Skip Navigation

Cardiovascular Research 1999 43(1):237-247; doi:10.1016/S0008-6363(99)00051-6
© 1999 by European Society of Cardiology
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Peuler, J. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Peuler, J. D.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Copyright © 1999, European Society of Cardiology

Opposing adrenergic actions of intravenous metformin on arterial pressure in female spontaneously hypertensive rats

Jacob D. Peuler*

Department of Pharmacology, Midwestern University, Downers Grove, IL, USA

* Corresponding author. Tel.: +1-630-515-6068; fax: +1-630-971-6414 jpeule{at}midwestern.edu

Received 2 July 1998; accepted 10 December 1998


   Abstract
Top
 Abstract
1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions and clinical...
 Acknowledgments
 References
 
Objective: Intravenous (iv) injection of the antidiabetic drug metformin rapidly lowers mean arterial pressure (MAP) in spontaneously hypertensive rats (SHR). However, if autonomic ganglia or {alpha}-adrenoceptors are first blocked then metformin rapidly raises MAP in SHR. This study was conducted to further characterize the adrenergic mechanisms of these opposing iv actions of the drug. Methods: Conscious, undisturbed female SHR with indwelling vascular catheters were used to measure acute effects of iv metformin (100 mg/kg; before and after sustained ganglionic blockade, GB, with chlorisondamine, 5 mg/kg) on: (1) circulating levels of catecholamines, (2) MAP after pharmacologic modulation of β- as well as {alpha}-adrenoceptors and (3) all the above in the absence as well as presence of the adrenal medulla. Results: Plasma norepinephrine (NE) and epinephrine (E) levels (pg/ml) were rapidly increased by iv metformin (8 SHR, p<0.05) both before GB ({Delta}NE=+146±41; {Delta}E=+119±31) and after GB ({Delta}NE=+79±24; {Delta}E=+120±32). Similar increases in plasma NE (though not E) were seen in SHR without adrenal medullae. Blockade of β-adrenoceptors with propranolol (pro; 3 mg/kg, 8 SHR) enhanced the rapid depressor response to iv metformin before GB ({Delta}MAP, mmHg: –38±4 with pro vs –17±3 without pro; p<0.05) and attenuated the rapid pressor response to iv metformin after GB ({Delta}MAP, mmHg: +8±3 with pro vs +30±4 without pro; p<0.05). Results were similar in SHR without adrenal medullae. Finally, if baseline MAP under GB was raised back to hypertensive levels with iv infusion of either NE or phenylephrine then iv metformin did not raise but rather reduced MAP in SHR. Conclusion(s): The acute depressor action of iv metformin in female SHR (1) is most likely due to a direct vasodilator action which includes inhibition of {alpha}-receptor-mediated vasoconstriction and (2) is buffered by an acute β-receptor-mediated pressor action likely due to a direct metformin-induced release of NE from postganglionic sympathetic nerve endings.

KEYWORDS Experimental; Regulatory systems; Pharmacology; Autonomic nervous system; Blood pressure; Diabetes; Hypertension; Vasoconstriction/dilation


   1 Introduction
Top
 Abstract
 1 Introduction
2 Methods
 3 Results
 4 Discussion
 5 Conclusions and clinical...
 Acknowledgments
 References
 
1.1 Clinical relevance and overall goal
While several human clinical studies have demonstrated that the antidiabetic agent metformin can reduce blood pressure after long-term oral administration [1–9], several others have not [10–19]. As might be expected, most of the former studies involved hypertensive patients [1–8] and many of the latter involved normotensive patients [14–19]. However, many of the latter also involved hypertensive patients [10–15], and the reason for variable antihypertensive effectiveness of the drug has not been identified.

To gain insight into potential mechanisms of metformin’s long-term oral effects on blood pressure, a number of investigators have turned to intravenous (iv) injection of the drug at relatively high doses (typically 100 mg/kg) which can immediately depress arterial pressures in both normal and spontaneously hypertensive rats, SHR [20–22]. Under certain experimental conditions some of these investigators have unmasked an apparently hidden pressor action of the drug in the SHR, which also occurs rapidly after such iv administration [22]. They have speculated that this acute pressor action could potentially buffer metformin’s acute depressor effect [22]. If a hidden pressor action of metformin should also occur chronically after oral administration of lower (more therapeutically relevant) doses, and particularly more so in some individuals than in others, then it could provide a reasonable explanation for the variable antihypertensive effectiveness of the drug as seen in the abovementioned clinical studies. Thus, the overall goal of the present study was to further characterize mechanisms responsible for the acute depressor and pressor actions of iv metformin in the SHR. In addition, unlike all the previous studies in which only male rats were used [20–22] we chose to study females since over half the patients used in the abovementioned clinical studies were women [1–19].

1.2 Acute depressor action of intravenous metformin
Muntzel et al. [22] reported that the rapid depressor action of iv metformin in male SHR could be abolished by pretreating them with either hexamethonium (a ganglionic blocker) or phentolamine (an alpha-adrenoceptor blocker). Unfortunately, unlike other ganglionic blocking agents (e.g. chlorisondamine) which completely block neurotransmission through all autonomic ganglia, hexamethonium fails to do so particularly in the rat [23,24]. Thus, it is somewhat difficult to accurately interpret the Muntzel et al. results [22] in light of this confounding variable. Nevertheless, when considered together, their results do raise the possibility that metformin’s acute iv depressor effect in the SHR is related to either (1) a rapid generalized decrease in postganglionic sympathetic neural release of norepinephrine (NE) onto vascular smooth muscle alpha-adrenoceptors and/or (2) a rapid generalized decrease in vascular smooth muscle contractile responsiveness to stimulation of alpha-adrenoceptors by NE.

