METHODS EMPLOYED TO STUDY THE INHERITANCE AND GENE FUNCTION IN STRONGYLOIDES SPP.

Most of the classical analyses about the modes of reproduction in Strongyloides spp. were based on cytological observations (see, for examples, Nigon and Roman, Reference Nigon and Roman1952; Zaffagnini, Reference Zaffagnini1973; Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977; Albertson et al. Reference Albertson, Nwaorgu and Sulston1979; Hammond and Robinson, Reference Hammond and Robinson1994). Owing to the methodology available at the time, these studies were done on fixed specimens, which make elucidating the dynamics of the processes difficult.

Studying certain traits (e.g. the ratios between male and female progeny or between homogonic and heterogonic development) over time and varying culture regimes provided hints about the modes of inheritance. Prominent among these are very extensive studies by Graham, for which he maintained Strongyloides ratti for many generations exclusively through the homogonic or the heterogonic cycle, even as successive single worm infections (Graham, Reference Graham1936, Reference Graham1938, Reference Graham1939a , Reference Graham b , Reference Graham1940a , Reference Graham b ). In agreement with Sandground (Reference Sandground1926), who had done similar but less extensive studies on Strongyloides papillosus and S. ratti, Graham (Reference Graham1939b ) noticed a ‘remarkable constancy of characteristics’ when Strongyloides spp. reproduced through the homogonic cycle while variability arose in cultures derived from the heterogonic cycle. This indicated that in the progeny of the free-living generation but not the parasitic generation new genetic combinations are created through recombination associated with sexual reproduction.

Later, with the emergence of PCR and relatively inexpensive sequencing, molecular markers including first micorsatellites and later single copy loci provided tools to study the passage of genetic information in mass matings of males and females of different strains or in single male – female crosses (examples of this type of study are Viney et al. Reference Viney, Matthews and Walliker1993; Viney, Reference Viney1994; Harvey and Viney, Reference Harvey and Viney2001; Grant et al. Reference Grant, Stasiuk, Newton-Howes, Ralston, Bisset, Heath and Shoemaker2006; Eberhardt et al. Reference Eberhardt, Mayer and Streit2007; Nemetschke et al. Reference Nemetschke, Eberhardt, Viney and Streit2010b ).

Unfortunately, as far as studying gene function in Strongyloides is concerned, minimal success can be reported at this time. DNA and RNA sequencing efforts have provided probably close to complete gene lists for multiple species of Strongyloides (Hunt et al. Reference Hunt, Tsai, Coghlan, Reid, Holroyd, Foth, Tracey, Cotton, Stanley, Beasley, Bennett, Brooks, Harsha, Kajitani, Kulkarni, Harbecke, Nagayasu, Nichol, Ogura, Quail, Randle, Ribeiro, Sanchez-Flores, Hayashi, Itoh, Denver, Grant, Stoltzfus, Lok and Murayama2016) and a number of microarray analyses (Evans et al. Reference Evans, Mello, Fang, Wit, Thompson, Viney and Paterson2008; Thompson et al. Reference Thompson, Barker, Hughes and Viney2008, Reference Thompson, Barker, Nolan, Gems and Viney2009; O'Meara et al. Reference O'Meara, Barber, Mello, Sangaralingam, Viney and Paterson2010; Ramanathan et al. Reference Ramanathan, Varma, Ribeiro, Myers, Nolan, Abraham, Lok and Nutman2011) and quantitative RNA sequencing experiments (Yoshida et al. Reference Yoshida, Nagayasu, Nishimaki, Sawaguchi, Yanagawa and Maruyama2011; Marcilla et al. Reference Marcilla, Garg, Bernal, Ranganathan, Forment, Ortiz, Munoz-Antoli, Dominguez, Pedrola, Martinez-Blanch, Sotillo, Trelis, Toledo and Esteban2012; Ahmed et al. Reference Ahmed, Chang, Younis, Langnick, Li, Chen, Brattig and Dieterich2013; Nagayasu et al. Reference Nagayasu, Ogura, Itoh, Yoshida, Chakraborty, Hayashi and Maruyama2013) using RNA extracted from different developmental stages and proteome analyses (Marcilla et al. Reference Marcilla, Sotillo, Perez-Garcia, Igual-Adell, Valero, Sanchez-Pino, Bernal, Munoz-Antoli, Trelis, Toledo and Esteban2010; Soblik et al. Reference Soblik, Younis, Mitreva, Renard, Kirchner, Geisinger, Steen and Brattig2011; Younis et al. Reference Younis, Geisinger, Ajonina-Ekoti, Soblik, Steen, Mitreva, Erttmann, Perbandt, Liebau and Brattig2011) provided a comparative overview of gene expression patterns and protein secretion. However, I would argue that so far such omics approaches have provided and can provide only very limited information about the functions of individual genes (see also Viney, Reference Viney2014).

