In vivo direct reprogramming of glial linage to mature neurons after cerebral ischemia

The therapeutic effect of in vivo direct reprogramming on ischemic stroke has not been evaluated. In the present study, a retroviral solution (1.5–2.0 × 107 /ul) of mock pMX-GFP (n = 13) or pMX-Ascl1/Sox2/NeuroD1 (ASN) (n = 14) was directly injected into the ipsilateral striatum and cortex 3 days after 30 min of transient cerebral ischemia. The reprogrammed cells first expressed neuronal progenitor marker Dcx 7 and 21 days after viral injection, then expressed mature neuronal marker NeuN. This was accompanied by morphological changes, including long processes and synapse-like structures, 49 days after viral injection. Meanwhile, therapeutic improvement was not detected both in clinical scores or infarct volume. The present study provides a future novel self-repair strategy for ischemic stroke with beneficial modifications of the inducer-suppressor balance.


Results
As the first pilot study, GFP-positive cells were located mainly in the ipsilateral cerebral cortex and lateral striatum. The number of GFP-positive cells was significantly larger in the mice brain that received a viral injection 3 days after tMCAO than in other groups (3 d, 96.2 ± 8.9; 7 d, 29.3 ± 19.1; 14 d, 4.0 ± 0.4; cell number/0.12 mm 2 , *p < 0.05 vs 7 d, # p < 0.05 vs 14 d, Fig. 1a-c). Thus, the appropriate time point for viral injection was set at 3 days after tMCAO in the following experiment.
At 21 days after viral injection (24 days after tMCAO), GFP/Dcx double-positive cells were found exclusively in the ipsilateral striatum of the pMX-ASN group (27 cells/141 GFP-positive cells of 12 brain slices from four mice, Fig. 4b), and no GFP/NeuN double-positive cells were found (0/201 GFP-positive cells of 12 brain slices from four mice, Fig. 4b right panel). At 49 days after viral injection (52 days after tMCAO), GFP/Dcx double-positive cells disappeared (0/441 GFP-positive cells of 21 brain slices from seven mice, Fig. 4c), but GFP-positive cells co-expressing mature neuronal marker NeuN were finally detected. These cells displayed elongated processes with synapse-like structures (18/232 GFP-positive cells of 21 brain slices from seven mice, Fig. 4c right two panels, arrows and arrowheads). Figure 5A,B show an infarct volume at 52 days after tMCAO with no significant difference between the mock pMX-GFP and pMX-ASN groups (pMX-GFP group, 14.8 ± 9.0 mm 3 ; pMX-ASN group, 16.2 ± 8.8 mm 3 ). In addition, there were no significant differences in body weight, Bederson's score and the corner test between the two experimental groups (Fig. 5C).

Discussion
The present study is the first report to show in vivo direct reprogramming of a stroke animal model. In vivo enforcement of transcriptional factors (Ascl1, Sox2 and NeuroD1) successfully induced ectopic neuronal cells in the ipsilateral cerebral cortex and lateral striatum of the post-stroke mice brain. The directly reprogrammed neuronal cells first expressed neuronal progenitor marker Dcx 7 and 21 days after viral injection, then expressed mature neuronal marker NeuN, processes that were accompanied by morphological changes such as long processes and synapse-like structures 49 days after viral injection (Figs 3 and 4). In the present study, a retroviral delivery system was selected for in vivo direct reprogramming because retroviruses only infect dividing cells such as microglia, astroglia, or some progenitor cells 16 . The cell types infected by the retroviral vector were not neuronal cells or mature oligodendrocytes, but were microglia (41.9% of GFP-positive cells), astroglia (40.5%) and oligoprogenitor cells (14.5%), suggesting, as one possibility, that the origin of induced neuronal cells is from microglia, astroglia, and/or oligoprogenitor cells (Fig. 2), a similar conclusion drawn in previous reports 9, 13,15 . However, another possibility is that undetected endogenous stem cells or other cell types might also take part in ectopic direct reprogramming to neurons, because 3.1% of infected cells types were unknown 48 h after viral injection (Fig. 2b).
Despite the success of ectopic neurogenesis, the present direct reprogramming methods did not therapeutically improve stroke animals, in terms of clinical scores and infarct volume (Fig. 5). This may be due to an insufficient number of induced neuronal cells that could contribute to the functional recovery of post-stroke injury. Enforced master transcriptional factors such as Ascl1, Sox2 and NeuroD1 worked as inducers to convert somatic cells into neuronal cells 12,13,17 . In contrast, multiple hurdles such as the p53-p21 pathway, CAF-1 complex, RE-1 transcription repressor complex (REST) and excessive oxidative stress served as suppressors to stabilize cell fate and prevent the reprogramming of somatic cells [18][19][20][21] . Data from these previous studies suggests that cell fate may finally be determined in vivo depending on the balance between inducers and suppressors.
Taken together, the present study successfully achieved, for the first time, in vivo direct reprogramming by enforced transcriptional factors (Ascl1, Sox2 and NeuroD1) in the post-stroke mouse brain. It is anticipated that the findings of induced neuronal cells described herein will be of fundamental importance to studying molecular mechanisms in order to modulate cell fate in the injured brain and for developing novel neuronal repair strategies.

