Experimental Support for the Ecoimmunity Theory: Distinct Phenotypes of Nonlymphocytic Cells in SCID and Wild-Type Mice

Abstract

Immune tolerance toward “self” is critical in multiple immune disorders. While there are several mechanisms to describe the involvement of immune cells in the process, the role of peripheral tissue cells in that context is not yet clear. The theory of ecoimmunity postulates that interactions between immune and tissue cells represent a predator–prey relationship. A lifelong interaction, shaped mainly during early ontogeny, leads to selection of nonimmune cell phenotypes. Normally, therefore, nonimmune cells that evolve alongside an intact immune system would be phenotypically capable of evading immune responses, and cells whose phenotype falls short of satisfying this steady state would expire under hostile immune responses. This view was supported until recently by experimental evidence showing an inferior endurance of severe combined immunodeficiency (SCID)-derived pancreatic islets when engrafted into syngeneic immune-intact wild-type (WT) mice, relative to islets from WT. Here we extend the experimental exploration of ecoimmunity by searching for the presence of the phenotypic changes suggested by the theory. Immune-related phenotypes of islets, spleen, and bone marrow immune cells were determined, as well as SCID and WT nonlymphocytic cells. Islet submass grafting was performed to depict syngeneic graft functionality. Islet cultures were examined under both resting and inflamed conditions for expression of CD40 and major histocompatibility complex (MHC) class I/II and release of interleukin-1α (IL-1α), IL-1β, IL-6, tumor necrosis factor-α (TNF-α), IL-10, and insulin. Results depict multiple pathways that appear to be related to the sculpting of nonimmune cells by immune cells; 59 SCID islet genes displayed relative expression changes compared with WT islets. SCID cells expressed lower tolerability to inflammation and higher levels of immune-related molecules, including MHC class I. Accordingly, islets exhibited a marked increase in insulin release upon immunocyte depletion, in effect resuming endocrine function that was otherwise suppressed by resident immunocytes. This work provides further support of the ecoimmunity theory and encourages subsequent studies to identify its role in the emergence and treatment of autoimmune pathologies, transplant rejection, and cancer.

Keywords

Major histocompatibility complex (MHC) class I Lymphocytes Pancreatic islets Autoimmunity Ecoimmunity

Introduction

Elucidating the mechanisms that facilitate “self” and “nonself” discrimination, as well as the maintenance of peripheral immune tolerance, is a long-standing challenge (15). While multiple models provide reasonable explanations of the conundrum of immune tolerance, based on particular conditions or within specific tissues, a unifying model has yet to emerge that defines which entities are treated by the immune system as an authentic threat and elicit a long-lasting immune memory (32). According to the current immunological view, immune regulation is achieved, inter alia, by clonal selection, which depends on antigen presentation in the context of potent costimulatory signals (33). This process is further extended to deletion of adaptive immune cells by repeated antigen stimulation and by active suppression of unwanted clones via regulatory T cells (11,28,30). However, this context leaves several questions unanswered: What exactly is being tolerated: is it merely an antigenic molecule that stimulates a response or rather a molecule within a specific context? Is peripheral reeducation of immune cells a lifelong process or rather a developmental one? How can a “unipole” interaction that is governed mainly by the immune system within an ever-changing antigenic realm result in a stable immune adaptation? And assuming the system is dynamic, how do nonimmune cells influence the immune system so that it can adjust tolerance levels when immunocytes encounter tissue cells?

Transplantation is an example of an inducible immune response that relies mostly on antigen presentation (12,22). While it appears that there is no evolutionary rationale for a host to launch a full-scale aggressive destruction of grafted tissues, an intact immune system will, in fact, readily destroy sterile donor tissues (15,20,26). Following the current immunological view dogma, the presence of “nonself” antigens determines the outcome of grafted tissues. However, there is evidence to suggest that prior to antigen presentation (approximately within the first 96 h after transplantation), the recipient mounts an unopposed inflammatory response that causes extensive tissue damage, regardless of antigen specificity (3). Importantly, this response is greater than that mounted against an autotransplant, although it occurs before antigen presentation (2, 8).

The theory of ecoimmunity postulates that the immune system and the individual's own tissues continuously interact with each other, fulfilling a predator and prey relationship and obeying the rules of coevolution (24). The theory suggests that immune tolerance to “self” is not merely the result of immune monitoring, but rather the consequence of a dynamic balance between immune cells that act to prey on tissue cells and the ability of tissue cells to avoid predation. The theory of ecoimmunity thus implies that interactions between immune cells and tissue cells coevolve symmetrically throughout ontogeny: while clonal selection shapes the population of immune clones and their specificities, selection of tissue–cell phenotypes defines the arsenal of nonimmune molecular “defense tools.” An analogous situation would be a macroscopic ecological system in which predators (such as lions) regulate prey populations (such as antelopes) and remove deficient individuals (e.g., slow and potentially diseased antelopes). The process is therefore beneficial for the tissue: immune cells remove cells that have failed to adapt their phenotype to one that allows evading the immune system (23). Predator–prey interactions are not limited to tissue– and immune–cell interactions in health; perturbation of this relationship might result in aberrant processes, such as tumorigenesis and autoimmune and degenerative diseases. Continuing the analogy of an imbalance within a perturbed ecosystem, the interaction can describe conditions that lead to disease. Immune tissue interactions that occur when degenerative or autoimmune diseases are present, for example, may be analogous to a scenario in which a population of antelopes that has outgrown its environment becomes extinct because of a transmitted disease. An alternative analogy to an interaction of the immune system with a diseased, aberrant population of cells (24, 25) is that of lions that predate on antelopes who are challenged by some biotic or abiotic conditions. This view may imply a universal observation encompassing immune cells, nonimmune cells, and pathogens and extends to pathologies as prevalent as autoimmunity, transplant rejection, and cancer (7).

