Wallerian degeneration: From mechanism to disease to imaging

Abstract

Wallerian degeneration (WD) was first discovered by Augustus Waller in 1850 in a transection of the glossopharyngeal and hypoglossal nerves in frogs. Initial studies suggested that the formation mechanism of WD is related to the nutrition of neuronal cell bodies to axons. However, with the wide application of transgenic mice in experiments, the latest studies have found that the mechanism of WD is related to axonal degeneration, myelin clearance and extracellular matrix. This review summarizes the discovery and research progress of WD and discusses the mechanism of WD from the perspective of molecular biology. In addition, this review combines the etiology, symptoms and imaging results of WD patients, and analyzes the clinical and imaging characteristics of WD, to provide the best perspective for future clinical research.

1. Introduction

In 1850, Augustus Waller described the histopathological degeneration of axons and myelin sheaths after transection of the hypoglossal and hypoglossal nerves in frogs, and first found that the distal portion of the nerves undergoes progressive degeneration after transection of the nerves, after which the phenomenon was commonly referred to as Wallerian degeneration (WD) [1], and was explained as a passive phenomenon in which the cytosol nourishes the axons. It was not until 1989 that Lunn et al. changed Wallerian degeneration by discovering a natural mouse strain that delays Wallerian degeneration (C57BL/6/Ola) to change this traditional concept (Lunn et al., 1989) [2].

2. What is Wallerian degeneration

In 1850, Augustus Waller described the histopathological degeneration of axons and myelin sheaths after transection of the hypoglossal and hypoglossal nerves in frogs, and first found that the distal portion of the nerves undergoes progressive degeneration after transection of the nerves, after which the phenomenon was commonly referred to as Wallerian degeneration (WD) [1], and was explained as a passive phenomenon in which the cytosol nourishes the axons. It was not until 1989 that Lunn et al. changed Wallerian degeneration by discovering a natural mouse strain that delays Wallerian degeneration (C57BL/6/Ola) to change this traditional concept (Lunn et al., 1989) [2]. When carrying the autosomal dominant inheritance "Wallerian Degeneration Slow" (Wlds) allele and wild-type mice were subjected to sciatic nerve transection, distal axons remained structurally and metabolically intact for up to two weeks in the Wlds mice, however, axons degenerated in less than two days in the wild-type mice. A subsequent study used quantitative and serial section electron microscopy to find out whether the onset and time course of presynaptic nerve terminal degeneration were delayed in the striatum of Wallerian degeneration (Wlds) mutant mice. Synaptic degeneration was observed within 48 h after cortical ablation in wild-type mice but was delayed by approximately 1 week in Wlds wild-type mice. This delayed synaptic degeneration in the CNS of Wlds mice after cortical lesions was therefore found [3]. Thus, it was confirmed laterally that the rate of degeneration (rate of axonal degeneration) is slower in the central nervous system than in the peripheral nervous system [4]. The discovery of the Wlds mouse became the opening of a new chapter in modern Wallerian degeneration, i.e., that axons, like the cell body, have genetically encoded active self-destructive capacity, and that the study of Wallerian degeneration had ushered in the move from passive to active research.

Since the concept of Wallerian degeneration was introduced, people have been working on its mechanism, i.e., in 1850, Augustus Waller roughly explained the phenomenon of WD as "neuronal cytosol trophoblastic axons". In 1850, Augustus Waller roughly explained the WD phenomenon as "neuronal cytotoxic axons". It was not until 1906 that Cajal proposed the "neuronal theory" to explain the WD phenomenon. Later, in 1948, Weiss and Hiscoe proposed axonal transport to explain this injury phenomenon [5]. However, Lassmann H et al. proposed a possible model of myelin degradation in 1978 by comparing ultrastructural results with histochemical and biochemical data. This meant that WD might be caused by myelin degradation. By 1981, Hofteig JH et al. built on the previous work by transecting the sciatic nerve of rabbits and demonstrated that nerve debris clearance was faster in young animals than in old animals, i.e., it was found that damage to peripheral nerves was more easily recovered in young animals, and here it introduced the concept of nerve debris clearance [6]. With the further development of ultrastructural studies, Ohara S. and Ikuta F. showed the early stages of Wallerian degeneration in 1985 by studying distal segments of the phrenic nerve in mice after crush injuries and found that axonal degeneration and myelin disintegration became evident from day 2 to day 6. In addition, most notably, they observed the detachment of neighboring endothelial cells accompanying macrophage invasion [7]. Soon after, two years later, in 1987, Röyttä M. and Salonen V. et al. investigated the changes in the endothelial nerve caused by Wallerian degeneration and regeneration in the rat sciatic nerve and demonstrated that endothelial fibroblasts form a rounded structure around the columns of Schwann cells, known as the Buengner band, which is conducive to axonal regeneration [8]. In the same year, Lindholm D et al. found that the level of IL-1 increased after axonal injury, which activated macrophages to increase the level of NGF-mRNA in sciatic nerve culture explants [9]. At this point, scientists knew about Wallerian degeneration at the molecular and cellular level. They believed that it happened because of axonal degeneration and myelin scavenging, which mostly involved Schwann cells and macrophages entering the nerve cells, but the exact way it happened was still not clear.

