Femoral nerve regeneration and its accuracy under different injury mechanisms.

Surgical accuracy has greatly improved with the advent of microsurgical techniques. However, complete functional recovery after peripheral nerve injury has not been achieved to date. The mechanisms hindering accurate regeneration of damaged axons after peripheral nerve injury are in urgent need of exploration. The present study was designed to explore the mechanisms of peripheral nerve regeneration after different types of injury. Femoral nerves of rats were injured by crushing or freezing. At 2, 3, 6, and 12 weeks after injury, axons were retrogradely labeled using 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (Dil) and True Blue, and motor and sensory axons that had regenerated at the site of injury were counted. The number and percentage of Dil-labeled neurons in the anterior horn of the spinal cord increased over time. No significant differences were found in the number of labeled neurons between the freeze and crush injury groups at any time point. Our results confirmed that the accuracy of peripheral nerve regeneration increased with time, after both crush and freeze injury, and indicated that axonal regeneration accuracy was still satisfactory after freezing, despite the prolonged damage.

Introduction

Five percent of patients in trauma centers are affected by peripheral nerve injury (Belkas et al., 2004; Taylor et al., 2008). Since most of these patients are of prime working age, such injuries pose a serious economic burden to society (Whitlock et al., 2009). Restoration of function after peripheral nerve injury remains one of the toughest challenges in surgery. Despite the development of microsurgical equipment, with which it is now possible to suture the damaged nerves accurately within a short time after injury, complete functional recovery after peripheral nerve damage has not yet been achieved (Eser et al., 2009). Various methods have been attempted (Vetter et al., 2010; Bosse et al., 2012; Seidel et al., 2013), but none have obtained complete success, mainly because the mechanisms underlying selective peripheral nerve regeneration remain poorly understood (Calvo et al., 2012; Jesuraj et al., 2012).

To achieve full recovery from nerve injury, high accuracy nerve innervation is just as important as robust axonal regrowth (Ruiter et al., 2011). Axons regenerated from proximal motor nerve stumps should grow into the motor nerve pathways in the distal stump, and vice versa. However, little is known about the mechanism underlying the misdirection of regenerating axons. To achieve accurate regeneration and full functional recovery, more research must be carried out to explore how the regeneration accuracy of axons changes after injury. Novel experimental approaches, such as retrograde labeling (Hoke et al., 2006), have provided us with new tools to observe this process. We have used this approach in the present study to investigate the regeneration of axons after crushing or freezing, two common types of peripheral nerve injury, with the aim of providing new insight into chemotactic regeneration.

Materials and Methods

Animals

A total of 92 healthy male Sprague-Dawley rats, aged 8 weeks and weighing 250 ± 20 g, were provided by the Experimental Animal Center of the Chinese PLA General Hospital. The study was approved by the same institute's Animal Ethics Committee in China.

Establishment of animal models

Animals were anesthetized by intraperitoneal injection of 10% chloral hydrate (0.3 mL/100 g) and placed on an operating table in the supine position. A surgical incision was made in the right groin to expose the femoral nerve. Animals were divided into three groups. In the control group (n = 40), the femoral nerve was transected 5 mm proximal to the bifurcation point (Brushart et al., 1993). In the crush injury group (n = 40), the nerve was transected 5 mm proximal to the bifurcation point, and the femoral nerve trunk was crushed using microforceps (WA3010; Head Biotechnology, Beijing, China) for 5 seconds (Brushart et al., 1993). In the freeze lesion group (n = 12), the femoral nerve was transected 7 mm proximal to the bifurcation point, and a cotton swab soaked in liquid nitrogen (Air Products, Beijing, China; stored in the Orthopedic Research Institute of the Chinese PLA General Hospital was used to freeze 5 mm of nerve trunk from 2 mm proximal to the bifurcation point. The incision was closed in layers after successful injury modeling.

Retrograde labeling

A 2% solution of True Blue (Sigma, St. Louis, MO, USA) was prepared in distilled water, and a 5% solution of 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (Dil; Sigma) was prepared in dehydrated alcohol. Both solutions were stored at 4°C until use.

Retrograde labeling was performed 2, 3, 6, and 12 weeks postoperatively (Ruiter et al., 2011). At each time point, 10 animals from the crush injury group, 10 from the freeze lesion group, and three from the control group were used. Animals were anesthetized by intraperitoneal injection of 10% chloral hydrate (0.3 mL/100 g), and placed on the operating table in the supine position. An incision was made in the right groin, to expose the right femoral nerve, its muscle branch, and saphenous nerve branches. The muscle branch was cut where it grows into the quadriceps femoris muscle, and the same length of saphenous nerve was cut to prepare for the next step. A 10 μL microsyringe was used to inject 5 μL of Dil or True Blue into aseptic plastic chambers. The nerve stumps of the muscle branches were inserted into the chamber filled with Dil, and the saphenous branch into the chamber of True Blue. The incision was closed in layers and animals were returned to their home cages with food and water ad libitum to recover for 3 days.

