Unlike Th1/Th17 cells, Th2/Th9 cells selectively migrate to the limbus/conjunctiva and initiate an eosinophilic infiltration process

Abstract

In this study we compared polarized mouse T-helper (Th) lymphocytes of four populations, sensitized against an ocular antigen, for their patterns of migration and induction of inflammatory processes in recipient mouse eyes expressing the target antigen. Th1, Th2, Th9 and Th17 cells transgenically expressing T-cell receptor (TCR) specific against hen egg lysozyme (HEL) were adoptively transferred to recipient mice expressing HEL in their eyes. Recipient eyes collected 4 or 7 days post injection were analyzed for histopathological changes. Th1 and Th17 cells induced moderate to severe intraocular inflammation in the recipient mouse eyes, but essentially did not migrate into the conjunctiva. In contrast, Th2 and Th9 cells invaded minimally the intraocular space of recipient eyes, but accumulated in the limbus and migrated into the conjunctiva of the recipient mice and initiated allergy-like inflammatory responses, as indicated by remarkable eosinophil involvement. These data thus shed new light on the differences between the migration patterns and ocular pathogenic processes mediated by Th1/Th17 and by Th2/Th9 populations.

Research in recent years has established the existence of different populations of lymphocytes, as well as their specific roles in the different functions of the immune response. In addition to their defined roles in defense against various infections, the known populations of T-helper (Th) cells were found to be responsible for mediation of immune responses with adverse effects. Thus, Th1 and Th17 populations are responsible for “cell-mediated” immune responses, such as initiation of certain autoimmune conditions (Jiang and Dong, 2013; Zhu et al, 2010), whereas Th2 and Th9 cells are involved in mediation of allergic responses (Rodriger and Weninger, 2015; Tan and Gery, 2012; Wynn, 2015).

The availability of transgenic mice that express T-cell receptor (TCR) specific to a single antigen made it possible to generate cell lines of the different Th populations and analyze their activities. We have used such a system to investigate the capacity of different Th populations to initiate immune-mediated ocular inflammation. In our system, transgenic Th lymphocytes, specific against hen egg lysozyme (HEL), are injected into syngeneic mice transgenically expressing HEL in their lens and induce ocular inflammation. (Cox et al., 2008; Kim et al, 2002; Lai et al., 1998). In the present study we compared HEL-specific polarized Th1, Th2, Th9 and Th17 cells for their migration patterns and capacity to induce inflammation in different tissues of recipient mouse eyes expressing HEL. It is of note that the transgenic expression of HEL damages the lens capsule and, as a result, HEL spills over to intraocular tissues of the recipient eye (Kim et al., 2002; Lai et al., 1998).

We found two different patterns of inflammation induction by the Th populations: Th1 and Th17 cells induced, as expected (Cox et al., 2008; Shi et al, 2009), moderate to severe intraocular inflammation that included inflammatory cell infiltration in both the anterior and posterior segments of the recipient eyes, including the anterior chamber and ciliary body, as well as the vitreous and retinal tissue (Fig. 1). Also of note is the typical retinal edema (Fig. 1) (Cox et al., 2008). Th1 and Th17 cells did not migrate, however, to the conjunctiva, despite the migration of a portion of these cells through the limbus toward the intraocular tissues (Fig. 1, arrows). In contrast, Th2 and Th9 cells induced inflammatory changes essentially only in the limbus and conjunctiva of the recipient mice (Fig. 1).

Figure 1.

Remarkable differences between the location and type of inflammation mediated by Th1 and Th17 cells and by Th2 and Th9 cells. CD4 cells from transgenic 3A9 mice, specific to HEL, were polarized toward these four phenotypes as detailed elsewhere (Cox et al., 2008; Shi et al., 2008; Tan et al., 2010) and injected (5 × 10⁶ per recipient) into recipient mice expressing transgenically HEL in their lens (Cox et al., 2008; Shi et al., 2008). The recipient eyes were collected at the peak of their inflammatory process, i.e., day 4 post-cell transfer for Th1 recipients and day 7 for the other groups (see Shi et al. 2009 for details on kinetics of ocular inflammation in Th1 and Th17 recipient mice). The Figure presents histological eye sections from representative recipients of the four Th populations at three magnifications. Th2- and Th9-mediated inflammation selectively accumulated in the limbi and conjunctivae of the recipient mouse eyes, whereas Th1 and Th17 cells mediated intraocular inflammatory process. Notably, Th1 and Th17 cells accumulated in the limbal area (black arrows), an entry site to the intraocular tissues/spaces, but did not invade the conjunctivae. The mouse eye sections in this Figure are representatives of 6–8 mice of each of the corresponding groups, tested in 3 separate experiments. H&E staining.

