Introduction

The endodermal germ layer includes the digestive tract, through which breakdown of food and absorption of nutrients and minerals occurs. Whereas the digestive tract of higher animals is subdivided into many regions, that of the nematode, Caenorhabditis elegans, comprises only a simple pharynx (esophagus), intestine, and hindgut (rectum). The intestine is a simple tube consisting of 20 cells that runs some 80% of the length of the animal. In addition to serving as the primary site of digestion and nutrient absorption, the intestine also fulfills the functions of the liver, while also having central roles in pathogen infection, immunity, longevity, and detoxification of metals (Block et al. 2015; Ezcurra et al. 2018; Peres et al. 2018; Pukkila-Worley and Ausubel 2012; Troemel et al. 2008). In hermaphrodites, the intestine is also the site of production of yolk lipoprotein, vitellogenin, that is secreted and taken up by developing oocytes (Grant and Hirsh 1999; Kimble and Sharrock 1983). The C. elegans intestine has also served as a model for the study of various human diseases, including fat storage disorders, diabetes, and metabolic syndrome (Campbell and Fares 2010; Hermann et al. 2012; Silverman et al. 2009; Zhang et al. 2011b; Zhu et al. 2016).

Unlike insect and vertebrate models, C. elegans lacks the capacity to replace somatic cells. The intestine must therefore serve the animal for its life span. By contrast, in Drosophila, the juvenile intestine is replaced by new cells during the pupal stage, while adults maintain a population of stem cells that can regenerate different intestinal cell types (Buchon and Osman 2015; Miguel-Aliaga et al. 2018). In vertebrates, intestinal crypts contain stem cells that regularly generate and shed cells from the villi lining the intestinal epithelium (Gehart and Clevers 2018; Spence et al. 2011). The lack of cellular replacement in C. elegans means that the gut is more prone to permanent cell damage or loss due to injury or infection, but it also means that more subtle effects resulting from experimental manipulation, or aging, are easier to discover because they persist over time.

C. elegans animals are found around the world, often on rotting fruit, where they likely consume nutrient-rich substrates and bacteria (Kiontke and Sudhaus 2006; Kiontke et al. 2011). In the laboratory, C. elegans are typically grown on Escherichia coli, a diet of ~ 55% protein (Brenner 1974; Yilmaz and Walhout 2016). Other bacteria used in laboratory studies include E. coli modified to express double-stranded RNA (dsRNA) for feeding-based RNA interference (Kamath et al. 2001), species that provide a richer food source (Laurent et al. 2001; Macneil and Walhout 2013) or pathogenic bacteria (Jiang and Wang 2018; Khan et al. 2018). The choice of food and C. elegans strain can have drastic effects on metabolism, gene expression, and growth phenotypes (Celen et al. 2018; Macneil and Walhout 2013; Reinke et al. 2010; Xiao et al. 2015; Zhao et al. 2018). Unlike mammals, but similar to Drosophila, C. elegans cannot synthesize cholesterol and it must be added to culture media (Hieb and Rothstein 1968; Rauthan and Pilon 2011; Stiernagle 2006; Vinci et al. 2008).

As a small metazoan model system, C. elegans is amenable to rapid characterization of genes by forward mutation, and reverse genetics, i.e., with CRISPR/Cas9 and RNA interference (Ahringer 2006; Jorgensen and Mango 2002; Waaijers and Boxem 2014), and its transparency enables live cell imaging (Hutter 2012). The intestine is particularly amenable to RNA interference by ingestion of dsRNA (Whangbo et al. 2017). Historically, genetic approaches identified small numbers of genes important for gut function by mutant phenotype. More recently, systems biology and “omics” work have begun to characterize the entire metabolic enzyme network and determine how the animal responds to changes in its diet. Several reviews cover various aspects of the C. elegans intestine in detail (Altun and Hall 2009; Kormish et al. 2010; Maduro 2017; McGhee 2007); here, we will describe C. elegans intestine structure and development and highlight key features of the underlying molecular genetics. We will then give an overview of recent work on intestinal cell biology as it relates to digestion, metabolism, physiology, and aging and give examples of emerging areas.

