|Home | About | Journals | Submit | Contact Us | Français|
Markers containing dyes such as crystal violet (CAS 548-62-9) are routinely used on the adventitia of vein bypass grafts to avoid twisting during placement. Since little is known about how these dyes affect vein graft healing and function, we determined the effect of crystal violet on cell migration and proliferation, which are responses to injury after grafting.
Fresh human saphenous veins were obtained as residual specimens from leg bypass surgeries. Portions of the vein that had been surgically marked with crystal violet were analyzed separately from those that had no dye marking. In the laboratory, they were split into easily dissected inner and outer layers after removal of endothelium. This f cleavage plane was within the circular muscle layer of the media. Cell migration from explants was measured daily as either 1) % migration positive explants, which exclusively measures migration, or 2) the number of cells on the plastic surrounding each explant, which measures migration plus proliferation. Cell proliferation and apoptosis (Ki67 and TUNEL staining, respectively) were determined in dye-marked and unmarked areas of cultured vein rings. The dose-dependent effects of crystal violet were measured for cell migration from explants as well as proliferation, migration, and death of cultured outer layer cells. Dye was extracted from explants with ethanol and quantified by spectrophotometry.
There was significantly less cell migration from visibly blue, compared to unstained, outer layer explants by both methods. There was no significant difference in migration from inner layer explants adjacent to blue-stained or unstained sections of vein, because dye did not penetrate to the inner layer. Ki67 staining of vein in organ culture, which is a measure of proliferation, progressively increased up to 6 days in non-blue outer layer and was abolished in the blue outer layer. Evidence of apoptosis (TUNEL staining) was present throughout the wall and not different in blue-stained and unstained vein wall segments. Blue outer layer explants had 65.9±8.0 ng dye/explant compared to 2.1±1.3 for non-blue outer layer explants. Dye applied in vitro to either outer or inner layer explants dose-dependently inhibited migration (IC50=8.5 ng/explant). The IC50s of crystal violet for outer layer cell proliferation and migration were 0.1 and 1.2 μg/ml, while the EC50 for death was between 1 and 10 μg/ml.
Crystal violet inhibits venous cell migration and proliferation indicating that alternative methods should be considered for marking vein grafts.
Vein grafts are frequently used for aortocoronary and peripheral vascular reconstructions. However severe luminal narrowing is a primary cause of graft failure1. This typically develops within the first year in approximately 25–30% of all grafts due to pathological remodeling and intimal hyperplasia. While much research is aimed at determining the pathological mechanisms of luminal narrowing, technical factors such as surgical graft preparation are also important2. For example, other factors such as conduit distension3, vein conduit storage4, and stretching during surgical harvest5 can adversely affect the function of vein wall cells.
Surgical marking pens containing dyes such as crystal violet are routinely used on the adventitia of vein bypass grafts to avoid twisting of the vein grafts during placement. It continues to be assumed that these pens are innocuous despite evidence that dyes used in these pens can inhibit endothelium-dependent vasodilation6, 7 and the smooth muscle contractile response8. To further explore possible deleterious effects of marking pens on vein grafts we have studied the effect of crystal violet on venous cell migration, proliferation, and apoptosis. These are the major biological activities that contribute to intimal hyperplasia in both large and small animal models of vein graft healing9–15.
Human saphenous vein (HSV) remnants were obtained from patients undergoing coronary artery bypass or peripheral vascular bypass surgeries under protocols approved by the University of Washington or the Benaroya Research Institute IRB and all subjects gave informed consent as indicated in these protocols. Veins were either endoscopically or openly harvested. All were distended to identify untied branches and were stored in heparinized buffered saline before being placed into Dulbecco’s Modified Eagles Medium with 10 mM Hepes, pH 7.4, for transport to the laboratory for dissection. The use of a surgical skin marker was decided by the surgical team. The HSV was dissected free of extraneous tissue and sections that were marked intraoperatively with blue dye were separated from unmarked sections. A visually distinct, white, lumenal layer was easily dissected from the remaining vein after removal of endothelium by gently wiping with a cotton-tipped swab. The identity of the natural cleavage plane of vein dissection was confirmed by immuno-histochemical staining for smooth muscle alpha actin of the dissected layers. This demonstrated that both outer and inner layers contain smooth muscle alpha actin-positive cells and that the cleavage plane goes through the circular muscle layer. The layers, hereafter called inner layer and outer layer, respectively, were cut into 2.5 mm2 explants using a McIlwain tissue chopper (customized to cut sizes > 1 mm2). Explants were placed into 25-cm2 flasks (15/flask) and 1.2 ml of DMEM with 20% fetal bovine serum (DMEM 20%FBS) were added. Medium was changed 3 times per week.