The possible functions of selected genes, normally selected because of the presence of homologous genes with known functions in other systems, were approached through detailed molecular characterization of the genes and their products including their temporal expression and localization (Tazir et al. Reference Tazir, Steisslinger, Soblik, Younis, Beckmann, Grevelding, Steen, Brattig and Erttmann2009; Peeters et al. Reference Peeters, Janssen, De Haes, Beets, Meelkop, Grant and Schoofs2011; Younis et al. Reference Younis, Geisinger, Ajonina-Ekoti, Soblik, Steen, Mitreva, Erttmann, Perbandt, Liebau and Brattig2011; Biewener et al. Reference Biewener, Welz, Khumpool, Kuttler and Schnieder2012). Thanks to the recent addition of transgenic techniques to the available tool kit for Strongyloides spp. research (Lok, Reference Lok2013), these approaches have been complemented with the use of reporter constructs (e.g. Yuan et al. Reference Yuan, Liu, Lok, Stoltzfus, Gasser, Lei, Fang, Zhao and Hu2014a , Reference Yuan, Lok, Stoltzfus, Gasser, Fang, Lei, Fang, Zhou, Zhao and Hu b ).

In order to study gene function in model organisms, mutations in genes obtained either by random mutagenesis followed by screening for phenotypes of interest (forward genetics) or by targeted knock out of molecularly known genes (reverse genetics) have been and still are tremendously useful tools for the investigation of gene function (Hodgkin, Reference Hodgkin2005; Kutscher and Shaham, Reference Kutscher and Shaham2014). However, as yet attempts at both forward and reverse genetics in Strongyloides spp. have been unsuccessful.

Isolating mutations for genetic analysis of endo-parasitic organisms is usually difficult because the adults, which are required for mutagenesis and later crossing, are located within the host. Strongyloides/Parastrongyloides with the free-living adult generations is an exception to this and appears much more suitable for this kind of approach; nevertheless, no success can be reported at this time. The only two reports of successful mutagenesis in Strongyloides spp. come from S. ratti, demonstrating that mutagenesis and screening for phenotypes of interest is possible (Viney et al. Reference Viney, Green, Brooks and Grant2002; Guo et al. Reference Guo, Chang, Dieterich and Streit2015). However, the main problem that has not yet been solved is the isolation of the molecular mutation causing this phenotype and with it the identification of the corresponding gene. In the model nematode Caenorhabditis elegans mutated genes have traditionally been identified by a process called positional cloning (Hodgkin, Reference Hodgkin and Hope1999; Fay, Reference Fay2006). For this strategy one needs a dense, high-quality genetic map for precise genetic mapping and a reliable physical map (ideally a full genome sequence) that is highly interlinked with the genetic map. In addition, transgenic technology is normally employed for gene verification after tentative identification. Positional cloning has so far been used successfully only in two nematodes other than C. elegans, namely in Caenorhabditis briggsae (Koboldt et al. Reference Koboldt, Staisch, Thillainathan, Haines, Baird, Chamberlin, Haag, Miller and Gupta2010) and in Pristionchus pacificus (Zheng et al. Reference Zheng, Messerschmidt, Jungblut and Sommer2005; Dieterich et al. Reference Dieterich, Roeseler and Srinivasan2006). Although a genetic map for S. ratti (Nemetschke et al. Reference Nemetschke, Eberhardt, Viney and Streit2010b ), a high-quality genome sequence (Hunt et al. Reference Hunt, Tsai, Coghlan, Reid, Holroyd, Foth, Tracey, Cotton, Stanley, Beasley, Bennett, Brooks, Harsha, Kajitani, Kulkarni, Harbecke, Nagayasu, Nichol, Ogura, Quail, Randle, Ribeiro, Sanchez-Flores, Hayashi, Itoh, Denver, Grant, Stoltzfus, Lok and Murayama2016) and transgenic technology (Shao et al. Reference Shao, Li, Nolan, Massey, Pearce and Lok2012) have been established recently, positional cloning will probably not be the method of choice for identifying mutations in Strongyloides spp. in the future. More likely this will rely on modern sequencing approaches (see conclusions and outlook).

Given the lack of success using forward or reverse genetic approaches for obtaining mutations in known genes in Strongyloides spp., attempts have been made to inactivate genes at least temporarily in order to study their functions. Double-stranded RNA interference (Fire et al. Reference Fire, Xu, Montgomery, Kostas, Driver and Mello1998), which was employed with great success in a number of organisms, unfortunately appears not to work for Strongyloides spp., as is the case for many other animal parasitic nematodes (Viney and Thompson, Reference Viney and Thompson2008). So far, two approaches to manipulate the function of genes were employed successfully in Stronglyoides spp. The first one is modulating the activity of proteins and pathways pharmacologically through the addition of certain chemicals whose activities had been characterized in other systems (Ogawa et al. Reference Ogawa, Streit, Antebi and Sommer2009; Wang et al. Reference Wang, Zhou, Motola, Gao, Suino-Powell, Conneely, Ogata, Sharma, Auchus, Lok, Hawdon, Kliewer, Xu and Mangelsdorf2009, Reference Wang, Stoltzfus, You, Ranjit, Tang, Xie, Lok, Mangelsdorf and Kliewer2015; Stoltzfus et al. Reference Stoltzfus, Massey, Nolan, Griffith and Lok2012a , Reference Stoltzfus, Bart and Lok2014). Second, the recently established methods for transgenesis (Shao et al. Reference Shao, Li, Nolan, Massey, Pearce and Lok2012) allowed the expression of mutant proteins with expected properties, like mimicking or preventing phosphorylation or acting as dominant negatives (Castelletto et al. Reference Castelletto, Massey and Lok2009). Inherently, both these approaches are limited to highly conserved proteins such that it can be assumed that the effects of the chemical compounds are the same in Strongyloides as they are in the organisms in which they had been previously analysed.

