Hindawi Publishing Corporation Neural Plasticity Volume 2016, Article ID 4969523, 13 pages http://dx.doi.org/10.1155/2016/4969523 Research Article Dose-Dependent Differential Effect of Neurotrophic Factors on In Vitro and In Vivo Regeneration of Motor and Sensory Neurons Daniel Santos, Francisco Gonzalez-Perez, Xavier Navarro, and Jaume del Valle Institute of Neurosciences, Department of Cell Biology, Physiology and Immunology, Universitat Autònoma de Barcelona and Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), Bellaterra, Spain Correspondence should be addressed to Jaume del Valle; jaume.delvalle@uab.cat Received 21 June 2016; Revised 7 September 2016; Accepted 27 September 2016 Academic Editor: Fengquan Zhou Copyright © 2016 Daniel Santos et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Although peripheral axons can regenerate after nerve transection and repair, functional recovery is usually poor due to inaccurate reinnervation. Neurotrophic factors promote directional guidance to regenerating axons and their selective application may help to improve functional recovery. Hence, we have characterized in organotypic cultures of spinal cord and dorsal root ganglia the effect of GDNF, FGF-2, NGF, NT-3, and BDNF at different concentrations on motor and sensory neurite outgrowth. In vitro results show that GDNF and FGF-2 enhanced both motor and sensory neurite outgrowth, NGF and NT-3 were the most selective to enhance sensory neurite outgrowth, and high doses of BDNF selectively enhanced motor neurite outgrowth. Then, NGF, NT-3, and BDNF (as the most selective factors) were delivered in a collagen matrix within a silicone tube to repair the severed sciatic nerve of rats. Quantification of Fluorogold retrolabeled neurons showed that NGF and NT-3 did not show preferential effect on sensory regeneration whereas BDNF preferentially promoted motor axons regeneration. Therefore, the selective effects of NGF and NT-3 shown in vitro are lost when they are applied in vivo, but a high dose of BDNF is able to selectively enhance motor neuron regeneration both in vitro and in vivo. 1. Introduction trophic support [5]. In fact, some authors argue that the key point for preferential attraction of axons to their adequate After peripheral nerve injury, transected axons can regenerate targets is the expression of trophic factors by their own target and reinnervate target organs. However, reinnervation of organ and the distal stump [2, 3]. distal organs and functional recovery are often deficient After nerve injury, in order to support neuronal survival because random regeneration of axons results in aberrant and enhance regeneration, motoneurons in spinal cord (SC) target reinnervation [1]. Thus, specificity of reinnervation is and sensory neurons in dorsal root ganglia (DRG) synthe- a key issue to improve functional recovery after peripheral size and secrete neurotrophic factors (NTFs). Furthermore, nerve injuries. within the proximal and the distal nerve stumps, denervated Although some studies described that motor axons pref- Schwann cells become growth supportive and secrete several erentially reinnervate muscular pathways [2], this accuracy NTF and cytokines that follow different patterns of expres- is compromised when optimal conditions such as pure sion, with an initial upregulation of NGF, BDNF, and GDNF, muscular and cutaneous branches, matching distal nerve whereas others, such as NT-3 or CNTF, are downregulated caliber, and short separation between nerve stumps are not [6]. However, changes in NTF levels are severalfold higher at met [3, 4]. It has been suggested that although motor or the lesion area than in the SC or the DRG where the somata sensory neurons tend to regenerate through their original of the axotomized neurons are located [7]. Moreover, it has pathway, axons sense the levels of trophic factors in distal been described that some NTF may influence the direction branches and then grow towards the target that offers more of regenerating axons on certain models of regeneration 2 Neural Plasticity [8]. Thus, modifying the regenerative microenvironment is primary antibody mouse RT97 (1 : 200, Developmental Stud- a promising approach to modulate and selectively enhance ies Hybridoma Bank) at 4∘ C. After three hours of washing, sensory and motor regeneration. the sections were incubated with secondary antibody AF594 However, high levels of NTF or a prolonged delivery over conjugated donkey anti-mouse (1 : 200, Life Technologies) time can in fact induce no improvement or even result in overnight at 4∘ C. After two more washes, samples were deleteriousness in terms of regeneration [9–11], inducing neu- mounted on slides with Mowiol. Cultures were visualized rotoxicity [12] and endoneurial sprouting and hyperalgesia with an Olympus BX51 fluorescence microscope; images of [13]. Therefore, as low levels of NTF may not reach ther- different areas were taken with a DP50 camera attached to apeutical action and high levels may disrupt regeneration, a computer with Cell A software (Olympus) and merged a comparative