degen - l REVIEW Degeneration, Trophic Interactions, and...

Info iconThis preview shows pages 1–19. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 16
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 18
Background image of page 19
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: l REVIEW Degeneration, Trophic Interactions, and Repair of Severed Axons: A Reconsideration of Some Common Assumptions GEORGE D. BITTNER, TIMOTHY SCHALLERT, and JEAN D. PEDUZZI We suggest that several interrelated properties of severed axons (degeneration, trophic dependencies, ini- tial repair, and eventual repair) differ in important ways from commonly held assumptions about those prop- erties. Specifically, (1) axotomy does not necessarily produce rapid degeneration of distal axonal segments because (2) the trophic maintenance of nerve axons does not necessarily depend entirely on proteins transported from the perikaryon—but instead axonal proteins can be trophically maintained by slowing their degradation and/or by acquiring new proteins via axonal synthesis or transfer from adjacent cells (e.g., glia). (3) The initial repair of severed distal or proximal segments occurs by barriers (seals) formed amid accumu- lations of vesicles and/or myelin delaminations induced by calcium influx at cut axonal ends—rather than by collapse and fusion of cut axolemmal leaflets. (4) The eventual repair of severed mammalian CNS axons does not necessarily have to occur by neuritic outgrowths, which slowly extend from cut proximal ends to possibly reestablish lost functions weeks to years after axotomy—but instead complete repair can be induced within minutes by polyethylene glycol to rejoin (fuse) the cut ends of surviving proximal and distal stumps. Strategies to repair CNS lesions based on fusion techniques combined with rehabilitative training and induced axonal outgrowth may soon provide therapies that can at least partially restore lost CNS functions. KEY WORDS Axotomy, Axonal degeneration, Axonal regeneration, Axonal proteins, Spinal cord lesions, Recovery of function, Nerve plasticity Many neuroscientists share an interrelated set of four assumptions (Box 1) that axotomy of PNS or CNS ax- ons produces (1) rapid degeneration of severed distal segments that (2) receive their trophic support (e.g., pro- teins) entirely from their perikaryon. Severed axons (3) initiate repair by collapse and fusion of axolemmal leaf- lets to seal their cut ends and may (4) eventually com- plete repair by the slow growth of neurites, which sometimes reestablish synaptic connections (1—4). How- ever, one problem with accepting as dogma this set of commonly held assumptions has been immortalized in the title of a song by Gershwin: “It Ain’t Necessarily So.” This review describes'an alternate set of four interre- lated assumptions that more completely account for the School of Biological Sciences (Neurobiology Section) and Institute of Neuroscience (GB and TS), Department of Pyschology (TS), The Uni- versity of Texas at Austin, Austin, Texas. School of Optometry, De- partment of Physiological Optics, Injury Control and Vision Science Research Centers, University of Alabama at Birmingham, Birming- ham, Alabama. (JP) Supported by NIH grant # NS31256 and ATP grant # 446 to GDB, NIH grant #NSZ3979 to TS, and a contract from the Spinal Cord So- ciety to JDP. Address reprint requests to: Dr. George D. Bittner, Department of Zoology, The University of Texas, Austin, TX 78712 (e-mail: bittner@mail.utexas.edu). 88 THE NEUROSCIENTIST Copyright © 2000 Sage Publications, Inc. ISSN 1073-8584 available data on degeneration, trophic interactions, and repair of severed axons in many invertebrates and verte- brates (including mammals), compared to the set of four commonly held assumptions (Box 1). In brief, 1. Axotomy does not necessarily produce rapid degeneration of severed distal segments. Severed distal segments naturally exhibit long-term sur- vival for weeks to years in invertebrates (15, 16, 25—30) and lower vertebrates (17, 31, 32)—-and can be induced to survive for weeks in mammals (33—36). 2. Axotomy does not necessarily remove all major sources of trophic support from the distal seg- ment, which can have a major trophic dependence on proteins received from various combinations of adjacent glia, neurons, or axonal protein synthesis (14—16, 25, 3740). (Slow turnover of existing axonal proteins may also be important [41—46].) 3. Initial repair of severed proximal (or distal) seg- ments does not occur by collapse and fusion of axolemmal leaflets, but rather by a seal formed by injury-induced accumulations of membranous structures (6—9, 47, 48) (e.g., endocytotic vesicles or myelin delaminations). 4. Eventual repair—natural or induced—of PNS or CNS axons does not necessarily occur just by extension of slowly growing neurites, but it also Degeneration, Trophic Interactions, and Repair of Severed Axons Box 1: Some Common Assumptions about the Degeneration, Trophic Interactions, and Repair of Severed Axons I. Axotomy induces rapid degeneration of distal axonal segments. Axotomy removes essentially all trophic sup- port to distal axonal segments. Initial axonal repair occurs by collapse and fusion of cut axolemmal leaflets to form a seal. Eventual axonal repair, when it occurs, is via slowly (1—2 mmfday) growing neurites that arise from proximal segments and may reestab- lish original functional connections and behav- iors weeks to years postseverance. According to this commonly held, internally con- sistent version of the “neuron doctrine,” neurons are individual cells that communicate with each other across intercellular spaces called synapses (l) as pro— posed by Ramon y Cajal (2)—and do not form a morphological or metabolic syncytium (reticulum), in which neurons are connected by cytoplasmic bridges, as proposed by Golgi (1). If neurons existed in a morphological syncytium, then axotomy should not produce rapid degeneration of anucleate distal segments (1—3). The common assumption that the cell body is the sole source of trophic support for a nerve axon is supported by observations that many distal segments degenerate rapidly (within 2—3 days) after axotomy (1—4), that proteins synthesized in the perikaryon are delivered to axons from all phyla by slow (1—2 mmlday) and fast (100-200 mmi'day) transport (1, 3. 5). and that axons lack rough endoplasmic reticulum as a defining morphological characteristic (1, 3. 5). Ultrastructural data from the early 19505 showing extracellular spaces of 20—30 can occur by a very rapid rejoining (fusion) of the cut ends of proximal and distal axonal segments. Furthermore, rehabilitative methods. such as envi- ronmental enrichment or treadmill training. may induce significant recovery of impertant behav- iors, even when original axonal connections are not reestablished by slow outgrowth or fusion. That is surviving distal axonal segments are natu- rally activated within days after severance by out- growths from surviving proximal segments in many invertebrates (15. 16, 25—49). Very rapid (within minutes) repair of severed invertebrate and/or mammalian PNS and CNS axons can now also be induced by specially formulated solutions containing polyethylene glycol (PEG), which rejoin (fuse) closely apposed ends (Box 2) of proximal and distal segments (51—54). Such PEG-fused axons conduct action potentials and allow intraaxonal dye diffusion across the lesion site. When a PEG-based hydrogel is subsequently Volume 6. Number 2. 2000 nm at synapses and biochemical data from the 19605 on slow and fast transport of proteins are commonly regarded as providing the crucial evidence needed to disprove Golgi and immortalize Cajal as the origina- tor of the neuron doctrine (l. 3). Although never carefully examined and challenged until recently (6—9). it is intuitively reasonable and widely assumed (I, 3. 10. 11) that initial axonal repair should occur by collapse and fusion of axolemmal leaflets to seal cut ends. Finally, it has been known for well over a century (2, 4) that proximal segments of PNS neu- rons may eventually complete axonal repair by slewly (1—2 mmfday) growing neurites that may reestablish original functional connections and behaviors weeks to years postseverance (1—3, 12). For more than a century, the surviving proximal seg- ments of CNS axons in mammals and birds were regarded as naturally incapable of eventual repair because neuritic outgrowths were, at best. abortive (1, 2, 12, 13). (CNS axons in fish and amphibians and many invertebrates often regenerate very well [14-17].) However. it is now generally accepted that CNS axons in mammals and birds can be induced to slowly regenerate neurites from proximal segments (13. 18—23), some of which may eventually establish connections that improve various motor behaviors (22). The acceptance of this set of interrelated assump- tions leads to therapeutic strategies focused on induc— ing severed proximal segments (or transplanted neurons) to slowly regenerate neurites to reestablish some of the original connections (1—3. 12, 13. 