 CLINICA-004 FINAL/Image_001.png){kind=small}
Invest Clin 62(1): 37 - 51, 2021 https://doi.org/10.22209/IC.v62n1a04
Negative correlation between virulence and multidrug resistance in intrahospital and community acquired infections
by Proteus mirabilis, in Eastern Venezuela.
Hectorina Rodulfo1,2, Milagro Horta2, Genova Mata2, Rafela Gutiérrez2, Yarmilia González2, Elvia Michelli2, Militza Guzman2, Dianny Martínez3, Ashutosh Sharma1 and Marcos De Donato1
1Tecnologico de Monterrey, Escuela de Ingenieria y Ciencias, Campus Queretaro, México.
2Laboratorio de Genética Molecular, Instituto de Investigaciones en Biomedicina
y Ciencias Aplicadas, Universidad de Oriente, Núcleo de Sucre, Cumaná, Venezuela. 3Laboratorio de Bacteriología, Hospital Universitario “Antonio Patricio de Alcalá”, Cumaná, Venezuela.
Key words: Proteus; MDR; virulence; resistance; antimicrobials.
Abstract. This is the first report for Venezuela of virulence/pathogenicity and resistance factors in intrahospital (HCAI) and community-acquired infec- tions (CAI) by P. mirabilis in two main hospitals from Eastern Venezuela. Viru- lence factors such as motility, biofilms, and resistance to serum killing (RSK) were determined. Antimicrobial susceptibility allowed classifying the isolates into resistant, multidrug resistant (MDR) and extensively drug-resistant (XDR).
P. mirabilis was identified in HCAI in both hospitals mostly from secretions, while some CAI were identified from urine and secretions. Twitching, swarm- ing, biofilm and RSK were identified in many isolates. Eleven antimicrobials showed resistance frequencies from 22-54% in one or both hospitals. A high frequency of MDR isolates was found in these hospitals (60.6 to 56.5%). Strains
carrying both bla
CTX-M
and bla
TEM
genes were found in one hospital in a frequency
of 27.0%. We also found that the frequency of MDR was lower in strains with
three or more virulence factors compared to those with fewer factors. Bacteria with swarming showed 5.85 times lower probability of being MDR, and those with twitching, 7.52 times lower probability. Infections by MDR/XDR P. mira- bilis strains in HCAI and CAI represent a public health problem that requires effective control and prevention measures to reduce their potential spread and persistence in the population.
 CLINICA-004 FINAL/Image_002.png){kind=small}
 CLINICA-004 FINAL/Image_003.png){kind=small}
Correlación negativa entre la viruluencia y la resistencia multidroga en infecciones intrahospitalarias y adquiridas en la comunidad por Proteus mirabilis, en el Oriente de Venezuela.
Invest Clin 2021; 62 (1): 37-51
Palabras clave: Proteus; MDR; virulencia; resistencia; antimicrobianos.
Resumen. Este es el primer reporte para Venezuela de virulencia/pato- genicidad y factores de resistencia en infecciones intrahospitalarias (IAAS) y adquiridas en la comunidad (IAC) por P. mirabilis en dos hospitales principales del oriente de Venezuela. Se determinaron factores de virulencia como la moti- lidad, formación de biopelícula y resistencia al suero humano normal (RSHN). La susceptibilidad a los antimicrobianos permitió clasificar los aislamientos en resistentes, multirresistentes (MDR) y extensivamente resistentes a fármacos (XDR). Se identificó P. mirabilis en IAAS en ambos hospitales principalmente a partir de secreciones, mientras que algunos IAC se identificaron en orina y secreciones. Se identificaron motilidades “twitching” y “swarming”, biopelícu- las y RSHN en muchos de los aislamientos. Once antimicrobianos mostraron frecuencias de resistencia del 22 al 54% en uno o ambos hospitales. En estos hospitales se encontró una alta frecuencia de aislamientos MDR (60,6 a 56,5%).
CTX-M
En un hospital se encontraron cepas que portaban genes bla
TEM
y bla
con
una frecuencia del 27,0%. También encontramos que la frecuencia de MDR fue menor en las cepas con tres o más factores de virulencia en comparación con aquellas con menos factores. Las bacterias con “swarming” mostraron una pro- babilidad 5,85 veces menor de ser MDR, y aquellas con “twitching”, una proba- bilidad 7,52 veces menor. Las infecciones por cepas de P. mirabilis MDR/XDR en IAAS y IAC representan un problema de salud pública que requiere medidas de control y prevención efectivas para reducir su potencial propagación y per- sistencia en la población.