An approach that is occasionally used to characterize genes of parasitic nematodes, among them Strongyloides spp. is to test if the parasite gene can rescue the corresponding mutation in the heterologous system C. elegans (Massey et al. Reference Massey, Bhopale, Li, Castelletto and Lok2006; Crook et al. Reference Crook, Grant and Grant2010; Hu et al. Reference Hu, Lok, Ranjit, Massey, Sternberg and Gasser2010). Normally only the coding region is taken from the parasite because the promoters do not function properly across species. Although useful, great caution must be exercised in order not to over interpret such experiments. If a protein derived from Strongyloides spp. can replace the endogenous one in C. elegans this only means that the biochemical properties of this protein are similar enough that the Strongyloides protein can perform the task of the C. elegans protein in C. elegans. However, this finding is completely uninformative about the function of this gene in Strongyloides spp.

MODES OF INHERITANCE IN THE PARASITIC AND THE FREE-LIVING GENERATIONS

Although in the literature various modes of reproduction had been postulated for the parasitic generations of Strongyloides sp. (Streit, Reference Streit2008) there is now wide agreement that at least in the relatively well-studied species of Strongyloides, reproduction is by mitotic parthenogenesis such that the progeny of a parasitic female are genetically identical with the mother. Ignoring new mutations, the only exception to this is the elimination of one copy of the X-chromosome or of the X-derived portion of a chromosome, in order to make males (see below). The arguments for this are summarized below and are reviewed in more detail in Streit (Reference Streit2008).

With the exception of two reports from the 1930s (Kreis, Reference Kreis1932; Faust, Reference Faust1933) describing the same isolates of Strongyloides spp. originally from various primates and dogs (presumably Strongyloides stercoralis and/or S. fuellebornei) all authors agree that no parasitic males exist in any species of Strongyloides analysed so far. Indeed, the presence of parasitic males was one of the decisive criteria for installing the new genus Parastrongyloides (Mackerras, Reference Mackerras1959). In S. ratti, at least, individual infective larvae (L3i) frequently lead to productive infections clearly demonstrating that males are not required for reproduction (Graham, Reference Graham1936, Reference Graham1938; Viney et al. Reference Viney, Matthews and Walliker1992; Viney, Reference Viney1994). After Sandground (Reference Sandground1926) proposed self-fertilization as a mode of reproduction, numerous authors working on multiple species of Strongyloides argued for mitotic parthenogenesis based on (i) cytological observations (Chitwood and Graham, Reference Chitwood and Graham1940; Nigon and Roman, Reference Nigon and Roman1952; Zaffagnini, Reference Zaffagnini1973; Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977); (ii) the observation that heritable traits remain rather stable through rounds of homogonic reproduction (Graham, Reference Graham1939b ), and (iii) molecular genetic observations (Viney, Reference Viney1994; Nemetschke et al. Reference Nemetschke, Eberhardt, Hertzberg and Streit2010a ). Interestingly, contrary to Strongyloides sp., Parastrongyloides trichosuri parasitic adults do reproduce sexually (Mackerras, Reference Mackerras1959; Grant et al. Reference Grant, Stasiuk, Newton-Howes, Ralston, Bisset, Heath and Shoemaker2006; Kulkarni et al. Reference Kulkarni, Dyka, Nemetschke, Grant and Streit2013).

Older literature on the mode of reproduction in the free-living generation is contradictory (Streit, Reference Streit2008). Although a few authors suggested that under certain circumstances free-living Strongyloides spp. females may reproduce in the absence of males (Sandground, Reference Sandground1926; Zaffagnini, Reference Zaffagnini1973), there is wide agreement that males are present and necessary in the free-living generations of all species of Strongyloides tested (Beach, Reference Beach1936; Premvati, 1958b ; Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977; Eberhardt et al. Reference Eberhardt, Mayer and Streit2007). However, most cytological studies on various species of Strongyloides concluded that males do not contribute genetically to the progeny but that reproduction occurs by sperm-dependent parthenogenesis (pseudogamy) (Nigon and Roman, Reference Nigon and Roman1952; Bolla and Roberts, Reference Bolla and Roberts1968; Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977; Hammond and Robinson, Reference Hammond and Robinson1994). Contrary to this, Graham (Reference Graham1939b ) noticed that heritable traits tended to be more variable in cultures maintained through the heterogonic cycle when compared with cultures passaged exclusively through the homogonic cycle, indicating that in contrast to the parasitic generation, recombination of genetic material does occur in the free-living generation. Furthermore, recent molecular genetic work argued clearly for sexual reproduction at least in S. ratti (Viney et al. Reference Viney, Matthews and Walliker1993; Harvey and Viney, Reference Harvey and Viney2001; Nemetschke et al. Reference Nemetschke, Eberhardt, Viney and Streit2010b ), S. papillosus (Eberhardt et al. Reference Eberhardt, Mayer and Streit2007) and Strongyloides vituli (Kulkarni et al. Reference Kulkarni, Dyka, Nemetschke, Grant and Streit2013).

