Introduction

Klotho gene was discovered in 1997 as an anti-aging gene (Kuro-o et al. 1997). The Klotho gene has two mRNA first variant encodes a longer transmembrane form (mKL) of the protein, while the second splicing option results in a truncated secreted one (sKL). The mKL is a single-pass transmembrane protein that includes 1012 amino acids (around 135 kDa). The mKL comprises a large extracellular domain with two homologous KL1 and KL2, a transmembrane domain and a short intracellular domain, while the sKL is twice shorter and includes 549 amino acids (around 65 kDa). Although there is an almost twofold difference in amino acid length, mRNAs of these proteins are poorly distinguishable, and the insertion of 45 bp is the mere of the established. The latter is due to alternative splicing and results in the nonsense (stop) codon; thus, the translated protein comprises three exons (Matsumura et al. 1998). It should be noted that mKL might also be shed, resulting in the shKL protein. The latter is capable of getting into the bloodstream, cerebrospinal fluid, and urine (Chen et al. 2014).

The involvement of Klotho in carcinogenesis has been shown in breast cancer (Rubinek et al. 2012; Wolf et al. 2010, 2008), lung cancer (Chen et al. 2012, 2010; Wang et al. 2013), liver cancer (Chen et al. 2013; Tang et al. 2016; Xie et al. 2013; Sun et al. 2015), and other oncological diseases. For example, Klotho expression in the tumor was examined and compared with similar healthy surrounding tissue (Rubinek et al. 2012; Chen et al. 2013; Tang et al. 2016; Yan et al. 2017). Also, the assessment of the methylation status of the promoter region is carried out (Rubinek et al. 2012; Xie et al. 2013; Lee et al. 2010). Moreover, its involvement in carcinogenesis was the scope of many studies with both gene knockout and overexpression. The means of genetic transfection or adding purified Klotho protein to the cell cultures are employed to achieve these tasks and are usually followed by evaluating the cell growth features (Chen et al. 2012, 2010; Tang et al. 2016; Yan et al. 2017; Lee et al. 2010).

In general, the data collected up to date justify the decrease of Klotho expression in tumor cells compared to the surrounding normal tissue. The latter corresponds with initial observation in breast cancer. It is also assumed that the methylation of the gene promoter and involvement of insulin and insulin-like growth factor 1 (IGF-1) play a partial role in the antitumor effect of the Klotho gene (Wolf et al. 2008). The following studies suggest the effect of insulin/IGF-1 on the canonical proteins of the Wnt family (Tang et al. 2016; Sun et al. 2015) and PI3K/AKT pathway (Li et al. 2014) and its possible contribution to more aggressive neoplasia. The involvement of Klotho in apoptotic proteins of the Bcl-2 and Bax family and many others was also determined (Chen et al. 2010). It is worth noting that the research endeavors used to reveal Klotho features have not attempted to research an isolated action of various forms of the protein and to compare them, rather predominantly focused on mKL(Wolf et al. 2008). Nevertheless, some previously conducted research showed data regarding features of recombinant Klotho lacking KL1 domain (516 amino acids) added to cancer cell lines (Chen et al. 2012; Tang et al. 2016; Lee et al. 2010).

Even though much data is available associating Klotho with carcinogenesis, there are no studies aimed to investigate the differently encoded Klotho proteins in neoplasia, yet it is known the mKL forms a complex with the FGF receptor and acts as a co-receptor for FGF23, thus affecting phosphate metabolism, whereas the functional characteristics of sKL remain unclear (Hu et al. 2013). Of the general assumptions that the sKL is released into the microenvironment and circulates as a humoral factor, causing auto- and paracrine effects on cells (Matsumura et al. 1998; Chen et al. 2014). Thus, the sKL may have broad targeting, affecting more than one cell.

Recent studies have demonstrated that Klotho is involved in carcinogenesis in glioblastoma cells. However, it revealed ambiguous causative patchwork between Klotho expression, methylation of the promoter region, and patients’ survival rate add value as itself and need to be deciphered (Peshes-Yeloz et al. 2019). It contributes to additional research aimed at clarifying different aspects of the Klotho on glioblastoma cells. There is evidence that Klotho can inhibit the viability of A-172 cells. In particular, some studies have shown that A-172 cells are sensitive to changes in the PI3K / AKT signaling pathway (Kalhori et al. 2019), Wnt expression (Kirikoshi et al. 2001), and Bax and Bcl-2 proteins (Chen et al. 2007; Dabili et al. 2019) affected by Klotho. Thus, A-172 might represent a significant interest in the consecutive study of the various Klotho forms.

