Breaching the Weismann barrier – a role of vitellogenin yolk proteins in nematode non-Mendelian inheritance

1 Introduction

August Weismann (1834-1914) is considered one of the most important biologists of the 19^th century because of his leading role in arguing for the acceptance of Charles Darwin´s theory of evolution. As pointed out first by Ernst Mayr and more recently by the historian Frederick B. Churchill in a magisterial Weismann biography, the Freiburg-based biologist convincingly argued i) for the continuity of the germ plasm within and between generations and ii) strongly against the inheritance of acquired characters (Churchill, 2015; Mayr, 1985). These ideas and many associated thoughts were formulated in dozens of contributions, but they were most strongly articulated in his 1892 monograph ‘Das Keimplasma: Eine Theorie der Vererbung – The germ-plasm: a theory of heredity’, and repeated in his final major opus ‘Vorträge über Deszendenztheorie’ published in three editions in 1902, 1904 and 1913 and translated into English by Arthur and Margaret Thomson under the title ‘The Evolution Theory’ (Weismann, 1892, 1904; Weismann et al., 1904). At a time, when neo-Lamarckism and vitalism dominated biological thought, Weismann has famously argued against the inheritance of acquired characters by building on his contemporary research on insect development, seasonal polyphenisms in butterflies (he used the term ‘polymorphism’ at the time), and germ cell and germ plasm development in hydromedusae. This huge body of work resulted in an integrative perspective – unique at its time – of development, heredity and evolution, which is also the title of Churchill´s 2015 biography.

Perhaps the most important contribution is Weismann´s argument for the continuity of the germ plasm, which is often referred to as the ‘Weismann barrier’. This concept is still central to evolutionary thought and modern evolutionary theory and gives little to no room for heritable information flow between generations other than that encoded in the DNA sequence. Nonetheless, many findings in modern biology and medicine strongly suggest environmental influence on trait inheritance and multiple studies in a diversity of model organisms from Arabidopsis to Caenorhabditis, Drosophila and mice indicate the widespread occurrence of long-lasting epigenetic effects that are not reset in the zygote (for recent reviews see, Baugh and Day, 2020; Chen and Rechavi, 2022; Frolows and Ashe, 2021; Perez and Lehner, 2019a; Santilli and Boskovic, 2023; Sengupta et al., 2023). Such ‘non-genetic or non-Mendelian inheritance’ is subject to intense molecular investigations but also a focus of modern philosophical debates (Bline et al., 2020).

In this review, we will consider recent findings that allow a novel and different view on the Weismann barrier. Focusing on the nematode model organisms Caenorhabditis elegans and Pristionchus pacificus, we will review studies on intergenerational and transgenerational epigenetic inheritance and a potential role of vitellogenin (yolk) proteins in memory transmission. The long-known yolk proteins represent a major exception from the Weismann barrier, although this aspect has only rarely been considered. We will discuss findings that suggest transgenerational inheritance and memory to rely on vitellogenin and will speculate on molecular mechanisms that might allow breaching the Weismann barrier.

2 An overview of epigenetic inheritance

Organisms are always threatened by environmental fluctuations and unpredictable perturbations. In all domains of life, they have evolved a nearly indefinite number of response mechanisms to adapt to such environmental changes. These responses can be physiological, morphological and behavioural in nature. In animals, these responses are usually reset in gametes and therefore, do not cross the generation barrier. In contrast, in haplodiplobionts, the maternal sporophyte influences seed development resulting in more complex interactions and parental effects (Dilkes and Comai, 2004). An increasing body of literature, largely from the last two decades, does however indicate that memory of the past can be transmitted for generations. Responses to diet, various stress conditions or odours are seen to be transmitted across generations with many examples from all major model organisms, i.e. Arabidopsis, Drosophila, C. elegans, mice and others. The memory of hunger after human famines are considered similar responses although the many confounding effects in humans make investigations substantially more complex than in model organisms (González-Rodríguez et al., 2023). In general, such phenomena are referred to as transgenerational epigenetic inheritance and are defined as ‘non-Mendelian inheritance encoded by epigenetic change’. There is consensus to distinguish three types of epigenetic inheritance: Parental and intergenerational effects describe transmission from individuals exposed to a certain stimulus to their immediate offspring (Frolows and Ashe, 2021; Sengupta et al., 2023). This includes nutritional provisioning, for example in utero in mammals. Mammals are special because a female foetus in utero already contains the germ cells of the next generation. In contrast to parental and intergenerational effects, ‘transgenerational epigenetic inheritance’ (TEI) lasts for at least three generations beyond the direct trigger or stimulus (Baugh and Day, 2020). Note that for such examples of TEI only the parental generation is exposed to the stimulus while the effect is seen for at least three further generations. Unfortunately, the distinct categories described above are not always properly distinguished. Below, we use the term ‘non-Mendelian’ inheritance for intergenerational and transgenerational but not for maternal effects. We refer the reader to a detailed distinction of these classes of inheritance (Sengupta et al., 2023). In addition, as many reviews on TEI have been published in the last years, we refer the interested reader to these more detailed overviews of TEI and its various facets in different model organisms (Chen and Rechavi, 2022; Frolows and Ashe, 2021; Perez and Lehner, 2019a; Santilli and Boskovic, 2023). In the following, we will briefly summarize TEI research in the roundworm C. elegans.

