NLRP 3 , IL-1 β , and Caspase-1 Gene Transcript Identification and Expression by QCM-D in Obese Children

Laboratorio de Biología Molecular, Escuela Superior de Medicina, Instituto Politécnico Nacional, Plan de San Luis y Díaz Mirón, C.P. 11340 Ciudad de Mexico, Mexico Centro Interdisciplinario de Ciencias de la Salud Unidad Santo Tomás, Instituto Politécnico Nacional, Avenida de los Maestros, C.P. 11340 Ciudad de Mexico, Mexico Centro de Nanociencias y Micro y Nanotecnologías, Instituto Politécnico Nacional, Av. Luis Enrique Erro, C.P. 07738 Ciudad de Mexico, Mexico Sección de Estudios de Posgrado e Investigación, Escuela Superior de Medicina, Instituto Politécnico Nacional, Plan de San Luis y Díaz Mirón, C.P. 11340 Ciudad de Mexico, Mexico Laboratorio de Biología Celular, Universidad Autónoma Metropolitana-Iztapalapa, Av. San Rafael Atlixco, C.P.09340 Ciudad de Mexico, Mexico


Introduction
Quartz crystal microbalance with dissipation monitoring (QCM-D) is a mass-sensitive transducer that provides the opportunity to analyze the interactions between molecules at solid/liquid interfaces.The advantages of the QCM-D on its application in the analysis of interaction mechanisms comprise high sensitivity, in situ, nearly real-time monitoring, and simultaneous mass and viscoelastic measurement [1].QCM-D advantages render it suitable for the development of biosensors, especially DNA and RNA biosensors.Currently, the techniques used for gene expression determination that have greater specificity are qualitative polymerase chain reactions (qPCR) and microarrays.However, these techniques are time-consuming and expensive.Affinity-based biosensors may represent an alternative tool able to quantify the expression of a set of selected genes by determining the amount of corresponding messenger RNA (mRNA) in complex mixtures.The advantages of this approach include avoiding the labelling, retrotranscription, and amplification steps while maintaining high sensitivity and specificity [2][3][4].Mass-sensitive biosensors have attracted considerable attention in recent years due to their simplicity, low instrumentation costs, high sensitivity, and the possibility of performing real-time and free labelling detection.The QCM-D system belongs to mass-sensitive biosensors.Its operating principle relies on the piezoelectric crystal properties: the resonant frequency shifts are directly proportional to the adsorbed mass [2,5].Quartz crystal microbalance (QCM) is suitable for biosensors, including DNA biosensors.Lucarelli and colleagues implemented a QCM biosensor for the detection of a singlebase DNA mutation, while the Jearanaikoon's group built a QCM-D biosensor to monitor conformational changes during DNA hybridization [4,6].Song et al. measured in real-time human papillomavirus 16 DNA through a loopmediated isothermal amplification-QCM biosensor [7].Bunroddith and colleagues developed a QCM sensor for the ultrasensitive detection of DNA with results comparable to the ones obtained using a SPR sensor platform [8].Ravi et al. reported a rapid screening method of Ehrlichia canis PCR amplification products based on PCR assay combined with QCM [9].Regarding RNA biosensors, Tedeschi and coworkers reported that they successfully achieved a QCM label-free RNA biosensor with high specificity, reusability, and the ability to provide quantitative information [2].There are reports about cDNA quantification and gene expression biosensor development.Han et al. reported a cDNA quantification platform using a giant magnetoresistive (GMR) biosensor array [10], meanwhile Papadakis et al. designed an electrochemical platform for direct detection of CYP707A1 gene expression from wheat leaves [11].Davis et al. developed a biosensor for SNP genotyping and gene expression quantification combining conventional PCR for target amplification and QCM for amplicon detection [12].Nevertheless, there are no reports of a biosensor that determine neither gene transcripts nor their expression in children with obesity.
Obesity is considered a risk factor for different comorbidities and has been related with high morbidity and mortality [13][14][15].It has been shown that obesity is a state of chronic inflammation characterized by macrophage infiltration and that the gene expression of molecules related with proinflammatory processes plays an important role during development of the disease [16].In Mexican population, obesity prevalence has increased very fast; it is estimated that in 2050, more than one half of the adult population will suffer from obesity (54%) [13,15,17].Immune response varies with respect to the life stage [18]: in children and adults with obesity, cytokines IL-1β and NLRP3 are elevated [19,20].It has been reported that adipose tissue in obese population influences the secretion of pro-and anti-inflammatory cytokines such as tumor necrosis factor alpha (TNF-α), interleukin 1 (IL-1), IL-8, caspase, and inflammasomes; the secretion of these molecules promotes comorbidities such as type 2 diabetes, steatosis, and cardiovascular diseases [13,16,21].Therefore, the goal of the present work was to identify NLRP3, IL-1β, and CASPASE 1 (CASP-1) gene transcripts, as well as their expression, from the blood samples of children with obesity by means of QCM-D biosensors.Precisely, the novelty of this biosensor lies in the identification of gene transcripts and their expression, which is of vital importance for the molecular diagnostic of different pathologies.