The present study aims to examine the effect of an isolated overexpression of the sKL on the growth features and apoptosis of the A-172 human glioblastoma cell line and provides a glimpse of a mechanism to investigate further profoundly.

Materials and Methods

Cell Culture and Transfection

Studies were carried out on the cell cultures of human glioblastoma line A-172 (ATCC CRL 1620) (Giard et al. 1973) obtained from the Institute of Cytology RAS (St. Petersburg, Russia). Cells were cultured under 95% humidity, 5% CO2, and 37 °C on Dulbecco’s Modified Eagle’s Medium/Ham F-12 (Sigma Aldrich, USA) mixed medium containing 10% bovine fetal serum.

Cultures were pre-sown in 25 cm2 vials (Orange Scientific, Belgium) or flat-bottomed 96-well plates (Orange Scientific, Belgium). Each group consisted of 7–10 replicates for both growth curves and apoptosis or 30 replicates for the MTT. Transfection was carried out with Lipofectamine 2000 polycationic lipid complex (Invitrogen, USA). Cell cultures reaching 60% of the confluence index were employed for further processing. The transfection mixture included 0.2 μg pDNA/sKL or pcDNA3.1/V5/His-TOPO for experiment and control series correspondingly. Lipofectamine 2000 in a volume of 0.5 μl per well of a 96-well plate and 8 μg DNA and 20 μl per 25 cm2 culture flask was used. The cells were incubated in the transfection mixture for 18 h, and then, the medium in cultures was changed into a standard growth medium (DMEM/Ham F-12 with 10% FBS).

pDNA Extraction, Qualitative and Quantitative Assays, Control Vector Development

The Klotho plasmid (secreted form) was kindly provided by the Hal Dietz laboratory of Johns Hopkins University (Addgene plasmid # 17,713) (Arking et al. 2002). It should be mentioned that the applied genetic construct included only three Klotho exons and was incapable of producing the full-length protein. The circular pcDNA3.1/V5/His-TOPO plasmid with the β-galactosidase blank sequence was made by inserting the corresponding gene into a linearized plasmid and used to transfect the control group (ThermoFisher, USA). E.coli were transformed with the blank genetic vector using the same heat shock approach as was used for the experimental group and subcloned under the standard conditions in LB media. Plasmid DNA (pDNA) was then isolated from the E. coli using the Miniprep Kit (Zymo Research, USA) according to the manufacturer’s protocol. The blank plasmid was then amplified and sequenced using the approach described below. The alignment of the blank incorporated lacZ gene to the reference sequence was performed using BLAST (Table S1, Suppl.).

Quantitative calculation of pDNA was performed on a spectrophotometer (BIO-RAD, USA) at a wavelength of 260 nm. Qualitative analysis was carried out according to the ratio of A260/A280. Samples were considered appropriate at A260/A280 = 1.8 ± 0.1. A series of dilutions were prepared for subsequent PCR in qualitative and quantitative formats with a known copy of 1 μl of solution: 1 × 104, 1 × 105, and 1 × 106. The plasmid DNA copy number of each dilution was also established using droplet digital PCR (BIO-RAD, USA) to quantify the exact number precisely. The least one was applied to qualitative PCR with custom-made primers spanning the junction of the first two exons of the Klotho (Table 1). The expected size of the product synthesized from these primers is 103 bp. Reactions were carried out in twofold buffer (Fermentas, USA), 25 μM of each primer, and 2.5 U Dream Taq polymerase (Fermentas, USA). The following thermal cycling parameters were used: 95 °C for 2 min and then 35 cycles: 95 °C for 15 s, 62 °C for 15 s, and final elongation at 72 °C for 7 min. The specificity of the reaction was evaluated by PAG electrophoresis. The main criterion was the absence of fragments of length other than 103 bp. Fragments of the desired size were cut out, and DNA was extracted and purified from the gel. Sanger sequencing was performed for three samples of both control and experimental vectors in forward and reverse directions.

