The principal mechanism of resistance to beta-lactams in Enterobacterales is the production of beta-lactamases – enzymes that hydrolyze the beta-lactam ring, inhibiting its antibacterial activity.2,3 Four classes of beta-lactamases have been established, based on their molecular structure (A, B, C and D).4 The enzyme activity depends on a serine group in classes A, C and D, and on one or two zinc ions in class B; as a result, the latter are also known as metallo-β-lactamases.5 On the other hand, the beta-lactamases can be classified into three functional groups (1, 2 and 3) according to their capacity to hydrolyze different substrates and their inhibition by different compounds (Table 1).6,7 From the clinical perspective, the enzymes of greatest interest in enterobacteria correspond to three groups: ESBL, class C enzymes and carbapenemases.2,3

The ESBL are class A enzymes that degrade penicillins, cephalosporins (except cephamycins) and monobactams. They are often coded for by plasmid genes.8 These enzymes are commonly inhibited by clavulanic acid, tazobactam, sulbactam and new inhibitors7 (Fig. 1). From a structural point of view there are a very large number of families, with TEM, SHV (related to an intrinsic chromosomal enzyme — without ESBL profile — of Klebsiella pneumoniae) and CTX-M (in particular CTX-M-15 and CTX-M-14) being the most important.9,10 Among others, the clones of Escherichia coli with sequence type (ST) ST131 or of K. pneumoniae ST11 and ST405 are implicated in the worldwide spread of CTX-M-15.11–13

Figure 1.

Extended-spectrum beta-lactamase (ESBL)-producing Klebsiella pneumoniae. In usual medium (without cloxacillin), the halos of ceftazidime (CAZ) and cefotaxime (CTX) are seen to be of smaller diameter than those of both cephalosporins combined with clavulanic acid (CCA and CTT, respectively). In medium with cloxacillin, which would inhibit the eventual additional presence of a class C beta-lactamase - plasmidic in the case of K. pneumoniae - no increase in the halos is observed, thus discarding this possibility.

(0.12MB).

The class C beta-lactamases include both AmpC enzymes coded for by chromosomal genes14 and variants coded for by plasmids (plasmidic cephamycinases).15 As a norm, they are not inhibited by clavulanic acid (neither tazobactam nor sulbactam), but are susceptible to new inhibitors such as avibactam.16 In many species (though not in E. coli), the production of chromosomal AmpC is regulated by a complex system of genes. Basal conditions are characterized by a low production of AmpC (repression state), though certain beta-lactams are more or less effective in inducing the production of the enzyme, and when the compound disappears, such induction ceases.16,17 Mutations in regulator genes17 can condition high-level production of enzyme even if there is no inducing beta-lactam; the corresponding strains are therefore referred to as derepressed strains.18 Most plasmidic cephamycinases are produced at a high level.15 The level of resistance of the microorganism to each beta-lactam depends on the amount of enzyme (beta-lactam induction capacity or enzyme derepression state) and the resistance of each compound to enzyme hydrolysis.14 The class C enzymes can degrade penicillins and cephalosporins such as cefotaxime and ceftazidime, and although they do not hydrolyze carbapenems effectively (with some exceptions, such as CMY-10 or ACT-28), when the enzyme is hyper-produced in strains with additional mechanisms (see below), clinically important resistance levels can be reached.19

The class A beta-lactamases with carbapenemase activity are mainly represented by the KPC family, and to a lesser extent by GES (not all their variants possess carbapenemase activity), SME, IMI and other variants.20–22 The KPC enzymes exhibit universal distribution, and generically hydrolyze carbapenems, penicillins, cephalosporins and monobactams. They are not inhibited by clavulanic acid (in fact, they hydrolyze the drug), but are inhibited by avibactam, vaborbactam and relebactam. To date, almost 100 variants of KPC are known, a number of which (e.g., KPC-31) produce resistance to ceftazidime-avibactam, though they do not efficiently hydrolyze carbapenems – thus generating a phenotype similar to that of the ESBLs23 (Fig. 2). The blaKPC gene forms part of transposon Tn4401 (of which several isoforms exist) and is vehiculized by different plasmids that also encode for genes of resistance to other families of antimicrobials. Several species of KPC-producing enterobacteria have been identified, though in view of their frequency, special mention must be made of the so-called high-risk clones of K. pneumoniae ST258, ST14, ST15, ST307, etc.24

Figure 2.

