Introduction

The central nervous system (CNS) is composed of neurons and glial cells, with astrocytes being the most abundant, accounting for 20–40% of mammalian CNS cells1. Neurons, which are highly differentiated and non-regenerative, are vastly outnumbered by glial cells, with a glia/neuron ratio of approximately 5.6–7.1:1 in the human spinal cord2,3. Due to the limited regenerative capacity of neurons, neurodegenerative diseases and neural trauma, characterized by neural cell loss and disrupted functional neural circuits, are difficult to cure with the current treatment methods and often cause severe neurological sequelae4,5,6,7. As a result, research has increasingly shifted toward strategies to replenish neurons and restore neurological function8,9 Cell replacement strategy is one of the most promising treatments for CNS degeneration, involving stem cell transplantation or direct in situ glia-to-neuron conversion. This process requires multiple steps including the transplantation/conversion of cells, neuronal survival, circuit integration, and proper connection establishment10. While stem cell transplantation shows therapeutic potential in many studies for replacing lost neurons, it faces challenges such as tumorigenesis, infection, and immune rejection8,11. Alternatively, direct neuronal trans-differentiation or reprogramming, triggered by small-molecule reprogramming drugs, the upregulation of neural transcription factors (e.g., NeuroD1, Dlx2, Ascl1), or Ptbp1 knockdown, appears to promote localized neuronal regeneration with fewer risks12,13,14,15,16,17. PTBP1 (also PTB) belongs to the heterogeneous nuclear ribonucleoprotein (hnRNP) family and is expressed across multiple tissues and cell types, including glial cells and neural stem cells, but it is uniquely silenced in neurons18. It plays a critical role in neurodevelopment by acting as a splicing inhibitor and regulating neuron-specific alternative splicing events19. It is intriguing that the knockdown of Ptbp1 alone can efficiently convert the abundant astrocytes into rare neurons. Moreover, compared to other in situ reprogramming strategies, Ptbp1 knockdown may pose a lower risk of inadvertently reprogramming pre-existing neurons or disrupting their normal function, due to the natural low expression of PTBP1 in these neurons. Astrocytes present a promising therapeutic target for SCI intervention due to their dual role: they can protect nerve tissue by neural protection and limiting damage spread, yet they may also hinder nerve regeneration through glial scar formation and chronic inflammation20,21,22. Due to their abundance in the CNS, unlimited self-renewal potential, and shared neuroepithelial origin23, astrocytes are considered a suitable and feasible option for glia-to-neuron conversion. By directly reprogramming astrocytes into neurons in situ, this approach reduces glial cell proliferation, alleviates neuroinflammation, prevents further neuronal necrosis and glial scar formation, and reverses neuronal loss by the generation of new neurons23. Xue et al. discovered that knocking down Ptbp1 converted both mouse and human astrocytes into neurons in vitro24,25,26. Furthermore, they successfully reconstructed the nigrostriatal circuit in a chemically induced Parkinson’s disease mouse model, leading to the recovery of dopamine levels and enhancement of motor function15. The therapeutic viability of Ptbp1 knockdown in CNS degeneration had been further confirmed in various disease models in vivo16,17,27.

However, it remains a controversy whether such conversion truly occurs as some laboratories have encountered challenges in replicating these results28. Much of the debate centers around distinguishing glia-converted neurons from pre-existing neurons, as in vivo studies may experience the “neuronal leakage” of AAV vectors and inevitably transduce neurons with marked virus, even when using the astrocyte-specific GFAP promoters28,29. Some studies had successfully knocked down Ptbp1 levels in vitro using antisense oligonucleotide (ASO), short hairpin RNA (shRNA), or CRISPR-CasRx techniques, yet failed to observe any neuronal trans-differentiation in lineage-traced astrocytes28,30,31,32. Moreover, the conversion efficiency and maturation of converted neurons to restore neural function, if achievable, remain unclear and require further investigation. The existing studies affirming the Ptbp1 knockdown-induced neuronal reprogramming in vitro employed a synergistic use of reprogramming agents including SB431542 and CHIR9902115,27,33. These agents are known to significantly enhance conversion efficiency, as they have been demonstrated to promote reprogramed astrocytes into mature neurons independently33,34,35. This raises the following question: is Ptbp1 knockdown alone enough to induce astrocyte-to-neuron conversion across different experimental settings? Moreover, astrocytes can rapidly respond to acute or chronic pathological insults, such as traumatic injury, neurodegenerative diseases, and CNS infections, transforming into reactive astrocytes21,36. This transformation is accompanied by changes in global transcriptomes, molecules, morphologies, and specific functions of cells36. These cells can express elevated levels of inhibitory factors and inflammatory markers, which can be detrimental to the survival of neurons and oligodendrocytes20,21 and may hinder the conversion process, highlighting the variability and complexity of this potential therapeutic approach.

In this study, we utilized shRNA and siRNA to knock down Ptbp1 expression in primary rat spinal astrocytes under different conditions and cellular states, free from other contaminating factors in vitro, aiming to confirm whether astrocyte-to-neuron conversion truly occurs in a relatively controlled environment. Additionally, we investigated methods to enhance the functional maturity of the converted neuron-like cells.

Results

Verification of astrocyte purity and experiment framework

Given the potential confounding factors in astrocyte-to-neuron (AtN) conversion in in vivo studies, it is essential to verify the identity and purity of cultured astrocytes before transduction in vitro, ensuring the exclusion of neurons and neural stem cells. By using distinct cell adhesion properties and mechanical shaking techniques, astrocytes can be efficiently isolated and purified from rat spinal cord tissue, providing an effective method for obtaining nearly pure astrocyte cultures37. In our study, astrocytes were isolated by a stringent protocol from neonatal rat spinal cord, followed by a further purification protocol37 including the differential adhesion method to eliminate contaminating cells. To ensure the absence of neural stem cells and neurons, astrocytes were passaged five times and cultured in astrocyte complete medium before being utilized in further neural reprogramming experiments, to promote cell proliferation and further enrich the astrocyte population. Before transduction with lentivirus (Fig. 1A) or transfection with siRNA, the cells exhibited a typical astrocyte-like flattened and polygonal morphology. Immunofluorescence staining confirmed that nearly all cells expressed various astrocyte-specific markers, including GFAP, S100β, and SOX9 (Fig. 1B–D). Conversely, no cells expressed neuron-specific markers, including Tuj1, NEUN, and MAP2, or the neural stem cell-specific marker Nestin (Fig. 1C, D), indicating that the presence of neurons or neural stem cells was negligible. Therefore, the possibility of interference by these cells can be effectively excluded in the subsequent AtN conversion study (Fig. 1E).

