Synthesis and characterization of novel ssDNA X-aptamers targeting Growth Hormone Releasing Hormone (GHRH)

Background Growth Hormone Releasing Hormone (GHRH), 44 amino acids containing hypothalamic hormone, retains the biological activity by its first 29 amino acids. GHRH (NH2 1–29) peptide antagonists inhibit the growth of prostate, breast, ovarian, renal, gastric, pancreatic cancer in vitro and in vivo. Aptamers, single-strand RNA, or DNA oligonucleotides are capable of binding to target molecules with high affinity. Our aim in this study is to synthesize and select X-aptamers against both GHRH NH2 (1–29) and GHRH NH2 (1–44) and demonstrate synthesized aptamers’ target binding activity as well as serum stability. Methods and results Aptamers against GHRH NH2 (1–44) and NH2 (1–29) peptides were synthesized, and binding affinity (Kd) of 24 putative X-aptamers was determined by the dot-blot method, co-immunofluorescence staining and, SPR analysis. The serum stability of TKY.T1.08, TKY1.T1.13, TKY.T2.08, TKY.T2.09 X-aptamers was 90–120 h, respectively. The dose-dependent binding of TKY1.T1.13, TKY.T2.08, TKY.T2.09 X-aptamers on GHRHR in MIA PaCa-2 was approved by co-IF assay results. Moreover, SPR analysis indicated the Kd (4.75, 1.21, and 4.0 nM) levels of TKY2.T1.13, TKY.T2.08, TKY.T2.09 putative X-aptamers, respectively. Conclusion Our results illustrate the synthesis of 24 putative X-aptamers against both GHRH NH2 (1–44) and NH2 (1–29) peptides and TKY1.T1.13, TKY.T2.08, TKY.T2.09 X-aptamers have high serum stability, high target binding potential with low Kd levels.

Bacterial His-GHRH (1-44) protein isolation and purification 1.5 x 10 8 cells/mL E. coli HB101 become competent by the CaCl 2 method [15] and then 500 μg/mL His-tagged and His-tagged GHRH plasmids transformed by the heat shock method [16]. Following the transformation, His-and His-GHRH expressing selective positive controls were grown at 37˚C overnight; the pellets were obtained by centrifugation and resuspended in lysis buffer (20 mM Tris-HCl pH: 7.5, 300 mM NaCl, 2M Urea, 2% Triton X-100, 1 mg/mL lysozyme and 1X EDTA-free Protease Inhibitor Cocktail). After the sonication step, the pellets were resuspended in the inclusion body solubilization solution (20 mM Tris-HCl, 0.5 M NaCl, 5 mM imidazole, and 6 M guanidine hydrochloride), and His-GHRH protein was purified by Dynabeads™ His-Tag Isolation and Pulldown magnetic beads (Invitrogen, Paisley, UK) according to the manufacturer's instructions [17].

RNA isolation and real-time PCR
Total RNA was extracted in HEK293 cells and putative His, His-GHRH expressing HEK293 cells by PureZol RNA isolation reagent (Bio-Rad, Hercules, California, USA). The first-strand cDNA was synthesized by the iScript cDNA synthesis kit (Bio-Rad). Primers for the GHRH gene were 5'-TATGCAGATGCCATCTTCAC-3' and 5'-TCATCCCTGGGAGTTCCTGC-3', and the 18S gene were 5'-CTACCACATCCAAGG AAGGCA-3' and 5'-TTTTTCGTCACTACCTCCCCG-3', respectively. Real-time PCR was performed in Bio-Rad CFX Connect Real-Time PCR Detection System with SYBR Green Supermix (Bio-Rad). 18S was used as a normalization control. The fold of change was calculated according to the PFAFFL method [18]. All reactions were run in triplicate.