In order to further analyze the differences between the two patterns of pathological changes, induced by Th1/Th17 and by Th2/Th9, we included in the recipient eye sections the lid with the complete conjunctival tissue. Fig. 2 shows representative eye sections of recipients of Th2 and of Th17 cells. Striking differences are seen between the histological changes in these two recipient eyes. Whereas no apparent intraocular changes are detected in the Th2 recipient eye (Fig. 2A), severe inflammatory changes are seen in all ocular tissues of the Th17 recipient eye (Fig. 2D). Clear differences were also noted between eyes of the two recipient mice in the pathological changes in their limbal/conjunctival areas. Thus, considerable accumulation of inflammatory cells is seen at the limbal and ciliary body areas of the Th17 recipient mouse (Figs. 2D and 2E), but there is only minimal “spill over” of the inflammatory cells from the limbus into other conjunctival areas in this Th17 recipient eye (Fig. 2E). On the other hand, inflammatory cells infiltrate both the limbal and the bulbar conjunctival areas of the Th2 recipient eye (Figs. 2B). Furthermore, whereas only very few eosinophils could be detected among the inflammatory cells in the limbus of the Th17 recipient (Fig. 2F), the lymphoid cell infiltration in the limbal/conjunctival areas in the Th2 eye consisted of remarkable proportions of eosinophils (Fig. 2C). Eosinophils were readily identified by routine H&E staining, based on their intense staining with eosin (red), as well as their typical bilobed nuclei.

Figure 2.

Extended analysis of differences between the pathogenic changes induced by Th2 and by Th17 cells in recipient mouse eyes. The experimental system is the same as that detailed in the legend for Fig. 1, with the exception that the Th17 cells were adoptively transferred at 10 × 10⁶/recipient. (A) An eye section of a recipient of Th2 cells. No intraocular changes are detected, but intense accumulation of inflammatory cells is seen at the limbal and conjunctival areas. (B) The inflammatory cells infiltrated only the bulbar portion of the conjunctiva; the white arrow indicates the end of the inflammatory cell infiltration. (C) The inflammatory cell population in the conjunctiva consists of a large proportion of eosinophils (white arrows). (D) Eye section of a recipient of Th17 cells. Severe intraocular changes, affecting all eye tissues. (E) In addition, inflammatory cells accumulate at the limbal area, but with just a minimal “spill over” into the conjuctival tissue (black arrow). Also of note is the intense accumulation of inflammatory cells at the limbal/ciliary body area (white arrow); these cells are assumed to be migrating toward the intraocular space. (F) Essentially no eosinophils are identified among the inflammatory cells at the limbal area of this eye. Panel G and H: identification by PAS staining of goblet cells in the conjunctivae of an unjected control mouse (G) and a recipient of Th2 cells (H). Goblet cells, that express mucin (red), are found in sections of these two eyes around the conjunctival sac. Panels A–F, eye sections stained by H&E, show representative eyes of mice, in 2 experiments, with a total of 6 mice of each treatment. Panels G and H (PAS staining) show representative eyes of a total of 4 mice of each group, in two separate experiments.

Analysis of the presence of eosinophils among other inflammatory cells in the affected conjunctivae revealed that the eosinophils are a major component of the inflammatory infiltration, presenting up to 50% of the total infiltrating cells (when counted within 200×100 μm portions of the conjunctival section). Notably, a correlation was seen between the intensity of the conjunctival inflammation and the percentage of eosinophils among the inflammatory cell population. Eosinophils were found throughout the affected areas of the recipient conjunctivae, but were “aggregated” in certain locations (e.g., Fig. 2C).

