Cardiovascular Research 47 (2000) 329–335
www.elsevier.com / locate / cardiores
www.elsevier.nl / locate / cardiores
HMG CoA reductase inhibition reduces sarcolemmal Na 1 –K 1 pump
density
David F. Gray a,b , Henning Bundgaard c , Peter S. Hansen a,b , Kerrie A. Buhagiar a,b ,
a
d
c
a,b ,
Anastasia S. Mihailidou , Wendy Jessup , Keld Kjeldsen , Helge H. Rasmussen *
a
Department of Cardiology, Royal North Shore Hospital, Sydney, Australia
b
The University of Sydney, Sydney, Australia
c
Department of Medicine, The Heart Centre, Rigshospitalet, National University Hospital, Copenhagen, Denmark
d
The Heart Research Institute, Sydney, Australia
Received 7 October 1999; accepted 4 April 2000
Abstract
Objectives: HMG CoA reductase inhibitors reduce cellular availability of mevalonate, a precursor in cholesterol synthesis. Since the
cholesterol content of cell membranes is an important determinant of Na 1 –K 1 pump function we speculated that treatment with HMG
CoA reductase inhibitors affects Na 1 –K 1 pump activity. Methods: We treated rabbits and rats for 2 weeks with the HMG CoA reductase
inhibitor lovastatin and measured Na 1 –K 1 pump current (Ip ) in isolated rabbit cardiac myocytes using the whole cell patch-clamp
technique, K-dependent p-nitrophenyl phosphatase (p-NPPase) activity in crude myocardial and skeletal muscle homogenates, and
vanadate-facilitated 3 H-ouabain binding in intact skeletal muscle samples from rats. Results: Treatment with lovastatin caused statistically
significant reductions in Ip , myocardial and skeletal muscle K-dependent p-NPPase activity and 3 H-ouabain binding in the myocardium
and skeletal muscle. The lovastatin-induced decrease in Ip was eliminated by parenteral co-administration of mevalonate. However, this
was not related to cardiac cholesterol content. Conclusions: Treatment with lovastatin reduces Na 1 –K 1 pump activity and abundance in
rabbit and rat sarcolemma.  2000 Elsevier Science B.V. All rights reserved.
Keywords: Cell culture / isolation; Cholesterol; Lipid metabolism; Membrane transport; Na / K-Pump
1. Introduction
Modification of the cholesterol content of cell membranes in vitro alters Na 1 –K 1 pump function (see Refs.
[2] and [26] for reviews). A possible physiological relevance of this is suggested by the finding that a modest
diet-induced increase in serum cholesterol above control
levels stimulates pump activity [9]. Since dietary cholesterol supplementation can enhance Na 1 –K 1 pump function it is reasonable to think that a decrease in cellular
cholesterol induced by inhibition of endogenous cholesterol synthesis may reduce pump activity.
Cholesterol synthesis can be inhibited by blocking the
conversion of 3-hydroxy-3-methyl-glutaryl CoA (HMGCoA) to mevalonate with specific HMG-CoA reductase
*Corresponding author. Tel.: 161-2-9926-8680; fax: 161-2-99266521.
E-mail address: [email protected] (H.H. Rasmussen).
inhibitors, and in vitro exposure to the HMG-CoA reductase inhibitor lovastatin causes a large reduction in the
cholesterol content of some cells [5]. As HMG-CoA
reductase inhibitors are widely used in clinical practice it is
important to establish if treatment with these drugs in vivo
can inhibit the Na 1 –K 1 pump. We have examined the
effect of treatment with lovastatin on the sarcolemmal
Na 1 –K 1 pump.
2. Methods
Male New Zealand White rabbits weighing 2.5–3.0 kg
and female Wistar rats weighing 125–150 g were used. We
gave rabbits a gelatin capsule containing 10 mg lovastatin
plus 230 mg lactose orally each day for 2 weeks (dose
Time for primary review 25 days.
0008-6363 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved.
PII: S0008-6363( 00 )00106-1
330
D.F. Gray et al. / Cardiovascular Research 47 (2000) 329 – 335
adopted from Ref. [19]). Control rabbits received capsules
containing 240 mg lactose for the same period. The dose of
lovastatin is |50-fold lower than the toxic dose for the
species [15]. We gave rats 10 or 30 mg lovastatin via a
gastric tube with 1 ml water daily (dose adopted from Ref.