KPC-31 producing Klebsiella pneumoniae resistant to ceftazidime-avibactam (CZA) and with a meropenem minimum inhibitory concentration (MIC) smaller than that which would correspond to the related conventional carbapenemase KPC-3.

(0.14MB).

The class B beta-lactamases, or metallo-β-lactamases, degrade carbapenems and other compounds (penicillins and cephalosporins, but not monobactams) via a Zn2+ dependent mechanism. These enzymes are not inhibited by the currently available inhibitors. Although they are inhibited with EDTA (or other binding agents), this is of no practical use from the therapeutic perspective.25 Of the different families known, the NDM enzymes (with plasmidic or chromosomal coding) in enterobacteria are geographically widespread in isolates of different species (E. coli ST101 or ST131, K. pneumoniae, etc.). There have also been descriptions of strains that produce VIM, IMP or other families.

The class D enzymes are generically known as oxacillinases, because in vitro they hydrolyze this compound more efficiently than the class A or C enzymes.26,27 Structurally, they constitute a heterogeneous group of beta-lactamases that may behave as penicillinases, extended-spectrum beta-lactamases and carbapenemases. With some exceptions, they are not inhibited by clavulanic acid, tazobactam or sulbactam. Some of them (e.g., the OXA-48 group) are inhibited by avibactam. In Enterobacterales, the most relevant carbapenemase among the oxacillinases is OXA-48, initially identified in K. pneumoniae in Turkey, but with a widespread distribution in the Mediterranean setting and in other countries. In this context, blaOXA-48 is usually associated with a conjugative plasmid and with variants of Tn1999 found in different ST of K. pneumoniae and other species. Although OXA-48 has little capacity to hydrolyze extended-spectrum cephalosporins, many strains with this enzyme also usually produce CTX-M-15; they consequently exhibit resistance to these compounds.28 Other variants related to OXA-48 have been described, such as OXA-163 (with activity against extended-spectrum cephalosporins) or OXA-181.

In addition to the production of beta-lactamases, Enterobacterales resistance to beta-lactams is related to other mechanisms. Loss or structural alteration of the porins (hydrophilic channels through which the antimicrobials reach the interior of the bacteria) in itself causes a minor increase in resistance which – in the absence of other mechanisms – would have a limited clinical impact, but which on acting synergically with the latter would increase resistance. The impact is greater when one same strain loses its different porins: many ESBL or carbapenemase-producing clones of K. pneumoniae already lack porin OmpK35; as a result, the additional loss of the second main porin (OmpK36) results in a marked increase in resistance.29 The loss of porins is also relevant in other species, such as Enterobacter spp.19 Curiously, in E. coli, although multiple details are known about the role of porins in resistance in the laboratory strain K-12, less information is available when it comes to clinical isolates. The active efflux pumps, which eliminate beta-lactams once they have penetrated the bacteria, also contribute to increase resistance – though their role in the case of the beta-lactams of greatest clinical interest is less relevant than in other antimicrobials or in other microorganisms (as will be commented on addressing P. aeruginosa).30 The combination of porin loss, hyperproduction of the AcrB pump, and the hyper-expression of KPC-2 has been related to resistance to the new combination meropenem-vaborbactam.

