Page 1 of 31 Diabetes
Activation of aldose reductase by interaction with tubulin and involvement of this mechanism in diabetic cataract formation
Juan F. Rivellia, Verónica S. Santandera, Sofía O. Perettia, Noelia E. Monesteroloa, Ayelen D. Nigraa, Gabriela Previtalia, Marina R. Amaidena, Carlos A. Arceb, Emiliano Primoa, Angela T. Lisaa, Juan Piec and César H. Casalea
a Departamento de Biología Molecular, Facultad de Ciencias Exactas, Físico Químicas y Naturales, Universidad Nacional de Río Cuarto, Río Cuarto, 5800 Córdoba, Argentina. b Centro de Investigaciones en Química Biológica de Córdoba (CIQUIBIC), UNC CONICET, Departamento de Química Biológica, Facultad de Ciencias Químicas, Universidad Nacional de Córdoba, Ciudad Universitaria, 5000 Córdoba, Argentina. c Departments of Pharmacology Physiology and Pediatrics, Medical School, University of Zaragoza, Zaragoza, Spain.
Corresponding author: César H. Casale. Departamento de Biología Molecular, Facultad de Ciencias Exactas, Físico Químicas y Naturales,ARTICLE Universidad Nacional de Río Cuarto, Río Cuarto, 5800 Córdoba, Argentina. Tel.: +54 358 4676422; fax: +54 358 4676232; E mail: [email protected]
ABSTRACT Our previous studies have shown that high levels of glucose induce inhibition of RETRACTEDNa+,K+ ATPase (NKA) via stimulation of aldose reductase (AR), polymerisation of microtubules, and formation of an acetylated tubulin/NKA complex. Inhibition of AR eliminated the effect of high glucose on NKA activity. In this study, we investigated the mechanism of regulation of AR activity by tubulin. Purified tubulin and AR were used. The results indicate that: (i) tubulin and AR interact with each other directly; (ii) tubulin/AR interaction results in a 6 fold increase of AR activity under microtubule growing conditions; (iii) AR interacts preferentially with tubulin that contains 3 nitro L tyrosine (3 NTyr); (iv) free tyrosine and 3 nitro tyrsine are able to block tubulin/AR interaction and thereby prevent AR activation; (v) exposure of cultured COS cells to high glucose concentrations promotes microtubule polymerisation and NKA inhibition, and both these promoting effects are inhibited by addition of free Tyr or 3 NTyr; (vi)
1 For Peer Review Only Diabetes Page 2 of 31
treatment of experimental (STZ induced) diabetic rats with 3 NTyr prevented cataract formation, suggesting that this complication of diabetes involves tubulin/AR interaction and AR activity. Taken together, these findings indicate that AR activity is controlled by association/dissociation of the tubulin/AR complex, that Tyr and 3 NTyr block such effect by preventing tubulin/AR complex formation, and that AR activity can be reduced by 3 NTyr or other compounds that inhibit tubulin/AR interaction.
INTRODUCTION The major pathogenic pathway in diabetes whereby hyperglycemia causes damage in tissues with insulin independent glucose transport is activation of the enzyme aldose reductase (AR) (1). AR activation is associated with various complications of diabetes (e.g., cataract formation, retinopathy) (2). In spite of over five decades of research, no AR inhibitor has been found that combines selectivity, efficacy, and safety for human therapeutic application (3). We demonstrated previously that acetylated tubulin (AcTub) is capable of associating with Na+,K+ ATPase (NKA) to form a complex, and that such association results in inhibition of NKA enzyme activity (4, 5, 6, 7, 8,ARTICLE 9). We demonstrated recently that high glucose concentrations induce polymerisation of microtubules, additional formation of AcTub/NKA complex, and inhibition of enzyme activity. The increase in microtubule content appeared to result from increased levels of sorbitol caused by AR activation. AR activity increased when AR was associated with microtubules, resulting in an "upregulation cycle" between microtubule formation and AR activation. On the RETRACTEDbasis of these findings, we proposed that glucose triggers a synergistic effect between tubulin and sorbitol that leads to AR activation (by association with tubulin), microtubule polymerisation, and consequent NKA inhibition (10). The objective of the present study was to elucidate the mechanism whereby tubulin/AR interaction regulates AR activity. We present evidence, using a diabetic rat model, that AR interacts directly with tubulin (primarily the 3 NTyr tub isotype) and that consequent AR activation occurs through tubulin polymerisation. Free Tyr or 3 NTyr inhibited tubulin/AR complex formation, AR activation, and various pathological events (including cataract formation) induced by high glucose concentrations. We propose a novel mechanism of NKA and AR regulation by tubulin under high glucose
2 For Peer Review Only Page 3 of 31 Diabetes
conditions, and demonstrate that AR activity can be regulated by drugs that inhibit tubulin/AR association.
RESEARCH DESIGN AND METHODS
Cell culture and treatment with glucose, Tyr, or 3-NTyr COS cells were grown in DMEM at 37 °C in a water saturated atmosphere of
air/CO2 (19:1). For treatment with glucose, Tyr, or 3 NTyr, cells cultured to 90% confluence were rinsed with HEPES FBS buffer (25 mM HEPES, pH 7.4, supplemented with 1 mM sodium pyruvate, 0.22% sodium carbonate, 10 mM glutamine, 100 mM NaCl, 10% FBS, 10 IU/ml penicillin, and 100 g/ml streptomycin) and incubated for 2 h.
Animals and in vivo experiments Induction of diabetes in rats by streptozotocin (STZ). Diabetes was induced in a rat model as described by Lin et al. (11). In brief, 7 wk old Wistar rats with ad libitum access to food were injected i.p. with a single dose of STZARTICLE (70 mg/kg body wt, in 0.1 mM citrate buffer, pH 4.5). In successful cases, the blood glucose concentration 7 8 days after injection was 300 400 mg/dl. Control animals were injected with an equal volume of citrate buffer. Tyr and 3-NTyr treatment. To investigate the effects of Tyr and 3 NTyr on diabetic cataract formation, rats were assigned randomly to one of the following groups: RETRACTEDcontrol, diabetic, diabetic treated with 3 NTyr (30 mg . kg 1. day 1), and diabetic treated with Tyr (90 mg . kg 1. day 1). Tyr and 3 NTyr were mixed with the water supply. Treatments were started on day 7 after STZ injection and continued until day 90. Rats were sacrificed, and blood was obtained by cardiac puncture for HbA1c assay. Eyes were examined with an ophthalmoscope once per wk. Pupils were dilated by a drop of a 1:1 mixture of 1% tropicamide and 10% phenylephrine hydrochloride. The degree of cataract maturity was classified as grade 0, clear; grade 1, peripheral vesicles and opacities; grade 2, central opacities; grade 3, diffused opacities; grade 4, mature cataract; grade 5, hypermature cataract.
Rat lens culture and analysis of lens opacity
3 For Peer Review Only Diabetes Page 4 of 31
For ex vivo examination of lens opacity, lenses were dissected from 6 wk old male rats. Each isolated lens was incubated in 2 ml of 199 medium with antibiotics and
20 mM glucose in 24 well plates under 95% air/5% CO2 atmosphere for 2 days at 37 °C. The medium was changed every day and supplemented with Tyr or 3 NTyr (500 M) in addition to 20 mM glucose. All the reagents used in lens culture were filtered (pore diameter 0.2 m). Lenses were examined for the development of generalized opacity under an optical microscope with a CCD camera. The opaque area intensities of the lens were quantified using the Scion imaging software programme (12). To determine AR activity, 20 lenses from each treatment were homogenized in 4 ml Tris HCl buffer (20 mM Tris HCl, pH 7) using a glass teflon homogenizer. The homogenate was centrifuged at 10,000 x g for 20 min at 4 °C. The resulting supernatant was used as an enzyme source (13).