2.1 The power of C. elegans

The self-fertilizing nematode C. elegans has multiple advantages for studying TEI. First, its generation time is short (3,5 days at 20 °C) (Figure 1A). Second, worms can be fed with a monoxenic diet, usually E. coli OP50, resulting in a simple, controlled husbandry that can be easily changed to other food sources (see examples below). Finally, self-fertilization results in isogenic cultures that simplify the analysis of TEI in comparison to outbreeding species with genetic variation in culture stocks. Not surprisingly therefore, research on C. elegans has been at the forefront of TEI investigations (Figure 1B). Early examples in C. elegans established stable, heritable transgenerational effects in the context of transgenes and experimentally induced RNA interference (RNAi) (Kelly and Fire, 1998; Vastenhouw et al., 2006), yet these paradigms involve transgene silencing, which represents laboratory-based conditions most remote from nature (Frolows and Ashe, 2021). Building on this foundation, subsequent studies have shifted toward more “natural” triggers that resemble ecologically relevant experiences. Odor-related stimuli can cause behavioural effects transmitted for multiple generations (Remy, 2010). Starvation was shown to induce transgenerational memory of small RNAs (Rechavi et al., 2014). Temperature can induce gene expression changes for at least 14 generations (Klosin et al., 2017). Also, worms can heritably adapt to pathogen infections (Burton et al., 2020). C. elegans exhibits systemic stress resistance, which is enhanced through intestine-to-germ line communication in an intergenerational manner (Nono et al., 2020). Finally, the avoidance behaviour of C. elegans to the pathogenic bacterium Pseudomonas aeruginosa PA14 was also shown to be transgenerationally inherited (Moore et al., 2019; Kaletsky et al., 2020), although these phenotypes are controversially discussed (Akinosho et al., 2025; Gainey et al., 2025). The above-mentioned list of examples of TEI in C. elegans is not exhaustive and we again refer the interested reader to more in-depth reviews indicated above.

The success of TEI research in C. elegans and other model organisms is largely due to the elucidation of associated molecular mechanisms (Figure 1C). In general, three modes of epigenetic inheritance have been identified. First, DNA methylation is a prominent epigenetic mark that plays an important role in mammals and plants (for recent review see Santilli and Boskovic, 2023). However, DNA methylation does not exist in C. elegans and most other nematodes, thereby simplifying the analysis of epigenetic processes in these organisms. The presence of DNA modifications in nematodes has only been established in phylogenetically basal taxa (Clade 1), and remains debated (Gao et al., 2012; Schiffer et al., 2013). Second, histone modifications are dominant activators and repressors of gene expression. A large body of literature has shown a key role for histone modifications at various sites, such as histone 3 lysin 4 (H3K4), H3K9, H3K27 and others to be involved in TEI (Fitz-James and Cavalli, 2022). Finally, small RNAs (sRNAs) represent another layer of gene regulation and silencing. Three types of sRNAs are usually distinguished, i.e. piwi-interacting (pi) RNAs, microRNAs (miRNAs) and various small-interfering (si) RNAs that can be further subdivided into various groups (Carthew and Sontheimer, 2009). Both, histone modifications and sRNA signalling were shown to be involved in TEI in C. elegans in response to a variety of stimuli (Greer et al., 2011; Rechavi et al., 2014; Schott et al., 2014; Jobson et al., 2015). Therefore, the identification of multiple associated molecular mechanisms has placed this nematode at the forefront of TEI research.

2.2 More worms: Pristionchus pacificus mouth-form plasticity

While highly successful in elucidating molecular principles associated with TEI, research in C. elegans often relies on transgene silencing, a laboratory-based condition that is artificial in nature. Therefore, the adaptive value of TEI, although largely taken for granted, remains often unclear (Baugh and Day, 2020).

One way to overcome these valid concerns is through the study of natural readouts of environmental influences. The most widespread and spectacular organismal readout of the environment is developmental (phenotypic) plasticity, the ability of a certain genotype to produce different phenotypes upon environmental changes (Sommer, 2020). Three monographs in the late 1990ies and early 2000s have laid the foundation for much of the ongoing studies on developmental plasticity in plants, vertebrates, insects and nematodes (Schlichting and Pigliucci, 1998; Pigliucci, 2001; West-Eberhard, 2003). Indeed, one nematode species that shares with C. elegans all of its technical advantages (short generation time, easy husbandry, self-fertilization of hermaphroditic animals), has become a new model system for research on developmental plasticity and associated traits (Schroeder, 2021). Pristionchus pacificus is a member of a distinct nematode family and is thought to be separated from C. elegans for around 250 million years (Qing et al., 2025). This species and its relatives (more than 50 known Pristionchus species from around the world), are most reliably found in association with scarab beetles (Herrmann et al., 2006; Kanzaki et al., 2021). Pristionchus nematodes remain in the arrested dauer stage as long as the beetle is alive and only when the insect dies in the soil, nematodes exit the dauer stage to feed on the developing microbes on the insect´s carcass (Renahan et al., 2021) (Figure 2A). The natural history of Pristionchus species, which is of interest in its own right, have recently been summarized and we again refer the reader to this more detailed account (Sommer, 2025).

Isogenic P. pacificus cultures contain two alternative mouth forms, resulting in distinct feeding strategies and feeding behaviours (Figure 2B). During post-embryonic development, individual worms have to decide to express one of two alternative morphs in an irreversible manner (Bento et al., 2010). Stenostomatous (St) animals have a single tooth and a narrow buccal cavity resulting in strict bacterial feeding, similar to C. elegans and many other free-living nematodes. In contrast, eurystomatous (Eu) animals form two teeth and a broad buccal cavity. Such Eu animals are omnivores, i.e. while they can still feed on bacteria like St individuals, they can also consume fungi, or kill other nematodes to supplement their bacterial diet. This mouth form dimorphism is discrete without intermediate forms (Theska et al., 2020), resulting in a polyphenism that provides many advantages for phenotypic, genetic and molecular analyses. Three aspects of the P. pacificus mouth-form polyphenism are crucial. First, the predatory behavior of P. pacificus adults against other nematodes on the insect carcass is an example of intraguild predation that also serves to remove competitors for bacterial food (Quach and Chalasani, 2020). Second, the mouth-form decision of individuals is to a certain extend stochastic in nature and is best described as a bistable epigenetic switch, similar to the lytic and lysogenic cycles of bacteriophage lambda (Kalirad and Sommer, 2024a). Specifically, each natural isolate shows a specific Eu: St mouth-form ratio under fixed laboratory culture conditions (Dardiry et al., 2023). For example, the wild type reference strain P. pacificus PS312 from Pasadena, California has a 90Eu:10%St ratio. Importantly, this bistable switch can easily be manipulated by genetic or environmental perturbations, both of which were essential for the elucidation of the gene regulatory network controlling the mouth-form decision (Ragsdale et al., 2013; Sieriebriennikov et al., 2020; Casasa et al., 2023; Levis and Ragsdale, 2023, 2025). Finally, in an ecological context, this mouth-form polyphenism is best characterized as a bet-hedging strategy given that the duration and longevity of an insect carcass, and thus food availability, cannot be easily predicted by any given nematode at any given time (Renahan and Sommer, 2021; Renahan et al., 2021). Indeed, ongoing modelling studies try to elucidate the interplay between plasticity and stochasticity on coexistence (Kalirad and Sommer, 2024b). Together, the natural readout of P. pacificus mouth-form development in a study system with its many technical simplicities allows feeding plasticity to be studied at the molecular and genetic level. Many associated investigations aim to elucidate the neurobiology of predation (Cook et al., 2025; Eren et al., 2026) and the ecological and evolutionary significance of mouth-form plasticity (Kalirad and Sommer, 2024b). Below, we summarize recent findings that mouth-form plasticity shows an element of TEI.

