|Home | About | Journals | Submit | Contact Us | Français|
Hypomagnesemia continues to be a significant clinical disorder that is present in patients with diabetes mellitus, alcoholism, and treatment with magnesuric drugs (diuretics, cancer chemotherapy agents, etc.). To determine the role of magnesium in cardiovascular pathophysiology, we have used dietary restriction of this cation in animal models. This review has highlighted some key observations which helped formulate the hypothesis that release of substance P (SP) during experimental dietary Mg deficiency (MgD) may initiate a cascade of deleterious inflammatory, oxidative, and nitrosative events, which ultimately promote cardiomyopathy, in situ cardiac dysfunction, and myocardial intolerance to secondary stresses. SP acts primarily through neurokinin-1 (NK-1) receptors of inflammatory and endothelial cells, and may: induce production of reactive oxygen and nitrogen species (superoxide anion, NO•, peroxynitrite, hydroxyl radical), leading to enhanced consumption of tissue antioxidants; stimulate release of inflammatory mediators; promote tissue adhesion molecule expression; and enhance inflammatory cell tissue infiltration and cardiovascular lesion formation. These SP-mediated events may predispose the heart to injury if faced with subsequent oxidative stressors (ischemia/reperfusion, certain drugs) or facilitate development of in situ cardiac dysfunction, especially with prolonged dietary Mg-restriction. Significant protection against most of these MgD-mediated events has been observed with interventions that modulate neuronal SP release or its bioactivity, and with several antioxidants (vitamin E, probucol, epicaptopril, d-propranolol). In view of the clinical prevalence of hypomagnesemia, new treatments, beyond magnesium repletion, may be needed to diminish deleterious neurogenic and prooxidative components described in this article.
Clincal hypomagnesemia has been associated with a higher incidence of arrhythmias,1 vasospasm,2 sudden death with congestive heart failure 3 and acute myocardial infarction. 3,4 In one study,5 45% of patients with acute myocardial infarction had this condition. Several animal models (hamsters,6,7 rats,8-10 mice,11,12 dogs,13 primates1) have been used to simulate clinical hypomagnesemia, and have exhibited histologic, electrical and/or functional abnormalities of the cardiovascular system. Moreover, pre-existing Mg-deficiency (MgD) was shown to amplify myocardial vulnerability to toxic agents1 and imposed stresses.14 When the imposed stress was ischemia / reperfusion (I/R), dogs on a Mg-restricted diet were shown to develop larger infarcts than those on a Mg-normal (or Mg sufficient, MgS) diet;14 and hearts from MgD rats exhibited lower recovery of cardiac function compared to the non-deficient group.10 Animals placed on Mg-restricted diets also displayed progressive cardiovascular lesion formation, heightened cardiac inflammatory cell infiltration,15 decreased levels of endogenous antioxidants (glutathione, vitamin E, ascorbate) 16,17 and higher plasma levels of pro-oxidant metals 18 and lipid peroxidation (LPO) products.19,20 Antioxidant treatment attenuated the severity of both cardiovascular inflammation in vivo21 and postischemic reperfusion injury in vitro,22 suggesting that dietary Mg-deficiency progresses into a pro-oxidant condition.
Mg-deficient rodents also exhibit temporal elevations in circulating inflammatory mediators which preceded significant leukocyte infiltration of cardiac tissue (≥ 3 weeks).19,23-25 Of further interest was our discovery that rodents experienced a neurogenic response during dietary MgD,24,26-28 and this became the underlying basis of our contention 26 that neuropeptides may initiate subsequent deleterious events in this model. In this review, experimental findings are described which were instrumental in formulating and refining of our hypothesis that MgD-mediated release of the neuropeptide, substance P, may trigger many of the inflammatory / oxidative / nitrosative events which eventually promote cardiomyopathy, in situ dysfunction, and loss of myocardial tolerance to secondary stresses.
