Summary Summary (text) Abstract Abstract (text) MEDLINE XML PMID List

#### Send to
jQuery(document).ready( function () {
jQuery("#send_to_menu input[type='radio']").click( function () {
var selectedValue = jQuery(this).val().toLowerCase();
var selectedDiv = jQuery("#send_to_menu div." + selectedValue);
if(selectedDiv.is(":hidden")){
jQuery("#send_to_menu div.submenu:visible").slideUp();
selectedDiv.slideDown();
}
});
});
jQuery("#sendto").bind("ncbipopperclose", function(){
jQuery("#send_to_menu div.submenu:visible").css("display","none");
jQuery("#send_to_menu input[type='radio']:checked").attr("checked",false);
});

File Clipboard Collections E-mail Order My Bibliography Citation manager

Format Summary (text) Abstract (text) MEDLINE XML PMID List CSV

- 1 selected item: 30126943
Format Summary Summary (text) Abstract Abstract (text) MEDLINE XML PMID List MeSH and Other Data E-mail Subject Additional text

Generate a file for use with external citation management software.

# Parkin is a disease modifier in the mutant SOD1 mouse model of ALS.

^{1}, Granatiero V

^{1}, Kawamata H

^{1}, Konrad C

^{1}, Kim M

^{1}, Arreguin AJ

^{1}, Zhao D

^{1}, Milner TA

^{1,}

^{2}, Manfredi G

^{3}.

### Author information

- 1
- Feil Family Brain and Mind Research Institute, Weill Cornell Medicine, New York, NY, USA.
- 2
- Harold and Margaret Milliken Hatch Laboratory of Neuroendocrinology, The Rockefeller University, New York, NY, USA.
- 3
- Feil Family Brain and Mind Research Institute, Weill Cornell Medicine, New York, NY, USA gim2004@med.cornell.edu.

### Abstract

Mutant Cu/Zn superoxide dismutase (SOD1) causes mitochondrial alterations that contribute to motor neuron demise in amyotrophic lateral sclerosis (ALS). When mitochondria are damaged, cells activate mitochondria quality control (MQC) mechanisms leading to mitophagy. Here, we show that in the spinal cord of G93A mutant SOD1 transgenic mice (SOD1-G93A mice), the autophagy receptor p62 is recruited to mitochondria and mitophagy is activated. Furthermore, the mitochondrial ubiquitin ligase Parkin and mitochondrial dynamics proteins, such as Miro1, and Mfn2, which are ubiquitinated by Parkin, and the mitochondrial biogenesis regulator PGC1α are depleted. Unexpectedly, Parkin genetic ablation delays disease progression and prolongs survival in SOD1-G93A mice, as it slows down motor neuron loss and muscle denervation and attenuates the depletion of mitochondrial dynamics proteins and PGC1α. Our results indicate that Parkin is a disease modifier in ALS, because chronic Parkin-mediated MQC activation depletes mitochondrial dynamics-related proteins, inhibits mitochondrial biogenesis, and worsens mitochondrial dysfunction.

© 2018 The Authors. Published under the terms of the CC BY 4.0 license.

#### KEYWORDS:

Parkin; SOD1; amyotrophic lateral sclerosis; mitochondria quality control; mitophagy

*n*= 8 mice (four males and four females). At 120 days, **

*P*= 0.0018 (Non Tg vs. SOD1‐G93A) and **

*P*= 0.0011 (SOD1‐G93A vs. wtSOD1) by paired one‐way ANOVA with Tukey's correction. No other statistically significant differences were found (paired Friedman's test with Dunn's correction at 30 days and paired one‐way ANOVA with Tukey's correction at 60 and 90 days).C, DRepresentative Western blot (C) and quantification (D) of Tim23 in homogenates from spinal cords using β‐actin as normalizer. Results are expressed as mean ± SEM and relative to Non Tg at 30 days;

*n*= 6 mice (three males and three females). ***

*P*= 0.0002 (Non Tg vs. SOD1‐G93A at 120 days) by paired one‐way ANOVA with Tukey's correction. No other statistically significant differences were found (paired Friedman's test with Dunn's correction at 30 days and paired one‐way ANOVA with Tukey's correction at 60 and 90 days). Tim23 is decreased in SOD1‐G93A spinal cord relative to Non Tg at 120 days.E, FWestern blot (E) and quantification (F) of COXI in homogenates from spinal cords. Protein levels are normalized by β‐actin. Results are expressed as mean ± SEM and relative to Non Tg at 30 days;

