Repeated Ivermectin Treatment Induces Ivermectin Resistance in Strongyloides ratti by Upregulating the Expression of ATP-Binding Cassette Transporter Genes

ABSTRACT.

Ivermectin (IVM) is a widely used anthelmintic. However, with widespread use comes the risk of the emergence of IVM resistance, particularly in strongyloidiasis. Adenosine triphosphate (ATP)-binding cassette (ABC) transporter genes play an important role in the IVM-resistance mechanism. Here, we aimed to establish an animal experimental model of IVM resistance by frequent treatment of Strongyloides ratti with subtherapeutic doses of IVM, resistance being evaluated by the expression levels of ABC transporter genes. Rats infected with S. ratti were placed in experimental groups as follows: 1) untreated control (control); 2) treated with the mutagen ethyl methanesulfonate (EMS); 3) injected with 100 µg/kg body weight of IVM (IVM); 4) treated with a combination of EMS and IVM (IVM+EMS). Parasites were evaluated after four generations. Extent of IVM resistance was assessed using IVM sensitivity, larval development, and expression of ABC genes. By the F4 generation, S. ratti in the IVM group exhibited significantly higher levels of IVM resistance than did other groups according to in vitro drug-sensitivity tests and inhibition of larval development (IC₅₀ = 36.60 ng/mL; 95% CI: 31.6, 42.01). Expression levels of ABC isoform genes (ABCA, ABCF, and ABCG) were statistically significantly higher in the IVM-resistant line compared with the susceptible line. In conclusion, IVM subtherapeutic doses induced IVM resistance in S. ratti by the F4 generation with corresponding upregulation of some ABC isoform genes. The study provides a model for inducing and assessing drug resistance in Strongyloides.

INTRODUCTION

Ivermectin (IVM), a member of the macrocyclic lactone drug class, is a drug of choice for the treatment of many parasitic diseases including strongyloidiasis, onchocerciasis, and cutaneous larva migrans.¹ The targets of IVM are glutamate-gated chloride-channel receptors (GluClRs), gamma aminobutyric acid (GABA), and glycine (Gly) receptors that are expressed in the neurons and muscle cells, leading to chloride influx–induced hyperpolarization and paralysis of the worm.² Although IVM remains an effective drug, a small proportion of patients did not respond after the first dose of IVM treatment of some parasites such as Strongyloides stercoralis,³ Onchocerca volvulus,⁴ and Haemonchus contortus.⁵ Furthermore, recent study has indicated that a single-dose IVM treatment (200 µg/kg) in some S. stercoralis–infected participants did not result in complete cure.⁶ Moreover, treatment failure was found after two doses of IVM (0.2 mg/kg) given 2 weeks apart for S. stercoralis–infected Australian Aboriginals with type 2 diabetes mellitus.⁷ Therefore, continuous surveillance for IVM resistance in parasites, especially S. stercoralis, is a necessary precaution.

The mechanisms of drug resistance can be classified as specific (e.g., altered drug receptors) and nonspecific (altered drug metabolism or increased drug efflux).⁸ ATP-binding cassette (ABC) transporters are involved in active efflux of anthelmintic drugs and are important nonspecific mediators of IVM resistance.⁸ ABC transporters are a family of transmembrane proteins important for transportation of various substrates across the cell membrane in an energy-dependent manner.⁹ There are seven distinct subfamilies of ABC transporters (ABCA to ABCG). P-glycoprotein (Pgp), encoded by the ABCB1 gene, is a member of this family and is reportedly involved in anthelmintic resistance.¹⁰ High expression levels of Pgp are associated with IVM resistance in some parasites such as Haemonchus contortus,¹¹ Onchocerca volvulus,⁴ and Teladorsagia circumcincta.¹² Several lines of evidence suggest that IVM is one of the activators of Pgp expression.¹³ Therefore, repeated mass administration with a drug might contribute to the emergence of drug resistance in the target parasite,¹⁴ probably via induction of ABC gene expression.¹⁵ However, this hypothesis has not yet been tested, especially in S. stercoralis.

