High-Fat Diet Elevates Blood Pressure and Cerebrovascular Muscle Ca2+ Current
Abstract—Dietary fat contributes to the elevation of blood pressure and increases the risk of stroke and coronary artery disease. Previous observations have shown that voltage-gated Ca2+ current density is significantly increased in hypertension and can be affected by free fatty acids (FAs). We hypothesized that a diet of elevated fat level would lead to an increase in blood pressure, an elevation of L-type Ca2+ current, and an increase in saturated FA content in vascular smooth muscle cell membranes. Male Osborne-Mendel rats were fed normal rat chow or a high-fat diet (Ob/HT group) for 8 weeks. Blood pressures in the Ob/HT group increased moderately from 122.5±0.7 to 134.4±0.8 mm Hg (P<0.05, n=26). Voltage-clamp examination of cerebral arterial cells revealed significantly elevated L-type Ca2+ current density in the Ob/HT group. Voltage-dependent inactivation of the Ob/HT L-type channels was significantly delayed. Total serum FA contents were significantly elevated in the Ob/HT group, and HPLC analyses of fractional pools of FAs from segments of abdominal aorta revealed that arachidonic acid levels were elevated in the phospholipid fraction in Ob/HT. No differences in vascular membrane cholesterol contents were noted. Plasma cholesterol was significantly elevated in portal venous and cardiac blood samples from Ob/HT rats. These findings suggest that an elevation of plasma FAs may contribute to the development of hypertension via a process involving the elevation of Ca2+ current density and an alteration of channel kinetics in the vascular smooth muscle membrane.
The ontogeny of hypertension in the spontaneously hypertensive stroke-prone rat involves significant alteration of the physiology of vascular smooth muscle cells (VSMCs), including alteration of the composition of the phospholipid bilayer.1 The change in membrane composition is accompanied by increased numbers of L-type Ca2+ channels2 and increases in membrane concentrations of long-chain fatty acids (FAs).3 This compositional change may contribute to increased membrane microviscosity and a reduced sensitivity to Ca2+-dependent membrane stabilization.3 Shifts in vascular smooth muscle membrane lipid (and protein) composition lead to increased 45Ca2+ uptake.4 Calcium homeostasis in rabbit aortic smooth muscle cells is sensitive to in vitro cholesterol enrichment,5 in which enrichment with cholesterol-laden liposomes has been shown to increase L-type channel current6 and norepinephrine-gated Ca2+ influx.4
Recent reports describe a reduction in the risk of ischemic stroke with increased dietary fat.7 Increased dietary intake of ω3-FAs has been associated with a reduced risk of sudden cardiac death.8 In addition, the content of dietary fat may play a direct role in the genesis of hypertension and atherosclerosis.9 Thus, the quantity and the type of dietary fat can have important effects on the development of cardiac disease, hypertension, and stroke. Transmembrane Ca2+ influx seems to be an important signal transduction mechanism affected by these membrane FA modifications. Diet-induced increases in membrane concentrations of linoleic acids have been shown to influence transmembrane Ca2+ flux in leukocytes.10 In hypertension, bilayer modification and increased density of voltage-gated Ca2+ channels influence vascular excitability, leading to increased vascular tone. Other studies have demonstrated that long-chain free FAs (FFAs) (eg, oleic, linoleic, arachidonic) increase L-channel current in ventricular myocytes and modulate dihydropyridine binding in cardiac myocytes, possibly through modification of the physicochemical properties of the lipid/protein interface.11
Data from our laboratory and others have demonstrated increased levels of L-type Ca2+ channel current and increases in the ratios of L-type to T-type current in various VSMCs from hypertensive rats.12 13 However, there has been no indication of a change in the kinetics of the L-channel population with the hypertensive state, nor have any relationships been established among increased membrane microviscosity, hypertension, and Ca2+ channel activity. We hypothesized that dietary hyperlipidemia-induced hypertension involves increases in VSMC membrane levels of long-chain FAs, which may contribute to increased membrane Ca2+ influx through the augmentation of L-type Ca2+ channel current. Using a rat model of central obesity, we discovered that a high-fat diet induces modest increases in blood pressure accompanied by large increases in inward Ca2+ channel current (ICa) and a rightward shift in the voltage-dependent inactivation of the channel population. To our knowledge, this is the first observation of a significant shift in channel population kinetics in a hypertension model and is one that increases the window current for Ca2+ in these cells. We simultaneously studied the FA composition of the VSMC membrane to determine whether a change in the abundance of any nonesterified FAs (NEFAs) in the membrane parallels the altered biophysical properties of the cells leading to these delayed kinetics.
