From the feces to the genome: a guideline for the isolation and preservation of Strongyloides stercoralis in the field for genetic and genomic analysis of individual worms

Strongyloidiasis is a soil-borne helminthiasis, which, in spite of the up to 370 million people currently estimated to be infected with its causing agent, the nematode Strongyloides stercoralis, is frequently overlooked. Recent molecular taxonomic studies conducted in Southeast Asia and Australia, showed that dogs can carry the same genotypes of S. stercoralis that also infect humans, in addition to a presumably dog-specific Strongyloides species. This suggests a potential for zoonotic transmission of S. stercoralis from dogs to humans. Although natural S. stercoralis infections have not been reported in any host other than humans, non-human primates and dogs, other as yet unidentified animal reservoirs cannot be excluded. Molecular studies also showed that humans carry rather different genotypes of S. stercoralis. As a result, their taxonomic status and the question of whether they differ in their pathogenic potential remains open. It would therefore be very important to obtain molecular genetic/genomic information about S. stercoralis populations from around the world. One way of achieving this (with little additional sampling effort) would be that people encountering S. stercoralis in the process of their diagnostic work preserve some specimens for molecular analysis. Here we provide a guideline for the isolation, preservation, genotyping at the nuclear 18S rDNA and the mitochondrial cox1 loci, and for whole genome sequencing of single S. stercoralis worms. Since in many cases the full analysis is not possible or desired at the place and time where S. stercoralis are found, we emphasize when and how samples can be preserved, stored and shipped for later analysis. We hope this will benefit and encourage researchers conducting field studies or diagnostics to collect and preserve S. stercoralis for molecular genetic/genomic analyses and either analyze them themselves or make them available to others for further analysis.


Background
Strongyloidiasis is a soil-transmitted helminthiasis (STH) and although as such it is included within the neglected tropical diseases (NTDs), it is often overlooked in comparison with other STHs and has therefore sometimes been described as (one of ) the most neglected NTDs [1][2][3]. Published estimates of the number of people currently infected with Strongyloides stercoralis, the species which causes the vast majority of human strongyloidiasis cases, vary between "30-100 million" [4,5], and "at least 370 million" [2]. Given the difficulties with diagnosis, [5][6][7][8] the actual number may be even higher. In general, S. stercoralis is more prevalent in tropical and subtropical regions, with a local prevalence of 10-40% reported in some endemic areas [9,10]. High temperature and moisture, poverty, poor sanitary conditions such as walking barefoot and open defecation are risk factors for S. stercoralis transmission [1]. However, strongyloidiasis is not limited to the tropical and subtropical regions and increasing numbers of S. stercoralis cases in areas with a