In the present study, we tested whether treatment with frequent, subtherapeutic doses of IVM could induce IVM resistance in vivo using Strongyloides ratti as a model.¹⁶ The expression of ABC transporter genes was also investigated by quantitative real-time polymerase chain reaction (RT-qPCR). The study may provide a means of assessing development of IVM resistance in Strongyloides species, using ABC transporter genes as a potential marker of this.

MATERIALS AND METHODS

Animal setting.

A culture of S. ratti was kindly provided by Prof. M Itoh, Parasitology Division, Department of Infection and Immunity, Aichi Medical University, Nagakute, Aichi-ken, Japan. The parasite life cycle was maintained in rats as described elsewhere.^16,17 In brief, rats were initially infected by a subcutaneous injection of 500 iL3s: the infections became patent from 4 to 5 days later. Agar was prepared as previously described.¹⁸ Then, 2–4 g of fresh rat feces was collected and cultured on agar plates to use as a source of individuals of the free-living generation.

Establishment of S. ratti IVM-resistant line.

The experimental protocol in rats was approved by the Animal Ethics Committee of Khon Kaen University, based on the Ethics of Animal Experimentation of the National Research Council of Thailand (ACUC-KKU-52/2559). Thirty-two male rats, 4–6 weeks of age, were obtained from the Animal Unit, Faculty of Medicine, Khon Kaen University. The protocol for the establishment of an IVM-resistant S. ratti line is shown in Figure 1. In brief, rats were allocated into four groups (two rats/group/generation): 1) S. ratti–infected without any treatment (Control); 2) S. ratti–infected and treated with 10 mM of ethyl methanesulfonate (EMS, Sigma-Aldrich, Darmstadt, Germany), a known mutagen; 3) S. ratti–infected and treated with 100 µg/kg body weight of IVM (IVM, Sigma-Aldrich, Darmstadt, Germany); 4) S. ratti–infected and treated with a combination of 100 µg/kg body weight of IVM and 10 mM EMS (IVM+EMS).

Figure 1.

Experimental model using rats for establishment of resistance to ivermectin in Strongyloides ratti. Control) S. ratti–infected without any treatment; EMS) S. ratti–infected and treated with 10 mM of ethyl methanesulfonate (EMS), a known mutagen; IVM) S. ratti–infected and injected with 100 µg/kg body weight of ivermectin (IVM); IVM+EMS) S. ratti–infected and treated with a combination of 10 mM EMS and 100 µg/kg body weight of IVM. The experiment was performed in duplicate in each experimental group (N = 32: two rats/group/generation). IVM = ivermectin; mFPC = modified filter paper culture technique. ^a S. ratti iL3 were incubated with 10 mM EMS for 2 hours before injection into a new rat. ^b S. ratti iL3 were cultured in IVM for 48 hours and then incubated with 10 mM EMS for 2 hours before injection into a new rat.

In the control group, rats were infected with 3,000–5,000 S. ratti iL3 larvae via subcutaneous injection. One week later, fecal samples were cultured to confirm S. ratti infection using a slightly modified Harada–Mori method (modified filter-paper culture [mFPC] technique), and then iL3 larvae were collected and counted for injection into a new rat. In the EMS-treated group (without IVM), S. ratti iL3 were incubated with 10 mM EMS in normal saline for 2 hours before subcutaneous injection into a rat to produce the next generation of worms as previously reported.¹⁹ In the IVM-treated group, 1 week after infection with iL3 larvae, fecal samples were collected and cultured to confirm S. ratti infection. Immediately after this, 100 µg/kg body weight of IVM was given via subcutaneous injection. After a further week, the fecal samples were collected and cultured with 12.5 ng/mL of IVM in water for 48 hours using the mFPC. Then, all iL3 worms were collected, counted, and injected into a new rat to produce the next generation. In the IVM+EMS group, iL3 larvae were cultured with 12.5 ng/mL of IVM in water for 48 hours using the mFPC and then immediately incubated in normal saline containing 10 mM EMS at room temperature for 2 hours before injection into a new rat. This continued until the fourth generation of worms. The iL3 larvae from the fourth generation were collected and used to evaluate IVM sensitivity, larval development, and expression of ABC genes.