Remark: Rather frequently, I meet colleagues who remember from their textbooks that free-living Strongyloides spp. are supposed to be diploid, while parasitic females are triploid. This information originated from a single reference (Chang and Graham, Reference Chang and Graham1957), in which the authors claimed that the sperm of free-living S. papillosus males contributes one set of chromosomes to the diploid egg produced by the females, leading to triploid individuals destined to become parasitic. Parasitic females, in turn, were proposed to produce triploid and diploid offspring forming the parasitic and free-living progeny, respectively. For several reasons, I believe that Chang and Graham (Reference Chang and Graham1957) should be disregarded. First, this reference is a meeting abstract, which does not contain any detailed description of data and the authors never published these findings in a full publication. Second, multiple authors, based on cytological observations like those by Chang and Graham (Reference Chang and Graham1957), concluded that the free-living and the parasitic females of S. papillosus (Zaffagnini, Reference Zaffagnini1973; Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977; Albertson et al. Reference Albertson, Nwaorgu and Sulston1979), S. stercoralis (Hammond and Robinson, Reference Hammond and Robinson1994) and S. ratti (Nigon and Roman, Reference Nigon and Roman1952) have equal numbers of chromosomes. Third, genetic experiments demonstrated that the progeny of free-living S. papillosus and S. ratti are diploid (Viney et al. Reference Viney, Matthews and Walliker1993; Viney, Reference Viney1994; Eberhardt et al. Reference Eberhardt, Mayer and Streit2007). Fourth, many authors found that female larvae of several Strongyloides species produced by parasitic females definitely commit to either parasitic or free-living live only after they became first-stage larvae (Streit, Reference Streit2008; Viney and Lok, Reference Viney and Lok2015). It is hard to imagine that the larvae going on to become L3i change their ploidy at this stage of development.

SEX DETERMINATION

A puzzling aspect of the life cycle of Strongyloides spp. is that parthenogenetic parasitic females produce two sexes while, with very few exceptions (Streit, Reference Streit2008), the progeny of the sexually reproducing free-living generation is exclusively female. For all species of Strongyloides where it has been studied it was found that different isolates produce very different sex ratios in the progeny of the parasitic generation, indicating that there exists a genetic pre-disposition for more or fewer males (Sandground, Reference Sandground1926; Graham, Reference Graham1939b ; Viney et al. Reference Viney, Matthews and Walliker1992; Viney, Reference Viney1996). However, there is clearly also an environmental effect on the sex ratio produced (Moncol and Triantaphyllou, Reference Moncol and Triantaphyllou1978; Gemmill et al. Reference Gemmill, Viney and Read1997; Harvey et al. Reference Harvey, Gemmill, Read and Viney2000; Crook and Viney, Reference Crook and Viney2005). In all cases studied, the stronger is the immune response of the host, the more males are produced. At the same time the females are pre-disposed but not fixed for the heterogonic cycle. Further, it has been noted that living in a permissive but suboptimal host species alters the sex ratio in the progeny of the parasitic generation, in most (Brumpt, Reference Brumpt1921; Sandground, Reference Sandground1926; Matoff, Reference Matoff1936; Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977) but not all (Crook and Viney, Reference Crook and Viney2005) cases towards more males. Among the different species of Strongyloides two different numbers of chromosomes have been found (Fig. 1). In S. ratti (Nigon and Roman, Reference Nigon and Roman1952; Bolla and Roberts, Reference Bolla and Roberts1968) and S. stercoralis (Hammond and Robinson, Reference Hammond and Robinson1994), the haploid chromosome number is three, namely two autosomes and one X chromosome (n = 3) and all chromosomes are roughly of equal size. In these species (diploid), females have two X chromosomes along with two pairs of autosomes (2n = 6) and males have only one X (2n = 5) resulting in an environmentally influenced XX/XO sex determining system (Harvey and Viney, Reference Harvey and Viney2001). In females of S. papillosus (Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977; Albertson et al. Reference Albertson, Nwaorgu and Sulston1979), Stronglyoides ransomi (Triantaphyllou and Moncol, Reference Triantaphyllou and Moncol1977), Strongyloides venezuelensis (Hino et al. Reference Hino, Tanaka, Takaishi, Fujii, Palomares-Rius, Hasegawa, Maruyama and Kikuchi2014) and S. vituli (Kulkarni et al. Reference Kulkarni, Dyka, Nemetschke, Grant and Streit2013) the diploid chromosome number is only four (2n = 4) and one pair of chromosomes is about twice the size of the other. Correspondingly, females have two large and two medium-sized chromosomes (2L2M). Based on molecular genetic experiments and whole-genome sequencing, it became clear that in these species (strictly shown for S. papillosus and S. venezuelensis) the genomic regions corresponding to the S. ratti chromosomes I and X are combined in the larger chromosome (Nemetschke et al. Reference Nemetschke, Eberhardt, Hertzberg and Streit2010a ; Hunt et al. Reference Hunt, Tsai, Coghlan, Reid, Holroyd, Foth, Tracey, Cotton, Stanley, Beasley, Bennett, Brooks, Harsha, Kajitani, Kulkarni, Harbecke, Nagayasu, Nichol, Ogura, Quail, Randle, Ribeiro, Sanchez-Flores, Hayashi, Itoh, Denver, Grant, Stoltzfus, Lok and Murayama2016). Sex determination in this group of species has been best characterized in S. papillosus. Triantaphyllou and Moncol (Reference Triantaphyllou and Moncol1977) concluded, based on cytological observations that in S. papillosus and S. ransomi males do not differ karyotypically from females. However, later authors described, based on cytology and molecular genetic evidence, that in S. papillosus a male-specific chromatin diminution event takes place in the mitotic oocyte maturation division (Albertson et al. Reference Albertson, Nwaorgu and Sulston1979; Nemetschke et al. Reference Nemetschke, Eberhardt, Hertzberg and Streit2010a ). In the process, the genomic region corresponding to the X chromosome in S. ratti is eliminated from one of the two long (L) homologous chromosomes (Albertson et al. Reference Albertson, Nwaorgu and Sulston1979; Nemetschke et al. Reference Nemetschke, Eberhardt, Hertzberg and Streit2010a ). This leads to individuals with two copies of the regions of the genome that correspond to S. ratti autosomes but only one copy of the genomic region corresponding to the S. ratti X. Because the eliminated portion is flanked by retained regions, which are not joined together upon chromatin diminution, the diploid chromosome number in males is five, namely one long (not diminished X-I fusion chromosome), three medium-sized (the pair M, like in females and one end of the diminished chromosome which is roughly equal in size) and one small (S, the other end of the diminished chromosome) leading to a 1L3M1S chromosomal configuration. Parastrongyloides trichosuri reproduces sexually in both generations and employs XX/XO sex determination with 2n = 6 in females, suggesting that within the genus Strongyloides the mode of sex determination in S. ratti is ancestral (Mackerras, Reference Mackerras1959; Grant et al. Reference Grant, Stasiuk, Newton-Howes, Ralston, Bisset, Heath and Shoemaker2006; Kulkarni et al. Reference Kulkarni, Dyka, Nemetschke, Grant and Streit2013; Streit, Reference Streit2014).

