Horizontal transfer of antibiotic resistance genes into microflora and blood cells in rats fed on GM-diet

Abstract

Background

With regard to the extensive production of genetically modified organisms, scientists focused on the safety of antibiotic resistance marker (ARM) genes present in GM-crops produced for the consumption of human or animal. We investigated the possible transfer of these genes to the microflora present in the gastrointestinal tract and blood cells of two groups of Male Wistar Albino rats fed on a transgenic diet containing the ARM genes nptII and aadA for 90 days. The conventional PCR was employed to screen for the presence of nptII and aadA genes in the experimental diets, and animals’ samples.

Results

The occurrence of DNA transfer of nptII and aadA genes from GM-diet to blood and bacterial cells has been unambiguously demonstrated. Results were further confirmed by sequencing and blast analysis, indicating that both transferred segments shared significant alignment with number of Binary and Cloning Vectors with variable similarities. The frequency of transfer of DNA segments from the GM-diet into blood DNA was unexpectedly higher than its transfer to bacterial DNA which may be attributed to a number of factors.

Conclusions

Our results unambiguously demonstrated the occurrence of DNA transfer of ARM genes (nptII and aadA) from GM plant diet to blood cells and enteric microflora in rats. Results may draw attentions to the importance of exploring the possible effects of transfer of ARM genes horizontally from GM products to consumers and to extend our attentions to the importance of a better understanding of the factors influencing HGT in the intestine of the GM-food consumers.

Background

Genetically modifying food technologies, like all new technologies, may have some known and unknown risks. Globally, great concern regarding genetically modified organisms (GMOs) is focusing on the safety of antibiotic resistance marker (ARM) genes in transgenic crops intended for human and animal consumption. Antibiotic resistance genes are known to produce enzymes to degrade antibiotics. Transfer of these genes to human or animal pathogens could make them resistant to available antibiotics (Bakshi 2003), which may, consequently, cause deleterious effects to public health.

The majority of antibiotic resistance genes used in biotechnology were isolated originally from bacteria (Miki and McHugh 2004). The neomycin phosphotransferase gene (nptII) conferring resistance to kanamycin and ampicilin is the most common selectable marker gene used in the transformation of various organisms as bacteria, yeast, plants and animals (Conner 1997). The source of nptII gene is the transposon (Tn5) from E. coli K12. The antibiotic hygromycin B is also used for selection as kanamycin (Miki and McHugh 2004). Resistance to hygromycin is confirmed mainly by aminoglycoside phosphotransferase (aphIV) which is also derived from E. coli (Day 2003; Miki and McHugh 2004). As for plastid transformation, the chimeric bacterial-derived antibiotic resistance marker, aminoglycoside 3″ adenyltransferase (aadA) gene, is the most widely used selectable marker. This gene (aadA) confers resistance to the antibiotics spectinomycin and streptomycin (Svab and Maliga 1993).

One of the hazards associated with the production of genetic modification is the probable spread of a DNA fragment from a donor cell/organism into unrelated recipient cells of another species through a process called horizontal gene transfer (HGT) (Keese 2008; Nielsen and Daffonchio 2007).

Several reports showed that transgenic or endogenous plant DNA could become incorporated into various tissues of animals (Oraby et al. 2014; Rizzi et al. 2012), leucocytes of rainbow trout (Chainark et al. 2008), goats (Tudisco et al. 2010), pigs (Mazza et al. 2005; Sharma et al. 2006) and human blood stream (Spisak et al. 2013), Mohr and Tebbe (2007) reported the transfer of segments of herbicide tolerance gene (pat-1) to gut bacteria in bees that pollinated GM Brassica. Another study (Netherwood et al. 2004) revealed evidence of low frequency transfer of a small fragment (180 bp) of EPSPS introduced gene derived from GM soybean to microorganisms within the small intestine of human ileostomists.