*n*= 6 mice (three males and three females). *

*P*= 0.017 (Non Tg vs. SOD1‐G93A at 120 days) by paired Friedman's test with Dunn's correction. No other statistically significant differences were found (paired Friedman's test with Dunn's correction at 30 and 60 days and paired one‐way ANOVA with Tukey's correction at 90 days). COX1 is decreased in SOD1‐G93A spinal cord relative to Non Tg at 120 days.GqPCR (fold change) of PGC1α mRNA normalized by β‐actin mRNA. Results are expressed as mean ± SEM and fold change of Non Tg at 30 days;

*n*= 6 (three males and three females) for 30 and 120 days,

*n*= 4 (two males and two females) for 60 days and

*n*= 5 (three males and two females) for 90 days. *

*P*= 0.035 at 90 days and ***

*P*= 0.0007 at 120 days. Paired Student's

*t*‐test (for 60, 90, and 120 days) and paired Wilcoxon's test (for 30 days) were used for comparisons.

Source data are available online for this figure.

*n*= 3, one male and two females), 90 days (

*n*= 4, two males and two females), and 120 days (

*n*= 4, two males and two females). Number of neurons imaged: 60 days,

*n*= 77 Non Tg and 87 SOD1‐G93A; 90 days,

*n*= 118 Non Tg and 173 SOD1‐G93A; and 120 days,

*n*= 110 Non Tg and 130 SOD1‐G93A. Comparisons within pairs showed that mitophagy is increased in SOD1‐G93A spinal cords at 60 and 90 days compared to Non Tg. Results are expressed as mean ± SEM. For 60 days, **

*P*= 0.0022; for 90 days, ***

*P*= 0.0001; and for 120 days,

*P*= 0.094, all by unpaired Mann–Whitney test. Comparisons along time showed that mitophagy increases over time in both genotypes. For Non Tg, by unpaired Kruskal–Wallis's test with Dunn's correction, **

*P*= 0.0046 and ***

*P*= 0.0001. For SOD1‐G93A, by unpaired Kruskal–Wallis's test with Dunn's correction, **

*P*= 0.0077 and ***

*P*= 0.0001.

Source data are available online for this figure.

*n*= 8 (four males and four females). Paired Wilcoxon's test at 90 days, **

*P*= 0.0078; and paired Student's

*t*‐test at 120 days, **

*P*= 0.0054. Paired Wilcoxon's test for 30 and 60 days showed no statistically significant differences. Parkin levels are decreased in SOD1‐G93A at 90 and 120 days compared to Non Tg.CqPCR (fold change) of Parkin mRNA normalized by β‐actin mRNA. Results are expressed as mean ± SEM and fold change relative to Non Tg at 30 days of age;

*n*= 6 (three males and three females) for 30 and 120 days,

*n*= 4 (two males and two females) for 60 days, and

*n*= 5 (three males and two females) for 90 days. No statistically significant differences were found by paired Student's

*t*‐test (90 days) and Wilcoxon's test (30, 60, and 120 days).D, ERepresentative Western blot (D) and quantification (E) of Parkin levels in spinal cord mitochondria at 120 days. Protein levels are normalized by Complex V. Parkin is reduced in spinal cord mitochondria of SOD1‐G93A mice at 120 days. Results are expressed as mean ± SEM and shown as percent of Non Tg;

*n*= 8 (four males and four females); **

*P*= 0.0078, by Wilcoxon's

*t*‐test.

Source data are available online for this figure.

*n*= 8 (four males and four females); **

*P*= 0.0078, by Wilcoxon's

*t*‐test.CQuantification of SOD1 at 130 days indicates that Parkin knockout does not affect the levels of transgenic mutant human SOD1 expressed in SOD1‐G93A spinal cord. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females); no statistically significant differences relative to SOD1 expression were found between G93A and PKO/G93A,

*P*= 0.214, by paired Student's

*t*‐test. Differences between Non Tg and G93A (by paired Friedman's test with Dunn's correction), *

*P*= 0.020 and PKO and PKO/G93A (by paired one‐way ANOVA with Tukey's correction), ***

*P*= 0.0001.D, ESOD1 filter‐trap assay (D) of spinal cord homogenates at 130 days and quantification of dot intensities (E), showing that PKO/G93A mice have less SOD1 aggregates than G93A. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females); **

*P*= 0.0078 by paired Wilcoxon's test.FRepresentative images of 130‐day‐old lumbar spinal cord sections immunostained for p62. Scale bar, 200 μm.GQuantification of the average number of large p62‐positive inclusions per section at 130 days indicates that PKO/G93A mice have less SOD1 aggregates than G93A. Results are expressed as mean ± SEM;

*n*= 7 (three males and four females) mice for Non Tg and PKO,

*n*= 10 (five males and five females) mice for G93A and

*n*= 9 (four males and five females) mice for PKO/G93A; *

*P*= 0.010 by unpaired Mann–Whitney test.

