Transplantation of glutamatergic neuronal precursor cells in the paraventricular thalamus and claustrum facilitates awakening with recovery of consciousness

Abstract Background Stem cells offer a promising therapeutic strategy for patients with disorders of consciousness (DOC) after severe traumatic brain injury (TBI), but the optimal transplantation sites and cells are not clear. Although the paraventricular thalamus (PVT) and claustrum (CLA) are associated with consciousness and are candidate transplantation targets, few studies have been designed to investigate this possibility. Methods Controlled cortical injury (CCI) was performed to establish a mouse model of DOC. CCI‐DOC paradigm was established to investigate the role of excitatory neurons of PVT and CLA in disorders of consciousness. The role of excitatory neuron transplantation in promoting arousal and recovery of consciousness was determined by optogenetics, chemogenetics, electrophysiology, Western blot, RT‐PCR, double immunofluorescence labeling, and neurobehavioral experiments. Results After CCI‐DOC, neuronal apoptosis was found to be concentrated in the PVT and CLA. Prolonged awaking latency and cognitive decline were also seen after destruction of the PVT and CLA, suggesting that the PVT and CLA may be key nuclei in DOC. Awaking latency and cognitive performance could be altered by inhibiting or activating excitatory neurons, implying that excitatory neurons may play an important role in DOC. Furthermore, we found that the PVT and CLA function differently, with the PVT mainly involved in arousal maintenance while the CLA plays a role mainly in the generation of conscious content. Finally, we found that by transplanting excitatory neuron precursor cells in the PVT and CLA, respectively, we could facilitate awakening with recovery of consciousness, which was mainly manifested by shortened awaking latency, reduced duration of loss of consciousness (LOC), enhanced cognitive ability, enhanced memory, and improved limb sensation. Conclusion In this study, we found that the deterioration in the level and content of consciousness after TBI was associated with a large reduction in glutamatergic neurons within the PVT and CLA. Transplantation of glutamatergic neuronal precursor cells could play a beneficial role in promoting arousal and recovery of consciousness. Thus, these findings have the potential to provide a favorable basis for promoting awakening and recovery in patients with DOC.


| INTRODUC TI ON
Disorders of consciousness (DOC) are a common complication after brain injury. Some patients with DOC do not recover their consciousness; instead, they remain in a vegetative state. Such patients are completely incapacitated and require continuous professional medical care. 1 The average survival time of patients with DOC is only 2-5 years. 2 The incidence of traumatic brain injury (TBI) is as high as 13/100,000; the mortality rate of severe TBI is approximately 50% and 10%-15% of patients with severe TBI develop DOC. 3,4 Thus, approximately 40,000 people worldwide develop DOC following TBI in a given year. At present, many treatments, including pharmacotherapy, hyperbaric oxygen therapy, and neuromodulation techniques (deep brain electrical stimulation, 5,6 spinal cord electrical stimulation, 7,8 and cortical nerve stimulation 9 ), are used to treat DOC following TBI, but their efficacy is highly variable. With the development of stem cell technology, stem cells have taken on an important therapeutic role as "seed cells 10 " or "carrier cells 11,12 " in clinical practice. In fact, stem cell therapy has the potential as a treatment for DOC 13 . At present, stem cell transplantation for the treatment of DOC is still in the exploratory stage and faces two main problems: the selection of transplanted cells and the identification of optimal transplantation targets.
In 2018, it was reported that the paraventricular thalamus (PVT) is a key nucleus for maintaining wakefulness and parsing the neural circuit mechanisms involving the PVT 14 ; almost at the same time, another study elaborated on the important role of the PVT in encoding information in the brain. 15 Anatomically, the PVT is located in the midline region of the thalamus, which receives a large number of neural afferents from the hypothalamus and brainstem, has close bidirectional neural projections to the insula and prefrontal lobes, and is the hub of top-down neural network regulation and information transmission in the brain. 16 The findings of these studies suggest that the PVT may be a key nucleus in the development of DOC.
Koch suggested that the claustrum is connected bidirectionally with most cortical regions, and the claustrum (CLA) was hypothesized to play a critical part in information integration leading to consciousness. 17 Koubeissi reported that the stimulation of the CLA by an electrical current reverses the state of consciousness (from awake to comatose and back again) in epileptic patients. The findings from another experiment also support the possible association of the CLA with consciousness. 18 In that study, scientists examined 171 veterans with traumatic cranial injury and assessed the effects of CLA damage on consciousness; they found that CLA damage was associated with the duration of loss of consciousness (LOC). 19 In light of the above studies, we hypothesized that there are neural network correlates of consciousness, 20 in which the PVT is one of the nuclei controlling awakening, a process that is closely related to the maintenance of consciousness; we also hypothesized that the CLA is the key channel of the consciousness circuitry and one of the hubs of connectivity among various parts, controlling the transmission and integration of information. Excitatory neurons are the smallest units involved in the neural network of consciousness.
Injury to excitatory neurons might affect the regulatory function of the CLA and PVT, leading to the dysfunction of the consciousnessrelated neural network, which eventually leads to DOC.
In this study, we explored the roles of the PVT and CLA in DOC and determined which cells play major roles in these nuclei, thus informing the choice of transplanted cells to treat DOC, clarifying whether the PVT and CLA can be used as targets for cell transplantation and explaining the role of excitatory neurons in the neural network of consciousness.