A functional manifestation of the ecoimmunity theory was originally examined by Hauben et al. (10). According to their study, grafted peripheral cells in a severe combined immunodeficiency (SCID) animal evolve differently from peripheral cells in a wild-type (WT) animal with an intact immune system. In other words, SCID islets represent antelopes that are not accustomed to living in the same environment as lions; when the antelopes are translocated to a niche inhabited by lions, these lions rapidly kill off these nonadjusted antelopes, even though lions normally can coexist with antelopes in this niche. Indeed, Hauben et al. observed that WT mice that were transplanted with pancreatic islets from syngeneic SCID mice displayed rapid graft failure, whereas mice that received grafts from syngeneic WT islets did not reject the graft. The choice of islets is highly justified in this context, both as targets of clinical autoimmunity and also in light of their pronounced (and readily measurable) susceptibility to inflammation and cell injury. In addition, pancreatic islets contain tissue cells, such as endocrine α and β cells, and resident immune cells such as macrophages and B lymphocytes (4, 5).

In the present study, we challenge the theory of ecoimmunity by examining the idea that the phenotype of different cells is shaped by the copresence of an adaptive immune system. We hypothesize that with the absence of interactions with lymphocytic immune cells, nonimmune cells will not be exposed to a selective pressure and thus enhance the activation state of immunoreactive tissues. We therefore compared pancreatic islets, spleen cells, and bone marrow (BM) cells in both SCID and WT mice under relevant challenges and studied the immune-relevant profile of subsequently distinct nonimmune cells. Finally, we examined the implications of intraislet resident immunocytes on the endocrine functions of the islet.

Materials and Methods

Animals

Male imprinting control region (ICR) mice 8 to 10 weeks old and SCID mice (ICR background) were used as islet donors (Harlan Laboratories Ltd., Rehovot, Israel). Male ICR mice 8 to 10 weeks old (Harlan Laboratories Ltd.) were used as islet graft recipients. All experiments were approved by the Ben-Gurion University of the Negev Animal Care and Use Committee.

Pancreatic Islet Isolation

Pancreatic islets were isolated as previously described (19). Briefly, donor mice were anesthetized with a combination of ketamine [80 mg/kg administered via intraperitoneal injection (IP); Vétoquinol, Lure, France] and xylazine (12 mg/kg IP; Eurovet Animal Health, Bladel, The Netherlands). Pancreata were inflated with collagenase (Type XI; 1 mg/ml; Sigma-Aldrich, Rehovot, Israel), excised, and incubated for 35 min at 37°C. Digested pancreata were vortexed and filtered through a 500-μm sieve, and the pellet was washed in Hank's balanced salt solution (HBSS; Biological Industries, Beit Haemek, Israel) containing 0.5% bovine serum albumin (BSA; Sigma-Aldrich). The pellet was resuspended in an RPMI-1640 medium (Biological Industries) supplemented with 10% fetal calf serum (FCS), 50 U/ml penicillin, and 50 μg/ml streptomycin (all from Biological Industries). Islets were then collected in a 100-μm cell strainer (BD Falcon, San Jose, CA, USA) and hand-picked under a stereomicroscope (Accu-Scope, Commack, NY, USA).

Islet Transplantation

Recipient mice were rendered hyperglycemic by a single dose of streptozotocin (225 mg/kg IP; Sigma-Aldrich) and grafted with 250 islets from either SCID or ICR mice under the renal capsule, as previously described (19). Briefly, prospective recipients were screened for nonfasting circulating glucose levels of ≤400 mg/dl. Hyperglycemic recipients were anesthetized, and islets were implanted under the renal subcapsular space, which was immediately sealed with a sterile absorbable gelatin sponge (Surgifoam; Ethicon, Somerville, NJ, USA). Posttransplantation blood glucose levels were monitored throughout the experiment by sampling of tail blood. Islet grafts were then harvested from recipient mice for further inspection.

Gene Expression Profiling

Explanted islet graft total RNA samples were processed for gene expression profiling by using Affymetrix Mouse Gene 1.0 ST Array, according to the manufacturer's instructions (Affymetrix, Santa Clara, CA, USA).

Histology and Immunostaining

Kidneys containing islet grafts were fixed in 10% formalin (Sigma-Aldrich). A cut was made through the center of the graft 24 h later, and the kidneys were then embedded in paraffin. For histological examination, hematoxylin and eosin (H&E; Dako, Carpinteria, CA, USA) staining and immunofluorescence staining were performed. Insulin immunostaining was performed with guinea pig anti-swine insulin (1:200; Dako Cytomation, Glostrup, Denmark), detected by Cy3 donkey anti-guinea pig (1:200; Jackson ImmunoResearch, West Grove, PA, USA). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (1 μg/ml; Sigma-Aldrich). Immunofluorescence was detected with an Olympus BX60 microscope (Olympus UK Ltd., London, UK).

Flow Cytometric Analysis

Renal, BM, and spleen cells and islet single-cell suspensions (1 × 10⁶ per sample) were stained with a combination of anti-CD11b-Pacific Blue, anti-F4-80-APC, and anti-major histocompatibility complex (MHC) class II-APC/Cy7, anti-CD40-PE/Cy7, anti-CD86-PE, anti-CD45-PerCp 5.5, anti-GLUT2-APC, and anti-MHC class I-FITC (1:100; BioLegend, San Diego, CA, USA). Nonspecific staining was excluded by using matching isotype control antibodies (BioLegend). Nonspecific Fc staining was minimized by the addition of Fc receptor blocking solution. Samples were analyzed with a Canto II FACS analyzer (BD Biosciences, San Jose, CA, USA), and data were analyzed with FlowJo (Tree Star, Ashland, OR, USA).

Islet Cultures

Thirty islets per well in quadruplicate were cultured in RPMI-1640 medium (Biological Industries) supplemented with 5% FCS, 50 U/ml penicillin, and 50 μg/ml streptomycin in the presence or absence of recombinant murine interleukin-1β (IL-1β) and interferon-γ (IFN-γ) (5 ng/ml each; PeproTech, Rehovot, Israel). Supernatants were collected 48 h later for analysis by a Q-Plex mouse cytokine chemiluminescence-based 9-p enzyme-linked immunosorbent assay (ELISA) (Quansys Biosciences, Logan, UT, USA). Each cytokine was quantified by densitometry with Quansys Q-View software (Quansys Biosciences). Nitric oxide levels were evaluated by Griess reagent assay (Promega, Madison, WI, USA). Insulin levels were determined by the mouse-specific insulin ELISA (Mercodia, Uppsala, Sweden).

