Microbiology (2015), 161, 2310–2315

DOI 10.1099/mic.0.000183

TEM-1-encoding small plasmids impose dissimilar fitness costs on Haemophilus influenzae and Haemophilus parainfluenzae Annette Søndergaard, Marianne Lund and Niels Nørskov-Lauritsen Correspondence

Department of Clinical Microbiology, Aarhus University Hospital, Aarhus N, Denmark

Niels Nørskov-Lauritsen [email protected]

Received 1 June 2015 Accepted 16 September 2015

Only two beta-lactamases, TEM-1 and ROB-1, have been observed in Haemophilus influenzae, while four different TEM but no ROB enzymes have been found in Haemophilus parainfluenzae. In order to investigate the mechanisms behind the dissemination of small beta-lactamaseencoding plasmids in H. influenzae and H. parainfluenzae, we assessed the fitness cost of three TEM-1- (pPN223, pA1209, pA1606), one TEM-15- (pSF3) and one ROB-1-bearing (pB1000) plasmid when expressed in either bacterial species. All plasmids were stable in H. influenzae and H. parainfluenzae except pB1000, which showed on average (sample mean) 76 % curing in H. parainfluenzae after 5 days of subculture. Competition assays between isogenic strains with and without plasmid showed no competitive disadvantage of pPN223 and pA1606 in H. influenzae, or of pA1209 in H. parainfluenzae. In contrast, pSF3 and pB1000 were associated with significant competitive disadvantages in both species. Some of the competitive disadvantages may be related to differences in plasmid copy number and mRNA expression of the beta-lactamase genes, as revealed by quantitative PCR analysis. In conclusion, plasmids encoding TEM beta-lactamases isolated from H. influenzae and H. parainfluenzae can be stably transferred between species. The fast curing of pB1000 in H. parainfluenzae observed in this study correlates to the fact that ROB-1 has never been reported for this species. TEM-1-encoding plasmids are associated with the lowest level of fitness cost, but different TEM-1 plasmids confer different levels of fitness cost on the two hosts.

INTRODUCTION Resistance to beta-lactam antimicrobial agents in Haemophilus influenzae is commonly mediated by production of beta-lactamases. Only two beta-lactamases, TEM-1 and ROB-1, have been described for H. influenzae, with TEM1 being by far the more prevalent (95 %) of the two (Farrell et al., 2005). Four different TEM beta-lactamases (TEM-1, TEM-15, TEM-34 and TEM-182) have been documented in H. parainfluenzae (Tristram et al., 2008; Garcı´a-Cobos et al., 2013). In H. influenzae, TEM-1 is primarily encoded on large integrative and conjugative elements (Leaves et al., 2000), and less commonly (14–17 %) on small plasmids (Søndergaard et al., 2012; Tristram et al., 2012; Fleury et al., 2014). Six different blaTEM-1-bearing plasmids of 4–6 kb have been characterized (Søndergaard et al., 2012; Tristram et al., Abbreviations: AMP, ampicillin; ESBL, extended spectrum betalactamase; sBHI, Brain Heart infusion broth supplemented with NAD and haemin. One supplementary table is available with the online Supplementary Material.

2310

2012). These plasmids can be separated into three groups: the two-gene plasmids encode blaTEM-1 and rep, the three-gene plasmids also encode mob/pre, and the four-gene plasmids also encode tnpR. The two-gene group is only represented by pPN223 (4304 bp), which also has a distinct backbone sequence. The three-gene group encompasses pLFS5 and pA1209 (5142 bp), while the four-gene group encompasses pLFH49, pLFH64 and pA1606 (5646 bp). Transcription of TEM-1 is regulated by two different promoters, namely Pa/Pb (pPN223 and pLFH49), and Prpt (pLFS5, pA1209, pLFH64 and pA1606). The plasmid encoding ROB-1 in H. influenzae is designated pB1000 and is 4.6 kb in size (San Millan et al., 2010). pB1000 has been detected in other Pasteurellaceae species of animal origin (San Millan et al., 2007, 2009, 2010; Tristram et al., 2010), but not in H. parainfluenzae. In H. parainfluenzae blaTEM-15 and blaTEM-34 have been found on small plasmids (3.7 kb and 5.5 kb, respectively), whereas blaTEM-1 and blaTEM-182 have been described on large integrative and conjugative elements (Tristram et al., 2008; Garcı´a-Cobos et al., 2013). Plasmid pSF3 encoding TEM-15 (an extended spectrum beta-lactamase – ESBL)