Circulating levels of the pro-inflammatory neuropeptide, substance P (SP), 29,30 were found to be significantly elevated in MgD rodents.9,26,31,32 Calcitonin gene-related peptide (CGRP) was also elevated, and probably emanated from sensory-motor neuron fibers which are rich in both neuropeptides.33 Significant early elevations of SP (dietary day 3) preceded the peak increases (between dietary days 7 and 21),9, 26, 29-32 in other inflammatory parameters (circulating interleukin-1 [IL-1], IL-6, tumor necrosis factor alpha [TNFα], histamine, PGE2, gamma-interferon, circulating white blood cells [WBCs] and cardiac tissue WBC infiltration, CD11b, CD14, ICAM, & VEGF expression), in oxidative indices (LPO products, endogenous antioxidant depletion, iron status changes) and in nitrosative stress markers (nitrosyl-Hgb, plasma NO• oxidation products, cardiac tissue nitrotyrosine, and NOS2 and NOS3 protein expression). Changes in most of these parameters were substantially attenuated by concurrent in vivo treatment with specific neurokinin-1 (NK-1, or SP) receptor antagonists or an inhibitor of the N-methyl-D-aspartate (NMDA) receptor/channel during MgD, 9,10,31,34-37
The NMDA receptor / channel complex enhances neuronal entry of substantial amounts of calcium, which is an important mediator of neuropeptide release (including SP). The NMDA channel is blocked by Mg2+ in a voltage-dependent manner.38 Dizocilpine (MK-801), a non-competitive NMDA receptor/channel blocker, inhibited cardiac inflammation in the Mg-deficient rat. The increases in endothelium adhesion molecule CD54 (iCAM) and of monocyte / macrophage surface protein CD11b in ventricular tissue were attenuated by MK-801. MgD also promoted apoptosis in rat hearts and this was ameliorated by NMDA receptor blockade.39,40 NMDA receptor/channel blockade effectively reduced circulating SP levels in MgD rats. Treatment of 3 wk MgD rats with a relatively low dose of MK-801 (0.5 mg/kg/day, s.c.) caused depression in plasma SP levels at all examined time intervals 41 and area integration of the 3 wk time-course revealed an overall 50 % reduction in total SP compared to untreated. The early elevation in circulating SP during MgD may involve SP-rich C-fibers of the dorsal root ganglia or the spinal cord.43 SP immunostaining of dorsal root ganglia showed a time-dependent (1-3 wks) depletion of SP in MgD rats, and this was prevented with MK-801 pretreatment.39
The involvement of neuronal SP in systemic oxidative/nitrosative stress during MgD was also evident from in vivo NMDA receptor blockade studies. Plasma NO• product levels were shown to increase 2.5-fold in 2 wk MgD rats, and were reduced 50% by MK-801.42 Circulating prostaglandin E2 (PGE2, cyclooxygenase [COX]-derived vasoactive mediator) levels were elevated 3.9-fold in 2-wk MgD rats suggesting endothelial activation and were dose-dependently reduced by MK-801 treatment.39,42 MK-801 significantly prevented loss of RBC glutathione in 2 and 3 wk MgD rats, resulting in 38 - 56 % higher levels compared to the untreated MgD group.39 Finally, MK-801 treatment substantially reduced (~ 70%) the 5.3-fold elevation in circulating neutrophil basal O2- production observed in 3 wk MgD rats.42 Collectively, these data implicate a central role of the NMDA receptor in modulating tissue and circulating SP levels during MgD.
Additional support for a critical involvement of SP in the pathology of experimental MgD comes from in vivo NK-1 (SP) receptor blockade. The actions of SP on inflammatory cells in vivo are mediated through NK-1 receptors 44 and this interaction can: induce degranulation of tissue mast cells and histamine release;45 induce T-cell proliferation 46 and leukocyte adhesion; 45 activate macrophages;47 prime neutrophils for oxidative metabolism (superoxide anion production);48 and induce production of nitric oxide by endothelial cells.49 In vivo treatment with the SP receptor antagonist, L-703,606, significantly attenuated the enhanced neutrophil production of O2— observed in 3 week MgD rats,35,50 suggesting a substantial role for SP in neutrophil activation. Three weeks of MgD also elicited a 13-fold elevation in circulating PGI2 levels (COX-derived vasoactive mediator), and this was dramatically reduced (80%) by L-703,606.35 In vivo treatment with another SP receptor antagonist, CP-96,345, also prevented most of the MgD-mediated loss of RBC glutathione, the rise in plasma levels of lipid peroxidation products (TBAR-reactive materials), and the formation of myocardial inflammatory lesions in this model. 10,31,36 Changes in these oxidative stress markers resulting from SP receptor blockade are consistent with our observation that NMDA receptor blockade (MK-801) also attenuated these oxidative stress parameters.
MgD also induced NO• production and NOS up-regulation in the rat model. A moderate (1.7-fold) increase of plasma NO• oxidation products (nitrate + nitrite) was evident at day 7 of MgD and levels became 2 to 2.5-fold higher during weeks 2 and 3, indicating activation of NO• synthesis. MgD was shown to significantly induce iNOS (NOS2) protein expression in rat ventricular tissue after only 7 days. 51,52 SP receptor blockade (L-703,606) in vivo inhibited the rise in plasma nitrate + nitrite by 75 %, 35 raising the possibility that excessive NO• could mediate some of the deleterious effects of SP. Collectively, these findings strongly suggest that SP promotes neutrophil activation and systemic oxidative/nitrosative stress during MgD and that interventions which reduce circulating SP levels or block its bioactivity at the NK-1 receptor are effective against MgD-induced injury.
