Chronically increased ciliary neurotrophic factor and fibroblast growth factor-2 expression after spinal contusion in rats

Abstract

Demyelination and oligodendrocyte loss following spinal cord injury (SCI) are well documented. Recently, we showed oligodendrocyte progenitor cell (OPC) accumulation and robust oligodendrocyte genesis occurring along SCI lesion borders. We have since begun investigating potential mechanisms for this endogenous repair response. Here, we examined ciliary neurotrophic factor (CNTF) and fibroblast growth factor-2 (FGF-2) expression, because both factors alter progenitor proliferation and differentiation and are increased in several CNS disorders. We hypothesized that CNTF and FGF-2 would increase after SCI, especially in regions of enhanced oligogenesis. First, CNTF protein was quantified using Western blots, which revealed that CNTF protein continually rose through 28 days post injury (dpi). Next, by using immunohistochemistry, we examined the spatiotemporal expression of CNTF in cross-sections spanning the injury site. CNTF immunoreactivity was observed on astrocytes and oligodendrocytes in naïve and contused spinal cords. Significantly increased CNTF was detected in spared white and gray matter between 5 and 28 dpi compared with uninjured controls. By 28 dpi, CNTF expression was significantly higher along lesion borders compared with outlying spared tissue; a similar distribution of phosphorylated STAT3, a transcription factor up-regulated by CNTF and to a lesser extent FGF-2, was also detected. Because CNTF can potentiate FGF-2 expression, we examined the distribution of FGF-2+ cells. Significantly more FGF-2+ cells were noted along lesion borders at 7 and 28 dpi. Thus, both CNTF and FGF-2 are present in regions of elevated OPC proliferation and oligodendrocyte generation after SCI and therefore may play a role in injury-induced gliogenesis.

Fig. 1

Increased CNTF expression 5–28 days post injury (dpi). A: Western blot analysis of 4 mm of SCI tissue centered on the lesion epicenter revealed a single CNTF band (22 kDa), which increased over time after injury; α-tubulin was used as loading control, and 10 µg of total protein was loaded per lane. B: Ratio of the relative optical density (ROD) of CNTF bands normalized to a-tubulin showed significantly higher CNTF expression at 14 and 28 dpi. One-way ANOVA followed by Tukey post hoc tests was used for statistical analysis and P < 0.05 was considered significant; n = 3/group.

Fig. 2

Increased CNTF immunoreactivity was detected in white matter (WM) and gray matter (GM) regions of the injured spinal cord. A,B: As expected, a low level of CNTF expression was observed in normal GM (A) and WM (B). Dashed line in A delineates dorsal horn (DH) gray matter and adjacent white matter. Note the high CNTF expression in the dorsal root in naïve tissue. C: WM at 14 dpi immunolabeled with CNTF antibody that had been preincubated with recombinant rat CNTF; no positive signal was detected. Markedly increased CNTF expression was observed after SCI in GM (D; 7 dpi) and WM (E,F; 28 dpi). CNTF expression was especially high in white matter bordering the lesion. Images were taken at the following locations: D, 1.2 mm rostral; C, 1.5 mm rostral; E, 1 mm rostral; F, 2.5 mm caudal to epicenter. Lesion is indicated by the asterisk; arrows indicate CNTF labeling. Scale bar = 20 µm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Fig. 3

CNTF expression is significantly elevated after SCI, especially at lesion borders. A: Quantification of CNTF immunoreactivity after SCI revealed a significant increase in gray matter by 7 days post injury (dpi) in sections 1–2 mm rostral to epicenter (0 on x-axis). Between 7 and 28 dpi, CNTF expression continued to increase at most distances compared with naïve gray matter and previous time points. B: CNTF increased modestly in white matter at 3 dpi. By 7 dpi, CNTF was significantly higher in white matter at all distances examined compared with naïve. By 28 dpi, CNTF expression in white matter was greater than naïve and 7 and 14 dpi at most distances examined. C: Diagram of an injured spinal cord section depicting sample box placement used to generate data in D. Sample box area = 0.0144 mm². White boxes represent spared white matter samples; gray boxes show gray matter samples; hatched boxes show samples along lesion borders. Boxes are not to scale. D: At 7 and 28 dpi, CNTF expression was significantly higher in spared white matter, white matter border, and gray matter border compared with naïve; at 28 dpi, CNTF along white matter and gray matter lesion borders was significantly elevated compared with the respective outlying spared tissue. Data spanning 4 mm of spinal tissue are collapsed across distance. For A and B, analysis included two-way repeated-measures ANOVA followed by Bonferroni post hoc analyses; for D, one-way ANOVA followed by Tukey post hoc tests was used. For all analyses, P < 0.05 was considered significant (n = 4–6/group). For A and B, sections were manually delineated into white matter or gray matter, and segmentation range was set to select only profiles positively labeled for CNTF. The proportional area labeled for CNTF was calculated using target area/total scan area. For D, the proportional area labeled for CNTF was similarly calculated for each sample box.