Fig. 1
figure 1

The verification of astrocyte purity and arrangement framework for experiments. (A) Schematic diagram of lentiviral plasmid. (B) Representative images of immunofluorescence staining of primary spinal cord astrocytes with astrocyte-specific markers GFAP (green, Alexa Fluor 488) and S100β (red, Alexa Fluor 647). Almost all cells were simultaneously stained positive for both the astrocyte-specific markers GFAP and S100β. Scale bar: 100 μm. (C) Representative merged images of immunofluorescence staining of primary spinal cord astrocytes with astrocyte-specific markers GFAP (green, Alexa Fluor 488), S100β (red, Alexa Fluor 647), and SOX9 (green, Alexa Fluor 488) , neuron-specific markers Tuj1 (red, Alexa Fluor 647), NEUN (red, Alexa Fluor 647), and MAP2 (red, Alexa Fluor 647), and neural stem cell-specific marker Nestin (red, Alexa Fluor 647). Scale bar: 100 μm. (D) Quantification of the number of immunofluorescence-positive cells stained by each marker among total cells. Nearly all cells expressed astrocyte-specific markers, and the presence of neurons and neural stem cells was excluded. Approximately 300 cells from a randomly selected field were used for quantification in each experiment. The experiments were performed three times. (E) Experimental timeline of this study, created by biorender. All data are presented as means ± SD. GFAP: glial fibrillary acidic protein; S100β: S100 calcium-binding protein β; SOX9: SRY-Box transcription factor 9; Tuj1: tubulin beta 3; MAP2: microtubule-associated protein 2; NEUN: neuronal nuclei.

Full size image

shPTB induced conversion of non-reactive rat spinal cord astrocytes into neuron-like cells

To confirm the trans-differentiation of non-reactive astrocytes to neurons, we transduced primary rat spinal cord astrocytes with lentiviruses expressing shRNA against Ptbp1 (shPTB-1 and shPTB-2). After 2 days of lentivirus exposure, approximately 90% of the cells of each were successfully transduced with the vector containing GFP, showing green fluorescence under fluorescence microscopy. The transduction efficiency was further validated by flow cytometry, which confirmed a transduction rate of 96.4% (Fig. 2A). The results of qPCR analyses and Western blotting confirmed that shPTB-1 significantly downregulated both Ptbp1 mRNA and protein levels, whereas shPTB-2 exhibited relatively lower knockdown efficiency (Fig. 2B, C). Additionally, within the first few days after transduction, the astrocytes’ flattened polygonal morphology began to transform into a thinner, more contracted neuron-like shape (Fig. 2D). By four weeks, astrocytes with Ptbp1 knockdown exhibited neuron-like characteristics, including small, spherical or oval-shaped cell bodies, and the development of elongated axons and extensive intercellular connections, resembling the multipolar neurons abundant in the spinal cord (Fig. 2E, F). This indicates that Ptbp1 knockdown may have converted spinal cord astrocytes into neuron-like cells.

Fig. 2
figure 2

shPTB induced the conversion of non-reactive astrocytes to neuron-like cells in vitro. (A) The transduction efficiency of lentiviral vectors was assessed 2 days post-transduction by flow cytometry, based on GFP fluorescence. Flow cytometry analysis revealed that 96.4% of the total cells were successfully transduced with GFP-shPTB. (B) The mRNA levels of Ptbp1 were measured 48 h after knockdown by shRNA, normalized to the control group (one-way ANOVA followed by Bonferroni post hoc test). (C) The protein expression of PTB was measured by Western blot after silencing with sh-PTB in astrocytes for 48 h. (D) Representative phase-contrast images of morphology changes of non-reactive astrocytes after Ptbp1 knockdown by shPTB-1 for 4 days, starting from day 1 post-transduction. The astrocytes’ flattened, polygonal morphology (yellow line) gradually transformed into a thinner and more contracted neuron-like shape (red line), whereas the control group exhibited no significant morphological changes. Scale bar: 100 μm. (E) Representative phase-contrast, GFP fluorescence, and merged images of morphology changes of non-reactive astrocytes after Ptbp1 knockdown by shPTB-1 after 4 weeks, compared to the control virus group. These cells exhibited neuron-like characteristics, including small spherical or oval-shaped cell bodies, elongated axons and extensive intercellular connections. Scale bar: 100 μm. (F) The maximum neurite (or process) length of reprogrammed cells after 4 weeks of conversion, compared to the shCtrl group (n = 8 random fields per group, with neurites measurements taken from typical cells in each field), and analyzed by two-tailed Student’s t-test. All data are presented as means ± SD. The experiments were performed three times. **P < 0. 01 and ***P < 0.001. PTB: polypyrimidine tract-binding protein; GFP: green fluorescent protein; D: day.

Full size image

One week after the transduction of shPTB-1 and shPTB-2, immunofluorescence staining analysis showed the percentage of Tuj1+/GFP+ cells reached approximately 18% and 16%, respectively (Fig. 3A, C). In contrast, astrocytes transduced with the control virus showed no Tuj1+ cells and were only positive for GFP (Fig. 3A). While most astrocytes were successfully transduced, the conversion process gradually intensified over time. As the culture period continued to the second week, the proportion of Tuj1+/GFP+ cells slightly increased, and the cells adopt a more typical neuronal morphology (Fig. 3B, C). By the fourth week, the conversion rate reached approximately 23% (Fig. 3C). Concurrently, the GFAP expression levels declined in GFP+Tuj1+ cells compared to Tuj1- cells, with a slight reduction in GFAP expression after 1 week (Fig. 3A, D) and a more pronounced reduction after 2 weeks (Fig. 3B, D). By 4 weeks, both knockdown groups exhibited a significant reduction in the mean fluorescence intensity of GFAP in GFP+ cells (Fig. 3E). The different initial levels of Ptbp1 knockdown seem to have no significant impact on the conversion rate, as no significant differences were observed between the lentivirus groups at each time point (Fig. 3C). The conversion process is gradual and progressive, characterized by a steady enhancement over time (Fig. 3C), rather than the rapid differentiation over a short period typically seen in neural stem cells. No GFP+Tuj1+ cells were observed in the control virus group, and no GFP-Tuj1+ cells were detected in both knockdown virus groups, which further affirmed the successful conversion of astrocytes to neuron-like cells rather than the presence of pre-existing neurons or neural stem cells. We also observed that some GFP+ cells exhibited no morphological or neural-specific fluorescent changes throughout the four-week period, suggesting that a considerable number of astrocytes failed to initiate the reprogramming process following Ptbp1 knockdown. In conclusion, these results demonstrate that Ptbp1 knockdown initiates a gradual conversion of astrocytes into neuronal-like cells, with noticeable changes emerging a few days after transduction and progressively intensifying over time.

