Early Effects in Perivascular Nerves and Arterial Media Following Renal Artery Denervation
To the Editor:
In the past years 2 different innovative methods of hypertension treatments were investigated. The first is the promising examination of the electrical activation of the carotid baroreceptors, which is now in the phase III studies,1,2 and second the usage of selective renal sympathetic denervation (RSD) as an alternative treatment in therapy-resistant hypertension. RSD is thought to be based on alteration of the sympathetic innervation of kidney and secondary effects on the renin–angiotensin–aldosterone system. The renal nerves are located in close vicinity to the renal artery wall. Their activation could be important for the progression of therapy-resistant hypertension. In this context, Krum et al3 showed the proof of principle of a percutaneous, catheter-based trail to destroy renal sympathetic nerves by the introduction of a RF catheter into the lumen of the main renal artery and its subsequent connection to a radiofrequency generator. Until recently, the renal denervation seemed to be a promising therapy option for antihypertensive treatment. However, the simplicity HTN-3 study as recently announced4 indicated a lack of effect of RSD. Nevertheless, at present, the underlying molecular mechanisms remain unclear during RSD. RF injury might damage the tissue by thermal coagulation but may also cause apoptosis as observed in colon cancer.5 Thus, we made up the hypothesis that RSD may induce apoptosis in the perivascular nerves. The initiation of apoptosis is still not entirely understood. One important pathway is the release of apoptosis-inducing factor (AIF) from the mitochondrial intermembrane space together with cytochrome c.6 Subsequently, AIF translocates to the nucleus and induces chromatin decondensation and DNA degradation, likely by endonucleases. Typically, when inducing apoptosis, AIF and caspases act together. However, AIF may induce cell death in a caspase-independent manner.7 The extrinsic pathway of apoptosis leads to an activation of the initiator caspase 8. However, within the instrinsic pathway various types of cell stress, including DNA damage and oxidative stress (and many more), cause mitochondrial release of cytochrome c and other factors, resulting in activation of caspase 9. These initiator caspases enable activation of so-called effector caspases, such as caspase 3, subsequently activating the apoptotic program.
Here, we wanted to investigate how acute RSD affects perivascular nerves and renal arteries, and whether apoptosis in nerves and vessel wall might be induced.
A detailed outline of the methods is given in the online-only Data Supplement.
All methods were approved by an institutional ethical board and are in accordance with the Guide for Care and Use of Laboratory Animals (National Institutes of Health).
In anaesthetized pigs (n=5; 76.2±2.13 kg; midazolam, thiopental, and fentanyl), radiofrequency ablation (4F electrode-tipped catheter, Symplicitiy Catheter System, Medtronic Ardian, Mountain View, CA) was performed. Under angiographic control the catheter was inserted within the main stem of right renal artery (first ablation: near aorta; second: near kidney; Figure S1 in the online-only Data Supplement). Ablations (90 s, 8 W, 60°C) were performed accordingly.3 Blood clotting was prevented by intravenous heparin. To allow manifestation of early apoptosis after RSD animals were kept for 3 hours under general anesthesia. Finally, renal arteries were excised (right renal artery with RSD; left as control without RSD).
Renal arteries (n=10 ablation points and n=10 pieces of control arteries) and perivascular nerves were fixed with neutrally buffered formalin, embedded in paraffin, cut in 4-μm thick sections, stained with hematoxylin and eosin. To investigate early apoptotic signals in nerves, immunohistochemical AIF staining with anti-AIF rabbit polyclonal (1:100, Santa Cruz) antibody was performed (detection: EnVision FLEX/AEC+Substrate-Chromogen [Dako]) with hematoxylin nuclear counterstaining. Activation of apoptosis was defined as the translocation of AIF into the nuclei. The same procedure was performed with anti-cleaved caspase 3 (Asp 175) rabbit polyclonal (1:250, Cell signaling) antibody to detect caspase-dependent apoptosis. Sections were analyzed by a blinded observer.
For statistics ANOVA and Student t test with Bonferroni correction were performed on level of significance P≤0.05 and data were presented as mean value±SEM.
In ablated arteries we always found AIF and cleaved caspase 3–positive perivascular nerves. However, not all nerves in a section were positive: after RSD 33 of 64 nerves were positive for AIF. AIF-positive nerves (between 1 and 10 nerves per sample) were located in tunica adventitia (n=31) or in the border zone between tunica adventitia and tunica media (n=2) and occupied an average area of 0.031±0.016 mm2 with an average diameter of 198.67 μm. Inside the perivascular nerves we found 18.6±1.25% AIF-positive nuclei that indicated early induction of apoptosis. In nonablated control vessels only 1 of 31 AIF-positive nerve was detectable (Figures 1A, 1B, and 2A). Interestingly, the majority (n=20) of these AIF-positive perivascular nerves was >1000 μm away from the endothelium, whereas only 1 was located <500 μm from the endothelium. The remaining perivascular nerves were found 500 to 1000 μm distant to the endothelium (n=12; Figure 2B).
