AAC Accepts, published online ahead of print on 23 June 2014 Antimicrob. Agents Chemother. doi:10.1128/AAC.00098-14 Copyright © 2014, American Society for Microbiology. All Rights Reserved.
1 2 3 4
Defining Daptomycin Resistance Prevention Exposures in Vancomycin Resistant Enterococcus (VRE) faecium and E. faecalis.
5
Murray4,5, W.E. Rose7, G. Sakoulas8, J. Pogliano9, C.A. Arias4,5,6, M. J. Rybak*1,2
B. J. Werth†1, M. E. Steed3, C. E. Ireland1, T.T. Tran4, P. Nonejuie9, B.E.
6 7
Anti-Infective Research Laboratory, Eugene Applebaum College of Pharmacy
8
and Health Sciences,1 and School of Medicine,2 Wayne State University, Detroit,
9
MI 48201 University of Kansas School of Pharmacy, Lawrence, KS.,3 Department
10
of Internal Medicine, Division of Infectious Diseases4 and Department of
11
Microbiology and Molecular Genetics5, University of Texas Medical School.,
12
Houston, TX., Molecular Genetics and Antimicrobial Unit, Universidad El Bosque,
13
Bogota, Colombia, 6 University of Wisconsin-Madison School of Pharmacy,
14
Madison, WI.7 University of California San Diego School of Medicine, La Jolla,
15
CA.8 University of California San Diego, Division of Biology, La Jolla, CA.9
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
† Present address: University of Washington School of Pharmacy, Department of Pharmacy, Seattle, WA, USA
31
Daptomycin Resistance Prevention Exposures in VRE
32
Keywords: daptomycin resistant enterococci, VRE, PK/PD, MPC, endocarditis
*Corresponding Author:
Michael J. Rybak Anti-Infective Research Laboratory Pharmacy Practice - 4148 Eugene Applebaum College of Pharmacy and Health Sciences Wayne State University 259 Mack Ave. Detroit, MI 48201 Tel: (313) 577-4376 Fax: (313) 577-8915 E-mail:
[email protected] 1
33 34 35
Abstract:
36
dosing targets for resistance prevention remain undefined. Doses of 4-6
37
mg/kg/day approved for staphylococci are likely inadequate against enterococci
38
due to reduced susceptibility. We modeled daptomycin regimens in vitro to
39
determine the minimum exposure to prevent daptomycin resistance (DAP-R) in
40
enterococci. Methods: Daptomycin simulations of 4-12 mg/kg/d (Cmax 57.8, 93.9,
41
123.3, 141.1 and 183.7 mg/L, t1/2 8h) were tested against 1 Enterococcus
42
faecium (S447) and 1 Enterococcus faecalis (S613) in a simulated endocardial
43
vegetation pharmacokinetic/pharmacodynamic model over 14d. Samples were
44
plated on media containing 3x the MIC of daptomycin to detect DAP-R.
45
Mutations in genes encoding proteins associated with cell-envelope homeostasis
46
(yycFG and liaFSR) and phospholipid metabolism (cardiolipin synthase [cls] and
47
cyclopropane fatty acid synthetase [cfa]) were investigated in DAP-R derivatives.
48
DAP-R derivatives were assessed for changes in susceptibility, surface charge,
49
membrane depolarization, cell wall thickness (CWT), and growth rate. Results:
50
S447 and S613 developed DAP-R after simulations of 4-8 mg/kg/day but not 10-
51
12 mg/kg/day. MICs for DAP-R strains ranged from 8-256 mg/L. Some S613
52
derivatives developed mutations in liaF or cls. S447 derivatives lacked mutations
53
in these genes. DAP-R derivatives from both strains exhibited slowed growth rate,
54
up to a 72% reduction in daptomycin-induced depolarization and up to 6nm
55
increases in CWT (P 4 mg/L.
