- Email: [email protected]
Preterm infant gut colonization in the neonatal ICU and complete restoration 2 years later
Accepted Manuscript Preterm infant gut colonization in the Neonatal ICU and Complete Restoration two years later Laura Moles, Marta Gómez, Esther Jiménez, Leónides Fernández, Gerardo Bustos, Fernando Chaves, Rafael Cantón, Juan Miguel Rodríguez, Rosa del Campo PII:
S1198-743X(15)00616-3
DOI:
10.1016/j.cmi.2015.06.003
Reference:
CMI 303
To appear in:
Clinical Microbiology and Infection
Received Date: 18 January 2015 Revised Date:
12 May 2015
Accepted Date: 5 June 2015
Please cite this article as: Moles L, Gómez M, Jiménez E, Fernández L, Bustos G, Chaves F, Cantón R, Rodríguez JM, Campo Rd, Preterm infant gut colonization in the Neonatal ICU and Complete Restoration two years later, Clinical Microbiology and Infection (2015), doi: 10.1016/j.cmi.2015.06.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Preterm Infant Gut Colonization in the Neonatal ICU and
2
Complete Restoration Two Years Later
3
Laura Moles1,2, Marta Gómez1, Esther Jiménez1,3, Leónides
5
Fernández1,3, Gerardo Bustos4,5, Fernando Chaves6,7, Rafael Cantón2,7,8,
6
Juan Miguel Rodríguez1,3, Rosa del Campo2,7,8
RI PT
4
7 8
1
9
Complutense de Madrid. 2Instituto Ramón y Cajal de Investigaciones Sanitarias
10
(IRYCIS), Madrid. 3ProbiSearch, S.L., Tres Cantos, Madrid. 4Servicio de Neonatología
11
Hospital Universitario 12 de Octubre, Madrid. 5Red de Salud Materno-Infantil y del
12
Desarrollo (SAMID), Instituto Carlos III, Madrid. 6Servicio de Microbiología, Hospital
13
Universitario 12 de Octubre, Madrid. 7Servicio de Microbiología, Hospital Universitario
14
Ramón y Cajal, Madrid. 8Spanish Network for Research in Infectious Diseases (REIPI),
15
Instituto de Salud Carlos III, Madrid. SPAIN.
SC
M AN U
TE D
EP
16
Departmento de Nutrición, Bromatología y Tecnología de los Alimentos, Universidad
Correspondence author: Rosa del Campo, Servicio de Microbiología, Hospital
18
Universitario Ramón y Cajal, Ctra. Colmenar, Km 9.1, 28034 Madrid, SPAIN.
19
Email: [email protected], Phone: +34 91 336 85 42, Fax: +34 91 336 88 09.
20 21
AC C
17
Running title: Gut Microbiota of Preterm Infants
22 23
Key words: Nosocomial High-Risk Clones, Antibiotic Resistance, Virulence Factors,
24
Human Milk, Gut Microbiota, Preterm Infants.
25
ACCEPTED MANUSCRIPT ABSTRACT
27
Preterm infants in a Neonatal Intensive Care Unit (NICU) are exposed to multi-drug
28
resistant bacteria previously adapted to the hospital environment. The aim of the present
29
study was to characterize the bacterial antibiotic-resistant high-risk lineages colonizing
30
preterm infants during their NICU stay and their persistence in faeces after 2 years.
RI PT
26
A total of twenty-six preterm neonates were recruited between October 2009 and
32
June 2010 and provided 144 faecal samples. Milk samples (86 mothers’ milk, 35 human
33
donor milk, and 15 formula milk) were collected at the same time as faecal samples. An
34
additional faecal sample was recovered in 16 infants at the age of two years. Samples
35
were plated onto different selective media and one colony per morphology was selected.
36
Isolates were identified by 16S rDNA nucleotide sequence and MALDI-TOF.
37
Antibiotic susceptibility (agar dilution), genetic diversity (RAPD, PFGE and MLST)
38
and virulence factors (only in enterococcal and staphylococcal isolates) were
39
determined by PCR.
TE D
M AN U
SC
31
A high proportion of antibiotic resistant high-risk clones was detected in both
41
faecal and milk samples during the NICU admittance. Almost all infants were colonized
42
by Enterococcus faecalis ST64 and Enterococcus faecium ST18 clones, while a wider
43
genetic diversity was observed for the Gram-negative isolates. Multi-resistant high-risk
44
clones were not recovered from the faecal samples of the 2-year-old infants.
AC C
45
EP
40
In conclusion, the gut of the preterm infants admitted in NICU might be initially
46
colonized by antibiotic-resistant and virulent high-risk lineages, which are later replaced
47
by antibiotic-susceptible community ones.
48
ACCEPTED MANUSCRIPT
INTRODUCTION
50
The gut microbiota ecosystem plays an important role in human health and its
51
alterations have been implicated in several systemic diseases, particularly in the early
52
stages of life [1]. After birth, an adequate gut microbiota establishment is essential for
53
both the innate immune system maturation and the creation of a barrier protection from
54
bacterial pathogens [2,3]. Moreover, these first bacterial cells colonizing our gut mucosa
55
might persist in symbiosis with our enterocytes throughout our complete life.
RI PT
49
Bacterial gut colonization starts during the intrauterine stage, increasing
57
drastically during the delivery and the first days of life [4-6]. This process can be
58
adversely affected by numerous factors, including antibiotic treatment, caesarean
59
delivery, early separation from the mother, formula feeding, invasive medical
60
treatments and long stays in Neonatal Intensive Care Units (NICU) [7,8]. Some of these
61
factors are usually associated with premature infants who in addition have a delayed or
62
aberrant gut microbiota characterized by poor global bacterial diversity and a low and
63
rather fluctuating Bifidobacterium population [9-11].
TE D
M AN U
SC
56
During their usually long NICU stays, preterm infants are exposed to multi-drug
65
resistant bacteria previously adapted to the hospital environment, that have been
66
recognized as high-risk clones. Most of these high-risk clones belong to the ESKAPE
67
group (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae,
68
Acinetobacter baumanii, Pseudomonas aeruginosa, and Enterobacter species) [12,13].
69
Furthermore, these preterm infants are usually treated with antibiotics which also select
70
the multiresistant bacteria that might colonize their guts.
AC C
EP
64
71
In the last years, the implementations of metagenomic tools based on 16S rDNA
72
massive sequence has provide large information about the complexity of the gut
73
microbiota ecosystem [14], including the identification of uncultured microorganisms.
ACCEPTED MANUSCRIPT 74
However, classical and extended techniques based on classical microbiological cultures
75
as culturomics, also provides a comprehensive approach for the human gut microbiota
76
study, and allowed us the further characterization of the bacterial isolates [15]. Although previous studies have assessed the gut colonization of preterm infants,
78
little is known about its evolution after hospital discharge and in their first years of life,
79
particularly around the two years of life, when a mature microbiota has been already
80
established. The aim of the present study is to characterize the bacterial lineages
81
colonizing premature infants during their NICU admittance and their persistence after 2
82
years, which is the age when the infant’s gut starts to contain an adult-like microbiota
83
[16,17].
84
M AN U
SC
RI PT
77
METHODS
86
Subjects and sampling
87
Twenty-six preterm infants (≤32 weeks of gestation and/or ≤1,500 g of weight) born
88
between October 2009 and June 2010 in the 12 de Octubre University Hospital in
89
Madrid were included in the study. Preterm infants with malformations, metabolic
90
diseases or severe conditions were excluded. Relevant demographic and clinical data are
91
summarized in Table 1. The local ethic committee approved the study (reference
92
09/157), and parents signed the fully informed consent. Following the routine NICU
93
feeding protocols, all infants were preferably fed with their own mother’s milk and
94
when this was not possible, with pasteurized human milk from the Milk Bank Unit.
95
When the weight of the infants was ≥1,500 grams and their own mother’s milk was
96
unavailable, they received adapted preterm formula. The fortifier FM 85 (Nestle) was
97
used approximately from the fourth week of life with an initial dose of 1 g per 50 ml of
98
milk that was increased to 1 g per 20 ml. Globally, the feeding patterns of the recruited
AC C
EP
TE D
85
ACCEPTED MANUSCRIPT 99 100
infants were very heterogeneous, a fact that prevented the formation of well-defined feeding groups. The first spontaneously evacuated meconium after delivery and faecal samples
102
were collected weekly during the first month of life and, then, every 15 days until
103
discharged from the NICU. Samples were collected and immediately stored at -80ºC.
104
Once the infants were 2-years-old, an additional faecal sample was also collected
105
although we only were able to contact 16 out of the 26 families.
RI PT
101
Mother’s milk was extracted using electric pumps and stored either refrigerated
107
for a maximum of 24 h or frozen for up to 6 months. Donor milk, pasteurized at 62.5ºC
108
after collection has a shelf life of 3 months. All milks were incubated for 10-15 min at
109
37-40ºC and administered to the infant through an orogastric or nasogastric feeding tube
110
using a syringe (manual feeding) or a pump (automatic feeding). Milk samples were
111
always collected after feeding from either the syringe or the extension set connected
112
with the orogastric or nasogastric feeding tube, at the same sampling times as the faecal
113
ones.
TE D
M AN U
SC
106
Bacterial strains causing a bacteraemia were recovered from the blood samples
115
after their identification was performed at the Department of Microbiology at the 12 de
116
Octubre University Hospital.
AC C
117
EP
114
118
Culture analysis of the samples
119
Suitable dilutions of milk and faecal samples obtained during the NICU stay were
120
spread onto Man Rogosa Sharpe (Oxoid, Basingstoke, UK) supplemented or not with L-
121
cysteine (0.5 g/L) (Sigma, St. Louis, USA) for lactic acid bacteria isolation;
122
MacConkey (BioMérieux, Marcy l’Etoile, France) for Enterobacteriaceae; Baird Parker
123
(BioMérieux) for staphyloccoci, Brain Heart Infusion (Oxoid), Wilkins-Chalgren
ACCEPTED MANUSCRIPT 124
(Oxoid) and finally in Columbia Nalidixic Acid (BioMérieux) as general agar media
125
plates for isolation of bacterial organisms. Persistence of selected antibiotic resistant strains acquired in the NICU stay was
127
evaluated in faecal samples of the 2-year-old infants; on the one hand with selective
128
agar media supplemented or not with antibiotics such as Manitol Salt Agar for
129
staphylococci (Difco Laboratories, Detroit, USA) and M-Enterococcus agar (Difco)
130
alone and supplemented with 16 µg/ml of ampicillin or 16 µg/ml of erythromycin. On
131
the other hand MacConkey plates were supplemented with 4 µg/ml of ciprofloxacin, or
132
2 µg/ml of imipenem or 1 µg/ml of cefotaxime.
SC
RI PT
126
Plates were aerobically incubated at 37ºC for up to 48 h, with the exception of the
134
Wilkins-Chalgren and Man Rogosa Sharpe-cysteine plates that were anaerobically
135
incubated (85% nitrogen, 10% hydrogen, and 5% carbon dioxide) in an anaerobic
136
workstation (Mini-MACS Don Whitley Scientific Limited, Shipley, UK) at 37ºC for 48
137
h. Bacterial counts were recorded as the colony forming units (CFU) per gram of
138
meconium or faeces, or CFU per ml of milk, and transformed to log10 values.
TE D
M AN U
133
139
Bacterial identification and genotyping
141
After the initial samples processing, ~4000 colonies were selected including at least one
142
colony per morphology, media and sample. Isolates were subcultured in Brain Heart
143
Infusion and Man Rogosa Sharpe media. In order to avoid repeated isolates, those with
144
the same morphology from the same media and sample were submitted to randomly
145
amplified polymorphic DNA polymerase chain reactions (RAPD-PCR) using the primer
146
OPL5 (5’-ACGCAGGCAC-3’) [18]. Finally, at least one representative isolate from
147
each sample, media, and infant was further selected. Bacterial identification was
148
performed
AC C
EP
140
by
matrix-assisted
laser
desorption/ionization
time-of-flight
mass
ACCEPTED MANUSCRIPT spectrometry (MALDI-TOF MS) (Bruker Daltonics, Bremen, Germany) combined with
150
16S rDNA PCR amplification and further nucleotide sequencing. Staphylococcal
151
species belonging to the species S. aureus, Staphylococcus epidermidis, or
152
Staphylococcus hominis were identified by a multiplex PCR method based on the dnaJ
153
genes [4].
RI PT
149
154
Genetic diversity
156
The genetic diversity of all selected isolates was assessed by pulsed field gel
157
electrophoresis (PFGE) in a CHEF DR II apparatus (Bio-Rad, Birmingham, UK). To
158
separate SmaI digested fragments of enterococci and staphylococci different protocols
159
were applied (2 to 28 s for 24 h; and 5 to 15 s for 10 h and then 15 to 60 s for 13 h,
160
respectively). The chromosomal DNA of K. pneumoniae isolates was digested with
161
XbaI enzyme and the electrophoresis conditions were 1 to 40 s for 20 h. Dendrograme
162
analysis of PFGE profiles was performed using the UPGMA method based on the Dice
163
similarity by the Phoretix 5.0 software (Nonlinear Dynamics Ltd., United Kingdom).
