Generation of Functional Dopaminergic Neurons from Reprogramming Fibroblasts by Nonviral-based Mesoporous Silica Nanoparticles

Direct-lineage conversion of the somatic cell by reprogramming, in which mature cells were fully converted into a variety of other cell types bypassing an intermediate pluripotent state, is a promising regenerative medicine approach. Due to the risk of tumorigenesis by viral methods, a non-viral carrier for the delivery of reprogramming factors is very desirable. This study utilized the mesoporous silica nanoparticles (MSNs) as a non-viral delivery system for transduction of the three key factors to achieve conversion of mouse fibroblasts (MFs) into functional dopaminergic neuron-like cells (denoted as fDA-neurons). At the same time, a neurogenesis inducer, ISX-9, was co-delivered with the MSNs to promote the direct conversion of neuron-like cells. Good transfection efficiency of plasmid@MSN allowed repeated dosing to maintain high exogenous gene expression analyzed by qPCR and the changes in neural function markers were monitored. To further validate the dopaminergic function and the electrophysiological properties of fDA-neurons, the results of ELISA assay showed the high levels of secreted-dopamine in the conditional medium and rich Na+/K+-channels were observed in the fDA-neurons on Day 22. The results demonstrated that MSN nanocarrier is effective in delivering the reprogramming factors for the conversion of functional dopaminergic neurons from adult somatic cells.


Results and Discussion
Characterizations of mesoporous silica nanoparticles. Mesoporous silica nanoparticles (MSNs) have been used extensively in nanomedicine, because of its good biocompatibility, high surface area (>1000 m 2 /g), tunable pore size and easiness surface functionalization. In this study, they were synthesized by our reported method 21 and functionalized with the amine group (APTMS) to give MSN-NH 2 to confer positive charge and then with biotin to give gated MSN-biotin. To develop a stimulation-response non-viral nanoparticle for loading and delivering ISX-9, the pores of MSN-biotin were capped by avidin with the strong biotin-avidin interaction to give MSN-avi. The three key differentiation factors, Ascl1, Brn2, and Myt1l, were adsorbed on the avidin-capped MSNs (abbreviated as MSN-avi) separately which was then delivered into the MFs at the same time. (Fig. 1) The average diameters of the nanoparticles, 3-aminopropyltrimethoxysilane (APTMS) modified MSNs (abbreviated as MSN-NH 2 ) and MSN-avi, are 80~120 nm with the hexagonally arranged pore structure (TEM shown in Fig. 2a). The average sizes of MSN-NH 2 and MSN-avi in aqueous solution are examined by dynamic light scattering (DLS) analysis. The average hydrodynamic diameters were determined by DLS (Fig. 2b), giving MSN-NH 2 (122.8 nm, zeta = 30.4 mV) and MSN-avi (383.5 nm, zeta = −16.8 mV). They were well-dispersed in water. The hydrodynamical sizes and the zeta potential analysis of the nanoparticles are listed in Table S2. The weight percentage of the biomolecules (-NH 2 -biotin-NHS and the avidin protein) modified on the mesoporous silica nanoparticles were 8.02% and 11.45%, respectively as determined by thermogravimetric analysis (TGA) (Fig. 2c). In addition, the -NH 2 -biotin and -iotin-avidin modification on the MSNs was examined by the FTIR absorption band at 1541 and 1557 − cm 1 (Fig. 2d,e). In testing the drug release, we used a small dye molecule rhodamine 800 (denoted as Rh800) as a probe to be released from MSN-avi during biotin-responsive drug release process (Fig. 2f). The Rh800 was loaded in the pores of MSN-NH2 and the biotin is added; after 1 hr they were capped with avidin. Figure 2f demonstrates that the dye Rh800 can be released upon addition of biotin competing for the avidin caps. From the release profiles, one can see that after an initial quick release (probably from the externally adsorbed Rh 800), there is little release of the Rh 800 afterwards. With uncapped MSN-biotin (no avidin), there is further release after 30 min to double the amount of total release. With the capped MSN-avi, there is a sudden release after triggering with added biotin (competing for avidin) and the total release is even higher than the uncapped MSN-biotin. The intracellular drug release could be triggered by the endogenous biotin molecule, for inducing the dissociation of avidin from MSN-avi 25 . The dissociation constant (K d ) of avidin toward biotin molecule (~1 × 10 −15 M) was lower than that of avidin and biotin-modified MSN (~3.5 × 10 −11 M), which indicates that the endogenous biotin molecule could compete as an uncapping reagent 26 .