Table 1 Sequences of oligonucleotides
Full size table

Total RNA Isolation and Reverse Transcription, Gene Expression Assay

The transfected cultures were washed three times with PBS solution and removed from the culture flasks according to a standard procedure using a 0.25% Trypsin solution. The resulting suspensions were centrifuged at 200 g for 5 min, and the pellet was re-suspended in 200 μl of PBS solution and then mixed with TRI Reagent solution (MRC, USA) in a ratio of 1: 4, and then, the total RNA was isolated according to the manufacturer’s instructions. Dry-precipitated RNA was dissolved in 10 μl of Rnase free water (Qiagen, Netherlands); the concentration was measured. A reverse transcription reaction was performed in 1 μg of total RNA at 37 °C for 60 min using random hexamers in 10 μl of a master mix (Promega, USA), consisting of 5X reverse transcription buffer, 10 mM dNTP, 100 μM random primer, 5 U MMLV reverse transcriptase, 1 U RNase inhibitor. For the subsequent qPCR reactions, 2.5 μl of cDNA was used per reaction.

Gene expression was studied by the RT-qPCR (reverse transcriptase quantitative PCR). The calibration curves were plotted using serial dilutions, prepared during the isolation step of pDNA/sKL with precisely characterized copy numbers. The sKL was detected by primers used for qualitative PCR, hydrolysis probe labeled with FAM (green channel) was added (Table 1). Reactions were performed in a uniplex format using a master mix (Promega) consisting of 5X buffer, 10 mM dNTP, 50 mM Mg2+, 6.5 µM of each primer, 5 µM probe, ddH20, and 2.5 U GoTaq polymerase and pDNA of the desired copy number. Amplification and fluorescence detection were performed on a LightCycler 96 instrument (Roche, Switzerland) with thermal cycling parameters: 95 °C for 2 min and then 45 cycles: 95 °C for 15 s, 62 °C for 15 s. The analysis of the curves raw data, the recording of Cq, and the plotting of calibration curves were performed in the LightCycler 96 SW 1.1 software.

Abelson gene (ABL1) was used as a reference gene due to its relatively stable expression in different cells and lack of dependency on various treatment factors. Abl1 detection was performed using the primers and FAM labeled hydrolysis probe (Table 1). Calibration curves for the ABL1 gene were plotted using commercial ABL1 Control gene plasmids with a characterized copy number of 1 × 104, 1 × 105, and 1 × 106(Qiagen, Netherlands). The efficiency of amplification of the sKL and Abl1 fragments was estimated by the formula: 10 − 1 / slope − 1. Gene expression assay in both experimental and control cultures was performed according to the described procedure. The absolute number of the sKL and Abl1 genes was found in copies per reaction volume. The results of sKL expression were expressed as relative to ABL1.

Protein Expression Assay

The concentration of Klotho protein in cell cultures was also measured to assess sKL overexpression directly. As the cells secreted the sKL into the microenvironment, we carried out an expression assay in both cell lysates and culture medium. The concentration of the sKL protein was determined using a kit for ELISA on the human Klotho protein (Cloud-Clone Corp., USA). The analysis was carried out by the manufacturer’s protocol. The optical density obtained at a wavelength of 450 nm was used to determine the protein concentration from the calibration data of standard dilutions represented by the regression equation.

MTT Test

MTT test was performed to evaluate the impact of sKL on the viability and proliferation; the test cells were cultured in 96-well plates at a seed concentration of 8 × 103 cells per well, followed by transfection. The viability evaluation was carried out after washing off the transfection mixture in 24, 48, and 72 h. The addition of MTT [3-(4,5-dimethylthiazol-2,5-diphenyltetrazolium bromide, 20 μl, 10 mg/ml; Sigma Aldrich, USA] was followed by incubation of cell cultures for 4 h. Then, the supernatant was removed from each well, and 200 µl of Me2SO/2-propanol solution (1:1 ratio) was added into each well for 10 min. The calorimetric assay was carried out on a Multiscan Go tablet spectrophotometer (Thermo Fisher Scientific, Finland). The optical density was measured at 570 nm.

