Keywords

1 Horizontal Gene Transfer of Resistance Genomic Islands

The horizontal gene transfer is crucial for the evolution and adaptation of bacterial species (Soucy et al. 2015; Koonin 2016). An important part of the horizontal gene transfer is facilitated by the large, discreet DNA segments called genomic islands (GIs) (Carraro et al. 2017b; Juhas et al. 2009). Some GIs are nonmobile and firmly integrated into the chromosomes of the host bacteria. Alternatively, GIs such as SGI/PGI/AGI-like GIs found in Salmonella enterica, Proteus mirabilis, Morganella morganii and Acinetobacter baumannii described in this review can be mobilized by other mobile genetic elements that are present in the host genome (Siebor et al. 2016, 2018; Kiss et al. 2015; Murányi et al. 2016; Carraro et al. 2017a, b; de Curraize et al. 2018; Hawkey et al. 2018; Schultz et al. 2017). Finally, some GIs encode means of their own horizontal transfer into the new host cells. These mobile GIs include a large group dubbed integrative and conjugative elements (ICEs). Although ICEs are often found integrated into the host chromosome, they are capable of excision, transfer into the new host cells by conjugation and reintegration into the chromosome of the new host. ICEs known to be involved in the spread of antibiotic-resistant genes include ICEs found in Haemophilus spp. described previously or the recently identified Pseudomonas aeruginosa ICEs presented in this review (Juhas et al. 2007a, b, 2008, 2009; Carraro and Burrus 2014; Juhas 2015c; Botelho et al. 2018; Roy Chowdhury et al. 2016, 2017).

In addition to genes crucial for self-propagation, GIs can carry a broad spectrum of accessory open reading frames involved in the virulence (pathogenicity islands), symbiosis (symbiosis islands), metabolism (metabolic islands), fitness (fitness islands) or resistance to antimicrobials (resistance islands) (Juhas et al. 2009). As resistance GIs are capable of carrying large numbers of variable antibiotic-resistant genes, their acquisition can provide the host bacterium with the resistance to the whole array of antimicrobials. The process of the rapid evolution of the multidrug-resistant (MDR) bacteria facilitated by GIs has therefore been dubbed the evolution by “quantum leaps” (Groisman and Ochman 1996; Hacker and Carniel 2001). This review provides an update on the contribution of GIs to the evolution of MDR bacteria, with the particular emphasis on GI3, which contributed to the increased pathogenicity of the Escherichia coli O104:H4 2011 outbreak strain and the most recent analyses of SGI/PGI/AGI-like GIs in MDR S. enterica, P. mirabilis, M. morganii and A. baumannii, SCCmec in MDR Staphylococcus aureus and GIs involved in the evolution of MDR P. aeruginosa.

2 GI3 in MDR Escherichia coli O104:H4

The outbreak of the highly pathogenic E. coli strain O104:H4 in 2011 which started in Germany resulted in nearly 5000 cases of severe infections associated with bloody diarrhoea, nearly 1000 cases of haemolytic uremic syndrome and 82 deaths (Rasko et al. 2011). The increased pathogenicity of the E. coli O104:H4 2011 outbreak strain was attributed to the acquisition of a battery of mobile genetic elements (Grad et al. 2013). These included Shiga toxin 2 (stx2)-encoding prophage typically found in enterohaemorrhagic E. coli strains, which in combination with the increased aggregative adherence typical for the enteroaggregative E. coli led to the increased absorption of Shiga toxin in the infections caused by the E. coli O104:H4 outbreak strain (Juhas 2015a; Rasko et al. 2011).

