Near‐Infrared Photothermal Manipulates Cellular Excitability and Animal Behavior in Caenorhabditis elegans

Near‐infrared (NIR) photothermal manipulation has emerged as a promising and noninvasive technology for neuroscience research and disease therapy for its deep tissue penetration. NIR stimulated techniques have been used to modulate neural activity. However, due to the lack of suitable in vivo control systems, most studies are limited to the cellular level. Here, a NIR photothermal technique is developed to modulate cellular excitability and animal behaviors in Caenorhabditis elegans in vivo via the thermosensitive transient receptor potential vanilloid 1 (TRPV1) channel with an FDA‐approved photothermal agent indocyanine green (ICG). Upon NIR stimuli, exogenous expression of TRPV1 in AFD sensory neurons causes Ca2+ influx, leading to increased neural excitability and reversal behaviors, in the presence of ICG. The GABAergic D‐class motor neurons can also be activated by NIR irradiation, resulting in slower thrashing behaviors. Moreover, the photothermal manipulation is successfully applied in different types of muscle cells (striated muscles and nonstriated muscles), enhancing muscular excitability, causing muscle contractions and behavior changes in vivo. Altogether, this study demonstrates a noninvasive method to precisely regulate the excitability of different types of cells and related behaviors in vivo by NIR photothermal manipulation, which may be applied in mammals and clinical therapy.


Introduction
In order to explore the complex nervous system and achieve precise therapy, a variety of neural manipulation techniques have DOI: 10.1002/smtd.20230084818][19] Among them, the most widely used neuromodulation technique is optogenetics, which can manipulate neurons by optical stimulation with spatiotemporal precision and sensitivity.However, implantation of optical fibers often causes tissue damage and immune response, and the penetration of visible light is too shallow to stimulate deep brain tissues. [20,21]The invasive problem also exists in traditional electrical stimulation, due to the electrodes are needed to be implanted into the mouse brain. [22,23]Then some non-invasive techniques emerged, such as the magnetothermal modulation, which can perform strong tissue penetration without brain implants concerns, has been applied to evoke motor behaviors in mice.[18] Another option for non-invasive manipulation is chemogenetics, using small chemical molecules to regulate target cells through ligand receptors.But the time resolution is coarse and hard to control, because the drugs have long residence time in vivo. [6,7]Ultrasound stimulation is also a non-invasive technique for neural modulation, which can pass through the skull to deep brain regions.[26][27][28] However, since the development of ultrasound stimulation is still in its infancy, the underlying mechanism and potential side effects are poorly studied, there are still many uncertainties in its specificity and safety. [29]herefore, it is desirable to develop a non-invasive technique to precisely modulate neural activity with high biosafety, especially in vivo.
Near-infrared (NIR) light modulation with a wavelength of 700-1880 nm, which is well known as "biological transparency window" due to its deep tissue penetration, has emerged as a promising and noninvasive technology for neuroscience research and disease therapy. [30,31]Especially, NIR-stimulated optogenetic tools have been used to directly regulate neural activity through NIR-responsive materials.[34] Another type of emerged NIR responsive materials is NIR-responsive photothermal materials, which can convert NIR light energy into heat and then trigger thermosensitive proteins expressed on cell membranes.Various kinds of photothermal agents have been synthesized and studied, including organic polymers, [35,36] gold nanoparticles, [37] carbon-based materials, [38,39] semiconductors, [40] and samll dyes. [41]Although some NIR-responsive materials have excellent photoconversion or photothermal conversion properties, and already be used to regulate cellular signaling pathways and gene transcription in a noninvasive manner.However, it should be noted that most of the NIR-responsive materials are inorganic or metal nanomaterials, which are difficult to apply in modulation of neural activity in vivo or clinical therapy, due to their biological toxicity and safety hazards.
Indocyanine green (ICG) dye is an FDA-approved photothermal agent, which can effectively transfer optical energy to thermal energy in response to NIR stimuli.And ICG can be metabolized by livers, has a short life time in blood circulation.[44][45] Therefore, ICG may be a promising candidate material for photothermal modulation of cellular activity in vivo.
The nematode Caenorhabditis elegans (C.elegans) is an extensively used model organism with numerous advantages for neuroscience research, including the small but well-characterized nervous system, complete connectome, amenability to genetic manipulation, short life span, and complex behavioral characteristics. [46,47]C. elegans offers a fast and mature in vivo experimental system to critically evaluate and optimize the effects and efficacy of neural modulation techniques.
In the present study, we developed a NIR photothermal system to regulate C. elegans cellular excitability and related behaviors in freely moving animals through thermosensitive TRPV1 channel with ICG in vivo.We successfully applied the photothermal modulation specifically in different types of neurons and muscle cells to mediate corresponding behaviors.Upon NIR stimuli at 793 nm with ICG, TRPV1 exogenously expressed AFD sensory neurons showed Ca 2+ influx and increase of neural excitability, and then induced backward locomotion.The GABAergic D-class motor neurons can also be activated by NIR irradiation, resulting in slower thrashing behaviors.Moreover, our system can also activate body wall muscles and vulva muscles, enhancing muscular excitability, leading to muscle contractions which induced paralysis and egg-laying behaviors respectively in C. elegans.Meanwhile, ICG showed no negative effects on survival, reproduction and lifespan of the nematodes.Altogether, we demonstrate a noninvasive and effective method to precisely regulate the excitability of different types of cells and animal behaviors in vivo by NIR photothermal manipulation.

