Inhibitors of the mammalian target of rapamycin (mTORi) can produce de novo proteinuria in kidney transplant patients. On the other hand, mTORi has been shown to suppress disease progression in several animal models of kidney disease. In the present study, we investigated whether glomerular permeability can be acutely altered by the mTORi temsirolimus and whether mTORi can affect acute puromycin aminonucleoside (PAN) or angiotensin II (ANG II)-induced glomerular hyperpermeability. In anesthetized Wistar rats, the left ureter was cannulated for urine collection, while simultaneously blood access was achieved. Temsirolimus was administered as a single intravenous dose 30 min before the start of the experiments in animals infused with PAN or ANG II or in nonexposed animals. Polydispersed FITC-Ficoll-70/400 (molecular radius 10–80 Å) and 51Cr-EDTA infusion was given during the whole experiment. Measurements of Ficoll in plasma and urine were performed sequentially before the temsirolimus injection (baseline) and at 5, 15, 30, 60, and 120 min after the start of the experiments. Urine and plasma samples were analyzed by high-performance size-exclusion chromatography (HPSEC) to assess glomerular sieving coefficients (θ) for Ficoll10-80Å. Temsirolimus per se increased baseline glomerular permeability to Ficoll50-80Å 45 min after its administration, a reactive oxygen species (ROS)-dependent phenomenon. PAN caused a rapid and reversible increase in glomerular permeability, peaking at 5 min, and again at 60–120 min, which could be blocked by the ROS scavenger tempol. mTORi abrogated the second permeability peak induced by PAN. However, it had no effect on the immediate ANG II- or PAN-induced increases in glomerular permeability.
- glomerular endothelium
- glomerular filtration
the glomerular filtration barrier (GFB) is a highly selective sieving barrier to macromolecules, yet being dynamic (4, 22, 38). Thus the permeability of the GFB can rapidly increase due to challenges induced by trauma (2), hyperglycemia (6), oxidative stress (44, 46), or due to systemic infusions of atrial natriuretic peptide (ANP) (7) or angiotensin II (ANG II) (3, 5). We have previously demonstrated that the rapid dynamic increases in glomerular permeability induced by systemic ANG II infusions in rats can be inhibited not only by ANG II receptor blockers (ARB) but also by scavengers of reactive oxygen species (ROS) and inhibitors of the intracellular Ca2+ signaling cascade in podocytes and endothelial cells (3). Furthermore, paricalcitol, which is supposed to interact with the Ca2+ entry into the cells via so-called transient receptor potential canonical 6 (TRPC6) receptor-operated Ca2+ channels, was found to be effective (3).
Inhibitors of the mammalian target of rapamycin, mTORi, such as sirolimus and everolimus, have been tried for replacing calcineurin inhibitors (CNI) as immunosuppressants in organ (kidney) transplantation, to avoid the profibrotic and glomerular filtration rate (GFR)-deteriorating effects of the latter (8, 9, 16). The use of mTORi has, however, been hampered by their antiproliferative actions; they should not be used during the first post-transplant month, and also by frequent reports of de novo proteinuria after conversion from CNI to mTORi (28). It has also been shown that long-term inhibition of mTOR may interact with pathways of autophagic flux in podocytes (11). Whether the proteinuric actions of mTORi are due to direct effects on the GFB or due to effects on the tubular reabsorption of proteins has also been under debate (10, 15).
mTOR comprises a complex of large serine/threonine kinases of the phosphoinositide kinase family that functions as part of two multimeric complexes, mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) (28, 29). mTOR is associated with a nonenzymatic scaffolding protein, regulatory-associated protein of mTOR (Raptor) in mTORC1, while rapamycin-insensitive companion of mTOR (Rictor) is associated with mTORC2. Rapamycin binds to FK506 binding protein of 12 kDa (FKBP12) to inhibit the function of mTORC1 (but not that of mTORC2). Activation of mTORC1 results among other things in phosphorylation of two downstream targets, the ribosomal S6 kinase (S6K) and 4EBP (eukaryotic translation initiation factor 4 E binding protein), which stimulate ribosome biogenesis and translation to increase cell mass. At large, mTOR integrates information on nutrient availability and growth factors to control protein synthesis and cell size.