Isolation of membrane fraction from COS cells Confluent cells from four 150 cm2 flasks were washed once with TBS PMSF (50 mM Tris HCl, pH 7.4, containing 150 mM NaCl and 0.1 mM PMSF), harvested in TBS PMSF, centrifuged at 2,000 x g for 10 min at 4 ºC, resuspended in hypotonic
solution (10 mM Tris HCl, pH 7.4, 1 mM MgCl2, 0.1 mMARTICLE PMSF), and stirred for 30 min on ice. The suspension was homogenized in a glass Dounce homogenizer (40 strokes), the homogenate was centrifuged at 100,000 x g for 30 min at 4 ºC, and the resulting pellet was resuspended in 3 ml TBS PMSF.
Isolation of cytosolic and cytoskeletal tubulin RETRACTED COS cells were washed at room temperature with microtubule stabilising buffer (90 mM Mes, pH 6.7, 1 mM EGTA, 1 mM MgCl2, 10% (v/v) glycerol) and extracted with 2.5 ml of the same buffer containing 10 M taxol, 0.5% (v/v) Triton X 100, and protease inhibitors (10 g/ml aprotinin, 0.5 mM benzamidine, 5 g/ml o phenanthroline, 0.2 mM PMSF) for 3 min at 37 ºC with gentle agitation. The detergent extract (cytosolic tubulin fraction including membrane fraction) was separated, and the diluted proteins were concentrated by the chloroform/methanol method of Wessel and Flügge (14). The cytoskeletal tubulin fraction (which remained bound to the dish) was washed with pre warmed microtubule stabilising buffer and resuspended in sample buffer (15). The cytosolic fraction was resuspended in an identical volume of sample buffer. Both fractions were immediately subjected to immunoblotting. 4 For Peer Review Only Page 5 of 31 Diabetes
Tubulin purification and incorporation of 3-NTyr Brains from 30 60 day old rats were homogenized at 4 ºC in one volume of MEM
buffer (0.1 M Mes/NaOH, pH 6.7, containing 1 mM EGTA and 1 mM MgCl2). The homogenate was centrifuged at 100,000 x g for 45 min, and the pellet was discarded. Tubulin was purified as described by Sloboda et al. (16). The concentration was adjusted to 1 mg/ml, and the tubulin was used immediately. In vitro incorporation of 3 NTyr into tubulin was performed as described by Bisig et al. (17).
Recombinant AR production AKR1B1 cDNA was isolated by RT PCR, starting from human adrenal total RNAs. cDNAs were inserted into the Nde1 and EcoRI sites of the PET 28a vector (Novagen, Tebu, Le Perray en Yvelines, France) to produce N terminal fusions with six histidine residues. This construct was kindly provided by Dr. Anne Marie Lefrancois Martinez, and recombinant AKR1B1 was expressed and purified as described by her (18). ARTICLE Interaction of tubulin with AR Recombinant AR was linked to Ni+2 NTA beads (Novagen) as described by Ganesan et al. (19). AR linked beads (0.2 ml) were washed with 1 ml of 10 mM Tris HCl (pH 7.6) buffer containing 20 mM NaCl, and incubated with various concentrations of tubulin in a total volume of 400 l for 30 min at 20 °C. The samples were RETRACTEDcentrifuged, and the precipitated materials were washed five times with the same buffer. Fractions (50 l) of packed beads were resuspended in 50 l Laemmli sample buffer, heated at 50 °C for 15 min, and centrifuged. Aliquots (20 l) of the soluble fractions were subjected to SDS PAGE.
Size exclusion chromatography A protein preparation (200 g of tubulin and 12 g of AR in a total volume of 50 l) solubilized in Tris HCl buffer (20 mM Tris HCl, pH 7) was subjected to molecular exclusion chromatography on an SE FPLC system (ISCO Inc., Lincoln, NE, USA), using a BioSep SecR S3000 column (300 7.8 mm; Phenomenex, Torrance, CA, USA) pre equilibrated with Tris HCl buffer at a flow rate of 1.5 ml/min. Running conditions:
5 For Peer Review Only Diabetes Page 6 of 31
sample volume, 50 l; flow rate, 1 ml/min; fraction volume, 1 ml; room temperature. Molecular weight standards (Sigma) were separated under identical conditions. Tubulin was detected by western blotting, and AR was detected by enzyme activity assay in the collected fractions.
AR assay Overexpressed and purified AR was assayed spectrophotometrically by the method of Tabakoff and Erwin (20). The reaction mixture for this assay contained 7 g protein, 0.1 mM NADPH, 1 mM D,L glyceraldehyde, and 0.1 M sodium phosphate, pH 7.0, in a total volume of 3 ml. The reaction was initiated by aldehyde addition. The rate of NADPH oxidation was assessed as the decrease in absorbance at 340 nm during 20 min at 25 °C.
NKA activity assay NKA enzyme activity of COS cells was determined by the method of Salvador and Mata (21). COS cell membranes (5 10 g protein) were added to the reaction
mixture (50 mM Tris HCl, pH 7.4, 20 mM KCl, 100 mM NaCl, 2.5 mM MgCl2, 0.5 mM EGTA, 0.16 mM NADH, 1 mM phosphoenolpyruvate,ARTICLE 2.5 IU pyruvate kinase, 2.5 IU lactate dehydrogenase) in a final volume of 340 l. The mixture was kept for 10 min at room temperature, and the reaction was initiated by addition of 1 mM ATP. NADH oxidation was measured for 15 min at room temperature using a recording spectrophotometer at wavelength 340 nm. Control cuvettes were prepared without enzyme or with heat denatured enzyme. Enzyme activity was estimated as the RETRACTEDdifference between samples incubated in the absence vs. presence of 1 mM ouabain.
Gel electrophoresis, immunoblotting, and protein determination Proteins were separated by SDS PAGE on 15% polyacrylamide slab gels as described by Laemmli (15) and transferred to nitrocellulose sheets. Blots were reacted with various antibodies as indicated. Sheets were reacted with the corresponding anti IgG conjugated with peroxidase and stained by the 4 chloro 1 naphthol method. Band intensities were quantified using the Scion imaging programme. Protein concentration was determined by the method of Bradford (22).
Immunofluorescence staining 6 For Peer Review Only Page 7 of 31 Diabetes
Cultured cells were fixed on coverslips with anhydrous methanol at 20 °C, washed, incubated with BSA in PBS buffer, and stained by indirect immunofluorescence as described by DeWitt et al. (23).
Statistical analysis Results were expressed as mean ± SD. Student’s t test was used for comparison of two populations. Analysis of variance (ANOVA) was used for comparison when sample sizes differed between groups. Differences between means were considered statistically significant for p values ≤ 0.05.