Fig. 3
figure 3

Immunofluorescence staining changes during shPTB-induced astrocyte-to-neuron-like cell conversion. (A) Representative images of immunofluorescence staining of non-reactive astrocytes after Ptbp1 knockdown by shRNA after 1 week, with GFP (green), astrocyte-specific marker GFAP (indigo, Alexa Fluor 555), and neuron-specific marker Tuj1(red, Alexa Fluor 647). The lowest panels show the enlarged views of the regions outlined by the white dashed boxes in the shPTB-2 group panels. The arrowheads represent Tuj1-positive cells with slightly reduced GFAP expression after 1 week, in contrast to the Tuj1-negative cells (arrows). Scale bar: 100 μm. (B) Representative images of immunofluorescence staining of non-reactive astrocytes after Ptbp1 knockdown by shRNA for 2 weeks, with GFP (green), astrocyte-specific marker GFAP (indigo, Alexa Fluor 555), and neuron-specific marker Tuj1 (red, Alexa Fluor 647). The lowest panels show enlarged views of the regions outlined by the white dashed boxes in the shPTB-2 group panels. The arrowheads represent Tuj1-positive cells with noticeably reduced GFAP expression after 2 weeks, in contrast to the Tuj1-negative cells (arrows). Scale bar: 100 μm. (C) Quantification of Tuj1+ cells within GFP+ cells at 1, 2, 3, and 4 weeks following shPTB-1 and shPTB-2 transduction, compared to the shCtrl group. The proportion of Tuj1+ cells among GFP+ cells increased over time following transduction in both the shPTB-1 and shPTB-2 groups, reaching approximately 23% at four weeks. (D) Quantitative analysis of GFAP fluorescence intensity in GFP+Tuj1+ cells compared to GFP-Tuj1- cells at 1 and 2 weeks post-transduction in the shPTB-2 group (two-tailed Student’s t-test). (E) Quantitative analysis of the mean fluorescence intensity of GFAP in GFP+ cells from shPTB-1 and shPTB-2 group at 4 weeks, compared to the shCtrl group (one-way ANOVA followed by Bonferroni post hoc test). The expression levels of GFAP, a key element of the astrocyte cytoskeleton, significantly decreased in both the shPTB-1 and shPTB-2 groups after 4 weeks of transduction. Approximately 100 cells from each of three randomly selected non-overlapping fields were used for quantification per experiment. All data are presented as means ± SD. The experiments were performed three times. ***P < 0.001. n.d.: not detected; PTB: polypyrimidine tract-binding protein; GFAP: glial fibrillary acidic protein; Tuj1: tubulin beta 3; GFP: green fluorescent protein; w: week.

Full size image

siRNA-PTB induced conversion of non-reactive rat spinal cord astrocytes into neuron-like cells

Small interfering RNA (siRNA) is a double-stranded RNA primarily used to bind and degrade target mRNA, thereby blocking its translation and regulating gene expression. It has gained significant attention for the treatment of various diseases, including neurological and neurodegenerative disorders, due to its extraordinary specificity, high-efficiency knockdown effects, relatively long-lasting but not permanent effects and reversable potential38. Although antisense oligonucleotides (ASOs) have been successfully applied to knock down Ptbp1 in both brain and spinal cord astrocytes, siRNA’s application in AtN conversion has not been yet explored. Thus, this study applied siRNA using the same sequences as those of lentiviral vectors to test siRNA’s efficacy in AtN conversion. As shown in Fig. 4A and C, the mRNA and protein expression of PTBP1 was effectively silenced following siRNA-PTB transfection after 2 days. In contrast to previous shRNA experiments, the knockdown efficiency of both siRNA vector series was significantly higher, highlighting the superior effectiveness of siRNA in the same vector sequence. Correspondingly, the mRNA and protein level of PTBP2 (nPTB) was markedly elevated in both the siRNA-PTB-1 and siRNA-PTB-2 groups, contrasting with its absence in the siRNA-Ctrl group (Fig. 4B, C).

Fig. 4
figure 4

siRNA-mediated Ptbp1 knockdown induces the conversion of astrocytes into neuron-like cells in vitro. (A) The mRNA levels of Ptbp1 were measured 48 h after knockdown by siRNA, normalized to the control group (one-way ANOVA followed by Bonferroni post hoc test). (B) The mRNA levels of Ptbp2 were measured 48 h after knockdown by siRNA, normalized to the control group (one-way ANOVA followed by Bonferroni post hoc test). (C) The protein expressions of PTB and nPTB (PTBP2) were measured by Western blot after silencing with siRNA-PTB in astrocytes for 48 h. (D) Representative images of immunofluorescence staining of non-reactive astrocytes 1 week after Ptbp1 knockdown by siRNA, with PTBP1 (green, Alexa Fluor 488) and the neuron-specific marker Tuj1 (red, Alexa Fluor 647). Many Ptbp1- cells exhibited positive Tuj1 staining and neuron-like morphology in the field. Scale bar: 100 μm. (E) Representative images of immunofluorescence staining of non-reactive astrocytes 2 weeks after transfected by siRNA-PTB-1 or siRNA-Ctrl, with Ptbp1 (green, Alexa Fluor 488) and the neuron-specific marker Tuj1 (red, Alexa Fluor 647). The upper and lower panels in the siRNA-PTB-1 group show the enlarged views of the regions outlined by the white dashed boxes in the middle overall panels. As a cluster of Ptbp1-Tuj1+ cells exhibited a neuron-like morphology, it should be noted that a considerable number of Ptbp1- cells did not show positive Tuj1 staining. Scale bar: 100 μm. (F) Quantification of Tuj1+ cells within Ptbp1- cells at 1 and 2 weeks following siRNA-PTB-1 and siRNA-PTB-2 transfection, compared to the siRNA-Ctrl group. The experiments were performed three times. All data are presented as means ± SD. ***P < 0.001. n.d.: not detected; PTB: polypyrimidine tract-binding protein; Ptbp1: polypyrimidine tract-binding protein; Tuj1: tubulin beta 3; w: week.

Full size image

After one week of transfection, Ptbp1-Tuj1+ cells were detected by immunofluorescence staining, making up approximately 15% of all Ptbp1- cells (Fig. 4D, F). Immunofluorescence staining analysis two weeks post-transfection showed a slight increase in the number of Ptbp1-Tuj1+ cells with a more typical neuron-like morphology (Fig. 4E, F), and these cells were still observed after three weeks, demonstrating the consistent and sustained conversion potential of siRNA-induced Ptbp1 silencing. Similarly, the conversion efficiency of Ptbp1- cells had no significant difference between the siRNA-PTB-1 and siRNA-PTB-2 groups. We observed the Ptbp1-Tuj1+ cells had a tendency to form clusters, although not all Ptbp1- cells exhibited positive Tuj1 staining or morphological changes (Fig. 4E). Additionally, rare Ptbp1+Tuj1+ cells were detected across the entire field, indicating that the Tuj1+ cells uniformly lost PTBP1 expression. In summary, siRNA-PTB also initiates the conversion of astrocytes to neuron-like cells, demonstrating its potential to induce neuronal reprogramming.

