1. Neuroscience

TIR-1/SARM1 inhibits axon regeneration and promotes axon degeneration

  1. Victoria L Czech
  2. Lauren C O'Connor
  3. Brendan Philippon
  4. Emily Norman
  5. Alexandra B Byrne  Is a corresponding author
  1. Department of Neurobiology, UMass Chan Massachusetts Medical School, United States
https://doi.org/10.7554/eLife.80856
Share this article
Cite this article
  1. Victoria L Czech
  2. Lauren C O'Connor
  3. Brendan Philippon
  4. Emily Norman
  5. Alexandra B Byrne
(2023)
TIR-1/SARM1 inhibits axon regeneration and promotes axon degeneration
eLife 12:e80856.
https://doi.org/10.7554/eLife.80856
  2. Download BibTeX
  3. Download .RIS

Abstract

Growth and destruction are central components of the neuronal injury response. Injured axons that are capable of repair, including axons in the mammalian peripheral nervous system and in many invertebrate animals, often regenerate and degenerate on either side of the injury. Here we show that TIR-1/dSarm/SARM1, a key regulator of axon degeneration, also inhibits regeneration of injured motor axons. The increased regeneration in tir-1 mutants is not a secondary consequence of its effects on degeneration, nor is it determined by the NADase activity of TIR-1. Rather, we found that TIR-1 functions cell-autonomously to regulate each of the seemingly opposite processes through distinct interactions with two MAP kinase pathways. On one side of the injury, TIR-1 inhibits axon regeneration by activating the NSY-1/ASK1 MAPK signaling cascade, while on the other side of the injury, TIR-1 simultaneously promotes axon degeneration by interacting with the DLK-1 mitogen-activated protein kinase (MAPK) signaling cascade. In parallel, we found that the ability to cell-intrinsically inhibit axon regeneration is conserved in human SARM1. Our finding that TIR-1/SARM1 regulates axon regeneration provides critical insight into how axons coordinate a multidimensional response to injury, consequently informing approaches to manipulate the response toward repair.

Editor's evaluation

This important study reveals a new role for TIR-1/SARM1 in inhibiting both axonal regeneration and promoting axon degeneration in a novel axonal injury model. The authors provide solid evidence to support their conclusions that TIR-1 mediates these processes by acting via distinct signaling pathways. This work will be of general interest to developmental neuroscientists.

https://doi.org/10.7554/eLife.80856.sa0

Introduction

In many injured neurons that are capable of repair, axons are broken into fragments with different fates. The axon segment that remains attached to the cell body regenerates and severed axon segments degenerate. However, in the central nervous system and in aged animals, regeneration of the proximal axon fragment is inhibited, resulting in a permanent loss of neuronal function (Cai et al., 2001; David and Aguayo, 1981). Understanding the molecular mechanisms that regulate both axon regeneration and degeneration is critical to developing therapeutic approaches to repair the injured nervous system.

On the proximal side of the injury, functional axon regeneration requires the synchronized execution of multiple processes, including growth cone formation, axon guidance, and synapse formation. Despite being recapitulations of well-characterized developmental processes, the precise molecular mechanisms that regulate axon regeneration remain incompletely understood. Those that have been identified are subdivided into two broad categories: intrinsic and extrinsic. Intrinsic regulators of transcription, transport, translation, and signaling are complemented and opposed by extrinsic regulators of myelin-dependent inhibition and scar formation (He and Jin, 2016; Kaplan et al., 2015; Mahar and Cavalli, 2018; Schwab and Strittmatter, 2014; Silver et al., 2014; Sutherland and Geoffroy, 2020). Manipulating either its intrinsic or extrinsic environment is sufficient to modestly increase the ability of an injured axon to regenerate (Curcio and Bradke, 2018; He and Jin, 2016; Schwab and Strittmatter, 2014; Tedeschi and Bradke, 2017; Tran et al., 2018).

On the distal side of the injury, the most notable endogenous regulator of axon degeneration is Sterile Alpha and toll/interleukin-1 resistance (TIR) motif containing 1 (SARM1/dSarm), which is essential for injury-induced axon degeneration of mouse and fly axons (Gerdts et al., 2013; Osterloh et al., 2012). SARM1/dSarm promotes axon degeneration cell-autonomously by depleting NAD+, interacting with the BTB/BACK domain protein Axundead, and interacting with mitogen-activated protein kinase (MAPK) signaling cascades, which in turn regulate calpain activation, SCG10/Stathmin-2 proteolysis, mitochondrial dysfunction, cytoskeletal disruption, and axon fragmentation (Coleman and Höke, 2020; Essuman et al., 2017; Gerdts et al., 2015; Gerdts et al., 2016; Neukomm et al., 2017; Shin et al., 2012; Summers et al., 2020; Walker et al., 2017; Yang et al., 2015).

The Caenorhabditis elegans SARM1/dSarm homolog, TIR-1, was first characterized as a determinant of neuronal asymmetry and the innate immune response, which it regulates in coordination with the NSY-1/ASK1 MAPK signaling pathway (Chuang and Bargmann, 2005; Couillault et al., 2004; Liberati et al., 2004). In the absence of injury, TIR-1 has been found to regulate neurodegeneration in response to disease and age, upon multimerization, and in the absence of mitochondria (Ding et al., 2022; Lezi et al., 2018; Loring et al., 2021; Vérièpe et al., 2015); however, whether TIR-1 regulates injury-induced axon degeneration is not known.

Here we report that in addition to promoting injury-induced axon degeneration, TIR-1/SARM1 functions cell-autonomously to inhibit axon regeneration. Specifically, we find that loss of TIR-1 increases axon regeneration and that the ability to inhibit axon regeneration cell-intrinsically is conserved in human SARM1. In response to injury, TIR-1 functions with the NSY-1/ASK1 MAPK signaling cascade and the terminal transcription factor DAF-19/RFX to inhibit regeneration of the proximal axon fragment. In parallel, TIR-1 functions independently from NSY-1/ASK1 signaling and with the DLK-1/DLK MAPK pathway to promote degeneration of the distal axon fragment. Together, our results indicate that TIR-1 inhibits axon regeneration cell-autonomously, and that it does so by regulating a downstream pathway that is separable from its role in degeneration. Consequently, our results reveal a mechanism to shift the injury response toward repair.

Results

TIR-1/SARM1 inhibits axon regeneration

In a candidate screen for genes that function on both sides of an injured axon, we asked whether tir-1, the highly conserved C. elegans homolog of Sarm1/dSarm (Mink et al., 2001), regulates the injury response (Figure 1A). To do so, we severed GABA motor neurons in tir-1 loss-of-function and wild-type animals with a pulsed laser and compared their morphology 24 hr later. We tested two putative null alleles of tir-1; the qd4 allele is a 1078 base pair deletion spanning exons 8–10 and the tm3036 allele is a smaller 269 base pair deletion in exon 8 that results in an early stop codon (Pujol et al., 2008; Shivers et al., 2009). Both the qd4 and tm3036 deletions disrupt the C-terminal toll/interleukin-1 receptor (TIR) domain, which is essential for TIR-1 function (Chuang and Bargmann, 2005; Figure 1A). Strikingly, we found that when GABA motor axons of tir-1 loss-of-function and wild-type animals were severed at the midline, significantly more axons regenerated in the absence of tir-1 (Figure 1B–D). These data reveal that TIR-1 inhibits motor axon regeneration.

Figure 1 with 1 supplement see all
Download asset Open asset
TIR-1 inhibits axon regeneration.

(A) Human SARM1, fly dSarm, and worm TIR-1 contain an N-terminal auto-inhibitory domain, two sterile alpha motif (SAM) domains and a toll-interleukin-1 receptor (TIR) domain. Null alleles qd4 and tm3036 disrupt the essential TIR domain. (B) C. elegans GABA motor neurons are axotomized to study the injury response in vivo and with single-axon resolution. (C) Representative GABA motor neurons before and 24 hr after single laser injury at the lateral midline of late larval stage (L4) animals (dashed line). Asterisks indicate distal fragments, arrowheads indicate retraction bulbs, and arrows indicate growth cones. Scale bar: 10 μm. (D) More axons regenerate in tir-1 mutants compared to wild-type animals, as quantified by the percentage of axons that form a growth cone 24 hr after laser surgery. N = 152, 99, 60. (E) Regeneration phenotypes were categorized according to whether regenerating axons reached landmarks M−, M, M+, F, or failed to form a growth cone, N. (F) More tir-1(-) axons initiate regeneration and reach the dorsal cord compared to wild-type axons, as indicated by a significant reduction in the number of axons that failed to form a growth cone (N) and a significant increase in the number of fully regenerated axons (Full). N = 158, 154. Significance relative to wild type is indicated by *p≤0.05, ***p≤0.001, Fisher’s exact test. Error bars represent 95% confidence intervals.

C. elegans have one other TIR domain-containing protein, which is TOL-1 (Liberati et al., 2004). To determine whether regulation of axon regeneration is a general function of TIR domain-containing proteins, we compared regeneration in the presence and absence of tol-1. We found no significant difference between the number of axons that regenerated in tol-1(nr2033) null mutants and wild-type animals (Figure 1—figure supplement 1). Therefore, TIR-1 is a TIR-domain-containing protein with a specific role in axon regeneration.

Successful regeneration of C. elegans motor axons includes growth cone formation, axon extension, and responsiveness to cues that guide the axon to the dorsal nerve cord. To determine which stage of the regenerative process is inhibited by TIR-1, we grouped individual injured axons into categories depending on whether they initiated a growth cone and how far they regenerated toward the dorsal nerve cord (Figure 1E and F). We found a dramatic decrease in the number of axons that did not initiate regeneration (N) and an increase in the number of axons that regenerated the full distance to the dorsal nerve cord (Full) in tir-1 mutants compared to wild-type animals (Figure 1F). Of note, C. elegans are capable of repairing injured mechanosensory axons both by regenerating and fusing severed proximal and distal axon fragments (Ghosh-Roy et al., 2010; Neumann et al., 2011). However, we did not observe fusion of severed motor axon fragments in either wild-type or tir-1 mutant animals. Together, these results indicate TIR-1 inhibits the initiation of axon regeneration and the successful extension of regenerating axons to their targets.

TIR-1 functions cell-autonomously in GABA motor neurons to regulate axon regeneration

We next asked how TIR-1 regulates axon regeneration and how it achieves specificity of function. TIR-1 functions in multiple cell types, including the nervous system, intestine, and epidermis, where it regulates neuronal development and the innate immune response (Chuang and Bargmann, 2005; Couillault et al., 2004; Liberati et al., 2004; Pujol et al., 2008). To determine where TIR-1 functions to regulate axon regeneration, we first CRISPR tagged the endogenous tir-1 locus with mCherry and analyzed its expression pattern. With this single-copy integration of mCherry, we observed faint TIR-1 expression in the nervous system, particularly in the nerve ring, select cells in the head and tail, and in the dorsal cord and ventral cord, as well as in the intestine (Figure 2A). While tagging the endogenous locus is perhaps the most accurate in vivo representation of a gene’s expression pattern, cells that express the gene at low levels may not be detected. Therefore, we also constructed a transgenic strain expressing multiple copies of tir-1b::mCherry under the endogenous tir-1 promoter. The tir-1b isoform encodes the SAM and TIR domains but not the autoinhibitory N-terminus of TIR-1, and is predicted to be an activated form of TIR-1 (Chuang and Bargmann, 2005). With this strain, we observed a similar pattern of TIR-1 expression in the intestine and nervous system. However, the increased copy number revealed TIR-1b::mCherry expression in the cell bodies and axons of GABA motor neurons (Figure 2B). The finding that TIR-1 is expressed in GABA motor neurons, albeit at relatively low levels, raises the possibility that TIR-1 could function cell-intrinsically to regulate axon regeneration.

Figure 2
Download asset Open asset
TIR-1 functions cell-autonomously to inhibit regeneration.

(A) Endogenous TIR-1 is expressed faintly in the nervous system, particularly in neurons along the dorsal and ventral (enlarged box) nerve cords. Note the intestinal expression significantly overlaps with autofluorescent gut granules and therefore is not an accurate representation of endogenous tir-1 expression in the intestine. Scale bar:25 μm. (B) Representative micrographs of transgenic animals expressing tir-1b::mCherry under its endogenous promoter indicates TIR-1b::mCherry is expressed in GABA neurons. Enlarged boxes show TIR-1b::mCherry expression in the cell bodies of GABA neurons. Scale bars: upper images, 25 μm; insets, 10 μm. (C,D) When expressed specifically in GABA neurons, mScarlet::tir-1b is expressed in a punctate pattern in the cell bodies and axons of GABA neurons before, immediately after and 7 hr after injury. Scale bars: 5 μm. (E) GABA-specific expression of tir-1b rescues axon regeneration in tir-1(qd4) animals but not wild-type animals. N = 197, 148, 62, 50, 69. (F) GABA-specific expression of human SARM1SAM-TIR lacking the N-terminal autoinhibitory domain rescues axon regeneration in single cut tir-1(qd4) mutants and not in wild-type animals. N = 91, 60, 53, 66, 50. Significance relative to wild type or tir-1(qd4) is indicated by *p≤0.05, **p≤0.01, ***p≤0.001, Fisher’s exact test. Error bars represent 95% confidence intervals.

We investigated where TIR-1 functions to regulate regeneration by asking whether tissue-specific expression of tir-1b cDNA is sufficient to inhibit axon regeneration of injured GABA motor neurons. We expressed a GABA-specific unc-47p::mScarlet::tir-1b construct in tir-1(qd4) animals and observed punctate mScarlet::TIR-1b in the cell bodies and axons of GABA neurons, both before and after injury (Figure 2C and D). Axon regeneration was fully suppressed in animals that expressed tir-1b specifically in their GABA neurons compared to tir-1(qd4) animals (Figure 2E). Therefore, cell-intrinsic tir-1 function inhibits axon regeneration.

The significance of the finding that tir-1 regulates regeneration is influenced by whether its function is conserved in homologous Sarm genes. TIR-1 and human SARM1 proteins are highly conserved, sharing 60% homology over 85% of the SARM1 sequence (Altschul et al., 1997; Mink et al., 2001). As a result, both TIR-1 and human SARM1 contain multiple HEAT/Armadillo repeats in the autoinhibitory N-terminus followed by two sterile alpha motifs (SAM) and one toll-like interleukin repeat (TIR) domain (Figure 1A). We investigated whether intrinsic inhibition of axon regeneration is a conserved function of TIR-1 by asking whether motor neuron-specific expression of human SARM1 is also capable of inhibiting axon regeneration. As we observed with tir-1, expressing an active form of human Sarm1 containing its SAM and TIR domains (hSarm1SAMTIR) in GABA motor neurons was sufficient to rescue the regeneration phenotype of tir-1(qd4) mutants (Figure 2F). Together, these results demonstrate that the ability to inhibit axon regeneration cell-autonomously is conserved in the human SARM1SAMTIR protein.

A novel model of injury-induced motor axon degeneration in C. elegans

The finding that TIR-1 and human SARM1 can function intrinsically to inhibit axon regeneration is unexpected since Sarm1 and dSarm promote the opposite response to injury, axon degeneration, in mice and flies (Gerdts et al., 2013; Osterloh et al., 2012). Whether TIR-1 also promotes injury-induced motor axon degeneration in C. elegans is not known. To answer that question, we needed to examine the consequences of mutating tir-1 in an injury paradigm that induces significant axon degeneration. Although the above axotomy assay, in which an axon is severed once, is extremely useful for investigating the genetic and cellular mechanisms that regulate axon regeneration (Byrne and Hammarlund, 2017), it is not as robust a model of injury-induced axon degeneration. The obstacle is that while proximal segments, which remain attached to the cell body, regenerate in the 24 hr following injury (Yanik et al., 2004), the distal segments, which are no longer attached to the cell body, remain largely intact (Nichols et al., 2016). This prompted us to develop an assay with which robust injury-induced motor axon degeneration and axon regeneration can be investigated simultaneously in vivo. Here, rather than severing axons once, we severed individual axons in two places, approximately 25 μm apart, which creates a middle fragment that degenerates (Figure 3A). The degeneration has similar morphological and temporal features as Wallerian degeneration, where after a brief latent period, the middle axon segment beads, fragments, and is eventually cleared (Figure 3B). We tested how reliably double injury induces degeneration of the middle fragment by quantifying the percentage of severed axons whose middle fragment either degenerated (complete clearance), partially degenerated (>65% of middle fragment degenerated) or largely remained intact (>80% of middle fragment remained). In wild-type axons, 90.2% of middle fragments degenerated 24 hr after double axotomy (Figure 3C). We did not observe a significant difference in the number of middle fragments when visualized with either cytosolic or myristoylated and membrane-bound GFP 1.5 hr and 24 hr after injury (Figure 3D), indicating that the observed degeneration is not caused by a release of cytosolic GFP from the severed fragments. Notably, significantly less axons degenerated 1.5 hr after injury compared to 24 hr after injury, suggesting the earlier time point represents a period of active degeneration (Figure 3E). Finally, we found that injured motor axons are less likely to degenerate in aged animals, indicating an active destruction process is either lost or suppressed with increased age (Figure 3F). Together, these data indicate that the observed degeneration is an active process, not caused by a leaky cytosolic reporter and that the double cut assay reliably induces axon degeneration that can be investigated in vivo and with single-neuron precision.

Figure 3 with 1 supplement see all
Download asset Open asset
TIR-1 promotes axon degeneration in injured GABA motor neurons.

(A) Single and double injury models. After being severed once, GABA axons do not degenerate. Severing an axon twice creates a middle fragment that degenerates. (B) Micrographs of intact and degenerated middle fragments 24 hr post axotomy. Boxes with dashed lines highlight intact (pink) and degenerated (red) middle fragments. Scale bar: 5 μm. (C) Most middle fragments degenerate 24 hr after double injury. N = 61. (D) No significant difference in middle fragment degeneration was observed between axons expressing membrane tagged- or cytosolic GFP 24 hr (N = 19, 14) and (E) 1.5 hr after injury. (F) Middle fragments degenerate less frequently in aged adult animals compared to animals in the fourth larval stage (L4). N = 70, 44. (G) tir-1 promotes middle fragment degeneration. N = 63, 68. (H) Expression of tir-1b specifically in GABA motor neurons restores regeneration to tir-1(qd4) animals. N = 211, 68, 34, 37. Significance relative to control is indicated by *p≤0.05, **p≤0.01, ***p≤0.001, Fisher’s exact test. Error bars represent 95% confidence intervals.

tir-1 is a functional ortholog of Sarm1 that promotes injury-induced axon degeneration

Sarm1/dSarm is a key regulator of injury-induced degeneration in multiple species, including mammals, flies, and zebrafish (Gerdts et al., 2013; Osterloh et al., 2012; Tian et al., 2020). To determine whether tir-1 is a functional ortholog of Sarm1/dSarm with a conserved role in injury-induced axon degeneration, we assayed its ability to regulate degeneration using the double-cut assay. We found that in tir-1 mutant animals, only 68% of middle fragments degenerated compared to 86% of middle fragments in wild-type animals (Figure 3G). Therefore, like Sarm1/dSarm, tir-1 positively regulates injury-induced axon degeneration. Expression of tir-1b specifically in GABA motor neurons restored middle fragment degeneration in tir-1(qd4) animals, indicating tir-1 functions cell-autonomously to regulate injury-induced axon degeneration (Figure 3H).

