Research ArticleNEUROPHYSIOLOGY

Multimodal mapping of neural activity and cerebral blood flow reveals long-lasting neurovascular dissociations after small-scale strokes

See allHide authors and affiliations

Science Advances  22 May 2020:
Vol. 6, no. 21, eaba1933
DOI: 10.1126/sciadv.aba1933
  • Fig. 1 The multimodal neural interface for simultaneous mapping of CBF and neuronal electrical activity.

    (A) Schematics of the multimodal measurement. The optical paths for targeted photothrombosis (532 nm, green) and the speckle imaging (685 and 785 nm, red) were sketched. DM, dichroic mirror. F1, filter. L1 to L3, lenses. M1 and M2, mirrors. Inset: The dashed circle shows the anatomical location of the cranial window relative to bregma (gray dot). Red dots indicate the implantation sites of the four-shank probes. A, anterior. L, lateral. (B) Photograph of a mouse on a treadmill under the imaging apparatus implanted with NETs and a cranial window. Inset: Photo of the cranial window at 10 weeks after surgery. Photo credit: Fei He, Rice University. (C) A representative LSCI of prestroke CBF, where the green shade highlights the arterioles for targeted photothrombosis and the white lines outline the NETs. (D) Photograph of a NET array (4 by 8 recording sites) in water showing its ultraflexibility. Photo credit: Fei He, Rice University. (E) Typical recording traces (300-Hz high-pass filtered) from eight recording sites on one NET shank. (F) Spatially resolved spike activities from a 5-min representative recording session using a 4 by 8 NET array. Squares sketch individual contacts; dots are color-coded to present spike rates, whose locations are estimated as the averaged locations of the contacts weighted by the spike magnitudes they detect. (G) Experimental procedure and timeline. Scale bars, 500 μm [(B) (inset), (C), and (D)], 400 μV [vertical in (E)], and 200 ms [horizontal in (E)]. D, day. WK, week.

  • Fig. 2 Neurovascular coupling weakens with ischemia in acute periods after stroke.

    (A) Sketch of the skull (left) and LSCI of CBF (middle) and sketch on a coronal section (right) showing the locations of the cranial window, four NET shanks, and 32 contacts in (B to D) and the targeted arteriole for photothrombosis (shaded in green). Red square marks the region of interest (ROI) used to quantify rCBF in (E and F). (B) Time series of LSCI of CBF at the acute stroke session. The overlaid color map shows the area of rCBF <50% baselined against prephotothrombotic images. Blue triangles mark the NET implantation sites. (C and D) Time series of LFP at 60 to 110 Hz and FR from all NET contacts at the acute stroke session. Time points match those of (B). Roman and Arabic numbers mark the NET shanks and contacts along the depth, respectively. One disconnected contact (II-8) is plotted in gray. (E and F) Simultaneous measurement of LFP and FR recorded by contact IV-4 and CBF from the ROI in (A) during and following photothrombosis. Normalized and 30-s averaged values were used in (F) to visualize the neurovascular dissociation. Dashed lines mark a PID event that was excluded from the following analysis. (G and H) Scatter diagrams of rCBF and rLFP at 60 to 110 Hz (G) and rCBF and rFR (H) for all locations (n = 39) and animals (N = 10) showing the positive linear correlation (darker color) breaks down as ischemia worsened (lighter color) after a PID occurred. (I and J) Correlation coefficients computed between rCBF and rLFP (I) and between rCBF and rFR (J). Pearson’s correlation was applied: ρ = 0.42, P = 4 × 10−94 (G); ρ = 0.12, P = 8 × 10−9 (H); ρ = 0.40, P = 3 × 10−6 (I); ρ = 0.35, P = 1 × 10−4 (J). Scale bars, 500 μm (A and B).

  • Fig. 3 Neurovascular disassociation lasts chronically after stroke.

