-creationist Wikipedia as early as 29 June 2011 stated: “A 2007 paper that described a series of studies by a team led by Seiji Negoro of the University of Hyogo, Japan, suggested that in fact no frameshift mutation was involved in the evolution of the 6-aminohexanoic acid hydrolase.”5 But ten years later, Biologos’ Dennis Venema is still pushing this false story.

The abstract of a 2007 paper by Negoro and colleagues states, “Present models illustrate why new activity against the nylon oligomer has evolved in an esterase with [beta]-lactamase folds, while retaining the original esterolytic functions.”6 In other words, it involved the modification of an existing enzyme, not something completely new (like from a frameshift mutation).

I have not found any of the researchers saying ‘Ohno was wrong’ about the frameshift idea, possibly because the researchers are Japanese and in a face-saving culture such would be considered rude (and Ohno was a respected scientist). But he was wrong and so are the creation skeptics, both atheistic and theistic, who have not kept up with the research, preferring to simply repeat what Thwaites asserted more than three decades ago, without any hard evidence.

Bacterial degradation of a man-made chemical bond

Figure 2: Pseudomonas aeruginosa bacterial culture.

How did the parent gene achieve increased enzymatic activity on nylon oligomers? Research has shown that this involved the substitution of just two amino acids (giving a 153-fold increase in activity). Furthermore, this increase is stepwise, so that one mutation can follow the other; it does not require them to be simultaneous. Yakashima et al., summarize this.7 It would be incredibly rare for two simultaneous changes to occur, even in a microbe, as Michael Behe has shown in his book, The Edge of Evolution. Behe presents strong evidence that two simultaneous mutations is about the limit of what can occur with natural processes, which puts a severe limit on what ‘evolution’ can achieve. More recently, Sanford has modelled this in human populations, showing that the ‘waiting time’ for new traits is an extreme problem for evolution.8 Unlike humans, however, microbes have large populations, rapid reproduction, and relatively small genomes, so they can tolerate high mutation rates,9 especially when this occurs in plasmids rather than the main chromosome. This is no threat to a creationist view of biology. The amount of potential change is still strongly limited by time, chance, and complexity.

Note that the nylon bonds attacked by the enzyme nylB are similar to the amide bonds in naturally occurring proteins (figure 3). Thus it is not at all surprising that a slight modification to an existing enzyme that can already break amide bonds can ‘tweak’ it to the point where it can break similar bonds in nylon. Carboxylesterases introduce a water molecule to break amide or ester bonds. There are many different carboxylesterases in living things and they are very important in metabolizing a wide range of drugs (e.g. beta-blockers, statins, cocaine, and aspirin in humans) and insecticides (e.g. organophosphates, pyrethrums, and carbamates).10 Nylon is a ‘polyamide’ and thus falls into the general class of molecules that can be attacked by carboxylesterases.

In my original response in 2003, I accepted at face value the old claim by the researchers that the nylon-degrading enzymes were de novo (brand new). They had based their claim on not having found a parent gene that was sufficiently similar to be the source. However, this was not true. As stated above, subsequent papers show that they were derived from existing enzymes and that the existing category of enzymatic activity is retained, just that the binding for the new substrate is enabled by a couple of small changes. So, in my original response I conceded far too much to the evolutionary view. I should have known better, and applied the apt adage that ‘the absence of evidence is not necessarily evidence of absence’. Grandiose claims of evolutionists should never be accepted at face value, even when they seem completely genuine. We need to check everything.

Creation implications

The research underlines once again the very limited capacity of mutations and natural selection to create the complex features that characterize all living things; such as metabolic pathways involving multiple enzymes or nano-machines such as helicases, kinesins, ATP synthase, the bacterial flagellum or even a truly novel enzyme. The sorts of mutational changes involved with these nylon-digesting bacteria that slightly modify an existing enzyme do not explain the origin of such things as brand new enzymes.

Organisms have clearly been designed to be able to adapt, but the sorts of adaptations we see give no support to the ‘big picture’ claim that all biological complexity on earth came about by mutations and natural selection.

Figure 3: The basic reaction that produces nylon (when repeated many times). The blue box shows the amide bond, which is common in biology (e.g. in proteins).

Kawai (2010) pointed out that while nylon degradation has been readily achieved in a range of microbes, degradation of polyethylene or polypropylene has not, probably because there is almost no existing catalytic activity for these compounds among the many known enzymes.11 Existing catalytic activity would give a target for mutations to tweak but with no target there is nothing to tweak. The implication is that such catalytic activity would require brand new enzymes, not just slight modifications of existing ones, and so is likely out of reach of natural processes to achieve.

However, scientists have certainly not fully explored the microbiome in the whole world. Thus, it is quite possible that enzymes do exist naturally that can break down the bonds found in polyethylene and polypropylene, for example, but they have just not been found yet.

Some thoughts about ‘new information’

Note also that modification of an existing enzyme to broaden its range of substrates represents an overall loss of specificity for that enzyme. Since specificity is related to information content, it can be argued that these changes represent a loss of information.

