Parkin, which tags other proteins with ubiquitin to mark them for destruction, has been linked to sporadic and some familial forms of Parkinson's disease (PD). In the August 25 Nature Neuroscience, scientists led by Ted and Valina Dawson, Johns Hopkins University School of Medicine, Baltimore, Maryland, report that parkin deficiency leads to a toxic buildup of a substrate that goes by the mouthful aminoacyl-tRNA synthetase complex interacting multifunctional protein-2. AIMP2 sets in motion a cell death cascade that specifically targets dopamine cells in transgenic mice, the authors found. Since AIMP2 and some of its downstream partners also appear at higher levels in postmortem tissue from patients with Parkinson's, this pathway could serve as a therapeutic target for the disease, the authors suggested.

AIMP2 stabilizes the aminoacyl-tRNA synthetase complex in the cytosol. It would ordinarily be ubiquitinated by parkin and then degraded in the proteasome. In PD, it instead turns up in Lewy bodies, protein inclusions that mark this disease (see Corti et al., 2003). The Dawson group previously reported that AIMP2 accumulates in animal models of PD and in post-mortem tissue from patients (see Ko et al., 2005). In the current study, they wanted to determine if too much AIMP2 would be toxic to dopamine cells, potentially explaining why parkin deficiency is so deadly.

First author Yunjong Lee and colleagues overexpressed human AIMP2 in mice and found that as the animals aged, they lost dopaminergic neurons and became physically uncoordinated. In 20-month-old mice, overexpression of AIMP2 killed more dopamine neurons in the substantia nigra zona compacta than in the ventral tegmental area, even though expression levels were similar, mirroring the pattern of cell death in PD.

To work out the mechanism of AIMP2 cell death, the researchers overexpressed the protein in SH-SY5Y neuroblastoma cells. According to subcellular fractionation experiments, a small portion of AIMP2 translocated from the cytosol to the nucleus, which happens when a cell is under stress. There, it directly activated poly(ADP-ribose) polymerase-1 (PARP1), a protein that senses damage to DNA and forms a poly(ADP-ribose) (PAR) polymer. PARP1 and PAR activate a type of cell death pathway called parthanatos (see diagram). In this pathway, the PAR polymer makes its way from the nucleus to mitochondria, where it binds apoptosis-inducing factor (AIF), a well-known instigator of that death pathway (see Wang et al., 2011).

To test whether parthanatos accounts for AIMP2-induced dopaminergic loss in vivo, the researchers knocked out or inhibited PARP1 in the AIMP2-overexpressing mice. Doing so protected the neurons, suggesting that AIMP2 does at least some of its damage through the cell death pathway. The authors acknowledged that it might wreak havoc through other mechanisms, as well.

Overexpression experiments are known to have led to artifacts in the past. To get a sense whether these findings might be relevant to human disease, the researchers compared levels of these proteins in diseased and control postmortem brains. AIMP2 was three-fold more abundant in dopaminergic cells from PD patients, while PAR was up 10-fold in the substantia nigra but not the cortex when compared to healthy controls. These results suggest that PARP1 inhibition could be a therapeutic target for PD, wrote the authors.

"The idea that a toxic substrate of Parkin, AIMP2, mediates activation of PARP-1 is novel and intriguing for genetic PD and probably sporadic PD," Imam, University of Texas Health Science Center, San Antonio, wrote to Alzforum in an email. "However, this pathway might be just part of a complex picture of progressive damage, and [inhibitors] might bring a minimal change." He noted that deactivating PARP1 could compromise DNA repair, and possibly increase the risk for cancer. Ted Dawson countered that other DNA repair mechanisms would likely compensate for a lack of PARP1, noting that PARP1 knockout mice live a normal lifespan. PARP1 inhibitors are currently being developed as breast, ovarian, and prostate cancer treatments, and some cross the blood brain barrier (for a review, see Rouleau et al., 2010).

Why does AIMP2 specifically kill dopamine neurons? Because these cells are particularly sensitive to oxidative stress (see ARF related news story), they are more likely to redirect AIMP2 from the cytosol to the nucleus to activate PARP1 and kick off parthanatos, Ted Dawson told Alzforum. Whether this happens in human dopaminergic cells is unclear.

