Abstract
Background Intimal hyperplasia (IH) remains a major limitation in the long-term success of any type of revascularization. IH is due to vascular smooth muscle cell (VSMC) dedifferentiation, proliferation and migration. The gasotransmitter Hydrogen Sulfide (H2S) inhibits IH in pre-clinical models. However, there is currently no clinically approved H2S donor. Here we used sodium thiosulfate (STS), a clinically-approved source of sulfur, to limit IH.
Methods Hypercholesterolemic LDLR deleted (LDLR-/-), WT or CSE-/- male mice randomly treated with 4g/L STS in the water bottle were submitted to focal carotid artery stenosis to induce IH. Human vein segments were maintained in culture for 7 days to induce IH. Further in vitro studies were conducted in primary human vascular smooth muscle cell (VSMC).
Findings STS inhibited IH in mice and in human vein segments. STS inhibited cell proliferation in the carotid artery wall and in human vein segments. STS increased polysulfides in vivo and protein persulfidation in vitro, which correlated with microtubule depolymerization, cell cycle arrest and reduced VSMC migration and proliferation.
Interpretation STS, a drug used for the treatment of cyanide poisoning and calciphylaxis, protects against IH in a mouse model of arterial restenosis and in human vein segments. STS acts as an H2S donor to limit VSMC migration and proliferation via microtubule depolymerization.
Funding This work was supported by the Swiss National Science Foundation (grant FN-310030_176158 to FA and SD and PZ00P3-185927 to AL); the Novartis Foundation to FA; and the Union des Sociétés Suisses des Maladies Vasculaires to SD.
Evidence before this study Intimal hyperplasia (IH) is a complex process leading to vessel restenosis, a major complication following cardiovascular surgeries and angioplasties. Therapies to limit IH are currently limited. Pre-clinical studies suggest that hydrogen sulfide (H2S), an endogenous gasotransmitter, limits restenosis. However, despite these potent cardiovascular benefits in pre-clinical studies, H2S-based therapeutics are not available yet. Sodium thiosulfate (Na2S2O3) is an FDA-approved drug used for the treatment of cyanide poisoning and calciphylaxis, a rare condition of vascular calcification affecting patients with end-stage renal disease. Evidence suggest that thiosulfate may generate H2S in vivo in pre-clinical studies.
Added value of this study Here, we demonstrate that STS inhibit IH in a surgical mouse model of IH and in an ex vivo model of IH in human vein culture. We further found that STS increases circulating polysulfide levels in vivo and inhibits IH via decreased cell proliferation via disruption of the normal cell’s cytoskeleton. Finally, using CSE knockout mice, the main enzyme responsible for H2S production in the vasculature, we found that STS rescue these mice from accelerated IF formation.
Implications of all the available evidence These findings suggest that STS holds strong translational potentials to limit IH following vascular surgeries and should be investigated further.
1. Introduction
Prevalence of peripheral arterial disease continues to rise worldwide, largely due to the combination of aging, smoking, hypertension, and diabetes mellitus (1-3). Vascular surgery, open or endo, remains the best treatment. However, the vascular trauma associated with the intervention eventually lead to restenosis and secondary occlusion of the injured vessel. Re-occlusive lesions result in costly and complex recurrent end-organ ischemia, and often lead to loss of limb, brain function, or life. Despite the advent of new medical devices such as drug eluting stent and drug-coated balloons, restenosis has been delayed rather than suppressed (4), and limited therapies are currently available.
Restenosis is mainly due to intimal hyperplasia (IH), a process instated by endothelial cell injury and inflammation, which induces vascular smooth muscle cell (VSMC) reprogramming. VSMCs become proliferative and migrating, secrete extra-cellular matrix and form a new layer called the neo-intima, which slowly reduces the vessel luminal diameter (5).
Hydrogen sulfide (H2S) is an endogenously produced gasotransmitter derived from cysteine metabolism (6). Circulating H2S levels are reduced in humans suffering from vascular occlusive disease (7, 8) and pre-clinical studies using water-soluble sulfide salts such as Na2S and NaHS have shown that H2S has cardiovascular protective properties(6), including reduction of IH in vivo in rats (9), rabbits (10), mice (11), and ex-vivo in human vein segments (12). However, the fast and uncontrolled release, narrow therapeutic range and high salt concentration of these compounds limit their therapeutic potential. Due to these limitations, H2S-based therapy are currently not available.
