MALDI-TOF identification of Campylobacter isolated from patients consulted in private laboratories in France

1 Laboratoire de Biologie Moléculaire, d’Épidémiologie et de Surveillance des Bactéries et Virus Transmissibles par les Aliments (LaBESTA)/ Centre de Recherche en Sciences Biologiques, Alimentaires et Nutritionnelles (CRSBAN)/Université Ouaga I Pr Joseph KI-ZERBO, 03 BP 7021 Ouagadougou 03, Burkina Faso. 2 Unité de Formation et de Recherche en Sciences Appliquées et Technologies (UFR/SAT)/Université de Dédougou, BP 176 Dédougou Burkina Faso. 3 INSERM, University of Bordeaux, UMR1053 Bordeaux Research In Translational Oncology, BaRITOn, 33000 Bordeaux, France. 4 French National Reference Center for Campylobacters and Helicobacters, Pellegrin Hospital, 33076 Bordeaux, France.


INTRODUCTION
Campylobacter enteritis was first identified by Butzer in the early 1970s. This pathogen is considered as one of the leading bacterial species of foodborne diseases in humans around the world (Abdi-Hachesoo et al., 2014;CNRCH, 2018). As a result, campylobacteriosis is a major public health concern in many developed countries (Wardak et al., 2007) and in developing countries infection has strikingly increased in recent years    Gwimi et al., 2015). Human campylobacteriosis has been linked to mishandling and consumption of contaminated poultry (Humphrey et al., 2007). Among the pathogenic species, Campylobacter jejuni and Campylobacter coli are leading causes of foodborne gastroenteritis and enteritis in humans worldwide (Friedman et al., 2000;Reddy and Zishiri 2007). In most European countries, the majority of Campylobacter infections are domestically acquired (EFSA, 2010). Thus, in 2017 several species of Campylobacter with a dramatic increase of C. jejuni were isolated in many blood cultures of patients. Nevertheless, C. jejuni along with C. coli and C. fetus, mostly isolated in stool, remain the common Camplylobacter of human pathogens (CNRCH, 2018). Similarly, other Campylobacter spp. such as Campylobacter lari and Campylobacter upsaliensis have been implicated in human gastrointestinal infections (Obeng et al., 2012;CDC, 2013).
Most Campylobacter enteric infections are self-limited and do not require antimicrobial drug treatment. However, because of severe or long-lasting Campylobacter infections their treatment may require antimicrobial drug therapy (Gallay et al., 2007). Therefore, macrolides as first-line therapy and fluoroquinolones as alternative therapy are recommended (Nachamkin and Blaser, 2000;Gallay et al., 2007). However, the resistance of Campylobacter to antimicrobial agents has substantially increased during the past two decades and become a matter of concern in severe human Campylobacter infections (Lucey et al., 2002;Nachamkin et al., 2002). The objectives of the present study were therefore 1) to identify Campylobacter species in patients consulted in private laboratories in France, using the MALDI-TOF method and 2) to test the bacteria's susceptibility to antimicrobials. With the low identification rates and continuous resistance of pathogenic bacteria species to antibiotics, an accurate and efficient analytical method such as MALDI-TOF MS will be a robust tool for controlling the extend of bacterial infections in both developed and developing countries.

Bacterial strains
A total of 197 Campylobacter-like strains were collected in September 2018 by CNR. Each strain was identified after being subcultured on a trypticase soy blood agar plate (bioMérieux, Marcy l'Etoile, France) and incubated overnight in a microaerobic atmosphere at 37°C. A typical Campylobacter colonies obtained were used for mass spectrometry identification.

Sample preparation
A part of a colony of each isolate, taken directly from the agar plate Dembélé et al. 95 after 18-24 h of incubation to obtain fresh bacteria, was deposited on a microtitre 384 target plate ground steel T F, (Bruker Daltonics, Bremen, Germany) in a single spot and allowed to dry at room temperature. One microlitre of matrix solution (saturated solution of a cyano-4-hydroxycinnamic acid in 50% acetonitrile) was added to the sample and was then crystallized through air-drying at room temperature for 5 min.

