Apamin inhibits renal interstitial inammation and brosis via suppressing TGF-β1 and STAT3 signaling pathway in vivo and in vitro

Background Renal brosis is a progressive and chronic process that inuences kidneys with chronic kidney disease (CKD), irrespective of cause, leading to irreversible failure of renal function and end-stage kidney disease. Among the signaling related to renal brosis, transforming growth factor-β1 (TGF-β1) signaling is a major pathway that induces the activation of myobroblasts and the production of extracellular matrix (ECM) molecules. Apamin, a component of bee venom (BV), has been studied in relation to various diseases. However, the effect of apamin on renal interstitial brosis has not been investigated. The aim of this study was to estimate the benecial effect of apamin in unilateral ureteral obstruction (UUO)-induced renal brosis and TGF-β1-induced renal broblast activation. Results This study revealed that obstructive kidney injury induced an inammatory response, tubular atrophy, and ECM accumulation. However, apamin treatment suppressed the increased expression of brotic-related genes, including α-SMA, vimentin, and bronectin. Administration of apamin also attenuated the renal tubular cells injury and tubular atrophy. In addition, apamin attenuated broblast activation, ECM synthesis, and inammatory cytokines such as TNF-α, IL-1β and IL-6 by suppressing the TGF-β1-canonical and non-canonical signaling pathways. Conclusions This study shown that apamin inhibites UUO-induced renal brosis in vivo and TGF-β1-induced renal broblasts activation in vitro. Apamin inhibited the inammatory response, tubular atrophy, ECM accumulation, broblast activation, and renal interstitial brosis through suppression of TGF-β1/Smad2/3 and STAT3 signaling pathways. These results suggest that apamin might be a potential therapeutic agent for renal brosis.


Introduction
Renal brosis is the ultimate common manifestation of progressive chronic kidney disease (CKD) leading to the irreversible destruction of kidney parenchyma and end-stage of renal failure [1,2]. Renal brosis is characterized by the demolition of renal tubules, tubular atrophy, in ltration of immune cells, accumulation of myo broblasts, and overproduction of extracellular matrix (ECM) resulting in renal tubular cell apoptosis and necrosis [3][4][5].
Among these characteristics, in ammatory response plays an important role in the progression of numerous acute and chronic renal injuries [6]. This reaction is induced as a protective response to a wide range of renal injuries, but ongoing in ammation promotes progressive renal brosis regardless of the underlying etiology [7]. To be more speci c, in ammatory cell in ltration in the circulation is localized to the damaged tissue, inducing a renal in ammatory response. This in turn leads to the generation and secretion of in ammatory mediators and pro-brotic cytokines and growth factors. In ammatory mediators mediate the cascade ampli cation and sustenance of in ammatory responses and cause apoptosis of the renal tubular cells, activation of myo broblasts, renal tubular atrophy and renal interstitial brosis. Therefore, inhibiting the in ammatory response may facilitate attenuating renal tubular epithelial cell apoptosis and interstitial brosis [8].
Fibroblasts are the main sources of interstitial ECM components. Fibroblasts become activated during brosis, sometimes irreversibly. Long-term activated broblasts are called myo broblasts and are characterized by the de novo expression of α-smooth muscle actin (α-SMA) [13]. Because of this reason, a massive increase in interstitial myo broblast activation is believed to play a central role in the pathogenesis of tubulointerstitial brosis [14,15].
The TGF-β1 signal is initiated when activated TGF-β1 binds to a TGF-β type receptor (TβR ), a constitutively active kinase, leading to phosphorylation of the TGF-β type receptor (TβR ). The activated TβR kinase then phosphorylates the downstream receptor-associated Smads, including Smad2 and Smad3. Then, the phosphorylated Smad2/3 forms a complex with Smad4, and translocate to the nucleus to control the transcription of target genes. [16,17]. On the other hand, Smad7, an inhibitory Smad, exerts its negative effect on the TGF-β signaling through consequently conpeting with the Smad2/3 [11].