Received: 31-10-2020 Accepted: 06-01-2021
P. mirabilis is recognized as an etiologi- cal agent of different infectious processes, and its colonization capacity is due to the presence of fimbriae that favor the adher- ence of this bacterium to the renal epithe- lium, the phenomenon of swarming (SW), hemolytic activity, hydrolysis of urea, deami- nation of amino acids, proteases, endotoxins and lipopolysaccharides (LPS), among other
virulence factors that allow species of the ge- nus Proteus to cause more than 40% of intra- hospital infections of the urinary tract, most in patients with a urinary catheter (1, 2).
In Proteus species, the combined action of adhesins, SW and urease promote entry into the urinary tract, and the bacterium can modulate the expression of specialized viru- lence factors for its survival. Furthermore, SW facilitates migration through different types of urinary catheters, suggesting that
this movement may play an important role in the initiation of catheter-associated urinary tract infections and along with the biofilm the establishment of the infection. P. mira- bilis shows the capacity to form biofilms on biological surfaces and abiotic environments (polystyrene, glass, latex, silicone), the lat- ter representing the main cause of 65% of intrahospital infections (3).
Although P. mirabilis is naturally resis- tant to penicillin G, oxacillin, macrolides, lincosamides, streptogramins, glycopeptides, rifampicin, and fusidic acid, this microorgan- ism is still susceptible to many categories of antimicrobials used in clinical practice (4). Data from the SENTRY antimicrobial suscep- tibility program in the United States and the European Union for the isolates collected in 2009–2011 reported that <10% of the iso- lates were resistant to amikacin, aztreonam, cefepime, ceftazidime, ceftriaxone, merope- nem, and piperacillin/tazobactam (5). How- ever, extended spectrum beta-lactamases (ESBL) constitute a major therapeutic and epidemiological problem, because the pres- ence of these enzymes leads to bacterial mul- tidrug resistance (6), and in P. mirabilis the detection of ESBL has increased, reporting in different countries in South America, includ- ing: Brazil, Argentina, and Peru (7–9).
The development of multidrug resis- tance (MDR) and the invasiveness of Pro- teus spp. is related to many virulence fac- tors, such as SW motility, biofilm formation, hemolysins, urease production, and LPS, among others, which can act independently or in complementary fashion, and allow spe- cies of this genus a better adaptation to the conditions of the hospital environment, as well as, evading the defense mechanisms of the host for its survival.
This is the first study in Venezuelan hospitals that focuses on evaluating the presence of virulence factors and their rela- tionship with antimicrobial susceptibility in Proteus mirabilis clinical isolates. This will allow the assessment of the extent of the vir- ulence and resistance problem that contrib-
ute to therapeutic failure, due to the lack of efficacy of antimicrobials; and to be able to establish effective prevention and control strategies in the current poor conditions of the Venezuelan health system.
Bacteria of the Enterobacteriaceae and Morganellaceae families were isolated from clinical samples from patients treated in the two central hospitals of Sucre state, East- ern Venezuela: Hospital Universitario “Anto- nio Patricio de Alcala” (HUAPA) of Cumana (423 isolates) and Hospital Universitario “Santos Anibal Dominicci” (HSAD) of Caru- pano (274 isolates), from January through December of 2018. The patients gave their written consent after they were informed of the risks and benefits of participating in the study and had answered a clinical-epidemio- logical questionnaire. The treatment of the patient’s data, the analysis of the isolates, and the information generated were con- ducted according to the bioethical and bio- safety guidelines set out by the Commission on Ethics, Bioethics, and Biodiversity of the Instituto de Investigaciones en Biomedicina y Ciencias Aplicadas “Dra. Susan Tai” at the Universidad de Oriente, Venezuela.
Each clinical isolate was conserved in LB medium, reactivated in brain heart infu- sion broth (BHI), incubating it for 1 hour at 37°C, and then plated on MacConkey agar, blood agar (BA) and CLED agar, incubating it for 18 hours at 37°C in an aerobic environ- ment. Colonies morphologically suggestive of P. mirabilis were selected, and confirmed biochemically by means of the convention- al identification protocols. The Urea Broth (Sigma-Aldrich) was used to test the ability to hydrolyze urea to ammonia and carbon di- oxide by the urease.
Infections were classified into a com- munity-acquired (CAI) and healthcare-asso-
ciated (HCAI) following Cardoso et al. (10). The different levels of strain resistance were defined following Magiorakos et al. (11) as: multidrug resistant (MDR), when a strain has acquired non-susceptibility to at least one agent in three or more antimicrobial categories; extensively drug-resistant (XDR), when it has acquired non-susceptibility to at least one agent in all, but two or fewer an- timicrobial categories; and pan drug-resis- tant (PDR), when it is non-susceptible to all agents in all antimicrobial categories. Resis- tant bacteria, non MDR, were those strains showing resistance to one or more antimi- crobials, but not being classified as MDR, XDR nor PDR.