Optical and Photothermal Characterization of ICG in vitro
ICG is a tricarbocyanine dye with a molecular weight of 751.4 Da.It is an FDA-approved clinical drug with high biocompatibility and has been widely used in angiography and fluorescence imaging. [44,45]First, we measured the optical and photothermal characterization of ICG.In our test, ICG showed an absorption spectrum ranging from 600 to 850 nm, the maximum absorption peak was observed at 796.5 nm, and the emission spectrum ranging from 700 to 1100 nm with a peak value at 865.2 nm (Figure 1a).These results suggested that ICG can be effectively excited by NIR stimuli at 793 nm, which would be used in our following experiments.Then, we tested the photothermal efficacy of ICG.After continuous irradiation by 793 nm (10 mW mm −2 ) near-infrared laser, the temperature increased in a dose-dependent manner with ICG concentration at 1, 3, 10, 30, 50, and 100 μg mL −1 (Figure 1c).The temperature of ICG reached plateau to 28.40 ± 0.10, 30.17 ± 0.23 °C, 40.40 ± 1.00 °C, 46.00 ± 0.95 °C, 50.97 ± 0.90 °C and 52.17 ± 0.12 °C, at the time of 0.67, 2.33, 3.00, 4.33, 4.91, and 5.33 min, respectively (Figure 1c-d), but did not change upon laser irradiation at the absence of ICG.The photothermal efficacy of ICG reached saturation level from the concentration of 50 μg mL −1 .
We then evaluated the temperature oscillation of ICG in fetal bovine serum (FBS) or in ultrapure water under a pulsed NIR irradiation (10 mW mm −2 , 4 min duration, 4 min intervals, 5 cycles).It was observed that the peak temperature of ICG dissolved in ultrapure water did not decrease significantly during the first three NIR exposures, but began to decrease significantly at the fourth irradiation.However, there was no significant change in the ICG solution mixed with FBS during the entire irradiation (Figure 1b).We also confirmed that pure FBS solution has no temperature oscillation under NIR irradiation (Figure S1, Supporting Information).These results showed a more stable photostability of ICG in FBS at NIR laser irradiation, which can be used for photothermal manipulation, so we performed our experiments with ICG dissolved in FBS.

In Vitro NIR Photothermal Activation of TRPV1 with ICG
We further evaluated the photothermal efficiency of ICG under NIR illumination in cultured cells.First, we examined whether ICG can be effectively taken up by cells.After 1 h incubation, both the SY5Y neuroblastoma cells (Figure S2a, Supporting Information) and HEK293T cells (Figure S2d, Supporting Information) were effectively labeled by ICG and sustained a detectable signalto-noise ratio fluorescence in NIR fluorescence microscopic imaging.
Then, to explore if the NIR-ICG photothermal system can modulate cellular activity, we transfected the mouse TRPV1 (mTRPV1) tagged with mCherry in HEK293T cells.The transient receptor potential cation channel subfamily V member 1 (TRPV1) is a well-known temperature-sensitive ion channel.Activating TRPV1 by high temperature (>43 °C) can induce Ca 2+ influx, which is a vital physiological process in neural firing. [48,49]The Ca 2+ influx was monitored using an intracellular calcium fluorescent indicator Fluo-8am in our study.When elevated both temperature to 45 °C, a calcium increase was observed in mTRPV1-mCherry transfected HEK293T cells (Figure S2b, Supporting Information), but not in control cells which were transfected with mCherry only (Figure S2c, Supporting Information).Importantly, the mTRPV1 transfected HEK293T cells can also be activated by NIR irradiation after ICG incubation, exhibited a significant calcium increase (Figure S2e, Supporting Information).When the cells were treated with ICG or NIR only, or not transfected with mTRPV1, there was no significant change in intracellular calcium signal.These cells could still be activated by capsaicin, the agonist of TRPV1, suggested the effective expression of TRPV1 and normal cellular physiological functions after treatment with the NIR and ICG.These data demonstrated that exogenously expressed TRPV1 could be activated by NIR stimuli with photothermal agent ICG, leading to a calcium response in HEK293T cells.

Photothermal Characterization of ICG in C. elegans In Vivo
Having demonstrated the effectiveness of NIR-ICG system in vitro, we asked whether the photothermal manipulation can be used in living animals.We chose Caenorhabditis elegans as the experimental model, and investigated the photothermal characterization of ICG in vivo.In order to test whether ICG could be introduced into the nematodes, we mixed ICG with Escherichia coli, the food of C. elegans.After ≈12 h of feeding, the wide-field nearinfrared fluorescence imaging was performed.And the ICG fluorescence signals could be firstly observed in the anterior pharynx in C. elegans after feeding with 0.2 μg mL −1 ICG.Moreover, when increasing the ICG concentration to 1 μg mL −1 and 2 μg mL −1 , the whole pharynx and intestine of the nematode progressively exhibited obvious ICG fluorescence signals (Figure 2a).These results indicated that ICG could enter the nematodes by feeding and distribute along the digestive system of the worms.
