Search
Close this search box.

Intrinsic and extrinsic actions of human neural progenitors with SUFU inhibition promote tissue repair and functional recovery from severe spinal cord injury – npj Regenerative Medicine

Generation and characterization of SUFU knockdown(KD) hNPCs derived from human pluripotent stem cells

SUFU is a highly conserved negative regulator of the SHH signaling pathway. Studies have shown that deletion or inhibition of SUFU increases SHH activity, promoting the expression of Olig2, a key regulator of motoneuron and oligodendrocytes differentiation15, and leading to an early onset of neurogenesis or glial differentiation with transiently enhanced proliferation in the CNS or PNS16,17,18,19. These findings suggest that the inactivation of SUFU could trigger intrinsic specification and differentiation programs without relying on extrinsic signaling. This prompted us to investigate whether genetic manipulation of SUFU levels in hPSCs-derived NPCs could trigger the precocious acquisition of therapeutic cell types to enhance their regenerative potential for treating SCI. To address this issue, we established an in vitro differentiation protocol as described previously20 (Supplementary Fig. 1a) to examine the transcript levels of SUFU and SHH effectors GLI1, PATCH1, and HIP1 during neural induction from three hPSCs lines. We used IMR90 for the rest of the studies, although three cell lines displayed similar results. The qPCR results showed that reduced SUFU expression coincided with upregulated SHH target gene expressions, indicating the elevation of SHH signaling in hNPCs during differentiation when compared with hiPSCs and embryoid body(EB)(Supplementary Fig. 1b). These findings suggest that reduced SUFU expression and activation of SHH signaling coincide with the acquisition of hNPC fate. To further explore the link between the reduction of SUFU levels and the specification/differentiation potential of hNPCs, we infected three hPSCs lines with two lentiviral-mediated short hairpin RNAs targeting SUFU (SUFU KD1 and KD2) or scramble as control by lentivirus(Fig. 1a and Supplementary Fig. 1d). Consistent with the inhibitory role of SUFU in SHH signaling, the expressions of SHH and its downstream targets (GLI1, GLI2, PATCH1 and HIP1) were significantly upregulated in SUFU KDs hiPSCs compared to their expression levels in scramble and wild-type (WT) hiPSCs(Supplementary Fig. 1c–e). However, SUFU KD hiPSCs with activated SHH pathway exhibited similar expression levels of pluripotent stem cell markers compared to scramble and WT hPSCs (Supplementary Fig. 1f). In addition, EdU incorporation and annexin V apoptosis assays showed no effects on proliferation and viability in SUFU KD hiPSCs(Supplementary Fig. 1g). These results are consistent with previous studies21 showing that activation of SHH signaling does not affect pluripotency, proliferation, and survival of hPSCs.

Fig. 1: Molecular characterization of SUFU KD hNPCs.
figure 1

a mRNA and protein expression levels of SUFU in hNPCs analyzed by qPCR and immunoblotting in scramble and SUFU KD treatment groups. b qPCR analysis of SHH and its effectors. One-way ANOVA followed by Tukey’s post-hoc test. c Representative bright-field(BF) images showing the formation and renewal of neurospheres in scramble and SUFU KD treatment groups. Quantification of the number (d) and size (e) of neurospheres from scramble and SUFU KD treatment groups, One-way ANOVA followed by Tukey’s post-hoc test. Scale bar = 100 μm. f Representative immunofluorescence images for OLIG2, SOX2, HuC/D, PAX6, and NKX6.1 in hNPCs from scramble and SUFU KD treatment groups. DAPI was used as a nuclei marker. Scale bar = 50 μm. g Quantification of (f). h qPCR analysis of dorsal-ventral markers (PAX6,OLIG2,PAX7), neural genes (SOX2/1, Nestin) and neuronal markers (MPA2,TUJ1,ISL1/2, and HB9) in hNPCs from scramble and SUFU KD treatment groups, One-way ANOVA followed by Tukey’s post-hoc test. All data are expressed as mean ± SEM. Three independent experiments.

