Diapause formation and downregulation of insulin-like signaling via DAF-16/FOXO delays axonal degeneration and neuronal loss

Abstract

Axonal degeneration is a key event in the pathogenesis of neurodegenerative conditions. We show here that mec-4d triggered axonal degeneration of Caenorhabditis elegans neurons and mammalian axons share mechanistical similarities, as both are rescued by inhibition of calcium increase, mitochondrial dysfunction, and NMNAT overexpression. We then explore whether reactive oxygen species (ROS) participate in axonal degeneration and neuronal demise. C. elegans dauers have enhanced anti-ROS systems, and dauer mec-4d worms are completely protected from axonal degeneration and neuronal loss. Mechanistically, downregulation of the Insulin/IGF-1-like signaling (IIS) pathway protects neurons from degenerating in a DAF-16/FOXO-dependent manner and is related to superoxide dismutase and catalase-increased expression. Caloric restriction and systemic antioxidant treatment, which decrease oxidative damage, protect C. elegans axons from mec-4d-mediated degeneration and delay Wallerian degeneration in mice. In summary, we show that the IIS pathway is essential in maintaining neuronal homeostasis under pro-degenerative stimuli and identify ROS as a key intermediate of neuronal degeneration in vivo. Since axonal degeneration represents an early pathological event in neurodegeneration, our work identifies potential targets for therapeutic intervention in several conditions characterized by axonal loss and functional impairment.

Figure 1. Functional and morphological time course of mec-4d mediated degeneration of C. elegans AVM neuron.

(A) Schematic representation of the six mechanosensory neurons of an adult C. elegans with an expanded representation of the AVM neuron (anterior part of the worm is to the left) and below a confocal image of the AVM expressing gfp under a touch neuron specific promoter (P_mec-17mec-17::gfp) in a wild type (N2) background merged with a Nomarski optic micrograph. Only the anterior portion is shown. Scale bar, 25 µm. (B) Morphological categories of the AVM axon in a wild type condition and during the progression of degeneration triggered by the expression of the MEC-4d degenerin channel. Schematic representations of the different axonal morphological categories are shown in the left panel, the right panel shows neurons as seen by confocal microscopy. GFP-expressing neurons have been pseudo-colored to match the morphological categories. Scale bar, 25 µm. (C–F) Temporal analysis of morphological categories from the time of hatching for AVM somas and axons in wild-type (C–D), and mec-4d mutants (E–F). (C) AVM somas in the wild type strain (P_mec-17mec-17::gfp) are almost fully present at 12 hours. (D) AVM axons are still growing at 12 hours but at 24 hours they have reached their full extension. (E) AVM somas in mec-4d mutants appear by 12 hours and therein steadily degenerate, going through a vacuolated stage. (F) AVM axons in mec-4d mutants appear and extend along the anterior region at 12 hours and thereafter they degenerate through beaded and then truncated intermediate stages (mean value for each category is shown; error bars [<0.1%] are not included; N = 3 of 30 worms each per group). (G) Anterior touch response (black bars) and percentage of morphologically wild-type axons (green bars) of mec-4d worms at 12, 24 and 48 hours post hatching. For the 12-hour time point of this graph, only AVM neurons with fully-grown axons were considered as wild-type (mean values are shown; error bars indicates SEM; N = 3 of 30 worms each per group). SoW: soma wild type; SoV: Soma vacuolated; So∅: soma absent; AxW: Axon wild type; AxB: Axon beaded; AxT: Axon truncated; Ax∅: Axon absent.