Downloaded from www.microbiologyresearch.org by 000183 G 2015 The Authors IP: 37.44.207.169 On: Tue, 17 Jan 2017 10:24:03

Printed in Great Britain

Fitness cost of small beta-lactamase plasmids

has not been fully sequenced, but TEM-15 is under control of the Prpt promoter. A plasmid bearing TEM-34 was not available for this investigation; the reported plasmid (p72322) is identical to pLFH64, except for the single amino acid substitution in TEM that renders the beta-lactamase resistant to clavulanic acid (Garcı´a-Cobos et al., 2013).

H. influenzae and H. parainfluenzae were routinely cultured in Brain Heart Infusion broth (Oxoid), supplemented with XV supplement (HTM supplement; Oxoid) (sBHI), and on chocolate agar plates (Statens Serum Institut). Inoculated agar plates were incubated at 37 uC in an atmosphere containing 5 % CO2. Broth cultures were incubated in ambient air at 37 uC with shaking (130 r.p.m.).

pB1000 has a competitive disadvantage of approximately 9 % per ten generations relative to the parental strain when transformed into H. influenzae strain Rd (San Millan et al., 2010). Fitness cost analyses of blaTEM-bearing plasmids have not, to our knowledge, been performed. To aid our understanding of the epidemiology of small beta-lactamase-bearing plasmids, we assessed stability and fitness cost, and performed expression analysis after introduction of TEM-1-, TEM-15- and ROB-1-bearing plasmids into recipient strains of H. influenzae and H. parainfluenzae.

Plasmid extraction and transformation. Plasmid DNA was

extracted using a QIAprep Spin Miniprep kit (Qiagen) according to the manufacturer’s instructions. Electrocompetent cells were prepared from H. influenzae strain Rd and H. parainfluenzae strain HK 23 and transformed with purified plasmid DNA using electroporation as described elsewhere (Ubukata et al., 2001). Transformants were selected on chocolate agar supplemented with 4 mg ampicillin (AMP) ml21. Electroporated controls with no added DNA were plated on selective agar to assess the selective quality of the agar. Selected colonies from each transformation experiments were subcultured twice on AMP-containing agar. Production of beta-lactamase was confirmed using nitrocefin dry slides (BBL DrySlide Nitrocefin; Becton, Dickinson and Company). Growth kinetics, plasmid curing and competition experiments.

METHODS Bacterial strains, culture conditions and susceptibility testing.

The characteristics of the strains and plasmids used in this study are listed in Table 1. Ampicillin-susceptible H. influenzae strain Rd (Fleischmann et al., 1995) and H. parainfluenzae strain HK 23 (CCUG 49489) (Hedegaard et al., 2001) were used as transformation recipients. Small beta-lactamase bearing plasmids were isolated from strains of H. influenzae (pPN223, pA1209 and pA1606) (Søndergaard et al., 2012) or H. parainfluenzae (pSF3) (Tristram et al., 2008). Plasmid pB1000 was kindly donated by Professor Bruno Gonzales-Zorn (San Millan et al., 2010). The blaTEM-1 plasmids used in this study represent each of the three TEM-1-bearing small plasmid types hitherto characterized in H. influenzae.

