Structural and Cellular Changes

Since the initial description by Waller (1850), structural and cellular changes and the time course of such changes that occur during wallerian degeneration have been described in increasing detail, especially in the PNS (Vial, 1958; Ohmi, 1961; Webster, 1962; Lee, 1963; Cravioto, 1969; Simon et al., 1969). For a short period after transection, most of the distal stump appears nearly normal. The segment just distal to the site of axotomy is an exception. Immediately after axotomy, there is a progressive accumulation of intra-axonal organelles at the proximal end of the disconnected nerve fibers (Zelena et al., 1968; Griffin et al., 1977) (Fig. 1). These organelles form a caplike pellet that increases in volume linearly with time until axonal breakdown supervenes (Zelena et al., 1968). This pellet results from continued retrograde transport of vesicular organelles, as well as some mitochondria, dense

Figure I Early structural changes during wwallerian degeneration (A) Twelve hours after transection of the sciatic nerve of an adult rat shows one of the earliest structural changes that take place during wwallerian degeneration, accumulation or organelles, including mitochondria (arrows) (original magnification x 10,000.) (B) Within 24 hours of the transection, granular disintegration of the cytoskeleton takes place. Notice that in panel A and B, the axons are swollen as evidenced by the decrease in g-ratio (ratio of the myelin thickness to the whole fiber diameter) (original magnification x 10,000). (C) At 7 days after transection of the nerve, Schwann cell proliferation, macrophage infiltration, and myelin degradation are all well underway (original magnification x 2,000).

Figure I Early structural changes during wwallerian degeneration (A) Twelve hours after transection of the sciatic nerve of an adult rat shows one of the earliest structural changes that take place during wwallerian degeneration, accumulation or organelles, including mitochondria (arrows) (original magnification x 10,000.) (B) Within 24 hours of the transection, granular disintegration of the cytoskeleton takes place. Notice that in panel A and B, the axons are swollen as evidenced by the decrease in g-ratio (ratio of the myelin thickness to the whole fiber diameter) (original magnification x 10,000). (C) At 7 days after transection of the nerve, Schwann cell proliferation, macrophage infiltration, and myelin degradation are all well underway (original magnification x 2,000).

bodies, and multivesicular bodies, producing a focal accumulation at the site of axonal interruption (Zelena et al., 1968; Ranish and Ochs, 1972). During this initial stage, there are only subtle changes elsewhere along the fibers. There is a tendency for particulate organelles to accumulate beneath nodes of Ranvier and Schmidt-Lantermann incisures (Ballin and Thomas, 1969), an exaggeration of their normal enrichment in those sites and presumably indicative of some early changes in fast bidirectional transport. However, on the whole, fast transport continues at normal rates and roughly normal abundance (Ranish and Ochs, 1972; Lubinska, 1977);

even uptake of exogenous materials at the synaptic terminals persists, and electrical conduction with synaptic transmission can be elicited normally with stimulation of the distal stump (Miledi and Slater, 1970). The duration of this initial stage of wallerian degeneration varies with the species, the length of distal stump (longer with a longer stump [Miledi and Slater, 1970]), the temperature (longer in cooler distal stump [Gamble and Jha, 1958; Sea et al., 1995; Tsao et al., 1999]), and possibly with the nature of the nerve (motor vs. sensory) and the location (PNS vs. CNS) (Pesini et al., 1999) (reviewed in detail by Chaudhry and colleagues, 1992).

This brief initial stage of wallerian degeneration is followed by breakdown of the axonal cytoskeleton. Granular disintegration of the cytoskeleton describes the abrupt conversion of the axoplasm into fine particulate and amorphous debris (George and Griffin, 1994), representing cleavage products of the cytoskeleton. This change must occur in an explosive fashion at any single level of the axon, as indicated by the rarity of finding partial stages in the process. In transverse sections, each fiber either appears normal or has complete granular disintegration, suggesting a virtual "all-or-none" change without persistent intermediate stages. In the PNS, the disintegration of particulate axonal organelles and axolemma is essentially simultaneous with the breakdown of the cytoskeleton. However, in the CNS, remaining axolemma and often mitochondria frequently are found in axons that have undergone recent disintegration of their cytoskeleton (Franson and Ronnevi, 1984; George and Griffin, 1994). Although this stage appears to last only for a few hours, it serves to emphasize the early structural changes in the cytoskeleton.

The direction of spread of granular disintegration of the cytoskeleton has been a controversial issue. In the garfish olfactory nerve, the process of axonal degeneration is extremely slow, and it clearly shows a centrifugal spread (Cancalon, 1982, 1983). In mammalian axons, the axonal breakdown is much more rapid, so that the spatiotemporal sequence has been debated. Some studies suggested that degeneration is virtually synchronous all along the distal stump or that degeneration starts distally and progresses proximally with time (Lunn et al., 1990). The benchmark analysis, however, remains the rigorous quantitative study of degeneration in the rat phrenic nerve, which showed a clear centrifugal spread of early changes of wallerian degeneration (Lubinska, 1977). The phrenic nerve remains at core temperature, thereby reducing the possibility of a proximal-distal temperature gradient that might affect the results. Another system with similar advantages is the degeneration of central projections of dorsal root ganglion sensory neurons after dorsal rhizotomy close to the cell body. In this system, spread of granular disintegration of the cytoskeleton starts at the site of transection and proceeds centrifugally (away from the cell body) at a rate of about 3 mm/hr (George and Griffin, 1994).

Several cellular changes follow promptly after the breakdown of the axon, including (in the PNS) opening of the blood-nerve barrier, degranulation of mast cells, recruitment of circulating macrophages, and early changes in the Schwann cells. The sequence of vascular and glial changes and the cell-cell interactions underlying these changes are only partially understood.

One of the first changes after granular disintegration of the cytoskeleton is opening of the blood-nerve barrier in the PNS. In the transected rat sciatic nerve, this immediately follows the breakdown of the cytoskeleton, occurring 24 to

48 hours after the injury (Powell et al., 1980; Bouldin et al., 1990). Of course, at the site of the axotomy, there is a prompt breakdown (Bouldin et al., 1991), but the barrier appears to remain intact along the rest of the distal stump. Shortly after dissolution of axoplasm, there is breakdown of the blood-nerve barrier, which can be triggered by degeneration of even a small number of fibers. In the rat sciatic nerve, restoration of the blood-nerve barrier occurs only after successful regeneration of axons through the segment (Bouldin et al., 1991). However, in the frog sciatic nerve, the blood-nerve barrier is at least partially restored even in the absence of successful regeneration (Weerasuriya, 1990). The basis for breakdown of the blood-nerve barrier is unknown; the effect of degranulation of mast cells (Powell et al., 1980) and cytokines (Griffin et al., 1992) have been proposed but not directly demonstrated experimentally. The CNS responds differently. There is less breakdown of the blood-tissue-barrier along the degenerating dorsal column axonal tracts than in the PNS (George and Griffin, 1994). This cannot be due to intrinsic differences between the central and peripheral axons; the same fibers are involved in the dorsal root and dorsal column when the central branches of dorsal root ganglion neurons are transected at the dorsal root level. This is likely to be due to differences in the intrinsic differences between the glial and endothelial cells of the CNS and the PNS.

0 0

Post a comment