|Home | About | Journals | Submit | Contact Us | Français|
Postural tachycardia syndrome (POTS) is associated with increased plasma angiotensin-II (Ang-II). Ang-II administered in the presence of NOS inhibition with nitro-L-arginine (NLA) and AT1R blockade with losartan produces vasodilation during local heating in controls. We tested whether this “Ang-mediated vasodilation” occurs in POTS and whether it is related to angiotensin converting enzyme 2 (ACE2), and Ang-(1–7). We used local cutaneous heating to 42°C and laser Doppler Flowmetry to assess NO-dependent conductance at four calf sites in 12 low flow POTS and in 12 control subjects aged 17.6–25.5 years. We perfused Ringer’s solution through intradermal microdialysis catheters and performed local heating. We perfused one catheter with NLA (10mM)+losartan (2 µg/L), repeated heating, and NLA+losartan+Ang-II (10 µM) repeating heating a third time. A second catheter received NLA+losartan+Ang-II, heated, perfused NLA+losartan+Ang-II+DX600 (1 mM, an ACE2 inhibitor), and reheated. A third catheter received NLA+losartan+Ang-II, heated, perfused NLA+losartan+Ang-II+angiotensin-(1–7) [100µM, Ang-(1–7)], and reheated. The fourth catheter received Ang-(1–7) then reheated a second time only. Ang-mediated vasodilation was present in control but not POTS. Ang-mediated dilation was eliminated by DX600 indicating an ACE2-related effect. Ang-mediated vasodilation was restored in POTS by Ang-(1–7). When administered alone during locally mediated heating, Ang-(1–7) improved the NO-dependent local heating response. ACE2 effects are blunted in low flow POTS and restored by the ACE2 product angiotensin-(1–7). Data imply impaired catabolism of Ang-II through the ACE2 pathway. Vasoconstriction in POTS may result from a reduction in Ang-(1–7) and an increase in Ang-II.
Chronic orthostatic intolerance is related to postural tachycardia syndrome (POTS) 1–5. POTS is defined by excessively increased heart rate during orthostatic challenge associated with symptoms of orthostatic intolerance 3 including dizziness, exercise intolerance, headache, fatigue, memory problems, nausea, blurred vision, pallor, and sweating which improve with recumbence. Findings have been ascribed to increased sympathetic activity 6;7. We described a subset of “low flow POTS” in which marked upright tachycardia is associated with supine pallor, acrocyanosis, and hypovolemia, tachycardia, decreased cardiac output, and increased peripheral resistance 8. We observed increased plasma angiotensin-II (Ang-II) 9;10 not accounted for by changes in angiotensinogen, renin, or angiotensin converting enzyme (ACE). Increased Ang-II and decreased nitric oxide (NO) increase central 11;12 and peripheral sympathetic activity 13;14.
Recently, we investigated a model of Ang-II - NAD/NADPH oxidase superoxide production which scavenges NO reducing its bioavailability 15;16. Experiments made extensive use of the vasodilation response of non-glabrous skin to local heating 17;18 which has three distinct phases: an initial peak, a nadir, and an increase to a plateau. All phases, especially the plateau phase, are NO dependent 17;18. The plateau has been used to test for bioavailability of NO during intradermal microdialysis 19–21. The local heating response is blunted in low flow POTS 22 and blunting is reversed by angiotensin-II type-1 receptor (AT1R) blockade 20.
We studied intradermal perfusion of exogenous Ang-II in healthy volunteers to replicate blunted vasodilation to local heating 22. Perfusion of Ang-II in the presence of NO synthase inhibition and AT1R blockade caused vasodilation during local heating shown in Figure 1. We denote this response as “Ang-mediated vasodilation”. We previously showed that Ang-mediated vasodilation was unrelated to angiotensin-II type-2 receptor stimulation 23. We hypothesized that the carboxydipeptidase angiotensin converting enzyme 2 (ACE2) could account for Ang-mediated vasodilation. ACE2 is the main catabolic enzyme for Ang-II, synthesizing the heptapeptide Mas receptor agonist angiotensin-(1–7) [Ang-(1–7)] 24. A deficit in ACE2 might account for increased Ang-II and decreased local heating response in POTS.