RESULTS
Tubulin-AR interaction: effect on enzyme activity We demonstrated previously that tubulin and AR can be associated to form a complex (10). To confirm that such complex formation results from direct interaction between the two proteins, we overexpressed and purified recombinant AR containing a polyhistidine tail (AKR1B1 human AR gene) and purifiedARTICLE rat brain tubulin (Fig. 1A, left panel). The absence of cross contamination was confirmed by western blotting (Fig. 1A, middle and right panels). Recombinant AR was linked to Ni+2 NTA beads and incubated with tubulin, and the precipitated material was analysed by western blotting to confirm the presence of tubulin. Recombinant AR linked to Ni+2 NTA beads was able to precipitate tubulin (Fig. 1B, lane AR Tub). Two controls were used in this RETRACTEDexperiment: (i) Ni+2 NTA beads without AR (Fig. 1B, lane Ctub) incubated with tubulin, and (ii) non specific protein (NSP; exopolyphosphatase of Pseudomonas aeruginosa). No tubulin precipitation was observed for either control, indicating that tubulin interacts with AR in a specific manner. If tubulin interacts directly with AR, such interaction would presumably affect AR enzyme activity, as we observed in living cells and preparations of rat brain AR (10). In the present study, the enzyme activity of recombinant AR was increased ~6 fold by polymerised purified tubulin but was unaffected by unpolymerised tubulin (Fig. 1C). AR enzyme activity was unaffected by treatment with nocodazol or taxol (data not shown). Similar experiments were performed using D glucose (150 mM) as a physiological substrate of AR. The activation of AR by growing microtubules was 7 For Peer Review Only Diabetes Page 8 of 31
similar regardless of whether enzyme activity was determined with D glucose or D,L glyceraldehyde as substrate (data not shown). These findings indicate that tubulin interacts directly with AR and that enzyme activity is enhanced by microtubule polymerising conditions. To further characterize the tubulin/AR complex, tubulin, AR, and previously formed complex were subjected to gel filtration chromatography. In each eluted fraction, tubulin was assessed by immunoblotting and AR was assessed by measurement of enzyme activity. Tubulin was eluted from the column in one peak at 4 ml, whereas AR was eluted in one peak at 6 ml (Fig. 1D). When tubulin/AR complex was analyzed, tubulin activity was detected in two eluted fractions: one at a 4 ml peak similar to that of tubulin alone and the other (with a smaller volume) at a 3 ml peak. Similarly, AR activity was detected in two peaks: the first at 3 ml, coinciding with the first tubulin peak, and the second at 6 ml, coinciding with the peak of AR alone. These results indicate that tubulin and AR are partners in the complex. A calibration curve with molecular mass standards (data not shown) was run under identical conditions.
AR activation results from binding of the enzyme to ARTICLEtubulin during microtubule growth To clarify whether AR is able to associate only with dimeric tubulin or with microtubules as well, we examined the ability of AR to interact with microtubules pre assembled in vitro. Recombinant AR linked to Ni+2 NTA beads was incubated with taxol pre assembled microtubules, and the insoluble material was collected. Tubulin RETRACTEDwas nearly absent in the precipitate, indicating th at AR did not associate with microtubules (Fig. 2A, line MT). In contrast, when recombinant AR linked to Ni+2 NTA beads was already present in the incubation system, and microtubule assembly was initiated by taxol addition, tubulin was present in the precipitate (Fig. 2A, line Tub). If AR were unable to associate with pre assembled microtubules, we would not expect to observe any effect on enzyme activity. When the enzyme was incubated in the presence of various amounts of pre assembled microtubules, there was no change in AR activity (Fig. 2B). On the other hand, if AR were able to associate with growing microtubules, we would expect to observe increased enzyme activity. In fact, enzyme activity was ~6 fold higher for incubation under microtubule growing conditions (Figs.
8 For Peer Review Only Page 9 of 31 Diabetes
1C, 2B). These findings indicate that AR can associate with microtubules under growing conditions and that such association causes enzyme activation.
Effects of 3-nitrotyrosinated tubulin, free 3-NTyr, and free Tyr on tubulin/AR complex formation and enzyme activity Changes in the redox state of cells have been shown to play an important role in regulating AR catalytic status by modifying a reactive cysteine residue, Cys298, present at the active site (24). Cys298 is sensitive to changes in the nitric oxide donor and can be nitrated. Oxidative stress in diabetes mellitus leads to “nitrosative” stress. The most common example is the addition of nitro groups to the Tyr or Cys residues of proteins (25). Peroxide nitrite (ONOO ) formed by the action of NO synthase can also nitrate free Tyr to produce free 3 NTyr. An alternative proposed mechanism whereby 3 NTyr is inserted in proteins involves the direct and specific incorporation of 3 NTyr into the C terminus of α tubulin by tubulin Tyr ligase, an enzyme that normally incorporates Tyr into that position through a post translational mechanism (26). The concentration of 3 NTyr in cells is very low under normal conditions but is often elevated under pathological conditions (27). In neural cells under high glucose conditions, the amount of nitrated Tyr in tubulin was elevated (28, 29, 30). ARTICLE In the present study, precipitated material was analyzed by immunoblotting to detect the presence of different tubulin isotypes following formation of tubulin/AR complex by incubation of recombinant AR linked beads with purified tubulin. 3 NTyr tub (detected with a specific polyclonal antibody) was enriched in comparison with the input, whereas other tubulin isotypes were not enriched (Fig. 3). This finding suggested RETRACTEDthe possibility that binding of tubulin that contains a 3 NTyr residue to the C terminal of the α chain plays a role in tubulin/AR association. To test this idea, we measured AR activity after interaction of the enzyme with two tubulin preparations with a low vs. high content of 3 NTyr. The activity promoting effect of the high 3 NTyr tub preparation was ∼4 fold higher than that of the low 3 NTyr tub preparation (Fig. 4A). Next, low and high 3 NTyr tub preparations were mixed in various proportions, added with AR, and AR activity was measured. AR activity was positively correlated with 3 NTyr tub content (Fig. 4B). These findings suggest that tubulin/AR complex formation requires binding of a 3 NTyr residue to the C terminus of the α tubulin chain. Because the 3 NTyr tub isotype is preferred for complex formation with AR, we considered the possibility that both Tyr and 3 NTyr affect tubulin/AR association and 9 For Peer Review Only Diabetes Page 10 of 31
AR enzyme activity. To test this idea, we measured tubulin/AR complex formation and AR activity in the presence of various concentrations of free 3 NTyr and Tyr. The presence of either compound reduced the stimulatory effect of tubulin on AR activity in a concentration dependent manner (Fig. 5, lower panel). A reduced stimulatory effect of tubulin on AR activity was well correlated with a reduced amount of tubulin co precipitated with AR linked to Ni+2 NTA beads (Fig. 5, upper panel). AR activity was not modified by treatment with 3 NTyr or Tyr (data not shown). When either compound (2 10 M) was added to previously formed tubulin/AR complex, there was no effect on the stimulatory effect of tubulin, AR activity, or the degree of tubulin precipitation with AR (data not shown). These findings indicate that 3 NTyr and Tyr prevent the association of tubulin with AR but are unable to dissociate previously formed complex.
Free Tyr and 3-NTyr inhibit microtubule formation and glucose-induced inhibition of NKA in COS cells We showed previously (10) that exposure of cultured COS cells to high glucose concentrations has two concomitant effects: (i) increased proportion of microtubules, and (ii) inhibition of NKA. Both of these effects were reduced by treatment with quercetin, an AR inhibitor, suggesting that AR is involvedARTICLE in the response to high glucose. Our present results indicate that AR is not activated in the presence of free Tyr or 3 NTyr in vitro because AR is unable to associate with tubulin or microtubules under these conditions. We therefore considered the possibility that Tyr and 3 NTyr are able to prevent increases in microtubule mass and NKA inhibition through an AR inhibitory effect. To test this idea, we pre incubated COS cells with Tyr or 3 NTyr, exposed them RETRACTEDto high glucose concentrations, determined tubulin content in the cytoskeleton by immunoblotting (Fig. 6A), and visualised microtubules by immunofluorescence (Fig. 6B). Microtubule mass was ~50% lower in cells treated with glucose in combination with Tyr or 3 NTyr than in cells treated with glucose alone. The immunofluorescence studies gave similar results. NKA activity and AcTub content in the membrane fraction were also determined (Fig. 6C). Treatment with high glucose alone resulted in increased membrane AcTub content and consequent inhibition of NKA activity. In contrast, pre incubation with Tyr or 3 NTyr followed by glucose treatment had no appreciable effect on membrane AcTub. NKA activity was not inhibited under these conditions; in fact, it increased ~40%.
