Longitudinal evaluation of cell graft-directed remodeling in stroke lesions using optical coherence tomography.

Introduction: Stroke is one of the leading causes of death and disability worldwide, however, there are limitations in available clinical treatments. Stroke lesions are unable to naturally repair and are filled with persistent immune cells and nonneural cells which do not lead to tissue repair. Cell grafts are a promising strategy for directing repair in stroke lesions, but we lack effective methods to longitudinally evaluate the angiogenic activity, graft cell fate and stroke lesion phenotype before and after the cell grafts1,2. Optical Coherence Tomography (OCT) is a non-invasive intravital imaging tool that collects backscattering information from label-free tissue after exposure to laser and uses dynamic information to construct angiograms that provide information on structure and velocity of blood vessels to the micron scale. OCT provides dynamic and functional information on the stroke lesion vasculature that can be tracked in the same animal and OCT image processing methods enable construction of attenuation maps to quantify volumes of stroke and contour3. The objective of this study is to longitudinally track cell grafting outcomes in cortical stroke lesions and make structure function correlations between OCT derived information and post-hoc immunohistochemistry (IHC). We hypothesize that OCT imaging will enable non-invasive and dynamic evaluations of graft-derived glia repair functions at cortical stroke lesions to provide critical information about wound repair outcomes after cell grafting that will correlate with IHC readouts.

Materials and

Methods: A chronic cranial window and head fixtured was implanted over the somatosensory cortex of C57BL/6 mice. A craniotomy was performed over one hemisphere and two circular glass windows of 3mm and 5mm were attached to the surrounding skull and allow to recover before subsequent procedures. For the stroke, a baseline Cerebral Blood Flow (CBF) was be obtained using Laser Speckle Contrast Imaging (LSCI) and Rose Bengal dye was injected retro-orbitally. A pial arterial vessel through the cranial window was targeted and illuminated by a 520nm laser. LSCI was be used to observe occlusion. Two days after stroke induction, Neural Progenitor Cells (NPC) derived from Mouse embryonic stem cells (mESC) that constitutively express the ribosomal protein Rp122 with a hemagglutinin (HA) tag (RiboTag) or saline were injected directly into stroke lesion cores. OCT data was collected at -1, 1-, 7- and 14-days post stroke. Mice were perfused at 2 weeks and brains processed by IHC and imaging analysis methods in ImageJ to visualize and quantify the HA-positive cell graft survival and the multicellular stroke lesion phenotype including responses of astrocytes, neurons, and inflammatory cells. For the OCT data, images were analyzed using established programs in MATLAB. OCT-derived angiograms, lesion contours, and tissue contraction metrics were obtained and correlated to histological metrics.

Results, Conclusions, and Discussions: OCT angiogram and scattering maps of strokes taken at acute time points showed defined stroke lesion regions that dynamically contracted over the 2-week evaluation period. Using OCT images, we calculated tissue-divot sizes, lesion area and surface vascular remodeling. Capillary density was significantly decreased in the non-neural lesion core and persisted at depleted levels through two weeks whereas preserved penumbra and healthy tissue showed vascular recovery by two weeks. We noted separation of the cortical surface from the imaging window that created defined divots detectable by OCT that increased over time while lesion area detected in the axial plane decreased over the same time. OCT and IHC analysis revealed that NPC grafts did not alter the lesion vasculature compared to saline injected control animals. We identified significant and increased anisotropic contraction in lesions receiving cell grafts that was detectable by both OCT and IHC analysis. In these animals, cell grafts showed preferential survival at lesion borders along the glia limitans and differentiated into Gfap-positive astrocytes that seemingly contributed to the anisotropic contraction of the lesion.
Here we optimize cell grafting techniques for cortical photothrombotic strokes and demonstrated proof of principle that OCT can be used as a non-invasive method to longitudinally track graft derived wound repair after stroke. Future studies will involve combining Oct with other intravital imaging modalities, such as two photon microscopy and ultrasound to improve characterization of graft derived wound repair in stroke.

Acknowledgements (Optional): Boston University Core Facilities and Neurophotonics Center. This work was supported by funding from Boston University (T.M.O), NINDS (R21NS128821) (T.M.O), and Schlumberger Foundation (H.A).