Optogenetic-induced Sympathetic Neuromodulation of Brown Adipose Tissue Thermogenesis

Abstract

The brown adipose tissue (BAT) is a thermogenic organ that plays a major role in energy balance, obesity and diabetes due to the potent glucose and lipid clearance that fuels its thermogenesis, which is largely mediated via sympathetic nervous system activation. However, thus far there has been little experimental validation of the hypothesis that selective neuromodulation of the sympathetic nerves innervating the BAT is sufficient to elicit thermogenesis in mice. We generated mice expressing blue light-activated channelrhodopsin-2 (ChR2) in sympathetic nerves innervating the BAT using two different strategies: injecting the BAT of C57Bl/6J mice with AAV6-hSyn-ChR2(H134R)-EYFP; crossbreeding tyrosine hydroxylase-Cre mice with floxed-stop ChR2-EYFP mice. The nerves in the BAT expressing ChR2 were selectively stimulated with a blue led light positioned underneath the fat pad of anesthetized mice while the BAT and core temperature simultaneously recorded. Using immunohistochemistry we confirmed the selective expression of EYFP in TH positive nerves fibers. Additionally, local optogenetic stimulation of sympathetic nerves induced significant increase in BAT temperature followed by an increase in core temperature in mice expressing ChR2, but not in the respective controls. The BAT activation was also paralleled by increased levels of pre-UCP1 transcript. Our results demonstrate that local optogenetic stimulation of the sympathetic nerves is sufficient to elicit BAT and core thermogenesis, thus suggesting that peripheral neuromodulation has the potential to be exploited as an alternative to pharmacotherapies to elicit organ activation and thus ameliorate type-2-diabetes and/or obesity.

1. Introduction

The brown adipose tissue (BAT) is a thermogenic organ that plays a major role in energy balance through potent glucose and lipid clearance required to fuel its thermogenesis (1–3), and has been demonstrated to be present and functional in adult humans (4–8). BAT activation is regulated by central nervous system neural networks which respond reflexively to thermal afferent signals from skin thermoreceptors (9). These networks mount an activation of the sympathetic nerves that innervate the BAT, classically in response to either cold acclimation, acute cold exposure (1, 2) or stress (10–12). Brown adipocyte activity is primarily regulated by the sympathetic nerve-derived norepinephrine (NE) signaling through β-adrenergic receptors (βARs) which results in activation of uncoupling protein 1 (UCP1) and consequent conversion of energy to heat (1, 9, 13, 14). Denervation leads to a dysfunctional BAT composed of UCP1-depleted adipocytes, low thermogenesis, cold intolerance, and vulnerability to obesity (1, 9, 15, 16), while electrical nerve stimulation elicits a thermogenic effect (9, 10). Adult humans reportedly experience only a modest basal activation of the BAT due to multiple factors, including higher body weight:surface area ratio, predominant exposure to thermoneutral comfort zone, low sympathetic tone to fat pads and low NE/βAR signaling compared to rodents (17, 18). In humans, BAT activity or amount is inversely associated with obesity and functional measures of diabetes (4, 17), thus suggesting that targeting BAT activation could be a viable therapeutic approach to address obesity (5, 19, 20). Based on the large body of evidence demonstrating a necessary and sufficient role for sympathetic nerves to regulate BAT differentiation and function (1, 9), and recent studies suggesting that optogenetic stimulation of brown and white adipose tissue can regulate organ function (21, 22), we used two independent approaches in which expression of the blue light-activated channelrhodopsin-2 (ChR2) is restricted to sympathetic neurons by either crossbreeding tyrosine hydroxylase (TH)-Cre mice with Rosa26-LSL-ChR2-EYFP (22), or by intra BAT injection of a retrograde AAV6-hSyn-ChR2-EYFP virus.