Podocytes are terminally differentiated cells with a limited ability to proliferate. Activation of mTORC1 is thought to play a key “protective” role in the adaptive hypertrophic podocyte responses to various noxious challenges, such as toxic, mechanical, or metabolic insults, whereby podocytes could increase in size to compensate for losses of nearby damaged podocytes in the glomerulus. Particularly in states of podocyte stress, e.g., hyperfiltration, mTORi treatment may therefore represent a “second hit,” worsening proteinuria. In hyperglycemia there is ample evidence of increased mTOR activity, conceivably by downregulation of the upstream negative mTOR regulators hamartin (TSC1) and tuberin (TSC2), which are maintained moderately active over time by the nutrient sensor 5′ AMP-activated protein kinase (AMPK). Activation of mTORC1 in diabetes following glucose-induced inhibition of AMPK seems to be able to induce podocyte loss, glomerular basement membrane (GBM) thickening, mesangial expansion, and proteinuria (14, 21).
In contrast to mTORC1, mTORC2 is largely rapamycin-insensitive, and phosphorylates cellular targets, such as protein kinase B (Akt), suppressor of kexz2 gas1 synthetic lethality (SKG1) and PKC to control cell survival and cytoskeletal (F-actin) organization (51). Interestingly, mTORC2, but not mTORC1, seems to be activated by certain insults, such as those induced by protamine sulfate (PS). Recent data thus suggest that the rapid changes in the glomerular barrier conformation and permeability by PS require increased Akt activity and that PS induces increases in Ca2+ signaling, which in turn activates Akt through mTORC2 (48). Since temsirolimus is primarily an mTORC1 inhibitor, it is not, however, expected to be protective in early, rapid permeability responses, such as those produced by PS. On the contrary, temsirolimus itself may adversely affect glomerular permeability, which was actually tested in this study.
Despite their reputation to induce proteinuria in human and animal models, the use of mTORi has been successful in ameliorating glomerular hypertrophy and albuminuria in the diabetic kidney (21, 30, 52) and to reduce cyst formation in animal models of polycystic kidney disease (41, 54). Given the context-dependent actions of mTORi (17, 24, 40, 47), implying that it may increase basal glomerular permeability, we first sought to assess its acute, direct actions on the glomerular permeability in rats. Second, since the effects of mTORi have been reported to confer protection of the GFB during conditions of glomerular hyperpermeability, we tested mTORi also on the acute, dynamic effects of puromycin aminonucleoside (PAN) and ANG II, which are known modulators of glomerular permeability (5, 25, 32, 34, 47). Furthermore, since oxidative stress and induction of reactive oxygen species (ROS) is an established mechanism of PAN-induced and ANG II-induced glomerular hyperpermeability, we also studied, for comparison, the actions of the superoxide (O2·−) scavenger tempol on PAN-induced glomerular permeability and also on the permeability effects of temsirolimus by itself.
To assess glomerular permeability in intact rats, we assessed the glomerular sieving coefficient (θ) i.e., the primary urine-to plasma-concentration ratios of FITC-Ficoll 70/400 (Mr 70,000 and 400,000, respectively) of Stokes-Einstein radii (ae) ranging from 10 to 80 Å after mTORi and PAN or ANG II infusions. Ficoll is a neutral copolymer of sucrose and epichlorohydrine, which is not significantly reabsorbed in the proximal tubules and therefore can be used as a direct probe of glomerular permeability. By contrast, albumin, for example, is normally almost fully reabsorbed by the proximal tubules. We found that mTORi per se increased baseline glomerular permeability to Ficoll50-80Å but that it blunted the actions of PAN-induced permeability changes, underpinning the complex dual and context-dependent interactions of mTORi with the GFB.
MATERIALS AND METHODS
Animals and general surgery.
