Chronic Inhibition of Mammalian Target of RapamycinSignaling Downregulates Insulin Receptor Substrates 1 and 2and AKT Activation: A Crossroad between Cancer andDiabetes?

Salvatore Di Paolo, Annalisa Teutonico, Domenica Leogrande, Carmen Capobianco, andPaolo F. SchenaDepartment of Emergency and Organ Transplants, Division of Nephrology, Dialysis and Transplantation, University ofBari, Policlinico, Bari, Italy

Overactivation of the mammalian target of rapamycin (mTOR) branch downstream of the phosphatidylinositol 3-kinase–AKTpathway critically modulates insulin and growth factor signaling by insulin receptor substrates (IRS). On the basis of in vitrostudies, the mTOR inhibitor rapamycin has been reported to lead to enhanced activation of AKT by relieving this feedbackinhibition on IRS function. In view of the critical role of AKT in insulin signaling and tumorigenesis, the in vivo expressionand activation of this kinase and of IRS-1 and IRS-2 were explored in PBMC of 30 patients who were treated long term withrapamycin. A marked decrease of basal and insulin-stimulated AKT phosphorylation, which correlated with the increase ofpatients’ insulin resistance, and a significant increase of IRS total protein expression, together with a lower (IRS-2) or absent(IRS-1) increase of insulin-induced tyrosine phosphorylation, were found. Therefore, contrary to the expectations, long-termexposure to rapamycin caused the impairment of IRS signaling and AKT activation, and this would help to explain theantiproliferative effect and the possible deterioration of glucose metabolism that are observed in rapamycin-treated patients.These findings may form a novel basis for improved understanding of the role of mTOR inhibition in human diseases, suchas diabetes and cancer.

J Am Soc Nephrol 17: 2236–2244, 2006. doi: 10.1681/ASN.2006030196

T he mammalian target of rapamycin (mTOR) is an evo-lutionarily conserved serine/threonine kinase that in-tegrates signals from nutrients (amino acids and en-

ergy) and growth factors to regulate protein translationthrough downstream effectors such as ribosomal S6 kinase(S6K) and the translation factor inhibitor 4EBP1 (1). This path-way is critical for proper cell growth, cell cycle progression, andregulation of organ size (1).

Most growth factors and cytokines activate mTOR and itsdownstream targets primarily through the phosphatidylinosi-tol 3-kinase (PI3-K)-AKT pathway (1). The identification of thecomponents of the tuberous sclerosis protein complex, a het-erodimer of the proteins hamartin and tuberin, has establishedthe link between upstream growth factor–signaling cascadesand mTOR, providing a mechanism for the crosstalk that oc-curs between these pathways (1,2). The tuberin–hamartin com-plex potently inhibits mTOR-dependent signaling; AKT-di-rected phosphorylation of tuberin relieves its inhibition ofmTOR via activation of the small G protein Rheb (3).

Building on studies that have examined the forward mode ofmTOR signaling that is activated by the growth factors–PI3-K–AKT pathway, recent studies have provided an insight into theparticular role of growth factor signaling via mTOR and S6K inthe reverse regulation of PI3-K activation and on the principalrole of insulin receptor substrate (IRS) proteins as recipients ofnegative regulation from kinases such as S6K (4). Therefore,physiologic conditions that lead to the activation of mTORcascade, such as excess nutrient supply or hyperinsulinemia,promote serine/threonine phosphorylation of IRS-1 and IRS-2that inhibits their function, promotes their degradation, and,possibly, inhibits their transcription (1–5). In this way, thehyperactivation of mTOR/S6K cascade would exert a signifi-cant negative effect on the activity of downstream componentsof the insulin/PI3-K pathway, such as AKT, and in this waywould contribute to or exacerbate the insulin resistance thatcauses metabolic disease and type 2 diabetes.

Several in vitro studies have shown that inhibitors of mTORpathway, such as rapamycin, relieve the repression of IRS-1–dependent PI3-K/AKT signaling in states of increased activityof mTOR/S6K pathway and lead to enhanced basal and stim-ulated AKT serine 473 phosphorylation (2,5–8), although onestudy reported that chronic repression of mTOR signaling byrapamycin in HEK 293 cells caused a significant reduction inAKT phosphorylation and catalytic activity (9). Paradoxic,these findings might represent a caveat for future use of mTOR

Received March 4, 2006. Accepted May 4, 2006.

Published online ahead of print. Publication date available at www.jasn.org.

Address correspondence to: Dr. Salvatore Di Paolo, Department of Emergencyand Organ Transplants, Division of Nephrology, Dialysis and Transplantation–University of Bari, Policlinico-Piazza Giulio Cesare, 11-70124 Bari, Italy. Phone:�39-080-5593231; Fax: �39-080-5593227; E-mail: s.dipaolo@nephro.uniba.it

Copyright © 2006 by the American Society of Nephrology ISSN: 1046-6673/1708-2236

inhibitors, because activation of AKT could provide cancer cellswith an increased survival signal that may be clinically unde-sirable (7,10,11). It is interesting that recent work indicates thattumors that are formed in mouse models of tuberous sclerosisprotein complex may be relatively nonaggressive because acti-vation of raptor–mTOR and S6K1 represses the PI3-K/AKTpathway (12,13).

Recent research, however, has furnished new insights intothe molecular mechanisms that regulate mTOR and their role ingrowth and disease. Sorbassov et al. (14) showed that mTORkinase and its associated protein, rictor, are necessary for serine473 phosphorylation of AKT and that a reduction in rictor ormTOR expression inhibited AKT. Also, mTOR inhibition byrapamycin was reported to impede insulin-induced phosphor-ylation of IRS-1 on tyrosine residues, resulting in decreasedinsulin sensitivity, in this way transforming human adipocytesinto a diabetes-like phenotype (15).

