High abundances of class 1 integrase and sulfonamide resistance genes, and characterisation of class 1 integron gene cassettes in four urban wetlands in Nigeria

There is little information about environmental contamination with antibiotic resistance genes (ARG) in Sub-Saharan Africa, home to about 1 billion people. In this study we measured the abundance of three genes (sul1, sul2, and intI1) used as indicators of environmental contamination with ARGs in the sediments of four urban wetlands in southwestern Nigeria by qPCR. In addition, we characterised the variable regions of class 1 integrons in sulfamethoxazole/trimethoprim (SMX/TRI)-resistant bacteria isolated from the wetlands by PCR and DNA sequencing. The indicator ARGs were present in all wetlands with mean absolute copy numbers/gram of sediment ranging between 4.7x106 and 1.2x108 for sul1, 1.1x107 and 1x108 for sul2, and 5.3x105 and 1.9x107 for intI1. The relative abundances (ARG/16S rRNA copy number) ranged from about 10−3 to 10−1. These levels of ARG contamination were similar to those previously reported for polluted environments in other parts of the world. The integrase genes intI1 and intI2 were detected in 72% and 11.4% SMX/TRI-resistant isolates, respectively. Five different cassette array types (dfrA7; aadA2; aadA1|dfrA1; acc(6’)lb-cr|arr3|dfrA27; arr3|acc(6’)lb-cr|dfrA27) were detected among 34 (59.6%) intI1-positive isolates. No gene cassettes were found in the nine intI2-positive isolates. These results show that African urban ecosystems impacted by anthropogenic activities are reservoirs of bacteria harbouring transferable ARG.


Introduction
There are increasing concerns about the presence of antibiotic resistant bacteria (ARB) and antibiotic resistance genes (ARG) in the environment [1,2] where they pose special challenges as environmental contaminants. ARB and ARGs can persist and multiply, and ARGs can spread by horizontal gene transfer (HGT) mediated by mobile genetic elements (MGE) within and beyond the original point of occurrence [2], making possible the exchange of resistance traits between environmental bacteria and human pathogens [3]. ARGs and MGE can be enriched in natural ecosystems impacted by human activities such as wastewater discharge [4], solid waste disposal [5,6], metal contamination [7,8], agriculture [9,10,11], aquaculture [12], and drug manufacture [13]. However, very little is known about environmental contamination with ARB and ARGs in developing countries, especially in sub-Saharan Africa. These are critical knowledge gaps since various risk factors favoring the development and spread of antibiotic resistance exist in these environments [14,15]. Aquatic ecosystems are considered important matrices for the release, mixing, persistence and spread of ARB and ARGs associated with horizontally transferable genetic elements [16,17]. In Nigeria, Africa's most populous nation, wetlands (both coastal and inland) are a key aquatic ecosystem covering about 2.6% of the country's total land surface [18] with the most extensive being the coastal wetlands found in the southern region including the Lagos and Lekki lagoons and the wetlands of the Niger Delta and Cross Rivers [19]. In addition to the coastal wetlands, several riverine wetlands which are extensively used for livestock grazing, farming and fishing activities are scattered across the country [20]. Presently, existing regulations give little attention to the protection and management of Nigerian wetlands [20], hence, they are constantly exposed to human excreta, raw sewage, untreated wastewater and other pollutants from diverse sources. These forms of anthropogenic impact make the urban wetland ecosystems of Nigeria potential reservoirs of ARB carrying ARGs that might spread to other bacteria through HGT mediated by MGE. However, few studies have investigated pollution of natural wetlands with ARGs and none of those studies emanated from Nigeria.
Integrons are important MGE involved in the capture, mobilization and spread of antibiotic resistance genes in bacterial species [21,22]. The integron platform consists of an integrase gene that can recombine discrete units of circularised DNA known as gene cassettes, primary recombination site attI, and a PC promoter that directs transcription of the captured genes [23,24]. Using the sequences of integrase proteins, several classes of integrons have been recognized out of which only a few are important for spreading multidrug resistance in bacteria [25]. Integrons also serve as platform of bacterial evolution [23] and vehicles of gene exchange between the environmental resistome and commensal and pathogenic bacterial species [26] through HGT. Class 1 integrase and sulfonamide resistance genes on has recently been proposed as indicator of pollution by ARB, ARGs and other anthropogenic pollutants [2,27]. In view of this and their association with HGT, an important process for the spread of resistance in environmental reservoirs, it is important to examine genes associated with integrons when investigating antibiotic resistance in the environment.
A few previous studies have reported the detection of clinically relevant ARG in cultivable bacteria isolated from environmental sources in Nigeria [28,29,30,31], but none that we know of used culture-independent quantification to assess the contamination of the Nigerian wetland ecosystems with ARGs. The purpose of this study was to quantify the copy numbers of sul1, sul2, and class 1 integrase gene (intI1) as markers for ARG contamination in the sediments of four polluted wetlands located in Lagos and Ibadan, two of Africa's most populous cities. Additionally, because integrons are commonly reported in members of the Enterobacteriaceae [32] and are found in human impacted ecosystems [25], we investigated the occurrence of class 1 and 2 integrons and their gene cassette contents in 79 sulfamethoxazole/ trimethoprim (SMX/TRI)-resistant bacteria of the family Enterobacteriaceae isolated from the wetlands.