The production of biofilm was assessed according to Kwiecińska-Pirog et al. (12). Overnight cultures of the strains (2 µL) were added to TSB supplemented with 1% glucose (198 µL). The biofilm was quantified by mea- suring the optical density after staining by crystal violet at 570 nm. The P. aeruginosa strain (M-PA01) was used as a positive con- trol and sterile broth culture as a negative control to verify sterility and non-specific media components. The cut-off point was
air-drying, stained with cristal violet and the stained area was measured (13). In all three cases, the Escherichia coli ATCC® 25922 strain was used as a negative control. Follow- ing the methodology used to determine the biofilm, strains with negative motility were considered when their measurement was < and positive when it was ≥.
0
3
The resistance to serum killing (RSK) was determined using normal human serum obtained after coagulation whole blood from healthy donors from the HUAPA Blood Bank. A pool of sera was prepared, aliquoted and stored in 200 µL portions at -80°C. Each bacterial inoculum, of approximately 1x107 CFU/mL, was mixed with an equal volume of serum and incubated for 3 hours at 37°C. RSK was determined immediately after mix- ing the bacteria with the serum (T ), and af- ter 3 hours of incubation (T ), performing the plate dilution quantification method. For this, different dilutions of each isolate (1x10-1 - 10-4) were prepared, spreading a 10-
µL aliquot of the dilution on the surface of a CLED agar plate, which was incubated at 36°C for 24 hours. Subsequently, the num- ber of colonies was quantified taking the
calculated using the negative control (NC),
0
number of T
colonies as 100%. The isolates
considering negative strains those with val-
3
were considered RSK when the number of T
ues < and producers ≥ .
0
colonies was ≥50% with respect to T
(14).
Swarming (SW) was determined with BA prepared with 0.7% agar, inoculating the isolates on the surface of the agar (10
µL). The presence of swimming (SM), was determined with BA containing 0.3% agar and the isolates (10 µL) were inoculated by puncturing the medium, while for twitching (TM), the BA contained 1% agar and isolates were inoculated with a micropipette below the agar layer (10 µL). All the media were incubated for 24 hours at 37°C, the diam- eters of the SM and SW zones were mea- sured. For the TM measurement, the agar was removed, and the plates were fixed by
Antimicrobial susceptibility was per- formed using the disc diffusion method, fol- lowing the guidelines proposed by the CLSI (15). The following antimicrobial categories were tested: Penicillins (ampicillin, AMP: 10
µg), Monobactams (aztreonam, ATM: 30 µg), Cephamycins (cefoxitin, FOX: 30 µg), 1st and 2nd generation Cephalosporins (cephalothin, CF: 30 µg; cefuroxime, CXM: 30 µg), 3rd and 4th generation Cephalosporins (cefotaxi- me, CTX: 30 µg; ceftriaxone, CRO: 30 µg; ceftazidime, CAZ: 30 µg; cefepime, FEP: 30
µg), Carbapenems (imipenem, IMP: 10 µg; meropenem, MEM: 10 µg; ertapenem, ETP:
10 µg), Betalactamase Inhibitors (amoxicil- lin / clavulanic acid, AMC: 30 µg; ampicil- lin/sulbactam, SAM: 20 µg; piperacillin/ tazobactam, TZP: 10 µg), Aminoglycosides (gentamicin, GM: 10 µg; amikacin, AK: 30
µg; netilmicin, NET: 30 µg; tobramycin, NN:
30 µg), Fluoroquinolones (ciprofloxacin, CIP: 5 µg), Inhibitors of Folic Acid Metabo- lism (sulfamethoxazole/trimethoprim, SXT:
1.25 µg/23.75 µg) and Fenicols (chloram- phenicol, C: 30 µg). Antimicrobial resis- tance profiles were established according to Magiorakos et al. (11). The quality control of biochemical tests and antimicrobial discs were verified with the control strains of E. coli ATTC® 25922 and P. aeruginosa ATTC® 27853, from the Venezuelan Center for Mi- crobiological Collections.
P. mirabilis strains were tested for the phenotypic production of extended spec- trum betalactamases (ESBLs) using the modified method proposed by Poulou et al. (16). For this, discs of cefotaxime (CTX, 30
µg) and ceftazidime (CAZ, 30 µg) were used, with or without clavulanic acid (CA, 10 µg), both discs with boronic acid (BA, 400 µg) and EDTA (292 μg). An increase of ≥5 mm in the diameter of the inhibitory zone in ei- ther CTX/CA or CAZ/CA was considered a positive result for ESBL production.