HGT of recombinant DNA from GMOs to bacteria is one of the unsettled biosafety issues (Nielsen et al. 2005). Scientists are concerned about transfer of such recombinant DNA which may promote the spread of antibiotic resistance (Beever and Kemp 2000; Chassy 2008; Ramessar et al. 2007). A study conducted by Woegerbauer et al. (2015) discovered that the nptII-load of the studied soils was low, rendering nptII a typical candidate as environmental pollutant upon anthropogenic release into these ecosystems.

Other studies (Wilcks and Jacobsen 2010) have suspected that DNA of genetically modified (GM) plants can be taken up by microbiota in the gut of animals fed on GM-plant diet. Chassy (2010) have also demonstrated that antibiotic resistance genes present in some transgenic crops have not added to the spread of antibiotic resistance in the environment.

The present work was planned to investigate the possible incorporation of transgenic DNA into enteric micro-flora of a group of experimental rats fed on GM-diet containing the antibiotic resistance marker (ARM) genes nptII and aadA for 90 days, using the polymerase chain reaction (PCR) assay. Transfer of ARM genes has also been investigated in blood cells of another group of experimental rats fed on the same GM-diet for the same duration (90 days). Samples were collected at three intervals (30, 60, and 90 days). The prevalence of the transferred ARM genes in blood cells of each of the experimental rats has been pursued all through the feeding duration. Further to confirm the transfer of these genes to the enteric microflora or blood cells, products of the PCR amplification of these genes were sequenced and analyzed through alignment with the GenBank Data Base.

Methods

Screening experimental diet samples for the presence of genetic modification

Animal feed samples were obtained from different animal feed suppliers in Cairo. None of these animal diets were labeled as genetically modified. Homogenous samples from each experimental diet were prepared by milling approximately 100–250 g of starting material. DNA was extracted from all samples applying a modified CTAB–based method (Aboul-Maaty and Oraby 2019).The presence of genetically modified contents in the purchased animal feed samples were investigated applying the conventional PCR assay, using primers for the Cauliflower Mosaic Virus -35S promoter (CaMVP-35S) and antibiotic resistance marker genes nptII and aadA (Table 1).

Table 1 List of primers used throughout the experimental duration, their sequences, amplicon lengths and annealing temperatures

The internal nucleotide sequencing of the PCR amplification products of primers for P-35S and nptII and aadA genes has been performed by MWG-Biotech AG, to confirm the presence of CaMV 35S promoter and ARM genes (nptII and aadA) in the chosen animal feed sample.

Blast analysis with the GenBank was further performed for sequencing data to test for alignment.

The animal diet, which was experimentally proven to contain the targeted GM ingredients (P35S, nptII and aadA), was chosen for feeding the experimental rats during this investigation.

The potential for horizontal gene flow of ARM genes from GM-diet into enteric microflora and blood cells of experimental rats.

Animals, housing, and feeding durations

For bacterial analysis Male Wistar Albino rats (30 rats) obtained from the animal house of the National Research Centre shortly before weaning, were divided into three groups ten animals each. Group 1 was immediately euthanized (shortly before weaning) and considered as a control group (G1). The other two groups of animals were housed in standard cages and under standard ambient conditions for 30 days (G2) or 90 days (G3).The two experimental animal groups (G2 and G3) were euthanized at the end of the two feeding durations (after 30 and 90 days, respectively). The protocol applied during this study fulfills all the requirements of the NRC Ethical committee’s guidelines (reference 12,142) as well as the ARRIVE guidelines, where all animals received humane care.

In order to study the potential for horizontal gene flow of antibiotic resistance marker genes from GM-diet into blood cells of investigated rats, an additional group of nine Male Wistar Albino rats were also purchased and were coded by the letters A, B, C, D, E, F, G, H, and K. A special approach was adopted for this part of our investigation, during which each animal is used as its own self-control (Table 2). Blood samples (1 ml each) were collected on EDTA (0.5%) from the coded experimental rats at four intervals. The first sample was collected shortly before weaning (i.e. 0-day sample) and before feeding them with the chosen GM-diet. Each of these samples (0-day samples) is used as a self-control for the three successive collected samples (after 30, 60, and 90 days) from the same corresponding rat (Table 2) during the whole feeding duration (90 days).