Fig. 1. Chromosomes in P. trichosuri, S. ratti and S. papillosus. Chromosomes and chromosomal regions present in two copies in both sexes are in black, Chromosomes and regions present in two copies in females but only one copy in males are in grey. I: autosome number 1; II: autosome number 2; X: X-chromosome; L: large chromosome; M: medium-sized chromosome; S: small chromosome. For detailed explanation and references see text.

At times when it was assumed that the free-living generation reproduced by pseudogamy the all-female progeny was easily explained because in such a scenario all progeny are genetically identical with the mother and therefore karyotypically female. In the case of sexual reproduction, as it was shown to occur in at least three species of Strongyloides (Viney et al. Reference Viney, Matthews and Walliker1993; Harvey and Viney, Reference Harvey and Viney2001; Eberhardt et al. Reference Eberhardt, Mayer and Streit2007; Nemetschke et al. Reference Nemetschke, Eberhardt, Viney and Streit2010b ; Kulkarni et al. Reference Kulkarni, Dyka, Nemetschke, Grant and Streit2013), this is more difficult to achieve. Several non-mutually exclusive mechanisms are imaginable. Genetically male (XO) embryos might be nonviable. Alternatively, sperm without an X chromosome might be inefficient or even incapable of fertilizing eggs. Alternatively, such sperm might never be formed in the first place. Two lines of evidence suggest that in S. papillosus mature male-determining sperm are never made. (i) For markers that are very closely linked with the region undergoing male-specific chromatin diminution only the allele present on the complete homologue of the autosome-X fusion chromosome and never the one present on the remnants of the diminished chromosome is present in mature sperm (Nemetschke et al. Reference Nemetschke, Eberhardt, Hertzberg and Streit2010a ). (ii) Quantitative DNA sequencing revealed that autosomal and X-derived chromosomal regions are present in equal amounts in mature S. papillosus sperm (Kulkarni et al. Reference Kulkarni, Holz, Rödelsperger, Harbecke and Streit2016). Contrary to this, when DNA isolated from mature S. ratti sperm was quantitatively sequenced, X-derived sequences were present in lower amounts than autosomal sequences, indicating that not all sperm contain an X chromosome (Kulkarni et al. Reference Kulkarni, Holz, Rödelsperger, Harbecke and Streit2016). Dying early embryos were observed consistently in S. ratti but not S. papillosus. However, the number of these dying embryos was lower than what would have been expected based on the number of nullo-X sperm suggested by the sequencing experiments (Kulkarni et al. Reference Kulkarni, Holz, Rödelsperger, Harbecke and Streit2016). This might indicate that in addition, nullo-X sperm fertilize eggs less efficiently than X-bearing sperm.

HOMOGONIC–HETEROGONIC SWITCH

The switch between homogonic and heterogonic development is the most extensively studied process in basic Strongyloides biology. First, the analysis of various isofemale/inbred lines of S. ratti and P. trichosuri demonstrated that there is a heritable component to this switch such that some isolates/lines are much more prone to heterogonic development than others (Viney et al. Reference Viney, Matthews and Walliker1992; Stasiuk et al. Reference Stasiuk, Scott and Grant2012). It was also shown that lines with more or less heterogonic development can be selected from a genetically heterogeneous population (Viney, Reference Viney1996; Guo et al. Reference Guo, Chang, Dieterich and Streit2015). However, environmental factors, in particular the immune status of the host, the population density and the temperature, also influence the switch (e.g. Viney, Reference Viney1996; Harvey et al. Reference Harvey, Gemmill, Read and Viney2000; Nolan et al. Reference Nolan, Brenes, Ashton, Zhu, Forbes, Boston and Schad2004; Minato et al. Reference Minato, Kimura, Shintoku and Uga2008; Stasiuk et al. Reference Stasiuk, Scott and Grant2012; Sakamoto and Uga, Reference Sakamoto and Uga2013; for more, older, references see Streit, Reference Streit2008). This inherent temperature dependence allowed identification of the late L1 early L2 stage as the time point when the decision is made by temperature shift experiments (Premvati, 1958a ; Arizono, Reference Arizono1976; Nwaorgu, Reference Nwaorgu1983; Viney, Reference Viney1996; Minato et al. Reference Minato, Kimura, Shintoku and Uga2008).