Intestine structure

The C. elegans digestive tract is essentially a tube (Fig. 1). At the anterior end, the self-contained pharynx, analogous to the esophagus, comprises 58 cells with its own epithelium, musculature and neurons (Mango 2009; Sulston et al. 1983). Pumping of the pharynx muscles, and peristalsis through its lumen, directs food to the intestine (Song and Avery 2013). At the posterior end of the pharynx is the grinder, a cuticular structure that mechanically breaks up food particles.

Fig. 1
figure 1

Origin and basic structure of the C. elegans intestine. a The gut originates as the E blastomere in the early embryo (Sulston et al. 1983). After approximately 12 h at 25 °C, in the newly hatched first-stage juvenile (L1 larva), the intestine is found between the anterior pharynx and the posterior hindgut. Further development takes approximately 45 h to progress through the L2, L3, and L4 larval stages to adulthood. The alternative dauer larva stage occurs during conditions of low food availability and/or high population density. In the adult, the intestine (outlined by dashed black lines) is visible by the granular contents of the gut cytoplasm in a light microscope image. In the fluorescence image directly below, expression of myo-2::GFP and acdh-1::GFP transgenes mark pharynx muscle and gut nuclei/cytoplasm, respectively (MacNeil et al. 2013; Okkema et al. 1993). The gut is outlined as in the light micrograph. b The basic structure of the intestine as a set of rings, consisting of four cells in the anterior-most int1 ring and two cells in int2 through int9. The lumen is located apically at the interface between cells in the rings (Asan et al. 2016)

Full size image

The intestine is posterior to the grinder, separated from it by the pharyngeal-intestinal valve. The intestine consists of 20 cells arranged as “rings” (Fig. 1b). At the anterior, four cells form the ring int1. The remaining 16 cells are arranged in pairs, with each pair forming rings int2 through int9. The outside surface of the intestine is surrounded by a basement membrane that faces the pseudocoelomic space (Leung et al. 1999). The intestinal basement membrane contains type IV collagen and most of the protein components found in basement membranes that surround other C. elegans tissues (Kramer 2005).

On the apical side of gut cells is the lumen, elliptical in cross-section, and which is lined with membranous microvilli that form the brush border (Fig. 2). The microvillar projections are supported by the terminal web, comprised primarily of an intracellular network of actin. Within the lumenal space and just outside of the microvilli, is the glycocalyx, a region rich in glycoproteins that provides a physical barrier from pathogens, serves as the interface between digestive enzymes and macromolecules, and likely allows filtering of lumen contents from entering gut cells (McGhee 2007). Besides stem cells alluded to earlier, the C. elegans gut lacks additional cell types such as phagocytes and gland cells that are found in other invertebrates (Karasov and Douglas 2013).

Fig. 2
figure 2

The active cytoplasm of C. elegans enterocytes. A pair of gut cells, part of the same int ring, are diagrammed shown as a transverse section. Cellular contents are shown as examples of the diverse structures and processes that are found in gut cells as cited in the text. Sizes of structures are only approximate. Abbreviations: AJ, adherens junction; ER, endoplasmic reticulum; LRO, lysosome-related organelle (gut granule)

Full size image

At its posterior end, the intestine is connected to the hindgut through the intestinal-rectal valve. The hindgut provides passage of waste to the environment through the anus, regulated by three muscle sets, including two muscles that wrap around the ventral posterior of the intestine (White et al. 1986). These ventral posterior muscles, and the attachment of the intestine to the pharynx and hindgut, are the only rigid connections that the intestine has to the body (Altun and Hall 2009).

Parasitic nematodes have variations of this basic structure, likely reflecting adaptations to unique environments. For example, whereas the C. elegans gut is cellularized, some plant and animal parasitic nematodes have a syncytial intestine (Byers and Anderson 1973; Colley 1970). Within gut cells, the terminal web is absent in plant parasitic nematodes, and may only appear in the adult stage of animal parasitic nematodes (Bruce 1966; Byers and Anderson 1973; Colley 1970). Finally, whereas the C. elegans intestine connects to a small number of muscles, that of the pinworm Aspiculuris tetraptera is surrounded by a muscle fiber network (Lee and Anya 1968).