10 For Peer Review Only Page 11 of 31 Diabetes
3-NTyr treatment prevents cataract formation in rats with induced diabetes Two well known diabetes associated pathologies are retinopathy and cataract formation. Blocking of AR activation in diabetes prevents the development of various disease complications (31, 3). However, no clinically useful AR inhibitor has been established. In view of the ability of Tyr and 3 NTyr to inhibit AR activity, we examined the ability of these compounds to prevent cataract formation in an STZ induced diabetic rat model. Diabetes was induced in Wistar rats by STZ administration, and the animals were then treated with Tyr (90 mg. kg 1. day 1) or 3 NTyr (30 mg. kg 1. day 1) (both introduced by mixing with the water supply). Treatments were started when the diabetic state of the rats was confirmed by a glucose blood test (on day 7 after STZ injection) and continued until day 90. The animals were divided into four groups: control (n=8), diabetic (DM; n=8), diabetic with Tyr treatment (+Tyr; n=8), and diabetic with 3 NTyr treatment (+3 NTyr; n=8). Characteristics of rats in the four groups after 3 months of treatment are summarised in Fig. 7A. Average blood glucose levels in STZ treated rats, either with or without Tyr or 3 NTyr treatment, were >4 fold higher than the level in control rats (data not shown), leading to higher levels of haemoglobin glycosylation (Fig. 7A). Survival of diabetic (DM) rats was 50% lower than that of control rats, whereas survival in the +Tyr and +3 NTyr ARTICLE groups was similar to that of controls. Tyr or 3 NTyr treatment had no significant effect on body weight, although increased A1C levels were characteristic of diabetic rats. Our daily inspections of diabetic and normal rats did not reveal signs or symptoms of adverse drug effects. Cataract formation was assessed by direct visualisation using an ophthalmoscope. In the DM group, cataracts began appearing 30 40 days after STZ administration (data RETRACTEDnot shown), and 100% of the animals had cataracts of grade 5 by day 90 (Fig. 7B). In the +3 NTyr group, only 12.5% of eyes had cataracts at day 80, and the cataracts observed at day 90 were of grade 1. The effects observed in the +Tyr group were intermediate even when the Tyr dosage was 3 fold higher than the 3 NTyr dosage. Cataracts in the +Tyr group began to appear at day 70, and at day 90 75% of the analysed eyes had cataracts of grade 3. Lenses in the control group remained transparent throughout the experimental period (Fig. 7B). To confirm the above findings, Tyr and 3 NTyr effects on development of high glucose induced cataracts were evaluated in cultured ex vivo lenses. For this purpose, lenses were dissected from normal rats and incubated with 20 mM glucose (360 mg/dl) to mimic a hyperglycemic state, or with 5 mM glucose (90 mg/dl) to represent the
11 For Peer Review Only Diabetes Page 12 of 31
normal range of glycemia (control). Lens opacity and AR activity were measured in both groups. High glucose treatment resulted in a ~300% increase in lens opacity. This effect was completely reversed by addition of 3 NTyr to the medium, in agreement with results of the in vivo experiments. Tyr treatment again had an intermediate effect; it produced only 15% reversion. High glucose treatment for 48 h caused an ∼150% increase of AR activity (Fig. 7C, panel Glc). Again, this effect was reversed by addition of 3 NTyr, which resulted in AR activity values similar to those of low glucose controls (Fig. 7C, panel 3 NTyr). Addition of Tyr to lenses treated with high glucose also reversed the effect on AR activity but to a lesser extent (Fig. 7C, panel Tyr). These findings indicate that both 3 NTyr and Tyr are able to penetrate lens tissues and (particularly in the case of 3 NTyr) to inhibit AR and prevent cataract formation under hyperglycemic conditions.
DISCUSSION There are a wide variety of published reports regarding the interactions of tubulin with glycolytic enzymes, many of which catalyse oxidation reduction processes. Interactions with these enzymes affect tubulin polymerisation;ARTICLE conversely, alterations of tubulin (microtubule) polymerisation affect the catalytic state of the enzymes (32, 33, 34, 35). The present study is the first to describe stimulation of AR catalytic activity via association with tubulin (or microtubules), and down regulation of the enzyme activity by dissociation of the tubulin/AR complex by free Tyr or 3 NTyr. Our findings indicate that: (i) AR interacts with tubulin to form a stable and specific complex (Fig. 1), (ii) RETRACTEDassociation of AR with tubulin and consequent AR activation occur during tubulin polymerisation (Figs. 1 and 2), (iii) 3 NTyr tub is the preferred isotype of tubulin for association with and activation of AR (Figs. 3 and 4), (iv) free Tyr and 3 NTyr inhibit association of AR with tubulin in vitro, thereby preventing AR activation (Fig. 5), (v) effects on COS cells exposed to high glucose concentrations (increased microtubule mass, increased membrane AcTub content, reduced NKA activity) were blocked to some extent by the presence of free Tyr or 3 NTyr in the incubation medium (Fig. 6). Several molecular mechanisms have been proposed to explain the damage caused by chronic hyperglycemia. The process is generally considered to include the following events: (i) accumulation of advanced glycosylation products, (ii) activation of the polyol pathway, (iii) activation of various pathways mediated by protein kinase C, and (iv) 12 For Peer Review Only Page 13 of 31 Diabetes
increased oxidative stress. These events are closely related to the generation of reactive oxygen species that produce chronic oxidative stress in diabetic patients (36, 37) and result in varying degrees of damage to various molecular structures in different tissues. Oxidative stress in diabetes mellitus in turn causes "nitrosative" stress, i.e., the covalent modification of proteins and DNA by peroxynitrite formed from interactions between reactive oxygen species and nitric oxide; the most common example is the addition of nitro groups to Tyr or Cys residues of proteins (25). Peroxide nitrite (ONOO ) formed by the action of NO synthase can also nitrate free Tyr to produce free 3 NTyr. An alternative proposed mechanism whereby 3 NTyr is inserted in proteins involves the direct and specific incorporation of 3 NTyr into the C terminus of α tubulin by tubulin Tyr ligase, an enzyme that normally incorporates Tyr into that position by a post translational mechanism (26). The concentration of 3 NTyr is very low under normal conditions, but is often elevated under pathological conditions (26). In neural cells under high glucose conditions, the amount of nitrated Tyr in tubulin was increased (28, 29, 30). We hypothesised that this phenomenon is due to an increased amount of 3 NTyr bound to the C terminus of α tubulin (3 NTyr tub). We were able to test this concept using an antibody that specifically recognises 3 NTyr tub. We found that AR forms a complex with tubulin, preferentially with the 3 NTyr tubARTICLE isotype (Fig. 3), and that the degree of enzyme activation by tubulin is correlated with the amount of polymerised tubulin in microtubules (Fig. 1) and with 3 NTyr tub content (Fig. 4). The fact that AR activity was increased by polymerised tubulin but not by soluble tubulin may explain why an excess of tubulin relative to AR was necessary to activate the enzyme; i.e., a minimum of 27 tubulin subunits is necessary to initiate microtubule RETRACTEDformation (38, 39). These findings indicate that the presence of 3 NTyr at the C terminus of α tubulin is important or essential for tubulin/AR interaction and the effect of tubulin on enzyme activity. The fact that the amount of microtubules with a high proportion of 3 NTyr tub is increased under high glucose conditions is consistent with our finding that a higher amount of AR is associated with tubulin during the polymerisation process (Figs. 1, 2, 3, and 4). On the basis of these results, we hypothesise that high glucose conditions act as a "trigger" of oxidative/nitrosative stress that generates reactive oxygen species, causing nitration of free Tyr to produce 3 NTyr which is subsequently incorporated into tubulin. The resulting 3 NTyr tub isotype is preferentially associated with AR. When this complex is polymerised into microtubules, the enzyme is activated (Fig. 8). The
