Differences in protein expression associated with ivermectin resistance in Caenorhabditis elegans

The indiscriminate administration of synthetic anthelmintics such as ivermectin contributes to the selection of subpopulations capable of resisting the drugs’ effects. To understand the mechanisms of ivermectin resistance in Caenorhabditis elegans, this study attempted to identify molecular targets. C. elegans lineages that were sensitive and resistant to ivermectin were used. Collected nematodes were added to an extraction buffer and macerated in liquid nitrogen for protein extraction. The extracted proteins were separated according to molecular weight by SDS-PAGE to verify their integrity. Subsequently, proteins from both lineages were separated using two-dimensional electrophoresis. The gels were analyzed and the relevant spots were excised and identified by mass spectrometry (NanoESI-Q-TOF and MASCOT) and subsequently assessed by GO enrichment and STRING analyses. The increased expression of proteins associated with high metabolic activity, such as ATP-2 and ENOL-1, which are responsible for ATP synthesis, was observed. Furthermore, proteins with involvement in mediating muscular function (MLC-1, ACT-1, and PDI-2), signaling (FAR-1 and FAR-2), and embryo development (VHA-2) were identified. Protein interaction analysis indicated that the majority of the identified proteins in the resistant lineages participated in the same reaction triggered by ivermectin.


Introduction
Anthelmintic resistance is a problem that plagues livestock production worldwide, causing a billion dollars in losses annually (CHARLIER et al., 2009).This problem is caused by the parasites' ability to survive the drugs used to treat helminthiasis (GOOLSBY et al., 2017).
There are several groups of drugs that are differentiated by their chemical formulas and mechanisms of action to control helminth parasites (KAPLAN & VIDYASHANKAR, 2012).Among the most heavily studied of these drugs, ivermectin is a semisynthetic compound derived from an actinomycete with a broad spectrum of action (CAMPBELL, 2012).Although ivermectin follows the parameters recommended for good anthelmintics, resistance to this drug was reported in the first years after its launch ( VAN WYK & MALAN, 1988).There is no explanation to date for ivermectin resistance, only hypotheses that consider changes in the chlorine receptor channels and p-glycoproteins, which are membrane proteins responsible for the cellular efflux of drugs (LESPINE et al., 2012).
Several protein changes occur in response to stress caused by various factors, such as temperature, food, radiation exposure, and chemicals (CEDERGREEN et al., 2016).Additionally, these changes can even be caused by genetic selection of the subpopulation to resist a stress environment promoted by drug compounds.It is possible to identify these proteins in proteomic studies, such as two-dimensional electrophoresis (2DE) associated with mass spectrometry, which emphasizes differences in expressiveness and concentrations (HART et al., 2015).
Although 2DE associated with mass spectrometry is a safe tool for identification, it is not possible to identify proteins when there is no information available in databases (TABUSE et al., 2005).Therefore, studies of model organisms are preferred because they are associated with large datasets, and findings obtained in these organisms can be extrapolated to other organisms that have been less heavily researched (ANTOSHECHKIN & STERNBERG, 2007).
For this study, we used the model Caenorhabditis elegans, a nematode that can be handled efficiently under laboratory conditions and compared with parasitic nematodes.Among this model's notable qualities is the ability to be manipulated genetically to develop parasite characteristics, such as ivermectin resistance (BRENNER, 1974;AUDHYA & DESAI, 2008).The aim of this study was to compare the main protein differences between lineages of C. elegans that are sensitive and resistant to ivermectin.The results may help to explain the mechanism of resistance to ivermectin in nematodes.

Strains of Caenorhabditis elegans
The wild-type lineage of C. elegans Bristol N2 was provided by Dr. Carlos Winter (Department of Parasitology/ICB/USP).The lineage was maintained on nutrient growth medium (NGM) agar in Petri dishes at 20 °C (BRENNER, 1974).The lineage resistant to 15 ng/mL ivermectin (IVR 15 ) was developed stepwise in plates containing NGM with the addition of doses (1 ng/ml) of ivermectin, according to James & Davey (2009).