It was not until the discovery of the Wlds transgenic mouse (C57BL/6/Ola) that the understanding of Wallerian degeneration was accelerated, and a new era of Wallerian degeneration was ushered in. It was found that Wallerian degeneration of distal axons in this Wlds mouse is active and a host-given trait, and that its neural regeneration is not impeded by the presence of essentially intact axons in the distal stump and by the absence of recruitment cells, myelin debris, and mitotic divisions of the Schwann cells [2,10]. The further studies have since revealed that genes affecting Wallerian degeneration were localized on the distal end of mouse chromosome 4 [11]. And wlds was found to be a chimeric gene encoding a protein consisting of two major components: the first four N-terminal amino acids of the E70 ubiquitin ligase Ube4b and the nicotinamide adenine dinucleotide (NAD) synthesizing enzyme nicotinamide mononucleotide adenylyl transferase 1 (NMNAT1). Unlike wild-type axons, which degenerate rapidly upon sciatic nerve transection, it causes axons to exhibit delayed degeneration and remain intact for weeks after injury [12,13].

3. What is the mechanism of Wallerian degeneration?

As research has progressed, it has been found that in the peripheral nervous system (PNS) and central nervous system (CNS), this Wallerian degeneration manifests itself in two distinct outcomes. This is because extensive Wallerian degeneration is recognized as axonal degeneration and myelin clearance. Before explaining this phenomenon, we introduce the concepts of axons and myelin.

3.1. Axonal

Axons degenerate when nerves are injured, and Kerschensteiner et al. found that proximal and distal axon segments degenerated by hundreds of micrometers just 30 min after surgical spinal cord injury, but that acute degeneration affected proximal and distal axon terminals equally. This phenomenon of very early axonal degeneration is therefore referred to as "acute axonal degeneration (AAD)" [14]. Alternatively, chronic axonal degeneration is also known as Wallerian degeneration. Distal axonal damage occurs over the following 8–24 h, and beading or swelling of the axonal membrane can be observed, with CNS axonal swelling occurring long before the loss of axonal continuity in the distal stump or disintegration of cytoskeletal granules. We found that the CNS responds very early by forming axonal spheroids within 6 h of axotomy in vivo, yet in contrast to the PNS, only minimal morphological changes in injury occur up to 37–44 h after trauma [15]. The addition of calpain inhibitors, calcium chelation [16], and the ubiquitin-proteasome system (UPS) did not prevent the disintegration of the axial membrane, and these studies suggest that the axial membrane is disrupted by a currently unknown calcium-independent mechanism after injury. With further growth in time, axonal cytoskeletal granule disintegration (GDC) occurred, which is a process by which microtubules, neurofilaments, and other cytoskeletal components are disassembled. In 1996, it was shown that increased assembly of microfilaments and microtubules provides axons with cytoskeletal components, with microtubulin and actin being the major proteins carried by axonal chronic transit (SCb), and that the resulting actin and loss of microtubulin subunits may play a role in the acceleration of SCb [17]. Subsequently, in a further study in 2002, it was learned that actin can polymerize, microtubulin does not, and that axon regeneration utilizes β-actin. And, in two consecutive axotomies, the accelerated axon growth seen after the second was associated with the ability of actin, already in transport, to polymerize in response to the first axotomy [18]. We also found two mechanisms involved in this process in GDC: a ubiquitin-proteasome system (UPS)-dependent process involving microtubule dismantling. The UPS is a multistep protein degradation mechanism, and in the nervous system, the UPS regulates the remodeling and degradation of neuronal processes and is implicated in WD(19); and a calpain-dependent process concerning the degradation of neurofilaments [20]. Calpain is responsible for the morphological degeneration of axons and synapses during WD [21]. It is known that UPS activation can lead to microtubule depolymerization and subsequent neurofilament degradation, which may also be related to Ca2+-dependent calpain [19]. However, some studies have also taken the opposite view, suggesting that m-calpain proteins are lost very early after axonal injury and may reflect activation and degradation of this protein long before cytoskeletal degradation. Calpain activation may be an early event in the protein hydrolysis cascade reaction, triggered by axonal injury and eventually accompanied by axonal degeneration [22]. It has also been found that increased intracytoplasmic Ca and K ions also prolong axonal degeneration [23], while Ca storage-mediated intra-axonal Ca release leads to secondary WD [24]. However, axonal GDC is also a critical step in Wallerian degeneration: in Wlds mice, myelin degeneration and macrophage influx begin only after axonal GCD(2), and thus axonal loss triggers subsequent events in WD. It has been shown that there is a lag period (one week in humans) between axonal degeneration and GCD [25], and it is hypothesized that if interventions were made during this lag period, it may be possible to reverse the emergence of WD in axons.

