- Email: [email protected]
Axonal regeneration stimulated by erythropoietin: An experimental study in rats
Journal of Neuroscience Methods 164 (2007) 107–115
Axonal regeneration stimulated by erythropoietin: An experimental study in rats Marios G. Lykissas a,∗ , Ekaterini Sakellariou b,1 , Marios D. Vekris a,1 , Vasilios A. Kontogeorgakos a,1 , Anna K. Batistatou c,2 , Gregory I. Mitsionis a,1 , Alexandros E. Beris a,1 a
Department of Orthopaedic Surgery, University of Ioannina, School of Medicine, Ioannina, P.C. 45110, Greece b Department of Plastic Surgery, Thriasio General Hospital, Athens, Greece c Department of Pathology, University of Ioannina, School of Medicine, Ioannina, Greece Received 14 March 2007; received in revised form 10 April 2007; accepted 10 April 2007
Abstract The aim of the present study is to evaluate the effects of erythropoietin to the collateral sprouting by using systemically delivered erythropoietin in an end-to-side nerve repair model. Forty-five rats were evaluated in four groups: (A) end-to-side neurorrhaphy only, (B) end-to-side neurorrhaphy and erythropoietin administration, (C) end-to-end neurorrhaphy and (D) nerve stumps buried into neighboring muscles. In all animals, the contralateral healthy side served as control. Functional assessment of nerve regeneration was performed at intervals up to 5 months using the Peroneal Function Index. Evaluation 150 days after surgery included peroneal and tibial nerve morphometric examination, and wet weights of the tibialis anterior muscle. During the first three weeks after surgery, when erythropoietin was regularly administered, functional evaluation showed that erythropoietin may facilitate peripheral nerve regeneration. However, there was rapid deterioration in the functional recovery when erythropoietin’s administration was discontinued. As a consequence, at the end of this study, erythropoietin failed to maintain its initial stimulating effect in axonal regeneration. The results of wet muscle weights revealed statistically significant differences between Groups A and C, and Group B. Furthermore, data on axonal counting showed significant difference between Groups A and C, and Group B. Erythropoietin appears to facilitate peripheral nerve regeneration at the initial phase of its administration. Further investigation will be necessary to optimise the conditions (dose, mode of administration) in order to maintain its effects. © 2007 Elsevier B.V. All rights reserved. Keywords: Erythropoietin; End-to-side neurorrhaphy; Nerve regeneration; Collateral sprouting
1. Introduction Erythropoietin (Epo) is a glycoprotein with a molecular weight of 30.400 Da that controls the production of red cells. Epo in humans is synthesized by renal peritubular cells in adults and by hepatic cells in the fetus, whereas a small amount is also synthesized in the adult’s liver (Jacobson et al., 1957; Fisher and
∗
Corresponding author. Tel.: +30 26510 97472; fax: +30 26510 34816. E-mail addresses: [email protected] (M.G. Lykissas), [email protected] (E. Sakellariou), [email protected] (M.D. Vekris), [email protected] (V.A. Kontogeorgakos), [email protected] (A.K. Batistatou), [email protected] (G.I. Mitsionis), [email protected] (A.E. Beris). 1 Tel.: +30 26510 97472; fax: +30 26510 34816. 2 Tel.: +30 26510 99415; fax: +30 26510 34816. 0165-0270/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jneumeth.2007.04.008
Birdwell, 1961; Erslev, 1974). Epo binds to its receptor (EpoR) on the red cell surface, which consists of a p66 protein that undergoes dimerization after activation, and two additional proteins (Livuah et al., 1999; Remy et al., 1999). Erythropoietin, as well as its receptor are present in the central nervous system, including the retina (Marti et al., 1996; Juul et al., 1998; Sir´en et al., 2001; Celik et al., 2002; Bocker-Meffert et al., 2002; Grimm et al., 2002). It has been proven that Epo is produced in the bodies and the axons of the normal ganglion of the rat’s dorsal root and is increased in Schwann cells after peripheral nerve injury (Campana and Myers, 2001). Epo-R is found on certain axons and neuron bodies of the dorsal root ganglion, on endothelial cells and Schwann cells of the normal peripheral nerve. Inside the dorsal root ganglion, the neuron type that expresses the Epo-R represents 40% of the ganglion cells. Moreover, systemic administration of rHuEpo (recombinant human
108
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
erythropoietin) reduces the apoptosis of the dorsal root ganglion cells and contributes to the recovery of mechanical allodynia (Campana and Myers, 2003). Epo’s neuroprotective effect is in fact synergic with that of insulin-like growth factor I (IGF-I). Thus, the combination of Epo and IGF-I exhibits neuroprotective effect with lower doses than each cytokine separately does (Digicaylioglu et al., 2004). Epo’s role as a neuroprotective agent is considered to be multifactorial, with a direct and an indirect action on nerve cells. Epo, in vitro, protects the nerve cells from hypoxia-induced glutamate toxicity, which is the main cause of nerve cells death as regards hypoxia (Morishita et al., 1997; Buemi et al., 2003). Besides, in a multiple sclerosis model, the systemic administration of erythropoietin alpha reduces the immune response and the inflammatory reaction, while it enhances nerve recovery after a spinal cord trauma (Agnello et al., 2002; Brines, 2002; Gorio et al., 2002). Epo is produced in the bodies and the axons of the normal ganglion of the rat’s dorsal root and is increased in Schwann cells after peripheral nerve injury (chronic constriction injury, CCI) (Campana and Myers, 2001). Furthermore, like in the central nervous system, in the peripheral as well, Epo presents anti-apoptotic after injury action. Although Epo’s anti-apoptotic mechanism has not been completely elucidated, it has been shown that Epo prevents apoptosis of CNS neurons by triggering cross-talk between Janus-tyrosine-kinase-2 (JAK2) and nuclear factor kappa B (NF-kB) signaling cascades (Digicaylioglou and Lipton, 2001). Moreover, Epo presents not only anti-apoptotic properties but also prevents axonal degeneration and enhances neurogenesis (Keswani et al., 2004; Tsai et al., 2006). Consequently, Schwann cell derived Epo activates an “axonoprotective” pathway. Alongside, Epo’s action on Schwann cells and on the inflammatory response after neurological trauma, may be effective in peripheral nerve regeneration after neurorrhaphy. Since the last 20 years end-to-side neurorrhaphy has been added to the surgical options, although this technique was already described at the beginning of the last century. Recent evidence suggests that after end-to-side neurorrhaphy nerve regeneration occurs by collateral sprouting (Edds, 1953; Slack et al., 1979; Hopkins and Slack, 1981; Zhang et al., 2006). It is certain that a great number of factors contribute to collateral sprouting, but the mechanism remains mostly unknown. Various growth factors have promoting effect in nerve regeneration and they are well documented for many different models of central or peripheral nervous system injury. However, the factors that stimulate collateral sprouting are listed only in few references in the literature (Caroni and Gramdes, 1990; Liuzzi and Tedeschi, 1991). Experimental work has demonstrated that Epo protects from glutamate-induced cell death in vitro (Digicaylioglou and Lipton, 2001) and in vivo (Brines et al., 2000), stimulating at the same time intracellular anti-apoptotic metabolic pathways. Systemic administration of Epo may inhibit the apoptosis of sensory and motor neurons after neurological damage. Epo’s action on Schwann cells and on the inflammatory response after neurological trauma, may be effective in peripheral nerve regeneration after end-to-side neurorrhaphy.