With the capping capability of MSN-avi checked, the neurogenesis drug ISX-9 was encapsulated in the MSN-avi, and the loading capacity of ISX-9 was determined to be 93.7 µmol/g MSN-avi, as analyzed by UV-vis spectrum (Table S3). We then used MSN-avi as a non-viral carrier for co-delivery of the three plasmid DNA, Ascl1, Brn2, and Myt1l, to achieve direct conversion of MFs into functional dopaminergic neuron-like cells. Herein, the pABM-I@M-induced functional dopaminergic neuron-like cell was denoted as iM-Neurons.

Mesoporous silica nanoparticles uptaken by MFs. We first study the interaction between nanoparticles
and MFs by examining the cellular uptake and the biocompatibility of nanoparticles. The MFs were incubated with 256 µg of the fluorescein isothiocyanate-labeled MSN-avi (FMSN-avi) for 1, 2, and 4 hrs. Flow cytometry data showed that the cellular uptake efficiency was 95.2% after 1 hr incubation (see Supplementary Fig. S2). The result indicated that FMSN-avi entered MF cells efficiently. Confocal microscopy analysis also showed that FMSN-avi was uptaken by MFs (see Supplementary Fig. S3) copiously. In Fig. S3, we also observed that some of the FMSN-avi (green channel) and the early endosome (red channel) could be co-localized in the merged image and some of the FMSN-avi have already escaped from the endosome and resided in the cytoplasm at 1hr (green dots in the merge channel). The data indicated that most of the FMSN-avi enter the cells through endocytosis mechanism.
Mesoporous silica nanoparticles are nontoxic to iM-Neurons. To evaluate the in vitro cytotoxicity, we compared cell proliferation after loadings of the nanoparticles pABM-I@M. For comparison, we also examine the same plasmid factors loaded with Lipofectamine 2000 (pABM-Lipo); the cell viability was determined by counting the cell number. The cells were incubated with 256 µg of the pABM-I@M and or 6 μL/mL Lipofectamine 2000 for 4 hrs and then washed with PBS twice. The same amount of plasmid and ISX-9 was used in Lipofectamine 2000 and the mesoporous silica nanoparticle. The results showed that there was no significant reduction in cell viability and long-term cellular proliferations with MSN as carriers (pABM-I@M). However, Lipofectamine 2000-mediated transfection of pABM-Lipo resulted however in 72~80% of cell death after 7 days. The results indicated that mesoporous silica is a biocompatible nanocarrier compared to Lipofectamine 2000 for long-term iM-Neurons formation (see Supplementary Fig. S4). mRNA expression levels of neuron-related factors. An equal amount of 2.0 µg pABM neural reprogramming factors in each of pABM-I@M (e.g. pA-I@M etc.) were delivered into MFs. Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) was used to examine the mRNA expression level of neuron-related genes enhanced by transfection of pABM-I@M. In Fig. 3, we show that the triple delivery of pAscl1, pBrn2, and pMyt1l was successfully transfected with I@M on Day 3, Day 7, and Day 11 and the dose-dependent and time-dependent effect of the three-corresponding mRNA were observed (Fig. 3d). As the time increased, the expression level of all the three-corresponding mRNAs also increased. On day 14 after the treatment of pABM-I@M, the qPCR analysis also shows the expression of neural function markers, Dat (Dopamine transporter, a symporter that moves dopamine across the cell membrane), Ngn2 (Neurogenin 2, played an important role in neurogenesis) 12 , Vmat (Vesicular monoamine transporter, acted to transport monoamine neurotransmitters), Map2 (Microtubule-associated protein 2, neuron-specific cytoskeletal proteins) and Shh (Sonic hedgehog, attracted commissural axons at the ventral midline of the developing spinal cord). Figure 3e shows the expression of Dat was 3.5-fold compared to control and the expression of Vmat was 2.2-fold compared to control after serial transfection of pABM-I@M on Day 14. However, the results for the Lipofectamine 2000 nanocarrier; the expression of Dat was only 1.9-fold compared to control, and the expression of Vmat was 1.3-fold compared to control after serial transfection of pABM-Lipo. The expressions of the other neuronal lineage markers, Ngn2, Map2, and Shh, were also shown that the serial transfection of pABM-I@M gave higher expressions of the mRNAs than that of Lipofectamine 2000 carrier. Thus, our MSNs carrier is more effective in promoting the gene transfection than using Lipofectamine 2000. Our data further supported that MSN-based nanocarrier system may provide a direct-conversion platform for adult somatic cells reprogramming into neurons.
