*3.1.3 Substrate and inhibitor profiles*

The results of the Carba NP test showed that all the Enterobacteriaceae identified in the biological samples use an enzymatic mechanism as a means of resistance to carbapenems. On the other hand, the study of the substrate and inhibitor profiles highlighted three cases, namely enzymatic activity implying resistance (R), decreased enzymatic activity leading to intermediate resistance (I) and a complete absence of enzyme activity implying sensitivity (S).

#### *3.1.3.1 Carbapenemase substrate and inhibitor profiles*

The carbapenemase KPC has described two different profiles defined as (P1 and P2). The P1 profile is observed with the microorganisms *Arizona* isolated from osteitis pus

*Phenotypic Characterisation of Carbapenemases Produced by Enterobacteria Isolated from… DOI: http://dx.doi.org/10.5772/intechopen.102969*


**Table 5.**

*Distribution of carbapenemases in biological samples.*

samples, *P. mirabilis* isolated from wound pus samples, *S. ficaria, E. gergoviae, E. asburiae, P. vulgaris* isolated from urine samples. The P1 profile, which is characterised by enzymatic activity on all substrates (ETP, IMP, MRP, AMX, CAZ, CTX, CFP) is not inhibited by the carbapenemase inhibitors used (AMC, CXC, AZT, EDTA, PIT). The MICs in the P1 profile are greater than 64 mg/L for ETP, 16 mg/L for IMP and 64 mg/L for MRP.

The second P2 profile was observed with *E. gergoviae* isolated from wound pus samples. Enzymatic activity towards all substrates was maintained in the presence of the inhibitors used, except for AMC, for which it was rather reduced. The MIC values here are above 64 mg/L for ETP and 16 mg/L for IMP and MRP respectively.

The identified OXA carbapenemases describe a single substrate and inhibition profile. This P3 profile is observed with *P. mirabilis* and *S. odorifera* 1 isolated from osteitis pus and blood samples respectively. The activity of the enzyme in this profile is observed on certain substrates (AMX, CFP) and in the presence of inhibitors (CXC, AMC, PIT). It is decreased on the substrates (ETP, IMP, MRP) as well as on one of the inhibitors, EDTA. Another characteristic of this profile is the absence of enzymatic activity towards the substrates CAZ, CTX and one of the inhibitors, AZT. The MICs for these microorganisms are greater than 64 mg/L for ETP, 4 mg/L for IMP, 4 mg/L and 8 mg/L for MRP in *P. mirabilis* and *S. odorifera* 1, respectively.

The P4 profile characterising the TND 1 carbapenemase was identified in *E. gergoviae* and *S. odorifera* 1, all isolated from urine samples. The activity of the enzyme is observed towards the substrates ETP, MRP, AMX, CAZ, CFP and in the presence of the inhibitors CXC, AMC, AZT, PIT. This enzymatic activity is diminished in the presence of IMP, one of the inhibitors, EDTA and is absent in the presence of the substrate CTX. The MICs are greater than 64 mg/L for ETP in both microorganisms, equal to 16 mg/L in *E. gergoviae* for IMP and MRP respectively, and 4 mg/L for IMP and 32 mg/L for MRP in *S. odorifera* 1.

Only *P. mirabilis* isolated from urine samples and osteitis pus expressed the P5 profile. This P5 profile associated with the production of the TND 2 carbapenemase shows an absence of enzymatic activity on the CAZ substrate. However, this activity is observed with regard to the substrates (ETP, IMP, MRP, AMX, CTX, CFP) and in the


*Phenotypic Characterisation of Carbapenemases Produced by Enterobacteria Isolated from… DOI: http://dx.doi.org/10.5772/intechopen.102969*




 *Maroua.*

#### *Enterobacteria*

### *Phenotypic Characterisation of Carbapenemases Produced by Enterobacteria Isolated from… DOI: http://dx.doi.org/10.5772/intechopen.102969*

presence of all inhibitors (AMC, CXC, AZT, EDTA, PIT). MICs are greater than 64 mg/L for ETP, equal to 32 mg/L for IMP and greater than 64 mg/L for MRP when isolated from urine and equal to 32 mg/L for MRP when isolated from osteitis pus.

The P6 profile is expressed by the *S. odorifera* 1 microorganism isolated from osteitis pus which produces TND 3 carbapenemase. The enzymatic activity of this profile is observed on the substrates ETP, AMX, CAZ and in the presence of the inhibitors CXC, AMC, PIT. This decreased enzymatic activity on the substrates IMP, MRP and in the presence of the inhibitors AZT and EDTA are absent with respect to CTX. The MIC values here are above 64 mg/L for ETP and 4 mg/L for IMP and MRP.