To explore whether the heat generated by NIR-ICG photothermal system was sufficient to regulate the physiological activities of C. elegans, we generated a transgenic strain carrying a transcriptional reporter gene gfp driven by the promoter of hsp-4 gene (heat shock protein).Using this transgenic strain, we can directly detect the expression of HFP-4 by observing the GFP fluorescence intensity changes.HSP-4 is expressed in several tissues including hypodermis, intestine, spermatheca, and some neuron cells.And the transcription of hsp-4 can be upregulated in response to endoplasmic reticulum stress induced by dithiothreitol (DTT) or tunicamycin (unfolded protein response) as well as in response to heat shock. [50,51]To test the thermal activation efficiency of Phsp-4::GFP strain, we incubated the nematodes at 33 °C for different lengths of time, then detected the GFP fluorescence intensity of the worms.We found that the GFP represented expression of HSP-4 increased obviously and progressively after heating from 1 to 30 min (Figure 2b,d).Then, we performed the heat shock experiment through our NIR-ICG photothermal system.Excitingly, the photothermal manipulation also induced significant heat shock effect.After feeding with 10 μg mL −1 ICG for 12 h, worms were irradiated by continuous NIR light (793 nm, 10 mW mm −2 ).We also applied gradient experiments at different stimulation times.With the extension of NIR irradiation time, the expression of HSP-4 also increased, in the presence of ICG.
It showed that the expression of HSP-4 was visibly upregulated under NIR irradiation for only 1 min.And treated worms with NIR-ICG photothermal stimulation for 5 min or 10 min was sufficient to fully activate HSP-4 expression, which demonstrated the high efficiency and effectiveness of our photothermal system (Figure 2c,e).NIR irradiation at 793 nm had no effect on temperature change of the NGM plates per se (Figure S3, Supporting Information).Collectively, these data suggested that ICG could successfully perform photothermal conversion response to NIR stimuli in vivo.Moreover, it is the first time to modulate the expression of heat shock proteins by a photothermal system, rather than direct heat in C. elegans.

Distribution of ICG in C. elegans Ingested by Feeding
To further test if our NIR-ICG photothermal system can regulate the activity of specific tissue cells in vivo, we constructed transgenic nematode strains expressing TRPV1 in different types of neurons and muscle cells.DsRed was used to label the target tissue cells.Confocal images showed the exogenously expression of mTRPV1-sl2-DsRed in AFD sensory neurons, the GABAergic D-class motor neurons (D-MNs), body wall muscles and vulva muscles in C. elegans driven by their own tissue-specific promoters, respectively (Figure S4, Supporting Information).Then to test if these tissue cells can be activated specifically by NIR with ICG, these transgenic worms were fed by Escherichia coli food mixed with ICG, as in the previous photothermal heat shock experiment indicated.After about 12 h of feeding, two-photon scanning fluorescence microscopic (TPFSM) imaging was performed to detect the expression of mTRPV1-sl2-DsRed and the location of ingested ICG on the same worm.The ICG fluorescence signals were widely distributed along the alimentary canal of C. elegans (Figure 3).We then measured the distance between ICG and TRPV1 expressed tissue cells (Figure 3b-d).We found that the ingested ICG has a distance ≈5 μm from AFD sensory neurons, 9 μm from D-class motor neurons, 8 μm from vulva muscles and 10 μm from body wall muscles, respectively.As the data showed, the distance between ICG and tissue cells was less than 10 μm, indicating a possible effective photothermal transfer to targeted cells.Next, we examined the activation of these specific tissue cells by NIR-ICG photothermal stimulation and detected some interesting corresponding behavioral output.

In Vivo NIR Photothermal Modulation of AFD Sensory Neurons Excitability and Reversal Locomotion
AFD neurons are the main thermosensors in C. elegans, which respond to thermal fluctuations at temperatures above its cultivation temperature (Tc). [52,53]The animal increases its reversal and turn frequency when detects a rise in temperature and then moves back down the gradient toward Tc (negative thermotaxis).This avoidance response requires the cell autonomous function of cGMP-dependent TAX-2/TAX-4 cation channels in AFD neurons. [54,55]We constructed a transgenic strain expressing mTRPV1 in AFD neurons in tax-2 mutant worms (tax-2; Pgcy-8::mTRPV1::sl2::Dsred, also shown as tax-2; AFD::mTRPV1).When performed NIR irradiation at 50 mW mm −2 power density, N2 worms fed with ICG showed a rapid retreat behavior, and the reversal rate increased with the concentration of ICG (Figure 4a-b).This result proved our NIR-ICG photothermal system can effectively induce thermal sensing retreat behavior.While there was no obvious behavioral response to the photothermal system in tax-2 mutant worms, which is consistent with previous research that tax-2 mutations lost the major heat sensing abilities.Excitingly, tax-2; AFD::mTRPV1 transgenic worms showed a rescued heat sensing phenotype, exhib-ited a ICG concentration-dependent (from 0 to 4 μg mL −1 ) photothermal reversal behavior as N2 worms (Figure 4a-b).What's more, both N2 and tax-2; AFD::mTRPV1 transgenic worms showed NIR power-dependent (from 0 to 16 mW mm −2 ) photothermal reversal behavior with ingested ICG concentration at 2 μg mL −1 , but not in tax-2 mutations (Figure 4a,c).These results showed that our NIR-ICG manipulation can effectively induce thermosensitive reversal behavior through the thermosensitive TRPV1 channel expressed in AFD sensory neurons in vivo.