[25]). Control rats were given the same amount of water.
Treatment with lovastatin was well tolerated and there was
no effect on body weight in either species.
Some rabbits treated with lovastatin received parenteral
mevalonate via osmotic minipumps (ALZET microosmotic minipump Model 1003D) for the second week of
the dosage period. We implanted two 100 ml minipumps in
the interscapular region under a brief general anaesthetic of
2% halothane with two parts nitrous oxide and one part
O 2 . Implantation of pumps containing only distilled water
had no effect on the Na 1 –K 1 pump (unpublished observations). Each pump contained 200 mg mevalonate dissolved
in water. The substance was administered at a rate of 1
mg / h from each pump. In addition, to ensure adequate
systemic levels, a dose of 120 mg mevalonate dissolved in
1 ml normal saline was injected daily via a marginal ear
vein. High cholesterol diets were prepared as published
previously [9]. Blood was taken from a marginal ear vein
for estimation of serum cholesterol in some rabbits [9].
At the end of the treatment period rabbits were anaesthetised with intramuscular xylazine hydrochloride (20
mg / kg) and ketamine hydrochloride (50 mg / kg). Once
deep anaesthesia was achieved the heart was excised and
single ventricular myocytes were isolated [10]. Myocytes
were used on the day of isolation only. Rats were
decapitated and the heart and gastrocnemius muscle excised. Heart weight was determined immediately after
excision and samples of left ventricular myocardium and
gastrocnemius muscle were taken for determination of
K 1 -dependent pNPPase activity and water and K 1 content.
All procedures were in accordance with guidelines of
Animal Care and Ethics Committees at our institutions and
conformed with the Guide for the Care and Use of
Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996).
For measurements of Na 1 –K 1 pump current (Ip ) in
rabbit myocytes at a fixed membrane voltage patch pipettes
were filled with a solution containing (in mM) 70 potassium glutamate, 1 KH 2 PO 4 , 5 N-2-hydroxyethylpiperazine-N9-2-ethanesulphonic acid (HEPES), 5
ethylene glycol-bis (b-aminoethyl ether)-N,N,N9,N9-tetraacetic acid (EGTA), 2 MgATP. In addition they contained
10 sodium glutamate plus 80 tetramethylammonium chloride (TMA.C1) or 80 sodium glutamate plus 10 TMA.C1.
The solutions were titrated to a pH of 7.0560.01 at 358C
with 1 M KOH. When the relationship between Ip and
membrane voltage was examined the compositions of
pipette filling solution and superfusate were designed to
eliminate K 1 -channel, Ca 21 -channel and Na 1 –Ca 21 exchange currents. The filling solution used in these experiments included Na 1 in a concentration ([Na] pip ) of 10
mM. The composition has been described in detail previously [9]. Filled pipettes had resistances of 0.8–1.1 MV.
Myocytes were initially superfused with modified Ca 21 containing Tyrode’s solution. When the whole cell configuration had been established the superfusate was changed to
one that was nominally Ca 21 -free. In some experiments
the K 1 concentration was varied between 0 and 15 mM.
The Na 1 concentration in these solutions was maintained
at 140 mM. TMA.C1 was used to maintain a constant
osmolality. The composition has been described in detail
previously [9]. Ip was defined as the shift in holding
current induced by 50 mM ouabain [11] unless indicated
otherwise. Membrane currents were recorded using the
continuous single electrode voltage clamp mode of an
Axoclamp-2A amplifier [9], and is reported normalised for
membrane capacitance [24] except where indicated. We
i
measured the intracellular Na 1 activity (a Na
) using ionsensitive microelectrodes. Details have been described
previously [10]. Tissue K 1 content was measured by flame
photometry as described [4].
K 1 -dependent pNPPase activity was determined in
crude homogenates with a tissue concentration of 10 mg /
ml. The K 1 -dependent pNPPase activity was determined
as the difference measured in buffers containing (in mM):
25 histidine, 15 MgCl 2 and 100 NaCl (pH 7.4) or 25
histidine, 15 MgCl 2 and 50 KC1 (pH 7.4). Details have
been described previously [16]. Vanadate-facilitated 3 Houabain binding to 2–4 mg (wet weight) intact samples
was performed as previously described [13]. Tissue water
content was determined as the relative reduction in weight
after heating samples at 908C until weight stabilisation,
and tissue protein was determined according to the method
of Lowry et al. [18].
