The Importance of ‘Mechnikov’s Thorn’ for an Improved Understanding of 21st Century Medicine and Immunology: A View from the Eye


Professor D. McGonagle, The Leeds Institute of Molecular Medicine, JIF Building, St. James’s University Hospital, The University of Leeds, Leeds LS9 7TF, UK. E-mail:


In 1908, Ehrlich and Mechnikov shared the Nobel Prize in Medicine for their independent studies that set the scene for the modern understanding of innate and adaptive immunity. However, 20th century immunology thinking was dominated by aberrant adaptive immunity but this never adequately explained the full spectrum of inflammatory disease. This article draws on medical observations, from where immunology originated, and uses the example of the eye to illustrate how the integration of medicine and immunology leads to an improved understanding of inflammation against self. The spectrum of ocular inflammation can be viewed as either predominantly adaptive immune mediated (mostly the realm of immunology), or predominantly due to ocular tissues factors that lead to regional innate immune activation (the realm of medicine), or a variable interaction between the two. Just as the thorns that Mechnikov inserted into molluscs lead to localized innate immune activation; ocular inflammation can likewise be driven by non-immune factors that include tissue degeneration or microdamage. The present article emphasizes the importance of such factors in the initiation or phenotypic expression of ocular immunopathology allowing different immunological dogmas including self–non-self discrimination, immunological tolerance and immunoprivilege to be viewed in a different light. This scheme also leads to an appreciation of how the innate immune system may be the sole perpetuator of some ocular immunopathologies. We propose that this integrated view of medicine and immunology is crucial for understanding immunology from a translational angle and has implications far beyond ocular disease.

‘The eyes are the windows to the soul’

The origins of this phrase from antiquity are obscure. The original recognition of autoimmunity was also in antiquity and is much clearer, being appreciated in the Hippocratic era with the description of an eye disease now termed sympathetic ophthalmia [1]. Another historical immunology milestone was the recognition in the 19th century that allogeneic tissue implanted into the eye was not rejected, which was the first description of immunoprivilege [2]. During the 20th century, the immunological basis for tissue inflammation was delineated. In the 1950s, the autoimmunity paradigm emerged and this broadly equated with the presence of potentially pathogenic autoantibodies and/or the ability to adoptively transfer disease with lymphocytes [3]. It is surprising that this paradigm was the only recognized theoretical means of conceptualizing self-directed tissue inflammation since the Nobel Prize in Physiology or Medicine shared by Ehrlich and Mechnikov in 1908, equally recognized the importance of innate immunity (Box 1) [4, 5]. However, Mechnikov’s late 19th century work into innate immunity was largely ignored [6]. He noted that thorns placed in the flesh of molluscs were associated with a tissue reaction characterized by the accumulation of ‘haemocytes’ which were eventually recognized as macrophages, which are key players in innate immunity.

Table Box 1.   Brief biography of Mechnikov and Ehrlich [4].
Ilya Mechnikov (1845–1916)Paul Ehrlich (1854–1915)
Share in 1908 Nobel Prize in Physiology or Medicine for the discovery of cellular immunity. This formed the foundation for an understanding of innate immunity.Share in 1908 Nobel Prize in Physiology or Medicine for the discovery of humoral immunity.
Had no medical training.Medically trained.
Worked in several Research Institutes including Odessa and Munich, before moving to Paris to work at The Pasteur Institute following talks with Louis Pasteur himself.Most famous student of Robert Koch. Worked in Berlin at Koch’s facility before moving to his own institute in Frankfurt.
Attempted suicide twice but outlook on life changed on the discovery of phagocytosis.Suffered from TB as young man but made a good recovery.
Lost first wife to Tuberculosis (TB) – a now somewhat ironic misfortune given that innate immunity is central to the control of TB. This is exemplified by the increased rate of TB following innate immune cytokine pathway blockade with tumour necrosis factor antagonists.Tremendous research activity in a number of areas including the use of histological staining techniques. The modern Ziehl and Neelson TB staining method adapted from his methods. Major contribution to microbial chemotherapy development.
Work not widely acclaimed during lifetime and met with some hostility. The ‘cellularist’ theory of immunology lost out to ‘humoralist’ one for many years.Set the scene for the recognition of autoimmunity but erroneously thought that such a ‘horror autotoxicus’ scenario was not compatible with survival.
It is now evident that these two great immunologist’s theorists defined different parts of the spectrum of inflammation against self but in 1908 the theories were considered mutually exclusive.