Preliminary evidence arising from clinical studies supportsthe use of rapamycin and its analogues in human cancers withconstitutively activated PI3-K/AKT pathway (16–19) but, con-versely, seems to disclose an increased association with hyper-glycemia (16,19–22). Accordingly, we recently reported thatrapamycin inhibits the progression of Kaposi’s sarcoma (23), ahuman malignancy in which AKT activation has been sug-gested to exert a central role (24). We also found that chronicmTOR inhibition caused a significant increase of peripheralinsulin resistance, along with an impairment of pancreatic �

cell response to glucose load, in a cohort of renal transplantrecipients who discontinued the calcineurin inhibitor cyclo-sporin A (CsA) and were converted to rapamycin (25).

Because AKT is a central player in the signal transductionpathways that are activated in response to growth factors orinsulin and the alteration of its activity may induce neoplasiaand transformation in some tissues or lead to aberrant glucosemetabolism and, eventually, diabetes, we designed this studyto evaluate the effect of long-term treatment with rapamycin onthe in vivo expression and activation of this protein kinase incirculating mononuclear cells (PBMC) of kidney transplantedpatients. Then, we extended this study to explore the effect ofchronic mTOR inhibition on IRS-1 and IRS-2 activation.

Materials and MethodsPatients

We enrolled 30 CsA-treated renal transplant recipients (16 men/14women; age 44.2 � 12.2 yr) who received the histologic diagnosis ofchronic allograft nephropathy 37.5 � 28.0 mo (range 11 to 119) aftertransplantation and were asked to be converted to rapamycin (trough8 to 12 ng/ml), without any further modification of the remainingimmunosuppressive therapy (low-dose steroids: prednisone 2.5 to 5mg/d and mycophenolate mofetil 1 to 2 g/d) (25). All patients wereasked to give their written informed consent to participate to the study,according to the Guidelines of the Local Ethical Committee.

All patients had anthropometric and laboratory parameters checkedat baseline and at the end of the 6-mo follow-up. Rapamycin exposurewas evaluated by predose (trough) monitoring. Blood concentration ofrapamycin was determined using an HPLC assay with ultravioletdetection, as described previously (26).

In vivo activation of IRS–AKT–mTOR pathway was attained by in-

travenous administration of 0.1 IU/kg regular insulin. Blood samples(25 ml) were drawn immediately before and 15 min after insulininjection and processed as detailed next.

Western BlottingHuman PBMC were isolated from freshly drawn blood by Ficoll-

Hypaque (Amersham Pharmacia, Little Chalfont, UK) gradient centrif-ugation, in the presence of 1% (vol/vol) protease inhibitor cocktail and1% (vol/vol) phosphatase inhibitor cocktail. Cells were lysed and sub-jected to blotting as described (27), with minor modifications. Briefly,aliquots that contained 50 �g of proteins from each lysate were sub-jected to SDS/PAGE on a 10% gel under reducing conditions and thenelectrotransferred onto a polyvinylidene difluoride membrane (Im-mun-Blot PVDF Membrane, 0.2 �m; BioRad, Milan, Italy) using asemidry Transblot apparatus (BioRad). After overnight blocking, thefilter was probed with either polyclonal rabbit anti–phospho-AKT(pAKT) antibody targeted to serine 473 (Saint Cruz Biotechnologies,Santa Cruz, CA) or mouse monoclonal anti–phospho-P70 S6K anti-body, raised against a peptide that contained the phosphorylated serine411 (Santa Cruz Biotechnologies), at 1:200 dilution in TBS at 24°C for2 h. The membranes then were washed in TBS and incubated for 1 hwith horseradish peroxidase–conjugated goat anti-rabbit or goat anti-mouse secondary antibody (BioRad) at 1:1500 dilution in TBS. The samemembranes then were stripped and immunoblotted with anti-humanAKT or anti-human P70 S6K mAb (Santa Cruz Biotechnologies).

The ECL enhanced chemiluminescence system (Amersham Bio-sciences, Little Chalfont, UK) was used for detection according to themanufacturer’s instructions. The intensity of signals that were detectedby ECL was quantitated by densitometric analysis, and results wereexpressed in arbitrary units (AU).

Cell Immunofluorescence and Confocal Laser ScanningMicroscopy

The low abundance of IRS docking proteins in human PBMC pre-cluded the use of Western blotting for the analysis of IRS expressionand activation in a clinical study. We therefore chose to perform adouble-fluorescence immunolabeling (27) to quantify the total expres-sion of IRS-1 and IRS-2 proteins and the levels of their activated(tyrosine phosphorylated) forms. In preliminary experiments, we usedPBMC from buffycoat byproducts of whole blood that was donated byhealthy volunteers to compare the quantification of basal and insulin-stimulated IRS tyrosine phosphorylation by immunoblotting with thequantification of signals that were detected by immunocytochemistry.Cells were incubated with 10 nM human insulin for 15 min. Afterward,the cells were extracted and subjected to immunoprecipitation withanti–IRS-1 or IRS-2 antibodies, followed by immunoblotting with anti-phosphotyrosine antibody (28), or underwent immunocytochemicalanalysis, as described next. By both methods, we found a markedintersubject variation in the total expression and in the extent of basaland insulin-stimulated tyrosine phosphorylation of IRS-1 and IRS-2proteins, which emphasizes the need to use each patient as his or herown control in the analysis. Importantly, signal quantification by im-munocytochemistry was in accordance with the results of Western blotanalysis. Finally, immunocytochemistry showed an intrasubject vari-ability (three healthy subjects) over a 2-wk period of only 12%.

