Abstract

The use of transgenic crops has become increasingly common in the United States over the last several decades. Increasing evidence suggests that DNA may be protected from enzymatic digestion and acid hydrolysis in the digestive tract, suggesting that crop-derived transgenes may enter into wastewater treatment plants (WWTPs) intact. Given the historical use of antibiotic resistance genes as selection markers in transgenic crop development, it is important to consider the fate of these transgenes. Herein we detected and quantified crop-derived transgenes in WWTPs. All viable US WWTP samples were found to contain multiple gene targets (p35, nos, bla and nptII) at significantly higher levels than control samples. Control wastewater samples obtained from France, where transgenic crops are not cultivated, contained significantly fewer copies of the nptII gene than US activated and digester sludges. No significant differences were measured for the bla antibiotic resistance gene (ARG). In addition, a nested PCR (polymerase chain reaction) assay was developed that targeted the bla ARG located in regions flanked by the p35 promoter and nos terminator. Overall this work suggests that transgenic crops may have provided an environmental source of nptII; however, follow-up studies are needed to ascertain the viability of these genes as they exit WWTPs.

INTRODUCTION

Transgenic crops have become increasingly popular relative to their conventional counterparts since their initial development in the 1980s (Clive 2012). As of 2012, 10 crops had been approved for cultivation and human consumption in the United States, including soybean, maize, cotton, canola, potatoes, alfalfa, papaya, sugar beets, squash and tomatoes (König et al. 2004). Many individual lines of each of these crops contain one or more genetically engineered traits (e.g., herbicide tolerance, pest resistance, drought tolerance, and virus resistance) conferred through genes of interest (GOI), amounting to hundreds of lines released for cultivation in the United States alone. The most popular engineered traits are pest resistance (Bt lines), herbicide tolerance (HT lines), or stacked pest and herbicide resistance. To date, more than 130 lines of transgenic crops have also been documented to contain one or more antibiotic resistance genes (ARGs), including β-lactamase (bla) and neomycin (nptII) resistance genes (ISAAA 2015). ARGs were first used as selection markers during the plant cell transformation process, and were commonly incorporated into first- and second-generation transgenic crops beginning in 1990 (Horsch et al. 1985; Moss et al. 2006). While some ARGs such as nptII continue to be used as selection markers in genetic engineering, other ARGs have fallen out of favor in recent years due to the controversies surrounding their potential to contribute to the rise in global environmental antibiotic resistance (Conner et al. 2003; Bennett et al. 2004; Dunfield & Germida 2004; König et al. 2004). Instead, herbicide resistance genes such as epsps and bar have become popular alternative selection markers (D'Halluin et al. 1992).

In addition to being sold in supermarkets as the crop product itself, transgenic crops may be further processed into other forms of food. For transgenic maize, these products include grits, cornmeal, flour, cereals, and some fermented products. Genetically modified crops are also further broken down to obtain derivative products that are incorporated into other processed foods. Some of these products include corn starches (e.g., maltodextrin, dextrose, high fructose corn syrup, and sugar alcohols), soy lecithin, sugar (from sugar beets), and vegetable oils (Padgette et al. 1995). These byproducts can then be incorporated into other foodstuffs, including baked goods, breads, baby foods, and other snack foods. Approximately 75% of processed foods sold in US supermarkets contain one or more transgenic crop-derived ingredients leading to frequent consumption of transgenic crops and crop byproducts in the United States (Millstone & Lang 2008; Kamle & Ali 2013). Once ingested, DNA from foods can conceivably behave in three ways: (1) DNA can be degraded by acids or DNase I and II enzymes found within the digestive system, (2) DNA can be taken up by cells in the gut (host or bacterial), or (3) DNA can pass through the digestive system and be excreted in a broken-down form or wholly intact (Heritage 2004). Because transgenic crops contain genetic cassettes containing foreign DNA (e.g., pest resistance, herbicide tolerance and/or ARGs), controversy exists regarding the ability of transgenes to be transferred to surrounding gut microflora within the host (Uzogara 2000). All transgenic crop lines intended for human or animal consumption are subject to rigorous review for approval – including tests for allergenicity, toxicity, and other acute adverse health outcomes. However, these assessments are often only performed on animal models and employ classic toxicological endpoints (e.g., no observed adverse effect level (NOAEL) or lowest observed adverse effect level (LOAEL)) (Food & Drug Administration 1992; Paarlberg 2002), which should detect most significant acute health outcomes but may not observe subtle health effects or those associated with chronic ingestion of transgenic crops (König et al. 2004). It has been widely established that free DNA in any environment is highly susceptible to degradation by DNase I and II enzymes – this susceptibility would only be heightened in gastrointestinal systems with the added presence of digestive acids (Mazza et al. 2005). However, some free DNA contained within food matrices is protected from both enzymatic digestion and acid hydrolysis and can pass intact through a host's digestive tract (Hohlweg & Doerfler 2001). As a result, it is possible that transgenes derived from transgenic crops may be introduced into municipal wastewater treatment plant (WWTP) systems.