Notably, the E. coli O104:H4 2011 outbreak strain was found to be resistant to multiple antibiotics. In addition to 90 kb IncI1 plasmid encoding the CTX-M-15 cephalosporinase, which was unique to the 2011 outbreak strain (Künne et al. 2012), comparative whole-genome analyses revealed the presence of the E. coli genomic island 3 (GI3) in the chromosomes of the enteroaggregative haemorrhagic E. coli O104:H4 outbreak strains. GI3, considered to be the hotspot for microevolutionary genetic events in the E. coli O104:H4 strains, harbours multiple antibiotic-resistant genes, namely, bla TEM-1, dfrA7, strAB, sul2 and tet(A)A encoding resistance to ampicillin, trimethoprim, streptomycin, sulfamethoxazole and tetracycline, respectively (Roy Chowdhury et al. 2015). The resistance genes in GI3 are located in the complex antibiotic-resistant gene loci (CRL), which also include ag43 encoding the virulence factor involved in the biofilm formation and mer genes encoding the resistance to mercury (Grad et al. 2013). Bioinformatics analysis revealed that the evolution of CRL in GI3 has been mediated by Tn6029 family of transposons carrying the bla TEM-1-sul2-strA-strAB gene cluster flanked by IS26 (Roy Chowdhury et al. 2015).

The presence of the multiple antibiotic-resistant genes in the genome of E. coli O104:H4 contributed to the 2011 outbreak. In addition to limiting the available treatment options, the multiple antibiotic-encoding GI3 in combination with the IncI1 plasmid-encoded CTX-M-15 cephalosporinase played a key role in the ability of the E. coli O104:H4 2011 outbreak strain to outcompete the other commensal E. coli strains in the guts of patients that were treated with antibiotics (Bielaszewska et al. 2011; Juhas 2015a).

Interestingly, although found to be integrated into the chromosomes of the host E. coli O104:H4 strains, GI3 was previously shown to be mobile (Roy Chowdhury et al. 2015). To integrate into the chromosome, GI3 targets a 23 bp genomic sequence found in a broad spectrum of Enterobacteriaceae. This highlights the potential risk of GI3 being involved in the emergence of a number of other MDR Enterobacteriaceae in the future (Roy Chowdhury et al. 2015).

3 SGI/PGI/AGI-Like GIs in MDR Salmonella enterica, Proteus mirabilis and Acinetobacter baumannii

Salmonella genomic island 1 (SGI1, 42.4 kb), its variants and related GIs, such as SGI2, Proteus genomic island (PGI1) and Acinetobacter genomic island (AGI1) are frequently found in a broad spectrum of Enterobacteriaceae (Siebor et al. 2016; Soliman et al. 2017; Hamidian et al. 2015b, c). Although usually stably integrated into the host cell’s chromosome, SGI1 can be mobilized by the helper plasmids of the incompatibility groups A and C (IncA and IncC) (Siebor et al. 2016). It was shown that mobilization of SGI1 is initiated by the helper plasmid-encoded master transcriptional activator complex AcaCD. AcaCD is crucial not only for the expression of genes involved in the excision of SGI1 from the chromosome but also for the expression of the three SGI1-encoded tra genes involved in the transfer of SGI1 into a new host cell (Kiss et al. 2015; Murányi et al. 2016). Recent studies showed that SGI1 can reshape the conjugative apparatus of the helper plasmids to promote its own transfer (Carraro et al. 2017a). Besides a conserved backbone comprised of 28 open reading frames involved in the life cycle, excision, transfer and reintegration of SGI1, SGI1 harbours a complex MDR region conferring resistance to multiple antibiotics, including ampicillin, tetracycline, sulfamethoxazole, chloramphenicol/florfenicol and streptomycin/spectinomycin (de Curraize et al. 2018) (Fig. 1).

Fig. 1
figure 1

Schematic representation of the SGI/PGI/AGI-like genomic islands. The figure shows the key features of the SGI/PGI/AGI-like genomic islands, which are involved in the evolution of MDR Salmonella enterica, Proteus mirabilis, Acinetobacter baumannii and Morganella morganii. SGI/PGI/AGI-like genomic islands are integrated into the chromosome in the 3′ end of the trmE gene and are flanked by the attL and attR attachment sites. In addition to the backbone region carrying open reading frames crucial for the life cycle (excision from the chromosome, transfer and integration into the new host’s chromosome), SGI/PGI/AGI-like genomic islands also harbour MDR region encoding resistances to a number of different antibiotics, such as ampicillin (AMP), chloramphenicol (CHLO), sulfamethoxazole (SUL), streptomycin/spectinomycin (STR/SPE) and tetracycline (TET)

Full size image

Proteus genomic island (PGI1) of 81.1 kb, found in P. mirabilis, has a gene backbone similar to that of SG1, in addition to the complex MDR region composed of transposons and IS elements and a number of antibiotic-resistant genes, including metallo-β-lactamase-resistant gene bla NDM-1 and extended-spectrum-β-lactamase-resistant gene bla VEB-6 (Siebor and Neuwirth 2014).