2. Material and Methods

2.1. Animals and experimental models

C57BL/6J male mice were purchased from Jackson Laboratories at 4–5 weeks of age. We generated TH-Cre X Rosa26-LSL-ChR2-EYFP mice as previously established (22) (referred to as ChR2^TH-Cre+ or ChR2^TH-Cre-). Founder mice of either strain (TH-Cre B6.Cg-Tg(Th-cre)1Tmd/J, JAX #008601 and Rosa26-LSL-ChR2-EYFP B6.Cg-Gt(ROSA)26Sor/J, JAX #024109) were purchased from Jackson labs in 2017. Our breeding scheme aimed at establishing the Rosa-LSL-ChR2-EYFP locus in homozygosity. Mice derived from this breeding were used in the next ~12 months in the experiments described below. Both males and females TH-Cre X Rosa26-LSL-ChR2-EYFP mice were used and, after confirming that sex did not affect the physiological effect, data from male and females were combined for each experimental treatment. Mice were housed at room temperature with 12-h light:dark cycle. All experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Minnesota.

2.2. Viral injections

C57BL/6J male mice were injected with optogenetic vectors under aseptic conditions. Mice were initially anesthetized with a ketamine/xylazine mixture (100 and 10 mg/kg, i.p.) and were given supplemental ketamine (~ 33 mg/kg, i.p.) to maintain them at an appropriate anesthetic depth for the duration of the surgery. Mice were injected with either AAV6-hSyn-ChR2(H134R)-EYFP or AAV6-hSyn-EYFP control virus (University of Minnesota Viral Vector Core; maps are presented in Suppl. Fig 1)) into the brown adipose tissue. The AAV6 serotype was chosen due to its documented ability to be taken up by neuronal terminals and transported in a retrograde fashion (23, 24). Each side of the tissue was injected with 3.0 µl of virus in 3–6 injection sites in order to maximize the spread of virus throughout the tissue using an approach similar to the one developed to perform local sympathectomy with chemical agents (25). Injections were delivered at a rate of 0.1 µl/min and needles were allowed to sit for 5 minutes after the end of each injection to allow for viral diffusion prior to needle removal. Mice began optogenetic experiments at least 6 weeks following injection to allow for viral transport and sufficient levels of opsin expression in the targeted terminal regions. In the weeks between surgery and test, the AAV6-ChR2 infected mice gained 2.4±0.3 grams on average, while the control AAV6-EYFP mice gained 2.4 ± 0.08 grams on average, suggesting no adverse effect of the viral injection.

2.3. Optogenetic stimulation of brown adipose tissue

A custom-made implant was used to deliver blue light onto opsin-expressing adipose tissue. The implant was made using a LED (455 nm, 0805 SMD, Chanzon) that was encased in a thin layer of polydimethylsiloxane (PMDS) and coupled to a thermoprobe (source) to allow for temperature measurements of the brown adipose tissue close (~0.1–1.0 mm) to the field of blue light illumination (irradiance ~ 10 mW/mm2). Control mice not expressing the ChR2 were also exposed to blue light pulses to confirm changes in temperature observed were not due to non-specific light-induced heating.

For testing, mice were anesthetized under isoflurane (4% induction, 2 to 2.5% maintenance) and placed on a heating pad (33–35° C). An incision was made on the back to expose the adipose tissue and an opening was carefully made in the mid-caudal region of the brown adipose tissue to allow for probe placement underneath the organ while avoiding the damage of nerves and vasculature, with the light directed upward towards the bottom ventral surface of the tissue in order to maximize tissue penetration. The probe was secured into place, the fat pad was gently lowered on top of the LED, and the skin was secured closely around the probe exit to prevent tissue drying during the experiment. An additional thermal probe was placed in the rectum to allow for simultaneous measurement of core body temperature. Next, baseline measurements were taken once a minute until both BAT and core temperatures were stabilized (<0.2°C change over 10 minutes). Mice were then exposed to pulsed blue-light stimulation (5 ms pulse width, 20 Hz, 1 sec on / 1 sec off) as described in the result section. Both BAT and rectal temperature were recorder once a minute during the blue-light stimulation and data are presented as delta °C change over baseline values.