In contrast to what is seen in gram-positive bacteria such as Staphylococcus aureus or Streptococcus pneumoniae, the role of alterations of the penicillin-binding proteins (PBPs) in Enterobacterales resistance to beta-lactams is small, though some studies have pointed to the interest of mutations in PBP-3.31

Resistance to quinolones in Enterobacterales is related to multiple mechanisms. Traditionally, special attention has focused on resistance derived from alterations of the type II topoisomerases (topoisomerase II or DNA gyrase and topoisomerase IV). These occur as a consequence of mutations in the so-called quinolone resistance determining region (QRDR) of the gyrA (DNA gyrase) and parC genes (topoisomerase IV). Less important are the mutations in gyrB or parE. A single mutation in gyrA causes an increase in resistance to non-fluorinated quinolones (nalidixic acid) and a low level of resistance to fluorinated quinolones (Fig. 3), though the level of resistance to the latter rises parallel to the increase in the number of new mutations in gyrA and parC.32

Figure 3.

Klebsiella pneumoniae with a low level of resistance to quinolones. Note the resistance to nalidixic acid (NAL) and the lesser involvement of the fluorinated quinolones, norfloxacin (NXN), levofloxacin (LVX) and ciprofloxacin (CIP).

(0.12MB).

As in the case of beta-lactams, alterations of the porins contribute to increasing the level of resistance to quinolones. It has also been shown that lipopolysaccharide modifications are related to resistance to these compounds. With regard to the active efflux pumps, special interest focuses on those of the family RND (resistance-nodulation-division), which are structurally integrated into systems of three proteins: the pump itself, an efflux channel in the external membrane and a coupling protein of the other two. The RND systems eliminate a range of compounds, including quinolones. The best-studied example is that of AcrA-AcrB-TolC, the expression of which contributes to the (low) basal level of resistance30; this system can be hyper-expressed due to mutations in other regulator genes (operon mar, acrR, etc.), consequently raising the level of resistance. Some of these regulators (mar) are capable of simultaneously leading to loss of a porin and to hyperproduction of the efflux pump.33 The great majority of strains of K. pneumoniae possess the chromosomal oqxAB genes, which encode for an active efflux mechanism that expels hydrophilic quinolones and other compounds.

Several families of Qnr proteins34 coded for by plasmids have been identified that produce resistance to quinolones via a mechanism that protects the target of these compounds (topoisomerases). In different species, intrinsic qnr type chromosomal genes have even been identified. Following the discovery of Qnr, other plasmidic resistance mechanisms have been characterized, related to the production of acetylase AAC(6′)-Ib-cr, and affecting both (some) quinolones and (some) aminoglycosides, or to QepA (active efflux pump) expression. These plasmidic mechanisms per se produce (very) low levels of resistance, but when they accumulate within one same strain or coincide with chromosomal mechanisms, the levels exceed the clinical resistance cut-off points.35

The two fundamental mechanisms of resistance to aminoglycosides in Enterobacterales are the enzymatic modification of these compounds or the modification of their target.36,37 Additional information is also available on disorders in penetration into the bacteria or active elimination by efflux pumps (AcrD).30

The aminoglycoside-modifying enzymes are grouped into three large families of nucleotidyl (adenyl)-transferases, phosphotransferases or acetyltransferases, each with an infinity of variants, that respectively transfer AMP, phosphate or acetyl coenzyme A to certain positions of the aminoglycoside molecule, canceling the antibacterial effect of the drug.36

Each concrete enzyme affects certain aminoglycosides, but not others. In addition, one same microorganism can produce multiple enzymes of the same or of different families. As a result, the resistance phenotypes are difficult to correlate to single proteins (genes). To further increase complexity, there are two different nomenclature systems for these enzymes (for the proteins and for their genes). The enzymes most commonly found in Enterobacterales (most studies refer to multiresistant E. coli and K. pneumoniae) are AAC(3)-IIa, AAC(3)-IVa, AAC(6′)-Ib (with variants that affect the quinolones, as commented above), ANT(2″)-Ia, APH(3′)-Ia, APH(3′)-IIa, APH(3″)-Ib, ANT(2″)-Ia and ANT(3″)-Ia. In any case, it must be taken into account that certain enzymes affect compounds that are currently of little clinical interest (e.g., ANT3″-Ia only affects streptomycin and spectinomycin).36 The new compound plazomicin escapes practically all of the modifying enzymes of this kind.38

