Download Resistance gene naming and numbering: is it a

J Antimicrob Chemother 2016; 71: 569 – 571
doi:10.1093/jac/dkv351 Advance Access publication 28 October 2015
Resistance gene naming and numbering: is it a new gene or not?
Ruth M. Hall1* and Stefan Schwarz2
1
School of Molecular and Microbial Biosciences, The University of Sydney, Sydney, 2006 New South Wales, Australia;
2
Institute of Farm Animal Genetics, Friedrich-Loeffler-Institut, Höltystr. 10, 31535 Neustadt-Mariensee, Germany
*Corresponding author. Tel: +61-2-9351-3465; E-mail: [email protected]
In the genomic era, studying the epidemiology of individual antibiotic resistance genes and resistance gene discovery are open to all. However, the identification and naming of resistance genes is not currently understandable
by all owing to a plethora of competing nomenclature systems, many of which do not comply with the basic rules
of bacterial gene nomenclature. Change is needed urgently. Here, we make a case for the resistance research
community to begin this process by agreeing on an answer to the question of when a new gene number should
be assigned. This cut-off is of necessity arbitrary and we suggest a threshold value of ≥2% difference in the
sequences of the DNA, predicted protein or both as a realistic boundary for assigning a new gene number.
This proposal can be a starting point for agreement or debate followed by renumbering of the affected gene
families.
Introduction
For quite some time now, the pace of discovery of new resistance
genes has been high and the need for gene nomenclature and
numbering systems that are clear to all and comply with the
basic rules of bacterial gene nomenclature is increasing.
Unfortunately, in the past there has been little agreement on
how this might be achieved.1 As interest in antibiotic resistance
increases and large numbers of genome sequences are being
analysed for the presence of resistance genes, the many different
systems used currently are causing confusion. Hence, the need for
rationalization and agreement on a more consistent central
ground is more pressing than ever. One important area that
needs to be addressed and rationalized is that the rules developed
for naming and numbering differ substantially and this is the
aspect we address here.
The problem
Currently, different systems are used for the classification of genes
conferring resistance to particular antibiotics, e.g. sulphonamides,
trimethoprim, b-lactams, tetracyclines or phenicols. Some systems do not have published rules and some, with or without published rules, are not in agreement with nomenclature rules that
are accepted by the broader bacterial genetics community. To
date, nomenclature systems have been proposed and published
for b-lactam resistance genes,2 – 6 tetracycline resistance genes,7
macrolide, lincosamide and streptogramin resistance genes,8 qnr
genes,9 mec genes10 and genes encoding 16S rRNA methylases
conferring extensive aminoglycoside resistance.11 However,
these nomenclatures are not consistent with each other in the
rules used either to name or to number genes. This situation
arose because informal networks, composed of leading scientists
in their respective field, each developed their own nomenclature
systems and the specifics of different fields have unfortunately
driven the direction of the system developed. Consequently,
there are disparate and essentially incompatible views on a
number of nomenclature issues such as how a gene name is composed and when and how a new number should be used for two
resistance genes that differ in their DNA sequence. The boundary
used to define a new gene is essentially arbitrary. However, currently, it ranges from a single base pair difference when it also
involves a difference in the amino acid encoded to a .20% difference at the amino acid level. Neither of these positions is suitable
for use in the genomic era and a more realistic threshold value is
needed. Moreover, polymorphisms in the gene sequence can
occur without necessarily changing the sequence of the encoded
product and non-identical DNA sequences can encode the same
product.
At one end of the spectrum, for genes conferring resistance to
b-lactam family antibiotics, genes are named after their protein
product instead of using conventional bacterial gene nomenclature and a single base pair change leading to a single amino
acid substitution is currently viewed as sufficient to assign a
new gene number.2 – 6 However, single amino acid substitutions
do not necessarily lead to a difference in the substrate profile
that is clinically relevant or to changes in the pI of the protein.