Recently, a novel resistance GI, GIPm1 of 55.8 kb, was identified in P. mirabilis inserted into the chromosome at the trmE site, as is typical for other GIs of this subfamily, such as SGI1 and PGI1. GIPm1 is composed of the gene backbone similar to GI found in Enterobacter cloacae DSM16690 and MDR region conferring resistance to multiple antibiotics (Siebor et al. 2018). The detection of the extrachromosomal circular form GIPm1 suggests that GIPm1 is mobile; however, its mobilization is not facilitated by the IncA and IncC helper plasmids but by another, yet unknown mechanism (Siebor et al. 2018).

Recently, two new variants of SGI1 named SGI1-K7 (55.1 kb) and SGI1-Pm2CHAMA (53.6 kb) were identified in P. mirabilis whose MDR regions carried resistance genes never before identified in this subfamily of GIs, namely, bla CTX-M-15 and bla CARB-2 (de Curraize et al. 2018). Particularly alarming is the identification of bla CTX-M-15 encoding an extended-spectrum-β-lactamase usually found in MDR E. coli clinical isolates among the open reading frames of SGI1-K7. Notably, CTX-M-15 was also found in the E. coli O104:H4 2011 outbreak strain (Künne et al. 2012; Rasko et al. 2011).

AGI1 with the gene backbone similar to SGI1, PGI1 and their variants was found in the chromosome of A. baumannii. AGI1 is integrated into the A. baumannii chromosome at the trmE site, similar to other GIs from this subfamily. Furthermore, AGI1 carries a MDR region harbouring genes bla PER, aadB, aadA13/2, aadA2, strAB and sul1 conferring resistance to cephalosporins, aminoglycosides and sulphonamides (Hamidian et al. 2015a). Besides AGI1, A. baumannii chromosome often harbours other types of resistant GIs, such as A. baumannii-resistant island (AbaR). These GIs, together with the plethora of other mobile genetic elements and the high intrinsic resistance of A. baumannii to a number of clinically used antimicrobials, such as the second-generation cephalosporins, aminopenicillins and chloramphenicol, make MDR A. baumannii one of the main public threats nowadays (Pagano et al. 2016; Blackwell et al. 2015; Kim et al. 2017; Gallagher et al. 2017; Hawkey et al. 2018).

SGI1-like variant dubbed SGI1-L harbouring bla CARB-2, dfrA15, floR, sul1 and tetA(G) conferring resistance to amoxicillin, trimethoprim, phenicols, sulphonamides and tetracyclines has been recently identified in a MDR Morganella morganii subsp. morganii (Schultz et al. 2017).

While the gene backbone region of the SGI/PGI/AGI-like subfamily of GIs is relatively well conserved, their MDR region is very plastic as a result of microevolution events, which include IS element-mediated genome rearrangements and homologous recombination (de Curraize et al. 2018). Taken together with the ability to be mobilized by the helper plasmids, SGI1 and its variants and related GIs, such as PGI1 and AGI1, present a high risk for the dissemination of the clinically relevant antibiotic-resistant genes among other Enterobacteriaceae (de Curraize et al. 2018).

4 SCCmec in MDR Staphylococcus aureus

S. aureus is a highly versatile Gram-positive opportunistic pathogen causing a variety of diseases, ranging from minor skin infections to endocarditis, osteomyelitis, pneumonia and septicaemia. Methicillin-resistant S. aureus (MRSA) is considered to be among the main threats to human health nowadays (Juhas 2015a). Alarmingly, MRSA strains have in the meantime acquired resistances to a number of other antibiotics, including β-lactams, lincosamides and macrolides and more recently also vancomycin (Jani et al. 2017). Consequently, MDR MRSA strains are among the leading causes of morbidity and mortality and are among the most significant sources of both nosocomial and community-acquired infections (Anderson et al. 2014).