DNA extraction was carried out using pure P. mirabilis overnight cultures. A Wiz- ard Genomic DNA purification kit (Promega Biotech, Spain) was used for extraction, ac- cording to the manufacturer’s specifications. ESBL producing genes were identified by PCR
The data were expressed in tables and figures. The relationship between virulence factors and resistance profiles of the isolates were analyzed by binary logistic regression with the SPSS statistical program, version 18.
From the 423 isolates of Enterobacteri- aceae y Morganellaceae analyzed from HUA- PA (Cumana), 46 strains were identified as Proteus mirabilis (11.1%), while out of 274 isolates of Enterobacteriaceae y Morganel- laceae from HSAD (Carúpano), 33 strains (12.4%) were found to be P. mirabilis.
In both hospitals, P. mirabilis were mainly found in long-stay hospital services such as medicine, trauma surgery, ICU, pedi- atrics, neonatology, etc, and were classified as HCAI (76.1% (35/46) and 72.7% (24/33),
in HUAPA and HSAD, respectively). The HCAI isolates were more frequently found in secretions of different types for both hospi- tals (76.1% (35/46) and 45.5% (15/33), in HUAPA and HSAD, respectively). P. mirabilis caused CAI in nine cases with strains isolat- ed from urine (6) and secretion samples (3) in HSAD, while in HUAPA, there were 11 CAI from strains isolated also from urine (6) and secretion samples (5).
No association was found with the type of infection and the clinical factors of the patients. Only three patients had kidney stones reported in HUAPA, while no patients had septic shock reported. Only one patient from the ICU in HSAD had a recurrent in- fection, with the strain isolated in day 10 of hospital stay being resistant to AMP, AMC,
for bla
, bla
and bla
using the prim-
IMP, GM, NN, CIP, and the strain isolated in
TEM
SHV
CTX-M
ers described previously (17-19). As a posi-
day 16 being resistant to the same antimi-
tive control, for the bla
, bla
, and bla
crobial, as well as to CF, STX and MEM:
SHV
TEM
CTX-M
genes, the K. pneumoniae strain ATCC 77915
was used. Amplified products were run in 2% agarose gels, and stained with GelGreen (Bio- tium, UK). The resulting bands were visual- ized and photographically documented in an iBright CL1000 (Invitrogen, USA).
The distribution of the SW, TW, and SW
measurements showed a pattern that allowed clear differentiation of positive and negative strains, according to the established cut- off points, taken from the measurements of the negative controls (Fig. 1). SW motility
Fig. 1. Distribution of the diameters (mm) shown by the P. mirabilis strains in the motility assays. The red lines represent the threshold calculated as described in the text. The diameters were converted into for a better display of the data.
in Proteus isolates showed a pattern com- posed of several concentric rings, where the swarming and consolidation phases are clearly identified (S and C, respectively), but with great morphological variability (Fig. 2). Interestingly, 22/46 isolates of P. mirabilis from HUAPA showed all three types of move- ment, while only 3/33 from HSAD simulta- neously showed all three.
About half of the P. mirabiles isolates showed RSK phenotypes. The serum bac- tericidal assay proved to be an easy test to apply and to discriminate among those who showed resistance or sensitivity (Fig. 3). Re- garding the distribution of virulence factors, the isolates from HUAPA presented higher frequencies of all the virulence factors evalu- ated, compared to those from HSAD (Fig. 4). In addition, the most frequent virulence fac- tor in HUAPA isolates was TM, followed by SW and biofilm production. Similarly, HSAD isolates show biofilm production as a more frequent factor, followed by TM, SW. Even though RSK was the least frequent factor in the strains from both hospitals, its frequency of higher than 40% is of clinical importance. The frequencies of SW, SM and TW were
statistically higher in HUAPA than in HSAD (Table I).
Isolates of P. mirabilis showed high resistance to AMP, CF, IMP, AMC, SAM, GM, NET, NN, CIP, SXT and C (Fig. 5), with
higher frequencies in isolates from HUAPA for CF, CTX, CRO, FEP, SAM, SXT, and C,
while HSAD isolates showed higher frequen- cies for GM and NN. Overall, 44 strains of P. mirabilis were classified as MDR and two as XDR. No statistically significant difference was found in the frequency of MDR/XDR between both hospitals (Fig. 6A, Table I); although no XDR isolates were detected in HSAD. No PDR isolates were found in either of the two hospitals. However, isolates from HSAD showed resistance to 4-6 categories, while those from HUAPA were mostly resis- tant to 7-10 antimicrobial categories. (Fig. 6B). Of the strains causing CAI, three and eight showed MDR phenotypes in HSAD and HUAPA, respectively. Additionally, three of the non-MDR resistant srtrains causing CAI were resistant to IMP and/or MEM in HSAD. When analyzing the presence of MDR strains in relation to the presence of viru- lence factors evaluated (Fig. 7A), a clear
Fig. 2. Different morphology seen in swarming by P. mirabilis in Luria Bertani agar (A, C and D), and blood agar (B). Swarming (S) and consolidation (C) phases are shown.