| Animals
All animal experiments followed the Guide for the Care and Use of Laboratory Animals (8th edition, revised in 2011) and were approved by the Animal Ethics Committee of Fudan University. C57BL6/J mice were used as the genetic background of our mouse models. Adult male mice aged 6-8 weeks were used in this study. awakening with recovery of consciousness, which was mainly manifested by shortened awaking latency, reduced duration of loss of consciousness (LOC), enhanced cognitive ability, enhanced memory, and improved limb sensation.

Conclusion:
In this study, we found that the deterioration in the level and content of consciousness after TBI was associated with a large reduction in glutamatergic neurons within the PVT and CLA. Transplantation of glutamatergic neuronal precursor cells could play a beneficial role in promoting arousal and recovery of consciousness.
Thus, these findings have the potential to provide a favorable basis for promoting awakening and recovery in patients with DOC.

| Glutamatergic neuronal precursor cells
Glutamatergic neurons were induced by inhibiting the SMAD pathway by adding 1 μM dorsomorphin and 10 μM SB431542 to the culture medium of iPSCs.

| Establishment of the CCI-DOC mouse model
Controlled cortical injury (CCI) is a commonly used and highly regarded model of brain trauma that induces reproducible and well-controlled injury. 21 Healthy male C57BL/6 mice 6-8 weeks of age were anesthetized with an animal gas anesthesia apparatus at an induction concentration of 4% isoflurane and a maintenance concentration of 2%. The mice were fixed on a controlled cortical impactor (Hatteras PinPoint™ PCI3000) stereotactic device in the prone position. The top of the head was shaved, the area was routinely disinfected with Anl iodine, the scalp was anesthetized with local infiltration of lidocaine, the skin was incised at the midline of the head (the incision was approximately 1 cm), the soft tissue and the outer membrane of the bone were bluntly peeled away, and the skull was exposed. A circular bone window with a diameter of 6 mm was created at the right side of the midpoint between the fontanelle and the herringbone point ( Figure 1A). The impingement device was set at 3 mm above the dura, with an impact speed of 3 m/s, and the duration was 180 ms. The anesthetic machine was withdrawn immediately after the impact, and a consciousness assessment was performed ( Figure 1B). In the Sham group, the bone flap was removed without cortical impact. All mice were singly housed after the operations, and the temperature was maintained at 25°C. The airway was unobstructed, and aspiration was performed when necessary. An appropriate amount of saline was injected intraperitoneally to reduce mortality.

| Consciousness level assessment
The modified consciousness level assessment criteria 22 ( Figure 1D) were as follows: Stage I, events as usual.

