SH-SY5Y cells were cultured in basic growth medium including Eagle’s Minimum Essential Medium (EMEM) (Sigma Cat#M4655), 15% hi-FBS (heat-inactivated fetal bovine serum) (LDP Cat#FBS-02 heat inactivated), 2 mM glutamine, and Pen/Strep and passaged every 3–5 days. For differentiation, cells were detached using Trypsin/EDTA 0.5% and cultured in the first differentiation medium containing EMEM, 2.5% FBS, 2 mM glutamine, 10 µM all-trans retinoic acid (ATRA) (Sigma Cat#R2625), and Pen/Strep. The cells were then moved to the second differentiation medium with 1% FBS. The cells were maintained in this medium for 3 days. Subsequently, the cells were transitioned into the third differentiation medium containing neurobasal medium, 1× B27, 20 mM potassium chloride (KCl), 2 mM GlutaMAX (Thermo Fisher Scientific Cat# 35050061), 50 ng/ml BDNF (brain-derived neurotrophic factor) (Sigma Cat# SRP3014), 2 mM dibutyryl-cAMP (db-cAMP) (Sigma Cat#28745-M), 10 µM ATRA, and Pen/Strep. In our modified protocol, we eliminated the initial ECM coating of the plates and observed better differentiation if we avoided the splitting of cells on day 7 and 10 of differentiation from previous existing protocols (17, 18). The occurrence of successful differentiation was confirmed by staining for SMI-31 with PhosphoDetect™ Anti-Neurofilament H mAb (called SMI-31 by Millipore Sigma Cat#NE1022), which is a cytoskeletal marker and used in the concentration of 1:1,000. Furthermore, for assessing the effects of TSH on neuronal characteristics, bovine thyroid stimulating hormone (TSH) was from Sigma (Cat# T8931) and supplemented in the medium in varying doses.

Since we have used differentiation protocol which can possibly alter the standard endogenous pool within the cells, we have used the absolute quantification of TSHR between undifferentiated and fully differentiated SH-SY5Y cells using the absolute digital PCR machine, which does not depend on the cycle threshold of the endogenous gene and the gene of interest. The data obtained by this method can be more reliable because the data reduced to negative and positive counts of the partitioned sample, thus eliminating assay efficiency and instrument dependent variability. The precision is controlled by the number of replicates done for each. In addition to comparing the data to that of the undifferentiated cells, we also used cells that overexpress the TSHR receptors as positive control (JPO9) (Chinese Hamster Ovary (CHO) overexpressing the human TSHR) (19). Total RNA was extracted using RNeasy Mini kit (Qiagen Ltd Cat#74134) in combination with ribonuclease-free deoxy-ribonuclease treatment to remove endogenous DNA contamination, and the concentration and quality of extracted RNA were measured by NanoDrop. 1 µg of total RNA was reverse transcribed into cDNA using RNA to cDNA EcoDry™ Premix (Double Primed) (Takara Bio, Cat# 639547). The PCR reaction for each sample contained 1.8 µl dPCR master mix (Thermo Fisher Scientific Cat# A52490), 0.45 µl PCR TSHR probe directed to the transmembrane region of the receptor (Thermo Fisher Cat number 4453320), and a fixed amount of cDNA samples, which were made to a final volume of 9 µl with DEPC water. The program was as follows: Cycle 1 Denaturation: 10“ at 96°C, Cycle 2 Amplification 5“ at 96° and 30“ at 60° 40×. Absolute digital PCR software uses Poisson statistics and presents the corrected data for each sample as copies/µl for each sample based on the threshold that is determined by the sample that is used as a reference, which shows anywhere 5%–85% negative wells in the partitioned matrix. We have used more than three replicates per sample, and data presented here are average of two such biological replicates.