Statistics

Analyses were performed with Prism 6.01 software (GraphPad, La Jolla, CA, USA).

Results were expressed as the mean or median ± standard error of the mean. The statistical significance of differences between groups was determined by a two-tailed Student's Mann–Whitney test. Results were considered significant at p ≤ 0.05.

Results

Inferior SCID/ICR Islet Graft Function in ICR Recipient Mice

Islet durability upon engagement with host immunocytes was examined by grafting hyperglycemic ICR mice with 250 islets from either SCID/ICR mice (n = 5) or ICR mice (n = 6). Graft function was monitored for 16 days. As shown in Figure 1A, mice grafted with SCID/ICR islets failed to control blood glucose levels throughout the follow-up, compared with mice grafted with ICR islets. According to histology, ICR mice grafted with SCID/ICR islets displayed infiltration of cells (Fig. 1B) and a low number of insulin-producing β cells (Fig. 1C). In contrast, ICR mice grafted with ICR islets displayed intact islet morphology and a greater number of insulin-producing β cells (Fig. 1B and C).

Figure 1.

Distinct response of imprinting control region (ICR) and severe combined immunodeficiency (SCID)/ICR islets to transplantation-related stress. ICR mice were rendered hyperglycemic by single-dose STZ (225 mg/kg) and then grafted with syngeneic islets (250 islets per graft) from ICR (n = 6; solid line) or SCID mice (n = 5; dashed line). (A) Graft glucose follow-up. Glucose follow-up was performed three times a week. (B) Histology. Representative graft histology. (C) Insulin content in graft site. Immunofluorescent staining. Red, insulin; blue, DAPI (4′,6-diamidino-2-phenylindole) nuclear counterstaining. Representative images.

SCID/ICR and ICR Islets Have Distinct Steady-State Gene Expression Profiles

Differences between steady-state gene expression in SCID/ICR and ICR islets (n = 3 in each group) were determined by using a gene array composed of 28,853 genes (Fig. 2). Pronounced differences in the expression profiles of SCID/ICR and ICR islets were shown in 59 genes; SCID/ICR islets displayed 41 underexpressed genes (Fig. 2A) and 18 overexpressed genes (Fig. 2B) compared with steady-state ICR islets. A computer-generated output for immune-related enriched processes is shown in Table 1.

Figure 2.

Gene expression profile: steady-state ICR and SCID/ICR islets. Total RNA was purified from freshly isolated islets (100 per sample) obtained from SCID/ICR and ICR mice (n = 3 for each). Affymetrix Mouse Gene 1.0 ST Array showing relative ICR to SCID/ICR (A) downregulated and (B) upregulated genes. Color index: red, positive fold change; green, negative fold change; black, no change. Numerical scale: degree of fold change: 1, most underexpressed; 12, most overexpressed; 6, no change. Presented 59 genes out of 89 genes selected by virtue of a large fold change between SCID/ICR and ICR.

Table 1.

SCID/ICR-Derived Genes Involved in Cellular Pathways

Gene Ontology (SCID Relative to ICR)	No. Genes/Total Input Genes	p Value
Phosphoinositide 3-kinase complex	2/41	0.024
Acute phase	2/41	0.046
Positive regulation of adaptive immune response	2/41	0.074
Regulation of cellular metabolic process	11/41	0.08
Positive regulation of leukocyte mediated immunity	2/41	0.086
Positive regulation of lymphocyte mediated immunity	2/41	0.086
Positive regulation of immune effector process	2/41	0.99

Gene list was obtained from the Affymetrix Mouse Gene 1.0 ST Array, and genes were analyzed by DAVID software for cellular pathway involvement.

SCID/ICR and ICR Cells Differ in Surface Expression of Immune-Related Markers

Splenocytes and BM cells were obtained from SCID/ICR and ICR mice so that their immunological profiles could be examined. Surface expression of MHC class I, MHC class II, and CD40^HI was examined by flow cytometry on a population of CD45⁺ cells, as well as on the specific subpopulations of CD45⁺CD11b⁺ cells and CD45⁺CD11b^– cells. In addition, the expression of MHC class I was examined in renal nonlymphocytic cells, representing a nonlymphoid compartment. As shown in Figure 3A, BM-derived CD45⁺ cells from SCID/ICR mice exhibited significantly lower surface levels of MHC class II and CD40^HI compared with cells from ICR mice (4.16 ± 0.02-fold and 9.99 ± 0.05-fold, respectively). A consistent trend was observed in spleen cells: SCID/ICR CD45⁺ splenocytes exhibited significantly lower levels of MHC class II and CD40^HI compared with splenocytes from ICR mice (2.1 ± 0.04-fold and 12.54-fold, respectively).

Figure 3.

Unique immune-related cell profiles depend on anatomical compartment and on cell subtype. Freshly isolated spleen cells, BM cells, and renal cells (1 × 10⁶ cells per sample) were obtained from SCID/ICR and ICR mice (n = 5 and n = 3 separate donors, respectively) and analyzed by flow cytometry. (A) Major histocompatibility complex (MHC) class II and CD40 surface expression, as gated from CD45⁺ cells. (B) MHC class II and CD40 surface expression, as gated from CD11b⁺ and CD11b^– BM cells. (C) MHC class I surface expression, as gated from CD45⁺ or CD11b⁺ cells. (D) MHC class I surface expression on renal cells. Representative results from two independent experiments. Mean ± SEM, *p <0.05, ***p <0.001.