11—24). Box 2: Polyethylene Glycol (PEG) as a Cellular Fusiogen PEG induces the plasmamembranes of closely apposed cell membrane to fuse (50). PEG is solu- ble in both lipids and water. It probably induces cell membranes to fuse by removing water bound to proteins and other macromolecules on the outer surface of the plasmalemma. thereby allowing the lipid layers of closely apposed cell membranes to flow into each other. (The hydrophilic groups and bound water prevent cells in metazoan tissues from spontaneously fusing to form a syncytium.) Molecular biologists often use PEG to produce multinucleate cell hybrids by suspending two or more cell lines in saline and exposing the suspen- sion to 40%—60% PEG for 0.5—3 minutes. The cells are then resuspended in saline and grown on a substrate that allows the harvesting of colonies selected for a desired nuclear composition. THE NEUROSCIENTIST 89 Fig. 1. Morphology and ultrastructure remain normal during long-term survivai of anucleate distal segments from vertebrates or inver- tebrates. A-C. Electron micrographs of rat ventral tail nerves (modified from Fig. 4 of ref. 36). A, intact. B. Severed for 5 days in a con- trol rat. C. Severed for 5 days in a rat injected with cyclosporin A for 7 days before axotomy and for 5 days after axctorny. Note most axons appear normal. m = myelinated, um = unmyelinated. ' = degenerated axons. s = Sehwann cell. D—E. Photomicrograph (D) and electron micrograph (E) of distal segments of crayfish medial giant axons at 10? days pcstseverance {modified from Fig. 1 of ref. 46). C-MGA = crayfish medial giant axon. sh or brackets = cytoplasmic glial sheath. F—G. Photomicrograph (F) and electron micrograph (G) of distal segment of goldfish Mauthner axon at 50 days postseverance. MA = Mauthner axon. msh = myelin sheath. applied at the lesion site to provide mechanical strength (54}, behavioral functions mediated by severed axons repaired by PEG-solutions are rap- idly and permanently restored in vivo. PEG-fusion techniques (54) could become an important strat- egy to rapidly repair severed PNS and CNS axons. particularly when combined with rehabiiitative strategies such as enriched environments or limb-use training (55—58). Distal (Anucleate) Segments Do Not Neces- sarily Degenerate Rapidly after Axotomy The distal segments of most mammalian PNS and CNS axons degenerate morphologically (loss of axolemmal 90 THE NEUROSCIENTIST Degeneration, Trophic Interactions. and Repair of Severed Axons Table 1. Sources other than the cell body for maintenance of cytoplasmic proteins (ii) (iii) (iv) Long half- Cytoplasmic Transfer from Preparation Survival Time lives Nonneuronal cells Algae Acetabularia (I, A) 290 days + invertebrates insect oocytes (I) N N Leech macroglia (l, A) 260 days N Vertebrates Mammalian red blood cells (A) 2120 days ++ Mammalian lens cells (I, A) oo ++ Neurons Unmyelinated invertebrate axons Crayfish medial giant axon (l, A) 2210 days + Crayfish lateral giant axon (I, A) oo N Squid giant axons (l) N N Crustacean motor axons (I, A) 2210 days N Myelinated invertebrate axons Earthworm medial giant axon (l, A) co N Unmyelinated vertebrate axons Garfish olfactory nerve (I, A) 260 days ++ Amphibian optic axons (I, A) 260 days N Myelinated vertebrate axons Goldfish mauthner axon (I, A) 270 days + Normal mammalian PNS or CNS (l) 2—3d + Mouse WLD mutant (I, A) 214 days N Mouse diabetic mutant (l, A) 214 days N synthesis adjacent cells References ++ 0 72 ++ ++ NC 5 N ++ NE 73, 74 0 0 5 0 0 5 _ ++ G 15, 16, 25, 27, 32, 40, 45, 46, 69, 85 N ++ G, PSN 15, 16, 25,28, 138 +, - ++ G 37, 38, 77, 79, 80, 81 — ++ G, PSM 14—16, 25, 26, 67, 68 107, 111 N ++ PSN 16, 25, 3o, 49 N N 60, 61 N N 17, 63 ++ + G 16,31,32,39,42—44162r 92, 93 ++ + G 1—4, 34, 36, 39, 59 N N 34, 64, 65, 91, 100 N N 66 Key: l = Data from intact cytoplasm in contact with the cell nucleus, A = Data from anucleate cytoplasm, oo = Indefinite survival, perhaps forthe life of the organism, + = Evidence for minor contribution, ++ = Evidence for major contribution, 0 = No contribu- tion is theoretically possible, — = Evidence against possible contribution, N = No published data exist, G = Glia, PSN = Postsynaptic neurons, PSM = Postsynaptic muscle fibers, NC = Nurse cells, NE = Neurons. and cytoskeletal integrity) and physiologically (loss of membrane potential and ability to conduct action poten- tials) within 2 to 3 days after axotomy (Box 1; Fig. 1A,B) when maintained in vitro or in vivo at normal (~38 °C) mammalian body temperature (1—4, 12, 34—36, 59). However, considering the entire animal kingdom, such rapid degeneration of severed distal segments is probably the exception rather than the rule (14—16, 25). Since an initial report that distal segments of crustacean motor neurons survive for at least 3 months after axotomy (26), it is now known (16, 25) that many (most?) anucleate invertebrate axons exhibit long-term survival for weeks to months (Table 1) at body tempera- tures (10—35 °C) experienced by organisms in many in- vertebrate phyla (platyhelminths, molluscs, annelids, arthropods). Furthermore, long-term survival for weeks to months of severed distal segments (Table 1) is not a curiosity confined to invertebrates, but is also observed at normal body temperatures in lower vertebrates (e.g., garfish olfactory axons [60, 61], goldfish Mauthner ax— ons [31, 32, 44, 62], amphibian unmyelinated optic ax- ons [17, 63]), and mammals (e.g., sciatic axons in WLD [33, 64, 65] or diabetic [66] strains of mice). Anucleate distal segments maintain cytoskeletal and axolemmal in- tegrity (Fig. 1), membrane potential, action potentials, and/or neurotransmitter release (14—17, 25—39, 60—70), even if artificially stimulated for months after axotomy (14). Morphological and Physiological Characteristics of Long—Term Survival in Different Axons Although distal segments of many axons from both in- vertebrates and vertebrates exhibit long—term survival after axotomy, various morphological and physiological phenomena associated with long-term survival are not Volume 6, Number 2, 2000 THE NEUROSCIENTIST 91 Omo. 2mo. time. 7mo. 0m 2mm 7m- 200’ 200-- . 97’ 97- 1—71 ._55 59v 59’ ._. 46’ 46-» «—33 "" -‘ - — _ 4-43kD . 30’ g _ so- ‘ . __ 21v ’ '1' * 21" * Fig. 2. Proteins are maintained for months despite short hall-lives in intact or anucleate invertebrate axons. (Modified from refs. 45. 45.) A—C. Protein banding patterns in paired intact crayfish medial giant axons versus anucleate medial giant axons i”) are very similar until ~7 months postseverance. A. Silverstained gel profile of proteins in paired anucleate (“i and contralateral intact medial giant axons. Banding patterns for the intact paired medial giant axons are the same before severance {e mo); intact and anucleate medial giant axons at 2 and 4 months postseverance (2 mo. 4 me) are also very similar. At 7 months postseverance (7 mo). there are notable decreases of some but not all protein bands in the anucleate medial giant axon versus its paired control. For example. proteins with apparent molecular weights of 33. 55. and 71 kD (indicated by arrows to the right of the figure) are decreased in the 7-month anucleate medial giant axon compared to the contralateral intact medial giant axon. Molecular weight markers are indicated by arrow- heads to the left of the figure. 8. C. Tubulin (C). but not actjn (8), decreases in anucleate medial giant axons from 2 to 7 months. as detected on immunoblots. lmmunobtots produced lrom 1D gels used to separate proteins in homogenates of individual anucleate (') or intact medial giant axons. probed with anti-actin (B) and anti-alpha tubulin (C). The level oi actin {43 kD) in the pair of intact medial giant axons is the same before severance [0 mo] and remains constant for intact versus anucleate medial giant axons at 2 and 7 months (2 mo. 7 mo) postseverance. The level of alpha-tubulin (55 k0) in the pair oi intact medial giant axons is the same before sev- erance. However. the level of or-tubulin decreases with time in an anucleate medial giant axon versus its paired intact control. the same for different axons (16, 32). These differences are well illustrated by comparing the crayfish medial gi- ant axon (Fig. 10.5; Table 1) as a model for many in- vertebrate axons and the goldfish Mauthner axon (Fig. IEG) as a model for many vertebrate (including mam- malian) axons. For example, long-term survival in inver- tebrate axons, such as the crayfish medial giant axon, is associated with a maintenance within control levels of membrane potential and action potential amplitude, whereas axonal diameter continuously decreases with time (15, 16. 27, 25. 32). In contrast, goldfish Mauthner axons maintain axonal diameter along most of their length within control levels for months (31, 32, 44. 62)—and then suddenly disintegrate along their entire length within a few days (44. 62)—whereas resting po- tential and action potential amplitude more gradually decrease after axotomy (32). Although the data are much less complete, long-term survival of mammalian distal axonal segments also appears to be terminated by a phase of very rapid disintegration (34). All of these morphological and physiological changes during long-term survival are not necessarily temperature de- pendent in anucleate distal segments of crustacean ax- ons. which can last for months at 25 °C (14—16. 25, 27. 32. 69). In contrast, morphological andi'or physiological measures of long-term survival are very temperature de- pendent in goldfish Mauthner axons (16) and in mam~ malian axons (34, 35, 70). The degeneration of rat peripheral axons in the tail can be delayed for at least 5 to 10 days by lowering the tail temperature to 13 “C (35). The rapid 2- to 3-day degeneration of mammalian distal segments can also be retarded at 38 “C for at least 92 THE NEUROSCIENTIST 5 to 20 days by cyclosporin A (Fig. lA—C) (36). antibodies to complement 3 receptors (33), or by ge- netic alterations. for example. WLD (64, 65, 71) 01' dia- betic (66) strains of mice. These seemingly disparate conditions for delayed degeneration of the distal seg- ment may have in common decreases in axonal trans- port. the immune response. levels of complement 3 or its receptor, andior microglial activation. At the very least, these data show that the rapid degeneration of anucleate distal segments in mammals and birds is not solely due to their higher body temperatures. Widespread Evolution of Long-Term Survivai The data discussed above demonstrate that, contrary to what is often assumed (Box 1). the rapid degeneration of severed anucleate distal segments is neither obliga- tory nor necessary. even in mammals. In fact, the evolu- tion of long-term survival for anucleate pieces of nonneuronal cytoplasm is rather common in many or- ganisms, including mammals (Table I). For example. anucleate pieces of cytoplasm survive for months in the marine algae Acerabuiaria (7‘2). glial cells of leeches (73. T4). and red blood cells of mammals (5}-—and sur- vive for years in the crystalline lens cells of mammals (5). Hence. data on mammalian axons should be reex- amined for answers to the following question: what might account for the unusually rapid degeneration of their anucleate distal segments? As presented in the fol- lowing section, such a reexamination shows that com- mon assumptions about rates of protein turnover and axonal trophic interactions are not necessarily correct. Degeneration. Trophic Interactions, and Repair of Severed Axons 200 I“ 97- ' as» 1WK 45* so» _' 14> 200 " 9}" 69b 3 WKS 46- 30-" 14- TCapp cts. ! total cts. weeks post-in]. Fig. 3. Newly synthesized proteins in the iast component of axonal transport have half-lives of 2 weeks in intact or anucleete inverte- brate axons, even though total protein is maintained for months in those same axons (see Fig. 2]. (Modified from refs. 45. 