Protein extraction
The lineages, containing the all larval stages and eggs, were collected in M9 buffer (Na 2 HPO 4 .7H 2 O, KH 2 PO4, NaCl, MgSO 4 .7H2O,ddH 2 O) and centrifuged at 1000 x g for one minute to remove the buffer.For each lineage, we added extraction buffer (1% CHAPS, 100mM DTT, 7 M urea, 2 M thiourea) in a 1:1 proportion to the nematode pool, followed by immersion in liquid nitrogen, maceration, and centrifugation at 13000 × g at 4 °C for 5 minutes; then, the supernatant was subjected to electrophoresis (SILVA et al., 2014).
The extraction quality was observed using polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS-PAGE).The samples were loaded, and the gels were stained with Coomassie brilliant blue R-250 solution (SAMBROOK & DAVID, 2006).

Two-dimensional electrophoresis and image analysis
The 2DE gels were made induplicate for each lineage using 50 μg of the protein pool.The protein samples were pooled with 250 µl of rehydration buffer (7 M urea, 2 M thiourea, 1% CHAPS, 1% DTT, 0.5% ampholytes, pH 3-10, trace of bromophenol blue) and subjected to 2DE electrophoresis.
The samples were subjected to isoelectric focusing (IEF) using Ettan IPGphor III (GE Healthcare).The linear IPG strips (pH 3-10, 13 cm) were equilibrated with solution I (6 M urea, 50 mM Tris, 30% glycerol, 2% SDS, 1% DTT and a trace of bromophenol blue) for 15 minutes with shaking.This step was repeated for equilibration solution II (6 M urea, 50 mM Tris, 30% glycerol, 2% SDS, 1% DTT, trace of bromophenol blue and 3% iodoacetamide).The strips were embedded on the top of 12% polyacrylamide gels and separated using the Hoefer SE 600 Ruby system (GE Healthcare).The gels were stained with Coomassie brilliant blue G-250 (CANDIANO et al., 2004).The gels were scanned (LabScan, GE Healthcare), and the images were saved in the TIFF format for spot detection, quantification, matching and differential analysis with the ImageMaster 2D Platinum program (GE Healthcare).Matched spots were compared for volume, with p-values less than 0.05 being considered significant, and histograms were generated by classifying the spots according to their quantitative expressiveness.

Two-dimensional gel excision and mass spectrometry
The differentially expressed and exclusively expressed spots were selected and excised manually, followed by digestion with trypsin according to Shevchenko et al. (2006).The tryptic peptide samples were analyzed by NanoESI-Q-TOF coupled with nanoUPLC.The data obtained were submitted to the MASCOT sequence query server (MÁDI et al., 2008) for identification against the NCBIprot database and the SwissProt database.The results were v. 28, n. 1, jan.-mar. 2019 107/112 107 evaluated by observing the best probability score and the number of peptides matched for each spot identified.

a. Gene ontology
Using the BLAST  option Blastp, the proteins identified in both lineages were compared with sequences in the NCBI database (MADDEN, 2013).The accession numbers were submitted to the tool Retrieve/ID Mapping of the UniProt database.The data were converted into Gene Ontology (GO) terms and grouped into categories: Molecular Function, Biological Process and Cellular Component.

b. Protein interactions
The sequences of the proteins exclusively expressed in lineage IVR 15 , as identified by MS, were acquired in BLAST and submitted to the protein-protein interaction program STRING, version 10.0, according to the standard protocol (SZKLARCZYK et al., 2017).The interactions were predicted per co-occurrence and co-expression and were determined experimentally.

SDS-PAGE and two-dimensional electrophoresis
The efficiency of the extraction method and the quality of the extracted proteins were confirmed by SDS-PAGE, enabling the analysis of protein expression changes between the lineages to be performed in two-dimensional gels (Figure 1).The replicate2DEgels were compared, and their matching generated a linear correlation coefficient of 0.853 for N2 and 0.806 for IVR 15 (Figure 2) .
The proteome profiles demonstrated that there were differences in the protein expression as a function of anthelmintic resistance (Figure 3).Although the number of spots remained relatively similar, the protein expression was 7% higher in the resistant lineage than in the sensitive lineage, which was beyond the changes observed in the protein distribution by pI and molecular weight.

Mass Spectrometry (MS)
Twenty-three spots of each lineage were chosen for analysis by MS.Among those spots, 10 were considered more expressed, and 13 were exclusively expressed in both lineages.The identified proteins of the N2 lineage and those of the IVR 15 lineage are listed in Table 1.After identification, two proteins found in the two lineages were similar: disulfide isomerase 2 (PDI-2) and ATP synthase subunit beta (ATP-2).Proteins that were uncharacterized but deposited in the UniProt data base were also found in each lineage.The intermediate filament protein (IFB-2) was identified in two spots of lineage N2, with each having a high score.Fatty acid and retinol-binding protein 1 (FAR-1) and fatty acid and retinol-binding protein 2 (FAR-2), belonging to the same family and possessing similar features, were found in IVR 15.