Open Access
Parasites & Vectors *Correspondence: adrian.streit@tuebingen.mpg.de 2 Department of Integrative Evolutionary Biology, Max Planck Institute for Developmental Biology, Tübingen, Baden-Württemberg, Germany Full list of author information is available at the end of the article moderate climate and comparably good sanitary conditions such as Europe, North America and Australia have been reported over the last years [11][12][13][14][15][16].
For various reasons, notably the absence of eggs from the stool and the often very low parasite burden, S. stercoralis remains frequently undetected in standard parasitological surveys [6]. However, since there is an increasing awareness of S. stercoralis and the difficulties with its diagnosis, increasing numbers of studies with specific methodology have been conducted [5][6][7][8].
Although PCR based methods are available [17,18], in most cases S. stercoralis diagnostics rely upon the direct observation of worms in stool using techniques such as stool smears, Kato-Katz, formalin ethyl acetate concentration, agar plate culture, Harada-Mori filter paper culture and the Baermann technique. Agar culture and Baermann were found to be fairly highly sensitive, when multiple stool samples from consecutive days are examined ( [1,5] and references therein).
Although the biology and pathology of S. stercoralis has been studied for more than 100 years [19], many questions remain open. An important one is to what extent strongyloidiasis is a zoonotic disease. In particular, dogs have been implicated as a possible animal reservoir; Brumpt [20] had proposed a species named S. canis as the natural Strongyloides species in dogs, which is different from S. stercoralis found in human and non-human primates. Although this was originally supported by some authors [21,22], later work [23] considered S. canis a nomen dubium, mainly due to the lack of a proper description. Given that no morphological differences could be observed between S. canis and S. stercoralis and that dogs are clearly suitable experimental hosts for human-derived S. stercoralis [24,25], the existence of a separate species S. canis was not generally accepted. Hence, it remained open whether Strongyloides spp. found naturally in humans and in dogs are actually the same species or not. Recent molecular taxonomic studies conducted in Southeast Asia [26,27] and Australia [18] showed that dogs carry at least two different populations (probably species) of Strongyloides, one of which appears to be the same S. stercoralis that also infects humans, underpinning the potential for zoonotic S. stercoralis transmission from dogs to humans. No comparable studies from other parts of the world are available. Interestingly, the very few Strongyloides worms isolated from dogs elsewhere in the world for which DNA sequence information was reported, appear to be of the human-infective type [28][29][30][31][32]. Although natural S. stercoralis infections have not been consistently reported in any host other than humans, non-human primates and dogs, other as yet unidentified animal reservoirs cannot be excluded.
Based on morphology, Strongyloides larvae or freeliving adults can normally be reliably identified to the genus (Strongyloides) but the species is frequently inferred from the host because of the lack of conclusive morphological differences, particularly in the developmental stages existing outside of the host [33,34]. Further, as humans carry rather different 18S rDNA and cox1 genotypes, it remains unclear what their exact phylogenetic relationship is and if they might differ in their pathogenic potential [26,27,35].
To clarify these issues, it would be most valuable to obtain molecular genetic/genomic information about S. stercoralis populations from around the world. One way of achieving this with little additional sampling effort would be that people encountering S. stercoralis in the process of their diagnostic work preserve some specimens for molecular analysis.
In this paper we provide a guideline for the genotyping and whole genome sequencing of single S. stercoralis, from worm isolation in the field, to the generation of the sequences in the laboratory (the bioinformatics analysis of the sequences is not part of this article). Because in many cases, a lack of equipment, time, funds or expertise may exist, or the full analysis is not possible, necessary or desired at the place and time where S. stercoralis are found, we emphasize here when and how samples can be preserved and shipped for later analysis. We hope this will benefit and encourage researchers conducting field studies or diagnostics to collect and preserve S. stercoralis for molecular genetic/genomic analyses and either analyze them themselves or send them to an appropriately equipped laboratory for further analysis.

Part 1. In the field: collection and preservation
In this first part we describe how S. stercoralis worms can be isolated, preserved and shipped. We describe our preferred method for S. stercoralis isolation, which includes culturing samples for 1-2 days followed by harvesting worms using the Baermann technique. This has the advantage that it allows the S. stercoralis worms to develop into infective larvae (homogonic development) or free-living females or males (heterogonic development), and therefore provides information on the sex and the developmental route that the analyzed worm underwent ( Fig. 1). In the case of adults, it also leads to more DNA per worm compared with young larvae. However, we do know that this method is frequently not practical in large-scale routine diagnostics or parasitological surveys. Therefore, we would like to stress that worms of any developmental stage isolated by any method can be used for molecular analysis.

Section 1: Fecal sampling from humans and animals
Soil nematodes are omnipresent and colonize deposited feces very quickly. While this causes no major problems for diagnostics using egg flotation, free-living nematodes are a disturbance in parasitic nematode cultures. Therefore, whenever possible, feces should be collected directly into sampling containers without contacting soil. This is usually the case for human samples. For animal samples, we suggest collecting feces directly from the rectum if possible, or to keep the animals on a clean surface and collect the fresh feces shortly after defecation.

