57

Molecular and Cellular Biochemistry 240: 57–65, 2002.© 2002 Kluwer Academic Publishers. Printed in the Netherlands.

Streptozotocin-induced diabetes impairs G-proteinlinked signal transduction in vascular smooth muscle

Shehla Hashim, Yi Yong Liu, Rui Wang and Madhu B. Anand-SrivastavaDepartment of Physiology and Groupe de recherche sur le Système Nerveux Autonome, Faculty of Medicine, University ofMontreal, Montreal, Quebec, Canada

Received 30 January 2002; accepted 30 April 2002

Abstract

The present studies were undertaken to examine if the impaired vascular function observed in diabetes is attributed to the al-tered levels of G-protein. Diabetes was induced in Sprague Dawley rats by a single intraperitoneal injection of streptozotocin(STZ) (60 mg/kg body wt) and after a period of 5 days, the aorta were used for adenylyl cyclase activity determination andprotein quantification. A temporal relationship between the expression of Giα proteins and development of diabetes was alsoexamined on day 1, 2, 3, 4 and 5 of injection of STZ. Blood glucose levels were significantly increased from day 1 in STZ-ratsas compared to their counterpart control rats and reached to about 20 mM on 3rd day and 30 mM on 5th day. The expressionof Giα-2 and Giα-3 proteins as determined by immunoblotting techniques was decreased by about 70 and 50% respectively inaorta from STZ rats compared to the control rats after 5 days of treatment, whereas 40% decrease in Giα-2 and Giα-3 wasobserved after 3rd day of STZ injection. On the other hand, the expression of Gsα was unaltered in STZ rats. In addition, thestimulatory effect of cholera toxin (CT) on GTP-mediated stimulation of adenylyl cyclase was not different in STZ as com-pared to the control group. However, the stimulatory effects of isoproterenol, glucagon, NaF and FSK on adenylyl cyclaseactivity were significantly enhanced in STZ rats as compared to control rats, whereas basal adenylyl cyclase activity was sig-nificantly lower in STZ-rats as compared to control rats. In addition, GTPγS inhibited FSK-stimulated adenylyl cyclase activ-ity in concentration-dependent manner (receptor-independent functions of Giα) in control rats which was completely attenuatedin STZ-rats. In addition, receptor-mediated inhibitions of adenylyl cyclase by angiotensin II, oxotremorine, atrial natriureticpeptide (ANP

99–126) and C-ANP

4–23 were also attenuated (receptor-dependent functions of Giα) in STZ-rats. These results indi-

cate that aorta from diabetic rats exhibit decreased levels of cAMP and decreased expression of Giα. The decreased expressionof Giα may be responsible for the altered responsiveness of adenylyl cyclase to hormonal stimulation and inhibition in STZ-rats. It may thus be suggested that the impaired adenylyl cyclase-Giα protein signaling may be one of the possible mechanismsresponsible for the impaired vascular functions in diabetes. (Mol Cell Biochem 240: 57–65, 2002)

Key words: adenylyl cyclase, aorta, diabetes, G-proteins, streptozotocin

Abbreviations: ANP – atrial natriuretic peptide (99–126); C-ANP4–23

– a ring deleted analog of atrial natriuretic peptide; C-ANP

4–23 – [des(Gln18,Ser19,Gln20,Leu21,Gly22)ANP

4–23-NH

2]; Gi – inhibitory guanine nucleotide regulatory protein; GTPγS –

guanosine 5′-0-(3-thiotriphosphate); FSK – forskolin; AngII – angiotensin II

Introduction

Cardiovascular complications of diabetes mellitus are respon-sible for most of morbidity and mortality associated with thedisease. The adenylyl cyclase/cyclic AMP (cAMP) system is

believed to be one of the biochemical mechanisms partici-pating in the regulation of cardiovascular functions. Theadenylyl cyclase system is composed of three components:receptor, catalytic subunit and stimulatory (Gs) and inhibi-tory (Gi) guanine nucleotide regulatory proteins [1, 2]. The

Address for offprints: M.B. Anand-Srivastava, Department of Physiology, Faculty of Medicine, University of Montreal, C.P. 6128, Succ. Centre-ville, Mon-treal, Quebec, Canada, H3C 3J7 (E-mail: anandsrm@physio.umontreal.ca)

58

stimulation and inhibition of adenylyl cyclase by hormonesare mediated by two distinct G-proteins, Gs and Gi respec-tively that couple the receptor to the catalytic subunit. TheG-proteins are heterotrimeric and are composed of α, β andγ subunits. The specificity of G-proteins is attributed to αsubunits. Molecular cloning has revealed four different formsof Gsα resulting from the differential splicing of one gene[3–5] and three distinct forms of Giα; Giα-1, Giα-2 and Giα-3 encoded by three distinct genes [6–8]. All three forms ofGiα (Giα

1–3) have been reported to be implicated in adenylyl

cyclase inhibition [8] and activation of atrial K+ channels [9].In addition, five different β subunits of 35–36 kDa and sevenγ subunits of 8–10 kDa have been identified by molecularcloning [10, 11]. Several functions of the complex βγ subunithave been reported such as anchoring the G-protein to themembrane [10], stimulation of type II and IV enzyme [12],inhibition of Ca2+/calmodulin or Gs-activated type I enzymeand activation of muscarinic-gated atrial K+channels [13].Molecular cloning has also revealed eight different types ofadenylyl cyclases but only types V and VI have been identi-fied in heart and aorta [14–16].