Cholesterol, phospholipid and protein contents were
measured in rabbit hearts that had been frozen previously
and stored at 2208C. After thawing, ventricular tissue was
dissected free of any visible connective tissue and chopped
into small pieces. One-gramme samples were homogenised
for 30 s in 5 volumes ice-cold 10 mM Tris–HCl (pH 7.4)
using a hand held homogeniser. The homogenates were
centrifuged for 10 min at 1000 g and samples of the
supernatant removed for analysis.
Protein was determined by the bicinchoninic acid assay
(Sigma Chemical Company, St. Louis, MO, USA). Samples of homogenate were extracted by the Folch method
into chloroform / methanol using [ 14 C]-cholesterol as internal standard. The chloroform extracts were evaporated to
dryness and either assayed for phospholipid phosphorous
[23] or redissolved in mobile-phase solvent and assayed
for cholesterol by HPLC in acetonitrile / isopropanol / water
(44:54:2) as previously described [8].
Lovastatin (pure substance) was a gift from Merck
Research Laboratories, Merck and Co, Inc, Rahway, NJ,
USA. Cholesterol, ouabain, TEA.C1, mevalonic acid and
pNPP were purchased from Sigma Chemical Company, St.
Louis, MO, USA. TMA.C1 was ‘purum’ grade and
D.F. Gray et al. / Cardiovascular Research 47 (2000) 329 – 335
331
purchased from Fluka (Switzerland). 3 H-ouabain was from
Amersham International, Buckinghamshire, UK. Vanadate
was from Merck, Darmstadt, Germany. Chemicals used to
measure tissue K 1 content, K 1 -dependent pNPPase activity and 3 H-ouabain binding site concentration were of
analytical grade and purchased from Bie and Berntsen
(Denmark). All other chemicals were analytical grade and
purchased from BDH (Australia)
Results are expressed as means6S.E. Statistical comparisons were made using Student’s t-test for unpaired
observations. Dunnett’s test was used when the same
control group was used for more than one comparison.
Non-linear regression was used for fitting of the Hill
equation to data. A P value of ,0.05 was regarded as
statistically significant.
3. Results
3.1. Effect of lovastatin on Ip
To examine the effect of treatment with lovastatin on Ip
we isolated myocytes from rabbits treated with lovastatin
and from control rabbits. The myocytes in these and most
other experiments were voltage clamped using patch
pipettes containing 80 mM Na 1 to nearly saturate intracellular Na 1 –K 1 pump sites. The test potential was 240 mV.
Examples of traces of holding currents from a myocyte
isolated from a control rabbit and a rabbit treated with
lovastatin are shown in Fig. 1. Mean Ip of myocytes from
the control rabbits and myocytes from rabbits treated with
lovastatin are included in Fig. 2. Mean Ip of myocytes
from the lovastatin-treated group was significantly lower
than mean Ip of myocytes from controls. We conclude that
Fig. 2. Effects of lovastatin, cholesterol and mevalonate on Na 1 –K 1
pump current (Ip ). Myocytes were dialysed with pipettes containing 80
mM Na 1 (data represented by the first 5 bars) or 10 mM (data represented
by the last 2 bars). Mean Ip in myocytes from lovastatin treated rabbits
(lov) was significantly lower than mean Ip in myocytes from control
rabbits. The effect of lovastatin was still apparent when the diet was
supplemented with cholesterol (lov1chol). Dietary cholesterol produced
no significant change in mean Ip compared to mean Ip in myocytes from
control rabbits. Parenteral mevalonate reversed the inhibitory effect of
lovastatin on the pump. Statistically significant differences for key
comparisons are indicated by an asterisk.
near-maximal Na 1 –K 1 pump activity is reduced by the
treatment.
To examine if a lovastatin-induced decrease in Ip also
occurs when [Na] pip is at levels near the physiological
levels for intracellular Na 1 we voltage clamped myocytes
with pipettes containing 10 mM Na 1 . As expected, mean
levels of Ip , shown in Fig. 2, were much lower than mean
levels measured using a [Na] pip of 80 mM. However, the
relative decrease in Ip induced by lovastatin was similar for
the two groups of experiments.