Resistance of S. aureus to methicillin was acquired via PBP2a-encoding gene mecA located on the staphylococcal chromosome cassette methicillin-resistant (SCCmec) island. SCCmec islands can be highly diverse ranging from 20 to 70 kb. Furthermore, SCCmec islands are classified into subtypes based on sequences of serine recombinases ccrAB and ccrC and regulators mecI and mecR1 (Ray et al. 2016). Recombinases ccrAB and ccrC play an important role in the excision, circularization and site-specific integration of SCCmec into the S. aureus chromosome (Smyth et al. 2011; Ray et al. 2016).

The origin of mecA in S. aureus remains still unclear; however, bioinformatics analyses suggest that it has been acquired horizontally from other staphylococcal species, such as Staphylococcus epidermidis, Staphylococcus fleuretti or Staphylococcus haemolyticus (Juhas 2015a; Xue et al. 2017; Smyth et al. 2011; Ray et al. 2016; Wipf et al. 2015; Tsubakishita et al. 2010; Hosseinkhani et al. 2018). There are several lines of evidence for this claim. First, SCCmec subtype IV has been found in the genome of S. epidermidis before it has been reported in S. aureus (Smyth et al. 2011). Second, the transfer of SCCmec island captured on staphylococcal conjugative plasmid into isogenic S. aureus and S. epidermidis has been demonstrated in vitro (Ray et al. 2016). Third, SCCmec elements have been identified in S. epidermidis carrying resistance genes against β-lactams, heavy metals and polyamines. It was shown that the resistance genes located on the S. epidermidis SCCmec elements originated from various bacteria, habitats and geographic regions (Xue et al. 2017). This highlights the role of S. epidermidis as a reservoir of resistance genes implicated in the evolution of MDR S. aureus. Fourth, S. fleuretti naturally harbours mecA in its chromosome and is therefore suspected to be the source of the mecA in the SCCmec island (Wipf et al. 2015; Tsubakishita et al. 2010). Fifth, homology searches showed the presence of the SCCmec-borne ccrB-encoding recombinase in the genomes of a number of other staphylococci (Fluit et al. 2013). Finally, highly diverse SCCmec islands were identified in the genome of S. haemolyticus recently, highlighting the role of S. haemolyticus in carrying the methicillin-resistant genes (Hosseinkhani et al. 2018).

Interestingly, the gene clustering approach has led to identification of a number of novel GIs which harbour a wide variety of resistance genes in the genome of S. aureus, recently. It has been suggested that these yet uncharacterized GIs can shed light on the evolution of MDR in S. aureus (Jani et al. 2017).

5 GIs in MDR Pseudomonas aeruginosa

P. aeruginosa is an opportunistic Gram-negative human pathogen frequently associated with the chronic nosocomial infections, particularly in immunocompromised individuals (those with AIDS, severe burns and cancer and cystic fibrosis patients) (Oliver et al. 2000; Azam and Khan 2018). The emergence and rapid spread of antibiotic-resistant P. aeruginosa led to the classification of the drug-resistant P. aeruginosa among the greatest threats to public health by WHO, CDC and ECDC (CDC 2013; ECDC 2015). P. aeruginosa infections are notoriously hard to eradicate due to the vast array of intrinsic, adaptive and acquired resistance mechanisms that confer resistances to a number of antimicrobials, including the last-generation carbapenems, which are usually used for the treatment of MDR bacteria (Azam and Khan 2018; Juhas et al. 2004, 2005; Wiehlmann et al. 2007; Juhas 2015b; Potron et al. 2015; Oliver et al. 2015).

Genes encoding resistance to carbapenems are often located on class 1 integrons; however, class 1 integrons are considered to be incapable of self-transfer. Thus for the transfer to the new host, class 1 integrons have to associate with other transferable elements, such as plasmids, transposons or GIs (Roy Chowdhury et al. 2016, 2017). Interestingly, class 1 integrons were found to be carried by the two globally dispersed GIs (GI1 and GI2) that are frequently found in P. aeruginosa multidrug-resistant clones ST235, ST253, ST111 and ST175. Both GI1 and GI2 harbour open reading frames conferring resistances to the entire range of antibiotics used to treat P. aeruginosa infections, including carbapenems (Roy Chowdhury et al. 2016, 2017). Besides resistance to multiple antibiotics, P. aeruginosa GI1 and GI2 harbour genes implicated in their conjugative transfer to the new host cells and reintegration into the chromosome, thus suggesting their evolution from a novel, yet uncharacterized ICE. This study also highlights the role GIs play in the capture and dissemination of antibiotic-resistant genes and in the evolution of MDR P. aeruginosa (Roy Chowdhury et al. 2017).