Observation of the spinal cord

Three days after retrograde labeling (Hoke et al., 2006), all animals were anesthetized with 10% chloral hydrate (0.3 mL/100 g). The dorsal spinal canal was opened and the spinal cord was exposed. The L2–4 segment was carefully removed, placed on a freezing microtome (CE1900; Leica, Germany) at −25°C until frozen solid, then embedded, and sagittal sections were cut using the microtome until the spinal gray matter was reached. Three sections (15 μm thickness) were taken and immediately observed and photographed under a fluorescence microscope (IX70; Olympus, Tokyo, Japan). Blue and red fluorescence was used to visualize True Blue and Dil, respectively, in the ventral horn of the spinal cord. Blue, red, and purple (colabeled) neurons were counted using Image-Pro Plus 6.0 image analysis software (Media Cybernetics, USA).

Statistical analysis

All data are presented as the mean ± SD. Data were analyzed by one-way analysis of variance and the least significant difference post-hoc test, using SPSS 17.0 software (SPSS, Chicago, IL, USA). P < 0.05 was considered statistically significant.

Results

Only single-labeled Dil-positive (red) neurons were seen in the anterior horn of the spinal cord in the control group. In the crush and freeze injury groups, the percentage of Dil single-labeled neurons in the anterior ventral horn increased over time, with significant increases at each time point after 2 weeks, after both crush and freeze injuries (P < 0.01). No significant differences between the crush and freeze injury groups were observed in Dil single-labeled cell counts at any time point (, , ).

Table 1

Number of labeled neurons in the anterior horn of rat spinal cords after surgery for crush injury or freeze lesioning

Table 2

Percentage of prechlorate (Dil) single-labeled neurons in the anterior horn of rat spinal cords after surgery for crush and freeze injuries

Figure 1

Retrograde labeling of neurons using Dil and True Blue, in the anterior horn of rat spinal cord 12 weeks after injury (immunofluorescence staining and fluorescence microscopy).

(A) Only Dil-labeled neurons were visible in the control group. (B, C) Double-labeled neurons in the crush (B) and freeze (C) injury groups. Red, Dil+; blue, True Blue+; purple, Dil+/True Blue+ double-labeled. Scale bars: 200 μm.

Discussion

Peripheral nerve injury is common in the trauma center and battlefield. The present study was designed to investigate regeneration after two types of common peripheral nerve damage: crush injury and freeze lesion. A series of pathophysiological changes is known to occur after peripheral nerve injury, in the neurons and axons at the injury site, as well as at target organs, and in the central nervous system (Lieberman, 1971). Nerve fibers at the proximal nerve stump begin to grow, while at the distal stump, axons break down, myelinolysis takes place and a new medullary sheath begins to form, providing a pathway to the growing axons from the proximal stump (Abdullah et al., 2013; Muheremu et al., 2013). Pathological changes after freeze injury include damage to the endoneurium, epineurium, and Schwann cells, at the injury site (Gaudet et al., 2011; Scheib et al., 2013). The perineurium remains intact. However, axon regrowth is impeded by the scar tissue created by the ruptured endoneurium and Schwann cells (Dubový et al., 2013; Suganuma et al., 2013; Gordon 2014), which impairs reinnervation to the distal target organs. Crush injury may also cause perineurial injury, leading to axonal misrouting.

With the development of microsurgical techniques, the timely and accurate suturing of peripheral nerve stumps is now possible, and considerable axonal regeneration can be achieved through part of the lesion. However, to date, it has not been possible to restore the function of the target organs to their pre-injury level. In the process of peripheral nerve regeneration, axonal misdirection may result in functional deficiency of original target organs (Madison et al., 2007; Vetter et al., 2010; Corriden et al., 2012; Yuan et al., 2013; Megan et al., 2014).

The theory of chemotactic nerve regeneration, proposed by Cajal (1928), assumes that nerve stumps tend to grow towards their original nerve pathways. Even before the 20th century, Bailey et al. (1993) found that nerve fibers regrowing from the proximal nerve stump tended to follow their original pathways, and called this phenomenon neurotropism, until Cajal (1928) renamed it chemotaxis. Cajal assumed chemotactic nerve growth to have histological, topographical, and target organ specificity. Although the theory of chemotaxis in nerve regeneration has been widely accepted, there is still much dispute about its specific characteristics and mechanisms. Furthermore, the phenomenon was denied by Weiss and Tailor with their famous experiment using Y tubes made from autologous blood vessels (Weiss et al., 1943, 1944), which discouraged research on chemotaxis for several decades. Lundborg (1986, 2000), Politis (1982, 1983, 1985), Seckel (1984), Mackinnon (1986), Bruneli (1987), Brushart (1993), and Chiu et al. (1982, 1988, 1990) demonstrated the existence of selective nerve regeneration and explained its chemotactic characteristics. It was presumed that chemotactic regeneration of axons after peripheral nerve injury promoted the accuracy of nerve reinnervation and the functional recovery of target organs (Jesuraj et al., 2012; Zhou et al., 2014).