Since MgD has a systemic impact in our rodent model, we considered whether this condition predisposed animals to changes in intestinal permeability, endotoxemia and resultant cardiac inflammation. The gut is rich in neuropeptides 53 and may contribute to the early rise in plasma SP during MgD. During week 1 of MgD, intestinal inflammation was evident and pronounced PMN infiltration occurred by week 3, when significant decreases in mucosal barrier function were observed. Gastrointestinal mast cells (MCs) express receptors for SP and secrete tryptase II, the major mast cell protease (MCP-II) in rats.54 MgD increased intestinal MCP-II expression in rats.55 Moreover, eNOS (NOS3) expression (mRNA by real time reverse-transcription PCR) in small intestinal tissue of 3 week MgD rats was shown to significantly decrease (65 %) compared to control and may be associated with increased vascular permeability.55 This is consistent with our findings that intestinal permeability (Evans blue dye) increased significantly (19-fold > control) in 3 week MgD rats and suggests the potential for endotoxin (LPS) translocation. LPS may induce systemic increases in TNFα, IL-1β and IL-6 levels,56 and can stimulate secretion of TNFα from adult rat cardiac myocytes. This process was dependent on the presence of LPS receptors (CD14).57 In this light, we have observed enhanced CD14 receptor expression in MgD rat cardiac and intestinal tissues. These findings, as illustrated in Figure 1, implicate a potential role for SP in the development of enhanced intestinal permeability, endotoxemia and cardiac dysfunction during prolonged MgD.
Our prior studies 22,24,34 showed that 3 weeks of severe MgD9 (9% recommended daily allowance [RDA] for Mg) causes reduced myocardial tolerance to in vitro postischemic stress: greater mechanical dysfunction, enhanced tissue injury, and greatly elevated lipid peroxidation (LPO-derived lipid hydroperoxide) and free radical production. There were no significant differences in pre-ischemic baseline functional / hemodynamic properties of 3 week MgD9 and MgS hearts prior to I/R.22,34 Thus, in vivo events during the 3 weeks of MgD9 predisposed hearts to I/R injury.
Treatment in vivo with antioxidants prevented much of the MgD9 -induced loss of myocardial tolerance to I/R stress. 22,24,32,34 Vitamin E improved recovery of function, and reduced total lipid alkoxyl radical levels (electron spin resonance spin trapping of LO•), tissue LDH release, and tissue protein oxidation in 40 min I/R hearts.22 However, antioxidant protection may be secondary to the in vivo actions of SP during MgD. I/R hearts from L-703,606 (NK-1 receptor antagonist) -treated 3 week MgD9 rats exhibited a significant improvement in cardiac work recovery, whereas similarly-stressed MgS hearts did not significantly benefit. 34 L-703,606-treatment of MgD9 rats led to significant reductions in postischemic tissue LDH loss, lipid hydroperoxide production, and alkoxyl radical formation. 34 MgD rat hearts also produced substantially more (2.4 x) NO• during reperfusion and in vivo SP receptor blockade significantly reduced this heightened NO• production.35 This further increase in NO• production from I/R-stressed MgD hearts is likely to involve prior SP-mediated enhancement of iNOS (NOS2) activity during MgD. Similar treatment in vivo did not provide additional protection to I/R MgS hearts with respect to the above parameters, and acute L-703,606 treatment in vitro proved ineffective.34 Thus, SP receptor blockade must inhibit in vivo SP-triggered inflammatory responses observed during MgD, and these responses must influence myocardial vulnerability to I/R injury.
Varying dietary Mg-intake in rats directly influenced plasma SP levels,58 the associated severity of systemic oxidative/nitrosative stress,59,60 and loss of myocardial tolerance to I/R stress. 58 We demonstrated that SP release also occurred in rats fed moderate MgD diets (MgD20 =20 % RDA; MgD40 = 40% RDA) 58 and the literature suggests that many of the same pathological characteristics observed in the severe MgD9 animal also occurred in these animals. 59,60 MgD9, MgD40, and Mg100 (100% RDA=Mg-Sufficient or MgS) rats 58 displayed proportional changes in their plasma Mg concentrations (% of Mg100 : MgD9 = 31; MgD20 = 50; MgD40 = 75). When placed on increasingly restricted Mg diets for 3 weeks, they exhibited proportionately greater SP levels: total SP levels for MgD40 and MgD9 were 2.43- and 5.14-fold higher than Mg100, respectively. This confirms that the magnitude of the SP elevation can be modulated by the extent of dietary Mg-restriction, and that neurogenic inflammation can occur at clinically-relevant levels of hypomagnesemia. 12,58
Changes in RBC glutathione and plasma malondialdehyde, were also monitored in these 3 week Mg-restricted rats. 58 An inverse relationship for plasma malondialdehyde formation versus Mg-dietary content was observed: levels from MgD40, MgD20, and MgD9 rats were elevated 20%, 60%, and 148% compared to Mg100 rats. Moreover, the decline in RBC glutathione was also directly proportional to the extent of Mg-restriction: levels from MgD40, MgD20, and MgD9 rats fell 20.6%, 29.4%, and 50% compared to the Mg100 group.
