Long-term effects of a lumbosacral ventral root avulsion injury on axotomized motor neurons and avulsed ventral roots in a non-human primate model of cauda equina injury

Effects of local and systemic treatment with human natural killer-1 mimetic peptide (HNK-1) after ventral root avulsion and reimplantation in mice

Background

Spinal ventral root injuries generate significant motoneuron degeneration, which hinders full functional recovery. The poor prognosis of functional recovery can be attributed to the use or combination of different therapeutic approaches. Several molecules have been screened as potential treatments in combination with surgical reimplantation of the avulsed roots, the gold standard approach for such injuries. Among the studied molecules, human natural killer-1 (HNK-1) stands out as it is related to the stimulation of motor axon outgrowth. Therefore, we aimed to comparatively investigate the effects of local administration of an HNK-1 mimetic peptide (mp-HNK-1) and systemic treatment with ursolic acid (UA), another HNK-1 mimetic, after ventral root avulsion and reimplantation with heterologous fibrin biopolymer (HFB).

Methods

Female mice of the isogenic strain C57BL/6JUnib were divided into five experimental groups: Avulsion, Reimplantation, mp-HNK-1 (in situ), and UA (systemic treatment). Mice were evaluated 2 and 12 weeks after surgery. Functional assessment was performed every four days using the Catwalk platform. Neuronal survival was analyzed by cytochemistry, and glial reactions and synaptic coverage were evaluated by immunofluorescence.

Results

Treatment with UA elicited long-term neuroprotection, accompanied by a decrease in microglial reactions, and reactive astrogliosis. The neuroprotective effects of UA were preceded by increased glutamatergic and GABAergic inputs in the ventral spinal cord two weeks after injury. However, a single application of mp-HNK-1 had no significant effects. Functional analysis showed that UA treatment led to an improvement in motor and sensory recovery.

Conclusion

Overall, the results indicate that UA is neuroprotective, acting on glial cells and synaptic maintenance, and the combination of these findings led to a better functional recovery.

Nonbinary 2D Distribution Tool Maps Autonomic Nerve Fiber Clustering in Lumbosacral Ventral Roots of Rhesus Macaques

Neuromodulation of the peripheral nervous system (PNS) by electrical stimulation may augment autonomic function after injury or in neurodegenerative disorders. Nerve fiber size, myelination, and distance between individual fibers and the stimulation electrode may influence response thresholds to electrical stimulation. However, information on the spatial distribution of nerve fibers within the PNS is sparse. We developed a new two-dimensional (2D) morphological mapping tool to assess spatial heterogeneity and clustering of nerve fibers. The L6-S3 ventral roots (VRs) in rhesus macaques were used as a model system to map preganglionic parasympathetic, γ-motor, and α-motor fibers. Random and ground truth distributions of nerve fiber centroids were determined for each VR by light microscopy. The proposed tool allows for nonbinary determinations of fiber heterogeneity by defining the minimum distance between nerve fibers for cluster inclusion and comparisons with random fiber distributions for each VR. There was extensive variability in the relative composition of nerve fiber types and degree of 2D fiber heterogeneity between different L6-S3 VR levels within and across different animals. There was a positive correlation between the proportion of autonomic fibers and the degree of nerve fiber clustering. Nerve fiber cluster heterogeneity between VRs may contribute to varied functional outcomes from neuromodulation.

Identification of Potential Ferroptosis Key Genes in the Pathogenesis of Lumbosacral Spinal Root Avulsion by RNA Sequencing and Bioinformatics Analysis

Objective: Ferroptosis is a type of cell death involved in various human diseases, including nerve injury. However, the role of ferroptosis in lumbosacral spinal root avulsion (LSRA) remains unknown. This study aims to investigate whether ferroptosis is induced after LSRA and the key ferroptosis-related genes and their potential function in LSRA.

Methods: The biochemical and morphological changes of ferroptosis were determined by detection of iron accumulation and by transmission electron microscopy in a rat LSRA model. The transcriptional expression profile following LSRA was investigated by RNA sequencing and ferroptosis-related genes were downloaded from FerrDb and used to identify ferroptosis differentially expressed genes (DEGs). The differential expressions of ferroptosis DEGs were confirmed by qRT-PCR analysis. The potential functions of ferroptosis DEGs were revealed by DAVID 6.8 and WebGestalt. A protein–protein interaction (PPI) network and gene–miRNA interaction network were further constructed to identify key modules in ferroptosis DEGs, and the results were verified by qRT-PCR and western blot analysis.