76
Most experts agree that the daptomycin doses of 4-6 mg/kg/day used for
77
staphylococcal infections are likely inadequate for VRE infections due to reduced
78
susceptibility observed in enterococci.(3-5) Daptomycin MIC50 and MIC90 are 2
79
and 4 mg/L for E. faecium, and 0.5 and 1 mg/L for Enterococcus faecalis vs. 0.25
80
and 0.5 mg/L for staphylococci.(6) Reports of daptomycin resistance are now
81
becoming more common emphasizing the need for improved daptomycin dosing
82
paradigms to help preserve the utility of this drug against these difficult to treat
83
pathogens.(7-10) Due to the substantial morbidity and mortality associated with
84
enterococcal infections, it is critical to identify characteristics that correlate with
Enterococcus spp. are among the leading causes of nosocomial-acquired
4
85
treatment failure and therefore improve antimicrobial optimization and patient
86
outcomes.
87 88
In order to derive daptomycin exposure targets for resistance prevention, we
89
evaluated simulated daptomycin regimens from 4-12 mg/kg/day in a 14-day
90
PK/PD model of simulated endocardial vegetations (SEV) against 2 daptomycin
91
susceptible clinical VRE strains (E. faecium S447 and E. faecalis S613).
92
Resistant strains derived from these models were then subjected to a variety of
93
phenotypic and genotypic analyses to evaluate characteristics associated with
94
development of daptomycin resistance.
95 96 97
Materials and Methods.
98
Bacterial strains.
99
Two daptomycin-susceptible and vancomycin-resistant enterococci strains, E.
100
faecalis (S613) and E. faecium (S447) recovered from the bloodstream and urine,
101
respectively, of patients before daptomycin therapy, were evaluated.(11, 12) Both
102
isolates developed daptomycin-resistant derivatives in vivo during daptomycin
103
therapy(11, 12) and have been previously characterized. Additionally, the draft
104
sequences of both genomes are available.(11, 12)
105 106
Antimicrobials and media.
107
Daptomycin was commercially purchased (Cubist Pharmaceuticals, Inc.,
108
Lexington, MA). Mueller-Hinton broth II (MHB; Difco, Detroit, MI) with 50 mg/L of
5
109
calcium and 12.5 mg/L magnesium was used for susceptibility testing. Due to the
110
dependency of daptomycin on calcium for antimicrobial activity and protein
111
binding of calcium, MHB supplemented to 75 mg/L of calcium was used for in
112
vitro SEV model experiments.(13) Colony counts were determined using brain
113
heart infusion agar (BHIA; Difco, Detroit, MI). Emergence of resistance was
114
assessed with BHIA supplemented with 50 mg/L of calcium and 3x MIC of
115
daptomycin of each original strain, E. faecalis S613 and E. faecium S447.
116 117
Susceptibility testing.
118
Daptomycin MICs were determined in duplicate by broth microdilution according
119
to the Clinical and Laboratory Standards Institute (CLSI) guidelines.(14) Any
120
colonies growing on resistance screening plates were evaluated for changes in
121
susceptibility by broth microdilution.
122 123
In vitro PK/PD model
124
A well-characterized in vitro PK/PD model of simulated endocardial vegetations
125
(SEVs) previously validated against an in vivo endocarditis model with
126
Staphylococcus and Enterococcus spp. and developed by our laboratory was
127
utilized.(15) The SEVs were prepared as previously described using human
128
cryoprecipitate as the source of fibrin, a platelet suspension, bovine thrombin in
129
siliconized microcentrifuge tubes and an organism suspension to reach a final
130
bacterial density of 109 CFU/g. The resultant SEVs were removed from the tubes
131
and suspended in the model via a sterile monofilament line. The model
6
132
apparatus was prefilled with growth media containing human albumin at a
133
concentration of 3.5 g/dL. Fresh medium was continuously added and removed
134
from the compartment along with the drug via a peristaltic pump set to simulate
135
the half-life of daptomycin. Daptomycin was administered once-daily via an
136
injection port. The following daptomycin exposures with a targeted half life (T½) of
137
8h were evaluated: 4 mg/kg q 24 h (peak 57.8 mg/L)(16), 6 mg/kg q 24 h (peak
138
93.9 mg/L)(17), 8 mg/kg q 24 h (peak 123.3 mg/L)(17), 10 mg/kg q 24 h (peak
139
141.1 mg/L)(17), 12 mg/kg q 24 h (peak 183.7 mg/L)(17) and a drug-free growth
140
control with media clearance set to mimic an 8 hour T1/2 similar to drug
141
experiments. All models were performed in duplicate.
