Clonal Spread and Intra- and Inter-Species Plasmid Dissemination Associated With Klebsiella pneumoniae Carbapenemase-Producing Enterobacterales During a Hospital Outbreak in Barcelona, Spain

Objectives: The study aimed to characterize the clonal spread of resistant bacteria and dissemination of resistance plasmids among carbapenem-resistant Enterobacterales at a tertiary hospital in Catalonia, Spain. Methods: Isolates were recovered from surveillance rectal swabs and diagnostic samples. Species identification was by matrix-assisted laser desorption ionization-time time of flight mass spectrometry (MALDI-TOF MS). Molecular typing was performed by pulsed-field gel electrophoresis (PFGE) and multi-locus sequence typing (MLST). Antimicrobial susceptibility was assessed by gradient-diffusion and carriage of bla genes was detected by PCR. Plasmid typing, conjugation assays, S1-PFGE studies and long-read sequencing were used to characterize resistance plasmids. Results: From July 2018 to February 2019, 125 Klebsiella pneumoniae carbapenemase (KPC)-producing Enterobacterales were recovered from 101 inpatients from surveillance (74.4%) or clinical samples (25.6%), in a tertiary hospital in Barcelona. Clonality studies identified a major clone of Klebsiella pneumoniae belonging to sequence type ST15 and additional isolates of K. pneumoniae, Escherichia coli and Enterobacter sp. from different STs. All isolates but one carried the blaKPC–2 allelic variant. The blaKPC–2 gene was located in an IncFIIk plasmid of circa 106 Kb in a non-classical Tn4401 element designated NTEKPC-pMC-2-1. Whole-genome sequencing revealed different rearrangements of the 106 Kb plasmid while the NTEKPC-pMC-2-1 module was highly conserved. Conclusion: We report a hospital outbreak caused by the clonal dissemination of KPC-producing ST15 K. pneumoniae but also the intra- and inter-species transmission of the blaKPC–2 gene associated with plasmid conjugation and/or transposon dissemination. To our knowledge, this is the first report of an outbreak caused by KPC-producing Enterobacterales isolated from human patients in Catalonia and highlights the relevance of surveillance studies in the early detection and control of antibiotic resistant high-risk clones.


INTRODUCTION
The rapid emergence of bacterial pathogens presenting resistance to multiple antimicrobial agents (MDR) together with the decreasing trend in the development of new antimicrobial compounds constitute an extremely serious threat to public health (Roca et al., 2015). Resistance rates are particularly alarming among Gram-negative bacteria (GNB) as available treatment options are severely impaired. In 2016 a report commissioned by the UK Prime minister, known as the "O'Neill report, " attributed 700.000 annual human deaths worldwide to microbial infections caused by antimicrobial resistant (AMR) pathogens (O'neill, 2016). The same report estimated that this figure would rise to 10 million annual deaths by 2050 unless there is a global effort to tackle AMR. Likewise, organizations such as the US Centers for Disease Control and Prevention (CDC), the European Centre for Disease Prevention and Control (ECDC) and the World Health Organization (WHO) are considering infections caused by MDR bacteria as an emergent global disease and a major public health problem (Roca et al., 2015). According to their mortality, health-care and community burden, prevalence of resistance and treatability, carbapenem-resistant Enterobacterales have been included into the 2017 WHO priority list of antibiotic-resistant bacterial pathogens and currently constitute one of the major public health threats worldwide (Tacconelli et al., 2017).
Among carbapenem-resistant Enterobacterales, Klebsiella pneumoniae stands out as a formidable nosocomial pathogen causing several infections associated with high mortality rates (Tumbarello et al., 2012). In Europe, carbapenem-resistant K. pneumoniae accounted for almost 16,000 infections and more than 2,000 deaths in 2015 (Cassini et al., 2018). Carbapenem-resistance is usually associated with the carriage of genes encoding carbapenem-hydrolyzing enzymes, most of which are commonly located within conjugative plasmids and, therefore, are easily disseminated among nosocomial pathogens (Bonomo et al., 2017). Predominant carbapenemases in Europe include OXA-48 oxacillinase, New Delhi metallo-β-lactamase (NDM) and Klebsiella pneumoniae carbapenemase (KPC), among others. In Spain, OXA-48 is currently the predominant mechanism associated with the dissemination of carbapenemresistant Enterobacterales (Grundmann et al., 2017). NDMproducing bacteria are reported less frequently in Spain but their presence is increasing and associated with a few clonal lineages (Marí-Almirall et al., 2021) while KPC-producing Enterobacterales are also sporadic in Spain, and only a few outbreaks have been reported in Madrid and Andalusia (Oteo et al., 2016).
Klebsiella pneumoniae carbapenemase is likely one of the most common mechanisms of resistance to carbapenems in K. pneumoniae, and the worldwide spread of the bla KPC gene was initially associated with the dissemination of a few K. pneumoniae clonal lineages within the ST258 clonal group (CG258), a particular plasmid backbone (pKpQIL) and a highly conserved Tn4401 transposon (Chen et al., 2014). More recently bla KPC has been found on a variety of clonal lineages and plasmids from different incompatibility groups (Wyres et al., 2020), and although it is usually associated with a composite Tn4401 transposon, some structural variations have also been described (Shen et al., 2009;Naas et al., 2012;Chmelnitsky et al., 2014). In this retrospective study we have examined the rapid dissemination of KPC-producing Enterobacterales within a tertiary hospital in Barcelona and we have investigated the inter-and intra-species spread of carbapenem resistance.