Next, we evaluated the effects of SUFU inhibition in hNPCs. Consistently, both SUFU KD1 and KD2 hNPCs(passage <5) exhibited significantly reduced SUFU mRNA and protein expressions concomitant with an upregulation of SHH target genes compared to scramble control (Fig.1a, b). We further examined whether reduced SUFU expression levels could affect the self-renewal capacity of hNPCs using a neurosphere assay in the absence of growth factors, EGF and FGF. We found that secondary neurospheres(P2) from SUFU KDs group exhibits more spherical and transparent multicellular complexes compared to scramble control(Fig. 1c, d). Importantly, SUFU inhibition significantly increased the number of neurospheres and reduced the formation of small-sized (<40μm) neurospheres (Fig. 1d, e). We next examined the impacts of SUFU inhibition on the fate of hNPCs. Compared with the scramble group, SUFU KD1 and KD2 hNPCs had much higher proportions of cells expressing ventral neural progenitor marker NKX6.1, motor neuron/oligodendrocytes progenitor marker OLIG2, and pro-neuronal marker HuC/D with subtle reduction of SOX2 that could be due to accelerated neurogenesis(Fig. 1f, g). Consistently, qPCR analysis revealed significant upregulated ventral neural patterning genes (NXK2.2, NKX6.1, and OLIG1/2) and motor neuronal genes(Tuj1, HB9, and ISLET1/2), along with downregulation of neural plate border markers PAX7(Fig. 1h). These data indicate that SUFU KD hNPCs exhibited enhanced activation of SHH signaling with the acquisition of pro-neurogenic potency bias toward motor neurons and oligodendrocyte lineages.

SUFU inhibition triggers a spontaneous differentiation program to promote the formation of motoneurons and oligodendrocytes

Cell intrinsic activation of SHH in SUFU KD hNPCs led us to investigate the potential of these modified hNPCs in regulating axonal projections/outgrowth, synapse formation, motoneuronal maturation, and oligodendrocytes differentiation, especially under the condition lacking supplementing growth signaling (e.g., EGF, FGF, GDNF, BDNF, and IGF) that is typically deficient in the injured spinal cord. Neurospheres (passage <5) derived from Scramble and SUFU KDs groups were cultured without adding neurotrophic factors. After 7 days of culturing, we observed a robust axonal outgrowth, labeled by the neurofilament marker (NF), in SUFU KDs cells, which exhibited an increased number of nerve fibers with much longer extension from the core of neurosphere compared to the scramble control (Fig. 2a, b). In addition, a substantial portion of the SUFU KD cells expressed ISLET1/2 (KD1:43.9% ± 6.3%; KD2:48.0±6.9% versus scramble:3.9±1.0%) among the MAP2+ pan-neuronal population on 7 days post-differentiation, indicating accelerated motoneuronal differentiation(Fig. 2c, d). At 21 days, more SUFU KDs cells expressed mature motoneuronal marker, choline acetyltransferase (ChAT) and HB9, which were associated with increased expression of the presynaptic marker synaptophysin (Syn) (Fig. 2e, f). The scramble exhibited a higher percentage of undifferentiated hNPCs (39.2± 2.6%) compared to SUFU KDs groups(KD1: 25.6± 1.9%; KD2: 24.8± 2.5%)(Supplementary Fig. 2a, b). Despite SUFU KD hNPCs primarily favouring motoneuronal fate, we observed a subtle increase in the percentage of postmitotic interneurons marked by Lim1/2+, indicating accelerated or enhanced neurogenesis(Supplementary fig. 2a, b). Consistently, qPCR analysis further confirmed significant downregulation of neural genes(SOX2 and PAX6) and upregulation of motoneuronal-associated genes(ISLET1/2, HB9, ChAT), ventral gene(NKX6.1) and pan-neuronal gene(TUJ1) upon SUFU KD (Fig. 2g). These results suggest that knockdown SUFU in hNPCs enhances cell intrinsic differentiation potency, promoting neurogenesis and maturation bias toward motor neuron fate in the absence of extrinsic neurotrophic factors support.

