Tirzepatide prevents neurodegeneration through multiple molecular pathways

The SHSY5Y human neuroblastoma cell line was purchased from DSMZ Cell Dive (ACC-209). Cell line has been tested and authenticated following the manufacturer's instruction and it was negative for mycoplasma contamination. The cell line was maintained in an incubator at 37 °C, 95% O₂, and 5% CO₂. The cells were grown in DMEM (Microgem cat# AL007) was used along with 15% fetal bovine serum (FBS, Euroclone cat# ECS0180L), 1% l-glutamine (Euroclone cat# ECB3000D), 1% penicillin–streptomycin (Euroclone cat# ECB3001D) and 1% non-essential amino acid (Microgem cat# ACL006). The cells were exposed to normal (NG) (25 mM) and high glucose (HG) (150 mM) [15] (Merck-KGaA, cat# G8644-100ML) for 7 days. Before selecting the appropriate concentration of TIR and initiating experiments, a comprehensive toxicity assessment using cell viability assays and cell counts were performed. Cells exposed to TIR concentrations ranging from 0 µM to 0.4 µM for 7 days exhibited no statistically significant differences in both cell viability percentage and cell numbers (Additional file 1: Fig. S1). Both NG and HG cells, after reaching 70–80% of confluence, were treated with Tirzepatide (Selleckchem, LY3298176 cat# P1206) at concentration of 0.2 μM, since the lower effective dose used in other cell line [26] was ineffective in our setting. Experiments were executed and repeated at least for 3 times. In all experiments, every 48 h cell culture media was replaced.

Cell proliferation was assayed by Ki-67 (Alexa Fluor® 488 conjugate) antibody (Cell Signaling cat#11882). According to the manufacturer’s instructions, the cells after 7 days treatment are detached and washed with PBS. The cells were then fixed with 4% formaldehyde followed by incubation for 15 min at room temperature. After washing with PBS, cells were permeabilized using 0.1% PBS-tween on ice for 15 min. According to datasheet 1 × 10⁶ cells were incubated with the antibody for 1 h at room temperature after washing with PBS. Before measurement the cells were again washed 2 times and resuspended in PBS. The measurements were carried out using BD Accuri C6 PLUS Personal Flow Cytometer (BD biosciences) and data processing was performed by FlowJo BD Accuri C6 Plus software for windows [27].

Cells were dissolved in lysis buffer containing protease inhibitors (Tris–HCl, pH 8, 10 mM; NaCl 150 mM; NaF 10 mM; NP40 1%; PMSF 1 mM). Then, the proteins were subjected to 8% or 10% sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to 0.22 μm polyvinylidene fluoride membranes. The membranes were blocked with 5% nonfat milk in TBS-T (Tris-buffered pH8/0.15% Tween 20) at room temperature for 1 h and then incubated with primary antibodies diluted in TBS-T (dilutions in according to data sheet), BDNF (1:1000) (Elabscience cat#e-ab-18244), CREB (1:1000) (Cell Signaling cat#9197), SORBS1(1:500) (Atlas antibodies cat# hpa027559), GAP43 (1:1000) (Abcam cat#ab16053), MAP2 (1:500) (Abcam cat#ab32454), GLUT4 (1:1000) (Abcam cat#ab33780), Bcl-2 (1:1000) (Elabscience cat#e-ab-15522), BAX (1:1000) (Abcam cat#ab32503), GLUT1 (1:1000) (Elabscience cat# e-ab-31556), GLUT3 (1:400) (Abcam cat#ab15311), AGBL4 (1:1000) (Gene Tex cat#n2c3), pAkt (1:1000) (Cell Signaling cat#4060) Akt (Cell Signaling cat# 2920) for overnight at 4 °C. For normalization of protein expression Actin (1:1000) (Abcam cat# ab8227) and Vinculin (1:10,000) (Abcam cat#ab129002) were used as internal control. After three washes in TBS-T, the membranes were incubated with corresponding secondary antibodies, with goat anti-rabbit IgG-h + HRP conjugated (Bethyl cat# A120-101p) and donkey anti-mouse IgG-h-I HRP conjugated (Bethyl cat# A90-137p) (1:5000) secondary antibodies, for 1 h at room temperature. Immunocomplexes were visualized by using Clarity Max Western ECL Substrate (Bio-Rad Laboratories cat#1705062) Immunocomplexes were observed through ChemiDOC Imaging System with Image LAB Software (Bio-Rad Laboratories, Version 6.1). The molecular weight of proteins was estimated with pre-stained protein markers (Opti-Protein-Marker abm cat# G623). Densitometry analysis was performed using ImageJ software.