Section 2: Sample culturing and worm isolation
(i) Set up cultures: If very young larvae should be isolated, the feces can be directly processed without culturing [see (iii)]. Otherwise, mix the fresh feces with an about equal volume of sawdust or charcoal bits in a clean container or bag. If the feces are too dry, add some water. The cultures should form a loose, moist, crumbly mass and not be soaking wet (Fig. 2). This serves to allow gas exchange and prevent anaerobic conditions. To avoid hypoxia, do not seal the container or bag but leave a gap to allow air exchange. If possible, use about 50-100 g of feces for one culture; however, we have successfully isolated S. stercoralis from much less (only a few grams) feces. (ii) Incubation conditions: Incubate the samples at 28 °C. If no incubator is available, samples can be kept at ambient temperature. Strongyloides stercoralis appears to be fairly temperature tolerant, as might be expected given that most S. stercoralis endemic regions have a tropical or subtropical climate. To maintain humidity, place some water in an open container in the incubator or cover the samples with a wet towel (Fig. 3). Do not refrigerate the samples or the cultures at any point. In our hands this is highly detrimental to the survival of various species of Strongyloides. (iii) Incubation time: S. stercoralis larvae hatch within the host. Therefore, to collect young larvae, fresh fecal samples can be processed either directly or after being cultured for a few hours. To collect free-living adults and differentiated infective larvae (iL3s) of the direct (homogonic) cycle, samples should be cultured for 1-2 days. To collect iL3s of the indirect (heterogonic) cycle, samples should be cultured for 6-9 days. (iv) Set up Baermann funnels: The simplest way is to place a piece of rubber tubing at the bottom of a glass funnel, close the end of the tube with a clamp and fill the funnel with tap water. Wrap the fecal sample with tissue paper and place it into the funnel (Fig. 4a). This can be done at ambient temperature. For a more sophisticated set up, see [25]. (v) Harvest worms: After the worms have crawled out of the feces and accumulated at the bottom of the funnel (allow at least two hours), place a second clamp 1-2 cm above the first clamp, remove the lower clamp and collect the worms into a watch glass or small Petri dish (Fig. 4b).
We recommend to transfer the worms into clean water and wash a few times prior to lysis or preservation in order to minimize the DNA derived from other organisms present in the fecal sample.

Section 3: Morphological observation
Strongyloides stercoralis in the watch glass can be directly observed under a dissecting stereomicroscope with illumination from below. Depending on the incubation time and temperature, different developmental stages of worms can be observed.   The L1 and L2 larvae can be observed from fresh feces and after a few hours of culture. They are approximately 0.18-0.54 mm in length. For non-specialists they are the most difficult stages to tell apart from, e.g. young hookworm larvae. However, S. stercoralis is normally the only nematode species of which larvae rather than eggs can be found in fresh human stool samples.
Free-living adults and differentiated infective iL3s of the direct (homogonic) cycle can be observed after one to two days of culture while the differentiated iL3s of the indirect (heterogonic) cycle can be observed after roughly 6 days (depending on the temperature). Strongyloides spp. are the only human gastrointestinal nematodes we are aware of that form reproductive adults outside of the host. The free-living adults are rhabditiform, approximately 0.8-1.2 mm in length and 40-70 μm in width. Females and males can be easily distinguished by their morphology (Fig. 5): Females are about 1.1-1.3 times larger than males. The female has a didelphic ovary and the vulva opens at mid-body length. The male has one testis, equal spicules and a ventrally curved tail.
The iL3s are filariform, approximately 0.4-0.8 mm in length and 14-19 μm in width and have a trifurcated tail. The most unique feature which sets Strongyloides (and Parastrongyloides) spp. apart from all other nematodes, is their clear oesophagus which extends over almost half of the body length followed by the much darker intestine.
Since we assume that most readers are accomplished diagnosticians who can recognize S. stercoralis, we do not provide further instructions on how to identify S. stercoralis based on their morphology. For a detailed morphological description of S. stercoralis we refer the reader to the classical paper [36] (reproduced in [37]).

Section 4: Sample preservation
After the steps described above, the S. stercoralis worms are alive in clean water. They can be kept in this stage at ambient temperature (in our hands tested from 25 °C to 32 °C) for several hours or in the case of iL3s several days. Consequentially, iL3s can be shipped in water by express Use (C) if: you want to store the worms as a batch and single out individuals later or process them as a batch.

A. Preservation of individual worms in water (frozen)
For each worm: (i) Add 9 μl of water into a PCR tube.
(ii) Transfer an individual worm in about 1 μl of water into the PCR tube using a mouth pipette or a micropipette (alternatively, use a worm pick to transfer the worm into a PCR tube containing 10 μl of water). (iii) Store frozen until analysis, preferably at −80 °C or lower, but −20 °C will do for a number of years (the oldest sample we tested had been stored for three years at −20 °C). (iv) If shipping is required, samples should be kept frozen (e.g. on dry ice).