Table 2 Blood sampling codes matrix

Sampling processing

Micro-flora in the rectum and caecal cavity of all animals groups were inoculated onto the surface of nutrient agar plates. These inoculated plates of all groups were incubated for 24–48 h at 37 °C, after which they were examined for colony characters, cellular morphology and the purity of the culture. Bacterial DNA extraction was performed by using GF-1 Bacterial DNA Extraction Kit (Vivantis) according to the manufacturer's procedure.

DNA was also extracted from blood samples collected from coded rats (Table 2) at the four intervals using DNA Purification Kit (Promiga Wizard Genomic DNA Purification Kit) following the manufacturer's manual. Determination of the quality and concentration of all DNA samples was performed using the Nano Drop 1000/Thermo Scientific spectrophotometer.

The transfer of transgenic DNA from the GM-diet into the extracted DNA samples was tested using the conventional polymerase chain reaction (PCR) assay.

At least duplicates of most of the Polymerase Chain Reactions (PCR) were conducted in TM Thermal cycler (MJ Research PTC-100 thermocycler). PCR conditions and profile were carried out as described by Oraby et al. (2014, 2021).

The presence of a fluorescent band of the expected level for the investigated segments (nptII 173 bp-target, and aadA 284 bp-target) in all PCR products was analyzed using SYNGENE Bio Imaging Gel Documentation System. Further to confirm the obtained results, total fragment DNA purification kit (MEGA quick-spin, iNtRON Biotechnology, Inc) was used for purification of some of these PCR products for the subsequent internal nucleotides sequencing analysis which was performed by MWG-Biotech AG. Following sequencing, BLASTN analysis with the GenBank was performed to test for alignment.

Results

Screening animal feed samples for the presence of genetic modification

The CaMVP35S promoter investigated in the present work (P-35S) was identified in all collected animal feed samples (1, 2, and 3) screened for the presence of genetic modification (Fig. 1). The figure also includes PCR products of amplified primers for nptII and aadA in diet samples.

Fig. 1

Detected genetic modifications in the collected diet samples and their nucleotide sequence alignment with GenBank

The presence of the amplified primers in DNA of the chosen animal feed (Animal diet 1) was further confirmed by DNA sequencing. The internal nucleotide sequencing analysis results of the PCR amplicon obtained from the amplification of primers for CaMV promoter (P-35S) and the antibiotic resistance marker genes nptII and aadA are included in the figure (Fig. 1), as well as results of Blast analysis with the GenBank. Alignment using NCBI-BLASTN program for the PCR amplicon of the investigated segments (P-35S, nptII and aadA) is also presented.

BLASTN analysis showed a 100% homology of the amplified segment with binary vector pGWB80 Acc. No. dbj|AB752377.1| at nucleotide (nt) coordinate 5559 to 5753. It also shared 100% sequence homology with CaMV-P35S promoter (emb|V00141.1|). The nptII retrieved segment (173 bp) shared 100% identity with Minitransposon mTn5-GNm, complete sequence (AF080390.1) at nt coordinate 2382–2554. The nptII retrieved segment also shared 100% identity with many other cloning, Binary and Gateway vectors. PCR amplification of the aadA primers gave a fragment of 284 bp in size. The aadA (284 bp) retrieved segment shared 100% identity with Binary vector pKT pKT-NB-H2BsfGFP, complete sequence (MH973511.1) at nt coordinate 1188–1471.

The potential for horizontal gene flow of ARM genes from GM-diet into enteric microflora and blood cells of experimental rats fed on GM-diet for three months

Bacterial DNA samples representing the three groups of animals were screened for the presence of CaMV 35S promoter, two segments of the antibiotic resistance marker gene nptII (101 bp-target and 173 bp-target) and a segment of the aadA gene(284 bp-target).

