The teneurin TEN-1 and its ligand LRON-11 support microbiota-induced neuroprotection and mechanosensory function in C. elegans

Summary

Neuroprotection preserves neuronal structure and function by limiting damage and cell death, and diet-associated microbiota produce metabolites capable of influencing neuronal survival^1–3. In C. elegans, GABA and lactate produced by E. coli HT115 protect touch receptor neurons (TRNs) from MEC-4(d)-induced degeneration², and comparative gene-expression analysis identified the teneurin ten-1 as required for this microbiota-induced neuroprotection³. Teneurins are evolutionarily conserved cell-adhesion molecules that regulate nervous system development, axon guidance, synapse formation, and neuronal connectivity primarily through trans-cellular interactions^4–22, and C. elegans, which contains a single teneurin gene, requires TEN-1 for normal nervous system development and tissue organization^5,13,14,23. Although vertebrate teneurins bind latrophilin G protein-coupled receptors^17,21,22,24, the C. elegans latrophilin homolog LAT-1 acts in parallel to TEN-1²⁵, leaving the receptor-ligand interactions that underlie TEN-1 function incompletely understood. Here, we identify Leucine-Rich Repeat Only 11 (LRON-11), a previously uncharacterized cell-adhesion molecule, as the ligand for TEN-1. The high-resolution cryo-EM structure of the TEN-1-LRON-11 complex reveals that the leucine-rich repeat domain of LRON-11 directly engages the TTR domain of TEN-1 through a previously unrecognized interaction interface. Tissue-specific genetic analyses demonstrate that TEN-1 and LRON-11 are required for microbiota-mediated neuroprotection, and structure-guided interface mutations disrupting TEN-1-LRON-11 binding impair normal touch responses in vivo. Together, our findings establish a conserved receptor-ligand signaling axis that supports microbiota-mediated neuroprotection and normal touch responses, revealing how a single receptor-ligand interaction can support distinct neuronal functions.

eTOC blurb draft

Teneurins wire the nervous system through cell-surface partners, but the ligand of the C. elegans teneurin TEN-1 was unknown. Li et al. identify LRON-11 as a TEN-1 ligand and solve their complex structure. Disrupting this complex impairs microbiota-induced neuroprotection and touch sensation, linking one receptor pair to two neuronal functions.

Graphical Abstract

Results and Discussion

TEN-1 interacts with LRON-11, a cell-surface receptor in C. elegans

We searched for TEN-1 ligands in C. elegans, in which many conserved cell surface receptor families have fewer members than in vertebrates, allowing for less extensive screens to cover the putative interaction space. There is only one C. elegans TEN gene, ten-1, which uses alternative promoters to produce two different protein isoforms, TEN-1 long (TEN-1L) and TEN-1 short (TEN-1S), and ten-1 mutants show major developmental phenotypes^13,23. TEN function is conserved between invertebrates and vertebrates, as the known nematode phenotypes match those in mice, including embryonic defects and axon guidance^9,13,23,26.

In recent work, we included the TEN-1 ectodomain in a protein interaction screen designed to discover pairwise extracellular interactions²⁷ (Figure 1A). The screen used the high-throughput Extracellular Interactome Assay (ECIA), an ELISA-like assay in which Fc-tagged baits capture oligomerized alkaline phosphatase (AP)-tagged preys²⁸ (Figure S1A), and tested TEN-1 against a library of 379 C. elegans extracellular regions (ECR) including protein domains known to be conserved and important in developmental functions across bilaterians²⁷. The TEN-1-Fc bait failed to capture LAT-1-AP prey, the C. elegans homolog of latrophilins (Figure 1B), and reciprocally LAT-1-Fc bait captured no TEN-1-AP signal (Figure 1C), suggesting that the C. elegans homologues of teneurin and latrophilin do not interact. Instead, the TEN-1-Fc bait captured LRON-11-AP prey with the highest significance across the entire library (Figure 1B), and LRON-11-Fc bait gave the strongest binding signal with TEN-1-AP (Figure 1D). LRON-11 is a relatively unstudied receptor in the LRR-only family of cell-surface receptors in C. elegans, and has an extracellular Leucine-Rich Repeat (LRR) domain, one transmembrane helix and a cytoplasmic tail (Figure 1A)²⁹.

Figure 1. TEN-1 binds to LRON-11 by a direct interaction.

(A) Domain architecture schematics of TEN-1 and LRON-11. Domain names of TEN-1 are indicated above scheme. TM, transmembrane domain; EGF, epidermal growth factor repeats; CRD, cysteine-rich domain; TTR, transthyretin-related domain; Ig-like, immunoglobulin-like domain. The horizontal lines below the diagrams show the domain boundaries of the constructs used for structure determination.

(B-D) Z-scores for AP signal observed against a TEN-1-Fc bait (B), LAT-1-Fc bait (C), and LRON-11-Fc bait (D) plotted from data in Nawrocka et al. (2026). The highest signals were observed for AP-tagged LRON-11 (B), TOL-1 (a LAT-1 ligand) (C), and TEN-1 (D). See Figure S1A for a schematic of the assay design.

(E) SPR sensorgrams for LRON-11 ECR binding to biotinylated TEN-1 ECR on a streptavidin-coated SPR chip.

(F) Binding isotherms for SPR data in (E), showing a K_D of 8.5 μM for WT LRON-11 and TEN-1 (95% confidence interval: 7–10 μM).

(G) SPR sensograms for LAT-1 ECR binding to biotinylated TEN-1 ECR on a streptavidin-coated SPR chip showing no detectable interaction.

See also Figure S1 and Figure S2.

Notably, the sole published study on lron-11 demonstrated it to be a critical regulator of ventral nerve cord (VNC) axon guidance, with mutants exhibiting robust axon navigation defects across multiple neuronal classes, including aberrant midline crossing of VNC interneuron and motor neuron axons as well as commissural polarity errors²⁹, suggesting that ten-1 and lron-11 may have closely related developmental functions.

The TEN-1–LRON-11 interaction is direct

To validate the TEN-1–LRON-11 interaction with an orthogonal, quantitative approach, we performed surface plasmon resonance (SPR) experiments in which various concentrations of purified LRON-11 ECR were flowed over purified, biotinylated TEN-1 ECR captured on streptavidin-coated SPR chips (Figure 1E). We observed a dissociation constant (K_D) of ~8.5 μM for wild-type LRON-11 and TEN-1 (Figure 1F), indicating that the TEN-1–LRON-11 interaction is direct and of moderate affinity. Repeating the experiment with various concentrations of LAT-1 ECR on the TEN-1 surface showed no direct interaction between C. elegans TEN-1 and C. elegans LAT-1 (Figure 1G), confirming the lack of TEN-1 interaction with LAT-1.