The expression of MHC class II and CD40^HI was also examined on both CD45⁺CD11b⁺ and CD45⁺CD11b^– BM-derived cells (Fig. 3B). As shown in Figure 3B, SCID/ICR CD45⁺CD11b⁺ cells exhibited surface levels of MHC class II similar to that of cells of ICR mice. However, the SCID-derived CD45⁺CD11b^– cell population displayed significantly lower levels of MHC class II, compared with the ICR group (2.56 ± 0.02-fold). The expression of CD40^HI on CD45⁺CD11b⁺ cells was slightly lower in the SCID/ICR group, compared with ICR mice (1.69 ± 0.08-fold). A consistent trend was observed in the CD11b^– cell population with respect to MHC class II, as SCID/ICR mice displayed significantly lower CD40^HI levels compared with ICR mice (6.19 ± 0.03-fold).

MHC class I expression is presented in Figure 3C. Both BM and spleen-derived CD45⁺ cells from SCID/ICR mice exhibited significantly higher levels of MHC class I expression compared with cells from ICR mice (4.54 ± 0.14-fold and 3.06 ± 0.02-fold, respectively). A similar expression pattern was observed in CD11b⁺ cells: SCID/ICR mice displayed considerably higher MHC class I expression levels in BM CD11b⁺ cells and in spleen-derived CD11b⁺ cells, compared with ICR mice (6.23 ± 0.07-fold and 1.8-fold, respectively).

Further examination of MHC class I levels in nonimmune renal cells revealed a similar trend to that of spleen and BM cells. As shown in Figure 3D, SCID/ICR renal cells displayed significantly higher MHC class I surface levels compared with cells from ICR mice (3.64 ± 0.12-fold).

Islets From SCID/ICR Mice Display an Immune-Stimulatory Profile and Low Tolerability to Inflammation In Vitro

To assess whether SCID/ICR and ICR pancreatic islets are similarly prone to aggravate the immune system, we addressed differences in the inflammatory behavior of islets (Fig. 4).

Figure 4.

Distinct inflammatory and insulin release profiles for ICR and SCID/ICR islets. Primary islets from ICR or SCID/ICR mice (n = 10 from each strain, 30 per well in triplicate) were incubated in the presence or absence of interleukin-1β (IL-1β) and interferon-γ (IFN-γ) (5 ng/ml each) for 48 h. Inflammatory mediators and insulin levels were determined, and cell injury was assessed by lactate dehydrogenase (LDH) release. (A) Resting islets and (B) stimulated islets. Representative results from three independent experiments. Mean ± SEM, *p < 0.05, **p < 0.01.

Leakage of lactate dehydrogenase (LDH) assay was used to determine the levels of released inflammatory products and islet function and viability. The latter were also assessed by examination of insulin accumulation levels in the supernatant.

As shown in Figure 4A, SCID/ICR nonstimulated (“resting”) primary islets displayed significantly higher basal nitric oxide levels compared with ICR islets (3.6 ± 0.26-fold). In addition, SCID/ICR islets released somewhat larger, although not statistically significant, amounts of IL-10 and IL-6 (1.7 ± 0.41-fold and 1.54 ± 0.29-fold, respectively). In contrast, IL-1Ra levels were similar in both groups. Cell injury appeared to be reduced in SCID/ICR islets as evidenced by a 2 ± 0.09-fold reduction in LDH release, compared with the ICR group. In addition, SCID/ICR islets had significantly higher accumulated insulin levels, compared with the ICR islets (2.73 ± 0.64-fold).

Stimulated islets were examined in the presence of IL-1β and IFN-γ (5 ng/ml each) (Fig. 4B). Supernatants were analyzed after 48 h of stimulation. As shown, no significant difference was observed in nitric oxide levels in the SCID/ICR and ICR mice: both groups reached the inducible levels shown in Figure 4A. However, stimulated SCID/ICR islets displayed significantly higher levels of IL-6 compared with ICR islets (8.14 ± 3.96-fold). It was also observed that following stimulation, no significant difference in the two groups was observed in IL-10 or in tumor necrosis factor-α (TNF-α) and IL-1α levels (not shown). Unexpectedly, SCID/ICR islets displayed significantly higher IL-1Ra levels compared with ICR islets (1.53 ± 0.13-fold). In the case of stimulated islets, LDH release was unchanged between the groups, but insulin release was significantly greater in the stimulated SCID/ICR group, compared with stimulated ICR islets (3.78 ± 1.25-fold).

Impact of Islet Resident Leukocytes on the Endocrine Function of SCID/ICR and ICR Islets

Islets contain resident leukocytes that are highly responsive to the microenvironment and are capable of communicating and amplifying inflammatory signals toward islet endocrine cells. Here we sought to characterize the profile of islet resident immune cells from ICR and SCID/ICR mice and also to assess their influence on the endocrine cells within their own hosting islets. The proportion of CD45⁺F4-80⁺ cells out of total islet cells was determined; as shown in Figure 5A, SCID/ICR islets contained more CD45⁺F4-80⁺ cells (2.42%) than did the ICR islets (0.75%). Surface expression of MHC class I, MHC class II, and CD40^HI is shown in Figure 5B. The expression of MHC class I was significantly higher in SCID/ICR mice compared with ICR islets (1.68 ± 0.05fold), while MHC class II levels were notably lower in SCID/ICR mice compared with ICR islets (1.81 ± 0.08fold). However, no difference was observed in CD40^HI expression levels between SCID/ICR and ICR mice. MHC class I expression on the surface of resident nonimmune islet cells demonstrated a similar trend to that of the immune cells (data not shown).

Figure 5.

Distinct MHC class I surface expression profile on resident pancreatic innate immune cells in ICR and SCID/ICR islets: implications for islet function. Freshly isolated islets were obtained from ICR or SCID/ICR mice (n = 5 from each strain, 100 per sample). Islets were enzymatically separated into a single-cell suspension, and resident macrophages were analyzed by flow cytometry. (A) F4-80⁺ surface expression. (B) MHC class I, MHC class II, and CD40 surface expression. (C) Resident leukocyte emigration: islets incubated for 72 h and then relocated into a new plate. CD45⁺ cell population. (D) Resident leukocyte emigration: insulin release levels. Mean ± SEM, *p < 0.05, **p < 0.01.

