Identification and antibiotic susceptibility profiles of Ab and Pa strains
A total number of 220 Ab and 84 Pa were isolated from clinical and water samples during the two consecutive years.
Two pilot sampling campaigns were performed, in Bucharest in 2018 from which 66 Ab and 50 Pa strains were isolated (Table 1).
In 2019, the sampling campaign was extended including, in addition to Bucharest, six other cities that are representative for the main country regions, i.e., North-East (Iași, Galați), Central-West (Cluj, Timișoara) and South (Târgoviște, Râmnicu Vâlcea). A total of 154 Ab and 34 Pa resistant isolates were recovered (see Table 1).
Within the same time frame, clinical Ab and Pa clinical strains were isolated in hospital units from which wastewater samples were collected. The hospital wastewater was collected and treated by the corresponding WWTPs from the same town.
Table 1. The distribution of Ab and Pa strains selected for this study.
Year
|
Romanian region
|
Isolation source
|
Number (n) of isolated species
|
Ab
|
Pa
|
2018
|
South
|
Bucharest hospital 1 (Clinical Institute Fundeni-CIF + Institute for Cardiovascular Emergencies Prof. C.C. Iliescu- IBC CCI)
|
n=10
|
n=15
|
Bucharest hospital (H) 1 EF
|
-
|
n=11
|
Bucharest hospital (H) 2 (NIIDMB)
|
n=4
|
n=3
|
Bucharest hospital 2 EF
|
n=49
|
n=21
|
Bucharest WWTP C EF
|
n=3
|
-
|
Total 2018
|
n=66
|
n=50
|
2019
|
South
|
Bucharest hospital 2
|
n=5
|
-
|
Bucharest hospital 2 EF
|
n=21
|
-
|
Bucharest WWTP - H
|
n=8
|
-
|
Târgoviște hospital
|
n=2
|
n=1
|
Târgoviște hospital EF
|
n=5
|
-
|
Targoviște WWTP – E
|
n=30
|
-
|
Vâlcea hospital
|
n=2
|
n=1
|
Vâlcea hospital EF
|
-
|
n=5
|
Valcea WWTP – J
|
n=10
|
-
|
Central-West
|
Cluj hospital
|
n=2
|
n=2
|
Cluj hospital EF
|
-
|
n=3
|
Cluj WWTP – I
|
n=5
|
n=2
|
Timișoara hospital
|
-
|
n=2
|
Timișoara WWTP – F
|
n=6
|
n=2
|
North-East
|
Iași hospital
|
n=1
|
-
|
Iași hospital EF
|
n=1
|
n=8
|
Iași WWTP – G
|
n=15
|
n=2
|
Galați hospital
|
-
|
n=4
|
Galați WWTP – D
|
n=41
|
n=2
|
Total 2019
|
n=154
|
n=34
|
The analysis of MDR rates from the hospital to the collecting WWTP in the first sampling campaign from 2018 has revealed the following aspects: 1) clinical isolates - all Ab and the majority of Pa strains (93.3-100%) were MDR (Supplemental Tables 2 and 3); 2) hospital EF - all Ab isolates were MDR (Supplemental Table 2), while the Pa strains exhibited various MDR rates (from 25% to 100%) (Supplemental Table 3); 3) WWTP C, collecting the two hospital EFs – all Ab isolates were MDR (Supplemental Table 2).
In 2019, the MDR rates from hospital to the collecting WWTP were as follows: 1) clinical isolates – with one exception, all Ab isolates were MDR (Supplemental Table 2), while the Pa strains have shown a high variation of MDR rate (from 0 to 100%) with the geographical location (Supplemental Table 3); 2) hospital EF – all Ab isolates were MDR (Supplemental Table 2), while the Pa strains exhibited various MDR rates (from 0% to 100%) (Supplemental Table 3); 3) WWTPs, collecting the sewages of the sampled hospital units effluents – the MDR resistance rates varied from 0 to 100% for both Ab and Pa strains.
Table 2 and 3. AR profiles in clinical and wastewater Ab and Pa strains.