Here, we focused on sodium thiosulfate (Na2S2O3), an FDA-approved drug used in gram-quantity doses for the treatment of cyanide poisoning(13) and calciphylaxis, a rare condition of vascular calcification affecting patients with end-stage renal disease(14). Pharmaceutical-grade sodium thiosulfate (STS) is available and has been suggested to release H2S through non-enzymatic and enzymatic mechanisms (15, 16).
We tested whether STS inhibit IH in a surgical mouse model of IH and in an ex vivo model of IH in human vein culture. NaHS, a validated H2S donor was systematically compared to STS. We observed that STS was at least as potent as NaHS to inhibit IH in mice carotids, and in human vein. Although STS did not release detectable amounts of H2S or polysulfides in vitro, it increased protein persulfidation and circulating polysulfide levels in vivo. STS inhibited apoptosis and matrix deposition associated with the development of IH, as well as VSMC proliferation and migration. We further observed that STS and NaHS induced microtubule depolymerization in VSMCs.
2. Methods
For details on materials and reagents please see the Supplementary Material files
2.1. Mouse treatment
8 to 10 weeks old male WT or LDL receptor knock out (LDLR-/-) mice on a C57BL/6J genetic background were randomly divided into control vs sodium thiosulfate (STS) or NaHS. Sodium Thiosulfate (Hänseler AG, Herisau, Switzerland) was given in mice water bottle at 4g/L (0.5g/Kg/day), changed 3 times a week. NaHS (Sigma-Aldrich) was given in mice water bottle at 0.5g/L (125mg/Kg/day), changed every day. LDLR-/- mice were put on a cholesterol rich diet (Western 1635, 0.2% Cholesterol, 21% Butter, U8958 Version 35, SAFE® Complete Care Competence) for 3 weeks prior to experiments initiation. Mice were euthanized after 7 or 28 days of treatment by cervical dislocation under isoflurane anesthesia (inhalation 3% isoflurane under 2.5 liter of O2) followed by PBS perfusion. Aortas, carotid arteries, livers and serum or plasma (via intracardiac blood collection with a 24G needle) were harvested.
2.2. Cse-/- mice
Cse knockout mice were generated from a novel floxed line generated by embryonic injection of ES cells containing a Cth allele with LoxP sites flanking exon 2 (Cth tm1a(EUCOMM)Hmgu). Both ES cells and recipient embryos were on C57BL/6J background. Mice that were homozygous for the floxed allele were crossed with CMV-cre global cre-expressing mice (B6.C-Tg(CMV-cre)1Cgn/J), which have been backcrossed with C57BL/6J for 10 generations to create constitutive whole-body Cse-/- animals on a C57BL/6J background. The line was subsequently maintained by breeding animals heterozygous for the Cse null allele. Mouse ear biopsies were taken and digested in DirectPCR lysis reagent (Viagen Biotech, 102-T) with proteinase K (Qiagen, 1122470). WT, heterozygous and knockout mice were identified by PCR using the forward primer 5’-AGC ATG CTG AGG AAT TTG TGC-3’ and reverse primer 5’-AGT CTG GGG TTG GAG GAA AAA-3’ to detect the WT allele and the forward primer 5’-TTC AAC ATC AGC CGC TAC AG-3’ to detect knock-out allele using the platinum Taq DNA Polymerase (Invitrogen, cat#10966-026)
2.3. Carotid artery stenosis (CAS) surgery
The carotid artery stenosis (CAS) was performed as previously published (17) on 8 to 10 week old male WT or LDLR-/- mice. Treatment was initiated 3 days before surgery and continued for 28 days post-surgery until organ collection. The day of the surgery, mice were anesthetized with an intraperitoneal injection of Ketamine (80mg/kg) (Ketasol-100, Gräub E.Dr.AG, Bern Switzerland) and Xylazine (15 mg/kg) (Rompun®, Provet AG, Lyssach, Switzerland). The left carotid artery was exposed and separated from the jugular vein and vagus nerve. Then, a 7.0 PERMA silk (Johnson & Johnson AG, Ethicon, Zug, Switzerland) thread was looped and tightened around the carotid in presence of a 35-gauge needle. The needle was removed, thereby restoring blood flow, albeit leaving a significant stenosis. The stenosis triggers IH proximal to the site of injury, which was measured 28 days post surgery (17). Buprenorphine (0.05 mg/kg Temgesic, Reckitt Benckiser AG, Switzerland) was provided subcutaneously as post-operative analgesic every 12 hours for 24 hours. Mice were euthanized under isoflurane anesthesia (inhalation 3% isoflurane under 2.5 liter of O2) by cervical dislocation and exsanguination, perfused with PBS followed by buffered formalin 4% through the left ventricle. Surgeons were blind to the group during surgeries.