The approach of assigning a new number when a single base
pair change alters a single amino acid serves to make closely
related genes appear to be different from one another. In addition, this system is vulnerable to sequence errors, as a single
error in an early sequence has since been identified and corrected
making oxa1 and oxa30 (also called blaOXA-30) identical12 and
blaOXA-24 was found to be identical to blaOXA-40.13
At the other end of the spectrum, nomenclature proposals for
genes that confer resistance to tetracyclines or macrolides,
# The Author 2015. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: [email protected]
569
Leading article
lincosamides and streptogramins have declared a .20% difference in the protein identity as the value for assignment of a designation (a letter or number) indicating a new gene.7,8 Using a
.20% amino acid difference as the threshold serves to make different genes, that in some cases cannot be detected with a single
PCR assay and are in different contexts,14,15 appear to be the
same. This threshold is from a time when hybridization was
often used to identify resistance genes, as sequencing was
more difficult. Two genes whose products differ by .20% would
usually provide even greater variation in their DNA sequences and
enough to permit distinct probes to be designed.8 However, the
days of hybridization as a tool for resistance gene identification
have long since passed and these nomenclature systems need
to be revisited in the light of current practice, which readily yields
DNA sequences resulting in PCR detection methods.
Three further published systems each have their own rules. The
qnr gene nomenclature system uses both of the aforementioned
approaches with genes/proteins sharing ≥70% identity being
assigned to a ‘family’ and a single amino acid difference allowing
a new ‘allele’ number to be assigned.9 In the case of mec genes,
an identity of ≥70% defines ‘allotypes’ of mec designated mecA,
mecB or mecC. Numbers are assigned to members of each allotype that differ by .5%.10 Nomenclature for 16S RNA methylases
causing extensive aminoglycoside resistance is complex.11 A gene
whose product has either an amino acid identity of ,50% with
the closest known 16S rRNA methylase or has been confirmed
to methylate a new residue of 16S rRNA may receive a brand
new gene designation, i.e. a different set of letters. A gene
whose product exhibits between 50% and 95% amino acid identity with the closest known 16S rRNA methylase gene will be
assigned a new alphabetical letter according to the existing
gene name of the closest gene. Finally, a variant number is
assigned to sequences with a difference of ,5%.11
a separation timescale measured in millions of years,16,17 and that
a 20% difference in the protein sequence is likely to reflect an even
larger difference at the DNA level, this value seems far too large. On
the other hand, a single base pair difference seems far too small, as
reported differences have previously been traced to sequencing
errors.
In the interests of moving this debate into the public arena, we
propose that a threshold of ≥2% difference in the DNA or protein
sequences or both be agreed as the point at which a new gene
number is assigned and that this threshold should apply to all
resistance gene families. This would reduce future complications
caused by sequencing errors, past or present, and allow a simple
designation system for genes that are distinct within the 20% difference range. Sequences with ,2% difference at either the DNA
or amino acid level or both, which can provide useful epidemiological markers, can be named as variants of an existing numbered gene using a variety of systems. For example, variants can
be assigned an additional number, e.g. oxa1-1, oxa1-2 etc., in
which case oxa30 might have been oxa1-1, creating far less confusion when the error was detected. For proteins, the standard
biochemical nomenclature for substitutions, which uses the single
letter codes for the amino acids coupled with a number corresponding to the position of the substitution in the protein
sequence (e.g. G172H), is simple and effective. A similar system
could be used for genes.
A proposal
This study was supported by internal funding from the University of Sydney
and from the Friedrich-Loeffler-Institut.
Some simple rules should be followed. First of these is that experimental confirmation that the gene in question confers the appropriate resistance phenotype is essential. This would be seen as an
increase in the MIC (or a reduction in the zone diameter) of the
respective antimicrobial agent(s) when the gene is present relative to the value in an identical genetic background when it is
absent. Second, variation in genes that are intrinsic to a particular
species, such as the blaSHV gene found in all Klebsiella pneumoniae
and the blaOXA-51-like gene found in Acinetobacter baumannii,
should not be assigned numbers when they are in their original
chromosomal location as this variation is simply equivalent to
that seen in the loci used for MLST. A name should be assigned
to each of these genes using standard bacterial gene nomenclature, e.g. ampC for genes encoding AmpC-type b-lactamases. The
genes become named resistance genes only if they have been
mobilized and are now part of a mobile element or found in
another species. Other basic requirements are: (i) only full-length
sequences should be assigned gene numbers; (ii) only naturally
occurring variants, not those created in vitro by mutation, should
be numbered; and (iii) variations in promoter sequences, which
alter the level of gene expression, should not be considered.