Stage II, decreased movement.
Stage III, ataxia and decreased limb movement.
Stage IV, ability to roll to their sides when placed on their backs but inability to rise up.
Stage V, inability to right themselves when placed on their backs but still with a response to pain, as evidenced by limb withdrawal when pinched.
Stage VI, absence of both righting reflex and response to pain.
Animals were considered unconscious if classified as stage V or VI. The duration of loss of consciousness was the time between the disappearance and the recovery of the righting reflex ( Figure 1B).
We assigned each level of consciousness a corresponding score for statistical purposes, where levels I, II, III, IV, V, and VI were scored 6, 5, 4, 3, 2, and 1, respectively.

| Immunofluorescence staining
The mice were fixed in the supine position after deep anesthesia, and the chest was opened rapidly to fully expose the heart. First, approximately 100 mL of prechilled 4°C physiological saline was perfused, and after the rat's two forelimbs and two lungs turned white, they were perfused with prechilled (4°C) 4% paraformaldehyde solution. The brains were bathed in 4% PFA for 90 min, dehydrated in 30% sucrose, and embedded in optimal cutting temperature (OCT; Sakura) compound. The brains were cut into 20 μm sections. The

| Western blotting
Total protein was isolated with radioimmunoprecipitation assay

| Stereotactic injection
The microsyringe (2.5 μL) was rinsed 3-5 times with PBS. One microliter of air was aspirated into the microsyringe first to fully inject 0.25 μL of diluted virus into the brain. The syringe needle was placed according to the location parameter ( Figure S1). The injection volume of ibotenic acid was 0.25 μL (10 μg/μL in saline), and the injection volume of viruses was 200 nL (30 nL/min). The Sham group was injected with the same amount of saline. The needle remained in place for an additional 8-10 min after injection.

| Optogenetics
The viruses used in this experiment were AAV-CaMKIIα-ChR2- Five-minute stimulation periods were separated by 15-min intervals.
No laser stimulation was performed in the control group.

| Chemogenetics
The virus used in this experiment was AAV-CaMKIIα-hM4D-mCherry with a titer of 2.34 × 10 13 vg/mL, provided by Hoyuan Biotechnology (Shanghai) Co. Clozapine-N-oxide (CNO, 1 mM) was injected intraperitoneally at 1 mg/kg 4 weeks after virus injection, and behavioral experiments were performed 45 min later. Sham group animals were injected with saline.

| EEG-EMG monitoring and analysis
The EEG recording electrodes were implanted after anesthesia, and a reference electrode was placed on the forehead. The EMG recording electrodes were implanted in the neck muscles. EEG-EMG recordings were performed with the recording system (Pinnacle), and the data were analyzed with EEGLAB (v2019.1). The EEG-EMG signals were amplified (Grass Link, Grass Technologies), filtered (EEG: The LOC and awake-state classifications were performed by a researcher who did not participate in the experimental manipulation ( Figure S2).

| Novel object recognition test
The experimental procedure was performed as described previously. 23 Three objects were used. Objects A and B were identical, while object C was completely different from the others. Objects A and B were placed in symmetrical left and right positions in the field to begin the training. The recording software was turned on once the animal entered the room, and the recording time was 5 min.
Object B was replaced with object C in a later test. The animal was put into the test room as above and recorded for 5 min to observe its exploration of object C. The exploratory preference was calculated as the percentage of time spent investigating the novel object in the total time spent exploring objects.