Growth Curves

Cultures were transferred into culture flasks with an area of 25 cm2 (Orange Scientific, Belgium) with a seed concentration equal to 4 × 103 cells/cm2. Cells were subjected to transfection (experimental group) or the transfection mixture with pcDNA3.1/V5/His-TOPO blank plasmid (control group). The cultures were consecutively removed from the experiment after transfection in 24, 48, and 72 h with the number of cells counted by the Scepter automatic cell counter (Millipore, USA).

Apoptosis

A-172 cells were cultured in 25 cm2 culture flasks followed by transfection. In 48 h, after transfection, the cells were stained with fluorescent dye using the FAM-FLICA (Fluorescent-Labeled Inhibitor of Caspases) Poly Caspase Assay Kit (Novus, USA) according to the manufacturer’s recommendations. Moreover, the apoptosis machinery was also assessed using fluorescent staining with the Annexin V Fitc Staining kit (Sigma Aldrich, USA).

The use of additional staining with propidium iodide solution enabled the differentiation of necrotic cells from apoptotic ones. The relative number of caspase-active cells and annexin V positive cells was determined by counting with a ZOE fluorescent cell imager (BIO-RAD, USA) via various fluorescent filters on the Fuchs-Rosenthal counting chamber.

Statistics

The data is presented as an arithmetic means of ± SD. Statistical analysis was carried out in the R version 4.0.2 (R Foundation for Statistical Computing, Austria, Vienna). Shapiro–Wilk’s test determined the distribution of numeric values. Bartlett’s test assessed the difference of the pairwise variance. The Student’s t test was used to determine the statistically significant differences between the two groups. The p < 0.05 was taken as statistically significant. The Y-intercept and slope coefficients for calculating gene copy numbers were obtained by the linear regression method. The copy numbers were calculated with the coefficients of determination (R2) ≥ 0.98.

Results

The designed oligos set’s specificity and optimal annealing temperature were evaluated after a one-round PCR, followed by visualizing heterogeneous reaction products in PAGE. It was shown that this custom-made set of both primers and probes achieved a lack of non-specific products with both plasmid and genomic DNA. Increased accumulation of the PCR product was observed at an annealing temperature of 62 °C in comparison with 58 °C and 64 °C. The resulting size of PCR products was similar to the expected of 103 bp (Fig. S1A, Suppl.)

The concordance of the obtained and expected products was established by direct sequencing. Two short sequences of 68 (sense) and 64 (anti-sense) bp were obtained from forward and reverse primers, respectively. The complete identity of the Klotho mRNA was established by comparing it with the reference sequences in both single and consensus variants. The consensus sequence was expectedly located at the junction of the first and second exons of the gene. It enabled detecting pDNA and mRNA of the sKL, plot calibration curves, and gene expression assay (Fig. S1B, Table S1, Suppl.).

The qPCR on plasmid DNA samples with a characterized copy number allowed us to plot standard calibration curves and create regression curves (R2 = 0.99) to calculate the absolute number of copies: Lgnumber of copies = (Ct-Y-intercept)/slope, where Ct is the value of the threshold reaction cycle (Fig. S1C, Suppl.).

The sKL and ABL1 fragments amplification efficiencies were commensurable and equal to 91 and 99%, respectively. Therefore, according to the MIQE criteria (Bustin et al. 2009), information regarding the qPCR technique and quantification data suggests that technology for the expression assay was proven as appropriate, so according to RT-qPCR data shown in Fig. 1A, the relative expression in the control group was on average 13.55 times less than in the experimental group (p < 0.001).

Fig. 1
figure 1

Induction of overexpression of the Klotho gene. A Relative Klotho gene expression in A-172 human glioblastoma culture cells, KL/ABL1 (n = 3, mean ± SD). B Concentrations of Klotho protein in cell lysates (Klotho, pg/total protein, μg; n = 3, mean ± SD). C Concentrations of Klotho protein in medium (Klotho, pg; n = 3, mean ± SD)

Full size image

Analysis of Klotho protein overexpression revealed concentration in cell lysates 74.22 ± 6.73 in the control group and 93.03 ± 6.65 in the experimental group (pg of Klotho protein/μg of total protein). The revealed differences also had a significance, p = 0.026 (Fig. 1B). Likewise, higher significant differences were detected in the culture medium, likely due to the release of the protein outwards (Fig. 1C). Thus, the concentration of the sKL in control was 7.82 ± 2.11 pg/mL compared to experimental: 115.1 ± 13.24 pg/mL. Therefore, the cell transfection per se facilitated an increased amount of sKL over 14 times (p < 0.001). The latter has almost the same magnitude as RT-qPCR overexpression (14.72 vs. 13.55) data pointing out that the total mRNA was translated into protein.