Results presented in Table 3 indicated that bacterial DNA extracted from GIT of the control group (G1) did not harbor any fragments from P-35S or from the antibiotic resistance marker genes (nptII or aadA). All tested primers showed no amplification in this group (G1). Table 3 also shows that tested primers for P-35S, nptII, and aadA, recorded variable amplification percentages in bacterial DNA extracted from GIT of the other two groups; G2 and G3 that were fed with GM-diet for 30 or 90 days, respectively. Primers for CaMV-35S promoter were amplified in about 89% and 78% of the bacterial samples in G2 and G3, respectively.

Table 3 Incidence of transfer of P35S, nptII and aadA into bacterial cells in GIT of rats fed on GM-diet

The primers pair for the smaller segment (101 bp) from nptII gene was amplified in 60% of each of the two investigated groups (G2 and G3), whereas primers pair for nptII gene (173 bp-target) was amplified in 40% of each of the two investigated groups (G2 and G3). Jointly, the two primers pairs for nptII gene (101 bp-target and 173 bp-target) recorded amplification percentage of 80% in each of the investigated groups G2 and G3.

Primers used for the amplification of segment from aadA gene (284 bp-target) were amplified in 90% and 100% of the bacterial cell in G2 and G3, respectively.

Sequencing and BLASTN analysis

It was not possible to obtain reliable sequencing results from the short segment (101 bp) of nptII gene. Therefore, in order to confirm the transfer of nptII fragment from the diet to microflora in GIT of rats fed on the GM diet for 90 days, the internal nucleotide sequencing of the amplified nptII longer segment (173 bp-target) was performed for two of these DNA samples (B14 and B15).

The retrieved sequences from both samples were 167 and 170 base pairs in size, respectively. Alignment of the two sequences (B14 and B15) revealed that similarity between the two sequences was only 96% (Fig. 2A).

Fig. 2

The internal nucleotide sequencing (for nptII and aadA amplimers), alignment and Blast analysis with gene bank. A Alignment of the internal nucleotide sequences of the amplified nptII in DNA of two of the microflora samples (B14 and B15) from GIT of rats fed on GM-diet for 90 days. B Variable similarities percentages of number of sequences that produced significant alignment with nptII in the two bacterial samples B14 and B15. C Percent Identity Matrix created by Clustal 2.1 for the internal nucleotide sequences of the PCR amplicons obtained from the amplification of primers for aadA gene in number of bacterial cells. OR is the original reference sequence of the corresponding segment of aadA (MH973511.1)

Blast analysis for sequencing results (Fig. 2B) showed that the distribution of the top 101 Blast Hits on 100 subject sequences were the same for both samples B14 and B15, yet with variable similarities percentages (97% and 100%, respectively).

The internal nucleotide sequencing of the amplified aadA amplimers (284 bp-target) was performed for a number of samples from both groups, samples B3, B4, B7, B8 from G2 and samples B12, B13, B15, B16, B18, B20 from G3 group. The retrieved sequences from the amplified segments of the aadA gene were 284 base pairs in size. Clustal Omega multiple sequence alignment analysis for these sequences along with the corresponding segment from the reference sequence (MH973511.1), showed that they were not completely similar.

The internal nucleotide sequences similarities among the bacterial DNA samples ranged from 74.91 to 100.00% as presented in Fig. 2C, whereas similarity between them and the original reference sequence (MH973511.1) ranged from 82.27 to 99.29%.

Screening DNA of blood samples of experimental animals for the presence of ARM genes

The prevalence of DNA segment from GM-diet containing fragments from nptII (173 bp) and aadA genes in blood cells of investigated rats is presented in Table 4. Results revealed that no amplification occurred in DNA extracted from blood samples of rats collected shortly before weaning (0-day samples). Results also indicated that nptII was detected in all DNA of blood samples collected from rats fed on GM-diet after 30, 60 and 90 days except sample D90 in which nptII and aadA were not detected. PCR amplification of nptII (173 bp) fragment in blood samples collected from rats after 90 days is presented in Fig. 3.