When dietary duration was extended to 7 weeks, hearts from MgD20 and MgD40 rats displayed normal baseline (control) hemodynamic properties (Fig. 2), 58 yet were more susceptible than Mg100 (MgD20 > MgD40 > Mg100) to in vitro I/R stress (40- minute global I/30-minute R). This was shown by the greater oxidative (alkoxyl radical production) and tissue (myocardial LDH loss) injury. Significant (p<0.05) reductions in cardiac output, and LV developed pressure, along with increased mean aortic diastolic pressure, were responsible for further decreases in I/R cardiac work observed with MgD20 and MgD40 hearts. 58
When exposure to the moderate MgD20 diet was extended to 8 weeks, perfused hearts from these rats displayed significant (p<0.05) declines (18%-21%) in baseline cardiac output and cardiac pressure-volume work (Fig. 2, top); such significant changes were not observed at 7 weeks (Fig 2, top). 58 Hearts from 8 week MgD20 rats also displayed greater I/R contractile dysfunction (13 % lower cardiac work recovery) and enhanced oxidative injury (15% more alkoxyl radical production) (Fig. 2, bottom) compared to the I/R-stressed 7 week group. This demonstrates that duration of moderate dietary Mg-restriction is also an important determinant of myocardial susceptibility to imposed stresses.
We previously showed that elevations in circulating SP during 3 weeks of MgD was transient. 58 After an initial rise in SP levels during week 1 of MgD, levels declined during week 2, and this was followed by substantially greater elevations during week 3.9,31,34-36 Changes in systemic oxidative stress markers occurred with or subsequent to the initial increase in SP and were largely attenuated by SP receptor blockade in vivo. The effects of extending severe dietary Mg-restriction to 5 weeks on circulating SP levels and indices of oxidative stress were determined in the rat (Fig. 3). Rats exposed to 5 weeks of MgD exhibited a 6.5-fold increase in circulating SP levels. This was associated with enhanced oxidative stress, as indicated by the 52% decline in RBC glutathione and elevations in plasma 8-isoprostane levels (LPO), which were 80 % higher at week 3, but 203 % higher at 5 week. Circulating PMNs isolated from 5 week MgD rats displayed a nearly 4-fold higher basal superoxide-generating activity compared to the MgS group, indicative of endogenous basal PMN activation. These findings suggest that in vivo oxidative stress was actively occurring even during later periods of MgD.
Extending the duration of severe dietary MgD to 5 weeks also had a negative impact on in situ rat heart function, as shown by noninvasive echocardiography.35 Three weeks of MgD did not significantly alter LV systolic function, but 5 weeks of MgD (Table 1) led to a significant (p<0.05) depression in LV systolic function (LV ejection fraction fell by 6.1%; % fractional shortening by 16.5%; maximum aortic pressure [Pmax] by 24.4 %, and maximum aortic velocity [Vmax] by 13.5 %). Interventricular septum diameter in systole decreased by 18.6%, possibly indicative of dilated cardiomyopathy, and the mitral valve E/A wave ratio fell by 27.7%, suggestive of LV diastolic dysfunction. These findings may represent the first echocardiographic demonstration of cardiac dysfunction in a rodent model of MgD. We anticipate that future studies with SP receptor antagonists may confirm a causal role for SP in the development of oxidative injury and cardiac dysfunction during long-term MgD.
Key findings are presented which support our proposal that neurogenic inflammation (SP) during early stages of dietary MgD in rodents is the principle trigger of inflammatory, oxidative, and nitrosative events, which promote the subsequent development of in situ cardiac dysfunction, cardiomyopathy, and the loss of myocardial tolerance to imposed stresses (I/R). This view is largely supported by observations that in vivo treatments which either alter neuronal SP release / bioavailability (NMDA receptor/channel blockade, varying dietary Mg content), reduced SP bioactivity (NK-1 receptor blockade), or limited oxidative/nitrosative stress injury (antioxidants, nitric oxide synthase inhibitors), afforded significant protection against most of the pathology associated with MgD. Moderate Mg-restricted diets (20% and 40% RDA) in rodents produced neurogenic inflammation, as well as levels of hypomagnesemia that are comparable 12 to those found in various clinical situations (cardiovascular disease, diabetes, alcoholism, AIDS/HIV, cancer, drugs with Mg wasting properties).1-3,61 Thus, the use of additional agents (antioxidants, NK-1 receptor blockers), which lessen the potential cardiac pathology caused by this mineral deficiency, might be worthy of future clinical applications in addition to Mg-replacement therapy.
Supported by: NIHRO1 HL-62282 and HL-65178