Confirmation for the existence of TSHR in neuronal cells and its localization in these neuronal cells was done with TSHR immunostaining using a TSHR-specific mouse monoclonal antibody MC1 directed against the hinge region of TSHR (epitope aa321-344) (20). Fisher rat thyroid cells (FRTL5) (which is a rat thyroid cell line) were used as the control. The FRTL-5 cells were grown in the six hormone medium (21) and were used as a positive control. For immunostaining, cells were cultured and differentiated in glass-bottom dishes at a density of 50,000 cells. After washing 1× with PBS pH 7.2, the cells were fixed with 4% paraformaldehyde in PBS at RT for 15 min. After fixation, cells were washed three times with PBS. The cells were blocked with 3% BSA and 5% horse serum in 1× PBS containing 0.3% Triton X-100 in the blocking buffer for 2 h. The fixed and blocked cells were treated with MC1 antibody (1:500), in a staining buffer containing 0.1% Tween 20 and 1% BSA in PBS overnight at 4°C. The next day, cells were washed three times with PBS and incubated with the anti-mouse conjugated to Alexa 488 (Cell Signaling Technology Cat# 4408) at 1:500 for 1 h at RT and then washed three times with PBS. Finally, the cells were mounted by mounting media containing DAPI (Vector Laboratories, Burlingame, CA) for nuclear staining. Fluorescence images were captured using ECLIPSE Ti2 inverted microscope (Nikon, Tokyo, Japan) equipped with NIS-Elements software (version 5.30.02, Nikon, Amsterdam, Netherlands) with 20×, 40×, and 100× objectives. ImageJ was used to adjust for brightness of all the images.

SH-SY5Y cells were grown in 96-well black bottom plates at 25,000 cells per well in the differentiation medium. After 18 days of differentiation, an assessment of cAMP for the functionality of TSHR in neuronal cells was performed using In-Cell Western, as described in our previous studies (22–24). Differentiated neuronal cells were treated with different doses of TSH (0, 1, 10, 100, and 1,000 micro-units per ml (µU/ml) for 24 h before the assessment of cAMP. PKA and PKC signaling molecules were also assessed using the same assay. To confirm if the effects were specific to TSH, blocking anti-TSHR-mAb (Ki-70) was used to reverse the observed effects. Ki-70 is a human TSHR blocking monoclonal antibody that binds to the leucine-rich repeat domain (LRD) region of the TSHR (25). The activation media were removed, and then the cells were fixed by adding 150 µl of fixation solution including 4% formaldehyde in 1× PBS for 20 min at room temperature. Cells were washed five times with 1× PBS containing 0.1% Triton X-100 (as a permeabilization agent) for 5 min each time, and 200 µl of washing solution was added and stayed for 5 min on the shaker and then removed; it was repeated four more times. Then, 50 µl of blocking buffer including BSA 5% in 1× PBS was added and stayed overnight at 4°. The blocking buffer was removed, and 25 µl of primary antibody diluted in the blocking buffer (1:100 dilution) was added to the wells and incubated overnight at 4° (primary antibodies were from Cell Signaling Technology, including Phospho-Akt (Ser473) (D9E) XP^® Rabbit mAb #4060, NF-κB p65 (D14E12) XP^® Rabbit mAb #8242, Phospho-PERK (Thr980) (16F8) Rabbit mAb #3179, and Phospho-PKCδ (Thr505) Antibody #9374). Cells were washed five times in 1× PBS containing 0.1% Tween 20 again five times each time for 5 min. Finally, 25 µl of infrared Fluorescent Dye (IRDye) subclass-specific antibody (IRDye 800CW LI-COR secondary antibody) in 1:800 dilution was added, in the dark for 60 min at room temperature on a shaker. Then, the cells were washed with PBS 1× containing 0.1% Tween 20 for five times and then visualized under the microscope.

As described for immunostaining, the cells were grown in 35-mm glass-bottom dishes at a seeding number of cells of 50,000. After staining the nuclei with DAPI, phase images and DAPI-stained images were acquired as two independent image files under the same settings. DAPI-stained nuclei cell bodies and a phase image were applied to define the extension threshold for quantification. We used a plugin named AutoNeuriteJ in ImageJ and counted the number of neurite outgrowths and their lengths both for primary (first neurite out of the cell body) and secondary/tertiary (S/T) outgrowths (which are defined as the further extensions from the primary neurite) (26). The measurement was done on four different images of each TSH treatment (n=4), and the result is shown here as the mean ± SD.