The adenylyl cyclase/cAMP system has been implicatedin the control of heart contractility [17, 18] and vascularsmooth muscle tone [19, 20]. Several abnormalities in theexpression of G-proteins and adenylyl cyclase regulationhave been demonstrated in various pathophysiological con-ditions, such as heart failure, hypertension and hypothy-roidism [21–25]. Mice deficient in Giα-2 have been shownto exhibit phenotype of insulin resistance [26]. In addition,recent studies showing that the overexpression of Giα-2ameliorates STZ-diabetes further suggest the involvementof Giα-2 protein in pathogenesis of diabetes [27]. Diabetes-induced alterations in G protein-adenylyl cyclase activity andits responsiveness to various hormones have been demon-strated in liver, adipose tissue, heart, skeletal muscle, cer-ebrum, cerebral microvessels and retina [28–34]. However,despite the fact that adenylyl cyclase signaling plays an im-portant role in the modulation of variety of vascular functions,a systematic evaluation of this pathway that may contribute tothe vascular complications in diabetes has not been performed.The present studies were undertaken to determine the levelsof G proteins (Gsα and Giα) and their relationship withadenylyl cyclase regulation in aorta from short-term STZ-rats.

Materials and methods

Materials

Streptozotocin, adenosine triphosphate (ATP), cyclic AMP(cAMP), isoproterenol, forskolin, glucagon and oxotremo-rine were purchased from Sigma Chemical Co., St. Louis,MO, USA. Creatine kinase, myokinase and GTPγS were pur-

chased from Boehringer Mannheim (Montreal, Quebec,Canada) [α-32P] ATP was from Amersham Corp. (Ontario,Canada). ANP

99–126 and C-ANP

4–23 were purchased from Pe-

ninsula Laboratories (Belmont, CA, USA). N-Ethylcarbox-amideadenosine (NECA) was from Research Biochemicals(Wayland, MA, USA).

Animal preparation

Male Sprague-Dawley (SD) rats (200 g) (6–8 weeks-old)were maintained on standard rat chow and tap water ad libi-tum with 12 h light/dark cycles in a quiet environment. Diabe-tes was induced by intraperitoneal injection of streptozotocin(STZ, 60 mg/kg body wt), dissolved in sodium citrate buffer(pH 4.5). Age-matched control rats were injected with anequal volume of buffer solution. Blood glucose levels weremonitored from day 1 to day 5 after the injection using adextrometer (Ames). Streptozotocin-injected rats with bloodglucose levels in excess of 26 mM were considered to bediabetic rats (STZ) and used in this study. The blood glucoselevel of control rats was 5.5 mM. All the protocols used inthe present study were approved by the comité de déontologiede l’expérimentation sur les animaux (CDEA), Canada.

Preparation of aorta participate fraction

Aorta washed particles were prepared as described previously[35]. The dissected aortae were quickly frozen in liquid N

2

and pulverized to a fine powder with a mortar and pestlecooled in liquid N

2 and were stored at –70°C until assayed.

After homogenization in a Teflon/glass homogenizer in abuffer containing 10 mM-Tris/HCl and 1 mM-EDTA (pH7.5), the homogenate was centrifuged at 16000 × g for 10 min.The supernatant fraction was discarded, and the pellet wasfinally suspended in 10 mM-Tris-HCl and 1 mM-EDTA andused for determination of adenylyl cyclase activity and G-protein expression.

Cholera toxin (CT) treatment

Aorta particulate fraction was treated with cholera toxin (CT)as described earlier [23]. CT (500 µg/ml) was preactivatedfor 20 min at 37°C in a mixture containing 20 mM KH

2PO

4

(pH 8.0). To study the effect of CT on adenylyl cyclase ac-tivity, aorta particulate fraction was pretreated with or with-out CT for 30 min at 30°C in a reaction mixture containing250 mM KH

2PO

4 (pH 6.8), 1 mM MgCl

2, 0.5 mM EDTA (pH

8.0), 5 mM ATP, 15 mM thymidine, 0.15 mM GTP, 20 mMdithiothretol and 1 mM NAD. The particulate fraction waswashed twice with buffer containing 10 mM Tris and 1 mM

59

EDTA (pH 7.5) and finally suspended in the same buffer foradenylyl cyclase activity determination.

Immunoblotting

Immunoblotting of G-proteins was performed as describedearlier [23]. After SDS/PAGE, the separated proteins wereelectrophoretically transferred to nitrocellulose paper(Schleicher and Schuell) with a mini transfer apparatus (Bio-Rad) at 100 V for 1 h or a semi-dry transblot apparatus (Bio-Rad) at 15 V for 45 min. After transfer, the membranes werewashed twice in phosphate-buffered saline (PBS) and wereincubated in PBS containing 3% BSA at room temperaturefor 2 h. The blots were then incubated with antisera againstG-proteins in PBS containing 1% BSA and 0.1% Tween-20at room temperature for 2 h. The antigen-antibody complexeswere detected by incubating the blots with goat anti-rabbitIgG (Bio-Rad) conjugated with horseradish peroxidase for2 h at room temperature. The blots were washed three timeswith PBS before reaction with enhanced-chemiluminescence(ECL) Western-blotting detection reagents from Amersham.Quantitative analysis of the G-proteins was performed bydensitometric scanning of the autoradiographs employing theenhanced laser densitometer (LKB Ultrscan XL) and quan-tified using the gel Scan XL evaluation software (version 2.1)from Pharmacia (Quebec, Canada).