2. Material and methods The procedures for this investigation were performed according to protocols approved by the Official Veterinary Services, General Directorate of Development, Prefecture of Ioannina (ref no. 2579). Forty-five male Wistar rats, weighting between 190 and 310 g, were used in the present study. The animals were divided in four groups according to the operative procedure: Group A: end-to-side repair (n = 12), Group B: end-to-side repair and Epo administration (n = 12), Group C: end-to-end nerve repair (n = 12), Group D: segmental nerve resection (n = 9). 2.1. Surgical procedure Just before the operation, all animals were individually numbered and given preconditioning trials on an 8 × 62 cm walking track, for comparison with normative data (Dellon and Dellon, 1991). Under anesthesia with a single 5 mg/kg intraperitoneal ketamin injection (Narketan 10, 100 mg/ml, Vetoquinolag, Switzerland), the right hind limb of all animals was used as the experimental limb, and the left hind limb as the control limb. The right sciatic, tibial and peroneal nerves were exposed and dissected through a semitendinous-biceps femoris muscle splitting incision, under magnification of an operating microscope. The common peroneal nerve was transversely divided, and the distal and the proximal stump were immediately repaired according to grouping. In Group A, the proximal peroneal nerve stump was sutured to the trunk of the intact tibial nerve, in a terminolateral fashion described by Viterbo et al. (1994). The distal segment was repaired end-to-side, 1.2 cm distal of the first end-to-side neurorrhaphy. In Group B (stimulate nerve regeneration with erythropoietin), the proximal and the distal stump of the severed peroneal nerve were sutured end-to-side to the trunk of the intact tibial nerve, as in Group A. In both Groups A and B, an epineural window was made on the tibial trunk before each end-to-side nerve coaptation took place. In Group C (primary repair), the peroneal nerve was repaired in a classic end-to-end fashion. In Groups A, B and C, the neurorrhaphies were performed under microscope magnification, using three epineural 10/0 nylon interrupted sutures, placed at 120◦ intervals. In Group D (denervated control), both proximal and distal nerve stumps were buried into the neighboring muscles and fixed with a single 8/0 nylon suture. In all groups wound closure was achieved by 4/0 vicryl interrupted sutures for the muscles, and 3/0 nylon interrupted sutures for the skin. Before and after the operation, the animals were kept in individual cages and maintained on standard rat chow and water ad libitum, with a 12-h light:12-h dark cycle. 2.2. Erythropoietin administration In Group B, subcutaneous administration of rHuEpo was chosen, as it is more effective than i.v. or i.p. administration (Kaufman et al., 1998). Nevertheless, rHuEpo injected subcuta-
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
neously has a half-life greater than 24 h compared with the 4 h half-life of rHuEpo injected intravenously (Egrie et al., 1988). According to several studies, rHuEpo preconditioning is more effective in preventing dorsal root ganglion apoptosis following peripheral nerve injury and neuron apoptosis during brain ischaemia (Morishita et al., 1997; Digicaylioglou and Lipton, 2001; Campana and Myers, 2003). Thus, in this report pretreatment with rHuEpo 1 day before the operation was followed. The rHuEpo administration was repeated every 24 h for the first 20 postoperative days. A middle dose of 2680 units/kg bw of rHuEpo was administered daily, as the effectiveness of a higher dose of 5000 units/kg bw was no better than the middle dose, while a low dose of 1000 units/kg bw was no effective (Campana and Myers, 2003; Sekiguchi et al., 2003). 2.3. Walking track analysis At 15, 30, 60, 90 and 150 days after the operation, all animals were analyzed using an 8 × 62 cm walking track darkened at one end (De Medinaceli et al., 1982). The paper strips containing the footprints were copied, the digitized footprints were analyzed, and the peroneal function index (PFI) was automatically calculated. The parameters measured in the footprints were print length (PL) and toe spread (TS), both in the normal (N) and experimental (E) pawprints. All measurements and PFI calculation were made by two blinded to the identity of the digitized footprints investigators at the end of the study according to the formula of Bain et al. (1989): PFI = 174.9
(EPL − NPL) (ETS − NTS) + 80.3 − 13.4 NPL NTS
109
Inc, Munich, Germany) connected with a computer using the public domain software NIH Image (http://rsb.info.nih.gov/nihimage/). Quantitative evaluation followed with myelinated nerve fibers counting in all nerve segments (axons/mm2 ). Morphometric analysis also included the calculation of percent neural tissue (100 × neural area/intrafascicular area). The same parameters were determined for the experimental limb, as well as the control limb. The final results were expressed as the ratio of the experimental side to the contralateral unoperated side. All calculations were made by a single blinded investigator. 2.5. Muscle weights On the 150th postoperative day, tibialis anterior muscle was harvested from the experimental and the contralateral control side. Wet muscle weights were measured and reported as the ratio of the experimental to the contralateral side. Afterwards, specimens from each muscle were immersed in 10% formalin and embedded in paraffin for histologic analysis. 2.6. Statistical analysis Peroneal function index, muscle mass ratios, and nerve fiber density values were analyzed by using the non-parametric Kruskal–Wallis and Mann–Whitney U tests. All tests were calculated with use of the SPSS, version 13.0 (SPSS Inc., Chicago, IL, USA) statistic package for personal computers. In all instances, p < 0.05 was regarded as statistically significant. 3. Results
2.4. Morphometric studies
3.1. Walking track analysis
Under the same anesthesia on the 150th day, the experimental side was exposed through a posterolateral approach to the thigh and peroneal and tibial nerves were dissected and removed. Afterwards, the animals were killed with a lethal dose of intraperitoneal sodium pentobarbital (100 mg/kg). In Groups A and B 0.5 cm-long sections were removed from the tibial nerve midway between the proximal and the distal endto-side neurorrhaphy, the distal peroneal nerve segment with the distal neurorrhaphy zone, and the tibial nerve below the site of neurorrhaphies. In Group C, section 0.5 cm in length of the peroneal nerve was removed distal to the site of end-to-end neurorrhaphy. In Group D, a 0.5 cm-long segment of the distal peroneal stump was harvested. In all animals, corresponding sections from the contralateral control sides were harvested. All sections were immersed in a solution of 2.5% glutaraldehyde buffered with cacodylate, washed in sodium cacodylated buffer (0.2 mol/L, pH 7.4), post-fixed in 1% osmium tetroxide, dehydrated in an ethyl alcohol solution of growing concentrations (50, 70, 95 and 100%), and embedded in epoxy resin for axonal counting. Cross-sections 1 m thick were cut using a Reichert-Jung ultracut E with a diamond knife (Reichert-Jung, Weiss, Austria) from the nerve segments, stained with toluidine blue, and examined with a light microscope (Axioscop, Zeiss,
Two hundred seventy paper strips containing at least six clearly marked footprints each, were used to evaluate peroneal nerve functional recovery. Analysis of walking tracks was done in all animals preoperatively, and at 15, 30, 60, 90 and 150 days, according to the formula of Bain et al. (1989) (Table 1). The preoperative PFI value was never 0, but oscillated between −22.36 and −7.57. Before the surgical procedure no significant difference obtained between the four groups (p = 0.739). Such was an essential condition to obtain intraanimal control data. Furthermore, PFI values carried out on subjects with the same age, weight and sex, in order to be comparable in many parameters (Dellon and Dellon, 1991). On the 15th postoperative day, significant difference was detected between Groups A and B (p < 0.001), with Group B revealing better results (Fig. 1). PFI values were not significantly different between Groups B and C (p = 0.184). On the 2nd postoperative evaluation, significant differences were recorded between Groups A and B (p < 0.001), as well as Groups B and C (p = 0.002), with Group B having better functional outcome. The 3rd evaluation on the 60th day after surgery revealed reversed significant differences between Groups A and B (p < 0.013), with Group B showing functional regression. Significant differences were also found to be reversed between Groups B and
110
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
Table 1 Average peroneal function index scores Group A Pre-op D 15 D 30 D 60 D 90 D 150
−14.05 −41.66 −36.63 −28.09 −19.93 −16.84
B ± ± ± ± ± ±
3.35 4.24 3.50 4.39 4.69 3.42
−13.32 −33.91 −23.17 −32.86 −29.93 −28.15
C ± ± ± ± ± ±
4.64 3.40 2.75 4.49 9.57 11.55
−14.33 −35.41 −28.41 −22.92 −18.01 −13.94
D ± ± ± ± ± ±
−14.35 −46.78 −46.89 −49.69 −62.81 −71.32
2.69 2.12 4.16 3.62 2.49 2.68
± ± ± ± ± ±
4.64 8.19 5.89 5.69 2.55 3.21
Average PFI scores in Groups A, B, C and D obtained at 0, 15, 30, 60, 90 and 150 days. All values are expressed as mean ± standard deviation.
C (p < 0.001). On the 90th postoperative day, difference between Groups A and C (p = 0.355) became insignificant, while differences between both Groups A and C compared with Group B remain significant (p < 0.05). The end scores at 150th day, were −16.84 ± 3.42, −28.15 ± 11.55 and −13.94 ± 2.68 for Groups A, B and C, respectively. Statistically, differences between all groups were significant (p < 0.05). 3.2. Morphometric studies At the end of the study, the ratios of peroneal nerve fiber density (axons/mm2 ) for Groups A, B and C were calculated (Table 2). Severe axonal degeneration was present in Group D. Morphometric evaluation could not be performed in this group. Morphometric evaluation demonstrated that there was statistically significant difference between Groups A and B (1.57 ± 0.62 and 0.65 ± 0.19, respectively), as well as between Groups B and C (0.65 ± 0.19 and 1.95 ± 0.46, respectively), with Group B having less sufficient results (p < 0.001). Statistically, the difference between Group A and Group C was insignificant (p = 0.083). The percentage neural tissue data is demonstrated in Table 3. In the end-to-side nerve repair group, the ratio of percentage
Table 2 Peroneal nerve fibre density Group
Axon density (n/mm2 )
Ratio of axon density
A B C
13,900 ± 2,900 (9,400–18,900) 13,100 ± 4,200 (5,700–19,000) 7,700 ± 4,200 (1,800–18,400)
1.57 ± 0.62 (0.67–2.68) 0.65 ± 0.19 (0.28–0.94) 1.95 ± 0.46 (0.97–2.47)
Average peroneal nerve fiber density reported as the ratio of the experimental to the contralateral side on 150th postoperative day. All values are expressed as mean ± standard deviation. Significant differences were noted between Groups A and B as well as Groups B and C both in axon density and axon density ratio (p < 0.05). Moreover, differences between Group A and C were insignificant (p > 0.05).
peroneal neural tissue of the experimental to the unoperated control side was 0.72 ± 0.17. For the Epo group, the ratio was 0.35 ± 0.01. These differences were statistically significant at p < 0.001. Moreover, the ratio of percentage neural tissue was significantly reduced in Epo group when compared with the endto-end nerve repair group (p < 0.001). No significant differences were noted between Groups A and C (p = 0.204). Macroscopic examination of all nerve segments was in accordance with quantity histomorphometry of nerves (Fig. 2). In Groups A and B, there was no histologic evidence of injury to the donor nerve (tibial nerve) after the intact nerve bridge procedure. 3.3. Muscle weights The tibialis anterior muscle, innervated by the peroneal nerve, was weighted (Table 4). Wet muscle weights were reported as the Table 3 The percentage neural tissue data
Fig. 1. Diagram of PFI values behaviour according to grouping.