Recent whole-genome sequencing analyses of the high-risk P. aeruginosa clone ST235 revealed the presence of a novel GI, ICEPae690 carrying bla GES-6 carbapenemase gene (Botelho et al. 2018). bla GES-6 was found on a class 1 integron In1076 located within ICEPae690. Besides In1076 integron, ICEPae690 carried gene-encoding integrase, type IV secretion system, relaxase and type IV coupling protein involved in the excision, conjugation and reintegration of ICEPae690 and other genes encoding maintenance functions. ICEPae690 was integrated into the chromosome next to a tRNAGly gene and flanked by 16 bp ICEclc-like attL and attR sequences. The flanking attL and attR sequences are probably a result of recombination between the attachment sites attB and attP on the chromosome and ICEPae690, respectively. This study also demonstrated that ICEPae690 is capable of the self-transfer by conjugation from the original P. aeruginosa clone ST235 to P. aeruginosa standard laboratory strain PAO1. It has been suggested that In1076 integron harbouring bla GES-6 “hitch-hiked” ICEPae690 to promote its own propagation to other host cells through ICE-mediated conjugation (Botelho et al. 2018).

Recent whole-genome sequencing analyses led to the identification of a number of novel GIs (PAGIs) in the genomes of P. aeruginosa, namely, PAGI-13, PAGI-14, PAGI-15 and PAGI-16 carrying a broad spectrum of resistance genes, including bla IMP-6, bla IMP-10 and bla GES-24. In PAGI-13, PAGI-14, PAGI-15 and PAGI-16, these resistance genes are located on the class 1 integrons and confer resistance to carbapenems (Hong et al. 2016; Silveira et al. 2016).

6 Conclusions

Work over the last years, in particular the recent bioinformatics and whole-genome sequencing analyses, highlighted the diversity of GIs and shed new light on the GI-mediated genome plasticity (Botelho et al. 2018).

It is now widely accepted that GIs play an important role in the evolution of a broad spectrum of bacterial species, including MDR bacteria, such as MDR E. coli, S. enterica, P. mirabilis, M. morganii, A. baumannii and P. aeruginosa, presented in this review but also in a number of other clinically relevant MDR pathogens. For instance, the master transcriptional activator complex AcaCD of the IncA/C helper plasmids originally found in the GIs of the SGI/PGI/AGI-like subfamily was shown to be involved in the mobilization of other unrelated resistance GIs, such as MGIVchHai6 of Vibrio cholerae which confers resistance to β-lactams, tetracycline, sulfamethoxazole, trimethoprim, streptomycin/spectinomycin and chloramphenicol (Carraro et al. 2016).

Furthermore, recent analyses revealed that GIs are undergoing constant microevolution and continue to accumulate novel resistance genes within their sequences (e.g. by capturing other resistance gene-harbouring integrons and transposons) (Roy Chowdhury et al. 2016, 2017). This development is alarming particularly in connection with the ability of some GIs to self-transfer into the new host as the acquisition of such multi-resistance GIs can provide the host bacterium with the resistances to the whole array of antimicrobials.

Future studies will be required to elucidate the molecular mechanisms involved in the life cycle of some GIs, in particular the role the GIs-borne regulatory genes play in their propagation. Furthermore, additional analyses will be needed to elucidate the details of the process of acquisition of accessory genes, such as novel antibiotic-resistant genes, by GIs. Additional analyses are also required to assess the prevalence of the recently identified GIs, such as SGI/PGI/AGI-like GIs and their variants, PAGI-13, PAGI-14, PAGI-15, PAGI-16 and ICEPae690 in other bacterial species. This will contribute to the better understanding of GIs and their role in the spread of antimicrobial-resistant genes and the evolution of MDR bacteria.