3.2. Myelin

• (1)OVERVIEW: Myelin is a lipid-protein complex derived from Schwann cells (SC) (in the PNS) and oligodendrocytes (in the CNS), a fatty membrane that wraps around axons, also known as myelin, and forms a sheath that supports and protects individual axons [26], which provides nutrition while allowing for the rapid conduction of action potentials. Myelin-providing (pre-myelinating) SCs are generated from precursors that undergo three developmental transitions, including the transition from cristae to SC precursors (SCPs), from SCPs to immature SCs, and from immature SCs to either pre-myelinating SCs or non-myelinating (Remak) SCs. The pre-myelinating SCs are then "radially sorted" (SCs only myelinate axons larger than 1 μm in diameter, a process known as "radial sorting"), and some of these cells go on to produce myelin [27].
• (2)Demyelination: Demyelination is an inflammatory response to peripheral nerve injury. Demyelination of injured peripheral nerves usually occurs through Wallerian degeneration (WD), which is an inflammatory response to injury to peripheral nerves. (WD), which both negatively regulates myelin formation and leads to myelin degradation. In peripheral nerves, this response accelerates myelin debris removal and nerve regeneration [28]. It has been found that the degenerative breakdown of myelin (i.e., demyelination) precedes the degeneration of the axonal cytoskeleton [29]. In addition, increased persistent sodium current is an ultra-early change in WD faster than conduction failure and axonal disintegration of action potentials, and thus sodium channel blockers may be considered as a therapeutic option [30].
• (3)Myelin formation: ① The c-Jun transcription factor c-Jun is located at the core of the AP-1 complex and is involved in myelin formation. ② Sox2 Sox2 transcription factor regulates the activity of Schwann cells [31].

3.3. Different causes affecting WD in the PNS and CNS

In PNS and CNS, Wallerian degeneration exhibits two distinct results, i.e., axons can be regenerated in PNS, but they are difficult to regenerate in CNS, and the reasons for this can be summarized as follows (Fig. 1).

Fig. 1.

Schematic illustration shows the ongoing process after Wallerian degeneration, which involves major cellular and inflammatory factors and promotes restoration of nerve integrity. (A) The nerve cell bodies are located in the central and peripheral areas, and the axons of the nerve are wrapped by myelin sheath, which can isolate and project signals to the distal axons. (B) Nerve injury disrupts the integrity of axons and induces Wallerian degeneration, with initially swollen, beaded axon cells. (C) disintegration and demyelination of cytoskeletal granules over time. (D) in the central nervous system, mainly involving the recruitment of oligodendrocytes (E) in the peripheral nervous system, mainly involving the recruitment of Schwann cells and Macrophages, driven by MCP-1 and IL-1β, are more likely to recruit blood vessels to the injury site to clear the damaged myelin sheath (F–G). These cells interact to promote axon regeneration and ultimately myelination.

3.3.1. The content of membrane proteins is different

Proteolipid protein (PLP) is a highly hydrophobic tetra-spanning protein that is the main protein constituting myelin in the CNS, accounting for 38 % of the total protein amount of myelin [32]. In contrast, the amount of PLP in the PNS is small [33,34]. The PLP1 gene encodes human PLP, which is expressed in oligodendrocytes, and by analyzing rodent and human brains, some scholars have found that axonal degeneration occurs in both humans and mice lacking the major myelin protein PLP1 [35], so that reduced expression of myelin-related proteins may also lead to rapid demyelination.