Figure 2. C. elegans dauer state prevents neuronal degeneration triggered by mec-4d.

mec-4d dauer worms were scored by morphological and functional readouts. Both AVM somas (A) and axons (B) were protected from degeneration when worms were kept in a dauer state for one week or one month (mean value for each category is shown; error bars [<0.1%] are not included; N = 3 of 30 worms each per group). (C) Morphologically conserved AVM neuron from a mec-4d mutant kept in diapause for 1 month. Scale bar, 10 µm. (D) Anterior touch response of mec-4d mutants kept in diapause for one month. Both wild-type (N2) and mec-4d groups have wild-type somas and axons as studied after the touch test by fluorescent microscopy. An equal fraction in both groups is unresponsive, which represents a dauer-dependent trait (mean values are shown, error bars indicates SEM, N = 3 of 20 worms each per group). (E–H) Degeneration of somas and axons during dauer exit by food reposition takes place in a rate that is independent to the dauer length. Line graphs in F and H (right side), represent the rate of decay of wild type somas (F) and axons (H) during dauer exit (mean value for each category is shown; error bars [<0.1%] are not included; N = 3 of 30 worms each per group; Student's t test at each time point was performed for SoW and AxW categories, no significant differences were found between groups). (I) In uIs58 wild-type non-dauers and dauers expressing MEC-4::GFP (P_mec-4mec-4::gfp), the GFP signal was expressed in a punctate pattern along axons, as in animals with the mec-4d (e1611) mutation. Scale bar, 10 µm. Quantification of MEC-4::GFP signal intensity (J) and interpuncta distance (K) reveal no significant differences between conditions (N = 9 worms per group; analyzed by Student's t test compared to wild-type non-dauer values).

Figure 3. Diapause entry prevents degeneration of the PVC neurons triggered by the deg-1(u38) mutation.

deg-1(u38) dauer worms were scored by morphological and functional readouts. Both PVC somas (A) and axons (B) were protected from degeneration when worms were kept in diapause for one week (mean value is shown; error bars indicate SEM; N = 3 of 10 worms each per group; *p<0.01 by Student's t test compared with 24 hours after hatching). (C) Morphology of the PVC neuron observed by expression of nmr-1::gfp on wild type (top), deg-1(u38) mutants at 36 hours (middle) and deg-1(u38) dauers (bottom). Scale bar, 10 µm. (D) Posterior touch responses of deg-1(u38) mutants after hatching, kept as dauers for one month and during dauer exit by food reposition. Only worms that responded to the anterior touch were included (mean values are shown, error bars indicate SEM, N = 3 of 33 worms each per group).

Figure 4. Downregulation of Insulin/IGF-1-like signaling (IIS) prevents neuronal degeneration triggered by mec-4d.

(A) DAF-2 negatively regulates both DAF-16/FOXO and SKN-1/Nrf2, which by different mechanisms increases the cellular antioxidative capacity. (B–C) The possible protective role of the IIS pathway activation in mec-4d degeneration was explored in daf-2(e1370ts) mutants. At the restrictive temperature both somas (B) and axons (C) were significantly protected respect to the controls not carrying the mec-4d(e1611) mutation or daf-2(e1370ts) mutants carrying mec-4d(e1611) and raised at 20°C. Below each bar graph, a line graph incorporating only the SoW and AxW of daf-2;mec-4d vs. mec-4d raised at 25 °C is shown. Notice that mec-4d-dependent degeneration of both somas and axons is delayed for about one day at 25°C compared to 20°C (mean value for each category is shown; error bars [<0.1%] are not included; N = 3 of 30 worms each per group; ^#p<0.05 and *p<0.01 by Student's t test compared with control at the same temperature for the SoW and AxW category). (D–E) TRN-autonomous daf-2(RNAi) was performed in the P_mec-17mec-17::gfp strain carrying the mec-4d(e1611) mutation (WCH6). Significant morphological protection of both AVM somas (D) and axons (E) is seen at 72 hours post hatching (mean value for each category is shown; error bars [<0.1%] are not included; N = 3 of 30 worms each per group; *p<0.001 by Student's t test compared with control at the same temperature for the SoW and AxW category).

Figure 5. DAF-16 contributes to inhibition of mec-4d–dependent AVM degeneration associated with reduced IIS signaling.