Group
Percentage neural tissue (%)
Ratio of neural tissue
A B C
35.98 ± 10.81 (19.70–52.30) 17.84 ± 5.12 (12.1–29.4) 48.86 ± 5.48 (37.9–54.7)
0.72 ± 0.17 (0.42–0.94) 0.35 ± 0.01 (0.25–0.54) 0.81 ± 0.01 (0.67–0.89)
The percentage neural tissue data of peroneal nerve reported as the ratio of the experimental to the contralateral side. All values are expressed as mean ± standard deviation. Significant differences were noted between Groups A and B as well as Groups B and C both in percentage neural tissue and percentage neural tissue ratio (p < 0.001). At the same time, difference between Groups A and C in percentage neural tissue ratio was insignificant (p = 0.204). However, difference between Groups A and C in percentage neural tissue was significant at p < 0.05.
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
111
Fig. 2. Transverse section of the distal part of the operated peroneal nerve 150 days postoperatively (toluidine blue, original magnification ×400). (A) Histologic sections in Group A revealed viable axons distal to the coaptation site. Axonal regeneration was similar to that obtained in end-to-end control group (Group C). However, there was lower number of axons in Group A compared with Group C. (B) In Group B histologic sections demonstrated few myelinated axons surrounded by connective tissue. (C) In end-to-end group (Group C), a normal pattern of nerve regeneration was present. (D) Wallerian degeneration was obtained in the distal stump of the unrepaired peroneal nerve in Group D.
ratio of the experimental to the contralateral control side (Fig. 3). The ratios were 0.80 ± 0.06, 0.56 ± 0.19 and 0.81 ± 0.09 for Groups A, B and C respectively. Statistically, the difference between Groups A and C was insignificant (p = 0.817). In some cases, muscle weights in the experimental side were almost the same with control side. On the other hand, the results revealed statistically significant differences between Groups A and B (p = 0.003), and Groups B and C (p = 0.002). Moreover, statistical difference was seen between the three experimental groups and Group D, in which the peroneal nerve was transected and buried into the neighboring muscles to prevent reinnervation (p < 0.001). Macroscopically, in Groups A and C there was only minimal atrophy in the tibialis anterior muscles of the experimental limb
(Fig. 4). However, in Groups B and D tibialis anterior muscle atrophy was obvious at the experimental side when compared with the normal contralateral side. It has been proven that the relative loss of tibialis anterior muscle weight closely reflects the degree of denervation (Trumble, 1992).
Table 4 Average wet muscle weights Group
Mean ± S.D. (m)
A B C D
0.80 0.56 0.81 0.19
± ± ± ±
0.06 0.19 0.09 0.04
Minimum
Maximum
0.71 0.31 0.68 0.13
0.92 0.84 0.97 0.28
Average wet muscle weights reported as the ratio of the experimental to the contralateral normal side on 150th postoperative day. All values are expressed as mean ± standard deviation.
Fig. 3. Tibialis anterior wet muscle weights expressed as the ratio of the experimental side to contralateral normal side. Data are shown as mean ± S.D.
112
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
Fig. 4. (A–D) Macroscopic appearance of the tibialis anterior muscles of control (left on each picture) and experimental (right on each picture) limbs. There was minimal atrophy in muscles of the experimental limbs in Groups A and C. In contrast, in Groups B and D the atrophy was evident.
3.4. Muscle histology
4. Discussion
The tibialis anterior muscles of the experimental and the contralateral control side were immersed in 10% formalin, embedded in paraffin, stained with hematoxylin-eosin, and evaluated histologically (Fig. 5). There was no significant difference between Groups A and C for the histologic appearance of the tibialis anterior muscles of the experimental limbs, which furthermore had a great similarity with the unoperated control limbs. In these groups transverse sections of muscle fascicles showed polygonal myofibers with mostly subsarcolemmal nuclei and little intervening endomysial connective tissue (Fig. 5A2 and C2). However, histologic examination of tibialis anterior muscles in Group B revealed focally small groups of atrophied fibers with increased endomysial connective tissue (Fig. 5B2). In Group D, transverse sections revealed small bundles of atrophic fibers with small diameter and centrally placed nuclei mingled with connective and adipose tissue (Fig. 5D2).
Prerequisite of functional recovery in an end-to-side nerve model is nerve regeneration as well as a successful muscle reinnervation. After peripheral nerve injury, the degree of motor recovery depends on the number of regenerated axons and the time length of muscle denervation. The longer the interval between muscle denervation and reinnervation, the poorer the degree of functional recovery (Kobayashi et al., 1997). It has been demonstrated that prolonged motor axonotmesis results in dramatic reductions in motor unit number (Fu and Gordon, 1995). This observation defines the presence of a “time window” after which the level of functional recovery is insufficient. According to experimental studies, in primates this time period does not appears to extend more than 100 days after nerve injury (Krarup et al., 2002). Consequently, any mechanism which decrease denervation time may contribute to a better functional result. In the present study, Epo administration as a potential
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
Fig. 5. Histologic appearance of the experimental tibialis anterior muscles in Groups A (A2) and C (C2) was almost normal (hematoxylin–eosin, ×100). In Groups B (B2) and D (D2), experimental tibialis anterior muscles were atrophic with small, angulated fibers. Tibialis anterior muscles of unoperated limbs in Groups A (A1), B (B1), C (C1) and D (D1) were used as control.
stimulating factor of nerve regeneration is investigated. If Epo provides early recovery of function after peripheral nerve injury and repair, its use may yield interesting results in large nerve gaps. According to the data of the present study, Epo’s action was satisfactory only during the first four weeks after surgery. The comparison between the four groups showed that during the first and the second walking track analysis on the 15th and 30th days there was significant difference between Groups A, C and D, and Group B, with Group B having clearly better results (p < 0.05). These observations are similar to those by Gorio et al. (2002), who reported enhanced neurological recovery in Epo group, during the first postoperative week. However, functional recovery in Group B was significantly affected by Epo interruption. Between the 30th and 60th postoperative day, motor recovery rate in Group B presented remarkable reduction, resulting in a diminished functional recovery when compared with Groups A and C at the end of the study.
113
The mechanisms responsible for these changes cannot be directly determined from this study. Only a few speculations can be mentioned. Already proven adequate results of Epo administration in the central nervous system may not be possible to be reproduced in peripheral nerves after injury. Schwann cells production of Epo during wallerian degeneration of the distal nerve segment, following peripheral nerve injury, is a necessary condition for nerve regeneration (Keswani et al., 2004). Exogenous Epo binding to its receptor Epo-R leads to activation of various proteins that inactivate enzymes-regulators of nerve cells apoptosis, while at the same time may antagonize endogenous erythropoietin action to the same receptors. Moreover, concentration of Epo-R on the Schwann cells does not change after peripheral nerve injury. Thus, exogenous Epo interruption may result in a diminished functional recovery, due to endogenous cytokine neuroprotective effect disturbance. In a first stage, the increased Epo level at the injury site may have stimulated Schwann cells proliferation, via an autocrine mechanism that includes JAK2 activation (Li et al., 2005). Likewise, during the first 20 days exogenous Epo may have presented not only anti-apoptotic effect, but also preventative mode of action to axonal degeneration (Keswani et al., 2004). In a second stage, when Epo’s administration was interrupted, regenerating axons may not had been longer recipients of Epo’s neurotrophic and axonoprotective effect. Many authors have mentioned erythropoietin’s role as a neuroprotective agent both in the central and the peripheral nervous system (Agnello et al., 2002; Brines, 2002; Gorio et al., 2002). However, there is not any study in the recent literature proving Epo’s long-standing neuroprotective effect. The most important observation from this study is that, there was a precipitous decrease in the functional recovery when Epo’s administration was discontinued. Five months later, Epo did not maintain its initial stimulating effect in axonal regeneration. Another factor that should be considered is Epo’s possible toxic effect. In the present study, a dose of 2680 IU/kg bw was used because this was shown to be efficacious (Campana and Myers, 2003; Sekiguchi et al., 2003). Our data indicated that systemic delivery of such dose may have acted inhibitory to collateral sprouting after end-to-side neurorrhaphy. The challenge for further studies in this area will be to determine the adequate dose for Epo administration after peripheral nerve injury. The initially increased functional recovery for the Epo group suggests a stimulating effect attributable to the presence of this agent. Whether the triggering of axonal regeneration observed is caused by Epo’s action directly on neurons, or results from increased collateral sprouting at the side of end-to-side neurorrhaphy is beyond the purpose of this study. It is possible that both mechanisms are contributing to the stimulation of the normal nerve regenerative capacity. Besides, it has been demonstrated that Epo presents not only anti-apoptotic properties but also prevents axonal degeneration (Keswani et al., 2004). We have further demonstrated that Epo is a promising substance that facilitates nerve regeneration after peripheral nerve injury. End-to-side neurorrhaphy can provide satisfactory functional recovery for the recipient nerve, without any deterioration of the donor nerve function. Viterbo et al. (1992) described the use of