In support of the first possibility, iv injection of metformin has been shown repeatedly to rapidly inhibit renal postganglionic sympathetic neuroaxonal action potentials in the rat [20,21] and some have suggested that it may exert a generalized, rapid sympatho-inhibitory effect [21]. If so, then it should also rapidly decrease systemic circulating levels of the postganglionic sympathetic neurotransmitter, NE. On the other hand, we and others have found that metformin in vitro (at concentrations likely to occur extracellularly immediately after such iv injection) can rapidly relax contractions produced by either NE or phenylephrine in tail arterial tissues freshly isolated from SHR and other rats [25–27]. If this rapid, direct vasodilator action of metformin also occurs in vivo (involving more vascular tissue than just tail artery) then like other direct-acting vasodilators (e.g. calcium channel blockers) it would be expected to reflexly increase systemic circulating levels of NE [28]. Thus, one goal of the present study was to measure plasma NE levels immediately after iv injection of metformin. In addition, since this rapid, direct vasodilator action of metformin on alpha-adrenoceptor-mediated vascular contractility has thus far only been seen in an isolated arterial preparation [25–27], we also sought evidence that the drug could exert this action systemically (in general) in the whole animal after iv administration.

1.3 Acute pressor action of intravenous metformin
Muntzel et al. [22] also found that pretreatment of SHR with hexamethonium not only abolished the rapid depressor effect of iv metformin but unmasked a rapid pressor action of the drug accompanied by an increase in heart rate. They were able to block the latter but not the former with propranolol in these same hexamethonium-treated rats [22]. Thus, they speculated that the increase in heart rate but not the increase in arterial pressure might be due to direct stimulation by metformin of epinephrine (E) release from the adrenal medulla [22]. In preliminary experiments we also examined effects of propranolol on metformin’s actions in ganglionically-blocked SHR but with chlorisondamine instead of hexamethonium to provide the ganglionic blockade [29]. We found that propranolol then not only abolished the increase in heart rate but also markedly attenuated the increase in arterial pressure [29]. Thus, we chose to further explore the effects of beta-adrenoceptor blockade in SHR, not only after but also before ganglionic blockade with chlorisondamine, to determine if it altered the depressor response to metformin. In addition, we measured plasma E levels in response to iv metformin, and also measured all hemodynamic and plasma catecholamine responses in SHR in which the adrenal medulla had been removed to determine its role in the iv actions of the drug.


   2 Methods
Top
 Abstract
 1 Introduction
 2 Methods
3 Results
 4 Discussion
 5 Conclusions and clinical...
 Acknowledgments
 References
 
2.1 Animals and vascular catheterizations
Female spontaneously hypertensive rats (SHR) were obtained from Harlan Sprague Dawley (Indianapolis, IN) at age 15 weeks (190–200 g). In some, the adrenal medulla was surgically removed prior to shipment. After arrival at Midwestern University, all procedures employed were performed in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23). Before experimentation, all rats were fed food and water ad libitum for 5–6 weeks after which body weights ranged between 230 and 250 g. During that time they were also housed individually with weekly bedding changes and in a room ventilated with more than 10 fresh-air changes per hour to desynchronize their estrous cycles, as described previously [30,31]. They were then anesthetized with ketamine/xylazine (50/5 mg/kg, intramuscularly) to allow implantation of vascular catheters for subsequent hemodynamic monitoring, blood sampling and drug administration. For each rat, a polyethylene catheter was inserted in the right femoral vein and a teflon catheter was inserted in the right femoral artery. In some rats, two polyethylene catheters were inserted in the right femoral vein to permit bolus drug injections in one while simultaneously infusing another vasoactive agent at a continuous rate through the other (Study 3, below). All catheters were exteriorized through the dorsal neck skin and locked with heparinized saline when not in use. All rats were allowed 3 days to recover from surgery before use in experiments.

The arterial catheters were fabricated from teflon because it is a very nonthrombogenic material and highly resistant to retention of particulates on the inner wall. They permit arterial blood sampling and hemodynamic monitoring repeatedly over a period of several days in conscious unrestrained rats [32,33]. When in use, the arterial catheter in each rat was connected to an arterial infusion-monitoring system consisting of a saline-filled extension catheter (suspended above the animal’s cage by a light weight spring) connected to a Sorenson model Intraflo (Abbot Laboratories, North Chicago, IL, USA) which in turn was attached directly to a pressure transducer. The Intraflo permitted a low rate of intra-arterial infusion of saline (≤10 µl per min) which does not influence arterial pressure recordings [34] but yet eliminates the need to vigorously flush catheters periodically with bolus injections of saline to maintain patency before and after blood sampling and while monitoring arterial pressures. Pressures and heart rates were recorded with Grass instrumentation (Quincy, MA, USA).

2.2 Preliminary studies
Preliminary experiments were conducted to determine (1) the completeness of adrenal medullectomy, (2) an iv dose of metformin that consistently produces detectable acute changes in MAP, (3) appropriate timing of arterial blood sampling for assessment of circulating catecholamines in conscious undisturbed rats and (4) appropriate iv blocking doses of chlorisondamine (for ganglionic blockade), propranolol (for beta-adrenoceptor blockade) and phentolamine (for alpha-adrenoceptor blockade).

Stress-induced increases in plasma E, as seen immediately after brief immobilization in SHR with intact adrenal medulla, were absent in preliminary tests with adrenal medullectomized SHR.

Preliminary iv injections of metformin at 10, 30, 100 and 300 mg/kg produced, on the average, graded reductions in MAP in conscious undisturbed SHR with intact autonomic ganglia. However, the reproducibility of these reductions was consistently better at 100 mg/kg than at either 10, 30 or 300 mg/kg. Also, at 300 mg/kg some rats appeared to be disturbed immediately after injection of the drug, as indicated by postural changes from one resting position to another. This in turn appeared to distort the recording of resting MAP in those rats. This was not observed at the lower doses of metformin including 100 mg/kg. Thus, only 100 mg/kg was used in all subsequent experiments. This is also the dose most often discussed in the previous studies with iv metformin [20–22].