TE D
M AN U
SC
155
Multilocus sequence typing (MLST) schemes were applied for PFGE unrelated
165
strains of Enterococcus faecalis, E. faecium and S. aureus (www.mlst.net); Escherichia
166
coli (http://mlst.ucc.ie/mlst/dbs/Ecoli), and K. pneumoniae (http://www.pasteur.fr
167
/recherche/genopole/PF8/mlst/Kpneumoniae.html). The E. coli phylogenetic group was
168
also determined by PCR [19]. MLST was represented by Minimum Spanning Tree
169
algorithm using the Phyloviz software (http://www.phyloviz.net)
AC C
EP
164
170
171
Antimicrobial susceptibility testing
172
Antibiotic susceptibility for ampicillin, cefotaxime, imipenem, gentamicin, kanamycin,
173
erythromycin, tetracycline, ciprofloxacin, and vancomycin was determined only in the
ACCEPTED MANUSCRIPT 174
selected isolates for PFGE and MLST typing by agar dilution following the Clinical and
175
Laboratory Standards Institute guidelines and breakpoints for each genus/specie [20].
176
All antibiotics were purchased from Sigma-Aldrich Co. (http://www.sigmaaldrich.
177
com/).
RI PT
178
Virulence factors carriage
180
Presence of genes encoding for sex pheromones (ccf, cpd, cad, cob), cell wall adhesins
181
(efaAfs), the aggregation protein (agg2), the protein involved in immune evasion (epsfs)
182
presence, the extracellular metaloendopeptidase (gelE) and the cytolisine (cylA) for
183
enterococci were assessed by the procedure described by Eaton and Gasson and
184
modified by Reviriego et al., [21,22].
M AN U
SC
179
Carriage of hemolysins (hla, hlb, hlg, hld) and adhesins (fnbA, clfa, sdrE, ebpS)
186
genes in the S. aureus isolates was determined by two multiplex PCRs [23], while a
187
single PCR scheme was used for the detection of the genes encoding the exfoliative
188
toxin A (eta), toxic shock syndrome toxin (tst), adhesion for collagen (cna) [24],
189
Panton-Valentin leukocidin components S and F (pvl-S/F) [25], Panton-Valentin
190
leukocidin components M and F (pvl-M/F) [26], staphylococcal enterotoxins (se-adej)
191
[27], polysaccharide intracellular adhesion (icaA) [28], methicillin resistance (mecA)
192
[29], β-lactamase (blaZ) [30], and iron siderophore transporter (sirB) [31]. At least one
193
isolate per infant and sample obtained during the NICU admission was evaluated for
194
virulence factors carriage.
AC C
EP
TE D
185
195
196
RESULTS
197
The main demographic and clinical data of the 26 preterm infants are described in Table
198
1. In all but one infant, the ampicillin-gentamicin combination was administered at birth
ACCEPTED MANUSCRIPT as empirical treatment, whereas teicoplanin, gentamicin, amikacin and ambisome were
200
used in the septic episodes (Figure 1). Seven infants (27%) had, at least, one
201
bacteraemic episode, and the foremost species involved were coagulase-negative
202
staphylococci (n=5), Serratia marcescens (n=3), K. pneumoniae (n=2), and P.
203
aeruginosa (n=2) (Figure 1).
RI PT
199
A total of 296 samples from infant’s faeces (144 during the NICU admission and
205
16 at 2 years-old), 86 from mother’s milk, 35 from donor’s milk, and 15 from formula’s
206
milk were processed. These samples yield ~4000 bacterial isolates and the most
207
numerous species was E. faecalis (n=564), followed by S. epidermidis (n=420), S.
208
marcescens (n=267), E. coli (n=252), K. pneumoniae (n=228), E. faecium (n=108) and
209
S. aureus (n=60) (Table 2). Other less abundant but still relevant species belonged to
210
Enterobacter (n=71), Pseudomonas (n=22), Acinetobacter genera (n=16), and others
211
(Suplementary Figure 1).
M AN U
SC
204
The number of strains with different PFGE profiles coexisting in the same infant
213
during NICU admittance oscillated between 1 and 5 for E. faecalis, 1 and 4 for E.
214
faecium, 1 and 10 for S. aureus, and 1 and 3 in K. pneumoniae. The final collection of
215
unrelated isolates was composed by 67 E. faecalis, 48 E. coli, 43 K. pneumoniae, 22 E.
216
faecium, and 17 S. aureus isolates, which were submitted to antibiotic susceptibility
217
testing and MLST typing. At this stage, S. epidermidis isolates were not studied further
218
due to their high number and the high genetic diversity observed; Serratia isolates were
219
excluded from the analysis due to the lack of an MLST scheme, while Acinetobacter,
220
Pseudomonas and Enterobacter isolates were scarcely represented in our population.
AC C
EP
TE D
212
221
The proportion of the high-risk clones detected in the faecal samples during the
222
NICU admittance was considerable: 5 (ST16-CC8, ST21-CC21, ST40-CC40, ST56-
223
CC30 and ST64-CC8) of the 8 STs found among the E. faecalis isolates; 2 (ST18-CC18
ACCEPTED MANUSCRIPT and ST132-CC18) of 3 in E. faecium; 6 (ST8-CC8, ST30-CC30, ST34-CC30, ST45-
225
CC45, ST121-CC121, and ST1490-CC8) of 7 in S. aureus, 5 (ST58, ST59, ST69,
226
ST393, ST2904) of 10 in E. coli, and finally 5 (ST25, ST29, ST35, ST45, and ST1547,
227
all of them fit into the CC292) of 12 in K. pneumoniae. Isolates belonging to these high-
228
risk clones exhibited resistance to most of the studied antibiotics and carried multiple
229
virulent factors (Figure 2).
RI PT
224
It is of note that identical STs with the same antibiotic resistance phenotype and
231
the same virulence factors were also detected in the milk and formula samples. In infant
232
14, E. faecalis ST56 and K. pneumoniae ST505 were simultaneously detected in faeces
233
and in blood samples during a bacteraemic episode. Infant 12 died in the NICU after a
234
complicated E. coli bacteraemia.
M AN U
SC
230
Almost all infants were colonized by both E. faecalis ST64 (19/24) and E.
236
faecium ST18 (9/12) clones, while a wider genetic diversity was observed for the Gram-
237
negative isolates. The predominant clones were ST393 for E. coli (8/20) and ST641 for
238
K. pneumoniae (8/16) (Figure 3).
TE D
235
Among the E. faecalis and E. faecium isolates, a high level of antibiotic resistance
240
was observed, especially for tetracycline (96.7% and 97.4%) and erythromycin (68.8%
241
and 92.9%); whereas ampicillin resistance was only detected in E. faecium isolates
242
(92.2%). E. faecalis ST16 and ST40 clones harbored the highest number of virulence
243
determinants. The most common virulent genes were gelE (84.3%), ccf (100%) and
244
efaAfs (95.7%) in E. faecalis isolates, and epsfs (24.8%), gelE (10.3%) and ccf (10.3%) in
245
E. faecium isolates (Figure 2).
AC C
EP
239
246
Globally, hld, fnbA, clfA, and icaA were the most frequent virulence genes in the
247
S. aureus isolates which were susceptible to the evaluated antibiotics, although 43% of
248
them exhibited resistance to gentamicin (Figure 2).
ACCEPTED MANUSCRIPT For the E. coli isolates, high percentages of tetracycline resistance (46%), and
250
ciprofloxacin (32%) were observed (Figure 2). As to K. pneumoniae, 21% of the
251
isolates were resistant to tetracycline while those belonging to the predominant K.
252
pneumoniae ST641 subgroup displayed resistance to ciprofloxacin, imipenem and
253
cefotaxime (Figure 2).
RI PT
249
In contrast, isolates from the faecal samples collected from the 2-year-old infants
255
were susceptible to the antibiotics and the proportion of high–risk clones was reduced:
256
3 (ST16-CC8, ST40-CC40, and ST64-CC8) of 7 in E. faecalis; none for E. faecium and
257
S. aureus, 5 (ST10-CC10, ST57, ST59, ST69, ST131, and ST3570-CC23) of 10 in E.
258
coli, and finally 1 (ST1486-CC292) of 2 in K. pneumoniae (Figure 2).
M AN U
SC
254
259
DISCUSSION
261
Recent studies have demonstrated that bacterial colonization starts during pregnancy
262
[4,32,33], and in fact the placenta harbours its own microbiota [34]. According to
263
general agreement, the first gut bacterial colonization constitutes the core of the future
264
microbiota that can influence the host’s gastrointestinal system and general health
265
throughout life. The definitive gut microbiota is usually established at the age of two
266
years, when the complete diet has introduced.
EP
AC C
267
TE D
260
In the present work, we analyzed the gut microbiota of the preterm infants during
268
their hospital admittance and its maintenance overtime. In contrast to the next
269
generation massive sequence tools, we preferred the classical techniques based on
270
microbiological culture, in order to collect the bacterial strains for further
271
characterization. A high proportion of high-risk clones linked to nosocomial infection
272
were detected both in the milk administered to the infants and in their faeces. As the
273
different milks had been processed and properly stored, the bacterial load could not be
ACCEPTED MANUSCRIPT attributed to a dramatic and continuous food contamination. In the NICU where the
275
preterm infants were recruited, single-use sterilized and ready-to-use preterm formula
276
are purchased while donor’s milk is pasteurized and, then, submitted to strict
277
microbiological controls. In relation to mother’s milk that has not been pasteurized, it
278
has been described that contains its own microbiota [35, 36], but its composition and
279
concentration differ considerably from the results obtained in this study. Milk sampling
280
process represents a limitation of our study, because we only collected the milk after the
281
tube or syringe pass, and we cannot ruled out that the milk might be also contaminated
282
previously.
SC
RI PT
274
The orogastric or nasogastric feeding tubes, irrespective of the feeding regime,
284
can act as locus for the bacterial attachment and multiplication of numerous
285
opportunistic pathogens [37, 38], and they constitute an important risk factor in the
286
neonatal infections [39]. In agreement with our results, the feeding system might act as
287
a reservoir for the hospital adapted high-risk bacterial clones and contaminate the
288
human and formula milk samples during their passage. After this study, the feeding
289
systems protocol in preterm infants of our NICU was subsequently modified and
290
nowadays the enteral tubes are replaced in a maximum of 48 hours. Other relevant
291
limitation was the lack of maternal samples to complete discharged the existence of
292
these multirresistant clones in their gut or skin microbiota. We hypothesized that the
293
clones have a nosocomial origin, although the maternal source also represents an
294
important option.
AC C
EP
TE D
M AN U
283
295
On the other hand, sepsis is one of the main causes of death in the NICU, and its
296
prevention and treatment considerably increase the sanitary costs [40]. The main
297
bacteria causing outbreaks in the NICU is S. epidermidis, followed by K. pneumoniae,
298
Serratia spp., E. coli and Enterococcus spp. Among them, hospital-adapted high-risk
ACCEPTED MANUSCRIPT clones that exhibit antibiotic resistance and contain virulence factors are of the
300
uppermost relevance [41,42]. In this study, the structure and main characteristics of the
301
E. faecalis, E. faecium, S. aureus, E. coli and K. pneumoniae populations were
302
investigated since they were considered to be the most representative species in the
303
analyzed samples.
RI PT
299
Epidemiological molecular studies based on MLST schemes have demonstrated
305
that the nosocomial pathogens usually grouped in particular lineages as CC10 in E. coli,
306
ST258 in K. pneumoniae [43], CC2/CC9 in E. faecalis, CC17 in E. faecium, and
307
CC5/CC8/CC30/CC45 in S. aureus [44]. In our study the proportion of these high-risk
308
clones detected in the faecal and milk samples was high, although the clonal
309
replacement observed (Figure 2) prevented an epidemic situation in the NICU (Figure
310
3).
M AN U
SC
304
Despite the elevated amount of enterococci isolates obtained during the NICU
312
stay (564 E. faecalis and 108 E. faecium), a low genetic diversity was observed (8 E.
313
faecalis and 3 E. faecium STs), whereas the remaining genera showed a higher
314
variability (7 STs of S. aureus from 7 infants, 10 STs of E. coli and 12 STs of K.
315
pneumoniae). Only E. faecalis ST64 and ST40 clones persisted during almost all the
316
studied period (Figure 3) which signified that the epidemiological measures implanted
317
routinely in this NICU are adequate including hand-washing protocols, aseptic
318
conditions for milk manipulation, early detection and further isolation of antibiotic
319
multi-resistant carriers, segregation of patients and staff during epidemic outbreaks, and
320
adequate antibiotic politics in consensus with the Microbiology Department.
AC C
EP
TE D
311
321
All but one infant received empirical antibiotic therapy after birth. Although the
322
prophylactic use of antibiotics immediately after birth usually exerts a negative
323
influence on early gut colonization and its protective effect has not been demonstrated
ACCEPTED MANUSCRIPT [43], a correct antibiotic therapy may be decisive to assure the survival of many preterm
325
infants. However, it is obvious that this routine practice may be a selector for antibiotic
326
resistant isolates that, subsequently, could be overrepresented, on a long-term basis, in
327
the host’s gut [44]. Therefore, persistence of the high-risk antibiotic-resistant clones
328
acquired during the NICU stay was evaluated in the same cohort two years after birth.
329
Interestingly, the initial antibiotic-resistant isolates and high-risk clones were scarcely
330
detected but, in turn, other clones, mainly observed in the community and initially non-
331
detected, such as E. coli ST131, ST10-CC10 or ST3570-CC23, were found.