To examine if sKL overexpression affects glioblastoma cells’ growth characteristics, we carried out three tests: MTT, growth curves, and apoptosis.

MTT

The results of the MTT test (Fig. 2A) did not reveal significant differences in the experimental and control groups after transfection in 24 h. The average OD570 in the experimental group was lower by more than 4% than in the control group (0.8 ± 0.09 vs. 0.83 ± 0.09, p = 0.15). However, the experimental group was already 4.35% lower than the control at the following point, indicating that significance was achieved (1.2 ± 0.1 vs. 1.26 ± 0.1, p = 0.04). Finally, the 72 h revealed an extra difference, so OD570 in the experimental group turned out to be 8% lower than in the control group (1.95 ± 0.27 vs. 2.12 ± 0.21, p < 0.01).

Fig. 2
figure 2

Effects of sKL overexpression on proliferation and viability of human glioblastoma cells. A MTT test (n = 30, mean ± SD). B Growth curves, cell count × 103/sm2 (n = 7, mean ± SD)

Full size image

Growth Curves

In general, growth curve plots (Fig. 2B) were well correlated with the MTT data. At 24 h, no significant differences were registered. The number of cells in the experimental group was lower only by 5.24% than in control (19.63 ± 1.39 vs. 20.71 ± 1.97, p = 0.26). However, 48 h after transfection in the experimental group, the number of cells decreased by 10.87% compared to control (27.89 ± 4.6 vs. 31.29 ± 2.21, p = 0.075). Finally, at 72 h, the number of cells in the experimental group was already 19.29% lower than in control (37.77 ± 6.39 vs. 46.8 ± 3.92, p < 0.01).

Apoptosis

The number of cells with activated caspases under the sKL overexpression significantly increased (Fig. 3A). In the experimental group, the number of caspase-active cells was equal to 6.71%. In the control group, this number was about 4.1% (6.71 ± 1.15 vs. 4.1 ± 0.91, p < 0.001). Thus, although the absolute count was moderate, higher apoptosis under sKL overexpression could be observed, with a 1.63-fold increase.

Fig. 3
figure 3

Effects of sKL overexpression on apoptosis. A The ratio of cells with increased caspase activity in the experimental and control groups (n = 10, mean ± SD). B The relative amount of Annexin-V-positive cells in the experimental and control groups (n = 10 control, n = 12 experimental, mean ± SD)

Full size image

In addition, a significant number of apoptotic cells were recorded using Annexin V staining. As shown in the diagram (Fig. 3B), the control group’s average percentage of apoptotic cells was 3.64 ± 0.42% compared to 4.82 ± 0.46% in the experimental group (1.32-fold higher, p < 0.001). Thus, as in the RT-qPCR vs. protein expression comparison mentioned above, the given numbers (1.32 vs. 1.63) correspond and show similar magnitude.

Discussion

In the present study, the liposomal transfection of a plasmid containing a secreted form of Klotho protein induced a 13.55 times higher expression compared to the control as detected by RT-qPCR. However, it is of particular importance to confirm this value by direct protein measurement. Therefore, we performed an ELISA as a verification tool. Eventually, revealed sKL mRNA overexpression is confirmed by direct protein measurements demonstrating an increase of sKL protein in the experimental group by more than 25% in cell lysates and by 14.72 times in culture medium. Meanwhile, the main idea of the research was to discover whether the isolated overexpression of the separate sKL was capable of inhibiting the growth of A-172 and of providing some glimpse of a possible mechanism to investigate further. To address this problem, we inspected the growth curves, MTT plots, and apoptosis (Fig. 2). On the one hand, we showed the consistency between tests conducted. By contrast, some absolute counts (caspase-cells) demonstrated slight differences, likely indicating that more complicated machinery is recruited in cells’ growth inhibition than apoptosis as sole. Nevertheless, we suggest providing a quick overview of the possible apoptotic mechanisms involved in antitumor action (based on the previous studies) and its consistency with our findings.