A period of at least two hours after handling SHR for the purpose of connecting their arterial catheters to the arterial infusion-monitoring system was often necessary in preliminary tests to allow arterial plasma NE and E levels to reach low, stable baseline (control) values. Thereafter, sampling at 15 min intervals consistently revealed similar values. Sample volumes were limited to 500 µl of whole blood each (per kg body weight). No more than four samples were withdrawn from any one rat per day in these preliminary experiments. In all other experiments described in Study 1 (below) no more than two samples were withdrawn from any one rat per experimental day.

In other preliminary tests, baroreflex-mediated changes in heart rate in response to bolus iv depressor doses of nitroprusside and iv pressor doses of phenylephrine, as seen in SHR with intact autonomic ganglia, were completely abolished for at least one hour after administration of the ganglionic blocker chlorisondamine at 5 mg/kg (iv) to the same rats. Finally, significant changes in arterial pressure in response to iv isoproterenol and NE, as seen in SHR with intact adrenoceptors, were essentially abolished for at least one hour after administration of propranolol at 3 mg/kg (iv) and phentolamine at 5 mg/kg (iv), respectively. Thus, these were the doses of these agents used below in Studies 1 and 2 as indicated.

2.3 Study 1. Acute effects of intravenous metformin on arterial plasma catecholamines in SHR before and after ganglionic blockade
Conscious, unrestrained SHR with the adrenal medulla either intact (n=8) or removed (n=8) were examined in this study. Three days after surgery (catheterizations) the arterial catheter in each rat was connected to the arterial infusion-monitoring system. After at least two hours, a bolus of metformin vehicle (isotonic saline) was injected iv (1 ml/kg). One min later, the arterial catheter was disconnected and a control arterial blood sample was withdrawn (500 µl/kg). The arterial catheter was then re-connected to the arterial infusion-monitoring system and 15 min later, a bolus of metformin was injected iv (100 mg/kg) in the same volume of saline. One minute later, another arterial sample was obtained similarly. Each sample was immediately heparinized and chilled on ice after its withdrawal, then centrifuged to separate plasma (within 5 min). Plasma in turn was immediately transferred to another tube and stored at –30°C for later analysis of catecholamines. Three days later, the same injections and sampling procedures were applied to the same rat but beginning 30–40 min after ganglionic blockade with chlorisondamine (5 mg/kg, iv). All arterial plasma samples were analyzed for NE and E by radioenzymatic assay as described previously [35]. Some were also analyzed for glucose by oxidation as described previously [33]. In preliminary work, metformin added to saline and to control plasma was subjected to analysis by the same methods and found not to interefere with the analysis of either catecholamines or glucose.

After completion of these blood sampling experiments (i.e. at least 3 days after recovery from the ganglionic blockade) a select number of these rats were further used to confirm previously reported effects of alpha-adrenoceptor blockade with phentolamine on the acute blood pressure responses to metformin in SHR with intact adrenal medullae [22] and also to test effects of phentolamine on pressure responses to metformin in our adrenal medullectomized rats. While continuously recording arterial pressure, phentolamine was injected iv at 5 mg/kg. After arterial pressures stabilized (about 40 min), a bolus of metformin was injected iv (100 mg/kg) and acute arterial pressure responses were recorded.

2.4 Study 2. Acute effects of intravenous metformin on arterial pressures and heart rates in SHR before and after ganglionic blockade and/or beta-adrenoceptor blockade
Conscious, unrestrained SHR with the adrenal medulla either intact (n=8) or removed (n=8) were examined in this study. Each rat was subjected to a series of 4 experimental conditions (A, B, C, D) administered on separate days (viz. days 3, 6, 9 and 12 after surgical implantation of vascular catheters). In condition A (autonomic ganglia intact; beta-adrenoceptors intact) each rat was injected with a bolus of metformin (100 mg/kg, iv) between 2 and 3 h after connecting the animal’s arterial catheter to the arterial infusion-monitoring system. In condition B (autonomic ganglia intact; beta-adrenoceptors blocked) each rat was pretreated with propranolol (3 mg/kg, iv) approximately 45 min before the scheduled injection of metformin (as in condition A). In condition C (autonomic ganglia blocked; beta-adrenoceptors intact) each rat was pretreated with chlorisondamine (5 mg/kg, iv) approximately 45 min before metformin. In condition D (autonomic ganglia blocked; beta-adrenoceptors blocked) each rat was pretreated with chlorisondamine and propranolol approximately 45 and 35 min before metformin, respectively. Then arterial pressure and heart rate responses to the iv metformin were recorded for later analysis.

The order in which rats were subjected to these 4 different experimental conditions was A, B, C, and D for half the animals in each group and the reverse (D, C, B and A) for the other half.

2.5 Study 3. Acute effects of intravenous metformin on arterial pressure elevations produced by intravenous phenylephrine or norepinephrine in SHR
The primary goal of this study was to determine whether the ability of metformin to rapidly relax vascular contractile reactivity to alpha-adrenoceptor agonists, as seen thus far only in vitro in isolated rat tail artery [25–27], could also be demonstrated in vivo (systemically) in the SHR. To achieve this goal, we first removed endogenous sympathetic neural support of peripheral vascular alpha-adrenoceptor activity in each rat (3 days after implantation of catheters). This was accomplished with the ganglionic blocking agent chlorisondamine (5 mg/kg, iv). Then arterial pressure in each rat was raised back to the resting hypertensive level (recorded before administration of chlorisondamine) by specifically re-activating the peripheral vascular alpha-adrenoceptors with phenylephrine (infused through a second iv catheter). After reaching this level, a bolus of metformin vehicle (isotonic saline, 1 ml/kg) was injected through the first iv catheter to confirm lack of vehicle effects on the phenylephrine-supported arterial pressure. Then metformin was injected iv at 100 mg/kg in the same volume of saline. Arterial pressure and heart rate responses were recorded. In other rats, NE was infused instead of phenylephrine. In all rats the same infusions of either phenylephrine or NE, plus the metformin injections, were repeated 4 days later but in the absence of ganglionic blockade. The rates of phenylephrine and NE infusions employed (per kg body weight per minute) were all in the low microgram range for phenylephrine and in the middle nanogram range for NE.