2.4. Tissue collection and immunohistochemical analysis

Mice were rapidly euthanized by cervical dislocation immediately after the post-stimulation period, and tissue was collected for further analysis. For BAT excision, each mouse was positioned prone facing down an incision was made with a pair of scissors at a 30° angle on the dorsal side along the midline from the neck to the lower abdomen. The skin was lifted carefully to expose the butterfly shaped BAT allowing to hold and lift the apical portion of the depot while carefully severing with scissors along the periphery of the entire area. Any superficial white adipose atop and around the butterfly was carefully removed. Half the tissue was snap frozen and stored at −80°C while half was fixed in 4% paraformaldehyde and processed for histological analysis (see below). Furthermore, an incision was made on the throat and muscle, bone, and connective tissue around the rib/clavicle area was dissected to expose the region where the brachiocephalic artery divides into the right common carotid and subclavian arteries. The sympathetic ganglia were located behind and below the subclavian artery at this division and could be identified by its star-shaped, pearly grey onion-like appearance. Once identified, the ganglion was held in place with forceps to be dissected with abundant surrounding tissue for anatomical reference.

Dissected stellate ganglia were immersion fixed in 4% PFA for 24 hours, then cryoprotected in 30% sucrose for 3 days. Tissue sections harvested from virus infected mice were obtained using a Leica CM 1900 cryostat and mounted onto Superfrost Plus microscope slides; they were washed two times in phosphate buffer (PB), then sections were incubated in 1:75 (vol/vol) normal donkey serum (Jackson Immuno Research; West Grove, PA) in PB for 20 minutes at room temperature to block nonspecific sites and then incubated overnight at 4 °C with a primary antibody against TH (AB1542 Chemicon Millipore, Milan, Italy; 1:300). The following day tissues were washed twice in PB for 15 minutes and then incubated with Alexa Fluor 568 nm (red) to visualize TH-positive adrenergic fibers. The tissue was finally counterstained with the nuclear dye TO-PRO3 633 nm (blue). Slides were mounted with Prolong Antifade (Invitrogen). Fluorescence was detected with a Leica TCS-SL confocal microscope (Leica Microsystems, Vienna, Austria) equipped with an Argon and He/Ne mixed gas laser. Dyes were imaged separately (1024×1024 pixels) and then merged. Images were obtained using a confocal pinhole of 1.1200 and stored as TIFF files. Brightness and contrast of the final images were adjusted using the Photoshop 6 software (Adobe Systems; Mountain View, CA, USA).

Brown Adipose tissue was immersion-fixed in modified Zamboni’s fixative (4% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer, pH 6.9), cryoprotected in 10% sucrose, and cryostat-sectioned into 30 mm section which were mounted on gel-coated slides. Sections were preabsorbed in blocking buffer (PBS containing 0.3% Triton-X 100, 1% BSA, 1% normal donkey serum) for 30 min, followed by incubation overnight at 4°C with appropriate primary antibodies: chicken anti-GFP (1:1000 Abcam, ab13970, Cambridge, UK), and sheep anti-TH (1:500, Millipore AB1542). After rinsing, the sections were then incubated for 1 hour at room temperature with appropriate secondary antibodies: Cy3 conjugated donkey anti-chicken and Alexa Fluor 488 conjugated donkey anti- goat secondary antibodies (1:1000, Jackson ImmunoResearch). Images were collected using an Olympus Fluoview 1000 imaging system, processed in FIJI (projection of optical sections and file format conversion) and adjusted for brightness and contrast in Adobe Photoshop CC.

To evaluate EYFP expression specificity exclusive from CD31, fresh snap-frozen brown adipose tissue was cryosectioned into 20μm sections and mounted onto Superfrost Plus microscope slides. The mounted sections were fixed in 4% PFA for 15 minutes then washed 3 times in PBS. Nonspecific binding was blocked in 2% normal donkey serum (NDS; Jackson Labs), 0.1% Triton-X 100 in PBS for 1 hour at room temperature then incubated with primary antibodies - sheep anti-TH (1:500, Millipore AB1542), rat anti-CD31 (Dianova, DIA-310) and rabbit anti-GFP (1:500, Invitrogen A11122) in blocking solution overnight at 4˚C. The following day the sections were washed 3 times with PBS, then incubated with secondary antibodies - donkey anti sheep Alexa Fluor 647 (1:500, Jackson Immuno Research), donkey anti rabbit 488 (1:500, Jackson Immuno Research), donkey anti-rat 568 (1:500; Jackson Immuno Research) in blocking solution for 1 hour at room temperature. After 3 washes in PBS, the slides were coverslipped with ProLong Diamond Antifade mountant (Invitrogen). The slides were imaged with a Nikon Eclipse microscope equipped with a 40X objective N.A.= 0.95, processed and adjusted for brightness and contrast with FIJI.

