The phosphoinositide 3‑kinase (PI3K)/protein kinase B (AKT)/ mammalian target of rapamycin (mTOR) signaling pathway is a crucial regulator of cellular processes like proliferation and metabolism, with mTOR kinase acting as its central hub [1, 2]. Its dysregulation is implicated in a wide spectrum of pathologies, ranging from progeria and tuberous sclerosis to neurological disorders and cancer, including breast cancer [1, 3–5]. Consequently, the study of mTOR signaling attracts significant interest from diverse scientific fields, including oncology, aging research, and neurobiology [6–8].

This central role makes mTOR a high‑priority target for novel therapeutics to restore pathway homeostasis. The broader scientific community actively pursues this direction [9]. In our previous work, miR‑162a suppressed osteosarcoma cell proliferation and viability, suggesting potential mTOR inhibition [10]. The lack of a reliable mTOR expression assay prevented direct confirmation of this mechanism, motivating the development of the assay presented.

The high interest in studying mTOR regulation and its functions drives the demand for reliable methods to assess its activity. While established methods like Western blot [11] and luminescence microscopy [12] are reliable, they suffer from high cost, complexity, and operator‑dependence.

An alternative method for evaluating protein activity is to assess the level of gene expression (mRNA quantity) using reverse transcription quantitative real‑time polymerase chain reaction (RT‑qPCR) with relevant reference genes, typically housekeeping genes which are stably expressed in most cells under normal and pathological conditions, for example Glyceraldehyde‑3‑Phosphate Dehydrogenase (GAPDH), BetaActin (ACTB), Beta‑2‑Microglobulin (B2M), as well as some of the most stable ones — TATA-binding protein (TBP) and Ribosomal protein P0 (RPLP0) [13–14].

A multiplex PCR assay for mTOR expression analysis was previously developed by Quidville et al. [15], utilizing the reference genes TBP and RPLP0, which are among the most stable known to date [14]. However, a fundamental flaw in its design — specifically, a probe‑primer overlap for the TBP gene — renders this system unreliable and highlights the critical need for a properly constructed alternative [13, 15, 16].

The aim of this study was to develop and validate a new multiplex RT‑qPCR assay for the relative quantitative analysis of mTOR expression normalized to the reference genes TBP and RPLP0.

METHODS

In silico development and analysis of the PCR assay

Target gene mRNA sequences (FASTA) were retrieved from the NCBI Nucleotide database (https://www.ncbi.nlm.nih. gov/nucleotide/) [Accessed 2024 Sep 23]. Transcripts were annotated using the Assemble by CDS tool in Geneious Prime 2019.2.1 software (Biomatters Ltd., USA). Primers and probes were designed using the PrimerQuest Tool web service (Integrated DNA Technologies, Inc., USA; https://www.idtdna. com/pages/tools/primerquest) [Accessed 2024 Sep 23]. The following selection criteria were used for primer design [17]:

• Specificity — primers complementary to a single region.
• Primer length of approximately 20 nucleotides.
• Guanine‑cytosine (GC) content of approximately 50%.
• At least one guanine‑cytosine base pair at the primer ends.
• No complementary regions between primers.
• Primer melting temperature (Tm) between 55 °C and 65 °C.

The following requirements were applied to the probes:

• Length of approximately 20 nucleotides.
• GC content of approximately 50%.
• No more than four consecutive guanine or cytosine repeats.
• The probe must be located as close as possible to the primers but must not overlap them.
• The probe Tm must be 8–10 °C higher than the primer Tm.

Analysis of primer specificity, flanking, and spanning was performed in Geneious Prime 2019.2.1 software using the BLAST algorithm. Primer dimer formation was checked using the OligoAnalyzer Tool web service (Integrated DNA Technologies, Inc., USA; https://www.idtdna.com/pages/tools/ oligoanalyzer) [Accessed 2024 Sep 23] [17].