Sample sites and sample collection
Sediment samples (n = 16) were collected monthly from four wetlands located in Lagos and Ibadan, southwestern Nigeria between October 2014 and January 2015 (Fig 1). The sampling period represented the two major Nigerian seasons "rainy" (sampling in October and November) and "dry" (sampling in December and January). The two wetlands sampled in Ibadan are Awba (AW) (07.4468˚N, 03.8763˚E) and Apete (AP) wetlands (07.4577˚N, 03.8828˚E) while in Lagos the two wetlands sampled are Abule-Agege (AA) (06.5145˚N, 03.4002˚E) and Ogbe Creek (OC) (06.5135˚N, 03.3937˚E). AW, AP and AA receive untreated wastewater and raw sewage from student hostel facilities of two universities. In addition, the wetlands receive waste streams from fish farms (AW and AA), a Zoological Garden (AW), and seepages from upland solid waste dumpsites (AP and OC). Triplicate sediment samples collected from the upper 1 cm portion of each wetland were pooled to form a composite and stored at -80˚C until processed for analysis.

DNA extractions
Total sediment community DNA was extracted using FastDNA Spin Kit for soil (MP Biomedicals, Ohio, USA) according to the manufacturer's instructions. DNA concentrations were estimated with a NanoDrop 1000 (NanoDrop Technologies, Wilmington, DE). Genomic DNA extraction from isolated SMX/TRI-resistant bacteria (see below) was performed by the microwave boiling method [33]. Extracted DNA was stored at -20˚C until used.
sample. The assay was run on a StepOne Plus Cycler (Applied Biosystems) in a 20 μl reaction mixture. The PCR conditions were 95˚C for 2 min, followed by 40 cycles of 95˚C for 20 s, 20 s at the respective melting temperatures and 72˚C for 20 s. Primers for sul genes were those reported by Wang et al. [34] and intI1 primers those by Mazel et al. [35]. Dilutions of template DNA were used for each sample to compensate for the effect of PCR inhibitors in the samples. The copy number of the 16S rRNA gene in the samples was determined by using Primers 519F and 909R [36] as a measure of total bacterial abundance. Standards were PCR-amplified fragments of the 16S rRNA gene from Escherichia coli as well as the sul genes and intI1 from environmental isolates obtained in this study. DNA concentrations of the standards (c) were measured by Nanodrop spectrophotometry and the copy number (CN) per gram of wetland sediment was calculated using the relation:

Statistical analysis
Measured copy numbers of the genes were log-transformed to normalise the distribution before the statistical significance of the differences in the copy numbers of the genes at each sampling site was determined by analysis of variance (ANOVA) at 5% level of significance (P<0.05). Pearson's Correlation analysis was used to test the association between measured copy numbers of the 16S rRNA gene, sul1, sul2, and intI1.