Animal and Cell Models
Adult, male Osborne-Mendel rats were fed ad libitum either a control rat chow (Purina 5001; L/NT group) or a high-fat diet (Teklad TD95407; Harlan; Ob/HT group) supplemented with AIN-76 mineral mix (Harlan) and Teklad 40060 vitamin mix (Harlan) for a period of 8 weeks. A comparison of these diets is shown in Table 1⇓. The Osborne-Mendel strain was chosen for its tendency to become obese when on a high-fat diet. Water was supplied ad libitum. The animals were maintained on a 12-hour day/night cycle. Blood pressures were monitored with tail-cuff plethysmography.
Animals were anesthetized with 3% halothane (volume percent in 95% O2/5% CO2) in an anesthesia chamber and decapitated. Segments of abdominal aorta were harvested, cleaned of adventitia, weighed, and frozen in liquid N2 for lipid extraction and analysis. Brains were removed and placed into chilled 0.1 mmol/L Ca2+ Hanks’ buffered salt solution containing (in mmol/L) NaCl 140, KCl 5.4, KH2PO4 0.44, NaH2PO4 0.42, NaHCO3 4.17, CaCl2 0.1, HEPES 5, and glucose 5.55 at pH 7.3. Single relaxed VSMCs were isolated from major brain arteries and subjected to voltage clamp as described previously.12
Isolation and Characterization of VSMC Membrane and Blood Lipids
Plasma NEFAs were assayed in fasted plasma samples assayed for total FFA through spectrophotometric analysis (NEFA kit; Wako). Rats in the control and high-fat diet groups were anesthetized with metofane. An abdominal incision was made, and the portal vein was exposed. Blood samples for total FFA determination were withdrawn. Cardiac blood was also withdrawn after a midline thoracotomy to expose the heart. Blood samples were stored in heparinized tubes and centrifuged, and the plasma was drawn off for FFA determination.
Membrane lipids were isolated from L/NT and Ob/HT VSMC aortic myocytes and plasma through separation into their major lipid constituents (phosphatidylcholine, phosphatidylinositol, phosphatidylserine, sphingomyelin, and cholesterol). The FA composition of each class was analyzed through the use of established HPLC techniques.14 15 Segments of cleaned thoracic aorta were opened, and the luminal surface was cleansed of endothelium with a swab. The tissue samples were blotted dry, immediately frozen, weighed, and stored in liquid nitrogen until analysis. Samples were continuously maintained under argon gas atmosphere to limit oxidation. FA analyses were performed as described previously.14
Plasma and membrane cholesterol levels were determined with a total cholesterol assay kit (Sigma Chemical Co). Vascular membrane cholesterol was assayed in 6 pairs of animals in either the L/NT or Ob/HT group. Portal venous and cardiac blood samples were obtained from metofane-anesthetized animals, before euthanasia. Blood samples were centrifuged to isolate plasma. Separation of membrane or plasma lipids with the use of TLC was performed as described for FA analyses. The cholesterol band was recovered from the silica with ether elution. The ether was evaporated, and the cholesterol samples were resuspended in PSS. Total cholesterol was determined spectrophotometrically. Recovery analyses of reference standards were 85% to 95% for the various compound classes. Phospholipid recovery was >95%; neutral lipids, including triglycerides, showed >85% recovery; and NEFAs exhibited >95% recovery.
Normalized data (eg, FA-to-phospholipid ratio) are expressed as mean±SEM and were compared with the use of ANOVA to assess differences in individual lipid species between aortas from L/NT and Ob/HT animals. A value of P<0.05 was considered significant. Values for n represent the number of vessels or plasma samples from individual animals unless otherwise specified.