Experiments were performed in 41 male Wistar rats (Möllegard, Lille Stensved, Denmark) with an average body weight of 260.8 ± 2.4 g. The rats were given water and standard chow ad libitum. The animal Ethics Committee at Lund University approved the animal experiments.
Anesthesia was induced by an intraperitoneal injection of pentobarbital sodium (60 mg/kg body wt) and maintained during the experiments through repeated intra-arterial injections. The rats were placed on a heating pad to maintain body temperature at 37°C. The tail artery was cannulated (PE-50 cannula) for administration of anesthesia and for continuous monitoring of blood pressure and heart rate (HR; MP 150 system, AcqKnowledge for MAC, Biopac Systems). A tracheotomy was performed to facilitate breathing. The left carotid artery and left and right jugular veins were cannulated (PE-50) for blood sampling and infusion purposes, respectively. After an intravenous bolus dose of furosemide (0.375 mg/kg body wt, Furix, Takeda Pharma, Solna, Sweden), and after a small abdominal incision (∼6–8 mm), the left ureter was cannulated (PE-10 connected to a PE-50) for urine-sampling purposes, after which the incision was closed by a small suture.
All experiments started with an initial resting period of at least 20 min following the cannulation of the left ureter. Following the resting period, a baseline measurement (time 0) of FITC-Ficoll in plasma and urine (see below) was done in all of the experiments just before the injection of mTORi.
Glomerular permeability after a 30-min preincubation period with temsirolimus.
In one group of animals temsirolimus (Torisel, Pfizer, Sollentuna, Sweden) was administered in one single dose systemically intravenously (iv; mTORi, n = 7, 3.8 mg/kg body wt) 30 min before the start of the experiment. Measurements of Ficoll concentrations in plasma and urine were performed sequentially before the start of the mTOR injection (baseline) and at 5, 15, and 60 min after the end of the 30-min mTORi incubation period.
Systemic PAN infusion.
PAN (P7130, Sigma-Aldrich, St. Louis, MO) was administered immediately after baseline (time 0), as an initial bolus dose of 4.4 mg, followed by a continuous iv infusion (PAN, n = 8, 37.5 mg/kg or 266 μg·min−1·kg−1). Measurements of glomerular Ficoll sieving were performed sequentially before the start (baseline) and at 5, 15, 30, 60, and 120 min of the PAN infusion.
Effects of mTOR inhibition with temsirolimus on PAN- and ANG II-induced glomerular hyperpermeability.
In separate groups of animals, temsirolimus was administered iv (3.8 mg/kg body wt) 30 min before the start of the administration of either PAN (n = 7, 37.5 mg/kg) or ANG II (n = 6; bolus dose of 50 ng followed by 16 ng·min−1·kg−1; Lo-Ang). PAN or ANG II infusions were thus started immediately after the end of the mTORi incubation period. Measurements of glomerular Ficoll sieving coefficients were performed sequentially at baseline (before the mTORi injection) and at 5, 15, 30, 60, and 120 min after the start of the PAN administration, and at baseline and 5 and 15 min after the start of the ANG II infusion, respectively.
Scavenging of ROS.
During either PAN or ANG II infusions, or, in some experiments, following the temsirolimus injection, the ROS scavenger and SOD mimetic compound 4-hydroxy-tempol (tempol; Sigma-Aldrich, St. Louis, MO) was administered in separate experiments. Tempol was given as a continuous infusion iv (15 mg·h−1·kg−1), starting 5 min before the start of either PAN or ANG II infusions, or 25 min after the temsirolimus injection (3.8 mg/kg), and was continued throughout the experiments: the tempol-PAN group (n = 7), tempol-ANG II group (n = 6), and tempol-temsirolimus group (n = 6), respectively. Measurements of Ficoll sieving coefficients were performed sequentially at baseline and at 5, 15, 30, 60, and 120 min after the start of the PAN infusion, and at baseline and at 5 and 15 min after the start of the ANG II infusion, or at 30, 35, and 45 min after the temsirolimus injection, respectively.
Sieving of FITC-Ficoll.