MATERIALS AND METHODS
This study included 125 carbapenem-resistant isolates of Enterobacterales collected from July 2018 to February 2019 at one tertiary hospital in Barcelona, Spain. Strains were recovered from surveillance and clinical samples (Supplementary Table 2). Identification at species level was performed by matrix-assisted laser desorption ionization-time time of flight mass spectrometry (MALDI-TOF MS) in a Microflex LT benchtop instrument (Bruker Daltonics) operated in linear positive mode.
Production of KPC, OXA-48-like, VIM, IMP or NDM carbapenemases was detected with the NG-Test R CARBA5 (NG-Biotech, France) and confirmed by PCR using previously described primers and conditions (Queenan and Bush, 2007;Solé et al., 2011;Bogaerts et al., 2013). Amplification products were purified from agarose gels and sent for Sanger sequencing (Genewiz, Germany) whenever necessary. The allelic identity of all genes was determined by sequence alignment with reference sequences retrieved from public repositories (PRJNA313047, last accessed June, 2021).

Clonal Relatedness
Clonality was studied by pulsed-field gel electrophoresis (PFGE) using XbaI (New England BioLabs Inc., United States) genomic digestions and a CHEFF-DRIII system (Bio-Rad, Spain) (Marí-Almirall et al., 2021). Molecular patterns were analyzed with InfoQuest TM FP-v.5.4 (Bio-Rad, Spain) and the unweighted pair group method with arithmetic mean (UPGMA) to create dendrograms based on Dice's similarity coefficient. Bandwidth tolerance and optimization values were set at 2% and isolates were considered within the same PFGE cluster (pulsotype) if their Dice similarity index was >85%.
Multi-locus sequence typing (MLST) was performed according to the Pasteur scheme for K. pneumoniae and the Achtman scheme for E. coli (Diancourt et al., 2005;Wirth et al., 2006).

Plasmid Analysis
Conjugation assays were carried out using E. coli MC1061 resistant to rifampicin and sodium azide as the recipient strain. Transconjugant strains were selected in LB agar plates supplemented with 1 mg/L of meropenem and 100 mg/L of sodium azide (Sigma-Aldrich, Spain). The location of resistant genes was determined by S1-nuclease digestion (New England BioLabs Inc., United States) followed by PFGE and Southern blot hybridization with digoxigenin-labeled PCR-probes against bla KPC (Bogaerts et al., 2013). Plasmid incompatibility groups were determined using the PBRT-2.0 kit (Diatheva, Italy) (Carattoli et al., 2005). The genetic environment of bla KPC was determined by PCR using primers listed in Supplementary  Table 1 and confirmed by long-read sequencing.