Significant cell morphology conversion. Significant changes in cell morphology, such as the neural cell body, axon, and dendrite, are common signatures of neurons which were observed by microscope after iM-Neurons formation (Fig. 4a). As can be seen in Fig. 4, the more days pABM-I@M cells were treated the more the neuron-outgrowth morphology was. In comparison, for the pABM-MSN treated and the positive control condition (pABM-Lipo + 10 µM ISX−9), there are little neurite outgrowths. Accordingly, the neurogenesis drug, ISX-9, loaded in MSN-avi, promotes transfection efficiency of pABM and enhanced the maturation of iM-Neurons. One can observe substantial neurite outgrowths.

The neural reprogramming factors promotes Tuj1+ and the expression of the neural lineage markers.
To determine whether the neural progenitor marker of iM-Neurons was induced by pABM-I@M, immunofluorescence (IF) staining was used to determine the protein-level of neural progenitor marker. The IF staining of neuron-specific class III beta-tubulin (Tuj1), expressed in the neuronal cell bodies and processes, was observed and is shown in Fig. 5a for treatment with pABM-I@M on Day 3, 7, 14 and 19. The quantification of the number of Tuj1-expressing cells was normalized to the number of cells which were counter-stained with 4′,6-diamidino-2-phenylindole (DAPI, blue) as shown in Fig. 5b. The data showed that the increase of Tuj1-expressing cells was dominating 19 days after iM-Neurons formation; 81% of MFs were directly converted into neural progenitor cells.
In order to examine the maturation of iM-Neurons, other neural lineage markers were also determined ( Fig. 6). IF staining of Tau (proteins that stabilize microtubules), the protein marker expressed in post-mitotic stage of neuron: NeuN, dopaminergic neuron marker: Th (Tyrosine hydroxylase, a dopamine-related enzyme), GABAergic neuron marker: GABA, Glutamergic neuron marker: VGluT2, and the proteins of synaptic vesicles: synapsin I, were shown in Fig. 6. The synapsin I-positive puncta adjacent to dendrites of iM-Neurons was marked with white arrows. The qPCR data and IF staining analysis showed that the MFs was direct converted into various neuron-like cells by pABM-I@M. The MSNs are shown to be a good non-viral transfection reagent for multiple plasmid delivery.

Enhanced expression of dopamine transporter and maturation of functional neurons.
To further examine whether the iM-Neurons present the characteristics of dopaminergic neurons, we monitored that the dopamine release ability from the iM-Neurons, which was analyzed by dopamine enzyme-linked immunosorbent assay kit (ELISA, LDN, Nordhorn, BA E-5300). One observes the neurotransmitter dopamine was released from neurons after iM-Neurons formation. In Fig. 7a, one observed increasing IF staining of the dopaminergic neuron marker on 7th, 14th and 22th days. The normalized percentage of Dat-expressing cells was 77% after 19 days iM-Neurons formation (Fig. 7b).