The fitness cost of each plasmid was assessed by growth kinetics, plasmid curing and competition between recipient strain and transformant. A saline suspension with an optical density of 1.0 McFarland was prepared for each strain or transformant and used for the simultaneous execution of the three assays. The assays were repeated on four separate occasions. For measurement of growth kinetics, the saline suspensions were diluted 1000 times in sBHI, and 200 ml was added to wells in triplicate in a 96-well flat bottomed plate. Absorbance at 620 nm was measured every 10 min for 24 h on a microplate spectrophotometer (Thermo Scientific Multiskan GO Microplate Spectrophotometer) set on 37 uC and 1 min of shaking every other minute. To determine the stability of the plasmids, saline suspensions of H. influenzae strain Rd and H. parainfluenzae strain HK 23

Table 1. Strains and plasmids used in this study Strain or plasmid Strains Rd KW20 HK23 (CCUG 49489) Rd/pPN223 Rd/pA1209 Rd/pA1606 Rd/pSF3 Rd/pB1000 HK23/pPN223 HK23/pA1209 HK23/pA1606 HK23/pSF3 HK23/pB1000 Plasmids pPN223 pA1209 pA1606 pSF3 pB1000

http://mic.microbiologyresearch.org

Description and characteristics

Reference

Amps H. influenzae recipient strain Amps H. parainfluenzae recipient strain Rd KW20 transformed with pPN223 Rd KW20 transformed with pA1209 Rd KW20 transformed with pA1606 Rd KW20 transformed with pSF3 Rd KW20 transformed with pB1000 HK23 transformed with pPN223 HK23 transformed with pA1209 HK23 transformed with pA1606 HK23 transformed with pSF3 HK23 transformed with pB1000

Fleischmann et al. (1995) Hedegaard et al. (2001) Søndergaard et al. (2012) Søndergaard et al. (2012) Søndergaard et al. (2012) This study This study This study This study This study This study This study

rep, blaTEM-1; isolated from H. influenzae rep, blaTEM-1, mob/pre; isolated from H. influenzae rep, blaTEM-1, mob/pre, tnpR; isolated from H. influenzae rep, blaTEM-15; isolated from H. parainfluenzae mobA, mobB, mobC, blaROB-1; isolated from H. influenzae

Søndergaard et al. (2012) Søndergaard et al. (2012) Søndergaard et al. (2012) Tristram et al (2008) San Millan et al. (2010)

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.169 On: Tue, 17 Jan 2017 10:24:03

2311

A. Søndergaard and others

transformed with each plasmid were diluted 1000 times in sBHI and incubated at 37 uC with continuous shaking (130 r.p.m.). Every 24 h for 5 days, the suspension was diluted 1000 times in fresh antibioticfree sBHI, and aliquots were plated on nonselective chocolate agar. The proportion of colonies expressing beta-lactamase was assessed by replica plating of 50 colonies on chocolate agar plates containing 4 mg AMP ml21. For competition experiments, 5 ml of the saline suspension of H. influenzae strain Rd or H. parainfluenzae strain HK 23 were mixed with equal amounts of the same strain transformed with the plasmid in question in 5 ml of antibiotic-free sBHI. The mixture was incubated at 37 uC with continuous shaking (130 r.p.m.). Every 24 h for 5 days, a total of 5 ml of the mixture were diluted 1000 times in fresh sBHI, and aliquots were plated on nonselective chocolate agar. The proportion of colonies expressing beta-lactamase was assessed by replica plating of 100 colonies on chocolate agar plates containing 4 mg AMP ml21. Plasmid copy number and beta-lactamase expression. Strains

were grown in sBHI to an OD600 of 0.2–0.3 at 37 uC and 130 r.p.m. Five millilitres of suspension was centrifuged and the pellet was resuspended in 1 ml RNAprotect Bacterial reagent (Qiagen) and kept at 220 uC until further use. RNA and DNA were extracted from the suspension using a Magna Pure Compact instrument using a MagnaPure Compact Nucleic Acid Isolation kit (Large Volumen) (Roche). The samples were split into an RNA sample for preparation of cDNA, and a DNA sample for assessment of the plasmid copy number. Residual DNA was degraded from the RNA sample using a Turbo DNA free kit (Ambion/Life Technologies) and 50 % more DNase than recommended by the manufacturer in a Veriti 96-well Thermal Cycler (Applied Biosystems) for 30 min at 37 uC, and 5 min at 95 uC. cDNA was prepared with TaqMan Reverse Transcription reagents (Life Technologies) in a Veriti 96-well Thermo Cycler for 10 min at 25 uC, 30 min at 48 uC and 5 min at 95 uC. The cDNA and purified DNA were mixed with primers, probes, and TaqMan Fast Advanced Master Mix (Life Technologies) and quantitative PCR was run in triplicates in a LightCycler 480 (Roche). RNA reverse transcriptase served as a negative control. Relative gene expression and plasmid copy number were quantified with LightCycler Relative Quantification software (Roche Applied Science) and normalized to the single-copy housekeeping genes adk, fucK and mdh (H. influenzae), or adk, mdh and recA (H. parainfluenzae). The PCRs were designed to have similar amplification efficiency (primers and probes are listed in Table S1, available in the online Supplementary Material).