study of the concentrations and delivery of using Adobe Photoshop CS3 (Adobe System). Whole culture NTF to specifically improve regeneration of motor or sensory images were analyzed with the Neurite-J plug-in [17] for neurons was undertaken. We analyzed the effect of GDNF, ImageJ software [18] and the number of neurites grown at FGF-2, NGF, NT-3, and BDNF at different concentrations on different distances from the explant was compared between motor and sensory axonal regeneration using in vitro and in vivo models. different conditions of the cultures. The area under the curve (AUC) for each group was calculated by the linear trapezoidal 2. Materials and Methods method using the amount of neurites every 25 𝜇m. 2.1. Ethical Guidelines. Both in vitro and in vivo experimental 2.4. In Vivo Study of Peripheral Nerve Regeneration. Only procedures were approved by the Ethical Committee of the NTFs which showed trophic selectivity for motor and the Universitat Autònoma de Barcelona in accordance with sensory outgrowth were further tested in vivo. Thus, NGF, the European Communities Council Directive 2010/63/EU. NT-3, and BDNF at 1, 2, and 10 𝜇g/mL each were added to a Adult rats were anaesthetized with pentobarbital sodium 2 mg/mL collagen solution prepared as above. These mixtures (40 mg/kg i.p.). P7 and adult rats were euthanized by pento- were used to fill silicone tubes (8 mm long, 3 mm wide, and barbital sodium overdose (200 mg/kg i.p.). 2 mm i.d.) that were maintained vertically for 12 h to promote 2.2. DRG and SC Organotypic Cultures. Organotypic cultures collagen fibril alignment during gel formation [19]. A collagen were prepared as previously described [14]. Briefly, collagen I matrix without NTF was used in the control group. (rat tail, #354236, Corning) diluted in basal Eagle’s medium Female Sprague-Dawley rats weighing 250–300 g (𝑛 = (Gibco) and sodium bicarbonate at 0.3 mg/mL was prepared 4 group) were anaesthetized, the sciatic nerve was exposed as control condition. NGF, NT-3, GDNF, and BDNF (Pepro- at the midthigh and sectioned 90 mm from the tip of the tech) were added to achieve concentrations of 5, 10, 50, or third toe, and a 6 mm nerve portion distal to the section 100 ng/mL. As FGF-2 has been reported to work at higher was resected. The prepared tube was then sutured with 10- concentrations [15], FGF-2 18 kDa (Peprotech) was prepared 0 monofilament sutures to each nerve end leaving a 6 mm at 25, 50, 250, and 500 ng/mL. Finally, 30 𝜇L single drops gap between nerve stumps. Animals were left to recover on of the prepared matrices were deposited on poly-d-lysine a warming pad and then housed with littermates. They were (1 g/mL, Sigma) coated 24-well multidishes (Iwaki, Asahi kept on standard laboratory conditions with a light-dark cycle Technoglass, Chiba, Japan) and were left to gel for 2 h at 37∘ C of 12 : 12 h and ad libitum access to food and tap water. All and 5% CO2 in the incubator. efforts were made to minimize pain and animal discomfort SC lumbar segments and lumbar DRG were harvested during surgery and recovery. from 7-day-old Sprague-Dawley rats, placed in 4∘ C Gey’s balanced salt solution (Sigma) enriched with 6 mg/mL glu- 2.5. Retrograde Labeling and Neuronal Counting. Rats were cose and cleaned from blood and meningeal debris. SCs anesthetized 20 days after operation with pentobarbital were cut with a McIlwain Tissue Chopper in 350 𝜇m thick sodium; the sciatic nerve was exposed and transected 8 mm slices. SC slices and DRG explants were placed on top of distal to the distal end of the silicone tube. Then, the tip of the collagen matrix and covered by a second 30 𝜇L drop of the severed nerve was soaked into 5 𝜇L Fluorogold (FG; 5%; the same solution. After 45 min in the incubator, samples Fluorochrome Inc.) for 1 h in a Vaseline well. After retrieval were embedded with 0.5 mL of neurobasal medium (Life of the well, saline was flushed to clean remnants of the tracer Technologies), supplemented with B27 (Life Technologies) before suturing the wound in planes. Seven days later, the rats and glutamine and penicillin/streptomycin (Sigma). After were deeply anesthetized and transcardially perfused with 4% one day in culture, the medium of SC cultures was removed paraformaldehyde in PBS. The lumbar segment (L3–L6) of and changed by a penicillin/streptomycin-free medium. DRG the SC and L4 and L5 DRG were removed, postfixed in the explants were cultured for 2 days and SC slices for 4 days. same fixative solution for 1 h, and transferred to 30% sucrose A detailed description of this protocol has been previously in PBS. The SC and DRG were cut longitudinally in 40 and reported [16]. 