Intestine development

The C. elegans gut originates as a single cell called E (for endoderm) during early embryogenesis (Sulston et al. 1983). A series of synchronous mitoses results in 16 E descendants, four of which undergo an additional division to result in 20 cells. The gut is generally invariant in cell number but occasionally there are 21 or 22 cells instead of 20 (Asan et al. 2016). Embryogenesis takes about 13 h at 25 °C (Sulston et al. 1983). Internalization of the gut occurs when the two E daughters migrate from the ventral part of the embryo into the interior (Leung et al. 1999). The gut primordium undergoes a set of carefully orchestrated cell movements and polarization to result in longitudinally arranged pairs of cells (with four at the anterior). The movements of these cells have been described in detail (Asan et al. 2016; Leung et al. 1999; Rasmussen et al. 2013). In the latter half of embryogenesis, the lumen of the intestine begins to form (Fig. 2) (Leung et al. 1999; Maduro 2017). The lumen forms first in discontinuous regions starting at the 16E stage, when gut cells begin to polarize their contents apicobasally (Leung et al. 1999). Formation of the lumen is coincident with formation of adherens junctions around the lumen and between int rings (Asan et al. 2016). The intestine is ready to function upon hatching. In the first larval stage, 14 additional nuclear divisions occur without cytokinesis, and a doubling of nuclear content accompanies each larval molt, resulting in 34 total nuclei of content 32C, in the same 20 cells that were present at hatching (Sulston et al. 1983).

The reproducible number of cells and nuclei in the C. elegans intestine suggests there has been strong selection for its maintenance. However, animals can accommodate slightly fewer numbers of intestinal E descendants or an excess with no apparent consequences to organogenesis and intestine function, at least in the laboratory (Choi et al. 2017; Clucas et al. 2002; Kostic and Roy 2002; Lee et al. 2016). For example, in mutant backgrounds that result in greater than 30 gut cells rather than the typical 20, a functional intestine still forms (Choi et al. 2017). Therefore, morphogenesis of the intestine does not require a fixed number of cells, and the nearly invariant number of 20 cells may represent a trade-off between resource allocation and maintaining enough cells to support integrity and flexibility of the lumen (Asan et al. 2016).

Genetics of gut development

Gut formation begins with specification of the E blastomere as the gut progenitor. An embryo in which E is not specified develops into an arrested larva, somewhat shorter than the wild type, and missing all intestinal tissue (Owraghi et al. 2010). As expected, the resulting larvae are inviable, presumably because they have no means by which to take in nutrients.

Specification of E occurs through the activation of a gene regulatory network of maternal and zygotic transcription factors that are expressed in a temporal sequence (Fig. 3). The earliest-acting factors that specify E are the paralogous end-1 and end-3 genes, which encode structurally similar GATA type transcription factors that are partially redundant (Maduro et al. 2005; Zhu et al. 1997). Upstream of the end genes, specification of E begins when the four-cell stage blastomere P2 contacts the mother of the E cell, called EMS (Goldstein 1992). After EMS divides, the side of EMS in contact with P2, its posterior side in normal embryos, becomes E (Goldstein 1993). The gene network that specifies gut was deduced through a combination of genetic and in vivo and in vitro DNA binding studies (Maduro 2017). Induction requires overlapping Wnt/MAPK/Src signaling that produces an endoderm-promoting state of the Wnt nuclear effector TCF/POP-1 in E (Rocheleau et al. 1997; Shetty et al. 2005; Shin et al. 1999; Thorpe et al. 2000). In this state, POP-1 interacts with the divergent β-catenin SYS-1 to become a weak transcriptional activator of the gene end-1. In parallel with the Wnt/β-catenin asymmetry pathway, maternally provided SKN-1, a bZIP/homeodomain-like transcription factor, activates expression of the med-1 and med-2 genes (Maduro et al. 2001). These encode divergent GATA-type transcription factors that together contribute to activation of end-1 and end-3 (Broitman-Maduro et al. 2005). The Sp1-like factor SPTF-3 also provides input into the end genes (Sullivan-Brown et al. 2016).

