Article Text

Download PDFPDF

Short communication
Changes in antimicrobial susceptibility of Mycoplasma bovis isolates from Great Britain
  1. R. D. Ayling,
  2. R. S. Rosales,
  3. G. Barden and
  4. F. L. Gosney
  1. Animal Health and Veterinary, Laboratories Agency, (AHVLA), Weybridge, Woodham, Lane, Addlestone, Surrey KT15 3NB, UK
  1. E-mail for correspondence: Roger.Ayling{at}

Statistics from

The UK has approximately 10 million cattle (Veterinary Medicines Directorate 2011), and each year approximately 1.9 million of these cattle are affected by respiratory disease. A disease that is often complex, caused by viruses and/or bacteria, with as many as one-third of these infected with the Mycoplasma bovis pathogen (Nicholas and Ayling 2003). The welfare and economic effects of clinical cases of M. bovis on farming are therefore substantial and costs may be even more when morbidity associated with respiratory infection is considered. Apart from respiratory disease, M. bovis is also associated with other clinical signs, including mastitis, arthritis, meningitis, infertility, abortion and keratoconjunctivitis (Nicholas and Ayling 2003). Elimination of M. bovis is difficult and treatment with antimicrobials often has limited effect unless animals are treated early in the course of disease. In recent years, as much as 11 tonnes per annum of active antimicrobials have been sold for use in the UK cattle antimicrobial products, which includes all intramammary products (Veterinary Medicines Directorate 2011). With no commercial vaccines available in Europe, although use of autogenous vaccines has shown some success (Nicholas and others 2006), it is important to effectively target antimicrobial treatment to ensure prudent use of antimicrobials and to reduce the development of antimicrobial resistance. Between 2004 and 2009, in vitro minimum inhibition concentration (MIC) data for 45 M. bovis isolates have been determined for up to 13 antimicrobials.

Materials and methods

Forty-five epidemiologically unrelated M. bovis isolates from cattle were tested, as well as the National Collection of Type Cultures (NCTC) type strain (NCTC 10131); details are given in Table 1. Samples were previously submitted to the Animal Health and Veterinary Laboratories Agency for mycoplasma detection and identification by culture and molecular methods that include PCR and denaturing gradient gel electrophoresis (McAuliffe and others 2005). Isolates were subsequently stored at below −50°C. The majority of isolates were from respiratory cases, but two were from cases of arthritis and seven from mastitis cases. M. bovis isolates were cultured in Eaton's broth medium (Nicholas and Baker 1998) at 37°C in air with 5 per cent carbon dioxide for between 24 and 96 hours.


Details of Mycoplasma bovis isolates and minimum inhibition concentration data in μg/ml for 13 antimicrobials

The method used to determine MICs essentially followed the guidelines for mycoplasma MIC testing (Hannan 2000) and the micro-broth dilution method described by Ayling and others (2000, 2005). ‘Sensititre’ plates that contained freeze-dried antimicrobials were designed to test the following antimicrobials—the macrolides: erythromycin, tilmicosin, tylosin and the modified macrolide, tulathromycin; the lincosamides, phenicols and aminocyclitols: clindamycin, lincomycin, spectinomycin; chloramphenicol and its derivative florfenicol; the fluoroquinolones: danofloxacin, enrofloxacin and marbofloxacin; and the tetracycline: oxytetracycline. Tylosin, marbofloxacin and tulathromycin were not initially included in the study of isolates from 2004 and 2005, but viable isolates have since been tested. Most antimicrobials were tested by doubling dilutions in the range 0.12–32 μg/ml with the exception of tulathromycin, which was tested at 0.25–128 μg/ml. The final antimicrobial concentrations were achieved when 200 μl of media/inoculum was added. In total, 190 μl of Eaton's media (Nicholas and Baker 1998) without antibiotics and phenol red were added to the plates before adding 10 μl of inoculum. The inoculum density was standardised using optical density measurement of the broth culture, adjusted to an OD450 of approximately 0.1, which is equivalent to 1×108 organisms per ml. This was then diluted 1 in 10 in fresh media to give a final concentration of cells equivalent to 5×105 organisms per ml in the ‘Sensititre’ plate. ‘Sensititre’ plates were sealed and incubated at 37°C for 48 hours, before being centrifuged and examined for growth using an inverted mirror.

