OUP user menu

Presence of erm gene classes in Gram-positive bacteria of animal and human origin in Denmark

Lars Bogø Jensen , Niels Frimodt-Møller , Frank M. Aarestrup
DOI: http://dx.doi.org/10.1111/j.1574-6968.1999.tb13368.x 151-158 First published online: 1 January 1999


A classification of the different erm gene classes based on published sequences was performed, and specific primers to detect some of these classes designed. The presence of ermA (Tn554), ermB (class IV) and ermC (class VI) was determined by PCR in a total of 113 enterococcal, 77 streptococcal and 68 staphylococcal erythromycin resistant isolates of animal and human origin. At least one of these genes was detected in 88% of the isolates. Four isolates contained more than one erm gene. ermB dominated among the enterococci (88%) and streptococci (90%) and ermC among staphylococci (75%) with ermA (Tn554) present in some isolates (16%). Variations in the presence of the different genes when comparing staphylococcal isolates of human and animal origin were observed.

  • Erythromycin resistance
  • Animal
  • erm gene classification

1 Introduction

The macrolide tylosin is the most commonly used antimicrobial agent in pig production in Denmark. Recent national surveys have found widespread resistance to macrolides in staphylococci, streptococci and enterococci isolated from pigs in Denmark [1, 2]. Macrolides are used for treatment of humans with erythromycin as first choice also as a substitute for penicillin in cases where patients are allergic to penicillin [3].

Resistance to macrolides is based on different mechanisms: target modification by point mutation or methylation of 23S rRNA inhibiting binding of macrolides so protein synthesis is not interfered with [4], hydrolysis of the lactone ring in the macrolide [5] and efflux pumps removing the antibiotic internally from the bacteria [68]. Resistance to macrolides can spread from animals to human, either by spread of the resistant bacteria or by horizontal gene transfer of mobile DNA elements. To determine whether a horizontal spread of resistance has occurred, a characterization of the mechanisms for resistance is needed.

According to the published literature [912] the most frequently found macrolide resistance genes in bacterial isolates from animals and humans are the erm genes. These genes encode a methyltranferase that has specific target residues in the 23S rRNA [4]. Methylation will inhibit binding of erythromycin. Several erm genes have been sequenced and named. However, the names associated with the genes have not been chosen according to homology with previously published genes, thus creating confusing names.

In this study a classification of the published genes based on sequence identity in the coding regions of the erm genes is presented. This classification was used to study the prevalence of selected erm gene classes by PCR in erythromycin resistant bacteria of animal and human origin in Denmark.

2 Materials and methods

2.1 Classification of erm gene classes

Several erm genes have been deposited in GenBank (Table 1). Among the published sequences names are not consistent. Using the DNASIS software the published sequences were aligned according to percent identity in the coding region and using the maximum likelihood method a phylogenetic tree was created. The minimal percentage of identity for a gene to be placed in a class was set at 95% in the sequenced area of the coding open reading frame.