13 For Peer Review Only Diabetes Page 14 of 31
increased sorbitol concentration generated by AR from excess glucose in turn enhances tubulin polymerisation (10). A positive feedback loop is thus generated between enzyme activation and microtubule polymerisation. Consequently, more AcTub is generated, and NKA activity is inhibited (Fig. 8). We demonstrated previously that this positive feedback loop is partially blocked by the AR inhibitor quercetin (10). In the present study, free 3 NTyr and Tyr also inhibited AR (by preventing its interaction with tubulin) and blocked the effects of high glucose (increase of AcTub and inhibition of NKA) (Fig. 7). The similarity of the effects of quercetin and 3 NTyr (or Tyr) supports the idea that AR activation is involved in the effects of high glucose. We also assessed NKA activity under high glucose concentration, or high glucose followed by addition of Tyr or 3 NTyr. In the +Tyr and +3 NTyr groups, membrane AcTub levels returned to values similar to that of the control group (no glucose treatment), whereas NKA activity not only returned to control value but increased ~40%. This apparent discrepancy (increased NKA activity but no reduction in membrane AcTub level) may be explained by a great (90%) decrease in detyrosinated tubulin content in the membrane. This tubulin isotype appears to be an important regulator of NKA activity (unpubl. data; MS in prep). Changes in cellular redox state play an important roleARTICLE in regulating the catalytic activity of AR (24). The high sensitivity to redox changes of the AR catalytic state has been attributed to a reactive Cys residue (Cys298) present at the active site of the enzyme. Although oxidation of Cys298 did not completely abolish enzyme activity, it did reduce the ability of the substrate and the inhibitor to bind to the active site. Cys298 is particularly sensitive to modifications by nitric oxide donors and can be nitrosylated. RETRACTEDThe high affinity of AR for nitrile groups suggests that nitric oxide may be a physiological regulator of AR (1). In view of the fact that AR is inhibited by nitric oxide but activated by 3 NTyr tub, we propose a mechanism of AR regulation that depends on the redox state of the medium and involves tubulin. According to our model: (i) under normoglycemic conditions, AR retains a nitric oxide molecule attached to its catalytic site (Cys298), resulting in an inactive state; (ii) exposure of the cell to high glucose concentration causes oxidative stress, whereby 3 NTyr tub is increasingly produced and displaces the nitro group attached to the Cys298 of AR; thus, a tubulin/AR complex is formed, and the enzyme is activated. Experiments to test this model are in progress. Both Tyr and 3 NTyr are able to prevent in vitro tubulin/AR complex formation and thus block tubulin
14 For Peer Review Only Page 15 of 31 Diabetes
induced enzyme activation (Fig. 5). This finding suggests that changes in the cellular redox state do not necessarily affect the catalytic state of AR but are necessary for 3 nitrotyrosination of tubulin. Numerous studies indicate that blocking of AR activation in diabetic conditions can prevent the development of various disease complications (31, 3). In spite of over five decades of research, no AR inhibitor has been found that combines high efficiency, AR selectivity, and safety of application over long periods in humans. An intensive search for new AR inhibitors applicable in human clinical therapy is ongoing (3). None of the previous studies showed that AR is physiologically regulated by tubulin, as is demonstrated here. The present results allow us to speculate that any drug that inhibits tubulin/AR interaction will be effective in regulating AR enzyme activity and consequently in preventing chronic complications of human diabetes. We propose a new therapeutic approach to AR regulation, with potential clinical application, based on the use not of enzyme inhibitor compounds (which have a history of failure in human studies) but rather of compounds that regulate tubulin/AR interaction. An important event in regard to the applicability of this approach in humans would be the determination of interaction domains between the two proteins. This knowledge would facilitate the design of new drugs to prevent AR activationARTICLE by tubulin under conditions of elevated glucose concentration. Based on the experimental results presented here, we propose the use of 3NT as an inhibitor of tubulin/AR association in humans. It is important to note that 3NT is a metabolite present under physiological conditions in various tissues, particularly those subject to oxidative stress. No study to date has indicated detrimental effects of 3NT per se to the body. RETRACTED In conclusion, the findings reported here provide the first clear evidence that treatment with 3 NTyr inhibits tubulin/AR association in vivo, leading to reduced AR activity and reduced cataract formation in an STZ induced diabetic rat model (Fig. 7), similar to previously described effects of other AR inhibitors (3). Our findings provide the basis for a new approach to regulating AR activity through the level of tubulin/AR complex formed; this level can be regulated by drugs that inhibit the tubulin/AR association.
ACKNOWLEDGMENTS This study was supported by grants from the FONCyT, CONICET, and Secretaría de Ciencia y Técnica of UNRC and UNC. The authors are grateful to Dr. S. Anderson
15 For Peer Review Only Diabetes Page 16 of 31
for English editing of the manuscript. The paper is dedicated to the memory of Marina Rafaela Amaiden, who will remain forever in our hearts. Juan F. Rivelli designed the principal research and analyzed data. Veronica S. Santander, Sofia O. Peretti, Emiliano Primo, Angela T. Lisa, Noelia E. Monesterolo, Ayelen D. Nigra, Gabriela Previtali, Marina R. Amaiden, and Juan Pie designed the secondary research and analyzed data. Cesar H. Casale performed the principal research, wrote the manuscript, and analyzed data. Carlos A. Arce performed the secondary research, wrote the manuscript, and analyzed data.
REFERENCES 1. Chandra D, Jackson EB, Ramana KV, Kelley R, Srivastava SK, Bhatnagar A. Nitric oxide prevents aldose reductase activation and sorbitol accumulation during diabetes. Diabetes 2002;51:3095 3101 2. Shim SH, Ryu J, Kim JS, Kang SS, Xu Y, Jung SH, Lee YS, Lee S, Shin KH. New lanostane type triterpenoids from Ganoderma applanatum. J Nat Prod 2004;67:1110 1113 3. Sun W, Oates PJ, Coutcher JB, Gerhardinger C, Lorenzi M. A selective aldose reductase inhibitor of a new structural class preventsARTICLE or reverses early retinal abnormalities in experimental diabetic retinopathy. Diabetes 2006;55:2757 2762 4. Arce CA, Casale CH, Barra HS. Submembraneous microtubule cytoskeleton: regulation of ATPases by interaction with acetylated tubulin. FEBS J 2008;275:4664 4674 5. Casale CH, Alonso AD, Barra HS. Brain plasma membrane NKA is inhibited by RETRACTEDacetylated tubulin. Mol Cel Biochem 2001;216:85 92 6. Casale CH, Previtali G, Barra HS. Involvement of acetylated tubulin in the regulation of Na+,K+ ATPase activity in cultured astrocytes. FEBS Lett 2003;534:115 118 7. Casale CH, Previtali G, Serafino JJ, Arce CA, Barra HS. Regulation of acetylated tubulin/NKA interaction by l glutamate in non neural cells: involvement of microtubules. Biochim Biophys Acta 2005;1721:185 192 8. Zampar GG, Chesta ME, Carvajal A, Chanaday NL, Díaz MN, Casale CH, Arce C. Na+, K+ ATPase associates with acetylated tubulin through its fifth cytoplasmic domain, and acts as an anchorage site for microtubules. Biochem J 2009;422:129 137