A mixture of FITC-Ficoll-70 (10 mg/ml) and FITC-Ficoll-400 (10 mg/ml; TdB Consultancy, Uppsala, Sweden) in a 1:24 relationship was administered as a bolus dose together with FITC-inulin (10 mg/ml, TdB Consultancy). The bolus dose (FITC-Ficoll-70, 40 μg; FITC-Ficoll-400, 960 μg; FITC-inulin, 500 μg; and 51Cr-EDTA, 0.3 MBq) was followed by a constant infusion of 10 ml·kg−1·h−1 (FITC-Ficoll-70, 20 μg/ml; FITC-Ficoll-400, 0.48 mg/ml; FITC-inulin, 0.5 mg/ml; and 51Cr-EDTA, 0.3 MBq/ml) for at least 20 min before sieving measurements, after which urine from the left kidney was collected for 5 min, with a midpoint (2.5 min) plasma sample collected. A high-performance size-exclusion chromatography (HPSEC) system (Waters, Milford, MA) was used to determine size and concentration of the Ficoll samples. Size exclusion was achieved using an Ultrahydrogel-500 column (Waters). The mobile phase was driven by a pump (Waters 1525), and fluorescence was detected with a fluorescence detector (Waters 2475) with an excitation wavelength at 492 nm and an emission wavelength at 518 nm. The samples were loaded to the system with an autosampler (Waters 717 plus), and the system was controlled by Breeze Software 3.3 (Waters). The column was calibrated with Ficoll and protein standards described in a previous paper (1).
The sieving coefficients (θ) of FITC-Ficoll 70/400 were determined as the fractional clearance from θ = (CFU·CIP)/(CFP·CIU), where CFU represents the Ficoll urine concentration, CIP represents the inulin concentration in plasma, CFP the Ficoll concentration in plasma, and CIU the inulin concentration in urine.
GFR was measured in the left kidney during the experiment using 51Cr-EDTA. A priming dose of 51Cr-EDTA (0.3 MBq in 0.2 ml iv, Amersham Biosciences, Buckinghamshire, UK) was administered and followed by a continuous infusion (10 ml·h−1·kg−1) of 51Cr-EDTA (0.3 MBq/ml) throughout the experiment. Urine was collected from the left ureter repeatedly and blood samples, using microcapillaries, taken to be able to calculate GFR, approximately every 5–10 min. Radioactivity in blood and urine was measured in a gamma counter (Wizard 1480, LKP, Wallac, Turku, Finland). Hematocrit was assessed throughout the experiments to be able to convert blood radioactivity into plasma radioactivity. During each FITC-Ficoll sieving period, GFR was also assessed from the urine clearance of FITC-inulin (data not shown). The urinary excretion of 51Cr-EDTA (and FITC-inulin) per minute (Ut·Vu) divided by the concentration of tracer in plasma (Pt) was used to calculate GFR, where Ut represents the tracer concentration in urine, and Vu the flow of urine per minute. Basal GFR was assessed at time −10 min and at time 0, and then followed at 5, 15, 30 and 60 min, and also at 120 min, during PAN infusion, tempol-PAN, and temsirolimus-PAN infusions, respectively.
A two-pore model was used to analyze the θ data for Ficoll (mol. radius 10–80 Å). A nonlinear least-squares regression analysis was used to obtain the best curve fit, using scaling multipliers, as described at some length previously (31, 44). The four major parameters of the two-pore model are: the small-pore radius (rs), the large-pore radius (rL), the unrestricted pore area over unit diffusion path-length (A0/ΔX), and the fraction of the glomerular UF-coefficient accounted for by the large pores (αL).
Values are presented as means ± SE. Differences among groups were tested using nonparametric analysis of variance with the Kruskal-Wallis test and post hoc testing using the Mann-Whitney U-test. Bonferroni corrections for multiple comparisons were made when applicable. Significance levels were set at *P < 0.05 and **P < 0.01. All statistical calculations were made using IBM SPSS 20.0 for Windows (SPSS, Chicago, IL).
Effects of temsirolimus (after 30 min of preincubation) on glomerular permeability.