3.2. Cholesterol, mevalonate and Ip
Fig. 1. Traces of holding currents from a myocyte isolated from a rabbit
treated with lovastatin (upper trace) and a myocyte isolated from a control
rabbit. Pump current (Ip ) is identified by the shift in holding current
induced by ouabain. The control myocyte was smaller than the myocyte
isolated from the lovastatin-treated rabbit as indicated by their membrane
capacitances. Despite this Ip of the control myocyte was larger than Ip of
the myocyte from the lovastatin-treated rabbit.
To examine if changes in Ip induced by treatment with
lovastatin might be related to changes in cholesterol status
we gave two rabbits a diet containing 1% cholesterol for 2
weeks. They were also given lovastatin. The serum cholesterol levels were 5.9 and 4.6 mmol / l (serum cholesterol of
control rabbits is |1 mmol / l, see Ref. [9]). Mean Ip of
myocytes isolated from these rabbits was similar to the
mean Ip of the rabbits given lovastatin and no dietary
cholesterol supplementation (Fig. 2). To examine the effect
of a wide range in serum cholesterol levels on Ip we gave
two additional groups of three rabbits each either 0.3%
cholesterol for 1 week or 1% cholesterol for 4 weeks.
These rabbits were not treated with lovastatin. The serum
cholesterol levels ranged from 5.3 to 21.2 mmol / l. There
was no correlation between serum cholesterol levels and Ip
(data not shown). The mean Ip of myocytes from all six
332
D.F. Gray et al. / Cardiovascular Research 47 (2000) 329 – 335
rabbits is included in Fig. 2. The mean Ip was similar to
mean Ip of myocytes from control rabbits fed a diet not
supplemented with cholesterol. Thus, wide variations in
serum cholesterol have no effect on Ip when intracellular
Na 1 is at a level expected to cause near-maximal pump
stimulation.
We also examined if lovastatin altered cardiac cholesterol content. We analysed left ventricular tissue from five
control rabbits and from five rabbits given lovastatin. The
diets were not supplemented with cholesterol. The cholesterol:phospholipid ratios (mol / mol) were 0.3560.02 and
0.3860.04 respectively. The difference was not significant.
Cholesterol contents normalised to total protein were
10.860.44 and 13.362.55 nmol / mg protein. The difference was not significant. Taken together, the absence of an
effect of dietary cholesterol on the lovastatin-induced
pump inhibition and the absence of an effect of lovastatin
on cardiac cholesterol content strongly suggest the effect
of lovastatin on Ip is independent of any effect of the drug
on cellular cholesterol.
To examine if the effect of lovastatin on Ip could be
reversed by mevalonate supplementation we treated a
group of rabbits with lovastatin for 2 weeks. Mevalonate
was given for the second week. Serum cholesterol levels
measured just before sacrifice were similar to levels of
control rabbits. As shown in Fig. 2 the reduction in Ip
induced by lovastatin was reversed by coadministration of
mevalonate.
3.3. Lovastatin and Na 1 –K 1 pump current–voltage
relationship
We next examined the effect of lovastatin on the voltage
dependence of Ip . After establishing the whole cell configuration myocytes from rabbits treated with lovastatin
and from controls were voltage clamped at a potential of
240 mV. We then applied voltage steps in 20 mV increments to test potentials (Vm ) from 2100 to 160 mV.
Details of the voltage clamp protocol and an example of
representative membrane currents have been published
previously [9]. The mean Ip –Vm relationships, normalised
to the Ip measured at 0 mV [9], for the myocytes from the
two groups of rabbits are shown in Fig. 3. The slopes were
similar.
Fig. 3. Mean Ip –Vm relationships for myocytes from rabbits treated with
lovastatin and for myocytes from control rabbits. Experiments on 9
myocytes from 2 rabbits treated with lovastatin and 18 myocytes from 4
control rabbits are summarised. Ip –Vm relationships are normalised to the
Ip recorded at 0 mV.
The relationship between the concentration of K 1 and Ip
are summarised in Fig. 4. The pump currents are normalised relative to the current recorded on pump activation
with 7 mM K 1 in the superfusate. When the Hill equation
was fitted to the data the K 1 concentration for half
maximal pump activation was 2.8 for myocytes from
rabbits treated with lovastatin and 2.9 for myocytes from
control rabbits. The Hill coefficients were 1.3 and 1.3.