The current study was therefore designed to explore the following aims:
We studied the effects of Ang-II, NOS inhibition and AT1R-blockade, ACE2 blockade and Ang-(1–7) administration on the local heating response in 12 healthy volunteers aged 18.1–25.5 years, median age 21.1 years (3 male and 9 female) and in 12 low flow POTS patients aged 17.6– 24.5, (2 male and 10 female). Subjects were excluded as controls for a history of orthostatic intolerance. All subjects were free from cutaneous, systemic, and cardiovascular diseases, were not taking medications, and refrained from alcohol and caffeinated beverages for at least 24 hours prior to study. There were no smokers or trained competitive athletes. There were no bed rested subjects. POTS patients were referred for chronic orthostatic intolerance lasting at least three months. Symptoms included day-to-day dizziness, exercise intolerance, headache, fatigue, memory problems, nausea, blurred vision, pallor, and abnormal sweating while upright, relieved by recumbence. The diagnosis of POTS was made during a 10 minute screening 70° upright. POTS was diagnosed when symptoms of orthostatic intolerance were associated with an increase in HR exceeding 30 beats per minute or to a rate exceeding 120 beats per minute during 10 minutes of tilt 4;25. POTS patients were subgrouped by supine calf blood flow measured with venous occlusion plethysmography 26 into “low flow POTS” with decreased blood flow and non-low flow POTS 8;27. Patients were retained if they belonged to the low flow POTS group. Female subjects were enrolled without regard to the phase of their menstrual cycle.
Experiments were performed on one day. Four microdialysis catheters were placed to infuse drugs locally into the intradermal space of the leg. Prior to microdialysis catheter insertion, laser Doppler flow (LDF) was measured over each of the 4 insertion sites to estimate baseline flows for later use in determining recovery from catheter insertion. Laser probes were removed and 4 microdialysis catheters were inserted. After recovery, LDF was measured for 10 minutes while perfusing lactated Ringer’s solution and again during local heating at each site. Subjects completely recovered from heating for 30–60 minutes until the preheat baseline flow was achieved. Thereafter, catheters were sequentially perfused with vasoactive drugs for 40 minutes, local heating repeated, heat recovery repeated, and catheters 1–3 were perfused with additional drugs for 40 minutes, and local heating repeated a third time. At the end of experiments all catheters were perfused with 28mM sodium nitroprusside for the determination of maximum cutaneous blood flow conductance (CVC max) 28. All chemicals were highly purified to USP standards and were produced under sterile conditions. The catheters were sterilized. Microdialysis membranes act as micropore filters preventing any possibility of infection. Informed consent was obtained. The IRB of New York Medical College approved all protocols.
To determine whether Ang-mediated vasodilation is present in POTS, we perfused the first catheter with the AT1R antagonist losartan combined with the non-isoform specific NOS inhibitor nitro-L-arginine (NLA) and repeated local heating. After recovery from the second heating response we perfused the same catheter with losartan + NLA + Ang-II and repeated local heating a third time.
To test whether ACE2 is involved in the Ang-mediated vasodilation in control subjects and POTS patients we perfused the second catheter with losartan+NLA+Ang-II and repeated local heating. After recovery we perfused the catheter with losartan+NLA+Ang-II to which the specific ACE2 inhibitor DX600 was added and repeated local heating repeated a third time.
To test whether perfusion with Ang-(1–7) restores Ang-mediated vasodilation in POTS we perfused the third catheter with losartan+NLA+Ang-II and repeated local heating. After recovery we perfused the catheter with losartan+NLA+Ang-II to which Ang-(1–7) was added and repeated local heating a third time.