We next sought to study the effect of leukocyte depletion on islet function (1). Islets from SCID/ICR and ICR mice were rendered leukocyte poor by a 72-h culture in vitro, an established condition for facilitating immune cell emigration from islet structures. As shown in Figure 5C, the CD45⁺ cell population size was markedly reduced after white blood cell emigration. Islets were then hand-picked and transferred into new wells, and after 48 h, supernatants were examined for accumulated insulin levels. As shown in Figure 5D, SCID/ICR islets that had lost the content of their resident immune cells accumulated 2.38 ± 0.12-fold more insulin compared with ICR islets that had undergone the same manipulation. Both, incidentally, surpassed the insulin release levels obtained with intact immunocyte content (Fig. 4A).

Discussion

In the current study, we presented data that provide further supportive experimental evidence to the ecoimmunity theory. Central to the design of the study is the main difference between WT and SCID mice: the presence of an adaptive immune system composed of fully functioning T and B lymphocytes (10). We thus examined tissue cells derived from SCID mice, which presumably had never encountered an adaptive immune system, in a syngeneic environment that contains an intact immune system. We first tested the theory in a transplantation model in which islets that were derived from SCID/ICR and ICR mice were grafted into syngeneic ICR mice. According to our findings, while ICR-derived islets rapidly adjusted to their new host, islets derived from SCID/ICR mice were inferior in function throughout the 2-week follow-up period, as if they were “overwhelmed” by the presence of an intact immune system, a description that is, of course, only used for the sake of discussion.

Because antigen presentation plays a crucial role in graft survival, we sought to examine cell surface immune-related molecules on nonimmune cells (27, 34). We focused on MHC class I and CD40 expression on various cell types, as well as on MHC class II expression on antigen-presenting cells.

On the basis of the classical immunological view, costimulatory molecules bridge between innate and adaptive immune cells, and both MHC class I and II mediate antigen presentation (16). We observed that BM and spleen CD45⁺ cells derived from SCID/ICR mice expressed lower levels of MHC class II and CD40 than cells from ICR mice. This finding could be explained by the lack of B-cell compartments in SCID/ICR mice. In addition, we looked at the CD11b⁺ subpopulation gated out of CD45⁺ cells. Less significant changes were depicted in MHC class II and CD40 expression levels in CD11b⁺ cells derived from SCID/ICR mice than from cells derived from ICR mice. Surprisingly, all cell types derived from SCID/ICR mice, including renal nonlymphoid cells, expressed higher MHC class I surface levels than the same cell types derived from ICR mice. We further examined resident islet macrophages in SCID/ICR and ICR mice and observed that resident islet macrophages derived from SCID/ICR mice expressed significantly greater levels of MHC class I than those derived from ICR mice, yet no significant change was observed in the expression of MHC class II and CD40. Interestingly, SCID/ICR islets contained more immunocytes, although the number was not statistically significant, than did ICR islets. We also described the inflammatory release products of resting and stimulated islets in both types of ICR mice; nonstimulated SCID/ICR islets released greater levels of nitric oxide, IL-6, IL-10, and insulin compared with ICR-derived islets. While a similar trend was observed in inflammation-stimulated pancreatic islets, IL-6 and IL-1Ra levels were markedly increased in stimulated SCID/ICR islets compared with levels in ICR islets.

According to the classical immunological view, syngeneic cells derived from a host that lacks an adaptive immune arm would have no particular difficulty in adapting to a new syngeneic host. Unexpectedly, however, our data and the outcomes reported by Hauben et al. both show that transplanted pancreatic islets derived from SCID mice are inferior in adapting to a syngeneic host, compared with islets derived from WT mice. The selection of pancreatic islets as the grafted entity was intentional as these islets are highly sensitive to inflammatory conditions and cellular injury and readily allow the depiction of their compromised endocrine function by simple measurement of nonfasting circulating glucose (10). Relying on the ecoimmunity interpretation, we can speculate that islets derived from SCID/ICR mice lack a selective ontogenic pressure that is activated by coevolution during the early and continuous interaction with lymphocytes as well as lymphocyte-responsive innate and tissue cells. This pressure thus shaped the phenotype of islets in the ICR mice but did not operate in the SCID mice. This speculation suggests a supplementary role for tissue cells and offers a new view on predator–prey interactions in immunology: “prey” tissue cells not only shape “predator” immune cells during their coevolution, but they are also probably sculpted by the presence of an intact immune system.

We next addressed potential immune-related differences between islet cells derived from SCID/ICR and those derived from ICR mice. A gene expression array revealed that tissue cells derived from SCID/ICR mice express higher levels of genes related to immune pathways, primarily those that address regulation of both innate and adaptive immune cells. Thus, gene expression data coincide with flow cytometric data. The flow cytometric data allow dissection of immune cell subtypes. Although CD45⁺CD11b⁺ cells derived from BM and spleen cells of SCID/ICR mice were not significantly different from cells from ICR mice in the expression of surface MHC class II and CD40 (spleen data not shown), the surface expression of MHC class I was significantly elevated in SCID/ICR cells across all tested cell types compared with results from ICR mice. This finding supports the possibility that SCID/ICR tissue cells might be superior to tissue cells from ICR mice in activating lymphocytes as the SCID/ICR cells harbor an overpresentation of “self.”

According to an ecoimmunity interpretation, the expression levels found in WT animals represent the “desired” degree of expression of different transmembrane proteins, such as MHC class I, to which tissue cells converge during ontogeny to survive the presence of an intact immune system. In the absence of an intact immune system, an unopposed, exaggerated expression potential appears to materialize. In classical immunological view, this overexpression may improve an immune–tissue interaction. This interpretation may be correct in many contexts, and yet ecoimmunity emphasizes the idea that this overexpression may be harmful as it might skew tissue immune evasion and compromise tissue function.