Adrenal glands harvested from transgenic mice were fixed overnight in 4% PFA then cryoprotected in 30% sucrose before they were cut into 10 µm cryosections and mounted onto Superfrost Plus microscope slides Tissue was rehydrated with 3 washes in PB. The tissue was blocked for 1 hour at room temperature in 2% normal donkey serum (NDS; Jackson Labs) in PB. They were then incubated with primary antibodies - sheep anti Tyrosine Hydroxylase (1:500, Millipore AB1542) and rabbit anti GFP (1:500, Invitrogen A11122) in 2% NDS in PB overnight at 4˚C. The following day the sections were washed 3 times in PB. Secondary incubation was for 1 hour at room temperature with donkey anti sheep Alexa Fluor 647 (1:200, Jackson Immuno Research) and donkey anti rabbit 568 (1:200, Jackson Immuno Research) in 2% NDS in 1X PB. After 3 additional washes in PB, the slides were coverslipped with ProLong Diamond Antifade mountant (Invitrogen). The sections were imaged with a Nikon A1R Confocal microscope, 20X objective N.A. = 0.75 (University Imaging Centers, University of Minnesota). FIJI was used to quantify colocalization of TH with EYFP.

2.5. Tissue clearing using the X-CLARITY system

The tissue was removed from the fixative, rinsed with 1× phosphate-buffered saline (PBS) several times and then incubated in 1× PBS at 4.0°C for 24 hours. Afterward the tissue was cleared using the X-Clarity System (Annandale, VA). In summary, the tissue was immersed in the X-CLARITY polymerization initiator solution and processed according to manufacturer’s instructions. The samples were incubated for 24 hours at 4.0°C, in the hydrogel monomer with accelerator and then polymerized for 3 hours at a vacuum (kPa) of −85, and a temperature of 37.0°C. The Tissue/gel was shaken for 1 minute and rinsed in 1× PBS. Subsequently samples were processed in the electrophoretic tissue clearing (ETC) solution for 20 hours (checked after 2 hours) at a current of 1.1 A, temperature of 37.0°C and pump speed of 40 rpm. Subsequently, samples were stored in 1× PBS at 4.0°C. In preparation for taking images, the samples were placed in X-CLARITY mounting solution (R.I. 1.46) for 2–4 hours. All fluorescent images were taken using the Ti-E microscope with A1R confocal Microscope using the Nikon 10X glycerol, 0.5 N.A., 5.0 mm WD objective (Nikon Instruments Melville, NY). Images were processed in Nikon Elements v4.6 or Imaris, 9.0.1 software (Bitplane, Oxford Instruments, concord, MA).

2.6. RT-qPCR

RNA from BAT was extracted using a PureLink RNA Mini Kit (Thermo Fisher Scientific) according to the manufacturer’s protocol. Transcription into cDNA was performed using the iScript cDNA synthesis kit (Bio-Rad) according to the manufacturer’s directions. Each cDNA sample was run in duplicate using the iTaq Universal SYBR Green PCR Master Mix (Bio-Rad) to a final volume of 15 µl in a CFX Connect thermal cycler and optic monitor (Bio-Rad). The expression of genes was normalized to the geometric mean of two housekeeping genes – actin and transcription factor II. Sequences of primers used were: actin (forward: 5’-GGCACCACACCTTCTACAATG-3’ reverse: 5’-GGGGTGTTGAAGGTCTCAAAC-3’), TFII (forward: 5’-TGG AGA TTT GTC CAC CAT GA-3’ reverse: 5’-GAA TTG CCA AAC TCA TCA AAA CT-3’), preUCP1 (forward: 5’-GGA TTG GCC TCT ACG ACT CA-3’ reverse: 5’-GGC CAG TTG GTT TTC ACA GA-3’) and UCP1 (forward 5’-GTC CCC TGC CAT TTA CTG TCA G-3’ reverse 5’-TTT ATT CGT GGT CTC CCA GCA TAG). Data was analyzed using the ∆∆Ct method.