Temsirolimus increased the glomerular permeability (θ) to Ficoll70Å at 15 min after the mTORi preincubation period, but this permeability increase was spontaneously reversed within the next 40–50 min (Fig. 1). θ for Ficoll70Å thus increased from 2.91 × 10−5 ± 1.18 × 10−5 at baseline (before the temsirolimus incubation) to 2.27 × 10−4 ± 5.02 × 10−5 (P < 0.01) at 15 min after the end of the preincubation period, after which the glomerular permeability returned to control (Fig. 1). Figure 2 shows the Ficoll θ vs. ae for animals treated with temsirolimus alone. An increase in θ for Ficoll50-80Å was observed at 15 min, but was not seen at 5 and 60 min. Tempol completely abrogated the permeability increment at 15 min. Thus θ for Ficoll70Å was 1.40 × 10−5 ± 3.80 × 10−6 (n = 6) 30 min after the temsirolimus injection but remained unchanged at 1.95 × 10−5 ± 5.25 × 10−6 (P = 0.42; n = 6) 15 min later in the tempol-temsirolimus group.
PAN rapidly increased the glomerular permeability to Ficoll50-80Å.
PAN rapidly and transiently induced increases in glomerular permeability (θ for Ficoll70Å), the initial permeability increase peaking at 5 min, after which θ was reduced, but later again increased at 60–120 min (Fig. 3). θ for Ficoll70Å thus increased from 1.52 × 10−5 ± 6.60 × 10−6 at baseline to 2.23 × 10−4 ± 6.69 × 10−5 after 5 min (P < 0.01). At 60 and 120 min, there was again an increase in θ for Ficoll70Å to 1.12 × 10−4 ± 2.98 × 10−5 and 1.15 × 10−4 ± 6.13 × 10−5 at 60 and 120 min, respectively (P < 0.01).
Effects of temsirolimus and tempol on PAN-induced glomerular hyperpermeability.
Systemic administration of temsirolimus significantly reduced the late permeability increase induced by PAN at 60 and 120 min to baseline values (P < 0.01) but did not affect the initial permeability peak at 5 min. Tempol, however, effectively reduced the early increase in θ induced by PAN as well as that at 120 min of PAN administration (P < 0.05) (Fig. 3). At 120 min, both mTORi and tempol effectively reduced the sieving coefficient for Ficoll70Å in PAN-treated animals. At 120 min, θ for Ficoll70Å had thus been reduced from 1.15 × 10−4 ± 6.13 × 10−5 to 1.01 × 10−5 ± 2.10 × 10−6 and 2.08 × 10−5 ± 6.76 × 10−6, for the mTORi-PAN group and the tempol-PAN group, respectively. Figure 4 shows the glomerular sieving coefficients for FITC-Ficoll-plotted vs. molecular radius (ae) at 120 min in the mTORi- and tempol-treated PAN groups. A marked increase in θ for Ficoll50-80Å was seen after PAN at 120 min. Both mTORi and tempol completely abrogated this increase in glomerular permeability.
Effects of mTORi on ANG II-induced glomerular hyperpermeability.
Temsirolimus did not significantly affect the acute increase in glomerular permeability induced by ANG II (Fig. 5). For comparison, the effect on glomerular permeability of ANG II (alone) from a previous study (5) is shown (hatched line) in Fig. 5.
There were no significant changes in GFR (Fig. 6) or mean arterial pressure (MAP) or HR (data not shown) during the course of any of these experiments, except for a significant decrease in HR, starting immediately after the start of the infusion of mTORi (data not shown) (P < 0.05).
The best curve fits of θ vs. ae for Ficoll according to the two-pore model were obtained using the parameters listed in the tables below. Temsirolimus alone caused (at 15 min) an increase in the fractional ultrafiltration coefficient accounted for by large pores (αL) (P < 0.05) and the rL (P < 0.01). Baseline data and data at 60 min are shown for comparison (Table 1). PAN administration increased the rL at all time points and also fractional fluid flow through large pores (JvL/GFR) and αL significantly at 5 min (P < 0.05) (Table 2), indicating an increase in the number of large pores in the glomerular filter at this time point.