To test whether perfusion with Ang-(1–7) restores the local heating response in POTS we studied whether Ang-(1–7) alleviates blunting of the local heating response. Therefore, we perfused the fourth catheter with Ang-(1–7) and repeated local heating a second time only.
We used a heat-reheat scheme in which the heating response was assumed unaffected by time or repeat measurements. We have verified this assumption in studies showing that intra-catheter differences of plateau phase measurements using heat-reheat are far smaller than inter-catheter differences of local heating plateaus in a given subject 10.
All testing was conducted in a temperature controlled room (approximately 25°C) at least two hours after a light breakfast. Supine subjects were instrumented in the dermal space of the lateral aspect of the left calf after hair was gently removed. The leg was always at heart level. Microdialysis insertion sites were cooled with ice-packs prior to insertion to reduce discomfort. Each probe (MD-2000 Linear Microdialysis Probes, Bioanalytical Systems, West Lafayette, IN) has a 10 mm microdialysis membrane section that is placed in the intradermal space using a 25 gauge needle as an introducer. Catheters were randomly designated. The molecular weight cutoff is nominally 10,000 Daltons.
Following placement, catheters were initially perfused with Ringer’s solution at 2 µl/min. A 7 element integrating LDF probe (Probe 413, Perimed, Stockholm) was placed directly over the center of each microdialysis catheter to measure LDF. LDF was thereafter recorded until values returned to pre-insertion baseline which indicated recovery from catheter placement trauma and usually occurred by 60–90 minutes 29.
Once returned to baseline subjects received perfusate containing lactated Ringer’s solution in all catheters and local heating was performed. Following recovery from local heating, 2 µg/L losartan+10 mM NLA was perfused through the first catheter; 2µg/L losartan+10mM NLA+10 µM Ang-II was perfused through the second and third catheters; and 100µM Ang-(1–7) was perfused through the fourth catheter. All drugs were dissolved in lactated Ringer’s solution and were perfused for at least 40 minutes. Heating was repeated. After recovery from local heating, 2µg/L losartan+10mM NLA+10 µM Ang-II were perfused through the first catheter ; 2µg/L losartan+10mM NLA+10 µM Ang-II+1mM DX600 were perfused through the second catheter, and 2µg/L losartan+10mM NLA+10 µM Ang-II+100µM Ang-(1–7) was perfused through the third catheter.
Doses of NLA and Ang-II were chosen on the basis of pilot studies showing this to be the minimum concentration yielding maximum attenuation of the NO-dependent local heating plateau 17;18. Losartan 2 µg/L was chosen based on prior human testing 30. Ang-(1–7) 100µM dose was chosen on the basis of pilot studies showing this to be the lowest concentration needed to restore the Ang-mediated vasodilation in POTS patients. DX600 1mM was chosen as the lowest concentration that suppressed the Ang-mediated vasodilation in control subjects.
Once baseline LDF values were obtained the areas under each laser were gradually heated at 1°C/10seconds to 42 °C until a plateau was reached. Heat was turned off to allow for recovery to baseline LDF. Investigators 17;18 have demonstrated that the first heating peak is mediated by neurogenic reflexes, NO, and neuropeptides 31;32. The first peak is followed by a nadir and then an NO-dependent plateau which is blunted by NOS inhibition.
Heart rate was monitored by electrocardiography and right extremity BP was measured by finger plethysmography (Finometer, Amsterdam) intermittently recalibrated against oscillometry. Mean arterial pressure (MAP) was obtained by averaging the signal over 5 minutes and compared against oscillometry (using the formula MAP= (SAP+2*DAP)/3. Finometer and oscillometric blood pressure were in agreement.