Cryo-EM structure of the complex reveals an unexpected topology

We next set out to determine the high-resolution structure of LRON-11 in a complex with TEN-1. To improve expression levels, we used a fusion construct in which the LRON-11 ECR is linked at its C-terminus to the TEN-1 ECR via a long flexible linker (Figure S1B), and collected single-particle cryo-electron microscopy (EM) images of the complex of the two full ECRs (Figure 2A–2C, Table S1 and Figure S2). Two-dimensional (2D) class averages showed an LRR-like horseshoe-shaped density protruding out from the bottom of the TEN-1 architecture (Figure 2A). We determined the cryo-EM structures of the TEN-1 ECR in isolation and in complex with the LRON-11 ECR at 2.6 Å and 3.4 Å resolution, respectively (Figure 2B–D, Table S1 and Figure S2). The structure of the TEN-1–LRON-11 complex revealed that LRON-11 binds to the transthyretin (TTR) domain of TEN-1 via its LRR domain, forming a heterodimer with approximate dimensions of 190 Å × 66 Å × 52 Å (Figure 2D). To our knowledge, LRON-11 represents the first identified ligand, other than teneurin itself, that binds to the TTR domain of a teneurin.

Figure 2. Cryo-EM structure of the TEN-1–LRON-11 complex reveals a previously unobserved mode of interaction.

(A) 2D cryo-EM averages of the TEN-1 ECR–LRON-11 ECR complex. The additional density corresponding to LRON-11 is indicated by an arrow.

(B) Three-dimensional cryo-EM density map of TEN-1 in isolation at 2.6 Å resolution. Different domains are shown in different colors.

(C) Three-dimensional cryo-EM density map of TEN-1–LRON-11 complex in isolation at 3.4 Å resolution.

(D) Cryo-EM structure of the complex of TEN-1 with LRON-11 (pink). Each domain of TEN-1 is colored and labeled.

(E-G) Superimposition of the TEN-1 (cyan) and LRON-11 (magenta) complex structure with (E) human TEN2–LPHN3 complex (orange) (PDB ID: 6VHH), or (F) Drosophila Ten-m/Ten-m dimer (orange) (PDB ID: 8FIA) or (G) TEN-1 in isolation (orange).

See also Figure S1, S2, and Table S1.

LRON-11 ECR is comprised of a large semicircular horseshoe-shaped N-terminal LRR domain with 14 leucine-rich repeat motifs, similar to the structure of Drosophila Capricious with the same number of LRR repeat numbers³⁰, flanked by N-terminal and C-terminal cap regions (Figure 1A). The TTR domain of TEN-1 binds to the N-terminal cap of the LRR domain of LRON-11 and makes several additional contacts with the first leucine-rich repeat within the LRR domain. Surprisingly, this interaction of the TTR domain with the LRON-11 LRR domain (Figure 2D) is different from other LRR domain-ligand complexes, where the ligand binding site of LRR domains is usually located in the concave surface of the LRR region, including the interaction of the D. melanogaster Toll receptor with its ligand Spätzle^31,32. Structurally, the TTR domain adopts a β-sandwich fold incorporating several loop regions that facilitate LRON-11 binding. No density was observed for the eight N-terminal EGF repeats of TEN-1, most likely because the region containing the EGF repeats of teneurins tends to be flexible.

LRON-11 binds to TEN-1 at a distant site than the vertebrate LPHN binding site

Superimposition of the TEN-1–LRON-11 complex with the vertebrate TEN2–LPHN3 complex structure¹⁷ revealed the mode of interactions of teneurin with its two different ligands: the LRON-11 binding site on TEN-1 is ~80 Å away from the LPHN-binding site on vertebrate teneurins, which is on the β-barrel domain (Figure 2E). Given that the two ligands of teneurin have distinct binding sites, the formation of a trimeric complex between teneurin, latrophilin and a yet-to-be-identified homolog of LRON-11 is very plausible in vertebrates, though such a complex would not form in C. elegans because of the lack of a LAT-1–TEN-1 interaction. Altogether, the LRON-11 interaction with TEN-1 is completely different from the TEN–LPHN interaction and is likely involved in additional teneurin functions.

The TEN-1–LRON-11 interaction is reminiscent of the asymmetric dimerization interface observed in Drosophila Ten-m, which is formed by the TTR domain of one Ten-m protomer and the β-propeller domain of another³³. Superimposition of the two structures shows that the C. elegans TEN-1–LRON-11 binding interface almost completely overlaps with the Drosophila Ten-m self-association interface, thus Drosophila Ten-m can either interact with the Drosophila LRON-11 homolog or interact with itself to homodimerize but cannot do both (Figure 2F). Comparison of the TEN-1 structure in isolation with the structure of the TEN-1–LRON-11 complex revealed no major conformational change upon ligand binding (Figure 2G).

Structure-guided point mutations break TEN-1–LRON-11 binding in vitro

The TEN-1–LRON-11 interaction is mediated by an extensive combination of polar and hydrophobic residues from both proteins, creating a large, buried interface area of 1257 Ų, and is formed by a major contact site mediated by non-canonical structural features in both proteins (Figure 3A–3B). The N-terminal cap of the LRR domain of LRON-11 is a cysteine-rich structure that connects β-strands with disulfide bridges; these entangled disulfide bridges provide rigidity and enable the presentation of the ligand binding site to TEN-1. On the TEN-1 side, numerous strangled loops that come out of the side of the β-sandwich of the TTR domain extend towards LRON-11 and form the majority of the LRON-11 binding interface (Figure 3B), with several hydrophobic residues protruding towards LRON-11. The binding interface is stabilized by a network of hydrogen bonds involving R47 with Y981, T982, and L1034, as well as interactions between I42 and R1026 and between L43 and Y981, together with a salt bridge formed between R47 and D983. Sequence alignment and conservation analysis of the TEN-1–LRON-11 structure suggests that the interface is conserved (Figure 3C and Figure S3A).

Figure 3. The binding Interface of the TEN-1–LRON-11 complex.

(A) Ribbon diagram of the TEN-1–LRON-11 complex showing the interface between TEN-1 and LRON-11. The empty red star shows the equivalent location of the Drosophila-specific splice insertion in the TTR domain on TEN-1 structure. In Drosophila, this splice insert regulates the interaction of teneurin with Capricious (see accompanying report, Ji et al.)

(B) A close-up view of the binding interface highlights residues involved in hydrogen bonds and salt bridges, indicated by yellow and red dashed lines, respectively. Residues at the binding interface are shown as sticks. Disulfide bridges formed between cysteine residues are shown as yellow sticks.

(C) Sequence alignments of LRON-11-binding interface on TEN-1 and TEN-1-binding interface on LRON-11. The red box highlights the splice site insertion in Drosophila teneurins. Black arrows mark residues chosen for mutation in C. elegans to selectively disrupt the TEN-1–LRON-11 interaction. Red stars indicate the key residues involved in the interaction. For the TEN alignment, sequences from Homo sapiens (Hs) TEN2 (UniProt: Q9NT68), Mus musculus (Mm) TEN4 (UniProt: Q3UHK6), Drosophila melanogaster (Dm) Ten-m isoforms PB and PD and Ten-a isoforms PL and PD (FlyBase), and Caenorhabditis elegans (Ce) TEN-1e (WormBase) were included. For the LRON-11 alignment, sequences from Homo sapiens (Hs) TLR2 (UniProt: O60603), Mus musculus (Mm) LINGO3 (UniProt: Q6GQU6) and R4RL2 (UniProt: Q7M6Z0), Anopheles gambiae (Ag) XP061518135.1 (UniProt: A0A1S4H6I9), Drosophila melanogaster (Dm) Tartan isoform PB and Capricious isoform PE (FlyBase), and Caenorhabditis elegans (Ce) LRON-11 isoform a (WormBase) were included. Sequence alignments were generated using PROMALS3D and visualized with Jalview ⁴².