The homogonic – heterogonic switch is believed to be evolutionarily related to the switch between the formation of fast developing L3s and dauer larvae in C. elegans (this so-called dauer hypothesis for the evolution of parasitism (Crook, Reference Crook2014) is discussed in more detail elsewhere in this special issue). Therefore, several studies used candidate approaches based on previous knowledge from C. elegans dauer formation and exit. A first approach was characterizing the structure and expression patterns of genes whose C. elegans homologues are known to control the dauer switch, and asking if the findings in Strongyloides are consistent with a similar role of the gene in the homogonic–heterogonic switch (Crook et al. Reference Crook, Thompson, Grant and Viney2005; Massey et al. Reference Massey, Castelletto, Bhopale, Schad and Lok2005, Reference Massey, Bhopale, Li, Castelletto and Lok2006, Reference Massey, Ranjit, Stoltzfus and Lok2013; Hu et al. Reference Hu, Lok, Ranjit, Massey, Sternberg and Gasser2010; Stoltzfus et al. Reference Stoltzfus, Massey, Nolan, Griffith and Lok2012a , Reference Stoltzfus, Minot, Berriman, Nolan and Lok b ).

Several attempts were made to further investigate the regulatory machinery controlling the homogonic–heterogonic switch at a more functional level. At the heart of the dauer switch is the nuclear hormone receptor DAF-12 (Antebi et al. Reference Antebi, Yeh, Tait, Hedgecock and Riddle2000). DAF-12 when free of ligand promotes dauer formation. When the ligand, dafachronic acid (DA) (Motola et al. Reference Motola, Cummins, Rottiers, Sharma, Li, Li, Suino-Powell, Xu, Auchus, Antebi and Mangelsdorf2006) is made no dauer larvae are formed and dauer formation can be prevented pharmacologically by the application of exogenous DA. Two groups demonstrated simultaneously and independently of each other that DA also prevents the formation of infective larvae in S. papillosus and S. stercoralis, respectively (Ogawa et al. Reference Ogawa, Streit, Antebi and Sommer2009; Wang et al. Reference Wang, Zhou, Motola, Gao, Suino-Powell, Conneely, Ogata, Sharma, Auchus, Lok, Hawdon, Kliewer, Xu and Mangelsdorf2009), indicating that there is a conserved endocrine regulatory module that controls dauer formation in C. elegans and L3i formation in Strongyloides spp. However, although a clear daf-12 orthologue is present in Strogyloides spp. (Wang et al. Reference Wang, Zhou, Motola, Gao, Suino-Powell, Conneely, Ogata, Sharma, Auchus, Lok, Hawdon, Kliewer, Xu and Mangelsdorf2009) the demonstration that the pharmacological effect of DA is through DAF-12 is pending and it is not known yet if the natural ligand of Strongyloides DAF-12 is DA. In fact, it is not even known if DA exists in Strongyloides spp. (but see below).

Based on these findings and the extensive knowledge about the genetic control of dauer entry and exit in C. elegans the Lok lab reported in multiple publications an extensive characterization of the L3i formation and activation in S. stercoralis, combining all currently available genetic tools in Strongyloides including RNA expression studies, transgenes encoding reporter constructs and wild-type and mutant versions of proteins (for example GFP tagged non-phosphorylatable, phospho-mimiking or dominant negative derivatives of the forkhead transcription factor type O (FOXO) FKTF-1b, the orthologue of C. elegans DAF-16) and pathway activating and inhibiting chemicals [phosphatidylinositol-3 (PI3) kinase inhibitors, 8-bromo-cGMP, cytochrome P450 inhibitors and DA] (Castelletto et al. Reference Castelletto, Massey and Lok2009; Stoltzfus et al. Reference Stoltzfus, Massey, Nolan, Griffith and Lok2012a , Reference Stoltzfus, Minot, Berriman, Nolan and Lok b , Reference Stoltzfus, Bart and Lok2014; Massey et al. Reference Massey, Ranjit, Stoltzfus and Lok2013; Albarqi et al. Reference Albarqi, Stoltzfus, Pilgrim, Nolan, Wang, Kliewer, Mangelsdorf and Lok2016). Due to the lack of mutations along with the candidate approach based on C. elegans gene function, which will inherently miss Strongyloides specific factors, and the much smaller number of man-hours spent on Strongyloides research, the picture is not as clear as in C. elegans. Nevertheless, the results are most interesting and indicate that dauer/L3i formation and exit are, at least in part, controlled by the same players in S. stercoralis and in C. elegans. In particular, the most recent of these papers Albarqi et al. (Reference Albarqi, Stoltzfus, Pilgrim, Nolan, Wang, Kliewer, Mangelsdorf and Lok2016) demonstrated that inhibition of cytochrome P450 activity by ketoconazole, which in C. elegans prevents biosynthesis of DA, has the opposite effect of DA addition in S. stercoralis and is suppressible by DA administration. This strongly indicates that in S. stercoralis DA or a closely related steroid hormone is involved in the process. Interestingly, however, differences in the expression patterns of several genes as well as varying epistatic relationships between regulatory modules strongly indicated that there are substantial differences in the regulatory logics of the two species.