Gene Ontology (GO)
The proteins identified were grouped according to GO terms and are listed in Tables 2 and 3. Lineage N2 had three proteins predicted for Molecular Function, nine for Cellular Component, and five for Biological Process.The distribution of the proteins was different between the IVR 15 lineage and the N2 lineage, principally regarding Biological Process with 11 proteins.

Protein interactions
The protein interactions exclusively expressed by the IVR 15 lineage are presented in Figure 4. ENOL-1, ACT-1, PDI-2, and ATP-2 have been predicted to interact by co-expression and textmining, and there are also interactions that were experimentally determined.Beyond the interaction described previously, PDI-2, ATP-2, and VHA-12 demonstrated homology according to curated databases.The FAR-1 and FAR-2 proteins were co-expressed and textmined.However, no interaction was predicted among MLC-1, C35B1.1 and the other proteins exclusively expressed as mentioned above.

Discussion
Ivermectin acts upon chlorine channels controlled by glutamate, causing hyperpolarization of the cellular membrane which paralyzes the muscular system.However, some nematodes are able to withstand the selection pressure promoted by the anthelmintics (IKEDA, 2003).To understand anthelmintic resistance to ivermectin, the two mechanisms of study were adopted: (I) Specific mechanism, result of the modification and/or change in the drug receptor and (II) Unspecific mechanism, characterized by cellular detoxification using proteins considered to be efflux pumps (RAZA et al., 2016).Protein differences associated with resistance to ivermectin v. 28, n. 1, jan.-mar. 2019 109/112 109 Therefore, to observe the protein alterations caused by ivermectin in the C. elegans-resistant lineage, it was necessary to use the wild-type lineage as a parameter because the IVR 15 was not precisely identical to the field conditions faced by parasite nematodes because it does not occur in a controlled and standardized form.However, the genetic selection of the IVR 15 lineage represents an excellent alternative to understanding the molecular puzzle associated with anthelmintic resistance against ivermectin in nematodes (JAMES & DAVEY, 2009).
Thus, comparing the protein expression between IVR 15 and N2, it was possible to identify the most highly expressed or exclusively expressed proteins (Table 1), which have not been associated with anthelmintic resistance in nematodes to date.The comparison was made initially by the classification of differentially expressed proteins in the N2 lineage for the respective GO terms.The molecular function group was divided into categories: (1) oxidoreductase, (2) catalytic, and (3) structural molecular.According to the Biological Process predictions, we inferred that the proteins were involved in the (1) cellular process, (2) metabolism and (3) stimuli responses.Janssen et al. (2013) highlighted that C. elegans possesses a greater demand for proteins responding to environmental interference than rodents and humans.C. elegans is part of a habitat full of ecological interactions, it is a free-living organism in direct contact  with the soil.Thus, it is observed that the expression of proteins that respond to environmental stimuli is maintained even when C. elegans develops in vitro and in ideal conditions.
The protein expression profile of the IVR 15 (Table 1) lineage shows variations in the distribution in their GO terms related to the N2 lineage (Table 2).The main differences were demonstrated in the classification according to the molecular function.The proteins were divided into (1) ligands, (2) catalytic activity and (3) transport activity.This result suggests an increase in metabolic rate and elimination of substances that can destabilize natural cell organization.
Proteins with increased expression in IVR 15 corroborate the results acquired in the classifications of the GO terms (Table 3).The maintenance of high metabolic rates requires proportional energetic production.Furthermore, the expression of such proteins as the ATP synthase β subunit (ATP-2) and enolase-1 (ENOL-1) is necessary to supply the nematode energetic needs to maintain cellular stability.ATP-2 is responsible for coding a part of ATP synthase, which is important inaerobic respiration and energetic production (DURAI et al., 2014).ENOL-1facilitates the ATP formation process during glycolysis, inducing the activation of heat-shock proteins and helping in the larval development and in the apoptosis inhibition (LUO et al., 2017).The findings of Chen et al. (2012) confirm the results of this study, suggesting a change in the amount of ATP due to the change in the production of ENOL-1.