45.} A. B. Fluorographs ot radiolabeled proteins in fast transport illustrating their short halt-life in individual crayfish medial giant axons. At 1 day following an iniection oi "S-methionine into supraesophageal ganglia. the right media] giant axons from two crayfish were severed close to their cell bodies. After 1 week (A) or 3 weeks {3). the anucleate medial giant axons were microdissected, cut into 3-mm seg- ments. and analyzed by SDS-PAGE and fiuorography. Note the decrease in radiolabeled proteins at 3 weeks compared to 1 week. 0. Graph showing the ratio of radioactivity associated with protein {i.e.. TCA-precipitable radioactivity. TCApp) to total radioactivity in indi- vidual anucteate medial giant axons severed 1 day after injection and sampled at 1. 2, or 3 weeks postinjection. These data show that proteins in last axonal transport have a half-life of about 2 weeks and are undetectable within 4 weeks. P = most proximal segment. D = most distal segment of the sampled axons. Intact and Anucleate Axons Do Not Neces- sarily Depend Exclusively on Their Perikaryon for Protein Maintenance The common assumption that the cell body is the sole source of trophic support for a nerve axon, which rap- idly degenerates in the absence of such support (Box 1), is consistent with many other observations that protein turnover times are measured in hours-to-days for many proteins in unicellular organisms (5, 75) or cells in tis- sue culture (5, 76), or for many essential enzymes {enolase. creatine phosphokinase. pyruvate kinase. etc.) in nonneuronal cells in vivo (75. 76). However, such data do not necessarily prove that the perikaryon is the sole source of axonal proteins (39)—and observations of long-term survival of severed distal segments contra- dict one assumption of any such “proof” (16). Further- Volume 5. Number 2, 2000 more, if the only source of axonal protein is slow or fast transport from the perikaryon and if the bulk of protein supplied to the axon from the cell body is carried by slow transport (41, 45, 46), then all proteins in slow transport essential for survival of intact axons must have unusually long half-lives compared to proteins in other cell types. In fact. intact axons from any organism might maintain various proteins supplied by slow or fast transport by any combination of four trophic sources: (i) slow and fast transport of proteins synthesized in the perikaryon, (ii) long half-lives of axonal proteins, (iii) axonal protein synthesis, andfor (iv) transfer to the axon of proteins synthesized in adjacent cells (glia, neurons. muscle fibers, etc.). Axotomy removes (i) the cell body as a trophic source for the anucleate distal segment. but (as summarized in Table 1) some combination of the other sources (ii-iv) might still be able to maintain vari- THE NEUROSCIENTIST 93 65p10254050 55 C 57 235—, q." hue» - :- ' I l . .. 145—» as - 132:: 3 3 a -- . 80-) IF - —» é ' 60—» - eh- _--*- .- 6 7 8 9 1O 11 12 13 .i. 235+ 1 2 3 4 5 I 1p 1- — t J ‘ - P - “" T 145 - r ; :g-P I i I 1 P .a. .. 80+ " 80-) i Fig. 4. A. B. Proteins are maintained for months in anucleate vertebrate axons. probably because phosphorylated proteins have long halt-lives. Phosphorylation of neurotilarnent proteins (NFPs) in distal segments oi severed goldfish M-axons (MAS) decreases lrom 10 to 40 days postseverance. From 50 to 65 days postseverance. phosphorytated NFPs are no longer detected with Stains-All (A), and novel bands (i.e.. NFP breakdown products) appear on silver stained gels (B) trom distal segments of MAs. (Data moditiad from Figs. 4-5 of ref. 44.) A, Stains-All-treated gel of control MA (lane 1}. 65-day proximal segment of MA (lane 2). and distal segments 01 MM. maintained in vivo for 10 to 65 postseverance days (as indicated by the number on top of each lane). Control MA (lane 1) and 65-day proximal segment (lane 2) have similar staining patterns in which NFPs (arrows, lane 1) at 235. 145. 123. 105. and 80 kDa stain blue (phosphorylated) and NFPs at 60 kDa stain pink (nonphosphorylated). As the postseverance time increases from 10 to 40 days (lanes 3—5). the staining ot NFP bands at 235. 145. 123, 105. and 80 kDa changes trom blue to purple, indicating a possible decrease in the phosphorylation state of these NFPs. At 50 through 65 days postseverance (lanes 5—15). NFPs in the distal segments of MAS do not stain at all, suggesting that NFPs are less phosphorylated than at 40 days postseverance. Although the lack of Stains-All reaction with NFPs after 50 postseverance days could be due to a complete degradation ot distal NFPs. the silver-stained gel below does not sup- port this latter hypothesis. 8. Silver-stained SDS gel ot Stains-All-treated gel in .4. Control MA (lane 1). 65-day proximal segment (lane 2). and distal segments of MAs severed tor up to 40 days (lanes 3~5) all have similar silver-stained banding patterns ot NFPs (arrows. lane 1). At 50 days postseverance (lane 6). the silver-stained banding pattern of NFPs in distal segments of Ms begins to change. that is. novel bands appear (arrowheads. lane 6). Although the NFPs are not as easily detected in the distal segments of MAS main- tained from 50 to 65 days postseverance. the NFPs can still be observed at 235. 145. 123, 105. 80. and 60 kDa for up to 65 days postseverance (arrows. lane 15}. 0—8 Confocal images oi severed earthworm medial giant axons (E-MGAs) showing dye barriers formed amid an accumulation of membranous structures. See rels. 47 and 48 for details about such experiments. 0. D. Severed E-MGA with its membranes labeled by a styryl dye (Flute-54. red or yellow false color). placed in saline containing the hydrophillc dye FlTC-doxtran (green false color} at 15 minutes postseverance. and confocally imaged at 20 {C} and 35 (D) minutes postseverance. Note that the dye barrier (arrow) forms amid an accumulation of FM-Iabeled membranous structures. which change their position and configuration trom 20 to 35 minutes postseverance. ' = original location of cut axona] end. E. E Severed E-MGA whose membranes are labeled with the styrl dye Fruit-43 (red or yellow in F). axoplasm is filled with the hydrophitic dye FlTC dextran (green in D and E ). and bath saline contains the hydrophilic dye Cy-5 (blue in E). The axon was contocally imaged at 25 minutes postseverance tor Cy-5 and FITC-dextran (E) and at 27 minutes postseverance for Cy-5 and Flint-43 (F). Note in E that the barriers coincide for the internally placed and externally placed hydrophilic dyes and that both barriers (arrow) are formed amid an accumulation of membranous struc- tures (F). “ = original location of cut axonat end. Scale bar for C—F = 10pm. 94 THE NEUROSCIENTIST Degeneration. Trophic Interactions, and Repair of Severed Axons B C D Sliver Stein Fluorograph Fluorogreph 200— 7—- is} 116— ' 97 _ 97— . ..... if: 69— .g_ E” — 66‘ Gib ' ’ ‘ —- -.-:. 1-— 43— r: 9- - __ _ n: 30— _ ‘- 1 2‘ I“ SH F 83 1 2' 3 4 5' 6 7' 8' 9‘ 10' Fig. 5. Silver-stained (total proteins) and fluorographed (newly synthesized proteins) gels showing giia-to-axcn transfer oi protein in intact and anucieate invertebrate axons. (Data from Figs. 3-4 of ref. 40.) A. Comparison oi silver»stained proteins in the axoplasm oi intact and anucieate crayfish MGAs showing that most proteins are still present in anucleate MGAs 6 months after contact with the perikaryon is removed. In this and subsequent figures. gel lanes labeied with asterisks designate anucieate MGAs (5 months postseverance in this figure). and gel lanes without asterisks designate intact MGAs. Numbers (200. 116. 97. 66. and 43) to the left of the gel lanes designate positions of molecular weight standards. Lines to the right of lens 2" indicate the most heavily silver-stained bands (70, 55. 43. 38. 27, and 25 kDe) in axoperiusate samples. B. Fluorograph oi radiolabeled proteins in axoperfusate (AX) and sheath (SH) samples showing that many. but not all. newly synthesized glial proteins are transferred to the axoplasm. Lines between lanes AX and SH indicate heavily radiolabeled bands (60. 4E, 55. and 34 kDa) common to axoperfusate and sheath samples. Open arrow (AX lane) indicates a band (30 kDa) in the axoperfusate sample. but not in the sheath sample. Solid arrow (SH lane) indicates a heavily radiolabeled band (200 kDa) in the sheath sample, which is not in the axoperiusate sample. C. Fluorograph (F) and silver stain (88) of the same axoperfusate sample showing that most. but not all. proteins in MGA axoplasm can be received from adjacent glia. Arrows designate heavily radiolabeled bands. which correspond to heavily silver-stained bands. D. Fluorograph oi radiolabeled axoperfusate proteins irom 10 different intact and anucieate MGAs (lanes 1—10') all electrophoresed on the same gel showing that both intact and anucieate MGAs obtain many radiolabeled (newiy synthesized) proteins from adjacent glia. Arrow at right designates the most heavily radiolabeled axoperfueate band (46 kDa). Lines at right designate other heavily radiolabeled axopertusate bands (96, 90. 74. 60. 44. 34. and 30 kDa). Brackets below lanes indicate that the two axoperfusate samples were irom the left and the right MGAs of the same animal. Anucleate MGAs were severed and maintained in vivo prior to sampling as follows: (2" = MGA severed 3 weeks). (5‘ = severed 6 weeks). {7' = severed 2 weeks). (8‘ = severed 9 weeks). (9“ = severed 5 weeks). {10' = severed 8 weeks). Lanes 1. 3. 4, and 6 were irom intact MGAS. ous proteins in the distal segment. Such data for intact (or recently severed) axons and anucleate distal seg- ments exhibiting long-term survival (Table 1) are avail- able primarily for crayfish medial giant axons (Figs. 2. 3. 5) and goldfish Mauthner axons (Fig. 4). Data from intact squid giant axons are also extensive (37—39. 