Fig. 3
figure 3

Integration of a developmental gene regulatory network with metabolic networks that modulate physiological responses to the environment. Environmental influences on gene expression, enzyme activity, and metabolic flux include temperature, oxygen levels, metabolites, ascarosides, and pathogens. Example regulators are shown activating shared and distinct target genes encoding enzymes that function at different steps in biochemical pathways. This figure was drawn by combining and modifying similar figures in other reviews (Maduro 2017; Watson and Walhout 2014)

Full size image

The result of these many inputs is to cause timely and robust activation of the elt-2 GATA factor gene, which is both necessary and sufficient to specify and maintain intestinal cell identity (Fukushige et al. 1998; McGhee et al. 2009). The paralogous elt-7 gene is also activated in the early E lineage and is expressed in parallel with elt-2, and the two factors have overlapping functions (Sullivan-Brown et al. 2016). When elt-2 is absent, gut is still specified in the embryo, but animals arrest as larvae with abnormal intestines (Fukushige et al. 1998). Consistent with overlapping function of elt-2 and elt-7, this phenotype is exacerbated when elt-7 is also absent, even though loss of elt-7 alone has no apparent phenotype (Sommermann et al. 2010). Hence, gut specification in C. elegans represents the collaborative input of multiple redundant factors working together to assure robust gut development (Choi et al. 2017).

Intestinal specification by an ELT-2-like factor appears to be widely conserved among nematodes. The parasitic nematode Haemonchus contortus, also known as the barber’s pole worm, has an intestinal ELT-2-like protein that can cause development of intestinal tissue when its coding region is overexpressed in C. elegans (Couthier et al. 2004). The genome of the parasite Ascaris suum also encodes a putative ELT-2-like factor (Rosa et al. 2014). Putative ELT-2 orthologs are also found among Pristionchus, Brugia, and Onchocera species as listed in WormBase (http://www.wormbase.org). Direct evidence that an ELT-2-like factor regulates intestinal fate in these other species has not yet been obtained.

Upstream of ELT-2, the MED and END genes have orthologs only in more closely related Caenorhabditis species (Coroian et al. 2005; Gillis et al. 2007; Maduro et al. 2005). These are absent in more distant Caenorhabditis species, suggesting that some other mechanism must initiate gut specification (Maduro 2017). It is possible that activation of elt-2 in E could specify gut directly, as overexpressed ELT-2 can achieve gut specification in C. elegans in the absence of end-1 and end-3 (Wiesenfahrt et al. 2015). Even among Caenorhabditis species that have MED and END orthologs, however, the upstream factors in the gut specification network may regulate gut specification differently. For example, in C. elegans, loss of pop-1 function results in an “extra endoderm” phenotype in which another cell, MS, makes gut as well as E (Lin et al. 1995). In C. briggsae, however, loss of the pop-1 ortholog results in the loss of endoderm specification (Lin et al. 2009). These differences illustrate the broader phenomenon of developmental system drift, common to such regulatory networks (Haag et al. 2018; True and Haag 2001).

The use of GATA transcription factors in specification and development of endodermal tissues is evolutionarily shared among protostomes and deuterostomes. For example, in Drosophila, gut specification involves the GATA factors Serpent and dGATAe (Murakami et al. 2005; Okumura et al. 2005), while among vertebrates, the GATA4/5/6 factors are important for development of endoderm and other tissues (Lentjes et al. 2016; Zorn and Wells 2009).

Assembly of the gut primordium results from the orientation of cell divisions that occur from one round of mitosis to the next within the E lineage, enforced by cell-cell signaling using the Van Gogh/VANG-1 pathway that keeps the primordial cells in a stereotypical alignment (Asan et al. 2016; Leung et al. 1999). Between the stages at which there are 8 E descendants and 20 descendants, specific pairs of cells undergo migrations and reorientations within the primordium, guided by signaling of the Ephrin, Notch, and Netrin pathways (Asan et al. 2016).

Formation of the gut lumen occurs by the apical targeting of components that support the terminal web, but the full details of how the microvilli become assembled are not yet understood. The Ezrin/Radixin/Moesin membrane-cytoskeleton linker protein, ERM-1, and the actin ACT-5 are two essential components that support the cytoskeleton supporting lumenal microvilli (Gobel et al. 2004; MacQueen et al. 2005). Apicobasal polarity of developing gut cells, and proper endosomal trafficking, are required for proper lumen formation. Loss of the Clathrin chc-1 results in mislocalization of these proteins and a disruption in lumen structure (Zhang et al. 2012). Loss of some enzymes in lipid biosynthesis, such as the sphingolipid biosynthetic enzyme LET-767, also disrupts apical localization of lumen components and hence lumen structure (Zhang et al. 2011a).