20 𝜇m thick sections, respectively, in a cryostat and mounted on slides. Sections were observed with an Olympus BX51 2.3. Neurite Outgrowth Analysis. SC and DRG cultures were fluorescence microscope under UV light and the number of fixed with 4% paraformaldehyde in PBS for 30 min. After- labeled neurons was counted in every third section following wards, SC and DRG samples were incubated for 48 h with the fractionator principle [20]. Neural Plasticity 3 2.6. Data Analysis. Data are presented as mean ± SEM. NT-3 at any of the concentrations compared to the control Results were statistically analyzed by using GraphPad Prism cultures (Figures 2(g)–2(j)). In contrast, addition of NGF (GraphPad Software, USA). Student’s 𝑡-test and one-way (Figures 2(k) and 2(m)) or NT-3 (Figures 2(l) and 2(n)) in ANOVA followed by Bonferroni’s post hoc test for compar- the collagen matrix yielded more sensory neurites at different ison between groups were used when applicable. Statistical distances and larger values of the AUC in comparison with significance was considered when 𝑝 value was < 0.05. controls (Figures 2(m) and 2(n)), the 50 ng/mL dose being the one that showed the highest values for NGF and NT-3. In agreement with previous studies [24], NGF and NT-3 are able 3. Results and Discussion to promote exclusively sensory neurite outgrowth, without BDNF, NGF, NT-3, GDNF, and FGF-2 at different concen- enhancing motor axon regeneration from SC slices (Figures trations were tested to elucidate the optimal concentration 4(c) and 4(d)). This selective effect may be explained because of each NTF that is able to enhance either motor or sensory motoneurons do not express TrkA receptor and after nerve neurite outgrowth without affecting the other neuronal pop- injury the expression levels of TrkC receptor remain relatively ulation. For that purpose, 3D organotypic cultures of SC slices unchanged [7]. and DRG were used (Figures 1–3). Taking advantage of the differential promotion of sensory but not motor neurite outgrowth, these NTFs were tested in 3.1. GDNF and FGF-2 Enhance Both Motor and Sensory vivo in a model of nerve regeneration. 20 days after section of Neurite Outgrowth. Almost all doses of both FGF-2 and the sciatic nerve and tube repair, all the rats showed evidence GDNF increased motor and sensory neurite outgrowth with of axonal regeneration, as judged by the retrograde labeling of respect to the control substrate (Figure 1). Regarding motor motor (Figures 5(a) and 5(b)) and sensory (Figures 5(c) and neurite outgrowth, FGF-2 showed a progressive dose effect 5(d)) regenerated neurons. Concerning motor axon regener- with the lowest concentrations not being able to reach ation, both NGF and NT-3 groups unexpectedly improved significant differences versus the control group (Figures 1(g) motor axon regeneration at all doses (Figure 5(e)). Indeed, and 1(i)), while GDNF enhanced motor neurite outgrowth although some studies have already described that NGF with no dose-dependency as all the groups exhibit similar administration in a fibrin depot improves motor regeneration curves (Figures 1(h) and 1(j)). On the other hand, sensory [25], the common belief is that NGF and NT-3 are prosensory neurites growth from DRG increased similarly at all the NTFs [6]. Regarding sensory neurons, as expected, NGF and tested concentrations of FGF-2 (Figures 1(k) and 1(m)), NT-3 groups showed a significantly increased number of while GDNF followed a dose-dependent increase showing regenerated neurons at all doses, except the lowest of NT-3 the high doses even significantly larger AUC values than (1 𝜇g/mL) (Figure 5(f)). the low ones (Figure 1(n)). Comparatively, using the data Thus, in vivo results with NGF and NT-3 seem to contra- of the AUC in each culture condition, FGF-2 promoted a dict the in vitro observations where only sensory neurite out- maximum increase of about 9 and 3 times in neurite growth growth was improved. Differences between the in vitro and in of both motor and sensory neurites, whereas GDNF induced vivo models should be first taken into account. Organotypic maximal increase of 11 and 8 times baseline, respectively cultures are multicellular in vitro models, in which neurons (Figures 4(a) and 4(b)). remain embedded in contact with accompanying glial cells. These results in vitro reveal that neither GDNF nor FGF- Although Schwann cells and fibroblasts migrate outside the 2 show a preference for motor or sensory neuritogenesis, slice in organotypic cultures interacting with and giving as both neuronal populations are enhanced at the different structural support to the newborn neurites [26], the amount concentrations used. While GDNF has already been reported and activation of these cells are probably lower than in vivo to promote neurite outgrowth of both populations, FGF- conditions in which cells from the cut nerve stumps migrate 2 was described to preferentially enhance motor neurite inside the tube to stimulate nerve regeneration. In addition, outgrowth when measuring the length of longest neurites the recruitment of hematogenous inflammatory cells that [15]. Since reliable measurements of neurite outgrowth anal- plays an important role during Wallerian degeneration in ysis in organotypic cultures are complicated [21], we used vivo is absent in vitro. Another important difference attains a semiautomatic analysis that works as an adaption of the the role of extracellular matrix components. In our in vitro Sholl method [17] to improve accuracy and reproducibility