(D) ECIA results show that TEN-1 mutations affecting the LRON-11 interface can diminish or abolish LRON-11 binding. Vice versa, LRON-11 mutations affecting the TEN-1 interface can diminish or abolish TEN-1 binding. The raw readout for the assay (absorbance at λ = 650 nm) is noted in each square.

(E-G) SPR sensorgrams to test the effects of mutations on the interaction of TEN-1 and LRON-11.

See also Figure S3.

To validate the binding interface, we used our cryo-EM structure and sequence analysis to design point mutations that specifically disrupt the interaction (Figure 3B–3C), using the ConSurf³⁴ server to identify conserved surface residues together with PDBePISA³⁵ to analyze all amino acids at the binding interface, combined with visual inspection of the interface (see STAR Methods). We engineered single- and multiple-point mutations on Fc-tagged TEN-1 and AP-tagged LRON-11 full ECRs and tested their effect on the TEN-1–LRON-11 interaction using ECIA. WT LRON-11 ECR robustly bound to WT TEN-1, whereas an LRON-11 ECR fragment carrying triple I42A, L43A, D64A mutations at the LRON-11 binding site did not bind to TEN-1 (Figure 3D and Figure S3B). Similarly, WT TEN-1 ECR robustly bound to WT LRON-11, whereas the TEN-1 ECR carrying F1037A, L1039A double mutations at the TEN-1 binding site did not bind (Figure 3D and Figure S3B). Western blots showed that the mutant ECR constructs successfully expressed and were secreted from insect cells (Figure S3C), and differential scanning fluorometry (DSF) showed that WT and mutant proteins had similar thermal stability, suggesting that the mutations are unlikely to cause misfolding or instability (Figure S3D–F). Thus, the LRR domain of LRON-11 and the TTR domain of TEN-1 are required for the interaction, and our cryo-EM structure accurately represents the complex.

To quantitatively measure how our engineered mutations break the LRON-11 and TEN-1 complex, we performed SPR experiments with the mutant proteins. Whereas wild-type proteins bound with the affinity described above (Figure 3E), binding of the TEN-1 F1037A, L1039A mutant or of the LRON-11 I42A, L43A, D64A mutant was decreased to undetectable levels (Figure 3F–G, Figure S3G–H), validating our structures and ECIA results. These experiments provide mutations that specifically diminish the TEN-1–LRON-11 interaction for use in functional in vivo assays.

ten-1 and lron-11 are both required for microbiota-mediated neuroprotection

We next asked whether TEN-1, together with its newly identified ligand LRON-11, is required for microbiota-mediated neuroprotection. We used the mec-4d degenerin as a model of neurodegeneration, in which the gain-of-function mec-4d mutation causes constitutive Na⁺ influx through the MEC-4 channel^36,37, leading to Ca²⁺ accumulation from both extracellular and intracellular sources and ultimately resulting in neuronal death, and in which bacterially produced metabolites, including GABA and lactate, protect neurons from neurodegeneration². Our comparative transcriptomic analyses of animals grown under protective and non-protective bacterial conditions previously identified ten-1 among the genes upregulated by the protective microbiota³.

We silenced each gene by feeding RNA interference (RNAi) and monitored touch receptor neuron (TRN) axonal integrity in the mec-4d neurodegeneration model (Figure 4A), comparing RNAi in two complementary strains to determine the tissue requirements for TEN-1 and LRON-11 in this response. In TU2773 animals, which are competent for systemic RNAi in non-neuronal tissues but largely refractory to neuronal RNAi, silencing of either ten-1 or lron-11 abolished the neuroprotective effect conferred by the microbiota (Figure 4B–C). In WCH6 animals, in which RNAi is restricted specifically to the TRNs, knockdown of either gene similarly eliminated microbiota-mediated neuroprotection (Figure 4B). Under these conditions, degenerating neurons exhibited pronounced axonal defects, including shortened or absent axons and abnormal axonal morphology (Figure 4C). Together, these findings indicate that both neuronal and non-neuronal expression of ten-1 and lron-11 contributes to microbiota-mediated neuroprotection.

Figure 4. TEN-1 and LRON-11 are required for microbiota-induced neuroprotection and normal touch responsiveness.

(A) Schematic representation of the RNA interference (RNAi) strategy used to assess the contribution of ten-1 and lron-11 to microbiota-induced neuroprotection.

(B) Effect of ten-1 and lron-11 knockdown on microbiota-induced neuroprotection. AVM morphology was quantified in mec-4d animals carrying either the systemic RNAi background or the TRN-specific RNAi strain WCH6 following RNAi against the indicated genes.

(C) Representative fluorescence micrographs showing AVM morphology following RNAi against ten-1 or lron-11 in the indicated RNAi background. Arrowheads indicate representative neuronal abnormalities. The abnormalities include neurites extending backwards and truncated forward neurites (ten-1, upper image), neurons with only one soma (ten-1, bottom), shorter anterior process (lron-11 upper image), discontinuous axons with truncation (lron-11, bottom). Animals were imaged 72 h after bleaching. Scale bar, 50 μm.

(D) Time-course analysis of microbiota-induced neuroprotection following RNAi against ten-1 or lron-11. The proportion of animals displaying wild-type AVM morphology was quantified at 24, 48, and 72 h after bleaching in the systemic and TRN-specific RNAi backgrounds.

(E) Schematic representation of the gentle-touch assay used to assess mechanosensory function.

(F) Tissue-specific RNAi analysis of ten-1 and lron-11 function in mechanosensation. Touch responsiveness was quantified following muscle-, neuron-, or intestine-specific RNAi against the indicated genes.

(G) Gentle-touch responses of wild-type animals, ten-1 and lron-11 CRISPR mutant strains, and endogenously tagged TEN-1::mNeonGreen and LRON-11::WrmScarlet reporter strains.

(H) Spatial distribution of TRN and LRON-11. Representative confocal imaging showing the spatial relationship between the endogenously tagged LRON-11::WrmScarlet reporter (magenta) and the TRN marker (green). Arrowheads indicate regions where the two fluorescent signals are observed in close spatial proximity (black spots), for a single confocal slice (top). For comparison, the maximum z-projection over the entire stack is shown (bottom). Scale bar, 50 μm.

Data are presented as mean ± SEM. Statistical analyses were performed using one-way ANOVA followed by Dunnett’s multiple-comparisons test (B) or Šidák’s multiple-comparisons test (F and G), or two-way ANOVA followed by Šidák’s multiple-comparisons test (D), in which each time point in treated animals was compared to the control in E. coli HT115. Unless otherwise indicated, comparisons were made against the corresponding control condition. Significance is indicated as ns, p ≥ 0.1234; p < 0032 (*); p < 0.0021 (**); p < 0.002 (***); and p < 0.0001 (****). Replication details are given in the STAR Methods.