Adenylyl cyclase activity determination

Adenylyl cyclase activity was determined by measuring [32P]-cAMP formation from [α32]ATP, as described previously [23].The assay medium containing 50 mM-glycylglycine, pH 7.5,0.5 mM-MgATP, [α-32P]ATP[1–1.5) × 106 c.p.m.), 5 mMMgCl

2 (in excess of the ATP) concentration), 100 mM NaCl,

0.5 mM-cAMP, 1 mM 3-isobutyl-1-methylxanthine, 0.1 mMEGTA, 10 µM-guanosine 5′-[γ-thio]triphosphate (GTPγS) (orotherwise indicated), and an ATP regenerating system con-sisting of 2 mM-phosphocreatine, 0.1 mg of creatine kinase/ml and 0.1 mg of myokinase/ml in a final volume of 200 µl.Incubations were initiated by addition of the membrane prep-aration (30–70 µg) to the reaction mixture, which had beenthermally equilibrated for 2 min at 37°C. The reactions, con-ducted in triplicate for 10 min at 37°C, were terminated byaddition of 0.6 ml of 120 mM-zinc acetate. cAMP was puri-fied by co-precipitation of other nucleotides with ZnCO

3, by

addition of 0.5 ml of 144 mM Na2CO

3 and subsequent chro-

matography by the double-column system, as described bySalomon et al. [36]. Under the assay conditions used, adenylylcyclase activity was linear with respect to protein concentra-tion and time of incubation. Protein was determined essen-tially as described by Lowry et al. [37] with crystalline BSAas standard.

Statistical analysis

Data are expressed as mean ± S.E.M. Comparisons betweengroups (control and STZ-treated rats) were made with stu-dent ‘t’-test for unpaired samples. Difference between groupswas considered statistically significant at p < 0.05.

Results

Blood glucose levels

The blood glucose levels were monitored daily using glucosemonitoring system from Roche diagnostics, Canada. Theblood glucose levels in control animals were in the range of5.0 ± 0.35 mM up to 5 days, whereas it was increased in atime-dependent manner in rats after single intraperitoneal in-jection of STZ and reached 30 mM on day 5 (Table 1).

Effect of GTPγS on adenylyl cyclase activity

Guanine nucleotides stimulate or inhibit adenylyl cyclaseactivity by interacting with G proteins. In order to investi-gate if G-proteins are impaired in aorta from STZ-rats, theeffect of GTPγS on adenylyl cyclase activity was examined andthe results are shown in Fig. 1. GTPγS stimulated adenylylcyclase activity in aorta from both STZ and control rats ina concentration-dependent manner; however, the extent ofstimulation was significantly greater in STZ than controlrats. GTPγS at 10 µM stimulated adenylyl cyclase activityby about 7-fold in control rats (CTRL) whereas about 12-foldstimulation was observed in STZ rats. However, the basaladenylyl cyclase activity was significantly decreased in STZ-rats as compared to control rats (61.0 ± 5.0 vs. 101.5 ± 8 pmolcAMP (mg protein.10 min–1).

Table 1. Effect of STZ injection on blood glucose levels

Day Control rats STZ rats(mM) (mM)

1 5.5 ± 0.25 13.5 ± 0.692 5.5 ± 0.15 13.8 ± 3.13 5.8 ± 0.36 19.9 ± 1.64 5.8 ± 0.21 25.3 ± 2.85 5.4 ± 0.28 30.1 ± 3.1

Diabetes was induced by single intraperitoneal injection of streptozotocin(60 mg/kg body wt) in sodium citrate buffer (pH 4.5) as described in ‘Ma-terials and methods’. Blood glucose levels were monitored from day 1 today 5 using glucometer. Values are means ± S.E.M. of 30–40 rats.

60

G-protein levels

In order to investigate if the increased responsiveness ofadenylyl cyclase to GTPγS in aorta from STZ as comparedto control rats was due to the increased levels of Gsα or de-

creased levels of Giα, the levels of G-proteins were deter-mined by immunoblotting using specific antibodies; AS/7 an-tibodies against Giα-1 and Giα-2, EC/2 antibodies againstGiα-3 and RM/1 antibodies against Gsα. As shown in Fig.2, AS/7 and EC/2 antibodies recognized a single protein of40 and 41 KDa respectively referred to Giα-2 (Giα-1 is ab-sent in aorta [38]) and Giα-3 respectively from both controland STZ rats, however, the relative amounts of immuno-detectable Giα-2 and Giα-3 were significantly decreased inSTZ as compared to control rats by about 90 and 50% respec-tively at day 5 as determined by densitometric scanning. Onthe other hand, the levels of Gsα protein were not altered inSTZ rats (in arbitrary units CTRL 1.4 ± 0.10, STZ, 1.5 ± 0.15(n = 3).

In order to investigate the relationship between decreasedexpression of Giα protein and development of diabetes, thelevels of Giα protein were determined in aorta at differentdays after STZ-injection. The results illustrated in Fig. 3demonstrate that the levels of Giα-2 and Giα-3 proteinswere decreased in STZ-rats as compared to control rats ina time-dependent manner. A small decrease of about 20% inGiα-2 and about 10% in Giα-3 levels in STZ-rats as com-pared to control was observed at day 1 and 2, however, at day5 the levels of Giα-2 and Giα-3 protein were decreased byabout 90 and 50% respectively.

To corroborate our results with the functions of Gs pro-teins, the effect of cholera toxin (CT) on adenylyl cyclaseactivity was investigated. The results shown in Table 2 indi-cate that CT stimulated adenylyl cyclase activity in the ab-sence or presence of GTP in aorta from both control and STZ

Fig. 1. Effect of GTPγS on adenylyl cyclase activity in the aorta from con-trol and STZ-induced diabetic rats. Adenylyl cyclase activity was deter-mined in aorta from control (CTRL) and STZ-induced diabetic rats (STZ)in the absence or presence of various concentrations of GTPγS as describedin ‘Materials and methods’. Values are mean ± S.E.M. of 3 separate experi-ments. Basal enzyme activities in control (CTRL) and STZ were 101.5 ± 8and 61 ± 5 pmol of cAMP/10 min per mg of protein respectively. Six ani-mals were utilized for each experiment. *p < 0.05.

