Antibiotics resistance in El Tor Vibrio cholerae 01 isolated during cholera outbreaks in Mozambique from 2012 to 2015

Rationale Mozambique has recorded cyclically epidemic outbreaks of cholera. Antibiotic therapy is recommended in specific situations for management and control of cholera outbreaks. However, an increase in resistance rates to antibiotics by Vibrio cholerae has been reported in several epidemic outbreaks worldwide. On the other hand, there are few recent records of continuous surveillance of antibiotics susceptibility pattern of V. cholerae in Mozambique. Goals The purpose of this study was to evaluate antibiotics resistance pattern of Vibrio cholerae O1 Ogawa isolated during Cholera outbreaks in Mozambique to commonly used antibiotics. Methodology We analyzed data from samples received in the context of surveillance and response to Cholera outbreaks in the National Reference Laboratory of Microbiology from the National Institute of Health of Mozambique, 159 samples suspected of cholera from cholera treatment centers of, Metangula (09), Memba (01), Tete City (08), Moatize (01), Morrumbala (01) districts, City of Quelimane (01), Lichinga (06) and Nampula (86) districts, from 2012 to 2015. Laboratory culture and standard biochemical tests were employed to isolate and identify Vibrio cholerae; serotypes were determined by antisera agglutination reaction in blade. Biotype and presence of important virulence factors analysis was done by PCR. Antibiotics susceptibility pattern was detected by disk diffusion method Kirby Bauer. Antibiotic susceptibility and results were interpreted by following as per recommendations of CLSI (Clinical and Laboratory Standards Institute) 2014. All samples were collected and tested in the context of Africhol Project, approved by the National Bioethics Committee for Health. Results Among isolates from of Vibrio cholerae O1 El Tor Ogawa resistance to Sulphamethoxazole-trimethropim was 100% (53/53) to Trimethoprim-, being 100% (54/54) for Ampicillin, 99% (72/74) for Nalidixic Acid, 97% (64/66) to Chloramphenicol, 95% (42/44) for Nitrofurantoin and (19/20) Cotrimoxazole, 83% (80/97) Tetracycline, 56% (5/13) Doxycycline, 56% (39/70) Azithromycin and 0% (0/101) for Ciprofloxacin. PCR analysis suggested strains of V. cholerae O1 being descendants of the current seventh pandemic V. cholerae O1 CIRS 101 hybrid variant. The V. cholerae O1 currently causing cholera epidemics in north and central Mozambique confirmed a CTXΦ genotype and a molecular arrangement similar to the V. cholerae O1 CIRS 101. Conclusion Although V. cholerae infections in Mozambique are generally not treated with antibiotics circulating strains of the bacteria showed high frequency of in vitro resistance to available antibiotics. Continuous monitoring of antibiotic resistance pattern of epidemic strains is therefore crucial since the appearance of antibiotic resistance can influence cholera control strategies.


Introduction
Vibrio cholerae toxin is the virulence factor causing cholera disease, which is characterized by a secretory acute diarrhea. Cholera can lead to severe dehydration and death within hours if not promptly treated. Cholera constitutes a serious public health problem in many parts the world [1,2].
The seventh pandemic of cholera reached Africa in the 70's, arrived in Mozambique in 1973 with cases reported until 1985. Cholera resurfaced in 1989 with over 3,600 cases reported, caused by V. cholerae O1 El Tor serotype Inaba (with some cases of Ogawa), susceptible to Tetracycline, Chloramphenicol and sulfadiazine. [3].
Mozambique continued on experiencing recurrent outbreaks of cholera, in different parts of the country, with different spatial pattern from year to year. Increase in antibiotic resistance including resistance to recommended antibiotic for treatment has also been reported [4,5]. The last decades have seen a growing trend in antimicrobial resistance in Mozambique, previous researches have reported V. cholerae Ogawa O1 El Tor drug resistance in the last decade for Maputo [4][5][6], Zambezia and Tete [5].
Antimicrobial resistance of V. cholerae O1 El Tor is of interest because it's becoming a serious public health problem for many African countries [7]. Therapy with effective antimicrobial agents significantly reduces the duration of diarrhea and hospitalization, reduces the volume of watery feces and need for maintenance fluids. The duration of fecal excretion of V. cholerae is also decreased, reducing transmission of infection to family members, as well as nosocomial infections [8].
Increasing drug resistance is well known and usually varies from one place to another. V. cholerae becomes drug resistant by exporting drugs through efflux pumps, chromosomal mutations or developing genetic resistance via the exchange of conjugative plasmids, conjugative transposons, integrons or self-transmissible chromosomally integrating sxt elements. In addition, V. cholerae, as an environmental organism, have means to acquire resistance genes from intimate contact with intrinsically resistant environmental bacteria, through mobile genetic elements. V. cholerae is able to share these antibiotic resistance genes with other bacteria; and once in the human gut, the bacteria may share these resistance traits with commensals or other enteric pathogens, what complicates antibiotic therapy of many infections [2].
Since antibiotic therapy is recommended in specific situations for management and control of cholera outbreaks, monitoring V.cholerae resistance is important for public health. The aim of this study was to describe spatiotemporally the antimicrobial resistance pattern of Vibrio cholerae O1 El Tor Ogawa isolated in patients admitted to cholera treatment centers (CTCs) and diarrheic disease treatment centers (CTDDs) during outbreak investigations in Mozambique, from 2012 to 2015.