B. Preservation of individual worms in ethanol (ambient temperature)
For each worm: (i) Add 9 μl of water into a PCR tube.

C. Preservation of a pool of worms in ethanol (ambient temperature)
For each pool:

Part 2. In the laboratory: molecular analysis
Below we describe our preferred method for DNA extraction from the single worms. While undoubtedly many other DNA extraction methods would also work, this method is rather simple, fast, does not require any commercial kits and is suitable for processing samples on a large scale. This method can also be used for extracting DNA from pools of worms. We also describe the protocol for PCR amplification and sequencing of genomic loci and we recommend several loci for Strongyloides species identification. Additionally, we provide a relatively simple and cost-effective solution for the whole genomic library preparation of single worms.

A. For worms preserved individually in water
(i) Freeze the sample by immersing the tube in liquid Nitrogen for a few seconds (or in a freezer at −80 °C for a few minutes) and thaw at room temperature; vortex and briefly centrifuge. Repeat the freeze-thaw cycle 3 times. (ii) Add 10 μl 2× lysis buffer and mix. (iii) Incubate at 65 °C for 2 h. Remark: at this step several published protocols, including some from our laboratory, call for inactivation of the proteinase K at 95 °C. We found this step to be unnecessary for subsequent PCR and even highly detrimental for the genomic library preparation using the tagmentation technique described in this paper (see Section 3 below), presumably due to partial denaturation of the DNA caused by the high temperature. (iv) The lysate can be used for PCR immediately or stored at −20 °C. This is also a stage at which the samples can be shipped. Preferentially, they should be kept frozen but the DNA is fairly stable and we have shipped successfully on wet ice.

B. For worms preserved individually in ethanol
(i) With a pipette, remove as much of the ethanol as possible without discarding the worm. (ii) Let the remaining liquid evaporate by leaving the sample with the lid open. Alternatively, the evaporation can be accelerated by using a vacuum concentrator. (iii) Add 10 μl water. (iv) Perform the lysis as described above in (A).

C. For a pool of worms preserved in ethanol
(i) Transfer the worms with ethanol into a watch glass. (ii) Remove as much of the liquid as possible without discarding the worms. (iii) To rehydrate the worms, add water into the watch glass. At first, the worms will float on the water surface. Remove as much of the liquid as possible without discarding worms, and add water again. Repeat this washing with water step at least two more times. (iv) Add 9 μl water in a PCR tube.
(v) Transfer individual worms into the PCR tube (one worm/tube) with a micropipette or using a worm pick. (vi) Perform the lysis as described above in (A).

Section 2: PCR amplification and sequencing of selected loci
As a basis for a first molecular classification of the Strongyloides species, we recommend to PCR amplify and sequence the nuclear 18S rDNA (SSU) hypervariable regions (HVR) I and IV and a portion of the mitochondrial gene cox1, as suggested by Hasegawa and colleagues [29,30,38]. Additionally, the worm lysates are also suitable for the amplification of single copy loci. As a well-working example, we describe here ytP274 (a polymorphic single copy locus) that was used by [26] to demonstrate sexual reproduction and can serve as a positive control.
The lysates of single worms can be directly used for PCR amplification. Normally 1-2 µl lysate is sufficient for amplification. Here we describe our PCR protocol using Taq DNA polymerase and ThermoPol Reaction Buffer (M0267S, New England BioLabs, Ipswich, USA). Other polymerases also work but the protocol may need minor modifications according to the instructions of the respective manufacturer (e.g. 72 °C instead of 68 °C for the extension step). The preparation of a 25 µl PCR reaction mix is described in Table 1. The primers and their respective annealing temperatures are listed in Table 2. The PCR cycling program is as follows: an initial denaturation step at 95 °C for 30 s, followed by 35 cycles of (denaturation at 95 °C for 20 s, annealing for 15 s, extension at 68 °C for 90 s), and a final extension step of 5 min at 68 °C.
The PCR product (0.5-1 µl) can be directly used for conventional Sanger sequencing with the primers listed in Table 2 using the BigDye Terminator v3.1 Cycle sequencing Kit (Applied Biosystems, Foster City, USA). Using other kits or submitting to a commercial sequencing facility might require additional purification of the PCR product.
For the SSU HVR-I, many studies in various nematodes used the primer pair SSU18A and SSU26R, because this primer pair is fairly universal and can also amplify the SSU HVR I from other nematodes such as hookworms [26,35,[39][40][41][42]. However, SSU18A is highly AT rich (37% GC) with a low Tm (49 °C) and does not match the S. stercoralis sequence perfectly. In our hands, PCR amplification from single worms, especially larvae, with this primer failed relatively frequently. Therefore, in cases where a worm is already known to be Strongyloides species, we recommend the use of ZS6492 as a forward primer. In our hands this primer works more reliably, compared with SSU18A, for Strongyloides spp. but does not work well for hookworms and presumably other nematodes.