cultures, the matrix was made of collagen I only, whereas as shown in other works [22, 23]. Thus, the differences in vivo reactive Schwann cells and fibroblasts secrete several in methods and variables to quantify neurite outgrowth in neurotropic molecules, including laminin and fibronectin these studies may in fact explain some controversies. Hence, [6]. It has been shown that neurite outgrowth in response in accordance with our results, we discarded GDNF and to neurotrophins, such as NGF, is modulated by the compo- FGF-2 for in vivo studies because of their lack of selective sition and density of the extracellular matrix modifying the effect. interactions with supporting Schwann cells [27]. After nerve damage, levels of NGF mRNA rise rapidly 3.2. NGF and NT-3 Selectively Enhance Sensory Regeneration in the nonneuronal cells of the damaged nerve [28]. It is In Vitro but Not In Vivo. Observation of cultures of SC slices also known that NGF and other neurotrophins interact not and DRG with NGF and NT-3 (Figures 2(a)–2(f)) revealed only with axons but also with Schwann cells and fibroblasts they were the most selective tested factors for sensory regen- within the regenerative microenvironment [29]. Taking into eration. Motor neurite outgrowth was not affected by NGF or account that Schwann cells dedifferentiate after injury and 4 Neural Plasticity (a) (d) (b) (e) (c) (f) Motor neurites outgrowth Motor neurites outgrowth 125 125 100 100 Number of neurites Number of neurites 75 75 50 50 25 25 0 0 250 500 750 1000 200 400 600 800 1000 Distance (𝜇m) Distance (𝜇m) Control FGF 250 ng/mL Control GDNF 50 ng/mL FGF 25 ng/mL FGF 500 ng/mL GDNF 5 ng/mL GDNF 100 ng/mL FGF 50 ng/mL GDNF 10 ng/mL (g) (h) Motor neurites outgrowth Motor neurites outgrowth 40000 40000 Area under the curve (neurite × 𝜇m) Area under the curve (neurite × 𝜇m) ∗ ∗ 30000 ∗ 30000 ∗ ∗ ∗ 20000 20000 10000 10000 Control FGF 25 ng/mL FGF 50 ng/mL FGF 250 ng/mL FGF 500 ng/mL GDNF 100ng/mL Control GDNF 5 ng/mL GDNF 50 ng/mL GDNF 10 ng/mL (i) (j) Figure 1: Continued. Neural Plasticity 5 Sensory neurites outgrowth Sensory neurites outgrowth 150 150 Number of neurites Number of neurites 100 100 50 50 0 0 250 500 750 1000 250 500 750 1000 Distance (𝜇m) Distance (𝜇m) Control FGF 250 ng/mL Control GDNF 50 ng/mL FGF 25 ng/mL FGF 500 ng/mL GDNF 5 ng/mL GDNF 100 ng/mL FGF 50 ng/mL GDNF 10 ng/mL (k) (l) Sensory neurites outgrowth Sensory neurites outgrowth # # 80000 80000 Area under the curve (neurite × 𝜇m) Area under the curve (neurite × 𝜇m) # ∗ 60000 # 60000 ∗ ∗ 40000 40000 ∗ ∗ ∗ ∗ 20000 20000 Control GDNF 5 ng/mL GDNF 50 ng/mL GDNF 10 ng/mL GDNF 100 ng/mL Control FGF 25 ng/mL FGF 50 ng/mL FGF 250 ng/mL FGF 500 ng/mL (m) (n) Figure 1: Representative images of RT97 stained neurites from spinal cord slices (a–c) and DRG neurons (d–f) cultured within a 3D collagen matrix alone (a, d) and with addition of 500 ng/mL of FGF-2 (b, e) or 100 ng/mL of GDNF (c, f). Quantification of the number of neurites grown at increasing distance from the spinal cord slices (g, h) and from DRG body (k, l) after the addition of FGF-2 and GDNF. Plots of the quantified AUC from (g), (h), (k), and (l) graphs for motor (i, j) or sensory (m, n) neurite outgrowth. Data expressed as mean ± SEM. ∗𝑝 < 0.05 versus control; # 𝑝 < 0.05. Scale bar: 100 𝜇m (a) and 200 𝜇m (b–f). change their phenotype to a proregenerative state [6, 30], with control values (Figure 3(g)), the 50 ng/mL being the secreting a variety of tropic and trophic molecules, both NGF concentration that produced the highest values in terms and NT-3 might enhance axonal regeneration indirectly and of neurites length (Figure 3(g)) and AUC (Figure 3(h)). indiscriminately by stimulating Schwann cell proliferation Interestingly, a higher concentration of 100 ng/mL did not and release of other NTFs such as BDNF or GDNF [31, 32]. promote neurite growth, suggesting a window of dose effect, in agreement with previous data [7, 9]. On the other hand, 3.3. BDNF Selectively Enhances Motor Regeneration In Vitro BDNF slightly enhanced sensory neurite outgrowth only at and In Vivo. BDNF was the NTF showing the most selective low concentrations of 5 and 10 ng/mL (Figures 3(e), 3(i), effect on motoneuron regeneration in vitro (Figure 3). Addi- and 3(j)) while higher levels reverted this increase. These tion of BDNF to the medium (Figures 3(b) and 3(c)) showed results suggest a preferential motor profile for BDNF with a an increase in motor neurite outgrowth in comparison high dose of BDNF (50 ng/mL) increasing about 10-fold the 6 Neural Plasticity (a) (d) (b) (e) (c) (f) Motor neurites outgrowth Motor neurites outgrowth 125 125 100 100 Number of neurites Number of neurites 75 75 50 50 25 25 0 0 250 500 750 1000 200 400 600 800 1000 Distance (𝜇m) Distance (𝜇m) Control NGF 50 ng/mL Control NT-3 50 ng/mL NGF 5 ng/mL NGF 100 ng/mL NT-3 5 ng/mL NT-3 100 ng/mL NGF 10 ng/mL NT-3 10 ng/mL (g) (h) Motor neurites outgrowth Motor neurites outgrowth 40000 40000 Area under the curve (neurite × 𝜇m) Area under the curve (neurite × 𝜇m) 30000 30000 20000 20000 10000 10000 