However, although chemotaxis is now a generally accepted notion, there is still no consensus on its characteristics. Politis (1985) used femoral nerve retrograde labeling to show that axons grown from proximal motor nerve stumps tend to follow distal motor nerve pathways, and proximal sensory nerve stumps tend to follow distal sensory pathways. He suggested that the specific factors produced by distal nerve stumps facilitate the selective growth of proximal nerve axons. Madison et al. (2007) proposed the “pruning hypothesis”, explaining that axons following the wrong pathway will be pruned back, gradually increasing the accuracy of nerve innervation.

Our results from the present study show that the type of neurons in the anterior ventral horn at 2 weeks is random, but that accuracy increased significantly with time. After injury, nearly 80% of regenerated motor axons grew into their original pathways. The gradual increase of the proportion of Dil-labeled neurons and the gradual decrease of True Blue labeled and True Blue + Dil labeled neurons in the anterior ventral horn supports the pruning hypothesis, indicating that misdirected nerve fibers gradually shrink back in the inappropriate microenvironment. Hoke et al. (2006) also suggested that chemotactic nerve growth may be affected by different Schwann cell phenotypes. In their research, they found that Schwann cells of motor and sensory nerves have different phenotypes and produce different neurotrophic factors. Tsubokawa et al. (1999) proposed that Schwann cells strongly promote the growth of sensory nerves, but have little effect on motor axons. In our experiment, although massive Schwann cell death occurs after the freeze lesion, axonal regeneration accuracy remains equal to that of the crush injury group, contrary to Tsubokawa's theory. Our results suggest that perineural suture may lead to more accurate reinnervation than epineural suture (Bonini et al., 2013; de Ruiter et al., 2014); and that relatively long nerve gaps may be bridged by tubes that simulate the stereochemical structure of nerves (Nectow et al., 2011; Xin et al., 2011; Kim et al., 2013).

Novel experimental methods can be powerful tools (Vyas et al., 2010; Lee et al., 2012; Pujic et al., 2013). In the present study, we made some improvements to the conventional retrograde labeling technique. Since we found that it was impossible to inject Dil and True Blue into the sephanous nerve and muscle branch of the femoral nerve with the smallest available microinjector (10 μL), because of the extremely small diameter of these branches, we used small plastic chambers filled with each drug, and put the nerve stumps into those chambers. Viewing sagittal sections of the spinal cord under a fluorescence microscope revealed neurons stained with the different dyes and confirmed that our method produced successful retrograde transport. This method can therefore be applied in other retrograde labeling studies of small-diameter nerves. Another modification we made is to eliminate the internal fixation step before removing the spinal cord. Instead we removed the spinal cords while the animals were under deep anesthesia and immediately put the cords into a freezing microtome precooled to −25°C, and embedded them once they became frozen. This modified procedure will not only protect the technician from the harmful effects of chemical fixatives such as paraformaldehyde, but will also help avoid false-positive fluorescence, which might occur with the use of an internal fixative.

Functional recovery after peripheral nerve injury cannot be further improved by traditional methods; there is therefore an urgent need to explore novel methods (Summa et al., 2010; Ladak et al., 2011; Gu et al., 2011; Fregnan et al., 2012; Napoli et al., 2012; Wang et al., 2012; Franz et al., 2013; Marquardt et al., 2013; Chan et al., 2014; Guaiquil et al., 2014; Xu et al., 2014). The results from the present study provide valuable information about selective peripheral regeneration mechanisms, and will be useful in the search for novel methods for the treatment of peripheral nerve injury.

However, as 50% of the muscle branch femoral nerve fibers were sensory, using this retrograde labeling technique, motor nerve axons that wrongly grew into sensory nerve pathways of the distal muscle branch were also considered to be correctly innervated (Politis et al, 1985). Future studies should be designed to avoid such interferences and find more accurate ways to evaluate correct nerve regeneration.

In summary, the accuracy of peripheral reinnervation rises gradually over time and may be the result of pruning at the injury site. Moreover, nerve stumps can regenerate with high accuracy over a relatively long distance if the integrity of perineurium is preserved.