Table 4 Prevalence of transfer of nptII and aadA genes into blood cells of rats fed on GM-diet

Fig. 3

PCR amplification products of nptII (173 bp) in blood samples of rats fed on GM-diet for 90 days

The prevalence of DNA segments from GM-diet containing fragments from aadA(284 bp) gene in blood cells of investigated rats are also presented in Table 4. Results showed that aadA segment was not detected in DNA extracted from blood samples of rats collected shortly before weaning (0-day samples).

Results also showed that aadA investigated segment was eliminated from some of the blood DNA samples, after 90 days (A90, C90 and D90). These samples were harboring the aadA segment at earlier sampling times, (A30, C30, and D60, respectively).

Further to confirm the transfer of fragments from nptII and aadA genes from the GM-diet to the blood of rats, the internal nucleotide sequencing of the obtained nptII and aadA amplimers was performed. The internal nucleotide sequencing of the amplified nptII amplimers (173 bp-target) was performed for selected samples after 60 days (K60) and 90 days (C90, E90, F90 and G90). The retrieved sequences from these samples were 173 bp in size. Alignment of these five sequences using the Clustal Omega multiple sequence alignment program along with the corresponding segment from the reference sequence (AF080390.1) is presented in Fig. 4A.

Fig. 4

Analysis of the results of amplification of nptII in blood cells of rats fed on GM-diet for 90 days. A Clustal Omega multiple sequence alignment of the internal nucleotide sequences of a number of blood samples along with the corresponding segment from the reference sequence mTn5 (AF080390.1). B Percent Identity Matrix—created by Clustal2.1—for the internal nucleotide sequences of the PCR amplicon (C90, F90, K60, G90 and E90) obtained from the amplification of primers for NPTII gene. mTn5 is the original sequence of the corresponding segment of mTn5 (AF080390.1)

CLUSTAL O (1.2.4) multiple sequence alignment (Fig. 4B) showed that samples C90, G90, F90 and K60 shared 100% identities with Minitransposon mTn5-GNm, complete sequence (AF080390.1). In case of E90, the identity percent was only 97.11%.

NCBI Blast analysis for NPTII (173 bp-target) sequencing results were carried out using BLASTN 2.9.0+ (Zhang et al. 2000). Blast analysis for C90, F90, K60, G90 internal nucleotide sequences gave 100% alignment with many cloning vectors (e.g. MK453498.1,MK448012.1, MK176935.1), Gateway binary vectors (e.g.AP018981.1, AP018954.1, AP018948.1) and other vectors (e.g. MH325111.1, MH325106.1, MK044344.1). Blast analysis for E90 internal nucleotide sequences gave 97.11% alignment with the same binary and cloning vectors.

Amplification of the aadA primers in DNA of blood samples of the experimental animals produced fragment of the size 284 bp. The internal nucleotide sequencing of the amplified aadA amplimers (284 bp-target) was performed for selected samples from the third sampling time group (E90, F90 and H90) fed on the GM-diet for 90 days.

Alignment of these sequences (E90, F90 and H90) using the Clustal Omega multiple sequence alignment program along with the corresponding segment from Binary vector pKT pKT-NB-H2BsfGFP and complete sequence (MH973511.1) are presented in Fig. 5A.

Fig. 5

Analysis of amplification results of aadA in blood cells of rats fed on GM-diet for 90 days. A Clustal Omega multiple sequence alignment of the internal nucleotide sequences of aadA (284 bp) in a number of blood samples along with the corresponding segment from the reference sequence (MH973511.1). B Percent Identity Matrix created by Clustal2.1 for the internal nucleotide sequences of the PCR amplicon (F90, E90, and H90) obtained from the amplification of primers for aadA gene. Original is the reference sequence of the corresponding segment of aadA (MH973511.1). C Variable similarities percentages of number of vectors that produced significant alignment with aadA amplemers

Results (Fig. 5B) showed that samples E90, F90 and H 90, respectively, shared 97.89%, 98.59 and 97.17% identities with the corresponding segment of aadA gene in the original Binary vector pKT pKT-NB-H2BsfGFP, complete sequence (MH9735111.1).