The *P. mirabilis* microorganism isolated from stool samples and producing the TND 4 carbapenemase expresses the P7 profile. This profile is characterised by an enzymatic activity towards each of the two substrates ETP, AMX and towards three inhibitors CXC, AZT, PIT. This enzymatic activity is decreased on the substrates IMP, MRP, CFP and in the presence of the inhibitor EDTA. Finally, no enzymatic activity was observed on the substrates CAZ, CTX and in the presence of the inhibitor AMC. The MIC values for this microorganism are greater than 64 mg/L for ETP and equal to 4 mg/L for IMP and MRP respectively.

The P8 profile described by TND 5 carbapenemase is observed with *S. odorifera* 1 isolated from stool samples. It is characterised by an enzymatic activity towards the substrates ETP, AMX, CAZ and in the presence of the inhibitors CXC, PIT. This enzymatic activity, which is diminished in the presence of the substrates MRP, CFP and the inhibitors AZT, EDTA, is absent on two substrates IMP, CTX and on an inhibitor AMC. The MICs here are greater than 64 mg/L for ETP, equal to 0.125 mg/L for IMP and 2 mg/L for MRP.

The *C. braakii* species producing the TND 6 carbapenemase isolated from stool samples express the P9 profile. The enzymatic activity here is observed on ETP and AMX substrates and in the presence of the inhibitors CXC and PIT. This enzymatic activity is decreased on both substrates MRP, CFP and in the presence of the inhibitor EDTA. It is absent on the substrates IMP, CAZ, CTX and in the presence of the inhibitors AMC, AZT. The MICs here are greater than 64 mg/L for ETP, equal to 0.5 mg/L for IMP and 4 mg/L for MRP.

The P10 profile is observed with the TND 7 carbapenemase produced by *E. gergoviae* isolated from stool samples and is characterised by enzymatic activity on the substrates ETP, AMX and in the presence of the inhibitors CXC, PIT. This enzymatic activity is decreased in the presence of two substrates MRP, CAZ and the inhibitor EDTA. There is no enzymatic activity on two substrates IMP, CTX and on two inhibitors AMC, AZT. The MICs are above 64 mg/L for ETP, 0.5 mg/L for ETP, and 4 mg/L for MRP.

Finally, the P11 profile is always found in *E. gergoviae* isolated from blood samples but which produces the TND 8 carbapenemase. It is characterised by enzymatic activity on three substrates ETP, IMP, AMX and on three inhibitors CXC, AMC, AZT. This enzymatic activity is decreased in the presence of the substrates MRP, CFP and in the presence of the inhibitor PIT. On the other hand, it is absent on the substrates CAZ, CTX and in the presence of the inhibitor EDTA. The MICs are greater than 64 mg/L for ETP, 16 mg/L for IMP and 0.5 mg/L for MRP (**Table 6**).

## **3.2 Discussion**

Biological samples containing carbapenem-resistant Enterobacteriaceae represented 5.97%. This percentage is distributed between urine samples (2.20%), osteitis pus (0.94%), wound pus (0.63%), blood (0.94%) and stool (1.26%). Carbapenem-resistant Enterobacteriaceae were not identified in urethral and vaginal swabs. The high proportion of carbapenem-resistant microorganisms in urine could be explained by the fact that this medium is potentially an extra-digestive reservoir for ESBL-producing Enterobacteriaceae [14]. The emergence of carbapenem resistance in some of the biological samples taken reflects the increasing complexity of the phenomenon in enterobacteria [15]. This complexification of the resistance phenomenon in the city of Maroua had already been observed in bacteria contaminating the food sold there [4, 5]. Several explanations can be found for the emergence of carbapenem resistance in the city of Maroua. The emergence of carbapenem resistance could be the consequence of exponential and uncontrolled use of antibiotics [6, 16–18]. The flow of populations between risk areas (Europe, Asia) and the city of Maroua could also contribute to the importation of strains expressing these types of resistance [19]. The opening of the University of Maroua, which contributes enormously to the migration of populations from various origins to the city, is also a major risk factor for the transport of multidrug-resistant strains of bacteria. The emergence of this type of resistance may finally be due to an exchange of the genes responsible for their expression between bacterial species from the digestive tract or the environment [20]. This exchange can take place via the phenomena of transduction [21], conjugation [22], or transformation [23].