In addition, TGF-β-induced renal brosis is mediated by Smad-independent signaling pathways. Signal transducer and activator of transcription 3 (STAT3) is a representative mediator in the Smad-independent signal. STAT3 is a signi cant member of the STAT family (STAT14, STAT5a/5b, and STAT6) and mediates cell proliferation and survival. Various growth factors and cytokines can phosphorylate STAT3 tyrosine. The activated STAT3 form a dimer and translocate into the cell nucleus to regulate the transcription of target genes [18][19][20]. It has been reported that STAT3 phosphorylation mediates the activation of myo broblast and the progression of renal interstitial brosis in unilateral ureteral obstruction (UUO) models [21,22]. Therefore, it is thought that inhibition of TGF-β1 and STAT3 signaling can alleviate renal brosis via suppression of myo broblast activation and ECM accumulation.
Bee venom (BV) therapy has been used to alleviate suffering and to treat a variety of in ammatory diseases since ancient times, including arthritis, bursitis, back pain, rheumatism, skin disease, and other chronic conditions [23,24]. Apamin, a component of BV, is well known speci cally a selective blocker of small conductance Ca 2+ -activated K + channel (SK channel) as it binds to the pore of the channel [24,25]. For several years, apamin has been studied as a speci c SK channel blocker in the central nervous system [26]. Recently, other therapeutic effects of apamin have been announced. The anti-in ammatory effect of apamin accompanied by a reduction of seromucoid and haptoglobin levels has been reported [27]. In addition, apamin attenuates in ammatory responses in THP-1-derived macrophages [28], suppresses in ammatory cytokines in TNF-α-and IFN-γ-induced keratinocytes [29], and inhibits hepatic brosis in CCl 4 -injected mice [30]. However, the effects and mechanisms of apamin in obstructive kidney injury and TGF-β1-induced broblast activation has not been reported. Therefore, this study investigated the anti-in ammatory and anti-brotic effect of apamin using a UUO-induced mice model and a TGF-β1induced NRK-49F cell model.

Results
Apamin attenuated renal interstitial brosis and improved kidney function in the UUO model Among experimental animal models, the UUO research model is mainly used to investigate obstructed renal interstitial in ammation and brosis [31]. Untreated urinary obstruction can lead to tubular atrophy, interstitial in ammation and brosis, and, nally, irreversible kidney injury [32,33]. To identify the bene cial effect of apamin on obstructive kidney injury, this study used a UUO-induced obstructive animal model. As shown in Fig. 1a, histological analyses showed that obstructive kidney exhibited interstitial immune cells in ltration, partial tubular expiation, and tubular atrophy. However, apamin treatment attenuated these morphology changes. Additionally, in representative trichrome images, apamin was able to suppress collagen accumulation caused by UUO-induced renal injury (Fig. 1a, d). In supplementary Fig. 2, we performed the western blotting analysis to investigate the bene cial effect of apamin on renal tubules injury. The expression of E-cadherin, as the epithelial cell marker, was decreased by obstructive kidney compared with normal kidney. In addition, the renal injury biomarker neutrophil gelatinase-associated lipocalin (NGAL) was increased in UUO mice. However, the expression of Ecadherin was increased and the expression of NGAL was decreased by apamin administration.
Furthermore, to clarify the protective effect of apamin on renal function, blood urea nitrogen (BUN) and serum creatinine were measured using mice serum plasma ( Fig. 1b and c). In UUO mice, the BUN and serum creatinine levels were increased, which means that UUO injury effectuated renal dysfunction. In contrast, apamin treatment indicated BUN and creatinine levels similar to normal kidney conditions.
Apamin inhibited renal interstitial in ammation response in UUO-injured mice After obstructive injury, macrophage and T lymphocytes were in ltrated and the in ammatory response was intensi ed in kidney tissue. In addition, continuous in ammation plays an important role in the initiation and development of renal interstitial brosis [34,35]. To examine the anti-in ammatory effect of apamin, the expression of TNF-α, IL-1β and IL-6 were analyzed by immune blot. The protein level of TNFα, IL-1β and IL-6 were markedly increased in UUO mice ( Fig. 2a-d). However, apamin treatment attenuated the UUO-mediated renal in ammatory cytokines.
Monocyte chemoattractant protein-1 (MCP-1) is widely known as a chemokine that has the ability to regulate in ltration and migration such as monocytes and macrophages. Monocytes collected by MCP-1 differentiate into macrophages and secrete in ammatory cytokines (TNF-a, IL-6, IFNs, etc.) to intensify the in ammatory response [36]. Based on this fact, to nd out the e cacy of apamin on renal in ammation, F4/80 and MCP-1 were stained using immunohistochemistry staining. The expression of MCP-1 was increased in obstructive kidney ( Fig. 2e, g). Furthermore, the F4/80 + macrophages were increased in UUO-injured kidney (Fig. 2f, h). Conversely, administration of apamin diminished MCP-1 expression, macrophage in ltration, and in ammatory cytokines.