In vitro validation of the PCR assay

Cell line

The study used the human bone marrow stromal cell line SCP‑1, which was obtained from the Cell Culture Collection of the Ural State Medical University (Yekaterinburg, Russia). Cells were cultured in T‑25 flasks (Sarstedt, Germany) with an adhesive coating in a medium containing 94% LoSera basal medium (HiMedia, India), 5% Fetal Bovine Serum (FBS) (HiMedia, India), 0.01% L‑glutamine (Servicebio, China), and 0.01% (1 : 1 : 1) solution of penicillin, streptomycin, and amphotericin B (Servicebio, China), in a CO₂ incubator (Panasonic (Sanyo) MCO‑15A, Japan) at 5% CO₂ and 37 °C [10]. Upon reaching 90–100% confluency, cells were passaged by removing the spent medium, washing twice with Hanks' Balanced Salt Solution (HBSS) without Ca and Mg ions (Servicebio, China), and then adding 2.5 mL of 0.25% trypsinEDTA solution (Servicebio, China). The solution was removed after 10 seconds of exposure. The flask was placed in the CO₂ incubator for 1 minute 30 seconds. After adding fresh medium with serum, detached cells were collected, centrifuged at 350g for 3 minutes, resuspended, and distributed to new flasks at a 1 : 3 ratio [18]. For this study, a cell suspension was prepared as described, and cells were counted in a Goryaev chamber using the trypan blue exclusion method.

Nucleic acid extraction

For total RNA extraction, containing the mRNA fraction, the cell suspension (1 million cells/mL) was placed in a 1.5 mL Eppendorf tube and titrated with twofold dilutions in normal saline to obtain samples with concentrations of 1 million, 500 thousand, 250 thousand, 125 thousand and 62,5 thousand cells/mL. RNA was extracted from all dilutions to calculate the analytical parameters of the PCR assay.

RNA extraction from cell culture samples was performed using the «Proba‑NK» reagent kit («DNA‑Technology» LLC, Russia) according to the manufacturer's protocol. After the RNA extraction, total RNA from the 1 million cells/mL suspension was serially twofold diluted in normal saline to obtain samples with the matrix concentrations of 1, 1 : 2, 1 : 4, 1 : 8 and 1 : 16. All RNA matrix concentration variants were used for amplification to validate the PCR assay and calculate PCR efficiency.

Amplification

To analyze the specificity of the amplification product, 10 separate RT‑qPCR technical repeats for each gene were performed using the BioMaster RT‑qPCR SYBR Blue reagent kits («Biolabmix» LLC, Russia). Undiluted RNA matrix from a suspension of 1 million cells/mL was used. Primers, according to the designed sequences, were synthesized by «DNK‑sintez» LLC, Russia. According to the manufacturer's protocol, reaction mixtures were prepared from 12.5 µl of “Reaction Buffer” (containing dNTP mix, Mg ions and SYBR Green), 1 µl of “Master Mix” (containing genetically modified Moloney Murine Leukemia Virus reverse transcriptase and recombinant Taq DNA polymerase inactivated by specific monoclonal antibodies), 5 µl of a mixture with probe, forward and reverse primers (each at 0.12 µmol/L, 3 pmol/reaction), and 6.5 µl of RNA matrix. The final reaction volume was 25 µl. Additionally, 10 multiplex technical repeats (amplification of the target gene and normalizers in one tube) were performed using the BioMaster RT‑qPCR reagent kits («Biolabmix» LLC, Russia). Reaction mixtures were prepared as above, except that the reaction buffer lacked an intercalating dye. Detection was performed by endpoint analysis using horizontal agarose gel electrophoresis.

The assay's analytical performance was assessed using the comprehensive dilution scheme as described in the section Nucleic Acid Extraction. This scheme incorporated both pre‑analytical (cell concentration) and analytical (RNA dilution) variables, enabling simultaneous evaluation of amplification efficiency, analytical sensitivity, and technical reproducibility. All resulting RNA samples — from the five different cell concentrations and the four serial dilutions — were analyzed in triplicate using the multiplex RT‑qPCR protocol described (fig. 1). Three negative control replicates were included. This design resulted in a total of 27 individual reactions (9 unique sample types × 3 replicates). The results from these 27 technical repeats formed the primary dataset for all subsequent calculations of PCR efficiency, reference gene stability, and technical variability.

The amplification protocol consisted of reverse transcription for 30 minutes at 45 °C, initial denaturation for 5 minutes at 95°C, followed by 45 cycles of denaturation (15 seconds at 95 °C) and primer annealing (30 seconds at 68 °C). The elongation step was excluded from the amplification protocol to limit DNA amplification [19]. For real‑time detection of mTOR, RPLP0, and TBP amplification products, the fluorescent channels FAM, HEX, and ROX were selected, respectively.