0
3
Fig. 3. Resistance to serum killing (RSK) according to the serum bactericidal assay, carried out in cystine–lac- tose–electrolyte-deficient agar in isolates of P. mirabilis. T : exposition of the bacteria to human serum without incubation. T : exposition of the bacteria after 3 hours of incubation. A: isolate of P. mirabilis resistant to serum killing. B: isolate of P. mirabilis sensitive to serum killing. C: Distribution of the percentage of reduction (T /T ) shown by the strains, converted into for a better display of the data.
3 0
 CLINICA-004 FINAL/Image_007.gif)
Fig. 4. Frequency of the virulence factors in P. mirabilis isolates from patients attending HUAPA and HSAD, Sucre state, Venezuela. HSAD: Hospital “Santos Anibal Dominicci”, Carúpano. HUAPA: Hospital Uni- versitario “Antonio Patricio de Alcalá”, Cumaná. RSK: Resistance to serum killing according to the serum bactericidal assay.
STATISTICAL ANALYSIS BY BINARY LOGISTIC REGRESSION TO COMPARE THE FREQUENCIES OF VIRULENCE FACTORS IN THE TWO HOSPITALS WHERE THE CLINICAL
ISOLATES OF P. mirabilis WERE FOUND.
Factor | HSAD | HUAPA | Total | P |
SW | 20 (60.6%) | 39 (84.8%) | 59 (74.7%) | 0.005* |
SM | 17 (51.5%) | 33 (71.7%) | 45 (63.3%) | 0.001* |
TM | 21 (63.6%) | 41 (89.1%) | 62 (78.5%) | 0.013* |
Biofilm | 26 (78.8%) | 37 (80.4%) | 63 (79.7%) | 0.476 |
RSK | 14 (42.4%) | 25 (54.3%) | 39 (49.4%) | 0.481 |
MDR/XDR | 20 (60.6%) | 26 (56.5%) | 46 (58.2%) | 0.435 |
P: Probability by binary logistic regression, * P ≤ 0.01, MDR: multidrug and XDR: extensively drug resistant. SW: swarming. SM: swimming. TM: twitching. RSK: Resistance to serum killing according to the serum bactericidal assay.
trend of higher frequency of virulence factors was observed in non-MDR isolates, except for SM, although only statistical significance was observed in SW and TW (Table II). The strains that presented SW or TM were 5.85 and 7.52 times less likely to have MDR/XDR, respec-
than two virulence factors in the same strain were significantly negatively associated to the presence of MDR (Fig. 7B); finding that such strains were 5.59 times less likely to have MDR (Table II). One strain showed ESBL phynotype in HSAD, with amplification of
tively. Furthermore, the presence of more
bla
CTX-M
gene, while one strain without ESBL
 CLINICA-004 FINAL/Image_008.gif)
Fig. 5. Frequency of antimicrobial resistance among the strains of P. mirabilis isolated from HUAPA and HSAD, Sucre state, Venezuela. Antimicrobial acronysms as described in the materials and methods section.
 CLINICA-004 FINAL/Image_009.gif)
P. mirabilis isolates in patients from HUAPA and HSAD, Sucre state, Venezuela. Resistant, Non-MDR: bacteria non-sensitive to two or less antimicrobial categories; MDR: multidrug resistant, non-sensiti- ve to more than two categories; XDR: extensively drug resistant, non-sensitive to all but two or fewer categories, as described in the text.
 CLINICA-004 FINAL/Image_010.gif)
Fig. 7. Relation of multidrug resistance to each of the virulence factors analyzed (A), and the number of viru- lence factors presented by the isolates of P. mirabilis from HUAPA and HSAD, Sucre state, Venezuela. MDR: multidrug and XDR: extensively drug resistant. RSK: Resistance to serum killing, according to the serum bactericidal assay described in the text.
phenotype amplified bla
TEM
gene in that hos-
distribution varies among service units in a
pital. Additionally, five strains showed ESBL
C-
phynotype in HUAPA, amplifying for both bla
hospital, with the patient casuistry, sites of
infection, antimicrobial administration pro-
TX-M
TEM
and bla
genes, while four strains with
tocols, types of infections, control practices
no-ESBL phenotype amplified both of those
genes and one all three genes studied.