Since loss of tir-1 incompletely suppressed axon degeneration, we investigated whether ced-1, a homolog of Draper and Megf10, also contributes to the degeneration phenotype. CED-1 functions in the hypodermis and muscle cells of C. elegans to remove axonal debris after injury in mechanosensory neurons (Chiu et al., 2018; Nichols et al., 2016). We found that injured motor axons in ced-1 loss-of-function animals degenerated as frequently as wild-type axons after double axotomy (Figure 3—figure supplement 1). Moreover, animals with loss-of-function mutations in both ced-1 and tir-1 phenocopied the amount of degeneration observed in tir-1 animals. These results demonstrate that middle fragment degeneration after double injury does not depend on ced-1 and that tir-1 functions independently from ced-1 to regulate injury-induced degeneration. Together with our finding that injury-induced degeneration is not completely dependent on tir-1 (Figure 3F), these data indicate C. elegans motor axons have evolved additional as yet unidentified regulators of injury-induced motor axon degeneration.

We hypothesized that perhaps distal fragments do not degenerate after single axotomy because TIR-1 is not sufficiently activated by this type of injury. TIR-1 and SARM1 functions are autoinhibited by intramolecular and intermolecular interactions with their N-terminal domains (Chuang and Bargmann, 2005; Figley et al., 2021; Gerdts et al., 2013; Horsefield et al., 2019; Jiang et al., 2020; Shen et al., 2021; Sporny et al., 2020; Summers et al., 2016). We asked whether low-level expression of an endogenous isoform lacking the autoinhibitory domain was sufficient to induce degeneration of GFP-labeled distal fragments, even after a single injury (Figure 4). Strikingly, we found that low levels of GABA-specific TIR-1b expression caused the majority of distal stumps to degenerate after a single axotomy (Figure 4A–E). Notably, degeneration was not observed in the proximal stumps, indicating that despite the lack of its autoinhibitory domain, TIR-1 specifically regulates degeneration distal to the site of injury. Therefore, in wild-type C. elegans, axons that have been injured once are likely incapable of degenerating, at least in part, because the injury response was of insufficient magnitude or identity to activate TIR-1. In contrast, a double axotomy may induce an enhanced injury response in the middle fragment, which is of sufficient magnitude or identity to activate TIR-1 or overcome its suppression.

Figure 4 with 1 supplement see all
Download asset Open asset
TIR-1 expression causes spontaneous axon degeneration in GABA motor neurons.

(A–C) Expression of mScarlet::tir-1b in GFP-labeled GABA motor neurons is sufficient to induce degeneration of the distal stump after a single axotomy. Micrographs and representative drawings demonstrate mScarlet::TIR-1b is localized in the cytoplasm in GABA cell bodies and along their axons in the ventral nerve cord, commissures, and dorsal nerve cord. Three timepoints are represented: (A) before injury, (B) immediately after injury, and (C) 7 hr after injury. At 7 hr, the distal stump has degenerated while the proximal stump remains intact. Scale bar: 5 μm. (D) Magnified image of a degenerating distal stump before injury, 7 hr after injury and 24 hr after injury. Scale bar: 5 μm. (E) Most distal stumps degenerate within 24 hr after a single injury in animals expressing mScarlet::tir-1b. (F, G) Individual GABA motor axons (seen in F) degenerate in transgenic animals expressing unc-47p::tir-1b::mCherry (G). Scale bar: 20 μm. (H) Magnified image of a wild-type axon and one that degenerates in animals that express unc-47p::tir-1b::mCherry. There is less (I) GFP intensity along the axon commissures (N = 18, 14), (J) GFP intensity along the dorsal nerve cord and (K) number of intact axons in animals that overexpress tir-1 in GABA neurons compared to wild-type animals (N = 6, 9), indicating tir-1 promotes axon degeneration. For categorical data, significance relative to wild type is indicated by ****p≤0.0001, Fisher’s exact test. Error bars represent 95% confidence intervals. For continuous data, significance relative to wild type is indicated by ***p≤0.001, Student’s t-test. Error bars represent SEM.

To further investigate the extent to which TIR-1 is able to promote axon degeneration, we asked whether overexpressing TIR-1 is sufficient to induce motor axon degeneration in the absence of injury. When we expressed tir-1b at high concentrations, we found that GFP-labeled motor axons were beaded, severely fragmented, and degenerated even in the absence of injury (Figure 4F–H). As a result, there was a significant decrease in the amount of GFP along the axon commissures and dorsal nerve cord of tir-1b overexpressing animals compared to the amount of GFP in wild-type controls (Figure 4I and J). The most obvious phenotype was the virtual absence of axon commissures that extended fully from the ventral to the dorsal nerve cord in tir-1 transgenic animals compared to wild-type animals (Figure 4K). Despite the absence of axons, the number of GABA neuron cell bodies was consistent between transgenic and non-transgenic animals, indicating tir-1 overexpression selectively induces axon degeneration and not cell death (Figure 4G). Importantly, the tir-1 transgene was expressed as an extrachromosomal array, resulting in mosaic GABA motor neuron expression. This mosaicism allowed us to find that the few non-transgenic GABA motor axons had a wild-type morphology, indicating the observed degeneration was caused by cell-autonomous overexpression of the tir-1b transgene and was not merely a secondary consequence of physically manipulating and imaging the animals (Figure 4—figure supplement 1). Therefore, elevated tir-1b expression within GABA neurons is sufficient to induce severe and chronic axon degeneration, even in the absence of injury. Together with the finding that loss of tir-1 function suppresses injury-induced axon degeneration, these data demonstrate that tir-1 positively regulates axon degeneration when it is constitutively activated or sufficiently activated by injury.

TIR-1 regulates axon regeneration independently of its NADase activity and with the NSY-1–PMK-1 mitogen-activated protein kinase signaling cascade

How does one gene regulate both seemingly opposite processes of repair and destruction in what was once the same cell? We proceeded to investigate how TIR-1 regulates both regeneration and degeneration using the double injury model. The ability to observe both axon regeneration and degeneration using the same injury model is essential for comparing how TIR-1 regulates the two seemingly opposite processes (1) within the same injured wild-type axon and (2) in the context of the same injury response. Indeed, the importance of comparing aspects of the injury response within the same injury model is manifest in the finding that axons regenerate significantly less frequently after a double injury compared to a single injury (60.53% of axons regenerate in the 24 hr following a single injury, and 37.38% of axons regenerate in the 24 hr following a double injury). For these reasons, unless otherwise indicated, the experiments below were performed using double axotomy.

SARM1 and dSarm execute several cellular functions in injured axons, including cleavage of NAD+, regulation of MAPK signaling and the Axundead protein, as well as regulation of calcium signaling, gene transcription, and autophagy (Coleman and Höke, 2020; Essuman et al., 2017; Neukomm et al., 2017; Walker et al., 2017; Yang et al., 2015). Early investigations of TIR-1 function revealed that it also signals through variations of a canonical downstream MAPK pathway in the innate immune response and in neuronal development (Chuang and Bargmann, 2005; Couillault et al., 2004; Liberati et al., 2004; Pujol et al., 2008; Sagasti et al., 2001). The molecular mechanisms that mediate TIR-1, SARM1, and dSarm functions informed a candidate approach to dissect the genetic pathways in which TIR-1 regulates both axon regeneration and degeneration of injured axons.

The NAD+ hydrolase activity of SARM1 is conserved in the C. elegans TIR domain, in a TIR-1 isoform lacking the N-terminal inhibitory domain, and in the full-length TIR-1 protein (Gerdts et al., 2016; Horsefield et al., 2019; Loring et al., 2021; Loring et al., 2020; Peterson et al., 2022; Summers et al., 2016). In SARM1, TIR domain function requires an active site containing the key catalytic residue E642, a structure specific to catalytically active TIR domains called a SARM-specific (SS) loop (residues 622–635), a BB loop (residues 594–605) thought to be important for intradomain interactions, and multiple residues important for interaction with the autoinhibitory domain and protein multimerization (Essuman et al., 2017; Horsefield et al., 2019; Loring et al., 2021; Loring et al., 2020; Summers et al., 2016). To directly address whether TIR-1 regulates regeneration by hydrolyzing NAD+, we CRISPR edited the endogenous tir-1 gene to create mutations in the key catalytic residue (tir-1(E788A)) and in the SS loop (tir-1(D773A)), which correspond with SARM1E642A and SARM1D627A mutants, respectively (Essuman et al., 2017; Loring et al., 2021; Summers et al., 2016). The conserved E788/E642/E1170 catalytic residue is required for the NADase activity of the TIR-1, SARM1, and dSarm TIR domains (Brace et al., 2022; Essuman et al., 2017; Herrmann et al., 2022; Horsefield et al., 2019; Hsu et al., 2021; Loring et al., 2021; Peterson et al., 2022; Summers et al., 2016). The D773/D627 aspartate is also conserved in the SS loops of C. elegans and human TIR domains. It is required for NAD+ catalysis by SARM1 and its ability to induce degeneration of injured DRGs but not for its ability to multimerize (Loring et al., 2021; Summers et al., 2016). We found that both tir-1(E788A) and tir-1(D773A) mutant animals displayed wild-type levels of axon regeneration, indicating that TIR-1E788A and TIR-1D773A are functionally expressed and that neither residue are required for TIR-1 to inhibit axon regeneration (Figure 5A). Therefore, TIR-1 regulates regeneration independently of its NADase activity.

Figure 5 with 2 supplements see all
Download asset Open asset
TIR-1 functions independently of its NADase activity and with the NSY-1 signaling pathway to inhibit axon regeneration.

(A) Conserved residues E788A/E642A, which is required for the NADase function of TIR-1 and SARM1, and D773A/D627 are not required to inhibit axon regeneration. (B) DLK-1 and TIR-1 function in the same or parallel pathways to regulate axon regeneration. N = 288, 68, 122, 41, 36, 30. (C) Axon regeneration is significantly increased in predicted null alleles of nsy-1/ASK1 and pmk-1/p38 following double injury, in both the presence and absence of tir-1 function. Double mutants are not statistically different from single mutants. N = 289, 68, 60, 50, 54, 71. (D) Null mutation of unc-43/CAMKII enhances axon regeneration after double injury in the presence and absence of tir-1. N = 89, 68, 39, 37. (E) Loss of the transcription factor daf-19/RFX1-3 increases axon regeneration after double injury in the presence and absence of tir-1. N = 142, 68, 41, 35. (F) Components of the TIR-1–NSY-1 signaling cascade. Significance relative to wild type is indicated by *p≤0.05, **p≤0.01, ***p≤0.001. Error bars represent 95% confidence intervals.

We began our search for NADase-independent mechanisms of TIR-1 function by investigating whether TIR-1 inhibits regeneration with established regulators of the injury response. Perhaps the most deterministic regulator of axon regeneration in C. elegans motor neurons is the mitogen-activated protein kinase kinase kinase DLK-1 (Dual leucine zipper kinase-1). DLK-1 function is absolutely required for the regeneration of otherwise wild-type GABA motor axons (Hammarlund et al., 2009; Yan et al., 2009). We found that loss of tir-1 function did not suppress the complete absence of regeneration in either the dlk-1 loss-of-function mutant or the downstream pmk-3/p38 MAPK loss-of-function mutant, indicating tir-1 either functions upstream of dlk-1 signaling or in a separate signaling pathway (Figure 5B). Physical interpretations of genetic interactions with genes that are essential for a given function can be difficult if that function requires the activity of multiple distinctly regulated processes. Therefore, since DLK-1 function is absolutely required for axon regeneration, these interactions do not distinguish between the possibilities that TIR-1 signaling regulates the activity of DLK-1 or that DLK-1 regulates an essential component of axon regeneration independently of TIR-1 signaling. However, the genetic interactions do indicate that TIR-1 does not function downstream of, or redundantly with, DLK-1 to regulate axon regeneration.

We next asked whether previously identified TIR-1 interactors mediate its role in axon regeneration. Initial studies of TIR-1 identified a shared MAPK signaling cassette that interacts with TIR-1 to regulate neuronal development and innate immunity (Chuang and Bargmann, 2005; Couillault et al., 2004; Liberati et al., 2004; Pujol et al., 2008; Sagasti et al., 2001). During neuronal development, TIR-1 interacts with the calcium-calmodulin-dependent protein kinase II (CaMKII) ortholog UNC-43 and the MAP kinase kinase kinase NSY-1/ASK1 signaling pathway to specify the asymmetric neuronal identity of AWC olfactory neurons. Specifically, in response to calcium, TIR-1 and UNC-43 form a postsynaptic protein complex that transiently interacts with NSY-1, which in turn regulates odorant receptor expression through its downstream MAP kinases (Chuang and Bargmann, 2005; Sagasti et al., 2001). In contrast, the TIR-1/NSY-1 signaling cassette functions independently of UNC-43 to regulate the innate immune response (Couillault et al., 2004; Liberati et al., 2004; Pujol et al., 2008). We first asked whether TIR-1 regulates axon regeneration by interacting with the NSY-1 signal transduction pathway and found that loss of nsy-1 or its downstream MAP kinase pmk-1/p38 enhanced axon regeneration after double injury (Figure 5C and F). These regeneration phenotypes were not significantly enhanced by deletion of tir-1. Because tir-1, nsy-1, and pmk-1 null mutants all inhibit axon regeneration to the same extent, and because the amount of regeneration in any pairwise combination of mutants is not additive or multiplicative compared to the single mutants, our data indicate that the three genes function in the same pathway to regulate axon regeneration after double injury (Figure 5F). Continuing our investigation of candidate interactors, we considered unc-43 because it was previously found to inhibit regeneration of sensory axons (Chung et al., 2016). We found that unc-43(e408) phenocopied the increased axon regeneration of tir-1(qd4) mutants and addition of the unc-43(e408) mutation to tir-1(qd4) mutants did not further enhance regeneration (Figure 5D). Therefore, tir-1 and unc-43 function together in a shared genetic pathway to inhibit axon regeneration after double injury. Since the nsy-1 and pmk-1 null phenotypes fully recapitulate the tir-1(qd4) phenotype and because they have the opposite phenotype to dlk-1 mutants, the nsy-1 and dlk-1 MAP kinase pathways do not function redundantly to inhibit axon regeneration downstream of tir-1. Together, these results indicate that tir-1 and the nsy-1–pmk-1 MAPK pathway function together to inhibit axon regeneration, which is dependent on dlk-1.

After a single injury, the regeneration frequency and genetic interactions between tir-1, nsy-1, and pmk-1 null mutants were similar to those after a double injury and wild-type regeneration was not affected by the lack of NADase activity in tir-1(E788A) animals (Figure 5—figure supplement 1). Interestingly, although wild-type axons regenerate less frequently after a double injury compared to a single injury, the magnitude of regeneration in TIR-1, NSY-1, and PMK-1 mutants is similar in axons that have been injured once (Figure 5—figure supplement 1A and B) or twice (Figure 5C). Therefore, the TIR-1–NSY-1–PMK-1 pathway not only inhibits regeneration after both types of injury, it may be epistatic to what differentiates the regeneration frequency between the two types of injury.

TIR-1 inhibits axon regeneration by regulating the RFX-type transcription factor DAF-19

CaMKII has previously been found to regulate axon regeneration and developmental axon outgrowth, in part by regulating transcription of effector genes (Chung et al., 2016; Ghiretti et al., 2014; Xi et al., 2019). However, the signaling pathways and set of transcription factors that mediate the effects of CaMKII activation on target gene expression during regeneration are not known. Given the genetic interactions between tir-1, unc-43, nsy-1, and pmk-1, we hypothesized that UNC-43 might regulate axon regeneration via TIR-1 and NSY-1-dependent transcription. We examined this possibility by testing the regenerative ability of animals with predicted null mutations in the cAMP-dependent transcription factor atf-7/ATF2/CREB5 and in the Regulatory Factor X transcription factor daf-19/RFX, both of which are regulated by nsy-1 signaling (Shivers et al., 2009; Vérièpe et al., 2015; Xie et al., 2013). Loss of atf-7 function did not affect GABA axon regeneration, consistent with previous studies of PLM mechanosensory axon regeneration (Figure 5—figure supplement 2; Ghosh-Roy et al., 2010). However, loss of daf-19 function increased regeneration to the same extent as loss of tir-1 and did not enhance the regeneration phenotype of the tir-1 null mutants (Figure 5D). Together, these data are consistent with a model in which an unc-43/tir-1/nsy-1/pmk-1/daf-19 signaling cascade (Figure 5E) inhibits axon regeneration by altering transcription of downstream regeneration-associated genes.

TIR-1 regulates degeneration in coordination with DLK-1 MAPK signaling

To determine how TIR-1 achieves specificity of function on either side of the injury, we also asked how it promotes axon degeneration. Since SARM1 and TIR-1 are established NAD+ hydrolases (Essuman et al., 2017; Horsefield et al., 2019; Loring et al., 2021; Loring et al., 2020; Peterson et al., 2022; Summers et al., 2016), we began by asking whether TIR-1 regulates regeneration by hydrolyzing NAD+ in injured axons. To do so, we quantified degeneration in TIR-1E788A animals (homologous to SARM1E642A), which renders the TIR domain of C. elegans TIR-1 catalytically dead (Essuman et al., 2017; Loring et al., 2020; Peterson et al., 2022). Interestingly, the E788A/E642A mutation did not suppress injury-induced axon degeneration (Figure 6A), indicating TIR-1 can regulate axon degeneration independently of its NAD+ hydrolase activity. This ability contrasts sharply with the requirement for NAD+ hydrolysis in dSarm and SARM1-mediated degeneration (Brace et al., 2022; Essuman et al., 2017; Gerdts et al., 2015; Herrmann et al., 2022; Hsu et al., 2021), indicating TIR-1 regulates a previously unexplored mechanism of axon degeneration.

Figure 6
Download asset Open asset
TIR-1 is capable of NADase-independent and -dependent axon degeneration.