(A–B) Double RNAi experiments in the TRN-autonomous RNAi strain carrying the mec-4d(e1611) mutation (WCH6) were performed to assess the role of DAF-16 and SKN-1 downstream DAF-2 in the protection of AVM neurons. In both somas (A) and axons (B), single daf-2(RNAi), but not skn-1 or daf-16, protects the AVM neuron from degeneration. When combined, only daf-16(RNAi) is able to revert the protective phenotype given by daf-2(RNAi) (mean value for each category is shown; error bars [<0.1%] are not included; N = 3 of 30 worms each per group; *p<0.05 by Student's t test compared to single daf-2(RNAi) for the SoW and AxW category). (C–D) Combinatorial RNAi experiments were performed to address the role of endogenous antioxidant enzymes. daf-2(RNAi) together with either sod-2, sod-4, ctl-1 or ctl-2 were fed to the TRN-autonomous RNAi strain carrying the mec-4d(e1611) mutation (WCH6). All double RNAi significantly decreased the daf-2(RNAi) protection of degeneration of somas and axons of the AVM neuron. (E) ROS was visualized using the fluorogenic dye CellROX (red) in mec-4d embryos. A strong signal of the dye co-localized with the endogeneous GFP expressed by degenerating TNRs. Scale bar, 5 µm. (F–G) mec-4d worms were treated from the time of hatching with the ROS scavenger trolox (100 µm) and ascorbic acid (AA, 50 mM); after 72 hours both somas (F) and axons (G) were protected several fold compared to control nematodes (mean value for each category is shown; error bars [<1%] not included; N = 3 of 30 worms each per group; *p<0.005 by Student's t test compared with control for the SoW and AxW category).

Figure 6. Dietary restriction inhibits mec-4d–dependent neuronal degeneration in C. elegans.

(A–B) The influence of dietary restriction (DR) in TRN degeneration was investigated by fasting mec-4d worms intermittently with a protocol consisting of 3 hours of ad libitum food regime followed by 9 hours of starvation. After 72 hours, the worms on DR show a significant protection of both AVM somas (A) and axons (B) compared to the 72 hours ad libitum fed worms (mean value for each category is shown; error bars [<1%] not included; N = 3 of 30 worms each per group; *p<0.005 by Student's t test compared with control for the SoW and AxW category). (C) AVM neuron with a morphologically conserved soma and axon from a mec-4d worm subjected to DR for 72 hours by means of intermittent fasting. Scale bar, 10 µm.

Figure 7. C. elegans and mammalian neurons share degenerative mechanisms triggered by diverse pro-degenerative stimuli.

The mechanisms associated to somatic and axonal degeneration of AVM-mec-4d neurons were studied by pharmacological and genetic means. Worms were grown for 72 hours after hatching in food supplemented with EGTA (50 mM) or cyclosporin A (CsA, 50 µM). Both drugs delay degeneration of AVM somas (A) and axons (B) several fold (mean value for each category is shown; error bars [<1%] not included; N = 3 of 30 worms each per group; *p<0.01 by Student's t test compared with control for the SoW and AxW category). (C) Fluorescent micrograph of a beaded axon (AxB) and a vacuolated soma (SoV) of a mec-4d-expressing AVM, contrasted with an AVM neuron treated with EGTA displaying a wild-type axon (AxW) and soma (SoW). Scale bar, 10 µm. (D) Touch neuron-specific expression of nmat-2::gfp (P_mec-18nmat-2::gfp), which is restricted to the cytosol in all TRNs. Scale bar, 10 µm. Overexpression of NMAT-2::GFP in TRN protects to a large extent somas (E) and axons (F) from mec-4d dependent degeneration at 72 hours post hatching (mean value for each category is shown; error bars [<0.1%] are not included; N = 3 of 30 worms each per group; *p<0.01 by Student's t test compared with control for the SoW and AxW category). (G) Anterior touch response of worms at 72 hours post hatching. The almost complete functional impairment triggered by mec-4d is significantly rescued by NMAT-2::GFP overexpression (mean values are shown, error bars indicates SEM, N = 3 of 30 worms each per group; p<0.05 by Student's t test for touch sensitivity values).