Fibrotic gene expression and myo broblasts activatin were reduced by apamin injection
A key step in renal brogenesis is the accumulation of myo broblasts and ECM molecules. Excessive ECM deposition is the end-result of increased matrix elements secretion and accumulation, and decreased degradation. There are many components in the ECM matrix, including bronectin, collagen I and others [13]. We next examined the expression of ECM in kidney tissues using western blot analysis. The results showed that the expression of α-SMA, vimentin and bronectin signi cantly increased in the brotic kidney compared to normal mice. However, these changes were reversed by apamin administration (Fig. 3a-d).
Myo broblasts show high proliferation and ECM secretion rates and play a key role in interstitial brosis in the UUO model [13]. To identify whether apamin suppresses myo broblast accumulation in UUO kidneys, this study investigated the protein levels of broblast-speci c marker-1 (FSP-1) using immunohistochemistry. The results showed that UUO-injured kidney increased the expression of FSP-1, as a myo broblasts marker, whereas apamin administration suppressed this expression (Fig. 3e, f).
Apamin inhibited TGF-β1/Smad signaling pathway and STAT3 signaling pathway TGF-β1 signaling is a key mechanism leading to broblast activation and renal interstitial brosis. This cytokine also induces Smad2/3 and STAT3 phosphorylation that regulate the transcription of target genes [12,37]. To con rm the molecular mechanism of apamin in obstructive kidney, we investigated the protein expression of TGF-β1/Smad signal mediators and STAT3 transcription factor through western blotting analysis. As shown in Fig. 4, the UUO group exhibited signi cantly increased expression of TGF-β1, p-STAT3 and p-Smad2/3, whereas these expressions were inhibited in apamin treatment group.
Smad7 is an inhibitory regulator in the TGF-β1/Smad signaling pathway, which blocks the signal transduction of TGF-β1 via its negative feedback loop. Moreover, Smad7 e ciently prevents Smad2 and Smad3 interaction, and Smad-related protein expression [12]. To further investigate the effect of apamin on Smad7 signal, we analyzed the Smad7 expression using western blotting analysis. In the immunoblotting study, the protein expression of Smad7 was signi cantly decreased by UUO-injury. However, apamin treatment recovered the Smad7 expression (Fig. 4a, e).
Anti-brotic effect of apamin in TGF-β1-treated renal broblast cells On the basis of the observation of apamin administration in UUO mice, we determined the anti-brotic effect of apamin on TGF-β1-induced kidney brosis in in vitro models. First, cell counting kit (CCK)-8 assay was executed to indicate the cytotoxicity of apamin at different doses. The normal rat kidney interstitial broblast cells (NRK-49F) were treated with 0.1, 0.5, 1, 2, 5, and 10 µg ml − 1 of apamin for 6, 24 and 48 h. In the 6 or 24 h apamin treatment, all concentrations of apamin treatment did not alter cell viability (Fig. 5a, b). In the case of 48 h apamin treatment, NRK-49F cell viability was reduced at 10 µg ml − 1 of apamin. However, 0.1, 0.5, 1, 2, and 5 µg ml − 1 of apamin did not affect NRK-49F cell viability (Fig. 5c). In accordance with this result and with a previous study [38], we selected the apamin concentrations (0.5, 1, and 2 µg ml − 1 ) and apamin treatment time (48 h) in the following experiments.
To determine the anti-brotic effect of apamin on TGF-β1-induced renal broblast activation, we investigated the protein expression of ECM products using immunoblotting analysis. After TGF-β1 treatment, the expressions of collagen and bronectin were signi cantly increased. However, collagen expression was decreased in a dose-dependent manner in the apamin treatment groups. TGF-β1-induced expression of bronectin was most decreased in the 0.5 µg/ml − 1 concentration of apamin (Fig. 5d-f). Furthermore, to con rm the molecular mechanism of apamin in TGF-β1-treated cells, we examined the expression change of TGF-β1 signal mediators. TGF-β1 stimulation induced phosphorylation of STAT3 and Smad2/3, and expression of TGF-β1. Conversely, apamin signi cantly diminished TGF-β1-induced STAT3, Smad2/3, and TGF-β1 expressions in a dose-dependent manner. In the case of Smad7, this expression was decreased in TGF-β1 treatment renal cells. However, apamin treatment recovered the expression of Smad7.