Fig. 2. Quantification of (a) Giα-2, (b) Giα-3 and (c) Gsα proteins by immunoblotting in aorta from control and STZ-induced diabetic rats (STZ). Themembrane proteins (50 µg from control and STZ) were separated on SDS-PAGE and transferred to nitrocellulose membrane, which was then immunoblottedusing AS/7, EC/2 and RM/1 antibodies for Giα-2, Giα-3 and Gsα proteins respectively as described in ‘Materials and methods’. The autoradiograph is arepresentative of 3 separate experiments utilizing 3 separate aorta preparations from control and STZ-rats (upper panel). Lower panel: Densitometric scan-ning of Giα-2 and Giα-3 and Gsα proteins from control and STZ-treated rats. The results are expressed as percentage of controls (taken as 100%). Valuesare means ± S.E.M. of 4 separate experiments utilizing 3–4 separate aorta preparations from control and STZ-rats.

61

rats, however, the percent stimulation was not significantlydifferent in two groups. These results suggest that the func-tions of Gsα in aorta were also not altered in STZ rats.

Hormonal stimulation of adenylyl cyclase

Since the levels and functions of Gsα were not altered inSTZ rats, it was of interest to investigate if Gs-mediated hor-monal stimulations were also unaltered in STZ rats. Resultsshown in Fig. 4A demonstrate that isoproterenol and glu-cagon, stimulated adenylyl cyclase activity in aorta fromboth the groups, however, the extent of stimulation was sig-nificantly augmented in STZ-diabetic rats as compared tocontrol rats. For example, isoproterenol and glucagon stim-ulated adenylyl cyclase activity by about 300 and 250%respectively in control and about 450 and 350% respectivelyin STZ rats. In addition, FSK- and NaF that stimulate adenylylcyclase activity by receptor-independent mechanisms alsostimulated enzyme activity to various degrees in both con-trol and STZ-rats, however, the extent of stimulation wasabout 150% higher in STZ-rats as compared to control rats(Fig. 4B).

Hormonal inhibition of adenylyl cyclase

Adenylyl cyclase activity is regulated by dual pathways; stim-ulatory and inhibitory, mediated by Gsα and Giα respectively.Since the levels of Giα proteins were decreased in STZ rats,it was of interest to examine if the functions of Giα were alsoattenuated in these rats. For this reason, the receptor-depend-ent and receptor-independent functions of Gi were studiedand the results are shown in Fig. 5. Angiotensin II (Ang II),oxotremorine (OXO), atrial natriuretic peptide (ANP

99–126)

and ring deleted peptide of ANP (C-ANP4–23

) that inhibitadenylyl cyclase through Giα proteins [39–41] inhibited theenzyme activity by about 20, 24, 21 and 24% in control rats;which was almost completely attenuated in STZ rats (Fig.5A). Similarly, GTPγS inhibited FSK-stimulated activity ina concentration-dependent manner in control rats which was

Fig. 3. Upper panel: Quantification of Giα-2 and Giα-3 proteins by im-munoblotting in aorta from control and STZ-treated rats at 1, 2, 3, 4 and 5day after a single injection of STZ. The aorta (particulate fraction) proteins(50 µg) from different groups (control and STZ-treated) were separated onSDS-PAGE and transferred to nitrocellulose, which was then immuno-blotted using AS/7 (A) and EC/2 antibodies (B) for Giα-2 and Giα-3 pro-teins respectively as described in ‘Materials and methods’. The detectionof Giα-2 and Giα-3 proteins was performed by using chemiluminescenceWestern blotting detection reagents from Amersham (upper panels). Theautoradiograms are representative of 3–4 separate experiments utilizing3–4 separate aorta preparations for control and STZ-rats. Lower panel:Densitometric scanning of Giα-2 and Giα-3 proteins from control and STZ-treated rats. The results are expressed as percentage of controls (taken as100%). Values are means ± S.E.M. of 4 separate experiments utilizing 3–4 separate aorta preparations from control and STZ-rats.

Table 2. Effect of cholera toxin treatment on adenylyl cyclase activity in aorta from control and STZ rats

Adenylyl cyclase activity(pmol cAMP (mg protein.10 min–1)

Control rats STZ-ratsAdditions –CT +CT Stimulation % –CT +CT Stimulation %

None 33 ± 5.1 69.5 ± 3.5 110 21.8 ± 5.5 41.8 ± 2.6 9210 µM GTP 46.4 ± 5.4 238.5 ± 11 410 36.6 ± 4.5 175 ± 9.5 386

Aorta particulate fractions from control and STZ-induced diabetic rats were incubated with CT at 30°C for 30 min as described in ‘Materials and methods’.Adenylyl cyclase activity was determined in the absence or presence of 10 µM GTP as described in ‘Materials and methods’. Values are means ± S.E.M. oftriplicate determinatins from 3 separate experiments. –CT – absence of cholera toxin (CT); +CT – presence of cholera toxin.

62

almost completely attenuated in STZ rats (Fig. 5B), suggest-ing a correlation between the decreased levels and decreasedfunctions of Giα proteins in STZ rats.

Fig. 5. (A) Effect of angiotensin II (Ang II), oxotremorine (OXO), atrialnatriuretic peptide (ANP) and ring deleted ANP (C-ANP4–23) on adenylylcyclase activity in aorta from control and STZ-induced diabetic rats. Adenylylcyclase activity was determined in the presence of 10 µM GTPγS alone orin combination with 10 µM Ang II, 50 µM OXO, 0.1 µM ANP and 0.1 µMcANP4–23 as described in ‘Materials and methods’. Values are means ±S.E.M. of 3 separate experiments. Six animals were utilized for each ex-periment. Adenylyl cyclase activities in the presence of GTPγS in con-trol and STZ rats were 831 ± 27 and 540 ± 21 pmol of cAMP (mg protein.10min–1) respectively. *p < 0.05. (B) Effect of GTPγS on Forskolin-stimulatedadenylyl cyclase activity in aorta from control and STZ-induced diabeticrats. Adenylyl cyclase activity was determined in the presence or absenceof 100 µM FSK alone or in combination with various concentrations ofGTPγS in aorta from control (O) and STZ-induced diabetic (�) rats as de-scribed in ‘Materials and methods’. Values are mean ± S.E.M. of 3 sepa-rate experiments. Six animals were utilized for each experiment. Adenylylcyclase activities in the presence of 100 µM FSK in aorta from CTRL andSTZ rats were 603 ± 25 and 381 ± 27 pmol of cAMP (mg protein.10 min–1)respectively. *p < 0.05.