In the mTORi-PAN group (Table 3), significant increases in the rL were observed at 15, 30, and 60 min (not shown) after the start of PAN administration, while there were no significant changes in either αL or JvL/GFR or the effective pore area over unit diffusion path length (A0/ΔX). The same pattern of unchanging A0/ΔX and JvL/GFR and rL was seen for tempol-PAN at 120 min, while the rL was significantly increased at 15 min (Table 4). In Table 5 both rL and αL increased significantly at 15 min in the mTORi-LoAng group, consistent with previous findings of the effects of LoAng (16.2 ng/min/kg) alone in this experimental setup.
In the present study, we tested the effects of systemic mTORi on glomerular permeability under normal resting conditions in vivo and during acute podocyte and endothelial stress evoked by PAN or ANG II in anesthetized rats. Although mTORi in the form of temsirolimus exhibited transient effects on glomerular permeability, evidently invoked by ROS generation, it ameliorated the nonimmediate increases in glomerular permeability induced by PAN. However, mTORi had no effect on the glomerular hyperpermeability induced acutely by ANG II. Tempol, a superoxide anion scavenger, was, however, efficient in abrogating the immediate increases in permeability induced by PAN and also by temsirolimus similar to its documented effects on ANG II-induced GFB hyperpermeability (3).
There is now accumulating evidence that both overactivation and underactivation of the mTOR complex(es) can lead to glomerular dysfunction and proteinuria, suggesting that a balanced regulation of mTOR activity and a tight control of its effectors are crucial for maintaining a normal function of the GFB. Aside from its immunosuppressant actions in organ transplantation, a number of nonimmunological effects of mTORi on podocyte dynamics (the F-actin cytoskeleton) have been described (24, 25, 48). In podocyte cell cultures, the exposure of the cells to PAN for 2 days induced smaller cells with a “polarized” shape and reduced adhesion, reminiscent of a migratory fibroblast phenotype with diminished central stress fibers and substantial accumulation of thin (and less organized) actin fibers in the cell periphery. In such an in vitro system of cultured podocytes, mTORi (everolimus) had a protective effect with respect to PAN-induced podocyte injury, leading to a more normalized cell shape and recovery of actin stress fibers and an enhanced cell adhesion (25).
By contrast, short-term exposure of human podocytes to mTORi (sirolimus) under nonstressed conditions was found to downregulate VEGF synthesis and Akt phosphorylation, thereby affecting podocyte survival and adhesion (27). Furthermore, in an anti-Thy1.1 nephritis model the rapamycin derivative SDZ RAD was demonstrated to produce proinflammatory effects and accelerated renal damage, while in the PAN model SDZ RAD significantly ameliorated the development of nephrosis (12).
In chronic renal disease, particularly in the nephrotic syndrome, there is evidence that antioxidative mechanisms are generally impaired (18). The PAN model has over the years become an experimental prototype of human minimal change disease and focal segmental glomerulosclerosis (39), and there is accumulating evidence that ROS generation and oxidative stress are main features of the pathogenesis of the glomerular barrier disruption induced by PAN (13, 20, 36, 42, 43, 50). Hence, an increased production of ROS will activate TRPC6 Ca2+ channels and intracellular Ca2+ signaling (50) to increase glomerular permeability and to eventually produce DNA damage, cell cycle arrest, and thereby cell injury (32). Actually, ROS generation will peak already after 5 min of PAN administration, with a second peak after approximately 9 days, when the nephrotic syndrome has become full blown (20). In the classic study by Yoshioka et al. (53) in which ROS, in the form of hydrogen peroxide (H2O2), was infused directly into the left renal artery of Munich-Wistar rats, acute massive, reversible ROS-induced proteinuria occurred. In the present study, there was an immediate glomerular permeability increase induced by PAN which was similar to that previously observed in, for example, isolated glomeruli (34), and it could be reversed by administration of tempol. In the same fashion, the acute permeability actions of temsirolimus itself were also found to be ROS dependent. This ROS dependence of GFB permeability is similar to that observed after several other challenges to the GFB, such as those induced by ANG II (3), TNF-α (33), transforming growth factor-β (45), or fetal Hb (46). It thus seems evident that ROS generation plays a key role in the initiation and maintenance of, at least, the initial Ca2+-signaling steps involved in acute cell shape alterations and glomerular permeability changes occurring after various insults to the GFB.