In Latin America, there are medical, social and ecological circumstances that favor a dynamic epidemiology of HCAI and CAI produced by Gram-negative bacteria, which are a public health problem, and their
and local resistance (20–22).
In the central hospitals of the Sucre state, Venezuela: HUAPA and HSAD, despite being geographically separated, showed sim- ilar behaviors for P. mirabilis infections. In this study, secretions were the main source of isolation of this species in patients with HCAI, while CAI-associated strains were isolated from urine and secretion samples. HCAIs were favored by long hospital stays,
STATISTICAL ANALYSIS BY BINARY LOGISTIC REGRESSION TO COMPARE MULTIDRUG RESISTANCE AND THE VIRULENCE FACTORS IN THE CLINICAL ISOLATES OF P. mirabilis.
Factor | Non MDR | MDR/XDR | OR | P |
SW | 30 (90.9%) | 29 (63.0%) | -5,85 | 0.009* |
SM | 17 (51.5%) | 33 (71.7%) | ̶ | 0.068 |
TM | 31 (93.9%) | 31 (67.4%) | -7,52 | 0.010* |
Biofilm | 28 (84.8%) | 35 (76.1%) | ̶ | 0.343 |
RSK | 19 (57.6%) | 20 (43.5%) | ̶ | 0.218 |
More than 2 Factors | 28 (84.8%) | 23 (50.0%) | -5,59 | 0.002* |
P: Probability by binary logistic regression, * P ≤ 0.01, OR = odds ratio. MDR: multidrug and XDR: extensively drug resistant. SW: swarming. SM: swimming. TM: twitching. RSK: Resistance to serum killing according to the serum bactericidal assay.
invasive care procedures, among many other factors, and constitute a health problem due to the high mortality they can cause. The higher frequency of P. mirabilis in secretions could be due to failures in the availability, use of antimicrobials, the lack of surveil- lance and epidemiological control of intra- hospital infections in these two evaluated health centers, due to poor clinical-epidemi- ological conditions of the Venezuelan health system. However, similar results have been reported in Ghana (23), who also report Pro- teus in 61.5% of hospital infections and in Bosnia and Herzegovina (24) where this ge- nus was reported as an important intrahospi- tal pathogen in intensive care units, isolated from secretions and urine samples.
Virulence factors in P. mirabilis from HUAPA and HSAD show a variable frequency, but the three most frequent factors were TM, SW, and biofilms in both hospitals, and with highly significant differences between HUA- PA and HSAD. SW was observed with varied intensity and morphology, and although it is a distinctive behavior of Proteus, the rel- evance of SW patterns is not clear and seems to be more as indicator of environmen- tal factors than a regulatory phenomenon (25, 26). However, the frequency of isolates of P. mirabilis not producing SW in strains
from both hospitals agree with the studies in the Czech Republic and Poland (13, 14), where 10 and 17%, respectively, of P. mira- bilis and P. vulgaris were found not produc- ing SW. In animal models, some strains of P. mirabilis require neither flagella nor SW to cause infection, and swarmer cells are rarely observed during infection, suggesting that their role in virulence and colonization may be tissue-specific (27–29).
The frequency of TM, as well as the si- multaneous presence of the three types of motility in 25 isolates of HUAPA and HSAD have also been reported in the Czech Repub- lic (13); but in catheter-associated urinary tract infections, suggesting that these differ- ences may be caused by the growth rates of each isolate, and its ability to adhere to a specific type of catheter or tissue.
Although in P. mirabilis isolates from HUAPA and HSAD, no association was shown between the production of SW, SM, and TM with biofilms, the latter being an important virulence factor. The expression of these types of motility, together with the forma- tion of biofilms, can act independently or be expressed simultaneously, allowing Proteus strains to a better adaptation to the condi- tions of the hospital environment, promot- ing infection in the host, acting as a defense
mechanism for survival and enhancing their virulence. The flagella in SW and TM are known to be involved in surface adhesion and bacterial colonization, which play a crit- ical role in the early stages leading to biofilm formation (30, 31).
Biofilm almost always leads to a sig- nificant decrease in antimicrobial sensitivity compared to cultures grown in suspension (32). This, added to the different resistance mechanisms that bacteria exhibit in them, could explain the high frequency of MDR strains in the two hospitals, where resistance affected many antimicrobial categories.
The O-antigen part of the LPS imparts antigenicity to the bacterial cell leading to the production of antibodies. Nonetheless, it should also be noted that RSK, although it was the least frequent virulence factor in
P. mirabilis for both hospitals, occurred with an important frequency; demonstrating that these isolates are highly pathogenic, be- cause of the variability in the length of the O-antigen chain of the LPS present in these bacteria, can help them avoid the lytic activ- ity of the complement (33). However, in a similar study on Proteus, no correlation be- tween the chemical structure of the specific LPS chains and the resistance, or presence of SW was found (14). However, LPS cause P. mirabilis to have intrinsic resistance to poly- myxins (34), which favors that these MDR strains can significantly increase mortality in patients with severe infections.