| Y-maze test
The experimental procedure was performed as described previously. 24 The Y-maze apparatus consists of three identical arms. The animal was placed on the endpoint of a random arm and allowed to explore freely

| Adhesive removal test
The experimental procedure was performed as described previously. 25 Two sticky papers were applied to the paws of the mice.
The mouse was gently placed in the test box, and two timers were started. The time when the mouse began shaking or licking its paw was recorded as the start time, and the time when the mouse removed the sticky paper was recorded as the stop time. Bilateral paws were recorded, and the total time required to remove the sticky papers was calculated.

| Statistical analysis
The data are presented as the mean ± SEM. Cell counting was performed using ImageJ (NIH). All data were analyzed with the independent-sample t test in Prism 6 (GraphPad). The data were analyzed for normality using the D'Agostino-Pearson test, and for data that did not conform to a normal distribution, we used the Kruskal-Wallis test. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 were considered to be statistically significant.

| A possible paradigm for a mouse model of DOC
Disorders of consciousness (DOC) are a common complication following traumatic brain injury (TBI) worldwide; however, there is a lack of rodent models. We attempted to generate a mouse model of DOC on the basis of controlled cortical injury (CCI), as we described in Section 2 ( Figure 1A,B). In the following text, we refer to this model as CCI-induced DOC (CCI-DOC).

| Consciousness score decreases with increasing traumatic impact depth
For more visual statistics of the degree of DOC, we assigned each level of consciousness a corresponding score for statistical purposes, and levels I, II, III, IV, V, and VI were scored 6, 5, 4, 3, 2, and 1, respectively ( Figure 1C). We found that the consciousness score of the CCI-DOC mice decreased with increasing impact depth ( Figure 1D). The consciousness score of mice decreased sharply once the impact depth exceeded 2.5 mm (2.5 mm group vs. Sham group, p < 0.001; 3 mm group vs. Sham group, p < 0.001; and 3.5 mm group vs. Sham group, p < 0.001). All of the consciousness scores in the 3, 3.5, and 4 mm groups were lower than 2 points.

| The survival rate of CCI-DOC mice decreases with increasing impact depth
To better ensure the successful establishment of our model, we also determined the survival rate of mice. We found that when the impact depth exceeded 2 mm, the survival rate of CCI-DOC mice decreased as the impact depth increased. The mortality rate of CCI-DOC mice increased sharply when the impact depth exceeded 3 mm, and the 14-day survival rate was significantly lower than 50% ( Figure 1E). Therefore, in combination with the state of consciousness score ( Figure 1D CCI-DOC, we performed the novel object recognition experiment, Y-maze experiment, and adhesive removal test. The results showed that the discrimination index of the CCI-DOC group was significantly decreased compared with that of the Sham group (p < 0.001; Figure 1H). In addition, animals in the CCI-DOC group showed a reduced frequency of exploration ( Figure 1H,K). The results of the Y-maze experiment showed that memory was impaired and the correct alternating reaction rate was decreased in the CCI-DOC mice (CCI-DOC group vs. Sham group, p < 0.001; Figure 1I). The adhesive removal performance showed that limb sensation was significantly reduced in CCI-DOC mice, and the time to removal was significantly prolonged (CCI-DOC group vs. Sham group, p < 0.001; Figure 1J).  Figure 2D). Thus, we speculated that the activity or function of glutamatergic neurons might be associated with impaired consciousness. Therefore, we performed the same assay on other brain regions except for the traumatic brain impact area, where PVT and CLA neurons had the most apoptosis and other brain regions had a nonsignificant increase in neuronal apoptosis. Next, we also used optogenetic techniques to further reveal the role of glutamatergic neurons in DOC (Figure 4), before which we first had to clarify how the two nuclei function in DOC.

| Arousal capacity and content of consciousness were also altered after stereotactic ibotenic acid destruction of the PVT and CLA
To further verify the relationship between these two nuclei and DOC, we performed neurotoxin destruction experiments on each of these two nuclei (PVT and CLA).