While tempol was able to reverse the initial permeability peak induced by PAN, and also the sustained permeability at 2 h, temsirolimus was only affective after 60. Thus mTORi showed a GFB-protective effect distinct from that characterizing its interactions with ROS, which actually caused a transient early increase in glomerular permeability. In a study on the microvessels of the hamster cheek pouch, rapamycin (sirolimus) inhibited the acute permeability increases induced by VEGF and platelet-activating factor (PAF) when given in doses at 0.1 and 0.5 mg/kg. However, at higher doses, 10 mg/kg, rapamycin increased the microvascular hyperpermeability induced by VEGF and PAF (26). The authors speculated that this opposite paradoxical effect exerted by high doses of rapamycin would potentially result from an activation of mTORC1, associated with Akt activation (via a negative feedback loop). Also, high concentrations of, or prolonged exposure to, sirolimus may inhibit both mTORC1 and mTORC2 pathways (see below), which would also lead to downstream interactions with Akt, affecting cellular actin dynamics. Again, this illustrates the complex, context-dependent actions of mTORi.
The interactions among cellular Ca2+ signaling, ROS generation, and mTORC activation are complex and have been only partly mapped out. An archetypic model of intracellular Ca2+ pathways involved in ANG II signaling was recently proposed, involving several different levels and two major routes of “channeling” of intracellular Ca2+-signaling (19). At the first level (level 1), ANG II interacts with its receptor (AT1R), resulting in the activation of TRPC5 and TRPC6 channels to allow Ca2+ influx into the cell (level 2), which, in turn, activates at least two separate, but highly interconnected, Ca2+-signaling cascades. At this level, membrane-bound NAD(P)H oxidase (Nox) complexes seem to be activated to generate ROS, which may reinforce or amplify the subsequent signaling steps. Next (level 3), a Ca2+-activated phosphatase, calcineurin, or kinases, such as PKA, are activated and are assumed to compete for downstream effects on synaptopodin, mainly a target of calcineurin, and on nuclear factor of activated T-cell (NFAT), mainly a target of protein kinases (level 4). At the next level (level 5), the small GTPases Rac-1, RhoA, and Cdc42 are suggested to compete for downstream effects on the actin cytoskeleton (level 6), inducing either cell contraction and “stiffness” via RhoA activation, or “hypermobility” and possibly foot process (FP) effacement in podocytes via Rac-1 activation. Note that Rac-1 will also bind to and activate Nox4 to increase O2·− generation (at level 2). It was hypothesized that in states of excess ANG II, TRPC5/Rac-1 overactivity may drive proteinuria, whereas under physiological conditions moderately active TPRC6 channels would be more important. In addition, also a “third” Ca2+-signaling cascade has been proposed, by which mTORC2 is activated (at level 3), which, in turn, induces Akt activity by phosphorylation of Akt at Ser473, resulting in reorganization of the cytoskeleton (48). However, as it seems, only prolonged treatment with or high doses of sirolimus would block mTORC2 signaling. Hence it is likely that in the present acute experiments the amelioration of the effects of PAN after 60 and 120 min occurred by inhibition of mTORC1 and not by mTORC2, and that the earliest stage of mTORC1 inhibition resulted in an acute activation of Nox and cellular Ca2+ signaling, transiently increasing glomerular permeability. The subsequent actions of temsirolimus are conceivably of a different nature. Thus more long-term effects of mTORi (prolonged sirolimus treatment) have been shown to include downregulation of the diaphragm slit protein, nephrin, and also of TRPC6 Ca2+ channels, as well as the expression of the cytoskeletal adaptor protein Nck, and to reduce podocyte adhesion and proliferation (49). Furthermore, a recent study demonstrated that mTORC1 activation can directly lead to increased Nox4 levels in podocytes, and that mTORi, in contrast to its acute and transient effects, may function as a major downregulator of ROS generation in the kidney (14).