    (A) Sketch of the skull and LSCI of CBF before stroke showing the location of the cranial window, the implanted NET shanks (highlighted by white solid lines), the targeted arteriole for photothrombosis (green shading), and the ROIs (red squares) used in Fig. 4A. Dashed squares show the area in (C). (B) Nissl-stained coronal sections showing the NET implantation locations (white dashed lines) and the infarct area (orange shading). (C) MESI of CBF at multiple temporal coordinates before and after stroke. The color map is baselined against the averaged prephotothrombotic values to highlight both hyperperfusion (red) and hypoperfusion (blue) regions. The yellow shades mark the regions under NETs, which were excluded from the quantification of CBF. Days after stroke are marked on top. BL, prestroke baseline. (D) Representative single-channel recordings from contact IV-5 at the same temporal coordinates as in (C). Black, raw trace; green, LFP at 60 to 110 Hz; blue, spiking activity. (E and F) LFP at 60 to 110 Hz (E) and FR (F) using all NET contacts. Two disconnected contacts are plotted in gray. Restoration of CBF occurred on day 1, while neural activity remained significantly suppressed several days after. Scale bars, 500 μm (A to C) and 200 ms [horizontal in (D)].

  • Fig. 4 Neurovascular disassociation is the most severe at the subacute phase.

    (A) rCBF, rLFP, and rFR at the four NET locations in Fig. 3 for 9 weeks baselined against prestroke values. Solid dots present the mean values, and shading shows the SD. Neurovascular disassociation is the most evident at the subacute phase, where the difference between rCBF and rLFP/rFR is the largest. (B and C) Scatter diagrams between rLFP and rCBF (B) and between rLFP and rFR (C) in three phases: baseline (BL), consecutive 7-day daily measurements before stroke; subacute (SA), days 0 to 14 after stroke; and chronic (CH), weeks 3 to 8 after stroke (location, n = 37 and animal, N = 10). (D to H) Box plots showing the relative values of CBF (D), LFP (E), FR (F), and their difference rCBF − rLFP (G) and rCBF − rFR (H) at the three phases: baseline (green), subacute (magenta), and chronic (blue). All data are plotted as dots. Significance levels are shown as follows: n.s., no significance; **P < 0.01; ***P < 0.001; ****P <0.0001.

  • Fig. 5 In vivo and histological measurements of infarct dimensions.

    (A) LSCI of CBF at the end of the acute session showing the spatial extent and severity of ischemia in all animals. Color map: Overlay of rCBF <50% baselined against prephotothrombotic images. The area under rCBF <20% is denoted under each image. (B) Nissl staining of the infarct tissue at 8 weeks after stroke for three animals as shown in (A). The coronal section templates indicate the infarct positions posterior to bregma. Numbers on the bottom are the relative distance to the templates. Scale bars, 500 μm (A and B).

  • Fig. 6 More severe ischemia leads to longer-lasting neural deficits and neurovascular disassociation.

    (A to F) Pearson’s correlation using acute values of CBF and LFP as predictors for LFP at week 6 after stroke [ρ = 0.41, P = 0.02 in (A); ρ = 0.64, P = 4.1 × 10−5 in (B)], for time-integrated neural deficits DLFP [ρ = −0.49, P = 0.004 in (C); ρ = −0.76, P = 2.3 × 10−7 in (D)], and for time-integrated neurovascular disassociation DCBFDLFP [ρ = −0.33, P = 0.06 in (E); ρ = −0.71, P = 3.3 × 10−6 in (F)]. The Pearson’s correlation coefficient is larger and the P value is smaller when using LFP0 as the predictor than using CBF0 for all three variables.

Supplementary Materials

  • Supplementary Materials

    Multimodal mapping of neural activity and cerebral blood flow reveals long-lasting neurovascular dissociations after small-scale strokes

    Fei He, Colin T. Sullender, Hanlin Zhu, Michael R. Williamson, Xue Li, Zhengtuo Zhao, Theresa A. Jones, Chong Xie, Andrew K. Dunn, Lan Luan

    Download Supplement

    The PDF file includes:

    • Figs. S1 to S7
    • Table S1
    • Legends for movies S1 and S2

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article