CP and ESBL encoding genes in clinical and environmental Ab and Pa strains
a. Profiles of CP and ESBL genes from the clinical to the aquatic environment in 2018 versus 2019
The Ab strains from the clinical settings to the WWTP effluent and receiving river exhibited different profiles of CP and ESBLs in the two consecutive years, i.e.: 1) clinical strains – Ab strains recovered in 2018 were OXA-23 and OXA-24 producers (64.28/42.85%) and only 7.14% were positive for blaTEM, while in 2019, 41.66% of all intra-hospital Ab strains were OXA-23 and OXA-24 producers, 25% were positive for blaSHV, 16.66% for blaVIM and blaVEB and 8.33% for blaTEM and blaGES (Supplemental Table 2); 2) hospital sewage – the identified carbapenem-hydrolyzing class D β-lactamases (CHLDs) were represented in the two consecutive years by OXA-23 (67.34/96.29%) and OXA-24 (2.04/0%), MBLs by VIM (4.08/14.81%) and ESBLs by TEM (20.41/0%), SHV and PER (each one 0/3.70%), GES (0/18.51%) (Supplemental Table 2); 3) WWTPs – some of the enzymes were common for the strains isolated in 2018 and 2019 [i.e. OXA-24 (57.14/10.25%); OXA-23 (42.85/53.84%); TEM (28.57/10.89%); SHV (14.28/0.64%)], while some other were different [i.e. NDM in Ab strains from 2018 (14.28%) and GES (19.87%); VEB (8.97%); CTX-M (2.56%) and PER (0.64%) in 2019].
The CP and ESBL genes identified in the Pa isolates during the transmission chain from the clinical sector to the WWTP effluent and receiving river in the two consecutive years were the following: 1) clinical strains – the Pa strains were positive for blaIMP (66.66/0%), blaVIM (33.33/20%) and blaVEB (38.88/30%); 2) hospital sewage – in the two consecutive years the following MBLs were identified in the Pa strains: VIM (9.37/18.75%); IMP (9.37/0%); NDM (0/12.5%); while the ESBLs were represented by GES (6.25/6.25%); VEB (43.75/12.5%) and TEM (0/31.25%); 3) WWTPs – only the Pa strains isolated in 2019 were positive for ESBL encoding genes [blaTEM (16.66%); blaGES and blaVEB (8.33% each)] (Supplemental Table 3).
- Geographic distribution of CP and ESBL genes in clinical and water Ab and Pa isolates
The 2018 pilot study was limited to the Bucharest region, and then it was extended during 2019 to other regions of the country, allowing us to perform a comparative analysis regarding the geographic distribution of the CP and ESBL encoding genes in the Ab and Pa strains.
Regarding the Ab strains, the Southern region shows the largest spectrum and positivity of CP and ESBL encoding genes, both in clinical [i.e., blaOXA-23 (44.44%); blaOXA-24 (33.33); blaVIM (22.22%); blaTEM and blaSHV (11.11%)] and the aquatic isolates [hospital sewage: blaOXA-23 (100%); blaGES (19.23%); blaVIM (15.38%); blaVEB and blaSHV (3.84%) and WWTPs: i.e. blaOXA-23 (44.44%); blaGES (32.09%); blaTEM (18.51%); blaOXA-24 (14.81%); blaVEB (13.58%); blaCTX-M (4.93%) and blaSHV (1.23%)].
Central-Western region revealed the presence of the following CP and ESBL encoding genes: 1) in clinical settings all Ab strains were blaVEB positive; 50% were blaOXA-23 and blaOXA-24 positive; 2) the aquatic isolates recovered from the two sampled WWTPs were blaOXA-23 and blaVEB positive (23.07% each of them).
In the North-Eastern region the CP and ESBL identified in Ab strains from the hospitals to the WWTP were the following: 1) all clinical Ab strains were OXA-24 producers; while in the sampled WWTPs the Ab strains harbored OXA-23 (72.58%); OXA-24 (6.45%); TEM (3.22%) and PER (1.61%) (Fig. 2).
The geographical distribution of the CP/ESBLs found in Pa strains isolated from intra-hospital infections, the hospital sewage tank and the sampled WWTP from the corresponding city: in the Southern region 50% of nosocomial Pa strains were VEB producers, while the wastewater Pa strains harbored blaVEB (40%), blaNDM (40%) and blaGES (20%); in the North-Eastern region, 50% of clinical Pa strains were VEB producers; 62.5% of the hospital sewages strains were positive for blaTEM; 50% respectively 25% of the WWTPs were blaVEB and blaTEM positive. The Pa strains from the Central-Western regions revealed different CP/ESBLs in clinical/hospital sewage (VIM) and in the sampled WWTPs (14.28% were GES positive) (Fig.3).
c. WGS analysis of clinical and wastewater Ab and Pa isolates
Acquired resistome
The WGS analysis allowed detailed insights in the acquired resistome of the selected strains.
