Colorectal cancer is the third most frequent type of cancer (after lung and breast cancers), with an incidence of 1,360,602 cases (9.7%) in 2012 all over the world, with a higher incidence for males (10.1%) compared to females (9.2%) (Global Cancer Observatory, International Agency for Research on Cancer, WHO, http://gco.iarc.fr). The estimated number of deaths due to colorectal cancer is 693,933 (8.5%).

Management of colorectal cancer is routinely performed through either surgeries, chemotherapy or radiotherapy (1). The advantage of radiotherapy is the localized delivery of radiation which allows, in the majority of cases, to avoid the systemic adverse reactions typical of chemotherapy. Although the systemic complications are reduced during radiotherapy, consequences can arise in the surrounding normal tissues, i.e., activation of inflammatory pathways, direct damage to healthy cells and non-targeted effects (2–4).

Among the effects elicited by radiation dose delivered during radiotherapy for colorectal cancers, two main issues deserve deep investigation, namely the interplay between tumoral cells and the immune system, and the changes in the intestinal permeability.

One of the most important functions of the intestinal epithelium is to create an impermeable barrier, in order to avoid paracellular passage of molecules and solutes. Impermeability is maintained thanks to the action of junctional complexes between epithelial cells (i.e., tight junctions and adherens junction). These complexes join together with the ones present on the plasma membrane of the surrounding cells creating a barrier, and are furthermore responsible of the polarity of the cellular monolayer (5).

The fundamental proteins that contribute in the formation of tight junctions are occludin and claudin-1, and all the proteins acting as adaptors between the junction and the cytoskeleton (ZO family members, afadin, CD2AP). Besides their function in creating junctions, it is well known that all these proteins play a crucial role in the transduction of signals inside the cell; a lot of these proteins are in fact misregulated in different types of cancer and are frequently responsible of signals able to orchestrate both proliferation and metastasis formation (6–8). As an example, claudin-1 expression was found to be reduced in breast (9, 10) and in colon cancer; in colon cancer it was also associated with recurrence and poor prognosis (11).

There is an increasing evidence that intercellular junction functioning and stability (e.g., tight junction, adherens junction, etc.) can be altered by irradiation (12). It has been demonstrated that the tight junction structure becomes disorganized when exposed to radiation, and this effect leads to an increased permeability of the monolayer (12). Deirò de Carvalho and colleagues showed that ionizing radiation (IR) exposure causes redistribution of the principal junctional proteins, leading to disassembly and loss of function of junctional complexes in Caco-2 cells (5). Moreover, Moyes and colleagues confirmed radiation-induced permeability changes, accompanied by an increased microparticle uptake (13).

Over the last 30 years, Caco-2 cell lines have been widely used as a model of intestinal barrier. Caco-2 cells derive from human colon adenocarcinoma: although they are tumoral cells, Caco-2 show the ability to differentiate in culture to create a functional polarized monolayer (14). Such ability to create a functional monolayer allowed the study of membrane functions when cells are grown on a porous support. This type of cell culture allows to create a difference among the two sides of the monolayer, improving differentiation and creating a gold standard for physiological intestinal transport and toxicity studies (14).

Being the culturing of Caco-2 cells on porous membrane an optimum in vitro model of intestinal barrier, an upgrade of this model is the coculture of Caco-2 cells with other cell types; this setup has been used in several studies to measure the interplay between different cell types (15), therefore adopted to highlight how Caco-2 response to exogenous stimuli is modified by coculture with respect to Caco-2 cells alone. For example, the Caco-2 coculture setup is common in such studies aiming to identify the complex cross talk between with the immune system due to the presence of non-pathogenic bacteria (16).

In their study, Pozo-Rubio and colleagues assessed the level of different cytokines in a Caco-2/peripheral blood mononuclear cells (PBMC) coculture with bifidobacteria stimulating the top layer of Caco-2 cells. They demonstrated that the cytokines secretion profile was completely different when compared with the one obtained after stimulation in absence of PBMC (17).

Other studies aimed especially at understanding the different response of Caco-2 cells to non-pathogenic and pathogenic bacteria. It was observed by Haller et al. that Caco-2 cells show a discriminative activation depending on treatment with lipopolysaccharide from enteropathogenic Escherichia coli spp. or non-enteropathogenic bacteria, i.e., E. coli spp., Lactobacillus johnsonii, and Lactobacillus sakei (18).

The aim of our work was to study the effect of X-ray irradiation in Caco-2 cells alone or cocultured with PBMC. In particular, we focused our study on modification of monolayer permeability, cell proliferation, and cytokine release. Finally, we tried to extrapolate the influence of cytokine spectra alteration on Caco-2 cell permeability and related tight junction pathways.