Although immunologists now fully appreciate the importance of Mechnikov’s work, the discipline still places a very strong emphasis on cells involved in aberrant adaptive immune responses. We believe that physicians and scientists alike have been over reliant on this and have lacked an appreciation of the full importance of Medicine for understanding tissue inflammation. To account for the emerging understanding of the genetics of inflammatory disorders, we proposed an immunological disease continuum (IDC) of innate and adaptive immunity as a way of understanding the full spectrum of self-directed non-infectious inflammation [7, 8]. Whilst disease localization in ‘classical autoimmunity’ can be related to autoantibodies that predate disease, it often appears that non-immune or tissue-specific factors may be pivotal for innate immune-mediated inflammation [7, 8]. Such non-immune factors, or tissue-specific factors, include tissue microdamage, mechanical or hypoxic stress, the presence of urate or pyrophosphate crystals and abnormal interactions with tissue flora (Fig. 1) [7, 9]. In a sense, these factors are the equivalent of a ‘Mechnikov’s thorn’ because, essentially, they perturb tissue homeostasis in a manner akin to an exogenous insult, such as a thorn. This perspective uses the example of eye inflammation to show how an understanding of tissue physiology and pathophysiological responses can be used to better conceptualize aberrant immunological responses. Inflammatory eye disease (IED) is used to illustrate the case for a number of reasons as set out in Table 1, but an identical scenario could be derived by considering diseases of other systems.

Figure 1.

 A combined innate and adaptive immune system view of eye disease. When viewing the spectrum of ocular inflammation, it is imperative to keep the focus on ocular specific factors as the key driving force in some diseases. For the classical autoimmune diseases, the primary and secondary lymphoid organs may play the major role. In many settings such as the MHC class I associated diseases and ocular immunoprivilege there may be an interaction between both of these. For innate immune-mediated inflammation, the events taking place in the eye determine disease expression. This knowledge is important for therapy development.

Table 1.   The relevance of eye inflammation to medicine and immunology.
Medicine perspectives
 Strong association with systemic diseases
 Subtle disease may manifest clinically due to eye complexity
 Direct visualization of pathology due to ocular anatomy
 Devastating clinical consequences of visual loss compared with  damage to other systems
 A direct extension of the brain
Immunological perspectives
 Eye inflammation associated with polygenic systemic diseases in up  to 50% of cases so an understanding of eye disease could have far  reaching consequences
 Arguably the first autoimmune diseases, sympathetic ophthalmia,  ocular based
 The case of ocular immunoprivilege where allogeneic tissue not  rejected
 Genetic association with varied monogenic and polygenic  inflammatory diseases

Traditional view of Eye Autoimmunity and Emerging Concepts

Based on the previously described IDC concept, which represented a generalized classification of inflammation against self, a more refined classification specific for eye disease is set out (Fig. 2). This classification scheme indicates that events taking place either in the lymphoid tissues or in the eye itself or an interaction between both may be the major determinants of eye disease expression (Fig. 1) [8]. This idea of lymphoid related tissue functional changes primarily underscoring autoimmunity and target tissue aberration or the equivalent of ‘Mechnikov’s thorn’ primarily underscoring innate immune-mediated disease, forms the basis for an evaluation of all ocular pathology without making the assumption that failure of self–non-self discrimination (SNS), or the inability to distinguish between foreign and self-antigens, underlies the immunopathology. The common ocular immunopathologies that epitomize different disease categories in this scheme are briefly discussed before their relevance for disease classification is considered (Figs. 2 and 3).

Figure 2.

 Proposed mechanistic classification of eye disease with examples. This table depicts eye disease classification with some examples. It is based on the immunological disease continuum concept, which uses underlying unifying immunopathogenetic mechanism for classification rather than classifying disease by site of ocular involvement. The classical autoimmune diseases, including Sjogrens syndrome and Wegener’s granulomatosis, with autoantibody and MHC class II associations are designated at the adaptive immune end of spectrum. A characteristic example is the autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) syndrome that results from mutations in the autoimmune regulator protein (AIRE) and is associated with keratitis and iridocyclitis. The innate immune-mediated diseases are placed at the opposite end of the spectrum. The latter are best exemplified by mutations in proteins such as NOD2 in Blau’s syndrome and cryopyrin in NOMID. As outlined in the text, the MHC class I associated diseases show strong clinical overlaps with predominant innate immune-mediated diseases and are intermediate diseases. From this classification scheme, it is possible to develop a better translational understanding of ocular disease.