3.3.2. The environment of the extracellular matrix is different

The existence of an extracellular matrix (ECM) in the central nervous system (CNS) was first generally recognized in 1971 [36,37].CNS axons are not incapable of regeneration but have a low regenerative capacity relative to the PNS. In some experiments, when the axons of CNS were put into the environment of PNS, the axon regeneration ability increased, proving that the environment of CNS is inhibitory and non-permissive [38]. It has been found that in the early stages of axonal injury, in the CNS, myelin debris from oligodendrocytes and myelin-associated inhibitors can limit axonal regeneration [39,40]. There are several known myelin-related inhibitors of axonal regeneration, including (i) phospholipase A2 PLA [2]: hydrolyzes phosphatidylcholine to Lys phosphatidylcholine and arachidonic acid. The former induces myelinolysis, and the latter stimulates inflammatory responses via arachidonic acid. It has been found that blockade of phospholipase A6 (cPLA(6)), which is made from a natural mutation of phospholipase A2 (sPLA(2)), in the transected sciatic nerve of C2BL/57 mice resulted in significant slowing down of myelin, axonal degradation, and phagocytosis of distal neural segments [41]. These results demonstrate the inhibitory effect of PLA(2) on myelin degradation. (ii) Nogo-A: Nogo-A is one of several neurite growth inhibitory components present in oligodendrocytes and myelin membranes of the central nervous system. Recent studies have shown that in vivo application of Nogo neutralizing antibodies, peptides blocking the Nogo receptor subunit NgR, or blockers of the post-receptor fractions Rho-A and ROCK, in the spinal cords of rats and mice, could induce long-distance axonal regeneration and compensatory germination while significantly enhancing functional recovery [42]. (iii) Myelin-associated protein MAP: In diseased PNS, axonal regeneration and functional recovery can be successfully performed after WD. However, there is no significant axonal regeneration in the diseased mammalian CNS. Myelin-associated protein (MAP) has been shown to play an important role in preventing axonal regeneration in the CNS [43]. (ⅳ) Hepcidin-A4: A study has shown that axonal hepcidin-A1 inhibits myelin formation through the use of axon-mimicking microfibers and a zebrafish model system, further validating that activation of the EphA1-RhoA pathway in oligodendrocytes by axonal hepcidin-A4 inhibits the stabilizing axonemal glia interactions required for myelin production [44]. (ⅴ) NgR1: NgR1 is a glycosylphosphatidylinositol (GPI)-linked membrane receptor that lacks transmembrane or cytoplasmic structural domains. Regeneration of corticospinal tracts and axons was not observed in transgenic mice knocked down for NgR1 after SCI [45]. Thus, NgR1 also inhibited axonal growth.

3.3.3. Cellular differences

The cells that predominantly influence myelin regeneration in the peripheral nervous system are Schwann cells and recruited macrophages; however, the slowed rate of myelin regeneration that occurs in the CNS of senescent animals is attributed to, on the one hand, delayed differentiation of oligodendrocyte precursor cells (OPCs) [46], and on the other hand, to the poor macrophage response and delayed clearance of myelin debris [47].

3.3.3.1. Oligodendrocytes

Oligodendrocytes can survive for long periods in the absence of axons, albeit in an inactive state, and then they are capable of functional regeneration when brought into contact with unmyelinated axons [48]. The survival of the majority of oligodendrocytes after WD suggests a possible reserve capacity for repair after CNS injury [49]. Oligodendrocyte progenitor cells have been found to contribute to demyelination, and recurrent episodes of demyelination may lead to the depletion of their progenitor cells, which may fail myelin regeneration [50]. One study, which caused Wallerian degeneration of the optic nerve in immature rats by removing one of their eyes, found that changes in myelin gene expression were not limited to the encoding proteolipid protein (PLP) mRNA, as the steady-state levels of myelin basic protein (MBP) mRNA parallel those of PLP mRNA in the developing brain and degenerating optic nerve. Thus, oligodendrocytes require axons to maintain their normal levels of PLP and MBP transcripts, as well as a high proportion of distally initiated PLP transcripts that characterize early myelin formation [51].