2.1.
Patients and Blood Samples.This protocol was approved by the Ethics and Research Committees of the Escuela Superior de Medicina (ESM-CE-01/03-12-2013).Written and informed consent was obtained from all participants.The trial was conducted in accordance with the ethical principles originating in the Declaration of Helsinki and it was consistent with Good Clinical Practice Guidelines (GCPG).Body mass index (BMI) was utilized to classify the patients.Patients were randomized in a group of children with obesity (n = 30).A sample of blood (4 mL) was obtained from the antecubital vein of the child's arm and used for RNA extraction.These samples were employed for determining the gene transcripts and expression of NLRP3, IL-1β, and CASP-1.
2.2.RNA Extraction and cDNA Synthesis.Human RNA was isolated using the TRIzol reagent according to the manufacturer's instructions (TriPure Isolation Reagent; Roche Applied Science, Indianapolis, IN, USA).The amount and purity of RNA was quantified nanophotometrically by measuring optical density (OD) at 260 and 280 nm.Integrity was checked for all samples by agarose gel electrophoresis.To avoid trace amounts of DNA contamination, RNA samples were treated with amplification-grade DNase I (Invitrogen, Carlsbad, CA, USA) before reverse transcription.All RNA samples were stored at −80 °C in RNA elution solution.The first-strand cDNA Synthesis Kit (Roche Diagnostics, GmbH Mannheim, Germany) was utilized, taking 1 μg of the total sample RNA; reverse transcription was performed with random hexamers primers following the instructions specified by the manufacturer, in an Eppendorf thermocycler.Finally, cDNA concentration, obtained from human RNA, was determined through nanospectrophotometric measurement.This cDNA was used as capture probe for the QCM-D biosensor experiments.
2.3.Surface Chemistry, cDNA Immobilization, and QCM-D Hybridization Assays.The reagents used for the surface chemistry over gold were 6-mercapto-1-hexanol (MCH), N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), and PBS 10X buffer, pH 7.4.All reagents for the surface chemistry were purchased from Sigma-Aldrich, Mexico.All chemicals were reagent grade and utilized without further purification.We used quartz crystals (Q-Sense sensors) coated with 2 Journal of Sensors 100 nm gold thin film for the QCM-D experiments, with a frequency of 4 95 MHz ± 50 kHz, cut AT, a diameter of 14 mm, a thickness of 0.3 mm, optically polished, and purchased from Biolin Scientific Inc., USA.QCM-D experiments were performed in a Q-Sense E4.The quartz crystals were washed with (type 1 water, ammonia, and hydrogen peroxide) HPA solution and placed into ultraviolet (UV) ozone.After this, they were incubated in a 1 mM MCH solution and isolated from light for 48 h at room temperature.
Once incubation was completed, the quartz crystals were washed with ethanol and dried with nitrogen flux.The quartz crystals were then incubated for 2 h in PBS 10X (2.5 mL) with 5 mM EDC (0.024 g) and 8 mM NHS (0.023 g), pH 8.The functionalized quartz crystals were placed in the QCM-D, and water was injected for 1 h to stabilize the frequencies and dissipations.Next, a previously denaturalized 3.5 ng concentration of cDNA corresponding to the genes NLRP3, IL-1β, CASP-1, 18S (positive control), and NLRP10 specific for rat (negative control) was injected into each of the QCM-D modules and incubated for 2 h.In order to achieve this concentration, a pool of the cDNA samples of the children with obesity was employed.After, the quartz crystal surfaces were exposed to UV light for 3 min by using a UV transilluminator (Stratalinker UV).Finally, specific oligonucleotides designed to hybridize with each of the capture probe cDNAs were injected in the corresponding QCM-D modules for hybridization for 1 h.The specific oligonucleotides used in the QCM-D hybridization assays were NLRP3, 5 ′ -CACCTGTTGTGCAATCTGAAG-3 ′ (forward) and 5 ′ -GCAAGATCCTGACAACATGC-3 ′ (reverse); IL-1β, 5 ′ -CTGTCCTGCGTGTTGAAAGA-3 ′ (forward) and 5 ′ -TTGGGTAATTTTTGGGATCTAC-3 ′ (reverse); CASP-1, 5 ′ -GCGAAGCATACTTTCAGTTTC-3 ′ (forward) and 5′-TCTCCTTCAGGACCTTGTCG-3′ (reverse); 18S, 5′ -CGAACGTCTGCCCTATCAAC-3 ′ (forward) and 5 ′ -TTGGATGTGGTAGCCGTTTC-3 ′ (reverse); and NLRP10 specific for rat, 5′-TCAACCAGGTTTGTCTGAATGA-3′ (forward) and 5 ′ -CTGTTTACACCCCAATCTTGC-3 (reverse).The standardized concentration of these oligonucleotides for the QCM-D hybridization assays was 3.5 ng/μL.This functionalization was specifically designed to allow bonding of cDNA obtained directly from human RNA without any label or chemical modification (patented by our group: MX/E/2014/057816).The corresponding data for mass and changes in frequency was registered continuously during the entire experiment.QCM-D hybridization assays for each of the genes were done in triplicate.The results obtained were analyzed with QTools software from Q-Sense.

