J Mol Cell Cardiol 29, 439–448 (1997)

The Role of Gi-proteins and b-Adrenoceptors in the Age-related Declineof Contraction in Guinea-pig VentricularMyocytesNicola Ferrara, Michael Bohm1, Oliver Zolk1, Peter O’Gara andSian E. HardingNational Heart and Lung Institute, Imperial College, Dovehouse Street, London, SW3 6LY, UK and1Klinik III fur Innere Medizin der Universitat der Koln, Koln 50924 Germany

(Received 19 December 1995, accepted in revised form 29 May 1996)

N. F, M. B, O. Z, P. O’G S. E. H, The Role of Gi-proteins and b-Adrenoceptors inthe Age-related Decline of Contraction in Guinea-pig Ventricular Myocytes. Journal of Molecular and CellularCardiology (1997) 29, 439–448. A decline in contractility in myocytes from ageing guinea-pig hearts wasdemonstrated, which is more pronounced for maximum b-adrenoceptor-stimulated activity than contraction inhigh Ca2+. In this study the role of the inhibitory G-proteins (Gi) in this process was investigated. Comparisonswere made between young (Y, <400 g, <4 weeks), adult (A, >600 g, >8 weeks) and senescent guinea pigs (S,53–65 weeks, 1136±30 g). Gia activity, detected by pertussis toxin-catalysed ADP ribosylation, was significantlyincreased in senescent compared to young animals, but immunodetectable levels of Gia were unchanged. b-adrenoceptor number was decreased by 27% in senescent compared with young animals (P<0.002). Pertussistoxin treatment increased the maximum response to isoproterenol in contacting myocytes so that there was nolonger any significant decline with age. Maximum contraction amplitudes (sarcomere length change, lm)with isoproterenol before pertussis toxin were 0.144±0.011 (Y, n=22 animals), 0.104±0.009 (A, 18) and0.098±0.009 (S, 14), P<0.01 by analysis of variance (ANOVA). Following toxin treatment amplitudes were0.140±0.012 (Y, 12), 0.117±0.010 (A, 10) and 0.117±0.018 (S, 8), P=.. Pertussis toxin treatment alsoreversed the effects of ageing on contraction and relaxation velocity in isoproterenol. In contrast, the effect ofage on contraction amplitude or velocity in maximum Ca2+ was more pronounced after toxin treatment. TheEC50 value for isoproterenol increased with age: pertussis treatment decreased the EC50 in each group, but the effectwas especially pronounced for senescent animals. There was no significant difference in the concentration–responsecurves for the negative inotropic effect of adenosine (in the presence of isoprotenerol) between the three agegroups before toxin treatment. All effects of adenosine were abolished after pertussis exposure. We conclude thatincreased Gia activity is likely to contribute to the decreased response to isoproterenol, but not to high Ca2+, inmyocytes from ageing guinea-pigs. 1997 Academic Press Limited

K W: Ageing; Guinea-pig; Cardiomyocyte; G-protein; Adenosine.

and senescence (Ferrara et al., 1995). The al-Introductionterations were detected in the contraction amplitudeand velocity of single myocytes isolated from theRecently, it was shown that there is a progressive

age-related decline in contractility in guinea-pig guinea-pig ventricle, thus precluding effects of tissuefibrosis or hormone/neurotransmitter levels. Con-heart, with changes occurring between young ad-

ulthood and maturity as well as between maturity traction stimulated by the b-adrenoceptor was af-

Please address all correspondence to: S. E. Harding, Cardiac Medicine, National Heart and Lung Institute, Dovehouse Street, LondonSW3 6LY, UK.

0022–2828/97/020439+10 $25.00/0 mc960397 1997 Academic Press Limited

N. Ferrara et al.440

fected to a greater extent than that by high Ca2+ renoceptor involvement. The aim was to discoverwhether treatment with pertussis toxin would beconcentrations, although both were significantly

reduced. In this the senescent guinea-pig resembles accompanied by a general improvement in theresponse to high Ca2+ in the adult guinea-pigs.the ageing human, where myocardial responses to

sympathetic stimulation are more strongly reduced Myocytes have the advantage that with well-washed, continuously superfused cells, the like-than the basal contractile parameters (Rodeheffer

et al., 1984). In single cells from failing human lihood of catecholamine contamination is muchreduced compared with muscle strips, and it isheart, both maximum responses to isoproterenol

(Harding et al., 1992) and relaxation velocity in possible to be reasonably certain that there is noresidual b-adrenoceptor stimulation in the coursehigh calcium (Harding et al., del Monte et al., 1995)

declined significantly with the age of the patient, of the experiment.but only the response to isoproterenol emerged asa significant independent variable when using amultiple regression analysis which included disease Materials and Methodstype and severity. For myocytes from non-failinghuman ventricle, a correlation was found between

Age of guinea-pigsmaximum response to isoproterenol (but not Ca2+)and the age of the subject (Davies et al., 1995).

