Fig 1.
General schema of hES cell differentiation.
Optogenetically engineered human neural precursors were differentiated into neurons with forebrain fate preference. The functionality of derived neurons was measured by their light-responsive action potentials and synaptic activity in vitro and the induction of the immediate early gene c-FOS in host neurons in vivo.
Fig 2.
Induction and differentiation protocol used to generate hNPs from H9 hES cells and to further advance them to a cortical neuronal fate.
ES, embryonic stem cell; KOSR, KnockOut™ Serum Replacement; NEAA, non-essential amino acids; 2-ME, 2-mercaptoethanol; EB, embryoid body; NP, neural progenitors; FGF, fibroblast growth factor 2; BDNF, brain-derived neurotrophic factor; GDNF, glial cell-derived neurotrophic factor.
Fig 3.
Critical steps in neural differentiation of human ES cells (Day 0) to NPs (Day 25).
On Day 0, H9 ES cell colonies (A) were detached using collagenase to form embryoid bodies (B; EBs on Day 2) that were further cultured in EB media with dorsomorphin and A-83. Beginning at day 4, medium was changed to the one promoting NP cell differentiation that was used to feed every other day to Day 22. On Day 7, EBs were attached to matrigel-coated plate; neural rosette structures were evident at day 8 (C) and became prominent on day 11 (D-E). On day 22, colonies with rosettes were detached manually to form neurospheres (F). On day 23, neurospheres were dissociated into single cells with accutase, plated on matrigel-coated plate, and fed with NP expansion media till they became confluent (G-H). Cells were passed every 4–5 days. Days 36 NPs were used for viral transduction and Day 41 NPs for transplantation. Scale bars: A-C, F, 500μm; D and G-H, 200μm; E, 100μm.
Fig 4.
Characterization of H9-derived hNPs by immunocytochemistry at an early stage.
Neuronal progenitors were plated on matrigel-coated ibidi chambers or PDL/Laminin coated-12 mm coverslips on day 23. Cells were fed with NDM every 2 days. At Days 13–28, most H9-derived cells expressed the primary progenitor and telencephalon marker FOXG1 (A) and the neural stem cell markers nestin (B) and PAX6 (C); cells immunoreactive for the stem cell marker Oct4 were rare (data not shown). On Day 41 (E-F), some cells were TUJ1, MAP2 and SMI-312 positive but vGAT, vGLUT1 and synapsin 1 signals were absent (data not shown). Scale bars: A, 25μm; B-D and F, 100μm; E, 50μm.
Fig 5.
Expression and localization of hChR2 in H9-derived hNPs.
Panels A-C depict typical neurons at various days of maturation in vitro: H9 hNPs, photographed here on day 47 in culture, had been transduced on Day 34 by lentivirus carrying hChR2 at moi 13. Human ChR2-YFP expression started to appear two-four days after transduction and progressively spread over the perikaryon and processes of maturing neurons. Panels D-F demonstrate the fine localization of hChR2 (YFP fluorescence), predominantly to the membranes of representative transduced nerve cells (arrows), seven days after transduction, by confocal microscopy. Scale bars: A-C, 100μm; D-F, 50μm.
Fig 6.
Characterization of hNP-derived neurons co-cultured with mouse astrocytes by immunocytochemistry at an advanced stage (Days 97 and 116).
The use of the co-culture system with CD1 astrocytes (A) greatly reduces aggregation and promotes neuronal differentiation in the form of nicely separated mature neurons expressing advanced neuronal signatures such as phosphorylated neurofilaments M and H (B: SMI312 immunoreactivity), MAP2 (C), and SYN1 (D, in conjunction with MAP2). Neurotransmitter markers consistent with excitatory or inhibitory neurotransmission such as vGLUT1 and vGAT are also abundantly expressed in these cultures (E and F). Note the vesicular nature of vGLUT1 and vGAT immunoreactivity. Scale bars: 100μm.
Fig 7.
Human NPs display glutamatergic and GABAergic spontaneous synaptic transmission in vitro (Days 90–167).