SC
RI PT
324
Previous theories generally assume that the first bacterial isolates reaching the
333
almost “microbial-free” gut during or after birth have a founder effect and, therefore,
334
they tended to persist as members of the gut microbiota throughout life [45]. In contrast,
335
our results showed that after the important selective pressures received early in life by
336
the preterm infants, the initial nosocomial and antibiotic-resistant high-risk lineages had
337
been replaced in their gut by antibiotic-susceptible community lineages when they were
338
2-year-old, when perturbations of the early microbiota decrease and become adult-like
339
microbiota [16,17].
EP
340
TE D
M AN U
332
Author Contributions
342
Laura Moles, Leónides Fernández, Rafael Cantón, Juan Miguel Rodríguez and Rosa del
343
Campo conceptualized and designed the study, drafted the initial manuscript, and
344
approved the final manuscript as submitted. Laura Moles, Marta Gómez, and Esther
345
Jiménez carried out the samples analyses, and the experimental data collection,
346
reviewed and revised the manuscript, and approved the final manuscript as submitted.
347
Gerardo Bustos, Fernando Chaves participated in the patient’s recruitment, the sample
348
collection and coordinated and supervised clinical data collection at the 12 de Octubre
AC C
341
ACCEPTED MANUSCRIPT 349
Hospital. Also, they critically reviewed the manuscript, and approved the final
350
manuscript as submitted.
351
Acknowledgments
353
We are grateful to the infant and their families and particularly to the staff of the
354
Department of Neonatology of the 12 de Octubre University Hospital.
RI PT
352
This work was supported by the PI13/00205 project (ISCIII-FIS), and by the
356
AGL2013-41980-P project, both from the Ministry of Economy and Competitiveness
357
(MINECO, Spain), and co-financed by the European Development Regional Fund “A
358
Way to Achieve Europe,” ERDF, Spanish Network for the Research in Infectious
359
Diseases (REIPI RD12/0015). L.M. and M.G. were the recipients of predoctoral
360
contracts from MINECO.
361
M AN U
SC
355
Transparency Declaration
363
The authors declare no conflicts of interest.
EP AC C
364
TE D
362
ACCEPTED MANUSCRIPT 365
REFERENCES
366
1.
microbiome in early life: health and disease. Front Immunol 2014; 5:427.
367
2.
gut microbiota. Science 2013; 341:1237439.
369 370
Faith JJ, Guruge JL, Charbonneau M, et al. The long-term stability of the human
3.
RI PT
368
Arrieta MC, Stiemsma LT, Amenyogbe N, Brown EM, Finlay B. The intestinal
Collado MC, Cernada M, Baüerl C, Vento M, Pérez-Martínez G. Microbial ecology and host-microbiota interactions during early life stages. Gut Microbes 2012;
372
3:352-365. 4.
actually sterile? Res Microbiol 2008; 159:187-193.
374 375
Jiménez E, Marín ML, Martín R, et al. Is meconium from healthy newborns
5.
M AN U
373
SC
371
DiGiulio DB, Romero R, Amogan HP, et al. Microbial prevalence, diversity and
376
abundance in amniotic fluid during preterm labor: a molecular and culture-based
377
investigation. PLoS ONE 2008; 3:e3056. 6.
Madan JC, Salari RC, Saxena D, et al. Gut microbial colonisation in premature
TE D
378 379
neonates predicts neonatal sepsis. Arch Dis Child Fetal Neonatal Ed 2012;
380
97:F456-462.
intestinal microbiota in early infancy. Pediatrics 2006; 118:511-521.
382 383
8.
Murgas Torrazza R, Neu J. The developing intestinal microbiome and its
relationship to health and disease in the neonate. J Perinatol 2011; 31(Suppl
384
1):S29–S34.
385 386
Penders J, Thijs C, Vink C, et al. Factors influencing the composition of the
EP
7.
AC C
381
9.
Magne F, Abély M, Boyer F, et al. Low species diversity and high interindividual
387
variability in faeces of preterm infants as revealed by sequences of 16S rRNA
388
genes and PCR-temporal temperature gradient gel electrophoresis profiles. FEMS
389
Microbiol Ecol 2006; 57:128–138.
ACCEPTED MANUSCRIPT 390 391
10. Westerbeek E, Berg A, Lafeber H, et al. The intestinal bacterial colonisation in preterminfants: A review of the literature. Clin Nutr 2006; 25:361–368. 11. Moles L, Gómez M, Heilig H, et al. Bacterial diversity in meconium of preterm
393
neonates and evolution of their fecal microbiota during the first month of life. PLoS
394
ONE 2013; 8(6):e66986.
395 396
RI PT
392
12. Pendleton JN, Gorman SP, Gilmore BF. Clinical relevance of the ESKAPE pathogens. Expert Rev Anti Infect Ther 2013; 11:297-308.
13. Boucher HW, Talbot GH, Bradley JS, et al. Bad bugs, no drugs: no ESKAPE! An
398
update from the Infectious Diseases Society of America. Clin Infect Dis 2009;
399
48:1-12.
401
M AN U
400
SC
397
14. Arumugam M, Raes J, Pelletier E, et al. Enterotypes of the human gut microbiome. Nature 201; 473:174-180.
15. Lagier JC, Hugon P, Khelaifia S, Fournier PE, La Scola B, Raoult D. The rebirth of
403
culture in microbiology through the example of culturomics to study human gut
404
microbiota. Clin Microbiol Rev 2015; 28:237-64.
TE D
402
16. Koenig JE, Spor A, Scalfone N, et al. Succession of microbial consortia in the
406
developing infant gut microbiome. Proc Natl Acad Sci USA 2011; 108(Suppl
407
1):4578-4585.
409 410 411 412 413
17. Yatsunenko T, Rey FE, Manary MJ, et al. Human gut microbiome viewed across
AC C
408
EP
405
age and geography. Nature 2012; 486:222-227.
18. Ruiz-Barba JL, Maldonado A, Jiménez-Díaz R. Small-scale total DNA extraction from bacteria and yeast for PCR applications. Anal Biochem 2005; 347:333-335. 19. Clermont O, Bonacorsi S, Bingen E. Rapid and simple determination of the Escherichia coli phylogenetic group. Appl Environ Microbiol 2000; 66:4555-4558.
ACCEPTED MANUSCRIPT 414
20. Clinical and
Laboratory Standards
Institute. Performance Standards for
415
Antimicrobial Susceptibility Testing: Twenty-first Informational Supplement
416
M100-S21. Wayne, PA, USA: CLSI; 2011. 21. Eaton TJ, Gasson MJ. Molecular screening of Enterococcus virulence determinants
418
and potential for genetic exchange between food and medical isolates. Appl
419
Environ Microbiol 2001; 67:1628-1635.
RI PT
417
22. Reviriego C, Eaton T, Martín R, et al. Screening of virulence determinants in
421
Enterococcus faecium strains isolated from breast milk. J Hum Lact 2005; 21:131-
422
137.
SC
420
23. Delgado S, García P, Fernández L, et al. Characterization of Staphylococcus aureus
424
strains involved in human and bovine mastitis. FEMS Immunol Med Microbiol
425
2011; 62:225-235.
427
24. van Leeuwen WB, Melles DC, Alaidan A, et al. Host- and tissue-specific pathogenic traits of Staphylococcus aureus. J Bacteriol 2005; 187:4584-4591.
TE D
426
M AN U
423
25. Jarraud S, Mougel C, Thioulouse J, et al. Relationships between Staphylococcus
429
aureus genetic background, virulence factors, agr groups (alleles), and human
430
disease. Infect Immun 2002; 70:631-641.
EP
428
26. Yamada T, Tochimaru N, Nakasuji S, et al. Leukotoxin family genes in
432
Staphylococcus aureus isolated from domestic animals and prevalence of lukM-
433 434
AC C
431
lukF-PV genes by bacteriophages in bovine isolates. Vet Microbiol 2005; 110:97103.
435
27. Martı́n MC, González-Hevia MA, Mendoza MC. Usefulness of a two-step PCR
436
procedure for detection and identification of enterotoxigenic staphylococci of
437
bacterial isolates and food samples. Food Microbiol 2003; 20:605-610.
ACCEPTED MANUSCRIPT 438
28. Peacock SJ, Moore CE, Justice A, et al. Virulent combinations of adhesin and toxin
439
genes in natural populations of Staphylococcus aureus. Infect Immun 2002;
440
70:4987-4996. 29. Mehrotra M, Wang G, Johnson WM. Multiplex PCR for detection of genes for
442
Staphylococcus aureus enterotoxins, exfoliative toxins, toxic shock syndrome toxin
443
1, and methicillin resistance. J Clin Microbiol 2000; 38:1032-1035.
RI PT
441
30. Martineau F, Picard FJ, Lansac N, et al. Correlation between the resistance
445
genotype determined by multiplex PCR assays and the antibiotic susceptibility
446
patterns of Staphylococcus aureus and Staphylococcus epidermidis. Antimicrob
447
Agents Chemother 2000; 44:231-238.
449
M AN U
448
SC
444
31. Dale SE, Sebulsky MT, Heinrichs DE. Involvement of SirABC in iron-siderophore import in Staphylococcus aureus. J Bacteriol 2004; 186:8356-8362. 32. Jiménez E, Fernández L, Marín ML, et al. Isolation of commensal bacteria from
451
umbilical cord blood of healthy neonates born by cesarean section. Curr Microbiol
452
2005; 51:270-274.
TE D
450
33. Rautava S, Luoto R, Salminen S, Isolauri E. Microbial contact during pregnancy,
454
intestinal colonization and human disease. Nat Rev Gastroenterol Hepatol 2012;
455
9:565-576.
457 458 459 460 461
34. Aagaard K, Ma J, Antony KM, Ganu R, Petrosino J, Versalovic J. The placenta
AC C
456
EP
453
harbors a unique microbiome. Sci Transl Med 2014; 6:237ra65.
35. Martín V, Maldonado-Barragán A, Moles L, et al. Sharing of bacterial strains between breast milk and infant feces. J Hum Lact 2012; 28:36-44. 36. Fernández L, Langa S, Martín V, et al. The human milk microbiota: origin and potential roles in health and disease. Pharmacol Res 2013; 69:1-10.
ACCEPTED MANUSCRIPT 462
37. Mehall JR, Kite CA, Saltzman DA, Wallett T, Jackson RJ, Smith SD. Prospective
463
study of the incidence and complications of bacterial contamination of enteral
464
feeding in neonates. J Pediatr Surg. 2002; 37:1177-1182. 38. Mehall JR, Kite CA, Gilliam CH, Jackson RJ, Smith SD. Enteral feeding tubes are
466
a reservoir for nosocomial antibiotic-resistant pathogens. J Pediatr Surg 2002;
467
37:1011-1012.
469
39. Hurrell E, Kucerova E, Loughlin M, et al. Neonatal enteral feeding tubes as loci for colonisation by members of the Enterobacteriaceae. BMC Infect Dis 2009; 9:146.
SC
468
RI PT
465
40. Bhutta ZA, Das JK, Bahl R, et al. Can available interventions end preventable
471
deaths in mothers, newborn babies, and stillbirths, and at what cost? Lancet 2014;
472
384:347-370.
M AN U
470
41. Willems RJ, Hanage WP, Bessen DE, Feil EJ. Population biology of Gram-positive
474
pathogens: high-risk clones for dissemination of antibiotic resistance. FEMS
475
Microbiol Rev 2011; 35:872-900.
TE D
473
42. Woodford N, Turton JF, Livermore DM. Multiresistant Gram-negative bacteria: the
477
role of high-risk clones in the dissemination of antibiotic resistance. FEMS
478
Microbiol Rev 2011; 35:736-755.
480
43. Tagare A, Kadam S, Vaidya U, Pandit A. Routine antibiotic use in preterm neonates: a randomised controlled trial. J Hosp Infect 2010; 74:332-326.
AC C
479
EP
476
481
44. Costello EK, Carlisle EM, Bik EM, Morowitz MJ, Relman DA. Microbiome
482
assembly across multiple body sites in low-birthweight infants. MBio. 2013;
483 484 485
4:e00782-13. 45. David LA, Materna AC, Friedman J, et al. Host lifestyle affects human microbiota on daily timescales. Genome Biol 2014; 15:R89.
ACCEPTED MANUSCRIPT 1
TABLE 1. Demographic data and clinical characteristics of the 26 preterm
2
infants included in this study Number of infants or Mean value (range)
Variable
28 (26.6-28.7)
RI PT
Gestational age (wk) Male/Female (n)
13/13
Birth weight (g)
1167 (987.3-1347.3)
Z SCORE
-0.23 (-0.50/0.04) 12/14
Sepsis
7
No Yes <3 days
25 12 13
Days with parenteral nutrition (n=22)
10 (3-12)a
Days with enteral feeding tube (n=26)
59 (46-72) 14 (0-10)a
Days with mechanical ventilation (n=16) Days with CPAPb (n=21)
18.5 (1.5-39)a
Days with oxygen therapy (n=20)
33.5 (1.5-39)a
EP
Days in NICU (n=26)
Days at Hospital (n=26) a
52 (18-81)a 69 (41-92)a
Median (IQR). bCPAP: Continuous positive airway pressure.
AC C
4
1
TE D
>3 days
M AN U
Prophylactic antibiotherapy at birth
3
SC
Vaginal/Cesarean section (n)
wk=weeks, n=number, g=gram.