Myocytes were isolated from young (<400 g, <4It was also found that age-related decreases inweeks) or adult (>600 g, >8 weeks) Dunkin–b-adrenoceptor response on rat heart correlate withHartley guinea-pigs. The body weight of adult, non-increases in inhibitory guanine-nucleotide bindingsenesecent animals was taken as a marker of age,proteins (Gi-proteins) (Bohm et al., 1993a). Increasebecause there is a linear relationship betweenin Gi may be a common mechanism whereby b-weight in grams and age from the birth to 15thadrenoceptor sensitivity is reduced in many situ-week of life in male guinea-pigs (Ferrara et al.,ations, including ageing (Bohm et al., 1993a), heart1995). The body weight was measured in eachfailure (Feldman et al., 1988; Brown and Harding,animal on the day of the experiment immediately1992; Bohm et al., 1994a), hypertension (Bohm etprior to death. A group of senescent animals wasal., 1992) and septic shock (Bohm et al., 1995).also studied; these were aged between 53 and 65The first aim of this study was to investigate b-weeks and weighed an average of 1136±30 g (n=adrenoceptor and Gi levels in the ageing guinea-22). Of the animals described, 18 of the young, 12pig, and then attempt to reverse age-related b-of the adult and nine of the senescent were notadrenoceptor desensitization in myocytes fromincluded in the previous work (Ferrara et al., 1995).guinea-pigs by treatment with pertussis toxin,

which inactivates Gi. It has been shown that ex-posure to pertussis toxin of myocytes from failinghuman heart or from hearts of noradrenaline- Preparation of isolated myocytestreated guinea-pigs reverses the existing b-ad-renoceptor desensitization in a matter of hours Isolated myocytes were prepared from the ventricle

of male guinea-pigs by enzymatic dissociation as(Brown and Harding, 1992). This experiment re-quires the use of isolated cells, since penetration of previously described (Brown and Harding, 1992).

The hearts were mounted on a Langendorff ap-pertussis toxin into tissue strips is poor.The second aim was to examine the effects of paratus and equilibrated with Krebs–Henseleit (KH)

solution of the following composition (in mmol/l):pertussis toxin treatment on contractile responsesin the absence of b-adrenoceptor stimulation. It has NaCl 119, KCl 4.2, MgSO4 0.94, KH2PO4 1.2,

NaHCO3 25, glucose 11.5, bubbled with 95% O2/been shown that low b-adrenoceptor responses infailing human or desensitised guinea-pig heart are 5% CO2, pH 7.4 containing 1 mmol/l calcium for

5 min at 37°C. Low calcium medium (in mmol/l):accompanied by low basal (unstimulated) cyclicAMP levels (Feldman et al., 1988; Danielson et al., NaCl 120, KCl 5.4, MgSO4 5, pyruvate 5, glucose

20, taurine 20, HEPES 10, bubbled with 100% 02,1989; von der Leyen et al., 1991; Bohm et al.,1994b; Harding et al., 1994), although there is pH 6.95, containing 12–15 lmol/l calcium was

then perfused through for 5 min at 37°C. This wascontroversy as to whether this occurs in vivo (Regitz-Zagrosek et al., 1994). The further possibility exists followed by a 2 min perfusion with the same low

calcium solution with 200 lmol/l calcium and 4 U/that the reduction of a tonically stimulating levelof basal cyclic AMP could produce a decline in ml (Sigma) type XXIV protease added. The solution

was then changed to one containing no addedcontraction parameters in the absence of b-ad-

Gi-proteins and Age-related b-Desensitization 441

protease but 0.3 mg/ml (Worthington) collagenase sarcomere length. Contraction and relaxation ve-locities are not shown for contraction in 1 mmol/land 0.6 mg/ml (Sigma) hyaluronidase for a further

10 min. The ventricles were then carefully removed Ca2+ due to the limitations of measurement ofvelocity at low contraction amplitudes.from the cannula, chopped and incubated in the

same enzyme solution for 2×4 min. At the endof each 4-min period the solution containing thedispersed cells was filtered through a 300 lm gauze Response to calcium, isoproterenol and adenosineand centrifuged at 40×g for 1–2 min. The cellswere washed twice by centrifugation in low calcium Calcium and isoproterenol concentration–response

curves were constructed in a cumulative manner.solution and resuspended in the same medium.Myocytes were incubated with pertussis toxin A maximum was judged to have been reached, for

either calcium or isoproterenol, when there was noimmediately after digestion. The incubation timeperiod was 2 h and the concentration of the toxin increase in contraction amplitude with increasing

concentrations, or when toxic signs such as phasicrequired to inactivate the inhibitory G-proteins was0.5 lg/ml. The incubation of the cells with the contractions or decreases in diastolic length oc-

curred. Myocytes which did not fully recover thepertussis toxin was carried out in the low Ca2+

medium at 35°C in order to accelerate the reaction. basal state following washout were excluded fromthe analysis.Comparisons were made with myocytes which had