| Assessment results in awakening ability
Five days after stereotactic injection of ibotenic acid, the arousal capacity and consciousness of mice were impaired ( Figure 3A).
After execution, their brain tissues were double labeled by TUNEL/NeuN immunofluorescence. The results showed that a large number of apoptotic neurons were severely damaged in the PVT ( Figure 3B) and CLA ( Figure 3C). The duration of LOC and the latency to wake were not significantly different (CLA-IA group vs. NS group, p > 0.05; Figure 3D). However, the PVT-IA group showed a statistically significant difference in the duration of LOC and the latency to wake (PVT-IA group vs. NS group, p < 0.05; Figure 3D).

| Assessment results in conscious content
The results of the novel object recognition experiment revealed that both the frequency of exploration and the discrimination index of the mice decreased significantly after the destruction of the CLA (CLA-IA group vs. NS group, p < 0.05; Figure 3E,H), but there was no significant decrease in the PVT-IA group (vs. NS group, p > 0.05; Figure 3E,H). The results of the Y-maze experiment revealed that the memory of the mice was significantly reduced after the destruction of the CLA (CLA-IA group vs. NS group, p < 0.05; Figure 3F), and the correct alternating reaction rate of the PVT-IA group was not significantly different (vs. NS group, p > 0.05; Figure 3F). The results of the adhesive removal test showed that the time to removal was significantly prolonged in the CLA-IA group (vs. NS group, p < 0.05; Figure 3G). The limb sensation of the mice in the PVT-IA group was not significantly different (vs. NS group, p > 0.05; Figure 3G).

| Optogenetic experiments
To determine the role of glutamatergic neurons in the regulation of consciousness, we performed an optogenetic experiment ( Figure 4A,B). Cells in the corresponding brain regions showed firing responses consistent with the stimulation frequency after stimulation with different frequencies of light ( Figure 4C). When the light stimulation frequency was greater than 20 Hz, the fidelity of action potential firing appeared to decrease significantly as the stimulation frequency increased ( Figure 4C

| Assessment results in awakening
The duration of LOC increased after 590 nm light inhibition of glutamatergic neurons in the PVT (vs. 470 nm group, p < 0.001; vs. Sham group, p < 0.001; Figure 5A), and the latency to wake was prolonged, but there was no statistically significant difference in the CLA (vs. 470 nm group, p > 0.05; vs. Sham group, p > 0.05; Figure 5A). The duration of LOC was significantly reduced after the activation of glutamatergic neurons in the PVT by 470 nm light (vs. 590 nm group, p < 0.001; vs. Sham group, p < 0.001), and the latency to wake was prolonged, but there was no significant difference in the CLA (vs. 470 nm group, p > 0.05; vs. Sham group, p > 0.05; Figure 5A).

| Chemogenetics experiments
Clozapine-N-oxide inhibition of the PVT was followed by an increased duration of LOC and prolonged awakening latency.
We performed follow-up experiments 4 weeks after hM4D virus injection in the PVT or CLA; 45 min prior to each experiment, mice were injected intraperitoneally with CNO ( Figure 5C). We found a significant increase in the duration of LOC in mice after the inhibition of the PVT, and we found similar results for the latency to wake (vs. NS group, p < 0.001; vs. CLA-CNO group, p < 0.001; Figure 5D).
The duration of LOC was not significantly altered after CLA inhibition, and the latency to wake was not significantly altered (vs. NS group, p > 0.05; Figure 5D).

| Cognitive decline, memory loss, and diminished limb sensation after CNO inhibition of the CLA
After the inhibition of glutamatergic neurons within the CLA by CNO, the mice showed a significant decrease in the discrimination index and the frequency of exploration (vs. NS group, p < 0.01; vs. PVT-CNO group, p < 0.05; Figure 5E), but no statistically significant difference was observed after the PVT was inhibited (vs. NS group, p > 0.05; Figure 5E). In addition, in the adhesive removal test and Y-maze test, the mice in the CLA-CNO group performed even worse after the inhibition of glutamatergic neurons within the CLA by CNO (vs. NS group, p < 0.01; vs. PVT-CNO group, p < 0.05; Figure 5F,G).