(A) Mutation of conserved residue E788A/E642A, which is required for the NADase function of TIR-1 and SARM1, does not affect axon degeneration after double axotomy. D773A/D627A modestly suppresses degeneration. N = 97, 106, 64. (B) Conserved residues E788A/E642A and D773A/D627A are required for chronic axon degeneration. Lines represent independently generated transgenic strains that were generated by injecting equal concentrations of the same tir-1b::mCherry expression plasmid into separate tir-1(qd4) animals and isolating transgenic lines from each animal. N = 34, 32, 33, 23, 1211, 17. (C) Equivalent amounts of wild-type tir-1b, tir-1b(e788a), and tir-b(d773A) transgenes are expressed, as quantified by total mCherry fluorescence in the cell bodies of tir-1(qd4) animals, which were identified by expression of an integrated GABA-specific GFP marker, oxIs12 (N = 63, 70, 70). Significance relative to wild type was determined by one-way ANOVA and Dunnett’s test and is indicated by ****p≤0.0001. Error bars represent 95% confidence intervals.

We then asked whether other components of the TIR domain are required for its ability to regulate axon degeneration. Specifically, we investigated residue D773, which is homologous to D627 in the SS-loop of SARM1 that is required for injury-induced degeneration of DRG axons yet promotes the TIR domain’s NADase activity through unspecified mechanisms (Loring et al., 2020; Summers et al., 2016). We found that axon degeneration in tir-1(D773A) mutants was slightly but significantly disrupted compared to wild-type animals and was similar to that seen in tir-1(qd4) animals (Figure 6A). These data indicate the SS loop is important for TIR-1’s ability to promote degeneration of injured axons. Collectively, the findings that mutation of the key catalytic glutamate E788A abolishes NADase activity (Peterson et al., 2022) but not degeneration, along with previous data showing the SS-loop residue D627 promotes the NADase activity of SARM1 through relatively unknown mechanisms (Loring et al., 2020; Summers et al., 2016), indicate that the D773A/D627A mutation suppresses degeneration by disrupting an as yet unidentified function of TIR-1. Notably, D773A/D627A is not a null mutation since it did not disrupt the ability of TIR-1 to inhibit regeneration (Figure 5A). It follows that TIR-1 regulates axon regeneration and degeneration by different mechanisms, potentially by directly interacting with components of the UNC-43-NSY-1-PMK-1 pathway on the proximal side of the injury and another signal transduction pathway on the distal side of the injury.

The surprising finding that TIR-1 is capable of NADase-independent degeneration in response to injury prompted us to ask whether TIR-1 is capable of inducing NADase-dependent axon degeneration in other contexts. We found that while expressing large amounts of wild-type tir-1b induced significant axon degeneration even in the absence of injury, expressing either tir-1b(E788A/E642A) or tir-1b(D773A/D627A) did not (Figure 6B). The absence of spontaneous degeneration in the presence of TIR-1b(D773A) and TIR-1b(E788A) is not likely attributed to differential expression of the mutant and wild-type transgenes, which was determined by quantifying expression of the fluorescent tag on each protein (Figure 6C). However, because a small but statistically insignificant number of damaged axons were observed, we cannot rule out the possibility that greater TIR-1b(D773A) or TIR-1b(E788A) expression might induce NADase-independent degeneration. Together with our previous results, we find that D773 is required for both injury-induced and spontaneous degeneration, while E788 is only required for spontaneous degeneration. These data indicate D773A disrupts multiple facets of TIR-1 function, and the NADase activity of TIR-1 is required to induce spontaneous degeneration when overexpressed in uninjured axons but is not required for degeneration of injured axon fragments.

Therefore, the question remains: how does TIR-1 regulate NADase-independent degeneration of injured axons? We first investigated whether TIR-1 signals with the same nsy-1–pmk-1 MAPK pathway that mediates its function on the proximal side of the injury by measuring the presence or absence of middle fragments 24 hr after double injury in the absence of the MAPK pmk-1. Loss of pmk-1 did not have a degeneration phenotype on its own and double tir-1; pmk-1 mutant animals phenocopied the degeneration observed in tir-1 mutant animals (Figure 7A). This lack of genetic interaction indicates that tir-1 functions independently from pmk-1 to promote axon degeneration after injury.

Figure 7 with 2 supplements see all
Download asset Open asset
TIR-1 functions with the DLK-1 signaling pathway to regulate axon degeneration.

(A) Loss of pmk-1/p38 does not affect degeneration with or without tir-1 function. N = 198, 68, 54, 71. (B) Loss of dlk-1 function suppresses the decreased degeneration of tir-1(qd4) animals and overexpression of dlk-1 in GABA motor neurons inhibits degeneration, indicating DLK-1 negatively regulates axon degeneration. Loss of pmk-3 function rescues degeneration in all backgrounds, indicating PMK-3 is epistatic to both dlk-1 and tir-1. N = 517, 140, 142, 41, 70, 82, 98, 45, 62, 23. Significance relative to wild type is indicated by *p≤0.05, **p≤0.01, ***p≤0.001, ****p≤0.0001. Error bars represent 95% confidence intervals. (C) Representative micrographs of double injured wild type and dlk-1(OE) axons in which the respective middle fragments did and did not (asterisk) degenerate. Scale bar: 10 μm. (D) TIR-1 inhibits axon regeneration and promotes axon degeneration in the same injured axon. In the proximal fragment, TIR-1 functions with the NSY-1 signaling cascade to modify gene transcription and inhibit axon regeneration. In the severed fragment, TIR-1 promotes degeneration by antagonizing the inhibitory DLK-1 signaling cascade. Significance relative to wild type or tir-1(qd4) is indicated by *p≤0.05, **p≤0.01, ***p≤0.001. Error bars represent 95% confidence intervals.

DLK MAPK-signaling functions with SARM1/dSarm to regulate axon degeneration in mice and flies (Miller et al., 2009; Walker et al., 2017; Welsbie et al., 2013; Xiong and Collins, 2012; Xiong et al., 2010; Yang et al., 2015). In our assay, loss of either dlk-1 or its downstream p38 MAP kinase pmk-3 did not affect axon degeneration on their own; however, they did rescue axon degeneration in tir-1 mutants (Figure 7B). This result suggests the decreased axon degeneration phenotype of tir-1 mutants depends on dlk-1 pathway activity. However, due to the high amount of axon degeneration in wild-type animals, it is possible that an upper threshold in the assay prevents us from seeing enhanced degeneration above wild-type levels. Therefore, to further investigate whether dlk-1 inhibits axon degeneration, we asked whether overexpression of dlk-1 would protect axons. We found that, indeed, significantly less severed axon fragments degenerated in dlk-1(oe) animals relative to wild-type animals (Figure 7B and C, Figure 7—figure supplement 1). Of note, DLK-1 function is not responsible for inhibiting degeneration after a single injury as neither loss or gain of dlk-1 function induced degeneration of singly injured distal axon fragments (Figure 7—figure supplement 2). Therefore, DLK-1’s ability to inhibit degeneration is context specific.

The suppression of degeneration in doubly injured dlk-1(oe) axons was not dependent on tir-1 function (Figure 7B). However, loss of pmk-3/p38 MAPK function rescued degeneration in both dlk-1(oe) and tir-1(-) animals, as well as in double dlk-1(oe); tir-1(-) mutants (Figure 7B). The simplest interpretation of these genetic epistasis analyses is that PMK-3 inhibits degeneration downstream of its canonical MAPKKK DLK-1, and also functions downstream of TIR-1. These findings support a model in which TIR-1 promotes degeneration by antagonizing or opposing the inhibitory role of DLK-1-PMK-3 signaling. However, the exact nature of the complex relationship between TIR-1, DLK-1, and PMK-3 are exciting future avenues of investigation, as is the question of how degeneration is effected in the absence of TIR-1. Together, our data indicate that TIR-1 promotes axon degeneration by directly or indirectly antagonizing DLK-1-PMK-3 function on the distal side of the injury and inhibits regeneration with UNC-43–NSY-1–PMK-1–DAF-19 signaling on the proximal side of the injury (Figure 7D).

Discussion

Understanding the mechanisms that determine whether axons regenerate and degenerate is critical to our ability to stimulate repair after neuronal injury (Girouard et al., 2018). Here, we identify TIR-1/SARM1/dSarm, a central regulator of axon degeneration, as a previously uncharacterized intrinsic inhibitor of axon regeneration that can be targeted to increase repair of injured motor axons. Our data indicate that to achieve this specificity of function, tir-1 inhibits axon regeneration with the NSY-1–PMK-1 MAPK pathway on the proximal side of the injury while it promotes degeneration by antagonizing the DLK-1–PMK-3 MAPK pathway on the distal side of the injury. Together, our findings reveal divergent mechanisms by which TIR-1 regulates the injury response.

TIR-1/SARM1 functions intrinsically to regulate both axon regeneration and axon degeneration

We found that TIR-1 inhibits axon regeneration and does so independently of its role in degeneration by investigating the injury response in a new model of C. elegans motor axon injury. C. elegans has been a powerful model of axon regeneration owing to its genetic tractability, the ability to carry out experiments in vivo with single-axon resolution, and because the majority of its genes are conserved with mammals (Byrne et al., 2011; Kim et al., 2018b; Yanik et al., 2004). In addition, the C. elegans motor nervous system lacks myelin and glia, enabling investigation of mechanisms that regulate axon regeneration independently of these extrinsic cues. The finding that both tir-1 and human Sarm1 can function cell-autonomously to inhibit axon regeneration indicates intrinsic regulation of axon regeneration may be a conserved function of SARM proteins.

C. elegans has not been a robust model of motor axon degeneration because in the standard injury model, in which motor axons are cut once, the distal segment remains largely intact in adult animals (Nichols et al., 2016). Why the distal segment of severed C. elegans motor axons remain intact has been a long-standing point of curiosity (Figure 7—figure supplement 2). Here, we found that severing an axon twice creates a middle fragment that completely degenerates and a proximal segment that is capable of regenerating. It follows that since double axotomy does induce degeneration, the resulting injury response may be more representative of the injury responses of other invertebrates and mammals compared to single axotomy. Practically, the double injury assay combines the tractability of C. elegans with the ability to investigate motor axon regeneration and degeneration on either side of the injury. The resulting finding that tir-1 functions cell-autonomously to simultaneously inhibit regeneration of the proximal axon fragment while also promoting degeneration of the distal axon fragment raises a number of questions. One of which is whether the enhanced regeneration in tir-1 mutants is not simply a secondary consequence of reduced degeneration. Either reducing the function of intrinsic inhibitors or changing the amount of extrinsic cues in the environment, such as degeneration and inflammation, can promote axon regeneration (Curcio and Bradke, 2018; He and Jin, 2016; Schwab and Strittmatter, 2014; Tedeschi and Bradke, 2017; Tran et al., 2018). Here, we found that axons lacking tir-1 regenerate more often than wild-type axons after a single injury even though a single injury does not cause degeneration in either wild-type or tir-1 mutants. Thus, the increased regeneration of tir-1 mutants is not a secondary consequence of a decrease in degeneration. Combined with our finding that tir-1 functions within GABA axons to regulate axon regeneration, our data suggest that tir-1 functions within the proximal fragments of injured axons to inhibit axon regeneration independently from its role in degeneration.

TIR-1/SARM1 inhibits axon regeneration with the NSY-1/ASK1 MAPK signaling pathway

Our finding that tir-1 regulates both axon regeneration and degeneration raises the question of how this single gene differentially regulates opposite functions in what was once the same axon. Our data indicate that TIR-1/SARM1 carries out these opposing functions by interacting with two divergent MAP kinase signaling pathways on either side of the injury. On the proximal side, TIR-1/SARM1 functions with UNC-43/CAMKII, the NSY-1/ASK1–PMK-1/p38 MAPK signaling cascade, and the transcription factor DAF-19/RFX to inhibit axon regeneration. The involvement of DAF-19 suggests that TIR-1 activation may change the transcriptional profile of the proximal fragment into one that does not support repair. Unlike the TIR-1–NSY-1–PMK-1 pathway that regulates the innate immune response and serotonin biosynthesis, DAF-19 does not require the co-transcription factor ATF-7 to regulate axon regeneration (Xie et al., 2013). Therefore, TIR-1 likely achieves specificity of function by regulating the activity or expression of particular transcription factors in a temporal and tissue-specific manner in response to injury. SARM1 has been found to regulate a c-Jun-mediated transcriptional program and an immune response in injured mammalian axons (Wang et al., 2018). Whether it could also regulate cytoskeletal dynamics and regeneration in specific cellular contexts through RFX-mediated transcription is an interesting question.

The additional finding that axon regeneration is not affected by mutation in the predicted NADase residue E788A/E642A suggests that the role of tir-1 in axon regeneration does not depend on its ability hydrolyze NAD+. TIR-1’s ability to regulate regeneration in coordination with NSY-1 signaling and independently of its NADase activity contrasts with recent findings that dSarm regulates neuromuscular junction formation and glial phagocytosis using both its NADase activity and interaction with Ask1 signaling (Brace et al., 2022; Herrmann et al., 2022). The NADase-independent function of TIR-1 also contrasts with the findings that dSarm, mSarm, and hSARM1 rely on their ability to hydrolyze NAD+ to regulate axon degeneration (Brace et al., 2022; Herrmann et al., 2022; Hsu et al., 2021). However, the TIR-1 mechanism parallels the mechanism by which dSarm regulates axon transport within uninjured bystander neurons, which occurs independently of its NADase activity and by interacting with the voltage-gated calcium channel Cacophony (Cac) and Ask1/NSY-1 MAPK signaling (Hsu et al., 2021). These data indicate that dSarm and tir-1 can regulate multiple components of the injury response independently of NAD+ hydrolysis. Whether UNC-43–TIR-1–NSY-1–PMK-1–DAF-19 signaling inhibits regeneration in C. elegans by suppressing axon transport and how the NSY-1/ASK1 pathway differentially regulates its downstream functions are interesting questions of future investigation.

Our finding that TIR-1 regulates regeneration also raises the question of whether TIR-1’s role in regeneration is dependent on NMNAT activity. NMNAT antagonizes SARM1 by maintaining NAD+ in the allosteric pocket of SARM1, which inhibits its NAD+ hydrolase activity (Bratkowski et al., 2020; Figley et al., 2021; Horsefield et al., 2019; Jiang et al., 2020; Sasaki et al., 2016; Shi et al., 2022; Sporny et al., 2020). In NMAT mutants, NAD+ is replaced by NMN in the allosteric pocket and the altered conformation of SARM1 activates its NADase activity (Figley et al., 2021; Sasaki et al., 2016; Shi et al., 2022). Previous work demonstrates that NMNAT2 inhibits axon regeneration in C. elegans mechanosensory axons (Kim et al., 2018a). This could lead to the prediction that NMAT-2 might also inhibit regeneration by inhibiting the NAD+ hydrolase activity of TIR-1. However, here we find that TIR-1 inhibits regeneration of motor axons independently of its function as an NAD+ hydrolase, indicating that NMNAT-2 must not inhibit regeneration by suppressing the NADase activity of TIR-1. Nonetheless, it remains possible that an NMNAT-dependent change in conformation determines whether TIR-1 can interact with downstream effectors of regeneration such as NSY-1. In this model, NMAT-2 would indeed facilitate TIR-1’s ability to inhibit regeneration. Alternatively, since many genes differentially regulate axon regeneration in a cell-type-dependent manner (Kim et al., 2018a), it is also possible that NMAT-2 and TIR-1 differentially regulate regeneration in the mechanosensory and motor axons.

TIR-1/SARM1 regulates injury-induced axon degeneration in coordination with the DLK-1 MAPK signaling pathway

On the distal side of the injury, our data suggest that TIR-1 regulates degeneration by interacting with an alternate MAPK pathway. Loss of pmk-1 function has no effect on axon degeneration, nor does it affect the degeneration phenotype of tir-1 loss-of-function mutants, indicating that TIR-1 regulates axon degeneration independently from the NSY-1–PMK-1 MAPK pathway. Instead, TIR-1 promotes axon degeneration through interaction with DLK-1–PMK-3 MAPK signaling. Specifically, our data suggest that in C. elegans, TIR-1 promotes injury-induced degeneration by inhibiting a protective function of DLK-1 and PMK-3. The protective function of DLK-1 and PMK-3 in axon degeneration appears to be conserved in specific contexts. For example, Wallenda (Wnd), the DLK-1 homolog in Drosophila melanogaster inhibits axon degeneration in motor axons that have undergone a conditioning lesion (Xiong and Collins, 2012). However, Wallenda also promotes degeneration in singly severed olfactory receptor axons and mammalian DLK promotes degeneration of severed dorsal root ganglion axons (Ghosh et al., 2011; Miller et al., 2009). While it is unlikely that the small amount of time between axotomies in the double injury model induces a conditioning lesion, the comparison demonstrates that the nature of MAPK activity can be differentially determined by specific types of injury and by cell type. In relation to SARM signaling, SARM1 and dSarm also regulate axon degeneration by interacting with the signaling pathways of DLK-1 homologs Wallenda and DLK (Summers et al., 2020; Walker et al., 2017; Yang et al., 2015). Separate studies have found that MAPK signaling functions downstream of SARM1 to regulate degeneration of crushed retinal ganglion cells and upstream of SARM1 to regulate degeneration of cultured mammalian dorsal root ganglia neurons and Drosophila motoneurons (Summers et al., 2018; Walker et al., 2017; Yang et al., 2015). Together with this multifaceted relationship between SARM1 and MAPK signaling, our data indicate the highly conserved TIR-1, dSarm, and SARM1 proteins regulate injury-induced degeneration through interactions with shared MAPK signaling components. However, the precise interactions between TIR-1/SARM1/dSarm and DLK-1/MAPK/Wnd signaling are complex and likely depend on cell type, form of injury, and species. Identifying the similarities and differences between the various interactions between SARM and MAPK signaling across species and types of injury will be important to understanding how these complex genes regulate the injury response.

In addition to regulating the injury response, TIR-1 is required for degeneration in a C. elegans model of ALS, functions in the epidermis to promote age-dependent degeneration of PVD dendrites, induces spontaneous degeneration when forced to multimerize, and inhibits spontaneous axon degeneration of PVQ interneurons in mutant ric-7(-) animals, which lack mitochondria (Ding et al., 2022; Lezi et al., 2018; Loring et al., 2021; Rawson et al., 2014; Vérièpe et al., 2015). Questions raised specifically from a comparison of the opposite destructive and protective roles of TIR-1 in response to injury and absence of mitochondria include whether a specific injury signal determines that TIR-1 takes on a destructive role and whether TIR-1 can protect multiple types of neurons against the spontaneous degeneration that is induced by the absence of mitochondria. Looking downstream, we found the map kinase pmk-3 is epistatic to tir-1 and inhibits injury-induced axon degeneration. Interestingly, pmk-3 was also found to inhibit spontaneous degeneration in the absence of mitochondria, suggesting the difference between promotion and suppression of degeneration may ultimately be attributed to differential regulation of PMK-3 in these two contexts.