77—81) and very similar to those obtained from crayfish medial giant axons (40. 45. 46. 69. 82—83), but no data exist for anucieate squid giant axOns (Table 1) because squid do not survive long in captivity. especially if injured. Sources of Trophi'o Support in invertebrate Axons Intact crayfish medial giant axons and squid giant axons appear to have a major trophic dependence (i) on their perikaryon and (iv) on glial-synthesized proteins trans- ferred to the adjacent axon, and a lesser dependence (ii) on long protein half~lives or (iii) on axonal protein syn« thesis. That is, (i) intact crayfish medial giant axons (Figs. 2. 3) and squid giant axons receive many radiolabeled (newly synthesized) proteins hours (via fast transport) and weeks to months (via slow axonal trans— port) after radiolabeled proteins are synthesized in the nerve cell body (45. 46. 77). (ii) The half—life of all pro- teins in slow transport is about 14 days and 17 days in fast transport (45. 46). At about 0.8 mmfday, these slowly transported proteins would take 50 to 160 days to reach the end of a medial giant axon. Hence. the maintenance of many slowly transported proteins almost certainly requires supplementation by (iii) or (iv). There is agreement that squid giant axons contain much trans- fer RNA (39. 79) and no disagreement that squid giant axons contain much messenger RNA (39, 79). There is. however. disagreement as to whether squid giant axons contain some (39) or no (79) ribosomal RNA and whether axonai protein synthesis contributes some (39) or undetectable (3?. 38) amounts of protein compared to glial-axonal protein transfer (Table 1). Considering all available data. (iii) axonal protein synthesis probably contributes few proteins to intact crayfish medial giant axons (40). In contrast. (iv) intact crayfish medial giant axons and squid giant axons begin to receive many newly synthesized glial proteins within 15 to 60 minutes Volume 5. Number 2, 2000 THE NEUROSCIENTIST 95 after their ensheathing glia synthesize radiolabeled pro- teins (37, 38, 40, 80—84). (The ensheathing glial layers contain much rough endoplasmic reticulum, which is the source of most protein synthesis for eukaryotic cells.) Many (but not all) species of proteins transferred to the axon from adjacent glia are proteins also deliv- ered to the axon by fast or slow transport (37, 38, 40, 45, 46, 80, 81, 84). Actin, tubulin, and heat shock pro- teins of the 70 kD family (HSP 70 isoforms) are received by transport from the cell body and by transfer from ensheathing glia. About 10% of the proteins syn- thesized in the sheath may be transferred to the axon and 10% to 50% of the proteins in the axon may arise from ensheathing glia; much more actin than tubulin appears to be transferred from glia in crayfish medial giant axons (40, 46). Intact medial giant axons contain high levels of constitutive heat shock proteins that may serve to protect against protein damage should an axonal stress occur (84). The mechanism of transfer for many glial-synthesized proteins appears to be by direct translation across the axolemma (81), perhaps by one or more isoforms of heat shock proteins synthesized in the glia (84). Axotomy (i) eliminates proteins synthesized in the cell body as a source of protein for anucleate crayfish medial giant axons (40). Nevertheless, anucleate medial giant axons exhibit long-term survival for about seven months (25,27,32,85), during which time many proteins are maintained at near-normal levels (40,45,46,84). The (ii) half-life of all proteins in slow transport is 17 days for anucleate medial giant axons (Figs. 2, 3), very simi- lar to the 14-day half-life of proteins in intact medial giant axons (45, 46). At 21 days after axotomy, many proteins in fast transport are completely degraded (Fig. 3), that is, are no longer detectable on silver stains or autoradiographs of 1D gels. Hence, an average protein half-life of about 2 weeks cannot easily account for maintenance of most proteins for about 7 months in anucleate medial giant axons (40, 45, 46). Similarly, (iii) axonal protein synthesis contributes few proteins to anucleate medial giant axons (40). In contrast, (iv) glial-to—axonal protein transfer makes a major contribu- tion to protein maintenance in anucleate crayfish medial giant axons (Fig. 5; 40, 81, 84), as reported for intact crayfish medial giant axons or squid giant axons. The axoplasmic levels of tubulin, but not actin, decline in crayfish medial giant axons during long-term survival of 7 months, perhaps because much more newly synthe- sized actin than tubulin is transferred from adjacent glia (Fig. 2; 45, 46). The cytoplasm of the ensheathing glial cells that surround these unmyelinated medial giant axons exhibits much hypertrophy (including prolifera- tion of rough endoplasmic reticulum) and hyperplasia following axotomy (14, 15, 27, 82, 85). Following heat stress (and presumably other stresses such as axotomy), anucleate and intact medial giant axons also receive inducible heat shock proteins from adjacent glia (84). The high levels of constitutive heat shock proteins and the rapid transfer from glia of inducible heat shock 96 THE NEUROSCIENTIST proteins may also help maintain many axonal proteins for months in anucleate distal segments. An effect of increased temperature to decrease protein half-life but to increase protein synthesis and transfer from adjacent nucleated cells, as well as the high levels of constitutive and inducible heat shock proteins, may also account for the rather small effect of temperature on the long-term survival of distal segments of crayfish medial giant axons (32, 84). A similar conclusion that (iv) adjacent glial cells are a major source of trophic input is also reached for other invertebrate axons that exhibit long-term survival for months to years (e.g., crustacean motor axons, crayfish lateral giant axons, and earthworm medial giant axons), although other adjacent cell types may also contribute proteins to intact or anucleate segments (14, 25). For example (Table 1), crustacean muscle fibers make a major trophic contribution to anucleate motor axons (15) and pre- or postsynaptic neurons make a major trophic contribution to crayfish lateral giant axons (28) and to earthworm medial giant axons (30, 49). However, some anucleate motor axons in rock lobsters have been reported to incorporate glial nuclei within the axoplasm (68), perhaps leading to increased levels of axonal pro- tein synthesis in anucleate lobster axons compared to crayfish medial giant axons or squid giant axons (37, 38, 68, 69). Sources of Trophic Support for Vertebrate (Including Mammalian) Axons Intact vertebrate axons (including goldfish Mauthner ax- ons) appear to have a major trophic dependence (i) on proteins newly synthesized in the perikaryon via slow and fast transport, (ii) on long half-lives of some pro- teins, and (iii) on axoplasmic protein synthesis—and a lesser trophic dependence (iv) on adjacent glial cells (Table 1). Specifically, in intact axons from vertebrate visual systems that show rapid (2- to 3-day) degenera- tion, (ii) total protein associated with slow and fast transport has a half-life of about 6 weeks in rats (86) and slow component proteins have a half-life of 14 days in rabbits (87) and 2 to 7 (88) or 14 (86) days in mice. Some cytoskeletal proteins can have half-lives of 20 to 55 days (neurofilaments: 41) or even 400 days (tubulin: 89). Furthermore, neurofilament proteins (90) are heavily phosphorylated (Fig. 4A,B) and calpain levels are low in intact Mauthner axons (42—44, 62) and intact axons from WLD mice (71, 91) whose anucleate distal segments exhibit long-term survival for weeks. (Phosphorylated neurofilaments are less easily degraded by calpains and other proteases [Fig. 4A,B; 40, 91].) Nevertheless, although some slowly transported proteins may have sufficiently long half-lives, the maintenance of many proteins in slow transport in intact vertebrate axons is not accounted for by the 2— to 14-day half-life of total protein in slow transport (39, 41—43). It is now apparent that some, perhaps many, proteins in intact goldfish Mauthner axons and mammalian axons can be supplemented by (iii) axonal protein synthesis, that is, Degeneration, Trophic Interactions, and Repair of Severed Axons commonly held assumptions that axons lack an endoge- nous ability to synthesize protein are not necessarily correct (39, 92—94). For example, all the necessary mo- lecular machinery (transfer, messenger, and ribosomal RNA) have been found in goldfish Mauthner axons and in various mammalian axons (39, 92—96)—and the amount of ribosomal RNA in Mauthner axons is several times the amount in the cell body or ensheathing glia (39, 92). Furthermore, incorporation of radiolabeled amino acids into protein has been reported for goldfish Mauthner axons and for mammalian axons (39, 92, 93). Although ultrastructural studies report that axons lack rough endoplasmic reticulum or free ribosomes as- sumed necessary for endogenous protein synthesis, RNA with high phosphorous content typical of ribo- somes is present in subaxolemmal plaques (39, 93). Finally, there is some evidence for (iv) transfer of some glial-synthesized proteins to vertebrate axons. For ex- ample, there are reports that radiolabeled glial proteins are transferred to cat CNS neurons, astrocytes transfer sodium channels at nodes of Ranvier (97, 98), and glial-synthesized B-galactosidase is transferred to adja- cent axons (99). However, the number and amount of glial-transferred proteins appear to be much less than that described above for invertebrate axons. This con- clusion is consistent with data showing that many axoplasmic proteins (especially neurofilament proteins) are heavily radiolabeled within minutes after Mauthner axons or sciatic axons are bathed in radiolabeled amino acids, but very little radiolabel is detected in the sheath within hours; the labeling of such proteins is blocked by inhibitors of eukaryotic protein synthesis (39, 92, 93; Bittner, unpublished observations). Axotomized vertebrate axons that exhibit long-term survival appear to maintain proteins by the same mecha- nisms that supplement axonally transported proteins in intact axons, that is, (ii) long protein half-lives and (iv) axonal