[58] To further test if our NIR-ICG photothermal system can regulate the excitability of AFD sensory neurons through TRPV1, we co-expressed TRPV1 and GCaMP5 in AFD neurons in the background of tax-2 mutation (tax-2; Pgcy-8::mTRPV1::sl2::Dsred+Pgcy-8::GCaMP5, also showed as tax-2; AFD::mTRPV1+ GCaMP5), which enabled us to measure the intracellular calcium changes in AFD neurons.When we rapidly raised the temperature of the bath solution in which the C. elegans were incubated to ≈45 °C, the AFD neurons showed an increase in calcium signal.Interestingly, a large increase in calcium was also observed in AFD neurons upon NIR irradiation at 25 mW mm −2 power density with ICG concentration of 10 μg mL −1 .But there was no response when NIR was performed alone (Figure 4d-f).Compared with direct heating, it took a longer time for AFD neurons to increase the intracellular calcium to the peak value under NIR-ICG photothermal stimulation, although the peak values were almost equal (Figure 4d).The different dynamic characteristics suggested that photothermal manipulation can activate AFD neurons in a slower and gentler manner, indicating a better controllability in vivo.Together, these results demonstrated our NIR-ICG manipulation can effectively activate AFD sensory neurons and regulate thermosensitive reversal behavior through TRPV1 in vivo.

NIR Activate D-Class Motor Neurons to Reduce Thrashing Frequency in Swimming Behavior
Next, we examined if our NIR-ICG photothermal system can regulate the activity of motor neurons which did not have temperature sensitivity.We constructed a transgenic strain expressing mTRPV1 in D-class motor neurons (tax-2; Pttr39::mTRPV1::sl2::Dsred, also shown as tax-2; D-MNs::mTRPV1).In C. elegans, D-MNs are inhibitory motor neurons which can release GABA neurotransmitter to muscle cells, causing muscle relaxation along the body, then slow down or halt locomotion. [59,60]We performed a thrashing assay, wherein nematodes generated body bends in M9 solution.NIR illumination (793 nm, 12.5 mW mm −2 ) induced a significantly reduced thrashing frequency with ICG (1 μg mL −1 ) in tax-2; D-MNs::mTRPV1 animals.And the motor behaviors can resume after the light was removed, which showed that the photothermal manipulation did not cause any tissue damage (Figure 5a).Neither N2 nor tax-2 mutant animals showed obvious slower thrashing upon NIR stimulation with ICG, suggested that TAX-2 did not function in D-MNs, it was the exogenously expressed mTRPV1 which be activated by the photothermal stimulation.And applying NIR or ICG alone also did not slow down the swimming behaviors (Figure 5a-b).What's more, the D-MNs regulated thrashing had NIR intensity dependence and ICG dose dependence (Figure 5c-d).These results suggested that NIR irradiation can activate mTRPV1 expressed in D-MNs to excite the motor neurons through photothermal conversion material ICG, and then induce motor behavior changes mediated by the motor neurons.In conclusion, our photothermal system regulates not only the activity of thermosensory neurons, but also the activity of motor neurons and related behaviors.

NIR Photothermal Modulation of Muscle Cells and Physiological Behaviors in C. elegans
So far, our NIR-ICG photothermal manipulation has been effectively applied in nervous system in vivo, then we wondered if it could also be used in other systems to manipulate the excitability of other excitable cells besides neurons in freely behaving animals.To explore the possibility, we expressed mTRPV1 in muscle cells driven by muscular system specific promoter Pmyo-3 (Pmyo-3::mTRPV1::sl2::Dsred).There are two main types of muscle in C. elegans: multiple sarcomere/obliquely striated muscles  and nonstriated muscles. [61]The 95 striated body wall muscle cells are the functional equivalents of vertebrate skeletal muscles.First, we examined the effects and efficacy of the NIR-ICG photothermal system on body wall muscles.Other studies have shown that activating body wall muscles with drugs or optogenetic tools will cause muscle contraction and induce nematodes to be stiff and paralyzed.As we expected, our photothermal manipulation (200 mW mm −2 NIR with 10 μg mL −1 ICG) could also induced stiffness and paralysis of C. elegans by excited the body wall muscles.And the mTRPV1 expressed worms (also shown as BM::mTRPV1) had a much higher ratio of paralysiscompared with N2 worms(Figure 6a).At the same time, the BM::mTRPV1 worms took significantly shorter time to be stiff respond to NIR irradiation (Figure 6b).These results proved the effectiveness of our photothermal system in activating body wall muscles in vivo.