MICs were determined for each isolate in duplicate, and the M. bovis NCTC-type strain was used as a control organism. Two antimicrobial-free control wells, containing inoculated medium, were used to show normal uninhibited growth. The MIC was determined as the minimum concentration that inhibited the growth of M. bovis. The criteria used for accepting the tests were that the MIC of the control organism had to be within one dilution of the expected result and the duplicate tests also within one dilution, with the higher MIC value being used. Statistical analysis to determine whether there is a significant increase or decrease in trend of MIC values was carried out using the Cuzick's test for trend (Cuzick 1985).


Following incubation and centrifugation, examination of the ‘Sensititre’ plates showed clear buttons of cells in the no antimicrobial control wells and where the antimicrobial had not been effective. The MIC results are given in Table 1, with the range and MIC50 and MIC90 results given in Table 2. The range of MIC values obtained for the M. bovis isolates covers the full range of dilutions of the antimicrobials tested, with the exception of erythromycin that had MICs of 32 or greater than 32 μg/ml, but that is not an antimicrobial that is used in cattle in the UK. The lowest MIC50 values were obtained for the fluoroquinolones and for tulathromycin. The MIC90 values for all antimicrobials ranged from 8 to greater than128 μg/ml.


Minimum inhibition concentration range, MIC50 and MIC90 values in μg/ml for all Mycoplasma bovis isolates tested and trends in MIC values from 2004 to 2009

A comparison of the MIC50 values for the respiratory isolates from 2004 and 2009 is given in Fig 1. There are significant trends using the Cuzick's test for tilmicosin, lincomycin and clindamycin, which have decreasing MICs over time, whereas chloramphenicol, oxytetracycline, danofloxacin, enrofloxacin and marbofloxacin have increasing trends in MIC values over time (Table 2). For lincomycin and clindamycin, the MIC50 values have decreased between 2004 and 2009, from 8 to 1 μg/ml and greater than 32 to 0.25 μg/ml, respectively, whereas for chloramphenicol, florfenicol, oxytetracycline, danofloxacin, enrofloxacin and marbofloxacin, the MIC50 value has increased, with the MIC50 for oxytetracycline increasing from 1 to 32 μg/ml. The MIC90 values for danofloxacin, enrofloxacin and marbofloxacin are also shown in Fig 1 and show an increase from 0.25, 0.25 and 1 μg/ml in 2004 to 8, greater than 32 and 32 μg/ml, respectively, in 2009.

FIG 1:

Comparison of the MIC50 values for the respiratory isolates from 2004 and 2009. MIC90 values are also shown for the fluoroquinolones. MIC, minimum inhibition concentration

The seven mastitis isolates were from 2007, 2008 and 2009. A comparison of the MIC range, MIC50 and MIC90 values for these mastitis isolates with the respiratory isolates from these years is given in Table 3. It is apparent from these data that the mastitis isolates' lowest range value and MIC50 value is higher for some macrolides, lincosamides and spectinomycin than from respiratory isolates.


Minimum inhibition concentration range, MIC50 and MIC90 values in μg/ml for Mycoplasma bovis comparing respiratory and mastitis isolates from 2007 to 2009


The lack of a cell wall in Mycoplasma species makes them intrinsically resistant to β-lactams and to all antimicrobials that target the cell wall, which substantially limits the range of antimicrobials that may be effective against them. Antimicrobial sensitivity testing in vitro is currently the only real way of monitoring the sensitivity of antimicrobials and likely effectiveness in vivo. Guidelines for MIC testing for veterinary mycoplasmas isolates were described by Hannan (2000), and more recently approved guidelines have been issued for testing three mollicutes that affect man: Mycoplasma pneumoniae, Mycoplasma hominis and Ureaplasma species (Clinical Laboratories and Standards Institute 2011). This details methods for MIC testing and details standard strains for the three species; however, these organism growth requirements are different from those for M. bovis. However, some antimicrobials also have additional properties, such as anti-inflammatory and immunomodulatory effects (Fischer and others 2013), as well as some possibly unknown attributes that may aid the animal's recovery from infection. Sweeney and others (2012) indicate that tulathromycin is effective in vivo even with MIC levels at greater than 16 μg/ml and other antimicrobials with high MIC levels may also be effective in vivo. Unlike other bacteria, breakpoints, an MIC level at which a bacterium is deemed to be susceptible or resistant, have not been determined for Mycoplasma species, therefore a precise definition of antimicrobial resistance is not possible based on MIC data. There is, therefore, a need to develop routine methods to determine whether veterinary Mycoplasma isolates are resistant to antimicrobials. This can be achieved for some antimicrobials using molecular methods. The presence of resistance by some Mycoplasma species has already been determined. M. pneumoniae is a cause of community-acquired pneumonia in man and is often tested for macrolide resistance in hospitals where samples are examined for mutations linked to macrolide resistance at point 2063 or 2064 (2058 or 2059 by Escherichia coli numbering) in the 23S rRNA (rrl) gene (Bébéar and others 2011). Use of real-time PCR assays that can detect these mutations in clinical isolates can help rapidly identify both susceptible and resistant genotypes, allowing clinicians to select appropriate treatment options more rapidly (Wolff and others 2008).