With approximately 16,000 in operation in the United States, municipal WWTPs are responsible for treating the wastewater produced by more than 75% of US households. The treatment schematics of individual plants may vary, but generally consist of primary treatment, secondary treatment, and tertiary treatment (Metcalf & Eddy 1972). As both secondary treatment (e.g., activated sludge) and additional biosolids production (e.g., aerobic or anaerobic digestion) rely on mixed bacterial communities to remove unwanted organic compounds, any partially or wholly intact free DNA contained within human waste may be exposed to a range of microbes within those treatment processes (Kinney et al. 2006; Wu et al. 2010). Additionally, more than 50% of biosolids produced by WWTPs in the USA are used for downstream agriculture or land applications (Hale et al. 2001). This suggests that the crop transgenes entering into a WWTP may ultimately also be exposed to diverse soil microbial communities. While studies have been performed to estimate the fate of transgenic crop-derived ARGs in soil environments as well as the related risks to soil microbes (Nielsen et al. 1997; Gebhard & Smalla 1999; Gardner & Gunsch 2017), there remains a significant gap in understanding how WWTP bacteria are exposed to these transgenic constructs.

Given the high rates of transgenic crop cultivation and consumption in the USA as well as the previously noted possibility of crop transgenes being able to survive the digestive process, it is possible that partial or full-length transgenes derived from transgenic crops may be present in WWTPs. To investigate this question, in the present study, activated and digester sludge samples from WWTPs in the USA and France were screened for the presence of ARGs located between promoter and terminator gene sequences commonly used in transgenic crop development. Specifically, we targeted genetic sequences that contained the p35 promoter sequence on the 5′ end and the nos terminator gene on the 3′ end and either the bla ARG or nptII ARG or both ARGs in between these two genetic elements. While not conclusively tying these ARGs to transgenic crops, the presence of ARGs in between the promoter and terminator sequences may be suggestive of transgenic crop origin as this expression cassette has been incorporated into a wide array of commercially available transgenic crop lines, including canola, wheat, tomato, maize, cotton, and soy (Hardegger et al. 1999). In addition, we quantified each transgene in the same samples.

METHODS

Wastewater sludge collection and DNA isolation

Activated sludge samples were obtained from 26 WWTPs located in North Carolina, Ohio, Oregon, Pennsylvania, Washington, West Virginia, Missouri, Kansas, and Iowa in the USA. Control samples were obtained from a WWTP in France where transgenic crop cultivation is banned. These WWTPs were selected based on geographic regions and the broad range of treatment capacities (0.3–155 million gallons per day (MGD)) (Table S1, Supplementary Information, available with the online version of this paper). Approximately 50 mL were collected in triplicate from each WWTP and samples were shipped overnight on dry ice and were stored at −80 °C upon receipt until DNA could be extracted. DNA was isolated from triplicate WWTP samples using the DNeasy PowerWater Kit (QIAGEN Carlsbad, CA, USA) according to kit manufacturer instructions. Eluted DNA was quantified using a QuBit 2.0 (ThermoFisher Scientific, Waltham, MA, USA) and stored at −20 °C for downstream analysis. Extracted DNA from three WWTPs appeared highly degraded and these samples were excluded from further analysis.

In addition, of the 26 WWTPs mentioned above, 10 facilities from North Carolina, Missouri, Kansas, Iowa, and Oregon also provided samples of digester sludge. Control digester samples were also obtained from France. For each WWTP, approximately 50 mL were collected in triplicate and samples were shipped overnight on dry ice and were stored at −80 °C upon receipt until DNA could be extracted. Aliquots of 5 mL were centrifuged in triplicate to obtain biosolids for DNA extraction using the DNeasy PowerLyzer PowerSoil kit (QIAGEN Carlsbad, CA) according to manufacturer instructions. Eluted DNA was quantified using a QuBit 2.0 (Life Technologies) and stored at −20 °C for downstream analysis.

End point PCR and qPCR assay conditions

First, extracted DNA was screened to detect the presence of p35, nos, bla, and nptII genes with end point polymerase chain reaction (PCR) using the primers shown in Table 1. All end point PCR reactions were performed using a Bio-Rad T100 Thermal Cycler (Hercules, CA, USA) under the following conditions: initial denaturation at 95 °C for 3 min; 35 cycles of 95 °C for 30 s, 60 °C for 60 s, and 72 °C for 30 s; and 72 °C for 5 min, before storing at 4 °C.