Control NT-3 5 ng/mL NT-3 10 ng/mL NT-3 50 ng/mL NT-3 100 ng/mL Control NGF 5 ng/mL NGF 10 ng/mL NGF 50 ng/mL NGF 100 ng/mL (i) (j) Figure 2: Continued. Neural Plasticity 7 Sensory neurites outgrowth Sensory neurites outgrowth 150 150 Number of neurites Number of neurites 100 100 50 50 0 0 250 500 750 1000 250 500 750 1000 Distance (𝜇m) Distance (𝜇m) Control NGF 50 ng/mL Control NT-3 50 ng/mL NGF 5 ng/mL NGF 100 ng/mL NT-3 5 ng/mL NT-3 100 ng/mL NGF 10 ng/mL NT-3 10 ng/mL (k) (l) Sensory neurites outgrowth 80000 Sensory neurites outgrowth 80000 Area under the curve (neurite × 𝜇m) # Area under the curve (neurite × 𝜇m) # 60000 60000 # ∗ # # ∗ ∗ 40000 40000 ∗ ∗ ∗ 20000 20000 Control NGF 5 ng/mL NGF 10 ng/mL NGF 50 ng/mL NGF 100 ng/mL Control NT-3 5 ng/mL NT-3 10 ng/mL NT-3 50 ng/mL NT-3 100 ng/mL (m) (n) Figure 2: Representative images of RT97 stained neurites from spinal cord slices (a–c) and DRG neurons (d–f) cultured within a 3D collagen matrix alone (a, d) and with addition of 50 ng/mL of NGF (b, e) or NT-3 (c, f). Quantification of the number of neurites grown at increasing distance from the spinal cord slices (g, h) and from DRG body (k, l) after the addition of NGF and NT-3. Plots of the quantified AUC from (g), (h), (k), and (l) graphs for motor (i, j) or sensory (m, n) neurite outgrowth. Data expressed as mean ± SEM. ∗ 𝑝 < 0.05 versus control; #𝑝 < 0.05. Scale bar: 200 𝜇m (a–f). growth of motor neurites from SC slices but without effect on The BDNF receptor TrkB is expressed constitutively in DRG neurite outgrowth (Figure 4(e)). motoneurons and its levels increase after injury [7], while Following the in vitro results, we tested if BDNF could intact sensory neurons show low levels of TrkB, although its selectively promote motor axon regeneration in vivo after expression is also upregulated after injury [33]. Similarly, the nerve section and tube repair. Quantification of retrolabeled endogenous expression of BDNF is also upregulated in the neurons revealed that 20 days after surgery all the groups distal stump after nerve injury [7]. On the other hand, it treated with BDNF had higher number of regenerated motor has been suggested that BDNF upregulates TrkB expression neurons than the control group (Figure 5(e)). In parallel at low doses [34], but high levels of this neurotrophin may to the in vitro results, BDNF only increased the number downregulate this receptor in neuronal cells [35]. Thus, it of regenerated sensory neurons at a low dose (1 𝜇g/mL) seems that BDNF can modulate motor axon growth under while higher doses gave similar result to the control group a bimodal profile. Indeed, Boyd and Gordon demonstrated (Figure 5(f)). a biphasic effect in which low doses of exogenous BDNF 8 Neural Plasticity (a) (d) (b) (e) (c) (f) 40000 Area under the curve (neurite × 𝜇m) 125 Motor neurites outgrowth Motor neurites outgrowth ∗ 100 30000 Number of neurites 75 ∗ ∗ 20000 50 10000 25 0 250 500 750 1000 BDNF 5 ng/mL BDNF 10 ng/mL BDNF 50 ng/mL Control BDNF 100 ng/mL Distance (𝜇m) Control BDNF 50 ng/mL BDNF 5 ng/mL BDNF 100 ng/mL BDNF 10 ng/mL (g) (h) 80000 Area under the curve (neurite × 𝜇m) Sensory neurites outgrowth 150 Sensory neurites outgrowth 60000 Number of neurites 100 # # 40000 ∗ 50 ∗ 20000 0 250 500 750 1000 BDNF 5 ng/mL BDNF 10 ng/mL BDNF 50 ng/mL Control BDNF 100 ng/mL Distance (𝜇m) Control BDNF 50 ng/mL BDNF 5 ng/mL BDNF 100 ng/mL BDNF 10 ng/mL (i) (j) Figure 3: Representative images of RT97 stained neurites from spinal cord slices (a–c) and DRG neurons (d–f) cultured within a 3D collagen matrix alone (a, d) and with addition of 10 ng/mL (b, e) and 50 ng/mL of BDNF (c, f). Quantification of the number of neurites grown at increasing distance from the cord slices (g) and from DRG body (i) after the addition of different doses of BDNF. Plots of the quantified area under each curve from (g) and (i) graphs for motor (h) or sensory neurite outgrowth (j). Data expressed as mean ± SEM. ∗ 𝑝 < 0.05 versus control; # 𝑝 < 0.05. Scale bar: 200 𝜇m (a–f). Neural Plasticity 9 FGF GDNF 14 14 ∗ 12 ∗ 12 Fold increase versus control Fold increase versus control 10 ∗ 10 8 8 6 6 4 4 2 2 0 0 25 50 250 500 5 10 50 100 (ng/mL) (ng/mL) MN MN SN SN (a) (b) NGF NT-3 14 14 12 12 Fold increase versus control Fold increase versus control 10 10 8 8 ∗ 6 6 ∗ ∗ 4 ∗ 4 ∗ ∗ 2 2 0 0 5 10 50 100 5 10 50 100 (ng/mL) (ng/mL) MN MN SN SN (c) (d) BDNF 14 ∗ 12 Fold increase versus control 10 8 6 4 2 0 5 10 50 100 (ng/mL) MN SN (e) Figure 4: Histogram of the fold increase of the AUC for motor and sensory neurite outgrowth induced by FGF (a), GDNF (b), NGF (c), NT-3 (d), and BDNF (e) at the concentrations tested compared to the control collagen matrix. Data expressed as mean ± SEM. ∗ 𝑝 < 0.05. 