Sample collection and transportation
Rectal swabs from suspected cholera cases were collected during Cholera outbreaks from CTCs and CTDD's from 2012 to 2015 by local laboratory technicians and transported to local laboratories in Cary Blair medium, prepared from dehydrated media. These samples were collected after patient's stabilization and before antimicrobial administration.

Culture and identification
Laboratory technicians, at provincial level, when possible, Vibrio cholerae were identified by standard culture methods; and sent to National Reference Laboratory of Microbiology (NRLM), at central level, for serological confirmation, Antimicrobial susceptibility test (AST) and a double multiplex PCR for mobilome profile analysis.
In the laboratory, samples were enriched in APA broth, then cultured in TCBS media, and suspected colonies were submitted to standard biochemical tests; when positive, colonies were submitted to serological tests for Vibrio cholerae TM Difco BD Poly and Vibrio cholerae TM Difco BD Ogawa (Denka Seiken, Tokyo, Japan).

PCR analyses
Representatively, positive isolates on serological tests were submitted to a double multiplex PCR essay, using five completely sequenced reference strains as positive controls (MJ1236. N1696, B33, CIRS101 and MO10), for the presence of RstR, ctxB, and tcpA genes, genetic markers for El Tor Biotype [10].

Ethical considerations
Analyzed data was from samples collected during Africhol Project, the African Cholera Surveillance Network multi-centric project consisting of an 11 African countries consortium and non-governmental organizations, aiming to collect epidemiological and microbiological information for the occurrence of cholera in Africa to advise for control and preventive measures (http://africhol.org/), which surveillance protocol was approved by the Mozambican National Bioethics Committee for Health. During the study, all participants gave an informed written consent.
We analyzed Africhol data from the laboratory records, anonymously (all samples were codified upon entrance), there was no direct intervention or interaction with human subjects and neither identifiable private information.

Results
Mozambique reported 27 outbreaks in 22 districts (of 145), on 07 of 10 provinces (Fig 1), from 2012 to 2015; and 1522 samples of suspected Cholera cases were received at the National Reference Laboratory of Microbiology, 510 confirmed as V. cholerae serogroup O1 biotype El Tor serotype Ogawa. From these 510, 159 were submitted to AST. Table 1 shows V. cholerae O1 El Tor Ogawa antibiotic resistance during epidemics, between 2012 and 2015. In 04 isolates tested on 2012's outbreaks we found no resistance to Ciprofloxacin and Azithromycin, low levels of resistance to Tetracycline (25%), and 100% resistance to Ampicillin, Nalidixic Acid, Chloramphenicol, SXT and Doxycyclin.
Tendency of resistance from 2012 to 2015 (Fig 2) indicates an increase in antimicrobial resistance for Tetracycline, Nitrofurantoin, and Azithromycin. A sudden decrease and subsequent increase of resistance in Chloramphenicol and SXT, decrease in Doxycycline from 2012 to 2013, 95% resistance in 2015 for Cotrimoxazole and 100% resistance along the years for Ampicillin and Nalidixic Acid. Laboratory analyses also shows that V. cholerae O1 El Tor Ogawa isolated during cholera outbreaks in Mozambique, have no resistance to Ciprofloxacin.
Genetic screening by PCR in 58 isolates revealed three important El Tor epidemic markers, ctxA, rstR2, and tcpA, and the presence of CTX ϕ on chromosome 1 instead of chromosome 2, confirming the profile found in V. cholerae O1 El Tor variants B33 and CIRS 101. ( Table 2)