See also Figure S4.

We next asked whether this phenotype results from the well-established developmental functions of TEN-1 or from a later role for the TEN-1/LRON-11 signaling axis during the neuroprotective response. We therefore examined TRN morphology at 24, 48, and 72 hours after bleaching in both TU2773 and WCH6 animals (Figure 4D). At 24 hours, when many animals had only recently hatched, the proportion of animals displaying wild-type axons remained very low in all treatments and both RNAi backgrounds. By 48 hours, however, control animals exhibited the expected increase in the proportion of wild-type axons, whereas knockdown of either ten-1 or lron-11 significantly impaired this neuroprotective response, and this difference persisted at 72 hours in both RNAi backgrounds, although the statistical significance of the individual knockdowns varied between strains. These findings suggest that the requirement for TEN-1 and LRON-11 becomes apparent as microbiota-mediated neuroprotection is established, rather than reflecting an early developmental defect. Finally, simultaneous TRN-specific knockdown of ten-1 and lron-11 did not enhance the degenerative phenotype beyond that observed following depletion of either gene alone (Figure 4B), consistent with TEN-1 and LRON-11 functioning within the same genetic pathway to mediate microbiota-induced neuroprotection.

The TEN-1-LRON-11 interaction is required for mechanosensory function

To investigate the roles of TEN-1 and LRON-11 within the mechanosensory circuit, we used the well-established C. elegans touch response assay³⁸ (Figure 4E), a behavior mediated by the TRNs, which detect gentle mechanical stimuli and relay signals to downstream interneurons and motor neurons³⁹. To identify the tissue(s) in which ten-1 and lron-11 are required for normal mechanosensory function, RNAi was restricted to muscle (NR350), intestine (IG1839), or neurons (TU3335), and gentle-touch responsiveness was quantified (Figure 4F). Neither muscle-specific nor intestine-specific knockdown of ten-1 or lron-11 altered touch responses compared to control animals. In contrast, neuronal RNAi against either gene significantly impaired touch sensitivity, with lron-11 producing the stronger behavioral defect (Figure 4F). Thus, neuronal expression of ten-1 and lron-11 is required for normal gentle-touch responsiveness.

We therefore next asked whether this neuronal function specifically depends on the physical interaction between TEN-1 and LRON-11. Guided by our high-resolution cryo-EM structure, we used CRISPR-Cas9 to selectively disrupt the TEN-1–LRON-11 binding interface while preserving the endogenous proteins. We generated strains homozygous for the ten-1(syb7771[ten-1_F1037A,L1039A]) allele, which carries two point mutations that disrupt the LRON-11 interaction surface, or for the lron-11(syb7811[lron-11_I42A,L43A,D64A]) allele, which carries three point mutations that disrupt the TEN-1-binding interface (Figure 3).

As a benchmark for assay performance, we examined animals carrying the mec-4(u253) allele, which lack functional MEC-4 channels and are therefore insensitive to gentle touch, consistent with a loss-of-function mutation⁴⁰; as expected, mec-4(u253) animals fail to respond to touch (Figure 4G). Animals carrying either the ten-1(syb7771) or lron-11(syb7811) alleles displayed pronounced touch insensitivity, with a significant reduction in responses to repeated gentle-touch stimuli (Figure 4G). Notably, mutation of the LRON-11 residues required for TEN-1 binding phenocopied key aspects of mec-4 loss-of-function animals, indicating that the TEN-1-binding surface of LRON-11 is required for normal TRN function; likewise, homozygous ten-1(syb7771) animals exhibited a comparable reduction in touch responsiveness, demonstrating that the reciprocal LRON-11-binding surface of TEN-1 is also required for this behavior (Figure 4G).

All mutant conditions differed significantly from wild type, confirming that disruption of either TEN-1 or LRON-11 impairs mechanosensory function. However, comparisons among mutant genotypes revealed a graded phenotype: the ten-1 point mutant showed an intermediate defect, whereas lron-11 point mutants and lron-11 loss-of-function animals exhibited more severe reductions that were indistinguishable from each other, and double mutants displayed the strongest impairment, although this defect was not significantly different from that observed in lron-11 mutant backgrounds. Together, these results suggest that TEN-1 and LRON-11 function within a shared pathway, in which disruption of either interaction surface is sufficient to impair touch sensitivity and combined disruption does not produce an additive phenotype. These structure-guided analyses identify specific residues within TEN-1 and LRON-11 as critical determinants of TRN function and support a model in which their interaction is required for proper mechanosensory behavior.

To determine whether LRON-11 is positioned to influence TRN function, we generated a double-transgenic strain (WCH55) carrying an endogenous LRON-11::WrmScarlet reporter together with the TRN marker uIs31 (Pmec-17::gfp). LRON-11::wrmScarlet fluorescence was detected broadly across multiple tissues, and portions of the LRON-11 signal were observed in close spatial proximity to the labeled TRN process (Figure 4H). These images do not establish that LRON-11 is expressed within the TRNs or demonstrate molecular colocalization at the TRN membrane. Nonetheless, this spatial association is consistent with the neuronal requirement revealed by tissue-specific RNAi and with the reduced touch sensitivity observed following disruption of the TEN-1–LRON-11 interaction, supporting a model in which LRON-11 influences mechanosensory function through local interactions with TRNs.

To test whether the behavioral defects of the interface mutants reflect disrupted binding rather than altered protein expression or localization, we used endogenous fluorescent reporters for both proteins. Animals carrying either the TEN-1::mNeonGreen or LRON-11::WrmScarlet reporters exhibited wild-type touch responses (Figure 4G), indicating that insertion of the fluorescent tags does not impair mechanosensory function. Introducing the F1037A/L1039A substitutions into the TEN-1::mNeonGreen reporter and the I42A/L43A/D64A substitutions into the LRON-11::WrmScarlet reporter altered neither the expression level nor the spatial distribution of the corresponding fluorescent fusion proteins (Figure S4A–D), indicating that the mechanosensory defects arise from disruption of the TEN-1–LRON-11 interaction itself.

A conserved receptor-ligand axis supporting distinct neuronal functions

The gut microbiome contributes to neuroprotection in part through metabolites that engage cell-adhesion receptor–mediated pathways, modulating immune responses and neuroinflammation⁴¹. However, the specific receptor–ligand interactions underlying microbiome-driven neuronal protection have remained largely unknown. Our results identify a direct interaction between TEN-1 and the leucine-rich repeat receptor LRON-11, resolve their high-resolution complex structure, and demonstrate that this receptor–ligand pair is required for microbiome-mediated neuroprotection and normal mechanosensory function, establishing a TEN-1–LRON-11 signaling axis that links microbiome-derived environmental cues to the preservation of neuronal integrity.