View this table:
Table 1

erm genes published in GenBank

Class 0 (ermE group)
S. erythraeaChromosomal DNA1257 bpermEX51891
S. erythraeaChromosomal DNA1113 bpermE2M11200 M11304
Class I (ermDK group)
B. anthracisChromosomal DNA864 bpermJL08389
B. licheniformisChromosomal DNA864 bpermDM29832
B. licheniformisChromosomal DNA864 bpermKM77505
Class II (ermF group)
B. fragilisConjugal element801 bpermFUM62487
B. fragilisTn4351801 bpermFM17124
B. fragilisTn4551801 bpermFSM17808
Class III (ermA2 group)
C. xerosisTn5432762 bpermCXU21300
C. diphtheriaepNG2855 bpermCdM36726
C. diphtheriaepSV5(pNG2)762 bpermAX57320
C. diphthteriaepNG2762 bpermAX51472
Class IV (ermB group)
L. fermentumpLEM3753 bpermU48430
S. pyogenespMD101750 bpermX66468
Enterococcusplasmid738 bperm2X82819
S. pyogenespBT233738 bperm2X64695
E. faecalisnot determined738 bpermBU86375
S. lentuspSES20738 bpermBU35228
S. agalactiaepIP501738 bpermX72021
C. perfringensChromosomal DNA738 bpermBPU18931
E. colipIP1527738 bpermBCM19270
E. hiraenot determined738 bpermAMX81655
S. sanguispAM77738 bpermAMK00551
E. faecalisTn917 (pAD2)738 bpermBM11180 M36722
S. pneumoniaeTn1545738 bpermBX52632
E. faecalispAMβ1283 bpermAMM20334
E. faecalispAMβ1161 bpermAMM20335
Class V (ermG group)
E. faecalisTn7853735 bpmetht.L42817
B. sphaericusChromosomal DNA735 bpermGM15332
Class VI (ermC group)
B. subtilispIM13735 bpermCM13761
S. aureusJ3356::POX7;1735 bpermCU36911
S. aureusJ3356::POX7;3759 bpermCU36912
S. chromogenespPV141735 bpermMU82607
S. simulanspV142735 bpermMAF019140
S. epidermidispNE131735 bpermMM12730
S. aureuspE194735 bpermCJ0175-8
S. aureuspE5735 bpermCM17990
S. aureuspT48735 bpermCM19652
S. equorumpSES6735 bpermCX82668
S. hominispSES5735 bpermCY09001
S. haemolyticuspSES4a735 bpermCY09002
S. hyicuspSES21735 bpermCY09003
S. aureuspRJ5283 bpermCL04687
S. aureuspA22226 bpermCX54338
Unique sequences
H. influenzaeChromosomal DNA1173 bpermAL45536/42023
B. fragilispBF41035 bpermFM14730
Arthrobacter sp.Chromosomal DNA1023 bpermAM11276
S. fradiaepSK101960 bpermSFM19269
C. perfringensChromosomal DNA774 bpermQL22689
LactobacilluspGT633735 bpermGTM64090
S. pyogenesChromosomal DNA732 bpermTRAF002716
S. aureusTn554732 bpermAK02987
  • Not totally sequenced.

  • Accession number in GenBan.

2.2 Bacterial isolates

A total of 258 erythromycin resistant isolates were tested. Among these 61 were of human origin and 197 from animals. All 44 human isolates of Staphylococcus aureus were collected in 1996 in Denmark from non-hospitalized patients. Isolates of several phage types were included indicating that these isolates were representatives of common S. aureus phage types of human origin found in Denmark. All S. aureus strains were susceptible to methicillin. All 16 human Enterococcus faecium isolates were isolated from faecal samples.

The animal isolates originated from the DANMAP surveillance program [1] and therefore reflect the number of isolates obtain from this project. They include 16 E. faecium isolates from broilers, 35 E. faecium, 36 E. faecalis and 16 S. hyicus isolates from pigs and eight staphylococcal (two S. aureus and six coagulase negative staphylococci) and nine enterococcal isolates (five E. faecium and four E. faecalis) from cattle. Furthermore, 77 Streptococcus suis isolates from a strain collection of diagnostic samples from pigs obtained from 1991 to 1996 were included.

2.3 PCR amplification of the erm genes

DNA extractions and PCR amplification were performed according to Jensen et al. [13]. From all isolates two single colonies were picked for isolation of total DNA and PCR performed. Strains were only considered positive if both amplifications were positive. If a positive and negative amplification was obtained two new single colonies were picked and a second round of amplification was performed. All PCR amplifications were run with a MgCl2 concentration of 1 mM.