Atomic Force Microscopy (AFM).
Finally, the topography of the functionalized, anchored, hybridized, and treated with DNase quartz crystals was explored using AFM.DNase in an enzyme degrades DNA exclusively; thus, DNase treatment allows verifying that the molecule attached is DNA.The AFM instrument used was a diMultiMode from Veeco (Santa Barbara, CA) in contact air mode with diNanoScope V controller, using RTESPA tips.All the images were visualized and analyzed using the NanoScope analysis software.
2.5.Real-Time Quantitative Polymerase Chain Reaction (qPCR).In this research, qPCR was used as a molecular biology validation technique; qPCR is a validated method used in molecular biology to determine genic expression [22,23].qPCRs were carried out using specific oligonucleotide primers that were originally generated by using online assay design software (https://qpcr.probefinder.com/organism.jsp).In order to compare and validate the results obtained in the QCM-D hybridization assays, the specific oligonucleotides used were the same in both methods.The reaction mixture was prepared according to the manufacturer's instructions (Roche Diagnostics, GmbH) utilizing 1 μL of standard cDNA at the appropriate dilution.The amplification was performed in LightCycler 8-Tube Strips (Roche Diagnostics) with a LightCycler nanoinstrument.Amplification conditions for the Universal ProbeLibrary-(UPL-) based assays were initially denaturized for 10 min at 95 °C, followed by 45 cycles of 10 s at 94 °C, of 20 s at 60 °C, and of 5 s at 72 °C.The qPCR assays included the standard curve of four serial dilution points for each gene, and mRNA levels were calculated by employing the comparative parameter threshold cycle (Ct) method and normalized to the endogenous control: 18S mRNA.qPCR experiments require a standardization of the reaction efficiency curves for gene expression.In the formula to standardize the results (2 ΔΔCt ), number two corresponds to the qPCR efficiency reaction (dynamic range curve or dilutions of the constitutive gene).
According to the guidelines and specifications of the Roche Nano Thermal Cycler software, the efficiency must be 2, so it is necessary to make serial dilutions of a sample chosen randomly from an experiment.These logarithmic dilutions estimate the accuracy of the concentration.Results were calculated as a percentage of the mean level found in control using the ΔCt method.
The qPCR primer pairs used for amplifying the NLRP3, IL-1β, and CASP-1 fragments were the same as the ones used for the QCM-D hybridization assays.Finally, the results of the genic expression are expressed as the mean ± standard deviation (SD).Gene expression and deposited mass data were analyzed with one-way analysis of variance (ANOVA), and the comparison between the anchored DNA and the hybridization assay was analyzed using the Student t-test from SigmaPlot ver.12.0 software (San Jose, CA, USA), while differences were considered statistically significant when p ≤ 0 05.