IVM-sensitivity assay.

Development of IVM resistance was evaluated by using an in vitro drug-sensitivity assay focusing on the motility of third-stage larvae in the presence of different concentrations of IVM. In brief, approximately 500 iL3 were placed into 1.5mL tubes containing different concentrations of IVM (ranging from 10 ng/mL to 100 ng/mL) in 1% dimethyl sulfoxide (DMSO) following a previous report.¹⁹ After incubation at room temperature in the dark for 48 hours, the proportion of live (motile) worms was recorded under a stereomicroscope and IC₅₀ was calculated.

Larval development inhibition assay.

To investigate the survival of iL3 or development to adults, modified agar plate culture (mAPC) was performed. Feces from infected rats in each generation were placed on nutrient agar prespread with 12.5 ng/mL of IVM. Plates were incubated at 19–22°C in the dark. After 5 days, worms were examined and counted under a stereomicroscope. Fecal samples cultivated without IVM were used as controls. All experiments were performed in duplicate.

RNA extraction and RT-qPCR analysis.

The filariform larvae (iL3) from the fourth generation that survived the larval development inhibition assay were collected for profiling of ABC transporter gene expression by RT-qPCR assay. Approximately 1,500 iL3 worms were washed several times with sterile phosphate-buffered saline (PBS) and total RNA was extracted using RNeasy mini kit (Qiagen^®, Hilden, Germany) following the manufacturer’s protocol. Briefly, the worms were homogenized in RNeasy lysis buffer for lysing cells and tissues (buffer RLT) using a bead-based homogenizer (Power-lyzer^® 24, Mo-Bio Laboratories, Carlsbad, CA) followed by washing, binding and elution of total RNA. The extracted RNA was treated with Turbo-DNase (Ambion^®, Huntington, UK) to remove genomic DNA. Approximately 3 µg of total RNA were reverse transcribed to complementary DNA (cDNA) using SuperScript^™ III first-strand synthesis kit (Invitrogen^®, Carlsbad, CA) according to the manufacturer’s instructions. cDNA synthesis was performed using oligo-dT primers for three distinct replicates of worms.

The primers targeting ABCA, ABCB, ABCD, ABCF, and ABCG were designed based on sequence information for chromosome 1 of S. ratti lodged in the National Center for Biotechnology Information (NCBI) nucleotide database (accession number LN609528) using the online Primer3 software (http://frodo.wi.mitedu/primer3/). Primer sequences are shown in Table 1. LightCycler^® 480 SYBR Green I master mix (Roche Applied Science, Mannheim, Germany) was used for the RT-qPCR assay according to manufacturer’s protocol. The reaction was carried out using LightCycler 480 RT-PCR System (Roche Applied Science). The following cycling conditions were used: incubation at 95°C for 5 minutes followed by 40 cycles of 95°C for 10 seconds, 60°C for 15 seconds, and 72°C for 30 seconds. Data were analyzed using Light Cycler 480 software release 1.5.0 (Roche Applied Science) with a cycle threshold (CT) in the linear range of amplification. The expression level of each gene was measured relative to the expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and then the average value of the control set to 1 using the $2^{-\mathrm{\Delta }\mathrm{\Delta }{C}_T}\mathrm{ }$ method.²⁰

Table 1.