Bacterial Isolation and Clonal Relatedness
In mid-July 2018, five carbapenem-resistant K. pneumoniae isolates were recovered from surveillance samples (rectal swabs) upon the implementation of a "zero-resistance" surveillance program to screen all patients admitted to intensive care units within the hospital. All five isolates produced a KPC-type enzyme. During the following weeks, additional surveillance samples and an increasing number of clinical samples flagged positive for KPC-producing Enterobacterales. By the end of August, 87 KPC-producing isolates had been recovered and infection control measures were initiated, including exhaustive room cleaning twice a day, reinforcing surveillance measures, skin cleaning with 2% chlorhexidine wipes, and oral decolonization of patients (amikacin, colistin, nystatin) (Pellicé et al., 2021). During the following months the number of KPC-producing isolates rapidly decreased and the outbreak was considered eradicated by November 2018, although some isolates were still recovered up until February 2019 (Figure 1). Overall, 125 isolates (112 K. pneumoniae, 12 E. coli, and 1 Enterobacter sp.) from 101 patients were collected and included for further studies. Thirty-two isolates (25.6%) were detected from diagnosis samples and 93 (74.4%) were isolated from surveillance samples (Supplementary Table 2). The outbreak involved the screening of 2031 patients. The Index case could not be identified.
One strain from each cluster of either K. pneumoniae or E. coli were selected for further phenotypic and genotypic characterization. A total of 24 isolates were eventually selected, 13 K. pneumoniae, 10 E. coli isolates, and 1 Enterobacter sp. (Table 1).

Phenotypic and Molecular Characterization of Resistance
The MICs of selected isolates are shown in Table 1. All isolates were non-susceptible to cephalosporins and carbapenems, but were susceptible to ceftazidime-avibactam, fosfomycin and colistin, except for MC-2-285 and MC-22-164, that were highly resistant to fosfomycin.
Sanger sequencing identified the bla KPC-2 variant in all isolates except for the MC-2-251 and MC-2-387 isolates, that carried bla KPC-3 (note that MC-2-387 also carried bla KPC-2 ). The genes for CTX-M-group 1 enzymes were also detected in 7 of the 24 strains, while CTX-M-group 2, 9, and 25 were only found in a few strains (Supplementary Table 2). CTX-M-group 8 was not detected. All isolates were negative for the presence of bla OXA-48 , bla NDM , bla VIM , or bla IMP . S1-nuclease-PFGE profiles of selected strains revealed the presence of different plasmids (according to size) that contained the bla KPC gene. Eight of the selected K. pneumoniae strains harbored bla KPC-2 within a plasmid of circa 110 kb in size (plasmid type A, Figure 2), including the ST15 strains. The two ST258 strains MC-2-251 and MC-2-387 carrying bla KPC-3 , however, showed hybridization signals with plasmids of circa 80 Kb and 120 Kb, respectively (plasmid types C and D, Figure 2), and the single ST307 strain (MC-2-230) carried bla KPC-2 within a plasmid of circa 300 kb (plasmid type E, Figure 2). Likewise, the Enterobacter sp. isolate harbored bla KPC-2 in a plasmid of approximately 400 Kb (plasmid type B, Figure 2). Among selected E. coli isolates, eight isolates also carried bla KPC-2 in plasmids of similar size as that of type A plasmids in ST15 K. pneumoniae (110 Kb), although the MC-2-315 isolate presented a second hybridization signal at circa 180 Kb (plasmid type F, Figure 2). Two additional strains, MC-2-240 and MC-2-304, carried bla KPC-2 in plasmids of different sizes. MC-2-240 showed a hybridization band at approximately 190 Kb, and MC-2-304 at 170 Kb (plasmid types G and H, respectively, Figure 2).
Notably, all E. coli isolates but one and the single Enterobacter sp. isolate recovered during the outbreak, originated from patients that previously carried a ST15 KPC-2-producing K. pneumoniae. On the other hand, only three out of the nine non-ST15 K. pneumoniae isolates were recovered from patients also co-carrying a ST15 isolate (Supplementary Figure 4 and Supplementary Table 2). Plasmid replicon typing identified the presence of multiple plasmids in some of the strains ( Table 1), but all strains were positive for a replicon belonging to the IncFIIk incompatibility group. Conjugation assays using the ST15 isolate MC-2-1 as donor and the E. coli strain MC1061 as recipient, showed that the acquisition of a bla KPC-2 gene was associated with the transfer of an IncFIIk plasmid (not shown).