Gene Transcript Identification and Expression by QCM-D.
The first step to achieving the goal of this work was to accomplish a stable functionalization based on MCH over the QCM sensors, obtaining covalent bonds between the gold surface and the cDNA, as shown in Figure 1.For covalent binding of cDNA target, the crystal surfaces were exposed to 254 nm UV light with a power of 23 mW/cm 2 for 3 min by using a UV transilluminator (Stratalinker UV).The positively charged carbon atom in carbodiimides rapidly reacts with the partially negatively charged oxygen in the 5 ′ -phosphate group forming an active phosphodiester.

Journal of Sensors
After the quartz crystals were functionalized, the hybridization assays of the NLRP3, IL-1β, and CASP-1 genes were performed by means of the QCM-D.In the hybridization assays, all the harmonics were used to keep track of the changes in frequency and dissipation changes.One of the sensors was utilized as a control sensor in which type 1 water was hybridized.The results of this sensor, shown in Figure 2, revealed no changes in frequencies and dissipations when the QCM-D module was filled with type 1 water (for the initial hybridization analysis, the technical specifications of the QCM-D suggested the use of the harmonics 5, 7, and 9 avoiding noise signals).This means that there was no deposition of nucleic acids interacting with the functionalized surface.
Once no interaction with the functionalized surface of the control sensor was proven, hybridization assays with the respective genes were performed.QCM-D hybridization assays for each gene were done in triplicate.First, the hybridization assay of the NLRP3 gene was carried out.The results of the NLRP3 gene hybridization assay are shown in Figure 3. cDNA was anchored during the first 2 h 46 min, reflecting changes in frequencies (2.5 to −0.3, 2.2 to 0.1, and −0.3 to 0.6 Hz, respectively).Later, the respective oligonucleotides were injected for hybridization with NLRP3 gene-specific oligonucleotides, obtaining a decrease in the frequencies (0.6 to −9.8, 0.1 to −8.7, and -0.3 to −7.5 Hz) and dissipations (0 to 0.38, 0.1 to 0.26, and 0 to 0.1). Figure 3 The second hybridization assay was conducted with the IL-1β gene.These results are shown in Figure 4. From time 0-2 h 46 min, cDNA immobilization (Figure 4(a)) was accomplished.A decrease in the respective frequencies is observed (ranging from 0.3 to −3.0 to −3 and 0 to-2 Hz) when anchoring the cDNA, as well as when hybridizing the oligonucleotides with the IL-1β gene (−3 to −14, −3 to −12, and −10.5 Hz).An increase in dissipation was also reported when anchoring the cDNA (0 to 0.22, −0.2 to 0.3, and 0.1 to 0.2), as well as when the specific oligonucleotides for hybridizing the IL-1β gene were injected (0.22 to 3.4, 0.3 to 2.8, and 0.2 to 2.2).An increase in mass deposition on the quartz crystal (p = 0 001) was observed when graphing the data, as well as an increase in viscoelasticity (p = 0 05) (S3 and S4 supplementary materials).
The results of the CASP-1 gene hybridization assay are presented in Figure 5. From time 0 to 2 h 46 min, immobilization of the cDNA demonstrates a decrease in frequencies in the different harmonics (4 to 2.5, 3 to 1.5, and 1.5 to −0.3).At the time of hybridization with the specific oligonucleotides for the CASP-1 gene, a decrease in frequencies is also observed (2.5 to −3, 1.5 to −5.5, and −0.3 to −8.5).On the other hand, the dissipations increased when anchoring the cDNA (0 to 0.22, −0.2 to 0.3, and 0.1 to 0.2) for each harmonic and when hybridizing with the CASP-1 gene      Journal of Sensors oligonucleotides (0.22 to 3.4, 0.3 to 2.8, and 0.2 to 2.2).In addition, an increase in mass concentration both when anchoring the cDNA and when hybridizing with the specific oligonucleotides was observed (Figure 5(b)), as well as an increase in viscoelasticity.By plotting the deposited mass values (S5 supplementary materials), significant differences in mass increase can be noted by hybridizing with the specific oligonucleotides for CASP-1 (p = 0 001).Otherwise, in Figure S6 supplementary materials, an increase in viscoelasticity can be noted (p = 0 012).
In order to prove the specific hybridization, the constitutive gene 18s of eukaryotic cells was tested as positive control and the NLRP10 rat gene that encodes exclusively for the rat genome as negative control.The results of the 18S gene hybridization assay are demonstrated in Figure 6(a).From time 0 to 2 h 30 min, cDNA bonding took place, and a decrease in the frequencies was observed (0 to −3.2, −0.2 to −3.8, and −0.2 to −3.2 Hz).Subsequently, specific oligonucleotides were injected to hybridize with the constitutive 18s gene, revealing a decrease in frequencies (−3.2 to −7.8, −3.8 to −7.8, and −3.2 to −8 Hz) and an increase in dissipations when anchoring the cDNA (−1.1 to −0.4, −0.14 to −0.2, and −0.24 to 0.02 Hz), as well as when hybridizing with the specific oligonucleotides for 18S (−0.02 to 0.08, −0.2 to 0.2, and −0.4 at 0.1 Hz). Figure 6(b) presents an increase in deposited mass concentration and low viscoelasticity.On the other hand, when plotting mass deposition on quartz (S7 supplementary materials), a statistically significant increase in mass deposition (p = 0 001) and an increase in viscoelasticity (p = 0 001) (S8 supplementary materials) were observed.Once realized the QCM-D hybridization assays of all tested genes (NLRP3, IL-1β, CASP1, 18S, and NLRP10-Rat), the number of molecules (N) on the surface of the quartz was calculated using the Sauerbrey equation [18], considering the frequency shifts and mass changes on the surface.As shown in Table 1, IL-1β presents the greatest number of immobilized and hybridized molecules, followed by 18S (positive control) and NLRP3, being CASP-1 the one with the least molecules immobilized and hybridized.On the other hand, the negative control NLRP10-rat showed no molecules attached.