Different states of astrocytes as starting cells resulted in distinct outcomes in trans-differentiation

In vitro, LPS-induced astrocyte proliferation and activation is a commonly used model to mimic the reactive astrocytes observed after CNS injury21. Dexamethasone (DEX), a potent glucocorticoid, is known for its anti-inflammatory properties and is often used to counteract the inflammatory state observed in LPS-activated astrocytes39. After LPS induction, significant changes were observed in astrocyte mRNA expression levels. Notably, the cytoskeletal protein GFAP was markedly increased (Fig. 5A). The reactive astrocyte marker C3 and pro-inflammatory factor CCL2 were significantly upregulated, while the anti-inflammatory cytokine IL-10 was markedly downregulated (Fig. 5B–D), indicating a shift toward an inflammatory reactive astrocyte phenotype. Following the addition of DEX, the mRNA levels of inflammatory markers in LPS-induced reactive astrocytes were partially reversed, as CCL2 and C3 expression levels were significantly lower than those in LPS-activated astrocytes, although the CCL2 expression remained higher than that in non-reactive astrocytes (Fig. 5B, 5C). Meanwhile, the expression of the anti-inflammatory cytokine IL-10 was significantly elevated (Fig. 5D). These results suggest that DEX partially reversed the inflammatory astrocyte reactivity induced by LPS.

Fig. 5
figure 5

LPS and DEX resulted in different inflammatory states of astrocytes. (A) Comparison of GFAP mRNA levels between non-reactive astrocytes and LPS-induced reactive astrocytes, with results normalized to the non-reactive astrocytes group (two-tailed Student’s t-test). (B) Comparison of C3 mRNA levels between non-reactive astrocytes (Lucid), LPS-induced reactive astrocytes (LPS), and DEX-reversed reactive astrocytes (LPS + DEX), with results normalized to the Lucid group (one-way ANOVA followed by Bonferroni post hoc test). (C) Comparison of CCL2 mRNA levels between Lucid, LPS, and LPS + DEX groups, with results normalized to the Lucid group (one-way ANOVA followed by Bonferroni post hoc test). (D) Comparison of IL-10 mRNA levels between Lucid, LPS, and LPS + DEX groups, with results normalized to the Lucid group (one-way ANOVA followed by Bonferroni post hoc test). All data are presented as means ± SD. The experiments were performed three times. **P < 0. 01 and ***P < 0.001. LPS: lipopolysaccharide; DEX: dexamethasone; ns: no significance.

Full size image

LPS-induced reactive astrocytes exhibit enlarged cell bodies, with thicker processes forming a web-like structure (Fig. 6A). After the transduction period of reactive astrocytes extended to four weeks with shRNA, the proportion of Tuj1+/GFP+ cells was notably lower compared to that observed in non-reactive astrocytes transduced for the same duration (Fig. 6A, B). Moreover, the GFP+Tuj1+ cells exhibited an enlarged cellular morphology that was distinct from the more compact, neuron-like appearance observed in the shPTB-1 group (Fig. 6A). This demonstrates that the knockdown of Ptbp1 in reactive astrocytes can start their conversion into neuron-like cells in vitro. However, LPS-induced astrocyte reactivity may negatively hinder the conversion process.

Fig. 6
figure 6

The different states of astrocytes following Ptbp1 knockdown resulted in distinct outcomes in astrocyte-to-neuron-like cell conversion. (A) Representative images of immunofluorescence staining of non-reactive astrocytes and LPS-induced reactive astrocytes after Ptbp1 knockdown by shPTB-1 for 4 weeks, with GFP (green), astrocyte-specific marker GFAP (indigo, Alexa Fluor 555), and the neuron-specific marker Tuj1(red, Alexa Fluor 647). Notably, after LPS induction, distinct from non-reactive astrocytes (yellow arrows), reactive astrocytes (yellow arrowheads) exhibited significant morphological changes, including enlarged cell bodies and web-like structures. While most cells displayed GFP fluorescence in the LPS group, only a small fraction stained positively for Tuj1, without exhibiting a typical neuronal morphology. Scale bar: 100 μm. (B) Quantification of Tuj1+ cells within GFP+ cells at 4 weeks following Ptbp1 knockdown by shPTB-1 in non-reactive astrocytes, compared to LPS-induced reactive astrocytes (two-tailed Student’s t-test). (C) Representative images of immunofluorescence staining of LPS-induced reactive astrocytes (LPS) and DEX-reversed reactive astrocytes (LPS + DEX), with Ptbp1 knockdown by siRNA-PTB-1 after 3 weeks, or Lucid control, showing Ptbp1 (green, Alexa Fluor 488) and neuron-specific marker Tuj1(red, Alexa Fluor 647). Scale bar: 50 μm. (D) Quantification of Tuj1+ cells within Ptbp1- cells at 3 weeks after transfection by siRNA-PTB-1 in DEX-reversed reactive astrocytes (LPS + DEX), compared to the LPS group (two-tailed Student’s t-test). All data are presented as means ± SD. The experiments were performed three times. ***P < 0.001. LPS: lipopolysaccharide; DEX: dexamethasone; ns: no significance; PTB: polypyrimidine tract-binding protein; Ptbp1: polypyrimidine tract-binding protein; GFP: green fluorescent protein; GFAP: glial fibrillary acidic protein; Tuj1: tubulin beta 3.

Full size image

To confirm whether the inflammatory state of astrocytes impedes astrocyte-to-neuron-like cell conversion, we conducted a reversal experiment by treating LPS-induced reactive astrocytes with DEX (LPS + DEX) prior to the addition of siRNA to initiate the conversion process. The morphology of astrocytes showed significant differences between those induced by LPS and LPS + DEX after 3 weeks of culture (Fig. 6C). No Ptbp1-Tuj1+ cells were observed in the Lucid astrocyte group following treatment with LPS or LPS + DEX (Fig. 6C), indicating that neither LPS nor DEX can induce astrocyte-to-neuron-like cell conversion in this study. After 3 weeks of Ptbp1 knockdown by siRNA, a few single Ptbp1-Tuj1+ cells were observed in the LPS group, but they lacked a typical neuronal morphology (Fig. 6C, D). In contrast, a significantly higher number of Ptbp1-Tuj1+ cells appeared and formed clusters in the LPS + DEX group, obviously more than that in non-reactive astrocytes (Fig. 6C, D). In summary, the potential for astrocyte-to-neuron-like cell conversion appears to be significantly influenced by the initial state of astrocytes. Specifically, the inflammatory reactive state induced by LPS impedes the capacity for successful trans-differentiation, while reversal with anti-inflammatory drugs can mitigate this inhibition, suggesting that modulating the inflammatory environment can improve the efficiency of the conversion process.