In addition, we used end point PCR to identify larger fragments originating from potential transformation plasmids containing ARGs used in transgenic crop development. As bla and nptII genes found in WWTPs originating from transgenic crops are likely to be located in gene sections flanked by p35 promoter and nos terminator genes, we first performed end point PCR with p35F/nosR primers to amplify that large DNA section and followed up with a nested PCR targeting the ARGs. The nested PCR primers used in this study are outlined in Table 1. Finally, extracted DNA was also screened for the presence of the cauliflower mosaic virus (CaMV) ORF (open reading frame) VII region to determine if the p35 genes detected in WWTPs were viral in origin. Negative and positive controls were incorporated into each end point PCR run.

Table 1

Primer and TaqMan™ probe sequences for qPCR assays

Target gene Primer sequence (5′-3′) TaqMan probe (5′-3′) Amplicon length (bp) Source 
p35 F: GCCTCTGCCGACAGTGGT
R: AAGACGTGGTTGGAACGTC TTC 
FAM-CAAAGATGGACCCCCAC CCACG-TAMRA 82 Waiblinger et al. (2008)  
nos F: GTCTTGCGATGATTATCAT ATAATTTCTG
R: CGCTATATTTTGTTTTCTATC GCGT 
FAM- AGATGGGTTTTTATGAT
TAGAGTCCCGCAA-TAMRA 
120 Mano et al. (2008)  
bla F: TGCGCGGAACCCCTATTTGT
R: GCGGCGACCGAG TTGCTCTT 
Not applicable 382 Badosa et al. (2004)  
nptII F: AGGAAGCGGTCAGCCCAT
R: GCGTTGGCTACCCGTGATAT 
FAM-CGCCGCCAAGCTCTTCA GCAA-TAMRA 60 Mason et al. (2002)  
16S F: TCCTACGGGAGGCAGCAG
R: GGACTACCAGGGTATCTAA TCCTGTT 
FAM- CGTATTACCGCGGCTGC TGGCAC-TAMRA 466 Nadkarni et al. (2002)  
CaMV ORF VII F: AGCGGTCAAAATATTGCTTA
R:AACTTACCGTATGCTAGATTACCT 
Not applicable 141 Love et al. (2005)  
Target gene Primer sequence (5′-3′) TaqMan probe (5′-3′) Amplicon length (bp) Source 
p35 F: GCCTCTGCCGACAGTGGT
R: AAGACGTGGTTGGAACGTC TTC 
FAM-CAAAGATGGACCCCCAC CCACG-TAMRA 82 Waiblinger et al. (2008)  
nos F: GTCTTGCGATGATTATCAT ATAATTTCTG
R: CGCTATATTTTGTTTTCTATC GCGT 
FAM- AGATGGGTTTTTATGAT
TAGAGTCCCGCAA-TAMRA 
120 Mano et al. (2008)  
bla F: TGCGCGGAACCCCTATTTGT
R: GCGGCGACCGAG TTGCTCTT 
Not applicable 382 Badosa et al. (2004)  
nptII F: AGGAAGCGGTCAGCCCAT
R: GCGTTGGCTACCCGTGATAT 
FAM-CGCCGCCAAGCTCTTCA GCAA-TAMRA 60 Mason et al. (2002)  
16S F: TCCTACGGGAGGCAGCAG
R: GGACTACCAGGGTATCTAA TCCTGTT 
FAM- CGTATTACCGCGGCTGC TGGCAC-TAMRA 466 Nadkarni et al. (2002)  
CaMV ORF VII F: AGCGGTCAAAATATTGCTTA
R:AACTTACCGTATGCTAGATTACCT 
Not applicable 141 Love et al. (2005)  

All samples yielding positive results by end point PCR were then further analyzed by quantitative real-time PCR (qPCR) using the Stratagene Mx3000P thermocycler (Agilent Technologies, Santa Clara, CA, USA). For each transgene assay, all triplicate WWTP samples were run in pseudo-triplicate (n = 9). qPCR amplification to detect the bla gene was performed using SYBR Green chemistry under the following conditions: initial denaturation at 95 °C for 3 min, followed by 40 cycles of 95 °C for 45 s, 55 °C for 60 s, and 72 °C for 30 s. Fluorescent measurements were taken during elongation. The dissociation curve was assessed at the end of the run under the following conditions: 95 °C for 60 s, 55 °C for 30 s, and 95 °C for 30 s. qPCR amplification using TaqMan™ Probe chemistry was performed for p35, nos, and nptII assays using the following conditions: initial denaturation at 95 °C for 60 s, followed by 40 cycles of 95 °C for 30 s and 60 °C for 30 s. The following conditions were used for 16S: initial denaturation at 95 °C for 10 min, followed by 40 cycles of 95 °C for 30 s and 60 °C for 60 s. Fluorescence measurements for each assay were taken during annealing steps for all qPCR assays.