The phenomenon of excessive surface expression of MHC class I has been reported in several cases of autoimmune type 1 diabetes, a pathology in which islet β-cell destruction is mediated primarily by CD8⁺ T lymphocytes (6, 31). Further studies should be conducted to fully verify whether antigen-presenting cells derived from SCID/ICR mice could activate ICR T and B lymphocytes in the context of islet transplantation and to determine whether this phenomenon is reproducible in other tissue cell types. In addition, as the common immunological view suggests, central antigen selection is exercised primarily in the lymphatic organs and not in peripheral tissues. Our data suggest that lymphocyte activity could also affect antigen presentation and self-regulation through a cross talk with peripheral tissue and innate cells. Additional studies should be conducted to examine this hypothesis.

Because the irregularity between SCID/ICR and ICR mice centers primarily on immune cells, we sought to further examine the function and immune reactivity of explanted, isolated pancreatic islets in vitro, with and without added cytokine stimulation. Stimulation consisted of the combination of IL-1β and IFN-γ, which impair islet function and exert a cytotoxic effect on islet β cells, representing an overt attack by an aggressive pan-immune system (16). As depicted in the present study, resting SCID/ICR islets released higher levels of insulin than did islets from ICR mice, as well as immune-mediating molecules such as nitric oxide, while demonstrating significantly less cellular injury than the ICR mice, according to an LDH leakage assay. In the ecoimmunity view, this observation suggests that SCID/ICR islets can afford free release of potentially immune-provoking factors such as nitric oxide, and cells that were shaped by continuous interaction with an immune system regulate such a necessary release of factors such as nitric oxide so as to minimize immune-mediated cell injury.

Therefore, nitric oxide production in both resting and stimulated SCID/ICR islets was relatively high, approximating levels that were achieved in ICR islets only upon stimulation. Similarly, IL-6 release was appropriately upregulated in stimulated ICR islets, yet was significantly more inducible in SCID/ICR islets. IL-6 may represent and facilitate an ongoing immune response, but also controls a variety of essential nonimmune cellular functions (29). Nonetheless, no differences in cellular injury were observed upon islet stimulation in both strains, while insulin release, which was fittingly suppressed by the inflammatory conditions, was markedly higher in SCID/ICR islets than in the ICR islets.

These in vitro culture studies suggest that SCID/ICR islets might be relatively more resistant to local inflammatory stimuli, somewhat in contrast to the in vivo findings. This difference is understandable considering the implication of excessive IL-6 in the whole animal. Moreover, unregulated release of IL-6 during the transplantation procedure might stimulate multiple host immune cells and markedly enhance the so-called process of “preying upon” tissue cells (14,29). With relevance to IL-6, we note that of the resident immune cells, which represent 3% of the total number of cells in the islet, 50% are B lymphocytes that possess a regulatory function that was not addressed in the present study due to scope considerations (21). This aspect is relevant to the outcomes of the present study; in SCID/ICR islets, the lack of resident lymphocytes might lead to an unbalanced immune response toward an inflammatory profile, with superiority of easily aggravated innate immune cells and functional implications that are more pronounced in vivo. Further studies should be conducted to determine islet resident immunocyte–tissue cell cross talk, as well as such cross talk within other peripheral tissues.

In examining resident immunocytes, we paid particular attention to islet-derived macrophages, which are highly inflammatory entities upon stimulation, yet also potentially islet-favorable cells under specific conditions (1). Levels of surface MHC class II and CD40 revealed an overall similarity between SCID and WT islets. Yet, further analysis showed that SCID/ICR resident islet macrophages expressed significantly greater levels of MHC class I compared with ICR islet cells. A possible explanation for this finding could be a compensation effect of macrophages “filling in” for lymphocytes within the resident immunocyte niche. Other experimental results in the present study that were collected from the spleen and BM compartments are consistent with islet data regarding the CD45⁺CD11b⁺ cell subtype; the surfaces of SCID innate cells display more MHC class I, relative to WT cells. Thus, our data suggest that the phenotype of SCID/ICR islets is not compatible with the new host environment and that the SCID/ICR cells have difficulty adapting to new conditions quickly enough. Slow-adapting cell types, such as islet β cells and nerve cells, might thus represent cells with somewhat heightened tendency for attracting autoimmune responses in cases where an organism experiences a significant immune flare. Moreover, SCID/ICR islets are able to activate lymphocytes with minimal additional stimuli. Lavin et al (17) and Hashimoto et al. (9) recently showed that resident tissue macrophages hold a specific transcription profile that matches local host tissue profile and that isolation of macrophages from each tissue reverses macrophage phenotypes toward an inflammatory profile. Indeed, our data suggest that tissue macrophages communicate with host tissue and that they might also be attentive to local lymphocyte signals within each given peripheral tissue. Further studies should be conducted to achieve a complete profile of SCID-derived tissue macrophage phenotypes, a phenomenon that holds immense relevance to multiple immune disorders.

The predator–prey relationship put forward in the ecoimmunity theory suggests the following analogy for our data: the ecological equivalent of an SCID islet introduced into an intact mouse would be the introduction of a herd of antelopes from a zoo, a predator-free environment, into a wild habitat. Although this herd carries the appropriate gene set for survival, it lacks the appropriate set of expressed behavioral patterns to avoid elimination by newly encountered local predators. In this analogy, some options for survival do exist. The herd could (a) quickly adapt to the new habitat or (b) lose all but a few “fit” members, those selected by genetic and phenotypic variations. We suggest that a similar phenomenon would occur in our experimental setting. Although we did not perform a complete screening of pancreatic islets that are derived from SCID/ICR mice, we were able to detect differences in costimulatory molecules in both immune and nonimmune tissue cells. High expression of an immune-mediating molecule, such as MHC class I, would assist the cells in evading natural killer cell responses but would also attract CD8⁺ T lymphocytes, a behavior demonstrated to occur in autoimmune pathologies, such as type 1 autoimmune diabetes (13).