Characteristics of the Ob/HT Rat Model
Systolic blood pressures, as measured with tail-cuff plethysmography after 8 weeks, averaged 122.5±0.7 mm Hg for the L/NT rats and 134.4±0.8 mm Hg for the high-fatOb/HT rats (P<0.05, n=26; Figure 1⇓). L/NT rats exhibited an average body weight of 477.8±17.6 g compared with 545.2±26.6 g for the Ob/HT rats (P<0.05). Total FFA content in plasma from L/NT and Ob/HT rats was 0.374±0.032 mEq/L for L/NT rats and 0.812±0.086 mEq/L (P<0.05) for Ob/HT rats (Figure 1⇓).
Hematocrits for each group were not significantly different, with a value of 50.4±1.6% for L/NT rats and 52.1±3.4% for Ob/HT rats (n=10). An increase in red blood cell hemolysis was noted in blood samples from Ob/HT animals. To measure the observed hemolysis, absorbance measurements of plasma were made at 414 nm with a spectrophotometer (Beckman Instruments). Plasma from L/NT rats had an average absorbance value of 0.281±0.018, whereas plasma from Ob/HT animals showed significantly higher levels of free hemoglobin, with an absorbance value of 0.837±0.247 (P<0.05, n=10 per group).
Passive Properties of L/NT and Ob/HT Cerebral VSMCs
Total membrane capacitance for cerebral VSMCs isolated from L/NT rats had a mean value of 16.16±0.98 pF. In VSMCs from Ob/HT rats, total cell capacitance averaged 16.04±0.94 pF (no significant difference from L/NT, n=34 cells). These values were taken from the null circuitry of the patch-clamp amplifier and were not significantly different from membrane capacitance values derived from the integrated area of the capacitative transient with the null circuitry turned off. These values may be considered to represent a rough measure of cell surface area and to indicate no differences in mean cell size after 8 weeks of a high fat or lean diet.
Voltage-Clamp Analyses of Ca2+ Channel Current
Inward voltage-gated Ca2+ channel current was significantly elevated in VSMCs from Ob/HT rats compared with L/NT control animals. In the lean VSMCs, ICa activated at −35 mV and reached a maximum inward amplitude of −50.6±6.2 pA at +20 mV with an −80 mV holding potential (Vh). In contrast, ICa in Ob/HT VSMCs activated at the same potential but achieved a maximum of −80.8±6.5 pA at +20 mV (P<0.05, n=27 cells). Normalization of inward current to cell capacitance yielded values of −4.0±0.8 pA/pF at +20 mV for L/NT cells and −6.1±0.8 pA/pF for Ob/HT cells (Figure 2⇓), which represents a significant increase in current density in Ob/HT cells (P<0.05, n=23). This strongly suggests an increase in Ca2+ channel conductance, an increase in Ca2+ channel number, or a change in channel kinetics. No evidence was obtained for the presence of T-type ICa in cells from either animal model based on digital subtraction of current records obtained at Vh of −80 and −40 mV. Current records obtained at Vh of −80 or −40 mV (Figure 2⇓) were robust and exhibited the typical slow decay of Ba2+ current.
We observed a significant rightward shift in the inactivation curve for OB/HT VSMCs compared with L/NT VSMCs (Figure 3A⇓). In the L/NT VSMCs, Boltzmann fits to plots of relative conductance versus conditioning voltage yielded a sigmoid plot showing a half-maximal inactivation (V1/2) of −14.0±1.9 mV. In contrast, V1/2 for Ob/HT VSMCs was −8.8±0.6 mV (P<0.05 compared with L/NT). Slope factors for the curves were not significantly different. Comparisons of residual noninactivated current reveal that Ob/HT cells exhibit significantly greater degrees of residual current activation in the voltage range of −30 to +10 mV. At 0 mV, for example, the residual current in Ob/HT is 197.5% of the L/NT value. In contrast to the inactivation data, no differences were observed in the voltage-dependent activation curves for ICa in the 2 cell types (Figure 3B⇓). Half-maximal activation occurred at −5.0±2.0 mV for L/NT cells and at 0.1±1.3 mV for Ob/HT cells (P>0.05). Slope factors for the Boltzmann fit curves were not different. These data indicate that the window current for ICa is larger in Ob/HT cells.