Plasmid Sequencing
Ten isolates, 5 K. pneumoniae, 4 E. coli, and 1 Enterobacter sp., representative of strains carrying all the different plasmid types identified with hybridization probes, were further selected for long-read plasmid sequencing. Plasmid sequencing identified an IncFIIk plasmid of 106,412 bp carrying bla KPC-2 in the ST15 K. pneumoniae MC-2-1 isolate, designated as pMC-2-1 (Supplementary Figure 3). The bla KPC-2 gene was the only antibiotic resistance gene present in the pMC-2-1 plasmid that also harbored several genes involved in conjugative transfer. The presence of plasmid-associated virulence genes was not detected in plasmid pMC-2-1 (Supplementary Figure 3). An almost identical plasmid of 106,462 bp was found in the E. coli strain MC-2-207, and highly similar IncFIIk plasmids of 109,489 and 101,915 bp were located in the K. pneumoniae and E. coli isolates MC-2-316 and MC-2-315, respectively, both recovered from the same patient ( Figure 3A), in good agreement with the carriage of bla KPC-2 in a plasmid of circa 110 Kb (plasmid type A), as identified by S1-digestion. Unfortunately, plasmid type F of circa 180 Kb, also present in strain MC-2-315 according to S1-digestion, could not be identified by long-read sequencing analysis.
The K. pneumoniae isolate MC-2-230, the Enterobacter sp. isolate MC-2-362 and the E. coli isolate MC-2-304, carried bla KPC-2 within IncFIIk plasmids of 281,298, 407,992, and 170,594 bp, respectively, much larger than that of ST15 strains and in good agreement with results from S1-nuclease digestion. Additional replicon types from IncN and/or IncU incompatibility groups were also identified in these plasmids. Interestingly, the entire 106 Kb sequence that made up for the pMC-2-1 plasmid was also found inserted within the plasmids harboring KPC in strains MC-2-304 and MC-2-362, and the 281 Kb plasmid from strain MC-2-230 also carried a 67 kb fragment from pMC-2-1 ( Figure 3B). On the other hand, the two ST258 K. pneumoniae isolates (MC-2-251 and MC-2-387) harbored bla KPC-3 within IncFIIk plasmids of 78,515 and 120,395 bp, and there was no similarity at all between these two plasmids and those of ST15 strains as there was no resemblance between them either. Arrows are proportional to the lengths of the genes and oriented in the direction of transcription. Red arrows represent resistance genes, orange arrows represent full-length transposon-related genes and ISs, dark blue arrows represent partial or truncated transposon-related genes and ISs, green arrows indicate genes involved in plasmid conjugation and light blue arrows shown genes related to plasmid replication/maintenance. Blue-gray arrows show putative or hypothetical genes and plum arrows show genes involved in other functions. The inner circle shows the GC content. The fully annotated sequence of plasmid pMC-2-1 is shown in Supplementary Figure 3. (B) Schematic representation of the larger IncFIIk plasmids carrying bla KPC-2 recovered from strains MC-2-230, MC-2-304, and MC-2-363. The location of the genetic structure NTE KPC -pMC-2-1 harboring bla KPC-2 is shown as well as the region matching the pMC-2-1 sequence. The inner circle shows the GC content.
In the E. coli strain MC-2-240 we only detected bla KPC-2 within a 25,058 bp plasmid showing 100% similarity with pMC-2-1, and a similar plasmid of 28,498 bp was also detected in strain MC-2-387, which also carried bla KPC-3 in the 120,395 bp plasmid ( Figure 3A).
Notably, in all plasmids bla KPC-2 was not located within the canonical Tn4401 element but inside an IS26-based composite transposon of roughly 17 Kb containing an ISKpn27-bla KPC-2 -ISKpn6-korC core structure as well as Tn3-associated sequences upstream from ISKpn27 (Figure 4). This rearrangement was tentatively designated as NTE KPC -pMC-2-1. The genetic structures surrounding the bla KPC-3 gene in both ST258 strains, however, did match that of a classical Tn4401 element (Figure 4). Specific primers to amplify the structures associated with NTE KPC -pMC-2-1 were designed (Supplementary Table 1 and Supplementary Figure 5) and used to verify the carriage of either the canonical Tn4401 transposon or the NTE KPC -pMC-2-1 variant in all 125 isolates of the outbreak. Interestingly, NTE KPC -pMC-2-1 was identified in all isolates but in the ST258 K. pneumoniae isolate MC-2-251.
In addition, all plasmid sequences were checked for the presence of plasmid-mediated virulence genes producing salmochelin, aerobactin, the hypermucoviscosity factor and/or other virulence genes (Russo et al., 2018), but none of the plasmid sequences carrying bla KPC harbored any known virulence determinant.