protein synthesis. That is, most proteins, includ- ing neurofilaments, are maintained at control levels for about 72 days after axotomy in all portions of anucleate distal segments of Mauthner axons from goldfish kept at 15 °C (Fig. 4A,B); the diameter of these distal segments does not obviously decrease during this time (Fig. 1F,G; 32, 44, 62). From 72 to 85 days after axotomy, most proteins are degraded at the same time that axonal diameter rapidly decreases (43, 62). Neurofilaments, in particular, remain heavily phosphorylated for about 72 days after axotomy, at which time their phosphorylation decreases and neurofilament breakdown products begin to appear (Fig. 4A,B). By 85 days, neurofilaments are almost completely degraded. Calpain levels in anucleate Mauthner axons are as low as in intact Mauthner axons, and then decrease from 80 to 85 days after axotomy at a time when the degradation of neurofilaments by calpain is increased. These data suggest that a reduced level of neurofilament phosphorylation results in their rapid deg- radation (even though axonal levels of calpain are decreased) and that neurofilament degradation may pro- duce a rapid collapse of the entire distal segment. These Volume 6, Number 2, 2000 data from goldfish Mauthner axons (44, 62) are also very consistent with data showing that neurofilaments in WLD mice whose anucleate distal segments exhibit long-term survival for weeks are more resistant to calpain degradation than are neurofilaments in control strains of mice (71, 91, 100). In WLD mice, the prop- erty for long-term survival is intrinsic to the axon (64, 65) and is inherited in autosomal dominant allele on chromosome 4. In normal mice or rats, axonal degener- ation of anucleate distal segments may be retarded for days to weeks by reduced temperatures (34, 35) or cyclosporin A (Fig. 1A—C; 36), both of which may slow protein turnover or inhibit phagocytosis, or antibodies to complement 3 receptor, which reduces the immune response (33). Finally, axoplasmic proteins (especially neurofilament proteins) are very heavily radiolabeled, whereas glial proteins contain very little label when dis- tal Mauthner axon segments axotomized for 7 to 50 days are bathed in radiolabeled amino acids (Bittner, unpublished observations). These data are very similar to data from intact Mauthner axons (39, 92, 93), which suggest that (iii) axonal protein synthesis, but not (iv) glia-to-axon protein transfer (although it probably exists), is a major source to supplement axonal proteins synthesized in the perikaryon (Table l). Conservative Evolution of Cellular/Molecular Mechanisms for Trophic Support Long-term survival of anucleate cytoplasm has evolved by various combinations of mechanisms (ii) through (iv) in many cells other than neurons (Table 1). As one ex- ample, the long-term survival for months to years of mammalian red blood cells and crystalline lens cells is presumably due to (ii) long protein half—lives (5). In fact, the commonly held assumption that proteins inher- ently have short half-lives is almost certainly incorrect. Rather, proteins are inherently stable molecules that have long half-lives unless they are actively degraded (5). As a second example, the 1- to 4—cm-long stalk and umbrella of the algae Acetabularia survive for months after enucleation because of (iii) local cytoplasmic pro— tein synthesis utilizing long-lived species of RNA (72),which could also occur in anucleate Mauthner ax- ons and other vertebrate axons—0r axonal RNAs could be supplemented by RNAs transferred from adjacent glia (39). As a third example, insect oocytes (iv) receive many proteins transferred from adjacent nurse cells (5). Given that many other cell types from many phyla have evolved various cellular/molecular sources and mechanisms for trophic support, it has always seemed counterintuitive that axons consisting of vulnerable threads of cytoplasm of exceedingly long length and small diameter in all phyla should all (i) necessarily depend entirely on the cell body to maintain axonal pro- teins (Box 1). On such a priori grounds, it seems most appropriate to find that there are now many data that show that such an obligatory dependence “ain’t neces- sarily so” and that (ii) axonal proteins supplied by slow or fast transport can have long half-lives or be THE NEUROSCIENTIST 97 Mechanisms for Sealing of Severed Ends of Proximal or Distal Axonal Segments Commonly assumed A. transection site B . damaged axolemma Fig. 6. Cartoon illustrating commonly assumed (A, B) versus actual mechanisms (C—F) for sealing of severed ends of proximal and distal segments. A, Seal assumed to form by collapse and fusion of axolemmal leaflets. B, Seal assumed to form by creation of new membranous partition. C, Seal observed to form amid single-layered vesicles formed by endocytosis in unmyelinated axons. D, Seal observed to form amid multilayered membranous structures formed by myelin delaminations in myelinated axons. E, Within 24 hours, the cut end is sealed by a continuous membrane and few injury-induced membranous structures are still present. F, Neuritic out- growths arise from cut ends after 24 hours when almost all injury-induced membranous structures have been resorbed. supplemented by proteins arising from (iii) axonal pro- tein synthesis or (iv) transfer from adjacent cells (Table 1). Hence, although Ramon y Cajal (2) was correct in concluding that the nervous system does not form a morphological syncytium (if we all agree to ignore gap junctions that contain cytoplasmic bridges 1—2 nm in diameter), the data and rationale that led to his conclu- sion are not correct (i.e., axons do not necessarily degenerate rapidly and axonal degeneration after axotomy is not a phenomenon that proves or disproves the existence of a syncytium). Furthermore, nervous 98 THE NEUROSCIENTIST tissue does form a metabolic syncytium in which large amounts of protein can be transferred between cells by various combinations of endocytosis, exocytosis, pinocytosis, membrane translocation, etc. All these data are also consistent with the suggestion given in an ear- lier section that anucleate distal segments of mammalian axons do not passively disintegrate, but instead their proteins are actively degraded by endogenous or exoge- nous mechanisms such as phosphatases, peroxidases, proteases, and phagocytosis. Reduced temperature and cyclosporin A each block one or more of these Degeneration, Trophic Interactions, and Repair of Severed Axons mechanisms for actively degrading anucleate distal stumps, and hence can be used to retard the rate that anucleate distal segments degenerate in mammals. Initial Repair of Severed Proximal or Distal Ends Occurs by the Formation of a Seal amid Accumulations of Membrane-Bound Structures Complete repair of severed (cut or crushed) axonal seg- ments requires the successful completion of many cellu- lar events (9), but the process begins with the successful formation of a barrier (seal) to ions and other sub— stances at the cut proximal and distal ends (69). With- out the rapid (minutes to hours) formation of a seal, axonal segments lose essential substances (e.g., pro- teins) and gain deleterious substances (e.g., Ca2+) by in— flow or outflow at the cut ends. Ca2+ inflow in particular can lead to rapid (hours to days) degeneration of axonal segments or of entire neurons (59) and thereby preclude eventual axonal repair by slow outgrowth of neuritic processes from the proximal end or the more rapid re- pair by activation or fusion of surviving proximal seg- ments with surviving distal segments. For decades, the repair of axolemmal damage was assumed to be assessable by measuring axonal mem- brane potential and/or input resistance (11, 101, 102), by determining the ability of the axolemma to exclude fluorescent dyes or other tracers (10) or by examining low-power ultrastructural sections (102a). However, mathematical and analog models demonstrate that elec- trical measures of membrane potential or input resis- tance by themselves provide a very ambiguous and inaccurate assessment of the status of a barrier because of its sensitivity to unknowable changes in axonal cable properties (103). A much more reliable assessment of the status of an axonal seal is given by combining mea- sures of injury current density (not dependent on mem- brane cable properties) and membrane potential with exclusion of dyes carrying different electrical charge (or having no net charge), sequential ultrastructural sec- tions, and serial confocal sections (6—9, 103). Morphological and Physiological Characteristics of Axolemmal Repair Data from this more reliable combination of measures show that the formation of a seal to initially repair axolemmal damage is not due to the collapse and fusion of axolemmal leaflets at cut ends (3, 11) (Fig. 6A) or the formation of a partition-like structure (102) (Fig. 63), as was once almost universally assumed (Box 1). Rather, injury-induced vesicles and other membranous struc- tures accumulate and interact in various (but not all pos— sible) ways to seal small holes or complete transections of the axolemma (Fig. 6C,D) at both distal and proxi- mal cut ends (7—9). In unmyelinated axons (Fig. 6C), small holes or complete transections are sealed primar- ily by Ca2+ influx at the lesion site, which induces sin- Volume 6, Number 2, 2000 gle-walled vesicles to arise from the axolemma and glialemma by endocytosis (7—9). These vesicles migrate rapidly to the lesion site and seal the cut end within 10 to 60 minutes after axotomy by forming a tightly packed plug or a continuous membrane that fuses with intact axolemma some tens of microns from the cut end. Cut axonal ends usually constrict and shorten within minutes after axotomy but do not completely collapse (close). Initial seal formation is very similar in myelinated axons (Fig. 6D), except that Ca2+ influx at the lesion site induces myelin delaminations, which rapidly form mul- tilayered membranous structures that also migrate to the lesion site to help form a barrier (6, 47, 48). This barrier is very dynamic in myelinated earthworm medial giant axons with respect to its location and configuration from 15 to 35 minutes postaxotomy (Fig. 4C—F) and is rather stable