Real‑time detection was performed, and the Crossing Point (Cp) value was determined by the instrument's software as the cycle number at the maximum of the second derivative of the fluorescence growth curve [20]. As this study focused on the relative quantification of gene expression using the comparative method, no external calibration curves were run alongside the experimental samples. Results were considered valid if the Cp value was detected before the 31^st amplification cycle. This threshold was established experimentally: when using a protocol that included an elongation step, additional amplification products appeared starting from cycle 31, which upon electrophoretic analysis corresponded in length to a fragment amplified from genomic DNA (574 bp, the size of the intron). Exclusion of the elongation step from the protocol eliminated this nonspecific amplification, and no signals were detected after the 31^st cycle. Thus, the 31‑cycle threshold was set as the limit beyond which no contribution from genomic DNA amplification is guaranteed to occur.

All runs used the «DTprime» thermal cycler («DNA‑Technology» LLC, Russia) with the manufacturer's software.

Agarose gel electrophoresis

A 2,5% agarose gel was prepared using 2,5 g agarose powder (Helicon, Russia) in 100 mL 1x TBE buffer («Biolabmix» LLC, Russia) with the addition of 3 µL ethidium bromide (Servicebio, China). Electrophoresis ran at 100 V/A in a Sub‑Cell GT Cell horizontal electrophoresis chamber (Bio‑Rad, USA) for 40 minutes. Results were analyzed on a UV table transilluminator (Vilber, France) using a 10–25 DNA marker (Servicebio, China) [25, 26]. Images were captured digitally and processed in Adobe Photoshop CC 26.11 software (Adobe Inc., USA).

Statistical analysis

The expression value was determined based on the Cp value using the "Fold Change" (FC) formula presented by Livak and Schmittgen [22]. The formula was as follows:

                 FC = 2–ΔCp, ΔCp = Cptarget – Cpreference , 

where Cp_target is Cp value of the target gene and Cp_reference is Cp value of the normalizer.

For the analysis, the ΔCp value was calculated using each normalizer individually (RPLP0 or TBP) and, additionally, using the geometric mean of both normalizers Cp values as a combined reference value.

PCR efficiency for the mTOR gene and reference genes was assessed using linear regression of the Cp value dependence on the log₂ of Cell Equivalent (CE). This metric, expressed in cell equivalents, represents the theoretical number of cells from which the RNA in the reaction was derived, accounting for both cell suspension dilution and RNA dilution. The correlation between the CE in the sample and the Cp value was assessed using Spearman's correlation coefficient (p, considered statistically significant at p < 0.05). The amplification efficiency (E) for each gene was calculated based on the slope (b) of the linear regression line plotted for the dependence of the Cp value on the log₂ of the CE.

Regression analysis yielded the following equation for each gene:

                                                                             y = a – b _• x,

where y is the predicted Cp value, a is the y‑intercept, b is the slope of the regression line, and x is log₂(N), with N being the calculated CE.

The CE was calculated as:

CE = initial cell concentration ÷ cell dilution ÷ matrix dilution.

The stability of the reference genes was assessed by analyzing the distribution of the mTOR expression value normalized to each reference gene (FC) and described by the median with the 0.25 and 0.75 percentiles.

Technical variability of the mTOR FC for a corresponding reference gene(s) was represented as median with the 0.25 and 0.75 percentiles of variability value, which had been calculated according to the formula: variability value = 1 – (mTOR FC sample value / mean mTOR FC value).

mTOR FC values were calculated for the corresponding normalization strategy (normalized to RPLP0, TBP, or the geometric mean of both).

All analyses and graphs were performed in the R environment, version 4.5.2.

RESULTS

In silico development and analysis of the PCR assay

Available methods to exclude genomic DNA amplification include "spanning" — positioning one of the primers or the probe to span two exons — and "flanking" — positioning primers on exons separated by a large intron.

To subsequently check the primers for meeting this condition, sequences (FASTA) of the mTOR (NM_004958.4), RPLP0 (NM_001002.4), and TBP (NM_003194.5) genes were retrieved. Data from human chromosomes 1 (NC_000001.11), 12 (NC_000012.12), and 6 (NC_000006.12), where the target genes are located, respectively, were used as reference sequences, which were then annotated to highlight introns and exons.

Using obtained exons sequences of the genes, the primer pairs and probes for the mTOR and RPLP0 genes, described previously [16], were verified and their effectiveness was confirmed. 10 primer‑probe sets for TBP were designed, and the most specific and suitable combination was selected (table). Primer and probe sequences and parameters are presented in Table.