Atomic Force Microscopy (AFM).
The images of the different quartz crystals in each of the experimental phases were taken with the diMultiMode AFM from Veeco (Santa Barbara, CA) in contact air mode with diNanoScope V controller.As shown in Figure 8(a), there is no visible structure deposited on the quartz, while in Figure 8(b), the deposited cDNA can be observed, and in Figure 8(c), it can be observed that the cDNA was hybridized with the specific oligonucleotides, which allowed to observe the deposition of nucleic acids with greater compaction than when only cDNA was deposited (Figure 8(b)).In Figure 8(d), the absence of material deposited on the quartz can be observed once it was treated with DNase I, an enzyme that digested the cDNA previously anchored and hybridized.
3.3.Real-Time Quantitative Polymerase Chain Reaction (qPCR).The gene expression levels of inflammatory genes NLRP3, IL-1β, and CASP-1 were determined from the blood samples of children with obesity.Expression of the genes encoding for NLRP3, IL-1β, and CASP-1 demonstrated differential expression in the study group.The expression of mRNA for NLRP3 was statistically increased when compared with that of IL-1β (p = 0 008) and CASP-1 (p = 0 041), as shown in Figure 9. On the other hand, IL-1β was increased compared to CASP-1, although no statistical differences were found between them, as observed.Figures 9(a) and 9(b) present the correlation between the gene expression results and the mass deposited on the quartz crystal of the NLRP3 gene (first bar), the IL-1β gene (second bar), and the CASP-1 gene (third bar), respectively.Proving that greater the gene expression determined by qPCR greater is the deposited mass detected by QCM-D.This same pattern was observed in the genes that presented lower genetic expression, which showed lower deposited mass.Determining the viscoelasticity for each of the genes allowed proving the hybridization between the DNA complementary strands, S11 (supplementary materials).
QCM-D is a mass-sensitive transducer that provides the opportunity to analyze the interactions among molecules at solid/liquid interfaces.The advantages of the QCM-D, applied in the analysis of interaction mechanisms, are high sensitivity, in situ, real-time monitoring, and simultaneous mass and viscoelastic measurement [1].QCM-D has been employed to analyze enzymatic reactions, detecting mass changes on the surface [24,25].On the other hand, experiments developed with QCM-D that involved nucleic acids focused on DNA binding, RNA transcription or elongation, and the relationship between RNA polymerase and DNA.Despite that the elongation process has not been well characterized, double-stranded DNA (dsDNA) was immobilized on the quartz surface in order to analyze the transcription reaction.The results showed nanogram-level mass increase and a decrease in frequency in real time [26][27][28].Regarding RNA biosensors, Tedeschi et al. successfully achieved a QCM label-free RNA biosensor with high specificity, reusability, and the ability to supply quantitative information.This sensor was able to bind the RNA under investigation through a complementary oligonucleotide probe by means of covalent immobilization [2].This previous work was very important because it was, to our knowledge, the first report of a RNA biosensor developed using QCM-D.However, synthetic RNA was utilized for the experiments and the authors only proved the binding ability of RNA probes using both a synthetic substrate and full-length "in vitro" transcribed mRNA.Regarding gene expression biosensors, Han et al. developed a 7 Journal of Sensors cDNA quantification platform using a giant magnetoresistive (GMR) biosensor array.The gene expression analysis relied on hybridization of complementary DNA strands.GMR chips were functionalized with amino-modified ssDNA probes and then spotted with these probes complementary to the GAPDH gene.Total mRNA from HeLa cells was isolated, reverse transcribed, and qPCR amplified with biotinylated primers to create cDNA, which is denatured and applied as ssDNA target to  8 Journal of Sensors the sensor and allowed to hybridize to the probes on the chip.After, streptavidin magnetic nanoparticles were added to the sample in order to amplify the signal.The authors concluded that cell line DNA had certain concentration through qPCR standard curve analysis, which was confirmed with GMR analysis [10].Meanwhile, Davis et al. reported an acoustic biosensor using, in order to bind DNA to the quartz crystal microbalance a neutravidin-modified device surface, which allows the binding of biotinylated-DNA molecules.They specifically designed sets of primers biotinylated at the 5′ end and used as template cDNA obtained from rat liver with and without treatment, in a multiplexed PCR reaction, which produced two different DNA fragments that were loaded on the neutravidin-covered acoustic device surface.Finally, the acoustic ratio was directly measured reflecting the mass ratio of the two immobilized fragments [12].Instead, in the present work, the RNA sample employed was obtained from the blood samples of children with obesity and transcribed to cDNA, which is a more stable molecule to work with.The cDNA used in our experiments as target in the biosensor is a faithful, freelabeled copy of the patients genomic DNA.A key feature in biosensor development is the target used.To our knowledge, this is the first report of a biosensor developed with human cDNA as target, as well as the first biosensor designed to identify gene transcripts and their expression, from children with obesity.This biosensor could be applied for the molecular diagnosis of several pathologies.Obesity is a state of chronic inflammation characterized by the expression of molecules related with proinflammatory processes such as the inflammasome.Inflammasome-type components depend on a specific antagonist or stimulation involved, but the final result of all inflammasomes is the assembly and activation or maturation of pro-IL-1b and pro-IL-18 active forms, through the Caspase 1 pathway  9 Journal of Sensors [29][30][31][32][33], finding differences among all components.It has been described that Caspase 1 presents high expression in adipose animal tissue and in humans with obesity [30,34], possessing a very important function in determining inflammatory characteristics in human adipose tissue [35].Due to the importance of these inflammatory molecules in human adipose tissue, in this work, QCM-D was utilized to determine the hybridization of five different genes; three of these were target genes (NLRP3, IL-1β, and CASP-1), a positive control (the 18s constitutive gene), and a negative control (NLRP10, which is exclusively expressed in rats), thus proving the functionalization previously