Primers were designed according to the published sequences and the classes for the erm genes defined in this work (see Section 3 and Table 1). All designed primers were tested for their specificity on several published strains (Table 2). For the ermA (Tn554) [14] gene the sequence for Tn554 was chosen for design of primer. For ermB (class IV) the sequence from Tn917 [15] was chosen and for ermC (class VI) the sequence from pE194 was chosen. The sequences of all primers and position on selected genes from the two classes and ermA (Tn554) are listed in Table 2. The primers were verified using strains listed in Table 3. The Tm values for the individual primers were calculated using the Tm DETERMINATION [16] available on INTERNET (http://alces.med.umn.edu/rawtm.html).

View this table:
Table 2

Sequence, position, class and reference for PCR primers used in this study

NameSequence (5′–3′)PositionClassReference
  • All primers used for PCR amplification were designed inside the coding regions. All numbers indicated refer to the sequences published in GenBank. The access numbers are: K02987 for the Tn554 containing ermA, M11180 for the Tn917 containing ermB and J01755 for pE194 for ermC.

View this table:
Table 3

Reference strains for erm genes

1206S. aureusermATn554[24]
RN1389S. aureus::Tn554ermATn554Dr. Courvalin, personal communication
ermE, class 0
S. lividans/pIJ702+ermEermE0[25]
E. coli/pIJ4026ermE0Dr. Vester, personal communication
ermB, class IV
JIR2220E. coli DH5α/pJIR599ermBPIV[26]
B. subtilis/pAM77ermAMIV[27]
JH2-2E. faecalis::Tn1545ermBIV[28]
S. lentusermBIV[11]
JM107E. coli/pSES20ermBIV[29]
CH116E. faecalis::Tn5384ermBIV[30]
BR-151B. subtilus/pAM77ermBIV[27]
ermC, class VI
B. subtilis/pE194ermCVI[31]
B.3HU104B. subtilis/pE194ermCVIDr. Courvalin, personal communication
RN4220S. aureus::pSES5ermCVI[11]
HB101E. coli/pKH80ermCVI[32]
L. reuteri/pGT633ermGTVI[33]
JIR2879E. coli DH5α/pJIR1120ermQ[26]

2.4 Sequencing

The nucleotide sequence of the amplification products was determined by cycle sequencing [17] using AmplitaqFS dye terminator kit and a 373A automatic sequencer (Applied Biosystems/Perkin Elmer, Foster City, CA, USA). The DNASIS software (Hitachi Software Engineering Co., Ltd) was used for sequence analysis.

3 Results

3.1 Classification of erm gene classes

On the basis of aligning the published sequences the erm genes were grouped into seven classes and some unique genes (Table 1). Class 0 contained genes from erythromycin producing strains while the remaining classes contained acquired genes for macrolide resistance in bacteria. Using the maximum likelihood method a phylogenetic tree was created verifying the defined classes (Fig. 1). The class number was assigned according to the length of the coding region and not due to the number of sequenced genes.

Figure 1

Phylogenetic tree of erm genes. The phylogenetic tree was created by the maximum likelihood method. Only fully sequenced genes are included in the tree and GenBank numbers and gene classes in parentheses are listed. For genes and organisms see Table 1.

3.2 Prevalence of selected erm genes among bacterial isolates of human and animal origin

All designed primers were tested on several reference strains (Table 2). Positive amplicons were only obtained from the reference strains containing the corresponding genes. No cross reaction towards other genes were seen. This is to our knowledge the first time so many reference strains have been used to check the specificity of designed primers.

The results of PCR amplification for selected gene classes in the tested isolates are given in Table 4. Using PCR the presence of at least one of the three genes were found in 88% of the isolates. For all strains amplicons of correct size was obtained and a selected number of amplicons was sequenced to verify that the correct target had been amplified (data not shown). Four isolates, three S. aureus of human origin and one E. faecium from a pig, contained more than one gene for macrolide resistance. The ermB (class IV) was the most common class found among enterococci (88%) and streptococci (90%). No amplicons of any size was obtained for ermA (Tn554) in enterococci and streptococci and only one E. faecium of animal origin contained the ermC (class VI).