142 143
Pharmacokinetic analysis
144
Pharmacokinetic samples were obtained and stored as previously described.(18)
145
Concentrations of daptomycin were determined using a validated high-
146
performance liquid chromatography assay that conforms to the guidelines set
147
forth by the College of American Pathologists.(17) This assay has demonstrated
148
an interday coefficient of variation between 0.6 and 7.3% for all standards. The
149
daptomycin peak concentrations, T½, and the 24-hour area under the
150
concentration time curve (AUC0-24h) were determined using PK Analyst software
151
(Version 1.10, MicroMath Scientific Software, Salt Lake City, UT). The
152
trapezoidal method was utilized to calculate AUC0-24h.
153 154
Pharmacodynamic analysis
7
155
Two SEVs were removed from each model at 0, 4, 8, 24, 32, 48, 56, 72, 96, 120,
156
144, 168, 192, 216, 240, 264, 288, 312, and 336 h. All models were run in
157
duplicate. The SEVs were homogenized and diluted in cold saline to be plated on
158
BHIA plates. For all samples, antimicrobial carryover was accounted for by serial
159
dilution of the plated samples. If the anticipated dilution was near the MIC,
160
samples were processed via vacuum filtration and washed through a 0.45 µm
161
filter (Pall Corporation, Ann Arbor, MI) with normal saline to remove any
162
daptomycin. The limit of detection for determination of colony counts was 1
163
log10CFU/g. Plates were incubated at 35°C for 24 h, and the viable CFU were
164
enumerated. Changes in log10CFU/g were plotted against time to construct
165
curves to describe the antibacterial activity of the simulated regimens.
166
Bactericidal activity (99.9% kill) was defined as a ≥ 3-log10 CFU/g reduction in
167
colony count from the initial inoculum. Bacteriostatic activity was defined as a
4mg/L were saved in glycerol tubes at -80oC
181
for further phenotypic and genetic workup.
182 183
Characterization of daptomycin resistant derivatives
184
Cationic peptide and daptomycin binding determination.
185
When daptomycin is bound with elemental calcium the daptomycin-calcium
186
complex is positively charged and its degree of binding with the cell membrane
187
has been postulated to be dependent on the attractive forces between the drug
188
and the cell surface.(19) Cytochrome C is a cationic peptide that has been used
189
as a surrogate marker for changes in surface charge. A cytochrome C binding
190
study was performed as described by Kraus et al.(20) Resistant derivatives from
191
each model were evaluated for changes in cytochrome C binding and compared
192
to the unexposed parent strain.
193 194
Binding of fluorescent daptomycin.
195
Bacteria were grown to an OD600 of 0.6, and then incubated with 16 µg/ml
196
daptomycin– boron-dipyrromethene (bodipy) for 10 min, washed three times in
197
medium to remove unincorporated label, stained with 2 µg/ml DAPI (4,6-
198
diamidino-2-phenylindole), and placed on a 1.2% agarose pad for imaging in an
199
Applied Precision deconvolution fluorescence microscope as described
9
200
previously.(21) For quantitation of daptomycin-bodipy fluorescence, images from
201
each sample were collected using identical camera exposures. Cell Profiler was
202
used to measure the average fluorescence intensity of individual pixels for >200
203
cells for the parent and resistant strains. The average fluorescence intensity of
204
individual pixels for the background was also measured and subtracted from the
205
cells to generate an accurate measurement of daptomycin-bodipy binding.
206
Daptomycin binding to the resistant strains was compared to the binding
207
observed with the parent strains.
208 209
Binding of fluorescent cathelicidin LL-37.
210
Bacteria were grown in LB broth to OD600nm 0.5-0.6, diluted 1:1 in phenol red-
211
free RPMI (Invitrogen) in a final volume of 200 μL, with TAMRA-labeled LL37
212
(American Peptide, Sunnyvale, CA) to a final concentration of 1 μM for E.
213
faecium and 32 μM for E. faecalis, and incubated for 45 minutes at 37°C with
214
shaking. These concentrations were chosen based on previous LL37
215
susceptibility data in order to maintain bacterial integrity for microscopy. Bacteria
216
were pelleted, washed 3 times with RPMI, counterstained with DAPI 2 µg/ml in
217
the final wash, and visualized using a Delta Vision Deconvolution microscope
218
(Applied Precision, Inc. Issaquah, WA).