Next, we examined its modulatory effect on another type of muscle, the vulva muscles, which belong to the nonstriated muscles and directly regulate egg-laying behavior in C. elegans.A young adult hermaphrodite generally has a store of 10-15 eggs in its uterus at any given time.Increasing the cellular excitability of the vulva muscles can result in muscle contractions, open the vulva and compress the uterus, then allow the eggs to be expelled into the environment. [62]In our previous study, we successfully used Channelrhodopsin-2 (ChR2) optogenetic tools to activate vulva muscles and induced nematodes egg-laying. [63]Interestingly, the egg-laying behavior driven by vulva muscles also be triggered by NIR-ICG photothermal system (50 mW mm −2 NIR with 7.5 μg mL −1 ICG) in mTRPV1 expressed worms (also shown as VM::mTRPV1), but not in N2 worms (Figure 6c-d).And the egg laying could not be stimulated by ICG alone (Figure 6d).So, our photothermal system can modulate not only straited muscles, but also nonstriated muscles in vivo.
Collectively, we successfully modulated two types of neurons (AFD sensory neurons and D-MNs motor neurons) and two types of muscles (BM straited muscles and VM nonstriated muscles) through our photothermal system in freely moving animals.As far as we know, it was the first time to achieve precisely photothermal modulation of different types of excitable cells and corresponding behaviors in different systems in vivo.

Biological Safety of ICG
ICG is an FDA-approved photothermal agent.It has been largely used in fluorescent imaging and photothermal therapy.In order to widely apply our photothermal manipulation in mammals or other experimental systems, we tested the biological safety of ICG.We incubated nematodes with different concentrations of ICG (up to 10 μg mL −1 ), then counted the survival rate of the worms after 24 h of feeding.In our experiments, almost all the worms were still alive with ingested ICG (Figure S5a, Supporting Information).To examine the long-term toxicity of ICG, we compared the lifespan and reproduction of C. elegans cultured with or without ICG.When worms were continually fed with ICG, the lifespan of adults showed no significant difference compared with the control animals.Interestingly, worms fed with ICG at a concentration of 10 μg mL −1 even had a slight increase in lifespan (Figure S5b, Supporting Information).The reproduction was examined by measuring the number of eggs.We found the nematodes did not exhibit any egg-laying defects when fed with different concentrations of ICG (0, 2, 10 μg mL −1 ).The number of eggs accumulated in uterus in D2 worms, which represented the ability of egg-laying behavior was almost equal (Figure S5e, Supporting Information).And the number of eggs laid per day also exhibited no difference (Figure S5c, Supporting Information), showing that ICG did not affect the brood size of the worms (Figure S5d, Supporting Information).These results all demonstrated the high biological safety of ICG.
Taken all together, we developed a 793 nm NIR light irradiation photothermal manipulation system with ICG to regulate cellular excitability in different systems through thermosensitive TRPV1 channel, and mediated corresponding physiological behaviors in vivo (Figure 7).

Discussion
Non-invasive neuromodulation techniques have attracted much attention in neuroscience research and disease therapy in recent years.Photothermal modulation offers a new approach to this field.However, due to the lack of research in vivo, most studies and applications were limited to the cellular level.Our study first achieved precise photothermal modulation of different types of excitable cells and related behaviors in freely moving animals with a non-invasive manner.
In our study, by activating the thermosensitive channel TRPV1 under NIR irradiation with photothermal agent ICG, which has excellent photothermal conversion efficiency and biosafety, we successfully achieved in vivo precise regulation of the cellular excitability in both nervous and muscular systems, and mediated a variety of related physiological behaviors.In order to fully demonstrated the effectiveness of our photothermal manipulation, we performed multiple experiments including animal behaviors, in vitro and in vivo calcium imaging, wide-field near-infrared fluorescence imaging, two-photon fluorescence imaging, confocal imaging, heat shock assay, biosafety test, and pharmacological assay.After demonstrated TRPV1 expressed HEK293T cells could be activated by NIR stimuli with photothermal agent ICG with a calcium response, we applied our photothermal modulation in C. elegans for in vivo detection.In our test, ICG showed widely distribution in C. elegans after 12 h of feeding, and performed photothermal conversion response upon NIR stimuli to induce the expression of heat shock protein HSP-4 in vivo.And then we successfully modulated two types of neurons (AFD sensory neurons and D-MNs motor neurons) and two types of muscles (BM straited muscles and VM nonstriated muscles) through NIR stimulation with ICG.We specifically excited these different types of excitable cells and induced reversal locomotion, slower swimming, paralysis and egg-laying behaviors, respectively.In addition, we tested the biosafety of ICG for survival, reproduction and lifespan, and found no negative effect.We developed a noninvasive photothermal manipulation system to regulate cellular excitability in different systems and mediated related physiological behaviors in vivo.
The wavelength of NIR we used in our experiments was 793 nm, which belongs to the range of NIR-I, due to ICG showed the highest absorption peak at this value (Figure 1a).Compared to visible light, NIR-I light exhibits a deeper tissue penetration.Compared to NIR-II light, the NIR-I light shows less water absorption, non-specific thermal effect and hyperthermia effect.
However, it cannot be avoided that NIR-I light possessed relatively lower tissue penetration depth compared to NIR-II, which might hinder deep NIR photothermal manipulation.In our in vivo experiments, the minimum NIR light intensity used was only 4 mW mm −2 , and the concentration of ICG just ranged from 0.2 μg mL −1 to 10 μg mL −1 .Both the NIR light intensity and ICG concentration we applied were at a very low level, almost the lowest values used for cellular activity modulation in vivo.The low light power density and low agent concentration requirement demonstrated the sensitivity and effectiveness of our photothermal manipulation.