Statistical analysis

The concentration of each transgene was determined by normalizing gene copy numbers to the volume of wastewater used to extract DNA. Gene copies for each target were also normalized to the gene copies of 16S (Supplementary Information, Figures S5–S8, available with the online version of this paper) to compare the relative abundance of each transgene to the total bacteria in each sample (Ma et al. 2011). Multiple linear regression analyses and correlation analyses were performed to investigate the relationships between gene concentration and WWTP characteristics. Averages and standard deviations of all data were calculated using Microsoft Excel 2007. All other statistical tests including t-tests and analyses of variance (ANOVA) were conducted in GraphPad PRISM (version 7). Strong statistical significance was defined as p < 0.05 and marginal statistical significance was defined as p < 0.10.

RESULTS AND DISCUSSION

End point PCR screenings

Activated sludge obtained from 26 WWTPs and digester sludge from 10 WWTPs were screened for the presence of crop transgenes using end point PCR. In the first step, we screened for individual transgenes (i.e., p35, nos, nptII, and bla) and all were detected during this preliminary screening step in both WWTP activated sludge and digester sludge (Tables S2 and S3, Supplementary Information, available with the online version of this paper). All but three samples contained at least one transgene, and 12 samples contained all four transgenes. In follow-up analyses, we found that the extracted DNA from the three WWTPs with no detected transgenes appeared highly degraded when run out on an agarose gel, both pre- and post-PCR. In addition, qualitative PCR screenings for 16S genes failed to amplify from these samples suggesting that the DNA within the sludge may have become damaged in transit and therefore we cannot conclusively state whether the transgenes were present or absent in these samples. These samples were excluded from further analysis.

The p35 promoter gene found in many transgenic crops is derived from the CaMV. To verify that the p35 promoter regions detected in this study originated from crop transgenes as opposed to CaMV, DNA extracted from WWTP activated sludge and digester sludge was also screened for CaMV ORF VII genes. This end point PCR screening was negative for all samples, suggesting that the detected p35 genes were not viral in origin. Rather, these data suggest that the p35 gene regions detected herein existed either as extracellular DNA or within the intracellular DNA of WWTP bacteria. However, the exact source was not identified herein.

Finally, we wanted to ascertain as to whether the fragments spanned by the p35 promoter region and nos terminator regions contained the bla ARG. This was accomplished by nested PCR using the forward p35 primer (p35F) and reverse nos primer (nosR) described in Table 1 (Supplementary Information, Figure S1, available online). PCR products obtained from the p35F/nosR screening step were then subjected to an additional PCR amplification using the bla primers detailed in Table 1. P35F/nosR primer screenings resulted in the amplification of multiple large gene fragments in all WWTP samples (Supplementary Information, Figure S2, available online). Some of these bands were found to contain the bla ARG within each WWTP sample. Others have shown that multiple copies of promoters, terminators, GOI, and selection markers are often incorporated into a single transformation plasmid (Taylor et al. 1999), and could therefore generate multiple bands using a single PCR primer set.

qPCR screenings

Samples that tested positive for potential transgenes during the end point PCR screening were then further analyzed using qPCR to determine the abundance of each transgene in each sample. Standard curves and efficiencies for each qPCR assay are shown in Supplementary Information (Table S4, Figures S3 and S4, available online). Copy numbers of the p35 promoter gene in US wastewater samples ranged from 2.92 × 102 to 3.61 × 104 copies per mL of wastewater, with an average of 1.12 × 104 copies/mL (Figure 1(a)). The abundance of p35 in activated sludge was not significantly different among wastewater treatment facilities sampled in North Carolina, Ohio, Pennsylvania, and West Virginia (WWTPs 1–14). The abundance of p35 detected in activated sludge from facilities sampled in Missouri, Iowa, and Kansas was more variable (WWTPs 15–21); p35 levels were significantly lower in WWTPs 15 and 18 (p < 0.05). Overall copy numbers among digester sludge were lower than those measured in activated sludge, but the same trends were observed. The abundance of p35 genes ranged from 2.53 × 102 to 2.28 × 103 copies per mL of sludge, with an average of 1.13 × 103 copies per mL. The levels of p35 were significantly higher in activated sludge compared to digester sludge obtained from US WWTPs (p < 0.05). In contrast, the concentration of p35 genes in control activated sludge obtained from France (WWTP 1) was significantly lower compared to that of US wastewaters (p < 0.05).

Figure 1

Gene copies of the (a) p35 promoter gene, (b) nos terminator gene, (c) nptII ARG, and (d) bla ARG in WWTP activated sludge and digester sludge. Bars representing activated sludge samples are denoted by black hatch patterns and digester sludge samples are represented by solid black. WWTPs have been reported anonymously, with control samples denoted by ‘C’. Error is reported as standard deviation (n = 9).