Results: LSRA was followed by ferroptosis-specific changes, such as shrunken mitochondria and increased iron accumulation, that can be alleviated by ferroptosis inhibitor deferoxamine (DFO). A total of 2,446 DEGs and 46 ferroptosis DEGs were identified after LSRA, and over 90% of the ferroptosis DEGs were confirmed to be differentially expressed following LSRA, which can also be eliminated by DFO treatment. Functional analysis demonstrated significant enrichment of the ferroptosis DEGs in pathways related to the oxidative stress response, the HIF-1 signaling pathway, and the tumor necrosis factor signaling pathway. PPI network analysis demonstrated that a set of key modules in ferroptosis DEGs were related to the HIF-1 signaling pathway: Il6, Nos2, Stat3, Hif1a, Vegfa, Cdkn1a, and Rela. Construction of a gene–miRNA network predicted miRNAs targeting four key ferroptosis DEGs—Stat3, Hif1a, Vegfa, and Rela, and further western blot analysis confirmed their upregulation after LSRA, which can be alleviated by DFO pretreatment.

Conclusion: The data revealed the induction of ferroptosis in a rat LSRA model and identified possible regulatory roles for ferroptosis-related genes in the molecular mechanisms of LSRA, which provides new insights into the pathogenesis and helps to find new molecular targets for the treatment of LSRA.

Long-Term Suppression of c-Jun and nNOS Preserves Ultrastructural Features of Lower Motor Neurons and Forelimb Function after Brachial Plexus Roots Avulsion

Brachial plexus root avulsions cause debilitating upper limb paralysis. Short-term neuroprotective treatments have reported preservation of motor neurons and function in model animals while reports of long-term benefits of such treatments are scarce, especially the morphological sequelae. This morphological study investigated the long-term suppression of c-Jun- and neuronal nitric oxide synthase (nNOS) (neuroprotective treatments for one month) on the motor neuron survival, ultrastructural features of lower motor neurons, and forelimb function at six months after brachial plexus roots avulsion. Neuroprotective treatments reduced oxidative stress and preserved ventral horn motor neurons at the end of the 28-day treatment period relative to vehicle treated ones. Motor neuron sparing was associated with suppression of c-Jun, nNOS, and pro-apoptotic proteins Bim and caspases at this time point. Following 6 months of survival, neutral red staining revealed a significant loss of most of the motor neurons and ventral horn atrophy in the avulsed C6, 7, and 8 cervical segments among the vehicle-treated rats (n = 4). However, rats that received neuroprotective treatments c-Jun JNK inhibitor, SP600125 (n = 4) and a selective inhibitor of nNOS, 7-nitroindazole (n = 4), retained over half of their motor neurons in the ipsilateral avulsed side compared. Myelinated axons in the avulsed ventral horns of vehicle-treated rats were smaller but numerous compared to the intact contralateral ventral horns or neuroprotective-treated groups. In the neuroprotective treatment groups, there was the preservation of myelin thickness around large-caliber axons. Ultrastructural evaluation also confirmed the preservation of organelles including mitochondria and synapses in the two groups that received neuroprotective treatments compared with vehicle controls. Also, forelimb functional evaluation demonstrated that neuroprotective treatments improved functional abilities in the rats. In conclusion, neuroprotective treatments aimed at suppressing degenerative c-Jun and nNOS attenuated apoptosis, provided long-term preservation of motor neurons, their organelles, ventral horn size, and forelimb function.

A shape-adjusted ellipse approach corrects for varied axonal dispersion angles and myelination in primate nerve roots

Segmentation of axons in light and electron micrographs allows for quantitative high-resolution analysis of nervous tissues, but varied axonal dispersion angles result in over-estimates of fiber sizes. To overcome this technical challenge, we developed a novel shape-adjusted ellipse (SAE) determination of axonal size and myelination as an all-inclusive and non-biased tool to correct for oblique nerve fiber presentations. Our new resource was validated by light and electron microscopy against traditional methods of determining nerve fiber size and myelination in rhesus macaques as a model system. We performed detailed segmental mapping and characterized the morphological signatures of autonomic and motor fibers in primate lumbosacral ventral roots (VRs). An en bloc inter-subject variability for the preganglionic parasympathetic fibers within the L7-S2 VRs was determined. The SAE approach allows for morphological ground truth data collection and assignment of individual axons to functional phenotypes with direct implications for fiber mapping and neuromodulation studies.