The proposed normalization strategies and oligonucleotide designs were verified in silico. For the mTOR and RPLP0 genes, primer flanking of exon‑exon junctions was confirmed (exons 21–22 and 2–3, respectively, flanking strategy), and for the TBP gene, probe spanning of the intron between exons 5 and 6 was confirmed (spanning strategy), which guarantees the absence of genomic DNA amplification. According to NCBI reference sequences (NM_004958.4, NM_001002.4, NM_003194.5), the primers and probes are localized on the CDS as follows: for mTOR — forward primer 3381–3402, probe 3411–3435, reverse primer 3437–3464; for RPLP0 — forward primer 95–115, probe 181–205, reverse primer 224–243; for TBP — forward primer 863–882, probe 902–926, reverse primer 930–951.

The primers for mTOR, RPLP0, and TBP did not form stable dimers, indicating a low risk of nonspecific amplification products.

In vitro validation of the PCR assay

When assessing the specificity of amplification products in all separate (n = 10) and all multiplex (n = 10) technical repeats, a single clear amplification product was obtained for each of the target genes — mTOR, RPLP0, and TBP. All amplicon lengths matched the calculated values (table). Side or nonspecific bands were not observed when analyzing amplification products by gel electrophoresis (fig. 2). The analytical specificity of the PCR assay was 100%. To assess analytical sensitivity, various cell dilutions were used; stable detection of all three genes was maintained at a minimum concentration of 125 thousand cells/mL.

A linear regression plot between Cp and the CE (log₂) was constructed for all three genes (fig. 3).

Spearman correlation coefficients for the relationship between Cp and the CE (log₂) were –0.81, –0.84, and –0.77 for mTOR, RPLP0, and TBP, respectively (p < 0.05 in all cases), indicating a strong and statistically significant inverse correlation between the number of investigated cells and the Cp value for each target.

Amplification efficiency was calculated for each gene from the slope of the linear regression. For the mTOR gene, the efficiency was 81%, for RPLP0 — 79%, for TBP — 73%. All efficiency values were obtained based on the deviation of the slope of the experimental regression from the theoretical slope corresponding to 100% efficiency (slope = –1). The obtained values indicate high amplification efficiency of the investigated primers, with the slopes of the experimental lines close to –1 (y = 39 – 0.81x for mTOR, y = 33 – 0.79x for RPLP0, y = 41 – 0.73x for TBP).

The stability of the reference genes was assessed in 27 multiplex technical repeats by analyzing FC of mTOR normalized to RPLP0 and TBP. The median FC value of mTOR normalized to RPLP0 was 0.02 (0.02 – 0.03), indicating a consistently higher expression level of the RPLP0 gene compared to mTOR in the studied samples. When normalized to TBP, the median FC of mTOR was 9.85 (8.57 – 12.13), indicating a consistently lower expression level of TBP relative to mTOR (fig. 4).

To assess variability between technical replicates, the mTOR FC deviation from the mean value was calculated for each sample. The median variability value for mTOR FC when normalized to RPLP0 was –2.3% (–21.5 – 26.4%). The median variability value for mTOR FC when normalized to TBP was 7.2% (–14.3 – 19.2%). When normalizing mTOR FC to both reference genes, the median mTOR FC value was 4.9% (–9 – 23.4%) (fig. 5). In this case, the narrower interquartile range indicates higher reproducibility of results when using the geometric mean of the two reference genes for normalization.

DISCUSSION

In this study, we successfully developed and validated a multiplex PCR assay for determining the expression level of the mTOR gene normalized to the reference genes TBP and RPLP0. The in silico analysis stage involved careful selection and verification of primers and probes based on specificity, thermodynamic parameters, and the absence of stable dimer formation. A crucial achievement was the elimination of a drawback of the previously published assay [15] — the overlap between the probe and primer sequences for the TBP gene, which had called into question the reliability of the results obtained with it. Our proposed assay completely solves this problem.

Preclinical validation of the PCR assay confirmed high analytical specificity: 100% analytical specificity was achieved with the detection of a single, clearly defined amplification product for each target gene in the complete absence of nonspecific products. Furthermore, the developed PCR assay demonstrated reliable analytical sensitivity, providing stable detection at a minimum concentration of 125 thousand cells/mL.

There was stability of the expression ratios of both reference genes RPLP0 and TBP, regardless of cell concentration or matrix RNA concentration, combined with low median variability between technical replicates. This indicates the high reliability and reproducibility of the developed multiplex PCR assay for analyzing mTOR expression levels.