Digestion: physiology and metabolism

Digestion begins with mechanical and enzymatic breakage of the cell walls and membranes of microbes in the pharynx grinder (McGhee 2007). Partially broken down food that enters the intestine lumen gets exposed to enzymes that break down membranes and their lipid constituents, such as lysozymes, saponins, and lectins, enabling access to the cellular contents and also serving as a defense against microbes (Gravato-Nobre et al. 2016; Kandasamy et al. 2012; Mallo et al. 2002; Tarr 2012). Experimental details about the subcellular compartmentalization of digestion in C. elegans are somewhat lacking, perhaps owing to a primary focus of most work on establishing cause-and-effect genetic relationships, coupled with optical challenges in determining protein localization due to the high autofluorescence of the gut. However, as fundamental digestive mechanisms are conserved, it is likely that in the C. elegans gut, macromolecules are partially hydrolyzed in the lumen prior to endocytosis (i.e., extracellular digestion), after which further breakdown occurs when endocytic vesicles internalize lumen contents and fuse with acidified lysosomes inside gut cells (intracellular digestion) (Karasov and Douglas 2013; McGhee 2007).

As would be expected, many digestive enzymes are expressed in the C. elegans intestine, the most abundant of which are proteases (McGhee et al. 2007). Examples of proteases include astacins (NAS), metalloproteases that are found within the digestive tract (Mohrlen et al. 2003); the aspartic protease ASP-1, found in both the lumen and intracellular lysosomes (Tcherepanova et al. 2000); and aminopeptidase P which likely functions in intracellular hydrolysis (Laurent et al. 2001). Enzymes for other macromolecules include an amylase-like enzyme, C50B6.7, that is orthologous to human amylases, and which is expressed in the intestine (McGhee et al. 2007; Mulder et al. 2003), and the deoxyribonuclease (DNase) NUC-1 (Hevelone and Hartman 1988). For information on digestion and metabolism of other macromolecules, the reader is referred to comprehensive reviews on lipid and carbohydrate metabolism and intermediary metabolism (Braeckman et al. 2009; Lemieux and Ashrafi 2015; Watts and Ristow 2017).

Among nematodes, core metabolic pathways are generally conserved, with missing or additional pathways apparent particularly in parasitic species, as predicted by genome sequence comparisons (Martin et al. 2012; Wylie et al. 2008; Yin et al. 2008). Recent genomics approaches have been used to infer the structure of the metabolome by identifying orthologous genes that encode enzymes and integrating information from metabolic pathway databases and published literature (Gebauer et al. 2016; Witting et al. 2018; Yilmaz and Walhout 2016). These types of studies, confirm that both intracellular and extracellular digestion occur in C. elegans. In one recent work, over 600 metabolic enzymes in almost 2000 reactions were identified and inferred to occur in the cytosol, mitochondria, or extracellularly based on prior work (Yilmaz and Walhout 2016). The resulting network, called iCEL1273, also incorporates transport proteins, supported in part by published literature. The resulting model can be used to integrate phenotype and genotype data to predict the biochemical flux of nutrients into C. elegans biomass and waste products (Yilmaz and Walhout 2016).

One class of biomolecules deserves special mention. Unlike DNA, dsRNA escapes digestion into monomers in C. elegans. The gut can internalize dsRNA through the lumen, resulting in systemic gene silencing by RNA interference (Timmons and Fire 1998). dsRNA is specifically imported into intestinal cells from the lumen by the transporter SID-2 (Winston et al. 2007). This is likely to be an adaptive form of immunity against RNA viruses, as C. elegans strains with defects in the RNAi pathway show increased susceptibility to virus replication (Ashe et al. 2013; Gammon et al. 2017; Lu et al. 2005). Susceptibility to RNAi by dsRNA ingestion varies across species related to C. elegans (Nuez and Felix 2012).