Discussion
Development of kidney brosis is the hallmark of most progressive CKDs, irrespective of the cause, and is thought to be a common pathway leading to end-stage renal diseases [3,4]. The development of endstage renal disease requires renal transplantation due to irreversible loss of tissue and impaired kidney function. For this reason, CKD has increasingly become a major global public health concern and portends high rates of morbidity and mortality [39]. Therefore, it is important to prevent kidney interstitial brosis to prevent or slow the devastating CKD sequelae and progression to end-stage renal disease [40,41].
While various drugs, mainly drugs targeting angiotensin II, slow the progression of CKD, the therapeutic armamentarium is still imperfect [41]. Hence, there is a critical need for new therapeutics to diminish kidney brosis and renal failure. Recently, drug discoveries for ghting kidney brosis have mainly focused on compounds that are speci c for a protein kinase or particular receptor. Given that kidney brosis is associated with increased production of multiple growth factors/cytokines and the following activation of their receptors and signaling pathways, it is supposed that inhibitors with wide speci city might provide improved therapeutic bene ts in renal brotic diseases [42].
Apamin is a speci c component of BV that is known as a greatly selective blocker of calcium-dependent potassium channel [43]. This channel connects intracellular Ca 2+ transients to changes of the membrane potential by inducing K + e ux following increases of intracellular Ca 2+ during an action potential [44]. Kim et al. [29] have shown that apamin has anti-in ammatory effects against TNF-α-and IFN-γstimulated keratinocytes. In addition, previous studies have shown that apamin suppressed Smaddependent and Smad-independent signaling pathways in liver brosis, and suppressed STAT signaling pathways in atopic dermatitis [29,30]. Following these studies, we hypothesized that apamin was going to have an anti-in ammatory and anti-brotic effect in renal interstitial brosis via suppressing the TGF-β1 and STAT3 signaling pathways.
To prove the hypothesis, this study investigated the therapeutic effect of apamin using UUO model and TGF-β1-treated broblast cells. Similar to other studies [32,33], obstructive kidneys show interstitial in ammation, tubular injury and death, tubular atrophy, renal failure, and interstitial brosis by UUO injury. Increasing evidence shows that tubular epithelial cells play various roles in renal repair or progression to CKD. Continuous injury of tubular epithelial cells promotes production and release of bioactive mediators that induce renal interstitial in ammation and renal brosis [45]. Thus, injury of tubular epithelial cells is an important indicator in renal diseases. In this study, we evaluated the renal tubular cell injury via investigative the expression of epithelial cell marker and kidney injury marker. In UUO kidney, E-cadherin was decreased and NGAL was increased compared with normal kidney, indicating the induction of tubular cells injury. On the other hands, apamin treatment remarkably attenuated the renal epithelial cells damage. In addition, administration of apamin signi cantly improves kidney function and suppresses tubulointerstitial brosis, as evidenced by a diminution in plasma levels of creatinine and BUN and histopathological changes such as tubular atrophy, collagen deposition, and interstitial brosis induced by obstructive injury. Taken together, these ndings suggest that, in mice, apamin protects from UUOinduced renal dysfunction, tubular cell injury and structural changes.
Renal interstitial in ammation is implicated as an important event in the initiation and progression of kidney brosis in CKD. The in ammatory response is characterized by in ltration of immune cells, activation of resident renal cells, excessive production of cytokines (including interstitial in ammatory, pro-brotic cytokines and growth factors), and renal tubular atrophy and interstitial brosis [46][47][48].
Several studies demonstrated that inhibiting the in ammatory response results in attenuation of renal tubular epithelial cell apoptosis and renal brosis [8,46,47]. In this study, we showed that the expression of TNF-α, IL-1β and IL-6 increased by obstructive injury, while administration of apamin reduced the in ammatory cytokines. Moreover, our previous research showed that apamin treatment suppressed in ammatory responses through inhibition of the NF-κB signal pathway in THP-1-derived macrophages [49]. Macrophages, a principal type of in ammatory cell, are recruited in all kidney disease. Recruited macrophages are associated with the induction of renal injury, repair, and brosis [40]. Furthermore, several studies have reported that macrophages induce the synthesize ECM molecules including bronectin and collagen [50,51]. In the current study, administration of apamin signi cantly suppressed the expression of F4/80, the macrophage marker, and MCP-1, the macrophage recruitment chemokine.