3.4. Lovastatin and apparent K 1 affinity of the Na 1 –
K 1 pump
To examine if treatment with lovastatin altered the
pump’s dependence on extracellular K 1 myocytes were
voltage clamped at a fixed holding potential of 240 mV
and exposed to K 1 concentrations in the superfusate
ranging from 0 to 15 mM. Patch pipettes contained 80 mM
Na 1 . Details of the experimental protocol used to study the
apparent K 1 affinity have been published previously [9].
Fig. 4. Effect of extracellular K 1 ([K] 0 ) on pump current in myocytes
from rabbits treated with lovastatin and in myocytes from control rabbits.
Experiments on 9 myocytes from 3 treated rabbits and 13 myocytes from
3 control rabbits are summarised. Pump currents have been normalised to
the current at [K] 0 of 7 mM. The Hill equation has been fitted to the data.
D.F. Gray et al. / Cardiovascular Research 47 (2000) 329 – 335
3.5. Lovastatin, 3 H-ouabain binding and K 1 -dependent
pNPPase
Maximal Ip correlates well with Na 1 –K 1 pump density
measured with H 3 -ouabain binding technique [12]. This
suggests that the lovastatin-induced decrease in Ip we
measured using a [Na] pip of 80 mM reflects a decrease in
pump density. To examine this we measured H 3 -ouabain
binding to myocardial samples from seven control rabbits
and eight rabbits treated with lovastatin. The mean H 3 ouabain binding site concentrations were 803622 and
727617 pmol / g wet wt. The difference was statistically
significant.
To examine if lovastatin can affect the sarcolemmal
Na 1 –K 1 pump in an animal that is more resistant to
HMG-CoA reductase blockade than rabbits we performed
experiments on rats. Rats have been used extensively in
previous studies on HMG-CoA reductase inhibitors. They
were treated with 10 or 30 mg lovastatin per day for 2
weeks. Controls had placebo treatment. We used 5–6 rats
in each treatment group. Treatment had no effect on heart
weight, myocardial or skeletal muscle protein or water
content. Because the rat cardiac Na 1 –K 1 pump has a low
affinity for cardiac glycosides we examined the effect of
lovastatin on the pump by measuring K 1 -dependent
pNPPase activity rather than Ip or 3 H-ouabain binding
concentration. Mean myocardial K 1 -dependent pNPPase
activities for control rats and rats treated with lovastatin
are shown in the Table 1. Activities in both treated groups
were significantly lower than the activity in the myocardium of control rats. Gastrocnemius K 1 -dependent
pNPPase activity of rats treated with 10 mg lovastatin per
day was not significantly different from controls (see Table
1). However, there was a significant reduction relative to
controls in the activity in muscles from rats treated with 30
mg per day.
3
H-ouabain binding concentration cannot be measured
in rat heart. However it can be measured in rat skeletal
333
muscle. We measured the effect of treatment with 10 or 30
mg lovastatin on 3 H-ouabain binding to intact gastrocnemius muscle. The mean values for either group were not
significantly different from the values measured in tissue
from control rats. However, when results from experiments
using either treatment schedule were pooled the difference
was significant.
3.6. Effect of lovastatin on a iNa and K 1 content
To examine if lovastatin-induced changes in Na 1 –K 1
pump activity cause a change in cytoplasmic Na 1 of intact
cardiac myocytes, we measured a iNa The mean levels were
8.961.0 mM in Na 1 papillary muscles from 10 rabbits
treated with lovastatin and 6.360.5 mM in papillary
muscles from 9 controls. The difference was statistically
significant. We also examined if treatment had an effect on
i
Na 1 influx by measuring the rate of rise in a Na
upon
sudden pump blockade with the fast-acting cardiac steroid
dihydroouabain as described previously [10]. There was no
difference between the two groups.
Myocardial K 1 content was 7561 mmol / g wet wt. in 12
rats given either 10 mg or 30 mg lovastatin per day and
7861 mmo1 / g wet wt. in 11 control rats. This difference
was significant. Gastrocnemius muscle K 1 content was
10561 mmol / g wet wt. in 10 rats given either 10 or 30 mg
lovastatin and 10761 mmol / g wet wt. in 10 control rats.
This difference was not statistically significant.
4. Discussion
Treatment with lovastatin induced a decrease in Ip ,
H-ouabain binding and K-dependent pNPPase activity.