For immunocytochemical analysis, PBMC (2 � 105/slide) wereseeded on a cover glass, fixed with 2% paraformaldehyde, and incu-bated first with 0.2% Triton X-100 in PBS then with a blocking solutionof 3% BSA in PBS. Subsequently, the cells were incubated with twoprimary antibodies: A polyclonal anti–IRS-1 or anti–IRS-2 antibody(1: 500 dilution in 3% BSA/PBS; Upstate Biotechnology, Dundee, UK)

J Am Soc Nephrol 17: 2236–2244, 2006 mTOR Signaling and Downregulation of IRS-1/2 and AKT 2237

and a monoclonal anti–phospho-tyrosine antibody (1:500 dilution; Up-state Biotechnology) for 2 h at 37°C in a humidified container. Afterwashing, the cells were incubated in a mixture of two secondaryantibodies: Alexa Fluor 488 goat anti-mouse IgG-FITC conjugate (1:300dilution; Molecular Probes, Eugene, OR) and Alexa Fluor 543 goatanti-rabbit IgG-TRITC conjugate (1:600 dilution; Molecular Probes).The slides then were mounted in Gel/Mount (Biomeda, Foster City,CA) and sealed. Finally, PBMC were analyzed by confocal laser scan-ning microscopy using the Leica TCS SP2 (Leica, Wetzlar, Germany)(27).

Statistical AnalysesDifferences between quantitative variables were tested by the Mann-

Whitney U test or Wilcoxon signed rank test, as appropriate. Therelationship between nonparametric variables was tested by Spearmanrank correlation. P � 0.05 was considered statistically significant. TheStatview software package (5.0 version; SAS Inc., Cary, NC) was usedfor all analyses.

ResultsPatients

Throughout the follow-up, care was taken to avoid any majormodification of pharmacologic therapy (25). None of the pa-tients studied had an acute rejection episode or clinically rele-vant infections; neither did they show significant modificationsof body mass index (start 25.7 � 4.7 k/m2; end 25.4 � 3.8 k/m2)and renal function, measured by serum creatinine (SCr; start1.52 � 0.47 mg/dl; end 1.60 � 0.71 mg/dl) and by creatinineclearance (start 63.9 � 20.8 ml/min; end 59.4 � 19.52 ml/min),throughout the follow-up period. Specifically, 56.66% of pa-tients had a decrease of SCr values (�0.31 � 0.21 mg/dl), 11(36.66%) showed a worsening of graft function (sCr: 0.39 � 0.22mg/dl), and two (6.66%) remained stable during the follow-upperiod.

Effect of Chronic Exposure to Rapamycin on AKTActivation

Six-month therapy with rapamycin was associated with a de-crease in the mean levels of both basal and insulin-stimulatedAKT phosphorylation (Figure 1, A and B), without any modifica-tion of total AKT content. In view of the large intersubject varia-tion in the expression and phosphorylation of the target kinase,we calculated the percentage change from baseline of pAKT, basaland stimulated, in each patient to ensure a more accurate assess-ment of drug-induced changes. This approach seems to havereinforced the evidence of the depression of AKT phosphorylationafter chronic mTOR inhibition by rapamycin (Figure 1C).

The decrease of insulin-stimulated activation of AKT, calcu-lated as percentage change from pretreatment values in eachpatient, turned out to correlate with the deterioration of periph-eral insulin sensitivity, as measured by the �-change in thekinetic decline of glucose after insulin infusion (25) (Figure 2,top). The �-change of insulin-stimulated serine 473 phosphor-ylation of AKT inversely correlated with rapamycin troughlevels: The higher the blood concentration of rapamycin, thelower the insulin-stimulated increase of pAKT after rapamycintherapy (Figure 2, bottom). Then, we wondered whether the

modifications of kidney graft function over time would affectAKT phosphorylation, but we did not find any significantrelationship between the depressed activation of the targetkinase and the �-change of either SCr or creatinine clearance.

Strong evidence suggests that intracellular lipid accumula-tion can adversely affect IRS-1–PI3-K–AKT signaling pathway,thereby decreasing insulin sensitivity (29). The search for arelationship between the � change of serum triglyceride levelsand the decrease of AKT phosphorylation revealed a trendtoward higher triglyceride levels in patients with lower levelsof AKT phosphorylation, which was not statistically significant,

Figure 1. Pharmacologic inhibition of mammalian target ofrapamycin (mTOR) inhibits AKT activation. (A) Representativeimmunoblots using antibody that recognizes only the phos-phorylated serine 473 form of AKT (pAKT) and antibody thatrecognizes total (both nonphosphorylated and phosphorylated)AKT in PBMC of renal transplant recipients, under basal andinsulin-stimulated conditions. (B) Quantification of pAKT as aratio of phosphorylated to total AKT, before and after 6 mo oftreatment with rapamycin. Results are the mean (�SEM) of theabsolute values measured in each patient. (C) Delta (�) changeof basal and insulin-stimulated AKT phosphorylation afterlong-term exposure to rapamycin. Data represent the mean(�SEM) of the percentage change from baseline measured ineach patient. *P � 0.0001.

2238 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 2236–2244, 2006

however (P � 0.27). This may reflect simply the lack of a strictcorrelation between serum lipid levels and intracellular triglyc-eride levels (30).

Finally, to rule out the possibility that the modification ofAKT activation could be accounted for, at least in part, by thewithdrawal of CsA, we performed a set of in vitro experimentsin which PBMC from control healthy subjects were cultured incomplete medium in the presence or in the absence of 200ng/ml CsA. Then the cells were challenged with 10 nM regularinsulin for 15 min, lysed, and subjected to blotting. Neithershort-term (2 h) nor long-term (48 h) exposure to the calcineurininhibitor seemed to influence the basal and insulin-stimulatedserine 473 phosphorylation of AKT, as already demonstrated invitro and in vivo (31,32).