2.6. Statistical analysis

All statistics were performed with Statistica software (Tibco). Data are expressed as mean +/- SEM unless otherwise specified. We first compared the response of males and females and, after excluding a significant sex effect, data were pooled according to the experimental treatment. Data were analyzed using repeated measures or one-way ANOVAs followed when significant by Fisher LSD post hoc tests for binary comparisons. The area under the curve was calculated using the trapezoid method. Differences were considered significant at p < 0.05.

3. Results

We engineered AAV6 to express the blue light-sensitive cation channel channelrhodopsin-2 (ChR2) fused to enhanced yellow fluorescent protein (EYFP), under the control of the pan-neuronal human synapsin-1 promoter (hSyn) (26). We injected retrograde hSyn-AAV6-ChR2-EYFP or hSyn-AAV6-EYFP control viruses in the BAT of C57BL/6J WT mice. Immunohistochemical analysis confirmed that AAV6-transfected TH-positive sympathetic nerves (23, 27, 28) and not adipocytes (29) in the BAT and sympathetic ganglia (Suppl. Fig. 2 A–I). In particular, co-localization of EYFP and TH staining was detected in correspondence of neuronal bodies within ganglia, and nerve fibers in the BAT (Suppl. Fig. 2 J–M).

Six weeks after the surgery, mice injected with AAV6-ChR2-EYFP or AAV6-EYFP or uninfected controls were anesthetized and stimulated with blue LED light with BAT and core temperature being simultaneously recorded (Fig. 1A). The LED and the thermal probe were positioned under the BAT taking care not to damage the nerves or the vasculature. Using a 30min light ON protocol we showed rapid and consistent rise in BAT and core temperature in ChR2 expressing mice but not in AAV6-EYFP or un-infected control mice (Fig. 1B–C,E–F), as highlighted by a statistically significant main effect of the virus during the stimulation time-course (BAT: F(2,13)=6.98, p<0.003 ; Rectal: F(2,13)=8.46, p<0.005) as well as of the area under the curve (BAT: F(2,13)=6.48, p<0.05 ; Rectal: F(2,13)=7.92, p<0.006). This stimulation protocol was also effective in inducing a significant increase in transcriptional activation of UCP1 (measuring of the nascent preUCP1 transcript (30, 31)) in ChR2 expressing mice when compared to controls (Fig. 1D). The optogenetic–induced thermogenic effect was confirmed in a second ON-OFF protocol with the duration of the rise in temperature corresponding to the duration of the tissue illumination (Suppl. Fig. 3A). To further corroborate this result we engineered the same construct described above using a second virus expressing tdTomato that led to a comparable rise of BAT temperature upon optogenetic stimulation (Suppl. Fig. 3B).

Figure 1. Brown Adipose Tissue (BAT) optogenetic stimulation in channelrhodopsin virally infected mice elicits thermogenesis.

A) Cartoon of the experimental procedure. BAT (B) and rectal (E) temperature change versus unstimulated preparation during 30 min (unstimulated average BAT temperatures +/- SEM: AAV6_ChR2_EYFP: 32.32 +/- 0.34°C; AAV6_EYFP: 32.8 +/- 0.30°C; No_virus: 30.81 +/- 0.69°C; unstimulated average core temperatures +/- SEM: AAV6_ChR2_EYFP: 34.24 +/- 0.81°C; AAV6_EYFP: 35.1 +/- 0.28°C; No_virus: 34.18 +/- 0.47°C). Cumulative BAT (C) and rectal (F) temperature change versus unstimulated preparation expressed as Area Under the Curve (AUC) during 30 min stimulation under anesthesia. D) preUCP1 RNA expression level from BAT collected at the end of the recording. AAV6-ChR2-EYFP, channelrhodopsin infected mice; AAV6- EYFP, virus control mice, No virus, uninfected mice. (N=5–7 males per group). The horizontal blue bar represents the duration of the LED stimulation. Data represent group averages +/- SEM. * indicates p < 0.05; ** indicates p < 0.01.

Overall, these data demonstrate that our viral construct can efficiently transfect sympathetic neurons in the BAT and that optogenetic neuromodulation of ChR2 expressing nerves is sufficient to elicit BAT activation. Importantly our two control groups exclude a role for the virus and the LED stimulation per se on BAT thermogenesis.