Fig. 4. (A) Effect of isoproterenol and glucagon on adenylyl cyclase ac-tivity in aorta from control and STZ-induced diabetic rats. Adenylyl cyclaseactivity was determined in the presence of 10 µM GTP alone or in combi-nation with 50 µM isoproterenol (ISO) or 1 µM glucagon (GLU) as describedin ‘Materials and methods’. Values are means ± S.E.M. of 3 separate ex-periments. Six animals were utilized for each experiment. Adenylyl cyclaseactivities in the presence of GTP in control and STZ rats were 198 ± 27 and192 ± 26.6 pmol of cAMP/10 min per mg of protein respectively. *p < 0.05.(B) Effect of forskolin and NaF on adenylyl cyclase activity in aorta fromcontrol and STZ-induced diabetic rats. Adenylyl cyclase activity was de-termined in the absence or presence of 50 µM forskolin (FSK) or 10 mMsodium fluoride (NaF) as described in ‘Materials and methods’. GTP orGTPγS was omitted from the reaction mixture. Values are means ± S.E.M.of 3 separate experiments. Six animals from each group were utilized foreach experiment. Basal adenylyl cyclase activities in control and STZ ratswere 66.3 ± 5.6 and 54.6 ± 1.4 pmol cAMP (mg protein.10 min–1) respec-tively. *p < 0.05.

63

Discussion

In the present studies, we demonstrate for the first time thatadenylyl cyclase activity, its responsiveness to various hor-mones as well as the levels of Giα but not of Gsα are impairedin aorta in very early stages of streptozotocin diabetes.

The rats treated with STZ showed enhanced blood glucoselevel within 2 days after injection with a concurrent decreasein the levels of Giα-2 and Giα-3 proteins, suggesting that thedecrease in the levels of Giα proteins is associated with thedevelopment of diabetes. Subsequent increase in the levelsof blood glucose through day 5 resulted in further decreasein the levels of Giα proteins suggesting a close relationshipbetween decreased levels of Gi and the severity of diabetes.The decreased expression of Giα proteins has been reportedin hepatocytes from human diabetics and STZ-diabetic rats[42, 43] whereas an increase in the levels and functions ofGiα was shown in diabetic adipocytes from a genetic modelof diabetes [29]. Livingstone et al. [44] have shown a de-creased expression of Giα proteins in platelets from diabeticsubjects as compared to non diabetic subjects. In addition,diabetic retina has been shown to exhibit decreased levels andfunctions of Gi [45]. However, Weber and McLeod [46] wereunable to observe any changes in the levels of Giα proteinsin aorta or caudal artery from 12–14 week STZ-diabetic ratsas compared to control rats. Our results are consistent withthe studies of Hattori et al. [47] who have also reported a simi-lar decrease in Gi protein in aorta from long term diabetic rats,however, these investigators did not examine adenylyl cy-clase Gi-protein signaling in their studies. Further support andinvolvement of Giα-2 protein in pathogenesis of diabetes hasbeen provided by the studies showing that the overexpres-sion of constitutively activated Giα-2 ameliorates STZ-in-duced diabetes in rats [27]. In addition, a complete knockoutof the Giα-2 gene that has been reported to produce a meta-bolic state resembling type II diabetes suggests the relation-ship between the decreased levels of Giα protein and diabetes[26]. On the other hand, an increased expression of Gqα indiabetic aorta has also been suggested to be responsible forthe increased vascular contraction to NaF through the acti-vation of Gqα-mediated Ca2+ channel [42].

The decreased levels of Giα-2 and Giα-3 were also re-flected in the decreased functions, as has been demonstratedby an attenuation of the inhibitory responses of Ang II, OXO,ANP and C-ANP on adenylyl cyclase activity. Our results arein accordance with the studies of other investigators, whohave also shown the attenuation of inhibitory functions of Gi-protein in various tissues [32, 42–45]. However, it appearsthat a partial decrease (~ 60%) in the levels of Giα-2 observedin aorta from STZ as compared to control rats may be enoughto uncouple the hormone receptors from adenylyl cyclasesystem or alternatively, some other mechanisms at the recep-tor level, such as receptor down regulation, may also be re-

sponsible for a complete attenuation of inhibitory responseson adenylyl cyclase. In this context, the down regulation ofANP receptors in various models of hypertension and con-gestive heart failure has been demonstrated in various tissuesincluding platelets [48, 49]. Taken together, it is suggestedthat the decreased levels of Giα-2 and Giα-3 in aorta fromSTZ rat may partly be responsible for the attenuated recep-tor-mediated inhibition of adenylyl cyclase by ANG II, OXO,ANP, C-ANP

4–23 and receptor-independent inhibition, i.e. the

inhibition of forskolin-stimulated adenylyl cyclase activityby low concentration of GTPγS.

We have also observed that basal adenylyl cyclase activ-ity was decreased in aorta from STZ-rats, suggesting thatdiabetes may induce alterations in the catalytic componentof adenylyl cyclase. Similar diabetes-induced reductions inbasal adenylyl cyclase activity have also been reported inadipocytes [50], hepatocytes [31] and retina [51].