2.4. Human tissue and VSMC culture
Static vein culture was performed as previously described (12, 18). Briefly, the vein was cut in 5 mm segments randomly distributed between conditions. One segment (D0) was immediately preserved in formalin or flash frozen in liquid nitrogen and the others were maintained in culture for 7 days in RPMI-1640 Glutamax I supplemented with 10 % FBS and 1% antibiotic solution (10,000 U/mL penicillin G, 10,000 U/mL streptomycin sulfate) in cell culture incubator at 37°C, 5% CO2 and 21% O2.
Human VSMCs were also prepared from these human great saphenous vein segments as previously described (12, 18). Vein explants were plated on the dry surface of a cell culture plate coated with 1% Gelatin type B (Sigma-Aldrich). Explants were maintained in RPMI, 10% FBS medium in a cell culture incubator at 37°C, 5% CO2, 5% O2 environment.
2.5. Carotid and human vein histomorphometry
After 7 days in culture, or immediately upon vein collection (D0), the vein segments were fixed in buffered formalin, embedded in paraffin, cut into 5 µm sections, and stained using Van Gieson Elastic Laminae (VGEL) staining as previously described (12, 19). Three photographs per section were taken at 100x magnification and 8 measurements of the intima and media thicknesses were made, evenly distributed along the length of the vein wall.
Left ligated carotids were isolated and paraffin-embedded. Six-µm sections of the ligated carotid artery were cut from the ligature towards the aortic arch and stained with VGEL for morphometric analysis. Cross sections at every 300 µm and up to 2 mm from the ligature were analyzed using the Olympus Stream Start 2.3 software (Olympus, Switzerland). For intimal and medial thickness, 72 (12 measurements/cross section on six cross sections) measurements were performed, as previously described (12).
Two independent researchers blind to the experimental groups did the morphometric measurements, using the Olympus Stream Start 2.3 software (Olympus, Switzerland) (12).
2.6. H2S and polysulfide measurement
Free H2S was measured in cells using the SF7-AM fluorescent probe (20) (Sigma-Aldrich). The probe was dissolved in anhydrous DMF at 5 mM and used at 5 μM in serum-free RPMI medium with or without VSMCs. Free polysulfide was measured in cells using the SSP4 fluorescent probe. The probe was dissolved in DMF at 10 mM and diluted at 10 μM in serum-free RPMI medium with or without VSMCs. Fluorescence intensity (λex = 495 nm; λem = 520 nm) was measured continuously in a Synergy Mx fluorescent plate reader (BioTek Instruments AG, Switzerland) at 37 °C before and after addition of various donors, as indicated.
Plasma polysulfides were measured using the SSP4 fluorescent probe. Plasma samples were diluted 3 times and incubated for 10 min at 37°C in presence of 10 μM SSP4. Plasma polysulfides were calculated using a Na2S3 standard curve. Liver polysulfides were measured using the SSP4 fluorescent probe. Pulverized frozen liver was resuspended in PBS-0.5% triton X-100, sonicated and adjusted to 0.5 mg/ml protein concentration. Lysates were incubated for 15 min at 37°C in presence of 10 μM SSP4 and fluorescence intensity (λex = 495 nm; λem = 520 nm) was measured in a Synergy Mx fluorescent plate reader (BioTek Instruments AG, Switzerland)
2.7. Persulfidation protocol
Persulfidation protocol was performed using a dimedone-based probe as recently described (21). Persulfidation staining was performed on VSMCs grown on glass coverslips. Briefly, 1mM 4-Chloro-7-nitrobenzofurazan (NBF-Cl, Sigma) was diluted in PBS and added to live cells for 20 minutes. Cells were washed with PBS then fixed for 10 minutes in ice-cold methanol. Coverslips were rehydrated in PBS, and incubated with 1mM NBF-Cl for 1h at 37°C. Daz2-Cy5.5 (prepared with 1mM Daz-2, 1mM alkyne Cy5.5, 2mM copper(II)-TBTA, 4mM ascorbic acid with overnight incubation at RT, followed by quenching for 1h with 20mM EDTA) was added to the coverslips and incubated at 37°C for 1h. After washing with methanol and PBS, coverslips were mounted in Vectashield mounting medium with DAPI and visualized with a 90i Nikon fluorescence microscope.