Figure 3.

 An innate and adaptive immunological glossary. The history of immunology has been mostly based on the autoimmune paradigm. There is some conflict amongst immunologists about certain dogmas and theories, as outlined in the text. By maintaining a focus on whether a human disease is predominantly innate or adaptive immune mediated could lead to an improved understanding of immunopathogenic mechanisms in vivo. It is clear that the classical immunology terminology is highly relevant for adaptive immune diseases but may not be equally relevant for the diseases, formerly regarded as autoimmune, e.g. Crohn’s disease, which is now known to have a major innate immune component. It is proposed that immunology theory and terminology which was developed to explain autoimmunity should be viewed along a continuum and that some paradigms, in fact, are more relevant for understanding innate immune related diseases.

Several polygenic eye diseases have autoantibody and MHC class II associations and represent classically recognized autoimmune diseases (Fig. 2). This primacy of adaptive immunity is well illustrated in the case of Graves’ disease and associated ophthalmopathy where transplacental passage of maternal autoantibodies, may lead to transient neonatal thyroid and eye disease that subsides with the disappearance of the causative antibodies [10]. This illustrates how an otherwise normal neonatal thyroid or eye tissue can be subject to autoimmunity. Several other autoimmune diseases including Sjogrens syndrome, systemic lupus erythematosus (SLE), the antiphospholipid syndrome, rheumatoid arthritis, dermatomyositis and the antineutrophil cytoplasmic antibodies (ANCA) associated vasculitides can involve the eye. Whilst the exact pathophysiologic role of autoantibodies in these systemic diseases remains unclear, it is evident that autoantibodies may predate discernible target tissue inflammation by several years [11–13]. Some of these autoantibodies including ANCAs may play direct pathogenic roles by mechanisms that are not fully defined [14].

However, even within these ‘classic’ autoimmune diseases, innate immune cells and what has been dubbed as ‘anatomical constraints’ may still play a key role in the eventual immunopathology and phenotypic expression of disease. For example, rheumatoid arthritis (RA) has been viewed as exclusively autoimmune in nature given its HLA-DRB1 ‘shared epitope’ and autoantibodies including rheumatoid factor and anti-CCP autoantibodies [12]. Several shared epitope and autoantibody associated RA ocular pathologies have been described [15–21]. A common pathogenic mechanism of RA is believed to be related to immune complex formation, which is thought to activate macrophages and other innate-adaptive immune bridging cells via Fc receptors, leading to ocular-specific inflammation, in a manner similar to that proposed for putative joint pathology mechanisms [22, 23]. In other ocular autoimmunity settings, it is evident that tissue-specific factors may also modulate the phenotypic expression of diseases. This is illustrated for thyroid eye disease where ocular muscle inflammation may be non-homogeneous, despite the diseased muscles having the same antigenic structure (Fig. 4). Whilst, autoimmune mechanisms may dominate the clinical picture, it is therefore clear that factors intrinsic to the target organs have a major, but secondary, impact on the phenotype.

Figure 4.

 Tissue-specific factors modulate ocular autoimmunity. This is a high-resolution microscopy coil MRI (HR-MRI) from a patient with thyroid eye disease where extraocular muscle inflammation is characteristic. This HR-MR image shows patchy involvement of the inferior and lateral recti muscles (arrows). The superior and medial recti are also affected (enlarged) (asterisks). This illustrates that even in a classically recognized autoimmune eye disease, local anatomical variations or unknown factors influence disease expression since inflammation is variable within the muscles.