ACCEPTED MANUSCRIPT
TABLE 2. Bacterial species and total counts in each type of sample Feces (n=144)a
Mother’s milk (n=86)
No.
CFU/g
E. faecalis
24
116
103-1010
49
101-108
E. faecium
12
47
105-1010
6
S. epidermidis
26
62
102-109
S. aureus
7
18
E. coli
19
K. pneumoniae S. marcenses
Formula Milk (n=15)
No.
CFU/ml
No.
CFU/ml
15
101-106
3
103-106
104-106
1
101
0
-
71
101-107
8
101-105
1
102
105-108
8
102-106
4
101-102
1
102
62
105-1010
10
102-105
2
102
1
102
16
63
105-1010
15
102-106
5
103-107
1
104
25
68
107-1010
103-108
10
102-107
4
102-106
56
6
a
7
CFU/g or CFU/ml, Colony forming units per gram or milliliter.
EP
Fecal samples recovered during the NICU admittance, the fecal samples of the 2 year-olds were not included.
AC C
8
CFU/ml
M AN U
No.
SC
(n=26)
Donor Milk (n=35)
RI PT
Infants
TE D
5
ACCEPTED MANUSCRIPT
31
2190
V
M
5
26
2
30
1150
C
M
26
42
3
27
1080
C
F
49
60
4
30
2030
C
M
8
27
5
30
1760
C
M
7
27
6
24
600
C
F
99
113
7
27
1540
V
M
35
68
8
26
790
V
F
68
84
Clo + Amk (4)
SC
1
RI PT
NICU Discharge GA Weight Infant (weeks) (gr) DM Gender (days) (days)
9
25
700
C
M
195
196
10
31
1890
V
F
5
19
11
32
1310
V
F
14
28
12
24
800
V
M
41
41
M AN U
K. pneumoniae + P. mirabilis
Amp + Clo + Amk + Mer
SCN 30
1350
C
F
18
44
14
26
920
V
F
97
102
Amb (4) + Fluc (1)
29
1040
C
F
32
47
16
24
740
V
M
97
116
SCN
17
25
720
V
M
140
144
Van + Amk (7)/ Amb (6)`/ Mica (11) + Fluc (9)
18
28
1100
C
M
43
73
19
31
1430
V
F
20
37
950
C
M
54
21
24
750
C
F
81
92
22
24
870
C
M
82
102
23
27
1040
V
F
43
68
24
29
680
C
M
60
70
25
30
1370
V
F
14
41
26
28
1150
C
F
70
31
52
S. marcenses S. marcenses + E. faecalis
Gen + Tei
EP
27
SCN + P. aeruginosa + K. pneumoniae
Van + Amk (7)
Gen + Van + Amb (5) / Cef (4) / Mero (28)
P. aeruginosa
Gen (4) / Amk (4) Tei (12)
AC C
20
Van + Amk (14) /Mer (3)
Gen (14) +Vanc (9) + Amk (13) +Mer (4)
15
Gen (36) + Clo (3) + Amk (8) + Mer (37) + Amb (43) + Fluoro (5)
E. coli
Gen + Van (3)
TE D
13
SCN
Ceft (7) Amb (8) Fluco (21)
Gen + Clo (1) / Van + Amk (6) / Ambi (8) Fluco (10)
S. marcenses SCN Gen (6) / Tei (13) / Ery (1) / Ambi (13) Vanc +Amk + Amb + Fluco
Gen + Clox
ACCEPTED MANUSCRIPT
Figure 1. Main characteristic of each preterm infant and their NICU admittance period, including the bacterial pathogens detected in feces
RI PT
represented in the different colors. The grey bar indicates the NICU stay of the infant, the different color bars inside the grey one are the bacterial species detected in the feces of the infant, and the sampling time is pointed in dark inside. Each bacteraemic episode and the corresponding
arrows. Infant 12 is highlighted in red as it was the patient that died.
SC
bacterial species is marked with a red arrow. The antibiotic treatment of each infant and the number of days are also represented with grey
M AN U
GA: Gestational Age; DM: Delivery mode, Amb: Ambisome, Amk: Amikacin, Amp: Ampicillin, Clox: Cloxacilina, Ery: Erhytromycin, Flu:
AC C
EP
TE D
Fluconazol, Fluoro: Fluorocytosine, Gen: Gentamicin, Mer: Meropenem, Mica: Micafungine, Tei: Teicoplanin, Vanc: Vancomycin.
ACCEPTED MANUSCRIPT
Enterococcus faecalis
ST81
ST21
ST16 1/2/0
ST179
1/0/1
5/3/0
ST64
0/0/1
RI PT
ST538
0/0/7
ST537 0/0/1
ST133
ST56
SC
0/0/1
ST30
ST40
5/4/0
15/13/0
7/5/0
M AN U
0/3/0
Fecal samples during NICU stance
TE D
Milk samples Fecal samples at 2 years
EP
Virulence factors
gelE cylA ccf
AC C
ST16 ST21 ST30 ST34 ST40 ST56 ST64 ST81
efa
cpd
esp agg2 cad
ST34
6/0/0
0% <50% >50% 100%
Antibiotic susceptibility HLR HLR cob Amp Gm Km Ery
Tet
Cip Van
ACCEPTED MANUSCRIPT
Enterococcus faecium
ST18 ST675
3/2/0
SC
0/0/1
RI PT
ST765
M AN U
ST132
9/1/0
0/1/0
Fecal samples during NICU stance
TE D
Milk samples Fecal samples at 2 years
Antibiotic susceptibility
EP
Virulence factors gelE cylL
ccf
efa
AC C
ST18 ST132 ST765
cpd
esp agg2 cad
0% <50% >50% 100%
cob
HLR HLR Amp Gm Km
Ery
Tet
Cip
Van
ACCEPTED MANUSCRIPT
Staphylococcus areus
ST121 ST34
ST30
RI PT
1/1/0 2/0/0
ST45 ST378
1/1/0
1/1/0
ST1490
SC
1/0/0
M AN U
ST8
0/2/0
Fecal samples during NICU stance Milk samples
TE D
Fecal samples at 2 years
Virulence factors
hlb
hld
hlg
tst
AC C
ST8 ST30 ST34 ST45 ST121 ST378 ST1490
hla
S/F
0% <50% >50% 100%
Antibiotic susceptibility
HLR HLR M/F fnbA clfA sdrE ebpS icaA mecA blaZ sirB Amp Gm Km Ery
EP
eta
0/2/0
Tet
Cip Lnz Van
ACCEPTED MANUSCRIPT
Escherichia coli
ST979
ST2904
4/2/2
ST4349 0/0/1
ST537
0/1/0
0/0/1
ST131
ST1978
ST10 ST1547
4/0/0
ST2660 0/0/1
1/0/0
ST58
ST59 1/0/0
7/2/0
M AN U
ST57
0/0/2
0/0/1
ST393
SC
1/0/0
ST375 1/0/0
1/0/0
ST95
ST538
0/0/1
RI PT
ST69
2/0/1
ST3570 0/0/1
0% <50% >50% 100%
TE D
1/0/1
Fecal samples during NICU stance Milk samples
AC C
EP
Fecal samples at 2 years
ST57 ST58 ST59 ST69 ST95 ST375 ST393 ST538 ST1547 ST1978 ST2904
Antibiotic susceptibility Phylogroup B2 B1/B2 D D B2 B2 D D B2 B2 B2
Amp Cef
Imi
Gen Tet
Cip
ACCEPTED MANUSCRIPT
Klebsiella pneumoniae
ST45
ST35
1/2/0
ST1573
1/1/1
ST1572
ST25
0/1/0
1/0/0
1/0/0
RI PT
ST1486
ST1574
1/0/0
ST505
ST29
M AN U
ST1547
SC
1/1/0
0/1/0
2/0/0
ST641
ST252 ST70 0/1/0
TE D
2/1/1
Fecal samples during NICU stance Milk samples
7/3/0
EP
4/4/0
Fecal samples at 2 years
Antibiotic susceptibility
Amp
Imi
Gen
T et
Cip
AC C
ST25 ST29 ST35 ST45 ST 70 ST 242 ST 505 ST 641 ST1486 ST1547 ST 1572 ST 1573 ST 1574
Cef
0% <50% >50% 100%
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2. Summary of the genetic diversity, antibiotic resistance profiles and virulence factors obtained for the different bacterial species. The upper part of the figure represents the genetic diversity obtained by MLST and using the minimum spanning tree algorithm. Each circle represents a MLST clone; its size depends on the number of isolates in that group and its clone number is indicated above the circle, whereas the number below indicates its frequency per origin (feces during NICU stay, milk samples or feces at 2 years). Clones considered as high-risk clones are marked in bold. STs that are genetically related are connected by black lines when they are strongly related and grey when the relation is weak. The antibiotic susceptibility of each clone is represented at the bottom of the figure. For this purpose all isolates belonging to the same ST were grouped, and the antibiotic resistances are represented in function of their frequency (see legend). The same scheme was followed to describe the virulence genes carriage for the gram-positive isolates.
ACCEPTED MANUSCRIPT
October
November
December
January
April
May
June E. faecalis
E. faecium S. aureus
E. coli
K. pneumoniae
AC C
EP
TE D
M AN U
SC
ST 18 ST 765 ST 132 ST 1490 ST 8 ST 30 ST 378 ST 34 ST 45 ST 121 ST 95 ST 69 ST 393 ST 59 ST 538 ST 1978 ST 1547 ST 58 ST 2904 ST 375 ST 1574 ST 641 ST 70 ST 1573 ST 252 ST 29 ST 25 ST 505 ST 1547 ST 45 ST 35 ST 1572
March
RI PT
ST 81 ST 34 ST 64 ST 21 ST 30 ST 40 ST 56 ST 16
February
Figure 3. Persistence and frequency of the different ST detected during the studied period in the NICU. The E. faecalis ST64 and ST40 persistent clones are marked.
ACCEPTED MANUSCRIPT
2
3
4
5
6
7
8
9
10
11
12
13
SC M AN U TE D
Actinobacteria
Bacilli
Bacteroidetes
Gammaproteobacteria
EP AC C
Bifidobacterium breve Bifidobacterium longum Corynebacterium sp. Dermabacter hominis Microbacterium oxydans Micrococcus luteus Rothia dentocariosa Bacillus sp. Enterococcus faecalis Enterococcus faecium Lactobacillus casei Lactobacillus fermentum Lactobacillus gasseri Lactobacillus salivarius Lactococcus lactis Leuconostoc citreum Staphylococcus aureus Staphylococcus epidermidis Other Staphylococcus Streptococcus agalactiae Streptococcus mitis Other Streptococcus Bacteroides fragilis Citrobacter sp. Enterobacter cancerogenus Enterobacter cloacae Enterobacter hormaechei Erwinia persicina Escherichia coli Escherichia fergusonii Klebsiella granulomatis Klebsiella oxytoca Klebsiella pneumoniae Klebsiella sp. Klebsiella variicola Morganella morganii Pantoea dispersa Proteus penneri Proteus sp. Proteus mirabilis Proteus vulgaris Pseudomonas fluorescens Other Pseudomonas Serratia marcescens Serratia nematodiphila Serratia ureilytica Shigella sp. Shigella dysenteriae Other Shigella Stenotrophomonas maltophilia Veillonella parvula Yeast
RI PT
0 7 14 21 0 7 14 21 28 42 0 7 14 21 28 35 42 56 0 7 14 21 0 7 14 21 0 14 21 28 35 49 63 70-84 0 7 14 21 28 35 49 63 0 7 14 21 28 35 56 84 7 14 21 28 42 56 77-210 0 14 0 7 14 21 14 21 7 14 21 42
1
Negativicutes
ACCEPTED MANUSCRIPT
15
16
17
18
19
20
21
22
23
24
25
26
SC M AN U TE D
Actinobacteria
Bacilli
Bacteroidetes
Gammaproteobacteria
EP AC C
Bifidobacterium breve Bifidobacterium longum Corynebacterium sp. Dermabacter hominis Microbacterium oxydans Micrococcus luteus Rothia dentocariosa Bacillus sp. Enterococcus faecalis Enterococcus faecium Lactobacillus casei Lactobacillus fermentum Lactobacillus gasseri Lactobacillus salivarius Lactococcus lactis Leuconostoc citreum Staphylococcus aureus Staphylococcus epidermidis Other Staphylococcus Streptococcus agalactiae Streptococcus mitis Other Streptococcus Bacteroides fragilis Citrobacter sp. Enterobacter cancerogenus Enterobacter cloacae Enterobacter hormaechei Erwinia persicina Escherichia coli Escherichia fergusonii Klebsiella granulomatis Klebsiella oxytoca Klebsiella pneumoniae Klebsiella sp. Klebsiella variicola Morganella morganii Pantoea dispersa Proteus penneri Proteus sp. Proteus mirabilis Proteus vulgaris Pseudomonas fluorescens Other Pseudomonas Serratia marcescens Serratia nematodiphila Serratia ureilytica Shigella sp. Shigella dysenteriae Other Shigella Stenotrophomonas maltophilia Veillonella parvula Yeast
RI PT
0 7 14 21 49 63 84-98 0 7 14 21 28 7 14 21 49 63 70-98 7 14 21 28 35 49 84-126 14 21 35 49 70 0 7 14 21 0 7 14 28 35 56 0 7 14 21 49 63 14 21 28 42 56 70 0 14 21 49 63 14 21 35 49 7 21 21 28
14
Supplementary Figure 1. Detection of minority species in each infant and sample.