| Acquisition of glutamatergic neuronal/precursor cells through the induction of iPSCs in vitro
The results of immunofluorescence staining showed that the cells induced from iPSCs ( Figure 6A

| Acquired glutamatergic neurons exhibit excitatory neuronal electrophysiological properties
Whole-cell patch clamp results showed that normal Na + currents and K + currents were successfully recorded in cortical neurons induced from iPSCs in vitro ( Figure 7A,B); Na + currents could be blocked by TTX and K currents could be blocked by 4-AP ( Figure 7A,B). In addition, cortical neurons induced from iPSCs showed strong, regular action potentials in response to step current injection ( Figure 7C,D). Thirty days after transplantation, we euthanized the animals, rapidly removed their brain tissue, and sectioned the tissue for brain slice patch-clamp assays. The patch-clamp results showed that normal Na + and Ca + currents and normal postsynaptic currents of excitatory neurons after transplantation were recorded after transplantation ( Figure 8F-J). The excitatory postsynaptic current (EPSC) could be inhibited by NBQX/AP-V ( Figure 8H). These results indicate that transplanted glutamatergic neuronal precursor cells survive and further develop in the brain and function as excitatory neurons.

| Transplanted glutamatergic neuronal precursor cells promote awakening and improve conscious content
To better demonstrate the effect of glutamatergic neuronal pre- In the CLA group, 30 days after iPSC-induced neuron transplantation, the conscious content was significantly improved (vs. NS group, p < 0.01; vs. NSC group, p < 0.05; Figure 9F-H). In addition, the expression of PSD-95 increased significantly ( Figure 9I). Although an association between these two nuclei and consciousness has been reported, 14,29 to demonstrate the relation- We propose that the mechanisms of DOC improvement may be as follows. First, transplantation directly replenished glutamatergic neurons, which released more excitatory neurotransmitters.

| DISCUSS ION
Glutamate is the most abundant neurotransmitter in the vertebrate nervous system, and more than 90% of excitatory synapses in neurosynaptic connections use glutamate as a neurotransmitter for information transmission. 36 Our study has considerable value but also some limitations. We identified two different functions of the PVT and CLA in consciousness generation, but we did not further elucidate whether there is some connection between the two nuclei; we believe that there may be some potential circuits linking them, which will be the next topic for our team to explore in depth.

| CON CLUS ION
In summary, we first constructed a feasible CCI-DOC model. We analyzed the data and wrote the manuscript.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

R E FE R E N C E S
F I G U R E 9 Glutamatergic neuronal precursor cell transplantation significantly improves wakefulness and facilitates recovery of conscious content. (A) Flow chart of the cell transplantation experiment. (B, C) At 7, 20, and 30 days after cell transplantation in the PVT, we performed the awaking ability assessments. The results showed that in the iPSC-induced neuron group, the duration of LOC was statistically reduced, and the latency to wake was shortened (n = 6 for each group). (D, E) We compared the effects of transplantation into the CLA and PVT on awaking ability, and the results showed that in the iPSC-induced neuron transplantation groups, the PVT group showed significantly enhanced arousal, while the CLA group did not show a statistically significant difference in awaking ability (n = 8 for each group). (F) After transplantation of iPSC-induced neurons, the discrimination index improved significantly (n = 6 for each group), and the frequency of exploration was enhanced (n = 8 for each group). (G, I) In the CLA group, 30 days after iPSC-induced neuron transplantation, the evaluation of conscious content was significantly improved (n = 7 for each group). (H) At 30 days after cell transplantation, elevated PSD95 expression was seen in the transplanted region, with significantly higher PSD95 expression in the iPSC-derived neuron group. Error bars: SEM. Significance determined by Student's t test: ***p < 0.001, n.s. p > 0.05. LOC, loss of consciousness.