Oscheius tipulae^*

Oscheius tipulae is a nematode of the same family as C. elegans (Rhabditidae), commonly found in soil. Its relationship to C. elegans is more distant than that of members of the Caenorhabditis genus, yet it is more closely related to C. elegans than Pristionchus pacificus and Brugia malayi are (Figure 1). Compared to C. elegans, Oscheius tipulae shares a common ancestor with the insect parasite Heterorhabditis and the vertebrate parasites Strongylida (Blaxter et al., 1998; see The phylogenetic relationships of Caenorhabditis and other rhabditids).

The Oscheius genus comprises two main subclades (Dolichura and Insectivora groups; see The phylogenetic relationships of Caenorhabditis and other rhabditids; Sudhaus, 1976; Sudhaus and Hooper, 1994; Figure 1). One distinctive character between these two clades appears to be the smaller size of the animals in the Dolichura clade (that includes Oscheius tipulae) (Flemming et al., 2000) (M.-A. F., unpublished). The Dolichura group comprises Oscheius guentheri, which is a species with a reduced posterior-gonadal arm (Sudhaus and Hooper, 1994), plus two subgroups, of which two representative species are Oscheius tipulae and O. dolichura, respectively (see The phylogenetic relationships of Caenorhabditis and other rhabditids). Historically, Oscheius dolichura (Schneider, 1866) was one of the first nematode species (before C. elegans) described to have an androdioceous mode of reproduction (Maupas, 1900; see English translation: http://www.wormbase.org/papers/1900-maupas/index.html).

Oscheius tipulae was first described by Lam and Webster (Lam and Webster, 1971). The species was brought to persisting live culture (strain SB128) and redescribed by Sudhaus (1993). It was called O. tipulae because these researchers found it associated with the larvae of the insect Tipula paludosa (a dipteran). The CEW1 strain was found by C.E. Winter in soil in São Paulo, Brazil around 1992 (Winter, 1992). Because of taxonomic confusions, the CEW1 strain has been variably named Dolichorhabditis sp. or D. brevesophaga or Oscheius sp. 1 before it was decided to be an isolate of Oscheius tipulae (Dichtel-Danjoy and Félix, 2004), which is the valid name.

O. tipulae and its two closely related species, provisionally called Oscheius spp. 2 and 3 (Figure 1), are hard to distinguish morphologically, but can be distinguished through mating experiments and molecular features (Félix et al., 2001).

Oscheius tipulae is one of the most common nematode species in soil of different (non-desert) parts of the world. It is both widespread and abundant in soil samples. It is mostly (or exclusively) found in the dauer stage (like C. elegans), raising the question of whether it proliferates in soil. It was originally isolated from Tipula larvae, but this may not be a particularly specific association. O. tipulae is more heat-resistant in the lab than O. spp. 2 and 3 (Félix et al., 2001) and has been found in tropical as well as temperate regions.

Oscheius tipulae is easy to isolate (see Isolation of C. elegans and related nematodes). Species of the Oscheius genus are characterized in particular by their long rectum (see Figure 5 in Isolation of C. elegans and related nematodes).

Available wild isolates can be found at the CGC (http://elegans.swmed.edu/CGC/), the NYU Rhabditidae collection (http://www.nyu.edu/projects/fitch/WSRN/) or in the Félix lab collection (http://www2.ijm.jussieu.fr/worms).

AFLP (Amplified Fragment Length Polymorphism) analysis on 53 wild isolates from all over the world suggest a genetic diversity at least three-fold higher than that measured with the same method in C. elegans (Barrière and Félix, 2005), with no genetic differentiation between continents, yet a clear structure at a smaller scale between three local populations sampled in France (100 km scale), and a relatively high local diversity (D. Baïlle and M.-A. F., unpublished).

Oscheius tipulae reproduces through XX hermaphrodites and XO males like C. elegans.

The X chromosome non-disjunction rate of the reference strain CEW1 is lower than that of C. elegans N2 and males are not easy to induce by heat-shock. However, other O. tipulae strains such as PS959 have a higher proportion of males. Males reproduce less well at 25°C and above than at lower temperatures; therefore, crossing plates are better maintained at 23°C (or below).

Mutagenesis and genetic analysis are basically performed as in C. elegans (Brenner, 1974). Because the germ line divides later than that of C. elegans, mutagenesis is best performed on young adults (rather than L4-stage animals). The concentration of EMS used for C. elegans (50 mM) is not as efficient nor as lethal as in C. elegans; therefore, mutageneses were performed using an EMS concentration of 100–150 mM, or using TMP-UV irradiation.

Screens for visible mutants and for mutants with a defective vulva (see below, Section 6.) have been performed. 79 visible markers were isolated: 24 with a Dumpy or Small phenotype, 5 Rollers, 2 Longs, 1 Constipated and 47 Uncs (twitchers, kinkers, sluggish worms, coilers including one dominant allele; Figure 2). Slightly surprisingly, 35 out of 74 mutants are X-linked (especially 26/43 Uncs compared to 3/23 Dpys, M.-L. Dichtel and M.-A. F., unpublished). Mutants with a Him phenotype (recessive or dominant) were also isolated, as well as one mutant with a partially penetrant left-right inversion of the body (cov-7; Delattre and Félix, 2001c; Louvet-Vallée et al., 2003).