Polygenic eye disease with putative innate immune mechanisms

Both Crohn’s disease and ulcerative colitis (UC) were formerly designated as autoimmune [7]. From the ocular perspective, Crohn’s disease is associated with anterior uveitis and acute iridocyclitis that is evident in up to 13% of patients [24, 25]. Acute iritis and iridocyclitis, typically synchronized with exacerbations of colitis, and associated with hypopyon formation (visible pus in anterior chamber) are also recognized in UC [26, 27]. At the genetic level, Crohn’s disease has been associated with mutations in the nucleotide-binding oligomerization domain-2 (NOD2) bacterial peptidoglycan sensing protein and in the autophagy-inducing IRGM gene both of which signify innate immune system involvement [28, 29]. In the intestine, the NOD2 protein is expressed in the intestinal Paneth cells and on monocytic lineage cells, which are abundant in the terminal ileum that is typically involved [7]. Likewise, UC is increasingly recognized as having a major innate immune component related to perturbations in the gut mucosal barrier system [26].

Crohn’s and UC related ocular inflammation shows striking clinical overlaps with ankylosing spondylitis (AS), psoriasis and reactive arthritis and all have a propensity for anterior uveitis (Table 2) [8]. We have shown that the arthropathy associated with these disorders localizes to joint insertions, which are sites of high mechanical stressing, microdamage, microfracture and tissue repair responses [30, 31]. However, given the MHC class I associations of some of these disorders, and especially the HLA-B27 AS association, they appear to represent true intermediates between adaptive immune perturbation and tissue specific dysregulation of target tissue sites (Table 2).

Table 2.   Link between ocular immunity, tissue microtrauma and immunoprivilege.
DiseaseMHC associationType of Tissue Specific Response
  1. These ocular diseases show a tendency for involvement of the anterior uvea where the ‘moving parts’ of the globe are found. With respect to their systemic manifestations, especially skin and joint associations, all show disease localization to sites of high mechanical stressing or tissue microdamage. The two most clinically recognized ones are the Koebner and pathergy responses in psoriasis and Behcet’s respectively [33, 39]. Also, the tendon and ligament insertions which are often involved in joint disease are sites of microdamage.

  2. The anterior chamber to where these diseases localize in man is also an immunoprivileged site where allogeneic tissue is not rejected. This phenomenon is called anterior chamber associated immunodeviation (ACAID). It is proposed that ocular immunoprivilege is an adaptation to minimize ocular inflammation in the ‘moving parts’ of the eye. As the MHC class I associated diseases and allied disorders show this is not fail proof. The recognition of the physiological requirements of the eye to minimize inflammation in the moving parts offers a novel perspective on ACAID.

PsoriasisHLA-Cw6 [32]Koebner response [33]
Ankylosing SpondylitisHLA-B27 [30]Microdamage at ligament and tendon attachment sites
Psoriatic arthritisControversial [30]Microdamage at ligament and tendon attachment sites
Crohn’s disease /ulcerative colitisNot known [24]Microdamage of moving parts in joints
Behcet’s diseaseHLA-B51 [36, 37]Pathergy response [39]

The same tissue microdamage principle holds true for the psoriasis related spectrum of disease where an association with uveitis and a strong MHC class I association, namely HLA-Cw6 exists [32]. Psoriasis is associated with skin trauma, termed the Koebner response (Table 2) [30, 31, 33]. Indeed, microdamage to the skin in psoriasis is specifically associated with the upregulation of the LL-37 cathelicidin protein, which exhibits a broad range of activities in innate immunity [31, 34, 35]. From the immunological perspective, Behcet’s disease appears to share a similar immunopathogenetic mechanism since 60% of Behcet’s patients are HLA-B51 positive [36, 37]. Heterozygosity for the familial Mediterranean fever (FMF) mutation is commoner in Behcet’s sufferers and the condition also overlaps strongly with innate immune diseases like Crohn’s and UC, where in, the latter, the clinical phenotype may be indistinguishable [38]. Like the aforementioned MHC class I related diseases, Behcet’s is also associated with tissue-specific injury, termed the pathergy response, where following venopuncture, pus-filled skin lesions, may develop [39].