To date, how SARM1 ultimately leads to axon degeneration after injury is not completely clear. Besides MAP kinase signaling, effectors of dSarm and SARM1 function include diminished levels of NAD+ and the downstream Axundead protein (Essuman et al., 2017; Neukomm et al., 2017). C. elegans does not have an identifiable Axundead homolog; however, the NADase activity of SARM1 and dSarm is conserved, albeit perhaps to a weaker extent, in the C. elegans TIR domain (Horsefield et al., 2019; Loring et al., 2020; Neukomm et al., 2017; Peterson et al., 2022; Summers et al., 2016). Our finding that TIR-1 can promote injury-induced degeneration despite having a E788A/E642A mutation in its key catalytic residue, along with a recent finding that E788A renders the C. elegans TIR domain catalytically dead and is required for TIR-1-mediated pathogen avoidance, indicates that although TIR-1 does function as an NADase, it regulates injury-induced axon degeneration independently of its NADase activity. How the D773A/D627A mutation disrupts TIR function remains an open question (Loring et al., 2020; Summers et al., 2016). It is possible that in addition to disrupting the enzymatic activity of TIR-1, D773A also disrupts one of the many complex intramolecular and intermolecular interactions of TIR-1, which suppresses degeneration by disrupting a yet unidentified mechanism of action. Knowing whether TIR-1 induces injury induced degeneration through direct interaction with DLK-1 will contribute significantly to our understanding of how TIR-1 promotes NADase-independent degeneration.

While TIR-1 can promote NADase-independent degeneration, dSarm and SARM1 depend on NAD+ hydrolysis to regulate injury-induced degeneration (Brace et al., 2022; Herrmann et al., 2022; Hsu et al., 2021). Interestingly, while overexpressing wild-type tir-1b in motor neurons induced spontaneous degeneration, expressing the same amount of mutant tir-1b, containing the D773A or E788A mutation, did not induce significant degeneration. Therefore, while TIR-1 does not require its NADase activity to promote injury-induced degeneration, it can use its enzymatic activity to induce spontaneous axon degeneration. These data suggest that TIR-1 has evolved dual mechanisms to regulate degeneration in response to different insults that it encounters in the wild.

The finding that TIR-1, a homolog of dSarm and SARM1, essential regulators of axon degeneration, also regulates the opposite response to injury, axon regeneration, is unexpected. In addition, TIR-1 was not thought to regulate injury-induced degeneration of C. elegans axons, likely because adult motor axons do not robustly degenerate using the single-injury assay and because degeneration of injured mechanosensory axons is regulated by apoptotic engulfment machinery in surrounding tissues, and not by TIR-1 (Byrne et al., 2011; Nichols et al., 2016). The identified dual role of TIR-1 in the injury response raises a number of compelling questions, including: Is TIR-1 differentially activated on either side of the injury to determine whether a fragment repairs itself or self-destructs? Does TIR-1 regulate regeneration throughout the nervous system, or, like other intrinsic regulators of regeneration, does it function according to cell type and age (Byrne et al., 2014; Duan et al., 2015; Geoffroy et al., 2016)? Recent studies support the idea that SARM orthologs might function intrinsically to directly regulate axon regeneration. Systemic inhibition of tir-1/Sarm1 increases axon regeneration after injury in zebrafish. In addition, a mammalian study suggests functional recovery and axon regeneration after injury is improved in the absence of SARM1 function in neurons and in astrocytes, in part due to reduced inflammation and increased protection of injured axons (Asgharsharghi et al., 2021; Liu et al., 2021). This data is supported by a preceding finding that SARM1 function within injured axons regulates an immune response that triggers inflammation (Wang et al., 2018). Another recent study demonstrated that SARM1 knockout mice have more collateral axon branching and increased cytoskeletal dynamics in postnatal sensory neurons, indicating that SARM1 inhibits axon growth even in the absence of injury (Ketschek et al., 2022). In light of these results, our finding that GABA-specific TIR-1 and hSARM1 expression inhibits axon regeneration even in animals that lack extrinsic inflammatory cells raises the intriguing question of whether SARM1 signaling can be manipulated to promote axon regeneration by directly driving cytoskeletal growth in injured mammalian axons.

In conclusion, we present a model where TIR-1 functions cell-autonomously in GABA motor neurons to regulate axon regeneration. We show that to achieve specificity TIR-1 directs the opposite processes of axon regeneration and axon degeneration by interacting with divergent downstream MAPK signaling pathways on either side of the injury. Understanding how TIR-1 regulates a balance between axon regeneration and degeneration may provide novel ways to push the nervous system toward regeneration in response to injury.

Materials and methods

C. elegans strains and culture

Request a detailed protocol

Unless otherwise noted, all C. elegans strains were maintained at 20°C on NGM plates seeded with OP50 Escherichia coli. Strains listed below were kindly provided by the Alkema, Francis, Hammarlund, Pukkila-Worley, Hanna-Rose labs, and the Caenorhabditis Genetics Center, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440). To visualize GABA motor neurons, strains were crossed into EG1285 (oxIs12 [unc-47p::GFP, lin-15(+)]). Mutants used in this study: tir-1(qd4), tir-1(tm3036), ufIs175(flp-13p::myristoylatedGFP), nsy-1(ag3), pmk-1(km25); unc-43(e408), daf-19(m86), dlk-1(ju476), pmk-3(ok169), pnc-1(ku212), tol-1(nr2033), ced-1(e1735), atf-7(gk715).

Laser axotomy, imaging, and quantification

Request a detailed protocol

Axotomy was performed as previously described (Byrne et al., 2011), unless otherwise stated below. Larval stage 4 (L4) worms were mounted on 3% agarose pads and immobilized with microbeads (Polysciences) or 0.1 mM levamisole. Single injuries were made at the midline of the axon commissure and 1–3 axons were cut per worm. Double injuries were made with two targeted cuts, one roughly halfway between the ventral nerve cord and the midline of the worm and the second roughly halfway between the midline of the worm and the dorsal nerve cord, creating an approximately 25 μm middle fragment. The two injuries are performed within seconds of each other in the same order (proximal-distal). One axon was cut per worm in the double injury experiments. Axon regeneration was quantified 24 hr after injury unless otherwise stated by placing animals immobilized with microbeads or 0.1 mM levamisole on a 3% agarose pad. Animals were imaged with a Nikon 100x 1.4 NA objective, Andor Zyla sCMOS camera and Leica EL6000 light source. Regeneration was scored as the percent of axons that regenerated fully to the dorsal cord (F), 75% of the distance to the dorsal cord (M+), to the midline (M), to below the midline or had 3+ filopodia extending from the proximal segment (M−). Axons that failed to regenerate or had two or fewer small filopodia protruding from the proximal segment were scored as N and NF, respectively. To control for day-to-day and instrument-to-instrument variability, experiments were carried out with same day controls on one axotomy rig.

CRISPR

Request a detailed protocol

CRISPR gene editing was conducted as previously described (Dokshin et al., 2018). crRNA, tracrRNA, and Cas9 were obtained from IDT and injected with the rol-6 co-injection marker. CRISPR mutants were genotyped and outcrossed before analysis.

TIR-1 expression

Request a detailed protocol

We generated transgenic lines expressing TIR-1 under its endogenous promoter by expressing bamEx102[tir-1p::tir-1b::mCherry::unc-54 3’UTR+ceh-22::gfp] in ABC16 (tir-1(qd4); oxIs12[unc-47p::GFP, lin-15(+)]) animals using standard microinjection (Mello et al., 1991). L4 stage animals were imaged on Perkin Elmer Precisely UltraVIEW VoX confocal imaging system. To determine the subcellular localization of TIR-1, we generated transgenic animals expressing unc-47p::mScarlet::tir-1b:: let-858 3’UTR and co-injection marker ceh-22::gfp in EG1285 (oxIs12[unc-47p::GFP, lin-15(+)]) animals. To determine the tissue specificity of TIR-1 function, we generated transgenic animals expressing unc-47p::mScarlet::tir-1b::let-858 3’UTR and co-injection marker pstr-1::gfp in ABC16 (tir-1(qd4); oxIs12 [unc-47p::GFP]) animals. To determine whether TIR-1 promotes chronic axon degeneration, we expressed unc-47p::tir-1b::mCherry::unc-54 3’UTR and co-injection marker pstr-1::gfp in EG1285 (oxIs12[unc-47p::GFP, lin-15(+)]) animals. Transgenic lines were established using standard protocols (Mello et al., 1991). L4 stage animals were imaged on Perkin Elmer Precisely UltraVIEW VoX confocal imaging system.

hSARM1 expression

Request a detailed protocol

To determine whether axon regeneration is a conserved function of SARM1, we built transgenic animals expressing mScarlet::SARM1 from the unc-47 GABA-specific promoter. The SARM1 plasmid was a kind gift from the Fitzgerald lab at UMass Medical School. The SAM-TIR domain of SARM1 was cloned into the unc-47p::mScarlet::tir-1b::let858 3’UTR plasmid. The resulting unc-47p::mScarlet::SARM1::let858 3’UTR plasmid was injected into ABC16 (tir-1(qd4) III; oxIs12 X) and oxIs12(unc-47p::GFP X) animals. The plasmid was injected at a final concentration of 10 ng/ul to generate bamEx4070, bamEx4072, bamEx4073. Further we amplified a 1.286 kb band spanning the end of the mScarlet sequence into the beginning of the let-858 3’ UTR using PCR to confirm expression within these animals (primers: GCCGACTTCAAGACCACCTACAAG and CAACGACGTTGGCGTCGATC).

Injury-induced degeneration

Request a detailed protocol

Middle fragment degeneration after double injury was quantified 24 hr after injury unless otherwise indicated and grouped into three categories: degenerated (complete clearance), partially degenerated (>65% of the middle fragment degenerated), or largely remained intact (>80% of middle fragment remained) 24 hr after double injury. Chronic degeneration in the absence of injury was assessed by immobilizing L4 stage worms with 0.1 mM levamisole on a 3% agarose pad. Images were taken on a Perkin Elmer Precisely UltraVIEW VoX spinning disc confocal imaging system using a 40x objective. Images were compressed and exported as TIFF files for processing in FIJI. The axon or dorsal nerve cord segment to be analyzed was traced using a segmented line ROI of fixed size and mean intensity was quantified.

NADase point mutations

Request a detailed protocol

To determine whether the NADase function of TIR-1 is required to induce chronic degeneration, we built transgenic animals expressing either unc-47p::tir-1b(E788A)::mCherry::unc-54 3’UTR or unc-47p::tir-1b(D773A)::mCherry::unc-54 3’UTR plasmids. Point mutations were generated using QuikChange site directed mutagenesis on the unc-47p::tir-1b::mCherry::unc-54 3’UTR plasmid. Primers used to create the E788A point mutation were GCACAAAGCGCTGAAATGTGCATTTG and CATTTCAGCGCTTTGTGCACCCAATCC. Primers used to generate the D773A point mutation were CAGTTTGGCTAGACTTTTAAATGATGATAACTG and GTCTAGCCAAACTGTTTGGTGTGAG. The resulting plasmids were injected into ABC16 (tir-1(qd4) III; oxIs12 X) animals at 50 ng/ul with ceh-22::GFP as a co-injection marker to generate bamEx4003[unc-47p::tir-1b(E788A)::mCherry::unc-54 3’UTR+ceh-22::GFP], bamEx4052[unc-47p::tir-1b(E788A)::mCherry::unc-54 3’UTR+ceh-22::GFP], bamEx4049[unc-47p::tir-1b(D773A)::mCherry::unc-54 3’UTR+ceh-22:GFP], bamEx4050[unc-47p::tir-1b(D773A)::mCherry::unc-54 3’UTR+ceh-22::GFP] animals. The number of intact, damaged, and missing GABAergic axons was quantified by mounting L4 stage animals on 3% agarose pads in 300 nM sodium azide. Axons were scored as damaged if they were broken or if there was significant beading. All other axons were scored as intact if they traversed completely from the ventral nerve cord to the dorsal nerve cord. The number of cell bodies was counted starting at the tail moving anteriorly toward the head. We observed an average of 15 axons as the 16th most anterior axon was either missing or redirected to the left side of the animal.

TIR-1b fluorescence intensity quantification

Request a detailed protocol

To determine TIR-1b levels in animals expressing unc-47p::tir-1b::mCherry::unc-54 3’ UTR, unc-47p::tir-1b(E788A)::mCherry::unc-54 3’ UTR, and unc-47p::tir-1b(D773A)::mCherry::unc-54 3’ UTR, we quantified fluorescent expression in the posterior seven GABAergic cell bodies. These cell bodies correspond with axons that were severed in the injury assays. Images were taken on a Perkin Elmer Precisely UltraVIEW VoX spinning disc confocal imaging system using a 40x objective. Images were compressed and exported as TIFF files for processing in FIJI. Cell bodies were then traced using the freehand tool in FIJI and the intensity of TIR-1b::mCherry was quantified. Background fluorescence was corrected by quantifying an additional trace next to each cell body.

RT-qPCR

Request a detailed protocol

Strains analyzed: oxIs12(Punc-47::GFP), tir-1(qd4);oxIs12, dlk-1(oe);oxIs12, dlk-1(oe);tir-1(qd4);oxIs12. RNA isolation: 100 L4 stage animals of each strain were transferred to unseeded NGM plates and transferred to 50 µl RNase-free dH2O. 500 µl of cold TRIZOL was added and samples were frozen in a –80°C freezer overnight. The following day, another 500 µl of cold TRIZOL reagent was added to each sample. Samples were subjected to three rounds of freeze-thaws that included flash-freezing in liquid nitrogen, thawing in 37°C water bath, and vortexing for 30 s. Samples were passed through 18-Gauge syringes 12 times and centrifuged at 14,000 rpm for 10 min at 4°C to remove insoluble material. Supernatant was transferred to fresh microcentrifuge tubes and 200 µl of CHCl3 was added. Tubes were inverted and vortexed for 30 s, incubated at room temperature for 5 min, and centrifuged at 14,000 rpm for 15 min at 4°C to separate phases. 500 µl of the clear, upper aqueous phase was transferred to fresh microcentrifuge tubes and 500 µl isopropanol was added and mixed by pipetting. RNA was isolated with QIAGEN RNeasy purification and eluted with 36 µl of RNase-free dH2O. cDNA synthesis: reverse transcriptase (RT) synthesized cDNA, and No-RT samples were prepared with NEB LunaScript RT SuperMix kit. The concentration of each was then quantified using a NanoDrop. RT-qPCR analysis: Procedures were followed according to the NEB Luna Universal qPCR kit protocol. Reactions were performed in triplicate, with RT+ template and No-RT template control. RT-qPCR was performed with a Bio-Rad CFX96 Real-Time System with a C1000 Touch Thermal Cycler.

Statistical analysis

Request a detailed protocol

Statistical analysis was performed using GraphPad QuickCalcs (https://www.graphpad.com/quickcals/) and Prism (GraphPad). Categorical data: Data from repeated assays comparing a mutant and corresponding control strain were pooled for statistical analysis. Bars represent 95% confidence intervals and significance determined with Fisher’s exact test, where *p≤0.05, **p≤0.01, ***≤0.001, n ≥ 30. Continuous data (fluorescence intensity of GFP along axon commissure) are presented as the mean with error bars representing the standard error of the mean. Significance was calculated with Student’s t-tests, where *p≤0.05, **≤0.01, ***≤0.001.

Materials availability

Request a detailed protocol

All strains and reagents used in the analysis are available upon request.

Data availability

All numerical data analysed during this study are included in the main text and figure legends. No additional datasets were generated.

References

Decision letter

  1. Piali Sengupta
    Senior and Reviewing Editor; Brandeis University, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Decision letter after peer review:

[Editors’ note: the authors submitted for reconsideration following the decision after peer review. What follows is the decision letter after the first round of review.]

Reviewer #1:

Julian and Byrne use the C. elegans interneurons as a model to test the involvement of TIR-1 in axon regeneration. TIR-1's function in axon degeneration is well established but its function in regeneration is novel. The authors show convincingly that TIR-1 inhibits axon regeneration through the NSY-1 MAPK pathway. Since TIR-1 is a NAD+ degradation enzyme, the authors should test if overexpression of NMnat can also increase regeneration. The following issue should also be addressed.

For figure 3, what causes the GFP levels to reduce in the pGABA::tir-1? mCherry protein fusion is often problematic because it forces proteins to aggregate. In Figure 2C, it seems that significant amount of GFP colocalized with the mCherry puncta. The authors should use GFP to fuse with tir-1 and use a cytosolic mCherry as a marker to repeat this experiment. At present, the evidence that TIR-1 overexpression causes degeneration is not strong. mCherry protein fusions sometimes cause aggregates and toxicity. Experiments in Figure 5D should be repeated with tir-1GFP or untagged tir-1.

Reviewer #2:

Byrne and colleagues examine whether the major executioner of Wallerian degeneration, Sarm-1, serves this function in C. elegans and has an additional role in axon regeneration. They make use of a well-established system to study axotomy-induced axon regeneration and present a novel paradigm to cause injury-induced degeneration in these same neurons. The authors present evidence that Tir-1, the presumed Sarm1 ortholog in C. elegans, has a novel role to inhibit axon regeneration upon injury and suggest that Tir-1 promotes degeneration and inhibits regeneration via distinct MAPK pathways. Inhibition of Sarm1 has garnered significant attention to block neurodegeneration in several contexts. Therefore, the implications of this result – that Sarm1 inhibition may have the previously unknown effect of enhancing an adaptive regenerative response – are large and would be of significant interest in the field. Yet much of the excitement about this work (and certainly the tone of the writing) assumes that Tir-1 functions in C. elegans as it does in flies and mammals, as the key executioner of Wallerian degeneration which is not resolved by the experiments presented.

I have several big-picture concerns about this work:

1) Does TIR-1 promote axon degeneration similarly to dSarm/Sarm1?:

– Are the authors studying bona fide Wallerian degeneration? Loss of dSarm/Sarm1 potently and completely protects severed axons; 100% protection, lasting for days. The authors use a novel two-cut paradigm to allow for injury-induced degeneration (since injured axons do not normally degenerate in C. elegans), but only observe a ~20% protective effect (Figure 3A) in the qd4 allele, and only at a single time point. This is very different from what would be expected of a true dSarm/Sarm1 ortholog, and a key point in interpreting this work in the context of the broader Sarm1 literature. Is this two-cut degeneration dependent on NAD levels as would be expected for degeneration regulated by dSarm/Sarm1? Or is this degeneration mechanistically distinct? This could easily be tested by overexpression of NMNAT enzymes or mutation of the conserved NAD-hydrolyzing domain of Tir-1, both of which should protect severed axons if the authors' model is correct.

