Blood substitution therapy rescues the brain of mice from ischemic damage

Acute stroke causes complex, pathological, and systemic responses that have not been treatable by any single medication. In this study, using a murine transient middle cerebral artery occlusion stroke model, a novel therapeutic strategy is proposed, where blood replacement (BR) robustly reduces infarctions and improves neurological deficits in mice. Our analyses of immune cell subsets suggest that BR therapy substantially decreases neutrophils in blood following a stroke. Electrochemiluminescence detection demonstrates that BR therapy reduces cytokine storm in plasma and ELISA demonstrates reduced levels of matrix metalloproteinase-9 (MMP-9) in the plasma and brains at different time points post-stroke. Further, we have demonstrated that the addition of MMP-9 to the blood diminishes the protective effect of the BR therapy. Our study is the first to show that BR therapy leads to profoundly improved stroke outcomes in mice and that the improved outcomes are mediated via MMP-9. These results offer new insights into the mechanisms of stroke damage.

(a) Experimental Design. (b) Representative TTC-stained coronal sections indicate brain infarction of mice. Viable brain tissue is stained red and infarct brain tissue remains white by TTC-staining. The filaments in MCAs were removed after 90 min occlusion. (c) Quantified infarct volumes revealed larger infarction from stroke mice at 23-hours (squares, n = 4) compared to 6hours (circles, n = 4). Data were presented as means ± SD; P-values were calculated using twotailed grouped analyses by Student's t test. Source data are provided as a Source Data file.

Supplementary Figure 2. A stroke alters cellular profiles in the blood at 6-hours stroke induction.
Blood obtained from control (circles, n = 6) and stroke mice (underwent tMCAO for 90 minutes) (squares, n = 6) were analyzed. (a) Stroke increased total white blood cell (WBC) counts in the blood. (b) Stroke increased total neutrophils and monocytes in the blood. Neutrophils were gated on PI -CD45 + CD11b + Gr1 + live cells and monocytes were gated on PI -CD45 + CD11b + Gr1live cells for the analyses detected by flow cytometry. (c) Stroke increased total CD4 + and CD8 + cells in the blood. CD4 + cells were gated on PI -CD45 + CD4 + live cells and CD8 + cells were gated on PI -CD45 + CD8 + live cells for the analyses. (d) Stroke reduced total CD19 + B-cells in the blood. B cells were gated on PI -CD45 + CD19 + live cells for the analyses. (e) Stroke increased total NK1.1 + cells in the blood. NK1.1 + cells were gated on PI -CD45 + NK1.1 + live cells. (f) Stroke did not change total number of CD11c + dendritic cells in the blood. Data were presented as means ± SD; P-values were calculated using two-tailed grouped analyses by Student's t test. Source data are provided as a Source Data file.

Supplementary Figure 3. Relative cerebral blood flow was not significantly different among randomized groups.
Mice underwent tMCAO for 90 minutes. The cerebral blood flow was recorded at four time points -prior to MCAO, 1 min post-MCAO, 5 min post-reperfusion, and prior to blood-replacement by Laser Speckle Imager. The occlusion and reperfusion were confirmed by Laser Speckle Imager then mice were randomized into three groups: sham control (circles, n = 8), 250µl (squares, n = 6), or 500µl (triangles, n = 6) of blood from naive healthy donor mice (3 months old). (a) Representative images of cerebral blood flow (CBF) from Laser Speckle Imager. MCA territories are outlined with dashed lines. (b) Quantified relative CBF (R/L) (ratio of CBF in right MCA territory: CBF in left MCA territory). The data confirmed that the MCAO reduced 70~80% relative CBF at the post-tMCAO time point. The reperfusion fully recovered the relative CBF at 5 min post-reperfusion and prior to blood-replacement in all stroke mice; however, no significant differences were detected among randomized groups of mice. Data were presented as means ± SD; One-way ANOVA followed by post-hoc Fisher's unprotected least significant difference multiple comparison tests. Source data are provided as a Source Data file.

Supplementary Figure 4. Blood transfusion or blood withdrawal alone does not protect stroke outcomes.
Mice underwent tMCAO for 90 minutes were randomized to three groups: sham transfusion/withdrawal procedure group, transfusion of 500µl blood group, and withdrawal of 500µl blood group. Brain infarct volumes were measured at 23 hours after ischemia induction.
(a) Quantified infarct volumes were not significantly changed by the blood transfusion or the blood withdrawal alone. Sham transfusion/withdrawal procedure group (circles, n=6), transfusion of 500µl blood group (squares, n=3), and withdrawal of 500µl blood group (triangles, n=3). (b) Representative TTC-stained coronal sections. (c) Neurological deficits were not significantly different in three groups. Sham transfusion/withdrawal procedure group (circles, n=8), transfusion of 500µl blood group (squares, n=9), and withdrawal of 500µl blood group (triangles, n=9). Data were presented as means ± SD; One-way ANOVA followed by post-hoc Fisher's unprotected least significant difference multiple comparison tests. 2 mice died in control group, 6 mice died in blood transfusion alone group, and 6 mice died in blood withdrawal alone group.
Source data are provided as a Source Data file.

Supplementary Figure 5. Blood obtained from stroke mice does not protect stroke outcomes.
Mice underwent tMCAO for 90 minutes and randomized to two groups: transfusion of 500µl blood obtained from naive donor mice (circles) and transfusion of 500µl blood obtained from stroke mice (received permanent MCAO for 6 hours) (squares). The same volume of blood was withdrawn from the recipient stroke mice during blood transfusion. Brain infarct volumes were measured at 23 hours after ischemia induction. (a) Blood obtained from stroke mice had significantly larger infarct volume in cortex, striatum, and total hemisphere than blood obtained from naive donors in stroke mice received blood replacement. (b) Representative TTC-stained coronal sections. (c) Blood obtained from stroke mice had significantly worse neurological deficits than blood obtained from naive donor mice in stroke mice received blood replacement.
Data were presented as means ± SD; P-values were calculated using two-tailed grouped analyses by Student's t test. N=3 or 6 per group for stroke infarction, n=6 or 9 per group for neurological deficits; No mice died in blood obtained from naïve donor group but 6 mice died in blood obtained from stroke mice group. Source data are provided as a Source Data file. Mice underwent tMCAO for 90 minutes and randomized to two groups: transfusion of 500µl plasma obtained from naive donor mice and a plain stroke group. The same volume of blood was withdrawn from the recipient stroke mice during plasma transfusion. (a) Plasma transfusion did not significantly change infarct volume in cortex, striatum, and total hemisphere compared to control mice. Control group (circles, n=6), plasma group (squares, n=6). (b) Plasma transfusion did not significantly change neurological deficits compared to control mice Control group (circles, n=8), plasma group (squares, n=8). Data were presented as means ± SD; Twotailed grouped analyses by Student's t test. Source data are provided as a Source Data file.