Consider the following strings of Roman characters:

  1. She has an automobile (21 characters)
  2. aushanaS he libeo mto (21 characters)
  3. Sue has a red Porsche (21 characters)

1 and 3 are English sentences, but 2 is a randomized set of characters with no meaning in English. Statistically, all take the same number of ‘bits’ to code them. In other words, it takes the same amount of information to describe the three sets of letters. This is the Shannon ‘information’ that evolutionists like to talk about. But you can see that just looking at the number of letters and their order does not measure the information content, for two of those sentences convey a higher meaning, and sentence 3 is even more specific than sentence 1. How do you gauge such things? This is a problem for evolution, because life is defined by its information content and information of this nature (1 and 3) does not arise from random events.

Note that increased complexity is not the same as increased information. String 2 is complex, but it contains almost no information (other than the fact that it includes 21 random English characters) because it does not specify anything. In the same way, a pile of sand is much more complex than a silicon chip (which is made from sand). It would take a huge number of bits to describe the size, shape, and location of each grain, but it contains very little specificity, or meaningful information, compared to the silicon chip. Evolution needs increased specified complexity, like the silicon chip, not just increased complexity, like a randomized pile of sand. This article explains some of these important principles.

However, the bacteria have a mechanism for not just modifying an enzyme, but for maintaining the original enzyme at the same time. The carboxylesterase to be modified resides on a plasmid, exterior to the main bacterial chromosome. Thus, when it is modified the original carboxylesterase still functions with its original specificity. In this case, a new function has indeed been added. Even if the direction of change is towards less specificity, overall there is an addition of genetic information.

This is one reason why we caution people to not use the ‘mutations never create new information, they only destroy information’ argument. The truth is much more subtle. Random mutations do tend to scramble the specified complexity in the genome, but this does not mean it is impossible for new traits to appear, especially when we are considering traits appearing in systems designed to produce them. I recommend this paper by Dr Rob Carter, about the subject of increased information through mutations: Can mutations create new information? This paper outlines the problem with simple statements about mutations and information. There is still an insurmountable problem for naturalism, but it is not a simple argument to get right—undoubtedly because biology is anything but simple!

In conclusion, I stand by my original suggestion (2003), which I backed up with evidence at the time, that what happened with these bacteria has resulted from a designed ability to adapt. This appears to be similar to the hypermutation of beta cells in the immune system that in turn leads to the production of new antibodies. In these bacteria, DNA sequences known as ‘transposable elements’ seem to be involved in stimulating controlled genetic recombination in the plasmid-based genes. Of course the origin of such a system would be a huge problem for evolution to explain, so no evolutionist would dream that such a mechanism could exist. This is one of the many ways that evolution harms science; those so afflicted by it are likely to not expect sophistication when they are studying living things.12

Published: 19 May 2017

References and notes

  1. Venema, D.R. and McKnight, S., Adam and the Genome: Reading Scripture after Genetic Science, Brazos Press, January 2017. Return to text.
  2. A frameshift mutation involves the insertion or deletion of bases (‘letters’) into a gene such that the reading of the 3-base codons (at least one for each amino acid in the coded protein) is upset ‘downstream’ of the mutation. For example, insertion of one base results in the ‘reading frame’ being shifted by one ‘letter’ and the codons will we read differently (different groups of three), producing a different amino acid sequence, or if a ‘stop’ codon is generated, the protein will be cut short (truncated). Return to text.
  3. Gauger, A., The Nylonase Story: When Imagination and Facts Collide; evolutionnews.org, 4 May 2017. Return to text.
  4. Batten, D., The adaptation of bacteria to feeding on nylon waste, TJ (now Journal of Creation) 17(3):3–5, 2003; creation.com/nylon. Return to text.
  5. See also Truman, R., Nylon-eating bacteria—part 2: refuting Ohno’s frame-shift theory, J. Creation 29(2):78–85. Return to text.
  6. Negoro, S., et al., Nylon-oligomer Degrading Enzyme/Substrate Complex: Catalytic Mechanism of 6-Aminohexanoate-dimer Hydrolase, J. Mol. Biol. 370(1):142–156, 2007, ISSN 0022-2836 | doi:10.1016/j.jmb.2007.04.043. Return to text.
  7. Kawashima, Y., et al., Molecular design of a nylon-6 byproduct-degrading enzyme from a carboxylesterase with a β-lactamase fold, FEBS Journal 276:2547–2556, 2009 | doi:10.1111/j.1742-4658.2009.06978.x. Return to text.
  8. Sanford, J., et al., The waiting time problem in a model hominin population, Theoretical Biology and Medical Modelling 12:18, 2015. Return to text.
  9. See Carter, R., Genetic entropy and simple organisms. Return to text.
  10. Esterases: Integrative molecular phenotying, metabolomics.se, accessed May 2017. Return to text.
  11. Kawai, F., The biochemistry and molecular biology of xenobiotic degradation by microorganisms, Biosci. Biotechnol. Biochem. 74(9):1743–1759, 2010. Update (April, 2018): A previously unknown bacterium has been found with enzymes that degrade polyethylene: Plastic-munching enzyme. Return to text.
  12. See Sarfati, J., Who’s really pushing bad science?, 2000, creation.com. Return to text.