Figure 1

Gene copies of the (a) p35 promoter gene, (b) nos terminator gene, (c) nptII ARG, and (d) bla ARG in WWTP activated sludge and digester sludge. Bars representing activated sludge samples are denoted by black hatch patterns and digester sludge samples are represented by solid black. WWTPs have been reported anonymously, with control samples denoted by ‘C’. Error is reported as standard deviation (n = 9).

Similar trends were observed in the detected copy numbers of the nos terminator genes in US wastewater samples (Figure 1(b)). Nos terminator copy numbers ranged from 2.26 × 102 to 4.78 × 104 copies per mL of wastewater, with an average of 1.15 × 104 copies per mL. Similar to the p35 promoter genes, the abundance of nos in activated sludge was not significantly different among wastewater treatment facilities sampled in North Carolina, Ohio, Pennsylvania, and West Virginia (WWTPs 1–14). The abundance of nos detected in activated sludge from facilities sampled in Missouri, Iowa, and Kansas was more variable (WWTPs 15–21); nos gene concentrations were significantly lower in WWTPs 15 and 18 (p < 0.05). Detected nos values in WWTP 20 were only marginally significantly lower than in other WWTPs (p = 0.093). Lower nos copy numbers were observed in digester sludge samples than in activated sludge samples. In these samples, nos abundance ranged from 4.19 × 101 to 7.95 × 103 copies per mL of wastewater, with an average of 2.77 × 103 copies per mL. Nos gene copy numbers were highest in WWTPs 2, 15, and 20, but these levels were not significantly different from other digesters (p = 0.052). Nos copy numbers were also not significantly different between activated and digester sludge (p = 0.058). Again, the concentration of nos genes in control activated and digester sludge samples was significantly lower compared with those from US wastewaters (p < 0.05).

Copy numbers of nptII ARGs in US activated sludge wastewater varied widely among sampling locations (Figure 1(c)). NptII concentrations ranged from 2.46 × 102 to 1.58 × 105 copies per mL of wastewater, with an average of 2.80 × 104 copies per mL. Unlike p35 and nos genes, nptII ARG abundance in activated sludge did not appear to be linked to the location of WWTPs (p = 0.091). The highest concentrations of nptII among activated sludge samples were found in WWTPs 3, 9, 19, and 21, and were significantly different from other WWTPs (p < 0.05). Overall copy numbers among wastewater digesters were similar to those observed in activated sludge samples (p = 0.086). NptII abundance in US digesters ranged from 8.04 × 102 to 2.20 × 104 copies per mL of wastewater, with an average of 5.29 × 103 copies per mL. The concentration of nptII ARGs in control activated and digester sludge (WWTP 1) was 6.30 × 101 and 3.20 × 101 per mL, respectively, and again was found to be significantly lower compared to that of US wastewaters (p < 0.05).

The detected copy numbers of the bla ARGs in US wastewater samples were similar to those of nptII in that they varied greatly among WWTPs (Figure 1(d)). In activated sludge, bla ARGs copy numbers ranged from 7.18 to 1.31 × 105 copies per mL of wastewater, with an average of 5.25 × 104 copies per mL. The lowest concentrations of bla among activated sludge samples were found in WWTPs 15, 18, 20, 22, and 23 and were significantly different from other WWTPs (p < 0.05). Overall copy numbers among digester sludge samples were also highly variable. Among these samples bla ARG abundance ranged from 2.29 × 102 to 9.71 × 103 copies per mL of wastewater, with an average of 3.67 × 103 copies per mL. The concentration of bla genes in control activated sludge was not significantly different from that of US wastewaters (p = 0.24).

Correlation analyses

Overall, gene copy numbers of the p35 promoter region and nos terminator region as well as nptII and bla ARGs varied widely among the sampled WWTPs. P35 promoter and nos terminator genes were less abundant than nptII and bla ARGs on average. This is not surprising given that p35 promoter and nos terminator genes are not known to exist in natural environments at high concentrations and may only be found in WWTPs as a result of ingestion and excretion of transgenic foods (Badosa et al. 2004). By contrast, ARGs have been measured at high concentrations in environmental bacterial populations (Ma et al. 2011; Laht et al. 2014; Mao et al. 2015). Additional statistical analyses suggest that some transgenes were positively associated with WWTP location and size. The concentrations of all potential transgenes (p35, nos, nptII, and bla) among individual WWTPs were found to be correlated to WWTP state locations (p < 0.05), determined by an ANOVA and multiple linear regression. P35 promoter gene copy numbers were most strongly correlated with WWTP location (p < 0.05), while nptII and bla copy numbers were marginally correlated with WWTP location (p = 0.051 and p = 0.099, respectively). Nos terminator and 16S gene copy numbers were not significantly correlated with WWTP location (p = 0.38, p = 0.27 respectively). Associations with WWTP locations were linked only to individual states and not entire geographic regions. For example, transgene abundance was linked to the US states in which they were located, as noted above. However, the correlations between entire geographic regions were not significant (e.g., Midwestern regions vs. Northeastern regions). This appears counterintuitive, given the large quantities of transgenic crops cultivated in Midwestern states. One explanation is that the consumption of transgenic crops may be a more important factor in assessing the prevalence and quantity of transgenes in WWTP sludge.