114
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
“double” end-to-side neurorrhaphy. This technique stimulates axonal growth by a supercharged effect compared with end-toend repair. Considering the adequate results of this procedure, it was our option to use “double” neurorrhaphy in end-toside repair groups. Nevertheless, significant variable affecting axonal regeneration after end-to-side nerve coaptation is the use of epineurial versus perineurial sutures (Al-Qattan and AlThunyan, 1998). Lundborg noticed that end-to-side anastomosis through an epineurial window results in a greater rate of axonal regeneration (Lundborg et al., 1994). Perineurial sutures are more likely to induce collateral sprouting than epineurial sutures. However, it should be noticed that the better axonal regeneration seen in repairs with perineurial sutures, might be due to a greater degree of donor nerve damage. In our study, epineurotomy was made by meticulous technique without damaging donor nerve fascicles. In conclusion, this investigation supports that systemic Epo administration stimulates axonal regeneration after axonotmesis. Much of the neurological benefit associated with Epo administration occurs within the first weeks after injury. Epo is an agent with axonoprotective properties and may be very helpful therapeutically. Moreover, the use of erythropoietin as a substance that promotes peripheral nerve regeneration after neurorrhaphy may yield interesting results. However, further investigation will be necessary to optimise the conditions (dose, mode of administration) in order to maintain its effects. References Agnello D, Bigini P, Villa P, Mennini T, Cerami A, Brines ML, et al. Erythropoietin exerts an anti-inflammatory effect on the CNS in a model of experimental autoimmune encephalomyelitis. Brain Res 2002;952:128–34. Al-Qattan MM, Al-Thunyan A. Variables affecting axonal regeneration following end-to-side neurorrhaphy. Br J Plast Surg 1998;51:238–42. Bain JR, Mackinnon SE, Hunter DA. Functional evaluation of complete sciatic, peroneal, and posterior tibial nerve lesions in the rat. Plast Reconstr Surg 1989;83:129–38. Bocker-Meffert S, Rosenstiel P, Rohl C, Warneke N, Held-Feindt J, Sievers J, et al. VEGF promote neural outgrowth from retinal explants in postnatal rats. Invest Ophthalmol Vis Sci 2002;43:2021–6. Brines M. What evidence supports use of erythropoietin as a novel neurotherapeutic? Oncology (Huntingt) 2002;16(Suppl. 10):79–89. Brines ML, Ghezzi P, Keenan S, Agnello D, de Lanerolle NC, Cerami C, et al. Erythropoietin crosses the blood–brain barrier to protect against experimental brain injury. Proc Natl Acad Sci U S A 2000;97:10526–31. Buemi M, Cavallaro E, Floccari F, Sturiale A, Aloisi C, Trimarchi M, et al. The pleiotropic effects of erythropoietin in the central nervous system. J Neuropathol Exp Neurol 2003;62:228–36. Campana WM, Myers RR. Exogenous erythropoietin protects against dorsal root ganglion apoptosis and pain following peripheral nerve injury. Eur J Neurosci 2003;18:1497–506. Campana WM, Myers RR. Erythropoietin and erythropoietin receptors in the peripheral nervous system: changes after nerve injury. Faseb J 2001;15:1804–6. Caroni P, Gramdes P. Nerve sprouting in innervated adult skeletal muscle induced by exposure to elevated levels of insulin-like growth factors. J Cell Biol 1990;110:1307–17. Celik M, Gokmen N, Erbayraktar S, Akhisaroglu M, Konakc S, Ulukus S, et al. Erythropoietin prevents motor neuron apoptosis and neurologic disability in experimental spinal cord ischemic injury. Proc Natl Acad Sci U S A 2002;99:2258–63.
Dellon ES, Dellon AL. Functional assessment on neurological impairment: track analysis in diabetic and compression neuropathies. Plast Reconstr Surg 1991;88:686–94. De Medinaceli L, Freed WJ, Wyatt RJ. An index of the functional condition of rat sciatic nerve based on measurements made from walking tracks. Exp Neurol 1982;77:634–43. Digicaylioglu M, Garden G, Timberlake S, Fletcher L, Lipton SA. Acute neuroprotective synergy of erythropoietin and insulin-like growth factor I. Proc Natl Acad Sci U S A 2004;101:9855–60. Digicaylioglou M, Lipton SA. Erythropoietin-mediated neuroprotection involves cross-talk between Jak2 and NF-kappa B signaling cascades. Nature 2001;412:641–7. Edds MV. Collateral nerve regeneration. Quart Res Biol 1953;28:260–76. Egrie JC, Eschbach JW, McGuire T, Adamson JW. Pharmacokinetics of recombinant human erythropoietin administered to hemodialysis patients. Kidney Int 1988;33:262. Erslev AJ. In vitro production of erythropoietin by kidneys perfused with a serum free solution. Blood 1974;44:77–85. Fisher JW, Birdwell BJ. Erythropoietin production by the in situ perfused kidney. Acta Haematol 1961;26:224–32. Fu SY, Gordon T. Contributing factors to poor functional recovery after delayed nerve repair: prolonged denervation. J Neurosci 1995;15:3886–95. Gorio A, Gokmen N, Erbayraktar S, Yilmaz O, Madaschi L, Cichetti C, et al. Recombinant human erythropoietin counteracts secondary injury and markedly enhances neurological recovery from experimental spinal cord trauma. Proc Natl Acad Sci U S A 2002;99:9450–5. Grimm C, Wenzel A, Groszer M, Mayser H, Seeliger M, Samardzija M, et al. HIF-1-induced erythropoietin in the hypoxic retina protects against lightinduced retinal degeneration. Nat Med 2002;8:718–24. Hopkins WG, Slack JR. The sequential development of nodal sprouts in mouse muscles in response to nerve degeneration. J Neurocytol 1981;10:537– 56. Jacobson LO, Goldwasser E, Fried W, Plzak L. Role of the kidney in erythropoiesis. Nature 1957;179:633. Juul SE, Anderson DK, Li Y, Christensen RD. Erythropoietin and erythropoietin receptor in the developing human central nervous system. Pediatr Res 1998;43:40–9. Kaufman JS, Reda DJ, Fye CL, Goldfarb DC, Henderson WG, Kleinman JG, et al. Subcutaneous compared with intravenous epoetin in patients receiving hemodialysis. Department of Veterans Affairs Cooperative Study Group on erythropoietin in hemodialysis patients. N Engl J Med 1998;339:578–83. Keswani SC, Buldanlioglu U, Fischer A, Reed N, Polley M, Liang H, et al. A novel endogenous erythropoietin mediated pathway prevents axonal degeneration. Ann Neurol 2004;56:815–26. Kobayashi J, Mackinnon SE, Watanabe O, Ball DJ, Gu XM, Hunter DA, et al. The effect of duration of muscle denervation on functional recovery in the rat model. Muscle Nerve 1997;20:858–66. Krarup C, Archibald SJ, Madison RD. Factors that influence peripheral nerve regeneration: an electrophysiological study of the monkey median nerve. Ann Neurol 2002;51:69–81. Li X, Gonias SL, Campana WM. Schwann cells express erythropoietin receptor and represent a major target for Epo in peripheral nerve injury. Glia 2005;51:254–65. Liuzzi FJ, Tedeschi B. Peripheral nerve regeneration. Neurosurg Clin N Am 1991;2:31–42. Livuah O, Stura EA, Middleton SA, Johnson DC, Jokiffe CK, Wilson TA. Crystallographic evidence for performed divers of erythropoietin receptor before ligand activation. Science 1999;283:987–90. Lundborg G, Zhao Q, Kanje M, Danielsen N, Kerns JM. Can sensory and motor collateral sprouting be induced from intact peripheral nerve by end-to-side anastomosis? J Hand Surg 1994;19B:227–82. Marti HH, Wenger RH, Rivas LA, Straumann U, Digicaylioglu M, Henn V, et al. Erythropoietin gene expression in human, monkey and murine brain. Eur J Neurosci 1996;8:666–76. Morishita E, Masuda S, Nagao M, Yasuda Y, Sasaki R. Erythropoietin receptor is expressed in rat hippocampal and cerebral cortical neurons, and erythropoietin prevents in vitro glutamate-induced neuronal death. Neuroscience 1997;76:105–16.
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115 Remy I, Wilson IA, Michnick SW. Erythropoietin receptor activation by a ligandinduced conformation change. Science 1999;283:990–3. Sekiguchi Y, Kikuchi S, Myers RR, Campana WM. Erythropoietin inhibits spinal neuronal apoptosis and pain following nerve root crush. Spine 2003;28:2577–84. Sir´en AL, Knerlich F, Poser W, Gleiter CH, Bruck W, Ehrenreich H. Erythropoietin and erythropoietin receptor in human ischemic/hypoxic brain. Acta Neuropathol (Berl) 2001;101:271–6. Slack JR, Hopkins WG, Williams MN. Nerve sheaths and motoneurone collateral sprouting. Nature 1979;282:506–7. Trumble TE. Peripheral nerve transplantation: the effects of predegenerated grafts and immunosuppression. J Neural Transplant Plast 1992;3:39–49.
115
Tsai PT, Ohab JJ, Kertesz N, Groszer M, Matter C, Gao J, et al. A critical role of erythropoietin receptor in neurogenesis and post-stroke recovery. J Neurosci 2006;26:1269–74. Viterbo F, Trindade JC, Hoshino K, Mazzoni Neto A. Latero-terminal neurorrhaphy without removal of the epineural sheath: experimental study in rats. Rev Paul Med 1992;110:267–75. Viterbo F, Trindade JC, Hoshino K, Mazzoni Neto A. End-to-side neurorrhaphy with removal of the epineurial sheath: An experimental study in rats. Plast Reconstr Surg 1994;94:1038–47. Zhang Z, Johnson EO, Vekris MD, Zoubos AB, Bo J, Beris AE, et al. Long-term evaluation of rabbit peripheral nerve repair with end-to-side neurorrhaphy in rabbits. Microsurgery 2006;26:245–52.