GHRH ELISA assay
wt, His, His-GHRH expressing HEK293 cells, MIA PaCa-2, HT-29, LNCaP, PC3, PNT1A cells were seeded at 100 mm Petri dishes with a seeding density of 3 x 10 6 cells/dish. Following overnight cell attachment, the media were concentrated via filtering by Amicon Ultra centrifugal filter (Merck, Massachusetts, USA). 100 μL of concentrated cellular media were used for GHRH ELISA assay (Elabscience, E-EL-H1146) according to the manufacturer's instructions and were measured at 450 nm wavelength by a microplate reader (Multiskan, Thermo Scientific). The secreted levels of GHRH concentration were calculated according to standards. Since we do not have unlabeled aptamers to compete with labeled aptamers, we used labeled aptamers in competitive ELISA Assay instead of detection antibodies.

Immunoblotting
Total proteins were extracted using M-PER Mammalian protein extraction reagent (Thermo Fisher Scientific). Concentrations of total proteins were quantified by Bradford assay [19]. 80 μg cellular proteins were separated on 15% SDS-PAGE and transferred to the PVDF membrane (Thermo Fisher Scientific). Membranes were blocked by 5% non-fat dried milk in TBS-T solution (1X TBS, 0.1% Tween-20) and probed with primary antibodies against GHRH and His-tag, then incubated with secondary anti-goat horseradish peroxidase-conjugated IgG, and developed with enhanced chemiluminescence solutions, images were taken by ChemiDoc Imaging System (BioRad, Hercules, California, USA).

X-aptamer selection and synthesis
According to the instruction, X-aptamer selection was performed by the X-aptamer selection kit (AM Biotechnologies, Houston, USA). At the first step, the X-aptamer library was prepared as follows: the library resuspended in a selection buffer [1X PBS pH:7.4, 1 mM MgCl 2 , 0.05% Tween20, 0.02% BSA], incubated at 95˚C water bath for 5 minutes, and cooled down slowly to room temperature. We performed two a bead based aptamer selection; first a bead based aptamer selection, we used bacterial/eukaryotic GHRH 1-44 His-tagged peptides, and second a bead based aptamer selection used GHRH 1-29 peptide (Sigma, G6771) as a target. Due to kit instructions, for the first a bead based aptamer selection, 10 μg His-tagged bacterial and 10 μg eukaryotic His-tagged GHRH 1-44 peptide was used to pulldown by magnetic beads. In addition, for the second bead based aptamer selection, biotinylated GHRH 1-29 peptide was used for streptavidin magnetic bead isolation. Next, His-tag/biotinylated isolation and pulldown magnetic beads were prepared for negative selection by washing with a selection buffer and incubated with the prepared library for one hour at room temperature. After incubation, the library which did not bind to magnetic beads (i.e., negatively selected library) was transferred to a clean tube with the aid of a magnetic stand. Then, his-tagged target protein [GHRH 1-44 (bacterial and eukaryotic)] or biotinylated GHRH 1-29 was mixed with His-tag/streptavidin coated-magnetic beads, incubated for 30 minutes at room temperature, and target proteinbound magnetic beads were washed with a selection buffer. Next, negatively selected library and target-bound magnetic beads were incubated at room temperature for 90 minutes. Magnetic beads were washed with a selection buffer and resuspended in the selection buffer. Putative aptamers were cleaved by incubating with NaOH at 65˚C for 30 minutes and neutralizing with Tris-HCl. Cleaved putative aptamers were transferred to a clean tube. Spin column buffer exchange was performed to eliminate NaOH, and a "cleaved oligonucleotide pool" was obtained. For the secondary selection process, four 1.5 mL centrifuge tubes were labeled Tube 1, Tube 2, Tube 3, Tube 7. Cleaved oligonucleotide pools and selection buffers were added to all four tubes. His-GHRH (1-44) protein was added to Tube 2 and Tube 3 and incubated with rotation at room temperature for one hour. His magnetic beads were added to Tube 2, Tube 3, and Tube 7 and incubated with rotation at room temperature for 30 minutes. 10 μL from the tubes used as a template for PCR reaction. The forward primer was common for all; however for tube 1, reverse primer 1, tube 2 reverse primer 2, tube 3 reverse primer 3, and tube 7 reverse primer 7 was used. PCR reaction prepared in total 100 μL volume as follows: 1X PCR Buffer, 2.5 mM MgCl 2 , 0.2 mM dNTP, 0.4 μM forward primer, 0.4 μM indicated reverse primer, 1U Taq polymerase. The temperature condition was 94˚C 1 minute, cycles of 94˚C 30 seconds, 50˚C 30 seconds, 72˚C 1 minute, and the last extension at 72˚C for 3 minutes. For each tube, 14, 18, 22 cycles of PCR were performed. PCR products were run on agarose gel, and images were obtained by UV Imaging System. For the four tubes, 25 cycles were performed, and all PCR products were mixed in one tube and resent back to the manufacturer for Illumina nextgeneration sequencing (AM Biotechnology, Texas, Huston, USA). Next-generation sequencing and synthesis of candidate x-aptamers with 5' biotin tag were performed by the corresponding company (AM Biotechnology, Texas, Huston, USA) [14].