DISCUSSION
The emergence of KPC-producing Enterobacterales has been increasing worldwide during the last decades (Lee et al., 2019). Since its first description in 1996 in the United States, KPC has spread through the globe, becoming endemic in areas such as the United States, Israel, Greece, and Italy, and sporadic spread being reported in additional European countries as well as in the Asia-Pacific region (Lee et al., 2019). Many of the descriptions from European countries have been related to patients traveling to endemic areas.
In Spain, the first KPC report dates back to 2009 and it was associated with the rapid spread of a ST384 K. pneumoniae strain carrying bla KPC-3 within a Tn4401a structure (Curiao et al., 2010). KPC-2 and KPC-3-producing Enterobacterales were later reported in the central region of Spain and Andalusia mainly associated with K. pneumoniae isolates belonging to the CG258 or ST101 (López-Cerero et al., 2014;Porres-Osante et al., 2014;Oteo et al., 2016). In Catalonia, a single KPC-3-producing K. pneumoniae isolate belonging to ST258 was recovered from a human patient in 2015 (Piedra-Carrasco et al., 2018) but, to our knowledge, our study constitutes the first hospital outbreak caused by KPC-producing Enterobacterales in this region.
Here we report an outbreak caused by KPC-producing K. pneumoniae and E. coli in a tertiary hospital in Barcelona that initiated during the summer of 2018. It involved 125 Enterobacterales isolates recovered from surveillance (74.4%) or diagnosis (25.6%) samples. Outbreak identification was possible upon the implementation of an active surveillance program in ICU patients, but it is likely that KPC-producing K. pneumoniae isolates had already been circulating in asymptomatic fecal carriers for some time. The outbreak was caused by the rapid spread of a carbapenem-resistant ST15 K. pneumoniae strain carrying bla KPC-2 in an IncFIIk plasmid of approximately 106 Kb and located within a non-Tn4401 genetic element (NTE KPC -pMC-2-1).
Nevertheless, upon the initial stage of clonal spread, we speculate that the IncFIIk plasmid was successfully transferred first to E. coli isolates in patients either infected or colonized with the original ST15 strain, but later to other strains of K. pneumoniae or even Enterobacter spp. (Figure 1 and  Supplementary Figure 4). This is supported by the fact that in at least 14 instances the same patient co-carried a bla KPC-2 -ST15 K. pneumoniae strain together with either a bla KPC-2 -K. pneumoniae belonging to a different sequence type or a bla KPC-2 -E. coli or Enterobacter sp. strain, but also because all plasmids carrying bla KPC-2 in K. pneumoniae or E. coli strains shared high similarity with the IncFIIk plasmids from the ST15 K. pneumoniae strains, including the NTE KPC -pMC-2-1 structure surrounding bla KPC-2 . Some genetic rearrangements, however, seem to have occurred, mainly in non-K. pneumoniae strains where the 106 Kb IncFIIk plasmid either co-integrated with other plasmids or suffered extensive genetic reduction. Therefore, intra-and inter-species dissemination of resistance also contributed to the spread of the outbreak. KPC transfer to other K. pneumoniae strains was not as heavily associated with co-carriage within the same patient, as opposite to inter-species dissemination, but such finding may also reflect a selection bias in the microbiology laboratory. Interestingly though, the clonal dissemination of non-ST15 strains was not detected. Recently, San Millán and co-workers proposed a similar hospital transmission dynamic for pOXA-48, where patient-to-patient transmission was tightly associated with the dissemination of a particular high-risk clone, while intra-and inter-species transmission of the plasmid was linked to concurrent gut colonization (León-Sampedro et al., 2021). As in the case of OXA-48, this genetic exchange represents an opportunity for the resistance gene to rearrange and shuffle into new plasmids and/or hosts, some of which may become more successful (Millan, 2018).
The genetic structures associated with the core structure of NTE KPC -pMC-2-1 identified in this work (ISKpn27-bla KPC-2 -ISKpn6-korC) had already been reported by other studies and seem to constitute a common rearrangement associated with bla KPC-2 mostly in isolates from China (Wang et al., 2015;Zhang et al., 2017). In Spain though, this structure has also been identified in IncP-6 plasmids from Citrobacter freundii, Enterobacter cloacae, and Klebsiella oxytoca but, to our knowledge, have never been associated with K. pneumoniae isolates of human origin, most likely suggesting a recent acquisition (Yao et al., 2017;Pérez-Vazquez et al., 2019).
Two sporadic K. pneumoniae strains belonging to the clonal group CG258 and carrying the bla KPC-3 gene within a canonical Tn4401 in two different non-related plasmids were also detected in the study and, most likely, constituted an independent event unrelated to the spread of the outbreak. Nevertheless, one of such strains also managed to acquire a partial plasmid sequence containing the entire NTE KPC -pMC-2-1 element.
The identification of a carbapenem resistant ST307 K. pneumoniae isolate in this study is also worth mentioning, as strains from this epidemic clonal group have recently been responsible for a hospital outbreak in Germany associated with multidrug resistance but also with the hypervirulent Klebsiella pneumoniae phenotype (hvKp). The hvKp phenotype was identified among ST307 isolates upon the acquisition of several plasmid-mediated virulence genes that merged with a resistance plasmid, hence creating a mosaic plasmid carrying both resistance and virulence genes (Heiden et al., 2020). In our study the ST307 strain (MC-2-230) carried bla KPC-2 in a plasmid of 280 Kb also likely resulting from the merging of two different plasmids and, hence, the potential carriage of plasmid-mediated virulence genes was investigated. Fortunately, none of the plasmid-mediated virulence factors associated with the hvKp phenotype were present in plasmid pMC-2-230 nor in any other plasmid sequenced in this study. Likewise, none of the K. pneumoniae strains showed a hypermucoid phenotype, which is characteristic (although not exclusive) of hvKp (Lan et al., 2021).
We acknowledge several limitations in our study. First, WGS was performed under a single long-read sequencing approach, and we acknowledge that the use of a hybrid approach would have allowed for additional and more accurate comparisons. In addition, only a selected group of isolates were sequenced and sequence similarity was, therefore, assumed for the remaining isolates on the basis of PFGE, MLST and conventional PCR data. Unfortunately, further WGS analyses were beyond our possibilities but we expect that results from this study will contribute to a better implementation of WGS pipelines in our institution.