To Further analyze the pathological process in the conjunctivae of experimental mice, we examined the presence of goblet cells in this tissue of experimental mice. Goblet cells were identified by PAS staining in conjunctival sections of Th2 recipient mice and of their uninjected controls. Considerable variability was noted among eyes of different mice in the location and number of goblet cells. Eye sections shown in Figs. 2G and 2H represent eyes of recipients and control mice in which goblet cells were localized mainly at the conjunctival sac; another location for goblet cell accumulation was found to be the palpebral section of the conjunctiva (not shown). Unlike in the “classical” form of ocular allergy, in which hyperplasia of goblet cells is a common feature (Dartt and Masli, 2014), we did not notice substantial differences between eyes of Th2 recipients and their controls in the number of goblet cells in their conjunctivae. This difference between the two types of inflammatory responses could be attributed to the participation of different mediators in these two pathogenic processes.

Differences between the populations of Th1/Th17 cells and Th2/Th9 cells in their biological functions are well established. (Eyerich and Zielinski, 2014; Jiang end Dong, 2013; Zhu et al, 2010). Little is known, however, concerning the migration patterns of cells of these populations to sites where their target antigens are expressed/located. The experimental system we used in this study thus made it possible to address this issue, since the eye is composed of various tissues. Significantly, Th1 and Th17 migrated into the intraocular spaces and tissues, whereas Th2 and Th9 migration was limited to the limbus/conjunctiva. It is conceivable that the migration patterns were determined by at least two local properties, i.e., (i) the concentration of HEL (the target antigen), and (ii) features provided by the different tissues. The HEL concentration is higher by far inside the eye ball than in the limbal/conjunctival tissues and it is striking that, unlike Th1 and Th17 cells, both Th2 and Th9 cells “avoided” the high concentrations of the target antigen in their migration pattern. This selective migration of Th2 and Th9 cells into the conjunctiva could be attributed to the unique “microenvironment” of this tissue, which is assumed to be determined by the profile of chemokines/chemokine receptors and adherence molecules that selectively attract cells with the “matched” profiles of these molecules.

The conjunctival tissue is a site commonly affected by allergic responses, as indicated by allergic conjunctivitis being a common eye disease in humans (Bielory and Friedlander, 2008; Ono and Abelson, 2005). The involvement of the conjunctiva in pathogenic allergic processes is likely related to the exposure of this tissue to the environment and to airborne allergenic molecules. Our findings now indicate that, in addition, the conjunctiva provides features that attract the migration of Th2 and Th9 cells, but not of Th1 and Th17 cells (Fig. 1). This observation is in line with the selective involvement of Th2 and Th9 cell populations in the pathogenic process of allergy (Kaplan et al., 2015; Wynn, 2015). The pro-allergy capacity of the Th2 and Th9 cells is also indicated in our study by the prominent involvement of eosinophils, the hallmark cells for allergic responses (Nadif et al., 2013; Rosenberg et al., 2013), in the inflammatory reactions initiated by these two Th populations in the limbi and cojunctivae of recipient mice (Fig. 2). We propose that the pathological changes in these recipient mice present a new animal model for “allergy-like” process, that differs from the “classical” allergic conjunctivitis by its not involving pathogenic components such as IgE antibodies, mast cells, or the specific allergy mediators. This animal model also differs from the “classical” disease by inflammatory process essentially affecting only the bulbar portion of the conjunctiva.

Our findings with the experimental system used here are in line with observations made with patients with intraocular inflammation (“uveitis”), or with allergic processes in the conjunctiva. Thus, similarly to the experimental systems, Th1 and Th17 cells were found to be dominant in the pathogenic process of uveitis (Horai and Caspi, 2011), whereas Th2 cells play major roles in allergic conjunctivitis (Reyes and Saban, 2014; Wynn, 2015). Importantly, however, Th1 and Th17 cells are also actively involved in the allergic process in humans, in particular in the more severe cases (reviewed by Reyes and Saban, 2014).

In summary, our data show, for the first time, the remarkable differences between sensitized Th1/Th17 cells and Th2/Th9 cells in their migration patterns and the inflammatory processes they evoke in tissues of the eye.

Legend for the Figure for the Rebuttal (attached separately).

Goblet cells in conjunctivae of an uninjected control mouse (A) and a recipient of Th2 cells. Please note the selective location of the goblet cells (PAS-positive) mostly in the palpebral sections of these conjunctivae (black arrows), with very few in the bulbar sections (white arrows)