Parkin and the mitochondrial kinase PINK1, whose gene is also linked to parkinsonism, are both known to regulate mitochondrial quality control through the process of mitophagy (see ARF related news story). Which is more important when parkin function is lost to the cell—mitochondrial abnormalities or parthanatos? Dawson could not say, but he pointed out that much of the mitochondrial data comes from flies and tumor cell lines, and research yet needs to show that mitochondrial abnormalities occur in the dopaminergic neurons of mice and humans. Dawson speculated that if they do, then parthanatos could act downstream of mitochondrial problems. Parkin dysfunction might dysregulate mitochondria and the cells’ response to oxidative stress, which would move AIMP2 toward the nucleus and set off the death cascade.—Gwyneth Dickey Zakaib


  1. This work by Lee and colleagues builds on a discovery we made several years ago related to the identification of AIMP2/p38 as a substrate of the multifunctional E3 ubiquitin ligase, parkin. This ligase is dysfunctional in nearly 30 percent of autosomal recessive Parkinson’s disease cases with early onset (PD) (Corti et al., 2003).

    AIMP2/p38 is a central scaffold subun it of the multiaminoacyl-tRNA synthetase complex which catalyzes the esterification of amino acids with their cognate tRNAs and therefore plays a crucial role in protein biosynthesis. Ten years ago, we showed that this protein forms inclusions and is toxic when overproduced in cell models and that Parkin provides protection against this toxicity. We also found that AIMP2/p38 accumulates in a subset of Lewy bodies in cases of sporadic PD. This data suggested a possible involvement of the protein in PD disease pathogenesis, although the underlying molecular mechanisms remained to be discovered.

    Ted Dawson’s team confirmed and added to our data already in 2005 (Ko et al., 2005). They showed that among a number of putative Parkin substrates published in the literature only AIMP2/p38 was upregulated at the protein level in Parkin-deficient mice. Importantly, they also found upregulation of AIMP2/p38 in the brains of cases of parkin-linked PD, as well as of sporadic PD and Lewy body dementia. This was a surprise to us, since our analyses did not reveal increased AIMP2/p38 abundance in parkin knock-out mice or in PD patients, but this discrepancy might be due to differences in experimental procedures ( Periquet et al., 2005; Hampe C & Corti O, unpublished work): the antibodies that we used to detect the protein in vivo were different from those used in Ted Dawson’s study. Dawson’s team also demonstrated that overproduction of AIMP2/p38 in the mouse substantia nigra caused significant dopaminergic neuron loss; such effects had been previously observed with similar approaches for alpha-synuclein (Kirik et al., 2002; Lo Bianco et al., 2002), for which a clear link with PD pathogenesis had been established by genetic studies, as well as through the involvement of the protein as an invariable Lewy body component. Based on all these observations, AIMP2/p38 was christened by Ko and colleagues as “the authentic Parkin substrate”, but the mechanisms responsible for its toxicity remained mysterious.

    With the present study, the Dawson team adds a new piece to the Parkin puzzle. By using an impressive combination of challenging complementary approaches and tools (inducible transgenic and knock-out mice; viral vector-mediated gene expression; immunochemical, biochemical and behavioral analyses; studies in cell models and PD patients), Lee and colleagues provide evidence for a direct and non-canonical role of AIMP2/p38 in a particular form of cell death termed parthanatos, involving the nuclear translocation of an apoptosis-inducing factor and caspase-independent fragmentation of chromosomal DNA. Altogether, the data suggest that accumulation of AIMP2/p38 in parkin-linked and possibly in sporadic PD leads to its nuclear translocation, followed by activation of PARP1, formation of PAR polymer and neuronal death. This mechanism of cell death appears to be specific to dopaminergic neurons, as PARP1 activation was not observed in the cortex of AIMP2/p38 transgenic mice or PD patients.

    Although the data are in general impressive and convincingly support a role for AIMP2/p38 and PARP1-dependent cell death in PD, they also raise a number of questions that will have to be addressed in future studies:

    What are the mechanisms behind the selective toxicity of AIMP2/p38 to dopaminergic neurons?
    AIMP2/p38 is a ubiquitous protein, and although researchers (Imam et al., 2011 , Ko et al., 2010 ) reported that brain areas relatively spared in PD, such as the cortex, do not accumulate AIMP2/p38, Ko et al showed increased abundance of AIMP2/p38 in the cingulate cortex of patients with PD and Lewy body dementia (Ko et al., 2005).