Dynamic acidification in the lumen during digestion

C. elegans is essentially a filter feeder (Seymour et al. 1983). Food transits the digestive tract in as little as a few minutes, suggesting it is adapted for the animal’s short life cycle and simple diet (Ghafouri and McGhee 2007). Enzymatic breakdown of macromolecules in animal digestive tracts is generally associated with a low-pH compartment. Recent work has elucidated the details of acidification of the intestinal lumen in C. elegans. The intestinal lumen is weakly acidic on average, around pH 4.4 (Allman et al. 2009). A region of higher acidity, 1–2 pH units lower, starts in the posterior third of the gut lumen, gets translocated to the very anterior over several seconds, and remains for several seconds more before raising in pH (Bender et al. 2013). During this time, the posterior lumen reacidifies slowly over 30 s, and the cycle repeats every 45 s (Allman et al. 2009; Bender et al. 2013). This wave of protons is tied to the defecation motor program (DMP), a set of regular body muscle contractions that propels food through the intestine posteriorly and out the anus (Thomas 1990). Movement of the acidic region in the opposite direction results from the dynamic activity of proton pumps along the lumen (Bender et al. 2013). The DMP itself is regulated by the propagation of a wave of intestinal calcium ions generated in gut cells (Nehrke et al. 2008; Teramoto and Iwasaki 2006). Consistent with a role for the defecation cycle in nutrient uptake, the DMP is required for the internalization of fatty acids from the lumen into gut cells (Sheng et al. 2015). Hence, C. elegans digestion relies on coordinated and rhythmical physiological changes along the digestive tract that produce cyclical compartments of high acidity.

Regulation of metabolism by transcription factor and signaling networks

The C. elegans intestine achieves regulation of catabolic and anabolic processes through multiple pathways that integrate response to nutrients and environmental conditions at multiple levels (Fig. 3). Transcription factors and regulatory pathways, many of which share conservation with similar pathways in humans, have been identified that modulate changes in biochemical flux and intestinal physiology (Watson and Walhout 2014). The SKN-1/Nrf factor, important for the early initial specification of the E fate in embryos, is active in the adult gut as a major regulator of the response to oxidative stress (An and Blackwell 2003). The nuclear hormone receptor (NHR) family of factors is highly amplified in C. elegans and many of these are known to regulate metabolic networks in the intestine (Arda et al. 2010; Watson and Walhout 2014). These include NHR-49, a central regulator of the C. elegans stress response, and the breakdown of fats by beta-oxidation (Goh et al. 2018; Hu et al. 2018; Van Gilst et al. 2005); NHR-8, which regulates cholesterol and bile acid metabolism (Magner et al. 2013); and the NHR DAF-12, which regulates enzymes in catabolism including those in the citric acid cycle (Hochbaum et al. 2011; McCormick et al. 2011; Watson and Walhout 2014). The mediator subunit MDT-15 and the sterol regulatory element-binding protein SBP-1 work with NHR-49 in regulation of fat accumulation, and MDT-15 itself has additional roles in response to oxidative stress (Goh et al. 2014; Moreno-Arriola et al. 2016; Taubert et al. 2006).

The environment also modulates metabolic functions during development. Young C. elegans larvae, exposed to conditions of limiting food or high population density, enter an alternative third larval stage known as the dauer larva (Fig. 1). In this stage, metabolism and physiology are altered, and development paused, until food becomes available (Androwski et al. 2017). A central component of both dauer formation and metabolism is Insulin/IGF-1 signaling, or IIS (Murphy and Hu 2013). Mutation of the gene encoding the insulin receptor DAF-2 results in constitutive dauer formation at higher temperatures (Kimura et al. 1997). In adults, mutation of daf-2 extends lifespan and increases stress resistance, connecting metabolic changes occurring in dauer larvae and in adults (Kenyon et al. 1993). DAF-2 functions through a conserved phosphatidylinositol-3-kinase/protein kinase B/Target of Rapamycin (PI3K/Akt/TOR) pathway (Chen et al. 2013; Paradis and Ruvkun 1998). The life extension mediated by loss of daf-2 depends primarily on the activity of the conserved FOXO transcription factor DAF-16 (Lin et al. 2001). DAF-16/FOXO and the helix-loop-helix factor HLH-30/TFEB both translocate to intestinal nuclei to regulate genes that promote longevity and stress resistance (Lin et al. 2018). Insulin-like peptides (ILPs), of which 40 are encoded in the C. elegans genome, are part of a dynamic signaling system involving DAF-16/FOXO that maintains metabolic homeostasis depending on nutrient availability (Kaplan et al. 2018). ILPs mediate communication between neurons and the intestine, explaining how nutritional signals in the animal affect metabolism in the gut (Hung et al. 2014).