Source data are available online for this figure.

*n*= 8 (four males and four females) mice per group. *

*P*= 0.039 by paired Wilcoxon's test.Quantification of p62 in the homogenates indicates that p62 is increased in PKO and PKO/G93A mice, independent of SOD1‐G93A expression, at 130 days of age. β‐actin was used for normalization. Results are expressed as mean ± SEM and percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between Non Tg and G93A by paired Friedman's test with Dunn's correction (

*P*= 0.99). ***

*P*= 0.0001 by paired Student's

*t*‐test (G93A vs. PKO/G93A).Representative Western blot of Parkin and p62 in liver homogenates at 130 days.Parkin protein levels were quantified at 130 days. β‐actin was used for normalization. Results are expressed as mean ± SEM and percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between Non Tg and G93A (

*P*= 0.546 by paired Wilcoxon's test).Quantification of p62 protein levels in liver at 130 days. Results are expressed as mean ± SEM and percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between Non Tg and G93A (

*P*= 0.546 by paired Wilcoxon's test). No statistically significant differences were found among the other groups.

Source data are available online for this figure.

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.987 by paired Student's

*t*‐test). **

*P*= 0.0058 by paired Friedman's test with Dunn's correction (for Non Tg vs. G93A), ***

*P*= 0.0002, by paired one‐way ANOVA with Tukey's correction (for PKO vs. PKO/G93A).

Source data are available online for this figure.

*n*= 4 (two males and two females) mice for Non Tg and PKO and

*n*= 6 (three males and three females) for G93A and PKO/G93A. Number of images recorded for each genotype:

*n*= 9 for Non Tg and PKO,

*n*= 16 for G93A, and

*n*= 12 for PKO/G93A. Results are expressed as mean ± SEM; *

*P*= 0.032, by unpaired one‐tailed Student's

*t*‐test.Representative Western blots of OPTN, p62, and VCP in spinal cord mitochondrial fractions at 130 days. Protein levels are normalized by Complex V. The quantifications in panels (D–F) show that mitochondria of PKO/G93A mice have less mitophagy adaptor proteins than G93A relative to Complex V.Quantification of OPTN relative to Complex V at 130 days. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females); *

*P*= 0.039 by paired Student's

*t*‐test.Quantification of p62, relative to Complex V at 130 days. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females); **

*P*= 0.0029 by paired Student's

*t*‐test.VCP was quantified using Complex V as loading reference at 130 days. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females); *

*P*= 0.044 by paired Student's

*t*‐test.

Source data are available online for this figure.

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.078 by paired Wilcoxon's test); ***

*P*= 0.0007 (Non Tg vs. G93A) and *

*P*= 0.037 (PKO and PKO/G93A) both by paired Friedman's test with Dunn's correction.DQuantification of OPTN accumulation in end‐stage mitochondria, with Complex V as protein loading control. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females) mice per group. **

*P*= 0.0078 (for G93A and PKO/G93A) by paired Wilcoxon's test; ***

*P*= 0.0003 (Non Tg vs. G93A) by paired Friedman's test with Dunn's correction. No other statistically significant differences were found.EQuantification of VCP in mitochondria at disease end stage. Complex V was used as normalizer. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females) mice per group. *

*P*= 0.039 (for G93A and PKO/G93A) by paired Wilcoxon's test; ***

*P*= 0.0007 (Non Tg vs. G93A) and *

*P*= 0.037 (PKO and PKO/G93A) both by paired Friedman's test with Dunn's correction.

Source data are available online for this figure.