been preincubated at 35°C in the absence of per- Concentration–response curves for adenosinewere cumulative, and were performed in the pres-tussis toxin for an equal time to those pretreated

with pertussis toxin. ence of concentrations of isoproterenol giving in-creases in amplitude that were 50–75% ofFor measurement of b-adrenoceptor and G-pro-

tein levels, myocytes were treated further to ensure maximum for that cell. The response was calculatedas the decrease in isoproterenol-stimulated con-the removal of non-myocyte cells from the pre-

paration. Cells were suspended in low Ca2+medium, traction amplitude, with basal contraction sub-tracted. For basal and maximum isoproterenol-layered on 10% Percoll in KH medium and cent-

rifuged at 40×g for 2 min. The myocyte pellet stimulated contraction, averaged values from beforeand after the adenosine exposure were used forwas resuspended and centrifuged through a second

Percoll gradient, after which the myocytes were the calculation, to compensate for any decrease incontractility over the duration of the experiment.washed in low Ca2+ medium, frozen in liquid N2

and stored at −80°C until required.

b-adrenoceptor bindingMeasurement of contraction

Membranes were incubated with [125I] cy-anopindolol (2000 Ci/mmol) for 1 h at 37°C in aCells in KH solution were placed in a perspex

chamber on the stage of a Zeiss IM inverted micro- total volume of 250 ll incubation buffer (50 mmol/lTris–HCl and 10 mmol/l MgCl2, pH 7.4). The re-scope and superfused with KH containing 1 mmol/l

Ca2+ at 2 ml/min and 32°C. A cell was selected action was terminated by rapid filtration throughWhatman GF/C filters, which were rapidly washedusing the following criteria: rod shaped, without

sarcolemmal blebs, no spontaneous contractions, three times with 6 ml ice cold incubation buffer.Non-specific binding was defined with 3 lmol/l pro-stable baseline contraction to electrical contraction

at 0.5 Hz and sarcomere length no shorter than pranolol.1.67 lm. The image of the cell was displayed on aTV monitor and the length change measured witha video motion detector. Sarcomere length was Measurement of Gia activity by pertussis toxin-induced

[32P]ADP-ribosylationmeasured at high magnification by counting thenumber of sarcomeres (>30) crossing a given por-

[32P]ADP-ribosylation of Gia by pertussis toxin wastion of the TV screen: it was assumed that sarcomerelength did not vary significantly along the length performed for 12 h at 4°C in a volume of 50 ll

containing 100 mmol/l Tris–HCl, pH 8.0 at 20°C,of the cell, and that the number of sarcomeres ina given cell did not vary during an experiment. 25 mmol/l dithiothreitol, 2 mmol/l ATP, 1 mmol/l

GTP, 50 nmol/l [32P]NAD (800 Ci/mmol) andContraction amplitude and velocity is expressed aschange in sarcomere length, calculated from the 20 lg/ml pertussis toxin that had been activated

by incubation with 50 nmol/l dithiothreitol for 1 hchange in length of the myocyte and its original

N. Ferrara et al.442

at 20°C prior to the labelling reaction as described Resultsearlier (Bohm et al., 1991). Samples were subjectedto sodium dodecyl sulfate polyacrylamide gel Myocyte contraction and age, control conditionselectrophoresis (SDS-PAGE) (10%) (w/v) acrylamide16 cm total gel length) as described by Lammli Contraction amplitude in the basal Ca2+ con-(1970). Gels were stained with Coomassie blue centration of 1 mmol/l was not altered with ageand dried before autoradiography was performed. (Table 1). However, there was a strong trend forProtein was determined according to Lowry et al. contraction amplitude and velocity to decrease with(1951) using bovine serum albumin as standard. age under conditions of maximal stimulation with

high Ca2+ or isoproterenol (Table 1). In this studythe effect in Ca2+ achieved significance for con-traction velocity but not amplitude. Pairwise com-parison of means showed that the significantImmunoblotting techniqueschange occurred in senescent animals comparedwith young. There was no difference in EC50 valuesImmunoblotting techniques were performed as pre-for Ca2+ between age groups.viously described (Gierschik et al., 1985). The poly-

The effect was more pronounced for iso-clonal antiserum (MB1) was raised in rabbitsproterenol, with significant decreases in both con-against the terminal decapeptide of retinal trans-traction amplitude and velocity, and a significantducin a (KENLKDCGLF) coupled to keyhole limpetincrease in the EC50 for the concentration–responsehaemocyanin as described (Goldsmith et al., 1987).curve (Table 1). The main effect on contractionThe antiserum recognized Gia1 and Gia2 (notoccurred between the young and adult groups,shown). Blots were stained with an alkaline phos-with little further change in senescence, while thephatase-labelled goat anti IgG antiserum.increase in EC50 occurred primarily during the pro-gression from adult to senescent. This is dem-onstrated in the concentration–response curves toisoproterenol [Fig. 1(a)]. Pairwise comparison of

Statistics each curve with the other by ANCOVA showedsignificant differences for young v senescent