The absence of an effect on heart weight, tissue water or
protein content indicates that lovastatin did not induce a
generalised reduction in protein synthesis. Treatment was
associated with a reduction in the total amount of Na 1 –K 1
3
Table 1
Effects of lovastatin treatment on rat Na 1 –K 1 pump a
Lovastatin
(10 mg)
Lovastatin
(30 mg)
Control
Myocardial K 1 dependent
pNPPase activity
(mmol / min / g wet weight)
1.3660.13*
1.4260.14*
1.7660.007
Gastrocnemius K 1 dependent
pNPPase activity
(mmol / min / g wet weight)
0.7560.07
0.6360.03*
0.8160.03
3
233620
204613
26167
H ouabain binding
(gastrocnemius)
(pmol / g wet weight)
a
*
Values are mean6S.E.M.
P,0.05 compared to control.
334
D.F. Gray et al. / Cardiovascular Research 47 (2000) 329 – 335
ATPase per heart, indicating that the decrease in Na 1 –K 1
ATPase concentration was not due to an increase in
myocardial mass relative to membranes. The absence of an
effect of treatment on the apparent K 1 affinity and on the
voltage dependence of Ip suggests that treatment did not
merely induce a functional change in existing pumps.
Taken together, the results of the studies in rabbits and rats
strongly support the conclusion that lovastatin caused a
reduction in the density of functional Na 1 –K 1 pump units
and an increase in a iNa .
4.1. Potential mechanisms for effects of lovastatin on
the pump
The effect of lovastatin was independent of effects on
known determinants of Na 1 –K 1 pump activity. There was
no difference in body weight between treated and control
animals in either rabbits or rats to suggest that semistarvation or protein deprivation caused pump inhibition [14].
We are unaware of studies indicating that lovastatin
increases renal K 1 excretion, and we observed no other
cause for K 1 depletion in either rabbits or rats. It is
therefore unlikely that the effect of lovastatin was secondary to changes in the K 1 balance, a known regulator of
the pump, and the decrease in myocardial K 1 content with
lovastatin treatment is therefore likely to be a consequence
rather than a cause of pump inhibition.
Lovastatin had no effect on Ip when the product of
HMG CoA reductase, mevalonate, was administered the
week before rabbits were sacrificed. This indicates that
HMG CoA reductase inhibition rather than a direct effect
of lovastatin on the pump caused the decrease in Ip . The
most likely cause for the decrease in Ip is a reduction in
levels of a product of the mevalonate pathway. Such a
product could have a direct effect on the pump or an effect
on one of its regulators.
Since lovastatin can inhibit a cAMP-activated process
[5,17] and since cAMP is reported to regulate the sar1
1
colemmal Na –K pump [6] one might speculate that the
lovastatin-induced pump inhibition in this study was
related to an effect on cAMP. However, cAMP-mediated
effects are more likely to affect functional properties of the
pump than influence pump density. Pump density was
affected by treatment while there was no effect on functional properties.
Cholesterol synthesis and glycosylation and isoprenylation of proteins depend upon the mevalonate pathway. The
role of these processes in the lovastatin-induced pump
inhibition should be considered. Cholesterol seems unlikely to be involved because there was no effect of treatment
on cardiac cholesterol content, and dietary cholesterol
supplementation had no effect on the lovastatin-induced
decrease in pump activity (Fig. 2). Since the b-subunit of
the Na 1 –K pump is heavily glycosylated [7] impaired
glycosylation might reduce expression of pumps in the
membrane. However, there is no evidence of a reduction in
glycosylation in cultured cells exposed to HMG CoA
reductase inhibitors in concentrations much higher than
those expected under the in vivo conditions of the present
study [3,5]. An effect on isoprenylation is thought to
underlie lovastatin-induced inhibition of Cl 2 secretion in
cultured colonic cells [5]. However, the lovastatin levels
required to inhibit isoprenylation are high, likely to cause
cytotoxicity [22] and not expected to be achieved with the
dosage schedule we used. An effective specific inhibitor of
protein isoprenylation is not available [3]. The same
applies to a large extent to the other steps in the mevalonate pathway. This has made it very difficult to firmly
establish which step in this complex pathway is important
in cellular effects of HMG CoA reductase inhibition, even
under in vitro conditions that allow selective control of
substrate concentrations. The same difficulty is encountered with an ex vivo model such as that used in the
present study, and it would be difficult to firmly establish
the mechanism for the lovastatin-induced pump inhibition.