MALDI-TOF mass spectrometry measurements
MALDI-TOF mass spectrometry measurements were performed with an Ultraflex III TOF/TOF mass spectrometer (Bruker Daltonics) equipped with a 200-Hz smartbeam 1 laser. The parameter settings were as follows: delay, 80 ns; ion source, 1 volt, 25 kV; ion source, 2 volts, 23.4 kV; lens voltage, 6 kV; and mass range, 0-20 137 kDa. Each run included an Escherichia coli control sample provided by Bruker Daltonics where the presence of eight specific proteins insured that the spectrometer was set properly. Raw spectra of the strains were analysed by MALDI BIOTYPER 2.0 software (Bruker Daltonics) using the default settings (all of the settings were potentially adjustable). The whole process from MALDI-TOF mass spectrometry measurement to identification was performed automatically without any user intervention. Briefly, the software generated a list of peaks up to 100. The threshold for peak acceptance was a signal-to-noise ratio of 3. After alignment, peaks with a mass-to-charge ratio difference of <250 ppm were considered to be identical. The peak list generated was used for matches against the reference library, by directly using the integrated pattern-matching algorithms of the software. All parameters were the same regardless of the presumptive bacterial species analysed. Concerning Campylobacter and related species, the BIOTYPER 2.0 database was composed of four Arcobacter butzleri, two A. cibarius, two A. cryaerophilus, one A. halophilus, one A. nitrofigilis, two A. skirrowii, three Campylobacter coli, five C. fetus, four C. helveticus, two C. hyointestinalis, six C. jejuni, four C. lari, one C. sputorum and four C. upsaliensis. The spectra were obtained in the positive linear mode after 1000 shots (size, 61 794 points; delay, 232 points). A score was attributed to each identification. When this score was >2.00, the identification was considered correct at the species level; between 1.7 and 1.999, the identification was considered correct at the genus level; and <1.7, the identification was not similar enough to a spectrum to draw a conclusion.

Antimicrobial susceptibility test
The antimicrobial susceptibility was performed by the agar diffusion method according to the criteria proposed by the CA-SFM and harmonized according to the criteria proposed by EUCAST (EUCAST, 2018): MH-F medium (Mueller-Hinton + 5% defibrinated horse blood and 20 mg/L β-NAD) (bioMérieux, Marcy l'Etoile, France) was used with an inoculum corresponding to 0.5 McFarland. Six antibiotics belonging to five families were tested: Ampicillin, Amoxicillin-clavulanic acid (beta-lactams), ciprofloxacin (quinolones), erythromycin (macrolides), tetracycline (cyclines) and gentamicin (aminosides). The plates were incubated in microaerobic atmosphere conditions at 35 ± 2°C, 24 h in microaerobic jar conditions (generation of atmosphere using a Anoxomat (Smart)). The reading at 24 h (or 48 h) was performed using the SIRScan system (I2A, Montpellier, France) then visual verification of the diameters read on the camera. A biologist always checks the values read. At the validation, any discrepancy with the result reported by the correspondent was verified and if necessary indicated on the final report. Campylobacter jejuni ATCC 33560 was used as quality control.

Data processing
Data were analyzed using the software package Epi Info 7.1.2.0 (Centers for Disease Control and Prevention [CDC], Atlanta). Multivariable logistic regression was used to estimate odds ratios (ORs) with 95% confidence intervals (95% CI) also calculated. The statistical significance was evaluated using the Fischer exact 2tailed p value, and a p ≤ 0.05 was considered significant.