In vertebrates, teneurins interact with latrophilins, an interaction that has been characterized structurally and functionally^17,21,22,24, whereas C. elegans TEN-1 does not appear to interact with the latrophilin homolog LAT-1, leaving the identity of its physiological binding partner unresolved; our findings identify LRON-11 as a direct TEN-1 ligand and provide a molecular framework for teneurin signaling in this system. Structural analysis revealed that LRON-11 binds to the TTR domain of TEN-1, a region distinct from the β-barrel domain that mediates latrophilin binding in vertebrates¹⁷. Notably, this interface overlaps with asymmetric teneurin homodimerization surfaces previously described in invertebrates³³, raising the possibility that heterophilic receptor binding competes with, or functionally replaces, homophilic teneurin interactions in specific biological contexts. Consistent with this structural model, structure-guided mutations that disrupt the TEN-1–LRON-11 interface impair mechanosensory behavior, demonstrating that this interaction is required for touch receptor neuron function in vivo.

Our findings further suggest that the TEN-1–LRON-11 interaction is conserved across invertebrates. In the accompanying manuscript, we show that mutations designed based on the C. elegans complex structure disrupt the interaction between Drosophila Teneurin (Ten-m) and the LRON-11 ortholog Capricious (Caps), impairing dendrite patterning (see accompanying manuscript). These mutations were carefully designed on the edge of the Ten-m–Caps binding interface to avoid interfering with the almost-overlapping Ten-m–Ten-m dimerization; thus, the dimerization-dependent function of Ten-m was maintained while the Caps-dependent dendrite patterning function of Ten-m was impaired. These results support a conserved role for this receptor-ligand pair in neuronal wiring across species, and suggest that the molecular architecture of the TEN-1–LRON-11 interaction has been conserved despite the diversification of teneurin function during evolution.

An unexpected finding of this study is that the TEN-1–LRON-11 interaction contributes to two distinct biological outputs: microbiome-mediated neuroprotection and mechanosensory behavior. While these processes differ in their physiological context, both depend on the same receptor-ligand interaction, suggesting that TEN-1 and LRON-11 form a conserved signaling axis that can be deployed in distinct neuronal functions. Interestingly, the tissue requirements for this signaling axis differ according to biological context: whereas neuronal TEN-1 and LRON-11 activity is required for normal mechanosensory behavior, microbiome-mediated neuroprotection additionally requires their activity in non-neuronal tissues. These findings point to a broader role for TEN-1-LRON-11 signaling in coordinating communication between neurons and surrounding tissues in response to environmental cues, and understanding how these tissue compartments cooperate to preserve neuronal integrity will be an important direction for future work.

More broadly, our study suggests that teneurins achieve functional diversity by engaging distinct molecular partners across species and biological contexts, and establishes the TEN-1–LRON-11 interaction as a conserved receptor-ligand axis with roles in mechanosensory behavior and microbiome-mediated neuroprotection.

Resource Availability

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Demet Araç (arac@uchicago.edu).

Materials Availability

C. elegans strains and DNA constructs generated in this study are available from the lead contact upon request.

Data and Code Availability

• All data needed to evaluate the conclusions in the paper are present in the paper and/or the supplemental information. Refined structural models and electron density maps have been deposited into the Protein Data Bank (http://www.rcsb.org) (PDB-2IN and PDB-12IM) and EMDB (EMD-76457 and EMD-76456) and are publicly available as of the date of publication. Accession numbers are also listed in the key resources table.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Key Resources Table.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
THE™ His Tag Antibody [iFluor 488], mAb, Mouse	GenScript	GenScript Cat# A01800 RRID: AB3064912
Bacterial and Virus Strains
E. coli Strain OP50	Caenorhabditis Genetics Center (CGC)	Cat# OP50; RRID:WB-STRAIN:WBStrain00041971
E. coli strain HT115	CGC	Cat# HT115; RRID:WB-STRAIN:WBStrain00041079
Chemicals, Peptides, and Recombinant Proteins
Insect-Xpress Cell medium	Lonza	Cat# 04351Q
SF900-III medium	Sigma-Aldrich	Cat# F0926
Cellfectin II reagent	Thermo Fisher	Cat# 10362100
Linearized baculovirus DNA	Expression Systems	Cat# 91-002
Bovine serum albumin (BSA)	Sigma-Aldrich	Cat# A3803
Fetal Bovine Serum (FBS)	Sigma-Aldrich	Cat# F0926
Petri dishes (60 mm)	Corning	Cat# 353002
Agar	Lab Scientific	Cat# A466
Gentamicin	Gibco	Cat# 15750-078
Critical Commercial Assays
Ni-NTA Agarose	QIAGEN	Cat# 30250
Superdex 200 Increase 10/300	Cytiva	Cat# 28990944
Superose 6 Increase10/300	Cytiva	Cat# 29091596
SYPRO Orange dye	Thermo Fisher	Cat# 4461146
Streptavidin-coated SPR chip	Cytiva	Cat# BR100531
Deposited Data
TEN-1–LRON-11 model	This study	PDB 12IN
TEN-1 model	This study	PDB 12IM
TEN-1–LRON-11 cryo-EM map	This study	EMD-76457
TEN-1 cryo-EM map	This study	EMD-76456
Experimental Models: Cell Lines
Sf9 cells	Thermo Fisher	Cat# 12659017 RRID:CVCL_0549
High Five cells; Trichoplusia ni	Thermo Fisher	Cat# B855-02 RRID:CVCL_C190
S2(Drosophila Schneider 2)	Expression System	Cat# 94-005F RRID:CVCL_Z232
Experimental Models: C. elegans
Wild-type Bristol N2	CGC	N2; RRID: WB-STRAIN:WBStrain00000001
ten-1(qy56[ten-1::mNG+loxP]) III	CGC	NK2502
Y71F9B.8(ok2333) I	CGC	RB1799 RRID:WB-STRAIN:WBStrain00032490
ten-1 (syb7771[ten-1_F1037A,L1039A])	This study (SunyBiotech)	PHX7771
ten-1(syb10574[ten-1_F1037A,L1039A]);NK2502 ten-1(qy56[ten-1::mNG+loxP]) III.	This study (SunyBiotech)	PHX10574
lron-11(syb7811[lron-11_I42A,L43A,D64A])	This study (SunyBiotech)	PHX7811
lron-11(syb10562[lron-11::3xLinker::WrmScarlet::linker::3xFLAG])	This study (SunyBiotech)	PHX10562
lron-11(syb10910[lron-11_I422A,L43A,D64A]; syb10562[lron-11::3xLinker::WrmScarlet::linker::3xFLAG])	This study (SunyBiotech)	PHX10910
ten-1(syb7771[ten-1_F1037A,L1039A]); lron-11(syb7811[lron-11_I42A,L43A, D64A])	This study	DEA20
mec-4 (u253) X	CGC	TU253 RRID:WB-STRAIN:WBStrain00035037
uIs31[Pmec-17mec-17::gfp]; mec-4d[e1611])	A gift from Martin Chalfie	TU2773 RRID:WB-STRAIN:WBStrain00045615
uIs31[Pmec-17mec-17::gfp]; lron-11(syb10562[lron-11::3xLinker::WrmScarlet::linker::3xFLAG])	This study	WCH55
uIs71[Pmec-18sid-1;Pmyo-2mcherry], uIs31[Pmec-17mec-17::gfp], sid-1[pk3321], mec-4d[e1611])	Previously made in A. Calixto lab43	WCH6
Oligonucleotides
ten-1e fwd: tcgttcttccgaatggtcat	This study	N/A
ten-1e rev: cagaactctatcaaaccgcc	This study	N/A
lron-11a fwd: tagccgaattgttatgttca	This study	N/A
lron-11a fwd: ccctattctttccactcaac	This study	N/A
Recombinant DNA
C. elegans TEN-1-LRON-11 fusion protein cloned onto a modified pAcGP67A vector suitable for expression in insect cells	This study	N/A
C. elegans TEN-1 ECR cloned onto a modified pAcGP67A vector suitable for expression in insect cells	This study	N/A
C. elegans LRON-11 ECR cloned onto a modified pAcGP67A vector suitable for expression in insect cells	This study	N/A
Software and Algorithms
MotionCor2	Li et al. 44	RRID:SCR_016499
CTFFIND4	Mindell and Grigorieff 45	RRID:SCR_016732
RELION 3.1	Zivanov et al.46	RRID:SCR_016274
Phenix	Adams et al.47	RRID:SCR_014224
Coot	Emsley et al. 48	RRID:SCR_014222
PyMOL	Schrödinger, LLC	RRID:SCR_000305
UCSF Chimera	Pettersen et al.49	RRID:SCR_004097
ConSurf server	Ashkenazy et al.34	RRID:SCR_002320
PISA	Krissinel et al.35	RRID:SCR_015749
Jalview	Waterhouse et al.42	RRID:SCR_006459
GraphPad Prism	GraphPad Software	RRID:SCR_002798
ImageJ	Schindelin et al.50	RRID:SCR_003070
Other
Quantifoil R 1.2/1.3 300 Mesh Cu grids	Quantifoil	https://www.emsdiasum.com/quantifoil-r1213-300-mesh-copper
Vitrobot Mark IV	FEI-Thermo Fisher	https://www.fei.com/