Isolation of SMX/TRI-resistant bacteria
Sediment samples (1 g) were suspended in saline and 100 μl of appropriate dilutions spread on Eosine Methylene blue (EMB) agar plates supplemented with SMX/TRI (56 and 8 μg/ml, respectively). The SMX/TRI combination (co-trimoxazole, Septrin) is commonly used in Nigeria. Bacterial colonies growing on the plates following incubation for 48 h at 30˚C were randomly selected, re-streaked on Mueller-Hinton Agar (MHA) plates with SMX/TRI until apparent purity, and the isolates stored in Müller-Hinton broth with 15% (v/v) glycerol at -80˚C until further molecular analysis.
The 16S rRNA gene of SMX/TRI-resistant isolates was PCR amplified with universal primers 27F and 1439R [37], amplicons were sequenced (GATC Biotech, Cologne, Germany), and the strains identified via BLASTn queries in the GenBank. The 16S rDNA sequences were deposited in the GenBank database under accession numbers MG859672-MG859730.

Literature survey for qPCR data on sul and intI1 genes in aquatic sediments
In order to compare abundances of sul and intI1 genes in the Nigerian wetlands with values reported in the literature we searched the ScienceDirect and Scopus databases using the keywords "antibiotic resistance genes" and "wetlands". We chose studies that reported the genes' relative abundances for sediments in natural aquatic environments. Eleven studies from China, Pakistan, Poland, USA, Sweden and Switzerland but none from Africa met our criteria [42][43][44][45][46][47][48][49][50][51][52][53]. In cases where more than one season was examined in a study, we extracted the values for summer in order to be closer to the climatic conditions in Nigeria. When data from different sample sites was provided, we took the highest values.

Ethical statement
There are no specific permits required for sample collection in the field studies. The wetlands not protected and not privately owned, hence there are no regulations that restrict collection of sediment samples for research purposes from the four wetlands. The field study does not involve any endangered or protected species, only sediment samples were taken from the sites.