Impact of Chronic Rapamycin Therapy on P70 S6KPhosphorylation

The P70 S6K represents one of the crucial targets of mTOR inthe response to nutrients and in the control of cell growth, and,conversely, its inhibition is a measure of the pharmacologiceffect of rapamycin. Therefore, to explore further whether cir-culating mononuclear cells would reflect the biologic effect ofthe mTOR inhibitor in vivo, we measured the activation of P70S6K in PBMC that were isolated from patients before and after6 mo of therapy with rapamycin. Chronic treatment with themTOR inhibitor caused a 32.4 � 5.3% decrease of basal P70 S6Kphosphorylation (from 3498.3 � 397.1 to 2187.5 � 294.7 AU;P � 0.0001) and a 62.3 � 3.3% decrease of insulin-stimulatedP70 S6K activation (from 7191.6 � 788.1 to 2667.9 � 352.5 AU;P � 0.0001).

Modulation of IRS-1 and IRS-2 Expression and TyrosinePhosphorylation

First, we found a significant increase of the total proteinexpression of both the docking proteins after 6 mo of exposureto rapamycin (Figure 3). This seemed to confirm the ability ofmTOR inhibition to prevent the proteasomal degradation ofIRS proteins and, possibly, the inhibition of their transcription(1–5).

Then we evaluated IRS-1 and IRS-2 tyrosine phosphoryla-tion. Chronic mTOR inhibition was associated with a signifi-cant increase in the basal phosphorylation of both of the dock-ing proteins (Figures 4 and 5) and with a complete lack ofinsulin-stimulated increase of IRS-1 tyrosine phosphorylation(Figure 4). Intravenous insulin injection increased IRS-2 phos-phorylation 2.1-fold at baseline and 1.5-fold 6 mo after theconversion to rapamycin; in other words, mTOR inhibition wasassociated with a significant depression of the incremental in-sulin activation of IRS-2 (Figure 5).

Last, to ascertain whether the effects of rapamycin on IRS-1/2–AKT–mTOR–P70 S6K pathway described herein were revers-ible, PBMC from three rapamycin-treated patients were restedovernight (18 h), then rechallenged in vitro with 10 nM insulinand finally processed to measure AKT, P70 S6K, and IRS-1 andIRS-2 expression and activation. In each case, the inhibition oftarget proteins was reversed to pretreatment levels by over-night incubation of the cells.

DiscussionThis is the first demonstration that chronic exposure to the

mTOR inhibitor rapamycin inhibits basal and insulin-stimu-lated activation of AKT in a cohort of renal transplant recipi-ents. The data presented herein suggest that rapamycin leadsalso to a dysregulation of tyrosine phosphorylation (i.e., activa-tion) of IRS-1 and IRS-2 proteins.

Previous in vitro studies demonstrated that the mTOR sig-naling pathway is crucial in promoting specific serine/threo-nine phosphorylation of IRS-1 and IRS-2 that inhibits theirfunction and promotes their degradation (1–5). Conversely,rapamycin is expected to relieve IRS from negative regulation,thereby restoring the activation of the PI3-K–AKT pathway.However, serine/threonine phosphorylation long has been

Figure 2. The inhibition of AKT activation correlates with theincrease of peripheral insulin resistance and with rapamycinblood levels. (Top) Correlation between the percentage changefrom baseline (�-change) of insulin-stimulated AKT phosphory-lation after 6 mo of treatment with rapamycin and the modifica-tion of insulin resistance, as measured by �-KITT (kinetic declinein glucose after intravenous insulin injection; negative values in-dicate an increase of insulin resistance). � � 0.670; P � 0.0003(Spearman rank correlation). (Bottom) Correlation between�-change of AKT activation and rapamycin blood concentration(trough). � � �0.679; P � 0.0003 (Spearman rank correlation).

J Am Soc Nephrol 17: 2236–2244, 2006 mTOR Signaling and Downregulation of IRS-1/2 and AKT 2239

suggested to have both a positive and a negative regulatory roleon tyrosine phosphorylation of IRS-1 and IRS-2 by insulin andIGF-1 receptors (33). Recently, Danielsson et al. (15) demon-strated that insulin-stimulated phosphorylation of serine 307 isrequired for efficient insulin signaling in human adipocytesand is inhibited by rapamycin, which reduced insulin sensitiv-ity and mimicked type 2 diabetes. The ability of rapamycin toinhibit serine 307 phosphorylation obviously points to themTOR pathway in relaying the phosphorylation. A potentialcandidate for serine 307 phosphorylation of IRS-1 is P70 S6K(2). Then, serine 307 is located in a consensus sequence forphosphorylation by AKT (34): the depressed activity of thiskinase, as reported here, therefore might impinge on IRS-1serine 307 phosphorylation.

Our findings showed that chronic treatment with rapamycinresulted, as expected, in the increase of IRS-1 and IRS-2 totalprotein expression (Figure 3). However, mTOR inhibitionmarkedly hampered the incremental increase of IRS-1 and, less,IRS-2 tyrosine phosphorylation after insulin stimulation, alongwith a significant increase in the basal phosphorylation of bothof the docking proteins (Figures 4 and 5). A closely similarpattern has been reported to occur naturally in patients withinsulin-resistant type 2 diabetes (35) as well as in glucose-intolerant first-degree relatives of patients with type 2 diabetes(36).

Figure 3. Long-term exposure to rapamycin increases insulinreceptor substrate-1 (IRS-1) and IRS-2 total protein content.(Top) The concentration of the docking proteins was analyzedby confocal microscopy using PBMC that were isolated fromrenal transplant patients before (�) and after (�) 6 mo oftherapy with rapamycin. (Bottom) Quantification of IRS-1 andIRS-2 total protein concentration. Results are expressed asmean � SD of IRS-1/2 concentration of approximately 100mononuclear cells in each patient.