SAG facilitated the maturation and electrophysiological function of converted immature neuron-like cells

It is crucial to evaluate the electrophysiological properties of newly converted neuron-like cells to determine whether they have successfully acquired functionality. Since the neural conversion process appeared to occur in a relatively small subset of Ptbp1-knockdown astrocytes, we employed microelectrode array (MEA) technology to sensitively measure the overall electrical activity of the entire cell population. Before detection, we transduced the cells with siRNA on the MEA plate as previously described and confirmed that the cells had covered all electrodes (Fig. 7A). Given that smoothened agonist (SAG) has been reported to effectively facilitate the reprogramming process40, we further investigated its role in the Ptbp1 knockdown-induced astrocyte-to-neuron-like cell conversion. One week post-transfection, we detected several single neuron spikes in the group of astrocytes treated with siRNA-PTB-1 and SAG (siRNA-PTB-1 + SAG), but no neuron bursts were observed (Fig. 7B). Meanwhile, neither neuron spikes nor neuron bursts were detected in the groups treated with siRNA-PTB-1, siRNA-Ctrl, or siRNA-Ctrl + SAG (Fig. 7B). After two weeks of transfection, the number of neuron spikes in the siRNA-PTB-1 + SAG group increased significantly, much more than those observed in the first week (Fig. 7B–D). Simultaneously, neuron bursts began to occur frequently (Fig. 7C, E). The waveform of these spikes displayed multiple and consistent patterns, resembling those of neurons under physiological conditions (Fig. 7C). In contrast, no neuron spikes or bursts were detected in the other three groups throughout the experiment (Fig. 7B, D, E). These results ruled out the presence of electrically detectable neurons in the siRNA-PTB-1, siRNA-Ctrl, and siRNA-Ctrl + SAG groups. Furthermore, treatment with the sodium channel blocker TTX significantly inhibited the firing of neuron spikes in the siRNA-PTB-1 + SAG group after two weeks of conversion, eliminating any potential system artifacts and verifying the involvement of voltage-gated sodium channel in the generation of neuron spikes in the converted cells (Fig. 7F, G). In conclusion, the progressively strengthened electrical signal in the siRNA-PTB-1 + SAG group over time suggested that SAG can facilitate the newly converted cells into a more mature and electrically functional neuronal state. However, neither SAG alone nor Ptbp1 knockdown successfully induced functional neuron-like cells.

Fig. 7
figure 7

The Shh signaling pathway agonist SAG facilitated the electrophysiological function of converted neuron-like cells. (A) Representative phase-contrast image of experimental cells on a 64-electrode MEA plate. (B) Heatmap of electrical activity in the siRNA-PTB-1, siRNA-PTB-1 + SAG, siRNA-Ctrl, and siRNA-Ctrl + SAG groups after 1 week (top) and 2 weeks (bottom). Only the siRNA-PTB-1 + SAG group detected electrical activity, with increased signals observed at 2 weeks compared to 1 week. (C) Graphs depicting neuron spikes represented as electrode waveforms detected by a single electrode (top) in the siRNA-Ctrl + SAG group, along with neuron spikes within a 1000 ms timeframe (bottom). Five or more consecutive spikes occurring within an inter-spike interval of 100 ms were identified as a single neuron burst (red arrowhead). (D) Quantification of the total number of neuron spikes detected by all electrodes per minute at one and two weeks. The number of neuron spikes in the siRNA-PTB-1 + SAG group increased significantly over time. The detection was repeated three times. (E) Quantification of the total number of neuron bursts detected by all electrodes per minute at one and two weeks. A series of neuron bursts began to occur at 2 weeks. (F) Representative neuron spike raster plots in cells from the siRNA-PTB-1 + SAG group recorded at week 2, before and after TTX application. The red dashed box indicates the neuron spikes detected from the active electrodes (spike rate > 0.1 Hz). Following TTX treatment, neuron spikes were nearly absent and no active electrodes were detected. (G) Quantification of the weighted mean firing rate in the siRNA-PTB-1 + SAG group recorded at week 2 before and after applying TTX. The experiments were performed three times. All data are presented as means ± SD. n.d.: not detected; PTB: polypyrimidine tract-binding protein; Ptbp1: polypyrimidine tract-binding protein; MAP2: microtubule-associated protein 2; w: week; TTX: Tetrodotoxin.

Full size image

To verify whether Ptbp1 knockdown alone tends to induce the formation of immature neuron-like cells, we conducted immunofluorescence staining using the mature neuron-specific marker MAP2. The results revealed that almost no GFP+MAP2+ cells were observed across the whole field following Ptbp1 knockdown by shRNA for 4 weeks (Fig. 8A), or Ptbp1-MAP2+ cells following Ptbp1 knockdown by siRNA for 3 weeks (Fig. 8B). Furthermore, the motor neuron-specific marker ChAT was expressed in approximately 37% of Ptbp1-Tuj1+ cells (Fig. 8C, D), suggesting that although the converted neuron-like cells were not fully mature, they had adopted the cholinergic neuronal phenotype and were capable of synthesizing acetylcholine. Following the addition of SAG, a significant number of Ptbp1-MAP2+ cells emerged after Ptbp1 knockdown by siRNA (Fig. 8E, F). Additionally, the co-staining of Tuj1 and MAP2 revealed that, while Ptbp1-Tuj1+ cells lacked MAP2 expression after sole Ptbp1 knockdown, the addition of SAG facilitated MAP2 expression in Ptbp1-Tuj1+ cells (Fig. 8G). In summary, neuronal reprogramming induced solely by Ptbp1 knockdown tends to generate immature neuron-like cells lacking electrophysiological properties. However, the addition of the Shh signaling pathway agonist SAG significantly promoted the maturation and acquisition of electrophysiological functions in these immature neuron-like cells.