ISP and PAP4 peptides promote motor functional recovery after peripheral nerve injury

Both intracellular sigma peptide (ISP) and phosphatase and tensin homolog agonist protein (PAP4) promote nerve regeneration and motor functional recovery after spinal cord injury. However, the role of these two small peptides in peripheral nerve injury remains unclear. A rat model of brachial plexus injury was established by crush of the C6 ventral root. The rats were then treated with subcutaneous injection of PAP4 (497 µg/d, twice per day) or ISP (11 µg/d, once per day) near the injury site for 21 successive days. After ISP and PAP treatment, the survival of motoneurons was increased, the number of regenerated axons and neuromuscular junctions was increased, muscle atrophy was reduced, the electrical response of the motor units was enhanced and the motor function of the injured upper limbs was greatly improved in rats with brachial plexus injury. These findings suggest that ISP and PAP4 promote the recovery of motor function after peripheral nerve injury in rats. The animal care and experimental procedures were approved by the Laboratory Animal Ethics Committee of Jinan University of China (approval No. 20111008001) in 2011.

GDNF Gene Therapy to Repair the Injured Peripheral Nerve

A spinal root avulsion is the most severe proximal peripheral nerve lesion possible. Avulsion of ventral root filaments disconnects spinal motoneurons from their target muscles, resulting in complete paralysis. In patients that undergo brachial plexus nerve repair, axonal regeneration is a slow process. It takes months or even years to bridge the distance from the lesion site to the distal targets located in the forearm. Following ventral root avulsion, without additional pharmacological or surgical treatments, progressive death of motoneurons occurs within 2 weeks (Koliatsos et al., 1994). Reimplantation of the avulsed ventral root or peripheral nerve graft can act as a conduit for regenerating axons and increases motoneuron survival (Chai et al., 2000). However, this beneficial effect is transient. Combined with protracted and poor long-distance axonal regeneration, this results in permanent function loss. To overcome motoneuron death and improve functional recovery, several promising intervention strategies are being developed. Here, we focus on GDNF gene-therapy. We first introduce the experimental ventral root avulsion model and discuss its value as a proxy to study clinical neurotmetic nerve lesions. Second, we discuss our recent studies showing that GDNF gene-therapy is a powerful strategy to promote long-term motoneuron survival and improve function when target muscle reinnervation occurs within a critical post-lesion period. Based upon these observations, we discuss the influence of timing of the intervention, and of the duration, concentration and location of GDNF delivery on functional outcome. Finally, we provide a perspective on future research directions to realize functional recovery using gene therapy.

Expression of Nogo-A in dorsal root ganglion in rats with cauda equina injury

OBJECTIVE

To investigate the expression of Nogo-A in dorsal root ganglion (DRG) in rats with cauda equina injury and the therapeutic effects of blocking Nogo-A and its receptor.

METHODS AND MATERIALS

Fifty-eight male Sprague-Dawley rats were divided randomly into either the sham operation group (n = 24) or the cauda equina compression (CEC) control group (n = 34). Behavioral, histological, and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) analyses were conducted to assess the establishment of the model. The dynamic expression change of Nogo-A was evaluated using real time-qPCR. Immunofluorescence was used to evaluate the expression of Nogo-A in the DRG and cauda equina. Furthermore, 20 male Sprague-Dawley rats were equally divided into 4 groups, including the sham group, the CEC group, the NEP1-40 (the NgR antagonist peptide) treatment group, and the JTE-013 (the S1PR2 antagonist) treatment group. Behavioral assessments and western blotting were used to evaluate the therapeutic effect of cauda equina injury via blocking Nogo-A and its receptor.