Fig. 4

Increased oligodendrocyte numbers were present along the lesion border at 28 dpi. For color figures, see the article online. Our previous work revealed that oligodendrocyte numbers are elevated along lesion borders through at least 14 dpi. Here, sections from 28 dpi double-labeled for oligodendrocytes (CC1; brown) and astrocytes (GFAP, black) reveal that elevated oligodendrocyte numbers are maintained along lesion borders (B) compared with outlying white matter (WM; A) through at least 28 dpi. Arrows indicate CC1⁺ oligodendrocytes; arrowheads denote GFAP⁺ astrocytes. Asterisk denotes lesion. Both the images were taken from the same section 1 mm caudal to epicenter. Scale bar = 20 µm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Fig. 5

Astrocytes, oligodendrocytes, and Schwann cells display CNTF immunoreactivity after SCI. For color figures, see the article online. A–C: Single-channel (A,B) and merged (C) confocal images (≤ 1 µm) reveal that many GFAP⁺ astrocytes (green) were colabeled with CNTF (red). D–F: CNTF immunoreactivity (red) was also observed on CC1⁺ oligodendrocytes (green); a merged image shown is in F. G–I: The oligodendrocyte-lineage identity was confirmed by colocalization of Olig2 (H) with CNTF (G). The single-labeled Olig2 cells in H are most likely OPCs; note that astrocyte-like CNTF⁺ cells in I do not show Olig2 labeling. J–M: A subset of p75⁺ Schwann cells expressed CNTF. Images G-I were taken at 14 dpi 2.5 mm rostral to epicenter; all other images were taken from 28 dpi sections: A–C, 1.2 mm caudal; D–F, 0.75 mm caudal; J,K, 0.45 mm caudal; and L–M, 1.5 mm caudal to epicenter. All sections were counterstained with DRAQ5, and the high red signal was converted to blue. Arrows indicate double-labeled cells; arrowheads indicate single-labeled cells. Asterisk denotes lesion in D–F. Scale bars = 20 µm in F (applies to A–F); 20 µm in I (applies to G–I); 20 µm in M (applies to J–M). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Fig. 6

CNTF receptor a (CNTFRα) expression initially declines then rises after SCI. A: CNTFRa was detected as a prominent 60-kDa band in spinal cord samples using Western blot; an additional, lighter, 55-kDa band was also noted at 3 and 5 dpi. a-Tubulin was used as loading control, and 10 µg of total protein was loaded per lane. B: Densitometric analysis of the 60-kDa band showed a dramatic and significant decrease in CNTFRa expression as early as 3 dpi, which rebounded to naïve levels by 28 dpi. One-way ANOVA followed by Tukey post hoc test was performed, and P < 0.05 was considered significant (n = 3/group).

Fig. 7

FGF-2 protein levels increase from 3 to 28 dpi. In naïve spinal cords, FGF-2 was detected as a faint band at 22 kDa. At 3 dpi, this band and a 24-kDa band was detected. Between 5 and 28 dpi, the 22-kDa FGF-2 band intensified, indicating increased FGF-2; a 20-kDa band also appeared, which intensified over time. α-Tubulin signal reveals comparable loading in all lanes. In total 12.5 µg protein was loaded per lane.

Fig. 8

FGF-2⁺ cell numbers increase after SCI. A–C: At 7 dpi, elevated numbers of FGF-2⁺ cells were present at the white matter (WM) lesion border (C) compared with outer spared WM (SWM) near meninges (B) and naïve WM (A). D–F: Many FGF-2⁺ cells were observed 28 dpi at the gray matter (GM) borders (F) compared with spared GM (SGM; E) and naïve GM (D). B,C, 1 mm rostral; E,F, 2 mm caudal to epicenter. Scale bar = 20 µm.

Fig. 9

FGF-2⁺ cell numbers increase at the lesion borders post-SCI. FGF2⁺ cells were counted every 1 mm in 4 mm of spinal tissue spanning the lesion using sample boxes (refer to Fig. 3C). naïve white matter had fewer FGF-2⁺ cells compared with naïve gray matter. At 7 dpi, significantly more FGF-2⁺ cells were present in white matter border, spared gray matter (SGM), and gray matter border compared with naïve tissue. Notably, white matter border also had significantly more FGF-2⁺ cells than spared white matter (SWM). At 28 dpi, elevated numbers of FGF-2⁺ cells were detected only in white matter and gray matter lesion borders compared with their respective spared regions as well as naïve tissue. One-way ANOVA was used to compare white matter and gray matter regions over time with their respective controls. P < 0.05 was considered significant (n = 4–6/group).

Fig. 10

NG2 cells with phosphorylated STAT3⁺ (pSTAT3) nuclei were common following SCI. For color figures, see the article online. NG2⁺ cells (grayish-black) with pSTAT3⁺ nuclei (brown) were seen in gray matter (A; 3 dpi, 1.8 mm rostral to epicenter) and white matter (B; 7 dpi, 1.5 mm rostral to epicenter) close to lesion cavity. Arrows indicate double-labeled cells; asterisk indicates lesion. Insets show higher magnification of double-labeled cells in A and B. Sections were counterstained with methyl green. Scale bar = 10 µm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Fig. 11

Oligodendrocytes and astrocytes along lesion borders display pSTAT3⁺ nuclei following SCI. For color figures, see the article online. A–C: By using confocal microscopy, many oligodendrocytes (CC1⁺, green; A) with pSTAT3⁺ nuclei (red; B) were detected along lesion borders. Higher magnification inset in B shows nuclear localization of pSTAT3. D–F: Astrocytes (GFAP, green; D) also contained pSTAT3⁺ nuclei (red; E,F). Nuclear labeling is shown in blue (DRAQ5). Arrows indicate double-labeled cells; arrowheads indicate single-labeled pSTAT3 cells. Asterisk denotes lesion, and the dotted line indicates lesion border. Images were taken from 14 dpi sections at 1.5 mm rostral (A–C) and 0.5 mm rostral (D–F) to the epicenter. Scale bar = 20 µm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]