developed by our working group (patent pending, MX/E/2014/057816).The novelty of the functionalization as well as the biosensor described in this study is precisely the ability to immobilize cDNA or DNA obtained directly from human samples, without any label or chemical modification.This allows the identification of specific genes for the molecular diagnostic of pathologies as cancer, hypertension, and diabetes among others.The cDNA used in our experiments as capture probe in the biosensor is a faithful copy of the patient genomic DNA.Once the cDNA is synthesized, it is used as capture probe without any label or chemical modification, allowing to spend less time in the development and use of the  Journal of Sensors biosensor.Previous works have reported the immobilization of double-stranded calf thymus DNA, purchased from a laboratory [36]; nonetheless, the majority of the probes used in biosensors are linear oligonucleotides, in the order of 18-25 bases synthesized in situ, presynthesized or hairpin oligonucleotides.Furthermore, peptide nucleic acids (PNAs) and locked nucleic acids (LNAs) have also been used as capture probes.PNAs are DNA mimics in which the bases are attached to a neutral N-(2-aminoethyl)-glycine pseudopeptide backbone, and LNAs are nucleic acid analogues in which the ribose ring is "locked" by a methylene bridge that shows increased thermal stability [4].Other capture probes used are chemically modified at one of the oligonucleotide ends, and usually, biotin, thiol, or amine residues are aggregated, biotin and thiol being the most used chemical modifications [7][8][9][37][38][39][40].Moreover, the surface chemistry used in this study enabled the immobilization of cDNA, which allowed developing a biosensor for gene expression determination from children with obesity.Besides being a proof of concept, this is the first biosensor of this type reported.
The results show that it is possible to detect the immobilized cDNA obtained from human RNA, since the QCM-D reports an increase in the deposited mass, as well as a decrease in the frequencies in nearly all of the genes studied, except for the negative control, in which no significant differences were found when hybridization with the specific oligonucleotides took place.Table 2 compares the frequency changes monitored during hybridization measurements for each of the assayed genes.These results are in agreement with the ones reported for QCM-D experiments; upon addition of a biomolecule, a rapid frequency decrease is observed, due to mass increase.Also, an increase in dissipation is reported [41].Analysis of the corresponding results of the mass detected for each of the genes demonstrated a major detected mass for IL-1β, followed by similar NLRP3 and constitutive gene (18s) concentrations and a smaller concentration for CASP-1.These results suggest that IL-1β could be the molecule with greatest expression because it has the greater amount of mass, followed by NLRP3.On the other hand, the increase in dissipation is related to a viscoelasticity increase.It has been reported that when DNA is hybridized with its complementary strand, an increase of dissipation due to the stiffening of the DNA is observed [42].This stiffening has to do with the hydrogen bond formation and hydration that occurs when complementary DNA strands are hybridizing, making the double-stranded DNA more stable [43].The dissipation results obtained in this study are in agreement with these results; the QCM-D assays showed an increase in dissipation, for NLRP3, IL-1β, and CASP-1, when the hybridization took place.The number of molecules anchored and hybridized of each of the assayed genes was also calculated according to Thanh Ngo 2014.The results obtained for each of the genes agree with the amount of mass reported by the QCM-D.Surprisingly, when comparing mRNA gene expression results for NLRP3, IL-1β, and CASP-1 obtained with qPCR and QCM-D, similar patterns were found, revealing greatest expression of IL-1β, followed by NLRP3, with CASP-1 being the molecule of least expression in the group of children with obesity.
These results are important, because gene expression in children with obesity was obtained with very similar patterns between two different methods: a novel cDNA QCM-D biosensor and a validated method used in molecular biology, qPCR.These results also demonstrate that qPCR is a validation technique for QCM-D when working with gene expression.A possible explanation for these results is that the transcription reaction was done with random hexamer primers, as specified under Methods; the use of these primers allows transcription of all the mRNAs present in the sample, making it possible to detect greater expression of a transcript.The probes that are being detected by QCM are the cDNA transcripts to which the specific oligonucleotides are completely binding.This union allows the identification of gene transcripts through frequency changes in the QCM, exploiting the principle of hybridization of complementary DNA strands, without the use of the enzyme Taq polymerase and the fluorophores that are normally employed in qPCR for gene expression quantification.Gene expression is identified when a higher number of oligonucleotides are attached to the cDNA transcripts bonded in the quartz when hybridizing.Nevertheless, the experiments were repeated three times for each assayed gene, the stability was proven at the beginning of each QCM-D experiment when the fundamental lectures were taken, and the sensitivity was determined since the QCM-D was able to find different expression levels for each of the analyzed genes and these corresponded with the expression levels determined by qPCR.
At present, AFM comprises a powerful tool employed in the structural study of biological macromolecules such as proteins and DNA.It allows for visualization of these molecules and its complexes deposited over a surface with nanometer resolution, preserving their native structure [44].AFM has been utilized to characterize protein-protein and protein-DNA interactions and conformations; it also allows determining the specificity of proteins binding to specific sites on DNA [45,46].Due to these features, AFM was used in this work to explore the changes occurring on the surface of the quartz crystal during DNA immobilization and hybridization processes.As well as in the work reported by Beckwitt's group, AFM imaging of anchored cDNA reveals "worm-like" features on top of the gold surface.These "worm-like" features disappear when the quartz crystal is treated with DNase I.Both in the present study and in those reported by other groups, AFM is a proven technique that provides advantages in terms of the imaging and characterization of DNA biosensors, allowing direct, label-