Conclusions and outlook

Over the years, how genetic information in Strongyloides spp. and P. trichosuri is passed from one generation to next has been elucidated in quite some detail (see above). Nevertheless, one has to remain open for the possibility that under certain circumstances or in particular species of Strongyloides alternatives from what has emerged as general rules for Strongyloides spp. are conceivable. Schad (Reference Schad and Grove1989), for example, has explicitly warned not to prematurely disregard the reports of parasitic males by Faust and Kreis (Kreis, Reference Kreis1932; Faust, Reference Faust1933).

The sequencing efforts over the last years have yielded a comprehensive catalogue of genes present in several species of Strongyloides and in P. trichosuri (Hunt et al. Reference Hunt, Tsai, Coghlan, Reid, Holroyd, Foth, Tracey, Cotton, Stanley, Beasley, Bennett, Brooks, Harsha, Kajitani, Kulkarni, Harbecke, Nagayasu, Nichol, Ogura, Quail, Randle, Ribeiro, Sanchez-Flores, Hayashi, Itoh, Denver, Grant, Stoltzfus, Lok and Murayama2016). In order to study the functions of these gens, methods to knock them out are highly desirable. No true success with this respect can be reported yet in Strongyloides spp. but there is hope. Over the last few years sequence-specific endonucleases such as Zn-finger nucleases, TALENs and the CRISPR/Cas9 system have been established for mutation induction and genome editing in various systems, among them the nematodes C. elegans, other species of Caenorhabditis and P. pacificus (Jinek et al. Reference Jinek, Chylinski, Fonfara, Hauer, Doudna and Charpentier2012; Wiedenheft et al. Reference Wiedenheft, Sternberg and Doudna2012; Lo et al. Reference Lo, Pickle, Lin, Ralston, Gurling, Schartner, Bian, Doudna and Meyer2013; Irion et al. Reference Irion, Krauss and Nüsslein-Volhard2014; Kim and Kim, Reference Kim and Kim2014; Sung et al. Reference Sung, Kim, Kim, Lee, Jeon, Jin, Choi, Ban, Ha, Kim, Lee and Kim2014; Waaijers and Boxem, Reference Waaijers and Boxem2014; Wei et al. Reference Wei, Shen, Chen, Shifman and Ellis2014; Witte et al. Reference Witte, Moreno, Rodelsperger, Kim, Kim, Streit and Sommer2015). In particular, the CRISPR/Cas9-based approach taken by (Cho et al. Reference Cho, Lee, Carroll and Kim2013) for C. elegans and (Witte et al. Reference Witte, Moreno, Rodelsperger, Kim, Kim, Streit and Sommer2015) for P. pacificus looks promising for Strongyloides/Parastrongyloides. In this approach, the components, namely the endonuclease Cas9 and a bipartite single guide RNA (sgRNA, one part recognizes a 20 bp target site by base-pairing and the other part binds Cas9) are synthesized and assembled into the active complex in vitro. The complex is then injected into the gonad of adult hermaphrodites (which in C. elegans and P. pacificus replace females). Contrary to the approaches taken by the other references mentioned above, which include expression of the RNA and/or Cas9 from transgenes or injected RNAs, this approach does not depend on the availability of promoters or untranslated RNA regions known to work efficiently in the gem line. The modified strategy for S. ratti could be as follows (Fig. 2). Inject the Cas9/sgRNA complex designed to recognize a particular gene into the gonads of free-living females. This is expected to introduce double-strand breaks at the recognition site in germ cells, some of which will be imperfectly repaired leading to progenies with small deletions/insertions. These mutations will usually only be present in one of the two copies of the gene such that carriers are phenotypically wild-type (assuming the mutation is recessive). The progeny of the injected mothers are then used to infect host animals and emerging larvae are first tested in batch by PCR and sequencing for the presence of mutations at the desired position. If such a mutation is present among the worms shed by a host individual, gravid adult free-living females are singled out and allowed them to reproduce. Once they have produced a number of progeny they are used for DNA preparation and tested for the presence of the mutation. Single infective larvae derived from heterozygous mutant mothers are then used to infect hosts and establish a culture of heterozygous mutant worms. Single worm infections with S. ratti are successful in roughly half of the attempts (Viney et al. Reference Viney, Matthews and Walliker1992). For Parastrongyloides the passage through the host is not necessary and the free-living progeny of the injected mothers can be tested directly after they produced a number of offspring, which can be used to secure the mutation.

Fig. 2. Proposed strategy for targeted gene knock out in S. ratti. This protocol is proposed based the one published for P. pacificus (Witte et al. Reference Witte, Moreno, Rodelsperger, Kim, Kim, Streit and Sommer2015) and has not yet been successfully used in Strongyloides spp. or Parastrongyloides spp. Further modifications, for example the co-injection of a DNA repair template in order to achieve a specific mutation (deletion, insertion or alteration) are possible. For details, see text.

