The nylon bug

Introduction

It is often said that random variation and selection cannot give rise to new genomic information. Let’s have a look at the nylon bug Arthrobacter sp. KI72 (formerly Flavobacterium sp.). Bacterial breakdown of cyclic nylon oligomers (e-caprolactam) was discovered in 1966 by Fukumura (double publishing mmm…) [[i],[ii]]. Since nylon was first synthesized in 1935 by Wallace Carothers at DuPont, the enzyme(s) responsible for the breakdown of nylon oligomers must have evolved within 31 years, possibly weeks, since P. aeruginosa can obtain the capability within 9 days [[iii]]. That means novel, new genetic information must have been added and fixed into the bacterial genome right? WRONG!! Actually that’s right hehe.

Background information:
So let’s look at how novel and new this information is. Nylon-6 is produced from caprolactam by ring cleavage polymerization and consists of more than 100 units of 6-aminohexanoate (Ahx). During the polymerization reaction, some molecules fail to polymerize and remain as linear oligomers, while others undergo head-to-tail condensation to form cyclic oligomers (Figure 1). These cyclic and linear oligomers (called nylon oligomers) are the byproducts from nylon factories and become nylon bug food. Three enzymes, 6-aminohexanoate-cyclic dimer hydrolase (EI), 6-aminohexanoate-dimer hydrolase (EII) and endo-type 6-aminohexanoate oligomer hydrolase (EIII), are found to be responsible for the degradation of the nylon oligomers (Figure 2) [[iv]]. The genes responsible for coding these proteins are located on plasmid pOAD2 (45,519 bp) in strain KI72 (not only K172 though) [[v]]. The end product of the breakdown of the nylon oligomers is 6-aminohexanoate, which is a source (alternate or sole source) of nitrogen and carbon for the bugs. I am unsure how 6-aminohexanoate is precisely metabolized, but I think it falls into the lysine degradation pathway.

Figure 1: 6-aminohexanoate and associated dimers

Figure 2: Breakdown of nylon oligomers by 6-Aminohexanoate dimer Hydrolases

Proposed Catalytic Mechanism of 6-Aminohexanoate-linear dimer Hydrolase (Negoro et al. 2007)
Negoro et al. (2007) wrote an interesting article and proposed a possible mechanism of how the 6-aminohexanoate linear dimer (Ald) is broken down by the E2 enzyme [[vi]].

Background information (from Negoro et al. (2007) [vi]):
1) Asp181 and Tyr170 form a hydrogen bond in the unbound state of E2 (Figure 3b). Asp181, and probably Ser112 and Ile345, are crucial in accepting the 6-aminohexaonate substrate, as upon binding, 6-aminohexaonate forms a stable electrostatic interaction with Asp181 and causes structural alterations that localizes Tyr170 to the catalytic center (Figures 3b and c).
2) The important amino acid residues involved in catalyzing the hydrolysis (catalytic center) of Ald into two 6-aminohexanoate molecules are (Fig 3c):
Tyr170
Asp181
Ile345
Ser112
Tyr215 and/or Lys115
Water molecules within the catalytic centre also play an important role.

Figure 3a: Catalytic center of the E2 substrate. (2dcf.pdb)

Figure 3b: Unbound conformation of EII. (1wyc.pdb)

Figure 3c: Bound conformation of EII. (2dcf.pdb)

Now for the mechanism (as proposed by Negoro et al. (2007) [vi]): Amide hydrolysis.
1) Upon binding, a conformational change occurs as a result of an electrostatic interaction between the Ald-amine and Asp181, causing in the Tyr170 to come into contact with the amide bond of Ald (1-2). Tyr170(phenolic hydroxyl) protonates the amide nitrogen (2)
2) A nucleophilic attack, facilitated by deprotonation of Ser112 by Ile345 and/or Lys115, to Ald by Ser112 results in the formation of a tetrahedral (3) intermediate and subsequent hydrolysis as a result a nucleophilic attack from water on the carbonyl carbon results in the hydrolysis of the amide bond and release of 1 Ald molecule (4).
3) Subsequent deacylation releases the second Ald molecule and influx of water molecules restores the activity of the unbound E2 (its more complicated than this, but it is covered in the paper) [[vi]] (5-8).
Where did the information come from for this novel adaptation?

Amide hydrolysis for other amides present in nature is quite common. Beta-lactamases, the enzymes responsible for the breakdown of… you guessed it… beta-lactams are present in many types of bacteria. Beta-lactamase breaks the 4-membered heteroatomic ring structure (three carbon atoms and one nitrogen atom) open by hydrolyzing the amide bond of a beta-lactam (Figure 5).

Figure 5: Amide bonds of Ald and betalactams

EII’ (nylB’) is an enzyme also encoded on plasmid OAD2 of Arthrobacter sp. KI72. The enzyme has B-lactamase folds and is also able to catalyze the breakdown of Ald. EII’ is a classical carboxylesterase with high activity towards carboxylesters with short acyl chains [[viii]]. The accession code for the the nylB' (EII') amino acid sequence is P07062 and the FASTA sequence can be used to search for similar sequences in other bacterial, archael and eukaryotic genomes at this site (with the BlastP program). After doing so, it can be seen that proteins with beta-lactam folds with 6-aminohexanoate-dimer hydrolytic activity (non-specific Ald amide hydrolysis) is spread throughout the bacterial and archaeal kingdoms.

EII’ is, therefore a pre-existing 6-aminohexanoate-dimer hydrolase with low activity (0.5% that of EII (nylB)) towards Ald that gained an increase in activity towards the Ald through amino acid substitutions in the catalytic cleft containing the “Ser-X-X-Lys” motive [[ix]]. The information needed to metabolize 6-aminohexanoate for energy was already present (presumably the lysine degradation pathway) and the useful esterase with B-lactam folds with minimal Ald hydrolytic activity allowed the bacteria to survive under stressful conditions where the sole energy source was Ald.