3.3.3.2. Sherwan cells (SC)

Schwann cells (SC) in the PNS play a major role in myelin debris removal and axon elongation. SC is divided into myelinated SC (i.e., mature SC) and non-myelinated SC. Mature SC is used to myelinate larger diameter fibers, whereas non-myelinated SC (also referred to as Remak SC) is wrapped around small sensory and autonomic axons (a single Schwann cell either wraps multiple small, unmyelinated axons to form a Myelinized axons are either wrapped around multiple small unmyelinated axons to form a Remak bundle [52]. The effects of SC are mainly: (a) Inhibition of myelination: SC reduces myelin lipid synthesis during the first 12 h after axotomy and stops myelin protein production within 48 h [53]. (b) Promotion of axonal regeneration: as early as the first week after neurectomy, Shewan's cells (SC) proliferate and form the so-called Bungner's band, which induces regeneration by connecting dissected nerve stumps [54]. The myelin sheath breaks down into characteristic myelin ovoid bodies. The SC actively retracts the cytoplasm from the myelin sheath and down-regulates the mRNA synthesis of myelin proteins [55]. (c) Recruitment of macrophages: However, SCs produce a variety of cytokines that recruit blood-derived monocytes through a paracrine signaling cascade to help remove degraded myelin. (d) Differentiation of autophagic SCs: Autophagy is an important pathway for the processing and eventual recycling of disordered proteins and discarded cellular structures, and is essential for maintaining cellular stability. It has been found that autophagic SC is required to facilitate the early stages of myelin removal, myelin regeneration, and scar reduction after peripheral nerve injury (PNI), and thus autophagy is critical for the timely removal of myelin after PNI [31].

3.3.3.3. Macrophages

Monocytes migrate in the peripheral blood to various organs or tissue systems and eventually differentiate into resident tissue macrophages. The resident macrophage population does not represent a constant population but is permanently exchanged by newly recruited blood monocytes, which form a stable population of cells in the central nervous system [36]. The predominant role of macrophages is phagocytosis, and many surfaces do not function through receptors while others become form chaperones [54]. Macrophages are involved in axonal regeneration in both peripheral nerves and the CNS, and it is clear that the results are not the same in the two, so we analyzed the reasons for the differences in macrophage recruitment in the PNS and the CNS as follows: (i) Different sites of presence: One of the main features of WD in the PNS is the strong gathering of recruited macrophages along the entire distal portion of the nerves. This is different from what happens after a spinal cord injury, where macrophage endocytosis mostly happens at the injury site. These immune cells may promote axonal regeneration in different ways, but mainly through phathegocytosis of degenerating myelin sheaths and axonal fragments. However, during CNS inflammation, macrophage responses are associated with deleterious reactions [56]. In addition, it has been further investigated that MCP-1, MIP-1α, and IL-1β are important regulators of macrophage responses, which can lead to rapid breakdown and clearance of myelin during WD [57]. (ii) Limitations of the blood-brain barrier (BBB): studies have shown that although resident macrophages still exert weak clearance, blood-borne macrophages constitute the pool of cells responsible for removing myelin and that CR3 mediates the myelin-clearing capacity of non-resident macrophages [58]. In contrast to the delayed macrophage response in the CNS [59], due to less restriction of the blood-brain barrier (BBB), in the PNS, blood-derived macrophages rapidly and in large numbers enter the damaged neural site after the injury caused by the rupture of the microvessel and remove myelin debris after WD of the distal ganglion occurs, thus favoring axonal regeneration. However, CNS lesions with blood-brain barrier disruption in the CNS: such as cerebral ischemia, brain abscesses, and stab wounds, cause rapid activation of microglia and concomitant blood-borne macrophage recruitment leading to debris clearance [60], which also laterally validates the limiting role of the blood-brain barrier in myelin clearance in the CNS. In addition, it has also been demonstrated for the first time that resident neuroendothelial macrophages (blood-borne macrophages) are quantitatively important in the rapid response to nerve injury, i.e., the greater the number exerts a clearing effect, and that localized macrophages (resident macrophages) contribute significantly to the total pool of neuroendothelial macrophages (blood-borne macrophages) during WD [61]. (iii) Complement involvement: the PNS complement is involved in macrophage recruitment and myelin clearance in the PNS [62], thus favoring axonal regeneration in the PNS. (iv) Interaction with other SC cells: macrophages can be functionally polarized to the M1 or M2 phenotype [63]. In an animal model of facial nerve injury in the PNS, the regulation of macrophage polarity and anti-inflammatory M2 macrophage polarity is also associated with monocyte chemotactic protein-1 expression and secretion by SCs [64]. In turn, In turn, M2-polarized macrophages that are recruited help the growth of SC bridges, which improves regeneration and functional recovery in the long run. In addition, it has been shown that anti-inflammatory M2 macrophages secrete microvesicles carrying a specific microRNA, miR-223, which is associated with altered proliferation and migration capacity of SCs, and thus macrophage-SC interactions in the PNS also facilitate regeneration of their axons [65].