In case of nine Ab strains recovered in 2018 in Bucharest from intra-hospital infections (n=2), hospital sewage EF (n=5) and the corresponding WWTP EF (n=2) the WGS demonstrated the presence of OXA-72 and OXA-23 in the IN and the EF of the collecting sewage tank and of genes encoding aminoglycoside modifying enzymes (AMEs) i.e. aph(3')-VIa, ant(3'')-IIa, sulphonamides (sul1) and class 1 integrons (qacE∆1), in all investigated samples (Table 4).
Table 4. ARGs in clinical and wastewater Ab strains isolated in Bucharest in 2018.
The WGS analysis of 12 Ab clinical strains isolated in 2019 from two hospital units (Bucharest H2 and Târgoviște), the collecting sewage tank of Bucharest hospital 2 and the sampled corresponding WWTPs from Bucharest (H) and Târgoviște (E) revealed the presence of blaOXA-23 in all sampled points from Bucharest. The presence of both blaOXA-23 and blaOXA-72 was shown in clinical Ab strains from Târgoviște hospital unit and in the corresponding WWTP (i.e., of blaOXA-23 in EF and of blaOXA-72 in the RS) (Supplemental Table 5). Similarly, the AMEs (i.e. aph(6)-Id, ant(3'')-IIa), sulphonamides resistance (sul 1) and class 1 integrons (qacE∆1) were present in all sampled sites from Bucharest. The WGS of nine Ab clinical and wastewater strains from Eastern and Northern Romania revealed a high diversity of ARGs, with differences between different cities, i.e.: blaOXA-23, aph(6)-Id, aph(3'')-Ib, armA, sul 1, tet(B), mph(E), msrE and qacE∆ in Galați WWTP (D) and blaOXA-72, aph(3')-Ia, aac(3)-I, aadA1, ant(2'')-Ia, sul1 and qacE∆1 in Iași WWTP (G) (Supplemental Table 5).
Table 5. ARGs in clinical and wastewater Ab strains isolated in 2019.
The WGS analysis of the acquired resistome of the 10 Pa strains recovered in 2018 from two Bucharest hospital units and their collecting sewage tanks demonstrated the dissemination of CP blaIMP-13 and of genes encoding AMEs (aph(3')-IIb, ant(2'')-Ia), fosfomycin (fosA), phenicols (catB7, bcr1) and trimethoprim-sulfamethoxazole (sul1) resistance in the case of one Bucharest hospital and its effluent (Table 6). In the case of the second investigated hospital and the corresponding sewage tank, there has been noticed the presence of genes encoding AMEs (aph(3')-IIb) and phenicols (catB) (Table 6).
Table 6. ARGs in clinical and wastewater Pa strains isolated in Bucharest in 2018.
The ARGs distribution in four Pa strains recovered in 2019 from Northern and Eastern Romania revealed the presence of ESBL encoding genes (blaTEM-40, blaVEB-9), AMEs encoding genes (aac(6')-II, aadA1, aph(3')-IIb) as well as determinants of resistance to fosfomycin (fosA), phenicols (catB7, bcr1), tetracycline [tet(A)] and class 1 integrons (qacE∆1) in clinical and wastewater samples. Regarding the six Pa clinical and wastewater strains from Central and Western regions of Romania, the presence of resistance genes encoding for fosfomycin (fosA) and phenicols (catB7, bcr1) has been observed in Pa strains from almost all sources (Table 7).
Table 7. ARGs in clinical and wastewater Pa strains isolated from North - Eastern and Central - Western Romania in 2019.
Ab and Pa strains molecular phylogeny
The Ab strains phylogeny, MLST profiles and SNP analyses highlight the relatedness between the strains which were divided in six groups.
In group I there were included clinical isolates from Iași, Târgoviște and Cluj and in the effluent and downstream of Bucharest WWTP C which belonged to ST636. The SNP analysis suggests the similarities between a clinical isolate sampled in 2019 and the two aquatic strains from this group, sampled in 2018 (harboring 51 and 60 SNPs of the clinical strain); the group II included most of the strains (clinical and wastewater isolates from all investigated regions) belonging to ST2, a successful widespread Ab clone. SNP analyses suggest the relatedness between clinical and hospital wastewater strains (189-201+202+203+204, thus less than 20 SNPs between clinical and hospital sewage), and more intriguing, the relatedness between strains sampled in clinics and urban WWTPs (169-184 =15 SNPs, 187-18033O3 =22 SNPs), suggesting the dissemination of clinical Ab ST2 strains in the wastewaters (Table 8); the group III was represented by strains isolated from Bucharest hospital and its collecting sewage tank isolates belonging to ST492, all of them being closely related (26 SNPs); in group IV clinical strains and aquatic strains isolated from Bucharest belonging to ST79 were included; these strains were less related (577 SNPs); the group V included wastewater strains isolated from South Romania belonging to ST1, the isolates being also less related (>700 SNPs); group VI included only environmental strains from the Northern region of the country belonging to ST155 and two related novel STs (Fig. 4).