Next, we aimed to validate our viral-mediated approach by using TH-Cre X Rosa26-LSL-ChR2-YFP mice as previously established (22) (referred to as ChR2^TH-Cre+ or ChR2^TH-Cre-) (Fig. 2A). ChR2^TH-Cre+ mice express Cre-recombinase under a TH promoter in catecholaminergic neurons thus driving selective expression of the blue light-sensitive ChR2 and EYFP in the sympathetic fibers innervating the BAT (Suppl. Fig. 4). Colocalization of TH and EYFP in the chromaffin cells in the adrenal gland for positive control was confirmed using IHC (Suppl. Fig. 5). EYFP fluorescence was also confirmed using X-Clarity (Suppl. Fig 6 and Suppl. Video 1). Furthermore, EYFP does not colocalize with the endothelial marker CD31 and there was no EYFP expression in TH negative regions (Suppl. Fig. 4), thus confirming the specificity of TH-driven expression, in the present experimental condition. Next, ChR2^TH-Cre+ or ChR2^TH-Cre- mice were tested in our acute optogenetic preparation with 30 min LED ON:OFF. Consistent with the data obtained with AAV6-ChR2 infected mice, optogenetic stimulation of the sympathetic nerves elicited BAT and Core thermogenesis in ChR2^TH-Cre+ but not in ChR2^TH-Cre- mice (Fig. 2). Specifically, 30 min optogenetic stimulation gradually and significantly elevated both BAT and rectal temperature in ChR2^TH-Cre+ mice, as indicated by the shape of the trajectory of the increase (BAT: F(1,9)=11.95, p<0.008; core: F(1,9)=5.98, p<0.05) (Fig. 2 C,D), as well as by the cumulative degree of increase for the whole 30 min stimulation (BAT: F(1,9)=11.82, p<0.008; core F(1,9)=6.18, p<0.05) (Fig. 2 E–H). After the LED was turned OFF, the BAT temperature trended to decrease while, as expected (32), the core temperature continued to rise (Fig. 2A,D). Finally, optogenetic excitation of the sympathetic nerves corresponded to a significant increase of the expression of preUCP1 RNA in ChR2^TH-Cre+ when compared to ChR2^TH-Cre- mice (p<0.05) (Fig. 2B). Consistent with previous data in cold-exposed mice (31, 33), 30 minutes optogenetic nerve stimulation was not sufficient to increase the expression of mature UCP1 mRNA (ChR2^TH-Cre-=1±0.28, ChR2^TH-Cre+ =0.54±0.12, NS). The 30 minutes optogenetic-induced increase in BAT and core temperature was larger in ChR2^TH-Cre+ when compared to the AAV6-ChR2 mice from the previous experiment (Suppl. Fig. 3C,D). Overall, our second strategy, where the expression of the genetic construct was driven in sympathetic neurons through breeding targeted genetic mouse lines, confirmed that optogenetic excitation of ChR2-expressing neurons is sufficient for BAT thermogenesis, where none was elicited in non-carrier control mice.

Figure 2. Brown Adipose Tissue (BAT) optogenetic stimulation in mice genetically expressing channelrhodopsin in sympathetic nerves elicits thermogenesis.

A) Cartoon of the experimental procedure. B) Optogenetic excitation increased preUCP1 RNA in ChR2^TH-Cre+ mice. N=5–7, half male half females per group. C) BAT and (D) rectal temperature change versus unstimulated preparation during 30 min stimulation and 30 min following the end of stimulation (unstimulated average BAT temperatures +/- SEM: ChR2-ThCre-: 30.9 +/- 0.86°C; ChR2-ThCre+: 31.4 +/- 0.53°C; unstimulated average core temperatures +/- SEM: ChR2-ThCre-: 33.6 +/- 0.49°C; ChR2-ThCre+: 33.6 +/- 0.52°C). Cumulative BAT (E) and rectal (G) temperature change versus unstimulated preparation expressed as Area Under the Curve (AUC) during 30 min stimulation under anesthesia. BAT (F) and Rectal (H) temperature change versus unstimulated preparation at the end of the stimulation phase. (N=4–7). The horizontal blue bar in C and D represents the duration of the LED stimulation. Data represent group averages +/- SEM. * indicates p < 0.05, ** indicates p < 0.01.