RESULTS

Tactile allodynia and heat hyperalgesia in the CEC model developed as soon as 1 day after surgery and recovered to normal at 7 days, which was followed by the downregulation of Nogo-A in DRG neurons. However, the locomotor function impairment in the CEC model showed a different prognosis from the sensory function, which was consistent with the expression change of Nogo-A in the spinal cord. Immunofluorescence results also demonstrated that Nogo A-positive/NF200-negative neurons and axons increased in the DRG and cauda equina 7 days after surgery. Surprisingly, Schwann cells, which myelinate axons in the PNS, also expressed considerable amounts of Nogo-A. Then, after blocking the Nogo-A/NgR signaling pathway by NEP1-40, significant improvement of mechanical allodynia was identified in the first 2 days after the surgery. Western blotting suggested the NEP1-40 treatment group had lower expression of cleaved caspase-3 than the CEC and JTE-013 treatment group.

CONCLUSION

Neuronal Nogo-A in the DRG may be involved in regeneration and play a protective role in the CEC model. Whereas Nogo-A, released from the injured axons or expressed by Schwann cells, may act as an inhibiting factor in the process of CEC repairment. Thus, blocking the Nogo-A/NgR signaling pathway can alleviate mechanical allodynia by apoptosis inhibition.

Up-regulation of heat shock protein 27 inhibits apoptosis in lumbosacral nerve root avulsion-induced neurons

Lumbosacral nerve root avulsion leads to widespread death of neurons in the anterior horn area of the injured spinal cord, which results in dysfunction in the lower extremities. Heat shock protein 27 (Hsp27) has been found to play cytoprotective roles under adverse conditions. However, the role of Hsp27 in neurons after lumbosacral nerve root avulsion is unknown. The aim of the present study was to investigate the effects and mechanism of action of Hsp27 on neurons after lumbosacral nerve root avulsion. It was found that Hsp27 expression was elevated in the anterior horn area of the injured spinal cord and the up-regulation of Hsp27 protected neurons against apoptosis after lumbosacral nerve root avulsion. In addition, Hsp27 plays an anti-apoptotic role by suppressing oxidative stress reactions. These findings indicated that Hsp27 may play a key role in resistance to lumbosacral nerve root avulsion-induced neuron apoptosis and may prove to be a potential strategy for improving prognosis after lumbosacral nerve root avulsion.

Neuregulin-1 Accelerates Functional Motor Recovery by Improving Motoneuron Survival After Brachial Plexus Root Avulsion in Mice

Brachial plexus root avulsion (BPRA) results in the complete loss of motor function in the upper limb, mainly due to the death of spinal motoneurons (MNs). The survival of spinal MNs is the key to the recovery of motor function. Neuregulin-1 (Nrg1) plays fundamental roles in nervous system development and nerve repair. However, its functional role in BPRA remains unclear. On the basis of our findings that Nrg1 is down-regulated in the ventral horn in a mouse model of BPRA, Nrg1 may be associated with BPRA. Here, we investigated whether recombinant Nrg1β (rNrg1β) can enhance the survival of spinal MNs and improve functional recovery in mice following BPRA. In vitro studies on primary cultured mouse MNs showed that rNrg1β increased the survival rate in a dose-dependent manner, reaching a peak at 5 nM, which increased the survival rate and enhanced the pERK levels in MNs under H₂O₂-induced oxidative stress. In vivo studies revealed that rNrg1β improved the functional recovery of elbow flexion, promoted the survival of MNs, enhanced the re-innervation of biceps brachii, and decreased the muscle atrophy. These results suggest that Nrg1 may provide a potential therapeutic strategy for root avulsion.

EMG characteristics of the external anal sphincter guarding reflex and effects of a unilateral ventral root avulsion injury in rhesus macaques ( Macaca mulatta)