In addition, the responsiveness of adenylyl cyclase toGTPγS stimulation in diabetic aorta that was significantlyincreased may be attributed to the increased sensitivity and/or increased levels of Gsα or decreased levels of Giα in aortafrom STZ rats. However, our results did not demonstrate anychanges in the levels or functions of Gsα in STZ rats and sug-gest that the Gsα may not be responsible for the increasedsensitivity of adenylyl cyclase to GTPγS stimulation. On theother hand, the decreased levels of Giα in diabetic aorta thatremoves the tonic GTP-dependent inhibitory function of Gimay be responsible for the enhanced stimulation of adenylylcyclase by GTPγS in STZ rats. In addition, decreased basaladenylyl cyclase activity in diabetic aorta may also contrib-ute to the enhanced stimulation of adenylyl cyclase to GTPγS.

The hyperresponsiveness of adenylyl cyclase to ISO andglucagon stimulation in diabetic aorta as compared to con-trol aorta may be attributed to the upregulation of hormonereceptors or increased levels of Gsα. However, several stud-ies have shown a down-regulation or no change in the numberof stimulatory hormone receptors in various cardiovasculardiseases [52]. Furthermore, since no alterations in the levelsand functions of Gsα were observed in STZ, the enhancedstimulation of adenylyl cyclase by ISO and glucagon cannotbe explained by Gsα. Based on our results in STZ rats, it ispossible that the augmented responsiveness of adenylyl cy-clase to ISO and glucagon in STZ rats may be attributed todecreased levels of Giα-2 proteins. In this regard, a relation-ship between decreased levels of Gi and augmented respon-siveness of adenylyl cyclase to stimulatory hormones hasbeen shown by previous studies where pertussis toxin (PT)and amiloride treatments which inactivate the Gi protein re-sulted in an augmentation of stimulatory responses of hor-mones on adenylyl cyclase [39, 53]. Furthermore, plateletsfrom spontaneously hypertensive rats [54] and patients [55]that exhibit decreased levels of Giα proteins elicited enhancedstimulation of adenylyl cyclase by N-ethylcarboxamide ad-

64

enosine and prostaglandins. Our results are consistent withthe observations made by other investigators showing that theloss of Gi functions in STZ-diabetic rats resulted in theaugmentation of glucagon and isoprenaline-stimulation ofadenylyl cyclase activity [31, 50]. Similarly, the enhancedstimulation of adenylyl cyclase by FSK and NaF in aorta fromSTZ rat as compared to control rats may be due to hypersen-sitivity or to the increased levels of the catalytic subunit ofthe adenylyl cyclase system per se or to the decreased expres-sion of Giα-2 or to the increased expression of Gsα or to thealterations in all the components of adenylyl cyclase system.Our results are in agreement with the previous studies show-ing an increased stimulation of adenylyl cyclase by FSK inplatelets from SHR [54] and hypertensive patients [55] thatexhibited decreased expression of Giα proteins. In addition,treatment of A10 smooth muscle cells with C-ANP

4–23 that

decreased the expression of Giα proteins showed an in-creased stimulation of adenylyl cyclase by FSK [56]. The Gi-mediated regulation of FSK-stimulated enzyme activity canbe further supported by the results of various studies show-ing an augmentation of FSK-stimulated adenylyl cyclaseactivity by PT treatment [39, 53]. On the other hand, therequirement of Gs and guanine nucleotides for the FSK ac-tivation of adenylyl cyclase has also been reported [53]. Sincethe present studies do not demonstrate any alteration in Gsα,the increased sensitivity of adenylyl cyclase to FSK stimu-lation in diabetes cannot be attributed to the Gs activity. Sincebasal adenylyl cyclase activity was decreased and not in-creased in STZ-aorta, it may be possible that the increasedpercent stimulation of adenylyl cyclase by FSK in STZ-ratsmay not be attributed to the increased levels of catalyticsubunit of adenylyl cyclase. Thus it may be suggested thatthe decreased levels of Giα proteins and decreased basaladenylyl cyclase activity in aorta from STZ rats may contrib-ute to the observed augmentation of the responsiveness ofadenylyl cyclase to FSK stimulation.

In conclusion, the expression of Giα is decreased in aortafrom STZ, whereas the levels of Gsα were not altered. Thedecreased expression of Giα proteins appears to explain, inpart, the attenuated responsiveness of adenylyl cyclase toinhibitory hormones and augmented responsiveness to stimu-latory hormones and agents that activate adenylyl cyclase byreceptor-independent mechanisms. It is suggested that thedecreased expression of Giα proteins in aorta from STZ maybe one of the possible mechanisms responsible for the im-paired cardiovascular functions in diabetes.

Acknowledgements

This work was supported by grant from Quebec Heart Foun-dation. We would like to thank Christiane Laurier for hervaluable secretarial help.

References

1. Neer EJ: Heterotrimeric G proteins. Organizers of transmembrane sig-nals. Cell 80: 249–257, 1995

2. Fleming JW, Wisler PL, Watanbe AM: Signal transduction by G-pro-teins in cardiac tissues. Circulation 85: 420–433, 1992

3. Murakami T, Yasuda H: Rat heart cell membranes contain three sub-strate for cholera toxin-catalyzed ADP-ribosylation and a single substratefor Pertussis toxin-catalyzed ADP-ribosylation. Biochem Biophys ResCommun 138: 1355–1361, 1986

4. Robishaw JD, Smigel MD, Gilman AG: Molecular basis for two formsof the G-protein that stimulate adenylate cyclase. J Biol Chem 261:9587–9590, 1986

5. Bray P, Caster A, Simons C, Guo V, Puckett C, Hamholz J, Spiegel A,Nirenberg H: Human cDNA clones for four species of Gsα signal trans-duction protein. Proc Natl Acad Sci USA 83: 8893–8897, 1986

6. Itoh H, Kozaka, T, Magata S, Nakamura S, Katada T, Ui M, Iwai S,Ohtsuka E, Kawasaki H, Suzuki K, Kaziro Y: Molecular cloning andsequence determination of cDNAs for α subunits of the guanine nu-cleotide-binding proteins Gs, Gi and Go from brain. Proc Natl AcadSci USA 83: 3776–3780, 1986