Colorectal cancer is one of the most important disease affecting industrialized society. Management of this type of tumor is routinely performed through surgery, chemotherapy or radiotherapy. Among these three methods, radiotherapy offers the opportunity to target the tumoral tissues limiting radiation exposure of healthy tissues.

In the last years, a plethora of data shed light on the so-called “bystander” effects, i.e., the radiation effects observed far from the irradiated target, and on the role played by the immune system in fighting against cancer concomitantly with radiotherapy, leading to radio-immunotherapy (2, 20–24).

Within this context, in this work we analyzed the effects of IR exposure with X-rays in colon adenocarcinoma Caco-2 cells, adopting a coculture experimental setup which allows to investigate the interplay between Caco-2 and PBMC.

First of all, we evaluated proliferation and mortality in Caco-2 cells after different doses of X-rays. Although proliferation (assessed with MTT, Figure 1A) did not decrease after radiation exposure, cell mortality (Trypan Blue, Figure 1B) was found to increase in a dose-dependent way up to 10 Gy of X-rays. Even though the increase was statistically significant, absolute values are very low. From the results of these assays, being MTT and Trypan Blue good complementary indicators of radiosensitivity (25), we evaluated the proliferation of viable cells normalizing the MTT results on the percentage of living cells obtained by Trypan Blue staining. Our calculations show an increased proliferation up to approximately 130% at 24 h after 5 Gy-irradiation. Considering all the irradiation doses and both time points, we can state that Caco-2 cells show a high radioresistance in this range of doses, with no severe effects on cellular proliferation or mortality.

The temporal dynamics of the trans-epithelial electrical resistance (TEER) of Caco-2 cell layer was then assessed (Figure 2). Being also present in sham-irradiated cells, the peak in TEER values observed within the first 2–3 h after irradiation can easily be ascribed to the stress induced by the transport to the irradiation facility. According to data available in the literature, we observed a decrease in TEER for the highest 10 Gy dose up to 24 h, while we did not observe changes in 2 Gy-irradiated cells with respect to the sham condition in the same time interval. After the 10 Gy-induced continuous decrease, a recovery process starts and such recovery appears to be completed within 48 h post-irradiation. TEER values for all irradiation conditions at 48 h are back to the pre-treatment level.

The presence of PBMC consistently changes the above described behavior. The same transient peak ascribable to environmental/procedural stress is observed and vanishes in few hours after irradiation. Our results demonstrate that a strong decrease in TEER values for all the irradiation conditions takes place over 30 h post-irradiation; after this time point, recovery starts. Interestingly, data show that the 10 Gy-irradiated cells recover to higher TEER values with respect to other conditions.

Comparing TEER values for Caco-2 with/without PBMC in coculture, we can conclude that the permeability of the Caco-2 layer seems to be strongly affected in presence of cells from the immune system. The mere presence of PBMC induces indeed a significant increase of permeability of the Caco-2 layer (decrease of TEER) over 48 h, which could be interpreted as due to the interplay between tumoral cells and the immune system. The situation is not changed if Caco-2 cells are exposed with 2 Gy X-rays, while a higher dose of 10 Gy seems to induce a shorter term (24 h) decrease in TEER in absence of PBMC, but a higher recovery to a less permeable layer at longer times (48 h) in presence of PBMC. A radiation effect appearing only at 10 Gy is consistent with the high radioresistance of Caco-2 derived from proliferation and mortality.

Expression of several cytokines from Caco-2 cells is differentially modulated depending both on the radiation exposure and the presence of PBMC (Figure 3). Concerning the family of interleukins, it is evident that the major effect observed is on IL-4, IL-5, and IL-15. Our results show that IR and coculture add their effects in the case of IL-4, while radiation only acts in reducing release of IL-5; in this case the presence of PBMC completely stops the release of this cytokine. In the case of IL-15, IR and the presence of PBMC taken alone increase the release, and all together the synergic effect is highlighted.