In this study, no permeability measurements were performed during the 30 min of preincubation with temsirolimus. For that reason, we cannot exclude that temsirolimus had altered glomerular permeability already early during the incubation period, after which glomerular permeability was again restored to baseline, to again transiently increase. Such a cyclic pattern of glomerular permeability has been observed after e.g., systemic ANP infusion (7). The rationale for having a “silent” preincubation period of 30 min before the starting of permeability measurements was to standardize the conditions for all the experiments, irrespective of the agent tested (ANG II or PAN). In all likelihood, the measured permeability changes for temsirolimus alone may thus represent a second peak of permeability increase coinciding with the assigned measurement period.
This group has previously demonstrated that a number of challenges to the GFB, such as anaphylaxis (4), acute hyperglycemia (6), systemic ANP infusions (7), or infusions of ANG II (5) or fetal Hb (46) can transiently open the GFB, conceivably by affecting the contractility of the podocytes and/or the endothelium. The exact nature of these acute permeability alterations is not known. However, podocytes may be crucial for maintaining or increasing barrier permeability. Although there is good evidence that the ultimate sieving barrier to proteins is not at the podocyte level (31), podocyte interactions with the rest of the GFB are important for its integrity. Podocytes are anchored to the GBM by integrins, and they seem to exert pressure on the GBM by their contractility, conceivably adapting the wall tension to the glomerular hydrostatic capillary pressure. Any changes in podocyte actin dynamics may thus directly affect the uphill components of the GFB, and hence, its function. Furthermore, the endothelial cells may also be involved in regulating the permeability of the GFB in a way similar to their role in peripheral capillaries, where paracellular gaps can open and close in a transient and cyclic fashion in response to various permeability challenges (35). At any rate, there is good evidence supporting cross talk between the glomerular endothelium and podocytes in the regulation of the permeability of the GFB (23).
Given the close similarity of θ for albumin and θ for Ficoll50-80Å, or more precisely, θ for Ficoll55Å, as demonstrated in several previous studies from this group (2, 4, 44), we employed in the present study glomerular θ for high-molecular-weight Ficoll as a surrogate marker for θ of albumin. Indeed, θ values for albumin and those for Ficoll55Å have been shown to be almost identical under various conditions, demonstrated also by other groups (37). Therefore, we feel confident that glomerular θ for Ficoll50-80Å are good indicators of glomerular albumin permeability.
In summary, the present study demonstrates that in the acute setting mTORi, in the form of systemic temsirolimus administration, has itself acute and dynamic effects on the GFB, causing ROS-dependent, transient glomerular permeability alterations. While temsirolimus was able to blunt the permeability actions of PAN after 60 and 120 min, it exerted no protective effects on the initial (0–30 min), ROS-dependent permeability peaks induced by either PAN or ANG II. The present study amply demonstrates the dual, opposing effects of mTORi on the GFB also in an acute experimental setting in vivo.
This study was supported by Swedish Medical Research Council (Grant 08285), the Swedish Heart and Lung Foundation, and the Medical Faculty at Lund University.
No conflicts of interest, financial or otherwise, are declared by the authors.
Author contributions: J.A., A.R., and B.R. provided conception and design of research; J.A. and A.R. performed experiments; J.A., A.R., and B.R. analyzed data; J.A. and B.R. interpreted results of experiments; J.A. prepared figures; J.A. and B.R. drafted manuscript; J.A., A.R., and B.R. edited and revised manuscript; J.A., A.R., and B.R. approved final version of manuscript.
The valuable advice from Dr. Javier de Arteaga, Catholic University of Cordoba, Cordoba, Argentina, at the initiation of this study is acknowledged. Kerstin Wihlborg is gratefully acknowledged for skillful typing of the manuscript.
- Copyright © 2015 the American Physiological Society