Notably, the abundance of the p35, nos, and nptII genes was markedly lower in control sludge samples as compared to US sludge samples. The significant differences in p35, nos, and nptII concentrations in French WWTP systems relative to US WWTP systems are probably due to the fact that France does not cultivate transgenic crops and human consumption is limited to imported processed foods containing transgenic crops (Kuntz 2014). The lower quantities of these transgenes in control samples further supports the idea that transgene presence in WWTPs is linked to human consumption and excretion of transgenic crops or their byproducts. The levels of p35, nos, and nptII genes in these control wastewater samples may be a reflection of the small amounts of foods consumed that contain ingredients derived from transgenic crops. Although France does not allow for the cultivation or human consumption of transgenic crops, France still imports processed foods containing transgenic byproducts from other countries, which may explain the baseline levels measured (Gruere 2006). Finally, as the abundance of the bla ARG was not found to differ significantly between control and US wastewater samples, these data may suggest that the bla genes detected in US wastewater are not derived from transgenic crops, but may instead be a reflection of naturally occurring antibiotic resistance. In fact, bla genes have been widely detected in WWTP systems across multiple countries in both biofilms and suspended bacterial communities (Marti et al. 2013).

The statistical association between WWTP size, in terms of typical daily flow rates in MGD, and transgene prevalence was also investigated. Daily flow rates in sampled WWTPs varied greatly, from 1.8 to 155 MGD (Table S1, Supplementary Information). P35 gene copy numbers were strongly correlated with WWTP size (p < 0.05). Nos gene concentrations as well as bla and nptII ARGs were not significantly correlated with WWTP size (p = 0.14, 0.33 and 0.67, respectively). The prevalence of 16S genes was most strongly correlated with WWTP size (p < 0.05). This is to be expected, as influent rate and sludge retention time are known to be strongly correlated with the carrying capacities, and therefore bacterial abundance, of WWTP sludges (Wéry et al. 2008; Xia et al. 2012).

CONCLUSIONS

In conclusion, all four selected transgenes (i.e., p35, nos, bla, and nptII) were detected in all viable samples obtained from WWTPs in the United States. Our data suggest that some of the transgenes detected in WWTPs are most likely derived from digested transgenic crops or foods containing their byproducts within human waste. The p35 promoter gene was detected at high concentrations in all US wastewater samples. Although this gene is naturally found within the CaMV the authors were unable to detect additional CaMV genes (e.g., ORF VII) in any WWTP sample, suggesting the p35 genes were derived from transgenic crop material rather than CaMV. The nos terminator gene was also detected at high copy numbers within sampled WWTPs. While this nos gene is derived from bacteria, it is specific to Agrobacterium tumefaciens, which is commonly found in soils, but has not been noted in the literature to reside in WWTP systems. Finally, bla ARG primers used by this study are crossborder primers, with the forward primer targeting a region 109 bp upstream of the bla gene and the reverse primer targeting a region 273 bp into the bla gene on the transgenic transformation plasmids. Utilizing crossborder primers, which include additional elements of the host plasmid, lends supports to the idea that the detected bla construct is derived from the transformation plasmid contained from transgenic crops. This is also supported by the strong differences in gene abundance between US and French control WWTPs, where the rates of transgenic crop cultivation and human consumption are significantly lower (Kuntz 2014). It is important to note that while our study suggests this association, additional in-depth metagenomic analysis is needed to confirm this association. Furthermore, additional studies are needed to determine if the DNA fragments amplified in this study originated from broken-down gene templates or full-length templates which may be transferred via horizontal gene transfer (HGT). HGT has been documented to occur within WWTPs at elevated rates, possibly as a function of the high concentration and mobility of aquatic bacteria as well as the elevated availability of dissolved Ca2+ and Mg2+ ions (Moura et al. 2007; Zhang et al. 2009a, 2009b; Caucci & Berendonk 2014). The potential risk of ARG transfer to bacteria may be heightened by the presence of promoter genes and terminator genes incorporated into transformation plasmids (e.g., p35 and nos, respectively); however, the integrity and viability of these larger fragments should be further verified. Transgenes detected within WWTP anaerobic sludges are of particular interest as more than 50% of these sludges are further utilized as agricultural fertilizers (Holzem et al. 2016). Upon land application of biosolids fertilizers, these transgenes may be able to adsorb to the surfaces of soil and clay particles, which are able to shield this DNA from physical and enzymatic degradation in additional environments (Gardner & Gunsch 2017). Therefore, the present study highlights the importance of verifying the sources of potential transgenes, as documented herein, with future metagenomic studies of WWTPs and evaluating their ability to be taken up by environmental bacteria and persist across multiple environments.