10 Neural Plasticity (a) (b) (c) (d) Spinal cord Dorsal root ganglia 1500 8000 ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ Regenerated sensory neurons ∗ ∗ Regenerated motor neurons ∗ ∗ ∗ ∗ 6000 ∗ 1000 4000 500 2000 0 0 Ctrl 1 2 10 1 2 10 1 2 10 Ctrl 1 2 10 1 2 10 1 2 10 (𝜇g/mL) (𝜇g/mL) NGF NGF NT-3 NT-3 BDNF BDNF (e) (f) Figure 5: Representative micrographs of neurons retrolabeled with FG in the spinal cord (a, b) and DRG (c, d) growing in control conditions (a, c), with 10 𝜇g/mL of BDNF (b) and 10 𝜇g/mL of NGF (d). Histogram of the number of regenerated motor (e) and sensory (f) neurons after sciatic nerve section and conduit repair with NGF, NT-3, and BDNF at different doses. Data expressed as mean ± SEM. ∗ 𝑝 < 0.05. Scale bar: 500 𝜇m (a–d). increased the number of chronically axotomized motoneu- such as tubulin or GAP-43 [36]. In contrast, high doses rons which regenerated their axons whereas high doses pro- would downregulate TrkB receptors, minimizing the effect of gressively reduced the number [9]. The low doses upregulate BDNF or even promoting an inhibitory effect via activation TrkB receptors and enhance both sensory and motor neurite of the p75 receptor [37]. The differential effect of the same growth promoting the expression of growth-associated genes doses on motor and sensory axonal regeneration might be Neural Plasticity 11 explained because of differences in the expression of these project NEBIAS under Contract no. 611687, FEDER funds, receptors and their contribution to neurotrophin transport and TERCEL (RD12/0019/0011) and CIBERNED (CB06/ in motor and sensory neurons [38, 39]. On the other hand, 05/1105) funds from the Instituto de Salud Carlos III of Spain. the fact that the high dose of BDNF used in our study in vivo The authors thank Monica Espejo and Marta Morell for their promoted motor axon regeneration can be explained because technical help. The RT97 antibody was obtained from the it is much lower than the doses shown to be inhibitory Developmental Studies Hybridoma Bank developed under when applied by continuous infusion [9]. Other studies have the auspices of the NICHD and maintained by the University proved that local release of BDNF increases the number of of Iowa, Department of Biology. regenerated axons through grafts or conduits [40, 41], and a selective increase in the number of motor axons was noted References with gene-induced BDNF overexpression [42]. At longer term after sciatic nerve injury and repair, administration of [1] A. Valero-Cabré and X. Navarro, “Functional impact of axonal exogenous BDNF enhanced motor functional recovery in misdirection after peripheral nerve injuries followed by graft or some studies [41, 43] but not in others [37, 42, 44]. Thus, the tube repair,” Journal of Neurotrauma, vol. 19, no. 11, pp. 1475– early administration of adequate concentrations of BDNF to 1485, 2002. axotomized motoneurons may be sufficient to sustain initial [2] T. M. E. Brushart, “Preferential reinnervation of motor nerves axon growth, but it may fail to support regeneration and by regenerating motor axons,” The Journal of Neuroscience, vol. target reinnervation for long time. 8, no. 3, pp. 1026–1031, 1988. Schwann cells from motor and sensory nerve branches [3] R. D. Madison, S. J. Archibald, and T. M. Brushart, “Reinner- express different molecular markers that may contribute vation accuracy of the rat femoral nerve by motor and sensory to the capacity of axons to specifically regenerate towards neurons,” The Journal of Neuroscience, vol. 16, no. 18, pp. 5698– 5703, 1996. appropriate pathways. Particularly, Schwann cells associated with motor but not with sensory axons express the HNK1 [4] G. A. Robinson and R. D. Madison, “Motor neurons can pref- erentially reinnervate cutaneous pathways,” Experimental Neu- carbohydrate epitope [45]. In addition, Schwann cells of rology, vol. 190, no. 2, pp. 407–413, 2004. sensory and motor nerves respond differently during den- [5] G. A. Robinson and R. D. Madison, “Manipulations of the ervation, by overexpressing different types of NTFs [46]. mouse femoral nerve influence the accuracy of pathway rein- However, such differences tend to decline with time after nervation by motor neurons,” Experimental Neurology, vol. 192, injury, suggesting that the endogenous production of factors no. 1, pp. 39–45, 2005. may not contribute enough to the sorting of different types [6] I. Allodi, E. Udina, and X. Navarro, “Specificity of peripheral of axons during nerve regeneration [6]. Interestingly, it has nerve regeneration: interactions at the axon level,” Progress in been shown that the upregulation of HNK-1 induced by Neurobiology, vol. 98, no. 1, pp. 16–37, 2012. electrical stimulation of the injured nerve is dependent on [7] J. G. Boyd and T. Gordon, “Neurotrophic factors and their BDNF and its receptor TrkB [47], which are also increased receptors in axonal regeneration and functional recovery after by the electrical stimulation [48]. These observations suggest peripheral nerve injury,” Molecular Neurobiology, vol. 27, no. 3, one mechanism by which exogenously modulating the local pp. 277–323, 2003. expression of BDNF may help to attract the regeneration of [8] A. Markus, T. D. Patel, and W. D. Snider, “Neurotrophic factors motor axons. and axonal growth,” Current Opinion in Neurobiology, vol. 12, no. 5, pp. 523–531, 2002. 