Negativicutes
S1198-743X(15)00616-3
DOI:
10.1016/j.cmi.2015.06.003
Reference:
CMI 303
To appear in:
Clinical Microbiology and Infection
Received Date: 18 January 2015 Revised Date:
12 May 2015
Accepted Date: 5 June 2015
Please cite this article as: Moles L, Gómez M, Jiménez E, Fernández L, Bustos G, Chaves F, Cantón R, Rodríguez JM, Campo Rd, Preterm infant gut colonization in the Neonatal ICU and Complete Restoration two years later, Clinical Microbiology and Infection (2015), doi: 10.1016/j.cmi.2015.06.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Preterm Infant Gut Colonization in the Neonatal ICU and
2
Complete Restoration Two Years Later
3
Laura Moles1,2, Marta Gómez1, Esther Jiménez1,3, Leónides
5
Fernández1,3, Gerardo Bustos4,5, Fernando Chaves6,7, Rafael Cantón2,7,8,
6
Juan Miguel Rodríguez1,3, Rosa del Campo2,7,8
RI PT
4
7 8
1
9
Complutense de Madrid. 2Instituto Ramón y Cajal de Investigaciones Sanitarias
10
(IRYCIS), Madrid. 3ProbiSearch, S.L., Tres Cantos, Madrid. 4Servicio de Neonatología
11
Hospital Universitario 12 de Octubre, Madrid. 5Red de Salud Materno-Infantil y del
12
Desarrollo (SAMID), Instituto Carlos III, Madrid. 6Servicio de Microbiología, Hospital
13
Universitario 12 de Octubre, Madrid. 7Servicio de Microbiología, Hospital Universitario
14
Ramón y Cajal, Madrid. 8Spanish Network for Research in Infectious Diseases (REIPI),
15
Instituto de Salud Carlos III, Madrid. SPAIN.
SC
M AN U
TE D
EP
16
Departmento de Nutrición, Bromatología y Tecnología de los Alimentos, Universidad
Correspondence author: Rosa del Campo, Servicio de Microbiología, Hospital
18
Universitario Ramón y Cajal, Ctra. Colmenar, Km 9.1, 28034 Madrid, SPAIN.
19
Email: [email protected], Phone: +34 91 336 85 42, Fax: +34 91 336 88 09.
20 21
AC C
17
Running title: Gut Microbiota of Preterm Infants
22 23
Key words: Nosocomial High-Risk Clones, Antibiotic Resistance, Virulence Factors,
24
Human Milk, Gut Microbiota, Preterm Infants.
25
ACCEPTED MANUSCRIPT ABSTRACT
27
Preterm infants in a Neonatal Intensive Care Unit (NICU) are exposed to multi-drug
28
resistant bacteria previously adapted to the hospital environment. The aim of the present
29
study was to characterize the bacterial antibiotic-resistant high-risk lineages colonizing
30
preterm infants during their NICU stay and their persistence in faeces after 2 years.
RI PT
26
A total of twenty-six preterm neonates were recruited between October 2009 and
32
June 2010 and provided 144 faecal samples. Milk samples (86 mothers’ milk, 35 human
33
donor milk, and 15 formula milk) were collected at the same time as faecal samples. An
34
additional faecal sample was recovered in 16 infants at the age of two years. Samples
35
were plated onto different selective media and one colony per morphology was selected.
36
Isolates were identified by 16S rDNA nucleotide sequence and MALDI-TOF.
37
Antibiotic susceptibility (agar dilution), genetic diversity (RAPD, PFGE and MLST)
38
and virulence factors (only in enterococcal and staphylococcal isolates) were
39
determined by PCR.
TE D
M AN U
SC
31
A high proportion of antibiotic resistant high-risk clones was detected in both
41
faecal and milk samples during the NICU admittance. Almost all infants were colonized
42
by Enterococcus faecalis ST64 and Enterococcus faecium ST18 clones, while a wider
43
genetic diversity was observed for the Gram-negative isolates. Multi-resistant high-risk
44
clones were not recovered from the faecal samples of the 2-year-old infants.
AC C
45
EP
40
In conclusion, the gut of the preterm infants admitted in NICU might be initially
46
colonized by antibiotic-resistant and virulent high-risk lineages, which are later replaced
47
by antibiotic-susceptible community ones.
48
ACCEPTED MANUSCRIPT
INTRODUCTION
50
The gut microbiota ecosystem plays an important role in human health and its
51
alterations have been implicated in several systemic diseases, particularly in the early
52
stages of life [1]. After birth, an adequate gut microbiota establishment is essential for
53
both the innate immune system maturation and the creation of a barrier protection from
54
bacterial pathogens [2,3]. Moreover, these first bacterial cells colonizing our gut mucosa
55
might persist in symbiosis with our enterocytes throughout our complete life.
RI PT
49
Bacterial gut colonization starts during the intrauterine stage, increasing
57
drastically during the delivery and the first days of life [4-6]. This process can be
58
adversely affected by numerous factors, including antibiotic treatment, caesarean
59
delivery, early separation from the mother, formula feeding, invasive medical
60
treatments and long stays in Neonatal Intensive Care Units (NICU) [7,8]. Some of these
61
factors are usually associated with premature infants who in addition have a delayed or
62
aberrant gut microbiota characterized by poor global bacterial diversity and a low and
63
rather fluctuating Bifidobacterium population [9-11].
TE D
M AN U
SC
56
During their usually long NICU stays, preterm infants are exposed to multi-drug
65
resistant bacteria previously adapted to the hospital environment, that have been
66
recognized as high-risk clones. Most of these high-risk clones belong to the ESKAPE
67
group (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae,
68
Acinetobacter baumanii, Pseudomonas aeruginosa, and Enterobacter species) [12,13].
69
Furthermore, these preterm infants are usually treated with antibiotics which also select
70
the multiresistant bacteria that might colonize their guts.
AC C
EP
64
71
In the last years, the implementations of metagenomic tools based on 16S rDNA
72
massive sequence has provide large information about the complexity of the gut
73
microbiota ecosystem [14], including the identification of uncultured microorganisms.
ACCEPTED MANUSCRIPT 74
However, classical and extended techniques based on classical microbiological cultures
75
as culturomics, also provides a comprehensive approach for the human gut microbiota
76
study, and allowed us the further characterization of the bacterial isolates [15]. Although previous studies have assessed the gut colonization of preterm infants,
78
little is known about its evolution after hospital discharge and in their first years of life,
79
particularly around the two years of life, when a mature microbiota has been already
80
established. The aim of the present study is to characterize the bacterial lineages
81
colonizing premature infants during their NICU admittance and their persistence after 2
82
years, which is the age when the infant’s gut starts to contain an adult-like microbiota
83
[16,17].
84
M AN U
SC
RI PT
77
METHODS
86
Subjects and sampling
87
Twenty-six preterm infants (≤32 weeks of gestation and/or ≤1,500 g of weight) born
88
between October 2009 and June 2010 in the 12 de Octubre University Hospital in
89
Madrid were included in the study. Preterm infants with malformations, metabolic
90
diseases or severe conditions were excluded. Relevant demographic and clinical data are
91
summarized in Table 1. The local ethic committee approved the study (reference
92
09/157), and parents signed the fully informed consent. Following the routine NICU
93
feeding protocols, all infants were preferably fed with their own mother’s milk and
94
when this was not possible, with pasteurized human milk from the Milk Bank Unit.
95
When the weight of the infants was ≥1,500 grams and their own mother’s milk was
96
unavailable, they received adapted preterm formula. The fortifier FM 85 (Nestle) was
97
used approximately from the fourth week of life with an initial dose of 1 g per 50 ml of
98
milk that was increased to 1 g per 20 ml. Globally, the feeding patterns of the recruited
AC C
EP
TE D
85
ACCEPTED MANUSCRIPT 99 100
infants were very heterogeneous, a fact that prevented the formation of well-defined feeding groups. The first spontaneously evacuated meconium after delivery and faecal samples
102
were collected weekly during the first month of life and, then, every 15 days until
103
discharged from the NICU. Samples were collected and immediately stored at -80ºC.
104
Once the infants were 2-years-old, an additional faecal sample was also collected
105
although we only were able to contact 16 out of the 26 families.
RI PT
101
Mother’s milk was extracted using electric pumps and stored either refrigerated
107
for a maximum of 24 h or frozen for up to 6 months. Donor milk, pasteurized at 62.5ºC
108
after collection has a shelf life of 3 months. All milks were incubated for 10-15 min at
109
37-40ºC and administered to the infant through an orogastric or nasogastric feeding tube
110
using a syringe (manual feeding) or a pump (automatic feeding). Milk samples were
111
always collected after feeding from either the syringe or the extension set connected
112
with the orogastric or nasogastric feeding tube, at the same sampling times as the faecal
113
ones.
TE D
M AN U
SC
106
Bacterial strains causing a bacteraemia were recovered from the blood samples
115
after their identification was performed at the Department of Microbiology at the 12 de
116
Octubre University Hospital.
AC C
117
EP
114
118
Culture analysis of the samples
119
Suitable dilutions of milk and faecal samples obtained during the NICU stay were
120
spread onto Man Rogosa Sharpe (Oxoid, Basingstoke, UK) supplemented or not with L-
121
cysteine (0.5 g/L) (Sigma, St. Louis, USA) for lactic acid bacteria isolation;
122
MacConkey (BioMérieux, Marcy l’Etoile, France) for Enterobacteriaceae; Baird Parker
123
(BioMérieux) for staphyloccoci, Brain Heart Infusion (Oxoid), Wilkins-Chalgren
ACCEPTED MANUSCRIPT 124
(Oxoid) and finally in Columbia Nalidixic Acid (BioMérieux) as general agar media
125
plates for isolation of bacterial organisms. Persistence of selected antibiotic resistant strains acquired in the NICU stay was
127
evaluated in faecal samples of the 2-year-old infants; on the one hand with selective
128
agar media supplemented or not with antibiotics such as Manitol Salt Agar for
129
staphylococci (Difco Laboratories, Detroit, USA) and M-Enterococcus agar (Difco)
130
alone and supplemented with 16 µg/ml of ampicillin or 16 µg/ml of erythromycin. On
131
the other hand MacConkey plates were supplemented with 4 µg/ml of ciprofloxacin, or
132
2 µg/ml of imipenem or 1 µg/ml of cefotaxime.
SC
RI PT
126
Plates were aerobically incubated at 37ºC for up to 48 h, with the exception of the
134
Wilkins-Chalgren and Man Rogosa Sharpe-cysteine plates that were anaerobically
135
incubated (85% nitrogen, 10% hydrogen, and 5% carbon dioxide) in an anaerobic
136
workstation (Mini-MACS Don Whitley Scientific Limited, Shipley, UK) at 37ºC for 48
137
h. Bacterial counts were recorded as the colony forming units (CFU) per gram of
138
meconium or faeces, or CFU per ml of milk, and transformed to log10 values.
TE D
M AN U
133
139
Bacterial identification and genotyping
141
After the initial samples processing, ~4000 colonies were selected including at least one
142
colony per morphology, media and sample. Isolates were subcultured in Brain Heart
143
Infusion and Man Rogosa Sharpe media. In order to avoid repeated isolates, those with
144
the same morphology from the same media and sample were submitted to randomly
145
amplified polymorphic DNA polymerase chain reactions (RAPD-PCR) using the primer
146
OPL5 (5’-ACGCAGGCAC-3’) [18]. Finally, at least one representative isolate from
147
each sample, media, and infant was further selected. Bacterial identification was
148
performed
AC C
EP
140
by
matrix-assisted
laser
desorption/ionization
time-of-flight
mass
ACCEPTED MANUSCRIPT spectrometry (MALDI-TOF MS) (Bruker Daltonics, Bremen, Germany) combined with
150
16S rDNA PCR amplification and further nucleotide sequencing. Staphylococcal
151
species belonging to the species S. aureus, Staphylococcus epidermidis, or
152
Staphylococcus hominis were identified by a multiplex PCR method based on the dnaJ
153
genes [4].
RI PT
149
154
Genetic diversity
156
The genetic diversity of all selected isolates was assessed by pulsed field gel
157
electrophoresis (PFGE) in a CHEF DR II apparatus (Bio-Rad, Birmingham, UK). To
158
separate SmaI digested fragments of enterococci and staphylococci different protocols
159
were applied (2 to 28 s for 24 h; and 5 to 15 s for 10 h and then 15 to 60 s for 13 h,
160
respectively). The chromosomal DNA of K. pneumoniae isolates was digested with
161
XbaI enzyme and the electrophoresis conditions were 1 to 40 s for 20 h. Dendrograme
162
analysis of PFGE profiles was performed using the UPGMA method based on the Dice
163
similarity by the Phoretix 5.0 software (Nonlinear Dynamics Ltd., United Kingdom).