2) Does the genetic pathway for Tir-1-dependent regeneration fit their data?

– Loss of Dlk-1 and Pmk-3 both suppress the enhanced regeneration seen in Tir-1 mutants (Figure 4C). Further, there appears to be no genetic interaction between Tir-1 and Nsy-1/Pmk-1 (Figure 4B), and the authors state that the nsy-1-pmk-1 MAPK pathway "does not act synthetically or in parallel" to the dlk-1-pmk-3 MAPK pathway in tir-1 function in axon regeneration. Despite these data, the authors' model (Figure 5E) is that Tir-1 inhibits regeneration through Nsy-1/Pmk-1, not via Dlk-1/Pmk-3. I do not clearly see how these epistasis experiments support their model.

Reviewer #3:

Julian and Byrne investigate the function of TIR1, the worm ortholog of SARM1, in the axon injury response. In worms TIR1 participates in a number of signal transduction pathways, while in mammals and flies SARM1 is best known for its essential role in Wallerian degeneration. Here, Julian and Byrne find that TIR1 inhibits axon regeneration using its canonical nsy-1/Ask1 signaling pathway. Worms do not undergo Wallerian degeneration, however the authors develop a double axon injury model where an internal axon segment is disconnected on both its proximal and distal sides. These axon fragments do degenerate, and TIR1 mildly promotes this degeneration. This TIR1-dependent component of degeneration requires a different MAP kinase cascade than does regeneration. The authors conclude that in worms TIR1 regulates both axon regeneration and axon degeneration via distinct mechanisms.

This is a solid manuscript that clearly demonstrates that TIR1 is a negative regulator of axon regeneration in worms. Many genes have been identified to regulate axon regeneration in this system, and so this is an interesting advance that will be useful to workers in the field attempting to build a comprehensive understanding of C. elegans axon regeneration. The significance of the findings with degeneration of doubly injured axon fragments is less clear. TIR1 does not trigger degeneration of distal axons in the worm as it does in mice and fish and flies, and even the effect seen here with the double-injured axon fragments is quite modest (86% of WT fragments degenerate while 68% of TIR1 mutant fragments degenerate). In contrast, SARM1 is absolutely required for degeneration in these other organisms. Moreover, the DLK signaling pathway identified is well known to participate in axon degeneration in other organisms. As such, the advance for the field of axon degeneration is modest. Indeed, it would be much more significant to understand why TIR1 is unable to trigger axon degeneration in worms rather than to show this modest effect in this double injury model.

There are a few additional issues for the authors to consider.

1) The abstract and introduction of the story is framed as a choice between axon regeneration and degeneration that must be balanced. However, there is no conflict between the two because they occur in different axonal compartments (proximal vs distal). Indeed, in the mammalian PNS, axon degeneration and regeneration work together for an effective injury response. Axon degeneration is necessary to clear the distal axon and allow for axon regeneration.

3) The authors mention that overexpression of SARM1 promotes axon degeneration. This is an interesting finding. The authors should assess whether overexpression phenotype induces cell death or is truly selective axon degeneration.

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "TIR-1/SARM1 Inhibits Axon Regeneration" for further consideration by eLife. Your revised article has been evaluated by Piali Sengupta (Senior/Reviewing Editor) as well as two of the previous reviewers.

The manuscript is much improved but there are some remaining issues that need to be addressed, as outlined below. We realize that you have already done a lot of experiments at this point, so it is up to you whether you would like to perform additional experiments as suggested below.

Essential revisions:

1 While a subset of experiments addressing degeneration are performed after the single axotomy, others are performed after a double cut. It is unclear whether the mechanisms underlying degeneration in these different experimental paradigms is identical or different. Specifically, the authors should compare the role of the MAPK components for degeneration between the single and double-cut protocols.

2. The role of DLK-1/TIR-1 in axon degeneration needs to be clarified with further experiments. The data do not always support the conclusions and in some cases, the two contradict each other – please see points 8-10 from Reviewer 2.

3. An NAD-independent function for TIR-1 in promoting degeneration is of great interest but this requires further substantiation, particularly in the context of being used in the somewhat unusual double-cut injury model. Does the expression of NADase-dead activated TIR-1 in the single cut model and overexpression of TIR-1 in the spontaneous degeneration model promote degeneration?

Reviewer #1 (Recommendations for the authors):

Czech and colleagues present work showing that TIR-1 inhibits axon regeneration in C. elegans while also promoting degeneration in response to a double axon injury. They have adequately addressed the questions raised in the first round of review, and I feel that their work will attract broad interest in the field. Yet as they have added new and fascinating data, new questions necessarily arise. Below are several points that I feel the authors should clarify:

1) In Figure 2E, F the columns in the graph seem to be mislabeled and prevent accurate assessment of the data. Both present rescue experiments comparing injury-induced regeneration in WT, tir-1(qd4), and then what appear to be rescue experiments (qd4 + tir-1 in 2E, qd4 + Sarm1 in 2F). The third and fourth bars in each are labelled the same. Inferring from the text, perhaps one is a full-length cDNA and the other is a constitutively active form? The Figure legend does not clarify this.

2) The authors show a fascinating result – that tir-1-dependent axon degeneration does not require the NAD hydrolayse activity, thus making tir-1's pro-degenerative role quite unique. However, they cite a Neuron paper from Marc Freeman as a precedent for NAD-independent pro-degenerative roles of dSarm. This citation is misleading. In the Freeman paper, they show that axons adjacent to those degenerating in a classical NAD-dependent manner have cell biological defects (trafficking, etc.) that are dSarm-dependent but NAD-independent. In other words, Freeman does not show that dSarm kills axons in a NAD-independent manner. The authors should clarify this citation, but also, importantly, clarify in their text that the pro-degenerative role of tir-1 has no current parallel in other systems.

3) Regeneration is initially assayed after a single injury, for example in Figure 1 where the original tir-1 qd4 phenotype is shown. But by the time pathway analysis is performed in Figure 5 (showing genetic interactions with the MAPK pathway), regeneration is now characterized after double injury. The authors make no attempt to reconcile this very strange switch in paradigm. Was it simply out of convenience that these were the same animals assayed for injury-induced degeneration after double-cut? Regardless, the authors should show that the regenerative response (timing, magnitude, genetic dependences) are the same (or different?) between single and double cuts.

Reviewer #2 (Recommendations for the authors):

I commend the authors on making extensive revisions. The demonstration of a role for Tir-1 functioning through the Ask1 pathway to regulate axon regeneration is clear, although would be solidified with the genetic rescue. Unfortunately, despite the new data, the findings on Tir-1 and axon degeneration remain difficult to follow. The double cut system is of unclear physiological relevance, key experiments yield mechanistically distinct results that are not clearly addressed by the authors, and whatever Tir-1 does do for axon degeneration in C. elegans is quite different than its role in other characterized organisms and so the broader significance is questionable.

2) In Figure 2E what is the impact on the regeneration of overexpressing tir-1 in a wild-type background? Overexpressing tir-1 is only shown in a Tir-1 mutant background. However, overexpression of tir-1 may cause inhibition of regeneration even in wild type, which would then mean the data in the mutant background cannot be cleanly interpreted as a rescue.

4) 2F: Is a deletion of the autoinhibitory domain of human Sarm1 being expressed? Expression of such a construct in flies or mammalian cells leads to cell death. Could this just be making the cell sick so it can't regenerate as well? What is the influence of expressing activated human Sarm on regeneration in WT?

5) Figure 3: The key demonstration that Tir-1 promotes axon degeneration in worms is the finding that double-cut middle axon fragments degenerate slightly less in tir-1 mutant than in wild type. This key result is not rescued and should be.

7) Figure 4: I am assuming that in this figure an activated Tir deletion mutant is being used. If so, this is well known to cause toxicity in other systems due to unrestrained enzymatic activity. The authors find that overexpression of this toxic protein promotes axon degeneration when the axons are cut. The authors make the claim that the degenerative mechanism in worms is not NADase-dependent. The assay here is somewhat similar to what has been done in other systems, where the effect of Sarm1 is always enzyme-dependent. Does this prodegenerative effect require the NADase activity? It is unclear whether the different axon degeneration models used in this paper are testing the same mechanism of Tir-1. Assessing the NADase activity here would help establish whether or not Tir-1 is having a consistent effect in these different assays.

8) Figure 6: The role of DLK/Tir-1 is unclear and does not agree with other data in the paper. The authors show that DLK OE leads to less AxD (Figure 6D), and this decrease in AxD requires Tir-1 (suppressed by loss of Tir-1, so in the absence of Tir-1 there is more degeneration). This means that DLK is inhibiting axon degeneration and this effect requires Tir-1, and so Tir-1 is functioning to decrease AxD. However, this is in direct contradiction to the wild-type case, where a loss of function Tir-1 mutant leads to less axon degeneration, arguing that Tir-1 promotes AxD. This highlights how unclear the findings are with Tir-1 and axon degeneration. In one case Tir-1 promotes and in the other case it inhibits degeneration.

[Final comments from Reviewers.]

Reviewer #1 (Recommendations for the authors): The authors have done a commendable job addressing the serious concerns of the reviewers. While the relevance of the double-cut model to the broader field of Wallerian degeneration remains to be determined, these data will nevertheless spark important discussions in the field.

Reviewer #2 (Recommendations for the authors): I commend the authors on their thorough revisions. I know that we asked for a lot in these multiple rounds of revision. While I am sure this was frustrating, I believe it has made for a much stronger paper. All of my original concerns have been fully addressed.

https://doi.org/10.7554/eLife.80856.sa1

Author response

[Editors’ note: the authors resubmitted a revised version of the paper for consideration. What follows is the authors’ response to the first round of review.]

Whether TIR-1 regulates axon regeneration by depleting NAD+ levels is an important question since TIR-1 and its SARM1 orthologs are well-characterized NAD+ hydrolases1-8. To directly address whether TIR-1 regulates regeneration by hydrolysing NAD+, we CRISPR edited the endogenous tir-1 gene to create E788A and D773A mutations, which correspond to E642A and D627A mutations in SARM11-3,7. E788/E642 is a key catalytic glutamate that is required for the NADase activity of both the SARM1 and TIR-1 TIR domains1-4,6,7. The D773/D627 aspartate also promotes NAD+ catalysis and is required for the degeneration activity of SARM11,3. We found that both tir-1(D773A) and tir-1(E788A) mutant animals displayed wild type levels of axon regeneration, indicating TIR-1D773A and TIR-1E788A are functionally expressed and that the NADase activity of TIR-1 is not required for its ability to inhibit axon regeneration. Together, our data reveal that TIR-1 inhibits axon regeneration by interacting with the UNC-43/CAMKII – NSY-1/ASK1 MAPK – PMK1/p38 MAPK – DAF-19/RFX1-3 pathway, and independently from its NADase activity. We have added this result to the manuscript at Line 229 and in Figure 5A.

Multiple lines of evidence indicate that overexpression of TIR-1 induces axon degeneration. As suggested by the reviewer, we expressed tir-1::GFP or untagged tir-1 in GABA neurons and found that these constructs also induce degeneration (see Author response table 1). Note that each line is derived from a separate injection of tir-1 and therefore each express unique extrachromosomal arrays and number of tir-1 coding sequences. In contrast, expression of either GFP or mCherry alone did not disrupt axon morphology. We have also ruled out the possibility that the observed degeneration is caused by manipulating and imaging the worms. In mosaic animals, tir1 expressing axons degenerate while non-tir-1 expressing axons maintain a wild type morphology, even though they are in the same animal (Extended Data Figure 4). We conclude that expression of tir-1 does induce axon degeneration.

Author response table 1
Motor axon degeneration is dependent on expression of tir-1, but not mCherry.

tir1(qd4) animals were injected with the indicated expression constructs and the number of degenerating GABA motor axons was quantified. Expression of GFP or mCherry in motor axons is not sufficient to induce axon degeneration on their own. However, expression of untagged tir-1 or GFP-tagged tir-1 does induce axon degeneration.

DescriptionExpression Plasmid% Damaged Axons
controlunc-47p::GFP0
controlunc-47p::mCherry0
untagged line 1unc-47p::tir-19
untagged line 2unc-47p::tir-16
untagged line 3unc-47p::tir-111
untagged line 4unc-47p::tir-117
GFP tagged line 1unc-47p::tir-1::GFP13
GFP tagged line 2unc-47p::tir-1::GFP31
GFP tagged line 3unc-47p::tir-1::GFP6

Thank you for bringing this to our attention. Wild type axons are less likely to regenerate when the middle segment has degenerated and are more likely to regenerate when the middle segment has not degenerated. We agree that based on this data alone, the mechanisms associated with protection of the middle fragment could promote regeneration in wild type animals or the mechanisms associated with degeneration of the middle fragment could inhibit axon regeneration. Because we cannot differentiate between the two possibilities, we have removed this section from the manuscript.

We thank the reviewer for their comments. We agree that the implications of our finding that TIR-1 regulates axon regeneration are large and are of significant interest to the field, particularly if TIR-1 and Sarm1 are functional orthologs.

Multiple lines of evidence indicate TIR-1 is indeed a functional ortholog of dSARM and Sarm1. The three genes have evolved through speciation and not duplication9, share extensive sequence homology10, domain conservation1,10-12 and overall structural identity1-4,10,11,13. TIR-1, dSarm and Sarm1 also share molecular interactions and functions. They interact with conserved MAPK signal transduction pathways4,11,12,14-19 and all three are NAD+ hydrolases1-5,7,13. In addition, the genes regulate a number of shared cellular processes, including the innate immune response11,19-28 and cell death1,29,30.

Our data adds the significant finding that, like dSarm and SARM131,32, TIR-1 positively regulates injury induced axon degeneration. Specifically, we found (1) loss of TIR-1 suppresses degeneration of injured axons (Figure 3), (2) expression of a TIR-1 isoform that lacks its autoinhibitory N-terminus induces axon degeneration (Figure 4), and () TIR-1 interacts with DLK MAP Kinase signaling to regulate injuryinduced axon degeneration (Figure 6). In addition to regulating degeneration in response to injury, TIR-1, dSarm and SARM1 also regulate degeneration induced by disease, age and toxicity18,33-37.

The reviewer’s comment inspired us to specifically address the question of whether TIR-1 and human SARM1 might have orthologous function as cell-autonomous regulators of axon regeneration. To determine whether human Sarm1 (hSarm1) could function cell-autonomously to inhibit axon regeneration, we expressed low levels of hSARM1 in C. elegans GABA motor axons and asked whether it was sufficient to rescue the regeneration phenotype of tir-1(qd4) null animals. We found that cell-intrinsic expression of hSARM1 does rescue the motor axon regeneration phenotype (Figure 2F). Therefore, hSARM1 is capable of regulating axon regeneration within injured neurons, which is an additional line of evidence that hSARM1 and TIR-1 function are evolutionarily conserved. These results have been added to the manuscript at line 166.

Together, the sequence homology, structural conservation, and ability to regulate multiple shared functions, including degeneration and regeneration, of TIR-1, dSarm, and SARM1 strongly indicate they are functional orthologs. We therefore believe that our data significantly contribute to our understanding of both TIR-1/SARM1 function and the injury response.

We agree with the reviewer that finding only 20% of degeneration is suppressed in tir-1(-) animals is different from what might be expected given the extensive conservation of function between tir-1, dSarm, and Sarm1. However, directly comparing the magnitude of the effect that SARM1 has in inducing degeneration with the magnitude of the effect of TIR-1 is not possible, as these are different organisms, different neurons and different injury paradigms, all of which could influence the potential of an injured axon to degenerate. Indeed, many one-to-one orthologs regulate conserved processes in a complex manner that usually reflects a degree of plasticity in the adaptation of organisms to their environment. For example, the NADase activity of dSarm in Drosophila differs from that of its human, mouse or zebrafish SARM orthologs, such that dSarm generates an inverse ratio of cyclic ADPR to ADP-ribose compared to its orthologs7.

Here, in addition to the loss of function phenotype of tir-1(-) animals, we found a strong relationship between expression levels of TIR-1 and the degree of injury-induced axon degeneration. Increasing TIR-1 expression levels and activity causes profound and penetrant injury-induced degeneration (Figure 4A-E), while decreasing TIR-1 function reduces injury-induced degeneration (Figure 3G). These data, together with multiple lines of evidence that the sequence, structure, and functions of TIR-1 are conserved with dSarm and SARM1 (outlined in the previous response), indicate that TIR-1 is indeed a true ortholog of dSarm and SARM1 that positively regulates injury-induced degeneration.

As the reviewer notes, this raises the question of whether TIR-1 promotes axon degeneration similarly to dSarm and SARM1? We have now found that the modest but significant and penetrant reduction in degeneration in the absence of tir-1 can be attributed, at least in part, to the following: (1) the presence of degeneration in the absence of tir-1 function indicates additional mechanisms function in parallel to tir-1 to regulate injury-induced degeneration in C. elegans (Figure 3G), (2) tir-1 regulates degeneration with non-canonical NADase independent, potentially conserved mechanisms (Figure 6A and see following response) and (3) while distal axon fragments do not degenerate after a single axotomy, expression of an active tir-1b isoform strongly induces injury-dependent degeneration specifically in the distal fragment of singly axotomized axons. Because we found that tir-1 is expressed in the non-regenerating axons and does promote degeneration of completely severed middle fragments, these data indicate tir-1 function may be insufficiently activated or suppressed in specific cellular contexts (Figure 4A-E and see response to reviewer 3).

Knowing both the similarities and the differences in how TIR-1 and its SARM orthologs regulate their shared functions is critical to understanding the multiple ways in which regeneration and degeneration can be regulated, which is a prerequisite to learning how to manipulate the identified mechanisms in different contexts to repair and protect the nervous system.

We thank the reviewer for this suggestion. To determine whether TIR-1 regulates injury-induced degeneration through its NAD+ hydrolase activity, we CRISPR edited the key catalytic glutamate E788/E6422,7,13 (Figure 6A). Interestingly, the E788A/E642A mutation did not suppress injury-induced axon degeneration (Figure 6A). Since E788A is reported to render the TIR domain of C. elegans TIR-1 catalytically dead4, these data indicate that TIR-1 regulates axon degeneration independently of its NAD+ hydrolase activity. Our finding is consistent with recent results from the Freeman lab, which show mutation of the corresponding catalytic residue in dSarm (E1170A/E788A/E642A) only partially protects axons from degenerating14. Together, these results indicate that the ability to regulate degeneration independently of NAD+ hydrolysis may be conserved between TIR-1 and dSarm. What the additional NADase-independent mechanisms are and whether they are shared between TIR-1 and dSarm remain exciting open questions. These data have been added to the manuscript at Line 305 and in Figure 6A.