ACKNOWLEDGEMENTS

This material is based upon work supported by the National Science Foundation under grant number CBET-1438612 and GRFP DGF-106401.

CONFLICT OF INTEREST STATEMENT

This material is based upon work supported by the National Science Foundation under grant number CBET-1438612 and GRFP DGF1106401. All decisions regarding study design, sample collection, and manuscript development were made solely by the authors of this work.

REFERENCES

REFERENCES
Caucci
,
S.
&
Berendonk
,
T. U.
2014
Environmental and public health implications of antibiotic-resistance genes in municipal wastewaters
.
Prävention und Gesundheitsförderung
9
(
3
),
175
179
.
Clive
,
J.
2012
Global status of commercialized biotech/GM crops: 2012
.
ISAAA Brief
44
,
1
2
.
Conner
,
A. J.
,
Glare
,
T. R.
&
Nap
,
J. P.
2003
The release of genetically modified crops into the environment
.
The Plant Journal
33
(
1
),
19
46
.
D'Halluin
,
K.
,
De Block
,
M.
,
Denecke
,
J.
,
Janssen
,
J.
,
Leemans
,
J.
,
Reynaerts
,
A.
&
Botterman
,
J.
1992
The bar gene has selectable and screenable marker in plant engineering
.
Methods in Enzymology
216
,
415
426
.
Dunfield
,
K. E.
&
Germida
,
J. J.
2004
Impact of genetically modified crops on soil- and plant-associated microbial communities
.
Journal of Environmental Quality
33
(
3
),
806
815
.
Food and Drug Administration
1992
Statement of policy: foods derived from new plant varieties
.
Federal Register
57
(
104
),
22984
.
Hale
,
R. C.
,
La Guardia
,
M. J.
,
Harvey
,
E. P.
,
Gaylor
,
M. O.
,
Mainor
,
T. M.
&
Duff
,
W. H.
2001
Flame retardants: persistent pollutants in land-applied sludges
.
Nature
412
(
6843
),
140
141
.
Hardegger
,
M.
,
Brodmann
,
P.
&
Herrmann
,
A.
1999
Quantitative detection of the 35S promoter and the NOS terminator using quantitative competitive PCR
.
European Food Research and Technology
209
(
2
),
83
87
.
Heritage
,
J.
2004
The fate of transgenes in the human gut
.
Nature Biotechnology
22
(
2
),
170
172
.
Holzem
,
R.
,
Gardner
,
C.
,
Stapleton
,
H.
&
Gunsch
,
C.
2016
Impact of emerging contaminants and biosolids aging on ecotoxicity in soils
.
Proceedings of the Water Environment Federation
2016
(
3
),
628
631
.
Horsch
,
R.
,
Fry
,
J.
,
Hoffman
,
N.
,
Eichholtz
,
D.
,
Rogers
,
S. A.
&
Fraley
,
R.
1985
A simple and general method for transferring genes into plants
.
Science
227
,
1229
1232
.
ISAAA
2015
GM Approval Database
.
International Service for the Acquisition of Agri-biotech Applications (ISAAA)
.
Kinney
,
C. A.
,
Furlong
,
E. T.
,
Zaugg
,
S. D.
,
Burkhardt
,
M. R.
,
Werner
,
S. L.
,
Cahill
,
J. D.
&
Jorgensen
,
G. R.
2006
Survey of organic wastewater contaminants in biosolids destined for land application
.
Environmental Science & Technology
40
(
23
),
7207
7215
.
König
,
A.
,
Cockburn
,
A.
,
Crevel
,
R.
,
Debruyne
,
E.
,
Grafstroem
,
R.
,
Hammerling
,
U.
,
Kimber
,
I.
,
Knudsen
,
I.
,
Kuiper
,
H. A.
&
Peijnenburg
,
A.
2004
Assessment of the safety of foods derived from genetically modified (GM) crops
.
Food and Chemical Toxicology
42
(
7
),
1047
1088
.
Ma
,
Y.
,
Wilson
,
C. A.
,
Novak
,
J. T.
,
Riffat
,
R.
,
Aynur
,
S.
,
Murthy
,
S.
&
Pruden
,
A.
2011
Effect of various sludge digestion conditions on sulfonamide, macrolide, and tetracycline resistance genes and class I integrons
.
Environmental Science & Technology
45
(
18
),
7855
7861
.
Mano
,
J.
,
Shigemitsu
,
N.
,
Futo
,
S.
,
Akiyama
,
H.
,
Teshima
,
R.
,
Hino
,
A.
,
Furui
,
S.
&
Kitta
,
K.
2008