Dot-blot assay
Putative 24 X-aptamers were renaturated (95˚C for 10 min), incubated with either GHRH NH2  or GHRH NH2 (1-44) peptide in dot blot binding buffer (15 mM Tris, 100 mM NaCl, 1 mM MgCl 2 ) overnight at +4˚C, blotted to nitrocellulose membrane (Invitrogen) as a small spot. As a negative control we used scramble aptamer which has a ΔG: -27.621 kJ (5'-TTTTTTTCAGACCAGCCGTGCACGACGAACCACAAGCAGGTGGGCCCA-3') Following air dry of the membrane, HRP-Streptavidin conjugate (Thermo Fisher Scientific) was added onto each membrane spot, incubated for two hours at room temperature in dark, TMB substrate solution (Thermo Fisher Scientific) added, and images were taken by LCD Camera. The same protocol was performed for K d determination of each putative aptamers by the dot-blot method. The K d levels were determined in each aptamer dot spot intensity level by Sigma Plot v14.0 (https://systatsoftware.com/products/sigmaplot/) program.

Serum stability of putative X-aptamers
5 μL from each 100 μM putative X-aptamer against GHRH was mixed with 5 μL human serum (Sigma Aldrich, P2918, Massachusetts, USA) for 0-120 h at 37˚C. Following each incubation period, 10 μL of the mixture loaded at 12% polyacrylamide gel electrophoresis and was visualized under the ChemiDoc MP imaging system (Bio-Rad Laboratories, Hercules, USA). DNaseI digested aptamers were used as positive controls.

Cyclic AMP assay analysis
To determine the impact of aptamers on cAMP levels, we performed Cyclic AMP XP1 Assay Kit (CST, 4339). Both HT29 colon cancer and MIA Paca-2 pancreatic cancer cells were seeded at 12 well-plate with a density of 2 x 10 4 . Following cell attachment, adherent cells treated with selected X-aptamers for 72 h, and cAMP assay performed according to the manufacturer's instructions.

Dose-dependent X-aptamer binding affinity by immunofluorescence
MIA PaCa-2 pancreatic cancer cells were seeded on coverslips with a density of 1 x 10 5 cells/ coverslip. Following cell attachment, x-aptamers were applied in a dose-dependent manner (0-500 nM), cells were fixed, permeabilized, blocked, and probed with anti-rabbit anti-GHRH primary antibody (1:50 dilution) at +4˚C overnight. Then, the anti-rabbit Alexa Fluor 568-conjugated secondary antibody (1:250 dilution) was applied at +4˚C for one hour, 1 μL streptavidin solution added, and incubated at 37˚C for 30 min, and probed with Streptavidin-Alexa fluor 488 conjugate (1:1000 dilution) for one hour at RT. The nucleus was stained with DAPI and images were visualized by fluorescence microscopy (Olympus, Japan). GHRH siRNA silencing with/without aptamer treatment was used as a negative control for aptamer binding affinity. 50 nM GHRH siRNA (Ambion, AM16708) with 6 μL FuGENE 6 (Promega, E2691) transfection reagent for 48 h performed in MIA Paca-2 cells. Following siRNA incubation period 500 nM aptamer treated and aptamers probed with Streptavidin-Alexa fluor 488 conjugate and GHRH probed with anti-rabbit Alexa Fluor 568-conjugated secondary antibody.