The external anal sphincter (EAS) is important for the maintenance of bowel continence and may be compromised by a variety of neuropathic conditions. However, large animal models for the study of EAS functions have been sparse. The EAS guarding reflex was examined by electromyography (EMG) in neurologically intact rhesus macaques ( n = 6) and at 4-6 wk after a unilateral EAS denervation from an L6-S3 ventral root avulsion (VRA) injury ( n = 6). Baseline EAS EMG recordings were quiescent in all subjects, and evoked responses showed an initial large-amplitude EMG activity, which gradually returned to baseline within 1-2 min. At 4-6 wk postoperatively, the EAS guarding reflex showed a significantly reduced EMG response duration of 47 ± 15 s and area under the curve (AUC) of 0.198 ± 0.097 mV·s compared with the corresponding evoked EAS EMG duration of 102 ± 19 s and AUC of 0.803 ± 0.225 mV·s ( P < 0.05) in the control group. Detailed time- and frequency-domain analysis of the evoked EAS EMG responses for the first 40 s showed no difference between groups for the maximum amplitude but a significant decrease for the mean amplitude across the study period and an early AUC reduction for the first 10 s in the VRA injury group. Time-frequency analysis and power spectrum plots indicated decreased intensity and a narrower midrange of frequencies in the VRA injury group. We conclude that the EAS guarding reflex in rhesus macaques shows characteristic EMG features in control subjects and signs of partial target denervation after a unilateral L6-S3 VRA injury. NEW & NOTEWORTHY The external anal sphincter guarding reflex showed initial large-amplitude peaks and a gradual return to a quiescent baseline after a rectal probe stimulus in rhesus macaques. At 4-6 wk after a unilateral ventral root avulsion (VRA) injury, the electromyography duration, mean amplitude, and area under the curve measurements were decreased. Time-frequency analysis and power spectrum plots indicated decreased intensity and a narrowed midrange of frequencies in the VRA injury cohort.

Microtubule stabilization promoted axonal regeneration and functional recovery after spinal root avulsion

A spinal root avulsion injury disconnects spinal roots with the spinal cord. The rampant motoneuron death, inhibitory CNS/PNS transitional zone (TZ) for axonal regrowth and limited regeneration speed together lead to motor dysfunction. Microtubules rearrange to assemble a new growth cone and disorganized microtubules underline regeneration failure. It has been shown that microtubule-stabilizing drug, Epothilone B, enhanced axonal regeneration and attenuated fibrotic scaring after spinal cord injury. Here, we are reporting that after spinal root avulsion+ re-implantation in adult rats, EpoB treatment improved motor functional recovery and potentiated electrical responses of motor units. It facilitated axons to cross the TZ and promoted more and bigger axons in the peripheral nerve. Neuromuscular junctions were reformed with better preserved postsynaptic structure, and muscle atrophy was prevented by EpoB administration. Our study showed that EpoB was a promising therapy for promoting axonal regeneration after peripheral nerve injury.

A ventral root avulsion injury model for neurogenic underactive bladder studies

Detrusor underactivity (DU) is defined as a contraction of reduced strength and/or duration during bladder emptying and results in incomplete and prolonged bladder emptying. The clinical diagnosis of DU is challenging when present alone or in association with other bladder conditions such as detrusor overactivity, urinary retention, detrusor hyperactivity with impaired contractility, aging, and neurological injuries. Several etiologies may be responsible for DU or the development of an underactive bladder (UAB), but the pathobiology of DU or UAB is not well understood. Therefore, new clinically relevant and interpretable models for studies of UAB are much needed in order to make progress towards new treatments and preventative strategies. Here, we review a neuropathic cause of DU in the form of traumatic injuries to the cauda equina (CE) and conus medullaris (CM) portions of the spinal cord. Lumbosacral ventral root avulsion (VRA) injury models in rats mimic the clinical phenotype of CM/CE injuries. Bilateral VRA injuries result in bladder areflexia, whereas a unilateral lesion results in partial impairment of lower urinary tract and visceromotor reflexes. Surgical re-implantation of avulsed ventral roots into the spinal cord and pharmacological strategies can augment micturition reflexes. The translational research need for the development of a large animal model for UAB studies is also presented, and early studies of lumbosacral VRA injuries in rhesus macaques are discussed.

Radiographic and Magnetic Resonance Imaging Identification of Thoracolumbar Spine Variants with Implications for the Positioning of the Conus Medullaris in Rhesus Macaques