Explants were examined daily for 8 days and cell migration was quantified as both the percentage of total explants with at least one migrating cell (a measure of only migration) and the average number of cells on plastic surrounding each explant (a measure of the combination of migration and proliferation)16.
To determine the amount of dye adsorbed to the different vein samples during the surgical marking, explants prepared as described above were divided into 4 groups: 1) visibly blue-stained outer layer explants, 2) inner layer explants luminally adjacent to blue-stained outer layer explants, 3) outer layer explants with no visible blue dye, and 4) inner layer explants luminally adjacent to outer layer explants with no visible blue dye. Crystal violet was quantified spectrophotometrically (OD590, the λmax of crystal violet) using crystal violet standards dissolved in ethanol. Pens obtained from the surgical services providing specimens for this study were confirmed to contain crystal violet, and dye extracted from vein graft specimens displayed a λmax of 590 nm. This assay is linear up to 50 μg/ml crystal violet. Blue dye was extracted by immersing 5 explants of each group in 100 ul absolute ethanol and incubating at room temperature for 30 minutes. Extraction efficiency of crystal violet added to unstained explants was 80% with a first extraction and >90% with a third extraction. While crystal violet may be tightly bound to tissue components such as DNA17, most dye is available for extraction.
To determine the effects of crystal violet on migration of cells from the explants, one μl of crystal violet or buffer alone (0 to 500 ng/μl) was added to explants (in the absence of culture medium) from unstained sections of vein and incubated for 5 minutes. Separate experiments using non-blue outer layer or adjacent inner layer explants were performed as indicated in the figure legends. Some explants were then extracted to measure bound dye while other explants were placed in a 25cm2 flask after one wash with DMEM/Hepes (15 explants per condition per vein for the number of veins indicated in the figure legends). Explants were given growth medium as before and the % of migration positive explants was determined at day 8.
After removal of loose connective tissue, veins were cut into rings ~3 mm in length and placed in 20% FBS/DMEM (one/2 ml). The medium was changed every other day and the tissue was fixed in 10% neutral buffered formalin at 0, 2, 4, and 6 days. Blue areas were marked with India ink before processing and embedding in paraffin because the blue dye is removed during processing. We measured cell proliferation by immunohistochemical staining of 5 μm sections with Ki67 (0.5 μg/ml SP6; Abcam, Inc.) using the avidin-biotin-peroxidase method (Vector Laboratories, Burlingame, CA) with rabbit IgG as a negative control. For measurement of apoptosis the terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay was performed using a kit (ApopTag; Millipore, Inc.) following the manufacturer’s instructions, but specifically using TdT at 1:32, citrate buffer (pH 4.5) rather than proteinase K, and omission of terminal deoxynucleotidyl transferase as a negative control. To compare blue dye stained sections to unstained sections, the percentage of Ki67 or TUNEL positive nuclei was determined in adventitia, outer media, and inner media/intima for the portion of the perimeter marked by India ink (i.e. blue dye) as well as the same perimeter length on the opposite side of the vein (unstained). Adventitia plus outer media is equivalent to the outer layer designation and inner media/intima is the same as the inner layer designation.
Cultured outer layer cells were obtained by the explant method from saphenous vein unstained outer layer explants in 20%FBS/DMEM. When cells near the explants become confluent we found the optimal medium is Smooth Muscle Cell Growth Medium (Cell Applications, Inc.), which was also used for subsequent growth of passaged cells. Cells were used at passage 6. For proliferation and death assays 48,000 cells were seeded in 5% FBS in bovine basal medium (Cell Applications, Inc.) in duplicate 6 well plates. The next day cells were counted using a hemocytometer and the medium in remaining wells was changed to Cell Applications SMC growth medium for proliferation studies, or transferred to 2% serum for studies of cell death. Cells were then exposed for three days to a range of concentrations of crystal violet (0, 0.01, 0.1, 1.0, 10 μg/ml), and then counted again. Preliminary experiments demonstrated that cell number is maintained but not increased with 2% FBS alone.