thereafter (48, 104). By 1 day postaxotomy, a continuous axolemma is usually discern- ible at the distal and proximal cut ends (Fig. 6E; 104). From 1 to 10 days postaxotomy, the vesicular structures associated with the seal gradually disappear (Fig. 6E) and neuritic outgrowths arise from sealed axonal ends whose axoplasm appears very normal (Fig. 6F; 104). Some of the morphological structures and physiologi- cal mechanisms involved in the initial formation of a seal have now been identified and have similar actions at severed proximal and distal ends of invertebrate and vertebrate axons. For instance, Ca2+ influx at the cut ends rather than injury per se or Ca2+ influx via mem- brane channels initiates vesicle formation by endocytosis, myelin delamination, and the sequence of dynamic changes associated with barrier formation (7—9, 48, 104). Ca2+ influx needs to increase internal Ca2+ to 100 uM to induce vesicle formation and Ca2+ is much more efficient than any other ion in inducing endocytosis (8). Ca2+ influx also activates endogenous calpains that act on vesicles to form a tightly packed plug or continuous membrane necessary to seal the cut ends (105). Consequently, inhibitors of calpain or other protease inhibitors prevent sealing, and addition of exogenous proteases can enhance scaling (105). In gen- eral, it appears that calcium plays multiple detrimental and beneficial roles in axolemmal sealing (8, 9, 105). That is, Ca2+ influx elevates intracellular Ca2+, which activates endocytosis or myelin delamination as well as proteases, which begin to degrade proteins that lead to cell death—but proteolytic modification is also neces- sary to promote vesicular interactions required for seal- ing. In contrast, some substances that affect cytoskeletal elements have different effects on invertebrate and mam- malian axons. For example, 20 mM taxol (microtubule sta- bilizer) or 6 ug/ml cytochalasin E (F-actin destabilizer) do not affect sealing of crayfish medial giant axons (9, 105) but do inhibit sealing of mammalian septal axons (10). THE NEUROSClENTIST 99 Natural Mechanisms of Axonal Repair Commonly Assumed \W/ I £13k“; B. {fiwmm W gm KL/ M} M it Site of fusion, gap junctions, synaptic contacts, etc. l 1 l. ,fl ®ml.é¥ gjmwwroflrwa c Fig. 7. Cartoon showing commonly assumed versus alternate natural mechanisms for axonal repair. A—D, Axonal repair after axotomy (A) is commonly assumed to occur exclusively by degeneration of the distal segment and sealing of the proximal segment by collapse of axolemmal leaflets (B) followed by slow outgrowth of neurites from the surviving proximal segment (C), which may reform its original synapses (D). 5—], Axonal repair after axotomy (E) and sealing of the proximal and distal segments by membranous vesicles (F) can occur by outgrowths from the proximal stump that contact (G) and activate the distal stump to generate action potentials. The proximal segment may fuse with the distal segment (H) or may make electrical, chemical, or ephaptic synapses with the distal segment (H). In the first instance, the axon is completely repaired. In the other cases, neurites arising from the proximal segment continue to grow alongside the surviving distal segment and make synapses on the original postsynaptic cells—at which time the surviving distal seg- ment rapidly degenerates (I). ’r’N 100 THE NEUROSCIENTIST Degeneration, Trophic Interactions, and Repair 01 Severed Axons Mechanism of Axonal Repair by PEG A. axonal ends sealed at transection site U*’ B. hypotonic solution with no Ca2+ D. Caz+ containing saline, no PEG WW E. PEG hydrogel l WWWM Fig. 8. Cartoon showing mechanism of axonal repair by polyethylene glycol (PEG). The closely apposed cut ends of axons sealed by accumulations of membranous structures (A) are placed in hypotonic saline lacking Ca“ to open the ends and prevent further vesicle formation (3). PEG is then applied to join the membranes of the two axons at regions where they are particularly closely apposed (C). The axon is then placed in Ca2‘-containing salines so that membranous structures will seal the holes in the incompletely fused cut ends (D). A PEG-based hydrogel is then placed around the axons fused by the PEG solution to give mechanical strength to the site of axolemmal repair (E). Volume 6, Number 2, 2000 THE NEUROSCIENTIST 101 Fig. 9. Morphological (A. B, D} and electrical (C) measures showing that a combination oi PEG solutions and PEG hydrogels can rap- idly and completely repair (fuse. join together] cut axonai ends in vitro and in vivo. A, Electronmicrograph of a sagittal (longitudinal) section through a cut crayfish medial giant axon {C-MGA) that was fused with PEG within 30 seconds in vitrc 6 hours prior to fixation. The C-MGA conducted action potentials through the site of membrane and axoplasmic fusion (arrow) for 6 hours. Sh = cytoplasmic glial sheath oi this unmyelinated axon. Modified from ref. 51. B. Photomicrograph of a Lucifer dye-filled earthworm MGA (E-MGA) that was cut and PEG fused within 2 minutes in vivo 20 days prior to sampling. The E-MGA was surrounded with a PEG hydrogel to give mechanical strength in vivo. The E-MGA was completely repaired by PEG because it conducted action potentials through the fusion site for 20 days. its stimulation evoked all the behaviors evoked by intact E-MGAs in unoperated earthworms, and its dye-filling showed the same morphology as that shown by nonsevered MGAs. Arrow = site of PEG-fusion of cut ends and site where the PEG-based hydrogel was applied. C. Individual compound action potentials (CAPS) stimulated in one end chamber and recorded in the other and chamber of a three-chambered device used for sucrose gap recordings (see ref. 54). The CAP was first recorded from intact control bundles of rat spinal axons (trace labeled 1) prior to replacing the physiological saline in the central chamber with hypotonic Ca"-free saline (trace labeled 2) at ~25 "C. The spinal axons were then completely cut in the central chamber to eliminate the CAP (trace labeled 3). PEG was applied to the apposed cut ends for 2 minutes, and the central chamber was again perfused with calcium-contain- ing physiological saline. within 15 minutes. the CAP again appeared in fused spinal axons (traces labeled 4) and remained for more than 80 minutes before the experiment was terminated by again cutting the spinal axons in the central chamber to eliminate the CAP, that is, to demonstrate that the CAP was not an artifact (trace labeled 5). 0, Electron micrographs of cross sections of PEG-fused rat spinal axons at the site of PEG-induced fusion. The axons were fixed 2 hours after they were shown to conduct action potentials through the site of PEG fusion. N = PEG-fused axons of near-normal morphology. P = PEG-fused axons of pathological morphology (disrupted myelin sheath and many membranous structures in the axoplasm). Conservative Evolution of Ceilulan’i'dolecular the configuration of the injured plasmalemma, the type Mechanisms for Piasmaiemmai Repair of injured cell and its relation to nearby cells. and the proximity of various membranous structures to the le- sion (7). Nevertheless. in many kinds of axons, muscle fibers. sea urchin eggs. and mammalian cell lines, plasmalcmmal damage is repaired by accumulations of membranous structures from the most readily available sources (Fig. 6C,D). Furthermore. many of these same molecular mechanisms responsible for plasmalemmal As discussed above for cellularlmolccular mechanisms of protein maintenance, it is now obvious there has also been a very conservative evolution of mechanisms for sealing plasmalemmal damage in axons and all other eukaryotic cells. The details of this common mechanism for plasmalemmal repair vary with the type of lesion, 102 THE NEUROSCIENTI ST Degeneration. Trophic Interactions, and Ftepair of Severed Axons repair by interactions of membranous structures are also responsible for interactions of membranous vesicles in the Golgi apparatus and in nerve terminals (6—9, 106). Eventual Functional Repair of Severed Axons Does Not Necessarily Have to Occur by Slowly Growing Neuritic Out- growths: Axonal Fusion and Rehabilitative Training in Enriched Environments Represent Alternate Strategies A Flejoining (Fusion) of Severed Proximal and Distal Segments Can Rapidly Produce Anatomi- cal and Physiological Repair The mechanism of eventual repair of PNS or CNS ax- ons in any organism does not necessarily have to occur by degeneration of the distal segment and slow out- growth of neuritic processes from the proximal segment (Box 1), which may eventually reestablish functional connections and behaviors (Fig. 7A—D). That is, as de- scribed in previous sections, sealed distal segments (Figs. 6C,D, 8A) can also survive for weeks to years in invertebrates and vertebrates (Table 1). In addition to providing long-term trophic support of postsynaptic muscle fibers in invertebrates (14, 15, 67, 107), long-term survival of distal PNS or CNS segments in in- vertebrates is also associated with a rapid (within days) and specific reconnection of the surviving distal seg- ment to the appropriate proximal segment (Fig. 7E—I). After axotomy (Fig. 7E), neuritic outgrowths (Fig. 70) extending from the sealed proximal segment (Fig. 7E) selectively and rapidly contact the appropriate surviving distal segment (Fig. 7G) and cause it to generate action potentials (“activate the distal segment”) by forming chemical (108, 109) or electrical (110) synapses, by ephaptic current spread (111), or by fusing with the sur- viving distal segment (29). In all but the last mechanism of distal segment activation, the'distal segment eventu— ally degenerates and the proximal segment eventually reforms its original synaptic connections (Fig. 71). If the outgrowths from the proximal segment fuse with the surviving distal segment (Fig. 7H), then morphological and functional repair has been completed much more rapidly compared to reformation of synaptic connec- tions on the denervated tissues by slowly growing [neurites (Fig. 7D,I). Axonal repair by fusion of the cut ends of severed axons can now be induced to occur very rapidly in sev- ered CNS or PNS axons of invertebrates or vertebrates (including mammals). That is, specially formulated