We chose to focus our analysis of aging on the biophysics of outgrowth of sensory neurons from neonatal and adult rats because their outgrowth is well characterized at the molecular and cellular levels 1718. This system has been of interest because adult sensory neuron axonal regeneration is enhanced by prior 'conditioning' lesions, which activate and inactivate the expression of groups of proteins linked to embryonic regeneration 7192021. Figure 1 shows an example of the outgrowth of unconditioned neonatal and adult sensory neurons. We found, as has been previously reported 17, that the elongation of adult sensory neurons was characterized by extensive axonal branching and a relatively slow rate of outgrowth (8.2 +/- 1.1 µm/hour; average +/- 95% confidence interval (CI)). Neurons from neonatal animals, also exhibited extensive branching, but had a significantly higher rate of axonal elongation than neurons from adult animals (11.5 +/- 1.4 µm/hour; average +/- 95% CI). These data are fully summarized in Table 1. The rates of growth were significantly different with p < 0.05 using a two tailed t-test. These results confirm the premise that the rate of axonal regeneration decreases during aging.

To determine how the mechanism of axonal elongation changes during aging, we stained the mitochondria along the axons and monitored the movement of the docked mitochondria during normal growth cone mediated axonal elongation (Figure 2). In the sensory neurons from neonatal and adult animals, we found that axonal elongation was coupled with axonal stretching, but with quantitative differences in the velocity profiles (Figure 2E, see Additional files 1 and 2). Analysis of the velocity profiles, reveals that the velocity of bulk movement along the neonatal axons was consistently higher as compared to the adult axons. The differences observed here suggested that sensory neurons from neonatal and adult neurons regenerate in vitro by the same fundamental mechanism (i.e. stretch and intercalated mass addition along the axon) as we previously reported for embryonic neurons 111516. It raised the question of whether the slower rate of growth of the adult neurons was due to a difference in axonal diameter, adhesion strength to the substrate, axonal viscosity, or force generation at the growth cone.

To determine the viscosity and the strength of adhesions, we applied known levels of forces to growth cones to experimentally stimulate elongation and analyzed the movement of docked mitochondria to measure the biophysical properties of neurons as previously described 11. We found that the overall viscosity (G) of the neurons from adult animals was significantly higher (p < 0.0001, two-tailed t-test) by a factor of 2 than neurons from neonatal animals (Figure 3A). In contrast, there was a small but significant decrease (p < 0.001, two-tailed t-test) in axonal diameter (Figure 3B). Because G is a function of axonal diameter, the underlying differences in the intrinsic viscosities (g) were even larger. The g value of the adult neurons was 4× higher (p < 0.0001, two-tailed t-test) than the neonatal and adult neurons (Figure 3C). We did not observe significant differences (p = 0.4, two-tailed t-test) in the levels of adhesion between the neonatal and adult neurons (Figure 3D). Together these results suggested that a substantially higher intrinsic viscosity was the key biophysical difference between the neonatal and adult neurons.

One possibility to explain the slower rate of adult axonal elongation is that the growth cones generate lower levels of force. To investigate, we estimated force generation at the growth cones of neonatal and adult neurons with Eq. 1.

F 0 = v G η .

Taking the average velocities (v) of axonal elongation (Table 1) and using the values of G and η from Table 2, we found the level of force generation (F₀) in the neonatal and adult growth cones to be 230 and 223 µdynes respectively (Table 3). As these data show the level of force generation in neonatal and adult growth cones is similar, a reduction in force generation seems unlikely to be the underlying factor that causes the reduction in regeneration during aging.

Changes in sensory neuron growth cone size have been observed during embryonic development. In particular, growth cones from embryonic day 7 neurons grown on polylysine are 146% greater in size than neurons from embryonic day 14 neurons 22. To determine if growth cone size is changing between postnatal and adult neurons, we measured their area. We found the average area of the neonatal growth cones to be 46 +/- 51 µm² (average +/- standard deviation, n = 53 growth cones) and the area of the adult growth cones to be 63 +/- 54 µm²(n = 46 growth cones). While there was a trend for adult growth cones to be larger, we did not find a significant difference in growth cone size using a two-tailed t-test.

Adult and neonate sensory neurons were cultured by a modification of a protocol developed by Lindsay et al., 1749, which uses a gentle enzymatic dissociation of the ganglion tissue. Supernumerary adult rats (> 200 g) were euthanized and their DRGs dissected and placed into Hanks Balanced Saline, without Ca²⁺ and Mg²⁺, buffered to pH 7.4 with 5 mM HEPES (HBSS-). After removal of ventral roots and ganglia capsules, the DRGs were dissociated for 10 minutes at 37°C in activated papain (Worthington Biochemical Corp, Lakewood, NJ) at a concentration of 50 U/ml in HBSS-. The ganglia from adult animals were transferred to a second enzyme solution containing 5 mg/ml dispase: 1 mg/ml collagenase (Life Technologies, Carlsbad, CA) in HBSS- for 10 minutes at 37°C, these enzymes were not used for the neonates. The ganglia were then triturated 10 times with a fire-polished Pasteur pipette, followed by another 10 min digestion and trituration in the same solution. Single cells were separated from larger chunks by gravity sedimentation. The single cells were then pelleted by brief centrifugation, and resuspended in culture medium for plating. Axonal outgrowth was supported in L-15 containing 10% fetal calf serum and 50 ng/ml 7S nerve growth factor (NGF) and N9 growth supplement 50 in plates treated with 0.01% polyornithine.