Fig. 2: hNPCs with SUFU inhibition exert cell intrinsic differentiation capacity in generating neurons and oligodendrocytes.
figure 2

a Representative immunofluorescence of neurofilament(NF) from scramble and SUFU KDs neurospheres at 7 days differentiation in the absence of neurotrophic factors. Scale bar = 100 μm. b Quantification of relative intensity of axon outgrowth in scramble and SUFU KDs neurosphere. c Representative immunofluorescence images of MAP2, and ISL1/2 in scramble and SUFU KD hNPCs at 7 days differentiation. The white box shows a magnified view with indicated markers. Scale bar = 100 μm. DAPI was used as a nuclei marker. d Percentage of ISL1/2-positive cells in scramble and SUFU KDs hNPCs after 14 days differentiation. e Representative immunofluorescence images of ChAT, HB9, MAP2, and synaptophysin (SYN) in scramble and SUFU KDs hNPCs at 21 days differentiation. The white box shows a magnified view with indicated markers, Scale bar = 100 μm. DAPI was used as a nuclei marker. f Quantification of (e). g qPCR analysis of HH effectors (SUFU, GLI1, and PATCH1), neural genes (SOX2 and PAX6) and neuronal markers (MPA2,TUJ1,ISL1/2, ChAT, and HB9) at 21 days differentiation of scramble and SUFU KDs treatment groups. h Representative immunofluorescence images of GFAP, NG2 and SOX10 after 3 weeks differentiation in scramble and SUFU KDs hNPCs, Scale bar = 100 μm. i Quantification of (h). Student’s t test. All data are expressed as mean ± SEM. *p < 0.01, **p < 0.05, ***p < 0.005 versus scramble. Three independent experiments.

Next, we evaluated the glial differentiation capacity, including astrocytes and oligodendrocytes in scramble and SUFU KDs hNPCs. After 35 days of culturing in glial differentiation media, GFAP+ astrocytes were more frequently observed in the Scramble hNPCs group compared to SUFU KD groups (Fig. 2h, i). In contrast, the SUFU KD group shows a significantly increased cell population expressing oligodendrocyte markers SOX10 and NG2 when compared to scramble (Fig. 2h, i), most likely due to increased OLIG2 progenitors induced by SHH activation (Fig. 1h). Consistently, the scramble had a higher percentage of undifferentiated hNPCs (26.6±2.4%) compared to SUFU KD (KD1:17.1±3.0%; KD2:16.9±1.7%)(Supplementary Fig. 2c, d). Together, these results suggest that inhibition of SUFU in hNPCs promotes the formation of therapeutic cell types for SCI, including motoneurons and oligodendrocytes.

Non-cell autonomous effects of SUFU KD hNPCs on cell survival and differentiation

SHH was shown to be neuroprotective and function as a survival-promoting factor for tissue repair. To determine whether SUFU KD hNPCs, with intrinsically activated SHH, could counteract the adverse effects from the injured spinal cord niche, hNPCs from scramble and SUFU KDs were cultured in the absence of neurotrophic factors and treated with cleared homogenate (100 µg/ml) from the injured spinal cord (SCI-H) for 2 weeks, as described before. Consistent with previous findings6, only a small portion of scramble cells differentiated into TUJ1+ neurons(16.7 ± 2.7%) with few axonal outgrowths when exposed to SCI-H. Additionally, we observed a significant increase in apoptosis marked by caspases 3 in scramble cells treated with SCI-H after 2 weeks of culture (Fig. 3a, b). In contrast, SUFU KD cells mitigated this effect, generating a substantial amount of neurons(41.7 ± 4.04% in SUFU KD1; 35.4 ± 2.8% in SUFU KD2) with thick nerve fibers and exhibiting a low percentage of caspases 3 expressions as compared to scramble cells upon treatment of SCI-H(Fig. 3a, b). Notably, the superior neurogenic effects observed in SUFUKD hNPCs under an aversive environment cannot be fully achieved through external Shh administration, despite a subtle increase in neuronal populations observed in the presence of SCI-H with SHH supplementation(Supplementary Fig. 2e, f). It is interesting to note that SUFU KD cells not only activated SHH signaling intrinsically but also increased SHH protein production in culture, implicating the non-cell autonomous effects via SHH exposure(Fig. 3c, d). To examine whether SUFU KD hNPCs could modulate other cell activities, we performed co-culture of SUFU KDs(GFP labeling) and scramble(GFP or Non-GFP labeling) (Fig. 3e). Non-GFP scramble NPCs co-cultured with GFP scramble hNPCs or SUFU KDs GFP hNPCs were treated with cleared homogenate (100 µg/ml) without supplementing neurotrophic factors. In accordance with the above findings, non-GFP scramble cells cultured with GFP-scramble cells exhibited increased apoptosis with few differentiating neurons(HuC/D+) after 14 days of culturing(Fig. 3f–h). Conversely, scramble(non-GFP) cells cultured with SUFU KDs(GFP+) hNPCs exhibited a significantly lower percentage of apoptosis and were able to form more HuC/D+ neurons after 14 days culturing(Fig. 3f–h). All these data demonstrated the non-cell autonomous effects of SUFU KD hNPCs in regulating cell survival and differentiation.

