Antibiotic resistance genes and class 1 integron: Evidence of fecal pollution as a major driver for their abundance in water and sediments impacted by metal contamination and wastewater in the Andean region of Bolivia

Water and sediment samples affected by mining activities were collected from three lakes in Bolivia, the pristine Andean lake Pata Khota, the Milluni Chico lake directly impacted by acid mine drainage, and the Uru-Uru lake located close to Oruro city and highly polluted by mining activities and human wastewater discharges. Physicochemical parameters, including metal compositions, were analyzed in water and sediment samples. Antibiotic resistance genes (ARGs), were screened for, and verified by quantitative PCR together with the mobile element class 1 integron (intl1) as well as crAssphage, a marker of human fecal pollution. The gene intl1 showed a positive correlation with sul1, sul2, tetA and blaOXA-2. CrAssphage was only detected in Uru-Uru lake and its tributaries and significantly higher abundance of ARGs were found in these sites. Multivariate analysis showed that crAssphage abundance, electrical conductivity and pH were positively correlated with higher levels of intl1 and ARGs. Taken together our results suggest that fecal pollution is the major driver of higher ARGs and intl1 in wastewater and mining contaminated environments.


42
Antibiotic resistance is considered a major threat to human health worldwide 43 (1). Antibiotic resistance genes (ARGs) have been classified as emergent 44 contaminants, with a significant impact in aquatic environments due to the 45 possibility to be acquired by pathogens, which could lead to public health issues 46 (2). Novel rearrangements of ARGs and mobile genetic elements (MGEs), that 47 favor their dissemination, are considered xenogenetic pollutants. These elements 4 72 different ARGs (14), whereas both Zn and Pb levels were correlated with the 73 abundance of erythromycin resistance genes in wastewater treatment plants (15). 74 These data suggest that, aquatic environments are important for the ecology and 75 evolution of ARGs. In particular, water bodies can be hotspots for the evolution of has a dam that captures water that supplies water to the Puchucollo drinking water 120 treatment plant, then water is distributed to the cities of La Paz and El Alto.

121
Mining activities were performed in the valley until 1990 by the company 122 COMSUR, water from the surrounding lakes was used in the mining activities. Acid 123 mining drainage (AMD) were discharged directly in the Milluni Chico lake, also 124 contaminating the downstream lake of Milluni Grande. As a consequence, these 125 two lakes acquired an extremely acid pH that favored the mobility of metals (Cd,126 Zn, As, Cu, Ni, Pb, Sn) in water and sediments (26). In contrast, the first lake Pata 127 Khota, is fed by water proceeding from the melting of Huayna Potosi mountain.

128
Anthropogenic activities are very limited around this site, and the lake is 129 considered an ecologically intact environment (26,27).   Samples were kept at 4°C with cold packs, and transported to the laboratory in La 165 Paz city, were they were rapidly stored at -70°C until their analysis. Surface water 166 samples from UP1, UP2 and UP3 were collected in triplicate, filtered (300 mL) 167 through 45 µm nitrocellulose filter membranes (Sigma-Aldrich), and the filters were 168 immediately stored at -70°C until their analysis.

169
Quantification of metals 170 Six elements were quantified in the sediment and acidified water samples: Cu,

171
Zn, Pb, Ni, Cd, and As. All the analyses were performed as previously described   predispensed primers for 85 different ARGs (S1 Table) (Table 1) were designed using Primer-BLAST (NCBI) (34). A six point calibration 211 curve was generated using serial dilution from 10 6 to 10 1 copies of the plasmid.

212
The 16S rRNA housekeeping gene was used for the normalization of the absolute 213 quantification of ARGs.  had pH values close to neutral (Fig 3a). showed a significant ten to twenty-fold increase compared to PK, with the 273 exception of UP1 (Fig 2b).  The correlation analysis of all the physicochemical parameters, and metal 315 levels ( Fig. 3) showed that metals levels were positively correlated among them.

316
EC and pH showed a negative correlation between them and no significant 317 correlation with other environmental parameters was observed. sampling sites with positive signal for crAssphage (Fig 6). To our knowledge, the present study is the first to explore the relationship 429 between ARGs, metal pollution, and wastewater discharges. In order to establish 430 these relationships, we analyzed the abundance of different ARGs and the class 1 431 integron (intl1) in three water bodies. PK site, is a glacier lake which could be 432 considered an ecological intact environment with very few anthropogenic activities 433 around. Like PK, MC, is also a glacier lake but heavily impacted by mining 434 discharges. UP, the third site, is a peri-urban lake with a long history of receiving host in water. In the case of Pata Khota, the absence of this fecal pollution marker 471 could reflect either, the absence of its host or the pristine character of this site.

472
In agreement with previous studies (28) sul2. All these genes presented positive correlations among them (Fig 5).

488
Commonly intl1 can be found in the environment positively correlated with several 489 ARGs (6) and its abundance is strongly correlated with the abundance of multi- ARGs abundance (18).

506
Some studies showed that metals such as Cu, Zn, Cd, and Ni can exert 507 stronger selection pressure over environmental microbial communities favoring the 508 selection of resistant bacteria, even more than antibiotics themselves (13, 45).

509
Even though metal levels in MC were higher than in PK, we found similar ARGs 510 abundance in both lakes. In fact, they clustered together according to its ARGs 511 abundance (Fig 4) and were very close to each other in the PCA analysis (Fig 7).

512
Samples collected in UP2, that drains sewage from Oruro city to UP3 presented 513 the highest abundance of ARGs and intl1. Remarkably, there are significantly 514 higher levels of all the measured metals in UP3. However, the ARGs abundance in 515 both sites grouped together in the hierarchical clustering analysis (Fig 4). These 516 results suggest that other parameters different from metal levels are explaining the 517 variation in ARGs abundance.

518
Previous studies reported positive correlations of ARGs and crAssphage (18, 519 46, 47). The hierarquical clustering analysis (Fig 4) revealed that higher levels of 520 crAssphage grouped most of the UP2 and UP3 samples, which at the same time 521 presented the highest levels of six out of the seven quantified genes. Also, we 522 showed that samples with and without fecal pollution presented statistically 523 significant differences in the abundance of intl1, sul1, sul2, bla OXA-2 and tetA (Fig 6).

524
All these genes presented a positive strong correlation among them, but no 525 individual correlation between the abundance of crAssphage and the other genes 526 were found (Fig 5). These findings could be explained by the low number of