There are several NIR stimulated photothermal materials have been synthesized and studied.Such as the semiconducting polymer nanoparticles have been applied to target TRPV1-expressed neuron cells response to NIR stimuli in vitro, or as a cargo encapsulating TRPV1 agonist capsaicin in cancer therapy. [64,65]The gold nanoparticles packed microcapsules were used to regulate Wnt signaling pathway upon NIR stimuli in Hydra. [37]And Cu2x Se-anti-TRPV1 nanoparticles were used to target the microglia and open TRPV1 channels under NIR irradiation to promote phagocytosis and degradation of -syn in PD model. [66]However, the biosafety of these inorganic or metal nanomaterials still needs to be considered and further tested, which resulting in the research and application of these materials still be limited to the cellular level in vitro.In our study, the photothermal conversion material ICG showed excellent photothermal conversion efficiency compared to these materials, as well as excellent photostability in FBS solution (Figure 1b).ICG showed higher photothermal stability when binding to FBS, because ICG could bind to the hydrophobic region of protein, avoiding aggregation induced quenching effect. [67,68]Due to the weak photodynamic character of pure ICG, we didn't observe ICG induced ROS production in C. elegans.Meanwhile, ICG exhibited high biocompatibility and high biosafety performance.Long term continuously fed by ICG had no negative effects on survival, reproduction and lifespan in C. elegans (Figure S5, Supporting Information).What's more, worms fed with ICG at a concentration of 10 μg mL −1 even had a slight increase in lifespan.As an FDA-approved agent, ICG has already been widely used in angiography and photothermal treatment of tumor.And clinical studies have shown that ICG has a short life time in blood circulation, which can be metabolized by the livers.Due to its excellent photothermal conversion efficiency and biosafety, ICG exhibited its advantages in photothermal regulation in vivo.
Recently, some other neural activity modulations by NIR were also performed in behaving animals.Gao's group has applied NIR to manipulate the upconversion nanoparticles (UCNPs) for motor regulation and neuron ablation in C. elegans.NIR at 808 nm stimulated Er-UCNPs and specifically controlled Chrimson expressed DVC interneuron and D-class motor neurons. [32]n addition, they used orthogonal emissive UCNPs excited by dual-NIR illumination (808 and 980 nm) to activate the optogenetic sensor BiPOLES expressed cholinergic motor neurons, which enabled bidirectional control of motor behaviors. [69]And they also developed an 808 nm NIR light-induced neuron ablation method by simultaneously activating miniSOG and Chrimson based on UCNPs. [70]Another work needs to be mentioned is made by Hong and Pu's group, TRPV1 ectopically expressed neurons in the hippocampus, motor cortex and ventral tegmental area of mice can be activated upon NIR-II illumination at 1064 nm, with macromolecular transducers consisting of a semiconducting polymer core and an amphiphilic polymer shell. [71]owever, in both works, only the neurons were modulated.In our study, in addition to in vivo manipulate neural activity and motor behaviors in two different types of neurons, the AFD thermosensitive neurons and the D-class motor neurons, we also successfully performed the NIR photothermal modulation in muscular system.Both the straited muscles (body wall muscle cells, functional equivalent to vertebrate skeletal muscles) and nonstriated muscles (vulva muscles) can be excited by NIR stimuli, induced paralysis and egg-laying behaviors respectively, which verified a broader application of our photothermal manipulation.So far, our approach appears to be the first photothermal modulation method to regulate different types of excitable cells and corresponding behaviors in vivo.In clinical treatment for neurodegenerative diseases and movement disorders (Parkinson's disease, Alzheimer's disease, Huntington's disease, amyotrophic lateral sclerosis, et al.), as well as in sports rehabilitation trainings, direct regulation of neurons often cannot achieve desired therapeutic effect, due to the complexity of neural circuits.Our photothermal manipulation can also regulate muscle cells directly in vivo, providing a new approach for clinical treatment.And the efficacy and specificity of our photothermal manipulation suggest it may also be applied in some other type of excitable cells or in other experimental systems.74] In conclusion, this study demonstrates a noninvasive and effective method to precisely regulate the excitability of different types of cells and mediate related animal behaviors in vivo by NIR photothermal manipulation.The approach appears to be the first photothermal modulation method to be applied in both nervous and muscular systems in freely moving animals with a noninvasive manner.And the manipulation system shows excellent effectiveness, specificity and biological safety, suggesting a wide range of application prospect.
NIR Fluorescence Microscopic Imaging In Vitro: NIR-I fluorescence microscopic imaging of SY5Y neuroblastoma cells were operated on confocal microscopy (STELLARIS 8 FALCON, Leica).NIR-II wide-field fluorescence microscopic imaging of HEK293T cells were operated on NIR-II-MS optical system (Sunnyoptical).The cells were incubated with ICG (50 μg mL −1 ) in HBSS solution for 1 h.The extra ICG was washed away or just left while imaging.
Confocal Imaging of C. elegans: The NaN 3 anesthetized young adult C. elegans were immobilized on a glass slide with an agar pad.After cover the agar pad with a coverslip, the fluorescence signals were recorded using Z axis scanning mode by a confocal microscope (FV1000, Olympus) with a Plan-Apochromatic 20× (Figure 2a,c) or 60× (Figure S4, Supporting Information) objective.