The anatomy of the vertebral column in mammals may differ between species and between subjects of the same species, especially with regards to the composition of the thoracolumbar spine. We investigated, using several noninvasive imaging techniques, the thoracolumbar spine of a total of 44 adult rhesus macaques of both genders. Radiographic examination of the vertebral column showed a predominant spine phenotype with 12 rib-bearing thoracic vertebrae and 7 lumbar vertebrae without ribs in 82% of subjects, whereas a subset of subjects demonstrated 13 rib-bearing thoracic vertebrae and 6 lumbar vertebrae without ribs. Computer tomography studies of the thoraco-lumbar spine in two cases with a pair of supernumerary ribs showed facet joints between the most caudal pair of ribs and the associated vertebra, supporting a thoracic phenotype. Magnetic resonance imaging (MRI) studies were used to determine the relationship between the lumbosacral spinal cord and the vertebral column. The length of the conus medullaris portion of the spinal cord was 1.5 ± 0.3 vertebral units, and its rostral and caudal positions in the spinal canal were at 2.0 ± 0.3 and 3.6 ± 0.4 vertebral units below the thoracolumbar junction, respectively (n = 44). The presence of a set of supernumerary ribs did not affect the length or craniocaudal position of the conus medullaris, and subjects with13 rib-bearing vertebrae may from a functional or spine surgical perspective be considered as exhibiting12 thoracic vertebrae and an L1 vertebra with ribs. Anat Rec, 300:300-308, 2017. © 2016 Wiley Periodicals, Inc.

Selected gene profiles of stressed NSC-34 cells and rat spinal cord following peripheral nerve reconstruction and minocycline treatment

The present study was conducted to investigate the effects of minocycline on the expression of selected transcriptional and translational profiles in the rat spinal cord following sciatic nerve (SNR) transection and microsurgical coaptation. The mRNA and protein expression levels of B cell lymphoma-2 (Bcl-2), Bcl-2-associated X protein (Bax), caspase-3, major histocompatibility complex I (MHC I), tumor necrosis factor-α (TNF-α), activating transcription factor 3 (ATF3), vascular endothelial growth factor (VEGF), matrix metalloproteinase 9 (MMP9), and growth associated protein-43 (GAP-43) were monitored in the rat lumbar spinal cord following microsurgical reconstruction of the sciatic nerves and minocycline treatment. The present study used semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR) and immunohistochemistry. As a PCR analysis of spinal cord tissue enabled the examination of the expression patterns of all cell types including glia, the motorneuron-like NSC-34 cell line was used to investigate expression level changes in motorneurons. As stressors, oxygen glucose deprivation (OGD) and lipopolysaccharide (LPS) treatment were performed. SNR did not induce significant degeneration of ventral horn motorneurons, whereas microglia activation and synaptic terminal retraction were detectable. All genes were constitutively expressed at the mRNA and protein levels in untreated spinal cord and control cells. SNR significantly increased the mRNA expression levels of all genes, albeit only temporarily. In all genes except MMP9 and GAP-43, the induction was seen ipsilaterally and contralaterally. The effects of minocycline were moderate. The expression levels of MMP9, TNF-α, MHC I, VEGF, and GAP-43 were reduced, whereas those of Bax and Bcl-2 were unaffected. OGD, but not LPS, was toxic for NSC-34 cells. No changes in the expression levels of Bax, caspase-3, MHC I or ATF3 were observed. These results indicated that motorneurons were not preferentially or solely responsible for SNR-mediated upregulation of these genes. MMP9, TNF-α, VEGF and Bcl-2 were stress-activated. These results suggest that a substantial participation of motorneurons in gene expression levels in vivo. Minocycline was also shown to have inhibitory effects. The nuclear factor-κB signalling pathway may be a possible target of minocycline.

Enhanced regeneration and functional recovery after spinal root avulsion by manipulation of the proteoglycan receptor PTPσ

Following root avulsion, spinal nerves are physically disconnected from the spinal cord. Severe motoneuron death and inefficient axon regeneration often result in devastating motor dysfunction. Newly formed axons need to extend through inhibitory scar tissue at the CNS-PNS transitional zone before entering into a pro-regenerative peripheral nerve trajectory. CSPGs are dominant suppressors in scar tissue and exert inhibition via neuronal receptors including PTPσ. Previously, a small peptide memetic of the PTPσ wedge region named ISP (Intracellular Sigma Peptide) was generated, and its capabilities to target PTPσ and relieve CSPG inhibition were validated. Here, we demonstrate that after ventral root avulsion and immediate re-implantation, modulation of PTPσ by systemic delivery of ISP remarkably enhanced regeneration. ISP treatment reduced motoneuron death, increased the number of axons regenerating across scar tissue, rebuilt healthy neuromuscular junctions and enhanced motor functional recovery. Our study shows that modulation of PTPσ is a potential therapeutic strategy for root avulsion.