219 220
Determination of cytoplasmic membrane depolarization.
221
The ability of daptomycin to depolarize the cytoplasmic membrane of parent and
222
resistant strains was carried out as previously described using a kinetic
10
223
spectrofluorometric assay.(22) Upon membrane potential dissipation from
224
daptomycin or another cationic antimicrobial peptide, DiSC3 is released from the
225
cell membrane into the surrounding medium causing an increase in the
226
fluorescence intensity at a wavelength of 670 nm. Daptomycin concentrations of
227
11.3 mg/L (correlating with an average free peak of a 10 mg/kg/day dose) were
228
utilized for each experiment. The membrane depolarization profiles over 1 hour
229
of the resistant strains and the parent strains were determined and compared.
230
Nisin was used as a positive control to demonstrate that the assay was
231
functioning properly.
232 233
Cell wall thickness
234
Cell wall thickness was determined by transmission electron microscopy (TEM)
235
as described previously.(23) Ultrathin sections were evaluated at a magnification
236
of 48,000X with a JEOL 100CX electron microscope, and images were captured
237
with a MegaView III side-mounted digital camera. The resulting images were
238
analyzed using ImageJ 1.39t software. Cell wall thickness was determined for at
239
least 15 cells per sample using four separate quadrants of each cell for a total of
240
≥100 measurements per sample.
241 242
Growth kinetics
243
Bacterial growth kinetics of daptomycin resistant strains derived from the final
244
time points of each model were evaluated and compared to the parent strains.
245
Briefly, using a 96 well, flat bottom plate, 200 µL of MHB was inoculated with 106
11
246
CFU/mL of each test organism in triplicate. The plate was incubated in a
247
spectrophotometer at 37o C for 8 hours and the optical density at 600nm (OD600)
248
was automatically mixed and measured every 15 minutes. Turbidity
249
measurements were plotted over time to construct growth curves. Daptomycin
250
resistant strains were compared to unexposed parent strains.
251 252
Mutation screening by PCR
253
Mutations in genes previously associated with daptomycin resistance were
254
evaluated using Sanger sequencing of PCR products obtained from the entire
255
open reading frame. The following genes were investigated: i) yycFG encoding a
256
predicted two-component regulatory system involved in cell-wall homeostasis
257
(yycF and yycG encode the putative response regulator and histidine kinase of
258
the system, respectively)(12); ii) liaFSR, encoding a putative three-component
259
regulatory system that is part of the cell envelope response to antibiotics and
260
antimicrobial peptides (11, 24) and iii) cls and cfa, encoding two enzymes
261
(cardiolipin synthase and cyclopropane fatty acid synthase, respectively) that are
262
likely to be involved in cell membrane phospholipid metabolism and have been
263
previously associated with daptomycin resistance in E. faecalis (11) and E.
264
faecium (11, 12, 25). A non-synonymous mutation was defined as a nucleotide
265
change that resulted in an amino acid substitution not present in daptomycin-
266
susceptible parental enterococci (S447 and S613), whose genomes are
267
sequenced and publicly available. E. faecalis V583 (26) and E. faecium DO (27)
12
268
(closed genomes) were used as template for sequence assembly. Primers
269
utilized to amplify the genes above are listed in Table 3.
270 271
Statistical analysis
272
Changes in CFU/g, cytochrome C binding, daptomycin-bodipy binding, TAMRA-
273
LL-37 binding, total daptomycin-induced depolarization and growth curve AUCs
274
were compared by one-way analysis of variance with Tukey's posthoc test or by
275
t-test as appropriate. A P value of ≤0.05 was considered significant. All statistical
276
analyses were performed using SPSS statistical software (release 21.0; SPSS,
277
Inc., Chicago, IL).
278 279 280
Results
281
Susceptibility testing.
282
Daptomycin MICs for S447 and S613 were 2 and 1 mg/L respectively. MICs of
283
daptomycin-resistant derivatives recovered from the in vitro models ranged from
284
8-256 mg/L.
285 286
In vitro PK/PD model
287
Observed PK parameters and achieved PK/PD ratios are summarized in table 1.