Cultured SH-SY5Y neuroblastoma cells remained in an undifferentiated status with no neurite outgrowth. However, these cells were amenable to differentiation using an 18-day protocol as outlined in Figure 1A . We developed a modified differentiation protocol where the medium contained 15% hi-FBS with no retinoic acid (RA) from day 0. As shown in the schematic diagram, undifferentiated cells cultured in high serum were subsequently transferred into a differentiation medium in the presence of 10 µM of RA followed by sequential decreases in serum concentration to 1% by day 8. On day 11, these cells were further subjected to additional factors including B27, BDNF, and db-cAMP until day 18. On redifferentiation, these cells developed easily visible neurite extensions ( Figure 1A phase image), which were visualized by staining their neurofilaments with antibody to SMI-31 shown by the green-stained axonal body and neurite outgrowth and blue DAPI-stained nuclei ( Figure 1B ) and demonstrated ~100% efficiency of differentiation into a neuronal phenotype.

We evaluated the expression of the TSHR in the differentiated cells at the mRNA and protein levels. Message levels of the TSHR were assessed by quantification of the absolute copy numbers of TSHR in undifferentiated versus differentiated cells using digital PCR (ddPCR). TSHR mRNA levels were quantified after 5, 10, and 18 days of differentiation. This time course assessment showed a significant increase in TSHR mRNA by day 18 and reached over 1,500 copies from a baseline of ~20 copies ( Figure 2A ). TSHR protein was detected by immunostaining with a TSHR-specific monoclonal antibody against the hinge region of the TSHR (MC-1) (20). As can be seen in Figure 2B , there was a surface expression of receptor protein in the differentiated cells. Optical sectioning of the stained cells showed the distribution of the TSHR expression throughout the surface of the differentiated cells ( Figure 2B upper).

Cyclic AMP generation by the differentiated SH-SY5Y cells after stimulation with TSH ( Figure 2C ) showed a dose-dependent increase, suggesting the presence of functional receptors. Since the TSHR is known to engage with different G proteins and beta arrestins for downstream signaling effects on cell proliferation and survival, we assessed other classical signals including pPI3K, pERK, p AKT, and NFKB and non-classical signals including pIKB, IRE1a, IL6, and PD1, on stimulation with increasing concentrations of TSH. Our previous studies have shown that activation of the TSHR with TSH or with stimulating TSHR autoantibodies leads to an array of signal pathway activity (23, 27, 28). Using TSH in a dose-response from 1 µU/ml to 1,000 µU/ml over 24 h, we measured these effects by In-Cell Western assays (29) and found AKT, NFkB, and PI3K signals in these cells increased in a dose-dependent manner ( Figure 3A ), confirming that some of the expected classical signals in addition to cAMP were increased by activation of the TSHR in these neuronal cells. In contrast, the non-classical responses were dampened rather than exacerbated ( Figure 3B ) as also seen with pERK, which after showing an early response was then suppressed ( Figure 3A ).

Having observed functional TSHRs on the differentiated neuronal cells, we then examined the effect of TSH on neurite outgrowth. The cells were supplemented at each differentiation medium change, with 50 µU/ml, 100 µU/ml, or 1000 µU/ml of TSH, and neurite growth was found to be stimulated by increasing concentrations of TSH ( Figure 4A ). The primary neurite outgrowth count showed a ~2.5-fold increase, and this effect of TSH was effectively blocked by pretreating the cells with a human TSHR blocking antibody (K1-70, 10 ug/ml). Similarly, quantification of S/T neurite outgrowth also showed a ~1.5-fold increase ( Figure 4B ). Figures 4C, D displays representative images analyzed for neurite count and neurite length in both untreated and TSH-treated cells. These results confirmed that TSH enhanced neurite outgrowth in these neuronal cells. For neurite length, in a low dose of TSH, a significant difference was observed, which did not show a dose response ( Figure 5 ).