NCBI Blast analysis for aadA (284 bp-target) sequencing results was carried out using BLASTN 2.9.0+ (Zhang et al. 2000). Blast analysis for F90 and E90 and H90 internal nucleotide sequences resulted in alignment with number of Cloning Vectors, and Binary Vector with variable identity percentages as presented in Fig. 5C.

Discussion

Despite all of the benefits of GMOs, there is great concern among scientists about the possible occurrence of horizontal gene transfer of antibiotic resistance marker (ARM) genes, employed in the production of transgenic plants, into different tissues of animal or human consuming these GM products.

The present work explored the potential for horizontal gene flow of ARM genes from GM-diet into enteric microflora and blood cells of experimental rats fed on the laboratory chow diet for 90 days. As reported during the present study, the chosen laboratory diet sample used for this experiment gave positive results when screened for the presence of segments from CaMV-35S promoter, neomycin phosphotransferase II (nptII) gene and aminoglycoside 3″ adenyltransferase (aadA) gene. The presence of these segments was further confirmed by DNA sequencing and BLASTN analysis (Fig. 1). Antibiotic resistance marker genes are mostly introduced to GM plant as part of the bacterial vectors which are used for the initial gene constructions for transformation purposes. The nptII gene is usually used as a selectable marker in the initial laboratory stages to select plants that were genetically modified (Jelenic 2003; Turrini et al. 2015). The choice of the selective agent is important and based on the plant species to be transformed. However, nptII is probably the most widely used selectable marker for plant transformation, whereas aadA gene is usually used in the laboratory prior to the production of the genetically modified plants to select for bacteria containing the modified DNA.

In the present work, PCR results demonstrated the positive occurrence of DNA transfer of the antibiotic resistance marker genes nptII and aadA from GM diet to the enteric microflora (Table 3) and blood cells (Table 4) of experimental animals fed on the GM-diet for 90 days. These results were further confirmed by performing the internal nucleotide sequencing of the amplified nptII and aadA amplimers in number of DNA purified from the amplification products (amplicons). Sequencing was followed by alignment with the corresponding segments from the Gene Bank and further BLASTN analysis was performed to test for sequences similarities percentages (Figs. 2A–C). As presented in Fig. 2B, blast analysis with the gene bank for nptII sequences gave variable similarities percentages with many cloning and binary vectors. These results confirm the transfer of segments of nptII gene from the GM-diet into the bacterial cells of rats fed on this diet for 90 days.

Clustal Omega multiple sequence alignment among each of the incorporated genes (nptII or aadA) in bacterial cells samples showed that they were not completely similar. Alignment of nptII sequences showed that similarity was only 96% (Fig. 2A), where similarity among aadA sequences ranged from 74.91 to 100.00% (Fig. 2C). These similarity variations within bacterial DNA samples may point out to the possibility that the aadA chimeric gene is having high mutations rates, as in the case of bla recombinant gene which is characterized by having high mutations rates (Ho 2014).

The microbial system in the gastrointestinal tract (GIT) is very dynamic on the genetic level. It is capable of rapid response at the genetic level (Lerner et al. 2017). Its ecosystem is extremely enriched by mobile genetic elements (Feld et al. 2008; Bahl et al. 2004) that make it ideal for a potentially extensive gene exchange (Aminov 2011). Horizontal gene transfer (HGT) of antibiotic resistance genes is pervasive among prokaryotes, especially bacteria (Soucy et al. 2015).