For the isolation and purification of total RNA, including small RNAs from SHSY5Y cells, in the first step, trypsin was used to detach the cells from the flask and collected in 1.5 ml tubes. Then, Trizol and bromocloropropane were added followed by vortexing for 15 s. The tubes were incubated for 15 min at room temperature. After incubation, they were centrifuged for 15 min at 12,000 rpm at 4 °C. The aqueous phase was collected and in which same quantity of chilled isopropanol was added, followed by vortexing. After incubating for more than 1 h at − 80 °C the tubes were again centrifuged for 30 min at 12,000 rpm at 4 °C. The supernatant was discarded, and the pallet was washed with 75% ethanol by centrifugation at 7500 rpm for 10 min. The pallet was resuspended with sterile water and detection of purity and concentration of the RNA were carried out through QIAexpert spectrophotometer (Qiagen cat#1038703). The complementary DNA (cDNA) was synthesized from 2 ng of the total purified RNA by TaqMan MicroRNA Reverse Transcription kit (Applied Biosystem Lithuania cat# 4366597) using specified RT primer (Applied Biosystem cat# 4427975 U6 sn RNA, cat# mir34, cat# mir212 cat# mir29c) following protocol of manufacturer. Rotor-gene Q (Qiagen cat# R0515102) was used to measure the expression of miRNA through utilizing the PrimeTime gene expression master mix (IDT cat#1055772) and the TM primers (Applied Biosystem cat#4427975 U6 snRNA, cat#4440887 mir34a-5p, mir212-3p, mir29c-5p). The PCR condition was polymerase activation at 95 °C for 3 min, followed by 40 cycles of amplification with denaturation at 95 °C for 5 s and annealing at 60 °C for 30 s. All the samples were run in duplicates. For each amplification cycle particular threshold cycle (C_t) value was obtained and the difference between C_t values of targeted miRNAs and U6 were used to calculate ΔC_t. Then to obtain ΔΔC_t difference between the ΔC_t of NG group and drug treated group were taken. Finally, for the analysis of fold change 2^−ΔΔCt was calculated and its average from three individual experiments was plotted in histogram [28].

cDNA was synthesized from 1 μg of total RNA isolated from SHSY5Y cells using QuantiTect Reverse Transcription Kit (Qiagen cat# 205310). mRNA levels were determined by qPCR with SsoAdvanced Universal SYBR Green Supermix (Bio-Rad cat#1725270) using Rotor-Gene Q (Qiagen).

Primer’s sequence:

BDNF

Primer 1: 5′-GCTCTGTGCGATTTCATTGTG-3′

Primer 2: 5′-GCCTTCATGCAACCAAAGTATG-3′

GLUT3

Primer 1: 5′-GAAGACTTGAATTAGATTACAGCGATG-3′

Primer 2: 5′-GAAAGAGCCGATTGTAGCAA-3′

GLUT1

Primer 1: 5′-GTGCCATACTCATGACCATCG-3′

Primer 2: 5′-GGCCACAAAGCCAAAGATG-3′

GLUT4

Primer 1: 5′-TCCAACAGATAGGCTCCGAA-3′

Primer 2: 5′-CCCAATGTTGTACCCAAACTG-3′

BAX

Primer 1: 5′-CAAACTGGTGCTCAAGGC-3′

Primer 2: 5′-AAAGATGGTCACGGTCCAAC-3′

BCL2

Primer 1: 5′-GATGACTGAGTACCTGAACCG-3′

Primer 2: 5′-AGCCAGGAGAAATCAAACAGAG-3′

CREB

Primer 1: 5′-GTCCAAACAGTTCAGTCTTCCT-3′

Primer 2: 5′-GTTACACTATCCACTGACTCCTG-3′

AGBL4

Primer 1: 5′-CTTGTAAGCCAGCACCCTAT-3′

Primer 2: 5′-CAGCCAGTGACGATTCAGAT-3′

SORBS1

Primer 1: 5′-ATCTTCCCACCACCTTAAACC-3′

Primer 2: 5′-GAACCACCATCACATTCAGAAC-3′

MAP2

Primer 1: 5′-TGAAGAACATCCGCCACAG-3′

Primer 2: 5′-ATCTTGACATTACCACCTCCAG-3′

GAP43

Primer 1: 5′-AGCCAAGCTGAAGAGAACATAG-3′

Primer 2: 5′-TTCTTAGAGTTCAGGCATGTTCT-3′

β-ACTIN

Primer: 5′-CATCCGCAAAGACCTGTACG-3′

Primer: 5′-CCTGCTTGCTGATCCACATC-3′

All samples were run in duplicate. For each amplification cycle, a threshold cycle (C_t) value was obtained, and ΔC_t was calculated as the C_t difference between target mRNA and housekeeping mRNA (β-Actin). Fold increase of mRNA expression compared with NG was calculated using 2^−ΔΔCt method.

The DNA was isolated with the help of QIAamp DNA Blood Mini kit (Qiagen cat#51104) following the protocols of manufacturer. 350 ug of DNA was taken for bisulfite conversion through EpiTect Fast DNA Methylation Kit (Qiagen cat#59824) according to the instructions from manufacturer. In the next step, this bisulfite converted DNA was amplified using Pyromark PCR kit (Qiagen cat#978703) where the PCR conditions were 1 cycle for 15 min at 95 °C and next 40 cycles for 30 s at 94 °C, 30 s at 56 °C and 10 min at 72 °C. For assurance the PCR products were run in 2% agarose gel electrophoresis (Amersham Bioscience). Then the biotinylate PCR products were investigated for pyrosequencing-based methylation utilizing PyroMark Q48 Advanced CpG Reagent (Qiagen cat# 974022) and commercially designed primers including SORBS1_03 (PM00042168) Chromosome 10, bp 97319534–97321622, AGBL4_03 (PM00094297) Chromosome 1, bp 50488739–50490934, CREB 1 (CRTC1_01) (PM00188804) Chromosome 19, bp 18793114–18795233 and BDNF_01 (PM00155491) Chromosome 11, bp 27721460–27723528. For the analysis PyroMark CpG SW 1.0 software (Qiagen) was used [29].

QIAGEN Ingenuity Pathway Analysis (IPA) software (QIAGEN, Milan, Italy) was used for enrichment analysis. The ‘‘core analysis’’ function was used to interpret the data based on biological processes, canonical pathways, and gene networks. Each gene identifier was mapped to its corresponding gene object in the Ingenuity Pathway Knowledge Base (IPKB). The p-value of 0.05 was set as the cutoff value for the enrichment. The top five enrichment results of the Molecular and Cellular Function were used as a focus point to connect all the available data using the tools “Connect” and “Path Explorer”.

The data were analysed by SPSS IBM Version 23 or GraphPad Prism 7.0 software and expressed as the means ± SEM. Differences between the mean values were considered significant at a p-value of < 0.05. GraphPad Prism software was used for drawing figures.

The effects of TIR on proliferation, and main markers involved in neuronal growth and death were first investigated in SHSY5Y exposed to normal (NG) and high glucose (HG) concentration.

The cell proliferation was detected by measuring Ki-67. Cells exposed to HG showed significant Ki-67 median of fluorescence reduction compared to NG (p < 0.05). However, TIR treatment did not show a significant effect on cell proliferation independently of glucose concentration (Fig. 1).

Moreover, TIR induced upregulation of mRNA expression and protein levels of both CREB and BDNF (Fig. 2A, B) compared to non-treated cells (NG) (p < 0.05 vs NG). HG induced downregulation of both CREB mRNA and protein levels compared to NG cells (p < 0.05) while TIR prevented such HG mediated effect (p < 0.05 vs HG) (Fig. 2A). No statistically significant differences between HG and NG were observed in BDNF mRNA level while HG induced a BDNF protein downregulation (p < 0.05 vs NG) (Fig. 2B). In contrast, TIR upregulated BDNF mRNA expression and prevented the HG-induced protein levels down-regulation (p < 0.05 vs HG) (Fig. 2B).