4. DISCUSSION

Using two independent approaches, we demonstrated that optogenetic neuromodulation is sufficient to elicit BAT thermogenesis and stimulate expression of pre-UCP1 RNA. More broadly, this research contributes to the field of optogenetic manipulation of the peripheral nervous system (PNS), which as a whole is far less developed than that of the central nervous system (CNS) (21, 22, 34).

For opsin virus mediated transfection, we used AAV6, which has been used for gene delivery through retrograde transport both in the periphery (35) and in the brain (24), and is of translational relevance since AAV vectors belong to the first gene therapy products approved by FDA and are currently tested in clinical trials for human diseases (36, 37). We engineered AAV6 to express the ChR2 fused to the fluorescent reporter EYFP, under the control of hSyn to restrict transgene expression from an adenoviral vector exclusively to neurons (38). In our hands, this strategy was successful as demonstrated by the specific staining of sympathetic neurons, both in cell bodies and in fibers proximal to the brown adipocytes (22, 39, 40). Furthermore, we functionally characterized the thermogenic response evoked in ChR2 infected mice by optogenetic stimulation both at the BAT and the systemic level. This result is further evidenced by the significant induction of the preUCP1 RNA (the nascent UCP1 transcript (30, 31)) in ChR2 infected mice when compared to controls. One limitation of our present study is that experiments were conducted under general anesthesia, which is known to cause hypothermia (41). This choice of gas anesthesia was consistent with previous protocols (21, 42) as well as the need to have a physiologically stable experimental preparation over an extended period of time, without the need for periodic re-dosing as required with chemical anesthetics.

It must be noted that the application of optogenetic stimulation of nerves outside the CNS has been limited due to intrinsic properties of the PNS. Opsin expression in the PNS can be reduced by immune responses induced by viral vector injection, a consequence of the higher degree of immune surveillance in the PNS compared to that of the CNS (26, 43). Furthermore, the distribution of excitable cells is sparsely scattered throughout the target tissue, necessitating broader reach of illumination, which can also be impaired by the opaqueness of the target tissue (34). To overcome these limitations and to aim at a more efficient optogenetic stimulation of the BAT, we adapted a previously validated approach to generate a genetic mouse model expressing the opsin under the control of a TH-Cre promoter that limited the opsin expression to the sympathetic fibers innervating the BAT adapting (21, 22). Many Cre-driver lines are still not fully characterized in the PNS. Thus, one additional value of the characterization we conducted of our ChR2^TH-Cre+ model is to elucidate the features of opsin expression in the sympathetic nerves innervating peripheral organs, but not CD31 positive cells, expanding on what already demonstrated for other adipose depots (21, 22). For example, a combination of intravital two-photon microscopy and local optogenetic stimulation of sympathetic inputs illustrated on how sympathetic nerve fibers establish neuro-adipose junctions, directly encasing adipocytes and locally induce lipolysis and loss of white adipose mass (22). Previous data showed that optogenetically induced thermogenesis was dependent on intracellular glycolysis, identifying glucose as the substrate that fuels acute BAT function (21). In line with these results, the optogenetic stimulation of the sympathetic nerves led to appreciably rapid temperature rises both at the BAT and at the systemic level in Ch2R^TH-Cre+ and AAV6-ChR2 infected mice while it failed to do so in Ch2R^TH-Cre- or the two controls groups used for the virus study. Consistent with sympathetic-mediated brown adipocyte activation in both virally-injected and genetically induced models, optogenetic excitation of the sympathetic nerves increased the expression of preUCP1 RNA in ChR2-expressing mice when compared to controls.

5. CONCLUSION

In conclusion, our experimental approach validated the use of retrograde viral transfection and of genetically-driven expression of genetic constructs to sympathetic nerves innervating BAT. We also demonstrated that targeted and selective optogenetic activation of the sympathetic nerves expressing the opsin is sufficient to elicit selective BAT and core thermogenesis, as well as transcriptional regulation of UCP1. Our results add to the existing literature supporting the notion that peripheral neuromodulation could be therapeutically exploited to manipulate the BAT potent catabolic function as an alternative therapeutic option for the treatment of obesity and obesity-associated diseases.

NONSTANDARD ABBREVIATIONS LIST

ChR2

channelrhodopsin-2

EYFP

enhanced yellow fluorescent protein

UCP1

uncoupling protein 1