Taken together, these data highlight the modulation of IR and the cross talk between the two different components of the coculture. Considering the effects of measured interleukins, we can speculate an enhanced effect on proliferation of NK cells due the synergic effect on IL-15. In addition, the decrease of IL-4 could lead to the deregulation of the Th1 and Th2 subset of CD4⁺ T cells, with the reduction of Th1 cells, as highlighted in the work done by Pellegrini et al. (26). Of particular interest is the behavior of IGFBP-4; this cytokine is known to cause an increase in apoptosis and Bax protein expression, and a decrease in tumor cellular mitosis (27). While radiation and coculture taken alone reduce the production of IGFBP-4, the presence of PBMC with irradiated cell translates into an increased production of this protein: also radiation itself causes an enhanced release in the coculture setup. Of great interest is also the modulation of IFN-γ and TNF-α. IR and coculturing seem to have a synergic effect in decreasing the release of IFN-γ, and, on the other hand, in increasing TNF-α. The modifying action of IFN-γ on tight junctions is recognized, although a pleiotropic effect has been demonstrated for this cytokine. In studies about inflammation, direct treatment with IFN-γ acts increasing the paracellular permeability of endothelial and epithelial monolayers. However, in airway epithelial cells, IFN-γ exposure has anti-inflammatory properties and promotes epithelial barrier function (28, 29). For the interpretation of the enhanced TNF-α release, we recall that Ma and colleague showed that TNF-α exerts late effect on Caco-2 cell permeability, with an increased small-molecule flux within 24 h of treatment, and TEER alteration at 48 h post-treatment (30).

On the other hand, studies performed on normal colon cells showed a significantly higher amount of IFN-γ and TNF-α in inflamed-induced murine model causing, as a consequence, an enhanced proliferation in intestinal epithelial cells (31–33).

Data obtained through western blotting showed that the typical tight junction proteins claudin-1 and occludin are not affected by both radiation exposure and coculture. On the contrary, we observed changes in the scaffold proteins ZO-1, ZO-2, and afadin (Figure 4). ZO-1 is a fundamental scaffold protein which provides a binding site for a plethora of other proteins, i.e., ZO-2, ZO-3, occludin, and F-actin; its function is then mandatory to connect cellular cytoskeleton to the junctional complexes. Differently from Youakim and Ahdieh (34), who observed a mis-localization of ZO-2 after IFN-γ treatment, we observed a reduction in the levels of this protein. An effect on afadin was also found. Changes in the amount of ZO-2 and afadin after coculturing suggest the influence of coculture on the scaffold system, whose function is mandatory for the tight junction efficiency. The alteration observed in our setup could be explained by the changes in the observed values of the scaffold proteins. We can suppose such effects being mediated by the cross talk between Caco-2 cells and PBMC, in particular when Caco-2 cells are irradiated. Two of the major player of this cross talk can be easily identified in the INF-γ and TNF-α.

For what concerns the PBMC activation in response to Caco-2 cells coculture, we studied NF-κB and XIAP. NF-κB is a family of pleiotropic transcription factors which can be found in almost all cell types and it is involved in a myriad of cell functions, among which inflammation, immune response, and apoptosis. Exogenous signals triggering the cell surface receptors activate the canonical NF-κB signaling pathway, which first leads to the phosphorylation of the NF-κB-bound inhibitor (IkB), the nuclear translocation of the NF-κB dimers, mostly p65-containing heterodimers, and finally binding to the DNA to allow the regulation of the transcriptional activity of the cell (35, 36). But NF-κB activity influences also other cross-linked pathways, such as the p53 and apoptotic pathway, through the up-regulation of genes encoding inhibitory proteins, e.g., A20 and XIAP. In particular, XIAP is a member of the inhibitor of apoptosis family of proteins (IAP), which arrests apoptotic cell death binding to caspase 3, 7, and 9.

In our in vitro model, western blot assays did not reveal any increment in the constitutive levels of NF-κB of PBMC in irradiated cocultures, but the increase in the XIAP expression levels allows us to hypothesize its induction by MDM2 (37) and the activation of the cross-linked NF-κB pathway (38), with a subsequent inhibition of the extrinsic apoptotic pathway through the inhibition of the caspase 3 and caspase 7 (39) (Figure 5).

In conclusion, our results show a weak effect of radiation on Caco-2 cells up to 10 Gy X-rays. The interplay with the immune system, addressed with a coculture setup with PBMC, induced increased permeability of the Caco-2 monolayer. Cocultured PBMC are activated, inhibiting the apoptotic pathway, therefore enhancing the release of cytokines in the culture medium. This could, in turn, affect tight junction scaffold proteins through IFN-γ- and TNF-α-mediated signaling (as illustrated in the simplified scheme of Figure 6).

JM, GBabini, GBaiocco, and AO: conceived the experiments. JM and GBabini: designed and performed all the experiments. JM, GBabini, GBaiocco, and SB: performed data analysis and data interpretation. JM, GBabini, GBaiocco, and SB: wrote and edited the manuscript. AO: critically read the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationship that could be construed as a potential conflict of interest.