To track axonal stretching during normal axonal elongation and while towing, mitochondria were labeled with Mitotracker 11 and observed with a Leica DM IRB inverted microscope and observed with a N Plan L 40/0.55 corr Ph2 with an adjustable collar infinity/0 - 2/c objective. Cells were illuminated with a 100 W Xenon lamp attenuated 98% with neutral density filters through a Texas Red cube (Chroma, Rockingham, VT) for visualization of MitoTracker 11.

Images were taken with Openlab (Improvision, Waltham, MA) using an Orca-ER digital camera CCD, model #CA742-95 (Hamamatsu, Bridgewater, NJ), converted into TIFFs and analyzed using ImageJ (NIH) as previously described 11.

Based on our prior work, axonal elongation in response to forces acting at the growth cone occurs in three stages 51: after an initial elastic stretch, there is delayed stretching, followed by elongation at a constant rate. This behavior is described by a three-element model consisting of a spring (k₁), a spring (k₂) and a dashpot (H) in parallel, and a dashpot (G) in series (Figure 5A). A dashpot is a mathematical construct for viscosity and obeys the relationship force = viscosity constant velocity. In our model 11, we treat the axon as a series of viscoelastic elements that interact with each other and the substrate (Figure 5B). Since the elastic components of the model are invariant under steady state conditions (such as elongation over the course of hours) we simplify the Dennerll model to a series of dashpots. Two factors that determine the velocity profile of an axon under tension are the axon's axial viscosity (G) and the constant of friction (η) that quantifies the interactions between the axon and the substrate (Figure 5C). Both of these parameters characterize resistance to flow and have dimensions of viscosity. The axial viscosity G is the amount of force needed to distend a unit amount of axon at unit velocity and is a function of intrinsic axonal viscosity (g) and the cross sectional area of the axon (A). Increasing axonal diameter has the effect of adding dashpots in parallel or, equivalently, increasing the dashpot constant. If an axon alters its diameter as a result of stretching or mass addition, but maintains its physiological properties, then varies while is unchanged. The coefficient of friction is characterized by the strength and the number of adhesions between axon and substrate (Figure 5C). Such adhesions have been shown to have effects on both axonal stretching and growth cone advance 52. η is assumed to be zero where the axon is unattached to the substrate and increases when adhesions form or strengthen. Our mathematical model, based on the elongation of embryonic sensory neurons, has been previously described in detail.

v [ x , L ( t ) ] = F 0 sinh [ x ( η / G ) 1 / 2 ] ( η G ) 1 / 2 cosh [ L ( t ) ( η / G ) 1 / 2 ] .

In brief, Eq. 2 gives the profile for velocity as a function of distance, x; axonal length, L; force at the growth cone, F_o; axonal viscosity, G; and adhesion strength, η. For the details on the derivation of this formula refer to 11.

Axons were towed as previously described 11. In brief, force calibrated towing needles previously coated in polylysine (1 mg/ml) and concanavalin A (1 mg/ml) were used to apply forces to growth cones. Force measurements were acquired from phase images throughout the experiment. To track bulk movement along the axon in response to forces, fluorescent images of docked mitochondria were analyzed using ImageJ (NIH). To calculate G and η, lines were fitted to the velocity data in the kymographs to calculate the rate of change of the velocity of the mitochondria. Using force measurements from the calibrated needles, a value of was found by dividing the average force over this interval by the slope of the fitted line. Once values of were determined (one value of per 30 minute) the Origin software package (OriginLab Corporation, Northampton, MA) was used with Eq. 2 to fit the best value of η to the data. For this calculation the velocities of mitochondria proximal to the point of adhesion were used. Empirical values of F_o,L , and G were fixed and a Levenberg-Marquardt algorithm was implemented in the Origin package to find the optimal value of η. The relationship G = gA (A =cross-sectional area) was used to calculate the intrinsic axial viscosity g for each axon. Phase images of each trial were analyzed using ImageJ to determine the axonal diameter at various times. For each phase image, the diameter was measured along the axon as described below.

To automate the process of axonal diameter measurement, we developed an ImageJ plugin (named 'Width Measurement') that measures pixel intensity across objects, finds the derivative to determine the steepest points, and then calculates the distance between these points 11. The plugin has previously been described in detail 16. In brief, the plugin uses the derivative of pixel intensity across the axon to find these steepest points and returns the distance between these points to calculate axonal width. To prepare images for analysis, images of axons acquired at 12-bit pixel depth were opened in ImageJ, converted to 32-bits, and straightened using the "Straighten"; plugin 53. The straightened images were then stretched 8× on the y-axis by interpolation using the ImageJ plugin TransformJ set to quintic B-spline 54. To remove high frequency noise, a Gaussian Blur filter with a radius 2 pixels was applied using the built-in ImageJ function. The 'Width Measurement' plugin was then run to determine axonal width at each pixel along the axon. The source code for the Plugin is available on request.

To determine the area of growth cones, the width and length of individual growth cones were measured from phase images using the line tool in ImageJ. These numbers were then multiplied to give growth cone area. While our goal was to simply determine if there is a change in growth cone size, we note that because growth cones are not perfectly square, the calculated numbers for growth cone area are an overestimate.

All animal studies were approved by the Michigan State University Institutional Animal Care and Use Committee.