Section 3: Whole genome sequencing of single worms
The worm lysate of single worms (as described above in Section 1) can be used to generate DNA libraries for whole genome sequencing. The methods for library preparation with very little DNA as input are currently evolving very rapidly and we recommend checking for improved methodology before starting with whole genome sequencing.
We have used several commercial low input kits, such as the Nextera DNA Library Prep Kit (Illumina, Dan Diego, USA), and the Low Input Library Prep Kit (Clontech, Kusatsu, Japan) [26]. Recently, we have developed a method based on Tn5 transposome tagmentation [45,  Below we describe a protocol, which worked for S. stercoralis and certain other nematodes. We use the Tn5 transposomes from the Illumina Nextera DNA Library Prep kit. The Tn5 (170 µl) of one kit is sufficient for the preparation of about 8500 single worm libraries of S. stercoralis making it cost effective. The required reagents including the Tn5 transposomes can also be prepared in the laboratory [46] (the Tn5 preparation is not part of this paper).  We have sequenced several genomic libraries prepared as described above of individual free-living adults and iL3s of S. stercoralis preserved in water or ethanol. When we mapped the raw reads to the reference genome (PRJEB528.WBPS11) with bwa [47] we noticed that the percentage of reads that did map to the reference as determined with Qualimap [48] was rather variable. The mapping ratios for the individuals included in our recent publication [35] (GenBank: PRJNA517237) ranged between 32.3 and 95.2% for adults (mean = 79.6%, median = 92.7%, n = 10) and between 3.5 and 58.4% for infective larvae (mean = 21.1%, median = 13.0%, n = 26). For comparison, we also examined the libraries of individual S. stercoralis in our earlier publication [26] (Gen-Bank: PRJEB20999), for which a commercial low input kit not relying on Tn5 tagmentation had been used to prepare the genomic libraries. The mapping ratios for adults (no iL3s were sequenced for this paper) ranged between 28.3 and 81.24% (mean = 55.9%, median = 60.1%, n = 23).

Reagents needed
While even the lowest proportion of mapped reads (3.5%) for the individual iL3 still provided us with sufficient sequence information for phylogenetic analysis, the variability we observed illustrates that there is room for methodological improvement. A higher mapping ratio will increase the number of samples that can be multiplex-sequenced per lane and thus further reduce the cost. We randomly looked at some of the un-mapped reads and most of them did not map to any bacterial genome or the host genome, indicating that the majority of unmapped reads were not derived from bacterial or host tissue contamination. Many un-mappable reads are probably the product of artefacts associated with the very low DNA input, such as PCR-induced errors in the process of the increased number of amplification cycles or overtagmentation of the DNA due to an excess of Tn5 [46].

Conclusions
We do appreciate that for many people whole genome sequencing is not a primary interest and may also not be within the limits for what the funds had been approved for. Also, it requires a rather sophisticated and time-consuming analysis. Therefore, we would like to stress that 18S and cox1 sequence information is most valuable in order to put a local S. stercoralis population into a global context. Above this, if specimens are preserved and stored properly, whole genome analysis can be performed later for selected, particularly interesting samples. In many areas where the conditions appear favorable for S. stercoralis transmission, no or very few studies have been conducted (e.g. for more than half of the African countries we could not find any S. stercoralis prevalence data in the published literature). Among the rather few available studies (e.g. case reports, clinical diagnosis, community surveys) most were solely based on morphology and did not provide any molecular information. Therefore, we hope this methodological article will encourage people around the world who encounter S. stercoralis in the process of their work to participate in a community effort to clarify the genetic structure, phylogenetic relationships and taxonomic status of S. stercoralis populations in different hosts and geographical locations.