Fig. 8
figure 8

The Shh signaling pathway agonist SAG facilitated the MAP2 expression of immature converted neuron-like cells. (A) Representative images of immunofluorescence staining of astrocytes transduced by shPTB-1 and shPTB-2 after four weeks, showing GFP (green) and the mature neuron marker MAP2 (red, Alexa Fluor 647). Scale bar: 100 μm. (B) Representative images of immunofluorescence staining of astrocytes transfected by siRNA-PTB-1 and siRNA-PTB-2 after three weeks, showing Ptbp1 (green, Alexa Fluor 488) and the mature neuron marker MAP2 (red, Alexa Fluor 647). Scale bar: 100 μm. (C) Representative images of immunofluorescence staining of astrocytes transfected by siRNA-PTB-1 or siRNA-Ctrl after two weeks, showing Ptbp1 (green, Alexa Fluor 488), the neuronal marker Tuj1 (red, Alexa Fluor 647), and the motor neuron-specific marker ChAT (indigo, Alexa Fluor 555). Some Ptbp1-Tuj1+ cells stained positive for ChAT. Scale bar: 50 μm. (D) Quantification of ChAT+ cells within Tuj1+ cells at two weeks in the siRNA-PTB-1 group and siRNA-Ctrl group. (E) Representative images of immunofluorescence staining of astrocytes transfected by siRNA-PTB-1 after two weeks, with or without SAG treatment, showing Ptbp1 (green, Alexa Fluor 488) and the mature neuron marker MAP2 (red, Alexa Fluor 647). Scale bar: 100 μm. Several Ptbp1- cells were positive for MAP2 in the siRNA-PTB-1 + SAG group, whereas almost no Ptbp1-MAP2+ cells were detected in the siRNA-PTB-1 group. (F) Quantification of MAP2+ cells within Ptbp1- cells at two weeks in the siRNA-PTB-1 + SAG group, compared to the siRNA-PTB-1 group (two-tailed Student’s t-test). (G) Representative images of immunofluorescence staining of astrocytes transfected by siRNA-PTB-1 after two weeks, with or without SAG treatment, showing Ptbp1 (green, Alexa Fluor 488), the neuronal marker Tuj1 (red, Alexa Fluor 647), and the mature neuron marker MAP2 (indigo, Alexa Fluor 555). Scale bar: 100 μm. Many Ptbp1- cells co-expressed Tuj1 and MAP2 in the siRNA-PTB-1 + SAG group, whereas they were positive for Tuj1 but negative for MAP2 in the siRNA-PTB-1 group. The experiments were performed three times. All data are presented as means ± SD. ***P < 0.001. n.d.: not detected; Ptbp1: polypyrimidine tract-binding protein; Tuj1: tubulin beta 3; MAP2: microtubule-associated protein 2; ChAT: Choline Acetyltransferase.

Full size image

Discussion

This study adds a new dimension to the understanding of Ptbp1 knockdown in rat spinal cord astrocytes, using both siRNA and shRNA, and its role in driving the neuronal reprogramming process. Our findings reveal that the newly converted neuron-like cells predominantly exhibit an immature phenotype and require the addition of the Shh signaling pathway agonist SAG to promote their maturation and acquisition of functional electrophysiological properties. Furthermore, this study demonstrates the significant influence of the starting state of astrocytes on the efficiency and outcome of the conversion process. These findings highlight the potential of Ptbp1 knockdown as a strategy for functional astrocyte trans-differentiation and provide a fresh perspective on the heated debate surrounding Ptbp1-induced astrocyte trans-differentiation.

The controversy surrounding the role of Ptbp1 in astrocyte phenotypic trans-differentiation primarily arises from challenges such as “neuronal leakage” of AAV vectors, the inability to replicate results in gene knockout animal models, and the inability of lineage tracing and single-cell sequencing experiments to reliably verify the conversion process in in vivo studies28,29. However, there is a lack of in vitro studies specifically investigating the AtN conversion process through Ptbp1 knockdown alone. Unlike previous research on Ptbp1-induced AtN conversion, which often combined reprogramming agents with Ptbp1 knockdown to promote conversion15,27, our study established a relatively pure cellular environment, free from the influence of exogenous factors, to accurately and thoroughly validate the AtN conversion process induced solely by Ptbp1 downregulation from both phenotypic and functional perspectives. This study illustrates the cellular transformation process following the downregulation of Ptbp1 using either shRNA or siRNA, observing phenotypic changes and a gradual increase in conversion efficiency over time, which provides direct evidence for Ptbp1-mediated astrocyte trans-differentiation.

Distinct from previous studies, our findings provide new insights while also revealing specific aspects and limitations of Ptbp1 knockdown-induced reprogramming that require further refinement. Firstly, while we observed morphological changes and expression of neuronal markers, the exact identity of the converted cells remains incompletely defined and would require further validation. Secondly, the converted cells exhibit functional immaturity, as evidenced by their absence of robust electrophysiological activity and insufficient expression of mature neuronal markers, such as MAP2. Thirdly, we observed a relatively lower conversion efficiency (23%) after 4 weeks of conversion, compared to the reported conversion efficiencies of 58%27 and 50–80%15 in previous studies, which had combined SB431542 and CHIR99021 with Ptbp1 knockdown. Since these reprogramming drugs can significantly promote the trans-differentiation process and improve conversion efficiency33,35, the efficiency of neuronal reprogramming observed in this study may be closer to the actual efficiency of neuronal reprogramming induced by Ptbp1 knockdown.

Interestingly, this study also reflects a degree of “randomness” in the ability of Ptbp1-knockdown astrocytes to enter the neural conversion process, suggesting that the sole downregulation of Ptbp1 may not be sufficient to initiate neuronal reprogramming in all astrocytes. We further found that the initial state of astrocytes significantly affects the conversion outcome. Astrocytes can rapidly transition into the reactive state in inflammatory and injury models, which often express elevated levels of inhibitory factors and inflammatory markers that not only hinder the conversion process but also are toxic to the survival of neurons and oligodendrocytes20,21. The altered phenotypic, molecular, and functional states of reactive astrocytes can reduce their plasticity of trans-differentiation, as observed in our study. This may partly explain why some studies failed to observe AtN conversion when using reactive astrocytes in disease models as starting cells30,31. In line with our expectations, the reversal of the LPS-induced reactive state by using DEX successfully restored the capacity of astrocyte trans-differentiation. Given that reactive astrogliosis is a common response of astrocytes in neurodegenerative diseases, incorporating anti-inflammatory treatments to mitigate this reactive state could be needed to enhance future AtN conversion strategies20. Such interventions may help overcome the inhibitory effects of the reactive astrocyte phenotype, thereby increasing the likelihood of successful AtN conversion. Moreover, the intrinsic heterogeneity of astrocytes likely contributes to variability in reprogramming potential, as only specific subtypes may be more prone to neuronal conversion. As reported previously by scRNA-seq analysis, astrocytes exhibit significant heterogeneity, with evidence pointing to the existence of a specific subclass of astrocytes that may have a higher propensity for neuronal reprogramming compared to other subtypes32,41. The inflammatory state, heterogeneity, and lineage-traced modifications may all potentially contribute to a high barrier to conversion29.

Additionally, a complete AtN process involves two key stages, i.e., neural induction and neural maturation, which are regulated by the sequential knockdown of Ptbp1 and Ptbp2 and related splicing regulators18,42,43. During neural induction, astrocytes shift toward a neuron-like identity, marked by Ptbp1 downregulation, Ptbp2 de-repression, and miR-124 induction25,44,45. In the maturation phase, the cells undergo morphological and functional development, including axon and dendrite outgrowth, synapse formation, and electrophysiological property establishment18,42,46,47. This phase is featured by the simultaneous reduction of Ptbp1 and Ptbp2 and the induction of miR-918,25,44,45,48. Our study observed a marked increase in PTBP2 protein levels following siRNA-mediated Ptbp1 knockdown, indicating the initiation of neural induction during AtN conversion. While Ptbp1 knockdown successfully converts mouse fibroblasts into functional neurons, it only generates neuron-like cells that lack full maturity in human fibroblasts24,25,45, as the PTB-REST-miR-124 and nPTB-BRN2-miR-9 regulatory loops, which automatically interconnect in murine fibroblasts, require independent activation in human fibroblasts25. Our study extends the model to rat spinal cord astrocytes and determines that the newly converted cells are, in fact, immature neuron-like cells. Importantly, this study is also the first to identify the synergistic role of SAG in the Ptbp1 knockdown-induced neuronal reprogramming process. SAG promotes neuronal differentiation, axonal growth, and synaptic maturation, thereby enhancing the functional maturity of newly converted neurons by activating the Shh pathway49,50. When combined with SAG, the conversion is more specifically directed toward a neuronal fate, improving the electrophysiological properties and overall functionality of the converted neuron-like cells. These findings indicate the need for maturation-promoting interventions in Ptbp1-induced AtN conversion protocols to achieve fully functional neurons.