Axonal regeneration stimulated by erythropoietin: An experimental study in rats Marios G. Lykissas a,∗ , Ekaterini Sakellariou b,1 , Marios D. Vekris a,1 , Vasilios A. Kontogeorgakos a,1 , Anna K. Batistatou c,2 , Gregory I. Mitsionis a,1 , Alexandros E. Beris a,1 a
Department of Orthopaedic Surgery, University of Ioannina, School of Medicine, Ioannina, P.C. 45110, Greece b Department of Plastic Surgery, Thriasio General Hospital, Athens, Greece c Department of Pathology, University of Ioannina, School of Medicine, Ioannina, Greece Received 14 March 2007; received in revised form 10 April 2007; accepted 10 April 2007
Abstract The aim of the present study is to evaluate the effects of erythropoietin to the collateral sprouting by using systemically delivered erythropoietin in an end-to-side nerve repair model. Forty-five rats were evaluated in four groups: (A) end-to-side neurorrhaphy only, (B) end-to-side neurorrhaphy and erythropoietin administration, (C) end-to-end neurorrhaphy and (D) nerve stumps buried into neighboring muscles. In all animals, the contralateral healthy side served as control. Functional assessment of nerve regeneration was performed at intervals up to 5 months using the Peroneal Function Index. Evaluation 150 days after surgery included peroneal and tibial nerve morphometric examination, and wet weights of the tibialis anterior muscle. During the first three weeks after surgery, when erythropoietin was regularly administered, functional evaluation showed that erythropoietin may facilitate peripheral nerve regeneration. However, there was rapid deterioration in the functional recovery when erythropoietin’s administration was discontinued. As a consequence, at the end of this study, erythropoietin failed to maintain its initial stimulating effect in axonal regeneration. The results of wet muscle weights revealed statistically significant differences between Groups A and C, and Group B. Furthermore, data on axonal counting showed significant difference between Groups A and C, and Group B. Erythropoietin appears to facilitate peripheral nerve regeneration at the initial phase of its administration. Further investigation will be necessary to optimise the conditions (dose, mode of administration) in order to maintain its effects. © 2007 Elsevier B.V. All rights reserved. Keywords: Erythropoietin; End-to-side neurorrhaphy; Nerve regeneration; Collateral sprouting
1. Introduction Erythropoietin (Epo) is a glycoprotein with a molecular weight of 30.400 Da that controls the production of red cells. Epo in humans is synthesized by renal peritubular cells in adults and by hepatic cells in the fetus, whereas a small amount is also synthesized in the adult’s liver (Jacobson et al., 1957; Fisher and
∗
Corresponding author. Tel.: +30 26510 97472; fax: +30 26510 34816. E-mail addresses: [email protected] (M.G. Lykissas), [email protected] (E. Sakellariou), [email protected] (M.D. Vekris), [email protected] (V.A. Kontogeorgakos), [email protected] (A.K. Batistatou), [email protected] (G.I. Mitsionis), [email protected] (A.E. Beris). 1 Tel.: +30 26510 97472; fax: +30 26510 34816. 2 Tel.: +30 26510 99415; fax: +30 26510 34816. 0165-0270/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jneumeth.2007.04.008
Birdwell, 1961; Erslev, 1974). Epo binds to its receptor (EpoR) on the red cell surface, which consists of a p66 protein that undergoes dimerization after activation, and two additional proteins (Livuah et al., 1999; Remy et al., 1999). Erythropoietin, as well as its receptor are present in the central nervous system, including the retina (Marti et al., 1996; Juul et al., 1998; Sir´en et al., 2001; Celik et al., 2002; Bocker-Meffert et al., 2002; Grimm et al., 2002). It has been proven that Epo is produced in the bodies and the axons of the normal ganglion of the rat’s dorsal root and is increased in Schwann cells after peripheral nerve injury (Campana and Myers, 2001). Epo-R is found on certain axons and neuron bodies of the dorsal root ganglion, on endothelial cells and Schwann cells of the normal peripheral nerve. Inside the dorsal root ganglion, the neuron type that expresses the Epo-R represents 40% of the ganglion cells. Moreover, systemic administration of rHuEpo (recombinant human
108
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
erythropoietin) reduces the apoptosis of the dorsal root ganglion cells and contributes to the recovery of mechanical allodynia (Campana and Myers, 2003). Epo’s neuroprotective effect is in fact synergic with that of insulin-like growth factor I (IGF-I). Thus, the combination of Epo and IGF-I exhibits neuroprotective effect with lower doses than each cytokine separately does (Digicaylioglu et al., 2004). Epo’s role as a neuroprotective agent is considered to be multifactorial, with a direct and an indirect action on nerve cells. Epo, in vitro, protects the nerve cells from hypoxia-induced glutamate toxicity, which is the main cause of nerve cells death as regards hypoxia (Morishita et al., 1997; Buemi et al., 2003). Besides, in a multiple sclerosis model, the systemic administration of erythropoietin alpha reduces the immune response and the inflammatory reaction, while it enhances nerve recovery after a spinal cord trauma (Agnello et al., 2002; Brines, 2002; Gorio et al., 2002). Epo is produced in the bodies and the axons of the normal ganglion of the rat’s dorsal root and is increased in Schwann cells after peripheral nerve injury (chronic constriction injury, CCI) (Campana and Myers, 2001). Furthermore, like in the central nervous system, in the peripheral as well, Epo presents anti-apoptotic after injury action. Although Epo’s anti-apoptotic mechanism has not been completely elucidated, it has been shown that Epo prevents apoptosis of CNS neurons by triggering cross-talk between Janus-tyrosine-kinase-2 (JAK2) and nuclear factor kappa B (NF-kB) signaling cascades (Digicaylioglou and Lipton, 2001). Moreover, Epo presents not only anti-apoptotic properties but also prevents axonal degeneration and enhances neurogenesis (Keswani et al., 2004; Tsai et al., 2006). Consequently, Schwann cell derived Epo activates an “axonoprotective” pathway. Alongside, Epo’s action on Schwann cells and on the inflammatory response after neurological trauma, may be effective in peripheral nerve regeneration after neurorrhaphy. Since the last 20 years end-to-side neurorrhaphy has been added to the surgical options, although this technique was already described at the beginning of the last century. Recent evidence suggests that after end-to-side neurorrhaphy nerve regeneration occurs by collateral sprouting (Edds, 1953; Slack et al., 1979; Hopkins and Slack, 1981; Zhang et al., 2006). It is certain that a great number of factors contribute to collateral sprouting, but the mechanism remains mostly unknown. Various growth factors have promoting effect in nerve regeneration and they are well documented for many different models of central or peripheral nervous system injury. However, the factors that stimulate collateral sprouting are listed only in few references in the literature (Caroni and Gramdes, 1990; Liuzzi and Tedeschi, 1991). Experimental work has demonstrated that Epo protects from glutamate-induced cell death in vitro (Digicaylioglou and Lipton, 2001) and in vivo (Brines et al., 2000), stimulating at the same time intracellular anti-apoptotic metabolic pathways. Systemic administration of Epo may inhibit the apoptosis of sensory and motor neurons after neurological damage. Epo’s action on Schwann cells and on the inflammatory response after neurological trauma, may be effective in peripheral nerve regeneration after end-to-side neurorrhaphy.