Materials and Methods
Animals and experimental groups. The data that support the findings of this study are available from the corresponding author upon reasonable request. All animal experiments were approved by the Institutional Animal Care and Use Committee of Okayama University (OKU-2017245), and performed in accordance with the guidelines of Okayama University on animal experiments. Adult male ICR mice (33-36 g, 8 weeks old) were used in this study. As the first pilot study to determine the appropriate time point of viral injection to the post-stroke www.nature.com/scientificreports www.nature.com/scientificreports/ brain, mice received an intracerebral injection of pMX-green fluorescence protein (GFP) 3, 7 or 14 days after 30 min of transient middle cerebral artery occlusion (tMCAO). Mice were sacrificed 7 days after each viral injection (n = 3 for each, Fig. 1). For the second pilot study to determine the original cell types infected by a retrovirus, mice received pMX-GFP injection 3 days after tMCAO, and were sacrificed 48 h after the viral injection (n = 3, Fig. 2). As a legitimate experiment to evaluate the therapeutic effect of this treatment, two mice groups received either mock pMX-GFP (n = 13) or in vivo direct reprogramming pMX-Ascl1/Sox2/NeuroD1 (ASN) (n = 14). In both cases, mice received a viral injection 3 days after tMCAO. To evaluate the expression of 3 genes in the post-ischemic brain, we performed double immunofluorescent analysis of retroviral vector (GFP) plus Ascl1, Sox2 and NeuroD1 at 7 days after viral injection (10 days after tMCAO), and confirmed that the subpopulation of GFP-positive cells expressed Ascl1, Sox2 and NeuroD1, respectively (40.8%, 16.7% and 10.2%) (See Supplementary Fig. S1). These mice were sacrificed 7, 21 and 49 days after the viral injection (pMX-GFP group; n = 3, 3, 7, pMX-ASN group; n = 3, 4, 7, respectively, see Figs 3, 4 and 5).
focal cerebral ischemia. During surgery, mice were anesthetized with a mixture of nitrous oxide, oxygen and isoflurane (69/30/1%). tMCAO was induced by the intraluminal filament technique 22 . In brief, the right carotid bifurcation was exposed, and the external carotid artery was coagulated distal to the bifurcation. A silicone-coated 8-0 filament was then inserted through the stump of the external cerebral artery and gently advanced (9.0-10.0 mm) to occlude the middle cerebral artery. After 30 min of occlusion, the filament was gently withdrawn, and the incision was closed.  www.nature.com/scientificreports www.nature.com/scientificreports/ Behavioral analysis. Just before tMCAO, 3, 10, 17, 24, 31, 38, and 52 days after tMCAO, mice were tested for behavioral changes and scored, as described by Bederson, but with minor modifications 25 , as follows: 0, no observable neurologic deficits; 1, failure to extend the right forepaw; 2, circling to the contralateral side; 3, falling to the right; 4, unable to walk spontaneously. A corner test was also carried out to detect the impairment of sensorimotor function 26 . Briefly, one mouse was placed between two boards, which were attached to an edge at a 30° angle to each other. After this test was repeated 10 times for each mouse, the number of right turns was recorded.

Quantitative and statistical analyses.
To evaluate the number of GFP-positive cells, stained sections were selected from three levels of the caudate putamen (1.0, 0.5 and 0 mm rostral to the bregma) 24 of each animal. Three areas around the site of viral injection were randomly selected in each section, and captured at ×200 magnification with a microscope (BX51; Olympus). For the cell-type-specific marker/GFP double-labeling immunohistochemistry (see Figs 2, 3, and 4), three areas around the site of viral injection were randomly chosen and captured at 100× magnification with a confocal laser microscope (LSM780; Zeiss). Values are expressed as means ± S.D. Differences in the number of GFP-positive cells and behavioral analysis were evaluated for statistical significance by non-repeated measures analysis of variance (ANOVA) and the Student-Newman-Keuls (SNK) test. In all statistical analyses, significance was assumed at p < 0.05.

Figure 5.
Clinico-pathological effects of in vivo direct reprogramming, showing (A) cresyl violet staining at 52 d after tMCAO, and no significant difference between mock pMX-GFP (n = 7) and pMX-ASN (n = 7) in terms of (B) infarct volume, (C) body weight, Bederson's score or the corner test.