4.2. Implications of lovastatin-induced Na 1 –K 1 pump
inhibition
The clinical usefulness of treatment with HMG CoA
reductase inhibitors for coronary artery disease is well
established. However, inhibition of the Na 1 –K 1 pump
could have some adverse clinical consequences. When
combined with a decrease in sarcolemmal Na 1 –K 1 pump
concentration in heart failure [1] and a decrease in pump
function induced by therapeutic use of cardiac glycosides
[20] inhibition induced by treatment with HMG CoA
i
reductase inhibitors might cause a large increase in a Na
.
This, in turn, would affect intracellular levels of several
ions because of the existence of Na 1 -dependent co- and
counter-transporters in the sarcolemma. Adverse consequences might include cardiac arrhythmias, a common
complication of heart failure and digitalis toxicity.
Na 1 –K 1 pump inhibition might also affect vascular
tone. Exposure of both rat and human resistance vessels to
lovastatin for 48 h in vitro causes an increase in the
intracellular Ca 21 concentration, enhances responsiveness
to vasoconstrictors and impairs responsiveness to vasodilators. These effects are reversed by coadministration of
mevalonate [21]. A lovastatin-induced increase in a iNa may
21
have caused an increase in the intracellular Ca
concentration and hence an increase in vascular tone. A
similar sequence of events in vivo might cause an increase
in blood pressure.
Effects of sarcolemmal Na 1 –K 1 pump inhibition are
usually attributed to changes in intracellular Ca 21 and the
contractile state of myocytes. However, pump inhibition
probably has a more pervasive impact. In cardiac myocytes
pump inhibition causes an increase in the production of
reactive oxygen species and activation of growth-related
genes [27]. A similar effect of pump inhibition might
D.F. Gray et al. / Cardiovascular Research 47 (2000) 329 – 335
contribute to the myopathy that can occur during treatment
with HMG CoA reductase inhibitors.
Acknowledgements
This study was supported by the North Shore Heart
Research Foundation and the Danish Heart Foundation. DF
Gray was the recipient of a National Health and Medical
Research Council of Australia Medical Postgraduate Research Scholarship. PS Hansen was the recipient of a
Postgraduate Medical Research Scholarship from the National Heart Foundation of Australia. W Jessup is the
holder of a Research Fellowship from the National Health
and Medical Research Council of Australia.
References
[1] Bundgaard H, Kjeldsen K. Human myocardial Na, K-ATPase
concentration in heart failure. Moll Cell Biochem 1996;163 /
164:277–283.
[2] Cornelius F. Functional reconstitution of the sodium pump. Kinetics
of exchange reactions performed by reconstituted Na / K-ATPase.
Biochim Biophys Acta 1991;1071:19–66.
[3] Davies JE, Ng LL. Simvastatin and intracellular pH regulation by
the Na 1 / H 1 antiport of SV40-virus-transformed human MRC5
fibroblasts. Clin Sci 1993;84:633–643.
[4] Dørup I, Skajaa K, Clausen T. A simple and rapid method for the
determination of the concentrations of magnesium, sodium, potassium and sodium, potassium pumps in human skeletal muscle. Clin
Sci 1988;74:241–248.
[5] Ecay TW, Valentich JD. Lovastatin inhibits cAMP- and calciumstimulated chloride secretion by T84 cells. Am J Physiol
1993;265:C422–C431, Cell Physiol 34.
[6] Gao J, Cohen IS, Mathias RT, Baldo GJ. Regulation of the
stimulation of the Na 1 –K 1 pump current in guinea-pig ventricular
myocytes by a cAMP-dependent PKA pathway. J Physiol
1994;477:373–380.
[7] Geering K. The functional role of the b-subunit in the maturation
and intracellular transport of Na. K-ATPase. FEBS Lett
1991;285:189–193.
[8] Gelissen IC, Brown AJ, Mander EL, Kritharides L, Dean RT, Jessup
W. Sterol efflux is impaired from macrophage foam cells selectively
enriched with 7-ketosterol. J Biol Chem 1996;271:17852–17860.
[9] Gray DF, Hansen PS, Doohan MM, Hool LC, Rasmussen HH.
Dietary cholesterol affects Na 1 –K 1 pump function in rabbit cardiac
myocytes. Am J Physiol 1997;272:H1680–H1689, Heart Circ
Physiol 41.
