Mechanical breaking of microtubules in axons during dynamic stretch injury underlies delayed elasticity, microtubule disassembly, and axon degeneration

Abstract

Little is known about which components of the axonal cytoskeleton might break during rapid mechanical deformation, such as occurs in traumatic brain injury. Here, we micropatterned neuronal cell cultures on silicone membranes to induce dynamic stretch exclusively of axon fascicles. After stretch, undulating distortions formed along the axons that gradually relaxed back to a straight orientation, demonstrating a delayed elastic response. Subsequently, swellings developed, leading to degeneration of almost all axons by 24 h. Stabilizing the microtubules with taxol maintained the undulating geometry after injury but greatly reduced axon degeneration. Conversely, destabilizing microtubules with nocodazole prevented undulations but greatly increased the rate of axon loss. Ultrastructural analyses of axons postinjury revealed immediate breakage and buckling of microtubules in axon undulations and progressive loss of microtubules. Collectively, these data suggest that dynamic stretch of axons induces direct mechanical failure at specific points along microtubules. This microtubule disorganization impedes normal relaxation of the axons, resulting in undulations. However, this physical damage also triggers progressive disassembly of the microtubules around the breakage points. While the disintegration of microtubules allows delayed recovery of the "normal" straight axon morphology, it comes at a great cost by interrupting axonal transport, leading to axonal swelling and degeneration.

Figure 1.

Dynamic mechanical stretch of isolated cortical axons. A, B) Schematic illustration of axonal stretch injury model. Axon-only region of the elastic membrane overlaps with a 2- × 15-mm slit at the bottom of an airtight chamber. A controlled air pulse deflects the elastic membrane downward, thus inducing a tensile elongation exclusively of axons. C) Phase-contrast photomicrograph of the axon-only region, formed by a silicone stamp that creates microchannels permitting only axon outgrowth across the channels. D) Fluorescence microscopic confirmation that the neurites in the microchannels were axons demonstrated by immunoreactivity to neurofilament protein (NF, green), while immunoreactivity to microtubule-associated protein 2 (MAP2, red), a specific marker for the dendrites, was found only outside the channels. Scale bar = 100 μm.

Figure 2.

Degree of axonal undulation is proportional to the applied mechanical strain. A) Fluorescence photomicrograph of axons loaded with fluo-4 displaying multiple undulations (arrow) immediately after stretch (1–30 ms duration, with a fixed strain rate of 44 s⁻¹), the extent of which is proportional to the applied mechanical strain levels. Scale bars = 5 μm. B) Quantitative analysis of total length of axons (normalized to prestretch lengths): 30% strain: CSS (see Materials and Methods), n = 18, taxol, n = 7; 50% strain: CSS, n = 18, taxol, n = 6; 75% strain: CSS, n = 15, taxol, n = 18. n = total number of axons examined. C) Quantitative analysis of undulation densities, defined as number of undulations per axon process of ∼150 μm length: 30% strain: CSS, n = 6, taxol, n = 3; 50% strain: CSS, n = 6, taxol, n = 3; 75% strain: CSS, n = 5, taxol, n = 6. n = total number of microscopic fields examined (field of 150 μm × 113 μm with average ∼50 axons). ⁺⁺P < 0.01 vs. control CSS; Student’s t test.

Figure 3.

Microtubule destabilization and axon relaxation after dynamic stretch injury. A) Phase-contrast and fluorescence (fluo-4) photomicrographs demonstrating that axons without treatment (CSS) displayed multiple undulations after stretch and gradually resumed their original straight orientation within 40 min. However, axons pretreated with taxol (T; 1 μM) maintained their undulations for at least 3 h, while removal of taxol at 40 min poststretch allowed relaxation of axonal undulations by ∼3 h postinjury. In contrast, axons pretreated with nocodazole (N; 0.1 μM) relaxed to their prestretch orientation immediately after injury. Scale bars = 5 μm. B) Quantitative analysis of the length of individual undulations over time demonstrated that only pretreatment with taxol (1 μM) maintained the undulations, while a lower pretreatment dose (10 nM) or poststretch treatment (1 μM) had no effect. Values of data points are listed in Supplemental Table S1.

Figure 4.

Quantitative analysis of undulation densities. CSS (see Materials and Methods), n = 5; nocodazole (N; 0.1 μM), n = 9; N (1 μM), n = 3; taxol (T; 1 μM), n = 3; T (1 μM) + N (1 μM), n = 3; T (1 μM) + N (50 μM), n = 8. n = total number of microscopic fields examined (field of 150×113 μm with average ∼50 axons). ⁺⁺P < 0.01 vs. control CSS; Student’s t test.

Figure 5.

Microtubule destabilization and axon degeneration after dynamic stretch injury. A) Simultaneous live imaging of fluo-4 fluorescence (top) and phase-contrast (bottom) photomicrographs show development of axonal swellings after stretch injury by 1 h and gradually enlarged in size. B) Axonal swellings shown with live imaging (left) correspond with β-tubulin immunoreactivity of the same axons following fixation (right). C) β-Tubulin immunostaining of fixed cultures shows axon degeneration over hours. In control injured axons (CSS), there was a progression of axonal swellings (arrows) leading to a complete loss of axons by 24 h. However, pretreatment with taxol (1 μM) substantially reduced the appearance of swellings after injury and prevented degeneration, with 50% of the axons remaining intact by 24 h postinjury. In contrast, pretreatment with nocodazole (0.1 μM) rapidly accelerated degeneration, with almost complete loss of axons by 3 h. Each photomicrograph represents ≥5 independent cultures. Scale bars = 5 μm.

Figure 6.

Microtubule breaking immediately after dynamic stretch injury. TEM (×2–9 × 10⁴) of 60-nm-thick longitudinal sections of axons demonstrate microtubule disruptions observed immediately after injury. A) In straight segments of axons, microtubules traversed the main axis of axons without regions of interruptions. Within undulations, microtubules demonstrated uniformly open spacing at the peak, displaying frayed free ends (asterisk). Remaining portions of microtubules were disorganized and twisted, with a curling configuration (arrow). B) Demonstration of progressive phases of undulation relaxation and acute formation of axonal swellings after stretch injury. C) Taxol (1 μM)-pretreated axons demonstrated similar microtubule disruptions in axons as control injured axons immediately after stretch injury (top). In contrast, nocodazole (Noc; 0.1 μM)-pretreated axons demonstrated almost complete loss of microtubules within the undulation (bottom). Scale bars = 500 nm.

Figure 7.

Progressive loss of microtubules after dynamic stretch injury. A) TEM cross-sections (60-nm thickness, imaged at ×2–9×10⁴) of axons demonstrate progressive microtubule loss. Axons without injury (sham), axons in CSS (control), or axons pretreated with taxol (1 μM) or nocodazole (0.1 μM) had similar numbers of microtubules (arrow). After stretch injury, control axons gradually diminished their microtubules. However, axons pretreated with taxol maintained their number of microtubules over time, and axons pretreated with nocodazole accelerated their loss of microtubules. B) Quantitative analysis showed relative proportions of surviving axons with cross-sections containing ≥1 microtubule. Refer to numeric values in Supplemental Table S2. Scale bars = 500 nm.