Table. 8. Matrix representation of calculated SNPs distances between the closely related ST2 strains.
Strain code/Isolation source
|
187
BL3 Ab11
|
169
DF0462 Ab
|
18033O3
WWTP C DO
|
182
19001 CNE1 Ab
|
183
19003 CNE4 Ab
|
184
19002 CNE1 Ab
|
185
19011 CNE6 Ab
|
186
19012 ENE2 Ab
|
189
BL3 Ab5
|
201
18003 CN10 Ab
|
202
18002 CN6 Ab
|
203
18001 CN7 Ab
|
187
BL3Ab11 – clinical Bals
|
|
|
|
|
|
|
|
|
|
|
|
|
169
DF0462 Ab - clinical Cluj
|
110
|
|
|
|
|
|
|
|
|
|
|
|
18033O3 WWTP C DO
|
75
|
140
|
|
|
|
|
|
|
|
|
|
|
182
19001 CNE1 Ab – WWTP D IN
|
580
|
140
|
521
|
|
|
|
|
|
|
|
|
|
183
19003 CNE4 Ab – WWTP D EF
|
589
|
91
|
542
|
64
|
|
|
|
|
|
|
|
|
184
19002 CNE1 Ab – WWTP D RS
|
616
|
109
|
555
|
73
|
129
|
|
|
|
|
|
|
|
185
19011 CNE6 Ab –Bals sewage IN
|
843
|
417
|
798
|
465
|
257
|
463
|
|
|
|
|
|
|
186
19012 ENE2 Ab –Bals sewage EF
|
844
|
418
|
800
|
466
|
257
|
463
|
2
|
|
|
|
|
|
189
BL3 Ab5 – clinical Bals
|
732
|
273
|
704
|
423
|
348
|
424
|
353
|
337
|
|
|
|
|
201
18003 CN10 Ab - Bals sewage IN
|
921
|
456
|
895
|
593
|
521
|
588
|
345
|
338
|
35
|
|
|
|
202
18002 CN6 Ab - Bals sewage IN
|
915
|
449
|
888
|
523
|
523
|
589
|
344
|
341
|
48
|
6
|
|
|
203
18001 CN7 Ab - Bals sewage EF
|
922
|
458
|
897
|
525
|
525
|
592
|
348
|
335
|
37
|
4
|
4
|
|
204
18004 CN10 Ab - Bals sewage EF
|
934
|
461
|
907
|
525
|
525
|
593
|
334
|
329
|
33
|
4
|
6
|
6
|
Conjugation assays revealed a clear dissemination of the same circulating clones from the hospital units into different aquatic compartments [i.e. ST2 encountered in Bucharest hospital unit and the corresponding sampled WWTP in 2019 carrying the same CP encoding gene (blaOXA-23 or blaOXA-72) in Ab strains; ST2 carrying blaOXA-72 gene in Ab strains from Târgoviște hospital unit and blaOXA-23 in the EF of the corresponding WWTP E].
The Pa strains (Fig. 5) were also grouped in six phylogenetic groups: group I included wastewater isolates from Timișoara, and one clinical strain from the Bucharest hospital that belonged to three singleton STs (ST252, ST254 and ST395); group II comprises clinical strains from Central Romania (Cluj hospital) and one collecting sewage tank from a hospital unit in Bucharest that belonged to the epidemic clone ST233; group III included strains isolated from Iași hospital sewage and its collecting WWTP G belonging to ST640; group IV (Bucharest hospital unit collecting sewage tank isolate belonging to ST621); group V (wastewater strains from central Romania that belonged to ST620); group VI contained the majority of the strains (clinical and wastewater isolates from South and East Romania) that belonged to the widespread ST357 and one unknown ST (Fig. 5).
The spread of the same epidemic clone has been noticed, from the hospital unit into the natural aquatic recipient from the East part of the country – Galați hospital unit and the receiving WWTP D (ST357 carrying the blaVEB-9 ESBL encoding gene). As revealed by the SNPs analyses, the two ST640 strains (Iași hospital and sludge form the Iași urban WWTP) represent the same clone (32 SNPs, the proposed threshold being 37 SNPs). The other Pa strains are more diverse, even within the same clone the strains being more distantly related (>100 SNPs); this fact was also suggested by the difference between the core genome (4716 genes) and the pan genome (12395 genes) calculated for all the Pa strains included in this study.