TE D
M AN U
SC
155
Multilocus sequence typing (MLST) schemes were applied for PFGE unrelated
165
strains of Enterococcus faecalis, E. faecium and S. aureus (www.mlst.net); Escherichia
166
coli (http://mlst.ucc.ie/mlst/dbs/Ecoli), and K. pneumoniae (http://www.pasteur.fr
167
/recherche/genopole/PF8/mlst/Kpneumoniae.html). The E. coli phylogenetic group was
168
also determined by PCR [19]. MLST was represented by Minimum Spanning Tree
169
algorithm using the Phyloviz software (http://www.phyloviz.net)
AC C
EP
164
170
171
Antimicrobial susceptibility testing
172
Antibiotic susceptibility for ampicillin, cefotaxime, imipenem, gentamicin, kanamycin,
173
erythromycin, tetracycline, ciprofloxacin, and vancomycin was determined only in the
ACCEPTED MANUSCRIPT 174
selected isolates for PFGE and MLST typing by agar dilution following the Clinical and
175
Laboratory Standards Institute guidelines and breakpoints for each genus/specie [20].
176
All antibiotics were purchased from Sigma-Aldrich Co. (http://www.sigmaaldrich.
177
com/).
RI PT
178
Virulence factors carriage
180
Presence of genes encoding for sex pheromones (ccf, cpd, cad, cob), cell wall adhesins
181
(efaAfs), the aggregation protein (agg2), the protein involved in immune evasion (epsfs)
182
presence, the extracellular metaloendopeptidase (gelE) and the cytolisine (cylA) for
183
enterococci were assessed by the procedure described by Eaton and Gasson and
184
modified by Reviriego et al., [21,22].
M AN U
SC
179
Carriage of hemolysins (hla, hlb, hlg, hld) and adhesins (fnbA, clfa, sdrE, ebpS)
186
genes in the S. aureus isolates was determined by two multiplex PCRs [23], while a
187
single PCR scheme was used for the detection of the genes encoding the exfoliative
188
toxin A (eta), toxic shock syndrome toxin (tst), adhesion for collagen (cna) [24],
189
Panton-Valentin leukocidin components S and F (pvl-S/F) [25], Panton-Valentin
190
leukocidin components M and F (pvl-M/F) [26], staphylococcal enterotoxins (se-adej)
191
[27], polysaccharide intracellular adhesion (icaA) [28], methicillin resistance (mecA)
192
[29], β-lactamase (blaZ) [30], and iron siderophore transporter (sirB) [31]. At least one
193
isolate per infant and sample obtained during the NICU admission was evaluated for
194
virulence factors carriage.
AC C
EP
TE D
185
195
196
RESULTS
197
The main demographic and clinical data of the 26 preterm infants are described in Table
198
1. In all but one infant, the ampicillin-gentamicin combination was administered at birth
ACCEPTED MANUSCRIPT as empirical treatment, whereas teicoplanin, gentamicin, amikacin and ambisome were
200
used in the septic episodes (Figure 1). Seven infants (27%) had, at least, one
201
bacteraemic episode, and the foremost species involved were coagulase-negative
202
staphylococci (n=5), Serratia marcescens (n=3), K. pneumoniae (n=2), and P.
203
aeruginosa (n=2) (Figure 1).
RI PT
199
A total of 296 samples from infant’s faeces (144 during the NICU admission and
205
16 at 2 years-old), 86 from mother’s milk, 35 from donor’s milk, and 15 from formula’s
206
milk were processed. These samples yield ~4000 bacterial isolates and the most
207
numerous species was E. faecalis (n=564), followed by S. epidermidis (n=420), S.
208
marcescens (n=267), E. coli (n=252), K. pneumoniae (n=228), E. faecium (n=108) and
209
S. aureus (n=60) (Table 2). Other less abundant but still relevant species belonged to
210
Enterobacter (n=71), Pseudomonas (n=22), Acinetobacter genera (n=16), and others
211
(Suplementary Figure 1).
M AN U
SC
204
The number of strains with different PFGE profiles coexisting in the same infant
213
during NICU admittance oscillated between 1 and 5 for E. faecalis, 1 and 4 for E.
214
faecium, 1 and 10 for S. aureus, and 1 and 3 in K. pneumoniae. The final collection of
215
unrelated isolates was composed by 67 E. faecalis, 48 E. coli, 43 K. pneumoniae, 22 E.
216
faecium, and 17 S. aureus isolates, which were submitted to antibiotic susceptibility
217
testing and MLST typing. At this stage, S. epidermidis isolates were not studied further
218
due to their high number and the high genetic diversity observed; Serratia isolates were
219
excluded from the analysis due to the lack of an MLST scheme, while Acinetobacter,
220
Pseudomonas and Enterobacter isolates were scarcely represented in our population.
AC C
EP
TE D
212
221
The proportion of the high-risk clones detected in the faecal samples during the
222
NICU admittance was considerable: 5 (ST16-CC8, ST21-CC21, ST40-CC40, ST56-
223
CC30 and ST64-CC8) of the 8 STs found among the E. faecalis isolates; 2 (ST18-CC18
ACCEPTED MANUSCRIPT and ST132-CC18) of 3 in E. faecium; 6 (ST8-CC8, ST30-CC30, ST34-CC30, ST45-
225
CC45, ST121-CC121, and ST1490-CC8) of 7 in S. aureus, 5 (ST58, ST59, ST69,
226
ST393, ST2904) of 10 in E. coli, and finally 5 (ST25, ST29, ST35, ST45, and ST1547,
227
all of them fit into the CC292) of 12 in K. pneumoniae. Isolates belonging to these high-
228
risk clones exhibited resistance to most of the studied antibiotics and carried multiple
229
virulent factors (Figure 2).
RI PT
224
It is of note that identical STs with the same antibiotic resistance phenotype and
231
the same virulence factors were also detected in the milk and formula samples. In infant
232
14, E. faecalis ST56 and K. pneumoniae ST505 were simultaneously detected in faeces
233
and in blood samples during a bacteraemic episode. Infant 12 died in the NICU after a
234
complicated E. coli bacteraemia.
M AN U
SC
230
Almost all infants were colonized by both E. faecalis ST64 (19/24) and E.
236
faecium ST18 (9/12) clones, while a wider genetic diversity was observed for the Gram-
237
negative isolates. The predominant clones were ST393 for E. coli (8/20) and ST641 for
238
K. pneumoniae (8/16) (Figure 3).
TE D
235
Among the E. faecalis and E. faecium isolates, a high level of antibiotic resistance
240
was observed, especially for tetracycline (96.7% and 97.4%) and erythromycin (68.8%
241
and 92.9%); whereas ampicillin resistance was only detected in E. faecium isolates
242
(92.2%). E. faecalis ST16 and ST40 clones harbored the highest number of virulence
243
determinants. The most common virulent genes were gelE (84.3%), ccf (100%) and
244
efaAfs (95.7%) in E. faecalis isolates, and epsfs (24.8%), gelE (10.3%) and ccf (10.3%) in
245
E. faecium isolates (Figure 2).
AC C
EP
239
246
Globally, hld, fnbA, clfA, and icaA were the most frequent virulence genes in the
247
S. aureus isolates which were susceptible to the evaluated antibiotics, although 43% of
248
them exhibited resistance to gentamicin (Figure 2).
ACCEPTED MANUSCRIPT For the E. coli isolates, high percentages of tetracycline resistance (46%), and
250
ciprofloxacin (32%) were observed (Figure 2). As to K. pneumoniae, 21% of the
251
isolates were resistant to tetracycline while those belonging to the predominant K.
252
pneumoniae ST641 subgroup displayed resistance to ciprofloxacin, imipenem and
253
cefotaxime (Figure 2).
RI PT
249
In contrast, isolates from the faecal samples collected from the 2-year-old infants
255
were susceptible to the antibiotics and the proportion of high–risk clones was reduced:
256
3 (ST16-CC8, ST40-CC40, and ST64-CC8) of 7 in E. faecalis; none for E. faecium and
257
S. aureus, 5 (ST10-CC10, ST57, ST59, ST69, ST131, and ST3570-CC23) of 10 in E.
258
coli, and finally 1 (ST1486-CC292) of 2 in K. pneumoniae (Figure 2).
M AN U
SC
254
259
DISCUSSION
261
Recent studies have demonstrated that bacterial colonization starts during pregnancy
262
[4,32,33], and in fact the placenta harbours its own microbiota [34]. According to
263
general agreement, the first gut bacterial colonization constitutes the core of the future
264
microbiota that can influence the host’s gastrointestinal system and general health
265
throughout life. The definitive gut microbiota is usually established at the age of two
266
years, when the complete diet has introduced.
EP
AC C
267
TE D
260
In the present work, we analyzed the gut microbiota of the preterm infants during
268
their hospital admittance and its maintenance overtime. In contrast to the next
269
generation massive sequence tools, we preferred the classical techniques based on
270
microbiological culture, in order to collect the bacterial strains for further
271
characterization. A high proportion of high-risk clones linked to nosocomial infection
272
were detected both in the milk administered to the infants and in their faeces. As the
273
different milks had been processed and properly stored, the bacterial load could not be
ACCEPTED MANUSCRIPT attributed to a dramatic and continuous food contamination. In the NICU where the
275
preterm infants were recruited, single-use sterilized and ready-to-use preterm formula
276
are purchased while donor’s milk is pasteurized and, then, submitted to strict
277
microbiological controls. In relation to mother’s milk that has not been pasteurized, it
278
has been described that contains its own microbiota [35, 36], but its composition and
279
concentration differ considerably from the results obtained in this study. Milk sampling
280
process represents a limitation of our study, because we only collected the milk after the
281
tube or syringe pass, and we cannot ruled out that the milk might be also contaminated
282
previously.
SC
RI PT
274
The orogastric or nasogastric feeding tubes, irrespective of the feeding regime,
284
can act as locus for the bacterial attachment and multiplication of numerous
285
opportunistic pathogens [37, 38], and they constitute an important risk factor in the
286
neonatal infections [39]. In agreement with our results, the feeding system might act as
287
a reservoir for the hospital adapted high-risk bacterial clones and contaminate the
288
human and formula milk samples during their passage. After this study, the feeding
289
systems protocol in preterm infants of our NICU was subsequently modified and
290
nowadays the enteral tubes are replaced in a maximum of 48 hours. Other relevant
291
limitation was the lack of maternal samples to complete discharged the existence of
292
these multirresistant clones in their gut or skin microbiota. We hypothesized that the
293
clones have a nosocomial origin, although the maternal source also represents an
294
important option.
AC C
EP
TE D
M AN U
283
295
On the other hand, sepsis is one of the main causes of death in the NICU, and its
296
prevention and treatment considerably increase the sanitary costs [40]. The main
297
bacteria causing outbreaks in the NICU is S. epidermidis, followed by K. pneumoniae,
298
Serratia spp., E. coli and Enterococcus spp. Among them, hospital-adapted high-risk
ACCEPTED MANUSCRIPT clones that exhibit antibiotic resistance and contain virulence factors are of the
300
uppermost relevance [41,42]. In this study, the structure and main characteristics of the
301
E. faecalis, E. faecium, S. aureus, E. coli and K. pneumoniae populations were
302
investigated since they were considered to be the most representative species in the
303
analyzed samples.
RI PT
299
Epidemiological molecular studies based on MLST schemes have demonstrated
305
that the nosocomial pathogens usually grouped in particular lineages as CC10 in E. coli,
306
ST258 in K. pneumoniae [43], CC2/CC9 in E. faecalis, CC17 in E. faecium, and
307
CC5/CC8/CC30/CC45 in S. aureus [44]. In our study the proportion of these high-risk
308
clones detected in the faecal and milk samples was high, although the clonal
309
replacement observed (Figure 2) prevented an epidemic situation in the NICU (Figure
310
3).
M AN U
SC
304
Despite the elevated amount of enterococci isolates obtained during the NICU
312
stay (564 E. faecalis and 108 E. faecium), a low genetic diversity was observed (8 E.
313
faecalis and 3 E. faecium STs), whereas the remaining genera showed a higher
314
variability (7 STs of S. aureus from 7 infants, 10 STs of E. coli and 12 STs of K.
315
pneumoniae). Only E. faecalis ST64 and ST40 clones persisted during almost all the
316
studied period (Figure 3) which signified that the epidemiological measures implanted
317
routinely in this NICU are adequate including hand-washing protocols, aseptic
318
conditions for milk manipulation, early detection and further isolation of antibiotic
319
multi-resistant carriers, segregation of patients and staff during epidemic outbreaks, and
320
adequate antibiotic politics in consensus with the Microbiology Department.
AC C
EP
TE D
311
321
All but one infant received empirical antibiotic therapy after birth. Although the
322
prophylactic use of antibiotics immediately after birth usually exerts a negative
323
influence on early gut colonization and its protective effect has not been demonstrated
ACCEPTED MANUSCRIPT [43], a correct antibiotic therapy may be decisive to assure the survival of many preterm
325
infants. However, it is obvious that this routine practice may be a selector for antibiotic
326
resistant isolates that, subsequently, could be overrepresented, on a long-term basis, in
327
the host’s gut [44]. Therefore, persistence of the high-risk antibiotic-resistant clones
328
acquired during the NICU stay was evaluated in the same cohort two years after birth.
329
Interestingly, the initial antibiotic-resistant isolates and high-risk clones were scarcely
330
detected but, in turn, other clones, mainly observed in the community and initially non-
331
detected, such as E. coli ST131, ST10-CC10 or ST3570-CC23, were found.