Shabat and others (2010) described a real-time PCR for the rapid detection of fluoroquinolone resistance by M. bovis based on a fragment of the parC quinolone resistance-determining region, but this is not in routine use. Earlier work by Lysnyansky and others (2009) showed a change in the parC gene at position 84, which occurred at an MIC of 2.5 μg/ml, thereby indicating that 2.5 μg/ml is a possible MIC level for identifying fluoroquinolone resistance. It is also known that some Mycoplasma species have active efflux systems that are also implicated in resistance (Raherison and others 2002), therefore testing for all possible resistance mechanisms would be demanding.

Nevertheless, from the data obtained here, the MIC values for the majority of antimicrobials are high and with the exception of the lincosamides appear to have increased during a five-year period. Of concern is the number of isolates that have MIC values of more than 8 μg/ml to the fluoroquinolones, whereas all isolates tested in 2004 and 2005 had values of 0.5 μg/ml or less. This significant increasing trend is likely to indicate that resistance is developing, or has developed, and over a relatively short period. The observation, although based on small numbers, that isolates from mastitis cases often have higher MIC values could suggest that older animals have been exposed to more antimicrobials. They may, therefore, have had more opportunities for the organism to develop resistance, although it could be that mastitis is harder to treat and organisms can survive at sub-MIC levels. However, fluoroquinolones have been reported to be distributed in the milk often at higher concentrations than in plasma (Kaartinen and others 1995). It is, however, encouraging that sensitivity levels to the lincosamides appear to have decreased, possibly indicating that the trend towards higher MIC values and the development of antimicrobial resistance is not inevitable.

This report covers a relatively small number of isolates from 2004 to 2009, with few isolates from arthritis and mastitis cases. This is not ideal for obtaining accurate data on antimicrobial sensitivities, but even with this limited number of samples trends are apparent, although not statistically significant, but are reported here.

Farms should have adequate cattle housing with good air circulation to reduce environmental stress and decrease the risks of developing respiratory disease. However, farms can remain free of M. bovis infection by ensuring that any stock bought in does not infect other animals. Ideally new stock should be tested for M. bovis as used when restocking Moorepark in Ireland (O'Farrell and others 2001). Once infected with M. bovis, correct diagnosis and early treatment with effective antimicrobials provide the best options for keeping the infection at a manageable level. Until alternative treatments such as vaccines are commercially available to help control M. bovis infections, antimicrobials remain the only available treatment. Therefore, prudent and specifically targeted use of antimicrobials that are likely to be effective is required.

This study has demonstrated that the MIC levels for many M. bovis isolates to most antimicrobials that may be used to treat M. bovis infections are high. Some isolates now show high MIC levels to the fluoroquinolones, which was not seen before 2007, thus indicating the ability of M. bovis to develop antimicrobial resistance. Successful treatment of M. bovis infections has always been difficult, but the in vitro data reported here would indicate that the choice of effective antimicrobials is becoming even more limited. Further work should be carried out to determine breakpoints for Mycoplasma species based on molecular mutations so that in vitro information can be used to provide advice on the selection of the best antimicrobial to use in vivo. Further monitoring of antimicrobial susceptibility by M. bovis should be continued until alternative disease treatment and prevention strategies such as vaccines are commercially available.


The authors would like to thank Defra for their continued support of the Mycoplasma Group at AHVLA (Weybridge) and Rachael Jinks (AHVLA) for the statistical analysis.


View Abstract


  • Provenance: not commissioned; externally peer reviewed

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.