2.8. BrdU assay
VSMC were grown at 80% confluence (5·103 cells per well) on glass coverslips in a 24-well plate and starved overnight in serum-free medium. Then, VSMC were either treated or not (ctrl) with the drug of choice for 24 hours in full medium (RPMI 10% FBS) in presence of 10µM BrdU. All conditions were tested in parallel. All cells were fixed in ice-cold methanol 100% after 24 hours of incubation and immunostained for BrdU. Images were acquired using a Nikon Eclipse 90i microscope. BrdU-positive nuclei and total DAPI-positive nuclei were automatically detected using the ImageJ software (12).
2.9. Flow Cytometry
VSMC were grown at 70% confluence (5·104 cells per well) and treated for 48 hours with 15mM STS or 10nM Nocodazole. Then, cells were tripsinized, collected and washed in ice-cold PBS before fixation by dropwise addition of ice-cold 70% ethanol while slowly vortexing the cell pellet. Cells were fixed for 1 hour at 4°C, washed 3 times in ice-cold PBS and resuspended in PBS supplemented with 20µg/mL RNAse A and 10µg/mL DAPI. Flow cytometry was performed in a Cytoflex-S apparatus (Beckmann).
2.10. Wound healing assay
VSMC were grown at confluence (104 cells per well) in a 12-well plate and starved overnight in serum-free medium. Then, a scratch wound was created using a sterile p200 pipette tip and medium was changed to full medium (RPMI 10% FBS). Repopulation of the wounded areas was recorded by phase-contrast microscopy over 24 hours in a Nikon Ti2-E live cell microscope. All conditions were tested in parallel. The area of the denuded area was measured at t=0 hours and t=10 hours after the wound using the ImageJ software by two independent observers blind to the conditions. Data were expressed as a ratio of the healed area over the initial wound area.
2.11. Apoptosis assay
Apoptosis TUNEL assay was performed using the DeadEnd™ Fluorometric TUNEL system kit (Promega) on frozen sections of human vein segments. Immunofluorescent staining was performed according to the manufacturer’s instruction. Apoptotic nuclei were automatically detected using the ImageJ software and normalized to the total number of DAPI-positive nuclei. In vitro VSMC apoptosis was determined by hoescht/propidium iodide staining of live VSMC and manually counted by two independent blinded experimenters (22).
2.12. Immunohistochemistry
Polychrome Herovici staining
was performed on paraffin sections as described (23). Young collagen is stained blue, while mature collagen is pink. Cytoplasm is counterstained yellow. Hematoxylin is used to counterstain nuclei blue to black.
Collagen III staining
was performed on frozen sections (OCT embedded) of human vein segments using mouse anti-Collagen III antibody (ab7778, abcam). Briefly, tissue slides were permeabilized in PBS supplemented with 2 wt. % BSA and 0.1 vol. % Triton X-100 for 30 min, blocked in PBS supplemented with 2 wt. % BSA and 0.1 vol. % Tween 20 for another 30 min, and incubated overnight with the primary antibodies diluted in the same buffer. The slides were then washed 3 times for 5 min in PBS supplemented with 0.1 vol. % Tween 20, and incubated for 1 hour at room temperature with anti-mouse AlexaFluor 568 (1/1000, ThermoFischer). Slides were visualized using a Nikon 90i fluorescence microscope (Nikon AG). Collagen III immunostaining area was quantified using the ImageJ software and normalized to the total area of the vein segment.
PCNA (proliferating cell nuclear antigen) and α-tubulin immunohistochemistry
was performed on paraffin sections as previously described (24). After rehydration and antigen retrieval (TRIS-EDTA buffer, pH 9, 17 min in a microwave at 500 watts), immunostaining was performed on human vein or carotid sections using the EnVision +/HRP, DAB+ system according to manufacturer’s instructions (Dako, Baar, Switzerland). Slides were further counterstained with hematoxylin. PCNA and hematoxylin positive nuclei were manually counted by two independent observers blinded to the conditions.