The use of ganglionically blocked rats precluded certain difficulties in interpretation of results from this particular study. Obviously, it excluded the possibility that effects of metformin on either the phenylephrine-induced or the NE-induced hypertension could be influenced in any way by changes in sympathetic neuroaxonal traffic, either reflex-mediated or directly drug-induced.

2.6 Data analysis
To determine acute effects of iv bolus injection of metformin on hemodynamic variables (mean arterial pressures and heart rates) in each rat, baseline (control) values of these variables averaged over several minutes immediately before injections were subtracted from values observed approximately 30–60 s after injection (which was typically when maximum changes occurred). To determine acute effects of iv bolus injection of metformin on arterial plasma concentrations of catecholamines, baseline (control) values of these variables obtained 15 min before injection were subtracted from values obtained approximately one minute after injection.

These acute effects of metformin were summarized as mean±SE per treatment and subjected to appropriate statistical analyses. Paired-t was employed to determine if acute changes produced by iv metformin injection in any variable (within any given treatment) were significantly different from zero change, with a probability of error less than 5% (p<0.05). Analysis of variance (ANOVA) with subsequent unpaired multiple mean comparisons was employed to assess influences of different treatments (e.g. adrenal medullectomy, β-blockade, ganglionic blockade) on the acute effects of metformin on any given variable (p<0.05).


   3 Results
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
4 Discussion
 5 Conclusions and clinical...
 Acknowledgments
 References
 
3.1 Study 1. Acute effects of intravenous metformin on arterial plasma catecholamines in SHR before and after ganglionic blockade
Baseline (control) levels of arterial plasma catecholamines (15 min before metformin) and acute changes in these levels as produced by iv metformin (1 min after metformin) in conscious, unrestrained female SHR are shown in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1 Effects of autonomic ganglionic blockade and adrenal medullectomy on acute changes in plasma catecholamine produced by bolus iv injections of metformin (100 mg/kg) in adult female spontaneously hypertensive rats

 
As expected, ganglionic blockade with chlorisondamine markedly reduced the control levels of both NE and E while adrenal medullectomy only reduced the control levels of E. Neither had any significant influence on control levels of plasma glucose (data not shown). Intravenous bolus injections of metformin acutely increased circulating levels of NE, and similarly so in SHR with and without intact adrenal medulla, but only increased circulating levels of E in SHR with intact adrenal medulla. Significant acute increases in plasma NE in response to iv metformin were also observed in the same rats after ganglionic blockade, although they were notably less than the increases observed before ganglionic blockade. Significant acute increases in plasma E in response to iv metformin were also seen after ganglionic blockade and to the same degree as observed before ganglionic blockade in those SHR with intact adrenal medulla. Metformin did not acutely alter plasma glucose levels (data not shown).

After recovering from the ganglionic blockade (3–4 days) a few of these rats, with and without intact adrenal medulla, were pretreated with phentolamine iv and then injected with metformin (100 mg/kg, iv). With or without adrenal medullae, they only showed acute pressor responses to the metformin (not depressor responses) similar in magnitude to the pressor responses observed in the ganglionically-blocked SHR of Study 2.

3.2 Study 2. Acute effects of intravenous metformin on arterial pressures and heart rates in SHR before and after ganglionic blockade and/or beta-adrenoceptor blockade
Baseline (control) values for hemodynamic variables (mean arterial pressures, MAP, and heart rates, HR) and maximal acute changes in these variable as produced by bolus iv injection of metformin in conscious, unrestrained female SHR are shown in Table 2. Fig. 1 illustrates chart recordings of the time course of these changes in a single SHR under the different experimental conditions.


View this table:
[in this window]
[in a new window]

  		Table 2 Effects of beta adrenoceptor blockade, autonomic ganglionic blockade and adrenal medullectomy on acute changes in mean arterial pressure (MAP) and heart rate (HR) produced by bolus iv injections of metformin (100 mg/kg) in adult female spontaneously hypertensive rats

 

Figure 1
View larger version (31K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 Chart recordings of effects of bolus injections of metformin (100 mg/kg, iv) on arterial pressures (AP), mean arterial pressures (MAP) and heart rates (HR) in a conscious, unrestrained female spontaneously hypertensive rat with intact adrenal medulla. The rat was subjected to separate experimental conditions (A,B,C,D; separated by 3-day intervals) involving blockade of either autonomic ganglia with chlorisondamine (5 mg/kg, iv) and/or beta adrenoceptors with propranolol (3 mg/kg, iv). The timing with which each condition was initiated prior to administration of the metformin is described in detail in Methods under Study 2. Maximal effects of metformin (seen within 1 min after its injection) are summarized in Table 2 for all rats in the study, with and without intact adrenal medulla.

 
Ganglionic blockade markedly reduced control levels of MAP and slightly reduced control HR in these rats (Table 2). Beta-adrenoceptor blockade only tended to reduce control HR (Table 2). Intravenous bolus injections of metformin rapidly increased HR and lowered MAP in rats with functionally intact autonomic ganglia and intact adrenergic beta receptors (Table 2). Ganglionic blockade did not significantly influence the acute tachycardic response to metformin but reversed the MAP response to that of an acute pressor effect. Beta receptor blockade greatly enhanced the depressor response to metformin in rats with intact autonomic ganglia and markedly attenuated the pressor response to metformin in rats with blocked autonomic ganglia. Beta receptor blockade also abolished the tachycardic effects of metformin both in the absence and in the presence of ganglionic blockade. Similar results were obtained in rats in which the adrenal medulla had been removed. There was no effect of the order in which these rats were subjected to the different experimental conditions.