1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Fitness cost of small beta-lactamase-encoding plasmids in H. influenzae and H. parainfluenzae None of the plasmids appeared to depress the growth rate of H. influenzae strain Rd (Fig. 1a). For H. parainfluenzae strain HK 23, plasmids pA1209 and pPN223 did not affect the growth kinetics, while plasmids pA1606, pSF3 and pB1000 reduced both the growth rate and the maximal cell density at stationary phase assessed after 24 h (Fig. 1b). Competition assays between recipient and transformant can reveal subtle disadvantages that are not readily disclosed by simple growth curve measurements (Andersson & Hughes, 2010). Equal numbers of bacteria with and without plasmid were mixed and subcultured every 24 h for 5 days, and the proportion of bacteria carrying the plasmid at each time point was monitored. Fig. 2 shows a significant cost imposed on both H. influenzae and H. parainfluenzae by TEM-15 (pSF3)- and ROB-1 (pB1000)-bearing plasmids: after only 24 h, H. parainfluenzae strain HK 23 harbouring plasmids pB1000 or pSF3 was almost completely outnumbered by the parental strain without plasmid, while the effects on H. influenzae were more protracted (Fig. 2g–j). The competitive disadvantage of plasmid pSF3 on H. parainfluenzae strain HK 23 may partly explain why TEM-15, an ESBL, has only sporadically been reported in this species (Tristram et al. 2008). However, it is a matter of concern that the fitness cost of this replicon is lower in the more pathogenic H. influenzae (Fig. 2g) than in its theoretical native host (Fig. 2h). The competitive disadvantage of plasmid pB1000 on H. influenzae strain Rd found in this study is in agreement with that reported by San Millan et al. (2010). In contrast, the three TEM-1-bearing plasmids were differentially tolerated by the two bacterial hosts: pPN223 and pA1606 did not impose a significant fitness cost to H. influenzae strain Rd, and pA1209 did not impose a significant fitness cost to H. parainfluenzae strain HK 23, while the reverse combinations resulted in significant disadvantages (Fig. 2a–f). The initial fitness

(b) Rd Rd/pPN223 Rd/pA1209 Rd/pA1606 Rd/pB1000 Rd/pSF3

OD620

OD620

(a)

RESULTS AND DISCUSSION

1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

HK23 HK23/pPN223 HK23/pA1209 HK23/pA1606 HK23/pB1000 HK23/pSF3

0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324

0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324

Time (h)

Time (h)

Fig. 1. 24 h growth curves of (a) H. influenzae strain Rd isogenic transformants, and (b) H. parainfluenzae HK 23 isogenic transformants. The curves represent the mean of three independent experiments. 2312

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.169 On: Tue, 17 Jan 2017 10:24:03

Microbiology 161

Fitness cost of small beta-lactamase plasmids

Relative percentage Relative percentage

24

48 72 Time (h)

96

Rd/pSF3 Rd

24

48 72 Time (h)

96

120

48 72 Time (h)

96

24

48

72

96

24

48 72 Time (h)

96

48 72 Time (h)

96

120

HK23/pA1606 HK23

0

120

24

100 90 80 70 60 50 40 30 20 10 0

120

HK23/pSF3 HK23

Rd/pA1209 Rd

0

(i)

0

100 90 80 70 60 50 40 30 20 10 0

120

Rd/pA1606 Rd

0

100 90 80 70 60 50 40 30 20 10 0

24

(f)

100 90 80 70 60 50 40 30 20 10 0

120 (h)

100 90 80 70 60 50 40 30 20 10 0

100 90 80 70 60 50 40 30 20 10 0

0

120

HK23/pA1209 HK

0 (j)