Taking into account the fact that 20% divergence in DNA
sequence is at the upper end of differences between corresponding
Escherichia coli and Salmonella enterica genes, which occurred over
570
Acknowledgements
We thank the many people who have drawn these problems to our
attention and requested changes.
Funding
Transparency declarations
None to declare.
References
1 Hall R, Partridge S. Unambiguous numbering of antibiotic resistance
genes. Antimicrob Agents Chemother 2003; 47: 3998.
2 Bush K, Jacoby G. Nomenclature of TEM b-lactamases. J Antimicrob
Chemother 1997; 39: 1– 3.
3 Jacoby GA, Bush K. b-Lactamase nomenclature. J Clin Microbiol
2005; 43: 6220.
4 Jacoby GA. b-Lactamase nomenclature. Antimicrob Agents Chemother
2006; 50: 1123– 9.
5 Lee SH, Jeong SH. Nomenclature of GES-type extended-spectrum
b-lactamases. Antimicrob Agents Chemother 2005; 49: 2148 –50.
6 Leflon-Guibout V, Heym B, Nicolas-Chanoine M. Updated sequence
information and proposed nomenclature for blaTEM genes and their
promoters. Antimicrob Agents Chemother 2000; 44: 3232 –4.
7 Levy SB, McMurry LM, Barbosa TM et al. Nomenclature for new tetracycline resistance determinants. Antimicrob Agents Chemother 1999; 43:
1523– 4.
Leading article
8 Roberts MC, Sutcliffe J, Courvalin P et al. Nomenclature for macrolide
and macrolide-lincosamide-streptogramin B resistance determinants.
Antimicrob Agents Chemother 1999; 43: 2823– 30.
9 Jacoby G, Cattoir V, Hooper D et al. qnr gene nomenclature. Antimicrob
Agents Chemother 2008; 52: 2297 –9.
10 Ito T, Hiramatsu K, Tomasz A et al. Guidelines for reporting novel mecA
gene homologues. Antimicrob Agents Chemother 2012; 56: 4997– 9.
11 Doi Y, Wachino J, Arakawa Y. Nomenclature of plasmid-mediated
16S rRNA methylases responsible for panaminoglycoside resistance.
Antimicrob Agents Chemother 2008; 52: 2287– 8.
12 Boyd D, Mulvey MR. OXA-1 is OXA-30 is OXA-1. J Antimicrob Chemother
2006; 58: 224–5.
13 Evans BA, Amyes SG. OXA b-lactamases. Clin Microbiol Rev 2014; 27:
241–63.
JAC
14 Li J, Li B, Wendlandt S et al. Identification of a novel vga(E) gene variant
that confers resistance to pleuromutilins, lincosamides and streptogramin
A antibiotics in staphylococci of porcine origin. J Antimicrob Chemother
2014; 69: 919–23.
15 Wendlandt S, Heß S, Li J et al. Detection of the macrolide-lincosamidestreptogramin B resistance gene erm(44) and a novel erm(44) variant in
staphylococci from aquatic environments. FEMS Microbiol Ecol 2015; 91:
pii: fiv090.
16 Ochman H, Wilson AC. Evolution in bacteria: evidence for a universal
substitution rate in cellular genomes. J Mol Evol 1987; 26: 74– 86.
17 Okoro CK, Kingsley RA, Quail MA et al. High-resolution single nucleotide
polymorphism analysis distinguishes recrudescence and reinfection in
recurrent invasive nontyphoidal Salmonella Typhimurium disease. Clin
Infect Dis 2012; 54: 955–63.
571