For migration assays, cells were preincubated for 30 minutes with crystal violet (0, 0.01, 0.1, 1.0, 10 μg/ml) and then seeded into the upper well of a 48 well microchemotaxis chamber at 30,000 per well (Neuroprobe, Inc.) in Bovine Basal Medium (Cell Applications, Inc) in the presence of the same concentrations of dye. Polycarbonate filters with 8 μm pores coated with 100 μg/ml monomeric bovine skin collagen were used, and bottom chambers contained 10% serum. After 5 hours at 37°, the cells that had migrated through the filters were fixed and counted at 400X.
Comparisons of measures between unstained and blue-stained explants, and between various concentrations of dye, were made with ANOVA with repeated measures or the Wilcoxon test with a Bonferroni correction as needed (SPSS, v20, Chicago). A P≤.05 was considered statistically significant. Data are presented as the mean ± SEM.
Cells began migrating from saphenous vein explants after a lag phase of about 3 days with outer layer cells initiating migration more quickly than inner layer cells. Migration was significantly less from blue-stained, compared with unstained outer layer explants by day 5. There was a reduction of >80% migration at day 8 by both measures of migration (Figures 1A and 1B). Inner layer explants separated into those adjacent to blue-stained outer layer and those adjacent to unstained outer layer exhibited equal cell migration (Figures 1C and 1D), which is consistent with the visual observation that the staining from the surgeon’s marking pain almost never penetrated into the inner layer. Of the hundreds of veins dissected in this laboratory only one vein had lightly visible staining of the inner layer from the surgeon’s marking pen. Indeed, while blue-stained outer layer explants had more than 30 fold the amount of dye seen in visibly unstained outer layer explants, there was no difference in amount of blue dye detected in inner layer explants, whether they were adjacent to a blue-stained outer layer, or an unstained outer layer (Table 1).
The effect of surgeon-applied blue dye on cell proliferation and apoptosis in vein rings was determined by comparing Ki67 and TUNEL staining in the blue stained perimeter of vein compared to the same perimeter length of unstained vein on the opposite side of the vein ring. These were whole, undissected vein segments. The percentage of Ki67 and TUNEL positive cells were counted separately in the inner media/intima, outer media, and adventitia separated as illustrated in Figure 2A. In the unstained portion of the vein, proliferation increased progressively between days 2 and 6 with adventitia showing the highest proliferation followed by the outer media then the inner media/intima. In contrast, there was no proliferation observed in the blue-stained side of the vein except in the inner media/intima, which is again consistent with a lack of penetration of blue dye into the inner layer (Figure 2B–D). In contrast, TUNEL positive cells were observed at all times primarily in the adventitia (Figure 3A), but no difference was noted between dye-positive and dye-negative areas of the vessel (Figure 3B–D).
To rule out the possibility of a difference in response to the dye between outer layer and inner layer cells we also performed crystal violet dose-response studies on unstained outer layer and adjacent inner layer explants (Figure 4). Inhibition of migration from both types of explants occurred with an IC50 of ~10 ng/explant. Dose-response curves for crystal violet for proliferation, migration and death of cultured outer layer cells were also determined. At the lowest concentrations it inhibited proliferation, and at the highest concentrations it induced cell death. The IC50 for proliferation was ~0.1 μg/ml, for migration it was ~1.2 μg/ml, and for death the EC50 was between 1 and 10 μg/ml (Figure 5A–C).