LDF was measured in arbitrary perfusion units (pfu). Continuous LDF data were sampled at 200 Hz. LDF data were multiplexed and interfaced to a personal computer through an A/D converter (DI-720, DATAQ industries, Milwaukee) using custom acquisition software. LDF data were converted to units of cutaneous vascular conductance (CVC) by dividing by the MAP. CVC measurements were then converted to a percent maximum conductance (%CVCmax) by dividing CVC by the maximum CVC achieved after the administration of 28mM sodium nitroprusside at the end of experiments. This fraction was converted to a percentile by multiplication by 100.
Changes in baseline LDF before and after drugs were compared by two-way ANOVA. Results are reported as mean ± standard error. Comparisons were made by repeated measures ANOVA to look at differences in the local heating response between pre and post drug infusion using the particular microdialysis catheter as the within factor. We also compared post-drug responses to NLA+losartan to responses to Ang-II+losartan+ NLA, and compared post-drug responses to Ang-II+losartan+NLA to post drug responses to Ang-II+losartan+NLA+DX600, and to Ang-II+losartan+NLA+Ang-(1–7) using catheters as the between factor and subjects as the within factor. Graphical representations comparing POTS and control subject data were obtained by averaging heat responses over all local heating curves for all subjects. All averaged numerical data ± standard error at baseline, first thermal peak, nadir, and plateau are tabulated in Table 2–Table 5. Results were calculated using SPSS software version 11.0. The minimum value for alpha was <0.05.
Data are shown in Table 1. POTS patients had reduced body mass index (BMI, P<.025) because they weighed less than control subjects (P<.025). Supine heart rate was increased in POTS (P<.01). Pulse pressure was reduced in POTS (P<.025). Systolic and diastolic pressures were not different. Upright HR was increased and calf blood flow was decreased in POTS by definition. Calf arterial resistance was increased in POTS patients (P<.001). Resting LDF (pfu) and resting %CVCmax were decreased in POTS compared to control subjects (P<.025). There was no difference in maximum LDF.
Table 2–Table 5 show results from individual catheter experiments. We compared the first peak, the nadir, and the plateau phase before and after drug treatments. Results are also depicted in Figure 2–Figure 5 which shows averages over all POTS patients and control subjects.
As shown in Table 2 and Figure 2, %CVCmax of the pre-drug baseline (P<.01), the first peak (P<.025), the nadir (P<.05) and the plateau phase (P<.001) were decreased in POTS compared to control subjects. After NLA+losartan were administered, control heating responses were decreased (P<.025 for first peak, P<.01 for nadir, P<.001 for plateau) compared to pre-drug. POTS and control heating responses became similar. After Ang-II was added, all heated phases increased in healthy control subjects (P<.05 first peak, P<.01 nadir, P<.0 01 plateau) but not in POTS in whom all heating phases remained less than control. Thus, an increase in NO and AT1R independent angiotensin-mediated vasodilation with local heating (Ang-mediated vasodilation) occurs in control subjects but not in POTS patients.
Table 3 and Figure 3 compare the response to DX600 following the administration of NLA+losartan+Ang-II. The addition of DX600 produced a significant decrease in the control heating response with decreased first peak (P<.05), nadir (P<.001) and plateau (P<.001). There was no significant change in the heating response for POTS. Following the addition of DX600, POTS and control heating responses were similar with no significant differences observed during any phase of heating. Thus, ACE2 inhibition causes a decrease in Ang-mediated vasodilation in control but not in POTS subjects.
Table 4 and Figure 4 compare the response to Ang-(1–7) following the administration of NLA+losartan+Ang-II. The combination of NLA+losartan+Ang-II produced a decrease in plateau flow for control subjects (P<.025). The addition of Ang-(1–7) produced a significant increase in the POTS heating response with increased first peak (P<.01), nadir (P<.05) and plateau (P<.025). There was no significant change in the control subjects, POTS and control subject heating responses were not different after Ang-(1–7) was added. Ang-(1–7) significantly increases Ang-mediated vasodilation in POTS but not in control subjects.