Two-photon Fluorescence Imaging of C. elegans: Worms were fed with ICG (0, 0.2, 1, 2 μg mL −1 ) for 12 h, then anesthetized and immobilized as described before.The images were taken by a two-photon fluorescence scanning microscope (Olympus, FV1200).A 1040 nm femtosecond (fs) laser (150 fs, 50 MHz) was introduced as excitation source.The signal of DsRed fluorescent probe was recorded by inner photomultiplier tube (PMT), while the fluorescence of ICG was collected by an outer photomultiplier tube (PMT, H7422-50, Hamamatsu) through NDD mode using a 750 nm long-pass filter.
Calcium Imaging: HEK293T cells were incubated with calcium indicator Fluo-8am for 30 min at 37 °C in the dark.Cells transfected with mTRPV1-mCherry or mCherry only were used to detect the fluorescence intensity changes under heat, capsaicin or NIR treatments.The heating process was controlled by perfusing preheated bath solution at 45 °C.And then 10 μM capsaicin was given after 5.5 min.The optical fiber was used to lead the light source of NIR stimulation (793 nm, 10 mW mm −2 ) to detect photothermal effect.HEK293T cells in ICG group were incubated with 50 μg mL −1 ICG for 1 h.After washing twice with HBSS, Fluo-8am was added for incubation.The extracellular solution (ECS) for cell calcium imaging contained 140 mM NaCl, 5 mM KCl, 1.2 mM MgSO 4 , 1.5 mM CaCl 2 , 10 mM HEPES, 10 mM D-glucose, and pH 7.4.
For calcium imaging in vivo, AFD::GCaMP5 expressed young adult C. elegans were immobilized with surgical glue (Gluture Topical Tissue Adhesive, Abbott Laboratories) on a glass pad, covered with bath solution (145 mM NaCl, 2.5 mM KCl, 1 mM MgCl 2 , 5 mM CaCl 2 , 10 mM HEPES, 20 mM glucose, 325≈335 mOsm, pH 7.3).Heat stimulation was performed by perfusing preheated bath solution at 45 °C.The worms in the NIR+ICG group were fed with 10 μg mL −1 ICG for 12 h and covered in a bath solution with 10 μg mL −1 ICG when recording.The optical fiber was used to lead the light source of NIR stimulation (793 nm, 10 mW mm −2 ).Calcium imaging was acquired on an inverted microscope (Olympus IX71) with a 40× objective lens.Raw images were acquired with a QImaging optiMOS Scientific CMOS (sCMOS) camera controlled by the micro-Manager 1.4 software.All image stacks were analyzed using the ImageJ software.
Optical System for Photothermal Manipulating of C. elegans Behaviors In Vivo: The optical system was based on a stereomicroscope (SZN71 Sunnyoptical).793 nm CW laser was focused on the plate through a focus lens (focus length = 160 mm) and control the spot diameter of 2 mm or as needed.The behavior changes of C. elegans was recorded by a VIS CCD (RETIGA-SRV, QImaging) with a white LED illumination.
Heat Shock of C. elegans: A total of 100 adult worms of YmIS6[Phsp-4::GFP] were placed on a seeded NGM plate to lay eggs for 2 hours, and the eggs were cultured at 20 °C until the young adult stage.For the heat group, enough nematodes were transferred to fresh NGM plates.Then the plates were heated in water bath at 33 °C for 0, 1, 10, 30 min, respectively.For the NIR+ICG group, nematodes were incubated on fresh NGM plates with ICG (10 μg mL −1 ) for 12 h to get the young adult stage.Then transferred enough nematodes to M9 droplets with ICG (10 μg mL −1 ) on glass pads.793 nm CW laser was focused on the droplets for 0, 1, 5, 10 min, respectively.Fluorescence intensity was detected 2 h after treatment.Worms were anesthetized and immobilized as described before.Imaging was performed using an upright fluorescence microscope (Olympus, BX51WI) with a 5× objective lens (Figure 2d,e).Raw images were acquired with a QImaging optiMOS Scientific CMOS (sCMOS) camera controlled by the micro-Manager 1.4 software.All image stacks were analyzed using the Im-ageJ software.
Reversal Assay: After fed with different concentrations of ICG for 12 h, young adult C. elegans were transferred to a fresh test plate without Escherichia coli food on it.After a few minutes of acclimatization, NIR irradiation (793 nm) was performed with different power densities as stated.The NIR spot with a diameter less than 2 mm was placed in front of the C. elegans when the worm moved forward.The response behaviors when worms moved into the NIR spot were recorded and classified as go forward and go backward (reversal behavior).A stereoscope (SZN71 Sunnyoptical) equipped with a high-sensitivity digital CCD camera (RETIGA-SRV-Qimaging, Fast1394) was used for recording.
Thrashing Assay: M9 solution with different concentrations of ICG were dropped (50 μL each) on glass pads.Young adult nematodes fed by the equal concentration of ICG were transferred into the droplets to acclimate for ≈1≈2 min.Then thrashing behaviors were recorded by a digital camera.For NIR irradiation experiments, animals were first recorded for 20 s as a baseline value, then illuminated under 793 nm NIR light for 20 s to detect the effect of photothermal stimulation, finally recorded for another 20 s after removing NIR to examine the recovery of swimming behaviors.