4. Conclusion [9] J. G. Boyd and T. Gordon, “A dose-dependent facilitation and inhibition of peripheral nerve regeneration by brain-derived The results of this study provide a comparative analysis of neurotrophic factor,” European Journal of Neuroscience, vol. 15, the optimal doses for stimulating motor and sensory axonal no. 4, pp. 613–626, 2002. regeneration both in vitro and in vivo for different NTFs [10] G. Piquilloud, T. Christen, L. A. Pfister, B. Gander, and M. Y. (GDNF, FGF-2, NGF, NT-3, and BDNF). Optimal concen- Papaloı̈zos, “Variations in glial cell line-derived neurotrophic trations of GDNF and FGF-2 show the highest potentiation factor release from biodegradable nerve conduits modify the of both motor and sensory neuron regeneration. On the rate of functional motor recovery after rat primary nerve other hand, NGF and NT-3 show a selective enhancement repairs,” European Journal of Neuroscience, vol. 26, no. 5, pp. of sensory neurite growth in vitro that is lost in our in vivo 1109–1117, 2007. model. Finally, BDNF at selected doses selectively promotes [11] M. R. Tannemaat, R. Eggers, W. T. Hendriks et al., “Differential motor axonal growth both in vitro and in vivo. effects of lentiviral vector-mediated overexpression of nerve growth factor and glial cell line-derived neurotrophic factor on regenerating sensory and motor axons in the transected Competing Interests peripheral nerve,” European Journal of Neuroscience, vol. 28, no. 8, pp. 1467–1479, 2008. The authors declare that they have no competing interests. [12] H.-J. Kim, J. J. Hwang, M. M. Behrens, B. J. Snider, D. W. Choi, and J.-Y. Koh, “TrkB mediates BDNF-induced potentiation of Acknowledgments neuronal necrosis in cortical culture,” Neurobiology of Disease, vol. 14, no. 1, pp. 110–119, 2003. This research was supported by the European Union FP7-NMP [13] G. Ruiz, D. Ceballos, and J.-E. Baños, “Behavioral and histologi- project MERIDIAN under Contract no. 280778, FPT-ICT cal effects of endoneurial administration of nerve growth factor: 12 Neural Plasticity possible implications in neuropathic pain,” Brain Research, vol. [30] K. R. Jessen and R. Mirsky, “The origin and development of glial 1011, no. 1, pp. 1–6, 2004. cells in peripheral nerves,” Nature Reviews Neuroscience, vol. 6, [14] D. Santos, G. Giudetti, S. Micera, X. Navarro, and J. del Valle, no. 9, pp. 671–682, 2005. “Focal release of neurotrophic factors by biodegradable micro- [31] A. Acheson, P. A. Barker, R. F. Alderson, F. D. Miller, and R. spheres enhance motor and sensory axonal regeneration in vitro A. Murphy, “Detection of brain-derived neurotrophic factor- and in vivo,” Brain Research, vol. 1636, pp. 93–106, 2016. like activity in fibroblasts and Schwann cells: inhibition by [15] I. Allodi, L. Casals-Dı́az, E. Santos-Nogueira, F. Gonzalez- antibodies to NGF,” Neuron, vol. 7, no. 2, pp. 265–275, 1991. Perez, X. Navarro, and E. Udina, “FGF-2 low molecular weight [32] R. Klein, V. Nanduri, S. Jing et al., “The trkB tyrosine protein selectively promotes neuritogenesis of motor neurons in vitro,” kinase is a receptor for brain-derived neurotrophic factor and Molecular Neurobiology, vol. 47, no. 2, pp. 770–781, 2013. neurotrophin-3,” Cell, vol. 66, no. 2, pp. 395–403, 1991. [16] A. Torres-Espı́n, I. Allodi, D. Santos et al., “Analysis of axonal [33] L. A. Karchewski, K. A. Gratto, C. Wetmore, and V. M. K. Verge, growth in organotypic neural cultures,” Protocol Exchange, 2016. “Dynamic patterns of BDNF expression in injured sensory neurons: differential modulation by NGF and NT-3,” European [17] A. Torres-Espı́n, D. Santos, F. González-Pérez, J. del Valle, and Journal of Neuroscience, vol. 16, no. 8, pp. 1449–1462, 2002. X. Navarro, “Neurite-J: an Image-J plug-in for axonal growth analysis in organotypic cultures,” Journal of Neuroscience Meth- [34] I. Ferrer, J. Ballabriga, E. Martı́, E. Pérez, J. Alberch, and E. ods, vol. 236, pp. 26–39, 2014. Arenas, “BDNF up-regulates TrkB protein and prevents the death of CA1 neurons following transient forebrain ischemia,” [18] C. A. Schneider, W. S. Rasband, and K. W. Eliceiri, “NIH Image Brain Pathology, vol. 8, no. 2, pp. 253–261, 1998. to ImageJ: 25 years of image analysis,” Nature Methods, vol. 9, [35] L. Frank, “BDNF down-regulates neurotrophin responsiveness, no. 7, pp. 671–675, 2012. TrkB protein and TrkB mRNA levels in cultured rat hippocam- [19] E. Verdú, R. O. Labrador, F. J. Rodrı́guez, D. Ceballos, J. pal neurons,” European Journal of Neuroscience, vol. 8, no. 6, pp. Forés, and X. Navarro, “Alignment of collagen and laminin- 1220–1230, 1996. containing gels improve nerve regeneration within silicone [36] N. R. Kobayashi, D.