2.3 Mouth-form plasticity is transgenerationally inherited

A recent study in P. pacificus introduced long-term environmental induction experiments through switching of the bacterial diet with later food reversal (Quiobe et al., 2025). This work used a wild isolate of P. pacificus (RSC011) that is preferentially non-predatory under standard E. coli OP50 growth conditions and has a mouth-form ratio of 30%Eu:70%St (Figure 2C). For this experiment, the progeny of a single mother was exposed to a Novosphingobium L76 diet, a bacterium previously isolated from a Pristionchus environment (Akduman et al., 2018). Exposure to Novosphingobium resulted in an immediate mouth form switch towards 100% Eu and thus, predatory animals. This response was not only immediate (in the first generation), but also systemic (in all lines) and permanent. Most interestingly, food reversal back to the standard E. coli diet revealed transgenerational memory of the predatory mouth form (Quiobe et al., 2025) (Figure 2D). For example, when worms were reverted back to E. coli after 15 generations on Novosphingobium, a high probability of expressing the predatory mouth form remained for four generations. Further studies revealed a correlation between the length of the dietary induction to Novosphingobium and the length of the TEI of the predatory mouth form. Different from many examples of TEI in C. elegans, the TEI of the predatory mouth form needs a minimal exposure to Novosphingobium of five generations (Quiobe et al., 2025) (Figure 2D). This example shows the potential power of developmental plasticity with its natural phenotypic readout to environmental influences for research on TEI. Together with the simplicity of dietary switching under laboratory conditions, the P. pacificus study system can be used to identify factors necessary for TEI of the predatory morph.

Indeed, genetic studies identified an essential role of the E3 ubiquitin ligase ebax-1 with Ppa-ebax-1 mutants being transgenerational inheritance-defective (tid) (Quiobe et al., 2025) (Figure 2E). EBAX-1 in P. pacificus and C. elegans, similar to Drosophila Dora and mouse ZSWIM8 are involved in target-directed miRNA degradation (TDMD), a process that reverts the logic of standard miRNA-mediated gene silencing (Han et al., 2020; Shi et al., 2020). Upon the binding of an AGO-miRNA complex to a trigger RNA, the ZSWIM8/EBAX-1 ligase is recruited and initiates the destruction of the AGO-miRNA complex through the ubiquitination of the AGO protein (Farnung et al., 2026). For the TEI of the P. pacificus predatory mouth form, a clustered miRNA locus with around 90 miRNAs has been identified that contains more than 40 copies of two unrelated miRNAs, i.e. miR-2235/mi-R35 and miR-8286 (Quiobe et al., 2025). Deletion of these clustered miRNAs results in precocious and extended TEI of the predatory morph indicating that under wild type conditions the miRNA acts as an inhibitor of transgenerational memory. Thus, Ppa-ebax-1 mutants are memory-defective, whereas deletions of the miR-2235 – miR-8286 cluster have the opposite phenotype, precocious and extended TEI. This study represents the first example of a miRNA in TEI. The fact that both Ppa-ebax-1 mutants and deficiencies that eliminate the complete miRNA cluster are viable, allows the molecular mechanisms associated with transgenerational memory to be elucidated. In particular, analysis of the trigger(s) and target(s) of the miRNA can provide novel insight into TEI, as well as TDMD (for a general review on TDMD see, Buhagiar and Kleaveland, 2024).

2.4 A role of vitellogenin in transgenerational memory

Further studies revealed that nutritional effects of Novosphingobium initiate the TEI of the predatory mouth form (Quiobe et al., 2026). Novosphingobium is a strong vitamin B12 producer, whereas the standard nematode food source E. coli OP50 is not (Bito et al., 2013; Watson et al., 2014; Akduman et al., 2020). Indeed, vitamin B12 supplementation of an E. coli diet can induce the predatory mouth form and its subsequent TEI. This and additional experiments indicated that vitamin B12 is necessary and sufficient for transgenerational memory (Quiobe et al., 2026). Transcriptomic studies revealed only a minimal overlap on P. pacificus gene expression changes upon Novosphingobium exposure and vitamin B12 supplementation, but identified vitellogenin yolk proteins as a potential candidate (Figure 2D). Studying vitellogenin yolk proteins is complicated by the fact that vitellogenin genes are usually amplified into larger non-clustered families, i.e. P. pacificus encodes nine vitellogenin genes in its genome (see below). Instead, most studies use manipulations of vitellogenin uptake receptors, which are present as single copy genes in nematodes and many other animals. In C. elegans, the receptor-mediated endocytosis locus 2 (rme-2) was identified as the sole receptor required for the uptake of vitellogenin into the gonad (Grant and Hirsh, 1999). The P. pacificus genome encodes a single Ppa-rme-2 locus and mutations in this gene result in the absence of transgenerational memory after Novosphingobium induction (Quiobe et al., 2026). While a knockout of all nine vitellogenin genes that are all located on the P. pacificus X chromosome has not yet been generated, we hypothesize that vitamin B12, supplied by a bacterial donor or through supplementation, induces stronger vitellogenin expression in the intestine (see below). Vitellogenin is specifically taken up by the germ line through the RME-2 receptor. In Ppa-rme-2 mutants, the TEI of the predatory mouth-form after Novosphingobium induction is no longer observed. Thus, vitamin B12 induces memory of predation through vitellogenin provisioning.