7. Itoh H, Toyama R, Kozasa T, Tsukamoto T, Matsuoka M, Kaziro Y:Presence of three distinct molecular species of Gi protein: A subunitstructure of rat cDNA and human genomic DNAs. J Biol Chem 263:6656–6664, 1988

8. Wong YH, Conklin BB, Bourne HR: G2-mediated hormonal inhibi-tion of cyclic AMP accumulation. Science 255: 339–342, 1992

9. Yatani A, Mattera R, Codina J: The G-protein gated atrial K+ channelsis stimulated by three distinct Giα subunits. Nature 336: 680–682, 1988

10. Simon MI, Strathmann MP, Gautam N, Science 252: 802–808, 199111. Cali JJ, Balcueva EA, Rybalkin I, Robishaw JD: Selective tissue dis-

tribution of G protein γ subunits, including a new form of the γ subunitsidentified by cDNA cloning. J Biol Chem 267: 24023–24027, 1992

12. Tang WJ, Gilman AG: Type-specific regulation of adenylyl cyclase byG protein βγ subunits. Science 254: 1500–1503 1991

13. Wickman KD, Iniguez-Liuhi JA, Davenport PA, Taussig R, KrapivinskiGB, Linder ME, Gilman AG, Clapham DE: Recombinant G-proteinβγ subunit activate the muscarinic-gated atrial potassium channel. Na-ture 368: 255–257, 1994

14. Tang WJ, Gilman AG: Adenylyl cyclase. Cell 70: 869–872, 199215. Premont RT, Chen J, Ma HW, Ponnapalli M, Iyengar R: Two mem-

bers of a widely expressed subfamily of hormone-stimulated adenylylcyclases. Proc Natl Acad Sci USA 89: 9809–9813, 1992

16. Katsushika S, Chen L, Kawabe JI, Nilakantan R, Halnon NJ, HomcyCJ, Ishikawa Y: Cloning and characterization of a sixth adenylyl cy-clase isoform: type V and VI constitute a subgroup within the mamma-lian adenylyl cyclase family. Proc Natl Acad Sci USA 89: 8774–8778,1992

17. Katz MA, Tada M, Kirchberger MA: Control of calcium transport inthe myocardium by the cyclic AMP-protein kinase system. Adv Cy-clic Nucleotide Res 5: 453–472, 1975

18. Keely SL Jr, Lincoln TM, Corbin JD: Interaction of acetylcholine andepinephrine on heart cyclic AMP-dependent protein kinase. Am JPhysiol 234: H432–H438, 1978

19. Triner L, Vulliemoz Y, Verosky M, Habif DV, Nahas VV: Adenylylcyclase phosphodiesterase system in arterial smooth muscle. Life Sci11: 817–824, 1972

20. Bär HP: Cyclic nucleotides and smooth muscle. Adv Cyclic nucleotideRes 4: 195–237, 1974

21. Feldman AM, Cates AE, Veazey WB, Hershberger RE, Bristow MR,Baughman KL, Baumgartner WA, Dop CV: Increase of the 40,000 molwt pertussis toxin substrate (G-protein) in the failing human heart. JClin Invest 82: 189–197, 1988

65

22. Malbon CC, Rapiejko PJ, Watkins DC: Permissive hormone regula-tion of hormone-sensitive effector systems. Trends Pharmacol Sci 9:33–36, 1988

23. Anand-Srivastava MB: Enhanced expression of inhibitory guaninenucleotide regulatory protein in spontaneously hypertensive rats: Re-lationship to adenylate cyclase inhibition. Biochem J 288: 79–85, 1992

24. Thibault C, Anand-Srivastava MB: Altered expression of G-proteinmRNA in spontaneously hypertensive rats. FEBS Lett 313: 160–164,1992

25. Anand-Srivastava MB, de Champlain J, Thibault C: DOCA-salt hy-pertensive rat hearts exhibit altered expression of G-proteins. Am JHypertens 6: 72–75, 1993

26. Moxham CM, Malbon CC: Insulin action impaired by deficiency ofthe G-protein subunit Gi alpha 2. Nature 379: 840–844, 1996

27. Zheng XL, Guo J, Wang HY, Malbone CC: Expression of constitutivelyactivated Gialpha 2 in vivo ameliorates streptozotocin induced diabe-tes. J Biol Chem 273: 23649–23651, 1998

28. Lynch CJ, Blackmore PF, Johnson EH, Wang RL, Krone PK, ExtonJH: Guanine nucleotide binding regulatory proteins and adenylate cy-clase in livers of streptozotocin- and BB/Wor-diabetic rats. Immuno-detection of Gs and Gi with antisera prepared against synthetic peptides.J Clin Invest 83: 2050–2062, 1989

29. Strassheim D, Palmer T, Houslay MD: Genetically acquired diabe-tes: Adipocyte guanine nucleotide regulatory protein expression andadenylyl cyclase regulation. Biochim Biophys Acta 1096: 121–126,1991

30. Allgayer H, Bachman W, Hepp KD: Increased dose response relation-ship of plasma membrane adenylate cyclase to glucagon stimulationin diabetic rats. A possible role of guanylyl nucleotide-binding regu-latory proteins. Diabetelogia 22: 464–467, 1982

31. Gawler D, Milligan G, Spiegel AM, Unson CG, Houslay MD: Aboli-tion of expression of inhibitory guanine nucleotide regulatory proteinGi activity in diabetes. Nature 327: 229–232, 1987

32. Pilkis SJ, Exton JH, Johnson RA, Park CR: Effects of glucagon oncyclic AMP and carbohydrate metabolism in livers from diabetic rats.Biochim Biophys Acta 20: 250–267, 1974