Statistical analysis
All the experiments were analyzed statistically by two-way ANOVA using GraphPad Prism 8 (GraphPad Software, La Jolla, CA, USA). p-values considered statistically significant were given as � p<0.05, �� p < 0.01, ��� p < 0.001, respectively. Error bars in the graphs generated using ± standard deviation (SD) values. The mean ± SD of the data representing ELISA assay and the qRT-PCR analysis was achieved from at least two experiments with three replicates.

Purification of His-tagged GHRH peptide
To obtain a target for X-aptamer selection, both bacterial and eukaryotic His-GHRH NH2 (1-44) expressing cellular models were generated. To yield a high range of GHRH 1-44 peptide, we selected E.coli HB101 cells and for post-transcriptional modifications of GHRH 1-44, we used HEK293 cells. E.coli HB101 cells expressing His-tagged and His-tagged GHRH protein determined by immunoblotting following plasmids transformed by heat-shock method (Fig  1a). Solubilized inclusion bodies and His-magnetic bead pull-down extracts containing His-GHRH determined by immunoblotting, respectively (Fig 1b). Since the bacterial protein synthesis mechanism lacks post-translational modifications, HEK293 cells were used for target protein production. For this purpose, His tagged (pCMV3-SP-N-His-NCV) and His-tagged GHRH (pCMV3-SP-His-ORF) plasmids were transfected to HEK293 human embryonic kidney cells, and stable His-tagged and His-GHRH-tagged expressing HEK293 cells generated following increased concentration of hygromycin application for selection. Transcriptional and translational GHRH 1-44 expression was determined by RT-PCR (Fig 1c), and immunoblotting (Fig 1d). Besides, the immunofluorescence method was confirmed the intracellular expression of GHRH majorly in HEK293-His-tagged GHRH stable cell line (Fig 1e). Moreover, the HEK293-His-GHRH stable cell line secreted 685,38 pg/mL GHRH was determined by the ELISA method (Fig 1f).

Synthesis and selection of x-aptamers against GHRH peptide
To synthesize and select X-aptamers against both GHRH NH2 (1-44) and NH2 (1-29), we performed two a bead based aptamer selection according to the kit protocol. First bead based aptamer selection, purified 10 μg bacterial and 10 μg eukaryotic His-tagged GHRH used, cycle-course PCR performed as 14 cycles, 18 cycles, 22 cycles, and 25 cycles of last PCR performed to obtain samples for sequencing (Fig 2a). As bacterial and eukaryotic His-tagged GHRH protein has 10 Histidine amino acids at N terminal region, GHRH 1-29 peptide also

PLOS ONE
Synthesis of Growth Hormone-Releasing Hormone (GHRH) X-aptamer used as a target for second bead based aptamer selection cycle-course and last PCR was also performed for GHRH (1-29) target (Fig 2b). All the samples were sequenced by Illumina nextgeneration sequencing via aligning 3' conserved sequences and putative x-aptamers were synthesized with a 5'-biotin tag by a commercial company, AM Biotechnologies (Houston, USA). The sequences of putative x-aptamers were given in

