Transduced Schwann cells promote axon growth and myelination after spinal cord injury

Abstract

We sought to directly compare growth and myelination of local and supraspinal axons by implanting into the injured spinal cord Schwann cells (SCs) transduced ex vivo with adenoviral (AdV) or lentiviral (LV) vectors encoding a bifunctional neurotrophin molecule (D15A). D15A mimics actions of both neurotrophin-3 and brain-derived neurotrophic factor. Transduced SCs were injected into the injury center 1 week after a moderate thoracic (T8) adult rat spinal cord contusion. D15A expression and bioactivity in vitro; D15A levels in vivo; and graft volume, SC number, implant axon number and cortico-, reticulo-, raphe-, coerulo-spinal and sensory axon growth were determined for both types of vectors employed to transduce SCs. ELISAs revealed that D15A-secreting SC implants contained significantly higher levels of neurotrophin than non-transduced SC and AdV/GFP and LV/GFP SC controls early after implantation. At 6 weeks post-implantation, D15A-secreting SC grafts exhibited 5-fold increases in graft volume, SC number and myelinated axon counts and a 3-fold increase in myelinated to unmyelinated (ensheathed) axon ratios. The total number of axons within grafts of LV/GFP/D15A SCs was estimated to be over 70,000. Also 5-HT, DbetaH, and CGRP axon length was increased up to 5-fold within D15A grafts. In sum, despite qualitative differences using the two vectors, increased neurotrophin secretion by the implanted D15A SCs led to the presence of a significantly increased number of axons in the contusion site. These results demonstrate the therapeutic potential for utilizing neurotrophin-transduced SCs to repair the injured spinal cord.

Figure 1. Medium conditioned by SCs transduced with AdV/GFP/D15A or LV/GFP/D15A contains biologically active neurotrophin that supports neurite outgrowth in DRG cultures

Embryonic DRG, incubated for three days in conditioned medium from (B) AdV/GFP/D15A or (D) LV/GFP/D15A SCs, exhibited robust neurite outgrowth when compared to (A) AdV/GFP and (C) LV/GFP SCs. These cultures may be compared to examples of similar cultures provided medium containing neutralizing NT-3 antibodies (E) and medium containing appropriate concentrations of added BDNF and NT-3 (F). Neurofilament staining. Scale bar = 1 mm.

Figure 2. Spinal cord tissue containing AdV/GFP/D15A or LV/GFP/D15A SC implants contains significantly higher levels of neurotrophin than non-transduced SC or GFP only controls

Spinal cord tissue was removed, homogenized and subjected to an NT-3 ELISA. Total protein was determined using a Bradford protein assay. At 2 days, a statistically significant difference was achieved in D15A-transduced SC implants compared to control non-transduced SC grafts; AdV/GFP/D15A implants generated significantly higher levels of neurotrophin than did LV/GFP/D15A SC grafts. By 1 wk, only the levels in AdV/GFP/D15A reached significance. The levels of neurotrophin secreted by control AdV/GFP and LV/GFP implanted tissue were similar to spinal cord implanted with non-transduced SCs. Error bars, SEM. ***P<0.001;**P<0.01.

Figure 3. D15A-secreting SC implant volumes are significantly increased at six weeks post-implantation

LV/GFP SCs are depicted in (A). In comparison, larger volumes are observed in LV/GFP/D15A SC grafts (B). (C) Volume quantification of implants, identified by GFP expression of LV vector-transduced SCs or p75 immunoreactivity of AdV vector and non-transduced SCs. A statistically significant difference was accepted at a P value of ***P<0.001 as compared to controls (one-way ANOVA). r, rostral. Scale bar = 1 mm. Error bars, SEM.

Figure 4. Myelinated axon counts in injury/graft centers are increased within D15A-secreting SC implants, at six weeks post-implantation

SC myelinated axons, evident as dark rings, were present in all implanted groups. Many more SC myelinated axons were present in transverse sections of the (B) AdV/GFP/D15A and (D) LV/GFP/D15A SC groups, when compared to (A) AdV/GFP, (C) LV/GFP and (E) non-transduced SC groups. (F) Quantification reveals significantly more axons myelinated by SCs in the epicenters of D15A-transduced SC grafts. A statistically significant difference was accepted at a P value of ***P<0.001 compared to controls and **P<0.01 comparing AdV/GFP/D15A with LV/GFP/D15A. Scale bar = 10 μm. Error bars, SEM.

Figure 5. D15A-secreting SC implants contain significantly increased lengths of 5-HT, DβH and CGRP-positive fibers at 6 weeks post-implantation

Confocal images at 400X magnification show AdV/GFP/D15A SC implants containing (A) 5-HT fibers and (C) CGRP fibers, compared to AdV/GFP SC implants (B,D). Significantly increased lengths of (E) 5-HT, (F) DβH and (G) CGRP immunostained fibers were observed within AdV/GFP/D15A SC grafts, when compared to AdV/GFP, LV/GFP/D15A, LV/GFP, and non-transduced SCs. (F) LV/GFP/D15A SC grafts also contained increased DβH fiber length. A statistically significant difference was accepted at a P value of **P<0.01 as compared to controls (one-way ANOVA). Scale bar = 50 μm. Error bars, SEM.

Figure 6. D15A-secreting SC implants did not improve regrowth of anterogradely-traced corticospinal or reticulospinal fibers into the implant

The rostral-most end of the lesion/implant site at the level of the central canal was designated as the zero point. Lines perpendicular to the longitudinal axis were drawn 500 μm apart, and labeled axons that crossed these lines were counted. (A) The presence of LV/GFP/D15A SC implants led to significantly increased numbers of BDA-traced CST fibers rostral to the zero point compared to non-transduced SC implants. AdV/GFP/D15A SC implants were elevated compared to controls, but not significantly. (B) No differences were observed in dextran-rhodamine-traced RST fibers among animal groups. A statistically significant difference was accepted at a P value of **P<0.01 as compared to controls (two-way ANOVA). Error bars, SEM.