Figure 4. Rapamycin inhibits insulin-induced tyrosine phos-phorylation of IRS-1. (Top) Tyrosine phosphorylation of IRS-1was analyzed under basal and insulin-stimulated conditionsbefore and after 6 mo of treatment with rapamycin. PBMC werefixed and stained with antibodies to phosphotyrosine, as wellas with anti–IRS-1 antibodies. Cells were visualized by confocalmicroscopy. Double-labeled cells then were optically merged sothat areas of co-localization of the signals could be quantified.In these combined images, phospho-tyrosine is green and IRS-1staining is red; the yellow speckles indicate co-localization.(Bottom) Quantification of basal (�) and insulin-stimulated (�)IRS-1 tyrosine phosphorylation.

2240 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 2236–2244, 2006

The elevated basal phosphorylation/activity of IRS proteinslikely stems from the removal of mTOR-S6K negative feedbackregulation. The incremental defect in insulin-stimulated ty-rosine phosphorylation of IRS proteins may be accounted for bythe elevated basal phosphorylation of the docking proteins,which possibly masks or inhibits any further increase in signaltransduction in response to insulin (36). Then it may result fromthe inhibition of serine 307 phosphorylation, which dampensthe steady-state level of IRS-1 tyrosine phosphorylation (15). Inconclusion, although a rapamycin-sensitive pathway is in-volved in the inhibitory mechanisms impinging on the insulin–IRS signaling pathway in response to prolonged insulin expo-sure and excess nutrient supply, chronic mTOR inhibition

would fail to improve but rather deteriorates IRS-1 and IRS-2function.

The main objective of this study, however, was to ascertainthe effect of chronic exposure to rapamycin on AKT activation,in view of its crucial role in the regulation of cell growth andsurvival, as well as in the cell response to nutrients and growthfactors. We found that chronic mTOR inhibition was associatedwith a marked downregulation of basal and insulin-inducedAKT phosphorylation. AKT is responsible primarily for manyof the metabolic actions of insulin (37). Not surprising, there-fore, the depression of AKT activation significantly correlatedwith the increase of insulin resistance in renal transplant recip-ients (Figure 2).

The inhibition of AKT activation may represent the logicalresult of the disruption of IRS protein function. Moreover, therictor–mTOR complex was identified recently as a hydrophobicmotif kinase for AKT phosphorylation on serine 473, mTORbeing the phosphorylating kinase (15). Although rapamycindoes not bind to a preformed rictor–mTOR complex (38), dur-ing long-term treatment, the drug eventually should sequestermany of the newly synthesized mTOR molecules within cells,thereby inhibiting rictor–mTOR activity and AKT phosphory-lation (14). In this way, the prolonged presence of rapamycinmay inhibit the kinase, and this, in turn, may impede serine 307phosphorylation of IRS-1 (15). Figure 6 summarizes the role ofmTOR protein complexes in insulin signaling and the interfer-ence of rapamycin both upstream and downstream of AKT.

Although in vitro experiments showed the lack of any effectof CsA on AKT phosphorylation in human PBMC, our studydoes not allow us to rule out the possibility that the downregu-lation of basal and insulin-induced AKT phosphorylation ob-served in vivo may be accounted for, at least in part, by thewithdrawal of CsA.

The model of the S6K1-deficient mouse has shown that low-ering S6K1 levels potentiates insulin-induced AKT phosphory-lation (39). However, mice that are deficient in S6K1, unlikerapamycin-treated subjects, are expected to exhibit a fully ac-tive rictor–mTOR complex and thus a preserved activation ofAKT. In addition, at variance with patients studied here, S6K1-deficient mice are markedly hypoinsulinemic (40), and thismight protect IRS proteins from uncontrolled basal tyrosinephosphorylation. Taken together, these data seem to suggestthat mTOR inhibition may affect cell responsiveness differentlyfrom specific S6K inhibition.

In healthy individuals, � cell mass adapts to an increasedmetabolic load caused by excess nutrient supply and insulinresistance, via the AKT–mTOR–P70 S6K signal transductionpathway, downstream of IRS-2/PI3-K activation (40–42). Wepreviously reported that 6 mo of treatment with rapamycin ledto a defect in the compensatory � cell response to increasedinsulin resistance (25). On the basis of the findings presentedhere, it might be inferred that pharmacologic manipulationsthat impair IRS-2 function, as well as AKT and S6K activation,could compromise the ability of � cells to compensate for anincreased metabolic demand.

AKT and likely also IRS-1 and IRS-2 have been reported to beoverexpressed or activated in a variety of human and experi-

Figure 5. Effect of rapamycin on basal and insulin-inducedtyrosine phosphorylation of IRS-2. (Top) Tyrosine phosphory-lation of the docking protein was analyzed by confocal micros-copy of PBMC using anti-phosphotyrosine and anti–IRS-2 an-tibodies. Signals that correspond to total IRS-2 are visualized inred, phosphotyrosine is in green, and overlapping signals are inyellow. (Bottom) Quantification of basal (�) and insulin-stim-ulated (�) IRS-2 tyrosine phosphorylation before and after 6mo of treatment with rapamycin.

J Am Soc Nephrol 17: 2236–2244, 2006 mTOR Signaling and Downregulation of IRS-1/2 and AKT 2241

mental malignancies (43–45). Consequently, the expected abil-ity of rapamycin to relieve the feedback inhibition on the IRSbranch of the growth factor signaling cascade and downstreamAKT activation could provide cancer cells with an increasedsurvival signal that may be clinically undesirable (7,10,11). Thedata presented herein would hamper the above concerns, dem-onstrating that chronic rapamycin treatment in humans is as-sociated with impaired activation of both IRS adaptor proteinsand AKT, and this may represent a novel mechanism in theanticancer effects of the mTOR inhibitor.