Voltage clamp recordings of spontaneous synaptic currents from ChR2- neurons. Holding potential -90 mV; at room temperature. (A) Example recording showing that in this cell, spontaneous synaptic currents are completely and reversibly blocked by 50 μM CNQX, a glutamate receptor blocker. (Aʹ) Top left panel shows a superimposed view of synaptic events from recording in A at an enlarged time scale (gray traces, n = 122 events). Average trace is shown in black. Bottom right panel shows an example of one postsynaptic event. Red line indicates monoexponential fit used to calculate the decay time constant. (B) Distribution of decay time constants for the recording shown in A, with a peak of the distribution at 0.79 ms. (C) Example recording showing that in this cell, spontaneous synaptic currents are completely and reversibly blocked by 20 μM Bicuculline, a GABA-A receptor blocker. (C´) Postsynaptic events from recording in C (n = 131 events, gray lines) were superimposed and averaged (black trace, top left). Bottom right panel shows an example of one postsynaptic current with monoexponential fit of decay time constant. (D) Distribution of decay time constants of the recording shown in C, with a peak of the distribution at 15 ms. Note that events at these inhibitory synapses appear as excitatory postsynaptic currents in this type of recording, as the reversal potential of chloride has been changed in the artificial in vitro recording environment. (E) Example recording showing a mix of glutamatergic and GABAergic synaptic events during spontaneous activity. In this cell, CNQX only partially abolished postsynaptic events (middle trace). The remaining events are blocked by bicuculline (bottom trace). (F) Distribution of decay time constants of postsynaptic events from the control trace before drug application in E shows a distribution with two peaks, at 1.6 ms and 8.5 ms (arrow). (G) Distribution of decay time constants of postsynaptic events during CNQX application for the recording shown in the middle trace in E. Note that compared to the distribution in F, the peak at 1.6 ms is abolished in CNQX (red dotted line); therefore it is representing the glutamatergic postsynaptic events. The blue line represents the events that are not blocked by CNQX, but sensitive to bicuculline, and therefore GABAergic. The arrow points to the same peak as in F, now appearing larger due to rescaling.
Fig 8.
Series of confocal images indicating the differentiation of transplant-derived cells into early neurons with hChR2 expression.
(A) These four panels indicate that a large percentage of transplant-derived HuNu+ cells (green) are also immunoreactive with antibodies against class III β-tubulin epitope TUJ1 (red). Blue nuclei are stained with DAPI. (B) These panels indicate that the majority of transplant-derived epitope SC121+ positive cells (red) also fluoresce in the green spectrum, which is indication of hChR2 expression. Cell nuclei are stained with DAPI (blue). Scale bars: 200μm.
Fig 9.
Neuronal differentiation of ChR2-transduced and non-transduced hNP transplants and pathfinding of axons from their neuronal progenies two months after transplantation (mice).
(A) These four panels are from representative cases visualized with IHC for the human-cell specific epitope SC121 and illustrate a consistent engraftment of hNPs in the primary motor area at a coronal level corresponding to caudal septum. Transplant engages deep cortical layers, medial corpus callosum, and dorsal neostriatum. In most cases, the transplantation needle track is also evident. Panels also show that SC121+ transplant-derived axons project along the fibers of corpus callosum and exit at the level of claustrum. Transplants in #21 and #22 contain transduced hNPs, whereas in #12 and #15 contain non-transduced cells. All cases are from injured mice. (B-D) SC121 immunohistochemical preparations illustrating the propensity of transplant-derived axons to exit the lateral corpus callosum at the level of the claustrum and then course in deep insular/piriform cortex. Case 15 involves non-transduced NPs. There is a striking concentration of axons at the claustrum, and an occasional further advancement towards the insular and piriform cortex. In some cases, bundles of axons extend all the way to layer II of piriform/insular cortex (B), commonly following a course oblique or vertical to the plane of view (C). Panel D shows that axon bundles branch into numerous single axons with boutons (arrows) in the deep insular/piriform cortex (Case 22, transduced NPs). (E-H) These human synaptophysin-immunostained preparations show the formation of early synaptic fields by transplant-derived neurons. Low-level synaptophysin expression often extends from the site of the transplant all the way to the lateral corpus callosum and insular/piriform area, with an apparent terminal field at the claustrum/piriform cortex (E). With higher magnification, terminal field contains putative synaptic boutons (F; photograph is taken from the claustral/piriform area indicated with asterisk on panel E). Bundles of axons that course obliquely or vertically to the coronal plane as in panels C-D also express diffuse synaptophysin immunoreactivity (G; bundles are indicated with arrows). Synaptophysin+ boutons surround such bundles (H; panel is a magnification of area indicated with an asterisk on panel G). M1, primary motor cortex M2, secondary motor cortex; S1, primary somatosensory cortex; cl, claustrum; ep, endopiriform nucleus; pir, piriform cortex; SYP, synaptophysin. Scale bars: A, 800μm; B-C, 200μm; D, 30μm; E, 400μm; F-H, 20μm; G, 100μm.