SC
RI PT
324
Previous theories generally assume that the first bacterial isolates reaching the
333
almost “microbial-free” gut during or after birth have a founder effect and, therefore,
334
they tended to persist as members of the gut microbiota throughout life [45]. In contrast,
335
our results showed that after the important selective pressures received early in life by
336
the preterm infants, the initial nosocomial and antibiotic-resistant high-risk lineages had
337
been replaced in their gut by antibiotic-susceptible community lineages when they were
338
2-year-old, when perturbations of the early microbiota decrease and become adult-like
339
microbiota [16,17].
EP
340
TE D
M AN U
332
Author Contributions
342
Laura Moles, Leónides Fernández, Rafael Cantón, Juan Miguel Rodríguez and Rosa del
343
Campo conceptualized and designed the study, drafted the initial manuscript, and
344
approved the final manuscript as submitted. Laura Moles, Marta Gómez, and Esther
345
Jiménez carried out the samples analyses, and the experimental data collection,
346
reviewed and revised the manuscript, and approved the final manuscript as submitted.
347
Gerardo Bustos, Fernando Chaves participated in the patient’s recruitment, the sample
348
collection and coordinated and supervised clinical data collection at the 12 de Octubre
AC C
341
ACCEPTED MANUSCRIPT 349
Hospital. Also, they critically reviewed the manuscript, and approved the final
350
manuscript as submitted.
351
Acknowledgments
353
We are grateful to the infant and their families and particularly to the staff of the
354
Department of Neonatology of the 12 de Octubre University Hospital.
RI PT
352
This work was supported by the PI13/00205 project (ISCIII-FIS), and by the
356
AGL2013-41980-P project, both from the Ministry of Economy and Competitiveness
357
(MINECO, Spain), and co-financed by the European Development Regional Fund “A
358
Way to Achieve Europe,” ERDF, Spanish Network for the Research in Infectious
359
Diseases (REIPI RD12/0015). L.M. and M.G. were the recipients of predoctoral
360
contracts from MINECO.
361
M AN U
SC
355
Transparency Declaration
363
The authors declare no conflicts of interest.
EP AC C
364
TE D
362
ACCEPTED MANUSCRIPT 365
REFERENCES
366
1.
microbiome in early life: health and disease. Front Immunol 2014; 5:427.
367
2.
gut microbiota. Science 2013; 341:1237439.
369 370
Faith JJ, Guruge JL, Charbonneau M, et al. The long-term stability of the human
3.
RI PT
368
Arrieta MC, Stiemsma LT, Amenyogbe N, Brown EM, Finlay B. The intestinal
Collado MC, Cernada M, Baüerl C, Vento M, Pérez-Martínez G. Microbial ecology and host-microbiota interactions during early life stages. Gut Microbes 2012;
372
3:352-365. 4.
actually sterile? Res Microbiol 2008; 159:187-193.
374 375
Jiménez E, Marín ML, Martín R, et al. Is meconium from healthy newborns
5.
M AN U
373
SC
371
DiGiulio DB, Romero R, Amogan HP, et al. Microbial prevalence, diversity and
376
abundance in amniotic fluid during preterm labor: a molecular and culture-based
377
investigation. PLoS ONE 2008; 3:e3056. 6.
Madan JC, Salari RC, Saxena D, et al. Gut microbial colonisation in premature
TE D
378 379
neonates predicts neonatal sepsis. Arch Dis Child Fetal Neonatal Ed 2012;
380
97:F456-462.
intestinal microbiota in early infancy. Pediatrics 2006; 118:511-521.
382 383
8.
Murgas Torrazza R, Neu J. The developing intestinal microbiome and its
relationship to health and disease in the neonate. J Perinatol 2011; 31(Suppl
384
1):S29–S34.
385 386
Penders J, Thijs C, Vink C, et al. Factors influencing the composition of the
EP
7.
AC C
381
9.
Magne F, Abély M, Boyer F, et al. Low species diversity and high interindividual
387
variability in faeces of preterm infants as revealed by sequences of 16S rRNA
388
genes and PCR-temporal temperature gradient gel electrophoresis profiles. FEMS
389
Microbiol Ecol 2006; 57:128–138.
ACCEPTED MANUSCRIPT 390 391
10. Westerbeek E, Berg A, Lafeber H, et al. The intestinal bacterial colonisation in preterminfants: A review of the literature. Clin Nutr 2006; 25:361–368. 11. Moles L, Gómez M, Heilig H, et al. Bacterial diversity in meconium of preterm
393
neonates and evolution of their fecal microbiota during the first month of life. PLoS
394
ONE 2013; 8(6):e66986.
395 396
RI PT
392
12. Pendleton JN, Gorman SP, Gilmore BF. Clinical relevance of the ESKAPE pathogens. Expert Rev Anti Infect Ther 2013; 11:297-308.
13. Boucher HW, Talbot GH, Bradley JS, et al. Bad bugs, no drugs: no ESKAPE! An
398
update from the Infectious Diseases Society of America. Clin Infect Dis 2009;
399
48:1-12.
401
M AN U
400
SC
397
14. Arumugam M, Raes J, Pelletier E, et al. Enterotypes of the human gut microbiome. Nature 201; 473:174-180.
15. Lagier JC, Hugon P, Khelaifia S, Fournier PE, La Scola B, Raoult D. The rebirth of
403
culture in microbiology through the example of culturomics to study human gut
404
microbiota. Clin Microbiol Rev 2015; 28:237-64.
TE D
402
16. Koenig JE, Spor A, Scalfone N, et al. Succession of microbial consortia in the
406
developing infant gut microbiome. Proc Natl Acad Sci USA 2011; 108(Suppl
407
1):4578-4585.
409 410 411 412 413
17. Yatsunenko T, Rey FE, Manary MJ, et al. Human gut microbiome viewed across
AC C
408
EP
405
age and geography. Nature 2012; 486:222-227.
18. Ruiz-Barba JL, Maldonado A, Jiménez-Díaz R. Small-scale total DNA extraction from bacteria and yeast for PCR applications. Anal Biochem 2005; 347:333-335. 19. Clermont O, Bonacorsi S, Bingen E. Rapid and simple determination of the Escherichia coli phylogenetic group. Appl Environ Microbiol 2000; 66:4555-4558.
ACCEPTED MANUSCRIPT 414
20. Clinical and
Laboratory Standards
Institute. Performance Standards for
415
Antimicrobial Susceptibility Testing: Twenty-first Informational Supplement
416
M100-S21. Wayne, PA, USA: CLSI; 2011. 21. Eaton TJ, Gasson MJ. Molecular screening of Enterococcus virulence determinants
418
and potential for genetic exchange between food and medical isolates. Appl
419
Environ Microbiol 2001; 67:1628-1635.
RI PT
417
22. Reviriego C, Eaton T, Martín R, et al. Screening of virulence determinants in
421
Enterococcus faecium strains isolated from breast milk. J Hum Lact 2005; 21:131-
422
137.
SC
420
23. Delgado S, García P, Fernández L, et al. Characterization of Staphylococcus aureus
424
strains involved in human and bovine mastitis. FEMS Immunol Med Microbiol
425
2011; 62:225-235.
427
24. van Leeuwen WB, Melles DC, Alaidan A, et al. Host- and tissue-specific pathogenic traits of Staphylococcus aureus. J Bacteriol 2005; 187:4584-4591.
TE D
426
M AN U
423
25. Jarraud S, Mougel C, Thioulouse J, et al. Relationships between Staphylococcus
429
aureus genetic background, virulence factors, agr groups (alleles), and human
430
disease. Infect Immun 2002; 70:631-641.
EP
428
26. Yamada T, Tochimaru N, Nakasuji S, et al. Leukotoxin family genes in
432
Staphylococcus aureus isolated from domestic animals and prevalence of lukM-
433 434
AC C
431
lukF-PV genes by bacteriophages in bovine isolates. Vet Microbiol 2005; 110:97103.
435
27. Martı́n MC, González-Hevia MA, Mendoza MC. Usefulness of a two-step PCR
436
procedure for detection and identification of enterotoxigenic staphylococci of
437
bacterial isolates and food samples. Food Microbiol 2003; 20:605-610.
ACCEPTED MANUSCRIPT 438
28. Peacock SJ, Moore CE, Justice A, et al. Virulent combinations of adhesin and toxin
439
genes in natural populations of Staphylococcus aureus. Infect Immun 2002;
440
70:4987-4996. 29. Mehrotra M, Wang G, Johnson WM. Multiplex PCR for detection of genes for
442
Staphylococcus aureus enterotoxins, exfoliative toxins, toxic shock syndrome toxin
443
1, and methicillin resistance. J Clin Microbiol 2000; 38:1032-1035.
RI PT
441
30. Martineau F, Picard FJ, Lansac N, et al. Correlation between the resistance
445
genotype determined by multiplex PCR assays and the antibiotic susceptibility
446
patterns of Staphylococcus aureus and Staphylococcus epidermidis. Antimicrob
447
Agents Chemother 2000; 44:231-238.
449
M AN U
448
SC
444
31. Dale SE, Sebulsky MT, Heinrichs DE. Involvement of SirABC in iron-siderophore import in Staphylococcus aureus. J Bacteriol 2004; 186:8356-8362. 32. Jiménez E, Fernández L, Marín ML, et al. Isolation of commensal bacteria from
451
umbilical cord blood of healthy neonates born by cesarean section. Curr Microbiol
452
2005; 51:270-274.
TE D
450
33. Rautava S, Luoto R, Salminen S, Isolauri E. Microbial contact during pregnancy,
454
intestinal colonization and human disease. Nat Rev Gastroenterol Hepatol 2012;
455
9:565-576.
457 458 459 460 461
34. Aagaard K, Ma J, Antony KM, Ganu R, Petrosino J, Versalovic J. The placenta
AC C
456
EP
453
harbors a unique microbiome. Sci Transl Med 2014; 6:237ra65.
35. Martín V, Maldonado-Barragán A, Moles L, et al. Sharing of bacterial strains between breast milk and infant feces. J Hum Lact 2012; 28:36-44. 36. Fernández L, Langa S, Martín V, et al. The human milk microbiota: origin and potential roles in health and disease. Pharmacol Res 2013; 69:1-10.
ACCEPTED MANUSCRIPT 462
37. Mehall JR, Kite CA, Saltzman DA, Wallett T, Jackson RJ, Smith SD. Prospective
463
study of the incidence and complications of bacterial contamination of enteral
464
feeding in neonates. J Pediatr Surg. 2002; 37:1177-1182. 38. Mehall JR, Kite CA, Gilliam CH, Jackson RJ, Smith SD. Enteral feeding tubes are
466
a reservoir for nosocomial antibiotic-resistant pathogens. J Pediatr Surg 2002;
467
37:1011-1012.
469
39. Hurrell E, Kucerova E, Loughlin M, et al. Neonatal enteral feeding tubes as loci for colonisation by members of the Enterobacteriaceae. BMC Infect Dis 2009; 9:146.
SC
468
RI PT
465
40. Bhutta ZA, Das JK, Bahl R, et al. Can available interventions end preventable
471
deaths in mothers, newborn babies, and stillbirths, and at what cost? Lancet 2014;
472
384:347-370.
M AN U
470
41. Willems RJ, Hanage WP, Bessen DE, Feil EJ. Population biology of Gram-positive
474
pathogens: high-risk clones for dissemination of antibiotic resistance. FEMS
475
Microbiol Rev 2011; 35:872-900.
TE D
473
42. Woodford N, Turton JF, Livermore DM. Multiresistant Gram-negative bacteria: the
477
role of high-risk clones in the dissemination of antibiotic resistance. FEMS
478
Microbiol Rev 2011; 35:736-755.
480
43. Tagare A, Kadam S, Vaidya U, Pandit A. Routine antibiotic use in preterm neonates: a randomised controlled trial. J Hosp Infect 2010; 74:332-326.
AC C
479
EP
476
481
44. Costello EK, Carlisle EM, Bik EM, Morowitz MJ, Relman DA. Microbiome
482
assembly across multiple body sites in low-birthweight infants. MBio. 2013;
483 484 485
4:e00782-13. 45. David LA, Materna AC, Friedman J, et al. Host lifestyle affects human microbiota on daily timescales. Genome Biol 2014; 15:R89.
ACCEPTED MANUSCRIPT 1
TABLE 1. Demographic data and clinical characteristics of the 26 preterm
2
infants included in this study Number of infants or Mean value (range)
Variable
28 (26.6-28.7)
RI PT
Gestational age (wk) Male/Female (n)
13/13
Birth weight (g)
1167 (987.3-1347.3)
Z SCORE
-0.23 (-0.50/0.04) 12/14
Sepsis
7
No Yes <3 days
25 12 13
Days with parenteral nutrition (n=22)
10 (3-12)a
Days with enteral feeding tube (n=26)
59 (46-72) 14 (0-10)a
Days with mechanical ventilation (n=16) Days with CPAPb (n=21)
18.5 (1.5-39)a
Days with oxygen therapy (n=20)
33.5 (1.5-39)a
EP
Days in NICU (n=26)
Days at Hospital (n=26) a
52 (18-81)a 69 (41-92)a
Median (IQR). bCPAP: Continuous positive airway pressure.
AC C
4
1
TE D
>3 days
M AN U
Prophylactic antibiotherapy at birth
3
SC
Vaginal/Cesarean section (n)
wk=weeks, n=number, g=gram.
ACCEPTED MANUSCRIPT
TABLE 2. Bacterial species and total counts in each type of sample Feces (n=144)a
Mother’s milk (n=86)
No.