Fig. 3: Non-cell autonomous effects of SUFU KD cells.
figure 3

a Representative immunofluorescence images showing increased SHH production from SUFU KDs hNPCs compared to scramble hNPCs. DAPI was used as a nuclei marker, Scale bar = 100 μm b Quantification of (a). c Representative immunofluorescence images of caspase 3, neuronal marker MAP2 and nuclei marker DAPI in Scramble and SUFU KD hNPCs after 14 days treatment of homogenate (100 μg/ml) from the injured spinal cord, Scale bar = 100 μm. d Quantification of caspase 3+ cells from (c). Student’s t test. e Schematic diagram showing co-culture of scramble and SUFU KD hNPCs with and without GFP. f Representative immunofluorescence images of caspase 3, HuC/D and GFP in co-cultures of Scramble+scramble(GFP), Scramble+SUFU KD1(GFP), and scramble+SUFU KD2(GFP), Scale bar = 100 μm. The white box shows a magnified view with indicated markers. g Percentage of caspase 3+ cells in GFP-positive or GFP-negative cells in co-cultures of scramble+scramble(GFP), Scramble+SUFU KD1(GFP), and scramble+SUFU KD2(GFP). One-way ANOVA followed by Tukey’s post-hoc test. h Percentage of HuC/D in total GFP cells in co-cultures of scramble+scramble(GFP), Scramble+SUFU KD1(GFP), and scramble+SUFU KD2(GFP). Student’s t test. All data are expressed as mean ± SEM. *p < 0.01, **p < 0.05, ***p < 0. 005. Three independent experiments.

SUFU KD hNPCs modulated the injured niche for survival and connection

After confirming the superiority of SUFU KDs NPCs in vitro, we further examine whether SUFU KDs hNPCs can provide beneficial effects for the injured spinal cord. We induced thoracic contusion injury at level T8 in rats, followed by cell transplantation at 2 weeks post-injury. A total of 2 × 105 GFP-expressing scramble control or SUFU KDs hNPCs were grafted into the lesion site (1 µL administered into the left and right sides of the injury site) without growth factors supplementation. Anatomical analysis showed that both scramble and SUFU KD1 hNPCs could expand and survive in grafted animals (Fig. 4a), indicating successful integration of host tissues at the lesion site and compensation for the dramatic tissue loss within the cavity. In both lesion control and scramble group, a substantial amount of caspase 3+ apoptotic cells were found in the periphery of the lesion epicenter or within the grafts, whereas SUFU KDs grafts significantly reduced caspases 3 expressions around lesion sites at 1 M post-grafting(Fig. 4a, b). Consistent with previous findings, the lesion site in the injured spinal cord is surrounded by a glial scar with robust expressions of GFAP, prohibiting nerve extension and reconnection from scramble grafts. Remarkably, the density of GFAP+ glial barrier around lesion sites significantly reduced in the injured spinal cord grafted with SUFUKD NPCs. These grafted cells exhibited massive axonal branches, penetrating the GFAP-expressing glial scar as early as 1-month post-graft(1 M)(Fig. 4c). Notably, there were no significant differences of the graft area within the lesion sites between scramble and SUFU KD1 graft (Fig. 4c).