In a few preliminary tests with two adult normotensive female Sprague Dawley (SD) rats (one with and one without intact adrenal medulla) definitive depressor responses to iv metformin (100 mg/kg) were observed before ganglionic blockade but they were notably smaller than the averages shown in Table 2 for the adult female SHR ({Delta}MAP=–10 mmHg in each female SD rat). After ganglionic blockade, both of these rats clearly demonstrated acute pressor responses to the same dose of metformin ({Delta}MAP=+33 mmHg and +25 mmHg, respectively).

3.3 Study 3. Acute effects of intravenous metformin on arterial pressure elevations produced by intravenous phenylephrine and norepinephrine in SHR
If baseline mean arterial pressures in SHR under ganglionic blockade (~100 mmHg) was raised back to hypertensive levels (typically≥160 mmHg) by stimulation of vascular smooth muscle alpha-adrenoceptors with phenylephrine then metformin did not raise MAP (as in Fig. 1C) but rather markedly reduced MAP (Fig. 2). The average MAP reduction for 5 SHR was –44±5 mmHg (p<0.05, paired-t). A depressor effect of metformin also occurred (–41±4 mmHg, 5 SHR, p<0.05, paired-t)) if NE was infused instead of phenylephrine to restore the pre-ganglionic blockade level of arterial pressure, independent of whether the infused NE increased heart rate. Nearly similar depressor effects were observed in the same rats if infused with the same levels of phenylephrine or NE after autonomic ganglionic functions had recovered four days later (data not shown).


Figure 2
View larger version (23K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 Chart recordings of effects of a bolus injection of metformin (100 mg/kg, iv) on arterial pressure (AP), mean arterial pressure (MAP) and heart rate (HR) in a conscious, unrestrained female spontaneously hypertensive rat in which autonomic ganglia were blocked but arterial pressure was supported at the hypertensive level via an infusion of phenylephrine through a second iv catheter. The phenylephrine infusion was initially set at 1 µg/kg/min but then adjusted to 5 µg/kg/min to achieve the desired level of arterial pressure.

 

   4 Discussion
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
5 Conclusions and clinical...
 Acknowledgments
 References
 
4.1 New findings
Several new findings regarding acute effects of iv metformin in SHR were obtained from the present study. First, bolus iv injections of metformin acutely increased NE and E levels in arterial plasma of conscious undisturbed female SHR, both before and after ganglionic blockade with chlorisondamine. Similar increases in NE, though not E, were observed in SHR in which the adrenal medulla had been removed several weeks before testing. Second, blockade of beta-adrenoceptors with propranolol greatly enhanced acute depressor responses to iv metformin in female SHR with intact autonomic ganglia and markedly attenuated acute pressor responses to the drug when autonomic ganglia were blocked with chlorisondamine. Similar results were observed in SHR in which the adrenal medulla had been removed. Third, if baseline arterial pressures in SHR under ganglionic blockade with chlorisondamine were raised back to the hypertensive levels observed before the blockade, by specifically re-activating only vascular smooth muscle alpha-adrenoceptors with iv phenylephrine (or NE), then metformin did not acutely raise but rather acutely reduced arterial pressures.

Although both depressor and pressor responses to iv metformin were also observed in a few preliminary tests with normotensive female rats in the present study, the depressor responses in particular were notably less than those observed in SHR females. Similar results have been observed after iv administration of metformin to normotensive male rats compared to SHR males [22]. Taken together, these differences somewhat reflect results of long-term oral administration of metformin to humans. Such long-term metformin lowers arterial pressures much more frequently in women and men with pre-existing hypertension [1–8] than in those who are normotensive [9].

4.2 Adrenergic mechanism(s) of the acute depressor action intravenous metformin
Like hexamethonium in the previous study in which metformin was given iv to male SHR [22], chlorisondamine in the present study abolished the acute depressor response to iv injection of metformin in female SHR while simultaneously unmasking an acute pressor response to the drug. As chlorisondamine’s ganglionic blocking action is essentially complete in the rat while that of hexamethonium’s is not [23,24], the present results establish more clearly than the previous study that the depressor mechanism of iv metformin is dependent on intact autonomic (presumably sympathetic) neural function while the pressor mechanism of the drug is not. In addition, we confirmed other results from the previous study [22] that pretreatment with phentolamine also abolishes the depressor response to iv metformin and unmasks the pressor response to the drug, further indicating that the depressor mechanism is also dependent on intact alpha-adrenoceptor functions (presumably vascular) while the pressor mechanism is not. Together, these results now provide strong evidence that the mechanism responsible for at least the acute depressor action of iv metformin in the SHR is related to either (1) a rapid generalized decrease in postganglionic sympathetic neural release of NE onto vascular smooth muscle alpha-adrenoceptors and/or (2) a rapid generalized decrease in vascular smooth muscle contractile responsiveness to stimulation of alpha-adrenoceptors by NE.

However, the present study contains additional information which clearly supports the latter possibility and not the former. The acute depressor response to iv metformin in our SHR with intact autonomic function was accompanied by an increase and not a decrease in circulating NE. Thus, it is not likely due to a generalized decrease in postganglionic sympathetic neural activity as previously postulated [21]. A more likely possibility, consistent with the increased circulating NE and with the abolishing effects of chlorisondamine and phentolamine, is that the depressor action of iv metformin is due to a direct vasodilator action which clearly includes inhibition of vascular smooth muscle contractile responsiveness to alpha-adrenoceptor stimulation. All the previously reported results from direct administration of metformin acutely to either phenylephrine-contracted or NE-contracted arterial tissues in vitro support this possibility [25–27] and the present results from acute iv administration of metformin to rats in which hypertension was maintained with infused phenylephrine or NE also support it (e.g. Fig. 2). Furthermore, a direct vasodilator action is also consistent with results from long term oral administration of metformin at lower doses to both rats and humans. In rats, such long term metformin decreases mesenteric arterial vascular reactivity to NE ex vivo [36]. In humans, long term metformin has been associated with reduced total peripheral vascular resistance [7,9] and reduced systemic (systolic) pressor responses to intravenously infused NE and angiotensin II in vivo [16].