288
Quantitative changes in log10 CFU/g for each regimen are illustrated in figures 1
289
and 2 for each organism. Against E. faecium S447, all regimens were
290
bactericidal within the first 24 hours but only 10 and 12mg/kg/day exposures
13
291
remained bactericidal for the duration of the models. These exposures produced
292
similar reductions in bacterial densities (P=0.55) by 336h but offered significantly
293
greater CFU/g reductions compared to 4-8mg/kg/day regimens (p=0.0005).
294
Against E. faecalis S613, 6-12mg/kg/day regimens were bactericidal within the
295
first 24 hours but only 12mg/kg/day exposures remained bactericidal for the
296
duration of the model. This regimen maintained statistically superior CFU/g
297
reductions compared to 4-10mg/kg/day regimens by 336h (P=0.0005).
298
Daptomycin resistance, defined as recovery of isolates with an MIC > 4mg/L,
299
emerged in both strains from models simulating 4, 6, and 8 mg/kg/day regimens
300
but not during simulated exposures of 10 and 12 mg/kg/day. Resistant isolates
301
were observed as early as 120 h and 144 h from the E. faecalis S613 and the E.
302
faecium S447, respectively. The peak/MIC and AUC0-24h/MIC ratios observed
303
with the 10 mg/kg/day simulations were 72 and 781 for S447 and 144 and 1562
304
for S613, respectively.
305 306
Characterization of daptomycin resistant derivatives
307
Cationic peptide and daptomycin-binding determination.
308
Cytochrome C binding studies revealed a significant reduction in the bound
309
fraction of cytochrome C in daptomycin-resistant isolates derived from our
310
models. On average, E. faecalis S613-derived resistant strains bound 42%
311
(median) less cytochrome C than the parent strain, with a range from 2%-94%
312
(P=0.01). E. faecium S447-derived resistant strains bound 28% (median) less
14
313
cytochrome C than the parent strain with a range of 13%-40%. This difference
314
was significant for some of the S447 derived resistant strains tested but not all.
315 316
Binding of fluorescent daptomycin.
317
Two representative daptomycin resistant derivatives were chosen from the SEV
318
model for microscopy experiments: i) a derivative of E. faecalis S613 that was
319
obtained on day 10 at a simulated dose of 4 mg/kg (E. faecalis S613D4 harboring
320
an insertion of isoleucine in position 177 of the putative LiaF protein, MIC 16
321
µg/ml), and ii) a derivative of E. faecium S447, obtained on day 14 at a simulated
322
dose of 4 mg/kg (E. faecium S447D4, MIC 128 µg/ml with no mutations in target
323
genes). Results of daptomycin-bodipy binding for selected parent and derivative
324
strains are depicted in figure 3. Mean fluorescent intensity of bound daptomycin-
325
bodipy at 16 µg/ml was 2.3x higher for E. faecalis S613 than for E. faecalis
326
S613D4 (671.29 ± 11.80 vs 282.16 ± 19.40; P < 0.0001). Similarly, mean
327
fluorescent intensity of bound daptomycin-bodipy for E. faecium S447 cells was
328
334.83 ± 5.65 but was below the level of detection for E. faecium S447D4, (±
329
represents standard error).
330 331
Binding of fluorescent LL-37.
332
Changes in fluorescent labeled LL-37 binding between E. faecalis S613 and E.
333
faecium S447 vs their daptomycin-resistant derivatives (E. faecalis S613D4 and E.
334
faecium S447D4, respectively) are illustrated in figure 4. Similar to daptomycin
335
binding, TAMRA-LL-37 was found to bind less to their daptomycin-resistant
15
336
derivatives obtained from the SEV model. Mean fluorescence intensity of bound
337
TAMRA-LL37 at 1 µM was 3.4x higher in E. faecium S447 than in E. faecium
338
S447D4 (601.99 ± 37.99 vs 177.30 ± 17.08; P < 0.0001). Similarly, mean
339
fluorescence intensity of bound TAMRA-LL37 at 32 µM in E. faecalis S613 cells
340
was 2.4x higher that those in E. faecalis S613D4 (154.35 ± 20.83 vs 65.09 ±
341
17.08; P