Table 5 and Figure 5 compare the response to Ang-(1–7) during a simple local heating experiment in control and POTS subjects. The addition of Ang-(1–7) produced a small non-significant increase in the control heating response. On the other hand, POTS patients’ baseline (P<.05), nadir (P<.001) and plateau phase (P<.001) increased significantly. POTS and control subject heating responses were similar after Ang-(1–7) was added although there remained a small difference in the plateau phase (P<.05). Thus, Ang-(1–7) significantly increases the local heating response in POTS but not in control subjects.
We studied the low flow POTS subgroup using the cutaneous microcirculation to investigate potential angiotensin-related defects. Our results suggest a deficit in ACE2 resulting in impaired metabolism of Ang-II, producing increased local concentrations of Ang-II and decreased concentrations of Ang-(1–7), and contributing to excessive vasoconstriction characteristic of this disorder. These conclusions are based on the following findings:
The first aim of this study was to determine whether NO and AT1R independent angiotensin-mediated vasodilation with local heating (Ang-mediated vasodilation) is present in POTS. The data in Table 2 and in Figure 2 show no effect of the addition of Ang-II to NLA+losartan in POTS while there is a significant effect in control subjects. We reasoned that the absent response may help to explain angiotensin-related abnormalities in POTS
Past work indicated that cutaneous Ang-mediated vasodilation was unrelated to AT2R stimulation. Our second aim, was to test whether ACE2 could be involved. Results shown in Table 3 and Figure 3 demonstrate that cutaneous ACE2 inhibition eliminates the Ang-mediated vasodilation in control subjects but not in POTS patients. The results of aims 1 and 2 suggest that Ang-mediated vasodilation is produced by cutaneous ACE2 in healthy control subjects and that the healthy response is absent in POTS. ACE2 metabolizes Ang-II producing Ang-(1–7) 24. ACE2 also produces Ang-(1–7) via angiotensin-(1–9) by the enzymatic degradation of angiotensin-I 33. The data imply a deficit of ACE2. If generalized this could account for increased Ang-II and reduced Ang-(1–7) in POTS patients.
Data from aims 1 and 2 predict aim 3, that restoring Ang-(1–7) with exogenous Ang-(1–7) could enhance Ang-mediated vasodilation with local heating in POTS. Data in Table 4 and Figure 4 demonstrate restoration of Ang-mediated vasodilation with local heating in POTS once Ang-(1–7) is perfused without significant effect in control subjects.
Aim 4 reinforces the importance of Ang-(1–7) and ACE2 because Ang-(1–7) is able to improve the blunted heating response in POTS. The data support prior observations of cutaneous nitric oxide deficiency in POTS 10 caused by excessive Ang-II 20.
ACE2 has assumed increasing importance as an enzyme which converts angiotensin-I to Ang-(1–9) which is thereafter converted to Ang-(1–7). ACE2 also removes angiotensin-II by catabolism to Ang-(1–7) 33;34. ACE2 is expressed in most tissues including skin 35. Ang -(1–7) is regarded as the principal counter regulatory mechanism for angiotensin-II 36. Binding of Ang-(1–7) to Ang-(1–7) receptors produces vasodilation, and antiproliferative -antihypertrophic effects. Ang (1–7) may interfere with Ang-II directly at the AT1R 37;38 or through other pathways `including antagonism of ACE 39. While Ang-(1–7) may exert dilator effects through bradykinin 40, its principal ligand is the recently discovered Mas receptor 41. During our experiments, NOS was inhibited and AT1R were blocked; actions mediated through ACE2 and Ang-(1–7) should have been independent of NO or AT1R. Ang-(1–7)-Mas receptors have multiple vasodilating effects: these include responses mediated by AKT-NO upregulation, direct Ang-II antagonism and reduction of cyclooxygenase-2 42;43. It is probable that the vasodilating effects of Ang-(1–7) repletion are not completely evidenced during Ang-mediated vasodilation.