Results from up to four myocytes from one animal (P<0.001), old v senescent (P<0.001) and a lowerwere pooled, and the data shown are averaged significance level for young v adult (P<0.05). Inresults (mean±...) where n=animals. Con- each case the covariance with concentration oftraction characteristics of young, adult and sen- isoproterenol was highly significant (P<0.001).escent animals were compared using one-wayanalysis of variance (ANOVA) with Tukey’s test forcomparison of adjacent means. Concentration–response curves were compared pairwise using Myocyte contraction and age, pertussis toxin treatmentanalysis of covariance (ANCOVA), where the con-centration of isoproterenol was the covariate. EC50 Inactivation of Gi by pertussis toxin was judged tovalues for isoproterenol were calculated using an be complete when there was no negative inotropiciterative curve fitting program for experiments effect of 10 lmol/l adenosine on the isoproterenol-where inotropic responses were obtained over at stimulated contraction. Control concentration–least 2 log units of concentration, and where the response curves to adenosine are shown in FigureHill coefficient was 0.8–1.2. 2, for young, adult and senescent animals. There

was no difference in effect of adenosine betweengroups, with 10 lmol/l adenosine giving markeddecreases in amplitude for all three prior to toxintreatment.Materials

Table 2 shows the same contractile parametersas Table 1 for cells treated with pertussis toxin.Analar water (BDH, Poole, Dorset, UK) was used

for all solutions. Salts (AnalaR or Aristar grade) In high Ca2+ conditions, the decline with age incontraction amplitude and contraction and re-were also from BDH. (−)-Isoproterenol hydro-

chloride and adenosine were obtained from Sigma laxation velocities was more marked than beforetreatment, reaching statistical significance in eachand pertussis toxin was obtained from Speywood

Pharmaceuticals Ltd. case. The magnitude of the effect was not greatly

Gi-proteins and Age-related b-Desensitization 443

Table 1 Young (<400 g), adult (>600 g) and senescent guinea-pigs, control conditions

Condition Sarcomere length Young Adult Senescent ANOVAchange

1 m Ca2+ Contraction amplitude 0.045±0.006 (24) 0.042±0.005 (21) 0.043±0.004 (14) ..(lm)

Max Ca2+ Contraction amplitude 0.172±0.014 0.149±0.012 0.125±0.011 ..(lm) (22) (19) (8)Contraction velocity 2.02±0.22 1.51±0.16 1.18±0.14∗ P<0.05(lm/s)Relaxation velocity 1.67±0.23 1.45±0.18 1.05±0.12 ..(lm/s)

Ca2+ EC50, m 2.33±0.15 (22) 2.46±0.31 (19) 2.17±0.21 (8) ..Max iso Contraction amplitude 0.144±0.011 0.104±0.009∗ 0.098±0.009∗ P<0.01

(lm) (22) (18) (14)Contraction velocity 2.00±0.22 1.21±0.11∗∗ 1.24±0.16∗ P<0.005(lm/s)Relaxation velocity 1.48±0.20 1.14±0.11 1.00±0.12 ..(lm/s)

Iso EC50, n 0.95±0.21 (11) 1.41±0.48 (7) 8.66±3.09 (12) P<0.05

Significantly different from young ∗P<0.05, ∗∗P<0.01.

increased: for contraction amplitude the senescent In previous work, it was found that incubationat 35°C without pertussis toxin was able partiallyvalue was 73% of the young before toxin treatment

and 62% after. Pairwise comparisons showed that to restore responses to isoproterenol in desensitisedmyocytes (Brown and Harding, 1992). Therefore,in each case the main change occurred between

the adult and senescent groups, with less of a myocytes which had been incubated in this wayfor 2 h were compared with those which haddifference between the young and adult. There was

no difference in EC50 for Ca2+ between young, adult been kept at room temperature and used within1 h of preparation. Incubation at 35°C reproducedand senescent animals after toxin treatment, nor

any significant difference between untreated and the slight increase in difference between myocytesfrom the three age groups at high Ca2+ duringtreated values for any group.

In contrast, for the effect of isoproterenol the toxin treatment (contraction amplitude, lmchange in sarcomere length, young, 0.168±0.015difference between age groups was less evident

after toxin treatment. This was mainly due to an (n=10); adult, 0.125±0.010 (n=10); senescent,0.111±0.009 (n=9), P<0.01). However, theincreased maximum response in the adult and sen-

escent groups, with little effect on the young (Table incubation at 35°C did not mimic the reversalof isoproterenol subsensitivity by pertussis toxin2). For contraction amplitude the senescent value

was 68% of the young before toxin treatment and treatment. It is therefore likely that the increaseddifference between groups for contraction para-84% after.