We also asked whether other components of the TIR domain are important for its ability to regulate axon degeneration. In addition to its key E788/E642 catalytic residue, TIR domain function requires a structure specific to catalytically active TIR domains called a SARM Specific (SS) loop (residues 622635), a BB loop (residues 594-605) thought to be important for intradomain interactions, and multiple residues important for protein multimerization and interaction with the autoinhibitory domain1-3,7,13. D627 is a conserved residue in the SS loops of C. elegans, human, and mouse TIR domains that is required for the NADase activity of SARM1 and for its ability to induce degeneration of injured DRGs, but not for its ability to multimerize1,2. Interestingly, we found that axon degeneration in tir-1(D773A) mutants was significantly disrupted compared to wild-type animals and was similar to that seen in tir-1(qd4) animals (Figure 6A). These data indicate the SS loop is important for TIR-1’s ability to promote degeneration of injured axons. Collectively with the findings that (1) mutation of the key catalytic glutamate E788A abolishes NADase activity in the TIR domain of TIR-14 but not degeneration (Figure 6A), and (2) the SS-loop residue D627 is required for degeneration and promotes the NADase activity of SARM1 through relatively unknown mechanisms1,2, our data indicate that D773A suppresses degeneration by disrupting an as yet unidentified function of TIR-1. Notably, D773A is not a null mutation, since it did not disrupt the ability of TIR-1 to inhibit regeneration (Figure 6A). It follows that TIR-1 regulates axon regeneration and degeneration by different mechanisms, potentially by directly interacting with components of the UNC-43-NSY-1PMK-1 and DLK-1-PMK-3 MAP kinase pathways on either side of the injury. These data have been added to the manuscript at Line 316 and in Figure 6A.

Future experiments will determine whether the amount of available NAD+ regulates injury-induced degeneration of C. elegans motor neurons independently of TIR-1’s NADase activity. Because this has not yet been determined, we have carefully avoided defining the injury-induced degeneration as Wallerian degeneration and instead noted that “The degeneration has similar morphological and temporal features as Wallerian degeneration, where after a brief latent period, the middle axon segment beads, fragments, and is eventually cleared (Figure 3B).” (Line 144)

As the reviewer points out, in non-injured chronically degenerating axons that overexpress the tir-1b isoform, our data agrees with previously published work15: DLK-1 functions downstream of TIR-1. As detailed below, our data also supports a bimodal relationship between TIR1 and DLK-1 that is dependent on the level of TIR-1 expression, the presence of the N-terminal autoinhibitory domain of TIR-1, and a traumatic injury event. While we believe this is a very interesting avenue of further investigation, we have removed this experiment from the manuscript to focus specifically on how the endogenous tir-1 gene regulates injury-induced degeneration.

A longstanding point of confusion in the Sarm1 literature is how MAPK signaling interacts with Sarm1. Separate studies found that MAPK functions downstream of Sarm1 to regulate degeneration of crushed retinal ganglion cells and upstream of Sarm115,17 to regulate degeneration of cultured mammalian dorsal root ganglia neurons and Drosophila motoneurons. These data indicate that the complex relationship between SARM1 and MAPK signaling may differ according to cell-type, assay, and species.

We found that TIR-1 interacts with DLK-1 MAPK signaling to regulate degeneration of injured motor axons. If, as the reviewer suggests, tir-1 promotes axon degeneration by inhibiting dlk-1 and pmk-3, one would expect that loss of dlk-1 and pmk-3 would increase axon degeneration relative to wild type. In the initial experiments, determining the nature of the relationship between tir-1 and dlk-1 was made difficult by a ceiling effect in the assay. Since almost all wild-type axons degenerate, it is difficult to determine whether a given mutant displays significantly more degeneration compared to wild type animals. This ceiling effect confounds interpretation of whether dlk-1 and pmk-3 suppress degeneration and if so, whether tir-1 functions upstream of dlk-1 and pmk-3 or whether tir-1 partially blocks dlk-1 and pmk-3 signaling.

To determine whether dlk-1 does inhibit axon degeneration and how it interacts with tir-1, we repeated the epistasis analysis using a gain of function allele of dlk-1. We found that overexpressing dlk-1 suppressed axon degeneration. This suppression is not likely to be a dominant negative effect since the opposite dlk-1 loss of function mutation does not suppress degeneration. We also found that dlk-1 induced protection was partially dependent on tir-1 function, indicating dlk-1 is capable of inhibiting degeneration of injured motor neurons by regulating tir-1 function and by regulating another as yet unknown mechanism of degeneration. Therefore, together with the multifaceted relationship between Sarm1 and MAPK signaling, our data indicate the highly conserved TIR-1, dSarm and Sarm1 proteins regulate injury-induced degeneration in coordination with shared MAPK signaling components. However, the precise interactions between TIR-1/Sarm1/dSarm, DLK1/MAPK/Wnd signaling, and degeneration are complex and likely depend on cell type, form of injury, and animal system. Identifying the similarities and differences between the various interactions is important to understanding how these complex genes regulate the injury response. We have added this data at Line 348 and included it in Figure 6.

We have modified this section to clarify the results. Since tir-1, nsy-1, and pmk-1 null mutants all inhibit axon regeneration to the same extent, and because the amount of regeneration in any pairwise combination of mutants is not additive or multiplicative compared to the single mutants, our data indicate that the three genes function in the same pathway to regulate axon regeneration. Line 274 and Figure 5D.

Our data indicate that dlk-1 function is required for axon regeneration in tir-1(-) and pmk-3(-) animals (Figure 5B). However, dlk-1 function is also absolutely required for regeneration of otherwise wildtype GABA motor axons38,39. Physical interpretations of genetic interactions with genes that are essential for a given function can be misleading if that function requires the activity of multiple distinctly regulated processes. In this experiment, since DLK-1 function is absolutely required for axon regeneration, we cannot distinguish whether TIR-1 signaling regulates the activity of DLK-1 or whether DLK-1 regulates an essential component of axon regeneration independently of TIR-1 signaling. However, since the enhanced regeneration in tir-1(qd4) mutants requires dlk-1, the genetic interaction does reveal that TIR-1 does not function downstream of, or redundantly with, DLK-1. Finally, the finding that loss of nsy-1 or pmk-1 causes the opposite regeneration phenotype to loss of dlk-1 or pmk-3 indicates the two sets of genes must not act redundantly. Together, our data indicate that tir-1 functions with nsy-1 and pmk-1 to inhibit axon regeneration, which is dependent on dlk-1. Line 249 and Figure 5B.

In the figures listed above, the amount of WT regeneration differs because they represent the frequency of regeneration after different types of injury. The frequency of regeneration in wild type animals is consistent within each particular type of assay.

We control against day-to-day and instrument to instrument variability in experiments by performing same day controls on the same axotomy rig, as is standard for C. elegans axon regeneration experiments. In addition to referring to the methods paper that describes these practices40, we have added a description of the axotomy protocol in the Materials and methods section.

Thank you for your positive comments.

We thank the reviewer for their positive comments about our regeneration data and its significance. We believe that the finding that TIR-1 regulates injury-induced degeneration is also significant because it demonstrates (1) that TIR-1 is a functional ortholog of the other SARMs (see response to reviewer two on page 4) and (2) that the mechanism by which TIR-1 regulates regeneration is distinct from how it regulates degeneration. We have revised the manuscript accordingly.

We also agree that understanding how degeneration is differentially regulated in C. elegans is a very interesting avenue of investigation. Our data contributes to our understanding of two exciting questions in this regard:

1) Why does TIR-1 only induce degeneration after double axotomy and not after a single axotomy?

We hypothesized that perhaps distal fragments do not degenerate after single axotomy because TIR-1 is not sufficiently activated by this type injury in C. elegans. TIR-1 and Sarm1 functions are autoinhibited by their N-terminal domains1,12,13,32,45-48. We found that when an endogenous isoform of tir-1 that lacks its N-terminal autoinhibitory domain is expressed at low levels, a single axotomy induces degeneration of distal stumps (Figure 4A-E). These data indicate that in wild type C. elegans, axons that have been injured once do not degenerate because the injury response was of insufficient magnitude or identity to activate TIR-1. In contrast, double axotomy induces an enhanced injury response of sufficient magnitude or identity to activate TIR-1 or overcome its suppression. In future experiments that are beyond the scope of the current manuscript we will investigate whether and how TIR-1 is differentially activated after single and double axotomy. We have added this data to the manuscript at Line 185 and Figure 4.

2) Why does loss of TIR-1 function not completely suppress axon degeneration?

Our finding that axons are only partially protected from degenerating in the absence of tir-1 (Figure 3G) reveals that multiple mechanisms regulate degeneration of injured motor axons in C. elegans. We asked whether ced-1, a homolog of Draper and Megf10, also contributes to the degeneration phenotype, since it functions in the hypodermis and muscle cells of C. elegans to remove axonal debris after injury in mechanosensory neurons49,50. Here, doubly axotomized motor axons in ced-1 loss of function mutant animals displayed wild type levels of axon degeneration (Extended Data Figure 3). Moreover, animals with loss of function mutations in both ced-1 and tir-1 phenocopied the degeneration observed in tir-1 animals. These results demonstrate that middle fragment degeneration after double injury does not depend on ced-1. Together with our finding that injury-induced degeneration is not completely dependent on tir1 (Figure 3G), these data indicate C. elegans motor axons have additional as yet unidentified regulators of injury-induced motor axon degeneration. We have added this data to the manuscript at Line 173 and Extended Data Figure 3.

We believe it will be difficult to definitively determine whether injuring an axon in two places is more or less artificial than injuring an axon precisely at the midline with a laser. However, since double axotomy induces degeneration of severed axons in C. elegans, we believe that the resulting injury response more accurately reflects that of other systems. This led us to ask whether the frequency of axon regeneration differs in the presence or absence of degeneration. The result is that, after double injury, axons regenerate less frequently when the middle fragment had degenerated compared to when the fragment remains intact. This data reveals a difference in the molecular mechanisms that regulate regeneration in the presence and absence of degeneration.

We agree with the reviewer that it is difficult to interpret this data and because we do not know how the interaction between regeneration and degeneration is regulated, we have removed figure 3H and mention of the relationship between regeneration and middle fragment degeneration from the manuscript.

References

  1. Summers, D. W., Gibson, D. A., DiAntonio, A. & Milbrandt, J. SARM1-specific motifs in the TIR domain enable NAD+ loss and regulate injury-induced SARM1 activation. Proc Natl Acad Sci U S A 113, E6271-E6280, doi:10.1073/pnas.1601506113 (2016).

  2. Loring, H. S., Icso, J. D., Nemmara, V. V. & Thompson, P. R. Initial Kinetic Characterization of Sterile Α and Toll/Interleukin Receptor Motif-Containing Protein 1. Biochemistry 59, 933-942, doi:10.1021/acs.biochem.9b01078 (2020).

  3. Loring, H. S. et al. A phase transition enhances the catalytic activity of SARM1, an NAD(+) glycohydrolase involved in neurodegeneration. ELife 10, doi:10.7554/eLife.66694 (2021).

  4. Peterson, N. D. et al. Pathogen infection and cholesterol deficiency activate the C. elegans p38 immune pathway through a TIR-1/SARM1 phase transition. ELife 11, doi:10.7554/eLife.74206 (2022).

  5. Gerdts, J., Summers, D. W., Milbrandt, J. & DiAntonio, A. Axon Self-Destruction: New Links among SARM1, MAPKs, and NAD+ Metabolism. Neuron 89, 449-460, doi:10.1016/j.neuron.2015.12.023 (2016).

  6. Horsefield, S. et al. NAD+ cleavage activity by animal and plant TIR domains in cell death pathways. Science 365, 793-799, doi:doi:10.1126/science.aax1911 (2019).

  7. Essuman, K. et al. The SARM1 Toll/Interleukin-1 Receptor Domain Possesses Intrinsic NAD(+) Cleavage Activity that Promotes Pathological Axonal Degeneration. Neuron 93, 1334-1343 e1335, doi:10.1016/j.neuron.2017.02.022 (2017).

  8. Essuman, K. et al. TIR Domain Proteins Are an Ancient Family of NAD(+)-Consuming Enzymes. Curr Biol 28, 421-430 e424, doi:10.1016/j.cub.2017.12.024 (2018).

  9. Thompson, J. D., Gibson, T. J. & Higgins, D. G. Multiple sequence alignment using ClustalW and ClustalX. Curr Protoc Bioinformatics Chapter 2, Unit 2 3, doi:10.1002/0471250953.bi0203s00 (2002).

  10. Mink, M., Fogelgren, B., Olszewski, K., Maroy, P. & Csiszar, K. A novel human gene (SARM) at chromosome 17q11 encodes a protein with a SAM motif and structural similarity to Armadillo/β-catenin that is conserved in mouse, Drosophila, and Caenorhabditis elegans. Genomics 74, 234-244, doi:10.1006/geno.2001.6548 (2001).

  11. Liberati, N. T. et al. Requirement for a conserved Toll/interleukin-1 resistance domain protein in the Caenorhabditis elegans immune response. Proc Natl Acad Sci U S A 101, 6593-6598, doi:10.1073/pnas.0308625101 (2004).

  12. Chuang, C. F. & Bargmann, C. I. A Toll-interleukin 1 repeat protein at the synapse specifies asymmetric odorant receptor expression via ASK1 MAPKKK signaling. Genes Dev 19, 270-281, doi:10.1101/gad.1276505 (2005).

  13. Horsefield, S. et al. NAD+ cleavage activity by animal and plant TIR domains in cell death pathways. Science 365, 793-799, doi:10.1126/science.aax1911 (2019).

  14. Hsu, J. M. et al. Injury-Induced Inhibition of Bystander Neurons Requires dSarm and Signaling from Glia. Neuron 109, 473-487 e475, doi:10.1016/j.neuron.2020.11.012 (2021).

  15. Yang, J. et al. Pathological axonal death through a MAPK cascade that triggers a local energy deficit. Cell 160, 161-176, doi:10.1016/j.cell.2014.11.053 (2015).

  16. Miller, B. R. et al. A dual leucine kinase-dependent axon self-destruction program promotes Wallerian degeneration. Nat Neurosci 12, 387-389, doi:10.1038/nn.2290 (2009).

  17. Walker, L. J. et al. MAPK signaling promotes axonal degeneration by speeding the turnover of the axonal maintenance factor NMNAT2. ELife 6, doi:10.7554/eLife.22540 (2017).

  18. Ding, C., Wu, Y., Dabas, H. & Hammarlund, M. Activation of the CaMKII-Sarm1-ASK1-p38 MAP kinase pathway protects against axon degeneration caused by loss of mitochondria. ELife 11, doi:10.7554/eLife.73557 (2022).

  19. Couillault, C. et al. TLR-independent control of innate immunity in Caenorhabditis elegans by the TIR domain adaptor protein TIR-1, an ortholog of human SARM. Nat Immunol 5, 488-494, doi:10.1038/ni1060 (2004).

  20. Hou, Y. J. et al. SARM is required for neuronal injury and cytokine production in response to central nervous system viral infection. J Immunol 191, 875-883, doi:10.4049/jimmunol.1300374 (2013).

  21. Gurtler, C. et al. SARM regulates CCL5 production in macrophages by promoting the recruitment of transcription factors and RNA polymerase II to the Ccl5 promoter. J Immunol 192, 4821-4832, doi:10.4049/jimmunol.1302980 (2014).

  22. Carty, M. et al. Cell Survival and Cytokine Release after Inflammasome Activation Is Regulated by the Toll-IL-1R Protein SARM. Immunity 50, 1412-1424 e1416, doi:10.1016/j.immuni.2019.04.005 (2019).

  23. Carty, M. et al. The human adaptor SARM negatively regulates adaptor protein TRIF-dependent Toll-like receptor signaling. Nat Immunol 7, 1074-1081, doi:10.1038/ni1382 (2006).

  24. Akhouayri, I., Turc, C., Royet, J. & Charroux, B. Toll-8/Tollo negatively regulates antimicrobial response in the Drosophila respiratory epithelium. PLoS Pathog 7, e1002319, doi:10.1371/journal.ppat.1002319 (2011).

  25. Baral, P. & Utaisincharoen, P. Sterile-α- and armadillo motif-containing protein inhibits the TRIF-dependent downregulation of signal regulatory protein α to interfere with intracellular bacterial elimination in Burkholderia pseudomallei-infected mouse macrophages. Infect Immun 81, 3463-3471, doi:10.1128/IAI.00519-13 (2013).

  26. Dinić, M. et al. Probiotic-mediated p38 MAPK immune signaling prolongs the survival of Caenorhabditis elegans exposed to pathogenic bacteria. Scientific Reports 11, doi:10.1038/s41598021-00698-5 (2021).

  27. Kurz, C. L., Shapira, M., Chen, K., Baillie, D. L. & Tan, M. W. Caenorhabditis elegans pgp-5 is involved in resistance to bacterial infection and heavy metal and its regulation requires TIR-1 and a p38 map kinase cascade. Biochemical and Biophysical Research Communications 363, 438-443, doi:10.1016/j.bbrc.2007.08.190 (2007).

  28. Pudla, M., Limposuwan, K. & Utaisincharoen, P. Burkholderia pseudomallei-induced expression of a negative regulator, sterile-α and Armadillo motif-containing protein, in mouse macrophages: a possible mechanism for suppression of the MyD88-independent pathway. Infect Immun 79, 29212927, doi:10.1128/IAI.01254-10 (2011).

  29. Hayakawa, T. et al. Regulation of anoxic death in Caenorhabditis elegans by mammalian apoptosis signal-regulating kinase (ASK) family proteins. Genetics 187, 785-792, doi:10.1534/genetics.110.124883 (2011).

  30. Kim, W., Underwood, R. S., Greenwald, I. & Shaye, D. D. OrthoList 2: A New Comparative Genomic Analysis of Human and Caenorhabditis elegans Genes. Genetics 210, 445-461, doi:10.1534/genetics.118.301307 (2018).

  31. Osterloh, J. M. et al. dSarm/Sarm1 is required for activation of an injury-induced axon death pathway. Science 337, 481-484, doi:10.1126/science.1223899 (2012).

  32. Gerdts, J., Summers, D. W., Sasaki, Y., DiAntonio, A. & Milbrandt, J. Sarm1-mediated axon degeneration requires both SAM and TIR interactions. J Neurosci 33, 13569-13580, doi:10.1523/JNEUROSCI.1197-13.2013 (2013).