96

(e)

100 90 80 70 60 50 40 30 20 10 0 0

(g)

48 72 Time (h)

Relative percentage

Relative percentage

(d)

24

HK23/pPN223 HK23

Relative percentage

0

(c)

100 90 80 70 60 50 40 30 20 10 0

Relative percentage

Relative percentage

Rd/pPN223 Rd

Relative percentage

(b)

100 90 80 70 60 50 40 30 20 10 0

Relative percentage

Relative percentage

(a)

100 90 80 70 60 50 40 30 20 10 0

24

48 72 Time Time(h) (h)

96

120

Rd/pB1000 Rd

0

24

48 72 Time (h)

96

120

HK23/pB1000 HK23

0

24

48 72 Time (h)

http://mic.microbiologyresearch.org

96

120

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.169 On: Tue, 17 Jan 2017 10:24:03

2313

A. Søndergaard and others

Fig. 2. Competition profiles of H. influenzae strain Rd isogenic transformants and H. parainfluenzae strain HK 23 isogenic transformants transformed with the beta-lactamase-encoding plasmid pPN223 (a, b), pA1209 (c, d), pA1606 (e, f), pSF3 (g, h) or pB1000 (i, j). The graphs represent the sample mean of four independent experiments with the standard deviation represented by error bars.

cost imposed by pA1606 on H. influenzae was not detectable after 5 days (Fig. 2e). We did not sequence the plasmids at the end of the experiment, and mutations leading to a decrease in fitness cost cannot be excluded. However, the effect was similar in four separate experiments. Another possible mechanism of amelioration of fitness cost of newly introduced plasmids is reduction of plasmid copy number (San Millan et al., 2015).

colonies) of the HK 23/pB1000 strains had lost pB1000. pB1000 has been detected in several Pasteurellaceae species (San Millan et al., 2007, 2009, 2010; Tristram et al., 2010), but not in H. parainfluenzae. The imposed fitness cost and the relatively fast curing of pB1000 may explain the absence of ROB-1 in H. parainfluenzae. Plasmid copy number and beta-lactamase expression in relation to fitness cost of small beta-lactamase-encoding plasmids

A recent study of H. influenzae AMP-resistance in Sweden found that 16 % of 278 beta-lactamase expressing isolates carried TEM-1 on small plasmids. Of the 44 small TEM-1bearing plasmids, four were pPN223, 19 were similar to A1209, while 21 were similar to pA1606 (Fleury et al., 2014). The low fitness cost of pA1606 could be implicated in the high prevalence of this plasmid type, while the low prevalence of pPN223 could be related to its lack of the mob/pre mobilization gene. However, our fitness cost analysis does not give any clues to the high prevalence of pA1209 in H. influenzae reported in that study.

Synthesis of plasmid copies, mRNA and encoded proteins impose an energetic burden, which may lead to a fitness cost for the recipient cell (Lenski, 1998). Therefore, plasmid copy number and beta-lactamase mRNA expression were assessed for the five plasmids in H. influenzae strain Rd and H. parainfluenzae strain HK 23. Levels of beta-lactamase mRNA (normalized to the expression of three housekeeping genes) are presented in Fig. 3(a). The results show that TEM-15 and ROB-1 were expressed at a higher level than TEM-1 in both species. Variation in blaTEM promoters has previously been reported and could affect TEM expression (Tristram et al., 2007, 2012). Plasmids pA1606 and pA1209 carry the Prpt promoter, whereas pPN223 carries the Pa/Pb promoter. Variation in promoter sequences did not affect transcription levels of TEM-1 (Fig. 3a).