Ribosomal modification as a cause of resistance to aminoglycosides may be due to alterations of the proteins of the ribosome or – more importantly, due to their frequency – to the modification of specific sites in 16SrRNA mediated by methyl-transferases (methylases) coded for by plasmids.37 There are two large families of these methylases: N7-G1405 (ArmA and variants of Rmt) and N1-A1408 (NmpA). The two families inactivate the aminoglycosides of clinical interest (gentamycin, tobramycin, amikacin and even plazomicin), resulting in high-level resistance, and differ from each other in inactivating or not inactivating other compounds that are rarely used in current clinical practice. Much of the information available on this mechanism of resistance refers to the study of multiresistant strains; in Spain, it has recently been found that 5.1% of all carbapenemase-producing enterobacteria (mainly K. pneumoniae and E. cloacae) produce methylases, in particular RmtF.39

A number of the enterobacteria most commonly isolated from clinical samples present intrinsic resistance to polymyxins, including Proteus spp., Morganella, Providencia spp., Serratia marcescens and Hafnia alvei.

Other species can develop resistance acquired through different mechanisms, of which the most important is the modification of lipopolysaccharide (LPS) due to the addition of different molecules – most commonly 4-amino-4-deoxy-l-arabinose or galactosamine (mediated by chromosomal genes) or phosphoethanolamine (mediated by chromosomal genes or by the recently discovered mcr type plasmid genes).40–42 Following this modification, the net negative charge of the LPS decreases, which complicates the interaction of the polymyxins with the bacteria. The comparative importance of other mechanisms, such as the production of capsule or active efflux pumps, is secondary.

The chromosomal genes that add the indicated compounds are activated as a consequence of mutations in double-component systems (composed of a sensor transmembrane histidine-kinase protein that undergoes auto-phosphorylation under certain conditions, and another cytoplasmic protein that on undergoing phosphorylation modulates the expression of different genes). These systems include PhoP-PhoQ, PmrA-PmrB and CrrA-CrrB.40–42 In K. pneumoniae, mutations of MgrB (a small transmembrane protein that under normal conditions negatively regulates the kinase activity of the PhoP/PhoQ system) are one of the main causes of resistance to polymyxins.43

The family of mcr plasmid genes codes for phosphoethanolamine transferases that modify LPS in a way analogous to other chromosomal genes. Their expression does not always result in levels that exceed the clinical resistance cut-off points.41

Stenotrophomonas maltophilia intrinsically presents a multiresistant phenotype related to the low permeability of its external membrane, which contains few porins,78 and to the presence of an SmeDEF efflux pump - affecting beta-lactams, quinolones and aminoglycosides.79 In addition, it is naturally resistant to aminoglycosides due to the presence of a chromosomal acetyltransferase, AAC(6′)-Iz.80 It also produces two inducible chromosomal beta-lactamases: carbapenemase L1, which confers intrinsic resistance to all the carbapenems, and cephalosporinase L2, which moreover also hydrolyzes aztreonam.78

The fluoroquinolones, and in particular levofloxacin, are active despite the low-level expression of a Qnr protein coded for at chromosomal level. High-level resistance to the fluoroquinolones may result from the selection of mutants with greater expression of the smQnr gene or of the SmeDEF efflux pump.81

Although S. maltophilia is very sensitive to the combination trimethoprim - sulfamethoxazole, resistance to the combination may appear through the acquisition of sul and dfrA genes, vehiculized by plasmids and class 1 integrons that code for dihydropteroate synthases with activity upon the sulfonamides.82