Using API 20 E galleries, *Arizona*, *C. braakii, E. gergoviae, E. asburiae, P. mirabilis, P. vulgaris, S. ficaria* and *S. odorifera* 1 were identified as the carbapenem-resistant Enterobacteriaceae in the specimens. These Enterobacteriaceae are variously distributed in the samples. The species *E. gergoviae, E. asburiae, P. mirabilis, P. vulgaris, S. ficaria, S. odorifera* 1 were identified in urine specimens. Those found in osteitis pus were *Arizona, P. mirabilis, S. odorifera* 1. Two microorganisms, *E. gergoviae and P. mirabilis* were isolated from wound pus samples. The microorganisms isolated from blood were *E. gergoviae, S. odorifera 1. Finally, C. braakii, E. gergoviae, P. mirabilis* and *S. odorifera* 1 were identified in stool samples. The proportions of carbapenemresistant Enterobacteriaceae in biological samples were 36.84% in urine samples, 21.05% in osteitis pus and stools respectively, and 10.53% in blood and wound pus samples. This distribution in biological samples shows that Enterobacteriaceae are likely to cause deleterious effects in the organism from a variety of environments [24]. The diversity of environments where these enterobacteria have been identified can be explained by the great power of adaptation that characterises them [25] and the multi-resistance to antibiotics that does not facilitate their elimination [16, 17].

The enzymatic mechanism of resistance to carbapenems was demonstrated in 100% of the Enterobacteriaceae that were identified. This observation is in agreement with the fact that enzymatic inactivation of carbapenems is the main mechanism used by enterobacteria to resist their bactericidal effects [26]. The yellow colour change of phenol red used as a colour indicator to show the presence of enzymatic activity on carbapenems has been interpreted as the result of acidification of the reaction medium [27, 28]. This acidification of the reaction medium is a consequence of hydrolysis of the -lactam ring at the amide bond which produces a carboxyl function [29]. The level of expression of this reaction confers certain characteristics to enterobacteria. These characteristics were assessed indirectly on culture media using the inhibition diameters-MIC relationship [9]. The inhibition diameters-MIC correlation for selected carbapenems (r = 0.578, p < 0.01 for IMP and r = 0.858, p < 0.01 for MRP) allowed three characteristics to be defined. The first characteristic is resistance to carbapenem, which indicates the presence of enzymatic activity (R). The second characteristic is

*Phenotypic Characterisation of Carbapenemases Produced by Enterobacteria Isolated from… DOI: http://dx.doi.org/10.5772/intechopen.102969*

intermediate resistance which is the result of decreased enzyme activity (I). The third characteristic, marked by an absence of enzyme activity (S), defines the susceptibility of the enterobacteria to carbapenem [9].

The interpretation of the characteristics expressed by the enterobacteria in the presence of the substrates and inhibitors defined by the algorithm used made it possible to highlight three types of carbapenemases in these enterobacteria isolated from biological samples. These are carbapenemases of the KPC, OXA-48 or OXA-181 type and TNDs. The dominant proportion of KPC carbapenemases (36.84%) can be explained by the fact that they are the most abundant and widespread among enterobacteria [30]. They are also characterised by the existence of several variants that differ only by the substitution of one or two amino acids [31]. In contrast, the low percentage of OXA-41 or OXA-181 carbapenemases (10.53%) in the samples can be justified by the fact that this is an enzyme produced from a single auto transferable plasmid that does not carry additional resistance genes [32]. The low proportion of each of the TNDs can be explained by the fact that they are new phenotypes of point synthesis due to the presence of integrons. Integrons sometimes contain transposons from which some transposase-containing Enterobacteriaceae can be naturally genetically engineered to form highly expressed resistance operons [33].

The types of carbapenemases identified are differently distributed in biological samples and between enterobacteria. This random distribution within species of Enterobacteriaceae could be justified by the ease with which resistance-conferring genes diffuse between microorganisms [11]. It is this random distribution that may explain the difficulty in effectively applying probabilistic and/or therapeutic antibiotic therapy in cases of infection with resistant carbapenem enterobacteria [34]. The enzymatic activity of carbapenemases, which is manifested by hydrolysis at the amide bond of the said ring, has made it possible to describe 11 different substrates and inhibition profiles.