Altogether, these results suggest that apamin has anti-in ammatory effect through inhibition of macrophage in ltration and cytokine production.
Based on these results, we thought that apamin would attenuate myo broblast activation and ECM accumulation by inhibiting in ammatory responses. Some study remarked that interstitial deposition of macrophages and myo broblasts is strongly associated with the progression of UUO injury [52]. In addition, proliferation of broblast with myo broblast transformation induce excessive accumulation of the ECM component in kidney brosis [53,54]. Similar to other studies, UUO mice observed the excess deposition of myo broblasts and ECM molecules. However, apamin administration showed that the accumulation of ECM, including α-SMA, vimentin and bronectin, was decreased and proliferation of myo brobalsts were diminished in vivo experiment.
TGF-β1, a key factor of the initiation and progression of renal brosis, induces tubular epithelial cell apoptosis, myo broblasts activation, and excessive production of ECM molecules by binding to the TβR receptor [55,56]. Furthermore, Liu et al. [57] showed that phosphorylation of Smad3 by TGF-β1 injury promotes STAT3 activation in the injured kidney. As previously reported, our results showed that TGF-β1, Smad2/3, and STAT3 were activated by obstructive injury, while apamin treatment reduced these signal mediators. In addition, TGF-β1 can induce phosphorylation of STAT3 which promotes the activation of renal broblasts and progression of renal brosis [37]. Based on these facts, we performed the in vitro experiments using TGF-β1 to investigate the molecular mechanism in more detail. Similar to other studies, TGF-β1 treatment increased the production of ECM molecules and the activation of TGF-β1 and STAT3 signaling pathways in renal broblasts cell. On the other hand, pre-treatment with apamin was shown to reduce the expression of TGF-β1, p-STAT3, and p-Smad2/3, and the production of collagen, and bronectin by TGF-β1 treatment.
Furthermore, we also investigated the expression of Smad7 in in vivo and in vitro models. Smad7, as an inhibitory regulator in TGF-β1 signaling, prevents Smad-related protein expression by suppressing Smad2 and Smad3 interaction [58]. In addition, hyperactivation of TGF-β1 and Smad3 was concerned with progressive degradation of Smad7. More importantly, the disproportion of Smad3 and Smad7 was determined to be one of the important mechanisms in mediating the brotic response [12]. In the current study, the expression of Smad7 was decreased in UUO-injured mice and TGF-β1-treated renal broblast cells, while apamin administration restored Smad7 expression like as normal condition. As mentioned in Meng et al. [12], our results show that the rebalancing of the Smad3/Smad7 ratio by apamin may contribute to the suppression of renal brosis. Taken together, apamin administration was observed to reduce the activation of TGF-β1 and Smad2/3 and to increase the expression of Smad7 in renal brosis in the in vivo and in vitro models. It is thought that suppression of the TGF-β1 and STAT3 signaling pathways by apamin may contribute to the attenuation of myo broblast activation and ECM production. Altogether, this study suggests that apamin inhibited kidney interstitial brosis by blocking various signaling pathways such as TGF-β1-canonical and TGF-β1-noncanonical signaling.

Conclusions
In conclusions, this study has demonstrated the anti-in ammatory and anti-brotic effect of apamin on UUO-induced renal brosis and TGF-β1-activated renal broblast models. This study shown that UUOinduced myo broblasts activation and ECM accumulation were inhibited by apamin treatment. In addition, UUO-induced interstitial in ammatory response in renal tissue was reduced by apamin, and the expression of pro-in ammatory cytokines was also decreased. These results suggest that apamin may have a protective effect against renal brosis. Based on the results of in vivo study, we con rmed the effect of apamin on broblast activation by TGF-β1 in vitro. Apamin administration suppressed the activation of broblasts by TGF-β1 and expression of brotic genes. In addition, this study showed that the expression of various brotic genes was signi cantly reduced through inhibition of Smad2/3 and STAT3 signaling. All the take together, the current study may be the rst proof that apamin can be used for an anti-in ammatory and anti-brotic effect on renal brosis. Although further examination will be required to clarify a more detailed mechanism, these results suggest that apamin might be a potential therapeutic strategy to prevent renal brosis.