2. Material and methods The procedures for this investigation were performed according to protocols approved by the Official Veterinary Services, General Directorate of Development, Prefecture of Ioannina (ref no. 2579). Forty-five male Wistar rats, weighting between 190 and 310 g, were used in the present study. The animals were divided in four groups according to the operative procedure: Group A: end-to-side repair (n = 12), Group B: end-to-side repair and Epo administration (n = 12), Group C: end-to-end nerve repair (n = 12), Group D: segmental nerve resection (n = 9). 2.1. Surgical procedure Just before the operation, all animals were individually numbered and given preconditioning trials on an 8 × 62 cm walking track, for comparison with normative data (Dellon and Dellon, 1991). Under anesthesia with a single 5 mg/kg intraperitoneal ketamin injection (Narketan 10, 100 mg/ml, Vetoquinolag, Switzerland), the right hind limb of all animals was used as the experimental limb, and the left hind limb as the control limb. The right sciatic, tibial and peroneal nerves were exposed and dissected through a semitendinous-biceps femoris muscle splitting incision, under magnification of an operating microscope. The common peroneal nerve was transversely divided, and the distal and the proximal stump were immediately repaired according to grouping. In Group A, the proximal peroneal nerve stump was sutured to the trunk of the intact tibial nerve, in a terminolateral fashion described by Viterbo et al. (1994). The distal segment was repaired end-to-side, 1.2 cm distal of the first end-to-side neurorrhaphy. In Group B (stimulate nerve regeneration with erythropoietin), the proximal and the distal stump of the severed peroneal nerve were sutured end-to-side to the trunk of the intact tibial nerve, as in Group A. In both Groups A and B, an epineural window was made on the tibial trunk before each end-to-side nerve coaptation took place. In Group C (primary repair), the peroneal nerve was repaired in a classic end-to-end fashion. In Groups A, B and C, the neurorrhaphies were performed under microscope magnification, using three epineural 10/0 nylon interrupted sutures, placed at 120◦ intervals. In Group D (denervated control), both proximal and distal nerve stumps were buried into the neighboring muscles and fixed with a single 8/0 nylon suture. In all groups wound closure was achieved by 4/0 vicryl interrupted sutures for the muscles, and 3/0 nylon interrupted sutures for the skin. Before and after the operation, the animals were kept in individual cages and maintained on standard rat chow and water ad libitum, with a 12-h light:12-h dark cycle. 2.2. Erythropoietin administration In Group B, subcutaneous administration of rHuEpo was chosen, as it is more effective than i.v. or i.p. administration (Kaufman et al., 1998). Nevertheless, rHuEpo injected subcuta-
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
neously has a half-life greater than 24 h compared with the 4 h half-life of rHuEpo injected intravenously (Egrie et al., 1988). According to several studies, rHuEpo preconditioning is more effective in preventing dorsal root ganglion apoptosis following peripheral nerve injury and neuron apoptosis during brain ischaemia (Morishita et al., 1997; Digicaylioglou and Lipton, 2001; Campana and Myers, 2003). Thus, in this report pretreatment with rHuEpo 1 day before the operation was followed. The rHuEpo administration was repeated every 24 h for the first 20 postoperative days. A middle dose of 2680 units/kg bw of rHuEpo was administered daily, as the effectiveness of a higher dose of 5000 units/kg bw was no better than the middle dose, while a low dose of 1000 units/kg bw was no effective (Campana and Myers, 2003; Sekiguchi et al., 2003). 2.3. Walking track analysis At 15, 30, 60, 90 and 150 days after the operation, all animals were analyzed using an 8 × 62 cm walking track darkened at one end (De Medinaceli et al., 1982). The paper strips containing the footprints were copied, the digitized footprints were analyzed, and the peroneal function index (PFI) was automatically calculated. The parameters measured in the footprints were print length (PL) and toe spread (TS), both in the normal (N) and experimental (E) pawprints. All measurements and PFI calculation were made by two blinded to the identity of the digitized footprints investigators at the end of the study according to the formula of Bain et al. (1989): PFI = 174.9
(EPL − NPL) (ETS − NTS) + 80.3 − 13.4 NPL NTS
109
Inc, Munich, Germany) connected with a computer using the public domain software NIH Image (http://rsb.info.nih.gov/nihimage/). Quantitative evaluation followed with myelinated nerve fibers counting in all nerve segments (axons/mm2 ). Morphometric analysis also included the calculation of percent neural tissue (100 × neural area/intrafascicular area). The same parameters were determined for the experimental limb, as well as the control limb. The final results were expressed as the ratio of the experimental side to the contralateral unoperated side. All calculations were made by a single blinded investigator. 2.5. Muscle weights On the 150th postoperative day, tibialis anterior muscle was harvested from the experimental and the contralateral control side. Wet muscle weights were measured and reported as the ratio of the experimental to the contralateral side. Afterwards, specimens from each muscle were immersed in 10% formalin and embedded in paraffin for histologic analysis. 2.6. Statistical analysis Peroneal function index, muscle mass ratios, and nerve fiber density values were analyzed by using the non-parametric Kruskal–Wallis and Mann–Whitney U tests. All tests were calculated with use of the SPSS, version 13.0 (SPSS Inc., Chicago, IL, USA) statistic package for personal computers. In all instances, p < 0.05 was regarded as statistically significant. 3. Results
2.4. Morphometric studies
3.1. Walking track analysis
Under the same anesthesia on the 150th day, the experimental side was exposed through a posterolateral approach to the thigh and peroneal and tibial nerves were dissected and removed. Afterwards, the animals were killed with a lethal dose of intraperitoneal sodium pentobarbital (100 mg/kg). In Groups A and B 0.5 cm-long sections were removed from the tibial nerve midway between the proximal and the distal endto-side neurorrhaphy, the distal peroneal nerve segment with the distal neurorrhaphy zone, and the tibial nerve below the site of neurorrhaphies. In Group C, section 0.5 cm in length of the peroneal nerve was removed distal to the site of end-to-end neurorrhaphy. In Group D, a 0.5 cm-long segment of the distal peroneal stump was harvested. In all animals, corresponding sections from the contralateral control sides were harvested. All sections were immersed in a solution of 2.5% glutaraldehyde buffered with cacodylate, washed in sodium cacodylated buffer (0.2 mol/L, pH 7.4), post-fixed in 1% osmium tetroxide, dehydrated in an ethyl alcohol solution of growing concentrations (50, 70, 95 and 100%), and embedded in epoxy resin for axonal counting. Cross-sections 1 m thick were cut using a Reichert-Jung ultracut E with a diamond knife (Reichert-Jung, Weiss, Austria) from the nerve segments, stained with toluidine blue, and examined with a light microscope (Axioscop, Zeiss,
Two hundred seventy paper strips containing at least six clearly marked footprints each, were used to evaluate peroneal nerve functional recovery. Analysis of walking tracks was done in all animals preoperatively, and at 15, 30, 60, 90 and 150 days, according to the formula of Bain et al. (1989) (Table 1). The preoperative PFI value was never 0, but oscillated between −22.36 and −7.57. Before the surgical procedure no significant difference obtained between the four groups (p = 0.739). Such was an essential condition to obtain intraanimal control data. Furthermore, PFI values carried out on subjects with the same age, weight and sex, in order to be comparable in many parameters (Dellon and Dellon, 1991). On the 15th postoperative day, significant difference was detected between Groups A and B (p < 0.001), with Group B revealing better results (Fig. 1). PFI values were not significantly different between Groups B and C (p = 0.184). On the 2nd postoperative evaluation, significant differences were recorded between Groups A and B (p < 0.001), as well as Groups B and C (p = 0.002), with Group B having better functional outcome. The 3rd evaluation on the 60th day after surgery revealed reversed significant differences between Groups A and B (p < 0.013), with Group B showing functional regression. Significant differences were also found to be reversed between Groups B and
110
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
Table 1 Average peroneal function index scores Group A Pre-op D 15 D 30 D 60 D 90 D 150
−14.05 −41.66 −36.63 −28.09 −19.93 −16.84
B ± ± ± ± ± ±
3.35 4.24 3.50 4.39 4.69 3.42
−13.32 −33.91 −23.17 −32.86 −29.93 −28.15
C ± ± ± ± ± ±
4.64 3.40 2.75 4.49 9.57 11.55
−14.33 −35.41 −28.41 −22.92 −18.01 −13.94
D ± ± ± ± ± ±
−14.35 −46.78 −46.89 −49.69 −62.81 −71.32
2.69 2.12 4.16 3.62 2.49 2.68
± ± ± ± ± ±
4.64 8.19 5.89 5.69 2.55 3.21
Average PFI scores in Groups A, B, C and D obtained at 0, 15, 30, 60, 90 and 150 days. All values are expressed as mean ± standard deviation.
C (p < 0.001). On the 90th postoperative day, difference between Groups A and C (p = 0.355) became insignificant, while differences between both Groups A and C compared with Group B remain significant (p < 0.05). The end scores at 150th day, were −16.84 ± 3.42, −28.15 ± 11.55 and −13.94 ± 2.68 for Groups A, B and C, respectively. Statistically, differences between all groups were significant (p < 0.05). 3.2. Morphometric studies At the end of the study, the ratios of peroneal nerve fiber density (axons/mm2 ) for Groups A, B and C were calculated (Table 2). Severe axonal degeneration was present in Group D. Morphometric evaluation could not be performed in this group. Morphometric evaluation demonstrated that there was statistically significant difference between Groups A and B (1.57 ± 0.62 and 0.65 ± 0.19, respectively), as well as between Groups B and C (0.65 ± 0.19 and 1.95 ± 0.46, respectively), with Group B having less sufficient results (p < 0.001). Statistically, the difference between Group A and Group C was insignificant (p = 0.083). The percentage neural tissue data is demonstrated in Table 3. In the end-to-side nerve repair group, the ratio of percentage
Table 2 Peroneal nerve fibre density Group
Axon density (n/mm2 )
Ratio of axon density
A B C
13,900 ± 2,900 (9,400–18,900) 13,100 ± 4,200 (5,700–19,000) 7,700 ± 4,200 (1,800–18,400)
1.57 ± 0.62 (0.67–2.68) 0.65 ± 0.19 (0.28–0.94) 1.95 ± 0.46 (0.97–2.47)
Average peroneal nerve fiber density reported as the ratio of the experimental to the contralateral side on 150th postoperative day. All values are expressed as mean ± standard deviation. Significant differences were noted between Groups A and B as well as Groups B and C both in axon density and axon density ratio (p < 0.05). Moreover, differences between Group A and C were insignificant (p > 0.05).
peroneal neural tissue of the experimental to the unoperated control side was 0.72 ± 0.17. For the Epo group, the ratio was 0.35 ± 0.01. These differences were statistically significant at p < 0.001. Moreover, the ratio of percentage neural tissue was significantly reduced in Epo group when compared with the endto-end nerve repair group (p < 0.001). No significant differences were noted between Groups A and C (p = 0.204). Macroscopic examination of all nerve segments was in accordance with quantity histomorphometry of nerves (Fig. 2). In Groups A and B, there was no histologic evidence of injury to the donor nerve (tibial nerve) after the intact nerve bridge procedure. 3.3. Muscle weights The tibialis anterior muscle, innervated by the peroneal nerve, was weighted (Table 4). Wet muscle weights were reported as the Table 3 The percentage neural tissue data
Fig. 1. Diagram of PFI values behaviour according to grouping.