Screening the binding affinity of putative X-aptamers to the target protein by dot blot assay and SPR analysis and illustrating the serum stability of each X-aptamer
To observe the binding affinity of X-aptamers to the target protein, the dot blot assay was performed. The dot intensities measured and fold of change of band intensities calculated by Graphpad Prism 8 program via using spot intensity. According to calculations, putative xaptamers against GHRH (Fig 3a and 3b). Then binding affinity of X-aptamers to target assessed in increasing aptamer doses. The fold of change was calculated according to the dot intensities and dissociation constant (K d ) was calculated by nonlinear regression in GraphPad Prism software (https://www. graphpad.com/scientific-software/prism/) (Fig 3c). TKY2.T1.08, TKY2.T1.13, TKY.T2.08, and TKY.T2.09 were selected for further experiments according to obtained data. To determine the non-specific binding of selected X-aptamers against its target, we performed GHRH competitive ELISA assay and dot-blot analysis (Fig 3d). According to GHRH ELISA results, instead

GHRH expression and secretion profile of MIA PaCa-2, HT-29, PC3 cancer cells and PNT1A epithelial derived normal cell
To determine the binding affinity of X-aptamers on cellular GHRH target, at first, GHRHexpressing cells were determined. Pancreatic cancer cell line MIA PaCa-2, colorectal cancer cell line HT-29, prostate cancer cell line PC3 and LNCaP, and normal human prostate epithelial cell line PNT1a used for this purpose. Primarily, GHRH expression in these cell lines determined by immunofluorescence assay with anti-GHRH antibody and it was observed that all these cell lines expressed GHRH endogenously (Fig 5a). GHRH expression of these cells was also investigated by immunoblotting (Fig 5b). GHRH concentration in media was also calculated by human GHRH ELISA assay. GHRH concentration in media was about 1500 pg/mL in MIA PaCa-2, HT-29, and PC3 cells while the value is about 700 pg/mL in PNT1a cells (Fig 5c).

Blocking activity of GHRH signaling of selected aptamers
To evaluate the selected aptamers' effect on GHRH signaling, we performed cAMP assay analysis and GH/GHRHR co-immunofluorescence staining (Fig 7a and 7b). Due to cAMP assay analysis, both TKY2.T1.08 and TKY2.T1.13 X-aptamers significantly depleted intracellular cAMP concentration as compared to untreated control cells in HT-29 cells. When we checked the results, no significant effect was observed for cAMP levels for TKY2.T1.08 and TKY2. T1.13 X-aptamers application in MIA Paca-2 cells. In addition, we measured a statistically significant decline in cAMP levels following TKY.T2.08 and TKY.T2.09 X-aptamers treatment in MIA Paca-2 cells (Fig 7a). To demonstrate the blocking activity of selected aptamers on GHRH signaling, we performed GH/GHRHR co-immunofluorescence staining (Fig 7b). Active GHRH signaling due to GH and GHRHR expression profile was detected in untreated and GHRH peptide treated MIA Paca-2 cells. However, a significant downregulation on GH/ GHRHR expression demonstrated following TKY2.T1.08, TKY.T2.08, and TKY.T2.09 treatment in MIA Paca-2 cells (Fig 7b).