Materials And Methods
Experimental animals and drug treatment After seven days of acclimatization, the UUO operation was performed after anesthetizing the mice. The animal's abdominal cavity was exposed by a small incision and the left ureter was isolated and ligated with 5 − 0 silk sutures at two different sites: upper and lower. Apamin treatment at a concentration of 0.5 mg/kg was given via intraperitoneal injection twice a week. Eight days after the UUO operation, the kidneys were collected for various experiments.
Cell culture and drug treatment Normal rat renal interstitial broblast cells (NRK-49F) were cultured in Dulbecco's Modi ed Eagle Medium (DMEM) supplemented with 5% fetal bovine serum (FBS) and 1% antibiotics at 37 ℃ in a 5% humidi ed Creatinine and blood urea nitrogen The blood samples were collected in tubes by cardiac puncture in all groups. All blood samples were coagulated for 1 h at room temperature (RT). Plasma was separated from whole blood using centrifugation (2,000 g, 20 min). The plasma samples were obtained from the supernatants after centrifugation method for blood urea nitrogen (BUN) and creatinine analysis. The plasma BUN was measured using a BUN-E kit (Asan Pharmaceutical, Seoul, Korea) and the serum creatinine was measured using a QuantiChrom™ creatinine assay kit (Bioassay Systems, Hayward, CA, USA). Analysis of the samples was carried out according to the manufacturer's recommended protocols. Scienti c, Waltham, Ma, USA) and transferred to nitrocellulose membranes (GE Healthcare, Chicago, IL, USA). After transfer, the membranes were blocked in 5% bovine serum albumin for 1 h at RT. The membranes were probed with a primary antibody overnight (4 ℃). Next day, the membranes were washed using a TBS-T buffer for 7 min on the shaker; the process was repeated three times. Then, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h 30 min at RT. Repeat the wash step. The signals were detected using enhanced chemiluminescence detection reagents (Thermo Fisher Scienti c). Signal intensity was analyzed using ChemiDoc™XRS + Imaging System (Bio-Rad Laboratories, Hercules, CA, USA) and quanti ed using the Image Lab software (Bio-Rad Laboratories). The protein expression levels were normalized to GAPDH (Cell Signaling, Beverly, MA, USA) and β-actin. The primary antibodies used were as follows: anti-α-SMA, TNF-α, TGF-β1, bronectin, collagen , IL-6 (Abcam, Cambridge, UK), anti-vimentin (BD Biosciences, San Jose, CA, USA), anti-IL-1β, Smad7, NGAL (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-GAPDH, p-Stat3, t-Stat3, p-Smad2/3, t-Smad2/3, E-cadherin (Cell Signaling), and anti-β-actin (Sigma-Aldrich).

Histological and immunohistochemistry
After harvesting, the kidney tissues were immediately xed in 10% formalin at RT and then embedded in para n. Thereafter, the para n-embedded tissues were cut into 4 µm sections. The thin sections were mounted on glass slides and depara nized. Kidney tissue sections were stained with hematoxylin and eosin (H&E) and Masson's trichrome according to standard protocol.
For the immunohistochemical stain, the para n-embedded sections on slides were depara nized with xylene and dehydrated in gradually decreasing concentrations of ethanol. The sections were incubated with a primary antibody (1:100 dilution) for 1 h at 37 ℃. The signal was visualized using an EnVision System (DAKO, Carpinteria, CA, USA) for 30 min at 37 °C; 3,3′-diaminobenzidine tetrahydrochloride was used as the coloring reagent, and hematoxylin was used as the counter-stain. Primary antibodies were as follows: anti-F4/80, MCP-1 (Santa Cruz Biotechnology), anti-FSP-1 (Cell Signaling). All sections were processed by an indirect immunoperoxidase technique using a commercial EnVision System kit (DAKO) and counterstained with hematoxylin. All slides were scanned using Pannoramic® MIDI slide scanner (3DHISTECH, Budapest, Hungary).

Statistical analysis
All data are presented as means ± SE. A Student's t-test was used to assess the signi cance of the independent experiments. Differences with p < 0.05 were considered signi cant.