According to the suggested principles of ecoimmunity, tissue cells (including innate immune cells) that are derived from SCID/ICR mice have actually never gone through selective pressure to adjust costimulatory molecule expression to the “desired” levels. However, from an ecological point of view, we suggest that lymphocyte maturation could also depend on the tissue's immune-related profile, suggesting a coevolution between lymphocytes and tissue cells. If this is the case, therapies directed at curing autoimmune conditions should be drastically extended to incorporate tissue adaptation, thus minimizing cell injury and advancing wound-healing pathways, rather than executing an effective ablation of immune cells by immunosuppression. Indeed, a compromised immune system is known to increase the risk of cancerous outbreaks and, according to ecoimmunity theory, would cause the tissue to gradually turn more immunogenic as its “predator” has left the equation. One may further speculate, based on the ecoimmunity theory, that the emergence of a spontaneous autoimmune condition in an individual might be the outcome of a successful combat against a subclinical cancerous growth or a viral infection, accompanied by the inadequate cell adaptation of particular nonimmune tissues, including the poorly regenerating islet β cells and nerve cells. Further studies should be performed to clarify this intriguing therapeutically relevant possibility.

Regulation of cellular injury was also examined in the context of ecoimmunity. Reduced exposure to damage could enhance tissue function and is usually accompanied by inflammatory and immune regulation (18,26). Surprisingly, in vitro tissue cells derived from SCID/ICR mice functioned similar to, or better than, ICR cells in both the presence and absence of an added inflammatory stimulus. In this regard, an apparent inflammatory paradox surfaced: IL-6, usually an inflammatory marker, was released at high levels from pancreatic islets, yet was associated with neither impaired islet cell survival nor reduced islet endocrine function. It is possible that islet-derived IL-6, in particular, serves as a tissue factor, as shown to occur in other tissue types (29). The absence of B lymphocytes is also relevant when cells are damaged. The findings suggest that the initial response to cellular damage is mediated by innate cells, tissue cells, and also local lymphocytes. Thus, the mere absence of B lymphocytes in SCID islets may result in the immune cells functioning better when there has been cellular injury than when the B lymphocytes are present.

Taken together, the theory of ecoimmunity and the present experimental work both imply that the principles underlying the ability of the immune system to coexist with tissues, without harming peripheral nonimmune cells, are to some extent less complex than once perceived. Our findings suggest that the complex view of self-tolerance originates from our assumed view of immune cells and tissue cells as separately evolving entities and since at the level of the host, the immune system helps tissue cells survive pathogens. However, as in most ecological systems, this benefit can be the product of a predator–prey interaction in which the tissue is always regarded as prey. In the absence of an overt pathogenic event, the prey (tissue) falls below the threshold of predation by immune activation. The array of molecular mechanisms involved in immune cells preying on tissue cells and in tissue cells evading immune cells is presumably vast, as in the analogous large ecosystems. The basic principles, however, are probably universal.

Footnotes

Acknowledgments

We are grateful to the Israeli Science Foundation (ISF) for facilitating this work with a Bikura (First) Grant No. 1601/10. U.N. was supported by the Colton Family scholarship for young scientists. U.N. originated and coordinated this study. U.N., E.C.L., and S.E. designed the experiments. D.E.O., B.M.B., R.S., and N.K. performed the experiments in the laboratory of E.C.L. at Ben-Gurion University of the Negev. R.B.-H., I.S., and S.E. performed the bioinformatical analysis. U.N., D.E.O., B.M.B., E.C.L., and S.E. wrote the article. The authors declare no conflicts of interest.

References

Arnush

; Scarim

A. L.

; Heitmeier

M. R.

; Kelly

C. B.

; Corbett

J. A.

Potential role of resident islet macrophage activation in the initiation of autoimmune diabetes. J. Immunol. 160(6): 2684–2691; 1998.

Azzi

; Geara

A. S.

; El-Sayegh

; Abdi

Immunological aspects of pancreatic islet cell transplantation. Expert Rev. Clin. Immunol. 6(1): 111–124; 2010.

Barshes

N. R.

; Lee

T. C.

; Goodpastor

S. E.

; Balkrishnan

; Schock

A. P.

; Mote

; Brunicardi

F. C.

; Alejandro

; Ricordi

; Goss

J. A.

Transaminitis after pancreatic islet transplantation. J. Am. Coll. Surg. 200(3): 353–361; 2005.

Brereton

M. F.

; Iberl

; Shimomura

; Zhang

; Adriaenssens

A. E.

; Proks

; Spiliotis II Dace

; Mattis

K. K.

; Ramracheya

; Gribble

F. M.

; Reimann

; Clark

; Rorsman

; Ashcroft

F. M.

Reversible changes in pancreatic islet structure and function produced by elevated blood glucose. Nat. Comm. 5: 4639; 2014.

Brissova

; Fowler

M. J.

; Nicholson

W. E.

; Chu

; Hirshberg

; Harlan

D. M.

; Powers

A. C.

Assessment of human pancreatic islet architecture and composition by laser scanning confocal microscopy. J. Histochem. Cytochem. 53(9): 1087–1097; 2005.

Coppieters

K. T.

; Dotta

; Amirian

; Campbell

P. D.

; Kay

T. W.

; Atkinson

M. A.

; Roep

B. O.

; von Herrath

M. G.

Demonstration of islet-autoreactive CD8 T cells in insulitic lesions from recent onset and long-term type 1 diabetes patients. J. Exp. Med. 209(1): 51–60; 2012.

Dunn

G. P.

; Old

L. J.

; Schreiber

R. D.

The three Es of cancer immunoediting. Annu. Rev. Immunol. 22: 329–360; 2004.

Emamaullee

J. A.

; Shapiro

A. M.

Factors influencing the loss of beta-cell mass in islet transplantation. Cell Transplant. 16(1): 1–8; 2007.

Hashimoto

; Chow

; Noizat

; Teo

; Beasley

M. B.

; Leboeuf

; Becker

C. D.

; See

; Price

; Lucas

; Greter

; Mortha

,; Boyer

S. W.

; Forsberg

E. C.

; Tanaka

; van Rooijen

; García-Sastre

; Stanley

E. R.

; Ginhoux

; Frenette

P. S.

; Merad

Tissue-resident macrophages self-maintain locally throughout adult life with minimal contribution from circulating monocytes. Immunity 38(4): 792–804; 2013.