STAR Methods

Experimental model and Study Participant Details

Animals

C. elegans hermaphrodite animals were grown at 20°C or 25°C on nematode growth media (NGM) plates supplied with E. coli OP50 (CGC, WBStrain00041971) or E. coli HT115 (CGC, :WBStrain00041079) as food source⁵¹. Genotypes for the point mutants lacking fluorescent reporters were confirmed via sanger sequencing (sequencing primer ten-1e (syb7771 = seq-s: tcgttcttccgaatggtcat, seq-a: cagaactctatcaaaccgcc; sequencing primer lron-11a (syb7811) = seq-s: tagccgaattgttatgttca, seq-a: ccctattctttccactcaac) and were independently verified by SunyBiotech upon generation. Genetic crosses were carried out using standard C. elegans procedures and progeny were screened and verified via sanger sequencing unless they could be identified by fluorophore expression such as WCH55. Synchronized L4 larvae and young adult hermaphrodite animals were used for all behavioral experiments. Synchronized animals (24, 48 or 72 hours post-hypochlorite treatment (all started as embryos) animals were used for morphological integrity assays. All strains used or generated for this study are listed in Key Resources Table (TEN-1 amino acid numbers are given for isoform e; LRON-11 amino acid numbers are given for isoform a).

Cell lines

Sf9 (Spodoptera frugiperda) insect cells were maintained in Sf-900 III SFM (Sigma-Aldrich) at 27°C with shaking and were used for baculovirus generation. High Five (Trichoplusia ni, ovarian) insect cells were cultured in Insect-XPRESS medium (Lonza) supplemented with 10 μg/mL gentamicin at 27°C with shaking and were used for recombinant protein expression. Schneider 2 (S2; Drosophila melanogaster) cells were maintained at 27°C with shaking and used for the expression of recombinant proteins for extracellular interactome assay (ECIA).

Microbe Strains

The OP50 and HT115 E. coli strains are used.

Method Details

Cloning and expression in insect cells

For structural studies, a fusion construct was generated in which Caenorhabditis elegans TEN-1 (Uniprot, H1ZUX0_CAEEL) ECR (residues R391-E2837) was linked to the C-terminus of LRON-11 (Uniprot, G5EGE7_CAEEL) ECR (residues G21-S405) via a flexible (GGS)×10 peptide linker. The coding sequences of LRON-11 and TEN-1 were joined in-frame using standard molecular cloning techniques, with the (GGS)×10 linker encoded by the nucleotide sequence GGTGGTTCAGGTGGTTCAGGCGGTTCCGGTGGTTCCGGCGGCTCTGGTGGTTCAGGAGGCTCAGGTGGCTCTGGTGGTTCCGGTGGTTCA, which was introduced through PCR amplification and overlap extension. The fusion gene was subsequently subcloned into the expression vector pAcGP67A using Gibson assembly, with 6×His-tag at C-terminal. Sf9 cells (Thermo Fisher, 12659017) were co-transfected with linearized baculovirus DNA (Expression Systems, 91–002) and the constructed transfer plasmid using the Cellfectin II (Thermo Fisher, 10362100) transfection reagent. Following transfection, initial baculoviral stocks were harvested and Sf9 cells were used to produce high-titer recombinant baculovirus. Baculoviruses were amplified in Sf9 cells in SF900-III medium containing 10% (v/v) fetal bovine serum (Sigma-Aldrich, F0926).

Large-scale protein expression was performed by infection of High Five cells (Trichoplusia ni, Thermo Fisher, B855–02) in Insect-Xpress medium (Lonza, 04351Q) supplemented with 10 μg/mL gentamicin at a cell density of 2.0×10⁶ cells/ml for 72 hours at 27°C. 72 h after viral infection of High Five cells, the medium containing secreted glycosylated proteins was collected and centrifuged at 900 g for 15 min at room temperature. After centrifugation, the supernatant was transferred into a beaker. The following were added to the supernatant (to final concentrations): 50 mM Tris pH 8.0, 5 mM CaCl₂ and 1 mM NiCl₂ and stirred for 30 min. After centrifugation at 8,000 g for 30 min, the supernatant was collected and incubated with nickel-nitrilotriacetic agarose resin (QIAGEN 30250) for 3 h at room temperature. The resin was transferred to a glass Buchner funnel and rinsed with HBS buffer containing 20 mM imidazole, then transferred to a poly-prep chromatography column (Bio-rad 7311550). Target protein was eluted with HBS buffer containing 200 mM imidazole and then applied to size-exclusion chromatography (Superose 6 Increase 10/300 column; GE Healthcare), purified in a final buffer comprised of 25 mM Tris pH 8.0, 150 mM NaCl.

For the Surface Plasmon Resonance assays, TEN-1 ECR (residues M911-E2837 in TEN-1e isoform, called in the text as TEN-1L) was cloned with C-terminal 6×His and Avi tags and purified as described above.