Group
Percentage neural tissue (%)
Ratio of neural tissue
A B C
35.98 ± 10.81 (19.70–52.30) 17.84 ± 5.12 (12.1–29.4) 48.86 ± 5.48 (37.9–54.7)
0.72 ± 0.17 (0.42–0.94) 0.35 ± 0.01 (0.25–0.54) 0.81 ± 0.01 (0.67–0.89)
The percentage neural tissue data of peroneal nerve reported as the ratio of the experimental to the contralateral side. All values are expressed as mean ± standard deviation. Significant differences were noted between Groups A and B as well as Groups B and C both in percentage neural tissue and percentage neural tissue ratio (p < 0.001). At the same time, difference between Groups A and C in percentage neural tissue ratio was insignificant (p = 0.204). However, difference between Groups A and C in percentage neural tissue was significant at p < 0.05.
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
111
Fig. 2. Transverse section of the distal part of the operated peroneal nerve 150 days postoperatively (toluidine blue, original magnification ×400). (A) Histologic sections in Group A revealed viable axons distal to the coaptation site. Axonal regeneration was similar to that obtained in end-to-end control group (Group C). However, there was lower number of axons in Group A compared with Group C. (B) In Group B histologic sections demonstrated few myelinated axons surrounded by connective tissue. (C) In end-to-end group (Group C), a normal pattern of nerve regeneration was present. (D) Wallerian degeneration was obtained in the distal stump of the unrepaired peroneal nerve in Group D.
ratio of the experimental to the contralateral control side (Fig. 3). The ratios were 0.80 ± 0.06, 0.56 ± 0.19 and 0.81 ± 0.09 for Groups A, B and C respectively. Statistically, the difference between Groups A and C was insignificant (p = 0.817). In some cases, muscle weights in the experimental side were almost the same with control side. On the other hand, the results revealed statistically significant differences between Groups A and B (p = 0.003), and Groups B and C (p = 0.002). Moreover, statistical difference was seen between the three experimental groups and Group D, in which the peroneal nerve was transected and buried into the neighboring muscles to prevent reinnervation (p < 0.001). Macroscopically, in Groups A and C there was only minimal atrophy in the tibialis anterior muscles of the experimental limb
(Fig. 4). However, in Groups B and D tibialis anterior muscle atrophy was obvious at the experimental side when compared with the normal contralateral side. It has been proven that the relative loss of tibialis anterior muscle weight closely reflects the degree of denervation (Trumble, 1992).
Table 4 Average wet muscle weights Group
Mean ± S.D. (m)
A B C D
0.80 0.56 0.81 0.19
± ± ± ±
0.06 0.19 0.09 0.04
Minimum
Maximum
0.71 0.31 0.68 0.13
0.92 0.84 0.97 0.28
Average wet muscle weights reported as the ratio of the experimental to the contralateral normal side on 150th postoperative day. All values are expressed as mean ± standard deviation.
Fig. 3. Tibialis anterior wet muscle weights expressed as the ratio of the experimental side to contralateral normal side. Data are shown as mean ± S.D.
112
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
Fig. 4. (A–D) Macroscopic appearance of the tibialis anterior muscles of control (left on each picture) and experimental (right on each picture) limbs. There was minimal atrophy in muscles of the experimental limbs in Groups A and C. In contrast, in Groups B and D the atrophy was evident.
3.4. Muscle histology
4. Discussion
The tibialis anterior muscles of the experimental and the contralateral control side were immersed in 10% formalin, embedded in paraffin, stained with hematoxylin-eosin, and evaluated histologically (Fig. 5). There was no significant difference between Groups A and C for the histologic appearance of the tibialis anterior muscles of the experimental limbs, which furthermore had a great similarity with the unoperated control limbs. In these groups transverse sections of muscle fascicles showed polygonal myofibers with mostly subsarcolemmal nuclei and little intervening endomysial connective tissue (Fig. 5A2 and C2). However, histologic examination of tibialis anterior muscles in Group B revealed focally small groups of atrophied fibers with increased endomysial connective tissue (Fig. 5B2). In Group D, transverse sections revealed small bundles of atrophic fibers with small diameter and centrally placed nuclei mingled with connective and adipose tissue (Fig. 5D2).
Prerequisite of functional recovery in an end-to-side nerve model is nerve regeneration as well as a successful muscle reinnervation. After peripheral nerve injury, the degree of motor recovery depends on the number of regenerated axons and the time length of muscle denervation. The longer the interval between muscle denervation and reinnervation, the poorer the degree of functional recovery (Kobayashi et al., 1997). It has been demonstrated that prolonged motor axonotmesis results in dramatic reductions in motor unit number (Fu and Gordon, 1995). This observation defines the presence of a “time window” after which the level of functional recovery is insufficient. According to experimental studies, in primates this time period does not appears to extend more than 100 days after nerve injury (Krarup et al., 2002). Consequently, any mechanism which decrease denervation time may contribute to a better functional result. In the present study, Epo administration as a potential
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
Fig. 5. Histologic appearance of the experimental tibialis anterior muscles in Groups A (A2) and C (C2) was almost normal (hematoxylin–eosin, ×100). In Groups B (B2) and D (D2), experimental tibialis anterior muscles were atrophic with small, angulated fibers. Tibialis anterior muscles of unoperated limbs in Groups A (A1), B (B1), C (C1) and D (D1) were used as control.
stimulating factor of nerve regeneration is investigated. If Epo provides early recovery of function after peripheral nerve injury and repair, its use may yield interesting results in large nerve gaps. According to the data of the present study, Epo’s action was satisfactory only during the first four weeks after surgery. The comparison between the four groups showed that during the first and the second walking track analysis on the 15th and 30th days there was significant difference between Groups A, C and D, and Group B, with Group B having clearly better results (p < 0.05). These observations are similar to those by Gorio et al. (2002), who reported enhanced neurological recovery in Epo group, during the first postoperative week. However, functional recovery in Group B was significantly affected by Epo interruption. Between the 30th and 60th postoperative day, motor recovery rate in Group B presented remarkable reduction, resulting in a diminished functional recovery when compared with Groups A and C at the end of the study.
113
The mechanisms responsible for these changes cannot be directly determined from this study. Only a few speculations can be mentioned. Already proven adequate results of Epo administration in the central nervous system may not be possible to be reproduced in peripheral nerves after injury. Schwann cells production of Epo during wallerian degeneration of the distal nerve segment, following peripheral nerve injury, is a necessary condition for nerve regeneration (Keswani et al., 2004). Exogenous Epo binding to its receptor Epo-R leads to activation of various proteins that inactivate enzymes-regulators of nerve cells apoptosis, while at the same time may antagonize endogenous erythropoietin action to the same receptors. Moreover, concentration of Epo-R on the Schwann cells does not change after peripheral nerve injury. Thus, exogenous Epo interruption may result in a diminished functional recovery, due to endogenous cytokine neuroprotective effect disturbance. In a first stage, the increased Epo level at the injury site may have stimulated Schwann cells proliferation, via an autocrine mechanism that includes JAK2 activation (Li et al., 2005). Likewise, during the first 20 days exogenous Epo may have presented not only anti-apoptotic effect, but also preventative mode of action to axonal degeneration (Keswani et al., 2004). In a second stage, when Epo’s administration was interrupted, regenerating axons may not had been longer recipients of Epo’s neurotrophic and axonoprotective effect. Many authors have mentioned erythropoietin’s role as a neuroprotective agent both in the central and the peripheral nervous system (Agnello et al., 2002; Brines, 2002; Gorio et al., 2002). However, there is not any study in the recent literature proving Epo’s long-standing neuroprotective effect. The most important observation from this study is that, there was a precipitous decrease in the functional recovery when Epo’s administration was discontinued. Five months later, Epo did not maintain its initial stimulating effect in axonal regeneration. Another factor that should be considered is Epo’s possible toxic effect. In the present study, a dose of 2680 IU/kg bw was used because this was shown to be efficacious (Campana and Myers, 2003; Sekiguchi et al., 2003). Our data indicated that systemic delivery of such dose may have acted inhibitory to collateral sprouting after end-to-side neurorrhaphy. The challenge for further studies in this area will be to determine the adequate dose for Epo administration after peripheral nerve injury. The initially increased functional recovery for the Epo group suggests a stimulating effect attributable to the presence of this agent. Whether the triggering of axonal regeneration observed is caused by Epo’s action directly on neurons, or results from increased collateral sprouting at the side of end-to-side neurorrhaphy is beyond the purpose of this study. It is possible that both mechanisms are contributing to the stimulation of the normal nerve regenerative capacity. Besides, it has been demonstrated that Epo presents not only anti-apoptotic properties but also prevents axonal degeneration (Keswani et al., 2004). We have further demonstrated that Epo is a promising substance that facilitates nerve regeneration after peripheral nerve injury. End-to-side neurorrhaphy can provide satisfactory functional recovery for the recipient nerve, without any deterioration of the donor nerve function. Viterbo et al. (1992) described the use of