The dopamine was determined by a cis-diol-specific affinity gel with which the dopamine was acylated to N-acyldopamine and converted into N-acyl-3-methoxytyramine enzymatically during the detection. To detect dopamine, the antigen bound to the surface of the microtitor plate was examined by the anti-rabbit IgG-peroxidase conjugate and used 3,3,5,5′-Tetramethylbenzidine (TMB) as a substrate. In Fig. 7c, the average amount of dopamine secretion released from dopaminergic neuron cells was determined to be 5.0 ng/mL after To examine whether the iM-Neurons perform signal transduction between cells, we perform a whole-cell patch-clamp assay to measure the electrophysiological properties within iM-Neurons. First, ovoid-shape iM-Neurons were applied to patch-clamp assay (Fig. 8a), and then we measure its putative process upon electric stimulation. The voltage response was observed in the recorded cells during the positive current injection (500 pA) (Fig. 8b). The average resting membrane potential (RMP) was −47.5 ± 8 mV (n = 16) and input resistance was 131 ± 56.85 MΩ (Fig. 8c,d). Besides, to investigate the properties of voltage-gated ion channels in the iM-Neurons, we gave the voltage step from 50 mV to −80 mV (Fig. 8e). Next, we added the voltage-gated Na + -channel blocker, tetrodotoxin (TTX) and the voltage-gated K + -channel blocker, 4-aminopyridine (4-AP) to confirm the signature of voltage-gated ion channels on iM-Neurons (Fig. 8f). These results indicated that there are voltage-gated Na + /K + -channels exist in the membrane of iM-Neurons, and MFs were directly converted into functional dopaminergic neuron-like cells induced by pABM-I@M 27 . In summary, these results indicated that iM-Neuron possess rich Na + /K + -channels and carried out electrophysiological transduction. We first demonstrated that MSN could be used to produce functional dopaminergic neuron-like cells from somatic cells.

Conclusion
In summary, a non-viral delivery system of drug and plasmid for direct conversion of MFs into functional dopaminergic neuron-like cells was reported. Herein, we used mesoporous silica nanoparticles for the transduction of three factors, Ascl1, Brn2, Myt1l, into MFs together with the delivery of neurogenesis drug, ISX-9 to enhance the conversion of MFs-to-neuron. Biotin was functionalized on the of surface MSNs, and the avidin protein was used to encapsulate the ISX-9 into the MSNs (denoted as I@M). The drug-release was triggered by endogenous biotin molecule which could prolong and promote the efficacy of ISX-9. Immunofluorescence staining, ELISA assay, and the voltage-clamp recording analysis showed that the MFs were efficiently converting and reprogramming by pABM-I@M into functional dopaminergic neuron-like cells. The maturation of neurons was demonstrated by detection of dopamine production, rich Na + /K + -channels observed by the voltage-clamp study and a good population of neuron dendritic morphology. In addition, the low cytotoxicity and efficient delivery of the plasmids of I@M were demonstrated, and MFs-neuron could survive longer than cells treated by Lipofectamine 2000. We have shown that MSNs can act as a good biocompatible and non-viral carrier for transfection, and it could be applied to direct conversion of other specific lineages. Because of the targeting ability, nontoxicity and non-viral nature of MSNs, further studies on animals may be worthwhile. In the long-term, direct conversion approach using a safe non-viral nanocarrier may provide safe regenerative therapies for neurodegenerative disorders, including the Parkinsonian's disease.   The release test was performed using a dialysis bag (Mw = 10 kDa). The released Rh800 could across the dialysis membrane but MSN-avi could not. The solution outside the dialysis membrane was collected every 30 minutes, and after 60 minutes, 0.6 mg of biotin was added into the solution. The released Rh800 from MSN-avi to the solution was examined by fluorescence emission at 710 nm. pABM-I@M Adsorption and Agarose Gel Electrophoresis. 0.33 μg of pAscl1 was mixed with several weight ratios of I@M: 1/1, 1/2, 1/4, 1/8, 1/16, 1/32, 1/64, 1/128 in Dulbecco's modified Eagle medium (DMEM) and incubated for 30 minutes, and the obtained conjugate is pA-I@M. Similar procedures were performed to load Brn2 and Myt1l to obtain pB-I@M and pM-I@M. And a final mixture of pA-I@M, pB-I@M and pM-I@M in the equal ratio is called pABM-I@M for the delivery of the three factors.