3 Vitellogenin genes and vitellogenesis

With the findings described above, we next turn to vitellogenin yolk proteins and their role in animal development and a potential function for breaching the Weismann barrier. First, we briefly summarize vitellogenin biology in animals, and then focus on the C. elegans yolk proteins and how vitellogenin might function during transgenerational inheritance.

3.1 What is known about vitellogenin – a brief overview

Oviparous animals, invertebrates and non-mammalian vertebrates alike, rely on maternally provided egg yolk proteins to provide nutrients for the development and growth of the embryo. Most egg yolk proteins are phosphoglycoproteins that are simultaneously a source of amino acids, phosphate and carbohydrates but also transport lipids. Vitellogenin is one of the major egg yolk proteins and is now known to be homologous in all egg-laying animals. Historically, however, yolk proteins were first characterized biochemically in a handful of classical model systems. In insects, these largely female-specific proteins were first found in the haemolymph of the silkmoth Hyalophora cecropia by Telfer in 1954 and shown to be involved in yolk formation (Telfer, 1954). The term vitellogenin itself was coined by Pan and co-workers only in 1969 (Pan et al., 1969). At that time, it was not clear that major yolk proteins would be homologous among major taxa. However, a rapidly growing body of evidence revealed that vitellogenin proteins belong to the broader large lipid transfer protein superfamily (Babin et al., 1999) and are indeed homologous between different insects and in other animals (Chen et al., 1997; Spieth et al., 1991; Tufail and Takeda, 2008). Many studies of vitellogenin yolk proteins and vitellogenin-encoding genes were performed between the 1970s and 1990s with an increasing involvement of tools from molecular biology. We refer the interested reader to detailed accounts on vitellogenin proteins and vitellogenesis for further information (Raikhel et al., 2005; Tufail and Takeda, 2008).

From the early days of vitellogenin research, it was known that these proteins are synthesized in somatic tissues before being transported into the oocyte (for review, see Byrne et al., 1989). In vertebrates, from fish to chicken, vitellogenin is synthesized in the liver. In contrast, in insects vitellogenin synthesis occurs in the fat body. Echinoderms seem to be an exception with vitellogenin being produced in the intestine, ovaries and testes. Finally, in nematodes, vitellogenin is synthesized in the intestine. Thus, in the majority of animals, vitellogenin proteins are synthesized in specific somatic tissues and then cross the Weismann barrier to be maternally provided to the oocyte. Interestingly, the soma – germ line transport has been noted since the early days of vitellogenin research in the 1950s and 1960s, but these findings have not been discussed in the context of the Weismann barrier. To the best of our knowledge, the fact that vitellogenesis and oocyte provisioning represents a massive transport of material from the soma to the germ line in nearly all oviparous animals has been largely overlooked, except for the extensive review of Perez and Lehner on vitellogenesis in C. elegans and a recent experimental study in C. elegans to be discussed in more detail below (Perez and Lehner, 2019a, 2019b; Aupérin et al., 2026).

3.2 The C. elegans vitellogenin gene set - an evolutionary outlier

Vitellogenin biology has been intensively studied in C. elegans for decades. Perez and Lehner (2019b) provided a detailed review on the subject, covering all areas of vitellogenin biology. While we refer the reader to this comprehensive account, we summarize below the most important aspects of vitellogenin biology, in particular in the context of a potential involvement in non-Mendelian inheritance.

3.2.1 Physiology

For invertebrate females of species with a large brood size, oocyte nutrient provisioning is the major challenge of adult individuals. In C. elegans hermaphrodites grown under standard laboratory conditions, embryo production within one day is equivalent to the body weight of the individual (Hirsh et al., 1976). Already early studies estimated that yolk comprises nearly 40% of the total amount of proteins in embryos (Kimble and Sharrock, 1983). Consequently, later gene expression studies revealed that vitellogenin genes are the most abundantly transcribed genes in nematodes. It is a remarkable fact that around 3% of all RNA-seq reads are from vitellogenin genes, although they are expressed only in the 20 intestinal cells, which are however, highly endo-reduplicated (Van Nostrand et al., 2013). Specifically, endomitosis in intestinal cells generate polyploid, binucleated cells, and specific inhibition of binucleation without changing total ploidy revealed that multinucleation per se is critical for proper intestinal gene expression (van Rijnberk et al., 2022). In particular, binucleation facilitates the rapid up-regulation of intestine-specific genes at the larval-to-adult transition, including vitellogenins, whose reduced expression in mononucleated intestinal cells causes developmental delays and reduced fitness (van Rijnberk et al., 2022). These studies support the notion that vitellogenesis represents the major metabolic cost of adult female and hermaphrodite worms. Rarely considered in molecular biology (in contrast to ecological research), these figures indicate that vitellogenesis might be at its biosynthetic capacity and upper limits. If so, and to what extent, this biosynthetic capacity is involved in regulating the total brood size of nematodes and thus, influencing fitness, has not been investigated. However, several studies in C. elegans have indicated that other intestinally-expressed genes are upregulated upon vitellogenin RNAi or in unc-62 mutants, the master regulator of vitellogenin gene expression (Van Nostrand et al., 2013; Sornda et al., 2019).

3.2.2 Biochemistry

Pioneering work by Sharrock and co-workers has characterized vitellogenin proteins in C. elegans. Four distinct yolk proteins were found: two larger polypeptides named YP170A and YP170B, and two smaller polypeptides named YP115 and YP88 (Sharrock, 1983, 1984; Sharrock et al., 1990) (Figure 3A). Note that these studies were done before the molecular biology of vitellogenin genes was known. Also, it was unknown at the time that C. elegans would turn out to represent the major ‘evolutionary outlier’, because its genetic composition of vitellogenin genes is unique and evolutionarily highly derived.