-P. Fan, K. M. Giehl, A. M. Bedard, S. J. tubes,” Restorative Neurology and Neuroscience, vol. 20, no. 5, Wiegand, and W. Tetzlaff, “BDNF and NT-4/5 prevent atrophy pp. 169–179, 2002. of rat rubrospinal neurons after cervical axotomy, stimulate [20] H. J. Gundersen, “Stereology of arbitrary particles. A review of GAP-43 and T𝛼1-tubulin mRNA expression, and promote unbiased number and size estimators and the presentation of axonal regeneration,” The Journal of Neuroscience, vol. 17, no. 24, some new ones, in memory of William R. Thompson,” Journal pp. 9583–9595, 1997. of Microscopy, vol. 143, no. 1, pp. 3–45, 1986. [37] J. G. Boyd and T. Gordon, “The neurotrophin receptors, trkB [21] H. Al-Ali, S. Beckerman, J. L. Bixby, and V. P. Lemmon, “In vitro and p75, differentially regulate motor axonal regeneration,” models of axon regeneration,” Experimental Neurology, 2016. Journal of Neurobiology, vol. 49, no. 4, pp. 314–325, 2001. [22] L. M. Parra, M. Hartmann, S. Schubach, Y. Li, P. Herrlich, and A. [38] M. Rende, C. Provenzano, G. Stipa, and P. Tonali, “Effects of sci- Herrlich, “Distinct intracellular domain substrate modifications atic nerve grafts on choline acetyltransferase and p75 expression selectively regulate ectodomain cleavage of NRG1 or CD44,” in transected adult hypoglossal motoneurons,” Neuroscience, Molecular and Cellular Biology, vol. 35, no. 19, pp. 3381–3395, vol. 81, no. 2, pp. 517–525, 1997. 2015. [39] R. Curtis, J. R. Tonra, J. L. Stark et al., “Neuronal injury increases [23] D. Tukmachev, S. Forostyak, Z. Koci et al., “Injectable Extracel- retrograde axonal transport of the neurotrophins to spinal lular matrix hydrogels as scaffolds for spinal cord injury repair,” sensory neurons and motor neurons via multiple receptor Tissue Engineering—Part A, vol. 22, no. 3-4, pp. 306–317, 2016. mechanisms,” Molecular and Cellular Neurosciences, vol. 12, no. 3, pp. 105–118, 1998. [24] G. Terenghi, “Peripheral nerve regeneration and neurotrophic factors,” Journal of Anatomy, vol. 194, no. 1, pp. 1–14, 1999. [40] D. S. Utley, S. L. Lewin, E. T. Cheng, A. N. Verity, D. Sierra, and D. J. Terris, “Brain-derived neurotrophic factor and collagen [25] M. Jubran and J. Widenfalk, “Repair of peripheral nerve tran- tubulization enhance functional recovery after peripheral nerve sections with fibrin sealant containing neurotrophic factors,” transection and repair,” Archives of Otolaryngology—Head and Experimental Neurology, vol. 181, no. 2, pp. 204–212, 2003. Neck Surgery, vol. 122, no. 4, pp. 407–413, 1996. [26] I. Allodi, M.-S. Guzmán-Lenis, J. Hernàndez, X. Navarro, and [41] E. Vögelin, J. M. Baker, J. Gates, V. Dixit, M. A. Constantinescu, E. Udina, “In vitro comparison of motor and sensory neuron and N. F. Jones, “Effects of local continuous release of brain outgrowth in a 3D collagen matrix,” Journal of Neuroscience derived neurotrophic factor (BDNF) on peripheral nerve regen- Methods, vol. 198, no. 1, pp. 53–61, 2011. eration in a rat model,” Experimental Neurology, vol. 199, no. 2, [27] C. M. Madl and S. C. Heilshorn, “Matrix interactions modulate pp. 348–353, 2006. neurotrophin-mediated neurite outgrowth and pathfinding,” [42] S. A. Hoyng, F. De Winter, S. Gnavi et al., “A comparative mor- Neural Regeneration Research, vol. 10, no. 4, pp. 514–517, 2015. phological, electrophysiological and functional analysis of axon [28] M. Meyer, I. Matsuoka, C. Wetmore, L. Olson, and H. Thoenen, regeneration through peripheral nerve autografts genetically “Enhanced synthesis of brain-derived neurotrophic factor in modified to overexpress BDNF, CNTF, GDNF, NGF, NT3 or the lesioned peripheral nerve: different mechanisms are respon- VEGF,” Experimental Neurology, vol. 261, pp. 578–593, 2014. sible for the regulation of BDNF and NGF mRNA,” The Journal [43] P.-R. Ho, G. M. Coan, E. T. Cheng et al., “Repair with collagen of Cell Biology, vol. 119, no. 1, pp. 45–54, 1992. tubules linked with brain-derived neurotrophic factor and [29] I. Matsuoka, M. Meyer, and H. Thoenen, “Cell-type-specific ciliary neurotrophic factor in a rat sciatic nerve injury model,” regulation of nerve growth factor (NGF) synthesis in non- Archives of Otolaryngology—Head & Neck Surgery, vol. 124, no. neuronal cells: comparison of Schwann cells with other cell 7, pp. 761–766, 1998. types,” The Journal of Neuroscience, vol. 11, no. 10, pp. 3165–3177, [44] M. S. Moir, M. Z. Wang, M. To, J. Lum, and D. J. Terris, “Delayed 1991. repair of transected nerves: effect of brain-derived neurotrophic Neural Plasticity 13 factor,” Archives of Otolaryngology—Head and Neck Surgery, vol. 126, no. 4, pp. 501–505, 2000. [45] R. Martini, Y. Xin, B. Schmitz, and M. Schachner, “The L2/ HNK-1 carbohydrate epitope is involved in the preferential outgrowth of motor neurons on ventral roots and motor nerves,” European Journal of Neuroscience, vol. 4, no. 7, pp. 628– 639, 1992. [46] A. Höke, R. Redett, H. Hameed et al., “Schwann cells express motor and sensory phenotypes that regulate axon regeneration,” Journal of Neuroscience, vol. 26, no. 38, pp. 9646–9655, 2006. [47] K. A. Eberhardt, A. Irintchev, A. A. Al-Majed et al., “BDNF/ TrkB signaling regulates HNK-1 carbohydrate expression in regenerating motor nerves and promotes functional recovery after peripheral nerve repair,” Experimental Neurology, vol. 198, no. 2, pp. 500–510, 2006. [48] A. A. Al-Majed, T. M. Brushart, and T. Gordon, “Electrical stimulation accelerates and increases expression of BDNF and trkB mRNA in regenerating rat femoral motoneurons,” Euro- pean Journal of Neuroscience, vol. 12, no. 12, pp. 4381–4390, 2000.