Pertussis treatment decreased the EC50 iso- meters in high Ca2+ is due to the incubationconditions, but the effect of pertussis toxin onproterenol for each age group, and the leftward

shift in the curves is evident in Figure 1. This responses to isoproterenol is specific.decrease in EC50 reached statistical significance forthe adult (P<0.05) and senescent (P<0.01) an-imals. EC50 values for cells from young animals arenot included in Table 2 because it was difficult toobtain full concentration–response curves to iso- b-adrenoceptor numberproterenol after pertussis toxin treatment. Evenwhere inotropic effects were obtained over 2 log Bmax for b-adrenoceptor binding was significantly

lower in myocytes from senescent animals com-units of concentration the Hill coefficient was oftenoutside the prescribed limits. Partial concentration– pared to young (16.1±1.3 cf. 22.0±0.9 fmol I-

cyanopindolol bound/mg protein, P<0.002, n=response curves from the young group are includedin Figure 1, and suggest that pertussis toxin in- 10 for both groups), indicating a reduction in b-

adrenoceptor number in the senescent group.creased sensitivity in this group also.

N. Ferrara et al.444

% in

itia

l val

ue

5

100

–10

–log [adenosine] mol/l

1020

0

7 6

30405060708090

Figure 2 Averaged concentration–response curves foradenosine, calculated as described in the Materials andMethods section, for young (n=10) (—Ε—), adult (n=7) (—Μ—), and senescent (n=7) (– –Ν– –) guinea pigs.

creases in myocyte response to both high Ca2+ andisoproterenol with increasing age in the guinea-pig(Ferrara et al., 1995). At least part of the decreasein the effect of isoproterenol is likely to be due tothe reduction in b-adrenoceptor number, whichwas 27% lower in senescent than young animals.No other data are available from guinea-pig hearts,but studies in rat have been inconclusive, with

7

100

00

% m

axim

um

ch

ange

50

75

25

11 10 9 8

(a)

Y

A

S

7

100

00

–log [Isoproterenol] mol/l

50

75

25

11 10 9 8

(b)

A

S

Y

both decreased (Gudmundsdottir et al., 1991) orunchanged (Bazan et al., 1994) b-adrenoceptorFigure 1 (a) averaged concentration–response curvesnumbers reported. A decrease in b-adrenoceptor(% maximum change) to isoproterenol in control myo-

cytes from young [Y, n=18, (Ε), solid lines], adult [A, number has been seen in ageing human ventriclen=17, (Μ), dashed lines] and senescent [S, n=13, (Ν), (White et al., 1994), but not in atria (Brodde et al.,dotted lines] guinea-pigs. (b) concentration–response 1995).curves following pertussis toxin treatment in myocytes

An additional mechanism for b-adrenoceptor de-from young (Y, n=6, (Ε), solid lines) adult (A, n=10,sensitisation, involving increased Gia with age, is(Μ), dashed lines) and senescent (S, n=8, (Ν), dotted

lines) guinea-pigs. indicated both by the increase in pertussis toxin-catalysed ADP-ribosylation and by the ability ofpertussis toxin treatment to reverse the sub-Gi protein levels and ADP-ribosylationsensitivity to isoproterenol. This resembles the au-thors’ findings in ageing rat heart, and suggestsComparisons were made between the young andthat increases in Gia might be a general mechanismsenescent groups for Gi levels by immunoblot, andfor decreasing responses to b-adrenoceptor stimu-for pertussis catalysed ADP-ribosylation. These ex-lation with age (Bohm et al., 1993a). It is alsoperiments were done in Percoll-gradient separated,similar to findings in failing human heart, wherewell-washed myocyte preparations, removing thepertussis catalysed ADP-ribosylation levels were in-possibility of contamination by non-myocyte cellcreased (Bohm et al., 1990), and b-adrenoceptortypes. There was a significant increase in ADP-desensitization could be reversed by incubation ofribosylation in the myocytes between young andmyocytes with pertussis toxin (Brown and Harding,senescent animals (Fig. 3) but no change in the1992). It is notable that pertussis toxin treatmentamount of immunoreactive Gia (Fig. 4).did not completely restore the maximum responseto isoproterenol, or produce superimposable con-centration–response curves for each group. TheDiscussionremaining differences between groups may beascribable to the fall in b-adrenoceptor number or,The present study confirms the general findings of

the authors’ previous paper, which showed de- as in failing heart, there may be multiple additional

Gi-proteins and Age-related b-Desensitization 445

Table 1 Young (<400 g), adult (>600 g) and senescent guinea-pigs, after pertussis toxin treatment

Condition Sarcomere length Young Adult Senescent ANOVAchange

1 m Ca2+ Contraction amplitude 0.055±0.007 (14) 0.063±0.010 (10) 0.038±0.012 (9) ..(lm)

Max Ca2+ Contraction amplitude 0.159±0.010 0.149±0.011 0.099±0.011∗‡ P=0.001(lm) (11) (9)Contraction velocity 2.06±0.15 1.90±0.27 1.11±0.09∗‡ P=0.005(lm/s)Relaxation velocity 1.60±0.15 1.63±0.27 0.91±0.08∗† P<0.05(lm/s)

Ca2+ EC50, m 1.92±0.25 (11) 1.57±0.26 (9) 2.65±0.61 ..Max iso Contraction amplitude 0.140±0.021 0.117±0.010 0.117±0.018 ..