Paralysis Assay: Young adult worms fed by ICG (10 μg mL −1 , 12 h) were transferred to a fresh plate without E. coli food on it.After a few minutes of acclimatization, the NIR (793 nm, 200 mW mm −2 ) light was placed onto the body of the nematodes.The motor behaviors upon NIR were recorded and classified as movement and paralysis.And the NIR irradiation time (within 10 s) required for animals to be stiff was also be counted.
Egg-Laying Assay: For the egg-laying behavior test under photothermal manipulation, nematodes in D2 stage which have been fed by 7.5 μg mL −1 ICG were transferred to fresh NGM plates with ICG.After a few minutes of acclimatization, laid eggs by each worm before or during NIR irradiation (793 nm, 50 mW/mm 2 ) were recorded by a digital camera (Figure 6c-d).
For biological safety test of ICG, the effects on reproduction were examined by counting the number of eggs laid per day by each worm in the whole life, and the unlaid eggs in D2 animals, respectively.Worms were incubated on NGM plates with different concentrations of ICG (0, 2, 10 μg mL −1 ), single worm on single plate.Then the worms were transferred to fresh ICG plates every day after D1 stage, until the egg-laying physiological behavior was over.The number of eggs laid per day was recorded and manually counted (Figure S5c, Supporting Information), and the number of progeny was calculated by accumulating all data (Figure S5c, Supporting Information).For egg retention assays, individual D2 worm was dissolved in a 10 μL droplet of 50% NaOCl solution and the number of retained eggs was counted after 5 min.
Lethality Assay: ICG at different concentrations was added to NGM plates mixed with E. coli.After dry for 1-2 h at room temperature, the L4 larva were allowed to transfer onto the plates.Number of dead worms was recorded after 24 h incubation and the viability rate of C. elegans was estimated after 3 times repetition.
Lifespan Analysis: For biological safety test of ICG, the effects on lifespan were performed by counting the number of survival nematodes cultured on NGM plates at 20 °C with different concentrations of ICG (0, 2, 10 μg mL −1 ) until all tested animals were dead.Every 15≈20 N2 worms at L4 stage were cultured on a 6 cm NGM plate with ICG, and needed to be transferred to a fresh plate every day in egg-laying period.The survival number was recorded from D1 stage.And the nematodes lost or died accidentally due to crawling off or drilling into the NGM plates were removed out from the test.The number of valid worms was more than 110 in each group.
Quantification and Statistical Analysis: All statistical tests in this study were performed using GraphPad Prism 8. Student's t test, or one-or twoway ANOVA were performed to determine statistical significance.If the variances were significantly different, the Welch's t test or Brown-Forsythe and Welch ANOVA tests were used.P values below 0.05 were considered significant.Data were presented as mean ± SEM.Details of the statistical analyses performed for each figure were provided in the figure legends.

Figure 1 .
Figure 1.Optical and photothermal characterization of ICG.a) The absorption and fluorescence spectra of ICG.b) The temperature oscillation of 100 μg mL −1 ICG in FBS or ultrapure water, within five irradiation-cooling cycles (793 nm, 10 mW mm −2 ).c) 793 nm NIR-induced photothermal effect of ICG.Data shown as mean ± SEM, n = 3. d) The temperature change of ICG with different concentration when reached the final plateau under NIR irradiation, as in (c).Data shown as mean ± SEM, n = 3. ns, not significant (p > 0.05), * p < 0.05, **** p < 0.0001 by one-way ANOVA.

Figure 3 .
Figure 3.The distribution of ICG ingested by feeding and the expression of mTRPV1 in specific tissue cells in C. elegans.Representative two-photon fluorescence images of ICG distribution (gray) after feeding (2 μg mL −1 , 12 h) and mTRPV1-sl2-DsRed (red) expression in C. elegans.a) tax-2(p691) was imaged as a control group.mTRPV1-sl2-DsRed were specifically expressed in AFD sensory neurons b), D-class motor neurons c), body wall muscles and vulva muscles d), respectively.Scale bar = 100 μm (a,c), 20 μm (b), 25 μm (d).White arrows, the position of tissue cells.Yellow line in Merge group, the position where fluorescence intensity was determined.Right panel in each row showed fluorescence intensity value along the line (yellow, in merge group) and the distances between mTRPV1 and ICG in each group.

Figure 7 .
Figure 7. NIR modulates C. elegans cellular excitability and animal behaviors through TRPV1 with ICG.NIR light (793 nm) irradiation can activate the thermosensitive TRPV1 channel through photothermal agents ICG to precisely regulate the excitability of different types of cells and related behaviors in C. elegans in vivo.Mouse TRPV1 channels are expressed exogenously in AFD sensory neurons, D-class motor neurons (D-MNs), body wall muscles and vulva muscles.In the presence of ICG, NIR illumination effectively activate TRPV1, cause Ca 2+ influx, lead to increased excitability of tissue cells, and then induce reversal, slower thrashing, paralysis and egg-laying behaviors, respectively.