3.2.3 Genes and phylogeny

Molecular studies revealed that there are six vitellogenin genes in the C. elegans genome, named vit-1 to vit-6. Of these, vit-2 is the best studied and a vit-2 reporter construct is used in many contemporary studies (i.e (Erdmann et al., 2024). Also, vit genes were a prime target for studying the intestinal transcription factors ELT-2 and provided an easy molecular readout (Goszczynski et al., 2016). Further studies revealed that vit-1 and vit-2 encode for the YP170B polypeptide, whereas vit-3 to vit-5, more than 95% similar in sequence, encode for YP170A (Blumenthal et al., 1984). The high sequence similarity among vit-1-vit-5, together with their clustered organization on the X chromosome, strongly suggests that these genes arose through relatively recent tandem duplication events from an ancestral vitellogenin gene (Spieth et al., 1985; Perez and Lehner, 2019b).

In contrast, vit-6 is the single most divergent member of the C. elegans vit family (Spieth et al., 1985) (Figure 3B). While vit-6 encodes for a polypeptide of around 180 kDa, similar to the other Cel-vit genes, it is post-translationally cleaved resulting in the two shorter polypeptides originally observed by Sharrock (1984). YP115 is derived from the C-terminal portion of VIT-6, whereas YP88 results from an N-terminal fragment. Final biochemical characterization revealed that these separate polypeptides are nonetheless, covalently bound to YP170A by disulfide bonds (Sharrock et al., 1990). Thus, VIT-6 combines features of the canonical C. elegans vitellogenins (similar precursor size, yolk association) with a cleavage pattern that resembles vitellogenins in many other animals (Tufail and Takeda, 2008). Importantly, Cel-vit-6 is also an outlier in its genomic context, as it is the only C. elegans vit gene that is not located on the X chromosome (Figure 3A).

Only through the comparative analysis of animal whole genome sequencing projects and the in-depth analysis of vitellogenin genes in other nematodes species, it became clear that C. elegans and its relatives with their vitellogenin gene arrangement represent a major outlier in the animal kingdom. Already early studies revealed that vitellogenin genes from cnidarians to birds are most similar to Cel-vit-6, which itself represents the most distant vitellogenin gene in C. elegans. Unfortunately, however, Cel-vit-6 is largely understudied. Work by the laboratory of C. Winther in Oscheius tipulae, a closely related species of rhabditid nematodes, indicated the presence of only two vitellogenin genes, most similar to C. elegans vit-1 and vit-6 (Almenara et al., 2013). Interestingly, the Oti-VIT-6 protein is again post-translationally cleaved, which also represents a common feature of non-nematode vitellogenin proteins.

Sequence comparison of vitellogenin genes in additional nematodes indicates that vit-6-like genes are evolutionarily conserved and that Cel-vit-1 to Cel-vit-5 resulted from recent evolutionary duplication events (Figures 3A, B). In P. pacificus, nine vitellogenin genes are present, all of which are most similar in sequence to Cel-vit-6 (Quiobe et al., 2026). This pattern suggests that P. pacificus independently expanded a vit-6-like ancestral gene. Interestingly, all P. pacificus vitellogenin genes are located on the X chromosome, while Cel-vit-6 is not. This contrast suggests that localization on the X chromosome might be important for vitellogenin gene expression in nematodes. The autosomal position of Cel-vit-6 suggests that it has secondarily escaped X-linked regulation, but why and how this relocation occurred, remain currently unknown. It is interesting to note that Cel-vit-6 has been reported to be involved in the protection against oxidative stress (Sornda et al., 2019), whereas another study suggested protection against bacterial pathogens through several vitellogenin proteins (Fischer et al., 2013). A broader comparison to beetles and moths revealed that vitellogenin expression can also rely on autosomal regulation (Yano et al., 1994; Sheng et al., 2011). While widespread autosomal retention in insects suggests that metabolic dosage constraints normally favour hormonal gating (Tufail and Takeda, 2008), the direct integration of Drosophila yolk proteins into the primary sex-determination cascade highlights how some lineages successfully transition to sex-specific chromosome-coupled regulation (Bownes, 1994). Together, the evolutionary diversity in vitellogenin gene organization and chromosomal context raises the question of how vitellogenin function is constrained, and whether these proteins are strictly required for embryonic development.

3.2.4 Vitellogenin proteins are dispensable for embryonic development

One of the most puzzling effects of nematode vitellogenin biology is that they are dispensable for embryonic development (Van Rompay et al., 2015; Ezcurra et al., 2018; Dowen, 2019). While it has long been assumed that vitellogenin is essential for adult fertility and embryogenesis, multiple mutant lines indicated that this is not so. Several mutants that largely abrogate vitellogenin expression are viable and have little effect on brood size and viability of offspring (Van Rompay et al., 2015). Similarly, the above-mentioned rme-2 mutant that prevents uptake of vitellogenin proteins into the oocyte, is viable. As already indicated by Grant and Hirsh, rme-2 null mutant mothers have a reduced brood size but around 80 produced progenies are viable without any obvious defects (Grant and Hirsh, 1999). These observations indicate that vitellogenin proteins and their uptake via RME-2 are important for efficient reproduction and maternal provisioning, but are not absolutely required for embryogenesis under standard laboratory conditions. Rather than functioning as indispensable determinants of embryonic development, the vitellogenin transport pathway primarily enhances offspring quality and resilience under stressful conditions without compromising embryonic viability (Chotard et al., 2011; Perez et al., 2017; Jordan et al., 2019; Sornda et al., 2019; Geens et al., 2023). This distinction is particularly important because it suggests that vitellogenins may have functions beyond simple nutritional support. In the context of transgenerational inheritance, their dispensability for development leaves the possibility that vitellogenins act as macromolecular vehicles for maternal signalling molecules or epigenetic information (Aupérin et al., 2026), rather than serving solely as nutritional yolk substrates.