To address the current limitations of Ptbp1 knockdown-induced neuronal reprogramming, future research should utilize diverse neuronal markers, single-cell transcriptomics/proteomics, and electrophysiological techniques such as whole-cell and single-channel patch-clamps to further confirm the identity of converted cells. Future trans-differentiation studies should also focus on exploring effective approaches to enhance the conversion process and improve efficiency, such as the synergistic use of other reprogramming methods and the identification of the most suitable cell states and subtypes of starting astrocytes. Furthermore, in vivo studies are necessary to validate these findings and evaluate the long-term survival, functional integration, and therapeutic potential of reprogrammed cells in models of spinal cord injury.

In summary, we found that Ptbp1 knockdown via shRNA and siRNA can both induce astrocyte-to-neuron-like cell conversion in rat spinal cord astrocytes in vitro, though it is insufficient to generate fully mature, functional neurons. The conversion efficiency of neuron-like cells gradually increases over time and then stabilizes, with the outcome being highly dependent on the initial state of the astrocytes. When combined with SAG, the newly converted neuron-like cells exhibit neuronal maturation in both phenotype and function. This study suggests that future research should focus on strategies to modulating astrocyte reactivity and optimize the initial state of astrocytes, making them more conducive to conversion. Overall, the combination of SAG and Ptbp1 knockdown holds great promise as a strategy for restoring lost neurons and promoting functional recovery after spinal cord injury.

Materials and methods

Primary culture of rat spinal astrocytes, activation, and reversal of activation

Primary rat spinal cord astrocytes were purchased from Procell (Wuhan, China, Cat# CP-R306) and obtained from neonatal Sprague Dawley rats. The astrocytes were harvested by using the shear force generated by horizontal shaking in an incubator at 280 rpm for 16 h and further purified using the protocol of Zarei et al.37. They were cultured in complete culture medium (Procell, Cat# CM-R306) formulated for rat spinal cord astrocytes, with medium changes performed every other day. These cells were passaged five times and confirmed by immunofluorescence staining before being used for further experiments.

To induce reactive astrocytes, which are abundant in spinal cord injury, lipopolysaccharide (LPS, Sigma-Aldrich, St. Louis, MO, USA, Cat# L2880) was added to the culture medium for 24 h at a concentration of 500 ng/mL21,51. To reverse the reactive state of astrocytes, following 24 h of LPS induction, the culture medium was refreshed, and 100 nM dexamethasone (DEX, MedChemExpress, Monmouth Junction, NJ, USA, Cat# HY-B1829A) was immediately added for an additional 24 h before subsequent experiments39.

RNA interference and viral vectors

Lentiviral shRNA targeting rat Ptbp1 (shPTB) was produced by Genmedicn Biopharma (Nanjing, China) using two specific sequences: shPTB-1 (5ʹ-GCG GGT GAA GAT CCT GTT CAA-3ʹ) and shPTB-2 (5ʹ-GCA CAG TCC TGA AGA TCA TCA-3ʹ). These sequences were subcloned into the pSF-SFFV-EGFP-MIR30a-shRNA-WPRE vector, with a specific non-targeting shRNA sequence (5ʹ-CAA CAA GAT GAA GAG CAC CAA-3ʹ) serving as the control group (shCtrl). The lentiviral product’s ultimate titers were shPTB-1 (pSF-SFFV-EGFP-MIR30a-shRNA (PTB-1)-WPRE, 1.46 × 109 TU/mL), shPTB-2 (pSF-SFFV-EGFP-MIR30a-shRNA (PTB-2)-WPRE, 2.15 × 109 TU/mL), and shCtrl (pSF-SFFV-EGFP-MIR30a-shRNA (NC)-WPRE, 1.90 × 109 transduction units (TU)/mL).

The shPTB-1 and shPTB-2 sequences were also utilized to separately construct two siRNA, siRNA-PTB-1 and siRNA-PTB-2, which were produced by Ruibiotech Co., Ltd. (Beijing, China). The siRNA-Ctrl was also used with the same vector backbone as the shCtrl to serve as a negative control.

Non-reactive, reactive, and reverse-reactive astrocyte trans-differentiation in vitro

To initiate astrocyte trans-differentiation in vitro, 12 mm glass coverslips were placed in 6-well culture plates and coated with PLL (poly-L-lysine, Procell, Cat# PB180523), with three coverslips per well. Primary non-reactive, reactive, and reverse-reactive spinal astrocytes were separately seeded onto the coated wells at a density of 40,000 cells per well in astrocyte complete medium supplemented with F108 (Sigma-Aldrich, St. Louis, MO, USA, Cat# 07579) and lentiviral vectors at the multiplicity of infection (MOI) = 20. Following 24 h of transfection or transduction, the culture medium was switched to a neural induction medium composed of 1:1 DMEM/F12 (Gibco, Grand Island, NY, USA, Cat# 12634028) and Neurobasal Plus medium (Gibco, Cat# A3582901), supplemented with 1% Glutamax (Gibco, Cat# 35050061), 1% N-2 (Gibco, Cat# 17502048), 2% B-27 (Gibco, Cat# A3582801), 1% penicillin/streptomycin (Gibco, Cat# 15140122), and 10 ng/mL each of GDNF (Stemcell, Vancouver, BC, Canada, Cat# 78058.1), BDNF (Stemcell, Cat# 78005.1), and neurotrophin 3 (Stemcell, Cat# 78074.1). The medium was half-changed every 3 days. Following a 48-hour transduction with GFP-lentivirus, transduction efficiency was quantified by flow cytometry, gating GFP-positive cells based on the blank control before analyzing the experimental groups.

For siRNA transfection, primary non-reactive, reactive, and reverse-reactive spinal cord astrocytes were seeded under the same conditions as described above (40,000 cells each well). After the cells had attained 70%–80% confluency, they were transduced with siRNA-PTB or siRNA-Ctrl (100 pmol/well), Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA, USA, Cat# 13778150), and Opti-MEM (Gibco, Cat# 31985070). After 8 h of transfection, the transfection medium was switched to neural induction medium, with the medium being half-replaced every 3 days thereafter. An additional 250 nM SAG (MedChemExpress, Cat# HY-12848B) was added into the neural induction medium in the neural function test to enable experimental comparison for neural maturation.