    In the present study, Lee and coworkers did not observe general neuronal degeneration in the cortex of AIMP2/p38 transgenic mice, which accumulate substantial amounts of the protein. However, they did not analyze specific non-dopaminergic neuronal subpopulations. It would have been particularly interesting to determine the effects of AIMP2/p38 overexpression in the noradrenergic neurons of the locus coeruleus, shown by Dawson’s team to be the only neuronal population subjected to degeneration in a constitutive parkin knock-out mouse model ( von Coelln et al., 2004 ). Would down-regulation of AIMP2/p38 or PARP1 protect these neurons from degeneration?

    Can other mechanisms underlying AIMP2/p38-induced degeneration be excluded?
    Lee and co-workers show that the overproduction of AIMP2/p38 does not affect gross basal or stress-induced protein translation in cell models. However, taken the canonical function of AIMP2/p38 into consideration, can moderate defects be excluded in vivo? Could they participate in enhancing neuronal vulnerability in the long-term? Deregulation of protein translation control under stress conditions has been suggested to unify mechanism in PD with distinct genetic causes ( Imai et al., 2008 ; Tain et al., 2009  ; Gehrke et al., 2010 ); could accumulation of AIMP2/p38 reflect disassembly of the multiaminoacyl-tRNA synthetase complex as a neuroprotective strategy aimed at arresting protein translation in stressed neurons?

    How relevant is AIMP2/p38 to other forms of PARP1-induced degeneration?
    In earlier studies Dawson’s team provided evidence for a role of parthanatos in dopaminergic neuron degeneration caused by the neurotoxin MPTP ( Mandir et al., 1999 ; Wang et al., 2003 ); conversely, they reported accumulation of AIMP2/p38 in the MPTP model ( Ko et al., 2010 ). Would down-regulation of AIMP2/p38, mediated for example by RNA interference, prevent MPTP-induced neurodegeneration?

    What is the relative contribution of AIMP2/p38 compared to other proteins/mechanisms suspected to be involved in dopaminergic neuron degeneration linked to loss of Parkin function?
    AIMP2/p38 is one of more than twenty putative Parkin substrates reported in the literature; a number of them have been shown to cause dopaminergic neurodegeneration in in vivo models ( for example work by several groups on Pael receptor/GPR37 or CDCrel-1); moreover, Parkin is a versatile ubiquitin ligase promoting different types of ubiquitylation, including monoubiquitylation and polyubiquitylation via various lysine residues of ubiquitin. Dawson’s team alone has provided considerable contributions in the field; they reported on other “authentic” Parkin substrates found to accumulate in park PD models and in patients with PD and on several possible mechanisms of neurodegeneration (see for example Ko et al., 2006, or Shin et al., 2011).

    In their recent study, Shin and coworkers demonstrated that Parkin regulates the abundance of PARIS, a transcriptional repressor of PGC-1α involved in mitochondrial biogenesis. This discovery put Parkin at the center of mitochondrial quality control with the proposal of a new mechanism by which this versatile ubiquitin ligase may preserve this organelle, probably central to the physiopathology of autosomal recessive PD (see Corti and Brice 2013).

    What is the relative contribution of these proteins to parkin-linked dopamine neuron degeneration and where do they meet on the intricate map of the multiple mechanisms suspected to play a role in PD?
    If downregulation of PARIS alone can prevent neurodegeneration in a conditional parkin knock-out mouse model, as demonstrated by Shin JH and coworkers, is there room left for a detrimental role of AIMP2/p38? Would downregulation of AIMP2/p38 also protect dopaminergic neurons? Lee and colleagues did not perform this experiment.


    . The p38 subunit of the aminoacyl-tRNA synthetase complex is a Parkin substrate: linking protein biosynthesis and neurodegeneration. Hum Mol Genet. 2003 Jun 15;12(12):1427-37. PubMed.

    . Accumulation of the authentic parkin substrate aminoacyl-tRNA synthetase cofactor, p38/JTV-1, leads to catecholaminergic cell death. J Neurosci. 2005 Aug 31;25(35):7968-78. PubMed.

    . Proteomic analysis of parkin knockout mice: alterations in energy metabolism, protein handling and synaptic function. J Neurochem. 2005 Dec;95(5):1259-76. PubMed.

    . Parkinson-like neurodegeneration induced by targeted overexpression of alpha-synuclein in the nigrostriatal system. J Neurosci. 2002 Apr 1;22(7):2780-91. PubMed.

    . alpha -Synucleinopathy and selective dopaminergic neuron loss in a rat lentiviral-based model of Parkinson's disease. Proc Natl Acad Sci U S A. 2002 Aug 6;99(16):10813-8. PubMed.