The decay phase of the inward ICa was described by a single exponential function with a distinct voltage dependence. Figure 3C⇑ shows the current-voltage relationships for L/NT and Ob/HT cell ICa decay. Although no differences in the time constant for current decay were observed at command steps to +20 mV (maximal activation voltage for ICa in both cell types), the Ob/HT cells exhibited slower decay time constants at more depolarized voltages. This suggests that an additional mechanism facilitates Ca2+ entry, which could result in an elevated intracellular Ca2+ concentration and increased vascular tone.
Serum and Vascular Smooth Muscle Lipid Profiles in L/NT and Ob/HT
Total plasma FFA content (Figure 1⇑) was significantly increased in the Ob/HT animals compared with the L/NT control animals. HPLC results revealed that only arachidonic acid levels in the phospholipid component were significantly elevated in the Ob/HT animals (n=28) (Table 2⇓). No other fractions exhibited significant changes in the Ob/HT animals.
In L/NT animals, membrane cholesterol level was 31.3±9.8 mg/mg tissue wet wt, but the Ob/HT animals exhibited cholesterol values of 25.4±2.7 mg/mg tissue wet wt (P>0.05 compared with L/NT, n=6). Portal venous serum cholesterol concentration was significantly elevated in the Ob/HT group (to 105.06±5.90 mg/dL) compared with 65.96±7.5 mg/dL in the L/NT control animals (P<0.05, n=8). Concentrations of serum cholesterol, which was taken from the heart via direct cardiac puncture, were 60.63±7.59 mg/dL in the L/NT rats and 95.74±5.00 mg/dL in the Ob/HT rats (P<0.05, n=8).
Ingestion of the high-fat diet elevated serum FA and cholesterol concentrations without inducing significant alterations in vascular smooth muscle membrane FA and cholesterol compositions. Despite a lack of effect on vascular membrane composition, the high-fat diet caused a modest elevation of systolic blood pressure and a significant increase in inward Ca2+ current density in the Osborne-Mendel rat. Most significant was the rightward shift in the L-type Ca2+ channel voltage-dependent inactivation curve in animals fed the high-fat diet. This suggests that a short period of high-fat diet intake may increase Ca2+ channel numbers or alter channel regulation, leading to increased transmembrane Ca2+ flux. In the cerebral arterial vessels used for these experiments, this process may have a significant impact on vascular reactivity. Although vascular reactivity was not examined in the cerebral vessels used in the present experiments, aortic responsiveness to the Ca2+ channel opener Bay K 8644 is elevated in animals fed the high-fat diet (J. Ritchey, personal communication, same animals as used in the present study, 1998).
It is unclear what molecular mechanism underlies the increased ICa density and shifted inactivation in the Ob/HT animals. Our working hypothesis that the high fat diet would increase VSMC levels of long-chain FAs, leading, perhaps, to a membrane that limited or slowed channel transitions from the open state to the closed state, has not been supported by the HPLC results from aortic muscle samples. The working hypothesis was centered on the notion that an increase in Ca2+ channel activity can contribute to increased tone in the vessels. The data show increased Ca2+ current density, delayed channel inactivation, and elevation of total plasma NEFA and arachidonic acid levels in the VSMC membrane, all in the face of only a modest increase in blood pressure. In other studies that demonstrate elevated Ca2+ current density in hypertension,12 13 the elevation in current density is associated with significantly elevated systolic pressures rather than the borderline hypertension observed in this study.
Research by Huang et al11 showed activation of Ca2+ channels in myocardial cells by long-chain NEFAs (including arachidonic, oleic, linoleic, and so on). The activation of ICa was independent of the activities of protein kinases A and C, G proteins, eicosanoid production, or nonenzymatic oxidation, strongly suggesting a direct effect of the FAs on the L-type channel. The proposed mechanism involved either the alteration of the local lipid domain of the channel or direct interaction of the FA with the channel. Although individual FAs may protect the myocardium in some cases and contribute to its injury in others, their role in vascular smooth muscle ion channel regulation is less clear. Our recent evidence demonstrates that linoleic acid, for example, can act as vascular smooth muscle hyperpolarizing factor by stimulating Na+/K+-ATPase activity.16 In rabbit coronary VSMCs, long-chain FAs were more effective than short-chain species in directly increasing maxi-K+ (BK) channel activity.17 Ω3-Polyunsaturated FAs (eg, eicosapentaenoic, docosahexaenoic) also inhibit receptor-mediated nonselective cationic currents in cultured A7r5 cells.18 Arachidonic and linoleic acids, although still inhibitory, showed much less effect in this system, whereas oleic and stearic acids showed no inhibition. Although evidence suggests an ion channel inhibitory action of individual FAs in vascular smooth muscle cells, the activation of L-type channels in myocardium leaves some room for speculation that these molecules may exert different actions at different sites. We have been unable to correlate changes in specific VSMC membrane FAs and cholesterol with the increased Ca2+ current density and altered inactivation properties in the Ob/HT rats. Thus, it may be that mechanisms secondary to the hyperlipidemia contribute to increases in Ca2+ channel current activity.