The plasmid curing assay measures the stability of plasmids within the host cells. This was performed by subculturing the transformants in non-selective broth every 24 h for 5 days and assessing the proportion of cells that had lost the plasmid and become susceptible to AMP. Curing was only observed for H. parainfluenzae strain HK 23 harbouring the ROB-1-encoding plasmid pB1000. After 5 days, 76+ 12 % (N53; 31–44 colonies out of 50 replicate plate

(a)

(b)

2 1 HK23_pB1000

HK23_pSF3

HK23_pA1606

HK23_pA1209

0 HK23_pPN223

HK23_pB1000

HK23_pSF3

HK23_pA1606

HK23_pA1209

HK23_pPN223

Rd_pB1000

Rd_pSF3

Rd_pA1606

Rd_pA1209

Rd_pPN223

0

3

Rd_pB1000

20

4

Rd_pSF3

40

5

Rd_pA1606

60

6

Rd_pA1209

80

7

Rd_pPN223

Beta-lactamase DNA copies

Beta-lactamase mRNA

100

Fig. 3. Quantitative PCR analysis of beta-lactamase mRNA and DNA in isogenic transformants of H. influenzae strain Rd and H. parainfluenzae strain HK 23. Data are normalized to housekeeping genes adk, mdh and fucK (H. influenzae) or adk, mdh and recA (H. parainfluenzae). (a) Average mRNA expression of the various beta-lactamases (arbitrary units); (b) Average plasmid copy number per cell. Each bar represents the mean quantitative PCR results of three (H. influenzae) or four (H. parainfluenzae) individual plasmid-transformants with the standard deviation represented by error bars. 2314

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.169 On: Tue, 17 Jan 2017 10:24:03

Microbiology 161

Fitness cost of small beta-lactamase plasmids

Plasmid quantification (normalized to single-copy housekeeping genes) showed a generally lower plasmid copy number per cell in H. parainfluenzae strain HK 23 compared to H. influenzae strain Rd (Fig. 3b). The plasmid copy number of pA1209 was much lower in both species, while pA1606, pSF3 and pB1000 were present in the highest copy numbers per cell (Fig. 3b).

Hedegaard, J., Okkels, H., Bruun, B., Kilian, M., Mortensen, K. K. & Nørskov-Lauritsen, N. (2001). Phylogeny of the genus Haemophilus

as determined by comparison of partial infB sequences. Microbiology 147, 2599–2609. Leaves, N. I., Dimopoulou, I., Hayes, I., Kerridge, S., Falla, T., Secka, O., Adegbola, R. A., Slack, M. P., Peto, T. E. & Crook, D. W. (2000).

Epidemiological studies of large resistance plasmids in Haemophilus. J Antimicrob Chemother 45, 599–604.

High beta-lactamase expression in transformants harbouring pSF3 and pB1000 could contribute to the increased fitness cost observed for these plasmids. However, the level of mRNA expression and plasmid copy number of the three TEM-1-encoding plasmids did not reflect the different profiles of fitness cost observed for the two bacterial hosts.

Lenski, R. E. (1998). Bacterial evolution and the cost of antibiotic

In conclusion, we show that small plasmids encoding TEM beta-lactamases isolated from H. influenzae and H. parainfluenzae can be stably transferred to recipient strains of the other species. TEM-1-encoding plasmids are associated with the lowest level of fitness cost in both species, but different TEM-1 plasmids confer different burden on the two hosts. The results are corroborated by the predominance of the TEM-1 beta-lactamase in H. influenzae, while more studies on the prevalence and type of small resistance plasmids in H. parainfluenzae are needed.

San Millan, A., Escudero, J. A., Gutierrez, B., Hidalgo, L., Garcia, N., Llagostera, M., Dominguez, L. & Gonzalez-Zorn, B. (2009).

resistance. Int Microbiol 1, 265–270. San Millan, A., Escudero, J. A., Catalan, A., Nieto, S., Farelo, F., Gibert, M., Moreno, M. A., Dominguez, L. & Gonzalez-Zorn, B. (2007). Beta-lactam resistance in Haemophilus parasuis is mediated

by plasmid pB1000 bearing blaROB-1. Antimicrob Agents Chemother 51, 2260–2264.

Multiresistance in Pasteurella multocida is mediated by coexistence of small plasmids. Antimicrob Agents Chemother 53, 3399–3404. San Millan, A., Garcia-Cobos, S., Escudero, J. A., Hidalgo, L., Gutierrez, B., Carrilero, L., Campos, J. & Gonzalez-Zorn, B. (2010).