We have found that crystal violet, the dye used in many surgical marking pens, has a strong inhibitory effect on migration and proliferation of cells in the vein ex vivo. These data add to a growing body of evidence that commonly used marking pens have deleterious effects on the veins used for grafting, such as inhibition of both vasoconstrictor and vasodilator function6–8. Surgical marking pens have dyes, such as crystal violet and methylene blue, dissolved in either ethanol or polyethylene glycol. In this study we focused on the effect of crystal violet, a dye in pens used in Seattle, and did not address effects of the diluent. While ethanol also has been shown to inhibit smooth muscle cell proliferation and migration18, 19, expected inhibition of cell migration from outer layer explants based on levels of dye measured in surgeon-treated veins was 99.9% (Figure 4) compared to observed inhibition of 86.1% (Figure 1A), suggesting little if any effect of diluents on cell migration in this case.
The inhibition of cell migration from explants is probably not directly linked to inhibition of cell proliferation or increased cell death. While the IC50 for adventitial fibroblast proliferation is 10 fold lower than that for migration (0.1 vs. 1.2 μg/ml, respectively), inhibitors of proliferation such as hydroxyurea do not appreciably alter migration from explants20 and non-proliferating vascular cells migrate after injury21. Nor does the inhibition of migration appear to relate to cell death. For if we assume that the 2.5 mm2 adventitial explants have a volume of ~1 mm3, then the measured amount of dye/explant achieves a concentration of nearly 70 μg /ml. This is well beyond the EC50 for death (<10 μg/ml) and one to two orders of magnitude higher than the IC50s for proliferation and migration. Based on this we expected extreme loss of cells from blue areas of vein organ culture specimens. However, neither adventitial nuclear number (data not presented) nor TUNEL staining were different on the blue-stained compared to the unstained side of the vein. One possibility for the lack of a difference is that the injury of surgical preparation and preparation of the veins for organ culture may have masked any additional effect of the dye on cell death. In addition, most crystal violet may bind to extracellular matrix molecules thus restricting access of the dye to cells, thus minimizing cell death which has a much higher EC50 than proliferation or migration.
Finally, these results do not inform us as to how surgical marking pens might affect vein graft patency. The biological effects of application of the surgical marking dye would have an effect limited to the outer layer of the marked vein section, since we observed that neither cells from the inner layer nor the adjacent outer layer segments were adversely affected by nearby staining with crystal violet. No difference in patency has been observed between reversed and in situ vein grafts, which are marked and unmarked, respectively22–25. While the different preparation of these vessels makes any effect of marking pens difficult to determine with certainty, even in a surgical service where both in situ and reversed vein grafts were marked (Dr. Tadahiro Sasajima, personal communication), no difference in one year primary patency was observed26. In addition, it is not clear how the effects of crystal violet on cell migration and proliferation and vein vasoactivity would impact graft patency. For example, based on results from animal models it is reasonable to expect that both cell migration and proliferation are required for the adaptation of the vein to the arterial circulation, which involves wall thickening and replacement of dead cells 27–29. From this perspective, crystal violet might inhibit healing and promote failure. On the other hand, since a possible source of intimal cells is the adventitia, as suggested by some but not all animal studies30–32, inhibition of migration of adventitial cells to the intima by crystal violet might increase graft patency by limiting intimal hyperplasia. In addition, since it is likely that the vein is only partially re-endothelialized at one month, a time when grafts that demonstrate reduced dilation are more likely to fail33 the predominate vasoactive effect of crystal violet may be to inhibit smooth muscle contraction rather than inhibit endothelium-dependent vasodilation, an effect that would also promote patency.
In conclusion, despite the uncertainties of the net effect of pen dyes on graft outcome, given the known effects of crystal violet on vasoactivity, proliferation, and migration as well as potential toxic effects, it may be prudent to find an alternative dye, such as the recently reported Brilliant Blue FCF34, for clinical use or to limit the areas of application.
NIH grants R01 HL30946 (AWC, MS) and R41 HL106967 (T.N.W.), Department of Veterans Affairs, Veterans Health Administration Merit Review Agency (MS)
We are very thankful to Dr. Mark Hill (Virginia Mason Hospital, Seattle, WA) and Sylvia Posso, Lisa Myers Bulmash, and Julieann Marshall, study coordinators, for providing vein graft specimens and enthusiastic support.
Part of the work described in this article was presented in a poster at the Arteriosclerosis, Thrombosis, and Vascular Biology 2014 meeting at Toronto, Canada.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.