*n*= 19 for Non Tg (nine males and 10 females),

*n*= 20 for PKO (10 males and 10 females),

*n*= 18 for G93A (eight males and 10 females), and

*n*= 19 for PKO/G93A (nine males and 10 females);

*P*= 0.0000285.Body weight changes over time, relative to body weight at 45 days that is set at 100% for each mouse. Parkin knockout delays body weight loss in SOD1‐G93A mice. Results are expressed as mean ± SEM;

*n*= 19 for Non Tg (nine males and 10 females),

*n*= 20 for PKO (10 males and 10 females),

*n*= 18 for G93A (eight males and 10 females), and

*n*= 19 for PKO/G93A (nine males and 10 females); *

*P*= 0.017 (at 150 days), *

*P*= 0.004 (at 164 days), *

*P*= 0.002 (at 171 days), and **

*P*= 0.001 (at 178 days), by two‐way ANOVA with repeated measures and with Holm–Sidak's correction for multiple comparisons.Clasping score changes over time. Parkin knockout does not affect clasping onset in SOD1‐G93A mice. An intermediate score of 2.5 is indicated by the dotted line. Results are expressed as mean ± SEM;

*n*= 19 for Non Tg (nine males and 10 females),

*n*= 20 for PKO (10 males and 10 females),

*n*= 18 for G93A (eight males and 10 females), and

*n*= 19 for PKO/G93A (nine males and 10 females).Representative images of NMJs in the soleus muscle at 130 days. Muscles were immunostained for neurofilament H (NF‐H, in red) and stained with α‐bungarotoxin (α‐BTX, in green). Scale bar, 20 μm.Quantification of the percentage of innervated NMJs at 130 days indicates that Parkin knockout decreases denervation in SOD1‐G93A mice. Results are expressed as mean ± SEM and were collected from

*n*= 6 mice per group (three males and three females);

*n*of images evaluated = 49 for Non Tg (with 483 NMJs counted), 46 for PKO (396 NMJs counted), 47 for G93A (413 NMJs assessed), and 38 for PKO/G93A (325 NMJs counted); ***

*P*= 0.0001, by unpaired Mann–Whitney test.Representative images of lumbar spinal cord sections at 130 days immunostained with ChAT. Scale bar, 200 μm.Quantification of ChAT‐positive MN in the spinal cord ventral horn (dotted areas) at 130 days shows that Parkin knockout decreases MN loss in the spinal cord of SOD1‐G93A mice. Results are expressed as mean ± SEM;

*n*= 10 mice per group in Non Tg, PKO, and G93A (five males and five females) and

*n*= 9 for PKO/G93A (four males and five females); *

*P*= 0.031, by unpaired Student's

*t*‐test.Representative images of lumbar spinal cord sections immunostained for GFAP at 130 days. Scale bar, 200 μm.Quantification of the average GFAP staining intensity at 130 days indicates that Parkin knockout decreases spinal cord astrogliosis in SOD1‐G93A mice. Results are expressed as mean ± SEM;

*n*= 10 mice per group (five males and five females); *

*P*= 0.033, by paired Student's

*t*‐test.

Source data are available online for this figure.

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.493 by paired Student's

*t*‐test).DQuantification of Tim23 at 130 days, using β‐actin as loading control, showed decreased levels of Tim23 in G93A mice. Results are expressed as mean ± SEM and as a percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.921 by paired Student's

*t*‐test). *

*P*= 0.035 (Non Tg Vs. G93A) by paired Friedman's test with Dunn's correction. No other statistically significant differences were found.E, FRepresentative Western blots of COXI (E) and Tim23 (F) in disease end‐stage homogenates.GQuantification of COXI at disease end stage using β‐actin as loading control. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 5 (three males and two females) mice per group. *

*P*= 0.047 for (G93A and PKO/G93A) by paired Student's

*t*‐test; **

*P*= 0.0018 by paired Friedman's test with Dunn's correction (Non Tg vs. G93A).HQuantification of Tim23 with β‐actin as normalizer at end stage showed that Parkin knockout can alleviate the turnover of mitochondrial proteins in G93A spinal cords. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.190 by paired Student's

*t*‐test). *

*P*= 0.037 (for PKO vs. PKO/G93A) by paired Friedman's test with Dunn's correction. No other statistically significant differences were found.

Source data are available online for this figure.