CC3 staining showed similar results (Figure 1C and 1D). Twenty of 55 nerves were positive for CC3 (control: 3 of 24 nerves CC3 positive). One to 5 positive perivascular nerves were detected within each sample, located in the tunica adventitia. The majority of CC3-positive nerves was larger than 5000 μm2 and >1000 μm far from the endothelium of the respective vessel. The area of the nerves averaged ≈0.020±0.009 mm2 with a diameter of 159.73 μm. About 12.38±1.9% of the nuclei within the nerves were CC3 positive (Figure 2).
In addition, we found that the integrity of the endothelium and the intimae was not destroyed or dissociated in all specimens. We also found minor areas of media with hypereosinophilic cells, which may be indicating damage (Figure 3). Further analysis of tunica media indicated AIF and CC3-positive fields in the media (AIF, 54.72±8.7%; CC3, 42.4±10.0%) in which 38.67±5.3% of counted cells were CC3 positive and 28.67±4.02% showed AIF translocation (Figure 4).
In hematoxylin and eosin and AIF and CC3 staining of control group, we did not found any destruction of media and induction of AIF translocation into nuclei in medial fibroblasts. In ≈12% of fields of vision, we found 15.18±3.76% CC3-positive medial fibroblasts.
As expected, blood pressure and heart rate remained unaltered after RSD (within 3 hours; Table S1 in the online-only Data Supplement).
To reduce renal sympathetic afferent and efferent nerves in patients with resistant hypertension, catheter-based radiofrequency ablation of renal sympathetic nerves was positively investigated.3,8,9 In contrast, recent statements on simplicity HTN-3 study indicated no significant difference in blood pressure between sham-treated and RSD-treated patients after 6 months.4 However, as long as the underlying mechanisms are mostly unclear, these data are difficult to interpret.
To the best of our knowledge, our data show for the first time that RSD initiates apoptosis in perivascular nerves, as indicated by AIF translocation and cleavage of caspase 3. According to the current understanding, after caspase 3 activation (cleavage) the cell is doomed to die within the next hours to days. Interestingly, in those clinical RSD studies which showed a positive effect, the blood pressure lowering was also not seen immediately after RSD but several days later.3 Another interesting next aspect is that mainly nerves in >1000 μm distance from the endothelium were affected. This may be because of local factors affecting the RF energy: at the catheter tip the blood flow absorbs parts of the RF energy thereby cooling intima and inner parts of the vessel wall. In outer parts (although energy decays with ) RF energy seems to be high enough to induce damage (Figure 5).
However, we found nuclear AIF translocation and CC3, early markers for initiated apoptosis in only ≈50% of the perivascular nerves 3 hours after ablation. Thus, there is no complete denervation (at least with the energy as used here, which is identical with the study by Krum et al3). Although it is tempting to speculate that such an incomplete denervation may be compensated by surviving nerves, a conclusion on this point requires a chronic detailed study including 6-month follow-up blood pressure measurements.
In consequence, one could consider the use of higher energy to optimize the RSD effect. In this case it is important that we also found AIF translocation into nuclei and CC3-positive cells in the media of renal arteries indicating beginning media apoptosis. Thus, further increase in RF energy probably will also cause more media damage.
Sources of Funding
This study was financed by Heart Centre Leipzig.
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.114.03241/-/DC1.
- © 2014 American Heart Association, Inc.
- Illig K,
- Levy M,
- Sanchez L,
- Trachiotis GD,
- Shanley C,
- Irwin E,
- Pertile T,
- Kieval R,
- Cody R
- Tordoir JHM,
- Scheffers I,
- Schmidli J,
- Savolainen H,
- Liebeskind U,
- Hansky B,
- Herold U,
- Irwin E,
- Kroon AA,
- de Leeuw P,
- Peters TK,
- Kieval R,
- Cody R
- Krum H,
- Schlaich M,
- Whitbourn R,
- Sobotka PA,
- Sadowski J,
- Bartus K,
- Kapelak B,
- Walton A,
- Sievert H,
- Thambar S,
- Abraham WT,
- Esler M
- Dougherty W,
- Warren J
- Clerk A,
- Sugden PH