Conclusions
In conclusion, QCM-D can be used for determining the gene expression of transcripts in children with obesity, with similar results as the ones obtained with qPCR, a validated method used in molecular biology to determine genic expression quantification.To our knowledge, this is the first work that identifies gene transcripts and their expression utilizing a biosensor such as QCM-D, in children with obesity.In addition, this work compares both the use of biosensors and a classical molecular biology technique to validate the former, obtaining amazingly similar values.The surface chemistry, which was previously designed by our working group, proved to be useful for cDNA immobilization.The QCM-D biosensor developed by our group was successful in gene expression determination, as proof of concept, using qPCR as validation method.This QCM-D biosensor despite being a proof of concept presents sensitivity, since expression level differences for each of the tested genes were detected, reusability, and stability, as well as in situ and nearly realtime monitoring.Furthermore, QCM-D decreases costs compared with a validated molecular biology technique, as qPCR.Further studies are needed to demonstrate the reproducibility of QCM-D as a novel technique for gene expression determination, using different samples and specific oligonucleotides for hybridization.Meanwhile, gene expression determination experiments with QCM-D require to validation with traditional molecular biology techniques, as qPCR.
(b) depicts an increase in the deposited mass concentration in each of the phases.On the other hand, the viscoelasticity of the sample increased as the amount of mass deposited increased.When plotting mass quantity at different times of the sample deposit (S1 supplementary materials), a significant increase (p < 0 001) in the deposited mass at the time of hybridizing the bonded cDNA with the NLRP3 oligonucleotides was observed.Viscoelasticity increased (S2 supplementary materials) upon hybridization with cDNA (p = 0 001).