3.3.3.4. Fibroblasts

Their role in nerve injury, regeneration, and SC function has not been extensively studied. After transection injury, fibroblasts are not only damaged but, together with dermal cells, also respond to the injury to initially remove tissue debris [65].

3.3.4. Nerve scarring and astrocytes

Over time, the recruitment of inflammatory cells and reactive astrocytes leads to neuroglial scar formation, usually accompanied by fluid-filled cysts [66]. This scarring process is associated with the extracellular matrix molecule chondroitin sulfate proteoglycan (CSPG), which is associated with increased release of the extracellular matrix molecule chondroitin sulfate proteoglycan (CSPG), further limiting regeneration. CSPG is strongly inhibitory to axon regeneration and is a major component of the glial scar tissue, and the slow decrease in CNS myelin sheaths and prolonged deposition of dense astrocyte scarring may present an environment that is unsupportive of axon regeneration [67]. Traditionally, however, this neuroglial scar was thought to be a simple mechanical barrier (Windle and Chambers, 1950), and later studies have shown that regeneration still fails even without the formation of dense neuroglial scarring. Multiple experiments later demonstrated that the molecular composition of the scar and the production of inhibitory molecules by astrocytes were contributing factors to regeneration failure [38]. In summary, these molecular inhibitors of the CNS glial environment provide an unfavorable environment for axonal repair, and myelin removal inhibits glial scar formation, providing an inhibitory extracellular matrix environment for the CNS [68]. Reactive astrocytes and oligodendrocyte precursor cells (OPCs) have been shown to up-regulate their expression of inhibitory CSPGs associated with scarring; thus, altering the inhibitory factors of the extracellular matrix leads to axon regeneration in the CNS [69].

4. Diseases and clinics of Wallerian degeneration

Through the collection of previous clinical cases, we found that the diseases that exhibit Wallerian degeneration on NMR include cerebral infarction, cerebral hemorrhage, brain tumors, traumatic brain injury, large arachnoid cysts, and cerebral cortical dysplasia [70], among which, in a retrospective study, we found that multiple sclerosis and encephalitis can also present with Wallerian degeneration [71]. Therefore, we can hypothesize that WD is a secondary pathological change, and such factors affecting axonal degeneration and myelin clearance capacity may cause WD. In addition, by studying the so-called abnormal pyramidal tracts in 150 consecutive autopsies of the human brain (APT) in 150 consecutive autopsies of human brains and found Wallerian degeneration in the APT ipsilateral to the brain lesion in 63 cases. These findings confirmed that the APT is a normal bundle of descending fibers that is part of the pyramidal tract [72].

In this paper, we focus on the case of cerebral infarction as an example, and from the known reports, we found that the WD that occurs after this type of cerebral infarction mainly involves the corticospinal tract (CST) [73,74], the pontine cerebellar tract [75], and dentato-red nucleus-inferior olivary nucleus pathway (dentato-rubral-thalamic-(cortical) bundle) [76], striatal fibers in the substantia nigra of the midbrain [77], corpus callosum [78,79], limbic system, and the spinal cord, and thus WD is associated with clinical neurological dysfunction, with varying degrees of cognitive psychiatric deficits, limbic motor deficits, and ataxia, which may occur. Subsequently, a series of retrospective clinical studies found that the time, size, and location of cerebral infarction directly influenced the degree of Wallerian degeneration [[80], [81], [82], [83]], i.e., the faster the WD appeared after the infarction, the larger the infarct area rather than the volume, and the closer the infarct was to the cortex, the more severe the degree of WD was. Subsequently, the degree of Wallerian degeneration was found to affect the prognosis of patients, and the severity of WD was associated with cognitive [71,75] and motor deficits [84,85], i.e., the greater the degree of WD, the worse the prognosis of limb motor function and related signs, but some reports illustrate that the greater the degree of WD, the slower the recovery of functional recovery of simple motor hemiparesis, does not limit the final recovery outcome of patients [86].

5. Development of WD in imaging

5.1. CT

In 1983, Stovring J studied the pathophysiology of WD of the corticospinal tracts of the brainstem in patients with old hemispheric infarcts by CT and found that there was a clear relationship between the location and size of the infarct and the presence of WD. That is, when most of the motor cortex was involved, degeneration of WD could accumulate at the level of the midbrain and even the pons, but WD was not seen in patients with small infarcts in the cortex [82].The following figure shows us the typical appearance of WD on CT [ Fig. 2].