This study allows us to explain better and reconcile ourprevious observations in the clinical setting, namely the regres-sion of Kaposi’s sarcoma lesions and the marked increase ofperipheral insulin resistance, together with the impairment ofpancreatic � cell response to glucose load, in patients who aretreated long term with rapamycin (23,25). However, we areaware that a potential limitation of this study arises from thecellular model that was used to detect the effects of mTORinhibition on intracellular signaling, i.e., circulating mononu-clear cells. Evidence exists that PBMC may represent a valuable

cellular model for the investigation of insulin signaling in hu-mans and for the study of postreceptor signaling defects ininsulin-resistant states (46,47). Moreover, in cancer patients, thedegree of S6K inhibition by the rapamycin-derivative CCI-779was reported to be identical in PBMC and simultaneously col-lected tumor tissue, suggesting that the PBMC are an adequatesurrogate tissue for studying the activity of mTOR inhibitors invivo (48). Regardless, it will be important to establish whether themechanisms described herein are operating in target tissues ofinsulin, such as muscle, liver, brain, and in islet � cells.

ConclusionHere we present evidence that chronic inhibition of the mTOR/

S6K pathway by rapamycin is associated with an impaired acti-vation of IRS-1, IRS-2, and AKT in PBMC of renal transplantrecipients who are challenged in vivo with insulin. This, on the onehand, may favor systemic insulin resistance and deteriorate glu-cose metabolism, whereas, on the other hand, may contribute tothe antiproliferative effects of rapamycin (Figure 6).

Figure 6. Schematic diagram illustrating the key role of the mTOR pathway in the regulation of cell metabolism and proliferationand the proposed interference of rapamycin (RAPA) on the activation of IRS-1–AKT–mTOR–P70 S6 kinase (S6K) cascade inresponse to insulin/growth factors. (Left) Normal signaling. Ligand binding to specific cell surface receptors activates IRS-1–dependent phosphatidylinositol 3-kinase (PI3-K), which, via increases in PI-3,4,5-triphosphate, leads to the activation of AKT. Thephosphorylation of IRS-1 on serine residues plays a dual role to either enhance or terminate insulin effects. Then the rictor–mTORcomplex directly phosphorylates AKT on serine 473. As a final result, insulin/growth factor signaling enhances cell growth andproliferation, inhibits cell apoptosis, and regulates metabolism. (Right) Long-term exposure to RAPA, by inhibiting the mTOR–raptor complex, blocks P70 S6K activation, and this, in turn, decreases inhibitory serine phosphorylation of IRS-1, favors basaltyrosine phosphorylation, and prevents IRS-1 degradation. Then, it may partially inhibit rictor-mTOR activity, thereby causing adecreased activation of AKT. Finally, the inhibition of the positive feedback loop impinging on IRS-1 serine 307 phosphorylationwould contribute to depress further the insulin-induced activation of AKT. Collectively, long-term treatment with RAPA inhibitscell growth and proliferation, favors cell apoptosis, and interferes with insulin-regulated metabolism. The arrows indicateactivation; bars indicate inhibition. G�L, G protein � subunit-like protein; p-S, phospho-serine; p-Y, phospho-tyrosine.

2242 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 2236–2244, 2006

References1. Hay N, Sonenberg N: Upstream and downstream of

mTOR. Genes Dev 18: 1926–1945, 20042. Harrington LS, Findlay GM, Gray A, Tolkacheva T, Wig-

field S, Rebholz H, Barnett J, Leslie NR, Cheng S, ShepherdPR, Gout I, Downes CP, Lamb RF: The TSC1-2 tumorsuppressor controls insulin-PI3K signaling via regulationof IRS proteins. J Cell Biol 166: 213–223, 2004

3. Manning BD, Cantley LC: Rheb fills a GAP between TSCand TOR. Trends Biochem Sci 28: 573–576, 2003

4. Harrington LS, Findlay GM, Lamb RF: Restraining PI3K:mTOR signalling goes back to the membrane. Trends Bio-chem Sci 30: 35–42, 2005

5. Shah OJ, Wang Z, Hunter T: Inappropriate activation of theTSC1-TSC2/Rheb/mTOR/S6K cassette induces IRS1/2depletion, insulin resistance, and cell survival deficiencies.Curr Biol 14: 1650–1656, 2004

6. Berg CE, Lavan BE, Rondinone CM: Rapamycin partiallyprevents insulin resistance induced by chronic insulintreatment. Biochem Biophys Res Commun 293: 1021–1027,2002

7. Shi Y, Yan H, Frost P, Gera J, Lichtenstein A: Mammaliantarget of rapamycin inhibitors activate the AKT kinase inmultiple myeloma cells by up-regulating the insulin-likegrowth factor receptor/insulin receptor substrate-1/phos-phatidylinositol 3-kinase cascade. Mol Cancer Ther 4: 1533–1540, 2005

8. Tremblay F, Gagnon A, Veilleux A, Sorisky A, Marette A:Activation of the mammalian target of rapamycin pathwayacutely inhibits insulin signaling to Akt and glucose trans-port in 3T3–L1 and human adipocytes. Endocrinology 146:1328–1337, 2005

9. Edinger AL, Linardic CM, Chiang GG, Thompson CB,Abraham RT: Differential effects of rapamycin on mamma-lian target of rapamycin signaling functions in mammaliancells. Cancer Res 63: 8451–8460, 2003

10. Manning BD: Balancing Akt with S6K: Implications forboth metabolic diseases and tumorigenesis. J Cell Biol 167:399–403, 2004