Abundances of intI1 and sul genes show medium to heavy pollution with antibiotic resistance genes in four urban Nigerian wetlands
The intI1, sul1, and sul2 genes were detected in all the samples analysed. The mean absolute copy numbers/gram of wetland sediment of the genes and the 16S rRNA gene are shown in Fig 2. The mean copy numbers of 16S rRNA gene ranged from 1.6x10 8 (December 2014, AA) to 5.7x10 9 (December 2014, AP) copies per g of sediment. The intI1 gene was detected in the wetlands at values ranging from 5.3x10 5 (January 2015, OC) to 1.9x10 7 (January 2015, AW) copies/gram sediment. These abundances were in the range reported for aquatic sediments in other regions of the world [25]. It should be mentioned that the respective studies cited in [25] used various primers for the quantification of the intI1 gene, which hampers the comparative assessment of the levels of intI1 contamination in the respective contaminated sites.
The copy number/gram sediment of sul1 in the four wetlands ranged from 4.7x10 6 (November 2014, OC) to 1.2x10 8 (November 2014, AW). The absolute abundances of sul2 was higher than that of sul1 at 1.1x10 7 (November 2014, OC) to 1x10 8 (December 2014, AP) copy number/gram sediment. There was no universal trend regarding the ratio of sul1 to sul2 in sediments. Greater abundance of sul2 over sul1 was found by Luo et al. [43] similar to what we found in our study sites, similar abundance of both genes were measured by Jiang et al. [54] and Lu et al. [55], while a higher abundance of sul1 was reported by Gao et al. [56] and Koczura et al. [48].
To the best of our knowledge, there are no similar studies of ARG contamination of the Nigerian environment available for the comparative analysis of our results. Overall, the absolute levels of sul contamination in the wetlands of this study are in the upper range of values reported for wetlands and river sediments in other parts of the world [42,46,48,51, [55][56][57][58][59]. The abundance of the sul genes as well as those of intI1 were significantly different (p<0.01) across the four wetlands. In particular the AW and AA sites had sul abundances mirroring those of heavily contaminated sites elsewhere [42,60]. These differences in ARG abundance in the four wetlands correlates with the different anthropogenic activities associated with the wetlands. AW receives direct input of untreated domestic wastewater, discharges from a fish farm and a Zoological Garden. The only known point source of anthropogenic input identified for OC, the wetland with the lowest ARG abundance, is an upland solid waste dumpsite which appears to be relatively new at the time of sample collection. There was no significant difference in the abundance of the ARGs measured for the samples collected in the rainy season (October and November) and those collected in the dry season (December and January).
The absolute abundance of the sul genes showed positive correlation with the copy number of intI1 (P<0.01 or P<0.05). The correlation coefficient of sul1 (0.794) is slightly higher than that of sul2 (0.729). This is expected as sul1 is typically located in the 3' conserved segments of class 1 integrons [32]. However, the mean absolute copy number of intI1 in all the samples was generally lower than the mean absolute copy number of sul1, which is similar to what has been reported in previous studies [47,48].
It should be mentioned that the intI1-directed primers [35] used in our study did not exclusively target clinical intI1 and thus may have resulted in an overestimation of anthropogenic pollution with the clinical class I integron. Yet that overestimation might not be substantial. Antelo et al. [61] used the same primers as in our study in a PCR-based survey on integron diversity with samples from King George Island, Antarctica, and found a prevalence rate of 42% of clinical class I integrons. On the other hand, the sul-targeting primers used here were designed based on the sequence of a single allele of the respective sul gene. This could have resulted in an underestimation of the sul gene abundances in our study sites as a recent study has shown that a wide variety of both clinical and non-clinical sul alleles are present in polluted natural ecosystems [62].
Essentially the same picture of the level of contamination of the wetlands investigated here is apparent from the relative abundance of intI1, sul1, and sul2 over 16S rRNA gene copy numbers (Fig 3). The relative abundance of intI1 was similar to the values reported for aquatic sediments elsewhere [25], while the relative abundance of sul genes indicate heavy contamination, especially of the AW and AA sites. These relative abundances are similar to values found at a pig farm where SMX was used [63], mariculture sites in China [64], and water bodies around a drug formulation facility in Pakistan [52].

SMX/TRI genes and integrons are prevalent in bacterial isolates from the wetlands
To understand the genetic context of the SMX/TRI resistance genes and identify other transferable genes associated with detected ARG markers, SMX/TRI-resistant Enterobacteriaceae were isolated from the four wetlands and screened by PCR for sulfonamide resistance genes sul1, sul2, sul3, the trimethoprim resistance genes dfrA1, dfrA5, dfrA7, dfrA12, dfrA17, and integrons class 1, 2 and 3. Despite its obvious limitations, data from cultivable bacteria are still important in the study of antibiotic resistance. They provide insight into the phenotypic and genotypic characteristics of bacteria isolates and are therefore integral to national and international antibiotic resistance surveillance and tracking efforts. Further, they also provide full insight into the resistome of natural ecosystems and the link between antibiotic resistances detected in the environment and clinical settings when combined with data generated from high throughput culture-independent methods [65].
Out  (1). Some of the isolates shared identical 16S rRNA sequences, however, the pattern of genes detected in the isolates were either different or where the isolates shared identical pattern of genes, they were isolated at different time periods. sul1 was the most frequently detected gene, found in 40/79 (51%) strains (AW = 18, AP = 16 and AA = 6), followed by sul2 found in 32/79 (41%) strains (AW = 10, AP = 10, AA = 10, OC = 2), and sul3 found in only 1 strain from AW also carrying sul1 and sul2. dfrA1, dfrA12 and dfrA7 were detected in 1, 2 and 10 isolates, respectively, dfrA5 and dfrA17 were not detected in any of the isolates. The gene dfrA7 occurred with sul1 in 4 isolates and with sul1 and sul2 in five isolates.
intI1 and intI2 were detected in 57 (72%) and 9 (11.4%) SMX/TRI-resistant isolates respectively, including eight bacteria (intI1 = 7, intI2 = 1) where none of the SMX/TRI genes were found. intI3 was not detected in any of the bacterial strains investigated. This is consistent with previous reports that class 1 and 2 integrons are the most frequently detected integrons in Gram negative bacteria [66] with class 2 integrons detected less frequently than class 1 integrons.