Cab-free solutions of polyethylene glycol (PEG) have been used in vitro to rapidly (within minutes after appli- cation) fuse the cut ends of severed unmyelinated cray— fish medial giant axons (Fig. 9A; 51), myelinated earthworm medial giant axons (Fig. 9B; 52—54), unmyelinated mammalian NG 108-15 neurites (51), and myelinated sciatic or spinal axons from rats, rabbits, and guinea pigs (Fig. 9C,D; 54). Successful fusion has Volume 6, Number 2, 2000 been documented by the restored conduction of action potentials through the lesion site (Fig. 9C), the intra—axonal diffusion of intracellularly located dyes through the lesion site (Fig. 9B), and the presence of intact axonal profiles that traverse the lesion site in elec- tron micrographs (Fig. 9A,D). Furthermore, the in vivo use of PEG solutions combined with the in vivo use of a hydrogel (synthesized from PEG) rapidly and perma- nently restores medial giant axon structure and function (Fig. 93), as well as all behaviors elicited by stimulating earthworm medial giant axons prior to their transection (54). The success of our PEG solutions to fuse cut axonal ends almost certainly depends on mechanisms that repair axolemmal damage (Fig. 8A), as described in the previous section (Fig. 6C,D). Specifically, PEG is a fusiogen (Box 2) soluble in both lipids and water. Its mechanism of action is probably to remove water from hydrophilic groups on the exterior of adjacent plasma membranes (50), thereby inducing them to fuse (Fig. 8C). The fusion of cut axonal ends is often aided by ini- tially bathing them in hypotonic Ca2*-free saline. (Sometimes the saline also contains 0.5—1 mM EDTA to bind any endogenous Ca2+.) The hypotonic saline (52—54) causes previously sealed cut ends that are par— tially constricted (Fig. 5A) to swell and open up (Fig. 8B). The elimination of Ca2+ prevents the formation of Ca2+—induced membranous structures at the cut ends (6—9, 47, 54). The cut ends are then bathed for 0.5 to 3 minutes in a hypotonic, Ca2+—free solution of 40% to 60% PEG (47, 52—54), which induces the fusion of the axolemma of the two axonal segments at points where they are especially close together, leaving some plasmalemmal discontinuities (Fig. 8C). The site of PEG—induced fusion is then bathed in Ca2+-containing physiological saline so that Ca2+ influx at the disconti- nuities can induce the formation of membranous struc- tures (7, 8, 54), which, in turn, seal the discontinuities (Fig. 8D). Within 2 to 30 minutes after PEG application, the plasmalemma is sufficiently repaired so that action potentials are transmitted across the fusion site. The membranous structures that seal membrane discontinu- ities are then resorbed over the next 12 to 24 hours (104; Fig. 6E). The mechanical strength of PEG—fused axons is very weak at the fusion site, and axons PEG-fused in vivo pull apart at the lesion site when the earthworms or rats recover from anesthesia and begin to move about (54). (The mechanical strength of bundles of intact axons in vivo is due to collagen and other material in their sheath, not to their plasma membranes.) To provide mechanical strength to PEG—fused axons, a PEG—based hydrogel (112, 113) that is noncytotoxic to invertebrate and mammalian tissue (54) is applied to the lesion site (Fig. 8E). When this PEG-based hydrogel is applied to severed earthworm medial giant axons whose cut ends are reconnected by PEG-solutions in vivo, then such PEG-fused medial giant axons continue for weeks to conduct action potentials, allow intracellular dyes to THE NEUROSCIENTIST 103 A. Increased dendritic arborizatlon B. Prevents Increased arborlzatlon In cortex opposite the lesion 18 13 g 15 Lesion-No Cast g 15 m'm'm ca“ o § Lesions-Contra Cast g 14 N E E 12 Leelon-r-Ipsi Cast 3 i 10 g 7.2 a o 5 s g 6 In h I: g 4 3 g 2 0 ° 2° 405° 30100120140150130 0 204060 80100120140160180200 C. Standard Housing D. Enriched Environment 15 0 § i 3 a m a: 3 to E 0 3-10 a _ g 3 fl = J: a: o .15 5 -20 Fig. 10. Enhanced arborization ot dendrites is use-dependent. A. Increased branching of basilar dendrites in Layer v pyramidal neu- rons in the homotopic area of the opposite hemisphere at 13 days after a lesion to the forelimb representation region of the eensorimotor cortex. Concentric circles. at 20 pm increments in diameter, were superimposed onto projections of the dendritic field, and the number of arbon'circle Intersections was counted as a function of distance from the neuronal soma. Sham-operated animals were compared to brain injured animals. Branching appeared only in animals permitted to freely use the forelimb corresponding to the enhanced arborization. B, Dendritic growth and functional improvement were prevented by fitting the animals with one-sleeve casts that immobilized the nonimpaired (intact) forelimb. Sham-operated animals fitted with one sleeve casts showed no enhancement of branching. which suggests that brain injury may initiate neural events that provide an opportunity for behavioral experience to promote maximal neural plasticity (see ref. 58}. (Data taken from Fig. 2 of ref. 129.) C. D, Comparison of functional deficits in rats with chronic spinal cord injury maintained in standard cages (C) or In an enriched environment {0). After receiving a spinal cord Injury using a weight drop device (10 g, 2.5 cm), 3 months after Intury, the rats were placed in standard cages (C) or In an enriched environment (D) for 2 weeks. Changes in the functional deficits were measured with tests that form a combined behavioral score (see ref. 130}. The enriched environment produced functional Improvement compared to the standard environment. (Date modified from ref. 55.) diffuse across the lesion site. and their stimulation elic- its the same behaviors as are elicited in unoperated ani- mals (54). Possible Therapeutic Use of PEG Solutions and PEG Hydrogels The data described in the previous section suggest that PEG solutions might also repair severed mammalian CNS or PNS axons in vivo and, if necessary, that PEG hydrogels could be used to provide mechanical strength 104 THE NEUROSCIENTIST at the lesion site to maintain axolemmal integrity of PEG-fused axons. Lore et a]. (54) have considered four problems that need to be solved for the successful in vivo therapeutic use of PEG to repair bundles of mam- malian PNS axons (e.g., sciatic or other peripheral nerves) or CNS axons (e.g.. corticospinal or other spinal tracts): First. severed distal stumps need to be induced to survive until they can be PEG-fused. At present, severed mammalian CNS or PNS axons can be induced to survive for days or Degeneration, Trophlc Interactions, and Repair of Severed Axons weeks by cooling (34, 35, 70), injections of cyclosporin A (Fig. 1A—C; 36), or antibodies to complement 3 receptors (33). Second, our PEG hydrogel (54) or some other tissue adherent needs to be applied to the site of PEG fusion to provide mechanical strength in vivo. Third, microsurgical procedures need to be devised to bring cut nerve ends into tight apposition (50, 54) because PEG solutions can only fuse membranes separated by a thin layer of bound water (Box 2). Fourth, if partial recovery of behavior is an acceptable goal, then the transected ends of a bundle of axons do not have to be perfectly aligned because the survival or regenera- tion of only 5% to 10% of CNS axons in mammals pro- duces substantial behavioral recovery (12, 114). Axons severed by crush injuries pose less of a problem in that the crushed ends often remain aligned (115). Rehabilitative Training in Enriched Environments Is an Underutilized Therapeutic Strategy to Produce Significant Recovery of Important Behaviors After CNS injury, neural events important for functional outcome can be shaped by experience. Just as in devel- opment of sensory perceptions and motor behaviors, maximum improvement following CNS injury must in- volve use-dependent changes in many brain structures (116, 117), including tissues that are not directly dam- aged (118—120). Thus, the nervous system has widely distributed self-integrating mechanisms that are exqui- sitely responsive to environmental and behavioral de- mands. Sensory feedback, limb use (Fig. 10A and 10B), enrichment (Fig. 10D), and other rehabilitative activities subsequent to CNS damage may produce substantial be- havioral improvement (121—123) by interacting with in- jury-induced factors such as changes in levels of neurotrophic or neurogenic factors, neurotransmitters, and alterations of CNS connections (124—128). For ex- ample, sensorimotor feedback to higher control centers might gradually recalibrate selected inputs and outputs and modulate the firing patterns of a small number of nonspecifically reconnected axons in a beneficial way. (Unlike a cut and respliced computer cable, the CNS is capable of remarkable rewiring and reinterpretation of abnormal signals.) Given CNS plasticities outlined above, rehabilitative training in enriched environments (a social environment where there is free access to novel items, including exercise-inducing equipment) might induce some recov— ery of behavioral function even if severed axons are not repaired. In fact, rats with a moderate degree of chronic spinal cord injury placed in an enriched environment show significantly more behavioral recovery compared to rats maintained in standard caging (55, 56, 128, 131). Even greater improvement is attained if the enriched environment is combined with exercise (132, 133). Yet greater improvement might be expected following sev- erance of CNS axons if some axonal connections regen- erate by outgrowth or if some axonal ends are PEG—fused, even if the new connections are not those Volume 6, Number 2, 2000 made by any axon that existed prior to the lesion. Fur- thermore, following CNS damage, complete disuse (e.g., by restraining the impaired or non-impaired extremity) can prevent expression of injury-evoked endogenous trophic factors, neuritic outgrowth, synaptogenesis, and functional improvement, each of which may depend on critical periods for optimal recov- ery (58, 117, 120, 134). On the other hand, extreme rehabilitative training too soon after injury can damage vulnerable tissue and have an adverse effect on recovery (135). Given such data, perhaps initial