114
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115
“double” end-to-side neurorrhaphy. This technique stimulates axonal growth by a supercharged effect compared with end-toend repair. Considering the adequate results of this procedure, it was our option to use “double” neurorrhaphy in end-toside repair groups. Nevertheless, significant variable affecting axonal regeneration after end-to-side nerve coaptation is the use of epineurial versus perineurial sutures (Al-Qattan and AlThunyan, 1998). Lundborg noticed that end-to-side anastomosis through an epineurial window results in a greater rate of axonal regeneration (Lundborg et al., 1994). Perineurial sutures are more likely to induce collateral sprouting than epineurial sutures. However, it should be noticed that the better axonal regeneration seen in repairs with perineurial sutures, might be due to a greater degree of donor nerve damage. In our study, epineurotomy was made by meticulous technique without damaging donor nerve fascicles. In conclusion, this investigation supports that systemic Epo administration stimulates axonal regeneration after axonotmesis. Much of the neurological benefit associated with Epo administration occurs within the first weeks after injury. Epo is an agent with axonoprotective properties and may be very helpful therapeutically. Moreover, the use of erythropoietin as a substance that promotes peripheral nerve regeneration after neurorrhaphy may yield interesting results. However, further investigation will be necessary to optimise the conditions (dose, mode of administration) in order to maintain its effects. References Agnello D, Bigini P, Villa P, Mennini T, Cerami A, Brines ML, et al. Erythropoietin exerts an anti-inflammatory effect on the CNS in a model of experimental autoimmune encephalomyelitis. Brain Res 2002;952:128–34. Al-Qattan MM, Al-Thunyan A. Variables affecting axonal regeneration following end-to-side neurorrhaphy. Br J Plast Surg 1998;51:238–42. Bain JR, Mackinnon SE, Hunter DA. Functional evaluation of complete sciatic, peroneal, and posterior tibial nerve lesions in the rat. Plast Reconstr Surg 1989;83:129–38. Bocker-Meffert S, Rosenstiel P, Rohl C, Warneke N, Held-Feindt J, Sievers J, et al. VEGF promote neural outgrowth from retinal explants in postnatal rats. Invest Ophthalmol Vis Sci 2002;43:2021–6. Brines M. What evidence supports use of erythropoietin as a novel neurotherapeutic? Oncology (Huntingt) 2002;16(Suppl. 10):79–89. Brines ML, Ghezzi P, Keenan S, Agnello D, de Lanerolle NC, Cerami C, et al. Erythropoietin crosses the blood–brain barrier to protect against experimental brain injury. Proc Natl Acad Sci U S A 2000;97:10526–31. Buemi M, Cavallaro E, Floccari F, Sturiale A, Aloisi C, Trimarchi M, et al. The pleiotropic effects of erythropoietin in the central nervous system. J Neuropathol Exp Neurol 2003;62:228–36. Campana WM, Myers RR. Exogenous erythropoietin protects against dorsal root ganglion apoptosis and pain following peripheral nerve injury. Eur J Neurosci 2003;18:1497–506. Campana WM, Myers RR. Erythropoietin and erythropoietin receptors in the peripheral nervous system: changes after nerve injury. Faseb J 2001;15:1804–6. Caroni P, Gramdes P. Nerve sprouting in innervated adult skeletal muscle induced by exposure to elevated levels of insulin-like growth factors. J Cell Biol 1990;110:1307–17. Celik M, Gokmen N, Erbayraktar S, Akhisaroglu M, Konakc S, Ulukus S, et al. Erythropoietin prevents motor neuron apoptosis and neurologic disability in experimental spinal cord ischemic injury. Proc Natl Acad Sci U S A 2002;99:2258–63.
Dellon ES, Dellon AL. Functional assessment on neurological impairment: track analysis in diabetic and compression neuropathies. Plast Reconstr Surg 1991;88:686–94. De Medinaceli L, Freed WJ, Wyatt RJ. An index of the functional condition of rat sciatic nerve based on measurements made from walking tracks. Exp Neurol 1982;77:634–43. Digicaylioglu M, Garden G, Timberlake S, Fletcher L, Lipton SA. Acute neuroprotective synergy of erythropoietin and insulin-like growth factor I. Proc Natl Acad Sci U S A 2004;101:9855–60. Digicaylioglou M, Lipton SA. Erythropoietin-mediated neuroprotection involves cross-talk between Jak2 and NF-kappa B signaling cascades. Nature 2001;412:641–7. Edds MV. Collateral nerve regeneration. Quart Res Biol 1953;28:260–76. Egrie JC, Eschbach JW, McGuire T, Adamson JW. Pharmacokinetics of recombinant human erythropoietin administered to hemodialysis patients. Kidney Int 1988;33:262. Erslev AJ. In vitro production of erythropoietin by kidneys perfused with a serum free solution. Blood 1974;44:77–85. Fisher JW, Birdwell BJ. Erythropoietin production by the in situ perfused kidney. Acta Haematol 1961;26:224–32. Fu SY, Gordon T. Contributing factors to poor functional recovery after delayed nerve repair: prolonged denervation. J Neurosci 1995;15:3886–95. Gorio A, Gokmen N, Erbayraktar S, Yilmaz O, Madaschi L, Cichetti C, et al. Recombinant human erythropoietin counteracts secondary injury and markedly enhances neurological recovery from experimental spinal cord trauma. Proc Natl Acad Sci U S A 2002;99:9450–5. Grimm C, Wenzel A, Groszer M, Mayser H, Seeliger M, Samardzija M, et al. HIF-1-induced erythropoietin in the hypoxic retina protects against lightinduced retinal degeneration. Nat Med 2002;8:718–24. Hopkins WG, Slack JR. The sequential development of nodal sprouts in mouse muscles in response to nerve degeneration. J Neurocytol 1981;10:537– 56. Jacobson LO, Goldwasser E, Fried W, Plzak L. Role of the kidney in erythropoiesis. Nature 1957;179:633. Juul SE, Anderson DK, Li Y, Christensen RD. Erythropoietin and erythropoietin receptor in the developing human central nervous system. Pediatr Res 1998;43:40–9. Kaufman JS, Reda DJ, Fye CL, Goldfarb DC, Henderson WG, Kleinman JG, et al. Subcutaneous compared with intravenous epoetin in patients receiving hemodialysis. Department of Veterans Affairs Cooperative Study Group on erythropoietin in hemodialysis patients. N Engl J Med 1998;339:578–83. Keswani SC, Buldanlioglu U, Fischer A, Reed N, Polley M, Liang H, et al. A novel endogenous erythropoietin mediated pathway prevents axonal degeneration. Ann Neurol 2004;56:815–26. Kobayashi J, Mackinnon SE, Watanabe O, Ball DJ, Gu XM, Hunter DA, et al. The effect of duration of muscle denervation on functional recovery in the rat model. Muscle Nerve 1997;20:858–66. Krarup C, Archibald SJ, Madison RD. Factors that influence peripheral nerve regeneration: an electrophysiological study of the monkey median nerve. Ann Neurol 2002;51:69–81. Li X, Gonias SL, Campana WM. Schwann cells express erythropoietin receptor and represent a major target for Epo in peripheral nerve injury. Glia 2005;51:254–65. Liuzzi FJ, Tedeschi B. Peripheral nerve regeneration. Neurosurg Clin N Am 1991;2:31–42. Livuah O, Stura EA, Middleton SA, Johnson DC, Jokiffe CK, Wilson TA. Crystallographic evidence for performed divers of erythropoietin receptor before ligand activation. Science 1999;283:987–90. Lundborg G, Zhao Q, Kanje M, Danielsen N, Kerns JM. Can sensory and motor collateral sprouting be induced from intact peripheral nerve by end-to-side anastomosis? J Hand Surg 1994;19B:227–82. Marti HH, Wenger RH, Rivas LA, Straumann U, Digicaylioglu M, Henn V, et al. Erythropoietin gene expression in human, monkey and murine brain. Eur J Neurosci 1996;8:666–76. Morishita E, Masuda S, Nagao M, Yasuda Y, Sasaki R. Erythropoietin receptor is expressed in rat hippocampal and cerebral cortical neurons, and erythropoietin prevents in vitro glutamate-induced neuronal death. Neuroscience 1997;76:105–16.
M.G. Lykissas et al. / Journal of Neuroscience Methods 164 (2007) 107–115 Remy I, Wilson IA, Michnick SW. Erythropoietin receptor activation by a ligandinduced conformation change. Science 1999;283:990–3. Sekiguchi Y, Kikuchi S, Myers RR, Campana WM. Erythropoietin inhibits spinal neuronal apoptosis and pain following nerve root crush. Spine 2003;28:2577–84. Sir´en AL, Knerlich F, Poser W, Gleiter CH, Bruck W, Ehrenreich H. Erythropoietin and erythropoietin receptor in human ischemic/hypoxic brain. Acta Neuropathol (Berl) 2001;101:271–6. Slack JR, Hopkins WG, Williams MN. Nerve sheaths and motoneurone collateral sprouting. Nature 1979;282:506–7. Trumble TE. Peripheral nerve transplantation: the effects of predegenerated grafts and immunosuppression. J Neural Transplant Plast 1992;3:39–49.
115
Tsai PT, Ohab JJ, Kertesz N, Groszer M, Matter C, Gao J, et al. A critical role of erythropoietin receptor in neurogenesis and post-stroke recovery. J Neurosci 2006;26:1269–74. Viterbo F, Trindade JC, Hoshino K, Mazzoni Neto A. Latero-terminal neurorrhaphy without removal of the epineural sheath: experimental study in rats. Rev Paul Med 1992;110:267–75. Viterbo F, Trindade JC, Hoshino K, Mazzoni Neto A. End-to-side neurorrhaphy with removal of the epineurial sheath: An experimental study in rats. Plast Reconstr Surg 1994;94:1038–47. Zhang Z, Johnson EO, Vekris MD, Zoubos AB, Bo J, Beris AE, et al. Long-term evaluation of rabbit peripheral nerve repair with end-to-side neurorrhaphy in rabbits. Microsurgery 2006;26:245–52.