There is a rightward shift in the inactivation kinetics for ICa in the Ob/HT animals. Because current activation kinetics were not different in the 2 groups, the delayed inactivation in the Ob/HT animals indicates a larger “window current” for Ca2+ entry. This shift in kinetic properties for the L-channel population may be important in that it is evident even at membrane potentials around −40 to −30 mV and therefore in the range of depolarization that these cells may experience in vivo. What is surprising in this study is that a relatively large increase in inward Ca2+ current density occurs with only a modest increase in systolic pressure. This hints at a lack of synchrony between pressure elevation and increased inward current. Recent data from our group show that ICa may be elevated in both Wistar-Kyoto and spontaneously hypertensive stroke-prone rats fed the high-fat diet, despite a reduction of blood pressure in both fat-fed groups compared with control animals fed normal chow (D.W.W. and D.F. Bohr, unpublished observations, 1998). The molecular mechanism for the increase in ICa density observed in the Ob/HT animals remains undiscovered. Our revised working hypothesis, based on current data, is that elevated serum FA levels have a direct influence on vascular smooth muscle membrane functions. It is possible that these effects are centered in the lipid annulus of the channel protein.
This work was supported by NIH Grant HL-18575 (R. Clinton Webb, administrator).
- Received July 27, 1999.
- Revision received August 3, 1999.
- Accepted October 14, 1999.
Dominiczak AF, McLaren Y, Kusel JR, Ball DL. Lateral diffusion and fatty acid composition in vascular smooth muscle membrane from stroke-prone spontaneously hypertensive rats. Am J Hypertens. 1993;6:1003–1008.
Storm DS, Turla MB, Todd KM, Webb RC. Calcium and contractile responses to phorbol esters and the calcium channel agonist, Bay K 8644, in arteries from hypertensive rats. Am J Hypertens. 1989;3:245S–248S.
Bialecki RA, Tulenko TN. Excess membrane cholesterol alters calcium channels in arterial smooth muscle. Am J Physiol. 1989;257:C306–C314.
Gleason MM, Medow MS, Tulenko TN. Excess membrane cholesterol alters calcium movements, cytosolic calcium levels, and membrane fluidity in arterial smooth muscle cells. Circ Res. 1991;69:216–227.
Sen L, Bialecki RA, Smith E, Smith TW. Cholesterol increases the L-type voltage-sensitive calcium channel current in arterial smooth muscle cells. Circ Res. 1992;81:1008–1014.
Bing RF, Heagerty AM, Swales JD. Dietary changes in membrane lipids and leucocyte calcium. J Cardiovasc Pharmacol. 1988;12:S110–S113.
Huang JM, Xian H, Bacaner M. Long-chain fatty acids activate calcium channels in ventricular myocytes. Proc Natl Acad Sci U S A. 1992;89:6452–6456.
Wilde DW, Furspan PB, Szocik JF. Calcium current in smooth muscle cells from normotensive and genetically hypertensive rats. Hypertension. 1994;24:739–746.
Ohya Y, Abe I, Fujii K, Takata Y. Voltage-dependent Ca2+ channels in resistance arteries from spontaneously hypertensive rats. Circ Res. 1993;73:1090–1099.
Massey KD, Burton KP. α-Tocopherol attenuates myocardial membrane-related alterations resulting from ischemia and reperfusion. Am J Physiol. 1989;256:H1192–H1199.
Pomposiello SI, Alva M, Wilde DW, Carretero OA. Linoleic acid induces relaxation and hyperpolarization of the pig coronary artery. Hypertension. 1998;31:615–620.