*n*= 8 (four males and four females) mice per group; *

*P*= 0.024, by paired Student's

*t*‐test. Activities are expressed as nmoles of substrate oxidized per minute per mg of protein.BCytochrome oxidase (COX) activity at end stage. Results are expressed as mean ± SEM;

*n*= 8 (four males and four females) mice per group; *

*P*= 0.032, by paired Student's

*t*‐test. Activities are expressed as nmoles of substrate oxidized per minute per mg of protein.CqPCR (fold change) of PGC1α mRNA normalized by β‐actin mRNA at 130 days of age. Results are expressed as mean ± SEM and as fold change of the Non Tg;

*n*= 8 (four males and four females). No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.250 by paired Wilcoxon's test); *

*P*= 0.035 paired Friedman's test with Dunn's correction (Non Tg vs. G93A). No other statistically significant differences were found.DqPCR (fold change) of PGC1α mRNA normalized by β‐actin mRNA at end stage. Results are expressed as mean ± SEM and as fold change of the Non Tg;

*n*= 8 (four males and four females). No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.613 by paired Student's

*t*‐test). **

*P*= 0.0058 by paired Friedman's test with Dunn's correction (for Non Tg vs. G93A comparison) and ***

*P*< 0.0001 by paired one‐way ANOVA with Tukey's correction (for PKO vs. PKO/G93A comparison). Parkin knockout improves PGC1α expression at 130 days (C) but the effect is lost at end stage (D).E, FRepresentative Western blots (E) and quantification (F) of PARIS at disease end stage. Protein levels are normalized by β‐actin, and results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females). No statistically significant differences were found between G93A and PKO/G93A by paired Wilcoxon's test (

*P*= 0.99). PARIS levels are unchanged in SOD1‐G93A mice relative to Non Tg at disease end stage.

Source data are available online for this figure.

*n*= 8 (four males and four females) mice per group; *

*P*= 0.011 by paired Student's

*t*‐test (for comparison G93A vs. PKO/G93A) and **

*P*= 0.003 by paired Friedman's test with Dunn's correction (for Non Tg vs. G93A). Parkin knockout increases the levels of Miro1 in SOD1‐G93A mice at disease end stage.C, DWestern blots (C) and quantification (D) of Mfn2 in spinal cord homogenates at disease end stage. Protein levels are normalized by β‐actin. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.078 by paired Wilcoxon's test); ***

*P*= 0.0007 by paired Friedman's test with Dunn's correction (for Non Tg vs. G93A) and *

*P*= 0.037 by paired Friedman's test with Dunn's correction (for PKO vs. PKO/G93A). Parkin knockout increases the levels of Mfn2 in SOD1‐G93A mice at disease end stage.EWestern blots of spinal cord mitochondria at disease end stage probed for lysine 48 (left panel) and lysine 63 (right panel) ubiquitin chains. Citrate synthase is used as loading control. Asterisks indicate ubiquitinated proteins with different abundance in SOD1‐G93A samples.F, GRepresentative Western blot (F) and quantification (G) of LC3 II in spinal cord mitochondria at end stage. Protein levels are normalized by Complex V. LC3 II is increased in both G93A and PKO/G93A mitochondria and Parkin knockout does not affect LC3 II levels. Results are expressed as mean ± SEM and as percent of Non Tg;

*n*= 8 (four males and four females) mice per group; No statistically significant differences were found between G93A and PKO/G93A (

*P*= 0.504 by paired Student's

*t*‐test). *

*P*= 0.035 by paired Friedman's test with Dunn's correction (for Non Tg vs. G93A) and **

*P*= 0.0014 by paired one‐way ANOVA with Tukey's correction (for PKO vs. PKO/G93A).HRepresentative Western blot of March5 and Mul1 in spinal cord mitochondria at end stage.IQuantification of March5 at disease end stage. Results are expressed as mean ± SEM and percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A by paired Wilcoxon's test (

*P*= 0.742). Protein levels are normalized by Complex V. March5 levels are unaffected by Parkin knockout.JQuantification of Mul1 at disease end stage. Results are expressed as mean ± SEM and percent of Non Tg;

*n*= 8 (four males and four females) mice per group. No statistically significant differences were found between G93A and PKO/G93A by paired Wilcoxon's test (

*P*= 0.94). Protein levels are normalized by Complex V. Mul1 levels are unaffected by Parkin knockout.

Source data are available online for this figure.

*n*= 40 fields per condition (from two independent experiments); for Mock NSC34 cells, GFP vs. Parkin transfected, treated with mtPQ, *

*P*= 0.013, by unpaired Student's

*t*‐test; for G93A NSC34 cells, GFP vs. Parkin transfected, treated with mtPQ, *

*P*= 0.029, by unpaired Mann–Whitney

*t*‐test.

Source data are available online for this figure.