These considerations do not rule out an inhibitory action of metformin on postganglionic sympathetic neuroaxonal action potentials. Rather, the results of the present study suggest that such an inhibitory action either (1) may not be generalized or (2) is opposed by an equally generalized action of the drug on the postganglionic sympathetic nerve endings making more, not less, NE available overall to cardiovascular tissues. Indeed, the present results clearly raise the possibility that the drug can rapidly and directly stimulate release of NE from the sympathetic nerve endings independent of neuroaxonal action potentials or adrenal medullary function, as it increased circulating NE in the absence of intact autonomic ganglionic neurotransmission and the adrenal medulla. Recently, we have also found in vitro evidence suggestive of direct stimulation of NE release. In rat tail arterial tissue which contains abundant sympathetic nerve endings, the experimental removal of these nerve endings facilitates metformin’s relaxation of underlying smooth muscle contractions independent of the contractile agonist [26]. The most likely explanation of this facilitation is that in tail arterial tissues without functionally intact nerve endings metformin cannot stimulate release of endogenous NE which otherwise buffers the drug’s ability to relax underlying smooth muscle responsiveness to contractile agonists.

4.3 Adrenergic mechanism(s) of the acute pressor action of intravenous metformin
As previously observed in male SHR [22] we found that propranolol abolished tachycardic reponses to iv metformin both before and after ganglionic blockade in female SHR. However, unlike the previous study [22], we also found that propranolol greatly enhanced the depressor response to iv metformin in our female SHR with intact autonomic ganglia and markedly attenuated the pressor response to metformin in the same rats after ganglionic blockade with chlorisondamine. It is not readily apparent why these particular effects of propanolol on metformin’s arterial pressure effects were not observed in the previous study. The following differences between the two studies might be relevant. First, in our SHR with intact autonomic neural functions neither metformin nor propranolol (followed by metformin) were administered until at least 2–3 h after handling the conscious rats to attach their arterial catheters to the hemodynamic monitoring system. In addition, metformin when scheduled to be administered after propranolol was not injected for at least 40–45 min after the beta-blocker. In the previous study, these two time periods were reportedly much shorter, viz. approximately 40 min and 15 min, respectively [22]. Conceivably, this could decrease the likelihood of detecting a stimulatory effect of metformin on pressor-like beta-adrenoceptor functions if, for example, control levels of such functions had not yet fully stabilized at low enough values. Second, in our SHR scheduled for blockade of both autonomic ganglia and beta-adrenoceptors before metformin, we co-administered only the propranolol with the ganglionic blocking agent (chlorisondamine) before the injection of metformin. In the previous study, phentolamine as well as propranolol was co-administered with the ganglionic blocking agent (hexamethonium) before the metformin [22], thus raising the chances of nonspecific drug interactions which might interfere with the specific effects of beta-blockade. Finally, as indicated, our SHR were all females while only male SHR were used in the previous work [22]. Whether any of these differences are meaningful remains to be determined. Obviously, the latter is most worthy of further exploration because of its potential clinical relevance.

Whatever the reason for the different effects of propranolol between our work and the previous study [22], its effects on metformin’s actions before and after ganglionic blockade are at least internally consistent within the present study. Propranolol enhanced the depressor action of metformin before ganglionic blockade to about the same extent that it attenuated the pressor effect of the drug after the ganglionic blockade. This suggests that metformin may be activating the same beta-adrenoceptor-related pressor mechanism in the presence as well as in the absence of intact autonomic nerve function. We had thought that the drug might do this by directly stimulating release of E from the adrenal medulla (independent of sympathetic neural activity) and thereby stimulate cardiac beta-adrenoceptors sufficiently to raise cardiac output. This would also be consistent with the increases in heart rate before and after ganglionic blockade and, in turn, their blockade with propranolol. However, even though we found such increases in circulating E (immediately after injection of metformin), removal of the adrenal medulla abolished them but did not abolish the effects of metformin on arterial pressures or heart rates. Thus, we suspect that the beta-adrenoceptor component of the pressor action of iv metformin is mediated by the drug’s direct stimulation of NE release from postganglionic sympathetic nerve endings, including those in cardiac tissues adjacent to the dense populations of beta1-adrenoceptors in the sino-atrial node and on ventricular contractile myocytes. We further suspect that unlike alpha-adrenoceptor-mediated contractions in vascular smooth muscle, which metformin clearly inhibits [25–27] despite its apparent stimulation of NE release from the vascular sympathetic nerve endings [26], beta-adrenoceptor-mediated contractions in ventricular myocytes are not inhibited by metformin. Indeed, metformin may even enhance contractile responsiveness of ventricular myocytes to activation of their beta-adrenoceptors with NE. Chronic oral administration of metformin to streptozotocin-diabetic rats improves their intrinsic ventricular contractility [37].