The important mutations that increased EII' specificity towards Ald are G181D, D370Y and H266N, and they are situated within the Ald catalytic centre [viii]. The G181D mutation is most likely due to a G:C->A:T transition as a result of cytosine deamination and subsequent mismatch repair by the uracil-DNA glycosylase (UDG) enzyme. The D370Y mutation is most likely the result of a G:C->T:A transversion and the H266N mutation due to a C:G->A:T transversion. It is known that during replication, certain polymerases are more error prone than others in creating transition and/or transversions. It will therefore be interesting to test whether the stringent response (triggered by nutrient downshifts) has any effect on cytosine deamination, UDG enzyme activity and polymerase III and/or IV activity (mutA cells have a higher proportion of transversions [x]). Gene expression analysis with microarrays with simultaneous sequencing of Arthrobacter sp. KI72 during various time intervals will be illuminating and would make it possible to determine whether the stringent response (and other stress responses) has any effect on the specificity of mutations. LexA, SpoT and recA activity and their downstream effect on specific polymerase activity will also be interesting.

It is known that certain environment factors and how bacterial cells respond to them (e.g SOS-response, stringent response, general stress response and heat-shock response [xi]) can have an affect on the specificity of mutations (e.g. cytosine deamination, transversions and transitions) because of the increased activity of specific polymerases induced by a stress response. LexA, SpoT and recA activity do play a role, and whether it may result in an adaptive mutation response (specific shift towards specific transversion and/or transitions) should be testable. And maybe there is a method to the madness... cytosine deamination from a teleological perspective.

Skepticism over the origins of EII'.
Is the pre-existing esterase the result of a frame-shift mutation? Could someone please provide the accession number and database of the PR.C sequence from Ohno (1984) [xii].? Was the classic esterase with beta-lactamase folds there from the start?
Ohno (1984) suggested that the coding sequence "originally" specified a 472-residue-long arginine-rich protein. However this "472-residue-long arginine-rich protein" has no significant homology to any known functional protein.

EI and EIII did not arise through a frame-shift mutation. EI has a high homology with amidases present in a lot of strains of bacteria, meaning a pre-existing enzyme was capable of breaking the amide bond of cyclic dimers. EII has a high homology with L-aminopeptidase/D-esterases present in many other bacteria.

The evidence points to (and seems like the simplest explanation - Occam's razor) a pre-existing classical carboxylesterase with B-lactamase folds (NylB'/EII') with low activity (0.5% that of EII (nylB) towards nylon oligomers, gained an increase in activity towards the the oligomers through amino acid substitutions in the catalytic cleft containing the “Ser-X-X-Lys” motive without affecting the activity towards its original substrates. Not a frame shift mutation.

References

[i]. Fukumura T. Hydrolysis of cyclic and linear oligomers of 6-aminocaproic acid by a bacterial cell extract. J Biochem (Tokyo). 1966 Jun;59(6):531-6.

[ii]. Fukumura T. Bacterial breakdown of e-caprolactam and its cyclic oligomers. Plant Cell Physiol 1966;7:93-104

[iii]. Prijambada ID, Negoro S, Yomo T, Urabe I. Emergence of nylon oligomer degradation enzymes in Pseudomonas aeruginosa PAO through experimental evolution. Appl Environ Microbiol. 1995 May;61(5):2020-2.

[iv]. Negoro S. Biodegradation of nylon oligomers. Appl Microbiol Biotechnol. 2000 Oct;54(4):461-6.

[v]. Kato K, Ohtsuki K, Koda Y, Maekawa T, Yomo T. et al. A plasmid encoding enzymes for nylon oligomer degradation: nucleotide sequence and analysis of pOAD2. Microbiology. 1995 Oct;141 ( Pt 10):2585-90.

[vi]. Negoro S, Ohki T, Shibata N, Sasa K, Hayashi H et al. Nylon-oligomer degrading enzyme/substrate complex: catalytic mechanism of 6-aminohexanoate-dimer hydrolase. J Mol Biol. 2007 Jun 29;370(1):142-56.

[vii]. Negoro S, Ohki T, Shibata N, Mizuno N, Wakitani Y et al. X-ray crystallographic analysis of 6-aminohexanoate-dimer hydrolase: molecular basis for the birth of a nylon oligomer-degrading enzyme. J Biol Chem 2005 Nov 25;280(47):39644-52

[viii] Ohki T, Wakitani Y, Takeo M, Yasuhira K, Shibata N, Higuchi Y, et al. Mutational analysis of 6-aminohexanoate-dimer hydrolase: relationship between nylon oligomer hydrolytic and esterolytic activities. FEBS Lett. 2006 Sep 18;580(21):5054-2058.

[ix]. Negoro S, Ohki T, Shibata N, Mizuno N, Wakitani Y et al. X-ray crystallographic analysis of 6-aminohexanoate-dimer hydrolase: molecular basis for the birth of a nylon oligomer-degrading enzyme. J Biol Chem 2005 Nov 25;280(47):39644-52

[x] Balashov S, Humayun MZ. Specificity of spontaneous mutations induced in mutA mutator cells. Mutat Res. 2004 Apr 14;548(1-2):9-18.

[xi] Foster PL. Stress responses and genetic variation in bacteria. Mutat Res. 2005 Jan 6;569(1-2):3-11.

[xii] Ohno S. Birth of a unique enzyme from an alternative reading frame of the preexisted, internally repetitious coding sequence. Proc Natl Acad Sci U S A. 1984 Apr;81(8):2421-2425.

Structures generated with Chimera