In addition, TIR treatment (NG-TIR) induced a statistically significant increase in Bcl-2 and decreased BAX protein levels, while no changes in mRNA expression were evidenced. Neither HG nor TIR-HG had any significant effect on Bcl-2 mRNA and protein expression (Fig. 3A).

HG increased BAX mRNA expression, but not protein levels (p < 0.05 vs NG). Addition of TIR (TIR-HG) prevented the mRNA increase and induced a significant decrease of BAX protein levels (Fig. 3B). So far, the protein ratio BAX/Bcl-2, an important apoptotic marker, peaked in presence of HG (p < 0.05 vs NG) but presence of TIR in medium smoothed both NG and HG effect (Fig. 3C).

Cell differentiation was evaluated by analyzing the main neurodifferentiation markers, such as pAkt, MAP2, GAP43, and AGBL4 (Fig. 4).

Along with NG, TIR treatment induced a statistically significant increase in pAkt, MAP2, GAP43, and AGBL4 protein levels (Fig. 4A–D, right panel). Nevertheless, only for GAP43 and AGBL4, we observed a parallel significant increase in mRNA and protein levels (Fig. 4C, and D, left panel). HG vs NG showed a significant decrease in pAkt and MAP2 protein levels (Fig. 4A, and B, right panel) and a decrease in MAP2 and AGBL4 mRNA expression (Fig. 4B, and D left panel). TIR-HG prevented the HG-induced changes in protein levels of pAkt, MAP2, GAP43 (Fig. 4A–C, right panel) and MAP2 and GAP43 mRNA expression (Fig. 4B and C, left panel). In cells exposed to HG, TIR failed to affect AGBL4 mRNA and protein levels (Fig. 4D).

To better understand the TIR role on neuronal glucose metabolism, the mRNA and protein levels of GLUT1, GLUT3 and GLUT4 and SORBS1 [30, 31], a major regulator of insulin-stimulated signaling and glucose uptake, were investigated (Fig. 5). TIR treatment induced a statistically significant increase in GLUT3 and GLUT4 mRNA expression (Fig. 5B and C, left panel) and in GLUT4 and SORBS1 protein levels [32] (Fig. 5C and D, right panel).

HG vs. NG showed a significant decrease in GLUT3 and GLUT4, and SORBS1 mRNAs Fig. 5B–D, left panel) and a significant decrease in GLUT1, GLUT3, GLUT4, and SORBS1 proteins (Fig. 5A–D, right panel). Treatment with TIR (HG-TIR) prevented HG-induced changes observed in mRNA expression and protein of glucose transporters and insulin sensitivity marker levels.

TIR did not affect DNA methylation in all genes investigated along with NG incubation. HG vs NG showed a statistically significant increase in total DNA methylation levels in the promoter region of CREB and BDNF (Fig. 6A, and B). HG-induced a statistically significant increase in methylation levels in positions 2 and 3 of CREB (p < 0.05 vs NG) and in position 3 of BDNF genes (Fig. 6A, and B, right panel). Such HG related changes were prevented by TIR addition (Fig. 5A and B).

No effect of HG and HG-TIR treatment on DNA methylation of AGBL4 promoter regions and SORBS1 gene body (Fig. 5C, and D) was observed.

TIR treatment vs NG induced a statistically significant decrease in miR-34a and an increase in miR-212 and miR-29c (Fig. 7A–C). HG showed a significant increase in miR-34a and a significant decrease in miR-212 and miR-29c expression levels. All the HG-related changes were reversed by TIR addition.

The interactive model confirmed the connections among analyzed markers and the main pathways in which they are involved (Additional file 1: Table S1). These analysis highlighted connections between already available targets in a dataset and uncovers relevant relationships between genes of interest enabling the identification of the shortest paths between molecules. To further explore the potential role of TIR, subnetworks, and canonical pathways were interrogated by adding to the interactive model TIR. The simulation showed and confirmed that TIR (Fig. 8) interfered with the main networks and pathways that were found activated in SHSY5Y human neuroblastoma cell line.