Haemophilus influenzae clinical isolates with plasmid pB1000 bearing blaROB-1: fitness cost and interspecies dissemination. Antimicrob Agents Chemother 54, 1506–1511. San Millan, A., Santos-Lopez, A., Ortega-Huedo, R., Bernabe-Balas, C., Kennedy, S. P. & Gonzalez-Zorn, B. (2015). Small plasmid-mediated

antibiotic resistance in Haemophilus influenzae is enhanced by increases in plasmid copy number and bacterial fitness. Antimicrob Agents Chemother 59, 3335–3341.

ACKNOWLEDGEMENTS

Søndergaard, A., San Millan, A., Santos-Lopez, A., Nielsen, S. M., Gonzalez-Zorn, B. & Nørskov-Lauritsen, N. (2012). Molecular

Line Nymann Penstoft is acknowledged for excellent technical assistance with quantitative PCR.

organization of small plasmids bearing blaTEM-1 and conferring resistance to b-lactams in Haemophilus influenzae. Antimicrob Agents Chemother 56, 4958–4960. Tristram, S., Jacobs, M. R. & Appelbaum, P. C. (2007). Antimicrobial

resistance in Haemophilus influenzae. Clin Microbiol Rev 20, 368–389.

REFERENCES Andersson, D. I. & Hughes, D. (2010). Antibiotic resistance and its

cost: is it possible to reverse resistance? Nat Rev Microbiol 8, 260–271. Farrell, D. J., Morrissey, I., Bakker, S., Buckridge, S. & Felmingham, D. (2005). Global distribution of TEM-1 and ROB-1 beta-lactamases

in Haemophilus influenzae. J Antimicrob Chemother 56, 773–776. Fleischmann, R. D., Adams, M. D., White, O., Clayton, R. A., Kirkness, E. F., Kerlavage, A. R., Bult, C. J., Tomb, J. F., Dougherty, B. A. & other authors (1995). Whole-genome random sequencing and assembly of

Haemophilus influenzae Rd. Science 269, 496–512. Fleury, C., Resman, F., Rau, J. & Riesbeck, K. (2014). Prevalence,

distribution and transfer of small b-lactamase-containing plasmids in Swedish Haemophilus influenzae. J Antimicrob Chemother 69, 1238–1242.

Garcı´a-Cobos, S., Arroyo, M., Campos, J., Pe´rez-Va´zquez, M., Aracil, B., Cercenado, E., Orden, B., Lara, N. & Oteo, J. (2013). Novel mechanisms

of resistance to b-lactam antibiotics in Haemophilus parainfluenzae: b-lactamase-negative ampicillin resistance and inhibitor-resistant TEM b-lactamases. J Antimicrob Chemother 68, 1054–1059.

http://mic.microbiologyresearch.org

Tristram, S. G., Pitout, M. J., Forward, K., Campbell, S., Nichols, S. & Davidson, R. J. (2008). Characterization of extended-spectrum beta-

lactamase-producing isolates of Haemophilus parainfluenzae. J Antimicrob Chemother 61, 509–514. Tristram, S. G., Littlejohn, R. & Bradbury, R. S. (2010). blaROB-1

presence on pB1000 in Haemophilus influenzae is widespread, and variable cefaclor resistance is associated with altered penicillinbinding proteins. Antimicrob Agents Chemother 54, 4945–4947. Tristram, S. G., Franks, L. R. & Harvey, G. L. (2012). Sequences of

small blaTEM-encoding plasmids in Haemophilus influenzae and description of variants falsely negative for blaTEM by PCR. J Antimicrob Chemother 67, 2621–2625. Ubukata, K., Shibasaki, Y., Yamamoto, K., Chiba, N., Hasegawa, K., Takeuchi, Y., Sunakawa, K., Inoue, M. & Konno, M. (2001). Association

of amino acid substitutions in penicillin-binding protein 3 with betalactam resistance in beta-lactamase-negative ampicillin-resistant Haemophilus influenzae. Antimicrob Agents Chemother 45, 1693–1699. Edited by: R. Lan

Downloaded from www.microbiologyresearch.org by IP: 37.44.207.169 On: Tue, 17 Jan 2017 10:24:03

2315