Fig. 4: SUFU KD grafts display efficient integration and modulate injured niche in the SCI model.
figure 4

a Representative images of caspase 3 positive cells in the injured spinal cord without grafting or with scramble and SUFU KD1 grafts at 1 month(1 M) post-graft, Scale bar = 200 µm. b Quantification of the caspase 3 positive cells from (a). One-way ANOVA. *p < 0.05, **p < 0.01. c Representative immunofluorescence images of GFP, MAP2 (red), and GFAP (Blue) in sagittal sections with scramble and SUFU KD1 grafts at 1 month(1 M) post-graft. The cystic lesion cavity (LC) formed with surrounding dense GFAP immunoreactivity (blue), whereas SUFU KD1 grafts (GFP-positive) crossed GFAP barriers. The white box shows a magnified view with indicated markers in lower panels. Scale bar = 200 µm. The dotted line indicates the dense astrocytic glia marked by GFAP. d Representative immunofluorescence images of GFP, CSPG (red), and DAPI (Blue) in spinal cord sagittal sections grafted with scramble and SUFU KD1 hNPCs at 1 month post-graft (1 M). The cystic lesion cavity (LC) formed with surrounding dense CSPG immunoreactivity (red). SUFU KD1 grafts with GFP expression attenuated CSPG graft/host interface. Scale bar = 200 µm. e Fluorescence intensity analysis of CSPG surrounding the lesion cavity (n = 4 rats per group). *p < 0.05. f GFP and Neurofilament(NF) immunolabeling in spinal cord sagittal sections revealed GFP-expressing SUFU KD1 grafts at injured sites generated robust axons extending into the host spinal cord caudally after 2 months post-graft(2 M). insets: a-a” and b-b” indicate higher magnification of NF-positive fibers in the graft at different regions from rostral to caudal. Scale bar = 100 µm. g Quantification of axon intercepts at specific distances from graft-host border in the injured cord grafted with scramble and SUFU KD1 hNPCs. **p < 0.01, ***p < 0.001, one-way ANOVA with Bonferroni. All data are expressed as mean ± SEM.

To determine whether SUFU KD grafts facilitate nerve outgrowth by modulating glial scar, we further examined the expression of CSPGs in the injured spinal cord. CSPGs are produced by reactive astrocytes and act as a key inhibitory component to limit axonal outgrowth and regeneration, oligodendrocyte replacement, and remyelination22. In agreement with the literature1,22, upregulated CSPG expression formed a complex barrier around the lesion sites in response to the injury, restricting the nerve regeneration and outgrowth of scramble grafts (Fig. 4d, e). In contrast, SUFU KD1 graft significantly inhibited CSPG deposition, which resulted in more dispersed CSPG expressions around lesion sites, leading to a number of GFP+ axonal outgrowth from the grafts for reconstituting neuronal connections (Fig. 4d, e). At 2 months post-grafting, we detected robustly projected neurofilaments(NF+) derived from SUFU KD1 graft, whereas scramble grafts were restricted within the lesion sites without obvious axonal outgrowth across the lesion site (Fig. 4f). Importantly, a large number of GFP+ axons co-expressing NF emerged from SUFU KD1 grafts into the injured spinal cord and extended much further caudally by more than 25 mm at 2 months post-graft (Fig. 4f, g), recapitulating the long nerve fibers extension observed in vitro. These findings indicate the transplanted SUFU KD1 hNPCs exert non-cell autonomous effects, which could modulate the injured niche to enhance the survival and the reconstitution of neuronal connections.

SUFU KD grafts promote robust neurogenesis intrinsically and extrinsically in the injured spinal cord

To further determine whether grafted SUFU KD1 hNPCs can surmount the injury environment that lacks growth factors and generate therapeutic cell types, we examined the differentiation capacity of scramble and SUFU KD1 hNPCs grafts. In line with previous findings3,5,23, scramble grafts gave rise to less mature neurons while SUFU KD1 grafts generate a much higher proportion of mature neurons, including motoneurons marked by ISL1/2(SUFU KD1: 27.8% ± 4.1% versus scramble: 9.8% ± 4.1%) and ChAT((SUFU KD1: 17.9% ± 2.1% versus scramble: 4.1% ± 0.7%), glycinergic inhibitory neurons(SUFU KD1: 26.5% ± 5.7% versus scramble: 13.5± 6.9%), CaMKII+ excitatory neurons (SUFU KD1: 32.9± 12.0% versus scramble: 10.9% ± 4.5%), and GABA+ neurons (SUFU KD1: 26.8% ± 5.26% versus scramble:8.4 ± 3.9%)(Fig. 5a–e, Supplementary Fig. 3a, b). The total number of GFP cells expressing glutaminergic marker, vGLUT1 was comparable between scramble and SUFU KD1 grafts. However, SUFU KD1 cells showed a higher density of vGLUT1 puncta(Supplementary Fig. 3c, d).