The qPCR primers used for detection of ABC transporter genes in Strongyloides ratti

Genes	Primer Sequences (Sense and Antisense)	Accession Numbers	Amplicon Size (bp)	Tm
GAPDH	5′-CAGTAACGTGTATGTCATTAACAAC-3′	LN609401	153	60.0
	5′-CAGCTGATACAGCAATTGAACTAA-3′		153	60.2
ABCA	5′-GATTATAATGTATATCCAGAAACACG-3′	LN609528	194	59.8
	5′-ATTAGATAGTCTTGAAAGTAATGCTG-3′		194	59.7
ABCB	5′-TCCTTGATTTATTAAATTACACCCA-3′	LN609528	163	59.9
	5′-GAATTAGCTGCTTATAAACAAAGAA-3′		163	60.2
ABCC	5′-GCAACATGCCTTCTTGATGA-3′	LN609528	188	59.8
	5′-TTTGCTCCAACATGAGCATC-3′		188	59.8
ABCD	5′-TGGTGGTTTACTTTCATATGGTATTC-3′	LN609528	183	59.5
	5′-AAACCAGCCATCTCTCCTGA-3′		183	59.8
ABCF	5′-ATGACACATTAGTTATTGATGCTG-3′	LN609528	154	60.3
	5′-ATTGCAACAAGACGTTCAC		154	60.1
ABCG	5′-TGTGATGAACCAACAACAGGA-3′	LN609528	191	59.9
	5′-GATGGTGGTCCTTGAAAAGC-3′		191	59.5

qPCR = quantitative polymerase chain reaction.

Statistical analysis.

Nonlinear regression was used to analyze IC₅₀. The proportions of live (motile) worms and IC₅₀ values from in vitro drug-sensitivity tests and the larval-development assay from the four experimental groups were compared using analysis of variance (ANOVA) as implemented in STATA package version 10.1 (StataCorp LLC, College Station, TX). One-way analysis of variance was used for the evaluation of ABC gene expression using GraphPad Prism version 7.00 for Windows, GraphPad Software (San Diego, CA). P values less than 0.05 were regarded as statistically significant.

RESULTS

Establishment of an S. ratti IVM-resistant line.

In the animal model, IC₅₀ values of S. ratti F4 isolates after subtherapeutic doses of IVM, EMS, or a combination of IVM+EMS are demonstrated in Figure 2. Treatment with subtherapeutic doses of IVM (0.1 µg/mL) in each generation produced the highest IC₅₀ values for iL3s obtained in the study. After four generations, IC₅₀ values for iL3s were 36.60 ng/mL (95% CI: 31.6, 42.01), which was significantly higher than in the untreated group (13.28 ng/mL; 95% CI: 11.52, 14.94, P < 0.001), EMS (18.04 ng/mL; 95% CI: 13.74, 22.31, P < 0.001), and IVM+EMS (12.16 ng/mL; 95% CI: 10.28, 13.87, P < 0.001) groups. In contrast, values for IC₅₀ among untreated, EMS and IVM+EMS groups were not statistically significantly different (P > 0.05). Unexpectedly, the combination of EMS and IVM yielded the lowest IC₅₀ value of any group.

Figure 2.

Effects of ivermectin (IVM) (concentrations ranging from 0.5 to 100 ng/mL) and ethyl methanesulfonate (EMS) on the viability of the third-stage larvae of the F4 generation of S. ratti using a drug-sensitivity assay. Control = Strongyloides ratti control group; EMS = EMS-treated group; IVM = IVM-treated group; IVM+EMS = group treated with a combination of IVM and EMS.

Larval development test of S. ratti IVM-resistant line.

The fecal samples of rats infected with each generation of worms were cultured and examined using mAPC and mFPC techniques with the inclusion of 12.5 ng/mL of IVM in the solution or agar media. The results demonstrated that S. ratti died or did not undergo further development in IVM+EMS, EMS, and control groups after exposure to 12.5 ng/mL of IVM. The iL3 worms of all four generations (F1–F4) survived after exposure to 10–25 ng/mL of either IVM or EMS for 48 hours. In the contrary in each generation (F1–F4) of the IVM group, worms were capable of development, indicating IVM resistance.

Increased expression of some ABC transporter genes in the S. ratti IVM-resistant line.