α-tubulin immunofluorescent staining
in human VSMC was performed as previously described. Cell were fixed at -20°C for 10min in absolut methanol. Then, cell were blocked/permeabilized in PBS-triton 0.2%, BSA 3% for 45 min at room temperature. Cells were incubated overnight at 4°C in the primary antibody diluted in PBS-0.1% tween, 3% BSA, washed 3 times in PBS and incubated for 1 hour at room temperature with the secondary antibody diluted in PBS-0.1% tween, 3% BSA, washed again 3 times in PBS and mounted using Vectashield mounting medium for fluorescence with DAPI. The microtubule staining was quantified automatically using FiJi (ImageJ, 1.53c). Image processing was as follows: Plugin, Tubeness/Process, Make Binary/Analyze, Skeleton. Data were summarized as filament number and total length, normalized to the number of cells per images. Data were generated from images from 3 independent experiments, 3 to 4 images per experiment per condition.
2.13. Western blotting
Mice aortas or human vein segments were flash-frozen in liquid nitrogen, grinded to power and resuspended in SDS lysis buffer (62.5 mM TRIS pH6,8, 5% SDS, 10 mM EDTA). Protein concentration was determined by DC protein assay (Bio-Rad Laboratories, Reinach, Switzerland). 10 to 20 µg of protein were loaded per well. Primary cells were washed once with ice-cold PBS and directly lysed with Laemmli buffer as previously described (12, 18). Lysates were resolved by SDS-PAGE and transferred to a PVDF membrane (Immobilon-P, Millipore AG, Switzerland). Immunoblot analyses were performed as previously described(18) using the antibodies described in supplementary Table S1. Membranes were revealed by enhanced chemiluminescence (Immobilon, Millipore) using the Azure Biosystems 280 and analyzed using Image J. Protein abundance was normalized to total protein using Pierce™ Reversible Protein Stain Kit for PVDF Membranes.
2.14. In vitro tubulin polymerization assay
The assay was performed using the In Vitro Tubulin Polymerization Assay Kit (≥99% Pure Bovine Tubulin; 17-10194 Sigma-Aldrich), according to the manufacturer’s instruction.
2.15. Statistical analyses
All experiments adhered to the ARRIVE guidelines and followed strict randomization. A power analysis was performed prior to the study to estimate sample-size. We hypothesized that STS would reduce IH by 50%. Using an SD at +/- 20% for the surgery and considering a power at 0.9, we calculated that n= 12 animals/group was necessary to validate a significant effect of the STS. All experiments were analyzed using GraphPad Prism 8. One or 2-ways ANOVA were performed followed by multiple comparisons using post-hoc t-tests with the appropriate correction for multiple comparisons.
2.16. Role of funding source
The funding sources had no involvement in study design, data collection, data analyses, interpretation, or writing of report.
2.17. Ethics Statement
Human great saphenous veins were obtained from donors who underwent lower limb bypass surgery (25). Written, informed consent was obtained from all vein donors for human vein and VSMC primary cultures. The study protocols for organ collection and use were reviewed and approved by the Centre Hospitalier Universitaire Vaudois (CHUV) and the Cantonal Human Research Ethics Committee (http://www.cer-vd.ch/, no IRB number, Protocol Number 170/02), and are in accordance with the principles outlined in the Declaration of Helsinki of 1975, as revised in 1983 for the use of human tissues.
All animal experimentations conformed to the National Research Council: Guide for the Care and Use of Laboratory Animals (26). All animal care, surgery, and euthanasia procedures were approved by the CHUV and the Cantonal Veterinary Office (SCAV-EXPANIM, authorization number 3114, 3258 and 3504).
3. Results
3.1. STS limits IH development in mice after carotid artery stenosis
We first assessed whether STS protects against IH, 28 days after carotid artery stenosis in mice (CAS)(17). STS treatment (4g/L) decreased IH by about 50% in WT mice (Figure 1A), as expressed as the mean intima thickness, the ratio of intima over media thickness (I/M), or the area under the curve of intima thickness calculated over 1 mm. To model the hyperlipidaemic state of patients with PAD, we also performed the CAS on hypercholesterolemic LDLR-/- mice fed for 3 weeks with a cholesterol-rich diet. As expected, the LDLR-/- mice developed more IH than WT mice upon CAS, and STS treatment lowered IH by about 70% (Figure 1B). Interestingly, STS did not reduce media thickness in WT mice (p= but significantly reduced media thickness in LDLR-/- mice (p=Figure 1B). Of note, the sodium salt H2S donor NaHS (0.5g/L) also significantly decreased IH following carotid stenosis in WT mice (Figure S1 in the online supplementary files).