(lm) (12) (10) (8)Contraction velocity 2.00±0.21 1.45±0.16 1.38±0.22 ..(lm/s)Relaxation velocity 1.74±0.19 1.35±0.18 1.26±0.24 ..(lm/s)

Iso EC50, n — 0.374±0.169 (6) 0.074±0.019 (7) ..

Significantly different from adult ∗P<0.05, significantly different from young †P<0.05, ‡P<0.01.

1500

0Young

guinea pig

Ct/

min

Senescentguinea pig

1200

900

600

300

* P<0.05 v SP

40 kDa

Transducin

Senescent

Senescent

Young

Young

Figure 3 Pertussis-catalysed [32P]ADP-ribosylation of G-protein a-subunits (>40 kDa) in membrane preparations ofmyocytes from young (n=7) and senescent (n=8) guinea-pigs. Membranes were treated with pertussis toxin plus[32P]NAD as described in the Materials and Methods section and separated by SDS-PAGE before autoradiography.Representative autoradiographs from two young and two senescent hearts are shown, together with averaged data.

N. Ferrara et al.446

1500

0Young

guinea pig

Arb

itra

ry u

nit

s

Senescentguinea pig

1200

900

600

300

40 kDa

Senescent

Senescent

Young

Young

Figure 4 Immunochemical detection (Western blots) of Gia in membrane preparations from the myocytes of sevenyoung and eight senescent guinea-pig hearts. Representative immunoblots from two young and two senescent heartsare shown, together with averaged data.

changes that occur in ageing myocardium. As with The mechanism of an increase of pertussis toxinsubstrates is likely mediated by chronic b-ad-failing human heart, the level of immunoreactive

Gia did not increase in line with that of pertussis renoceptor stimulation. In neonatal rat car-diomyocytes incubated with norepinephrinetoxin substrate (Feldman et al., 1991). It was sug-

gested that a post-translational modification sim- (Reithmann et al., 1990) and in rats treated withisoprenaline in vivo (Eschenhagen et al., 1992),ultaneously increased Gi activity and facilitated

more complete ADP-ribosylation during heart fail- the increase of pertussis toxin substrates has beenobserved to be mediated by b-adrenoceptor stimu-ure.

Pertussis toxin treatment failed to reverse the lation. Muller et al. (1993) observed an increasedtranscription rate of the Gia2 gene in the heartdecreased response to high Ca2+ that was seen with

increasing age. The differences were in fact slightly following treatment of rats with isoprenaline. InS49 cyc− (genetically lacking Gsa) and in S49greater following toxin treatment, although this is

likely to be a result of the incubation because it kin− (genetically deficient in the cAMP-dependentprotein kinase) cells, unlike in wild-type S49 mousewas mimicked by keeping the cells at 35°C for 2 h

without toxin. This result shows that the reversal lymphoma cells, pertussis toxin substrates werenot increased following b-adrenoceptor stimulationof b-adrenoceptor desensitization was unlikely to

be due to a general increase in contractility of the (Clark et al., 1988). These observations show thatenhancement of pertussis toxin substrates involvescells following exposure to pertussis. It also indicates

that the decreased contraction in high Ca2+ is b-adrenoceptor stimulation and requires intact b-adrenoceptor coupling as well as an intact cAMP-probably unrelated to low basal cyclic AMP in the

cells from ageing heart. Pertussis toxin treatment dependent phosphorylation cascade. In this respectit is interesting that in ageing individuals (Zieglerhas been shown to reverse the decrease in basal

adenylate cyclase activity in membranes from fail- et al., 1976) as well as in adult rats (Galbo etal., 1977) sympathetic activation as judged froming heart (Feldman et al., 1988), and therefore

might have been expected to have a similar effect circulating norepinephrine levels occurs. These ob-servations are in favour of the notion that sym-in the senescent myocytes in the present study.

Gi-proteins and Age-related b-Desensitization 447

pathetic activation with consequent stimulation of receptors have been observed to be coupled pre-dominantly through Gia3, while in the heart Gia2the b-adrenoceptor–adenylyl cyclase system could

also be involved in the increase of pertussis toxin is the main subtype (Bohm et al., 1990; 1994a).In conclusion, the present study provides strongsubstrates in cardiomyocytes from ageing guinea-

pigs. However, the 35°C incubation in the absence evidence for a role of decreased b-adrenoceptornumber and increased Gi in the b-adrenoceptorof toxin was unable to reverse b-adrenoceptor de-

sensitisation to any appreciable extent. In myocytes subsensitivity associated with ageing in the guinea-pig. The evidence supports the hypothesis that anfrom both failing human heart and from nor-

adrenaline-treated guinea-pigs the 35°C incubation increase in inhibitory guanine nucleotide bindingproteins may provide a common pathway for b-alone was sufficient to produce some reversal of b-

adrenoceptor subsensitivity, although the effect was adrenoceptor desensitization under a number ofphysiological and pathological conditions. The gen-significantly less than in the presence of pertussis

toxin (Brown and Harding, 1992). This was at- eral decline in contraction with age in the absenceof b-adrenoceptor stimulation is not Gi-mediated.tributed to the removal of the cells from the in vivo

environment, where norepinephrine levels werehigh, and the consequent reversal of the de-sensitization process. The lack of reversal of de-