3.2.5 A role of vitellogenin proteins in oogenesis

Another recent study indicated a function of vitellogenin in oocyte maturation through interactions with the conserved LMD-3/OXR1 protein (Zhai et al., 2025). In C. elegans, like in many other animals, reproductive capacity declines with hermaphrodite (female) age. This decline is increasingly linked to impaired proteostasis caused by oxidative stress. The LysM domain-containing protein Cel-LMD-3 was shown to regulate proteostasis in C. elegans germ cells with Cel-lmd-3 mutans exhibiting a reduced brood size and increased germ cell apoptosis. Interestingly, Cel-LMD-3 is localized in germ cell lysosomes where it interacts with vitellogenin yolk proteins and the proton-pumping V-type ATPase. This allows proper acidification of lysosomes, which in turn is crucial for yolk protein sorting and allows the lysosomal degradation of vitamin B12-binding proteins (Zhai et al., 2025). This study is in line with previous reports that vitellogenin accumulation in later adult stages leads to reproductive senescence by impairing lysosomal function (Tang et al., 2023; Zhai et al., 2025). While increased vitellogenin allocation enhances reproductive investment during early adulthood, continued vitellogenin production beyond the reproductive period becomes maladaptive. Zhai et al. (2022) demonstrated that aging drives extensive fusion and expansion of intestinal vitellogenin vesicles, leading to pathological yolk accumulation within the intestine and pseudocoelom, contributing to age-associated intestinal degeneration. In the germ line, accumulation of vitellogenin results in the impairment of lysosomes. If vitellogenin accumulation is inhibited, extended reproduction is observed, which requires lysosomal function (Tang et al., 2023). Together, these studies add to the recent expansions of vitellogenin functions in C. elegans and suggest pleiotropic functions during oogenesis and embryogenesis.

3.2.6 Expression in males

The regulation of vitellogenin gene expression in C. elegans has been intensively studied. The role of Insulin, TGF-ß, the TOR pathway and other factors has been reviewed in detail by Perez and Lehner (2019b). Besides tissue and stage-specific regulation, vitellogenin expression is also largely restricted to hermaphrodites. Surprisingly, however, vitellogenin expression has also been reported in males. Shi and co-workers demonstrated that mating affects vitellogenin gene expression in both hermaphrodites and males and results in shrinkage and death (Shi et al., 2017; Shi and Murphy, 2014). More broadly, these findings fit with the idea that vitellogenin production is tightly coupled to reproductive state and metabolic allocation, and that sex-specific regulation of yolk synthesis may be particularly sensitive to physiological stress and mating status. To our knowledge, no similar experiments have been performed in other nematode species.

3.2.7 Vitellogenin amino acid composition

Finally, we want to draw the attention of the reader to what might represent another aspect of vitellogenin biology that has so far been largely overlooked (Figure 3C). As indicated above, vitellogenesis biosynthesis capacity might be at its limit in C. elegans and similarly, other nematodes. One additional aspect to the pure expression level of vitellogenin proteins is their amino acid composition (Figure 3C). Vitellogenin yolk proteins are often huge with C. elegans, O. tipulae and P. pacificus vit genes encoding polypeptides of around 180 kDa, before potential cleavage. More detailed sequence analysis reveals that these proteins are enriched in repeated stretches of amino acids that are metabolically costly, making yolk production a particularly demanding investment for the mother. Such ‘expensive’ amino acids likely increase the nutritional burden of vitellogenesis, because their synthesis draws heavily on limited resources and their accumulation in yolk represents a substantial transfer of maternal biomass to the next generation. This consideration highlights that vitellogenesis is a high-output process, representing the major energetic investment and biosynthetic burden for adult nematode females and hermaphrodites. Similarly, many other small-scale invertebrates with large brood sizes face related challenges, where egg provisioning must balance reproductive output against maternal resource allocation. Therefore, although only rarely considered, vitellogenesis represents the major cost to adult females (hermaphrodites) that is important for both, ecology and evolution.

3.3 Nematode vitellogenin and its potential role in memory transmission

Two features of vitellogenin proteins and their biosynthesis make them candidates for non-Mendelian inheritance. First, being synthesized in the intestine, vitellogenin yolk proteins have to be transported to the germ line and thus, breach the Weismann barrier. Given the enormous amount of yolk protein synthesized during adulthood, the potential of vitellogenin proteins for information flow from the soma to the germ line should not be underestimated. Second, forming yolk complexes as glycoproteins with lipids, vitellogenin might carry other molecules during the transport from the intestine through the pseudocoelom to the germ line and finally into the oocyte. In at least three instances, vitellogenins have been considered for soma – germ line communication in C. elegans, all of which represent examples of intergenerational inheritance (Figure 4).

3.3.1 Maternal age effects

Perez and co-workers (2017) showed that vitellogenin provisioning of oocytes in C. elegans increases with maternal age. Young mothers (day 1 adult) provide less vitellogenin to oocytes than older mothers (day 2 and particular day 3 adults) (Figure 4A). Strikingly, these maternal age effects were shown to influence larval growth and development (Perez et al., 2017). Early hatching larvae (from young mothers) are smaller, slower in reaching adulthood and are more starvation sensitive when compared to larvae from older mothers that have been provisioned with more vitellogenin. Starvation sensitivity can be studied under laboratory conditions, and it has been shown that prolonged L1 starvation induces postembryonic developmental defects, including reduced fertility and sterility, with offspring from young mothers being particularly susceptible (Perez et al., 2017). However, if such effects are relevant in natural ecological conditions remains currently unknown. Ongoing studies investigate an intergenerational lipid memory from parent to offspring (Wilhelm et al., 2025). Importantly, however, all of these effects are intergenerational and are not seen in further generations without new stimuli.

3.3.2 Extracellular RNA transport and systemic RNAi

RNAi was first described in C. elegans in 1998 and represents a powerful mechanism of sequence-specific gene silencing (Fire et al., 1998). The introduction of double-stranded RNA (dsRNA), first shown in worms by injection, can trigger the degradation of homologous transcripts in multiple organisms (Fire et al., 1998; Hamilton and Baulcombe, 1999; Zamore et al., 2000). In C. elegans, RNAi is systemic, that is, dsRNA introduced by other methods, such as feeding or soaking, will spread through the organism causing gene silencing (Tabara et al., 1998; Timmons and Fire, 1998). For a broader discussion of soma-to-germline RNA communication across species, we refer readers to a more detailed account (Conine and Rando, 2022). RNAi can last for multiple generations and thus, represents a sub-type of TEI. For example, expression of dsRNA in neurons was shown to silence a gene of matching sequence in the germ line for at least 25 generations (Devanapally et al., 2015). Systemic silencing requires the dsRNA channel encoded by the gene sid-1 and was shown to involve dsRNA transfer from mothers to offspring (Winston et al., 2002). This study also showed that in addition to Cel-SID-1, other dsRNA transport systems exist in C. elegans.