Intestinal endosomes and storage granules

Different types of granules or membrane-bound storage organelles have been identified in C. elegans enterocytes as shown diagrammatically in Fig. 2. The most visible of these are gut granules, which contain birefringent material that is visible under polarized light (Hermann et al. 2005). Gut granules are membrane-bound, lysosome-related organelles (LROs) enriched in various substances including zinc (Coburn and Gems 2013; Hermann et al. 2012; Roh et al. 2012). The granules also fluoresce blue under UV light due to the presence of glycosylated anthranilic acid (Coburn et al. 2013). Gut cells also contain conventional acidified lysosomes that are enriched in chloride and calcium ions (Chakraborty et al. 2017; Narayanaswamy et al. 2018). Lipid droplets are a major site of fat storage that can be detected by different methods, including staining with Oil Red O and dark field microscopy (Fouad et al. 2017; Lapierre et al. 2013; O'Rourke et al. 2009). As in other animals including humans, lipid droplets are surrounded by perilipin, a lipid storage regulatory protein that recognizes the phospholipid monolayer surrounding the droplet and regulates lipid metabolism (Beller et al. 2010; Chughtai et al. 2015; Copic et al. 2018; Hashani et al. 2018). Gut cells also contain peroxisomes, which serve important functions in fatty acid metabolism (Artyukhin et al. 2018; Yokota et al. 2002).

Because of the role of endosomes in digestion, the C. elegans gut is also a model for the study of endosomal trafficking, which relies on conserved endosomal proteins such as the GTPases RAB-5 and RAB-10 (Chen et al. 2006; Liu et al. 2018; Sato et al. 2014). Maintenance of the lumen over time also relies on proper apical targeting of endosomes, which relies on the conserved adapter proteins Clathrin and AP1 (Zhang et al. 2012). Enterocytes in hermaphrodites contain vesicles carrying yolk lipoprotein from the endoplasmic reticulum to the basement membrane (Grant and Hirsh 1999). Finally, a novel endosome type of unknown function has been recently identified that is distinct from LROs, lipid droplets, and vesicles carrying yolk and is associated with the ABC transporters HAF-4 and HAF-9 (Tanji et al. 2016).

Dietary restriction, lipid storage, autophagy, and life span

In C. elegans, a complex relationship exists among diet, mobilization or storage of lipids, autophagy, and life span (Lapierre et al. 2011; Lemieux and Ashrafi 2015; Seah et al. 2016). Dietary (caloric) restriction (DR) can be achieved in C. elegans by diluting E. coli cultures, growing worms in liquid culture, or using an eat-2 mutant which has decreased pharyngeal pumping (Cypser et al. 2013; Gelino et al. 2016; Lenaerts et al. 2008; Palgunow et al. 2012). DR during the larval stages slows development and extends post-reproductive lifespan (Palgunow et al. 2012). The intestine and epidermis (hypodermis) are major storage sites for lipid droplets, and animals undergoing DR show an apparent increase in their number (Palgunow et al. 2012; Zhang et al. 2010). Loss of germline stem cells also increases life span, and this increased longevity correlates with increased lipids in the intestine (Hsin and Kenyon 1999; Wang et al. 2008). Consistent with this, DR induces fatty acid synthesis gene expression in a manner dependent on DAF-16/FOXO (Amrit et al. 2016), and in response, fat catabolism genes are activated by NHR-49 and SKN-1 (Ratnappan et al. 2014; Steinbaugh et al. 2015). Autophagy, which delivers lipid droplets to lysosomes for lipophagy in the intestine, is a necessary component of this fat breakdown (Lapierre et al. 2011). Autophagy also mediates the life-span extension seen in animals undergoing dietary restriction (Gelino et al. 2016; Jia and Levine 2007).