Synthesis of APTMS Conjugated Mesoporous Silica Nanoparticles, MSN-NH
After mixing of the pA-I@M, pB-I@M and pM-I@M, pABM-I@M complexes were loaded in 1% agarose gel. After gel electrophoresis (at 110 V for 30 minutes), DNA bands were separated and visualized by ethidium bromide staining. The size standards, pAscl1, pBrn2, pMyt1l, and I@M were used as references.
Flow cytometry analysis. The uptake of MSN-avi by MF was examined by flow cytometry as follows. MFs were seeded at 2 × 10 5 cells per well in six-well plates and allowed to attach for 24 h. To determine the amount of FITC-conjugated MSN-avi (FMSN-avi) uptake by MFs, the cells were incubated with 256 μg/mL of FMSN-avi suspension in serum-free medium for 1, 2 and 4 h. Treated cells were then washed twice with phosphate-buffered saline (PBS), and then harvested by trypsinization. The green emitting fluorescein isothiocyanate dye in FMSN-avi serves as a marker to quantitatively determine their cellular uptake. In addition, trypan blue (ratio of volume, PBS:trypan blue = 6:1) was utilized to quench the FMSN that adsorbed on the membrane of MFs. FM4-64 staining of endosome after pABM-I@M delivery. The MFs were seed at a density of 1.5 × 10 5 cells per cover slide well. After treatment of 256 µg of FMSN-avi for 1 hr, the cells were washed with PBS. Then, the cells were fixed in 1% paraformaldehyde and permeabilized with 0.1% NP-40. Finally, the nucleus was stained by 4′,6-diamidino-2-phenylindole (DAPI), and observed by confocal microscopy.

MFs Transfection and generation of iM-Neurons.
Before gene delivery, MFs were trypsinized into single cells and cultured in six-well plates. Next, delivery of 256 μg pABM-MSN and pABM-I@M (weight ratio: pABM/MSN = 1/128) complexes into1 mL DMEM was carried out. Lipofectamine 2000 in DMEM (Invitrogen) was used as a positive control. Before delivery, 0.67 μg of plasmid (pAscl or pBrn2 or pMyt1l) was mixed with 85.33 μg nanoparticle (MSN or I@M) in DMEM for 3 hrs, and the plasmid-nanoparticle complexes were formed. Finally, the pAscl1-MSN (or pAscl1-I@M), pBrn2-MSN (or pBrn2-I@M) and pMyt1l-MSN (or pMyt1l-I@M) were mixed together. The solution was added to six-well plates and the incubation time was 4 hrs. After washing with PBS, MFs were cultured in DMEM supplemented with 10% FBS, 1% of penicillin-streptomycin at 37 °C. Then 2 μg/mL of doxycycline were added into the culture medium after transfection for 12 h to induce the expression of the plasmids. After five-doses transfection for 11 days, the culture medium was changed to neuronal induction medium at Day 12. The neuronal induction medium contained Neurobasal Medium, N2, and B27 supplements, GlutaMAX, 1% of penicillin-streptomycin (all from Life technologies), and bFGF (100 ng/mL). To enhance the survival rate and improve the efficiency of direct conversion, the ROCK inhibitors Y27632 (2 μM), the P38 MARK inhibitors SB203580 (1 μM) and the BET family bromodomain inhibitors I-BET 151, (1 μM), sonic hedgehog (Shh, 500 ng/mL) and FGF8 (100ng/mL) were used. The medium (neuronal induction medium plus the small molecules) was refreshed every 3-4 days during the induction period.