11. Fisher TL, White MF: Signaling pathways: The benefits ofgood communication. Curr Biol 14: R1005–R1007, 2004

12. Manning BD, Logsdon MN, Lipovsky AI, Abbott D, Kwi-atkowski DJ, Cantley LC: Feedback inhibition of Akt sig-naling limits the growth of tumors lacking Tsc2. Genes Dev19: 1773–1778, 2005

13. Ma L, Teruya-Feldstein J, Behrendt N, Chen Z, Noda T,Hino O, Cordon-Cardo C, Pandolfi PP: Genetic analysis ofPten and Tsc2 functional interactions in the mouse revealsasymmetrical haploinsufficiency in tumor suppression.Genes Dev 19: 1779–1786, 2005

14. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM: Phos-phorylation and regulation of Akt/PKB by the Rictor-mTOR complex. Science 307: 1098–1101, 2005

15. Danielsson A, Ost A, Nystrom FH, Stralfors P: Attenuationof insulin-stimulated insulin receptor substrate-1 serine307 phosphorylation in insulin resistance of type 2 diabe-tes. J Biol Chem 280: 34389–34392, 2005

16. Atkins MB, Hidalgo M, Stadler WM, Logan TF, Dutcher JP,Hudes GR, Park Y, Liou SH, Marshall B, Boni JP, Dukart G,Sherman ML: Randomized phase II study of multiple doselevels of CCI-779, a novel mammalian target of rapamycin

kinase inhibitor, in patients with advanced refractory renalcell carcinoma. J Clin Oncol 22: 909–918, 2004

17. Guertin DA, Sabatini DM: An expanding role for mTOR incancer. Trends Mol Med 11: 353–361, 2005

18. Vignot S, Faivre S, Aguirre D, Raymond E: mTOR-targetedtherapy of cancer with rapamycin derivatives. Ann Oncol16: 525–537, 2005

19. Chan S, Scheulen ME, Johnston S, Mross K, Cardoso F,Dittrich C, Eiermann W, Hess D, Morant R, Semiglazov V,Borner M, Salzberg M, Ostapenko V, Illiger HJ, BehringerD, Bardy-Bouxin N, Boni J, Kong S, Cincotta M, Moore L:Phase II study of temsirolimus (CCI-779), a novel inhibitorof mTOR, in heavily pretreated patients with locally ad-vanced or metastatic breast cancer. J Clin Oncol 23: 5314–5322, 2005

20. Hricik DE, Anton HA, Knauss TC, Rodriguez V, SeamanD, Siegel C, Valente J, Schulak JA: Outcomes of AfricanAmerican kidney transplant recipients treated with siroli-mus, tacrolimus, and corticosteroids. Transplantation 74:189–193, 2002

21. Ciancio G, Burke GW, Gaynor JJ, Mattiazzi A, Roth D,Kupin W, Nicolas M, Ruiz P, Rosen A, Miller J: A random-ized long-term trial of tacrolimus/sirolimus versus tacroli-mus/mycophenolate mofetil versus cyclosporine (NE-ORAL)/sirolimus in renal transplantation. II. Survival,function, and protocol compliance at 1 year. Transplantation77: 252–258, 2004

22. Sulanc E, Lane JT, Puumala SE, Groggel GC, Wrenshall LE,Stevens RB: New-onset diabetes after kidney transplanta-tion: An application of 2003 International Guidelines.Transplantation 80: 945–952, 2005

23. Stallone G, Schena A, Infante B, Di Paolo S, Loverre A,Maggio G, Ranieri E, Gesualdo L, Schena FP, GrandalianoG: Sirolimus for Kaposi’s sarcoma in renal-transplant re-cipients. N Engl J Med 352: 1317–1323, 2005

24. Sodhi A, Montaner S, Patel V, Gomez-Roman JJ, Li Y,Sausville EA, Sawai ET, Gutkind JS: Akt plays a centralrole in sarcomagenesis induced by Kaposi’s sarcoma her-pesvirus-encoded G protein-coupled receptor. Proc NatlAcad Sci U S A 101: 4821–4826, 2004

25. Teutonico A, Schena FP, Di Paolo S: Glucose metabolism inrenal transplant recipients. The effect of calcineurin inhib-itors withdrawal and conversion to Sirolimus. J Am SocNephrol 16: 3128–3135, 2005

26. Cattaneo D, Perico N, Gaspari F: Assessment of sirolimusconcentrations in whole blood by high-performance liquidchromatography with ultraviolet detection. J Chromatogr BAnalyt Technol Biomed Life Sci 774: 187–194, 2002

27. Loverre A, Ditonno P, Crovace A, Gesualdo L, Ranieri E,Pontrelli P, Stallone G, Infante B, Schena A, Di Paolo S,Capobianco C, Ursi M, Palazzo S, Battaglia M, Selvaggi FP,Schena FP, Grandaliano G: Ischemia-reperfusion inducesglomerular and tubular activation of proinflammatory andantiapoptotic pathways: Differential modulation by rapa-mycin. J Am Soc Nephrol 15: 2675–2686, 2004

28. Dib K, Whitehead JP, Humphreys PJ, Soos MA, Baynes KC,Kumar S, Harvey T, O’Rahilly S: Impaired activation ofphosphoinositide 3-kinase by insulin in fibroblasts frompatients with severe insulin resistance and pseudoacro-megaly. A disorder characterized by selective postreceptorinsulin resistance. J Clin Invest 101: 1111–1120, 1998

29. Morino K, Petersen KF, Dufour S, Befroy D, Frattini J,

J Am Soc Nephrol 17: 2236–2244, 2006 mTOR Signaling and Downregulation of IRS-1/2 and AKT 2243

Shatzkes N, Neschen S, White MF, Bilz S, Sono S, PypaertM, Shulman GI: Reduced mitochondrial density and in-creased IRS-1 serine phosphorylation in muscle of insulin-resistant offspring of type 2 diabetic parents. J Clin Invest115: 3587–3593, 2005