After IVM and EMS treatment of four generations, the results from qPCR showed that the expression of ABCA, ABCF, and ABCG genes was significantly greater (P < 0.05) in the S. ratti IVM-resistant (IVM) group relative to all other groups. On the other hand, the expression levels of ABCB and ABCD genes did not differ significantly different among all experimental groups (P > 0.05) (Figure 3).

Figure 3.

mRNA expression of ABC subfamily genes of the F4 generation of Strongyloides ratti for each experimental group (Control, EMS, IVM, and IVM+EMS). The iL3 worms were washed with PBS and used for investigation of mRNA expression of ABCA to ABCG genes by real-time polymerase chain reaction (PCR) as described in Materials and Methods. Experimental animal groups and abbreviations are as in the legend to Figure 1. Triplicate experiment was performed. *, **, and *** indicate P < 0.05, P < 0.01, and P < 0.001, respectively. n.s. = not significant.

DISCUSSION

In this study, we used the S. ratti-rat model to establish IVM-resistant strains of S. ratti in vivo. Resistance was associated with upregulation of ABC transporters. We induced resistance by repeated treatment of S. ratti with subtherapeutic doses of IVM and studied expression of genes involved in IVM-resistance mechanisms. The F4 generation exhibited the highest drug tolerance after exposure to IVM alone (IC₅₀ 36.60 ng IVM/mL). This contrasted with the EMS group (IC₅₀ 18.04 ng IVM/mL), EMS+IVM group (IC₅₀ 12.16 ng IVM/mL), and the control group exposed to normal saline (IC₅₀ 13.28 ng IVM/mL).

In a previous study, Viney et al (2002) used a different protocol to establish IVM resistance in S. ratti. They exposed L4 larvae to 75–95 mM EMS and cultured them in agar plates to obtain iL3 larvae of the F1 generation. These iL3 were injected into another rat, fecal samples from which were cultured by APC containing 10 ng/mL IVM. Subsequently, worms of the F2 generation could develop to iL3s in the presence of 10 ng/mL IVM.¹⁹ In addition to S. ratti, in vitro experiments revealed that Caenorhabditis elegans can grow and develop in nematode growth media containing IVM at concentrations up to 10 ng/mL. IVM-resistant strains of C. elegans produced in this way have been used for evaluating drug resistance in vitro.¹⁵ In a larger host, the dog, infected by Dirofilaria immitis, a monthly IVM challenge dose of 24 µg/kg led to the development of drug-resistant worms.²¹ This was a higher degree of IVM resistance than in our S. ratti–infected rat model. Thus, it is likely that the difference of method and of host–parasite system may have affected the degree of IVM resistance that could be induced.

As mentioned earlier, previous workers have found that either IVM or EMS treatment can induce IVM resistance.^15,19 In this study, we initially hypothesized that the combination of IVM and EMS would induce the highest levels of resistance. Unexpectedly, we found that a combination of EMS and IVM was less effective at inducing IVM resistance than was either of these agents alone. It is possible that EMS may interact with IVM so as to hinder the development of IVM resistance.²²

IVM resistance may be due to an increased expression of ABC transport proteins.¹⁵ Our approach to establishing S. ratti IVM-resistant lines has provided a model for investigating transcription levels of each ABC transporter gene. We found that levels of ABCA, ABCF, and ABCG genes were significantly higher in the IVM F4 isolates than in the other groups. Exposure of S. ratti more frequently to subtherapeutic doses of IVM may cause ABC genes involved in xenobiotic metabolism and transport to increase constitutively. Upregulation of subfamilies ABCA, ABCF, and ABCG in the IVM-resistant F4 larvae implies that efflux pumps contribute to the observed drug tolerance. Likewise, in vitro exposure of larvae of resistant isolates of Haemonchus contortus to IVM significantly increased the expression of ABC transport proteins and the subfamilies pgp-1, pgp-9.1 and pgp-9.2 compared with a susceptible isolate.²³ Also, Pgp homologues belonging to the ABC transporter family could enhance IVM efficacy in sensitive isolates, and also restore IVM sensitivity to resistant strains, suggesting that ABC transporters play an important role in resistance to IVM of gastro-intestinal nematodes.²⁴ In addition, in the C. elegans model, resistance to low levels of IVM (≤ 6 ng/mL) was associated with increased expression of multidrug-resistance protein-1, and P-glycoprotein-1 (pgp-1) and decreased glutathione, while higher-level resistance (10 ng/mL) was primarily associated with the increased expression of Pgps.¹⁵ Overall, the expression levels of ABC subfamily genes significantly increased after the pressure of repeated selection with either IVM or moxidectin, reflecting that ABC transport genes might play a general role in anthelmintic resistance by preventing drugs from reaching their target.