   5 Conclusions and clinical relevance
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions and clinical...
Acknowledgments
 References
 
We conclude from this study that in female SHR the acute depressor action of iv metformin (1) is due to a direct vasodilator action which includes inhibition of alpha-adrenoceptor-dependent vasoconstriction and (2) is buffered by an acute beta-adrenoceptor-dependent pressor action, likely due to a direct metformin-induced release of NE from sympathetic nerve endings. Whether similar opposing actions of this drug are operative after its chronic oral administration at lower yet effective antidiabetic doses remains to be evaluated in future clinical studies. One recent study revealed higher 24-h heart rates in a group of type 2 diabetics when given metformin for 1 month than when crossed-over to glibenclamide therapy which provided the same level of glycemic control over a similar period of time [16]. Interestingly, 24-h blood pressure measures did not differ between the two treatments despite lower systemic pressor responses to intravenously infused NE and angiotensin II during the metformin [16]. In an identical 1 month crossover study with the same two drugs (again providing the same glycemic control) but in patients from an entirely different population of type 2 diabetics, heart rates were not higher and arterial pressures as well as systemic vascular resistances were lower during the metformin compared to either the glibenclamide or pretreatment periods [9]. Conceivably, in the first group of patients, a hidden pressor action of the metformin (marked by higher heart rates) was able to prevent reduction in 24 h pressures [16]. Based on our results, it would now seem appropriate to include circulating catecholamines along with more heart rate measures in future long-term clinical studies. Perhaps acute hemodynamic responses to a beta blocker should also be included. It is possible that elevated circulating NE (and even E), as well as elevated heart rates, might mark those particular individuals who are unresponsive to a long-term lowering of blood pressure by metformin and/or that a beta blocker might unmask the antihypertensive potential of the drug in these subjects.

Finally, our results suggest that the impact of gender on these measures should be given attention in future clinical studies. Gender has already been shown to influence the relationship of body weight to blood pressure in certain metformin-treated patients. Generally, long-term metformin is expected to lower body weight. Unfortunately, weight measures were often ignored in the clinical studies in which blood pressures were recorded. However, among one group of hyperlipidemics with hypertension, a greater weight loss and a stronger positive correlation of that loss to reduction in blood pressure was observed in women compared to men after 6 months of oral metformin therapy. Yet, their average reduction in blood pressure was similar [1]. One potential explanation for this paradox is that other actions of the drug may counterbalance weight-related reductions in blood pressure to a greater degree in women compared to men. Whether the pressor action uncovered in the female SHR of the present study plays a role in this regard remains to be determined but it is noteworthy that this action was greater in these animals ({Delta}MAP=+30±4 mmHg; Table 2, post ganglionic blockade) than Muntzel et al observed in male SHR given the same dose of metformin ({Delta}MAP=+24±6 mmHg) [22]. Moreover, with their autonomic ganglia intact, the depressor action of the same dose of metformin was smaller in the female SHR of the present study ({Delta}MAP=–17±3 mmHg; Table 2, no ganglionic blockade) than Muntzel et al observed in male SHR ({Delta}MAP=–26±3 mmHg) [22].

Time for primary review 27 days.


   Acknowledgments
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions and clinical...
 Acknowledgments
References
 
The author acknowledges the excellent secretarial assistance of Victoria L. Sears of Midwestern University. This study was funded by the NIH (#1-R15-HL/OD 56331-01).


   References
Top
 Abstract
 1 Introduction
 2 Methods
 3 Results
 4 Discussion
 5 Conclusions and clinical...
 Acknowledgments
 References
 