Fig. 5: SUFU KD hNPC grafts promote beneficial differentiation and regneration intrinsically and extrinsically in the SCI model.
figure 5

a Representative immunofluorescence images of GFP, GlyT2(red), and ISLET1/2(blue) in sagittal sections of injured spinal cord with scramble and SUFU KD1 grafts at 2 month(1 M) post-graft. The empty arrow shows the indicated markers expression in GFP- positive cells. The white arrow shows the indicated markers in GFP negative cells. The white box shows a zoomed-in view of the co-localization of indicated markers. Scale bar = 100 µm. b Quantification of the percentage of GlyT2 and ISLET1/2 in grafts or non-grafts cells from(a). c Representative immunofluorescence images of GFP, GABA(red) and CaMKII (purple) in sagittal sections with scramble and SUFU KD1 grafts at 2 month(1 M) post-graft. The empty arrow shows the indicated markers expression in GFP positive cells. The white arrow shows the indicated markers in GFP negative cells. The white box shows a zoomed-in view of the co-localization of indicated markers. Scale bar = 50 µm. n = 4–5 rats per group. *p < 0.05, **p < 0.01, ***p < 0.005. Quantification of the percentage of CaMKII (d) and GABA (e) in grafts or nongrafts cells from(c). n = 5–6 rats per group. *p < 0.05, **p < 0.01, ***p < 0.005. (f) Representative immunofluorescence images of GFP, SOX10(red) and HNFEL(blue) in sagittal sections with scramble and SUFU KD1 grafts at 2 month(1 M) post-graft. The white box shows a zoomed-in view of the co-localization of indicated markers. Scale bar = 50 µm. (g) Quantification of the percentage of SOX10 in GFP-positive or -negative cells in the injured cord grafted with scramble and SUFU KD1 hNPCs. n = 5–6 rats per group, 4–5 sections/rats. Scale bar = 50 µm, **p < 0.01. (h) GFP-positive grafts from scramble and SUFU KD1 immunolabeled with HNEFL and myelination marker (MBP, red) at 3 months post-graft. Insets show the distribution of GFP-positive grafts in the injury site. Scale bar = 50 µm. All data are presented as mean ± SEM.

Strikingly, we observed significantly increased non-GFP host cells around SUFU KD1 grafts expressing mature neuronal markers compared to those around scramble grafts, including glycinergic inhibitory neurons (SUFU KD1:27.4% ± 10.2% versus Scramble:12.1 ± 7.3%), CaMKII+ excitatory neurons (SUFU KD1:23.4 ± 5.5% versus scramble12.10% ± 7.01%), motoneurons(ISL1/2, SUFUKD1: 29.3% ± 2.3% versus scramble: 15.0% ± 1.9%; ChAT, SUFU KD1: 10.6% ± 1.4% versus scramble: 1.1% ± 0.4%) and GABA+ neurons (SUFU KD:37.3 ± 7.5% versus scramble:10.6% ± 6.3%) (Fig. 5a–e). All these data suggest that SUFU KD1 grafts not only exhibit enhanced neurogenic potential intrinsically but also exert beneficial effects on the injured environment by promoting neurogenesis of host cells. In a study by Erceg et al.8, transplantation of both motoneurons and oligodendrocytes into the SCI of rat models led to better locomotor recovery than transplantation of individual cell types alone. To further investigate whether SUFU grafts also promote oligodendrocytes formation in the injured spinal cord, we examined markers of SOX10 and myelin basic protein (MBP) for oligodendrocytes in the grafts. We found that a substantial amount of SOX10 expression can be detected in SUFU KD1 grafts, whereas SOX10 expression was barely detectable in the scramble grafts(SUFU KD1:18.6 ± 4.2% versus scramble:1.6% ± 1.8%)(Fig. 5f, g). The percentage of non-GFP host cells expressing SOX10 was comparable in both groups, suggesting SUFU grafts did not exert non-cell autonomous effects on regulating oligodendrocyte properties in host. Consequently, SUFU KD1 hNPCs-derived axons were myelinated, as evidenced by the co-localization of MBP-labeled myelin sheath with HNEFL+ axons, whereas no MBP expression was detected with axons emerging from the scramble grafts (Fig. 5h, i). It’s worth noting that the absence of myelination in human axonal fibers in rodent/primate hosts has been reported24, which may be attributed to the lack of inter-species recognition. Altogether, these results demonstrate the intrinsic and extrinsic superiority of SUFU KD hNPCs in generating therapeutic cell types and modulating injury niche for the treatment of SCI.