Live cell imaging of astrocytes following Ptbp1 knockdown

Morphological changes in astrocytes following Ptbp1 knockdown were observed using non-reactive astrocytes transduced with lentivirus. The cells were cultured on PLL-coated 6-well plates and transduced according to a previously reported protocol. The following day, neural induction medium was added to refresh the previous medium, and the plates were moved to the IncuCyte S3 live-cell system (Sartorius, Göttingen, Germany). The phase-contrast images of the cells were captured for 4 days with snapshots taken every 24 h for live cell imaging. After 4 weeks, the cells were further analyzed under the inverted microscope Leica DMi8 in both phase-contrast and fluorescence modes (Leica, Wetzlar, Germany).

Western blot analysis and quantitative reverse transcription-polymerase chain reaction

Protein samples were collected 48 h after transduction with shRNA or transfection with siRNA, by lysing cells in RIPA lysis buffer (Beyotime, Cat# P0013B) supplemented with 0.1 mM PMSF and a protease/phosphatase inhibitor cocktail (Beyotime, Cat# P1045). After protein quantification, proteins from each group of converted cells were loaded in equal amounts and resolved using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were subsequently transferred onto a polyvinylidene difluoride (PVDF) membrane and blocked with 5% BSA for 1 h. The membrane was incubated at 4 °C overnight with the primary antibodies: rabbit anti-Ptbp1 (1:1000; cell signaling technology, Danvers, MA, USA, Cat# 57246S, RRID: AB_2799528), rabbit anti-Ptbp2 (1:1000; CST, Cat# 15719S), and mouse Anti-β actin (1:3000; ZSGB-Bio, Cat# TA-09, RRID:AB_2636897), followed by incubation with the second antibody for 1 h at room temperature. Protein bands were visualized using an ECL kit (Thermo Fisher Scientific, Waltham, MA, USA, Cat# 35050). Signals were developed using the Tanon 5200 chemiluminescence imaging analysis system (Tanon, Shanghai, China).

Reprogrammed cells were lysed using Trizol reagent (Thermo Fisher Scientific, Cat# 15596018CN) to extract total RNA, which was then reverse-transcribed into cDNA with the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China, Cat# R312-02). Taq SYBR® Green qPCR Premix (Vazyme, Cat# Q712-03) was used to perform quantitative RT-PCR (RT-qPCR) on the multi-block PCR thermal cycler (Longgene, Hangzhou, China) following the manufacturer’s instructions. Relative gene expression levels were normalized to GAPDH expression using the 2–ΔΔCt method52. Primer sequences are listed in Table 1.

Table 1 Primer sequences utilized for quantitative reverse transcription polymerase chain reaction.
Full size table

Immunofluorescence staining

Reprogrammed cells were harvested for immunocytochemistry at 1–4 weeks following shRNA-mediated Ptbp1 knockdown, as well as 1 and 2 weeks following siRNA-mediated Ptbp1 knockdown. The cells were fixed with 4% paraformaldehyde for 10 min at room temperature and washed three times with PBS for 5 min each. Then, the cells were permeabilized with PBS containing 0.2% Triton X-100 for 5 min, followed by three washes of PBS of 5 min each. Blocking was conducted with 3% BSA in PBS for 30 min, after which the cells were incubated with primary antibodies in PBS with 3% BSA overnight at 4 °C. The primary antibodies are listed in Table 2.

Table 2 The primary antibodies used for immunofluorescence staining.
Full size table

The cells were then washed three times with PBS for 10 min each and incubated at room temperature with the appropriate secondary antibodies diluted in PBS for 1 h. The secondary antibodies included Alexa Fluor 488 goat anti-mouse IgG (1:1000; Abcam, Cat# ab150117, RRID: AB_2688012), Alexa Fluor 555 goat anti-mouse IgG (1:1000; Abcam, Cat# ab150114, RRID: AB_2687594), Alexa Fluor 555 goat anti-chicken IgG (1:1000; Abcam, Cat# ab150174, RRID: AB_2864276), Alexa Fluor 555 goat anti-sheep IgG (1:1000; Abcam, Cat# ab150178), and Alexa Fluor 647 goat anti-rabbit IgG (1:1000; Abcam, Cat# ab150083, RRID: AB_2714032). After washing the cells three times with PBS, DAPI (Beyotime, Shanghai, China, Cat# C1002) was used to stain their nuclei, and the Leica stellaris 5 confocal microscope (Leica, Wetzlar, Germany) was utilized to acquire images.

Microelectrode array (MEA) recording

Neuronal spike and burst activity were recorded using the Maestro Pro MEA system, powered by AxIS Navigator software 3.11.1 (Axion Biosystems, Atlanta, GA, USA). Each cell group was cultured separately on CytoView six-well MEA plates (Axion Biosystems), pre-coated with PLL as previously described. Each well included 64 electrodes arranged in an 8 × 8 grid, with a recording area of 2.1 mm × 2.1 mm. The cells were transduced with siRNA, cultured in neural induction medium with or without the addition of Shh signaling pathway agonist SAG, and checked under a microscope to confirm that they were fully covering the microelectrodes. Spontaneous neuron activity recordings were conducted after one and two weeks of induction for 5 min at 37 °C and 5% CO2. Following the baseline recording of neuron activity after two weeks of conversion, 1 nM TTX (Must Bio-Technology, Chengdu, China, Cat# A0224) was added into the wells and incubated for 5 min before recording the neuron activity for an additional 5 min.

The AxIS Navigator software was utilized for data collection and analysis. For the identification of spikes by spike detection, the adaptive threshold crossing method was used, with a threshold set at 6 × standard deviation and a hold-off of 2.16 ms. Single burst detection identified bursts containing a minimum of 5 spikes, with a maximum inter-spike interval of 100 ms. An active electrode was defined as having a spike rate greater than 0.1 Hz (6 spikes per minute). Spike raster plots were generated using the Neural Metric Tool software (Axion Biosystems).

Statistical analysis

The data were all presented as the mean ± standard deviation (SD). Prior to statistical testing, normality was assessed using the Shapiro–Wilk test, and homogeneity of variance was evaluated using Levene’s test. Only data that met these assumptions were analyzed using parametric tests. For comparisons between two groups, unpaired two-tailed Student’s t-tests were performed. For comparisons involving more than two groups, one-way ANOVA was used followed by Bonferroni post hoc test to assess statistical significance between individual group pairs. The software GraphPad Prism 9.0.0 (GraphPad Software, Boston, MA, USA) was used to conduct statistical analyses and data visualization. A P value of less than 0.05 was deemed statistically significant.