Acknowledgementssensitization by 35°C incubation in the myocytesfrom the adult and senescent guinea-pigs may

Michael Bohm is supported by Deutsche For-therefore argue against a role of high circulatingschungsgemeinschaft (Heisenberg and Gerhard-norepinephrine levels in ageing.Hess programs). Nicola Ferrara is a Research FellowThere was no alteration in the ability of adenosinefrom Cattedra di Geriatrica, University of Naplesto decrease the response to isoproterenol (anti-Federico II (Italy).adrenergic effect) in the myocytes from the adult or

senescent guinea-pig hearts: neither the maximumeffect of adenosine nor its EC50 was changed. Thepattern is similar to that seen in failing human Referencesheart, where the concentration–response curvesfor adenosine are unaltered despite a significant B A, V D V E, F N, 1994. Effect of

age on beta-receptors, Gs alpha- and Gi alpha-proteinsincrease in Gi and pertussis toxin-catalysed ADP-in rat heart. Biochem Pharmacol 48: 479–486.ribosylation (Bohm et al., 1993b). This is surprising,

B M, G P, J KH, P B, S P,because the adenosine A1-receptors which mediateU M, E E, 1990. Increase of Gi alpha

this effect are coupled through Gi and their effects in human hearts with dilated but not ischemic cardio-are blocked by the treatment with pertussis toxin. myopathy. Circulation 82: 1249–1265.

B M, L K, E E, C M, J K,However, it should be remembered that the con-G P, 1991. Failure of [32P]ADP-ribosylationcentration curves to adenosine are carried out atby pertussis toxin to determine Gi alpha content inthe level of b-adrenoceptor stimulation that givesmembranes from various human tissues. Improved

an effect between 50 and 75% of maximum. This radioimmunological quantification using the 125I-is likely to occur at a higher isoproterenol con- labelled C-terminal decapeptide of retinal transducin.

Biochem J 277: 223–229.centration in myocytes from senescent guinea-pigsB M, G P, K A, L K, W(or failing human hearts) than in their respective

K, E E, 1992. Desensitization of adenylatecontrols. It is well-known that the anti-adrenergiccyclase and increase of Gi alpha in cardiac hypertrophy

effect of adenosine is less evident at high iso- due to acquired hypertension. Hypertension 20: 103–proterenol levels. The increased sensitivity to ad- 112.

B M, D H, H P, L H, P D, E-enosine because of increased Gi-coupling might E, 1993a. Effects of exercise on myocardialtherefore be offset by a decreased response due toadenylate cyclase and Gi alpha expression in sen-the higher levels of isoproterenol that it is requiredescence. Am J Physiol 264: H805–H814.

to oppose. It should also be noted that Gia protein B M, U M, E E, 1993b. Adenosinelevels were not increased: the change in bioactivity receptors in the human heart: pharmacological char-

acterisation in nondiseased and cardiomyopathic tis-that causes an increase in the ADP-ribosylated statesue. Drug Dev Res 28: 268–276.may not itself alter coupling of the A1 receptor.

B M, E T, G P, L K, LAlternatively, the A1-receptor may be coupled toH, M U, S W, S P, S H, S-

an additional Gia subtype in the heart as well as M, E E, 1994a. Radioimmunochemicalthe one that mediates inhibition of adenylyl cyclase. quantification of Gia in right and left ventricles from

patients with ischemic and dilated cardiomyopathyIn this respect it is interesting that A1-adenosine

N. Ferrara et al.448

and predominant left ventricular failure. J Mol Cell G P, C J, S C, B L, SAM, 1985. Antisera against a guanine nucleotideCardiol 26: 133–149.

B M, R B, S RHG, E E, 1994b. binding protein from retina cross-react with the b-subunit of the adenylyl cyclase-associated guaninecAMP concentrations, cAMP dependent protein kinase

activity, and phospholamban in non-failing and failing nucleotide binding protein, Ns and Ni. Proc Natl AcadSci USA 82: 727–731.myocardium. Cardiovasc Res 28: 1713–1719.

B M, K R, G P, E E, 1995. G P, G P, M G, U CG, V- R, M HI, S AM, 1987. AntibodiesIncrease of myocardial inhibitory G-proteins in ca-

techolamine-refractory septic shock or in septic multi- direct against synthetic peptides distinguish betweenGTP-binding proteins in neutrophils and brain. J Biolorgan failure. Am J Med 98: 183–186.