SUFU KD grafts effectively establish the integration into host neural circuits

The effectiveness of grafts in integrating with the host neural circuitry and promoting tissue repair for locomotion recovery was evaluated by injecting rAAV-hSyn-CRE + AAV-DIO-mCherry into the motor cortex for antegrade tracing of trans-synaptic connectivity after injury25. Histological analysis showed a strong mCherry+ signal in the brain injection sites and injury sites rostrally in the lesion control, scramble and SUFU KD1 graft groups at 2 weeks post-injections(Fig. 6a and Supplementary Fig. 2). In the lesion control group, the trans-synaptic transmission was severely blocked by the cavity in the injured spinal cord, resulting in absence of mCherry+ cells in the caudal region of the lesion sites, which demonstrated disrupted neural circuits following contusive SCI (Supplementary Fig. 4). In the scramble hNPCs group, only a few mCherry+ cells were observed in the lesion sites and caudal to the lesion site, indicating less effective trans-synaptic transmission. In contrast, strong mCherry signals were detected in the lesion sites overlapping with GFP+ cells and several millimeters caudal to the lesion site with SUFU KD1 graft, confirming trans-synaptic spread of mCherry from host neurons to the graft (Fig. 6a–c). In the lumbar spinal cord, a larger number of mCherry+ cells were present in ventral and dorsal horns of the spinal cord grafted with SUFU KD1 hNPCs, which co-localized with ChAT+ motoneurons and GABA+ sensory interneurons, probably representing long descending propriospinal tracts in the spinal cord(Fig. 6d, e). These findings indicate that SUFU KD1 grafts can effectively integrate with both long-projecting and local spinal cord circuitries. In addition, some 5-HT-positive nerve fibers grew into the injury sites and formed synaptic contacts with SUFU KD1 grafts, whereas few synaptic connections were established between 5-HT-positive nerve fibers and in the area caudal lesion site of scramble graft (Fig. 6f). These results demonstrate that SUFU KD1 graft effectively promotes the re-establishment of synaptic connectivity with the major host neural circuitry that normally projects to the spinal cord.

Fig. 6: Graft-initiated trans-synaptic AAV virus antegrade labeling of host connectivity.
figure 6

a Sagittal section showing antegrade, trans-synaptically traced host mCherry-expressing cells in the injured spinal cord with Scramble and SUFU KD1 graft. Scale bar=500 μm. Inset, image showing injection sites in the brain region. Inset (a’-a’’’ and b’-b’’’), high-magnification view of boxed area. b Representative immunofluorescence images showing mCherry+;GFP+ trans-synaptically connection from motor cortex to the grafts in the lesion sites. Scale bar=50 μm. c Quantification of the proportion of mCherry-labeled cells/axons in scramble of SUFU KD1 grafts, normalized to the total number of mCherry-labeled axons located 0.5 mm rostral to the lesion site. n = 7 Scramble recipients and n = 6 SUFU KD1 recipients. One-way ANOVA with Tukey’s multiple comparisons; **p < 0.01, ***p < 0.001, ****p < 0.0001 compared with Scramble grafts. d Transverse sections labeled for mCherry and GFP at L2 host spinal cord levels, showing that host neurons monosynaptically connected to grafts were detected over long lengths of the rat spinal cord. Scale bars, 250 μm. left panel showing synaptically connected mCherry+ host neurons included GABA sensory interneurons and ChAT+ motor neurons at the lumbar level. Scale bars=50 μm. The white arrowheads indicate mCherry traced motoneurons(ChAT+) in SUFU KD1 grafts. e Quantification of mCherry traced neurons in L1-L2 in scramble and SUFU KD1 grafts. n = 7 Scramble recipients and n = 6 SUFU KD1 recipients. ***p < 0.001. f Triple labeling for GFP, 5-HT, and human synapsin(hSYN) revealed colocalization of regenerating 5-HT axon terminals with hSYN, suggesting synaptic connectivity. Scale bar=25μm. All data are presented as mean ± SEM.