    . Novel regulation of parkin function through c-Abl-mediated tyrosine phosphorylation: implications for Parkinson's disease. J Neurosci. 2011 Jan 5;31(1):157-63. PubMed.

    . Phosphorylation by the c-Abl protein tyrosine kinase inhibits parkin's ubiquitination and protective function. Proc Natl Acad Sci U S A. 2010 Sep 21;107(38):16691-6. Epub 2010 Sep 7 PubMed.

    . Loss of locus coeruleus neurons and reduced startle in parkin null mice. Proc Natl Acad Sci U S A. 2004 Jul 20;101(29):10744-9. PubMed.

    . Phosphorylation of 4E-BP by LRRK2 affects the maintenance of dopaminergic neurons in Drosophila. EMBO J. 2008 Sep 17;27(18):2432-43. PubMed.

    . Rapamycin activation of 4E-BP prevents parkinsonian dopaminergic neuron loss. Nat Neurosci. 2009 Sep;12(9):1129-35. PubMed.

    . Pathogenic LRRK2 negatively regulates microRNA-mediated translational repression. Nature. 2010 Jul 29;466(7306):637-41. PubMed.

    . Poly(ADP-ribose) polymerase activation mediates 1-methyl-4-phenyl-1, 2,3,6-tetrahydropyridine (MPTP)-induced parkinsonism. Proc Natl Acad Sci U S A. 1999 May 11;96(10):5774-9. PubMed.

    . Apoptosis inducing factor and PARP-mediated injury in the MPTP mouse model of Parkinson's disease. Ann N Y Acad Sci. 2003 Jun;991:132-9. PubMed.

    . Identification of far upstream element-binding protein-1 as an authentic Parkin substrate. J Biol Chem. 2006 Jun 16;281(24):16193-6. PubMed.

    . PARIS (ZNF746) repression of PGC-1α contributes to neurodegeneration in Parkinson's disease. Cell. 2011 Mar 4;144(5):689-702. PubMed.

    . Mitochondrial quality control turns out to be the principal suspect in parkin and PINK1-related autosomal recessive Parkinson's disease. Curr Opin Neurobiol. 2013 Feb;23(1):100-8. PubMed.

    View all comments by Olga Corti

Make a Comment

To make a comment you must login or register.


News Citations

  1. Can Calcium Channel Blockers Save Stressed-Out Dopaminergic Neurons?
  2. Abnormal Mitochondrial Dynamics—Early Event in AD, PD?

Paper Citations

  1. . The p38 subunit of the aminoacyl-tRNA synthetase complex is a Parkin substrate: linking protein biosynthesis and neurodegeneration. Hum Mol Genet. 2003 Jun 15;12(12):1427-37. PubMed.
  2. . Accumulation of the authentic parkin substrate aminoacyl-tRNA synthetase cofactor, p38/JTV-1, leads to catecholaminergic cell death. J Neurosci. 2005 Aug 31;25(35):7968-78. PubMed.
  3. . Poly(ADP-ribose) (PAR) binding to apoptosis-inducing factor is critical for PAR polymerase-1-dependent cell death (parthanatos). Sci Signal. 2011;4(167):ra20. PubMed.
  4. . PARP inhibition: PARP1 and beyond. Nat Rev Cancer. 2010 Apr;10(4):293-301. PubMed.

External Citations

  1. diagram

Further Reading


  1. . Accumulation of the authentic parkin substrate aminoacyl-tRNA synthetase cofactor, p38/JTV-1, leads to catecholaminergic cell death. J Neurosci. 2005 Aug 31;25(35):7968-78. PubMed.
  2. . Poly(ADP-ribose) (PAR) binding to apoptosis-inducing factor is critical for PAR polymerase-1-dependent cell death (parthanatos). Sci Signal. 2011;4(167):ra20. PubMed.
  3. . PARP inhibition: PARP1 and beyond. Nat Rev Cancer. 2010 Apr;10(4):293-301. PubMed.
  4. . The p38 subunit of the aminoacyl-tRNA synthetase complex is a Parkin substrate: linking protein biosynthesis and neurodegeneration. Hum Mol Genet. 2003 Jun 15;12(12):1427-37. PubMed.

Primary Papers

  1. . Parthanatos mediates AIMP2-activated age-dependent dopaminergic neuronal loss. Nat Neurosci. 2013 Aug 25; PubMed.