Regulation of the erm(C) gene in Staphylococci from reservoir with different usage of macrolides.

A high prevalence of macrolide resistant isolates has been found among staphylococci isolated from animals [4]. The erm(C) gene is the most common gene encoding macrolide resistance in staphylococci [8]. It is well known that expression of the erm(C) gene is normally regulated by formation of hairpin structures upstream for the erm(C) gene [5,9] rendering the start codon of erm(C) gene non-accessible. Only 14- and 15-membered macrolides like erythromycin can induce expression of the gene and induce resistance while 16-membered macrolides cannot activate expression of erm(C) [7]. If deletions from 16 to 116 bp occur in the regulatory area, expression of the erm(C) gene becomes constitutive [10]. Constitutive expressed erm(C) genes give resistance not only to 14- and 15-membered macrolides, like erythromycin, but also to 16-membered macrolides like spiramycin, tylosin and streptogramin B [6]. Deletions are believed to be the result of high concentration of non-inducible macrolides like tylosin in the environment, selecting for constitutive expression of the macrolide resistance. In the presence of macrolides, like tylosin, this could give staphylococci with constitutive expressed erm(C) a selective advantage not only to sensitive staphylococci but also to staphylococci containing regulated erm(C) genes. In this article we have investigated the ratio of regulated and constitutive expressed erm(C) genes in human and animal reservoirs (cattle and pigs) with differences in uses of the 16-membered macrolide tylosin.

Large amounts of the macrolide tylosin have been used for pig production in Denmark for growth promotion and therapy [1]. In 1996 68,350 kg of tylosin was used for growth promotion and 1,350 kg for therapy. No macrolides have been used for growth promotion for cattle but spiramycin and tylosin have been used therapeutically for treatment of mastitis [3]. A total of 644 kg macrolides, primarily tylosin, was used for cattle in 1996 in Denmark. Local variations in treatment strategies exist depending on the choice of the veterinarian but due to the used strain collection this effect will be minimal. At the same time 5,934 kg of penicillin was used (Erik Jacobsen, personal communication). The usage of macrolides for treatment of infections in human in general practice constitutes approximately 20–25 percent of the total usage of antibiotics in humans. However, in human medicine 16-membered macrolides are not used. The macrolides used in human medicine in Denmark are primarily erythromycin (14-membered) and azithromycin [2].

A total of 185 macrolide resistant staphylococci were tested, twenty-nine staphylococci from cattle (8 Staphylococcus aureus and 21 coagulase negative staphylococci (CNS)), 111 Staphylococcus hyicus isolates of porcine origin and 45 S. aureus from non-hospitalized humans [4]. All animal isolates were obtained from the DANMAP surveillance program with one isolate per herd hereby representing a broad spectrum of farms in Denmark. Human isolates were obtained from individuals of both sex and from different age groups. All human, bovine and 96 porcine isolates were collected from 1995 to 1998. The remaining 15 porcine isolates were collected in 2001, two years after the discontinued usage of growth promoters in Denmark.

The presence of erm(C) was confirmed using previous described primers [4]. Among the animal isolates from 1995–98, all except one porcine isolate contained the erm(C) gene (Table 1). erm(C) was found in 23 (69%) of the human isolates and 7 (47%) of the porcine isolates from 2001. PCR for erm(A) and erm(B) was performed for porcine isolates from 2001. No positive amplicons were obtained (data not shown). A set of PCR primers (RegermC-1: 5'-TAAACCGTGTGCTCTACGAC-3' and RegermC-2: 5'-CCTTTTCCTGAGCCGATTTC-3') was constructed spanning the regulatory region upstream the erm(C) gene and PCR amplification was performed. Fourteen amplicons from selected strains from the three different reservoirs were sequenced. Results are presented in Figure 1.

Table 1

Identification of presence and regulation of the erm(C) gene was done using PCR. Classification of genes as regulated or constitutive was based on size of the obtained amplicon. Consumption of antimicrobial agents in the three reservoirs is indicated.

		Presence and regulation of erm(C) among staphylococci of human and animal origin
	Origin	Human	Cattle	Pigs
		S. aureus	staphylococci	S. hyicus
	Year	1995–1998	1995–1998	1995–1998	2001
	Usage	low	moderate	high	low
	n =	45	29	96	15
erm(C)	positive*	69	100	99	47
	regulated	81	31	9	43
	constitutive	19	69	91	57

* All numbers are given in percentage

Figure 1

Regulation of expression of the erm(C) gene. Deletions in the regulatory region of erm(C) in staphylococci from animal and human origin were identified by sequencing PCR amplicons obtained using primers RegermC-1 (5'-TAAACCGTGTGCTCTACGAC-3') and RegermC-2 (5'-CCTTTTCCTGAGCCGATTTC-3'). Origins of strains are indicated as well as Shine-Delgano (SD-1 and SD-2) sequences, sequence of the leader peptide (by amino acid translation) and start of erm(C) (Met...). Underlined bases indicate position of hairpin II and III.

Deletion of 16 bp, 107 bp, 109 bp and 111 bp was found in the regulatory region of erm(C). Based on the obtained sequences, the size of the PCR amplicons could be used to determine whether an erm(C) gene was expressed constitutive or regulated. Results on regulation of the erm(C) gene in the three reservoirs are presented in Table 1.

The differences in occurrence of regulated erm(C) between isolates from the different reservoirs were statistically significant (chi-square test). Significant difference could be demonstrated between S. hyicus from pig from 1995–98 and 2001 (p = 0.034) and between staphylococcal isolates from pigs and cattle (p = 0.013), isolates from cattle and humans (p < 0.001) and isolates from humans and pigs (p < 0.001).

In a reservoir with high usage of tylosin constitutive expressed erm(C) genes were dominant (91% in porcine isolates from 1995–98). In a reservoir with moderate usage of tylosin constitutive expressed genes was still most prevalent (69% in cattle and 57% in pigs from 2001) while in a reservoir with no usage of tylosin regulated erm(C) genes was most prevalent (81% in human isolates). When comparing porcine erm(C) positive S. hyicus isolates from 1995–98 with isolates from 2001 a change in the ratio could be observed between constitutive and regulated genes. This change to a higher prevalence of regulated erm(C) genes could reflect the changes in usage of tylosin introduced by the discontinuous usage of growth promotion in 1998 in Denmark. Results presented here indicate that the ratio of constitutive to regulated erm(C) genes could be related to the amount of tylosin used in the different reservoirs. Statistically significant differences in occurrence of constitutive and regulated erm(C) genes were demonstrated for reservoirs with different usage of tylosin. This indicates that not only have the usage of tylosin selected for macrolide resistant staphylococci [2] but regulation of expression of the erm(C) gene has also been changed. Since regulated erm(C) do not give resistance to tylosin and only very limited amount of spiramycin and tylosin has been used for human therapy, the higher prevalence of constitutive expressed resistance genes in animal isolates compared to human isolates could be associated to the usage of tylosin as growth promoter and prevalence of constitutive expressed erm(C) in the human reservoir could indicate an animal origin of the resistance.