The apparent haploid number of chromosomes observed by DAPI staining of the germ line is six, as in C. elegans. G. Maro, M.-L. Dichtel and M.-A. Félix defined (by genetic complementation or linkage data; unpublished) 6 dpy loci, 7 unc loci and 2 lon loci. The unc, dpy, and lon loci are provisionally designated by a letter to avoid confusion with C. elegans loci. Rough linkage data gave six linkage groups, which may or may not cover the six chromosomes:

A few mutants are at CGC. Others can be found at http://www2.ijm.jussieu.fr/worms. The gene nomenclature is temporary, since these loci are not homologs to the corresponding C. elegans genes. Once their molecular identity is determined, the correct nomenclature is, for example, Oti-lin-39 for the Oscheius tipulae lin-39 ortholog.

Injection of morpholinos results in a partial inactivation of the corresponding gene product, for example Oti-lin-39 (Louvet-Vallée et al., 2003) or Oti-let-60 (Dichtel-Danjoy and Félix, 2004). However, as in many non-Caenorhabditis nematodes, RNA interference does not appear to work efficiently: injection of double-stranded RNA for various genes has not produced any phenotype.

DNA-mediated transformation has been unsuccessful so far, but recent studies in Strongyloides stercoralis suggest that using a 3′UTR from the species may be crucial for transgene expression (Li et al., 2006). This remains to be tested in Oscheius tipulae. Plasmids injected into the mother's female germ line have been detected by PCR in the F1 progeny (M.-L. Dichtel, unpublished).

Gonad development and morphology of O. tipulae are overall similar to those of C. elegans, except that germ line divisions are delayed. However, O. guentheri presents an interesting case of variable partial reduction of the posterior gonadal arm, which appears to result from a lesser activity of the posterior distal tip cell in its roles in arm migration and germ line proliferation (Félix and Sternberg, 1996).

Anchor cell determination is stochastic and unbiased in O. tipulae as in C. elegans, but biased towards Z4.aaa becoming the anchor cell in Oscheius guentheri (Félix and Sternberg, 1996; Sudhaus and Hooper, 1994).

Oscheius tipulae has been mostly studied for its vulva development (see Evolution of development in nematodes related to C. elegans). Several features make it a particularly interesting comparison to C. elegans. Its vulval cell lineage is different, and simpler than that of C. elegans. In particular, it is much easier to distinguish the 1° and 2° precursor cell fates by the number of division rounds they undergo (three and two, respectively) (Sommer and Sternberg, 1995; Figure 3). The vulval competence group is composed of P(4-8).p: P3.p is not competent (unlike in C. elegans). The adult vulva is formed of six superposed rings, instead of seven in C. elegans (Louvet-Vallée et al., 2003). Most strikingly, anchor cell ablations demonstrate that the anchor cell is required after P6.p division for its daughters to acquire a 1° fate: if the anchor cell is ablated in the mid-L3 stage, P(5-7).p all adopt a 2° lineage pattern. The same spatial pattern of fates is thus induced by different mechanisms in O. tipulae and C. elegans (Félix and Sternberg, 1997), and recent results suggest that the two-step induction mechanism found in O. tipulae is ancestral (Kiontke et al., in preparation).

O. tipulae has been used for extensive genetic screens for Egl (Egg-laying defective) and Pvl (Protruding vulva) mutants (Dichtel-Danjoy and Félix, 2004; Sommer, 2000). Those with abnormal vulva development were divided in three categories corresponding to successive steps in vulval development:

Interestingly, the spectrum of phenotypes obtained in vulva mutant screens in O. tipulae strongly differs from that obtained in C. elegans. For example, the Dov mutations have no equivalent in C. elegans, or are at least not found at the same rate. In contrast, Hypoinduced phenotypes have proven difficult to isolate in O. tipulae. Also, the 1°/2° fate pattern of Hyperinduced mutants differ between the two species. This comparison of parallel mutagenesis screens on a homologous developmental system shows that the spectrum of phenotypes that can be reached by random mutagenesis evolves (Dichtel-Danjoy and Félix, 2004).

A particularly interesting phenotype of Oscheius species is the number of division rounds of P4.p and P8.p: zero in Oscheius guentheri, one in O. dolichura and O. dolichuroides (as in C. elegans), but usually two in O. tipulae, spp. 2 and 3. However, whereas P4.p and P8.p divide twice in the CEW1 reference strain of O. tipulae, this trait is highly polymorphic within an isogenic strain or between wild isolates of O. tipulae (Delattre and Félix, 2001a; Delattre and Félix, 2001b). From the analysis of recombinant inbred lines between the two isolates CEW1 and PS959, the difference between them (two versus one division round) is caused by variations at several segregating loci (Delattre and Félix, 2001a). P(4,8).p division is a trait that is easily mimicked by mutation in CEW1, as shown by the isolation of several mutants with one or zero division round of these cells (Dichtel et al., 2001).