The HGT of antibiotic resistance genes in the intestines of humans (Spisak et al. 2013) was reported. Its widespread in the human-associated microorganisms (Huddleston 2014; Liu et al. 2012; Smillie et al. 2011) was suggested to be due to the close physical proximity and increased cell-to-cell contact within the human body (e.g. gastrointestinal tract). Jeong et al. (2019) also suggested that ‘phylogenetic effect’ can significantly increase HGT activity among closely-related microorganisms. Transfer of these antibiotic resistance genes between these microorganisms, as suggested by Licht et al. (2003) may increase the possibility of acquisition of resistances by human pathogens through the use of antibiotics as additives in agricultural animal feed. Yau and Stewart (2013) revealed that the use of antibiotics may not only select for resistant populations but also may enhance the formation of new resistant strains by HGT.

On the contrary, transfer of transgenic DNA from the diet into the microflora found in GIT of rats fed on transgenic cucumber was not detected (Kosieradzka et al. 2001). However, as Nielsen and Townsend (2004) and Nielsen and BøhnT (2014) clarified that the different methodological approaches used have many limitations that may reduce the possibility to estimate the occurrence and impact of horizontal gene transfer in limited time study. Methodologically, they explained, the obtained bacteria from plate screening are only representing a tiny proportion of the bacterial populations of the tested habitats.

The prevalence of the transferred ARM genes in blood cells of each of the investigated rats has also been pursued all through the feeding duration (90 days). As mentioned above, a special approach was adopted in this investigation, during which each animal, of the second group of rats, was used as its own self-control (Table 2). Therefore, each blood sample collected shortly before weaning was considered as control sample (0-day samples) for the other three samples collected from the same animal. Screening results showed that nptII and aadA segments were not detected in DNA extracted from blood samples collected from the animals shortly before weaning (0-day samples). These finding revealed that the control samples (0-day samples) were free from these segments (nptII and aadA) before the beginning of the feeding duration.

Results also showed that nptII was detected in all DNA of blood samples collected from rats fed on GM-diet for 30, 60 and 90 days except sample D90 in which nptII seemed to be eliminated (Fig. 3).

The aadA investigated segment was not detected in some of the blood DNA samples, after 90 days (A90, C90 and D90). In view of the fact that these aadA segments were detected in the same samples at earlier sampling times (A30, C30, and D60, respectively), suggests that these missing transferred segments seemed to be eliminated too from these samples. These findings imply that incorporation of both nptII and aadA segments were not within the nucleus of the blood cells at the earlier sampling times A30, C30 and D60 (Table 4), which is supported by the fact that a foreign genetic material usually must be incorporated in the nucleus of the recipient cell in order to be permanently added to eukaryotic cells (Moses 1987). Our findings also suggest the possible occurrence of a repair mechanism which may result in the elimination of the transferred segment at that latter stage of sampling time.

More than 80% of genetically modified plants (Cankar et al. 2008) contain the CaMVP-35S promoter, therefore, we investigated the transfer of a segment of P-35S (195 bp-target) into bacterial cells in rats fed for 30 and 90 days on the GM-diet. Table 3 demonstrates that the rate of transfer of P-35S (195 bp-target) was 88.8% and 77.7% in bacterial samples after 30 and 90 days, respectively. We also explored the transfer of another segment (Cf3Cr4, 123 bp-target) from CaMV-P35S promoter into blood cells of rats fed on the GM-diet for 90 days. Results (Fig. 6) indicated a 100% presence of this fragment in all DNA samples collected from rats fed on GM-diet for 90 days. In a previous study, Oraby et al. (2014) reported that ingested fragments from the CaMV-35S promoter (195 bp-target) were incorporated into blood, liver and brain tissues of experimental rats and the affinity of incorporation of different transgenic fragments from the ingested GM-diet into the different tissues of rats varied from one target sequence to the other. In some cases, only short DNA fragments derived from diet containing GM-maize were detected in blood lymphocytes of cows (Einspanier et al. 2001). Variation in rate of incorporation of the two investigated segments (101 bp-target and 173 bp-target) from nptII gene into bacterial cells were also reported in the present investigation.