The first substrate and inhibition profile, P1, is characterised by enzymatic activity on all carbapenems including monobactam (AZT) used. The fact that this enzymatic activity is not influenced by the presence of EDTA proves that the enzyme does not need a heavy metal to hydrolyse the substrates. These characteristics are unique to KPCtype class A carbapenemases produced from plasmids [35]. It was also observed that the activity of this enzyme is maintained in the presence of its inhibitors PIT and AMC. This observation highlights a synergy of action between the carbapenemase KPC and an ESBL. Indeed, in the presence of a "suicide" inhibitor that serves as a decoy, such as clavunate or tazobactam, the bacteria compensate for the enzymatic deficit by amplifying the synthesis of ESBLs [6, 36]. This hyperproduction can be mediated by mutations in the promoter of the gene and/or by an increase in the number of plasmids carrying the bla gene. These ESBLs would therefore play the known role of multiplying the targets of antibiotics to limit their effectiveness [37]. From the above, it appears that bacteria of the P1 profile have the capacity to produce both KPC-type carbapenemases and ESBLs, all of which are class A.

Measurement of MICs for this profile using the E-test showed that variations are only observable between *P. vulgaris*, *E. asburiae*, *E. gergoviae* and *P. mirabilis*. From 16 mg/L for *P. vulgaris* and *E. asburiae*, it increases to 64 mg/L for *E. gergoviae* and *P. mirabilis*. The fluctuations obtained with the MIC values for carbapenems in these microorganisms could be explained by the existence of two KPC variants between these identified enterobacterial species [31].

The second substrate and inhibition profile (P2) is associated with the carbapenemase identified in *E. gergoviae* isolated from wound pus. The enzymatic activity here shows several similarities with the P1 profile. The only difference is the decrease in enzyme activity in the presence of AMC. The MIC values do not differ from those obtained with *E. gergoviae* isolated from urine samples. This slight variation in MIC suggests that the same KPC is produced in the P1 profile by this microorganism in both urine and wound pus. The decrease in enzyme activity in the presence of clavunate may be due to insufficient ESBL production to contain all the suicide inhibitor molecules. The consequence is a decrease in the number of enzyme molecules available for substrate hydrolysis which would then lead to a decrease in enzyme activity.

The third substrate and inhibition profile (P3) is expressed by OXA-type carbapenemases (48 or 181) produced by *P. mirabilis* and *S. odorifera* 1 isolated from osteitis pus and blood samples respectively. This profile is characterised by enzymatic activity on CXC, decreased on the three carbapenems and not observed at all on AZT. EDTA has no discernible influence on this activity. All these characteristics are consistent with the description of a class D carbapenemase [35, 38]. Another observation on this profile is that the activity of the enzyme resumes on AZT in the presence of CTX. The resumption of enzyme activity on AZT in the presence of CTX illustrates the theory that the combination of two -lactams can be antagonistic if one of them is an -lactamase inducer. CTX would therefore induce the production of ESBLs that could hydrolyse AZT. This illustrates the fact that *P. mirabilis* and *S. odorifera* 1 are likely to produce inducible ESBLs in addition to OXAs. Analysis of the MICs obtained in these two species shows that there are no differences in the activity of this enzyme either at the level of the microorganisms or the samples. This suggests that the OXA produced by these microorganisms originates from the same plasmid that has migrated from one species to another [39].

The P4 profile is only found in *E. gergoviae* and *S. odorifera* 1 isolated from urine samples. It is characterised by an enzymatic activity on ETP and MRP but diminished with respect to IMP. The inhibitors clavunate and tazobactam have no effect on this enzymatic activity. This observation can be explained by the fact that these bacteria produce class B carbapenemases or, a combination of ESBL and chromosomal type A and/or B carbapenemases [16, 17]. The decrease in enzymatic activity in the presence of EDTA validates the hypothesis of the presence of a class B carbapenemase [38]. The combination of these observations leads us to believe that the genes coding for the synthesis of both class A and B carbapenemases, both chromosomal, are present in these bacteria. It is, therefore, the inhibition of class B carbapenemase by EDTA that would be at the origin of the decrease in enzymatic activity. In this context, the decrease in enzymatic activity would then be due to the reduction in the quantity of carbapenemases potentially active on carbapenems. In view of the above, it is possible that the bacteria *E. gergoviae* and *S. odorifera* 1 possess in their chromosomes both genes coding for the synthesis of class A and B carbapenemases. The MIC measurements for these microorganisms did not show any differences apart from that obtained with IMP (16 mg/L and 4 mg/L in *E. gergoviae* and *S. odorifera* 1 respectively). This difference in MICs can be explained by mutations that may occur in the amino acid sequence homology or by the level of production of one or the other of these carbapenemases.

