Transient impedance changes in venous endothelial monolayers as a biological radiation dosimetry response

HUVECs were seeded on collagen coated ECIS arrays at values near confluence at the beginning of the experiment. The individual endothelial cells required time to mature into a sealed monolayer with competent barrier function. ECIS recordings of HUVEC monolayers showed cells making contact with the substrate and spreading over an initial 4-6 h period followed by a slower increase in transmonolayer resistance associated with establishment of lateral contacts and monolayer sealing (Fig 1A). After reaching peak resistance after 14 h of culture, resistance began to gradually decline. Following irradiation, two phenomena are observed. There was a small increase in resistance reaching a maximum at 1 h after irradiation and a larger, significant decrease (n = 3; p = 0.017) in monolayer resistance with a maximum occurring roughly 3 h after irradiation (Fig. 1B). In a separate, independent iteration of the experiment, this maximum shifted later by 0.5 h. Different current paths beneath and between cells, dependent on frequency, inform a modeling algorithm which permits discernment of electrical contributions arising from lateral contacts and cell-substrate contacts (34, 35). There was no statistically significant perturbation in the substrate-interaction parameter, ^α, when models were applied to the monolayers (Fig. S1). The majority of the observed change in resistance is commensurate with the Rb modeling parameter that reports the resistance arising from lateral contacts between cells (Fig. 1C). Given the suggestion of a decrease in resistance, a single tailed t-test was used and found the depression in the Rb parameter at 3 h after irradiation to be statistically significant (n = 3; p = 0.038).

Fig.1

To verify that electrical changes observed in HUVEC monolayers were associated with changes in permeability and not overt killing or removal of the cells, a time-lapse course of video microscopy was performed on irradiated and sham-irradiated controls. Cells were seeded on collagen-coated cultureware at area-corrected densities equivalent to ECIS studies and allowed to mature into competent monolayers before being irradiated. In response to radiation, cells did not appear to retract from neighboring cells in the monolayer. No overt lifting-off or gross changes were observed in the monolayer three hours post irradiation. Later in culture, dead cells were seen to be extruded from the monolayers (Suppl. Movies S2,S3, Fig. 2).

Fig.2

To conclude, HUVEC monolayers, widely used as a model system by many researches, are similar to coronary arterial monolayers in the magnitude and kinetics of radiation induced perturbations to monolayer permeability. Both endothelial monolayer types exhibit a transient 7-8 % decrease in resistance.

Prior work in radiation biology has shown that vascular barriers can be perturbed by irradiation and that this perturbation can be assessed by impedance methods [6, 7, 28]. The significant and unique finding of this paper, using an impedance method, is that a primary human venous endothelium also responds to gamma irradiation with 5.0 Gy with comparable kinetics to those observed in coronary vascular endothelium. Imaging studies revealed no major changes to the cell monolayer which contrasted with findings in other endothelia which retract more robustly [36, 37].

While a number of endothelial functions are conserved throughout the systemic vascular network, other functions differ depending on location. The endothelium throughout the body varies widely, from delicate sinusoids in the marrow to large capacitance veins in the legs, which, in turn, have a different oxygen and hemodynamic shear exposure. Oxidative species are known to be generated in endothelial cells in response to flow [38]. Oxygen synergizes with ionizing radiation to form damaging reactive oxygen species that damage the cell and so locally elevated oxygen concentration together with greater localized flow conditions could result in a larger effective radiation dose.

Endothelial cells sourced from various sites in the body respond differently to radiation. The most robust response is seen in pulmonary endothelial cells which exhibit significant depolarization of the cytoskeleton and a resultant increase in the wet weight of the lung [36, 37]. Such dramatic responses are not seen in other endothelia [39]. Common endothelial functions do exist and, as seen in the case of rolling of leukocytes across immunologically activated endothelial barriers, these can be reproduced ex vivo [40]. Furthermore the induced relaxation of endothelintreated microvessels seems to be able to report pathologies of the vascular system as a whole [29].

Since oxygen content and hydrodynamic flow conditions differ among venous and arterial vessels, it is important to learn whether primary human umbilical venous endothelial cell (HUVEC) monolayers responded in a different way as compared to an arterial endothelium that would otherwise function at a site of greater oxygen demand. In at least two monolayer systems, sourced from different body sites, impedance changes in response to radiation are very similar and hint at a more widely applicable physiological response that could be utilized by radiation dosimetry.

The changes in vascular permeability in response to radiation demonstrated here can support two possible dosimetry strategies. The first strategy involves continuous, painless, noninvasive bioelectric surveillance of an individual’s physiology. An individual who is under such continuous monitoring could have sensors that report a dose of radiation by sensing vascular permeability changes. The wide use of portable electronic devices together with advances in miniaturization of biomedical sensing technologies could make such a strategy feasible in the future. A more feasible strategy at present is to assess the results of a transient change in vascular permeability with a bioimpedance approach. In particular, changes in blood content, fluid content or other tissue architecture changes that are not limited to a seemingly early and transient signal. Such assessment is not without difficulty. It can be argued that the heterogeneity and quality of skin in the human population is significant ̶ varying with age and body site [41, 42]. Strategies to overcome this heterogeneity do exist. Indeed, normalization against high frequency impedance values has been used in previous studies [17].

This work is limited to a single radiation dose of 5.0 Gy in a demonstration of the response in a second type of endothelium. Further validation of this response as a dosimetry modality must necessarily include a greater range of doses to establish the threshold of distinction and the threshold of detection. Prior work demonstrates that the leaking of blood from superficial vessels in rodents occurs more rapidly as the radiation dose increases [43]. As radiation doses increase, it is unclear whether the vascular barrier will become opened to a greater extent (amplitude of response) or will remain open longer (duration of response). In future dose-response studies, this question will assuredly be answered. This work supports the idea that sourcing endothelial cells from more amenable sites in the body other than the coronary circulation is possible to allow for assessment of this response in future experiments. The exvivo platform could also provide a means to further probing the molecular mechanisms behind radiation-induced change in the function and shape of endothelial cells.