Fig. 6

PCR amplification of Cf3Cr4 in some blood samples of rats fed on GM-diet for 90 days. Cf3Cr4 (123 bp-target) is another segment from CaMV promoter, (con) is a negative control (0-day sample)

Spisak et al. (2013) revealed that large meal-derived DNA fragments carrying a complete gene can avoid degradation and enter the human circulation system. Other studies reported the presence of DNA from M13 virus, GFP and even ribulose-1, 5- bisphosphate carboxylase (Rubisco) genes in the blood and tissue of experimental animals (Guertler et al. 2009; Brigulla and Wackernagel 2010).

Jelenic (2003) suggested that the presence of fragments from the diet-DNA harboring segments of CaMV promoter (P-35S, and Cf3Cr4) may cause a risk to the consumer. Studies conducted by Oraby et al. (2015a) suggested health hazards accompanying the ingestion of GM-diets containing CaMV promoter. They reported deleterious histopathological and histochemical impacts as well as biochemical alterations in different tissues of rats in association with the GM-diet consumption. Genotoxicity of the GM diet was previously reported in somatic and germ cells (Oraby et al. 2015b).

It has also been reported that CaMV-P35S promoter can function in a wide range of organisms (plant and animal). Expression of the bacterial nptII gene in Escherichia coli (Assaad and Signer 1999) can also be initiated by CaMV-P35S promoter. Despite the fact that, the bacterial promoter controlling the expression of aadA gene is not active in GM-plants (Moses 1987), yet CaMV-35S promoter sequence can alter the expression pattern not only in adjacent genes (Yoo et al. 2005), but also it can convert other specific gene promoter in the adjacent tissues and organs into a globally active promoter (Zheng et al. 2007).

It is well known that bacteria (Prokaryotes) lack an organized nucleus and usually accept new DNA more easily than other tissues (Moses 1987), in the present study, the collective rate of transfer (%) for both genes (nptII and aadA) were unexpectedly, higher in blood cells than that in the bacterial cells (Table 5). This may be explained knowing that horizontal gene transfer which occurs naturally in bacteria (OECD 2010), directs the bacteria to impose DNA repair mechanisms in order to adapt to changing environmental conditions and to generate genetic diversity without losing too much genomic stability (Fall et al. 2007). Transformation frequencies varied in association with the recipient strain and the position on the bacterial genome (Fall et al. 2007). Jelenic (2003) assumed the occurrence of a more or less significant transfer of foreign food DNA into some types of consumer's cells depending on the animal species and the type of food. He concluded that in both mammals and birds, transfer of food-derived DNA fragments into the resident microflora, in the GIT, has been indicated to occur in the presence of sites for homologous recombination, or when the exogenous fragments are part of replicating plasmids. This was the case in our results, where sequencing and BLAST analysis for the amplified nptII (173 bp-target) gene in DNA of blood or bacterial samples showed 97% alignment with some synthetic construct clone Sp. (e.g. MK371206.1 and MK371204.1). The other amplified aadA gene in DNA of bacterial samples showed alignment percents ranged from 88 to 96% with some plasmids from different bacterial strains complete sequences (e.g. CP031295.1, CP028419.1 and LN830952.1).

Table 5 Rates of transferred segments from the ingested GM-diet into blood or bacterial samples of rats after 90 days

These results in its turn, confirm that these amplified segments were transferred into blood or bacterial cells horizontally from the GM-diet. Genetic engineering technology is known to depend on designing vectors for cloning and transferring genes involving artificially recombining and manipulating genes from unrelated species and their viral pathogens, which is known to improve and increase the probability for horizontal gene transfer and recombination (Ho et al. 1998).

Conclusions

The occurrence of DNA transfer of antibiotic resistance marker genes nptII and aadA from GM plant material to blood cells and enteric microflora has been unambiguously demonstrated in rats fed on GM-diet. These results may draw attentions to the importance of exploring the possible effects of transfer of ARM genes horizontally from GM products to consumers and to extend our attentions to the importance of a better understanding of the factors influencing HGT in the intestine of the GM-food consumers.