Also, many patients with anterior uveitis do not have a systemic disease association but, nevertheless, 30% of such cases are HLA-B27 positive [40]. Like the joint, the anterior uvea is a dynamically functioning structure with the iris and ciliary body moving to change the pupil size depending on light exposure and to allow for near-distant accommodation respectively. Unlike the classical autoimmune ocular disease, the onset of HLA-B27 related disease is usually rapid and attacks are self-limiting [39]. Furthermore, successive attacks tend to move from side to side suggesting that ocular factors are pivotal in disease expression [40]. Whether or not this is due to microtrauma of the ‘moving parts’ of the iris and ciliary body in a manner suggested for HLA-B27-linked joint disease has not yet been established, but the concept of tissue-specific factors activating innate immunity in the anterior chamber offers a new perspective on the intermittent pattern of inflammation noted at this site. A second factor in the HLA-B27 related diseases that favours innate immune activation is bacterial access to the circulation, usually due to either a breach or the gastrointestinal or genitourinary mucosal barriers [41]. The role of innate immunity, in the pathogenesis of anterior uveitis, is further supported by studies where the systemic administration of lipopolysaccharide alone leads to experimental anterior uveitis [42]. What is intriguing from the clinical perspective is that all of these related diseases, both innate immune and MHC class I associated, have a propensity for the neutrophilic inflammation in the anterior chamber of the eye, termed hypopyon (Fig. 5). Therefore, a mechanistic classification of eye disease can be directly linked to the type of immune perturbation and can be recognised at the bedside (Figs. 1 and 5).

Figure 5.

 Bedside recognition of a major innate immune component in eye disease. This drawing illustrates how ocular diseases with a major innate immune contribution can be discerned on clinical grounds alone, which integrates with a mechanistic immunology disease classification. Hypopyon is the medical term for visible pus (neutrophils, an innate immune cell) in the anterior chamber of the eye. Characteristic examples of systemic diseases with hypopyon uveitis are HLA-B27 related arthropathy associated uveitis, Crohn’s disease, ulcerative colitis, psoriasis and Behcet’s disease. As depicted in the classification scheme in Fig. 2, all of these sit near the innate immune end, or intermediate location of the classification spectrum. In keeping with the key role for innate immunity, all of these diseases may show striking responses to anti-TNF therapy, TNF being a key cytokine in innate immune responses. Conceptually, this neutrophilic inflammation may reflect rapid innate immune activation in response to local tissue-specific factors. Such factors may include access of bacterial molecules, which may occur in UC, Crohn’s and in AS, due to the abnormal gut wall permeability. Extending the analogy with Mechnikov’s thorn, it is also noteworthy that penetrating injuries to the eye also lead to hypopyon.

Immunological Implications of Ocular Disease Classification

When ocular inflammatory disease is classified using a combined medical and immunology perspective, rather than the autoimmunity concept alone, the immunology of eye can be reassessed. For example, immunologists have suggested that their science was that of distinguishing between self–non-self (SNS). However, Matzinger argued that the immune system was, not so much geared towards SNS discrimination, but was more concerned with responding to danger signals [43]. According to this and our work in the HLA-B27 associated arthropathies, it is ultimately tissue specificity that determines pathology, since dysregulated tissue homeostasis results in danger signals that trigger innate immune activation [41, 43]. It has been argued that danger is essential for adaptive immune responses, and by extension for autoimmunity [43]. However, based on the present scheme, SNS and danger are not competing theories but rather belong to the duality concept –with SNS suiting the paradigm for traditional autoimmune diseases and danger the paradigm for innate immune related disease [7]. The idea that danger is not necessary for autoimmunity has been elegantly shown in AIRE-deficient mice where autoimmunity still develops when danger is removed by breeding mice in a sterile environment or where Toll receptor sensing danger pathways are knocked out [44].

However, a mechanistic classification of ocular inflammation has relevance far beyond SNS and danger. We propose that immunology definitions and terminology should be placed along an ‘innate and adaptive immunological glossary’ with different terminology being applicable for the various categories of inflammatory disease (Figs. 2 and 3). As an example of this, the present scheme suggests distinct, rather than overlapping mechanisms for the avoidance of self-reactivity. For example, immunological tolerance has been viewed as either central or peripheral with both mechanisms working in concert to prevent auto-reactivity. Whilst, this is undoubtedly the case in health, the present perspective strongly suggests that, in autoimmune disease, tolerance failure is either predominantly due to primary and secondary lymphoid organ dysfunction or perturbations within the target tissues themselves. The two major mechanisms of immunological tolerance failure according to this scheme are:

  • 1Aberrant immunological responses in thymus, marrow or lymphoid tissues. This is clinically recognizable by the MHC class II and/or autoantibody associations, the latter of which may predate clinical disease or target tissue damage [45]. In this setting, the pathology is principally due to failure to regulate T and B cells [7]. The tissue immunopathology may ultimately be related to dysregulation of apoptosis of lymphoid cells within these tissues.
  • 2Aberrant tissue responses leading to secondary adaptive immunity at sites of microdamage or perturbed tissue homeostasis due to other regional factors. This is typically seen in the MHC class I related diseases. According to this scheme, following dendritic cell priming in target tissue and migration to regional nodes, pathogenic CD8 T cell clones home back the sites of microdamage, gaining access at sites of altered vascularity, thus aggravating inflammation [30, 31]. Therefore, a combination of dendritic cell co-stimulation and activated vascular endothelium at the disease prone sites are the major players in the breakage of tolerance and disease development. Tissue microdamage including necrosis may be the key events in the target tissues that influence immune activation.

An appreciation of two major types of tolerance failure that underscores autoimmunity is not just of academic relevance. In the latter category of autoimmune disease, the innate immune pathway cytokine blockade is likely to be associated with good clinical response. Proof of principle for this comes from several MHC class I associated diseases where TNF blockade has shown good efficacy, this cytokine being a key player in innate immunity [46].

Tissue Destruction Without Failure of Tolerance

However, it must be emphasized that the worst case scenario in ocular inflammation, namely blindness, can occur without evidence for failure of either central or peripheral immunological tolerance, but where disease is solely the result of apparent chronic innate immune activation. Whilst, innate immune-mediated organ destruction is well recognized in septicaemic shock, its full appreciation outside the setting of sepsis has considerable implications for therapy development [47]. Blau’s syndrome and neonatal-onset multisystem inflammatory disease (NOMID) attest to the fact that severe end-organ damage may occur without discernible adaptive immune mechanisms, but these are very rare [48]. However, age related macular degeneration (ARMD) is a very common cause of blindness that was historically viewed as a purely degenerative disorder [49]. In this condition, retinal pigment epithelium (RPE) debris accumulates resulting in the local activation of macrophages and associated angiogenesis, which eventually leads to choriocapillary injury and blindness [50]. Recently ARMD has been linked to innate immunity at the genetic level, namely mutations in the complement factor H protein, which prevents the pathways unrestrained activation [49]. The indirect targeting of innate immune-mediated inflammation via angiogenesis blockade offers a novel therapeutic approach in ARMD [51, 52].

Taken together these findings indicate that different types of immunological tolerance govern the propensity for different types of disease and that, in some cases, adaptive immune system immunological tolerogenic mechanisms may have little or no role. Therefore, immunologists must consider that, sometimes in vivo, tissue damage may be independent of the classically described immune hypersensitivity reactions. Indeed Mechnikov was describing the forerunner of these ‘innate autoimmunity’ reactions a quarter of a century before the recognition of immune hypersensitivity.

A Novel Perspective on Anterior Chamber Associated Immune Deviation

It has long been recognized that some tissues including the eye and the brain are relatively protected from inflammatory responses. This is termed immunoprivilege. A better understanding of these phenomena could have major implications for the development of new therapies for the prevention of organ transplant rejection – a major ongoing 21st century challenge to medicine. Anterior chamber associated immune deviation (ACAID), whereby allogenic tissue implanted into the front compartment of the eye is not rejected, was recognized nearly 130 years ago and is a critical component of immune privilege of the normal eye [53]. Immunologists have recognized that the significance of ACAID extended far beyond the fortuitous lack of rejection of allogeneic tissue, but its physiological role has remained enigmatic.

The immunological basis and clinical significance of ACAID, or indeed its failure in MHC class I associated diseases including AS, psoriasis and Behcet’s related anterior uveitis, can be conceptualized by assessing tissue-specific factors in the MHC class I associated diseases (Table 2). As previously outlined, the systemic component of these MHC class I related disorders are associated with tissue microdamage (Table 2). However, unlike the joints and the skin, tissue microdamage in the moving parts of the eye could have devastating consequences for vision. Accordingly, we propose that ACAID represents an intrinsic anti-inflammatory strategy in the anterior chamber to minimize immune activation at this site of movement of ocular components. However, three different MCH class I associated eye disease, including HLA-B27 related conditions, develop anterior uveitis. This shows that the anterior chamber immunomodulatory pathway, not infrequently, fails in the clinical setting.