The amplification efficiency for all three genes ranged from 73 to 81%. This range is attributable to two key methodological choices made to ensure the assay’s clinical relevance. First, we validated the protocol using a widely adopted RNA extraction kit to reflect real‑world diagnostic conditions; the resulting RNA purity, while representative, can modestly impact efficiency [23]. Second, to mitigate the risk of amplifying contaminating genomic DNA — a residue not fully removed by even advanced kits — we omitted the elongation step from the amplification protocol [19]. This precaution is particularly relevant given the high processivity of modern Taq polymerases [24]. Despite these necessary trade‑offs, assay specificity remained uncompromised [22]. Future work may focus on fine‑tuning primer concentrations to further improve efficiency.

Despite good median values, the interquartile ranges remain relatively wide. This is explained by the use of the Fold Change metric, based on an exponential formula, which amplifies the influence of even small changes in expression levels on the magnitude of percentage variability, which is an expected result for relative quantitative PCR analysis. The original Cp differences varied within a one‑cycle range.

Such ranges of analytical FC variation fully correspond to the data from most published studies on the development of relative quantitative PCR kits without the use of absolute standards and are consistent with the results of fundamental works [22, 25]. In the study by Gentle et al. [26], the authors report technical FC variation values from 23 to 52%, depending on the number of technical replicates.

A key advantage of this study is the performance of rigorous analytical validation on a standardized cell line. The use of SCP‑1 cells provided a homogeneous and controlled system, free from the pre‑analytical variables inherent in clinical samples (such as tissue heterogeneity or variable RNA quality). This allowed us to accurately determine the intrinsic analytical parameters of the PCR assay — specificity, sensitivity, and reproducibility — without the influence of confounding biological factors. The demonstrated robustness of this PCR assay is a prerequisite for its successful application in the analysis of more complex biomaterial types, such as human tumor tissues, which will be the next step in our research.

Our multiplex RT‑qPCR assay provides a direct and reproducible tool for assessing mTOR expression, capturing the integrated activity of the PI3K/AKT/mTOR pathway [27] and representing a promising objective alternative to operatordependent methods like immunohistochemistry [28]. Unlike detecting mutations in genes like BRCA1/2 or PIK3CA — which, despite their clinical utility, are limited to specific patient subgroups and provide information on predisposition or disease aggressiveness rather than serving as true diagnostic tools — our method quantifies pathway output directly. Given the role of mTOR overexpression in breast cancer [1, 6, 28], this assay holds significant potential for integration into future diagnostic and research strategies.

A limitation of this study is the use of a standardized cell line (SCP‑1), which does not reflect the full heterogeneity of clinical tumor samples and does not account for the influence of pre‑analytical variables [30]. Furthermore, assay performance is inextricably linked to the specific reagents and protocols employed in this work. Therefore, the clinical applicability of the developed test system for assessing mTOR expression in breast cancer diagnosis and stratification must be established through future studies on representative tissue cohorts. Such studies will determine its diagnostic accuracy (sensitivity, specificity, AUC) and evaluate the potential for implementation into routine clinical practice. It should also be noted that increased mTOR expression is observed not only in oncological but also in other diseases (e.g., rheumatic pathologies); therefore, the interpretation of results will depend on the type of biological material studied and the clinical study design. Depending on the specific task, future applications may require either the establishment of diagnostic thresholds for particular sample types (e.g., tumor biopsies) or the introduction of additional markers (e.g., RILP) to differentiate conditions with similar mTOR expression profiles. Despite the need for further research, the developed PCR assay represents a promising foundational tool for advancing molecular diagnostics in breast cancer.

CONCLUSIONS

In this study, we designed and validated a novel multiplex RT‑qPCR assay for the relative quantification of mTOR gene expression normalized to RPLP0 and TBP. A critical flaw of the the previously described design, which consisted of overlapping sequences of the probe and primer for the TBP gene was eliminated in the developed assay through the correct design of oligonucleotides, verification of the absence of primer‑dimer interactions, and the use of optimal strategies to prevent genomic DNA amplification. The assay demonstrated 100% analytical specificity and a sensitivity of 125 thousand cells/mL. The amplification efficiency was 73% for TBP, 79% for RPLP0, and 81% for mTOR. In the SCP‑1 cell line, the expression of mTOR was substantially higher than that of TBP but lower than RPLP0. Normalization to the geometric mean of both reference genes yielded the highest reproducibility, with a median Fold Change deviation of 4.9% between technical replicates. The developed PCR assay allows to overcome a critical design flaw of a previously published assay and, given the role of mTOR overexpression in carcinogenesis, is a promising tool for advancing molecular diagnosis in breast cancer.