View this table:
Table 4

Prevalence of selected erm genes in enterococci, streptococci and staphylococci of animal and human origin in Denmark

BacteriaS. aureusE. faeciumE. faeciumstaphylococcienterococciS. suisE. faecalisE. faeciumS. hyicus
ermB, IV01715086931280168
ermC, VI3600300011252
  • All numbers indicate a positive amplicon of correct size. n=number of isolates tested, N.D.=number of isolates where the genetic background was not determined.

  • Since four isolates contained more than one erythromycin resistance gene the total number of positive reactions will be greater than the number of isolates.

  • Two S. aureus and six coagulase negative staphylococci.

  • Five E. faecium and four E. faecalis.

Among human S. aureus both ermA (Tn554) (23%) and ermC (VI) (82%) were found. In staphylococci isolated from animals ermA (Tn554) (5%) and ermC (VI) (63%) were found. No significant difference in prevalence of these genes in staphylococci of animal (24 isolates) and human (44 isolates) origin could be detected. The ermB (class IV) was not found in staphylococci.

4 Discussion

In this study a re-classification of the erm genes was suggested and the prevalence of selected erm gene classes in bacteria of animal and human origin was detected by use of PCR. On the basis of the re-classification the first published sequence for ermA (Tn554) does not belong to class III in which two other genes named ermA are placed. Especially for the ermA genes the published names are not consistent with the classes proposed in this study. In the phylogenetic tree the ermA (Tn554) was grouped together with the ermTR from Streptococcus pyogenes. These two sequences were 82% homologous in the coding region and were for that reason not defined as a class. Several genes belonging to class IV are named ermAM and in many cases almost identical genes are called ermB. We propose that the published names are kept but that additional to these the class to which the gene belongs should be defined and noted after the name. As an example Tn917 contain the ermB (class IV) gene.

The presence of two classes of genes as well as the ermA (Tn554) was tested among selected erythromycin resistant isolates. By limiting the detection to two classes of genes and the Tn554 at least one of these genes was found in 88% of the tested isolates. In the remaining 11 percent the genotype was not determined but genes of other erm classes or other mechanism for erythromycin resistance could be present.

ermA (Tn554) was found in staphylococci predominantly isolated from humans in accordance with previously published studies [18]. ermB (class IV) dominated among enterococci and streptococci as found in previously studies [10, 19, 20, 11, 21, 4, 18]. ermC (VI) dominated among staphylococci of human and animal origin [10, 18].

In the study differences in the prevalence of erm gene classes in enterococci/streptococci and staphylococci were observed. For different staphylococci and for enterococci and streptococci identical erm gene classes were observed. S. pyogenes and S. pneumoniae of human origin have previously been found to harbor ermB (class IV) [22, 20, 23, 8] and transfer of genes between enterococci and S. suis of animal origin to these bacteria could take place. However, in Denmark the frequency of macrolide resistance among these bacteria is low, making it difficult to obtain resistant isolates.

Identical genes in different bacteria can be a result of horizontal transfer but could also indicate a common reservoir for resistance or evolution from the same ancestor. Proof of horizontal transfer would be the presence of identical mobile DNA elements in different bacterial species of human and animal origin. Further studies of the position and the mobility of the different erm genes are needed to determine whether horizontal transfer takes place. Such studies are ongoing at present.


We would like to acknowledge the following persons for their technical assistance: René Hendriksen, Mette Juul, Lissie Kjær Jensen, Karina Kristensen, Inge Hansen, Dorthe Nielsen, Anne Lykkegaard Lauritsen and Christina Aaby Svendsen. Special thanks to Flemming Bager, DVL for creating lists of resistant bacteria from the DANMAP database, Thomas D. Leser for creating the phylogenetic tree and Birte Vester, Copenhagen University, Knud Bϕrge Pedersen, Anders Meyling and Henrik C. Wegener, DVL, for helpful comments in the preparation of the manuscript.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
View Abstract