CFU/g
E. faecalis
24
116
103-1010
49
101-108
E. faecium
12
47
105-1010
6
S. epidermidis
26
62
102-109
S. aureus
7
18
E. coli
19
K. pneumoniae S. marcenses
Formula Milk (n=15)
No.
CFU/ml
No.
CFU/ml
15
101-106
3
103-106
104-106
1
101
0
-
71
101-107
8
101-105
1
102
105-108
8
102-106
4
101-102
1
102
62
105-1010
10
102-105
2
102
1
102
16
63
105-1010
15
102-106
5
103-107
1
104
25
68
107-1010
103-108
10
102-107
4
102-106
56
6
a
7
CFU/g or CFU/ml, Colony forming units per gram or milliliter.
EP
Fecal samples recovered during the NICU admittance, the fecal samples of the 2 year-olds were not included.
AC C
8
CFU/ml
M AN U
No.
SC
(n=26)
Donor Milk (n=35)
RI PT
Infants
TE D
5
ACCEPTED MANUSCRIPT
31
2190
V
M
5
26
2
30
1150
C
M
26
42
3
27
1080
C
F
49
60
4
30
2030
C
M
8
27
5
30
1760
C
M
7
27
6
24
600
C
F
99
113
7
27
1540
V
M
35
68
8
26
790
V
F
68
84
Clo + Amk (4)
SC
1
RI PT
NICU Discharge GA Weight Infant (weeks) (gr) DM Gender (days) (days)
9
25
700
C
M
195
196
10
31
1890
V
F
5
19
11
32
1310
V
F
14
28
12
24
800
V
M
41
41
M AN U
K. pneumoniae + P. mirabilis
Amp + Clo + Amk + Mer
SCN 30
1350
C
F
18
44
14
26
920
V
F
97
102
Amb (4) + Fluc (1)
29
1040
C
F
32
47
16
24
740
V
M
97
116
SCN
17
25
720
V
M
140
144
Van + Amk (7)/ Amb (6)`/ Mica (11) + Fluc (9)
18
28
1100
C
M
43
73
19
31
1430
V
F
20
37
950
C
M
54
21
24
750
C
F
81
92
22
24
870
C
M
82
102
23
27
1040
V
F
43
68
24
29
680
C
M
60
70
25
30
1370
V
F
14
41
26
28
1150
C
F
70
31
52
S. marcenses S. marcenses + E. faecalis
Gen + Tei
EP
27
SCN + P. aeruginosa + K. pneumoniae
Van + Amk (7)
Gen + Van + Amb (5) / Cef (4) / Mero (28)
P. aeruginosa
Gen (4) / Amk (4) Tei (12)
AC C
20
Van + Amk (14) /Mer (3)
Gen (14) +Vanc (9) + Amk (13) +Mer (4)
15
Gen (36) + Clo (3) + Amk (8) + Mer (37) + Amb (43) + Fluoro (5)
E. coli
Gen + Van (3)
TE D
13
SCN
Ceft (7) Amb (8) Fluco (21)
Gen + Clo (1) / Van + Amk (6) / Ambi (8) Fluco (10)
S. marcenses SCN Gen (6) / Tei (13) / Ery (1) / Ambi (13) Vanc +Amk + Amb + Fluco
Gen + Clox
ACCEPTED MANUSCRIPT
Figure 1. Main characteristic of each preterm infant and their NICU admittance period, including the bacterial pathogens detected in feces
RI PT
represented in the different colors. The grey bar indicates the NICU stay of the infant, the different color bars inside the grey one are the bacterial species detected in the feces of the infant, and the sampling time is pointed in dark inside. Each bacteraemic episode and the corresponding
arrows. Infant 12 is highlighted in red as it was the patient that died.
SC
bacterial species is marked with a red arrow. The antibiotic treatment of each infant and the number of days are also represented with grey
M AN U
GA: Gestational Age; DM: Delivery mode, Amb: Ambisome, Amk: Amikacin, Amp: Ampicillin, Clox: Cloxacilina, Ery: Erhytromycin, Flu:
AC C
EP
TE D
Fluconazol, Fluoro: Fluorocytosine, Gen: Gentamicin, Mer: Meropenem, Mica: Micafungine, Tei: Teicoplanin, Vanc: Vancomycin.
ACCEPTED MANUSCRIPT
Enterococcus faecalis
ST81
ST21
ST16 1/2/0
ST179
1/0/1
5/3/0
ST64
0/0/1
RI PT
ST538
0/0/7
ST537 0/0/1
ST133
ST56
SC
0/0/1
ST30
ST40
5/4/0
15/13/0
7/5/0
M AN U
0/3/0
Fecal samples during NICU stance
TE D
Milk samples Fecal samples at 2 years
EP
Virulence factors
gelE cylA ccf
AC C
ST16 ST21 ST30 ST34 ST40 ST56 ST64 ST81
efa
cpd
esp agg2 cad
ST34
6/0/0
0% <50% >50% 100%
Antibiotic susceptibility HLR HLR cob Amp Gm Km Ery
Tet
Cip Van
ACCEPTED MANUSCRIPT
Enterococcus faecium
ST18 ST675
3/2/0
SC
0/0/1
RI PT
ST765
M AN U
ST132
9/1/0
0/1/0
Fecal samples during NICU stance
TE D
Milk samples Fecal samples at 2 years
Antibiotic susceptibility
EP
Virulence factors gelE cylL
ccf
efa
AC C
ST18 ST132 ST765
cpd
esp agg2 cad
0% <50% >50% 100%
cob
HLR HLR Amp Gm Km
Ery
Tet
Cip
Van
ACCEPTED MANUSCRIPT
Staphylococcus areus
ST121 ST34
ST30
RI PT
1/1/0 2/0/0
ST45 ST378
1/1/0
1/1/0
ST1490
SC
1/0/0
M AN U
ST8
0/2/0
Fecal samples during NICU stance Milk samples
TE D
Fecal samples at 2 years
Virulence factors
hlb
hld
hlg
tst
AC C
ST8 ST30 ST34 ST45 ST121 ST378 ST1490
hla
S/F
0% <50% >50% 100%
Antibiotic susceptibility
HLR HLR M/F fnbA clfA sdrE ebpS icaA mecA blaZ sirB Amp Gm Km Ery
EP
eta
0/2/0
Tet
Cip Lnz Van
ACCEPTED MANUSCRIPT
Escherichia coli
ST979
ST2904
4/2/2
ST4349 0/0/1
ST537
0/1/0
0/0/1
ST131
ST1978
ST10 ST1547
4/0/0
ST2660 0/0/1
1/0/0
ST58
ST59 1/0/0
7/2/0
M AN U
ST57
0/0/2
0/0/1
ST393
SC
1/0/0
ST375 1/0/0
1/0/0
ST95
ST538
0/0/1
RI PT
ST69
2/0/1
ST3570 0/0/1
0% <50% >50% 100%
TE D
1/0/1
Fecal samples during NICU stance Milk samples
AC C
EP
Fecal samples at 2 years
ST57 ST58 ST59 ST69 ST95 ST375 ST393 ST538 ST1547 ST1978 ST2904
Antibiotic susceptibility Phylogroup B2 B1/B2 D D B2 B2 D D B2 B2 B2
Amp Cef
Imi
Gen Tet
Cip
ACCEPTED MANUSCRIPT
Klebsiella pneumoniae
ST45
ST35
1/2/0
ST1573
1/1/1
ST1572
ST25
0/1/0
1/0/0
1/0/0
RI PT
ST1486
ST1574
1/0/0
ST505
ST29
M AN U
ST1547
SC
1/1/0
0/1/0
2/0/0
ST641
ST252 ST70 0/1/0
TE D
2/1/1
Fecal samples during NICU stance Milk samples
7/3/0
EP
4/4/0
Fecal samples at 2 years
Antibiotic susceptibility
Amp
Imi
Gen
T et
Cip
AC C
ST25 ST29 ST35 ST45 ST 70 ST 242 ST 505 ST 641 ST1486 ST1547 ST 1572 ST 1573 ST 1574
Cef
0% <50% >50% 100%
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2. Summary of the genetic diversity, antibiotic resistance profiles and virulence factors obtained for the different bacterial species. The upper part of the figure represents the genetic diversity obtained by MLST and using the minimum spanning tree algorithm. Each circle represents a MLST clone; its size depends on the number of isolates in that group and its clone number is indicated above the circle, whereas the number below indicates its frequency per origin (feces during NICU stay, milk samples or feces at 2 years). Clones considered as high-risk clones are marked in bold. STs that are genetically related are connected by black lines when they are strongly related and grey when the relation is weak. The antibiotic susceptibility of each clone is represented at the bottom of the figure. For this purpose all isolates belonging to the same ST were grouped, and the antibiotic resistances are represented in function of their frequency (see legend). The same scheme was followed to describe the virulence genes carriage for the gram-positive isolates.
ACCEPTED MANUSCRIPT
October
November
December
January
April
May
June E. faecalis
E. faecium S. aureus
E. coli
K. pneumoniae
AC C
EP
TE D
M AN U
SC
ST 18 ST 765 ST 132 ST 1490 ST 8 ST 30 ST 378 ST 34 ST 45 ST 121 ST 95 ST 69 ST 393 ST 59 ST 538 ST 1978 ST 1547 ST 58 ST 2904 ST 375 ST 1574 ST 641 ST 70 ST 1573 ST 252 ST 29 ST 25 ST 505 ST 1547 ST 45 ST 35 ST 1572
March
RI PT
ST 81 ST 34 ST 64 ST 21 ST 30 ST 40 ST 56 ST 16
February
Figure 3. Persistence and frequency of the different ST detected during the studied period in the NICU. The E. faecalis ST64 and ST40 persistent clones are marked.
ACCEPTED MANUSCRIPT
2
3
4
5
6
7
8
9
10
11
12
13
SC M AN U TE D
Actinobacteria
Bacilli
Bacteroidetes
Gammaproteobacteria
EP AC C
Bifidobacterium breve Bifidobacterium longum Corynebacterium sp. Dermabacter hominis Microbacterium oxydans Micrococcus luteus Rothia dentocariosa Bacillus sp. Enterococcus faecalis Enterococcus faecium Lactobacillus casei Lactobacillus fermentum Lactobacillus gasseri Lactobacillus salivarius Lactococcus lactis Leuconostoc citreum Staphylococcus aureus Staphylococcus epidermidis Other Staphylococcus Streptococcus agalactiae Streptococcus mitis Other Streptococcus Bacteroides fragilis Citrobacter sp. Enterobacter cancerogenus Enterobacter cloacae Enterobacter hormaechei Erwinia persicina Escherichia coli Escherichia fergusonii Klebsiella granulomatis Klebsiella oxytoca Klebsiella pneumoniae Klebsiella sp. Klebsiella variicola Morganella morganii Pantoea dispersa Proteus penneri Proteus sp. Proteus mirabilis Proteus vulgaris Pseudomonas fluorescens Other Pseudomonas Serratia marcescens Serratia nematodiphila Serratia ureilytica Shigella sp. Shigella dysenteriae Other Shigella Stenotrophomonas maltophilia Veillonella parvula Yeast
RI PT
0 7 14 21 0 7 14 21 28 42 0 7 14 21 28 35 42 56 0 7 14 21 0 7 14 21 0 14 21 28 35 49 63 70-84 0 7 14 21 28 35 49 63 0 7 14 21 28 35 56 84 7 14 21 28 42 56 77-210 0 14 0 7 14 21 14 21 7 14 21 42
1
Negativicutes
ACCEPTED MANUSCRIPT
15
16
17
18
19
20
21
22
23
24
25
26
SC M AN U TE D
Actinobacteria
Bacilli
Bacteroidetes
Gammaproteobacteria
EP AC C
Bifidobacterium breve Bifidobacterium longum Corynebacterium sp. Dermabacter hominis Microbacterium oxydans Micrococcus luteus Rothia dentocariosa Bacillus sp. Enterococcus faecalis Enterococcus faecium Lactobacillus casei Lactobacillus fermentum Lactobacillus gasseri Lactobacillus salivarius Lactococcus lactis Leuconostoc citreum Staphylococcus aureus Staphylococcus epidermidis Other Staphylococcus Streptococcus agalactiae Streptococcus mitis Other Streptococcus Bacteroides fragilis Citrobacter sp. Enterobacter cancerogenus Enterobacter cloacae Enterobacter hormaechei Erwinia persicina Escherichia coli Escherichia fergusonii Klebsiella granulomatis Klebsiella oxytoca Klebsiella pneumoniae Klebsiella sp. Klebsiella variicola Morganella morganii Pantoea dispersa Proteus penneri Proteus sp. Proteus mirabilis Proteus vulgaris Pseudomonas fluorescens Other Pseudomonas Serratia marcescens Serratia nematodiphila Serratia ureilytica Shigella sp. Shigella dysenteriae Other Shigella Stenotrophomonas maltophilia Veillonella parvula Yeast
RI PT
0 7 14 21 49 63 84-98 0 7 14 21 28 7 14 21 49 63 70-98 7 14 21 28 35 49 84-126 14 21 35 49 70 0 7 14 21 0 7 14 28 35 56 0 7 14 21 49 63 14 21 28 42 56 70 0 14 21 49 63 14 21 35 49 7 21 21 28
14
Supplementary Figure 1. Detection of minority species in each infant and sample.
Negativicutes