To maintain balanced physiological processes, the increase in the expression of the β subunit of V-type ATPase (VHA-12) in the C. elegans-resistant lineage contributes to embryo development under adverse conditions.Ernstrom et al. (2012) silenced the VHA-12 gene in C. elegans and verified that it does not influence oogenesis or spermatogenesis in the germinative line of the adult phase, although it can assist in the initial embryo development process.VHA-12 promotes acidification in synaptic vesicles, acting in neurotransmitter normalization using a positive membrane potential in the cellular interior in adult nematode nervous cells and suppressing cellular necrosis (ROST et al., 2015).This function allows ivermectin-resistant nematodes to maintain nervous impulse transmission, thereby ensuring normal muscle contraction among other benefits.
In addition, to complement the neuromuscular effect of VHA-12, there was an increase in the production of protein myosin regulatory light chain 1 (MLC-1) and disulfide isomerase 2 (PDI-2).Usually, MLC is found in the entire body of the nematode (RUSHFORTH et al., 1998).Li et al. (2010) observed an increase in MLC-1 expression that affects the cytoskeleton of nematodes subject to stress by hypoxia.Conversely, PDI-2 is important in collagen matrix formation and is located in the hypodermis, where it helps to maintain nematode physiology, cellular homeostasis and protect against stress in the endoplasmic reticulum, the site responsible for the unfolding of secreted proteins (DURAI et al., 2014).Among the exclusively expressed proteins of the IVR 15 lineage, actin-1 (ACT-1) joins the PDI-1 and MLC-1 proteins to ensure nematode muscular viability because ACT-1 is responsible for cytoskeletal formation in addition to muscular regions and myofilaments (MACQUEEN et al., 2005).
The other exclusively expressed proteins in IVR 15 were the fatty acid and retinol-binding proteins 1 and 2 (FAR-1 and FAR-2) and C35B1.5 protein.The FAR-1 and FAR-2 proteins have six other isoforms described in C. elegans, which are localized in the  cell membrane and are related to cell signaling and the immune response.Previous studies showed the bound capacity of these proteins in anthelmintics, such as albendazole, levamisole, abamectin, and praziquantel (REY-BURUSCO et al., 2015;QIAO et al., 2016).Thus, it is possible to suggest that FAR-1 and FAR-2 promote ivermectin resistance mechanisms by triggering biological signaling in addition to their possible role in recruiting fatty acids to contribute to the maintenance of energetic balance during the stress induced by the anthelmintic agent.Moreover, new studies detailing the process and the possible signaling pathways of FAR-1 and FAR-2 proteins are warranted.Furthermore, a database enrichment of C35B1.5 protein is needed because there is not enough information available at present.It is known that this protein is detected in excretory cells, germ cells, the intestine and GABA-like excitatory neuronal receptors (SHAYE & GREENWALD, 2011).The proteins identified in the IVR 15 lineage were mapped and there was no identification of p-glycoprotein as expected for the resistant lineage.We suggest that this fact is a consequence of the technique used, since Hart et al. (2015) using the same technique reported similar results.Figueiredo et al. ( 2018) identified p-glycoproteins using gene expression mapping.The proteins identified when compared with the database showed co-expression among themselves.This finding suggests that co-expression is triggered as a response to stress caused by ivermectin, highlighting the series response that makes C. elegans lineage resistant.
In conclusion, we can observe that the results obtained by identifying new proteins exclusively and differentially expressed associated ivermectin resistance in C. elegans, as well as the studies of molecular interactions, provided new data regarding the defense mechanism utilized by the nematode during resistant lineage selection.Based on the collected data, we observed a possible contraposition to ivermectin action in all affected tissues through the augmented expression of proteins related to cellular and structural viability, energy production, survival and nematode reproduction under the tested conditions.In addition, these results point to proteins that can be employed to help characterize and monitor anthelmintic resistance to ivermectin.

Figure 3 .
Figure 3. Two-dimensional electrophoresis gels with protein samples from the N2 (A and B) and IVR 15 (C and D) lineages of C. elegans.Notes: pI -Isoeletric point; WM -Weight Molecular; kDa -Kilodalton.

Table 1 .
Proteins with modified expression in ivermectin-sensitive lineage, proteins more expressed and proteins exclusively expressed ivermectin-resistant lineages from C. elegans were identified by mass spectrometry NanoESI-Q-TOF.

Table 3 .
Predicted functions of the proteins identified in the C. elegans lineage resistant to ivermectin by gene ontology annotation.

Table 2 .
Predicted functions of the proteins identified in the C. elegans lineage sensitive to ivermectin by gene ontology annotation.