Recent studies point to a causative role of yolk lipoprotein production in the regulation of autophagy and remodeling of lipids, which is proposed to promote long lifespan. In animals undergoing dietary restriction, longevity results from decreased expression of vitellogenin genes, and activation of DAF-16/FOXO (Seah et al. 2016). In germline-less animals, the retention of yolk causes activation of SKN-1, which in turn promotes remodeling of lipids to maintain lipid homeostasis (Steinbaugh et al. 2015). As animals age, the intestine undergoes progressive detrimental changes (McGee et al. 2011). Recently, these aging-related pathologies have been proposed to result from “run-on” vitellogenesis that consumes intestine cell contents, from within, by autophagy (Ezcurra et al. 2018). In this model, the conversion of intestinal biomass into yolk into late adulthood results in the reduction of volume of intestinal cells and widening of the lumen (both of which are seen in senescing adults) and extracellular buildup of yolk pools. Interestingly, autophagy has a role in promoting senescence by converting gut cell biomass into yolk; hence, the role of intestinal autophagy in life span depends on the role it plays in promoting lipoprotein synthesis or lipid breakdown (Ezcurra et al. 2018; Gelino et al. 2016). The complex interplay among different mechanisms for metabolic homeostasis continues to be an active area of study.

Future directions

The C. elegans gut continues to be an important system for study. The field now has access to systems-level descriptions of how the intestine functions in time and space at a finely detailed level, incorporating environmental signals, biochemical pathways, and gene regulation (Fig. 3). Some emerging areas, which we did not have time to discuss in detail above, include the following:

Transgenerational inheritance

Transgenerational epigenetic inheritance is observed by changes in gene expression when prior generations are exposed to dietary extremes such as starvation or high-glucose diets (Rechavi et al. 2014; Tauffenberger and Parker 2014). These necessarily involve transmission of information from the intestine to the germline, and understanding this process may shed light on mechanisms that may mediate human epigenetic inheritance influenced by diet (Kadayifci et al. 2018).

Metabolic network models

As alluded to above, recent studies have systematically identified all metabolic enzymes encoded in the C. elegans genome to generate metabolic network models (Gebauer et al. 2016; Witting et al. 2018; Yilmaz and Walhout 2016). Coupled with metabolomics analyses, these models are a computational framework for predicting changes in metabolic flux that can be validated or tested experimentally, for example, by whole-animal metabolomics studies (Witting et al. 2018). Future studies that could integrate high-resolution protein localization studies may add new insights into subcellular compartmentalization of digestive functions in this species.

The microbiome

Recent work has begun to analyze the microbiota found in the C. elegans natural environment, with a view toward analyzing host-microbe interactions (Dirksen et al. 2016; Gerbaba et al. 2017; Tan and Shapira 2011; Zhang et al. 2017a). Future studies could reveal many new insights into interplay among environmental metabolites, metabolic flux, cellular physiology, and immunity.

Ascarosides as signaling molecules

Ascarosides are a nematode-specific group of diverse glycolipids based on the sugar ascarylose, and they are known to modulate behavioral, developmental and environmental responses (von Reuss 2018). These are synthesized by enzymes that also function in fat breakdown in the intestine (Panda et al. 2017). The relationship of C. elegans metabolism to ascaroside synthesis and the effects of some 100 known ascarosides on physiology are areas of active study (von Reuss and Schroeder 2015; Zhang et al. 2017b; Zhou et al. 2018).

The glycocalyx

Although the glycocalyx is the main interface between the lumen and the apical surface of intestine cells, the molecular biology of its structure and function is not well understood (McGhee 2007). Development of new methods to study the glycocalyx may be able to reinvigorate interest in the digestive processes that occur within the lumen, particularly now that the dynamics of lumen acidification are understood (Bender et al. 2013).

Caenorhabditis inopinata, a sister species of C. elegans

The newly described species Caenorhabditis inopinata, genetically a close relative of C. elegans, spends most of its life cycle in the interior of the fruit of the fig Ficus septica and disperses through its pollinating wasp Ceratosolen (Woodruff and Phillips 2018). Because of these differences in life histories between C. elegans and C. inopinata, a study of the intestine and digestion between the two is likely to reveal key evolutionary adaptations of the latter species to a very different environment than that of C. elegans. The genome of C. inopinata has been sequenced, and transgenic and RNAi methods similar to those of C. elegans have been developed, which should enable functional genetic studies between the two species (Kanzaki et al. 2018).

Conclusion

The C. elegans intestine continues to be a valuable model in which to study organ structure and function, from small-scale studies investigating small numbers of genes, to systems biology approaches integrating metabolomics, genetics, behavior, and environment. With so many dimensions of C. elegans biology now being actively studied, future studies are likely to continue to reveal new insights about intestine development and function that are relevant across animals.