  1. Montaguti U., Celin D., Ceredi C., Descovich G.C. Efficacy of the long-term administration of metformin in hyperlipidaemic patients. Res. Clin. Forum. (1979) 1:95–103.
  2. Haupt E., Knick B., Koschinsky T., Liebermeister H., Schneider J., Hirche H. Oral antidiabetic combination therapy with sulphonylureas and metformin. Diabet Metab (1991) 17:224–231.[ISI][Medline]
  3. Landin K., Tengborn L., Smith U. Treating insulin resistance in hypertension with metformin reduces both blood pressure and metabolic risk factors. J. Int. Med. (1991) 229:181–187.[ISI][Medline]
  4. Calle-Pascual A.L., Garcia-Honduvilla J., Martin-Alverez P.J., et al. Comparison between acarbose, metformin and insulin treatment in type-2 diabetic patients with secondary failure to sulphonylurea treatment. Diabet. Metab. (1995) 21:256–260.[ISI][Medline]
  5. Giugliano D., De Rosa N., Di Maro G., et al. Metformin improves glucose lipid metabolism, and reduces blood pressure in hypertensive, obese women. Diabetes Care (1993) 16:1387–1390.[Abstract]
  6. Giugliano D., Quatraro A., Consoli G., et al. Metformin for obese, insulin-treated diabetic patients. Improvement in glycaemic control and reduction of metabolic risk factors. Eur J Clin Pharmacol (1993) 44:107–112.[CrossRef][ISI][Medline]
  7. Mendoza S.G., Faieta A., Carrasco H., Hamer T., Glueck C.J. Metformin lowers blood pressure, insulin resistance, triglycerides and endogenous estradiol in hypertensive men [abstract]. Clin. Res. (1994) 42:336A.
  8. Fanghanel G., Sanchez-Reyes L., Trujillo C., Sotres D., Espinosa-Campos J. Metformin’s effects on glucose and lipid metabolism in patients with secondary failure to sulfonylureas. Diabetes Care (1996) 19:1185–1189.[Abstract]
  9. Chan J.C.N., Tomlinson B., Critchley J.A.J.H., Cockram C.S., Walden R.J. Metabolic and hemodynamic effects of metformin and glibenclamide in normotensive NIDDM patients. Diabetes Care (1993) 16:1035–1038.[Abstract]
  10. Snorgaard O., Kober L., Carlsen J. The effect of metformin on blood pressure and metabolism in nondiabetic hypertensive patients. J Int Med (1997) 242:407–412.[CrossRef][ISI][Medline]
  11. Semplicini A., Del Prato S., Giusto M., et al. Short-term effects of metformin on insulin sensitivity and sodium homeostasis in essential hypertensives. J. Hypertens. (1993) 11(suppl_5):S276–S277.[ISI]
  12. Gudbjörnsdottir S., Friberg P., Elam M., Attvall S., Lönnroth P., Wallin B.G. The effect of metformin and insulin on sympathetic nerve activity, norepinephrine spillover and blood pressure in obese, insulin resistant, normoglycemic, hypertensive men. Blood Pressure (1994) 3:394–403.[Medline]
  13. Dorella M., Giusto M., Da Tos V., et al. Improvement of insulin sensitivity by metformin treatment does not lower blood pressure of nonobese insulin-resistant hypertensive patients with normal glucose tolerance. J. Clin. Endocrinol. Metab. (1996) 81:1568–1574.[Abstract]
  14. Hermann L.S., Schersten B., Bitzen P., Kjellström T., Lindgarde F., Melander A. Therapeutic comparison of metformin and sulfonylurea, alone and in various combinations: a double blind controlled study. Diabetes Care (1994) 17:1100–1109.[Abstract]
  15. Robinson A.C., Burke J., Robinson S., Johnston D.G., Elkeles R.S. The effects of metformin on glycemic control and serum lipids in insulin-treated NIDDM patients with suboptimal metabolic control. Diabetes Care (1998) 21:701–705.[Abstract]
  16. Sundaresan P., Lykos D., Daher A., Diamond T., Morris R., Howes L.G. Comparative effects of glibenclamide and metformin on ambulatory blood pressure and cardiovascular reactivity in NIDDM. Diabetes Care (1997) 20:692–697.[Abstract]
  17. Campbell I.W., Duncan C., Patton N.W., Broadhead T., Tucker G.T., Woods H.F. The effects of metformin on glycemic control, intermediary metabolism and blood pressure in non-insulin-dependent diabetes mellitus. Diabet. Med. (1987) 4:337–341.[ISI][Medline]
  18. Fontbonne A., Charles M.A., Juhan-Vague I., et al. The effect of metformin on the metabolic abnormalities associated with upper-body fat distribution. Diabetes Care (1996) 19:920–926.[Abstract]
  19. Wright A.D., Cull C.A., Holman R.R., Turner R.C. UKPDS 28 – A randomized trial of efficacy of early addition of metformin in sulfonylurea-treated type 2 diabetes. Diabetes Care (1998) 21:87–92.[Abstract]
  20. Petersen J.S., DiBona G.F. Acute sympathoinhibitory actions of metformin in spontaneously hypertensive rats. Hypertension (1996) 27:619–625.[Abstract/Free Full Text]
  21. Petersen J.S., Liu W., Kapusta D.R., Varner K.J. Metformin inhibits ganglionic neurotransmission in renal nerves. Hypertension (1997) 29:1173–1177.[Abstract/Free Full Text]
  22. Muntzel M.S., Abe A., Petersen J.S. Effects of adrenergic, cholinergic and ganglionic blockade on acute depressor responses to metformin in spontaneously hypertensive rats. J. Pharmacol. Exp. Ther. (1997) 281:618–623.[Abstract/Free Full Text]
  23. Abdel-Rahman A.R. Inadequate blockade by hexamethonium of the baroreceptor heart rate response in anesthetized and conscious rats. Arch Int Pharmacodyn Ther (1989) 297:68–85.[ISI][Medline]
  24. Santajuliana D., Hornfeldt B.J., Osborn J.W. Use of ganglionic blockers to assess neurogenic pressor activity in conscious rats. J. Pharmacol. Toxicol. Meth. (1996) 35:45–54.[CrossRef][ISI][Medline]
  25. Miller J.A., Peuler J.D. Rapid vasorelaxation by high levels of metformin [abstract]. FASEB J (1996) 10:A700.
  26. Lee J.M., Peuler J.D. Acute vasorelaxant effects of metformin and attenuation by stimulation of sympathetic agonist release. Life Sci. (1998) 64(4):PL57–PL63.[ISI]
  27. Chen X.-L., Panek K., Rembold C.M. Metformin relaxes rat tail artery by repolarization and resultant decreases in Ca2+ influx and intracellular [Ca2+]. J. Hypertens. (1997) 15:269–274.[CrossRef][ISI][Medline]
  28. Imai K., Higashidate S., Prados P.R., Santa T., Adachi-Akahane S., Nagao T. Relation between blood pressure and plasma catecholamine concentration after administration of calcium antagonists to rats. Biol. Pharm. Bull. (1994) 17:907–910.[ISI][Medline]
  29. Peuler J.D. Adrenergic depressor and pressor actions of metformin in spontaneously hypertensive rats [abstract]. Am. J. Hypertens. (1997) 10:107A.
  30. McClintock M.K. Estrous synchrony and its mediation by airborne chemical communication (Rattus norvegicus). Hormones and Behavior (1978) 10:264–276.[CrossRef][Medline]
  31. McClintock M.K. Estrous synchrony: Modulation of ovarian cycle length by female pheromones. Physiol. Behav. (1984) 32:701–705.[CrossRef][Medline]
  32. Peuler J.D., Schelper R.L. Partial protection from salt-induced stroke and mortality by high oral calcium in hypertensive rats. Stroke (1992) 23:532–538.[Abstract/Free Full Text]
  33. Peuler J.D., Johnson B.A.B., Phare S.M., Sowers J.R. Sex specific effects of an insulin secretagogue in stroke-prone hypertensive rats. Hypertension (1993) 22:214–220.[Abstract/Free Full Text]
  34. Peuler J.D., Edwards G.L., Schmid P.G., Johnson A.K. Area postrema and differential reflex actions of vasopressin and phenylephrine in rats. Am. J. Physiol. (1990) 258:H1255–H1259.[ISI][Medline]
  35. Peuler J.D., Johnson G.A. Simultaneous single isotope radioenzymatic assay of plasma norepinephrine, epinephrine, and dopamine. Life Sci. (1977) 21:625–636.[CrossRef][ISI][Medline]
  36. Verma S., Bhanot S., McNeill J.H. Decreased vascular reactivity in metformin-treated fructose-hypertensive rats. Metabolism (1996) 45:1053–1055.[CrossRef][ISI][Medline]
  37. Verma S., McNeill J.H. Metformin improves cardiac function in isolated streptozotocin-diabetic rat hearts. Am J Physiol (1994) 266:H714–H719.[ISI][Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Peuler, J. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Peuler, J. D.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?