33. Garber AJ: The impact of streptozoticin-induced diabetes mellitus oncyclic nucleotide regulation of skeletal muscle amino acid metabolismin the rat. J Clin Invest 65: 478–487, 1980

34. Palmer GC, Wilson GL, Chronister RB: Streptozotocin-induced dia-betes produces alterations in adenylate cyclase in rat cerebrum, cer-ebral microvessels and retina. Life Sci 24: 365–374, 1983

35. Marcil J, de Champlain J, Anand-Srivastava MB: Overexpression ofGi-proteins precedes the development of DOCA-salt-induced hyper-tension: Relationship with adenylyl cyclase. Cardiovasc Res 39: 492–505, 1998

36. Salomon Y, Londos C, Rodbell MA: Highly sensitive assay. AnnalBiochem 58: 541–548, 1974

37. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ: Protein measurementwith folin-phenol reagent. J Biol Chem 193: 265–275, 1951

38. Jones DT, Reed RR: Molecular cloning of five GTP-binding proteincDNA species from rat olfactory neuroepithelium. J Biol Chem 262:14241–14249, 1987

39. Anand-Srivastava MB, Srivastava KA, Cantin M: Pertussis toxin at-tenuates atrial natriuretic factor-mediated inhibitin of adneylyl cyclase.Involvement of inhibitory guanine nucleotide regulatory protein. J BiolChem 262: 4931–4934, 1987

40. Anand-Srivastava MB, Sairam MR, Cantin M: Ring-deleted analogs ofatrial factor inhibits adenylyl cyclase/cAMP system: Possible couplingof clearance atrial natriuretic factor receptors to adenylate cyclase/cAMPsignal transduction system. J Biol Chem 265: 8566–8572, 1990

41. Anand-Srivastava MB: Angiotensin II receptors negatively coupled toadenylate cyclase in rat myocardial sarcolemma: Involvement of in-hibitory guanine nucleotide protein. Biochem Pharmacol 38: 489–496,1989

42. Bushfield M, Griffiths SL, Murphy GJ, Pyne NJ, Knowler JT, MilligenG, Parker PJ, Mollner S, Houslay MD: Diabetes-induced alterationsin the expression, functioning and phosphorylation state of the in-hibitory guanine nucleotide regulatory protein Giα-2 in hepatocytes.Biochem J 271: 365–372, 1990

43. Caro JF, Raju MS, Caro M, Lynch CJ, Poulos J, Exton JH, ThakkarJK: Evidence for altered ‘cross-talk’ between the insulin receptor andGi proteins. J Cell Biochem 54: 309–319, 1994

44. Livingstone C, McLellan AR, McGregor M-A, Wilson A, ConnellJMC, Small M, Milligan G, Paterson KR, Houslay MD: Altered G-protein expression and adenylate cyclase activity in platelets of non-insulin-dependent (NIDDM) male subjects. Biochim Biochem Acta1096: 127–133, 1991

45. Hadjiconstantinou M, Ou ZX, Moroi-Fettersse, Neff NH: Apparent lossof Gi protein activity in diabetic retina. Eur J Pharmacol 149: 193–194,1988

46. Weber LP, Macleod KM: Influence of streptozotocin diabetes on thealpha-1 adrenoceptor and associated G-protein in rat arteries. JPharmacol Exp Ther 283: 1469–1478, 1997

47. Hattori Y, Matsuda N, Sato A, Watanuki S, Tomioka H, Kawasaki H,Kanno M: Predominant contribution of the G protein mediated mecha-nism to NaF-induced vascular contractions in diabetic rats: Associa-tion with an increased level of G (q alpha) expression. J Pharmacol ExpTher 292: 761–768

48. Swithers SB, Stewart RE, McCarty R: Binding sites for atrial natriu-retic factor (ANF) in kidneys and adrenal glands of spontaneouslyhypertensive (SHR) rats. Life Sci 1673–1681, 1987

49. Schiffrin EL: Decreased density of binding sites for atrial natriureticpeptide on platelets of patients with severe congestive heart failure. ClinSci 74: 213–218, 1988

50. Strassheim D, Palmer T, Houslay MD: Diabetes abolishes the GTP-dependent, but not the nucleotide regulatory protein expression andadenylate cyclase regulation. Biochim Biophys Acta 1096: 121–126,1991

51. Kowluru A, Kowluru RA, Yamaraki A: Functional alterations of G-pro-teins in diabetic rat retina: A possible explanation for the early visualabnormalities in diabetes mellitus. Diabetologia 35: 624–631, 1992

52. Limas C, Limas CJ: Reduced number of α-adrenergic receptors in themyocardium of spontaneously hypertensive rats. Biochem Biophys ResCommun 83: 710–714, 1978

53. Anand-Srivastava MB: Amiloride interacts with guanine nucleotideregulatory proteins and attenuates the hormonal inhibitin of adenylatecyclase. J Biol Chem 264: 9491–9496, 1989

54. Anand-Srivastava MB: Rat platelets from spontaneously hypertensiverats exhibit decreased expression of inhibitory guanine nucleotide regu-latory protein: Relationship with adenylate cyclase activity. Circ Res73: 1032–1039, 1993

55. Marcil J, Schiffrin EL, Anand-Srivastava MB: Aberrant adenylyl cy-clase/cAMP signal transduction system and G protein levels in plate-lets from hypertensive patients: Improve with antihypertensive drugtherapy. Hypertension 28: 83–90, 1996

56. Anand-Srivastava MB: Downregulation of atrial natriuretic peptideANP-C receptor is associated with alterations in G-protein expressionin A10 smooth muscle cells. Biochemistry 39: 6503–6513, 2000

57. Hildebrandt JD, Hanoune J, Birnbaumer J: Guanine nucleotide inhi-bition of cyc-S49 mouse lymphoma cell membrane adenylyl cyclase.J Biol Chem 257: 14273–14275, 1982

66