Fig 10.
Differentiation of hChR2-hNP transplants in rat motor cortex and patterns of innervation of host neurons.
(A-B) These YFP-immunostained preparations show the distinct features of hChR2+ neural tissue in the transplant (A) versus a hChR2+ terminal field in cingulate cortex (B). YFP expression in (A) is restricted to the membrane of the neurons and their processes. The latter form a dense neuropil with webs of processes filling the space between neuronal cell bodies. Some of these hChR2-hNP-derived neurons, for example the one indicated with the arrow and enlarged in the inset have cortical features. The terminal field of hChR2-hNP-derived neurons in cingulate cortex (B) has a very different appearance than the “neuropil” in the transplant and it obeys the cytoarchitecture of host cortex. No perikaryal-type staining is encountered, evidence that graft-derived neurons have not migrated into host cortex. (C) This dually stained preparation for two transplant-selective neuronal markers (YFP immunoreactivity for hChR2 and human synaptophysin immunoreactivity) is from layer II of host cingulate cortex and demonstrates both the dense terminal field and the extensive colocalization of the two markers in transplant-derived axons and their processes (asterisks; double labeling is white here). A larger panel showing more of this terminal field is in S6 Fig. (D) These dually stained preparations from host motor and cingulate cortex with MAP2 for neurons and human synaptophysin for transplant-derived terminals show a very large number of terminals apposing dendrites and dendritic branches of host neurons in layer 2–3 of motor cortex (top left), layer 5 of motor cortex (top right), deep layer 2–3 of cingulate cortex (middle) and superficial layer 2–3 of cingulate cortex (bottom). In many cases, dendrites are sectioned transversely. There are occasional terminals on the perikarya of large neurons (see an example on a pyramidal neuron labeled with #). Scale bars: A-B, 50μm; C, 10μm; D, 20μm.
Fig 11.
Partial characterization of neurotransmitter identity of differentiated hChR2-hNP transplants and their terminals in host brain.
(A-B) These are dually immunostained preparations for YFP (hChR2 marker to label the transplant and transplant-derived structures) and either vGAT, a presynaptic marker of GABAergic neurotransmission (A) or vGLUT1, a presynaptic marker of glutamatergic neurotransmission (B). Color panels show both transplant and neurotransmission marker immunoreactivities, whereas black and white panels underneath the color ones illustrate transplant and neurotransmitter markers separately. Note the dense GABAergic neurotransmission in the graft (orange neuropil in A) as contrasted with the very low glutamate marker immunoreactivity (B). (C-E´) Representative images illustrating the GABAergic or glutamatergic differentiation of transplant-derived human synaptophysin+ (hSYP+) terminals in rat motor cortex. Differentiation was assessed with IHC for GABAergic markers such as vGAT (C-C´) or GAD1 (D-D´) and glutamatergic markers such as a mixture of antibodies for vGLUT 1 and 2 (vGLUT) (E-E´). Images in C-C´ are taken at a lower magnification than D-E´. There are multiple vGAT+ or GAD1+ transplant-derived hSYP+ terminals (yellow color), most of them on non-identifiable host structures, probably dendrites (asterisks in C-C´) but quite a few also on somata (double asterisk on C´ as part of a basket-type GABAergic innervation of a rat cortical neuron; arrows on D-D´). At the contrary, we were not able to find any colocalization of vGLUT in hSYP+ terminals (D-D´); even when there is double labeling as in E´, the relationship is that of juxtaposition, not colocalization. Asterisks on E-E´ are unlabeled rat (host) structures, probably dendrites. Scale bars: A-B, 100μm; C-E, 10μm.
Fig 12.
Myelinated human axons in dually immunostained preparations with YFP antibodies (for hChR2+ axons) MBP (for myelin) through the border of a transplant.