3.2. STS limits IH development in a model of ex vivo human vein segment culture
Both STS (15mM) and NaHS (100µM) inhibited IH development in our validated model of static ex vivo human vein segment culture(12) as measured by intima thickness and I/M ratio (Figure 1C). The polysulfide/H2S donors diallyl trisulfide (DATS), cysteine-activated donor 5a and GYY4137 also prevented the development of IH in human vein segments (Figure S2).
3.3. STS is a biologically active source of sulfur
Overall, STS and “classical” H2S donors similarly inhibit IH. To measure whether STS releases detectable amounts of H2S or polysulfides, we used the SF7-AM and SSP4 probes, respectively. We could not detect any increase in SSP4 or SF7-AM fluorescence in presence of STS with or without VSMCs. Na2S3 was used as a positive control for the SSP4 probe and NaHS as a positive control for the SF7-AM probe (Figure S3). The biological activity of H2S is mediated by post-translational modification of reactive cysteine residues by persulfidation, which influence protein activity (21, 27). As a proxy for H2S release, we assessed global protein persulfidation by DAZ-2-Cy5.5 labelling of persulfide residues in VSMC treated for 4 hours with NaHS or STS. Both STS and NaHS similarly increased persulfidation in VSMC (Figure 2A). Using the SSP4 probe, we further observed higher polysulfides levels in vivo in the plasma of mice treated for 1 week with 4g/L STS (Figure 2B). Similarly, polysulfides levels tended to be higher in the liver of mice treated with STS (p=0.15). As a positive control, mice treated with 0.5g/L NaHS had significantly higher polysulfides levels in the liver (Figure 2C). Cse is main enzyme responsible for endogenous H2S in the vasculature and Cse-/- mice have been shown to develop more IH (11, 28, 29). Here, we generated a new Cse knock-out mouse line. Consistent with a role as an H2S donor, STS fully rescued Cse-/- mice from increased IH in the model of carotid artery stenosis (Figure 2D).
3.4. STS limits IH-associated matrix deposition and apoptosis in human vein segments
We further assessed matrix deposition and apoptosis in human vein segments. Concomitant with IH formation, ex vivo vein culture (D7) resulted in de novo collagen deposition compared to D0, as assessed by polychrome Herovici staining (Figure 3A). Immunoanalysis of immature Collagen III levels confirmed that vein culture (D7) resulted in de novo collagen deposition compared to D0, as assessed by Western blot (Figure 3B) and immunostaining (Figure 3C). STS and NaHS treatment tended to reduce collagen deposition as assessed by Herovici staining, collagen III immunostaining and Western blotting (Figure 3A-C). TUNEL assay revealed that STS, and to a lesser degree NaHS, decreased the percentage of apoptotic cells observed after 7 days in culture (D7) (Figure 3D). STS also attenuated pro-apoptotic protein Bax overexpression observed after 7 days in culture, while NaHS had a non-significant tendency to decrease Bax level (p=.11; Figure 3E). STS also tended to increase the protein level of anti-apoptotic protein Bcl-2 (p=.06), while NaHS significantly increased it (p=.04; Figure 3E).
3.5. STS blocks VSMC proliferation and migration
IH is driven by VSMC reprogramming toward a proliferating, migrating and ECM-secreting phenotype (5). Both STS and NaHS significantly reduced the percentage of proliferating cells (defined as PCNA positive nuclei over total nuclei) in vivo in CAS-operated carotids in WT mice (Figure 4A) and ex vivo in human vein segments (Figure 4B). In vitro, STS dose-dependently decreased primary human VSMC proliferation as assessed by BrdU incorporation assay (Figure 5A). Of Note, NaHS as well as DATS, donor5A and GYY4137, also decreased VSMC proliferation (Figure S4). STS and NaHS also decreased VSMC migration in a wound healing assay (Figure 5B). Further evaluation of cell morphology during the wound healing revealed that STS and NaHS-treated cells lost the typical elongated shape of VSMC, as measured through the area, Feret diameter and circularity of the cells (Figure 5C).