  33. Lezi, E. et al. An Antimicrobial Peptide and Its Neuronal Receptor Regulate Dendrite Degeneration in Aging and Infection. Neuron 97, 125-138 e125, doi:10.1016/j.neuron.2017.12.001 (2018).

  34. Veriepe, J., Fossouo, L. & Parker, J. A. Neurodegeneration in C. elegans models of ALS requires TIR1/Sarm1 immune pathway activation in neurons. Nat Commun 6, 7319, doi:10.1038/ncomms8319 (2015).

  35. Henninger, N. et al. Attenuated traumatic axonal injury and improved functional outcome after traumatic brain injury in mice lacking Sarm1. Brain 139, doi:10.1093/brain/aww001 (2016).

  36. Massoll, C., Mando, W. & Chintala, S. K. Excitotoxicity upregulates SARM1 protein expression and promotes Wallerian-like degeneration of retinal ganglion cells and their axons. Invest Ophthalmol Vis Sci 54, 2771-2780, doi:10.1167/iovs.12-10973 (2013).

  37. Geisler, S. et al. Prevention of vincristine-induced peripheral neuropathy by genetic deletion of SARM1 in mice. Brain 139, 3092-3108, doi:10.1093/brain/aww251 (2016).

  38. Hammarlund, M., Nix, P., Hauth, L., Jorgensen, E. M. & Bastiani, M. Axon regeneration requires a conserved MAP kinase pathway. Science 323, 802-806, doi:10.1126/science.1165527 (2009).

  39. Yan, D., Wu, Z., Chisholm, A. D. & Jin, Y. The DLK-1 kinase promotes mRNA stability and local translation in C. elegans synapses and axon regeneration. Cell 138, 1005-1018, doi:10.1016/j.cell.2009.06.023 (2009).

  40. Byrne, A. B., Edwards, T. J. & Hammarlund, M. in vivo laser axotomy in C. elegans. J Vis Exp, doi:10.3791/2707 (2011).

  41. Wang, Q. et al. Sarm1/Myd88-5 Regulates Neuronal Intrinsic Immune Response to Traumatic Axonal Injuries. Cell Rep 23, 716-724, doi:10.1016/j.celrep.2018.03.071 (2018).

  42. Neumann, B., Nguyen, K. C., Hall, D. H., Ben-Yakar, A. & Hilliard, M. A. Axonal regeneration proceeds through specific axonal fusion in transected C. elegans neurons. Dev Dyn 240, 1365-1372, doi:10.1002/dvdy.22606 (2011).

  43. Sasaki, Y., Nakagawa, T., Mao, X., DiAntonio, A. & Milbrandt, J. NMNAT1 inhibits axon degeneration via blockade of SARM1-mediated NAD(+) depletion. ELife 5, doi:10.7554/eLife.19749 (2016).

  44. Kim, K. W. et al. Expanded genetic screening in Caenorhabditis elegans identifies new regulators and an inhibitory role for NAD(+) in axon regeneration. ELife 7, doi:10.7554/eLife.39756 (2018).

  45. Shen, C. et al. Multiple domain interfaces mediate SARM1 autoinhibition. Proc Natl Acad Sci U S A 118, doi:10.1073/pnas.2023151118 (2021).

  46. Jiang, Y. et al. The NAD(+)-mediated self-inhibition mechanism of pro-neurodegenerative SARM1. Nature 588, 658-663, doi:10.1038/s41586-020-2862-z (2020).

  47. Figley, M. D. et al. SARM1 is a metabolic sensor activated by an increased NMN/NAD(+) ratio to trigger axon degeneration. Neuron 109, 1118-1136 e1111, doi:10.1016/j.neuron.2021.02.009 (2021).

  48. Sporny, M. et al. Structural basis for SARM1 inhibition and activation under energetic stress. ELife 9, doi:10.7554/eLife.62021 (2020).

  49. Nichols, A. L. A. et al. The Apoptotic Engulfment Machinery Regulates Axonal Degeneration in C. elegans Neurons. Cell Rep 14, 1673-1683, doi:10.1016/j.celrep.2016.01.050 (2016).

  50. Chiu, H. et al. Engulfing cells promote neuronal regeneration and remove neuronal debris through distinct biochemical functions of CED-1. Nat Commun 9, 4842, doi:10.1038/s41467-018-07291-x (2018).

[Editors’ note: what follows is the authors’ response to the second round of review.]

Essential revisions:

1. While a subset of experiments addressing degeneration are performed after the single axotomy, others are performed after a double cut. It is unclear whether the mechanisms underlying degeneration in these different experimental paradigms is identical or different. Specifically, the authors should compare the role of the MAPK components for degeneration between the single and double-cut protocols.

In response to the reviewers’ comments, we have added significantly to our finding that DLK-1 and PMK-3 MAP kinase components regulate injury induced axon degeneration in wild type animals. Further investigation of double cut axons has revealed that DLK-1 and PMK-3 inhibit wild type degeneration of middle axon fragments that are severed from both the cell body and synaptic partner (detailed below in Essential revision 2). Whether the DLK-1 MAP kinase pathway also inhibits wild type degeneration of distal fragments after a single injury is an interesting question. Previously published work demonstrates the distal fragments of singly injured axons do not readily degenerate in wild type or tir-1(-) backgrounds 24 hours after injury (Nichols et al., 2016; Yanik et al., 2004). Here, we asked whether that inability to degenerate is determined by DLK-1 by investigating whether the severed distal fragments of dlk-1(-) or dlk-1(oe) axons degenerate after a single injury. We found no significant difference in the number of axons that degenerate between either of the dlk-1 mutants and wild type animals, indicating DLK-1 does not promote or inhibit degeneration of distal fragments in this injury model (Figure 7—figure supplement 2). Very rarely, a severed distal fragment was absent from wild type or dlk-1(oe) animals; however, whether or not this represents an active form of degeneration, it is not regulated by DLK-1 MAPK signaling. Similarly, we did not observe distal fragment degeneration in singly injured pmk-3(-), nsy-1(-), or pmk-1(-) axons, indicating these MAP kinases do not inhibit degeneration of wild type distal axon fragments after a single injury.

In summary, our analysis of the role of DLK MAPK in C. elegans across diverse models of degeneration reveals: (1) DLK-1 inhibits degeneration of middle axon fragments that are severed from both the proximal and distal axon fragments by a double injury (See response to Essential revision 2 and Figure 7B,C), and (2) DLK-1 does not regulate degeneration of distal axon fragments after a single injury (Figure 7—figure supplement 2). We conclude that degeneration of C. elegans motor axons is differentially regulated by DLK-1, according to type of insult.

The dichotomous role of DLK-1 in axon degeneration agrees with the duality of MAPK signaling in established models of injury-induced degeneration. For example, Wallenda (Wnd), the DLK-1 homolog in Drosophila melanogaster, inhibits axon degeneration in motor axons that have undergone a conditioning lesion (Xiong & Collins, 2012). However, Wallenda also promotes degeneration in singly severed olfactory receptor axons and mammalian DLK promotes degeneration of severed dorsal root ganglion axons (Ghosh et al., 2011; Miller et al., 2009). While it is unlikely that the small amount of time between axotomies in the double injury model induces a conditioning lesion; the comparison demonstrates that Wnd and DLK1 activity is differentially determined by specific types of injury and by cell type. In relation to SARM signaling, separate studies have found that MAPK signaling functions downstream of SARM1 to regulate degeneration of crushed retinal ganglion cells, and upstream of SARM1 to regulate degeneration of cultured mammalian dorsal root ganglia neurons as well as Drosophila motoneurons (Summers et al., 2018; Walker et al., 2017; Yang et al., 2015). Identifying the similarities and differences between the various interactions across species and types of injury will be important to understanding how these complex genes regulate the injury response.

These data have been added to the manuscript as Figure 7—figure supplement 2 and are described at Line 424.

2. The role of DLK-1/TIR-1 in axon degeneration needs to be clarified with further experiments. The data do not always support the conclusions and in some cases, the two contradict each other – please see points 8-10 from Reviewer 2.

We performed additional experiments to clarify the role of TIR-1 and DLK-1 in injury-induced axon degeneration. We previously found that loss of tir-1 function reduces the degeneration frequency of middle axon fragments that are severed from both the proximal and distal fragments (Figure 3G). Therefore, TIR-1 promotes degeneration of severed GABA motor axon fragments. Our analysis of loss of function mutations in dlk-1 and its canonical downstream MAPK, pmk-3, revealed that both DLK-1 and PMK-3 inhibit degeneration of severed middle axon fragments. Specifically, we found that loss of dlk-1 or pmk-3 function alone did not significantly affect degeneration of severed middle fragments compared to wild type animals; however, loss of dlk-1 or pmk-3 function increased the frequency of middle fragment degeneration in tir-1(qd4) mutants (Figure 7B). Therefore, wild type DLK-1 and PMK-3 inhibit axon degeneration, at least in the absence of tir-1(qd4). As outlined in the above response to Essential revision 1, the conclusion that DLK-1 inhibits axon degeneration is analogous to the finding that Wallenda, its Drosophila homolog, also inhibits degeneration of injured motor neurons that have undergone a conditioning lesion (Xiong & Collins, 2012). However, it differs from the finding that Wallenda promotes degeneration in singly severed olfactory receptor axons and that DLK promotes degeneration of severed dorsal root ganglion axons (Ghosh et al., 2011; Miller et al., 2009).

In response to the reviewers’ comments, we further investigated the epistatic relationship between tir-1 and dlk-1 using animals that overexpress dlk-1 (dlk-1(oe)). We first outcrossed the tir-1(qd4) mutation from dlk-1(oe); tir-1(qd4) animals because we found that the integrated high copy dlk-1 array was being silenced over successive generations. Therefore, rebuilding the strains allowed us to compare the phenotypes of animals expressing the same transgenic allele, which originated from the same strain and had not been passaged for multiple generations. We verified that the newly isolated dlk-1(oe) and dlk-1(oe); tir-1(qd4) animals contained and expressed similar amounts of mRNA from the integrated dlk-1 transgene using qRT-PCR (Figure 7—figure supplement 1). Axons in tir-1(-) and dlk-1(oe) animals degenerated significantly less frequently than in wild type animals (Figure 7B,C). These data reinforce that tir-1 promotes degeneration and dlk-1 inhibits degeneration. The suppression of degeneration in doubly injured dlk-1(oe) axons was not dependent on tir-1 function (Figure 7B). These data could be interpreted in at least two ways. First, dlk-1 inhibits degeneration downstream of tir-1 and therefore removing tir-1 function has no consequence on a gene that is already activated. Alternatively, dlk-1 inhibits degeneration at least in part by suppressing the function of tir-1. If so, loss of tir-1 function does not affect the dlk-1(oe) phenotype because tir-1 has already been inhibited.

We next asked whether blocking dlk-1 signaling in the dlk-1(oe) strain would reveal the nature of the genetic relationship between dlk-1 signaling and tir-1(qd4). We previously found that loss of function mutations in dlk-1 and its canonical downstream MAPK pmk-3 cause similar degeneration phenotypes, both in the presence and absence of tir-1 function (Figure 7B). The phenotypic similarity prompts the hypothesis that pmk-3 functions downstream of dlk-1 to inhibit axon degeneration. We asked whether the inhibition of degeneration observed in dlk1(oe) animals does depend on pmk-3 function by quantifying degeneration in dlk-1(oe); pmk3(-) animals. Loss of pmk-3 function rescued degeneration in animals that overexpress dlk-1 (Figure 7B); therefore, dlk-1 functions with its downstream MAPK pmk-3 to inhibit degeneration.

With the ability to block dlk-1 function in dlk-1(oe) animals, we could further investigate how dlk-1–pmk-3 signaling interacts with tir-1. Loss of pmk-3 function rescued degeneration in dlk-1(oe); tir-1(-) animals, indicating DLK-1–PMK-3 signaling does not inhibit degeneration by inhibiting TIR-1 (Figure 7B). Instead, DLK-1–PMK-3 signaling inhibits an as yet unknown effector of degeneration that can function even in the absence of TIR-1. Given the ceiling effect in the assay, we cannot rule out the possibility that pmk-3 also functions downstream of another inhibitor of degeneration; however, we can conclude that DLK-1-PMK-3 signaling is genetically epistatic to TIR-1.

These data have been added to the manuscript as Figure 7 and Figure 7—figure supplement 1, and is described in Lines 413-427.

3. An NAD-independent function for TIR-1 in promoting degeneration is of great interest but this requires further substantiation, particularly in the context of being used in the somewhat unusual double-cut injury model. Does the expression of NADase-dead activated TIR-1 in the single cut model and overexpression of TIR-1 in the spontaneous degeneration model promote degeneration?

We found that in wild type axons, injury-induced regeneration and degeneration is regulated by non-canonical NADase-independent TIR-1 activity after a double injury (Figures 5A and 6A). In response to the reviewers’ comment, we asked whether the NADase activity of TIR-1 was required for its ability to promote degeneration in two additional models.

A) Spontaneous axon degeneration induced by overexpression of TIR-1b. We first asked whether overexpressing TIR-1b is capable of inducing NADase dependent axon degeneration in the absence of an acute injury. To do so, we mutated E788A/E642A and D773A/D627A in the TIR-1b expression construct, injected each into tir-1(qd4) animals, and isolated multiple independent transgenic strains. Expressing TIR-1b(E788A/E642A), which encodes a mutation that abolishes the NADase activity of TIR-1 and SARM1, did not induce spontaneous axon degeneration (Figure 6B) (Essuman et al., 2017; Horsefield et al., 2019; Loring et al., 2021; Peterson et al.; Summers et al., 2016). Nor did expression of TIR-1b(D773A/D627A), which encodes a mutation that disrupts multiple aspects of TIR-1 function, including NADase activity (Loring et al., 2021; Summers et al., 2016). The absence of spontaneous degeneration in the presence of TIR-1b(D773A) and TIR1b(E788A) is not attributed to differential expression of the respective transgenes compared to overexpression of the wild type tir-1b transgene, which was determined by quantifying expression of the same fluorescent tag on each protein (Figure 6C). However, because a small but statistically insignificant number of damaged axons were observed, we cannot rule out the possibility that greater TIR-1b(D773A) or TIR-1b(E788A) expression might induce NADase independent degeneration. Together with our previous results, we find that D773/D627 is required for both injury induced and spontaneous degeneration, while E788/E642 is only required for spontaneous degeneration. These data indicate D773A disrupts multiple aspects of TIR-1 function, and the NADase activity of TIR-1 is required to induce spontaneous degeneration in uninjured axons.

We note that these results do not substantiate or confound the finding that TIR-1 regulates middle fragment degeneration independently of its NADase activity. Rather, our data strongly support the conclusion that TIR-1 is capable of actively inducing degeneration of middle axon fragments without its enzymatic activity. The strength of this conclusion is owed to: (1) The nature of the E788A mutation. E788A is a CRISPR generated mutation of the endogenous tir-1 locus, and therefore is a highly specific and system-wide mutation. (2) Previously published biochemical evidence that E788A does abolish the NADase activity of TIR-1 (Peterson et al.). And (3) The finding that middle fragment degeneration in tir-1 mutants can be genetically suppressed by the loss of dlk1 or pmk-3 function (Figure 7B) indicates that TIR-1 regulates degeneration of severed middle fragments by inhibiting DLK-1–PMK-3 MAPK signaling. The ability to drive degeneration independently of NADase activity is potentially unique to the TIR-1 homolog, as the enzymatic activity of dSarm and SARM1 is required for their ability to regulate injury-induced degeneration (Brace et al., 2022; Herrmann et al., 2022; Hsu et al., 2021). However, dSARM has been found to regulate axonal transport after injury in an NADase independent manner demonstrating a key role for dSARM not requiring its NADase function (Hsu et al., 2021).

B) Degeneration of the distal axon fragment after a single injury in neurons that overexpress TIR-1b. In response to the reviewers’ comments, we also asked whether specifically disrupting the NADase activity of TIR-1b would block its ability to induce distal axon fragment degeneration after a single injury. We found that the distal stumps of axons that have been severed once do not degenerate significantly more frequently in axons that overexpress the tir-1(E788A) transgene compared to wild type axons. The significance of this data is diminished because the experiment combines injury and overexpression of tir-1, which we have found are regulated by different mechanisms.

Together, our finding that TIR-1 promotes degeneration independently of its NADase activity in wild type GABA motor axons that have been severed twice, along with the finding that TIR1 is capable of inhibiting degeneration of PVQ interneurons in the absence of mitochondria (Ding et al., 2022), and the inability of TIR-1 to promote spontaneous motor axon degeneration in the absence of its NADase activity, indicate TIR-1 regulates degeneration using multiple context specific mechanisms, at least two of which differ from how dSarm and SARM1 have been shown to regulate degeneration.

These data have been added to the manuscript as Figure 6 and described in Lines 389-404.

4. As you are no doubt aware, Marc Hammarlund recently showed that TIR-1 inhibits axon degeneration (albeit in a different neuron type) in contrast to the findings described here. This work is referenced only briefly; the contradictory findings and possible underlying reasons should be discussed in more detail.

We have included a more detailed description of the contrast between the protective and destructive functions of TIR-1: The Hammarlund lab found that TIR-1 can inhibit spontaneous axon degeneration of PVQ interneurons in mutant ric-7(-) animals (Ding et al., 2022), which lack mitochondria (Rawson et al., 2014). Together with our finding that TIR-1 promotes injury-induced degeneration of wild type GABA motor axons, along with previous findings that TIR-1 is required for degeneration in a C.elegans model of ALS, functions in the epidermis to promote age-dependent degeneration of PVD dendrites, and induces spontaneous degeneration when forced to multimerize, the data indicate endogenous TIR-1 differentially regulates degeneration in C. elegans in response to specific insults, contexts and potentially cell types (Ding et al., 2022; Lezi et al., 2018; Loring et al., 2021; Veriepe et al., 2015). Questions raised specifically from a comparison of the opposite roles of TIR-1 in response to injury and absence of mitochondria include whether a specific injury signal determines that TIR-1 takes on a destructive role and whether TIR-1 can protect multiple types of neurons against the spontaneous degeneration that is induced by the absence of mitochondria. Looking downstream, we found the map kinase pmk-3 is epistatic to tir-1 and inhibits injury-induced axon degeneration. Interestingly, pmk-3 was also found to inhibit spontaneous degeneration in the absence of mitochondria, suggesting the difference between promotion and suppression of degeneration may ultimately be attributed to differential regulation of PMK-3 in these two contexts.

Reviewer #1 (Recommendations for the authors):

Czech and colleagues present work showing that TIR-1 inhibits axon regeneration in C. elegans while also promoting degeneration in response to a double axon injury. They have adequately addressed the questions raised in the first round of review, and I feel that their work will attract broad interest in the field. Yet as they have added new and fascinating data, new questions necessarily arise. Below are several points that I feel the authors should clarify:

Thank you for your comments and enthusiasm for our findings.