The presence of these virulent strains and MDR that are causing HCAI and CAI in patients treated in these hospitals are cause for concern. Mainly those that involve risks of community spread, which could be relat- ed to the indiscriminate use of antimicrobi- als that cause selective pressure on bacteria, leading to a higher prevalence of resistance, which is very common in developing coun- tries like Venezuela.
The Center for Disease Control of USA (CDC) has reported the high increase in an- tibacterial resistance as one of the most im- portant reasons threatening human health
worldwide, and HCAIs are two to three times higher in developing countries, in compari- son with Europe or the USA (11, 22). The results found in the P. mirabilis isolates show the significant increase in antimicrobial re- sistance of this genus in the hospitals evalu- ated in eastern Venezuela. Although the fre- quency of strains showing ESBL phenotypes in these hospitals was low, the presence of two or more genes that confer resistance to beta-lactams in HUAPA strains, in addition to the high frequency of resistant strains, which was caused by many molecular mecha- nisms, is a very important finding, with seri- ous repercutions for the health of the popu- lation in this region. In fact, according to the SENTRY reports (35), increase in ESBL frequencies are reported for USA (10.9%), Europe (16.2%) and Latin America (22.4%).
When correlating virulence with MDR in Proteus isolates from the studied hospitals, virulent and resistant isolates were identified, however, in general, MDR strains presented fewer virulence factors. These results could be explained by the fact that, in some bacte- rial species, the acquisition of a phenotypic advantage such as antimicrobial resistance (beta-lactamases, porins, efflux pumps, or PBP, among others), is associated with a met- abolic or fitness cost, leading to a decrease in virulence and vice versa (36). Although re- sistance is essential for pathogenic bacteria to overcome antimicrobial therapies, adapta- tion and survival in challenging hospital en- vironments, it is often associated with a fit- ness cost motivated by the additional energy expenditure represented by maintenance of these resistance mechanisms or the effects on other essential functions of the cell. Vir- ulence mechanisms, on the other hand, are necessary to overcome host defense systems, favoring the best adaptation of the pathogen, but they also involve a metabolic cost. It can be understood then, that because both adap- tation mechanisms require high-energy costs, the strains tend to establish an equilibrium in the number of mechanisms of each type (37). In Proteus, it has been reported that the de-
crease or inhibition of SW expression may be associated with the presence of MDR in some strains (38).
This study, carried out on strains of P. mi- rabilis in the hospitals of the state of Sucre, represents the first report for Venezuela of the mechanisms of resistance and virulence, allowing evidence of the prevalence of HCAI by virulent isolates and MDR for this spe- cies, which represents a high risk of spread in these hospitals. Furthermore, the presence of strains producing CAI can spread more rapidly in the population, and the isolation of these strains showing MDR phenotypes are great risk for the entire population. It would be very important to be able to deepen the study of possible associations of the factors involved in virulence and antimicrobial resis- tance, to understand how microorganisms de- fend themselves during the infection process, allowing the development of effective control and prevention measures to be implemented to prevent the spread of these strains.
We want to thank the staff of the SA- HUAPA Laboratory of Bacteriology for their valuable collaboration in collecting samples and data.
1. Rózalski A, Sidorczyk Z, Kotełko K. Poten- tial virulence factors of Proteus bacilli. Mi- crobiol Mol Biol Rev 1997; 61:65-89.
Jacobsen SM, Shirtliff ME. Proteus mirabi- lis biofilms and catheter-associated urinary tract infections. Virulence 2011; 2:460-465. https://doi.org/ 10.4161/viru.2.5.17783.
Stock I. Natural antibiotic susceptibility of Proteus spp., with special reference to P. mi- rabilis and P. penneri strains. J Chemother 2003; 15:12-26. https://doi.org/10.1179/ joc.2003.15.1.12.
78:443-448. https://doi.org/10.1016/j.diag microbio.2013.11.025
Bedenić B, Firis N, Elveđi-Gašparović V, Krilanović M, Matanović K, Štimac I, Lux- ner J, Vraneš J, Meštrović T, Zarfel G, Gri- sold A. Emergence of multidrug-resistant Proteus mirabilis in a long-term care facili- ty in Croatia. Wien Klin Wochenschr 2016; 128:404-413. https://doi.org/10.1007/s00 508-016-1005-x
L. Diversidad de β-lactamasas en aislamien- tos clínicos de enterobacterias. Acta Bioquim Clin Latinoam 2012; 46(3):405-412.