Fig. 2.

This is a CT of a case of WD following left temporal lobe infarction, which shows varying degrees of atrophy in the cerebral peduncles of the midbrain.

5.2. MRI

In 1987, Cobb SR et al. confirmed by magnetic resonance (MR) imaging the first finding of WD in corticospinal tracts in patients with Schilder's disease. The histochemical stages of myelin breakdown were summarized, and these processes could be verified by MRI [87]. Subsequently, it has also been found that MRI seems to be an effective method for early detection of WD [88, 89] and that in the adult CNS, T2-weighted images are unchanged for the first 4 weeks, a period referred to as stage 1 Wallerian degeneration. A study compared MR images and histologic images of people who had spinal cord injuries. The histologic images showed early Wallerian degeneration, which is marked by myelin and axonal rupture, starting at a time when T2-weighted MR images were normal. At this point, however, degeneration is only showing up physically in the axon. The myelin sheath hasn't changed much biochemically, so regular MR images don't show any problems with the signal intensity. Even though the myelin sheath breaks down into ellipsoids and spheres, the fragmented sheath retains the staining properties of myelin [90]. From 4 to 14 weeks, myelin proteolysis happens without lipolysis. This changes the protein-lipid ratio and causes T2-weighted images to show less signal intensity. This is called phase 2. Subsequently, increased edema and further lipolysis lead to an increase in T2-weighted signal intensity; this is described as phase 3. Finally, months to years later, volume loss occurs due to atrophy in phase 4(88). Studies in adult patients have shown that Wallerian degeneration detected by T2-weighted MR imaging is strongly associated with the presence of persistent dysfunction and therefore has prognostic value [71]. The following three cases show that the appearance of WD on MRI is formed along the distribution of corticospinal tracts. [Fig. 3, Fig. 4, Fig. 5].

Fig. 3.

The above MRI shows WD after left periventricular infarction. MRI shows hypointense on T1, hyperintensity on T2 and FLIAR, with varying degrees of atrophy in the cerebral peduncle.

Fig. 4.

The above MRI shows WD occurring after infarction in the right basal ganglia and periventricular, still showing hypointense on T1 and hyperintensity on T2 and FLIAR, with simultaneous atrophy involving the pons and the midbrain and cerebral peduncle.

Fig. 5.

The MRI above shows WD that developed after a large infarct in the left temporal lobe, still showing hypointense on T1 sequences and hyperintensity on T2 and FLIAR sequences and involving simultaneous atrophy of the pons and midbrain.

It was not until 1999 that Castillo M, by performing DWI in 72 patients with cerebral infarction within 1 h of symptom onset, found that the signal intensity at the location of corticospinal tracts with Wallerian degeneration could be predicted by DWI [91]. Later in 2010, DeVetten G. et al. found that in children with stroke, poor motor outcome was associated with early corticospinal tracts visible on diffusion-weighted (DWI) Wallerian degeneration [92]. He therefore hypothesized that post-stroke adults with corticospinal tracts (CST) also undergo early diffusion changes after stroke and that these lesions are associated with a poor prognosis. As a result, he found that apparent diffusion coefficient (ADC) measurements were much more accurate than other MRI sequences in detecting CST degeneration. ADC decreased in a time-dependent manner in patients with poor motor outcomes, but not in patients with better outcomes [92]. In studies of axons, information on the structural integrity of axonal white matter was provided by measuring the effect of myelin and axonal membrane integrity on the local diffusion properties of water [93]. The diffusion of water molecules is limited by cellular structures that provide a free movement of water molecules, thus exhibiting orientation dependence. White matter in the brain and spinal cord is highly restricted in diffusion due to its structural components, such as axons and myelin [94]. Therefore, in the previous work, some scholars have added four stages of WD [95], the first stage being characterized by the physical disintegration of axons and myelin sheaths. This acute white matter bundle injury is associated with an energy-dependent cessation of axoplasmic transport and cytotoxic edema. DWI identifies the earliest stage of WD as a high signal in the context of a low signal on the ADC map. However, conventional imaging will be normal at this stage. The second stage still occurs 4–14 weeks after stroke and is characterized by rapid destruction of myelin. It is seen as a low signal on T2 and proton density images. Glial proliferation occurs in stage 3, which gives a high signal on T2 and fluid attenuation inversion recovery (FLAIR) sequences. Stage 4 or the terminal stage occurs after several years and shows volume loss due to atrophy of white matter bundles. The following figure illustrates the WD appearance on DWI [Fig. 6].