16 For Peer Review Only Page 17 of 31 Diabetes
9. Santander VS, Bisig CG, Purro SA, Casale CH, Arce CA, Barra HS. Tubulin must be acetylated in order to form a complex with membrane Na+,K+ ATPase and to inhibit its enzyme activity. Mol Cel Biochem 2006;291:167 174 10. Rivelli JF, Amaiden MR, Monesterolo NE, Previtali G, Santander VS, Fernandez A, Arce CA, Casale CH. High glucose levels induce inhibition of Na,K ATPase via stimulation of aldose reductase, formation of microtubules and formation of an acetylated tubulin/Na,K ATPase complex. Int J Biochem Cell Biol 2012;44:1203 1213 11. Lin Y, Berg AH, Iyengar P, Lam TK, Giacca A, Combs TP, Rajala MW, Du X, Rollman B, Li W, Hawkins M, Barzilai N, Rhodes CJ, Fantus IG, Brownlee M, Scherer PE. The hyperglycemia induced inflammatory response in adipocytes: the role of reactive oxygen species. J Biol Chem 2005;280:4617 4626 12. Lee J, Jang DS, Kim NH, Lee YM, Kim J, Kim JS. Galloyl glucoses from the seeds of Cornus officinalis with inhibitory activity against protein glycation, aldose reductase, and cataractogenesis ex vivo. Biol Pharm Bull 2011;34:443 446 13. Brubaker AN, DeRuiter J, Whitmer WL. J Med Chem. Synthesis and rat lens aldose reductase inhibitory activity of some benzopyran 2 ones. 1986;29:1094 1099 14. Wessel D, Flügge UI. A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal BiochemARTICLE 1984;138:141 143 15. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680 685 16. Sloboda RD, Rosenbaum JL. Purification and assay of microtubule associated proteins (MAPs). Methods Enzymol 1982;85:409 416 17. Bisig CG, Purro SA, Contín MA, Barra HS, Arce CA. Incorporation of 3 RETRACTEDnitrotyrosine into the C terminus of alpha tubulin is reversible and not detrimental to dividing cells. Eur J Biochem 2002;269:5037 5045 18. Lefrançois Martinez AM, Bertherat J, Val P, Tournaire C, Gallo Payet N, Hyndman D, Veyssière G, Bertagna X, Jean C, Martinez A. Decreased expression of cyclic adenosine monophosphate regulated aldose reductase (AKR1B1) is associated with malignancy in human sporadic adrenocortical tumors. J Clin Endocrinol Metab 2004;89:3010 3019 19. Ganesan AK, Vincent TS, Olson JC, Barbieri JT. Pseudomonas aeruginosa exoenzyme S disrupts Ras mediated signal transduction by inhibiting guanine nucleotide exchange factor catalyzed nucleotide exchange. J Biol Chem 1999;274:21823 21829
17 For Peer Review Only Diabetes Page 18 of 31
20. Tabakoff B, Erwin VG. Purification and characterization of a reduced nicotinamide adenine dinucleotide phosphate linked aldehyde reductase from brain. J Biol Chem 1970;245:3263 3268 21. Salvador JM, Mata AM. Purification of the synaptosomal plasma membrane (Ca2+Mg2+) ATPase from pig brain. Biochem J 1996;315:183 187 22. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 1976;72:248 254 23. DeWitt ND, dos Santos CF, Allen KE, Slayman CW. Phosphorylation region of the yeast plasma membrane H+ ATPase. Role in protein folding and biogenesis. J Biol Chem 1998;273:21744 21751 24. Chandra A, Srivastava S, Petrash JM, Bhatnagar A, Srivastava SK. Active site modification of aldose reductase by oxide nitric donors. Biochem Biophys Acta 1997;1341:217 222 25. Myatt L. Reactive oxygen and nitrogen species and functional adaptation of the placenta. Placenta 2010;31:S66 S69 26. Barra HS, Arce CA, Argaraña CE. Posttranslational tyrosination/detyrosination of tubulin. Mol Neurobiol 1988;2:133 153 ARTICLE 27. Eiserich JP, Estévez AG, Bamberg TV, Ye YZ, Chumley PH, Beckman JS, Freeman BA. Microtubule dysfunction by posttranslational nitrotyrosination of alpha tubulin: a nitric oxide dependent mechanism of cellular injury. Proc Natl Acad Sci USA 1999;96:6365 6370 28. Gadau S, Lepore G, Zedda M, Manca P, Chisu V, Farina V. D glucose induces RETRACTEDmicrotubular changes in C1300 neuroblastoma cell line through the incorporation of 3 nitro L tyrosine into tubulin. Arch Ital Biol 2008;146:107 117 29. Gadau S, Lepore G, Zedda M, Mura A, Farina V. Different nitrosative induced microtubular modifications and testosterone neuroprotective effects on high D glucose exposed neuroblastoma and glioma cells. Neuro Endocrinol Lett 2009;30:515 524 30. Cappelletti G, Maggioni M, Ronchi C, Macia R, Tedeschi G. Protein tyrosine nitration is associated with cold and drug resistant microtubules in neuronal like PC12 cells. Neurosci Lett 2006;401:159 164 31. Lee AY, Chung SK, Chung SS. Demonstration that polyol accumulation is responsible for diabetic cataract by the use of transgenic mice expressing the aldose reductase gene in the lens. Proc Natl Acad Sci USA 1995;92:2780 2784
18 For Peer Review Only Page 19 of 31 Diabetes
32. Volker K, Reinitz C, Knull H. Glycolytic enzymes and assembly of microtubule Networks. Comp Biochem Physiol 1995;112B:503 514 33. Cassimeris L, Caruso S, Miller E, Ton Q, Molnar C, Fong J. Fueled by microtubules: Does tubulin dimer/polymer partitioning regulate intracellular metabolism? Cytoskeleton 2012;69:133 143 34. Vértessy BG, Bánkfalvi D, Kovács J, Löw P, Lehotzky A, Ovádi J. Pyruvate Kinase as a microtubule destabilizing factor in vitro. Biochem Biophys Res Commun 1999;254:430 435 35. Marmillot P, Keith T, Srivastava D, Knull H. Effect of Tubulin on the Activity of the Muscle Isoenzyme of Lactate Dehydrogenase. Arch Biochem Biophys 1994;315:467 472 36. Maeda Y, Inoguchi T. Oxidative stress. Nihon Rinsho 2010;68:814 818 37. Coppey L, Gellett J, Davidson E, Dunlap J, Lund D and Yorek M. Effect of Antioxidant Treatment of Streptozotocin Induced Diabetic Rats on Endoneurial Blood Flow, Motor Nerve Conduction Velocity, and Vascular Reactivity of Epineurial Arterioles of the Sciatic Nerve. Diabetes 2001;50:1927 1937 38. Wade RH, Chrétien D, Job D. Characterization of microtubule protofilament numbers. How does the surface lattice accommodate? J MolARTICLE Biol 1990;212:775 786 39. Chrétien D, Wade RH. New data on the microtubule surface lattice. Biol Cell 1991;71:161 174. Erratum in: Biol Cell 1991;72:284
FIGURE LEGENDS RETRACTED Fig. 1. Tubulin interacts directly with AR and activates the enzyme under microtubule-polymerising conditions. Tubulin was purified from rat brain and recombinant AR was obtained by expression of the AKR1B1 (human AR) gene in E. coli strain BL21a as described in Research Design and Methods. (A) Aliquots (20 g protein) of tubulin (Tub) and AR were subjected to SDS PAGE, gel stained with Coomassie brilliant blue, and analyzed by immunoblotting using anti AR or anti Tub (DM1A). (B) Tubulin (200 g) was precipitated with Ni+2 NTA beads linked to AR (recombinant AKR1B1 protein; lane AR Tub), NEP (exopolyphosphatase of Pseudomona aeruginosa [gene name ppx]; lane CI) or Ni+2 NTA beads without ligand (lane Ctub). The presence of tubulin was analyzed in the sedimented material by
19 For Peer Review Only Diabetes Page 20 of 31
immunoblotting using DM1A as primary antibody. The volume of the analyzed sample was calculated to represent the same amount of protein in each case. (C) Recombinant AR (7 g) was mixed with tubulin at the indicated concentrations and incubated for 30 min at 37 ºC in TBS buffer (final volume 1 ml) in the presence of 5 M taxol (□) or 50 M nocodazole (■). The samples were processed for determination of AR activity. The values presented are mean ± SD from three independent experiments. (D) AR, rat brain tubulin, and tubulin/AR complex preparations were subjected to molecular exclusion chromatography. Fractions corresponding to various elution volumes were collected, and tubulin and AR were analyzed by immunoblotting with DM1A and AR activity measurement, respectively. Upper panel: bands indicate tubulin elution volumes. Lower panel: AR activity was quantified and plotted as a function of elution volume. The values presented are mean ± SD from three independent experiments.