B OE, Z HR, S D, B-S A, Chem 262: 14683–14688.G E, B VE, GM-R M, S-B E, P

JA, O H, 1995. Age-dependent changes in the S, 1991. Combined effects of age and dietary fat onbeta 1-receptors and Ca2+ channels in rat hearts. Ambeta-adrenoceptor-G-protein(s)-adenylyl cyclase sys-

tem in human right atrium. J Cardiovasc Pharmacol 26: J Physiol 260: H66–H72.H SE, J SM, O’G P, D M F, V G,20–26.

B LA, H SE, 1992. The effect of pertussis P-W PA, 1992. Isolated ventricular myocytesfrom failing and non-failing human heart; the relationtoxin on b-adrenoceptor responses in isolated cardiac

myocytes from noradrenaline-treated guinea-pigs and of age and clinical status of patients to isoproterenolresponse. J Mol Cell Cardiol 24: 549–564.patients with cardiac failure. Br J Pharmacol 106:

115–122. H SE, D CH, W DG, P-W PA,1994. Contractile characteristics and response to agon-C RB, K MW, F J, G TJ, J JA,

1988. Activation of cAMP-dependent protein kinase is ists in myocytes from failing human heart. Eur HeartJ 15 (Suppl D): 35–36.required for heterologous desensitization of adenylyl

cyclase in S49 wild-type lymphoma cells. Proc Natl L UK, 1970. Cleavage of structural proteins duringthe assembly of the head of bacteriophage T4. NatureAcad Sci USA 85: 1442–1446.

D W, L H, M W, N J, 227: 680–685.L OH, R NJ, F AL, R RJ,S W, S H, S J, S B, D

V, K P, 1989. Basal and isoprenaline-stimulated 1951. Protein measurement with the Folin phenolreagent. J Biol Chem 193: 265–270.cAMP content in failing versus nonfailing human car-

diac preparations. J Cardiovasc Pharmacol 14: 171–173. M FU, B KR, E T, S W,S H, 1993. Isoprenaline stimulates gene tran-D CH, F N, H SE, 1995. Beta-ad-

renoceptor function changes with age of subject in scription of the inhibitory G protein alpha-subunit Gialpha-2 in rat heart. Circ Res 72: 696–700.myocytes from non-failing ventricle. Cardiovasc Res 31:

152–156. R-Z V, H R, S C, HA, F E, 1994. Myocardial cyclic AMP and nor-D M F, O’G P, P-W PA, Y MH,

H SE, 1995. Cell geometry and contractile ab- epinephrine content in human heart failure. Eur HeartJ 15 (Suppl. D): 7–13.normalities of myocytes from failing human left vent-

ricle. Cardiovasc Res 30: 281–290. R C, G P, M U, W K, JKH, 1990. Pseudomonas exotoxin A presents beta-E T, M U, D M, N M, S

W, S H, S AM, E J, W A, adrenoceptor-induced upregulation of Gi protein alpha-subunits and adenylyl cyclase desensitization in ratS H, 1992. Long term beta-adrenoceptor-me-

diated up-regulation of Gi alpha and G(o) alpha mRNA heart muscle cells. Mol Pharmacol 37: 631–638.R RJ, G G, B LC, F JL,levels and pertussis toxin-sensitive guanine nucleotide-

binding proteins in rat heart. Mol Pharmacol 42: 773– W ML, L EG, 1984. Exercise cardiacoutput is maintained with advancing age in healthy783.

F AM, C AE, V WB, H RE, human subjects: cardiac dilatation and increased strokevolume compensate for a diminished heart rate. Cir-B MR, B KL, B WA, V

D C, 1988. Increase of the 40,000-mol wt pertussis culation 69: 203–213. L H, M U, M W, N J,toxin substrate (G protein) in the failing human heart.

J Clin Invest 82: 189–197. N M, S W, S H, S J, S B,W H, D V, K P, H A, 1991.F AM, J DG, B MR, C AE, V

D C, 1991. Immunodetectable levels of the inhibitory Mechanism underlying the reduced positive inotropiceffects of the phosphodiesterase III inhibitors pi-guanine neucleotide-binding regulatory proteins in fail-

ing human heart: discordance with measurements of mobendan, adibendan and saterinone in failing ascompared to nonfailing human cardiac muscle pre-adenylate cyclase activity and levels of pertussis toxin

substrate. J Mol Cell Cardiol 23: 439–452. parations. Naunyn Schmiedebergs Arch Pharmacol 344:90–100.F N, O’G P, W DG, B LA, D M

F, P-W PA, H SE, 1995. Decreased W M, R R, M W, K MF, L P,W M, P JD, A F, C D,contractile responses to isoproterenol in isolated car-

diac myocytes from ageing guinea-pigs. J Mol Cell F AM, et al., 1994. Age-related changes inbeta-adrenergic neuroeffector systems in the humanCardiol 27: 1141–1150.

G H, R EA, H JJ, C NJ, 1977. heart. Circulation 90: 1225–1238.Z MG, L CR, K IJ, 1976. Plasma nor-Diminished hormonal responses to exercise in trained

rats. J Appl Physiol 43: 953–958. adrenaline increases with age. Nature 261: 333–335.