The ABC transporter family of nematodes is similar to that of other eukaryotes, with some variation.²⁵ In our model, ABCB was expressed at lower levels than ABCA, ABCF, and ABCG genes in IVM-resistant isolates of S. ratti, but its expression level was increased in adult C. elegans during experimental development of resistance against IVM.²⁶ The different expression patterns of the ABC transporter family between S. ratti and C. elegans suggests that expression of ABC isoforms varies according to the stage and species of parasite.

In our study, new mutations have likely not arisen in S. ratti in the nonmutagenized IVM treatment: mutations are therefore likely not the cause of increased resistance. Therefore, if the resistance has a genetic basis, preexisting alleles must have been selected. Alternatively, previous study has suggested that the mechanism of drug resistance might involve genetic mutation and/or DNA damage.²⁷ New mutations might be expected in the treatments including a mutagen, although, attempts at mutagenesis might not be effective in iL3s. At the iL3 stage the germ line consists of few cells and represents a small target. In previously reported mutagenesis experiments on S. ratti,^19,28 adults were mutagenized and their progeny from the indirect cycle used for infection. This approach would have generated a chance for mutations to become homozygous. Moreover, mutagenesis might make selection less efficient. The stress imposed by DNA damage and the activation of the repair systems might partially erase epigenetic marks and therefore make epigenetic inheritance less efficient.²⁹ However, we did not assess whether DNA methylation, histone modification, or micro-RNA increase occurred, suggesting avenues for future research.

Strongyloides ratti is closely related to S. stercoralis,³⁰ suggesting that establishment of IVM-resistant S. ratti in a rat model could be further applied to establishing S. stercoralis–resistant lines. However, S. ratti can infect and maintain a complete life cycle in mouse or rat, whereas S. stercoralis cannot. Other susceptible hosts for S. stercoralis infection should be used such as dogs, cats, and primates that are possible reservoir hosts for transmission of human strongyloidiasis.³¹ Moreover, our establishment of S. ratti IVM-resistant lines might be useful for evaluating the efficacy of drugs proposed for treatment of S. stercoralis infection by determining the expression of ABC isoform genes. Although mass administration of IVM could diminish the prevalence of S. stercoralis infection,³² this approach cannot prevent the reinfection and frequent treatment might lead to drug resistance. Multiple doses of IVM did not increase efficacy and were tolerated less than a single dose; thus, a single dose should be preferred for the treatment of nondisseminated strongyloidiasis.³³

CONCLUSIONS

In conclusion, we successfully established a strain of S. ratti resistant to IVM in an animal model and investigated ABC genes involved in the IVM-resistance mechanism. Repeated exposure of S. ratti to subtherapeutic doses of IVM could induce a greater level of resistance than could a mutagenic chemical (EMS) or a combination of IVM and EMS. Resistance in S. ratti in the F4 generation was associated with upregulation of mRNA expression of ABCA, ABCF, and ABCG. Expression levels of these genes might be indicators for IVM resistance. Our study provides background information concerning drug resistance development in Strongyloides, which may be valuable for optimizing efficacy of the current anthelmintic or of new drugs, developing markers to detect anthelmintic resistance and as a tool in designing new therapeutic agents.

Note: Supplemental materials appear at www.ajtmh.org.