Discussion
GHRH is a neuropeptide that is known to be ectopically expressed in many cancer types such as the pancreas [20], prostate [21], ovarian [22] and lung cancer [23]. Due to the expression of GHRH/GHRHR and biologically active signaling leads the potential essential role of .09 x-aptamers were investigated by immunofluorescence. Streptavidin-Alexa Fluor 488 conjugate was used to specify biotin-labelled x-aptamers. Cells were treated with X-aptamers in a dose-dependent manner (0-500 nM). Following incubation, cells were examined under a fluorescence microscope (Olympus) and images were taken and analyzed. The binding position of B TKY2.T1.08, D TKY2.T1.13, F TKY.T2.08, H TKY.T2.09 x-aptamers investigated by co-immunofluorescence. Streptavidin-Alexa Fluor 488 conjugate was used to specify biotin-labelled X-aptamers. Streptavidin-Alexa Fluor 588 conjugated anti-GHRH was used to specify GHRH. Cells were treated with x-aptamers in a dose-dependent manner (0-500 nM). Following incubation, cells were examined under a fluorescence microscope (Olympus) and images were taken and analyzed. GHRH siRNA was used as a negative control.
https://doi.org/10.1371/journal.pone.0260144.g006 therapeutic regimes targeting GHRH. For this purpose, many GHRH peptide antagonists have been designed and synthesized by modifications on GHRH NH2 (1-29) peptide [24]. It was shown that these antagonists compete with GHRH for binding to the GHRH receptor and block receptor activation [25]. In this way, these antagonists were demonstrated to inhibit proliferation and induce apoptosis, via a direct effect on tumor cells in many cancer types such as lung, breast, prostate, ovarian, colorectal, and pancreas cancers [6,10]. With a similar strategy, we tried to target GHRH by synthesizing capturing aptamer as a ligand to block the GHRH signaling. Although GHRH antagonists were studied to block GHRH signaling, the peptide nature of these antagonists limits their use in vivo. Also, there are no reported aptamers against GHRH in literature, thus in this study, we aimed to select an aptamer against GHRH NH2 (1-44) and NH2 (1-29) peptide as a GHRH antagonist and characterize its binding affinities to target.
Aptamers can be easily degraded by plasma nucleases because of their oligonucleotide structure. This challenge may be overcome by some modifications to different positions of nucleotides, phosphodiester bonds, or oligomerization by using L-conformation nucleotides which are not recognized by endonucleases [11]. Aptamers are selected conventionally by a method called SELEX (systemic evolution of ligands by exponential expansion) [26,27]. Nucleic acid aptamers are single-stranded synthetic DNA or RNA oligonucleotides that may fold into three-dimensional structures and interact with many types of targets with a high affinity and specificity [11,13]. Many molecules including small molecules, proteins, peptides, toxins, and even cells can be targets of aptamers [28]. Although generally aptamers were preferred as a capturing key molecule in a biosensor and conjugated with chemotherapeutic agents, aptamer-based hormone targeting technology approaches still maintained. As aptamers are smaller, less immunogenic, with low cost, more stable, more prone to chemical modifications, and renaturation capability after non-physiological conditions, they are more preferable to peptides or antibodies [29]. Due to their structural advantages, aptamers against some target molecules were preferred for metabolic disease treatment. Among these aptamers, Macugen is the first FDA-approved VEGF targeting RNA-aptamer for age-related macular degeneration treatment [12]. Aptamers selected by the SELEX method might be assumed as a time-consuming method due to having a limited number of modifications, and also performing 8-15 selection rounds [29]. Therefore, new methods providing more functional groups and more stability have been developed. One of the newly-developed technologies is X-aptamers, which contain proteinlike or drug-like side chains providing both more stability against nucleases and more affinity to target molecules [14]. Although X-aptamer technology and selection is a newly-emerged system, it is being used in different studies. Recently, potential biomarkers have been identified for schizophrenia by using X-aptamer technology [30]. X-aptamers against two immune checkpoint proteins PD-1 and PD-L1 have also been selected by this technology [31]. In this study, we selected x-aptamers against the growth hormone-releasing hormone by using an Xaptamer selection kit.