3.6. STS interferes with microtubules organization
Given the impact of STS on cell morphology, we examined in more details the effect of STS on the cell cytoskeleton. α-tubulin levels were increased in the carotid wall of CAS-operated mice, which were reduced by STS, as demonstrated by immunohistochemistry (Figure 6A). α-tubulin levels were also decreased in the native aorta of mice treated with STS for 7 days (Figure 6B). Similarly, total α-tubulin levels were decreased in ex vivo vein segments after 7 days of STS treatment (Figure 6C-D). Looking further at α-tubulin by immunofluorescent staining showed a loss in microtubule in VSMCs treated with STS or NaHS for 8 hours (Figure 6 E-F). To study the effect of H2S on microtubule formation, an in vitro tubulin polymerization assay was performed in presence of 15mM STS, 100 µM NaHS or 10µM Nocodazole, an inhibitor of microtubule assembly. As expected, Nocodazole slowed down microtubule assembly, as compared to the control. Surprisingly, both NaHS and STS fully blocked microtubule assembly in this assay (Figure 6G). Further studies of the cell cycle in VSMC revealed that 48 hours of treatment with STS or Nocodazole resulted in accumulation of cells in G2/M phase (Figure 6H).
Discussion
Despite decades of research, intimal hyperplasia remains one of the major limitations in the long-term success of revascularization. In addition, in December 2018, Katsanos and colleagues reported, in a systematic review and meta-analysis, an increased risk of all-cause mortality following application of paclitaxel-coated balloons and stents in the femoropopliteal artery (30). This recent report had tremendous repercussions on the community, urging vascular societies and policy makers to further investigate the use of paclitaxel-coated balloons, stents and other devices for the treatment of peripheral arterial disease. Here, we demonstrate that exogenous sulfur supplementation in the form of STS limits IH development in vivo following mouse carotid artery stenosis. Furthermore, STS reduced apoptosis, vessel remodeling and collagen deposition, along with IH development in our ex vivo model of IH in human vein segments. We propose that STS limits IH by interfering with the microtubule dynamics, thus VSMC proliferation and migration.
An ever increasing number of studies document the protective effects of H2S against cardiovascular diseases (31), including studies showing that H2S reduces IH in preclinical models (9-11, 32). The administration of H2S in those studies relies on soluble sulfide salts such as NaHS with narrow therapeutic range due to fast and uncontrolled release. Thiosulfate is an intermediate of sulfur metabolism that can lead to the production of H2S (15, 33, 34). Importantly, STS is clinically approved and safe in gram quantities in humans. Although STS yields no detectable H2S or polysulfide in vitro, we observed increased circulating and liver levels of polysulfides in mice, as well as increased protein persulfidation in VSMC. Overall, STS has protective effects against IH similar to the H2S salt NaHS and several other “classical” H2S donors, but holds much higher translational potential.
Mechanistically, we first observed that STS reduces cell apoptosis and matrix deposition in our ex vivo model of human vein segments. This anti-apoptotic effect of STS and NaHS is in line with known anti-apoptotic effects of H2S (31). STS also reduces IH in vivo following carotid artery stenosis. In this model, fibrosis plays little role in the formation of IH, which relies mostly on VSMC proliferation (5). Therefore, although ECM and especially collagen deposition are major features of IH in human (35, 36), reduced apoptosis and matrix deposition is not sufficient to fully explain the protection afforded by STS in carotids in vivo.
STS, similarly to the H2S donor NaHS, also inhibits VMSC proliferation and migration in the context of IH in vivo in mouse stenotic carotids, as well as ex vivo in human vein segments, and in vitro in primary human VSMC. These findings are in line with previous studies demonstrating that “classical” H2S donors decrease VSMCs proliferation in pre-clinical models (10, 12, 37). In mouse VSMC, exogenous H2S has been proposed to promote cell cycle arrest (28), and regulate the MAPK pathway (9) and IGF-1 response (38). Regarding mouse VSMC migration, H2S may limit α5β1-integrin and MMP2 expression, preventing migration and ECMs degradation (11, 28).