Marré and co-workers described extracellular RNA transport in C. elegans. By injecting fluorescently labelled dsRNA into the pseudocoelom, dsRNA would eventually accumulate in embryos (Marré et al., 2016) (Figure 4B). This study claimed that this dsRNA transport is sid-1-independent and instead requires the vitellogenin germ line uptake receptor Cel-RME-2. However, further analysis revealed multiple independent paths of maternal dsRNAs to be transported into the oocyte (Wang and Hunter, 2017). Maternal and embryonic sid-1 function contribute independently to parental RNAi and maternal RME-2 is only required in the absence of sid-1. Wang and Hunter (2017) could also show that this parental RNAi phenotype represents a strictly intergenerational effect.

Finally, it should be noted that while RNAi in its various forms is conserved between some of the major model organisms, not all animals and plants are susceptible to RNAi. In P. pacificus for example, RNAi does not work using any method, injection, feeding or soaking. This might be due to the absence of a SID-1 ortholog in the P. pacificus genome (note that the Wormbase annotation for Ppa-sid-1 represents the paralogous Ppa-chup-1 gene). Nonetheless, RNAi by injection worked in P. pacificus for one gene, the master sex determination regulator tra-1 (Pires-daSilva and Sommer, 2004). It is currently unknown why tra-1 is amenable to RNAi whereas other genes are not, and what restricts the successful application of this method for other genes.

3.3.3 Soma-to-germ line miRNA inheritance through vitellogenin

The last example from C. elegans is based on a recent study that started from the assumption that somatically synthesized vitellogenin is a potential carrier for lipids and other molecules for transport into progeny (Figure 4B). Using fluorescently labelled VIT-2::GFP and tissue-specific depletion of the RME-2 receptor in oocytes using a degron inactivation system, vitellogenin yolk proteins were shown to carry specific classes of miRNAs (Aupérin et al., 2026). Interestingly, miRNAs transported by vitellogenin are core regulators of intestinal gene expression. Moreover, the transfer of miRNAs is modulated by maternal age and certain environmental stressors. Importantly, these findings are again limited to intergenerational gene regulation, similar to the other examples discussed above. Together, several independent evidences from C. elegans strongly support a role of vitellogenin in the transmission of non-genetic information from mother to offspring.

3.3.4 Evolutionary divergence of C. elegans and P. pacificus at multiple levels

Various aspects discussed above indicate the rapid evolutionary diversification of molecular processes involved in epigenetic inheritance. Vitellogenin plays a role in intergenerational and transgenerational inheritance in C. elegans and P. pacificus, respectively, but in completely different processes with unrelated phenotypes. Similarly, the composition of vitellogenin genes diverged substantially between nematode species with C. elegans exhibiting the most distinct pattern and sequence variation of vitellogenin genes in all animals. The miR-2235/miR-35 clusters also show different composition in P. pacificus vs. C. elegans and the role of these miRNAs in the regulation of TEI is restricted to P. pacificus. Finally, RNAi-related phenomena, a particular sub-type of TEI, are widespread in C. elegans but are largely absent in P. pacificus, which misses a key parental RNAi factor SID-1.

A major take home message therefore is that TEI is among the most rapidly evolving processes, which is in line with other epigenetic factors that were also shown to evolve extremely rapid. For example, P. pacificus has lost the histone methyltransferase complex PRC2 although the histone modifications normally produced through this complex (H3K27me3) are still present (Werner et al., 2018; Brown et al., 2024). Most strikingly in this context, the loss of the PRC2 complex is not correlated with a larger phylogenetic taxon (i.e. genus or family level), but instead occurred within the genus Pristionchus between species that can still hybridize (Brown et al., 2024). While all of these examples are experimentally proven, evolutionary explanations and studies aiming towards the elucidation of the underlying evolutionary mechanisms lack behind. For example, one outstanding question is to what extent these differences are the result of natural selection vs. drift, and to which extend the associated phenotypes and traits represent adaptations to the ecological niches of the organisms. While an adaptive status of any pattern seen in nature is often taken for granted (Milo, 2019), further research into the ecological significance of any transgenerationally inherited trait is necessary to provide actual proof (Baugh and Day, 2020). In this context, it might be important to distinguish ‘the arrival of the fittest’ from later evolutionary phases of ‘the survival of the fittest’. It is well possible that the TEI of the predatory mouth-form differs among P. pacificus isolates at the microevolutionary level and studies on this subject are currently ongoing. Therefore, studies towards the adaptive value of TEI and comparative work at the microevolutionary level represent some of the major challenges of TEI research in the years to come with P. pacificus and its mouth-form plasticity offering potential assets.

4 Opportunities and challenges

The TEI of the predatory mouth form in P. pacificus holds the promise to link laboratory-based molecular and experimental findings of transgenerational memory to ecology and evolution. Three major aspects represent challenges but also opportunities for future research. First, this review, focusing on nematode model organisms with their striking characteristics enabling large-scale laboratory studies, clearly indicates that material other than DNA is transmitted between generations and can influence the phenotype and behavior of offspring, sometimes over multiple generations. While such claims have been made for decades, i.e. the many examples of several human famines in the 19^th and 20^th century (Painter et al., 2008; Huang et al., 2010; Veenendaal et al., 2013), the severe confounding effects and the long generation times in humans have made it difficult to reach conclusive results. Therefore, model system approaches are crucial to achieve definite insight. Learning about the rapid evolutionary diversification of epigenetic processes and mechanisms, it will be crucial to involve a diversity of organismal from many domains of life, to elucidate the forms and varieties of TEI and their underlying molecular mechanisms.