Characterisation of integrons gene cassettes
Because of the important role of integrons in the spread of resistance in natural ecosystems, we carried out PCR analysis together with DNA sequencing of the variable regions of detected intI1 and intI2 among isolated SMX/TRI-resistant bacteria to identify additional ARGs that may potentially spread by HGT. In all intl2-positive isolates, the primer pair attl2 and orfX failed to yield any useful amplification product in all cases indicating that they are likely to be novel integrons or integrons without gene cassettes. In contrast, variable regions containing gene cassettes were amplified in 34 of 57 (59.6%) intI1-positive isolates including three isolates where none of the tested SMX/TRI genes was detected. The remaining intI1-positive isolates either lack gene cassettes or might be derivatives of Tn5090-like class 1 integrons that lack the 3' conserved end [67,68]. Seven different types of genes arranged in five different arrays were found within the variable regions of the class 1 integrons in the 34 isolates (Table 1). The genes encode resistance to trimethoprim (dfrA1, dfrA7, dfrA27), aminoglycosides (aadA1, aadA2), rifampicin (arr-3) and fluoroquinolones (acc(6')-lb-cr). The dfrA genes were the most frequently detected gene cassettes occurring in (82%) of the gene cassette-bearing isolates. This is similar to previous reports by Lin et al. [69] and corroborated previous reports that class 1 integrons are important for the dissemination of SMX/TRI resistance genes [70].

Diversity of promoters of class 1 integrons
We analysed the promoter types in the intI1-positive isolates. Two different Pc promoter types, the weak promoter PcW and the Hybrid Type 1 promoter PcH1 were detected in 26 and 8 isolates, respectively, of 34 intI1-positive isolates. In all cases, the detected promoters were associated with an inactive second promoter type P2 where the -35 and -10 hexamers of the respective promoters were separated by 14 bp instead of the 17 bp expected with an active P2 [91]. Jové et al. [40], Vinue et al.
[41] and Moura et al. [92] similarly reported PcW as the most frequently detected promoter type in bacteria from clinical and environmental sources. Since promoter strength and integrase excision activity are inversely correlated, thus affecting the propensity for the dissemination of resistance genes [40]; the prevalence of weak PC variants in this study suggests that there is a high propensity for further dissemination of the antibiotic resistance genes carried on the class 1 integrons among bacteria in the wetlands.
In summary, indicator ARGs sul1, sul2, and intI1 were detected in all four Nigerian wetlands at high absolute and relative (vs. the 16S rRNA gene) gene frequencies. Class 1 and 2 integrons were found in 72% and 11.4% of 79 culturable bacteria from the wetlands with five gene cassette arrays containing 7 different resistance genes found in 59.6% of detected class 1 integrons. No gene cassette was detected in the class 2 integrons. Our results demonstrate that the Nigerian wetlands are contaminated with ARGs at levels similar to those usually reported for highly contaminated sites in other parts of the world. Although only a limited number of ARGs and wetlands were selected, this study provides the first data on the status of ARG contamination in any Nigeria's aquatic ecosystem and suggested a role for anthropogenic activities in environmental contamination with ARGs. Larger studies are needed for an overall evaluation of the magnitude of ARG contamination in the Nigerian aquatic ecosystem and its associated risks.