Experiments using direct intradermal administration of ACE2 would restore endogenous Ang-(1–7) and the local heating response. Unfortunately, ACE2 was not available.
Skin can be a representative microvascular surrogate in illnesses such as low flow POTS, hypertension, hypercholesterolemia, heart failure and diabetes where there are widespread and generalized pathophysiological microvascular effects44. It has the advantages of ubiquity, easy access, and experiments lack systemic perturbation. Our work indicates that vascular abnormalities occur throughout the circulation in POTS; local flow abnormalities in skin can be generalized. However, the cutaneous circulation has unique autonomic control and may not be the best representative organ for understanding angiotensin-II relationships.
We studied females without regard to menstrual cycle. The phase of the menstrual cycle can affect nitric oxide dependent mechanisms. However, NO was blocked throughout experiments. Also, we found directionally consistent and similar results across all subjects. There was no evidence suggesting a relationship between menstrual cycling and changes in POTS symptoms or signs.
Endogenous angiotensin was not considered. Many tissues produce Ang-II but there are no skin data. However, data from skeletal muscle microvessels suggest local concentrations on the order of 100 pmoles/liter 45. This was similar to our lowest dose of angiotensin-II administered during prior dose-response measurements but was far less than exogenous administered angiotensin-II in current experiments. Alternatively, we could have sought to eliminate local Ang-II production with an ACE inhibitor. However, since ACE is also bradykininase, use of an ACE inhibitor enhances bradykinin release altering microvascular conductance.
Microdialysis is invasive and alters the interstitial milieu. The work of Anderson et al (2) suggests that flow responses return to baseline levels within approximately 1 hour. In pilot experiments we measured baseline flows, removed the LDF probes, instrumented the same site with microdialysis catheters, replaced the probes, waited at least an hour and repeated the LDF measurements with (on average) similar results.
The local cutaneous heating model has been useful in dissecting biochemical pathways, but its relevance to POTS is uncertain. Specifically, all of the results entail vasodilating responses to heat, not at baseline. POTS patients are often intolerant to heat. Also, it is unclear how vasoconstriction can lead to POTS. However, euthermic skin is highly vasoconstricted with very small changes occurring with further vasoconstriction. Experiments concerning constrictor mechanisms often include a predilation step. Local heating is particularly useful because it produces a near maximum increase in LDF in control subjects but a much smaller increase in POTS patients. Local skin heating is distinct from core heating which diverts blood from the systemic circulation into the skin, worsening orthostatic tolerance. Local heating produces only local effects primarily due to NO. Our findings suggest that an ACE2 deficit explains increased Ang-II in POTS. ACE2 deficit also decreased Ang-(1–7). The increase in Ang-II and decreased Ang-(1–7) are associated with increased vascular resistance, and decreased vascular capacitance, blood volume, and cardiac output which respond poorly to orthostasis causing orthostatic intolerance akin to the physiology of pheochromocytoma.
Postural tachycardia syndrome is associated with a “hyperadrenergic state” of sympathoexcitation and widespread vasoconstriction. Our earlier work showed that cutaneous and systemic angiotensin-II is increased in low flow POTS in which nitric oxide bioavailability is reduced. We have now shown that there is blunted activity of angiotensin converting enzyme 2 which is the central pathway for Ang-II degradation. This may be the first human illness in which this deficit has been detected. The findings transcend the tachycardia syndrome and point to mechanisms for angiotensin-II excess and sympathetic activation that are potentially present in diverse systemic illnesses of vascular regulation.
Sources of Funding
Supported by R01HL074873, R01HL087803, R21HL091948 from the National Heart Lung and Blood Institute of the National Institutes of Health.
Conflicts of Interest/Disclosures
The authors have nothing to disclose to the Publications Office concerning any potential conflict of interest (e.g., consultancies, stock ownership, equity interests, patent-licensing arrangements, lack of access to data, or lack of control of the decision to publish).