The extracellular interactome assay (ECIA)

The ectodomains of TEN-1 and LRON-11 were cloned into ECIA expression plasmids to generate both Fc-tagged bait constructs (TEN-1) and AP-tagged prey constructs (LRON-11). These constructs were subsequently expressed in Drosophila S2 cells following standard protocols. Bait capture and prey binding were detected colorimetrically using chromogenic alkaline phosphatase substrates, with absorbance measured at λ = 650 nm.

The ECIA was performed essentially as previously described²⁸. The assay’s high sensitivity is critically dependent on two design features: the high-density display of the Fc-tagged bait protein on a solid plastic support, and the oligomerization of the AP-tagged prey achieved through a pentameric coiled-coil domain. This combined approach significantly enhances avidity, thereby increasing the sensitivity of the interaction detection^28,52 . For quantitative binding assessments involving engineered TEN-1 and LRON-11 constructs, precise normalization of protein expression levels was crucial. Expression levels for each mutant series and their corresponding wild-type versions were quantified via Western blot analysis. This was performed using an anti-His-Tag Antibody coupled with iFlour 488 (Genscript, A01800, 1:5000 dilution), targeting the 6xHis tags engineered at the C-termini of all constructs. This blotting also confirmed the successful expression and secretion of the mutant ectodomains (Figure S3A). Fluorescent signal quantification was performed using a Biorad Chemidoc system and ImageLab version 6 (Biorad). Normalization of protein quantities across samples was achieved by diluting higher-expressing variants with fresh Schneider’s culture media to match the expression levels of lower-expressing variants.

Cryo-EM data acquisition

3.5 μl purified LRON-11–TEN-1 full-ectodomain complex (0.45 mg/ml) was applied on glow-discharged holey carbon grids (Quantifoil R1.2/1.3, 300 mesh), and vitrified using a Vitrobot Mark IV (FEI Company). The specimen was visualized using a Titan Krios electron microscope (FEI) operating at 300 kV and equipped with a K3 direct electron detector (Gatan, Inc). Images were recorded with a nominal magnification of 81,000x in super-resolution counting mode, corresponding to a pixel size of 0.5325 Å on the specimen level. To maximize data collection speed while keeping image aberrations minimal, image shift was used as imaging strategy using one data position per hole with four holes targeted in one template with one focus position. In total, 4,535 images with defocus values in the range of −1 to −2.5 μm were recorded. Images were recorded in 40 subframes to give a total dose of roughly 60 electrons per Ų. Data collection and processing details are provided in Figure S2. All structural depictions were prepared using PyMOL v. 3.1.6 or Chimera v. 1.11.

Image processing and 3D reconstruction

Stack images were subjected to beam-induced motion correction using MotionCor⁴⁴. CTF parameters for each micrograph were determined by CTFFIND4⁴⁵. In total, 4,965,253 particle projections were selected using automated particle picking and subjected to reference-free two-dimensional classification to discard false-positive particles or particles categorized in poorly defined classes, resulting in 4,234,550 particle projections for further processing. Initial 3D maximum-likelihood-based classification was performed on a binned dataset (Bin4; 4.26 Å/pixel) using a previously reported structure of TEN2 (PDB: 6FB3) as the reference model using RELION-3⁴⁶. The detailed data processing workflow is presented in Supplementary Figures. Briefly, 1,397,007 particles exhibiting well-defined density for teneurin were selected after initial 3D classification rounds. A second round of 3D classification (Bin2) further segregated the dataset into two major classes: teneurin with Domain 3 and teneurin lacking Domain 3. Particles from the class lacking Domain 3 underwent an additional 3D classification to identify the most homogeneous subsets, which were subsequently refined using iterative 3D auto-refinement, Bayesian polishing, and CTF refinement, yielding a final map at 2.6 Å resolution. Particles from the class containing Domain 3 were re-extracted using a box size of 288 and subjected to further 3D classification, during which the TEN-1–LRON-11 complex was identified. These particles were also refined through the same pipeline, resulting in final reconstructions at 2.6 Å for TEN-1 with Domain 3 and 3.4 Å for the TEN-1–LRON-11 complex (Figure S2). The atomic model for the complex was refined against the experimental and sharpened maps using AlphaFold2 models ⁵³ for TEN-1 and LRON-11 as starting models, followed by manual building and regularization in Coot ⁵⁴ and automated real-space refinement in Phenix ⁵⁵. The map output by RELION were deposited to the EMDB (Electron Microscopy Data Bank).

Binding interface analysis and design of interface mutations

To identify mutations that specifically disrupt the TEN-1–LRON-11 interaction, the binding interface of the cryo-EM structure was analyzed with the ConSurf server to identify conserved surface residues and with PDBePISA to analyze all amino acids at the binding interface. These analyses were combined with visual inspection of the binding interface, in which we considered the position of the residues within the binding interface, the secondary structure elements they belong to (helix, strand or loop), the identity of the amino acid and whether they are involved in any hydrogen bonds or salt bridges. Single- and multiple-point mutations selected in this way were introduced into Fc-tagged TEN-1 and AP-tagged LRON-11 full ECR constructs and tested by ECIA and SPR.

Surface Plasmon Resonance

For SPR experiments, we created baculoviruses for expressing TEN-1 ECR with a C-terminal Avi-tag followed by a His-tag. Purified WT and point mutants of TEN-1 ECR were biotinylated using E. coli BirA biotin ligase and immobilized on streptavidin-coated SPR chip (Cytiva, Series S Sensor Chip SA, BR100531). Purified (unbiotinylated) WT and mutants of LRON-11 ECR and LAT-1 ECR were used as analytes at concentrations from 0.625 μM to 160 μM. Binding curves were obtained using a Biacore 1K (Cytiva) with HBS with 0.05% Tween-20 and 0.1% BSA as running buffer. The K_D values were determined using steady-state affinity analysis in Biacore Evaluation software (Cytiva). Binding data were fit by nonlinear regression using a one-site model that accounts for both total and nonspecific binding.

Differential Scanning Fluorimetry

Protein thermal stability was assessed using differential scanning fluorimetry (DSF), also known as a thermal shift assay, as previously described⁵⁶. Purified proteins were diluted to final concentrations in a final buffer comprised of 25 mM Tris pH 8.0, 150 mM NaCl. SYPRO Orange dye (Thermo Fisher Scientific, 4461146) was added at a final 1X concentration. Each reaction had a final volume of 20 μl. Reactions were assembled on ice in 96-well PCR plates, sealed with optically clear adhesive films, and centrifuged briefly at 4°C to remove air bubbles. Non-protein control reactions containing dye and buffer were included to account for background fluorescence. All measurements were performed in triplicate.

Fluorescence was monitored using a real-time PCR instrument while the temperature was increased from 4°C to 95°C, with fluorescence readings collected at multiple points per degree Celsius. Protein unfolding was detected as an increase in fluorescence due to binding of SYPRO Orange to exposed hydrophobic regions. Fluorescence–temperature curves exhibited a sigmoidal transition characteristic of cooperative unfolding.