The P5 substrate and inhibition profile is found in *P. mirabilis* isolated from urine samples and osteitis pus. This profile is characterised by enzymatic activity on all substrates except CAZ. This activity is maintained in the presence of all inhibitors. This suggests a most likely plasmid hyper production of KPC associated with cephalosporinase. The different profiles for this microorganism (P1 when derived from wound pus and P5 when derived from either urine samples or osteitis pus),

### *Phenotypic Characterisation of Carbapenemases Produced by Enterobacteria Isolated from… DOI: http://dx.doi.org/10.5772/intechopen.102969*

although suspected of producing all the KPCs, would be the result of the difference in the enzyme that accompanies the production of these KPCs.

The P6 profile is found in *S. odorifera* 1 isolated from osteitis pus. It is characterised by an enzymatic activity towards ETP. This activity decreases on IMP and MRP. The presence of inhibitors has no visible effect on the enzymatic activity. The analysis of this P6 profile shows several similarities with the P3 profile. The same is true for the MIC values, which are close to those of the P3 profile. The great similarity observed between the P6 and P3 profiles suggest that *S. odorifera* 1 and *P. mirabilis*, both isolated from osteitis pus samples, produce carbapenemases of types OXA-48 or OXA-181. However, the increase in enzymatic activity observed with the P6 profile of *S. odorifera* 1 is thought to be the result of possible mutations in the OXA carbapenemases and the production of a cephalosporinase that activates the hydrolysis of CAZ [9].

The P7 profile expressed by *P. mirabilis* isolated from stools is characterised by an enzymatic activity on ETP and decreased on IMP and MRP substrates. The presence of clavunate shows inhibition of the enzymatic activity. This enzymatic activity, the extent of which varies from one carbapenemase to another, can be explained by the fact that it is the product of genes carried by the chromosomes [35, 40, 41]. The inhibition of the latter by clavunate validates the hypothesis of a class A carbapenemase.

The P8 profile identified in *S. odorifera* 1 isolated from stools shows enzymatic activity on ETP. This activity decreases on MRP and disappears on IMP. The inhibitor clavunate causes a loss of enzyme activity while EDTA has no effect on this activity. The MIC values show that the enzyme activity is distinct from one substrate to another. The fact that the enzyme activity is distinct on carbapenems and cephalosporin (CTX) sensitivity shows that this bacterium produces a chromosomal carbapenemase [35]. The inhibition of enzyme activity in the presence of clavunate supports the hypothesis of a class A carbapenemase [38]. Suspected carbapenemases may be SME, IMI-1 [40, 41]. The multiple similarities observed between the P7 and P8 profiles suggest that the carbapenemase produced in profile P8 may be a mutated form of that produced by *P. mirabilis* isolated from stool samples.

The P9 profile observed with the *C. braakii* microorganism isolated from stools is characterised by enzymatic activity on the ETP. This activity decreases with respect to MRP and disappears with respect to IMP and cephalosporins (CAZ, CTX). The presence of the inhibitor EDTA has no effect on the enzymatic activity contrary to clavunate and AZT which inhibit this activity. The strong similarity between the P9 and P8 profiles suggests that the same carbapenemase is mutated between *C. braakii* and *P. mirabilis* isolated from stool samples.

The P10 profile found in *E. gergoviae* isolated from stools always shows an enzymatic activity that varies from one carbapenem to another. With a few exceptions, this P10 profile is similar to the P9 profile. The observations show that the two profiles are similar and the few differences observed could reflect the presence of mutations in the genes producing these enzymes.

The last profile P11 is the fourth substrate and inhibition profile obtained with *E. gergoviae* isolated from blood. It is characterised by an enzymatic activity on ETP and IMP. This activity is diminished in the presence of MRP. The disappearance of this activity in the presence of EDTA indicates that the activity of this enzyme requires the presence of heavy metal [42]. No inhibition of the enzyme activity is observed with classical class A carbapenemase inhibitors. All these observations point to a class B carbapenemase [43]. The difficulty in typing this carbapenemase from this substrate and inhibition profile is the demarcation observed with other classical class B

carbapenemases. This demarcation comes from the fact that the enzymatic activity here decreases towards MRP whereas class B carbapenemases exhibit enzymatic activity on all carbapenems [43]. The decreased enzymatic activity of this P11 profile on MRP can be explained by the presence of mutations in the primary amino acid sequence homology at the active site [44]. The presence of these mutations may be a consequence of being produced from integrons carrying 'cassette' genes from which several genes can be assembled [10]. The fact that this P11 profile shows activity on AZT and towards CAZ and CTX suggests the presence of an ESBL.