Discussion
The current study found increasing antibiotic resistance in Vibrio cholerae O1 El Tor Ogawa isolated from Cholera outbreaks from 2012 and 2015 in Mozambique to Tetracycline, Trimethropim-sulphamethoxazol, Chloramphenicol and Nitrofurantoin.
Increasing resistance to tetracycline was consistent with data from Zambia [11], urban and rural Bangladesh [12]; DRC [13]; Nigeria [14] and South Mozambique [4,6]. However, decreasing resistance to tetracycline have been reported in other places, such as Calcutta [15], East Delhi [16] and Ghana [17] and susceptibility to tetracycline, in Puduchuary in India [18], Haiti [19] and north India [20]. Our results indicating resistance to Nitrofurantoin and SXT, match to those observed in earlier studies in Zambia [11], and DRC [13]; and SXT resistance also in a Mozambique rural area [4], and in north India [20].
SXT resistance pattern was in accord with Folgosa et al ten years ago in South and Central Mozambique, having along studied years, all isolates resistant to this antibiotic disc except for one, that was sensitive to it [5].
Another important finding was that isolates were 100% sensitive to Ciprofloxacin activity throughout the studied years; In agreement with Folgosa [5] in 3 provinces in Mozambique (Maputo, Zambezia and Tete), Gujral in Maputo city and province; although these results differ from some published studies in Calcutta [15], East Delhi [16], Nigeria [14], Haiti [19], Iran [21], urban and rural Bangladesh [12], Ghana [17] and Indonesia [22] with resistance or increasing resistance to Ciprofloxacin.
Along the years, isolates in our study were 100% resistant Ampicillin, which is consistent with results from Calcutta [15], Zambia [11] and in contrast with Mozambique in the last decade, North India [20] and Indonesia [22].    Resistance to Nalidixic acid for tested isolates in the study was 100%, being consistent with data obtained from Iran [21], Haiti [19], Mozambique South rural area [4], and an increasing resistance in Calcutta [15,17]. Conversely, in East Deli [16] and urban South Mozambique [6] was found susceptibility to Nalidixic acid in tested isolates.
In our study, isolates were 100% resistant to Chloramphenicol in 2012, registered a sudden drop of nearly 50% in 2013 with subsequent increase; this behavior was seen in Calcutta [15] and Zambia [11]; while susceptibility to Chloramphenicol was found in East Delhi [16]; and Mozambique [4,6].
Regarding Doxycycline and Azythromycin, recomended drugs for Cholera control by WHO, like in Zambia [11] and Ghana [17], there was some resistance for Doxycyclin and like in urban and rural Bangladesh [12] and Haiti [23], 2012-2013 isolates were susceptible to Azythromycin, while 2015's like Ghana [17] presented with some resistance.
PCR screening revealed the same virulence genes found in Mozambique by Folgosa [5] and [24], Cholera toxin ctxA and toxin-coregulated pilus, tcpA, contained by the majority of V. cholerae O1 strains, confirming the profile found in V. cholerae O1 El Tor variants B33 and CIRS 101 (CIRS 101 from Bangladesh). Diferent mobilome profiles have been reported for other African countries where V cholerae occurs and unique Mozambique's profile may be associated with different antibiotic resistance profile.
Due to the plasticity of V. cholerae resulting in the constant emergence of variants, surveillance and characterization of outbreak strains, and their antibiotic resistance determinants, is essential on defining the complex scenario of cholera in this continent as well as worldwide [14].
These findings may be somewhat limited since AST was done according to laboratory discs availability at the moment, usual outbreak scenario where not all antibiotics are tested for each isolate; that could be different on an established antibiotics resistance surveillance scenario, where you predict your demand and therefore your supply. This disabled not only the evaluation of susceptibility profiles for different affected areas and testing for all affected areas but the evaluation of the resistance pattern throughout studied years for Cotrimoxazole, Nitrofurantoin, Azithromycin and Doxycyclin, last two antibiotics specially important because of their reported antimicrobial activity on V. cholerae in previous years and currently in other countries where this pathogen occurs.
Although, serological testing and PCR analyses shows for all strains the same serotype and same genetic similarity, a genetic profile coming from the same clonal origin, it is possible, therefore, to assume a similar resistance pattern for each year.

Conclusion
In general, therefore, it seems that antibiotic resistance profile of V. cholerae regarding same serotype and same year varies in different countries and this study strengthens the importance of having local antibiotic choice based on an updated AST local report.
With a highly frequent and increasing resistance, the current data highlights the importance to monitor antimicrobial resistance in epidemic strains, since the appearance of antimicrobial resistance to commonly used and recommended antibiotics will influence Cholera national control strategies.
as well as Provincial field teams, Cláudio Muianga and Dr. Sérgio Chicumbe are acknowledged for their technical support to this study and on writing this manuscript.