Quantitative Reverse Transcriptase Polymerase Chain Reaction. Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) was used to determine the messenger ribonucleic acid (mRNA) expression of cells after gene delivery. 1 mL of Trizol reagent (Invitrogen) was used to homogenize the cell lysate and the sample was left standing for 5 minutes at room temperature. Then, 200 μL chloroform was added into the mixture and centrifuged (13200 g, 5 minutes at 4 °C) to separate the phases. The transparent aqueous phase was transferred to a new tube and mixed with 500 μL isopropyl alcohol (Sigma-Aldrich, St. Louis, MO, USA). After the solution was vortex vigorously, RNA was collected by centrifugation (13500 g, 10 minutes at 4 °C). Then, the precipitate was washed with 1 mL of 75% ethanol. Next, the ethanol was removed by centrifugation, and the pellet was dissolved in 24 μL diethylpyrocarbonate-treated H 2 O (Life technologies, Carlsbad, CA, USA). The RNA was quantized by UltroSpec 3100 Pro (Amersham), and 1 μg of RNA was reversely transcribed with a SuperScript III reverse transcriptase kit (Invitrogen). qRT-PCR was performed in real time using an ABI 7900 Fast System and SYBR Green Master Mix. The real-time PCR cycling condition used was: 94 °C for 3 minutes, 35 thermal cycles (denaturing the DNA at 94 °C for 30 seconds, annealing with primers at 60 °C for 5 seconds, extending the length of the product at 72 °C for 1 minute), and a final extension of the product at 72 °C for 10 minutes. The reactions were carried out by specific primers and shown in Table S1. Moreover, glyceraldehyde3-phosphate dehydrogenase (Gapdh) was used to normalize the raw data of qPCR, and the error bar indicated the standard deviation of triplicate measurements.

Measurement of dopamine release by enzyme-linked immunosorbent assay (ELISA).
The supernatant liquid of iM-Neurons was collected at Day 7 th , 14 th , and 22 th . The following description was the experimental process.
Extraction and acrylation. Dopamine standards, controls and samples (750 μL/well) was added into coated plates, followed by 25μL TE buffer, and then shake for 60 minutes. After washed with wash buffer, 150 μL acrylation buffer and 25 μL acrylation reagent. After 20 minutes, it was washed with wash buffer again. Then 100 μL of hydrochloric acid was added into the plate and incubated for 10 minutes.
Enzymatic conversion. 90 μL of standards, controls and samples were added into the microtitor plate, and 25 μμL of enzyme solution was added. Then, the microtitor plate was incubated at 37 °C for 2 hrs.
Dopamine ELISA. 100 μL of standards, controls, and samples were added into dopamine microtitor plate, followed by dopamine antiserum and incubated at 2-8 °C for overnight (15-20 hrs). Then, it was washed with wash buffer and in which 100 μL of enzyme conjugate solution was added. Finally, 100 μL of substrate liquid was added for 20-30 minutes, and then 100μL stop solution was pipetted into each well. The absorbance of the solution in the wells was measured by a microplate reader at 450 nm within 10 minutes and the reference wavenumber was set between 620-650 nm. MFs-neuron cells were placed in a submerged chamber with artificial cerebrospinal fluid (ACSF) containing the following (in mM): 125 NaCl, 25 NaHCO 3 , 1.25 NaH 2 PO 4 , 2.5 KCl, 25 glucose, 2 CaCl 2 , and 1 MgCl 2 . For the whole-cell patch-clamp recordings, recording electrodes (4-8 MΩ) were pulled from borosilicate glass tubing (outer diameter, 1.5 mm; inner diameter, 0.86 mm; Harvard Apparatus) filled with low Cl − internal solution, containing the following (in mM): 136.8 K-gluconate, 7.2 KCl, 0.2 EGTA, 4 MgATP, 10 HEPES, 7 Na 2 -phosphocreatine, 0.5 Na 3 GTP (pH 7.3 with KOH). Individual cells were selected for the recordings based on a small round or ovoid cell body (diameters, 5-10 μm). The voltage-clamp recording was measured at a holding potential of −70 mV. The current-clamp was performed at −70 mV by using a holding current to adjust the membrane potential compatible. The tetrodotoxin (TTX, a voltage-gated Na+ -channel blocker, 1 μM, Sigma) and 4-aminopyridine (4-AP, a voltage-gated A-type K + -channel blocker, 1 mM, Tocris) were used in some experiments.