CONCLUSION
We report a hospital outbreak caused by the clonal dissemination of KPC-producing ST15 K. pneumoniae mainly among colonized carriers but also by the intra-and inter-species transmission of the bla KPC-2 gene associated with plasmid conjugation and/or transposon dissemination. The ST15 clonal lineage is considered a high-risk clone and has been associated with KPC-2-producing isolates in Bulgaria and Vietnam (Markovska et al., 2015;Berglund et al., 2019), and it has also been reported in Portugal and Italy (Rodrigues et al., 2016;Fasciana et al., 2019). In Spain, ST15 has been reported but only associated with the production of OXA-48 (Madueño et al., 2017). Nevertheless, ST15 has been attributed with a high potential for horizontal gene acquisition and dissemination and it is critical that active surveillance strategies are prolonged over time to allow for the rapid detection and eradication of these highly resistant and virulent clones (Berglund et al., 2019).

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary Material.

AUTHOR CONTRIBUTIONS
MM-A, NF, and MF contributed to the conception, design, and implementation of the study, acquisition of laboratory and clinical data, analysis of the results, drafting the manuscript, and approval of the final version of the manuscript. CC, JViñ, ER, AC, and LM contributed to the acquisition of laboratory data, analysis of the results, and review and approval of the final version of the manuscript. MP, AV, IC, LR-S, GS, and AD contributed to the design of the study, acquisition of clinical data, analysis of the results, and review and approval of the final version of the manuscript. OF, PC, FB, FM, JM, and ÁS contributed to the design of the study, analysis of the results, and review and approval of the final version of the manuscript. CP, JVil, and IR contributed to the conception, design, and implementation of the study, analysis of the results, drafting the manuscript, and approval of the final version of the manuscript. All authors critically revised the manuscript for intellectual content and read and approved the final manuscript.

FUNDING
This study was supported by Plan Nacional de I+D+i 2013-2016, Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía y Competitividad, Spanish Network for Research in Infectious Diseases (REIPI RD16/0016/0010); the 2017 Call for Strategic Action on Health (PI17/01932), co-financed by European Development Regional Fund "A way to achieve Europe" and the operative program Intelligent Growth 2014-2020; and grant 2017 SGR 0809 from the Departament d'Universitats, Recerca i Societat de la Informació, of the Generalitat de Catalunya. MM-A and CC were supported by grants FPU 14/06357 and FPU 13/02564, respectively, from the Spanish Ministry of Education, Culture and Sports. IR was supported by the Department of Health, Generalitat de Catalunya, grant SLT002/16/00349. We also acknowledge support from the Spanish Ministry of Science, Innovation and Universities through the "Centro de Excelencia Severo Ochoa 2019-2023" Program (CEX2018-000806-S), and support from the Generalitat de Catalunya through the CERCA Program." The funders had no role in the study design, data collection, analysis and interpretation of data, decision to publish, or preparation of the manuscript.