Fig. 2. AR is associated with growing microtubules but not with pre-assembled microtubules. Tubulin and recombinant AR were obtained as described in Fig. 1. (A) Ni+2 NTA beads linked to AR (recombinant AKR1B1 protein; lane Tub) were incubated with pre assembled taxolated microtubules (lane MT) or with soluble tubulin plus taxol (lane Tub) as described in Research Design andARTICLE Methods. After incubation, samples were precipitated by centrifugation. The presence of tubulin in the precipitated materials was analyzed by immunoblotting with DM1A. The volume of the analysed sample was calculated to represent the same amount of protein in each case. (B) AR (7 g recombinant AKR1B1 protein) was incubated for 30 min at 37 °C in TBS buffer (final volume 1 ml) in the presence of various concentrations of tubulin. The tubulin RETRACTEDwas added in pre assembled form in taxolated microtubules (■) or in soluble form in the presence of taxol (□). The samples were processed for determination of AR activity. The values presented are mean ± SD from three independent experiments.
Fig. 3. Tubulin isotypes associated with AR. Ni+2 NTA beads linked to AR (recombinant AKR1B1 protein) were incubated with tubulin plus taxol as described in Research Design and Methods. The samples were precipitated by centrifugation. Tubulin preparation (Input) and precipitated materials (Precipitate) were analysed by immunoblotting for determination of total tubulin (α tubulin) and various tubulin isotypes (acetylated, Ac ; tyrosinated, Tyr ; detyrosinated, Glu ; 3 nitrotyrosinated, 3 NTyr). Bands were quantified by densitometry using Scion Image software. Input ratio 20 For Peer Review Only Page 21 of 31 Diabetes
= ratio between tubulin isotype and α tubulin in preparation. Precipitate ratio = ratio between tubulin isotype and α tubulin in precipitate. R = ratio between precipitate ratio and input ratio. The values presented are mean ± SD from three independent experiments.
Fig. 4. Stimulation of AR activity by tubulin is dependent on the proportion of 3- NTyr-tub isotype in the tubulin preparation. AR (7 g recombinant AKR1B1 protein) was incubated for 30 min at 37 °C in TBS buffer (final volume 1 ml) with various concentrations of tubulin. AR enzyme activity was determined as described in Research Design and Methods. (A) Various concentrations of tubulin preparations with low (□; L) and high (■; H) 3 NTyr tub isotype content were used. Insert: identical amounts of tubulin from the two preparations were analysed by immunoblotting for total and 3 NTyr tub. (B) Two tubulin preparations as above were mixed at the indicated percentages of 3 NTyr tub (% of H). A fixed amount of tubulin and various amounts of 3 NTyr tub were mixed with AR and processed as in (A). Insert as in (A). For both panels, the values presented are mean ± SD from three independent experiments, and the bands shown are from a representative experiment. ARTICLE Fig. 5. Tyr and 3-NTyr prevent the association of AR with tubulin and the stimulation of AR activity by tubulin. (A) Purified rat brain tubulin (200 g) was mixed with Ni+2 NTA beads linked to recombinant AR in the presence of Tyr or 3 NTyr at the indicate concentrations. Precipitated materials were analyzed by immunoblotting with DM1A. The volume of the analyzed sample was calculated to represent the same RETRACTEDamount of protein in each case. (B) AR (7 g) was i ncubated for 30 min at 37 °C in TBS buffer (final volume 1 ml) in the presence of the indicated concentrations of Tyr (■) or 3 NTyr (□) and 5 M taxol. The samples were processed for determination of AR activity. The values presented are mean ± SD from three independent experiments.
Fig. 6. Effects of Tyr and 3-NTyr on microtubule content, membrane tubulin, and NKA activity. COS cells were incubated for 2 h in glucose free HEPES FBS buffer at 37 °C (control; Ctrl) and then for 1 h in the presence (+) or absence ( ) of 0.5 mM Tyr or 3 NTyr (+3 NTyr). Glucose (Glc; final concentration 25 mM) was added, and the cells were incubated for another 2 h. (A) The cytoskeletal fraction was obtained as
21 For Peer Review Only Diabetes Page 22 of 31
described in Research Design and Methods, and α tubulin was quantified by immunoblotting with DM1A. Bands were quantified by densitometry using Scion Image software. The values presented are mean ± SD from three independent experiments, expressed as a percentage of the control value. (B) Cells were fixed on coverslips and subjected to immunofluorescence with DM1A. (C) The membrane fraction was obtained as described in Research Design and Methods, and an aliquot (10 g protein) was processed for measurement of NKA activity by the NADH oxidation method (lower panel, white bars). A second aliquot was processed for estimation of the amount of AcTub by immunoblotting with mAb 6 11B 1. Upper panel: regions containing relevant bands of the blot from a representative experiment. Bands were quantified using Scion Image software. The values presented are mean ± SD from three independent experiments, expressed as a percentage of the control value (lower panel, black bars). The differences between values from different treatments were statistically significant (p< 0.01).
Fig. 7. Effects of Tyr and 3-NTyr on cataract formation in rats with STZ-induced diabetes in vivo and in rat lenses exposed to high glucose conditions ex vivo. 30 day old rats (n=24) were i.p. injected with 70 mg STZ per kgARTICLE body weight. As a control, 8 rats were injected with vehicle (citrate buffer). After 1 wk, one group (n=8) from the STZ treated rats was orally administered Tyr (90 mg . kg 1. day 1) for 3 months, a second group (n=8) was administered 3 NTyr (30 mg . kg 1. day 1), and a third group (n=8) was administered vehicle (water) only. Tyr and 3 NTyr were mixed with the water supply. (A) Characteristics of the four experimental groups at the end of the 3 RETRACTEDmonth period. HbA1c: glycosylation of haemoglobin e xpressed as % (mmol/mol). The values presented are mean ± SD from three independent experiments. (B) Representative photographs of lenses from rats in the four experimental groups at the end of the 3 month period, showing the effects of Tyr and 3 NTyr on cataract formation in vivo (magnification x4). (C) Upper panel: representative images of lenses from the four groups (magnification x25). Middle panel: lens opacity for each group was analyzed; data are expressed as Lens Diameter (LD) vs. Opacity (Arbitrary Units). Lower panel: AR activity in lenses incubated ex vivo. The values presented are mean ± SD from three independent experiments.
22 For Peer Review Only Page 23 of 31 Diabetes
Fig. 8. Proposed mechanism of regulation of NKA and AR by tubulin under high glucose conditions.
ARTICLE
RETRACTED
23 For Peer Review Only Diabetes Page 24 of 31
ARTICLE
RETRACTED Fig. 1. Tubulin interacts directly with AR and activates the enzyme under microtubule polymerising conditions. Tubulin was purified from rat brain and recombinant AR was obtained by expression of the AKR1B1 (human AR) gene in E. coli strain BL21a as described in Research Design and Methods. (A) Aliquots (20 µg protein) of tubulin (Tub) and AR were subjected to SDS PAGE, gel stained with Coomassie brilliant blue, and analyzed by immunoblotting using anti AR or anti Tub (DM1A). (B) Tubulin (200 µg) was precipitated with Ni+2 NTA beads linked to AR (recombinant AKR1B1 protein; lane AR Tub), NEP (exopolyphosphatase of Pseudomona aeruginosa [gene name ppx]; lane CI) or Ni+2 NTA beads without ligand (lane Ctub). The presence of tubulin was analyzed in the sedimented material by immunoblotting using DM1A as primary antibody. The volume of the analyzed sample was calculated to represent the same amount of protein in each case. (C) Recombinant AR (7 µg) was mixed with tubulin at the indicated concentrations and incubated for 30 min at 37 ºC in TBS buffer (final volume 1 ml) in the presence of 5 µM taxol (□) or 50 µM nocodazole (■). The samples were processed for determination of AR activity. The values presented are mean ± SD from three independent experiments. (D) AR, rat brain tubulin, and tubulin/AR complex preparations were subjected to molecular exclusion chromatography. Fractions corresponding to various elution volumes were collected, and tubulin and AR were analyzed by immunoblotting with DM1A and AR activity measurement, respectively. Upper panel: bands indicate tubulin elution volumes. Lower panel: AR activity was quantified and plotted as a function of elution volume. The values presented are mean ± SD from three independent experiments.