[10] Hool LC, Whalley DW, Doohan MM, Rasmussen HH. Angiotensinconverting enzyme inhibition, intracellular Na 1 and Na 1 –K 1
pumping in cardiac myocytes. Am J Physiol 1995;268:C366–C375,
Cell Physiol 37.
335
[11] Hool LC, Gray DF, Robinson BG, Rasmussen HH. Angiotensin
converting enzyme inhibitors regulate the Na 1 –K 1 pump via effects
on angiotensin metabolism. Am J Physiol 1996;271:C172–C180,
Cell Physiol 40.
[12] Jaunin P, Horisberger J-D, Richter K, Good PJ, Rossier BC, Geering
K. Processing, intracellular transport and functional expression of
endogenous and exogenous a ]b 3 Na K-ATPase complexes in
xenopus oocytes. J Biol Chem 1992;267:577–585.
[13] Kjeldsen K. Complete quantification of the total concentration of rat
skeletal-muscle Na 1 1K 1 -dependent ATPase by measurement of
[ 3 H]ouabain binding. Biochem J 1986;240:725–730.
[14] Kjeldsen K, Everts ME, Clausen T. Effects of semi-starvation and
potassium deficiency on the concentration of [ 3 H]ouabain-binding
sites and sodium and potassium contents in rat skeletal muscle. Br J
Nutr 1986;56:519–532.
[15] Kornbrust DJ, MacDonald JS, Chennekatu PP et al. Toxicity of the
HMG-Coenzyme A reductase inhibitor lovastatin to rabbits. J
Pharmacol Exp Ther 1989;248:498–505.
[16] Larsen JS, Kjeldsen K. Quantification in crude homogenates of rat
myocardial Na 1 , K 1 - and Ca 21 -ATPase by K 1 - and Ca 1 -dependent
pNPPase. Age-dependent changes. Basic Res Cardiol 1995;90:323–
331.
[17] Lasa M, Chiloeches A, Garcia N, Montes A, Toro MJ. Lovastatin
decreases prolactin and growth hormone gene expression in GH 4 C 1
cells through a cAMP dependent mechanism. Mol Cell Endocrinol
1997;130:93–100.
[18] Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with folin phenol reagent. J Biol Chem 1951;193:265–275.
¨
[19] Ma PTS, Gil G, Sudhof
TC, Bilheimer DW, Goldstein JL, Brown
MS. Mevinolin, an inhibitor of cholesterol synthesis, induces mRNA
for low density lipoprotein receptors in livers of hamsters and
rabbits. Proc Natl Acad Sci USA 1986;83:8370–8374.
[20] Rasmussen HH, Okita GT, Hartz RS, Ten Eick RE. Inhibition of
electrogenic Na 1 pumping in isolated atrial tissue from patients
treated with digoxin. J Pharmacol Exp Ther 1990;252:60–64.
[21] Roullet J-B, Xue H, Pappu AS et al. Mevalonate availability affects
human and rat resistance vessels function. J Clin Invest
1995;96:239–244.
[22] Sinensky M, Beck LA, Leonard S, Evans R. Differential inhibitory
effects of lovastatin on protein isoprenylation and sterol synthesis. J
Biol Chem 1990;265:19937–19941.
[23] Sokaloff L, Rothblatt GH. Sterol to phospholipid molar ratios of L
cells with qualitative and quantitative variations of cellular cholesterol. Proc Soc Exp Biol Med 1974;146:1166–1172.
[24] Whalley DW, Hool LC, Ten Bick RE, Rasmussen HH. Effect of
osmotic swelling and shrinkage on Na–K 1 pump activity in
mammalian cardiac myocytes. Am J Physiol 1993;265:C1201–
C1210, Cell Physiol. 34.
[25] Willis PA, Folkers K, Tucker JL, Ye CQ, Xia U, Tamagawa H.
Lovastatin decreases coenzyme Q levels in rats. Proc Natl Acad Sci
USA 1990;87:8928–8930.
[26] Yeagle PL. Modulation of membrane function by cholesterol.
Biochimie 1991;73:1303–1310.
[27] Xie Z, Kometiani P, Liu J, Li J, Shapiro JI, Askari A. Intracellular
reactive oxygen species mediate the linkage of Na 1 / K 1 -ATPase to
hypertrophy and its marker genes in cardiac myocytes. J Biol Chem
1999;274:19323–19328.