Fig. 6.

The above MRI showed that WD occurred after left periventricular cortical infarction, MRI showed low signal on DWI and high signal on ADC, and at the same time, varying degrees of atrophy of the cerebral peduncle were also involved.

5.3. MRS

In 2006, Kubas B. et al. discovered that after cerebral ischemia, there were no visible problems on a standard MRI, even though Wallerian degeneration was killing neurons in intact pyramidal tracts. Conventional MRI is not an acute or subacute period Wallerian degeneration because it is unchanged during the first week, so he assessed the appearance by resonance spectroscopy (MRS). (MRS) to assess metabolite changes in normal-appearing corticospinal tracts and cortical screen cluster bundles to better quantify the extent or severity of fiber damage [96].

5.4. DTI

However, there is now growing evidence that noninvasive DTI assays can quantify the microstructural changes in the pyramidal bundles after an ischemic event. The CST originates in the cerebral cortex, and the vertebral fibers converge via the radial crown in the posterior limb of the internal capsule and the cerebral peduncle, cross over in the medulla, and finally terminate in the gray matter of the spinal cord [97]. The DTI sequences provide a clear visualization of the changes associated with the CST. The cone bundle is also visualized to build out the intracerebral cone bundle pathway and obtain relevant data on the whole cone bundle pathway [98]. It has been shown that water diffusion parallel to the pyramidal bundles is faster than perpendicular diffusion, which leads to anisotropic diffusion [99], and this anisotropic diffusion is related to the integrity of the pyramidal bundles, the degree of restriction on water diffusion by Ca2+, axonal membranes, and different tissues. By calculating the anisotropy fraction (fractional anisotropy, FA) as a scalar of diffusion direction dependence, data on water molecules in different diffusion directions can be obtained quantitatively. It has been shown that in the acute phase after stroke, a decrease in FA of CST can be detected, but it has less predictive value for the 3-month motor outcome. In contrast, in the subacute or chronic phase after stroke, significantly reduced FA in affected CST was associated with greater motor deficits and poorer motor outcomes at 3 and 6 months [100,101]. It has also been stated that the recovery of motor function is mainly related to the number of CSTs involved in the cerebral infarction lesion: the greater the number of CST cone bundles involved, the worse the recovery of motor function; and it is not related to the size of the cerebral infarction [102,103].

6. Summary and outlook

This review provides us with a systematic account of the mechanisms, possible diseases, and imaging findings of Wallerian degeneration. Since WD was first reported, our understanding of WD has been incomplete. With the study of the pathophysiological mechanism of WD and molecular changes in related pathways, the different processes of WD in the PNS and CNS were gradually recognized, and the influencing factors, i.e., whether it may be caused by axonal degeneration or reduced myelin clearance ability, were further analyzed. At the same time, the discovery of Wlds in transgenic mice broke the stereotype of WD as a passive process and led to the belief that there might be some active program. After that, from the molecular level to the clinical research, the retrospective study of WD patients found that the clinical manifestations and prognosis were related to the course of WD in the central nervous system. Although WD lacks characteristics on MRI, the lesions have characteristic distribution characteristics. Based on primary brain injury, we should pay attention to the observation of the degeneration along the direction of the nerve fiber bundles, and through the observation of the relationship between primary brain injury and secondary degeneration, we can see the relationship between primary brain injury and secondary degeneration. By observing the relationship between primary brain injury and secondary degeneration, combined with the corresponding MRI manifestations and clinical manifestations, it is easier to identify other lesions. Although WD may differ in the time course of these changes according to the nature of different nerve bundles, the sequence of MRI changes and the alteration of signal characteristics are basically the same. More and more new imaging techniques have been shown to not only show early pyramidal tract degeneration in WD but also predict how well patients will recover their motor function over time. This could help doctors treat and intervene with WD patients more quickly in the future.

CRediT authorship contribution statement

Ruiqi Tian: Writing – review & editing, Writing – original draft. Yingying Zhou: Data curation. Yuan Ren: Investigation. Yisen Zhang: Conceptualization. Wei Tang: Writing – review & editing.

Data availability statement

Applicable.

Statement of ethics

Patient data such as imaging images presented in this chapter were ethically reviewed and informed consent was obtained from the patients.

Funding

Not applicable.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.