For Peer Review Only Page 25 of 31 Diabetes
182x194mm (127 x 127 DPI)
ARTICLE
RETRACTED
For Peer Review Only Diabetes Page 26 of 31
Fig. 2. AR is associated with growing microtubules but not with pre assembled microtubules. Tubulin and recombinant AR were obtained as described in Fig. 1. (A) Ni+2 NTAARTICLE beads linked to AR (recombinant AKR1B1 protein; lane Tub) were incubated with pre assembled taxolated microtubules (lane MT) or with soluble tubulin plus taxol (lane Tub) as described in Research Design and Methods. After incubation, samples were precipitated by centrifugation. The presence of tubulin in the precipitated materials was analyzed by immunoblotting with DM1A. The volume of the analysed sample was calculated to represent the same amount of protein in each case. (B) AR (7 µg recombinant AKR1B1 protein) was incubated for 30 min at 37 °C in TBS buffer (final volume 1 ml) in the presence of various concentrations of tubulin. The tubulin was added in pre assembled form in taxolated microtubules (■) or in soluble form in the presence of taxol (□). The samples were processed for determination of AR activity. The values presented are mean ± SD from three independent experiments. 88x60mm (150 x 150 DPI)
RETRACTED
For Peer Review Only Page 27 of 31 Diabetes
Fig. 3. Tubulin isotypes associated with AR. Ni+2-NTA beads linked to AR (recombinant AKR1B1 protein) were incubated with tubulin plus taxol as described in Research Design and Methods. The samples were precipitated by centrifugation. Tubulin preparation (Input) and precipitated materials (Precipitate) were analysed by immunoblotting for determination of total tubulin (α-tubulin) and various tubulin isotypes (acetylated, Ac-; tyrosinated, Tyr-; detyrosinated, Glu-; 3-nitrotyrosinated, 3-NTyr). Bands were quantified by densitometry using Scion Image software. Input ratio = ratio between tubulin isotype and α-tubulin in preparation. Precipitate ratio = ratio between tubulin isotype and α-tubulin in precipitate. R = ratio between precipitate ratio and input ratio. The values presented are mean ± SD from three independent experiments.
ARTICLE
RETRACTED
For Peer Review Only Diabetes Page 28 of 31
ARTICLE
RETRACTED
Fig. 4. Stimulation of AR activity by tubulin is dependent on the proportion of 3 NTyr tub isotype in the tubulin preparation. AR (7 g recombinant AKR1B1 protein) was incubated for 30 min at 37 °C in TBS buffer (final volume 1 ml) with various concentrations of tubulin. AR enzyme activity was determined as described in Research Design and Methods. (A) Various concentrations of tubulin preparations with low (□; L) and high (■; H) 3 NTyr tub isotype content were used. Insert: identical amounts of tubulin from the two preparations were analysed by immunoblotting for total and 3 NTyr tub. (B) Two tubulin preparations as above were mixed at the indicated percentages of 3 NTyr tub (% of H). A fixed amount of tubulin and various amounts of 3 NTyr tub were mixed with AR and processed as in (A). Insert as in (A). For both panels, the values presented are mean ± SD from three independent experiments, and the bands shown are from a representative experiment. 88x160mm (300 x 300 DPI)
For Peer Review Only Page 29 of 31 Diabetes
ARTICLE
RETRACTED
Fig. 5. Tyr and 3 NTyr prevent the association of AR with tubulin and the stimulation of AR activity by tubulin. (A) Purified rat brain tubulin (200 g) was mixed with Ni+2 NTA beads linked to recombinant AR in the presence of Tyr or 3 NTyr at the indicate concentrations. Precipitated materials were analyzed by immunoblotting with DM1A. The volume of the analyzed sample was calculated to represent the same amount of protein in each case. (B) AR (7 g) was incubated for 30 min at 37 °C in TBS buffer (final volume 1 ml) in the presence of the indicated concentrations of Tyr (■) or 3 NTyr (□) and 5 M taxol. The samples were processed for determination of AR activity. The values presented are mean ± SD from three independent experiments.
For Peer Review Only Diabetes Page 30 of 31
ARTICLE
RETRACTED
Fig. 6. Effects of Tyr and 3 NTyr on microtubule content, membrane tubulin, and NKA activity. COS cells were incubated for 2 h in glucose free HEPES FBS buffer at 37 °C (control; Ctrl) and then for 1 h in the presence (+) or absence ( ) of 0.5 mM Tyr or 3 NTyr (+3 NTyr). Glucose (Glc; final concentration 25 mM) was added, and the cells were incubated for another 2 h. (A) The cytoskeletal fraction was obtained as described in Research Design and Methods, and α tubulin was quantified by immunoblotting with DM1A. Bands were quantified by densitometry using Scion Image software. The values presented are mean ± SD from three independent experiments, expressed as a percentage of the control value. (B) Cells were fixed on coverslips and subjected to immunofluorescence with DM1A. (C) The membrane fraction was obtained as described in Research Design and Methods, and an aliquot (10 g protein) was processed for measurement of NKA activity by the NADH oxidation method (lower panel, white bars). A second aliquot was processed for estimation of the amount of AcTub by immunoblotting with mAb 6 11B 1. Upper panel: regions containing relevant bands of the blot from a representative experiment. Bands were quantified using Scion Image software. The values presented are mean ± SD from three independent experiments, expressed as a
For Peer Review Only Page 31 of 31 Diabetes
percentage of the control value (lower panel, black bars). The differences between values from different treatments were statistically significant (p< 0.01). 192x416mm (300 x 300 DPI)
ARTICLE
RETRACTED
For Peer Review Only Diabetes Page 32 of 31
ARTICLE
Fig. 7. Effects of Tyr and 3-NTyr on cataract formation in rats with STZ-induced diabetes in vivo and in rat lenses exposed to high glucose conditions ex vivo. 30-day-old rats (n=24) were i.p. injected with 70 mg STZ per kg body weight. As a control, 8 rats were injected with vehicle (citrate buffer). After 1 wk, one group RETRACTED(n=8) from the STZ-treated rats was orally administered Tyr (90 mg . kg-1. day-1) for 3 months, a second group (n=8) was administered 3-NTyr (30 mg . kg-1. day-1), and a third group (n=8) was administered vehicle (water) only. Tyr and 3-NTyr were mixed with the water supply. (A) Characteristics of the four experimental groups at the end of the 3-month period. HbA1c: glycosylation of haemoglobin expressed as % (mmol/mol). The values presented are mean ± SD from three independent experiments. (B) Representative photographs of lenses from rats in the four experimental groups at the end of the 3-month period, showing the effects of Tyr and 3-NTyr on cataract formation in vivo (magnification x4). (C) Upper panel: representative images of lenses from the four groups (magnification x25). Middle panel: lens opacity for each group was analyzed; data are expressed as Lens Diameter (LD) vs. Opacity (Arbitrary Units). Lower panel: AR activity in lenses incubated ex vivo. The values presented are mean ± SD from three independent experiments. 176x170mm (600 x 600 DPI)
For Peer Review Only Page 33 of 31 Diabetes
ARTICLE
Fig. 8. Proposed mechanism of regulation of NKA and AR by tubulin under high glucose conditions. 182x182mm (96 x 96 DPI) RETRACTED
For Peer Review Only