DISCUSSION
The identification of Campylobacter species and related organisms at the species level has always been difficult using phenotypic methods because of their low metabolic activity, whereas molecular methods are more reliable but time-consuming (Bessède et al., 2011). The development of MALDI-TOF MS, a rapid and cost effective analytical method, has profoundly improved the bacterial identification process (Mandrell et al., 2005;Kolinska et al., 2008;Alispahic et al., 2010). In overall, we isolated about 88.8% of a combined species of Campylobacter. This is consistent with reported percentages of Campylobacter isolated in Nigeria (62.7%; Ewnetu and Mihret, 2010) and in Ethiopia (72.7%; Gwimi et al., 2015). The relative high percentage of Campylobacter's identification may be attributed to the use of mass spectrometry by the CNR. Several studies, (e.g. 20%, Coker et al., 2002;5.8 -9%, Girgis et al., 2014;17.3%, Karikari et al., 2017), have also reported much lower identification rates of Campylobacter than those recorded in this study. The incidence of Campylobacter-associated food poisoning has gradually increased, and this organism is now considered as the leading cause of bacterial gastroenteritis worldwide (Bessède et al., 2011). A study has shown that campylobacteriosis incidences have been globally in rise in the past decade. Thus, the numbers of campylobacteriosis incidences have increased in North America, Europe and Australia (Kaakoush et al., 2015).   Incidences and number of cases may substantially vary among countries or regions or within a given country (Kubota et al., 2011;Sadkowska-Todys and Kucharczyk 2012). These variations are attributable to several factors such as sensitivity of detection methodologies, geographic locations, target population, differences in the standard and stringency of biocontrol protocols, surveillance bias, food practices as well as the availability of natural reservoirs of Campylobacter species (Kaakoush et al., 2015).
In contrast, the identified species were C. jejuni (72.59%), C. coli (14.21%), Arcobacter butzleri (1.02%), C. fetus and C. lari (0.51% each). These results are similar to those previously reported in France: C. jejuni (78%), C. coli (14%) and C. fetus (4%) (Bessède et al., 2011). Because of the low identification rates of the analytical methods, incidences of C. jejumi and C. coli infections are likely to be underestimated (Wagenaar et al., 2013). In most industrialized countries, Campylobacter organisms are, along with Salmonella, the most common cause of foodborne bacterial gastroenteritis (Allos and Blaser, 1995;Frost, 2001). Furthermore, consumption of undercooked (raw, rare, or "pink") chicken and beef was the most important food-specific risk factor for Campylobacter infection in France (Gallay et al., 2008;Berthenet et al., 2019). In the present study, high rates (22%) of Campylobacter were identified in children under 10 years old. This result was consistent with those reported in 2017 by French National Reference Centre for Campylobacter and by developing countries (Coker et al., 2002). While infection with C. jejuni or C. coli can occur in patients of all ages, a recent study in Denmark showed that infection is more prevalent in toddlers (1 to 4 years) and young adults (15 to 24 years) than in other age groups (Nielsen et al., 2013).
Our study also showed that Campylobacter infections were more prevalent in male (59.4%) than in female (40.6%) patients. This is corroborated by previous studies, which reported high campylobacter prevalent in male patients compared to female patients (Friedman et al., 2000;Fitzgerald et al., 2011). However, some studies found that females have higher risk of getting infected by Campylobacter than males (Gillespie et al., 2006;Karikari et al., 2017).The isolated strains in our study showed considerable resistance to ampicillin (26.4%), to tetracycline (30.5%) whereas the resistances to ciprofloxacin, erythromycin and gentamicin were relatively low (0.5-1%). These results are in agreement with 30.7% resistance to tetracycline, 26.9% to ciprofloxacin, 1.7% to erythromycin and 0.9% to gentamicin previously reported in France (Gallay et al., 2008). A similar resistance to tetracycline (22%) was documented in Ethiopia (Ewnetu and Mihret, 2010). Furthermore, our results are lower than 48.0% resistant to ampicillin reported by Gallay et al. (2008). High resistance rates to tetracycline have been described in Ghana (92.3 -100%, Karikari et al., 2017), in Spain (72%; Prats et al., 2000) in human isolates. At the global scale, the tetracyclines are a heavily used class of antibiotics both in human and in veterinary medicine (Iovine, 2013). The discovery in 1950 that the addition of antibiotics to animal feed at sub therapeutic levels could lead to increased growth rates of these animals. This results in research into methods to improve or stabilize meat supplies to the consumer (Kaakoush et al., 2015). For instance, by the turn of the 20th century, the majority of antibiotic used in the United States was for agricultural purposes. This approach has led to a dramatic increase in antibiotic resistance in several human pathogens that originate from domesticated animals, including Campylobacter species (Barton, 2014). Since the beginning of the 1990s, the resistance of Campylobacter organisms to antibiotics has increased (Gallay et al., 2008). Thus, a strong relationship has been observed between the amounts of fluoroquinolone in animal feed and the presence of pathogen strains of Campylobacter in humans (Kaakoush et al., 2015). Although the Campylobacter infection is self-limiting, the extraintestinal infection or septicaemia may occur, and thereby requiring treatment using appropriate antibiotics.

Conclusion
The identification by MALDI-TOF mass spectrometry is particularly efficient for the identification of campylobacters (should be campylobacters, low cas for c) and makes it possible to identify genera and species difficult to access by traditional identification tests (phenotypic and molecular methods): C. lari, C. upsaliensis, C. "anaerobic" "(C. ureolyticus in particular), Arcobacters and enterohepatic Helicobacters, some of whose pathogenesis is close to Campylobacter (especially Helicobacter pullorum, Helicobacter cinaedi). The systematic identification by mass spectrometry, a quick and inexpensive method, of several colonies (including of atypical aspect) pushing on the selective media increases the rate of detection. Therefore, it seems that the results using mass spectrometry correlates with the profiles of the control strains using mass spectrometry but further confirmation is needed.