Figure 1 :Figure 2 :
Figure1: cDNA immobilization on the QCM gold surface.In Figure1, the quartz crystal surface chemistry based on a MCH SAM formation step by step is shown.Once the quartz crystal was functionalized, the cDNA obtained from human RNA was immobilized and hybridized with each of the oligonucleotides designed specifically for each of the genes tested in the QCM-D.

Figure 3 :
Figure 3: Monitoring of anchorage of the cDNA anchorage and hybridization with NLRP3 oligonucleotides by means of the QCM-D.The first phase reveals changes in dissipation (ΔD) and frequency (ΔF) when cDNA was anchored, followed by hybridization with NLRP3-specific oligonucleotides (a).In graph (b), the deposited mass is plotted in each of the phases.

Figure 4 :
Figure 4: Monitoring the anchoring and hybridization of cDNA and IL-1β (oligonucleotide) by means of QCM-D.In a first phase, changes in dissipation (ΔD) and frequency (ΔF) are reported on anchoring cDNA, followed by hybridization with IL-1β.

Figure 5 :
Figure 5: Monitoring the anchoring and hybridization of cDNA and CASP-1-specific oligonucleotides by means of QCM-D.In a first phase, changes of dissipation (ΔD) and frequency (ΔF) when anchoring cDNA are depicted, followed by hybridization with CASP-1 (a).In (b) deposited mass is plotted.

Figure 7 (
b) demonstrated low concentration in the mass detected after 2 h.No significant changes were observed at the time of plotting mass and viscoelasticity (S9 and S10 supplementary materials).

Figure 6 :
Figure 6: Monitoring the anchoring and hybridization of cDNA with oligonucleotide 18S by QCM-D.In the first phase, changes in dissipation (ΔD) and frequency (ΔF) when immobilizing cDNA are shown, followed by hybridization with 18s (a).In (b) deposited mass is plotted.

Figure 7 :
Figure 7: Monitoring the anchoring and hybridization of cDNA and NLRP10-rat oligonucleotides by QCM-D.In a first phase, dissipation (ΔD) and frequency (ΔF) changes are shown when anchoring cDNA, followed by hybridization with NLRP10 (a).In (b) deposited mass is plotted.

Figure 8 :
Figure 8: Gold surface substrate AFM images: (a) functionalized gold substrate, (b) gold substrate with immobilized cDNA, (c) gold substrate with immobilized cDNA and hybridized with the specific oligonucleotides, and (d) after digestion of cDNA with DNase I.

Figure 9 :
Figure 9: NLRP3, IL-1β, and CASP-1 gene expression.Each bar represents the average of 20 individuals.Graph (a) reveals the results of NLRP3, IL-1β, and CASP-1 gene expressions.Graph (b) depicts the deposited mass obtained from the QCM-D experiments.
free imaging at the single-molecule level of the process being undergone[46].

Table 1 :
Number of molecules anchored and hybridized for each of the genes tested on the surface of the QCM.

Table 2 :
Average frequency changes ± standard deviation monitored during hybridization measurements.