Four independently carried out tests demonstrated a statistically significant effect with a high degree of reliability (in all cases, the p value was < 0.01). At the same moment, the most pronounced effect was registered at caspases activity assessment and annexin V (in both cases, p < 0.001). However, in general, the increase of the cell numbers, supposedly in apoptosis, turned out to be moderate (4% vs. 6.7%). It is noteworthy that we obtained comparable data upon analyzing annexin V. It might be due to some intracellular mechanism capable of partially blocking the sKL-induced apoptosis in glioblastoma cells. However, the given hypothesis could be considered assumptive and require a more detailed study. Likewise, it is known that an increase in the activity of caspases does not always indicate the initiation of programmed cell death. For example, activated caspases may not contribute to cell death if the action of these proteases is blocked by apoptosis inhibitor proteins (PAIs) (Shi 2004; Vaux and Silke 2005). The published research data indicate the involvement of some proteins in the initiation of the mitochondrial pathway of apoptosis in klotho-mediated mechanisms for the suppression of cell viability. For instance, it was shown the participation of proteins of the Bcl-2 and Bax family in klotho-mediated antitumor activity using lung cancer as an example (Chen et al. 2010). The activity of well-studied initiator and effector caspases and caspases − 1, − 4, and − 5 possibly associated with inflammatory processes or unknown functions of these compounds are discussed. Therefore, the induction of apoptosis in cells is only one of the possible options for suppressing the viability of a human glioblastoma cell culture.

The next issue is differences in significance upon analyzing similar plot points shown in Fig. 2. As it can be seen, the MTT test showed significant differences already at the second and third-time marks (48, 72 h) of the experiment, while the growth curves showed significance only on the third day (72 h). We assume that this may be explained by various ways to statistical data analysis in these experiments. The growth curves were plotted based on data of 7–10 cultures in each group, while the sample during the MTT test had 60 independent observations in both groups. Although the same parametric approaches were performed upon statistical calculations, it should be noted that the applied statistical power of any test relies, among the others, on the sample size. Therefore, we argue that the MTT test provides more reliable values ​of the effect of sKL on the proliferation of A-172.

Meanwhile, overexpression of Klotho can induce the development of other regulatory disorders in the cells. According to several studies, the effect of Klotho on the signal interactions of insulin and insulin-like growth factor 1 was determined to possibly influence cell antioxidant protection (Li et al. 2014; Shu et al. 2013; Lu et al. 2008). It is also assumed that there is a correlation between various miRNAs that may be involved in the development of carcinogenesis. In particular, some studies have shown the correlation between the expression of the Klotho gene and miR-339 and miR-556 (Mehi et al. 2014), miR-199a-5p (He et al. 2014), and miR-10b (Pan et al. 2015).

It should be mentioned that the current research has demonstrated for the first time the effect of the gene overexpression of the secreted (3-exon) Klotho protein on the growth of human glioblastoma cells. In the previously reported studies, the effect of the full, i.e., a 5-exon form of the protein, was investigated (Wang et al. 2013; Chen et al. 2013; Shu et al. 2013). Meanwhile, to assess Klotho gene expression by the RT-qPCR method, we recruited ddPCR for precise calibration of sKL copies and designed our own primers and probe, which considerably increases the specificity of the assay. The addition of mentioned above, as well as detection of expression of a reference gene, makes an assay more accurate and sensitive as well as enabled to observe almost complete concordance between qPCR and ELISA (13.55 and 14.72 times). It is noteworthy that the existing approaches applied to detect the mKL use primers targeting gene regions of 4 and 5 exons and cannot be used for the secreted form since these regions are absent in the given (sKL) genetic construct. For example, the quantification approach described above, as well as full-length plasmid, was used by Chen et al. (Chen et al. 2010).

In general, the current research results correlate with the results described in the articles on the subject. The methods of growth curves plotting and the MTT test demonstrated inhibition of cell proliferation combined with significant activation of caspases. More detailed research is required to form the hypothesis of Klotho involvement in the mechanisms of carcinogenesis in glioblastoma cells.