In this study, we further document that STS and NaHS disrupt the formation of microtubules in human VSMC in vitro. Our findings are in line with previous studies showing that Diallyl trisulfide, a polysulfide donor, inhibits microtubule polymerization to block human colon cancer cell proliferation (39). NaHS also depolymerizes microtubules within Aspergillus nidulans biofilms (40). The α/β tubulin dimer has 20 highly conserved cysteine residues, which have been shown to regulate microtubule formation and dynamics (41). In particular, thiol™disulfide exchanges in intrachain disulfide bonds have been proposed to play a key role in microtubule assembly (42). Several high throughput studies of post-translational modification of protein cysteinyl thiols (-SH) to persulfides (-SSH) demonstrated that cysteine residues in α- and β-tubulin are persulfidated in response to H2S donors in various cell types (43, 44). Given the prominent role of cytoskeleton dynamics and remodelling during mitosis and cell migration, we propose that STS/H2S-driven microtubule depolymerisation, secondary to cysteine persulfidation, contributes to cell cycle arrest and reduces migration in VSMC.
Our findings suggest that STS holds strong translational potential to limit restenosis following vascular surgeries. The dosage of STS used in this study is comparable to previous experimental studies using 0.5 to 2□g/Kg/day (15, 16, 33, 34). In humans, 12.5 and 25□g of STS has been infused without adverse effects (45) and short-term treatment with i.v. STS is safely used in patients for the treatment of calciphylaxis (14). Of note, the pathophysiology of calciphylaxis, also known as calcific uremic arteriopathy (CUA), is caused by oxidative stress and inflammation, which promote endothelial dysfunction, leading to medial remodelling, inflammation, fibrosis and VSMC apoptosis and differentiation into bone forming osteoblast-like cells. Although the main effect of STS on CUA is certainly the formation of highly soluble calcium thiosulfate complexes, our data certainly support the positive effects of STS in the treatment of calciphylaxis. Plus, STS infusions have been shown to increase distal cutaneous blood flow, which could be beneficial in the context of vascular occlusive disease.
Here, we propose that STS treatment results in persulfidation of cysteine residues in the tubulin proteins, which lead to microtubule depolymerization. However, further studies are required to test this hypothesis and demonstrate the STS-induced persulfidation of tubulin cysteine residues. In addition, although we show in vitro that H2S directly affect microtubule formation in a cell-free environment, other proteins involved in the microtubules dynamics in vivo may also be modified by H2S, and contribute to the effect of STS on microtubule polymerization and cell proliferation. In this study, STS was administered in the water bottle. However, a validated oral form of the compound has not been developed yet. Case reports and case series suggest that intravenous STS administration is safe, even for relatively long periods of time (46, 47). However, randomized controlled trials testing long-term administration of STS are lacking (47) and the long-term safety and effects of STS administration should be further explored.
In summary, under the conditions of these experiments, STS, a FDA-approved compound, limits IH development in vivo in a model of arterial restenosis and ex vivo model in human veins. STS most likely acts by increasing H2S bioavailability, which inhibits cell apoptosis and matrix deposition, as well as VSMC proliferation and migration via microtubules depolymerization.
Contributors
FA, AL and SD designed the study. FA, DM, MMA, ML and SU performed the experiments. FA, DM, MMA, ML and SD analyzed the data. FA, DM, MMA, AL and SD wrote the manuscript. JMC critically revised the manuscript. FA, AL, SD and DM finalized the manuscript.
Funding
This work was supported by the following: The Swiss National Science Foundation (grant FN-310030_176158 to FA and SD and PZ00P3-185927 to AL); the Novartis Foundation to FA; and the Union des Sociétés Suisses des Maladies Vasculaires to SD.
Data sharing statement
The data that support the findings of this study are available from the corresponding author, Florent Allagnat, upon request.
Declaration of Competing Interest
The authors declare no competing interests.
Acknowledgments
We thank the mouse pathology facility for their services in histology (https://www.unil.ch/mpf).
Abbreviations
- CAS
- carotid artery stenosis
- H2S
- hydrogen sulfide
- IH
- intimal hyperplasia
- NaHS
- sodium hydrogen sulfur
- PCNA
- proliferating cell nuclear antigen
- SBP
- systolic blood pressure
- STS
- Sodium Thiosulfate
- VSMC
- vascular smooth muscle cells
- VGEL
- Van Gieson elastic lamina