Single optical sections are taken with a confocal microscope. Human ChR2-hNP-derived axons are in green and myelin in red. At least a third of these axons is myelinated. Inset on top right is a 3D rendition of profile 1 in main frame from z-stack. Insets on bottom left are optical sections of profiles 3–5. Note the non-continuous myelination in the form of serial MBP+ “cuffs”. Graft is on the left. Rat corpus callosum is on top right. Scale bar: 10μm.
Fig 13.
Myelinated human axons deeply into host cortex.
This section was immunostained and imaged exactly as section in Fig 12. Here we illustrate a terminal field of transplant-derived neurons in cingulate cortex. Note the occasional double-labeled profiles, suggesting a low degree of myelination of transplant-derived axons away from the graft. Insets are magnifications of profiles in main frame linked with arrows. Scale bar: 20μm.
Fig 14.
An example of ensheathing-type contacts made between a transplant-derived axon and several host-derived (rat) oligodendrocytic processes in host cingulate cortex.
Maximum projection confocal image was captured from a section triply immunostained for YFP for transplant-derived axons (green), CNPase for oligodendrocytes (red) and SC121 for transplant-derived cells (white). (Top panel). Although four oligodendrocyte cell bodies are identified (asterisks), it appears that many of the contacts are made by the centrally located cell. None of the red oligodendrocyte profiles was double labeled with SC121, indication that they were of host origin (the lack of SC121 immunostaining of the central profile is shown in inset, against the SC121 immunoreactivity of the human axon). For the sake of conveying a sharper color, white was omitted from the panel. Examples of ensheathing contacts are indicated with numbers 1–4. (Bottom panels). Here we performed 3D reconstructions of ensheathing profiles 1–4 to show the detailed configuration of the human axon and the host oligodendrocytic processes enveloping it. Ensheathing profile 3 has unusual morphology, but this may be related to the immaturity of the process. Scale bars: Top panel, 20 μm; insets, 5μm; bottom panels, 1μm.
Fig 15.
Electrophysiological recordings and optogenetic light stimulation of hChR2-hNP-derived neurons differentiated in vitro.
(A-C) Representative voltage clamp traces in response to blue light stimulation (485 nm) at 5 Hz (A), 10 Hz (B) and 20 Hz (C) obtained in hChR2-hNP-derived neurons. The amplitude of the 1st pulse response was similar at each tested stimulation rate (for example, 269.8 ± 42.2 pA at 5 Hz). However, response amplitudes decreased during the 10-pulse stimulation, by a total of 29–36% (from 1st to 10th pulse), with an increasing level of adaptation at higher stimulation rates (asterisks). (D-F) Representative current clamp traces in response to a train of light stimulation at 5 Hz (D), 10 Hz (E) and 20 Hz (F). Before light exposure, the membrane potential was preset to -55 mV by current injection. When TTX is added to the extracellular solution AP generation is blocked (E´). (G) Comparison of spiking fidelity in response to different rates of 10-pulse optogenetic stimulations. (H) During a single 2.5s-long light stimulation, the AP firing rate initially increases and then adapts. (I) During a 2.5 s-long single light stimulation (+light), AP firing rate significantly increased from 0.7 ± 0.2/s (-light) to 3.6 ± 0.6/s (n = 10; p < 0.001, paired t-test).
Fig 16.
Light stimulation of hChR2-hNP-derived neurons triggers postsynaptic events in ChR2- and ChR2+ neurons.
(A) Confocal image showing ChR2+ neurons (green) in co-culture with ChR2- neurons after 119 days in vitro. (B) Example of a voltage clamp recording from a ChR2- neurons in co-culture with ChR2+ neurons; same cell as shown to be recorded in A. In response to light stimulation (blue bar), a slow inward current and synaptic events were triggered in control conditions (top trace). Inset shows an extended time scale of the recording time. The postsynaptic responses were blocked by the presence of CNQX and bicuculine suggesting that they were due to light stimulation of presynaptic hChR2+ neurons. (C) Synaptic rate before and during light stimulation plotted for individual recordings of ChR2- neurons (n = 8) and ChR2+ neurons (n = 3). Solid black squares represent mean of combined ChR2- and ChR2+ neurons, respectively (mean ± SD). For ChR2- neurons, the synaptic rate increased about 4-fold during light stimulation, from 2.2 ± 2.4 to 8.9 ± 8.4 events/s (n = 8; p < 0.05, paired t-test).