Published December 1, 2011 - More info
Stroke is the leading cause of long-term disability and the third leading cause of death in the United States. While most research thus far has focused on acute stroke treatment and neuroprotection, the exploitation of endogenous brain self-repair mechanisms may also yield therapeutic strategies. Here, we describe a distinct type of stroke treatment, the naturally occurring extracellular matrix fragment of perlecan, domain V, which we found had neuroprotective properties and enhanced post-stroke angiogenesis, a key component of brain repair, in rodent models of stroke. In both rat and mouse models, Western blot analysis revealed elevated levels of perlecan domain V. When systemically administered 24 hours after stroke, domain V was well tolerated, reached infarct and peri-infarct brain vasculature, and restored stroke-affected motor function to baseline pre-stroke levels in these multiple stroke models in both mice and rats. Post-stroke domain V administration increased VEGF levels via a mechanism involving brain endothelial cell α5β1 integrin, and the subsequent neuroprotective and angiogenic actions of domain V were in turn mediated via VEGFR. These results suggest that perlecan domain V represents a promising approach for stroke treatment.
Boyeon Lee, Douglas Clarke, Abraham Al Ahmad, Michael Kahle, Christi Parham, Lisa Auckland, Courtney Shaw, Mehmet Fidanboylu, Anthony Wayne Orr, Omolara Ogunshola, Andrzej Fertala, Sarah A. Thomas, Gregory J. Bix
Original citation: J. Clin. Invest. 2011;121(8):3005–3023. doi:10.1172/JCI46358.
Citation for this corrigendum: J. Clin. Invest. 2012;122(2):777. doi:10.1172/JCI61899.
In the legends for Figures 3, 5, and 9, it was not explicit that the same data were used for multiple figures.
The correct figure legends appear below.
DV is neuroprotective. (A and B) Mean ischemic lesion volumes measured from brain sections stained with TTC (PSD 1–3) or H&E (PSD 7 and 15) in WT mice treated with different doses of DV (A) or in WT and Pln–/– mice treated as indicated (B) (*P < 0.05, **P < 0.01, n = 15 per treatment group per PSD). Mean ischemic lesion volumes for WT and Pln–/– on PSD 1–3 are repeated from Figure 1E, as these volumes were obtained in the same experiments as the volumes obtained with DV treatment of WT and Pln–/– mice. (C) WT or Pln–/– mouse brain TTC staining at PSD 3, or WT H&E staining on PSD 15, after animals received i.p. PBS or DV injections (1 mg/kg). Yellow asterisks and red circles indicate ischemic lesions. PBS-treated WT and Pln–/– brain images are repeated here from those shown in Figure 1F, as they were obtained from the same experiments shown in that figure. (D) Cresyl violet, cleaved caspase-3, and TUNEL staining, with propidium iodide (PI) nuclear counterstain, in the peri-infarct area in WT mice treated with PBS or with DV. Scale bars: 5 μm (cresyl violet) and 10 μm (caspase-3 and TUNEL).
DV neuroprotection is VEGF and VEGFR mediated. (A) Anti-VEGF Western blot analysis of ipsilateral stroke hemispheres as labeled, with GAPDH as internal loading control. (B) Densitometry analysis of VEGF Western blot as shown in A as normalized to corresponding GAPDH bands (**P < 0.01, n = 15 per treatment group, per PSD). (C) Plot of VEGF ELISA ipsilateral stroke brain tissue treated as labeled (#P < 0.01 as compared with corresponding PBS-treated WT control or as labeled, n = 3 per treatment group per PSD). (D) Mean ischemic lesion volumes of stroke WT mice on PSD 1–3 treated as labeled (**P < 0.01, n = 15 per treatment group per PSD). (E) Vibrissae-elicited forelimb placement test on WT mice treated as labeled. DV had no effect in animals also treated with PTK787/ZK 222584 (P = NS). Stroke PBS and stroke DV (1 mg/kg) results are repeated from the identically labeled data in Figure 4B, as these experiments were performed in parallel using the same groups for comparison. (F) NeuN and VEGFR2 co-immunohistochemistry of PSD 5 peri-infarct brain tissue of mice treated as labeled. White arrows indicate cells that were positive for both NeuN and VEGFR2. Scale bar: 50 μm. (G) Number of NeuN- and VEGFR2-positive cells per mm2 in the peri-infarct regions as labeled (**P < 0.01, n = 10 images per animal, 5 animals per treatment condition).
DV effects are mediated via the α5β1 integrin in vivo. (A) Anti-α5β1 Western blot analysis from PSD 3 mouse brain tissue treated as labeled, with GAPDH as internal control. (B) α5β1 immunohistochemistry of mouse PSD 3 peri-infarct brain tissue with or without DV treatment. Scale bar: 10 μm. (C) Quantification of mean ischemic lesion volumes of stroke WT mice on PSD 1–3 as labeled (*P < 0.05, n = 15 per treatment condition per PSD). (D) Cresyl violet staining, caspase-3 17- to 20-kDa cleavage product immunostaining, and TUNEL staining with PI of peri-infarct brain regions as labeled. Scale bars: 10 μm. (E) Vibrissae-elicited forelimb placement test on WT mice treated as labeled (n = 15 mice per condition from 3 separate experiments with 5 mice each). WT and WT + DV values are repeated here from Figure 5E (there labeled stroke PBS and stroke DV, 1 mg/kg), as these experiments were all performed in parallel using the same groups for comparison. (F) Von Willebrand factor immunohistochemistry (green) on PSD 5 from WT mice treated as labeled. Scale bar: 10 μm. Representative WT stroke + PBS and WT stroke + DV images are repeated here from the identically labeled images in Figure 8A, to ease their visual comparison with animals treated with α5 antibody or DV + α5 antibody. (G) Peri-infarct blood vessel quantification as labeled (*P < 0.05, **P < 0.01 compared with PBS + IgG on the same day, n = 20 images analyzed per animal, 10 animals per experimental condition). (H) Anti-VEGF Western blot analysis of mouse stroke hemispheres with internal GAPDH as control. (I) Optical density quantification of VEGF Western blot analysis as shown in H (**P < 0.01, n = 5 per experimental condition).
The authors regret the error.