1423-1431. https://doi.org/10.1590/S151
7-83822014000400037
Hola V, Peroutkova T, Ruzicka F. Virulen- ce factors in Proteus bacteria from biofilm communities of catheter-associated uri- nary tract infections. FEMS Immunol Med Microbiol 2012; 65:343-349. https://doi. org/10.1111/j.1574-695X.2012.00976.x
Kwil I, Kaźmierczak D, Różalski A. Swar- ming growth and resistance of Proteus pen- neri and Proteus vulgaris strains to normal human serum. Adv Clin Exp Med 2013; 22(2):165-175.
Galani I, Xirouchaki E, Kanellakopoulou K, Petrikkos G, Giamarellou H. Trans- ferable plasmid mediating resistance to multiple antimicrobial agents in Klebsiella pneumoniae isolates in Greece. Clin Micro- biol Infect 2002; 8:579-588. https://doi. org/10.1046/j.1469-0691.2002.00391.x.
Aarestrup FM, Lertworapreecha M, Evans MC, Bangtrakulnonth A, Chalermchaikit T, Hendriksen R, Wegener H. Antimicrobial susceptibility and occurrence of resistance genes among Salmonella enterica serovar Weltevreden from different countries. J Antimicrob Chemother 2003; 52:715-718. https://doi.org/10.1093/jac/dkg426.
Edelstein M, Pimkin M, Palagin I, Edelstein I, Stratchounski L. Prevalence and molecu- lar epidemiology of CTX-M extended spec- trum betalactamase producing Escherichia coli and Klebsiella pneumoniae in Russian hospitals. Antimicrob Agents Chemother
2003; 47(12):3724-3732. https:// doi. org/10.1128/aac.47.12.3724-3732.2003.
Alabi OS, Mendonça N, Adeleke OE, da Silva GJ. Molecular screening of antibiotic- resistant determinants among multidrug- resistant clinical isolates of Proteus mira- bilis from SouthWest Nigeria. Afri Health Sci 2017; 17(2), 356-365. https://doi. org/10.4314/ahs.v17i2.9
Miryala SK, Anbarasu A, Ramaiah S. Gene interaction network approach to elucida- te the multidrug resistance mechanisms in the pathogenic bacterial strain Proteus mirabilis. J Cel Physiol 2021; 236: 468-479 https://doi.org/10.1002/jcp.29874.
Feglo P, Gbedema S, Quay S, Adu-Sarkodie Y, Opoku-Okrah C. Occurrence, species distribution and antibiotic resistance of Proteus isolates: A case study at the Komfo Anokye Teaching Hospital (KATH) in Gha- na. Int J Pharma Sci Res 2010; 1:347-352.
Matsuyama T, Takagi Y, Nakagawa Y, Itoh H, Wakita J, Matsushita M. Dynamic as- pects of the structured cell population in swarming colony of Proteus mirabilis. J Bacteriol 2000; 182:385-393. https://doi. org/10.1128/jb.182.2.385-393.2000.
Morgenstein RM, Szostek B, Rather PN. Re- gulation of gene expression during swarmer cell diferentiation in Proteus mirabilis. FEMS Microbiol Rev 2010; 34:753-763. https://doi. org/10.1111/j.1574-6976. 2010.00229.x
Zunino P, Piccini C, Legnani-Fajardo C. Flagellate and non-flagellate Proteus mira- bilis in the development of experimental uri- nary tract infection. Microb Pathog 1994; 16:379-385. https://doi.org/10.1006/mp at.1994.1038
Legnani-Fajardo C, Zunino P, Piccini C, Allen A, Maskell D. Defined mutants of Proteus mirabilis lacking flagella cause as- cending urinary tract infection in mice. Mi- crob Pathog 1996; 21:395-405. https://doi. org/10.1006/mpat.1996.0070.
Armbruster CE, Mobley HL, Pearson MM. Pathogenesis of Proteus mirabilis infection. EcoSal Plus 2018; 8(1):10.1128/ecosalplus. ESP-0009-2017. https://doi.org/10.1128/ ecosalplus.ESP-0009-2017.
org/10.1016/j.addr.2005.04.007
Thiriard A, Raze D, Locht C. Diver- sion of complement-mediated killing by Bordetella. Microbes Infect 2018; 20(9-
10):512-520. https://doi.org/10.1016/j.mi- cinf.2018.02.002.
2013.11.002.
Chouduri A, Wadud A, Islam A. Exten- ded spectrum multi-drug resistance versus pathogenic factors- swarming, proteases, and urease- of Proteus species. Int Res J Microbiol 2014; 5(1):8-15. https://doi.org/ 10.14303/irjm.2013.052.