Second, nutrient provisioning through vitellogenin seems to represent a recurring theme for information and material flow from the parental soma to the germ line, thereby breaching the Weismann barrier. Its biochemical features, make these glycoproteins prime candidates for the transmission of additional material. Vitellogenin is long known to form larger complexes with lipids, but only few preliminary studies approach the potential role of lipids as information carrier (Wilhelm et al., 2025). Small molecules, such as vitamins, neurotransmitters or neuropeptides, might also carry information across the generation barrier. Work on neuromodulation in defined circuits, especially in C. elegans, shows that internal state and sensory context reshape how environmental cues are processed: neuromodulators such as serotonin, dopamine, tyramine, octopamine, and neuropeptides select which sensory neurons and synaptic pathways are recruited for the same stimulus, activating only a subset of anatomically possible circuits (Bargmann, 2012). In this view, neuromodulatory state determines not only how information is represented, but also which environmental features are available to be encoded and potentially transmitted to the germ line.

Finally, RNAs, either double-stranded or single-stranded, and of course any type of sRNA, might be transferred with vitellogenin from the soma of the mother to the germ line and the oocyte. In C. elegans, specific microRNAs have been shown to physically associate with yolk proteins, to originate predominantly from the intestine and to depend on receptor-mediated endocytosis for delivery to oocytes (Aupérin et al., 2026). Such studies represent a unique opportunity into novel modes of inter-organismal communications.

Third, the adaptive value of any type of non-Mendelian inheritance for ecology and evolution is highly disputed and contentious. However, we would argue that this controversy stems in part from the limited number of model systems to study natural readouts of environmental influence, developmental plasticity and transgenerational effects. Indeed, studying such natural readouts under controlled laboratory conditions with its build-in opportunity will help to identify underlying molecular principles. Therefore, many of the examples described in this contribution have the potential to overcome current shortcomings and concerns. However, this also requires the awareness of the scientific community and practitioners of this research that not all patterns and processes observed in nature must necessarily be the result of natural selection and thus, not all such patterns must be adaptive. In an insightful contribution, Baugh and Day (2020) have carefully laid down which pre-requirements have to be fulfilled to make such claims. To investigate the adaptive consequences and the potential adaptive value of non-genetic inheritance, experimental studies are best combined with modelling approaches. In P. pacificus mouth-form plasticity and TEI research, an array of studies has recently begun to investigate these essential questions (Dardiry et al., 2023; Kalirad and Sommer, 2024a, 2024b, 2025). Similarly, the combination of molecular, experimental work with modelling has also been applied in C. elegans (Houri-Zeevi et al., 2020). Therefore, insight from modelling in the context of evolutionary theory will likely help to overcome current resistance.

5 Conclusions: the importance of transgenerational inheritance for developmental plasticity as facilitator of novelty

The studies summarized above, as well as other related works indicate that the time is ripe to consider non-genetic information flow from parents to offspring as common processes in biology with potential importance for evolution. This exciting research field will likely benefit from the interactions and comparisons between various study systems going well beyond nematodes, which were the focus of this article. How the Weismann barrier is breached and what type of information is transmitted - lipids, small molecules and/or various types of RNAs - will be the focus of future research.

In addition to be an interesting phenomenon in its own right, we want to conclude by highlighting the significance of transgenerational inheritance for developmental plasticity and its role for evolution (Figure 5). Developmental plasticity has emerged as a significant facilitator of phenotypic evolution and for the generation of novelty (Sommer, 2020). The latter is of extreme importance, as it is difficult to explain how natural selection should favour the development of new structures in the first place, a critique of Darwin that has been continuously expressed ever since the publication of the Origin (Mivart, 1871). Developmental plasticity might have a central function in this context. Proof that novel structures evolve in association with pulses of plasticity has been provided in empirical case studies with examples from insects (Hunt and Amdam, 2005), nematodes (Susoy et al., 2015) and vertebrates (Levis and Pfennig, 2019). Building on a strong theoretical foundation through the magisterial monograph of West-Eberhard (2003), these studies and many others provide strong support for ‘plasticity-first’ or ‘plasticity-led’ evolution (for a recent monograph see, Pfennig, 2021)). Four consecutive steps have to be envisioned for plasticity-associated evolution (Sommer, 2020) (Figure 5). First, a monotypic trait comes under the influence of an environmental signal resulting in a plastic readout or a polyphenic trait associated with novelty, i.e. plasticity-first evolution. Second, environmentally induced developmental switches regulate the expression of alternative phenotypes. Indeed, several such developmental switches have recently been identified (for recent review see, Wighard et al., 2025). Third, existing polyphenisms and other plastic traits can be subject to natural selection resulting in further diversification of associated structures, a process often referred to as ‘genetic accommodation’ (West-Eberhard, 2005). Finally, the environmental cue that originally generated the plastic character in the first place will become genetically encoded resulting in a fixation event. This termination of plasticity has been referred to as genetic assimilation, genetic fixation or canalization, as originally used by Waddington (1942). Given the different meaning of ‘canalization’ in the robustness and plasticity literature, we use the term genetic assimilation in the context of plasticity.

In such theoretical models of plasticity-associated evolution, three important processes need to be characterized at the molecular level. First, what are the molecular mechanisms associated with plasticity-first evolution? Second, what are the gene regulatory networks that allow developmental switching? And third, what type of molecular mechanisms result in genetic assimilation? We hypothesize that transgenerational inheritance holds promise as a contributor to the third, but potentially also the first question; genetic assimilation and the origin of plasticity, respectively. The fact that environmentally-induced phenotypes can be transmitted for multiple generations indicates that epigenetic processes can have long-lasting effects on phenotypic plasticity. Several studies in C. elegans have reported that transgene silencing upon environmental stressors became indefinite (Klosin et al., 2017; Devanapally et al., 2021). How the original epigenetic phenomena can be permanently manifested and become genetically encoded remains currently unknown. While such examples after experimental laboratory-based manipulations are rare, they have the potential to induce evolutionary changes. Therefore, the true challenge for future research will be to identify molecular mechanisms associated with genetic assimilation and similarly, the origin of plasticity.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Takayuki Onai, University of Fukui, Japan

Reviewed by: Alessandro Minelli, University of Padua, Italy

Rokhana Faizah, Indonesian Oil Palm Research Institute (IOPRI), Indonesia