Classical forward genetic approaches with the random introduction of mutations followed by screening for a phenotype of interest would also be most useful. This approach has several advantages. It does not rely on prior assumptions about which genes may be involved in the process of interest and it leads not only to loss of function mutations or alterations with already known consequences (e.g. dominant negatives) but also to hypomorphic (reduced function), hypermorphic (enhanced function) or neomorphic (new function) alleles, which can be highly informative, as is illustrated by one of the best-known mutations in the fruit fly. The phenotype of the mutation nasobemia in the gene antennapedia, which eventually led to the discovery of the conserved homeobox (McGinnis et al. Reference McGinnis, Levine, Hafen, Kuroiwa and Gehring1984; Gehring et al. Reference Gehring, Affolter and Burglin1994) is caused by the mis-expression of the gene (Schneuwly et al. Reference Schneuwly, Klemenz and Gehring1987). Such a mutation would not have been found in the context of a systematic gene knock out analysis. Protocols for the induction of mutations in S. ratti are available and mutant worms were isolated successfully (Viney et al. Reference Viney, Green, Brooks and Grant2002; Guo et al. Reference Guo, Chang, Dieterich and Streit2015). But, so far forward genetic studies in Strongyloides have been hampered by the formidable obstacles to identifying the mutated genes causing the phenotype. However, there is great hope that this will change in the near future. With the recent progress in sequencing technology, requiring less material and becoming more affordable, in the model nematodes C. elegans (Sarin et al. Reference Sarin, Prabhu, O'Meara, Pe'er and Hobert2008; Doitsidou et al. Reference Doitsidou, Poole, Sarin, Bigelow and Hobert2010) and P. pacificus (Ragsdale et al. Reference Ragsdale, Muller, Rodelsperger and Sommer2013) it has become possible to identify mutations by sequencing the genomes of mutant animals and comparing them with the wild-type. In particular, the approach by (Doitsidou et al. Reference Doitsidou, Poole, Sarin, Bigelow and Hobert2010) looks to be very promising at least for S. ratti where an excellent reference genome is now available (Hunt et al. Reference Hunt, Tsai, Coghlan, Reid, Holroyd, Foth, Tracey, Cotton, Stanley, Beasley, Bennett, Brooks, Harsha, Kajitani, Kulkarni, Harbecke, Nagayasu, Nichol, Ogura, Quail, Randle, Ribeiro, Sanchez-Flores, Hayashi, Itoh, Denver, Grant, Stoltzfus, Lok and Murayama2016). In this strategy (Fig. 3), the mutant line (parental line 1′, which is a mutagenized derivative of a line 1) is first crossed with a different strain (parental line 2) with several thousands of known sequence differences compared with line 1. Theoretically parental lines 1 and 1′ differ only at the positions that have been altered by the mutagenesis treatment. The resulting F₁ animals are all heterozygous at all different loci, including the locus of interest and therefore, assuming the mutation is recessive, phenotypically wild-type. The F₁ animals are then crossed among themselves. The resulting F₂ progeny is divided into two pools containing the mutant and all the phenotypically wild-type animals, respectively. From these pools DNA is isolated and quantitatively sequenced. Around the position of the mutation causing the phenotype of interest, all mutant animals carry only alleles derived from parental line 1′. At all positions not genetically linked the allele frequency for both alleles is expected to be 50%. Within the region that in mutants is all parental line 1′ derived, only very few positions will differ between parental line 1′ and line 1. These are the interesting candidates. In order to make this approach workable in S. ratti suitable parental lines 1 and 2 need to be established. Since the genomic sequence of the currently most commonly used standard laboratory isolate ED321 has been determined and published (Hunt et al. Reference Hunt, Tsai, Coghlan, Reid, Holroyd, Foth, Tracey, Cotton, Stanley, Beasley, Bennett, Brooks, Harsha, Kajitani, Kulkarni, Harbecke, Nagayasu, Nichol, Ogura, Quail, Randle, Ribeiro, Sanchez-Flores, Hayashi, Itoh, Denver, Grant, Stoltzfus, Lok and Murayama2016), this strain is a prime candidate for parental line 1. However, ED321 has been maintained in several laboratories for many years and the populations have accumulated rare alleles, which are undetectable by sequencing genomic DNA isolated from large numbers of worms (Guo et al. Reference Guo, Chang, Dieterich and Streit2015). Selection of individuals with the desired mutant phenotypes represents a very dramatic population bottleneck (the mutant population is derived from the one originally mutant individual and its mates). This will make visible all the rare alleles present in the founding individuals. These variants will appear as differences from the wild-type along with the mutations induced by the mutagen, thereby increasing the number of candidate mutations (Guo et al. Reference Guo, Chang, Dieterich and Streit2015). Creating two more strongly inbred laboratory strains should therefore be a priority for S. ratti geneticists.

Fig. 3. Proposed strategy for mapping and cloning of mutations in S. ratti. This protocol is proposed based on the one published for C. elegans (Doitsidou et al. Reference Doitsidou, Poole, Sarin, Bigelow and Hobert2010) and has not yet been successfully used in Strongyloides spp. or Parastrongyloides spp. It is assumed that the mutation of interest is recessive and worms with the mutation in the homozygous state are viable. Generating mutations and homozygous mutant lines require at least one host passage but is abbreviated as one step. Red and blue lines represent chromosomes derived from parental line 1 (red) or parental line 2 (blue). Parental line 1 is the one used for mutagenesis. Parental line 2 is a different isolate, whose genome sequence differs from that of parental line 1 at thousands of positions. The mutation to be mapped is indicated by a yellow asterisk. Only one of the two chromosomes not containing the mutation is shown. In the example the chromosome not shown is the X chromosome. The strategy also works for mutations on the X chromosome provided that hemizygous mutant males are fertile. Wt, wild-type.