C. elegans maintenance and growth

Wild-type (N2) and mutant strains TU2773 (uIs31[P_mec-17mec-17::gfp]; mec-4d[e1611]), WCH6 (uIs71[P_mec-18sid-1;Pmyo-2mcherry], uIs31[Pmec-17mec-17::gfp], sid-1[pk3321], TU3335 [lin-15B(n744) X; uIs57 [unc-119p::YFP + unc-119p::sid-1 + mec-6p::mec-6], NR350 [rde-1(ne219) V; kzIs20 [hlh-1p::rde-1 + sur-5p::NLS::GFP], IG1839 [frSi17 II (mtl- 2p::rde-1 3’UTR) II; frIs7 (nlp-29p::GFP + col-12p::DsRed IV); rde-1(ne300)] , WCH55 [uIs31[P_mec-17mec-17::gfp]; lron-11::wrmScarlet] and strains generated via CRISPR/Cas9 gene editing were grown at 20°C. All nematode strains are maintained on E. coli OP50 strain prior to feeding with other bacteria. Unless otherwise noted, all plates were incubated at 20°C.

C. elegans strains generated via CRISPR/Cas9 gene editing

CRISPR/Cas9 genome editing was performed by SunyBiotech following standard procedures⁵⁷. The ten-1 endogenous reporter [ten-1(qy56[ten-1::mNG+loxP]) (NK2502) was ordered from CGC. The original description states it was generated via CRISPR/Cas9 by inserting 18 amino acid flexible linker (6×glycine-alanine-serine) and mNeonGreen immediately upstream of the endogenous ten-1 stop codon, producing a C-terminal TEN-1::mNeonGreen fusion protein. A loxP site (a short DNA sequence recognized by Cre recombinase) was left after the stop codon. The endogenous lron-11 reporter allele syb10562 (PHX10562) was generated by inserting 3xLinker::WrmScarlet::linker::3xFLAG immediately before the stop codon in the lron-11 locus. The syb10574 ten-1 (F1037A, L1039A) point mutations (PHX10574) was generated via CRISPR/Cas9 by mutating the tagged ten-1 strain; and the syb10910 lron-11(I42A, L43A, D64A) point mutation (PHX10910) was generated by CRISPR/Cas9 by mutating the tagged lron-11 strain.

Gentle touch assay

Mechanosensory responses were assessed using a gentle touch assay as previously described³⁸. Briefly, animals were stimulated by stroking the head and the tail using an eyelash hair mounted on a toothpick. Each animal was subjected to 10 consecutive touches in a head-to-tail fashion, and responses were scored based on the presence or absence of backward (head touch) or forward (tail touch) movement upon stimulation. Animals capable of movement but defective in gentle touch responses were scored as touch insensitive. All experiments were performed in a blinded manner. For each strain, 10 animals were analyzed per biological replicate, with at least three independent biological replicates derived from distinct hermaphrodite progenitors and conducted on different days. As controls, wild-type animals and mec-4 loss-of-function mutants were included and scored under the same blinded conditions. All assays were performed at 20°C.

Imaging and fluorescence quantification

Nematodes were mounted on ~100 μm-thick agarose pads, immobilized with levamisole, and covered with #1 square coverslips. Imaging was performed using a spectral confocal Stellaris 5 inverted microscope (Leica) equipped with a 40× 1.3 NA and 63× 1.4 NA oil immersion objectives. Fluorescence signals were acquired by excitation with 488 nm and 561 nm laser lines for mNeonGreen (or EGFP) and WrmScarlet, respectively. Emission was collected using spectral windows of 493–530 nm (green channel) and 563–650 nm (red channel). Fluorescence quantification was performed using ImageJ/FIJI. Acquisition was made by taking 1024×1024 px² images and z-stacks spanning 10–30 μm in total thickness, separated by 0.3–2 μm intervals. Pictures were generated as maximum intensity projections along the z-axis from z-stacks, with the exception of strain WCH55, for which a single confocal slice was compared to the total z-stack.

Strain WCH6 was used to image TRN morphology expressing cytoplasmic EGFP (excitation at 488 nm). Strains NK2502, syb10574, and N2 were imaged using 488 nm excitation to detect mNeonGreen-tagged TEN-1 variants and assess autofluorescence background. Strains syb10562 and syb10910 were imaged using 561 nm excitation to detect WrmScarlet-tagged LRON-11 variants. Strain DEA11 was imaged using both 488 nm and 561 nm excitation to simultaneously detect mNeonGreen and WrmScarlet signals.

RNAi treatment

RNA interference (RNAi) was performed as previously described². E. coli HT115 bacteria expressing dsRNA targeting ten-1 or lron-11 were thawed from glycerol stocks and streaked onto LB agar plates supplemented with tetracycline (15 μg/mL) and carbenicillin (25 μg/mL). Control HT115 bacteria carrying the empty vector were streaked on LB plates supplemented with tetracycline only. Plates were incubated overnight at 37°C. The following day, bacterial lawns were inoculated into liquid LB medium without antibiotics and grown overnight at 37°C with shaking. For double RNAi experiments, equal volumes of ten-1 and lron-11 bacterial cultures were mixed prior to seeding. Cultures were then used to seed NGM agar plates supplemented with IPTG (1 mM) and carbenicillin (25 μg/mL) by spreading approximately 300 μL of bacterial culture per plate. Seeded plates were allowed to dry overnight at room temperature to induce dsRNA expression. The next day, 50–100 embryos of the WCH6 strain were transferred onto RNAi plates and allowed to develop at 20°C.

AVM touch receptor neuron integrity was assessed 72 hours after plating. For RNAi experiments during development, neuronal integrity was assessed at 24, 48 and 72 hours post hypochlorite treatment to release embryos. Experiments were performed in three independent biological replicates. Each replicate consisted of three plates per condition, with 30 animals per plate (90 animals per replicate), yielding a minimum total of 270 animals analyzed per condition.

Neuronal integrity was scored as previously described^2,58, and the functional category corresponding to wild-type axons was quantified. The wild-type axon of the AVM touch receptor neuron extends from the mid-body to the anterior region of the animal, reaching the nerve ring. Animals displaying shorter, broken, or absent axons, or lacking identifiable neuronal somas, were classified as degenerated.

Quantification and Statistical Analysis

Statistical analyses of data were performed using GraphPad Prism 10, and data are presented as mean ± SEM. For the neuronal integrity and behavioral assays, at least three independent biological replicates were performed for each condition; each biological replicate consisted of three technical replicates, with at least 30 animals scored per biological replicate. Unless otherwise indicated, comparisons were made against the corresponding control condition, and the statistical test used for each panel is indicated in the figure legends. Quantification methods and tools used are described in each relevant section of the methods or figure legends.

Supplementary Material

Supplemental Information

Document S1. Figures S1–S4, Table S1, and supplemental references

Highlights.

LRON-11 is a newly identified ligand of the C. elegans teneurin TEN-1

Cryo-EM structure of the TEN-1–LRON-11 complex reveals a distinct structural assembly

ten-1 and lron-11 are both required for microbiota-induced neuroprotection

The TEN-1-LRON-11 interaction is required for mechanosensory function