30. Gan SK, Kriketos AD, Poynten AM, Furler SM, ThompsonCH, Kraegen EW, Campbell LV, Chisholm DJ: Insulin ac-tion, regional fat, and myocyte lipid: Altered relationshipswith increased adiposity. Obes Res 11: 1295–1305, 2003

31. Bodine SC, Stitt TN, Gonzalez M, Kline WO, Stover GL,Bauerlein R, Zlotchenko E, Scrimgeour A, Lawrence JC,Glass DJ, Yancopoulos GD: Akt/mTOR pathway is a cru-cial regulator of skeletal muscle hypertrophy and can pre-vent muscle atrophy in vivo. Nat Cell Biol 3: 1014–1019,2001

32. Lungu AO, Jin ZG, Yamawaki H, Tanimoto T, Wong C,Berk BC: Cyclosporin A inhibits flow-mediated activationof endothelial nitric-oxide synthase by altering cholesterolcontent in caveolae. J Biol Chem 279: 48794–48800, 2004

33. Greene MW, Garofalo RS: Positive and negative regulatoryrole of insulin receptor substrate 1 and 2 (IRS-1 and IRS-2)serine/threonine phosphorylation. Biochemistry 41: 7082–7091, 2002

34. Obenauer JC, Cantley LC, Yaffe MB: Scansite 2.0: Pro-teome-wide prediction of cell signaling interactions usingshort sequence motifs. Nucleic Acids Res 31: 3635–3641, 2003

35. Cusi K, Maezono K, Osman A, Pendergrass M, Patti ME,Pratipanawatr T, DeFronzo RA, Kahn CR, Mandarino LJ:Insulin resistance differentially affects the PI 3-kinase- andMAP kinase-mediated signaling in human muscle. J ClinInvest 105: 311–320, 2000

36. Storgaard H, Song XM, Jensen CB, Madsbad S, BjornholmM, Vaag A, Zierath JR: Insulin signal transduction in skel-etal muscle from glucose-intolerant relatives of type 2 di-abetic patients. Diabetes 50: 2770–2778, 2001

37. Whiteman EL, Cho H, Birnbaum MJ: Role of Akt/proteinkinase B in metabolism. Trends Endocrinol Metab 13: 444–451, 2002

38. Sarbassov DD, Ali SM, Kim DH, Guertin DA, Latek RR,Erdjument-Bromage H, Tempst P, Sabatini DM: Rictor, anovel binding partner of mTOR, defines a rapamycin-insensitive and raptor-independent pathway that regulatesthe cytoskeleton. Curr Biol 14: 1296–1302, 2004

39. Um SH, Frigerio F, Watanabe M, Picard F, Joaquin M,Sticker M, Fumagalli S, Allegrini PR, Kozma SC, Auwerx J,

Thomas G: Absence of S6K1 protects against age- anddiet-induced obesity while enhancing insulin sensitivity.Nature 431: 200–205, 2004

40. Pende M, Kozma SC, Jaquet M, Oorschot V, Burcelin R, LeMarchand-Brustel Y, Klumperman J, Thorens B, ThomasG: Hypoinsulinaemia, glucose intolerance and diminishedbeta-cell size in S6K1-deficient mice. Nature 408: 994–997,2000

41. Bernal-Mizrachi E, Fatrai S, Johnson JD, Ohsugi M, OtaniK, Han Z, Polonsky KS, Permutt MA: Defective insulinsecretion and increased susceptibility to experimental dia-betes are induced by reduced Akt activity in pancreaticislet beta cells. J Clin Invest 114: 928–936, 2004

42. Rhodes CJ: Type 2 diabetes: A matter of beta-cell life anddeath? Science 307: 380–384, 2005

43. Chang Q, Li Y, White MF, Fletcher JA, Xiao S: Constitutiveactivation of insulin receptor substrate 1 is a frequent eventin human tumors: Therapeutic implications. Cancer Res 62:6035–6038, 2002

44. Nagle JA, Ma Z, Byrne MA, White MF, Shaw LM: Involve-ment of insulin receptor substrate 2 in mammary tumormetastasis. Mol Cell Biol 24: 9726–9735, 2004

45. Luo Y, Shoemaker AR, Liu X, Woods KW, Thomas SA, deJong R, Han EK, Li T, Stoll VS, Powlas JA, Oleksijew A,Mitten MJ, Shi Y, Guan R, McGonigal TP, Klinghofer V,Johnson EF, Leverson JD, Bouska JJ, Mamo M, Smith RA,Gramling-Evans EE, Zinker BA, Mika AK, Nguyen PT,Oltersdorf T, Rosenberg SH, Li Q, Giranda VL: Potent andselective inhibitors of Akt kinases slow the progress oftumors in vivo. Mol Cancer Ther 4: 977–986, 2005

46. Goldfine AB, Simonson DC, Folli F, Patti ME, Kahn CR:Metabolic effects of sodium metavanadate in humans withinsulin-dependent and noninsulin-dependent diabetesmellitus in vivo and in vitro studies. J Clin Endocrinol Metab80: 3311–3320, 1995

47. Renard E, Grigorescu F, Lavabre C, Kahn CR: Insulin-dependent phosphatidylinositol 3�-kinase activity co-pre-cipitates with insulin receptor in human circulating mono-nuclear cells. Biochem Biophys Res Commun 209: 234–241,1995

48. Peralba JM, DeGraffenried L, Friedrichs W, Fulcher L,Grunwald V, Weiss G, Hidalgo M: Pharmacodynamicevaluation of CCI-779, an inhibitor of mTOR, in cancerpatients. Clin Cancer Res 9: 2887–2892, 2003

2244 Journal of the American Society of Nephrology J Am Soc Nephrol 17: 2236–2244, 2006