training should be moderate, followed by incremental increases in intensity, as sublethally damaged tissue recovers and can sustain more extensive activity (57). In any case, the optimal regimen needs to be based on behavioral, ana- tomical, and other markers. A Combination of Strategies Is Likely to Produce the Best Therapeutic Approach Of all types of injuries in mammals, the most difficult to repair is damage to nervous tissue, especially in the cen- tral nervous system (brain and spinal cord). Until about 20 years ago, essentially nothing could be done to repair CNS lesions and almost no experimental data provided much hope for developing effective therapeutic strate- gies. While no treatment currently in use is specifically directed at repair of damage to CNS axons or cell bod- ies, this may change in a few years. Most strategies cur- rently under consideration assume that repair of severed CNS axons requires the induction of slow regeneration of neurites from surviving proximal stumps (or trans- planted cells to slowly grow out neurites) that may rees- tablish some of the severed connections (Box 1). For example, transplants of Schwann cells (18), olfactory sheath cells (136), genetically engineered cell lines (22), fetal tissues (21), growth guides (137), fibrin or other extracellular matrix materials (138, 139), trophic factors (23, 24), reduction of the basal membrane (140), and antibodies to oligodendritic inhibitors of axonal out- growth (20) can all induce neurite outgrowth. We do not doubt the potential usefulness of strategies based on a commonly assumed set of interrelated assumptions (Box 1), even if those assumptions “ain’t necessarily so.” Rather, our alternate set of assumptions suggests that at least two novel or relatively underutilized approaches should also be added to the above list of potential therapeutic strategies, (i) rapid rejoining (fusion) of severed proximal and distal segments induced to survive until a fusiogen can be applied and the lesion site stabilized with hydrogels or other tissue adherents (54) and (ii) the use of exercise and other enriched-environment strategies (55, 58, 120, 116—135). In fact, significant advances in the repair of CNS lesions may not come dramatically by the use of a single “molecular magic bullet,” but rather may come incrementally by the continuously improved use of vari- ous combinations of many currently available strategies. For example, cortical spinal neurons may respond best to the trophic factor NT—3, and rubrospinal axons to THE NEUROSCIENTIST 105 BDNF; inhibition of the outgrowth of some spinal axons may be reduced primarily by antibodies to oligodendritic proteins, and others primarily by antibod- ies to the basal lamina. As a result of such combined therapies, perhaps we will soon be in the situation where a physician in conference with the family of a patient with severe spinal cord or head injury will no longer have the problem immortalized in the title of another song from Porgy and Bess by Gershwin: “I Got Plenty o’ Nuttin’.” References 10. ll. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. . Peters A, Palay S, Webster H de F. The fine structure of the ner- vous system. 2nd ed. New York: University Press; 1991. Ramon y Cajal S. Degeneration and regeneration of the nervous system. London: Oxford University Press; 1928. . Kandel ER, Schwartz JH, editors. Principles of neural science. Amsterdam: Elsevier/North Holland; 1985. . Waller A. Experiments on the section of glossopharyngeal and hypoglossal nerves of the frog, and observations of the alter- ations produced thereby in the structure of their primitive fibres. Philos Trans 1850;140:423-9. Alberts B, Bray D, Lewis J, Raff M, Roberts K, Watson JD. The cell. 3rd ed. New York: Garland; 1994. . Krause TL, Fishman HM, Ballinger ML, Bittner GD. Extent and mechanism of sealing in transected giant axons of squid and earthworms. J Neurosci 1994;14:6638—51. . Eddleman CS, Ballinger ML, Smyers ME, Godell CM, Fishman HM, Bittner GD. Repair of plasmalemmal lesions by vesicles. Proc Natl Acad Sci U S A 1997;94:4745—50. . Eddleman CS, Ballinger ML, Smyers ME, Fishman HM, Bittner GD. Endocytotic formation of vesicles and other membranous structures induced by Ca” and axoplasmic injury. J Neurosci 1998;18:4029—41. . Bittner GD, Fishman HM. Axonal sealing following injury. In: Murray M, Ingoglia, N, editors. Nerve regeneration. New York: Marcel Dekker; 2000 (in press). Xie X, Barret JN. Membrane rescaling in cultured rat septal neu— rons after neurite transection: evidence for enhancement by Can-triggered protease activity and cytoskeletal disassembly. J Neurosci 1991;11:3257—67. Spira ME, Banbassat D, Dormann A. Resealing of the proximal and distal cut ends of transected axons: electrophysiological and structural analysis. J Neurobiol 1993;24:300—16. Das GG, Wallace RB. Neural transplantation and regeneration. New York: Springer; 1986. Schwab ME, Bartholdi D. Degeneration and regeneration of ax- ons in the lesioned spinal cord. Physiol Rev 1996;76:319—70. Bittner GD. Degeneration and regeneration in crustacean neuromuscular systems. Am Zool 1973;13:379-408. Bittner GD. Trophic interactions of crustacean neurons. In: Hoyle G, editor, in honor of Professor CAS Wiersma. Identified neurons and behavior. New York: Plenum; 1977. p. 507—32. Bittner GD. Long term survival of anucleate axons and its impli- cations for nerve regeneration. Trends Neurosci 1991;14:188—93. Beazley L. Central nerve regeneration in amphibians and rep- tiles. In: Murray M, lngoglia, N, editors. Nerve regeneration. New York: Marcel Dekker; 2000 (in press). David S, Aguayo AJ. Axonal elongation into peripheral nervous system “bridges” after central nervous system injury in adult rats. Science 1981;241:931—3. Murray M. Strategies and mechanisms of recovery after spinal cord injury. Adv Neurol 1997;72:219—25. Schnell L, Schwab ME. Axonal regeneration in the rat spinal cord produced by an antibody against myelin-associated neurite growth inhibitors. Nature 1990;343:269—72. Giovanini MA, Reier PJ, Eskin TA, \Vrrth E, Anderson DK. Characteristics of human fetal spinal cord grafts in the adult rat 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. spinal cord: influences of lesion and grafting conditions. Exp Neurol 1997;148:523—43. Liu Y, Kim D, Himes BT, Chow SY, Schallert T, Murray M, Tessler A, Fischer 1. Transplants of fibroblasts genetically modi- fied to express BDNF promote regeneration of adult rat rubrospinal axons and recovery of forelimb function. J Neurosci 1999;19:4370—87. Kim D, Adipudi V, Shibayama M, Giszter S, Tessler A, Murray M, Simansky KJ. Direct agonists for serotonin (5-HT2) recep- tors enhance locomotor function in rats that received neural transplants after neonatal spinal transection. J Neurosci 1999;6213—24. Olson L. Regeneration in the adult central nervous system: ex- perimental repair strategies. Nat Med 1997;12:1329—35. Bittner GD. Long term survival of severed distal axonal stumps in vertebrates and invertebrates. Am Zool 1988;28:1165—79. Hoy RR, Bittner GD, Kennedy D. Regeneration in crustacean motoneurons: evidence for axonal fusion. Science 1967;156: 251—2. Ballinger ML, Bittner GD. Ultrastructural studies of severed me- dial giant and other CNS axons in crayfish. Cell Tissue Res 1980;208:123—33. Bittner GD, Ballinger ML. Ultrastructural changes at gap junc- tions between lesioned crayfish axons. Cell Tissue Res 1980;207:143—53. Deriemer SA, Elliot El, Macagno ER, Muller KJ. Morphologi- cal evidence that regenerating axons can fuse with severed axon segments. Brain Res 1983;272:157—61. Birse SC, Bittner GD. 1981. Regeneration of earthworm giant axons following transection or ablation. J Neurophysiol 1981;45:724—42. Zotolli SJ, Marek L, Agostini M, Strittmatter S. Morphological and physiological survival of goldfish Mauthner axons isolated from their somata by spinal cord crush. J Comp Neurol 1987;255:272—82. Blundon JA, Sheller RA, Moehlenbruck JW, Bittner GD. Effect of temperature on long-term survival of anucleate giant axons in crayfish and goldfish. J Comp Neurol 1990;297:377—91. Lunn ER, Perry VH, Brown MC, Rosen H, Gordon S. Absence of Wallerian degeneration does not hinder regeneration in pe- ripheral nerve. Eur J Neurosci 1989;1:27—33. Tsao JW, George EB, Grifin JW. Temperature modulation re- veals three distinct stages of Wallerian degeneration. J Neurosci 1999;19:4718—26. Sea T, Ballinger ML, Bittner GD. Cooling of peripheral myelinated axons retards Wallerian degeneration. Exp Neurol 1995;133:85—95. Sunio A, Bittner GD. Cyclosporin retards the Wallerian degener- ation of peripheral mammalian axons. Exp Neurol 1997;146: 46—56. Gainer I-I, Tasaki I, Lasek RJ. Evidence for the glia-neuron pro- tein transfer hypothesis from intracellular perfusion studies of squid giant axons. J Cell Biol 1977;74:524—30. Lasek RJ, Gainer H, Barker JL. Cell-to-cell transfer of glia pro- teins to the squid giant axon. The glia-neuron transfer hypothe- sis. J Cell Biol 1977;74:501—23. Koenig E, Giuditta A. Protein-synthesizing machinery in the axon compartment. Neurosci 1999;89:5—15. Sheller RA, Bittner GD. Maintenance and synthesis of proteins for an anucleate axon. Brain Res 1992;580:68—80. Nixon RA, Logvinenko KB. Multiple fates of newly synthesized neurofilament proteins; evidence for a stationary neurofilament network of retinal ganglion cell neurons. J Cell Biol 1986;14:289—98. Raabe TD, Nguyen T, Bittner GD. Calcium activated proteolysis of neurofilament proteins in goldfish Mauthner axons. J Neurobiol 1995 :26 :253—61. Raabe TD, Bittner GD. Phosphorylation of neurofilament pro- teins in isolated goldfish Mauthner axoplasm. J Neurochem 1996;66:1214—21. Raabe TD, Nguyen T, Archer C, Bittner GD. Mechanisms for the maintenance and eventual degradation of neurofilament pro- / 106 THE NEUROSCIENTIST Degeneration, Trophic Interactions, and Repair of Severed Axons ...
View Full Document

This note was uploaded on 10/03/2009 for the course BIO 53665 taught by Professor Iverson during the Fall '07 term at University of Texas at Austin.

Page1 / 19

degen - l REVIEW Degeneration, Trophic Interactions, and...

This preview shows document pages 1 - 19. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online