Real-time PCR array as a universal platform for the detection of genetically modified crops and its application in identifying unapproved genetically modified crops in Japan
.
Journal of Agricultural and Food Chemistry
57
(
1
),
26
37
.
Mao
,
D.
,
Yu
,
S.
,
Rysz
,
M.
,
Luo
,
Y.
,
Yang
,
F.
,
Li
,
F.
,
Hou
,
J.
,
Mu
,
Q.
&
Alvarez
,
P.
2015
Prevalence and proliferation of antibiotic resistance genes in two municipal wastewater treatment plants
.
Water Research
85
,
458
466
.
Mason
,
G.
,
Provero
,
P.
,
Vaira
,
A. M.
&
Accotto
,
G. P.
2002
Estimating the number of integrations in transformed plants by quantitative real-time PCR
.
BMC Biotechnology
2
(
1
),
1
.
Mazza
,
R.
,
Soave
,
M.
,
Morlacchini
,
M.
,
Piva
,
G.
&
Marocco
,
A.
2005
Assessing the transfer of genetically modified DNA from feed to animal tissues
.
Transgenic Research
14
(
5
),
775
784
.
Metcalf and Eddy
1972
Wastewater Engineering: Collection, Treatment, Disposal
.
McGraw-Hill
,
New York
.
Millstone
,
E.
&
Lang
,
T.
2008
The Atlas of Food: Who Eats What, Where and Why
.
Earthscan
,
London
.
Moss
,
C. B.
,
Schmitz
,
A.
&
Schmitz
,
T. G.
2006
First-generation genetically modified organisms in agriculture
.
Journal of Public Affairs
6
(
1
),
46
57
.
Moura
,
A.
,
Henriques
,
I.
,
Ribeiro
,
R.
&
Correia
,
A.
2007
Prevalence and characterization of integrons from bacteria isolated from a slaughterhouse wastewater treatment plant
.
Journal of Antimicrobial Chemotherapy
60
(
6
),
1243
1250
.
Nadkarni
,
M. A.
,
Martin
,
F. E.
,
Jacques
,
N. A.
&
Hunter
,
N.
2002
Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set
.
Microbiology
148
(
1
),
257
266
.
Nielsen
,
K.
,
Gebhard
,
F.
,
Smalla
,
K.
,
Bones
,
A.
&
Van Elsas
,
J.
1997
Evaluation of possible horizontal gene transfer from transgenic plants to the soil bacterium Acinetobacter calcoaceticus BD413
.
Theoretical and Applied Genetics
95
(
5–6
),
815
821
.
Padgette
,
S. R.
,
Kolacz
,
K.
,
Delannay
,
X.
,
Re
,
D.
,
LaVallee
,
B.
,
Tinius
,
C.
,
Rhodes
,
W.
,
Otero
,
Y.
,
Barry
,
G.
&
Eichholtz
,
D.
1995
Development, identification, and characterization of a glyphosate-tolerant soybean line
.
Crop Science
35
(
5
),
1451
1461
.
Taylor
,
N. B.
,
Fuchs
,
R. L.
,
MacDonald
,
J.
,
Shariff
,
A. R.
&
Padgette
,
S. R.
1999
Compositional analysis of glyphosate-tolerant soybeans treated with glyphosate
.
Journal of Agricultural and Food Chemistry
47
(
10
),
4469
4473
.
Waiblinger
,
H.-U.
,
Ernst
,
B.
,
Anderson
,
A.
&
Pietsch
,
K.
2008
Validation and collaborative study of a p35s and T-nos duplex real-time PCR screening method to detect genetically modified organisms in food products
.
European Food Research and Technology
226
(
5
),
1221
1228
.
Wu
,
C.
,
Spongberg
,
A. L.
,
Witter
,
J. D.
,
Fang
,
M.
&
Czajkowski
,
K. P.
2010
Uptake of pharmaceutical and personal care products by soybean plants from soils applied with biosolids and irrigated with contaminated water
.
Environmental Science & Technology
44
(
16
),
6157
6161
.
Xia
,
S.
,
Jia
,
R.
,
Feng
,
F.
,
Xie
,
K.
,
Li
,
H.
,
Jing
,
D.
&
Xu
,
X.
2012
Effect of solids retention time on antibiotics removal performance and microbial communities in an A/O-MBR process
.
Bioresource Technology
106
,
36
43
.
Zhang
,
X.-X.
,
Zhang
,
T.
&
Fang
,
H. H.
2009a
Antibiotic resistance genes in water environment
.
Applied Microbiology and Biotechnology
82
(
3
),
397
414
.
Zhang
,
Y.
,
Marrs
,
C. F.
,
Simon
,
C.
&
Xi
,
C.
2009b
Wastewater treatment contributes to selective increase of antibiotic resistance among Acinetobacter spp
.
Science of the Total Environment
407
(
12
),
3702
3706
.

Supplementary data