10.

Hauben

; Roncarolo

M. G.

; Draghici

; Nevo

The role of tissue adaptation and graft size in immune tolerance. Transpl. Immunol. 18(2): 122–125; 2007.

11.

Hogquist

K. A.

; Baldwin

T. A.

; Jameson

S. C.

Central tolerance: Learning self-control in the thymus. Nat. Rev. Immunol. 5(10): 772–782; 2005.

12.

Hornick

Direct and indirect allorecognition. Methods Mol. Biol. 333: 145–156; 2006.

13.

In't Veld

Insulitis in human type 1 diabetes: A comparison between patients and animal models. Semin. Immunopathol. 36(5): 569–579; 2014.

14.

Itoh

; Nitta

; Nishinakamura

; Kojima

; Mera

; Ono

; Kodama

; Yasunami

HMGB1-mediated early loss of transplanted islets is prevented by anti-IL-6R antibody in mice. Pancreas 44(1): 166–171; 2015.

15.

Janeway

C.A.

Jr. , Medzhitov

Innate immune recognition. Annu. Rev. Immunol. 20: 197–216; 2002.

16.

Kalis

; Kumar

; Janciauskiene

; Salehi

; Cilio

C. M.

Alpha 1-antitrypsin enhances insulin secretion and prevents cytokine-mediated apoptosis in pancreatic beta-cells. Islets 2(3): 185–189; 2010.

17.

Lavin

; Winter

; Blecher-Gonen

; David

; Keren-Shaul

; Merad

; Jung

; Amit

Tissue-resident macrophage enhancer landscapes are shaped by the local microenvironment. Cell 159(6): 1312–1326; 2014.

18.

Lewis

E. C.

; Dinarello

C. A.

Responses of IL-18- and IL-18 receptor-deficient pancreatic islets with convergence of positive and negative signals for the IL-18 receptor. Proc. Natl. Acad. Sci. USA 103(45): 16852–16857; 2006.

19.

Lewis

E. C.

; Mizrahi

; Toledano

; Defelice

; Wright

J. L.

; Churg

; Shapiro

; Dinarello

C. A.

Alpha1-antitrypsin monotherapy induces immune tolerance during islet allograft transplantation in mice. Proc. Natl. Acad. Sci. USA 105(42): 16236–16241; 2008.

20.

Lewis

E. C.

; Shapiro

; Bowers

O. J.

; Dinarello

C. A.

Alpha1-antitrypsin monotherapy prolongs islet allograft survival in mice. Proc. Natl. Acad. Sci. USA 102(34): 12153–12158; 2005.

21.

Marino

; Walters

S. N.

; Villanueva

J. E.

; Richards

J. L.

; Mackay

C. R.

; Grey

S. T.

BAFF regulates activation of self-reactive T cells through B-cell dependent mechanisms and mediates protection in NOD mice. Eur. J. Immunol. 44(4): 983–993; 2014.

22.

Moore

D. J.

; Markmann

J. F.

; Deng

Avenues for immunomodulation and graft protection by gene therapy in transplantation. Transpl. Int. 19(6): 435–445; 2006.

23.

Nevo

; Golding

; Neumann

A. U.

; Schwartz

; Akselrod

Autoimmunity as an immune defense against degenerative processes: A primary mathematical model illustrating the bright side of autoimmunity. J. Theor. Biol. 227(4): 583–592; 2004.

24.

Nevo

; Hauben

Ecoimmunity: Immune tolerance by symmetric co-evolution. Evol. Dev. 9(6): 632–642; 2007.

25.

Nevo

; Kipnis

; Golding

; Shaked

; Neumann

; Akselrod

; Schwartz

Autoimmunity as a special case of immunity: Removing threats from within. Trends Mol. Med. 9(3): 88–93; 2003.

26.

Ochayon

D. E.

; Mizrahi

; Shahaf

; Baranovski

B. M.

; Lewis

E. C.

Human alpha1-antitrypsin binds to heat-shock protein gp96 and protects from endogenous gp96-mediated injury in vivo. Front. Immunol. 4: 320; 2013.

27.

Rescigno

Functional specialization of antigen presenting cells in the gastrointestinal tract. Curr. Opin. Immunol. 22(1): 131–136; 2010.

28.

Reth

; Wienands

; Schamel

W. W.

An unsolved problem of the clonal selection theory and the model of an oligomeric B-cell antigen receptor. Immunol. Rev. 176: 10–18; 2000.

29.

Rincon

Interleukin-6: From an inflammatory marker to a target for inflammatory diseases. Trends Immunol. 33(11): 571–577; 2012.

30.

Sakaguchi

; Sakaguchi

; Asano

; Itoh

; Toda

Immunologic self-tolerance maintained by activated T cells expressing IL-2 receptor alpha-chains (CD25). Breakdown of a single mechanism of self-tolerance causes various autoimmune diseases. J. Immunol. 155(3): 1151–1164; 1995.

31.

Sarikonda

; Pettus

; Phatak

; Sachithanantham

; Miller

J. F.

; Wesley

J. D.

; Cadag

; Chae

; Ganesan

; Mallios

; Edelman

; Peters

; von Herrath

CD8 T-cell reactivity to islet antigens is unique to type 1 while CD4 T-cell reactivity exists in both type 1 and type 2 diabetes. J. Autoimmun. 50: 77–82; 2014.

32.

Tauber

The biological notion of self and non-self, Summer 2012 ed: Stanford University; 2012.

33.

Van Parijs

; Abbas

A. K.

Homeostasis and self-tolerance in the immune system: Turning lymphocytes off. Science 280(5361): 243–248; 1998.

34.

Williams

N. A.

; Wilson

A. D.

; Bailey

; Bland

P. W.

; Stokes

C. R.

Primary antigen-specific T-cell proliferative responses following presentation of soluble protein antigen by cells from the murine small intestine. Immunology 75(4): 608–613; 1992.