In the case of GHRH, it was indicated that, N-terminal 29 amino acids of the 44 amino acids-long GHRH peptide show full biological activity [7]. To study the difference, if any, we used both GHRH NH 2 (1-44) and GHRH NH2 (1-29) peptides for aptamer selection. Further, for purification of the target protein, we needed large-scale protein synthesis in the cells and we used E. coli expressing His-tagged GHRH protein for large-scale protein isolation. Also, since the bacterial protein synthesis system lacks post-translational modifications, we transfected human embryonic kidney HEK293 cells with His-GHRH vector stably and used the protein isolated from these cells as a target. Moreover, increasing technology on nucleic acidbased drug discovery, new generation aptamers synthesis by magnetic bead loaded DNA library was designed. The trademark name for this biotechnological approach was X-aptamer. Thus, we synthesized X-aptamers via targeting both GHRH NH2 (1-44) and NH2 (1-29) by using an X-Aptamer selection kit. Following cloning and sequencing of magnetic bead-based PCR products, 24 putative X-aptamers were synthesized by AM BioTechnologies (Houston, USA) (Fig 2c). Mfold program is generally preferred for the determination of ssDNA Aptamer's schematic profiles and ΔG levels under physiological conditions. As we used new generation technology for synthesizing aptamers; X-aptamer, they are containing modified nucleotides such as phosphorodithioate, indole-or phenol-modified deoxyuridine. So there is not any tool that illustrates the potential secondary structure predictions of X-aptamers. To determine the binding affinity of aptamers against its synthesized ligands, Sypabekova et. al, 2017 and preferred dot-blot methods [32,33]. As our 24 putative X-aptamers have 5' biotinylated labeling, we performed the dot-blot method via targeting GHRH NH2 (1-44) and GHRH NH2 (1-29) peptide (Fig 3a and 3b). According to the dot-blot spot intensity relative folding with aptamer absent spot intensity, TKY2. .09 X-aptamers were significantly high binding affinity against target peptide. When we compared the binding affinity of GHRH NH2 (1-44) x-aptamers with GHRH NH2  peptide, similar binding potential illustrated by GHRH NH2 (1-44) X-aptamers against GHRH (1-44) peptide in dot-blot analysis. Although aptamers against luteinizing hormone [34], thyroxine hormone [35], cortisol hormone [36] synthesized and characterized, there is not any study on GHRH aptamer synthesis. Thus, this is the first study reporting the X-aptamer synthesizing against both GHRH NH2 (1-44) and GHRH NH2 (1-29) peptide and also characterize each synthesized GHRH X-aptamers by dot-blot, immunofluorescence and, SPR analysis.
To evaluate the efficiency of aptamer and the incline impact of aptamers against their target molecules, the selection of appropriate targets is one of the most important steps in the aptamer selection process [29]. Generally, the dot-blot technique is used to determine the K d levels of aptamers with ligands [32]. Besides the dose-dependent aptamer-ligand dot blot analysis, SPR analysis is also preferred to determine the dissociation fold of aptamers. From this point of view, for each putative X-aptamer dose-dependent dot-blot performed. Following Graph-Pad Prism non-linear regression analysis, TKY2.  (Fig 3c). To confirm the results of the dot-blot technique, SPR analysis was performed for selected three putative aptamers (TKY2.T1.13, TKY.T1.08, and TKY.T1.09). SPR mediated K d determination was generally preferred for illustrating the aptamer-ligand binding affinity [37]. In our study, three novel putative X-aptamers against GHRH were TKY2.T1.13, TKY.T2.08, TKY.T2.09 before SPR analysis (Fig 3f). Even if, aptamers are used as a therapeutic agent, the stability of aptamers under specific conditions such as temperature and nuclease-containing conditions. Generally, the serum stability of aptamers was evaluated by mixing an equal volume of aptamer and human serum in a time-dependent manner [38] .09 X-aptamers serum stability was prolonged to 120 h time points (Fig 3e). From so on, when we combine the K d level, coimmunofluorescence profile, and putative secondary structure, TKY2. T1.08, TKY2.T1.13, TKY.T1.08, and TKY.T1.09 X-aptamers selected for the competitive targeting of GHRH-GHRHR or GHRH-aptamer heterocomplexes. Co-immunofluorescence techniques are generally preferred for the binding profile of two ligands. As our putative X-aptamers were synthesized with 5'-biotinylated labeling, we can use streptavidin Alexa Fluor 488 to demonstrate the cell surface binding of X-aptamers on MIA PaCa-2 pancreatic cancer cell line ( Fig  5). Co-immunofluorescence results illustrate the membrane binding of TKY2.T1.08, TKY2. T1.13, and TKY.T2.08 X-aptamers starting from the 50 nM concentrations. Concomitantly, this effect was increased by dose-dependent X-aptamers. However, clear fluorescence intensity was performed following 100 nm TKY.T2.09 treatment in MIA PaCa-2 cell by immunofluorescence staining. These results guide us to confirm the GHRH binding with GHRHR simultaneously with cell surface attachment of X-aptamers. According to co-immunofluorescence results, merge figures illustrate the aptamer binding with GHRH on the surface of MIA PaCa-2 cells. Especially, a significant binding profile was detected for TKY2.T1.08, TKY.T2.08, and TKY.T2.09 only after 250 nM aptamer treatment. Following the binding affinity of GHRH Xaptamers against GHRH secreted from MIA PaCa-2 cells, we demonstrate the blocking impact of GHRH signaling via cAMP levels and GH/GHRHR expression profile. Due to cAMP assay kit results, both TKY2.T1.08 and TKY2.T1.13 X-aptamers declined cAMP levels compared to untreated controls in HT-29 colon cancer cells. In MIA PaCa-2 cells, both TKY.T2.08 and TKY.T2.09 X-aptamers decreased cAMP levels (Fig 6a). Moreover, each selected X-aptamer treatment sharply decreased the expression of both GH and GHRHR expressions as compared to untreated control cells in MIA Paca-2 cells (Fig 6b). Thus, TKY2.T1.08, TKY2.T1.13, TKY. T208, and TKY.T209 X-aptamers block GHRH signaling through suppression cAMP levels and inhibition on GH/GHRHR expression profiles. Finally, further studies may need to figure out the binding site of TKY2.T1.08, TKY2.T1.13, and TKY.T2.08 aptamers on GHRH epitope regions and determine the docking site for GHRHR and GHRH during GHRH-aptamer complex formation. Besides, in silico analysis, in vitro analysis might highlight the inhibitive effect of TKY2.T1.08, TKY2.T1.13, and TKY.T2.08, TKY.T2.09 X-aptamers on GHRH signaling.
In conclusion, this is the first report demonstrating the synthesis, selection, and characterization of four putative X-aptamers against GHRH peptide. Although various peptide antagonists targeting GHRH developed, there is not any X-aptamer targeting GHRH. We synthesized anti-GHRH aptamers by using X-aptamer technology via targeting both GHRH NH2 (1-44) and NH2 . Due to new generation sequencing, 24 putative X-aptamer against GHRH synthesized. Among these x-aptamers, TKY2.T1.01, TKY2.T1.02, TKY2.T1.03, TKY2.T1.04, TKY2.T1.05, TKY2.T1.12, TKY2.T1.13, TKY.T1.01, TKY.T1.02, TKY.T2.07, TKY.T2.08, TKY.T2.09 were demonstrated to be a high binding affinity and low K d levels by dot-blot analysis. Due to serum stability and potential K d levels predictions; TKY2.T1.13, TKY.T2.08, TKY.T2.09 were selected for SPR analysis. In addition, novel two anti-GHRH NH2 (1-44) and two anti-GHRH NH2 (1-29) X-aptamers were characterized according to their binding ability on in vitro systems by using high GHRH expressing and secreting cell lines. Two anti-GHRH NH2  x-aptamers have a high binding affinity, low K d levels, and elevated binding potential in low doses compared to anti-GHRH NH2 (1-44) x-aptamers. Similar to GHRH peptide antagonists, 1-29 amino acid residue of GHRH peptide is more essential for the targeting and synthesizing efficient GHRH binding aptamers. However, new studies might be performed to determine the potential inhibitive effect of these x-aptamers on GHRH signaling and the determination of docking sites for x-aptamers on GHRH ligands. In addition, further in vitro, in vivo, and in silico analysis may need to highlight the impact of these three-novel X-aptamers efficiencies on GHRH signaling.