- Regeneration is initially assayed after a single injury, for example in Figure 1 where the original tir-1 qd4 phenotype is shown. But by the time pathway analysis is performed in Figure 5 (showing genetic interactions with the MAPK pathway), regeneration is now characterized after double injury. The authors make no attempt to reconcile this very strange switch in paradigm. Was it simply out of convenience that these were the same animals assayed for injury-induced degeneration after double-cut? Regardless, the authors should show that the regenerative response (timing, magnitude, genetic dependences) are the same (or different?) between single and double cuts.

We have clarified the text to highlight the rationale for using the double injury model to investigate how TIR-1 regulates axon regeneration. We used the double injury model because we found that a double injury induces both degeneration and regeneration of wild type motor axons in the 24 hours following injury. In contrast, the distal segments of axons that have been severed once remain largely intact (Nichols et al., 2016; Yanik et al., 2004). The ability to observe both processes using one injury model is essential for comparing how TIR-1 regulates axon regeneration and degeneration within the same injured wild type axon and in the context of the same injury response.

We have added this clarification to the manuscript at Line 239-249.

We agree that it is also interesting to determine whether the regenerative responses are the same or different between axons that have been cut once or twice. We compared the timing, magnitude, genetic dependencies, and requirement for the NADase activity of TIR-1 between the two injury models. Each set of experiments is described below.

Magnitude and timing: Significantly less axons regenerate when they are injured twice, compared to when they are injured once. For example, in Figure 1D, 60.53% of singly cut wild type axons regenerate in the 24hrs following injury, compared to Figure 5B, in which 37.38% of doubly cut wild type axons regenerate in the same amount of time. Loss of tir-1(-) significantly rescues regeneration after either a single or double injury (Figure 1D and 5B). Therefore, while axons are less capable of regenerating after a double injury compared to a single injury, tir-1 inhibits axon regeneration in multiple contexts. In response to the reviewers’ original comments, we removed the discussion of what might be responsible for the difference in regenerative ability from the previous version of the manuscript.

Genetic dependencies: TIR-1 regulates regeneration using similar mechanisms after different types of injury. We found that regeneration was increased relative to wild type in tir-1(-), nsy1(-), and pmk-1(-) animals after a single injury (Figure 5—figure supplement 1). Therefore, TIR1, NSY-1 and PMK-1 inhibit regeneration after either a single or double injury. In addition, we found that the increased regeneration phenotypes are not additive in tir-1(-); nsy-1(-) or tir1(-); pmk-1(-) double mutants compared to any of the single mutants, indicating the two MAP kinases function in the same pathway as TIR-1 to inhibit axon regeneration (Figure 5—figure supplement 1). Together with our previous data, these data reveal that TIR-1 inhibits regeneration with NSY-1 and PMK-1 in response to multiple types of injury.

Reviewer #2 (Recommendations for the authors):

I commend the authors on making extensive revisions. The demonstration of a role for Tir-1 functioning through the Ask1 pathway to regulate axon regeneration is clear, although would be solidified with the genetic rescue. Unfortunately, despite the new data, the findings on Tir-1 and axon degeneration remain difficult to follow. The double cut system is of unclear physiological relevance, key experiments yield mechanistically distinct results that are not clearly addressed by the authors, and whatever Tir-1 does do for axon degeneration in C. elegans is quite different than its role in other characterized organisms and so the broader significance is questionable.

Thank you for your comments. We have addressed each one below.

- In Figure 2E what is the impact on the regeneration of overexpressing tir-1 in a wild-type background? Overexpressing tir-1 is only shown in a Tir-1 mutant background. However, overexpression of tir-1 may cause inhibition of regeneration even in wild type, which would then mean the data in the mutant background cannot be cleanly interpreted as a rescue.

We asked whether overexpressing tir-1b in wild type GABA neurons causes an additive phenotype with tir-1(qd4) or if it rescues the regeneration phenotype of tir-1(qd4) animals. We found that overexpression of tir-1b did not inhibit regeneration in wild type animals (Figure 2E). Therefore, we conclude that cell-autonomous expression of tir-1b in GABA neurons is sufficient to inhibit axon regeneration in animals that otherwise lack tir-1.

We have added the above data to Figure 2E.

- 2F: Is a deletion of the autoinhibitory domain of human Sarm1 being expressed? Expression of such a construct in flies or mammalian cells leads to cell death. Could this just be making the cell sick so it can't regenerate as well? What is the influence of expressing activated human Sarm on regeneration in WT?

Although ‘sick’ is difficult to define molecularly, we found that expressing a low copy array of hSarm1 lacking its autoinhibitory N-terminal, herein referred to as hSARM1SAMTIR, in otherwise wild type animals resulted in wild type amounts of axon regeneration (Figure 2F). The severed proximal axon fragments remained intact, whether they did or did not regenerate. Note the wild type + pGABA::hSARM1SAMTIR strain was derived from the tir-1(qd4) + pGABA::hSARM1SAMTIR (Line 1) strain; therefore, they express the same extrachromosomal array. Together, our data indicate hSARM1SAMTIR is sufficient to inhibit axon regeneration cell-autonomously in C. elegans motor neurons.

We have added the above data to Figure 2F.

- Figure 3: The key demonstration that Tir-1 promotes axon degeneration in worms is the finding that double-cut middle axon fragments degenerate slightly less in tir-1 mutant than in wild type. This key result is not rescued and should be.

Axon degeneration was restored in tir-1(qd4) animals that express tir-1b specifically in their GABA neurons. The same extrachromosomal tir-1b expression array has no significant effect on middle fragment degeneration in wild type animals. The ceiling effect in this assay and the use of extrachromosomal overexpression arrays prevent us from conclusively determining whether expressing multiple copies of tir-1b could activate more degeneration than observed in wild type animals or if a single copy of tir-1b in GABA neurons would restore degeneration to wild type levels. However, these data demonstrate that expressing multiple copies of tir-1b in GABA neurons is sufficient to restore degeneration in tir-1(qd4) animals.

We have added the above data to Figure 3H.

- Two recent PLos Genetics papers (Herrmann et al. and Brace et al.) demonstrated that Sarm1 Nadase activity is necessary to regulate the Ask1 MAP kinase pathway in Drosophila. This is in contradiction to the findings here and should be discussed. Ideally, the authors would test the requirement of the Tir-1 NADase in the classic left-right asymmetry paradigm in C elegans (known to require Ask1), but that might be beyond the scope of this paper.

As the reviewer indicates, our finding that TIR-1 regulates regeneration in coordination with NSY-1 signaling and independently of its NADase activity contrasts with recent findings that dSarm regulates neuromuscular junction formation and glial phagocytosis using both its NADase activity and interaction with Ask1 signaling (Brace et al., 2022; Herrmann et al., 2022). However, the TIR-1 mechanism does parallel the mechanism by which dSarm regulates axon transport within uninjured bystander neurons, which occurs independently of its NADase activity and with ASK1 MAP kinase signaling (Hsu et al., 2021). These data, along with the findings that dSARM regulates degeneration independently of Ask1 signaling, that TIR-1 can induce degeneration independently of its NADase activity when overexpressed and that TIR-1 can promote degeneration by antagonising DLK-1-PMK-3 MAPK, comprise only a subset of the varied interactions between dSARM/SARM1/TIR-1, their NADase activity and MAPK signaling. The mix and match nature of these interactions indicate that dSarm and TIR-1 have evolved to intricately and specifically regulate a number of diverse processes, including development, cell death, the injury response, neuronal maintenance and innate immunity using variations of common themes. We agree that determining whether TIR-1 regulates cell fate determination and its other developmental functions independently of its NADase activity is also interesting but beyond the scope of this paper.

References

Brace, EJ, et al. Distinct developmental and degenerative functions of SARM1 require NAD+ hydrolase activity. PLoS Genet 18, e1010246, doi:10.1371/journal.pgen.1010246 (2022).

Ding, C, et al. Activation of the CaMKII-Sarm1-ASK1-p38 MAP kinase pathway protects against axon degeneration caused by loss of mitochondria. Elife 11, doi:10.7554/eLife.73557 (2022).

Essuman, K, et al. The SARM1 Toll/Interleukin-1 Receptor Domain Possesses Intrinsic NAD(+) Cleavage Activity that Promotes Pathological Axonal Degeneration. Neuron 93, 1334-1343 e1335, doi:10.1016/j.neuron.2017.02.022 (2017).

Ghosh, AS, et al. DLK induces developmental neuronal degeneration via selective regulation of proapoptotic JNK activity. J Cell Biol 194, 751-764, doi:10.1083/jcb.201103153 (2011).

Herrmann, KA, et al. Divergent signaling requirements of dSARM in injury-induced degeneration and developmental glial phagocytosis. PLoS Genet 18, e1010257, doi:10.1371/journal.pgen.1010257 (2022).

Horsefield, S, et al. NAD+ cleavage activity by animal and plant TIR domains in cell death pathways. Science 365, 793-799, doi:10.1126/science.aax1911 (2019).

Hsu, JM, et al. Injury-Induced Inhibition of Bystander Neurons Requires dSarm and Signaling from Glia. Neuron 109, 473-487 e475, doi:10.1016/j.neuron.2020.11.012 (2021).

Lezi, E, et al. An Antimicrobial Peptide and Its Neuronal Receptor Regulate Dendrite Degeneration in Aging and Infection. Neuron 97, 125-138 e125, doi:10.1016/j.neuron.2017.12.001 (2018).

Loring, HS, et al. A phase transition enhances the catalytic activity of SARM1, an NAD(+) glycohydrolase involved in neurodegeneration. Elife 10, doi:10.7554/eLife.66694 (2021).

Miller, BR, et al. A dual leucine kinase-dependent axon self-destruction program promotes Wallerian degeneration. Nat Neurosci 12, 387-389, doi:10.1038/nn.2290 (2009).

Nichols, ALA, et al. The Apoptotic Engulfment Machinery Regulates Axonal Degeneration in C. elegans Neurons. Cell Rep 14, 1673-1683, doi:10.1016/j.celrep.2016.01.050 (2016).

Peterson, ND, et al. Pathogen infection and cholesterol deficiency activate the C. elegans p38 immune pathway through a TIR-1/SARM1 phase transition. Elife 11, doi:10.7554/eLife.74206 (2022).

Rawson, RL, et al. Axons degenerate in the absence of mitochondria in C. elegans. Curr Biol 24, 760-765, doi:10.1016/j.cub.2014.02.025 (2014).

Summers, DW, et al. SARM1-specific motifs in the TIR domain enable NAD+ loss and regulate injury-induced SARM1 activation. Proc Natl Acad Sci U S A 113, E6271-E6280, doi:10.1073/pnas.1601506113 (2016).

Summers, DW, Milbrandt, J, & DiAntonio, A. Palmitoylation enables MAPK-dependent proteostasis of axon survival factors. Proc Natl Acad Sci U S A 115, E8746-E8754, doi:10.1073/pnas.1806933115 (2018).

Veriepe, J, Fossouo, L, & Parker, JA. Neurodegeneration in C. elegans models of ALS requires TIR-1/Sarm1 immune pathway activation in neurons. Nat Commun 6, 7319, doi:10.1038/ncomms8319 (2015).

Walker, LJ, et al. MAPK signaling promotes axonal degeneration by speeding the turnover of the axonal maintenance factor NMNAT2. Elife 6, doi:10.7554/eLife.22540 (2017).

Xiong, X, & Collins, CA. A conditioning lesion protects axons from degeneration via the Wallenda/DLK MAP kinase signaling cascade. J Neurosci 32, 610-615, doi:10.1523/JNEUROSCI.3586-11.2012 (2012).

Yang, J, et al. Pathological axonal death through a MAPK cascade that triggers a local energy deficit. Cell 160, 161-176, doi:10.1016/j.cell.2014.11.053 (2015).

Yanik, MF, et al. Neurosurgery: functional regeneration after laser axotomy. Nature 432, 822, doi:10.1038/432822a (2004).

https://doi.org/10.7554/eLife.80856.sa2

Article and author information

Author details

  1. Victoria L Czech

    Department of Neurobiology, UMass Chan Massachusetts Medical School, Worcester, United States
    Contribution
    Conceptualization, Formal analysis, Validation, Investigation, Visualization, Writing – original draft, Writing – review and editing
    Contributed equally with
    Lauren C O'Connor
    Competing interests
    No competing interests declared

  2. Lauren C O'Connor

    Department of Neurobiology, UMass Chan Massachusetts Medical School, Worcester, United States
    Contribution
    Conceptualization, Formal analysis, Validation, Investigation, Visualization, Writing – review and editing
    Contributed equally with
    Victoria L Czech
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1966-6932

  3. Brendan Philippon

    Department of Neurobiology, UMass Chan Massachusetts Medical School, Worcester, United States
    Contribution
    Formal analysis, Investigation
    Competing interests
    No competing interests declared

  4. Emily Norman

    Department of Neurobiology, UMass Chan Massachusetts Medical School, Worcester, United States
    Contribution
    Formal analysis, Investigation, Writing – review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-1575-3473

  5. Alexandra B Byrne

    Department of Neurobiology, UMass Chan Massachusetts Medical School, Worcester, United States
    Contribution
    Conceptualization, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing – original draft, Project administration, Writing – review and editing
    For correspondence
    alexandra.byrne@umassmed.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-7449-9188

Funding

National Institute of Neurological Disorders and Stroke (5R01NS110936)

  • Alexandra B Byrne

National Institute of Neurological Disorders and Stroke (1F31NS124338)

  • Lauren C O'Connor

The Dan and Diane Riccio Fund for Neuroscience

  • Alexandra B Byrne

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank A Zeamer for technical contribution, W Joyce and M Liu for construction of reagents, Craig Mello and Krishna Ghanta for help with CRISPR design and injections, K Fitzgerald for the hSARM1 plasmid, P Thompson, H Loring, J Icso, R Pukkila-Worley, N Peterson, as well as the M Francis and M Alkema labs for helpful discussions, along with M Alkema, P Emery, and V Budnik for helpful comments on the manuscript.

Senior and Reviewing Editor

  1. Piali Sengupta, Brandeis University, United States

Publication history

  1. Preprint posted: June 23, 2020 (view preprint)
  2. Received: June 7, 2022
  3. Accepted: January 31, 2023
  4. Version of Record published: April 21, 2023 (version 1)

Copyright

© 2023, Czech, O'Connor et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 634
    Page views
  • 120
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Victoria L Czech
  2. Lauren C O'Connor
  3. Brendan Philippon
  4. Emily Norman
  5. Alexandra B Byrne
(2023)
TIR-1/SARM1 inhibits axon regeneration and promotes axon degeneration
eLife 12:e80856.
https://doi.org/10.7554/eLife.80856

Categories and tags

Research organism

Further reading

    1. Neuroscience

    Cholinergic and noradrenergic axonal activity contains a behavioral-state signal that is coordinated across the dorsal cortex

    Lindsay Collins, John Francis ... David A McCormick
    Research Article Updated

    Fluctuations in brain and behavioral state are supported by broadly projecting neuromodulatory systems. In this study, we use mesoscale two-photon calcium imaging to examine spontaneous activity of cholinergic and noradrenergic axons in awake mice in order to determine the interaction between arousal/movement state transitions and neuromodulatory activity across the dorsal cortex at distances separated by up to 4 mm. We confirm that GCaMP6s activity within axonal projections of both basal forebrain cholinergic and locus coeruleus noradrenergic neurons track arousal, indexed as pupil diameter, and changes in behavioral engagement, as reflected by bouts of whisker movement and/or locomotion. The broad coordination in activity between even distant axonal segments indicates that both of these systems can communicate, in part, through a global signal, especially in relation to changes in behavioral state. In addition to this broadly coordinated activity, we also find evidence that a subpopulation of both cholinergic and noradrenergic axons may exhibit heterogeneity in activity that appears to be independent of our measures of behavioral state. By monitoring the activity of cholinergic interneurons in the cortex, we found that a subpopulation of these cells also exhibit state-dependent (arousal/movement) activity. These results demonstrate that cholinergic and noradrenergic systems provide a prominent and broadly synchronized signal related to behavioral state, and therefore may contribute to state-dependent cortical activity and excitability.

    1. Neuroscience

    Early language exposure affects neural mechanisms of semantic representations

    Xiaosha Wang, Bijun Wang, Yanchao Bi
    Research Article Updated

    One signature of the human brain is its ability to derive knowledge from language inputs, in addition to nonlinguistic sensory channels such as vision and touch. How does human language experience modulate the mechanism by which semantic knowledge is stored in the human brain? We investigated this question using a unique human model with varying amounts and qualities of early language exposure: early deaf adults who were born to hearing parents and had reduced early exposure and delayed acquisition of any natural human language (speech or sign), with early deaf adults who acquired sign language from birth as the control group that matches on nonlinguistic sensory experiences. Neural responses in a semantic judgment task with 90 written words that were familiar to both groups were measured using fMRI. The deaf group with reduced early language exposure, compared with the deaf control group, showed reduced semantic sensitivity, in both multivariate pattern (semantic structure encoding) and univariate (abstractness effect) analyses, in the left dorsal anterior temporal lobe (dATL). These results provide positive, causal evidence that language experience drives the neural semantic representation in the dATL, highlighting the roles of language in forming human neural semantic structures beyond nonverbal sensory experiences.

    1. Neuroscience

    Two conserved vocal central pattern generators broadly tuned for fast and slow rates generate species-specific vocalizations in Xenopus clawed frogs

    Ayako Yamaguchi, Manon Peltier
    Research Article Updated

    Across phyla, males often produce species-specific vocalizations to attract females. Although understanding the neural mechanisms underlying behavior has been challenging in vertebrates, we previously identified two anatomically distinct central pattern generators (CPGs) that drive the fast and slow clicks of male Xenopus laevis, using an ex vivo preparation that produces fictive vocalizations. Here, we extended this approach to four additional species, X. amieti, X. cliivi, X. petersii, and X. tropicalis, by developing ex vivo brain preparation from which fictive vocalizations are elicited in response to a chemical or electrical stimulus. We found that even though the courtship calls are species-specific, the CPGs used to generate clicks are conserved across species. The fast CPGs, which critically rely on reciprocal connections between the parabrachial nucleus and the nucleus ambiguus, are conserved among fast-click species, and slow CPGs are shared among slow-click species. In addition, our results suggest that testosterone plays a role in organizing fast CPGs in fast-click species, but not in slow-click species. Moreover, fast CPGs are not inherited by all species but monopolized by fast-click species. The results suggest that species-specific calls of the genus Xenopus have evolved by utilizing conserved slow and/or fast CPGs inherited by each species.