SUFU KD grafts improve hindlimb function after contusive SCI

To evaluate the effect of scramble and SUFU KD grafts on functional recovery after SCI, we performed a series of motor function tests during the 16-week post-injury period. Hindlimb locomotor activity in the lesion control, receipts with scramble and SUFU KD grafts was assessed weekly by the Basso, Beattie, and Bresnahan (BBB) locomotor scale, starting 7 days before and after injury. At 1 and 2 weeks post-injury, all treatment groups experienced a dramatic loss of locomotor function, which is attributed to the “spinal shock” associated with the transient shutdown of sensorimotor function in the spinal cord acutely after SCI26. From 3 to 4 weeks onwards, the BBB score of all groups reached a relatively stabilized level(around 7) during the subacute and chronic phases of injury, showing the similar level of locomotion activity reported previously27,28,29. Scramble recipients started to show significantly improved locomotor function from 12 weeks (10 weeks post-graft) compared to lesion control(#p < 0.05, ##p < 0.01), leading to an increase in BBB score from 7(extensive movement of all three joints of the HL) to 11, which aligns with the effects reported in other studies using unmodified NPCs30,31. Notably, SUFU KD recipients demonstrated quicker improvement in hindlimb motor function starting from 10 weeks (8 weeks post-graft) until the end of the experiment (16 weeks post-injury), as compared to lesion control (Fig. 7a). In addition, SUFU KD1 recipients showed much greater improvement in hindlimb motor function compared to scramble recipients animals from 11 weeks post-injury (*p < 0.05,**p < 0.01) (Fig. 7a), resulting in a significant recovery of the BBB score from approximately 7 to 13. To further evaluate skilled locomotor function and coordination, the grid-walking test was employed by counting the percentage of correct steps out of paw replacements and foot faults as rats traversed the metal grid32. After injury, rats lost the ability to place their hind paw correctly on the metal grid. Starting from 10 weeks post-injury, the SUFU KD1 grafted rats exhibited much better performance in placing their affected hind paw (left and right) correctly on the grid with less misdirected steps compared to the scramble recipients, which showed a mild improvement by week 12 compared to lesion control(Fig. 7b, c). Subsequently, the gait of the transplanted animals was assessed by the footprint test, in which the pressure exerted by the feet during locomotor activity was converted into a digital image of the plantar surface by a force sensor, indicating limb stepping ability and coordination. Quantitative analysis of stride length in grafted animals further confirmed a marked improvement in SUFU KD grafted animals with values higher than those of scramble and lesion controls (Fig. 7d). These findings demonstrate the enhanced therapeutic potential of SUFU KD1 grafts for the restoration of locomotor function in a rodent model of contusion SCI.

Fig. 7: Significant functional improvement after transplantation of SUFUKD1 hNPC grafts into contusive SCI.
figure 7

a BBB scores of lesion control, and pre-and post-grafting with scramble and SUFU KD1 hNPCs. Two-way repeated-measures ANOVA followed by post-hoc Fisher’s exact test. *p < 0.05, **p < 0.01 SUFU KD1 versus scramble; #p < 0.05, ##p < 0.01 scramble versus lesion control. b Grid walk quantitative analysis measured as a percentage of hind limb placement. One-way ANOVA with Tukey’s post-hoc test; *p < 0.05, **p < 0.01. c Foot fault score analysis of hind limb measured by rating scale for foot placement in the skilled ladder rung walking test (correct placement = 6 points; partial placement = 5 points; placement correction = 4 points; replacement = 3 points; slight slip = 2 points; deep slip = 1 point; and total miss = 0 points). One-way ANOVA with Tukey’s post-hoc test; *p < 0.05, **p < 0.01. d Quantification of stride length in sham, SCI(lesion control) and SCI rats with scramble and SUFU KD1 grafts. Student’s t test. *p < 0.05, **p < 0.01. All data are expressed as mean ± SEM. n = 7(sScramble); n = 5(lesion Control); n = 9(SUFU KD1); n = 6 (Sham).

Latest Intelligence