Figure 8. Systemic antioxidant treatment and intermittent fasting delay axonal degeneration in mice.

Wild-type mice (C57BL/6J strain) were subjected to systemic treatments of the antioxidant ascorbic acid (AOX) or to dietary restriction (DR) for 10 days. After this, a sciatic nerve injury was performed unilaterally at the sciatic notch level. The degree of axonal degeneration was quantitatively assessed by immunofluorescence using neurofilament antibodies (A–D) and also studied by ultrastructural examination (E–J). (A–C) Transverse sections of sciatic nerves stained for NFM (green) and NFH (red) isoforms in control (upper panels) and 3, 5 and 7 days after injury, distal to the injury region (lower panels). A nerve crush was performed for 3 days experiments and a nerve cut for experiments lasting for 5 and 7 days, to avoid the interference of axonal regeneration in the analysis. The neurofilament signal decrease progressively at 3, 5 and 7 days after nerve injury in non-treated mice (A). Both AOX or DR treatments (B and C, respectively) strongly delays axonal degeneration triggered by nerve injury when observed 3 and 5 days after damage and AOX treatment strongly delays degeneration even at 7 days post injury. Scale bar, 20 µm. (D) Quantification of NFH-positive axons in nerve cross sections as shown in (A–C), expressed as axonal density. In injured nerves, statistically significant protection from axonal degeneration was seen after AOX and DR treatment compared to non-treated mice in all days post-damage (mean values are shown, error bars indicate SEM, N = 4 per group; #p<0.001 by Student's t test compared with non-injured controls; *p<0.005 by Student's t test compared with the respective day post-injury in non-treated mice). Representative electron micrographs (EM) of sciatic nerves from non-treated (E), AOX-treated (F) and DR-treated (G) wild-type mice. Micrographs of non-injured nerves are shown in the upper panels and from nerves 3 days after sciatic nerve crush are shown in the lower panels. In non-injured nerves from all conditions, axons are rounded, myelin sheaths are well preserved and the tissue is well organized. After nerve crush, most axons are degenerated and the myelin sheaths appear collapsed (E, bottom panel). In contrast, crushed nerves from mice systemically treated with AOX or with a restricted dietary intake by intermittent fasting, display preserved axons with conserved myelin sheaths (F and G, lower panels). Scale bar, 5 µm. (H) Diameters of axonal mitochondria were measured in EM transverse sections. The almost three-fold increase in mitochondrial diameter in crushed and non-treated nerves was prevented by both AOX and DR treatment (mean values are shown, error bars indicate SEM, N = 75–100 mitochondria/nerve of 3 mice per condition; #p<0.0001 by Student's t test compared with non-injured controls; *p<0.0001 by Student's t test compared with 3 days after crush in non-treated mice). (I) ROS-dependent lipoperoxidation measured in untreated sciatic nerve explants at 0, 12, 24, 48 and 72 hours post-injury and in nerves treated with the ROS scavenger trolox (1 mM) for 72 hours. Lipoperoxide levels progressively increase after nerve damage and AOX treatment strongly inhibit injury induced lipid peroxidation (mean values are shown, error bars indicate SEM, N = 4 per group; #p<0.05 and *p<0.01 by Student's t test compared with time 0; the AOX values are compared statistically with 72 hours non-treated nerves). (J) In situ ROS detection in YFP-expressing axons (green) using the vital dye CellROX (red). Uninjured axons have ROS levels not detectable by this technique, which increases considerable at 24 hours post damage. Scale bar, 5 µm.

Figure 9. Model of ROS dependent neuronal degeneration.

Opening of the MEC-4d channel results in an increase in intracellular calcium, which together with calcium contribution from other sources leads to mitochondrial dysfunction, ATP depletion and ROS generation, which enters into positive feedback loops, finally leading to activation of proteases in somas and axons. In dauer and daf-2 mutants, anti-ROS systems are elevated, decreasing mitochondrial dysfunction and blocking neuronal degeneration in dauers.