A re-evaluation of Sympathetic Ophthalmia as an archetypal autoimmune disease

Sympathetic ophthalmia (SO) was recognized in antiquity and was also one of the first diseases to be classified as autoimmune [1]. Whilst recognizing that immunopathologic mechanisms may eventually involve both autoantibodies and CD8 T cell cytotoxicity, the crux of the present viewpoint is that autoimmunity falls into two main categories based on initiating factors. These are the classic autoantibody associated disease where autoantibody production predates discernible tissue damage and a second category of MHC class I associated diseases where tissue damage predates secondary adaptive immune response. The need for injury to one eye that leads to autoimmunity in the fellow eye (sympathizing eye) argues for the latter type of ocular autoimmunity in SO [53]. Also, whilst the classic autoimmune diseases are characterized by progressive destruction, the sympathizing eye in SO can be saved by surgical removal of the injured eye strongly suggesting that tissue-specific factors in that eye drive autoimmunity [54]. The demonstration that bovine melanin associated antigen (BMAA) effectively stimulates peripheral lymphocytes in SO patients combined with the absence of specific autoantibodies supports the concept that CD8 T cells could orchestrate the immunopathology [55, 56].

The immunopathogenetic basis of SO can be better understood from looking the related disease called the Vogt–Koyanagi–Harada syndrome (VKH), which is a multisystem disorder with bilateral granulomatous uveitis, skin and central nervous system disease [57, 58]. It is thought to be due to autoimmunity against melanocytes that are found at all of these sites [56]. Unlike SO, strong MHC class II associations have been shown in Mexican, Hispanic and Japanese patients [59, 60]. The widespread target tissue distribution of VKH also supports the idea of an autoantibody gaining access to multiple sites and thus initiating tissue damage. The presence of IgG anti-lens epithelium-derived growth factor (LEDGF) autoantibodies in patients with VKH was significantly higher than that in healthy controls, but not in patients with panuveitis of other aetiology, suggesting that the humoral immune response against LEDGF plays a role and is not a mere secondary phenomena caused by uveal tissue damage [61]. These findings point towards divergent immunopathogenetic mechanisms in these two clinically related conditions and ultimately, this could have important implications for therapy development.

Challenging Contentious Autoimmune Diseases Paradigms – Multiple Sclerosis

We have previously drawn attention to the importance of inflammatory disease at two different sites that is not clearly autoantibody associated, especially in the case of psoriasis and psoriatic arthritis, where a common reaction to biomechanical stress, rather than a common autoantigen, could lead to disease [7, 62]. In the absence of an alternative immunological paradigm, multiple sclerosis (MS) has invariably been viewed as autoimmune. In favour of a traditional autoimmune hypothesis for MS is the MHC class II (HLA-DR15) association for both CNS and uveal disease and the presence of potentially pathogenic autoantibodies in established disease, although the latter has been recently challenged [63–65]. The proposed immunopathogenetic mechanisms in MS have always emphasized the importance of autoreactivity to myelin components or other CNS lipid-associated proteins [66]. This latter theory is firmly supported by the experimental allergic encephalomyelitis (EAE) model which is said to mimic human MS [67].

An evaluation of ocular microanatomy in relationship to MS associated uveitis raises serious questions about the prevailing paradigms for MS immunopathology. The commonest ocular manifestation of MS is intermediate uveitis and is present in 3–27% of patients [68, 69]. The uveal tract contains a limited amount of myelinated neurons compared with the brain and axons of the nerve fibre layer of the retina are unmyelinated [70]. Given the relative paucity of myelin related proteins at all of these sites, it is unclear why inflammation is so common here and why it shows such a propensity for the intermediate uvea (Fig. 6). Therefore, considering MS from its ocular micro-anatomical facet raises questions about the importance of myelin autoimmunity in the EAE model as a representative human model. Furthermore, the presence of autoantibodies that pre-date end-organ damage, a recently recognized hallmark of true autoimmunity, has not yet been demonstrated in MS [71].

Figure 6.

 The pathogenesis of human MS has been viewed predominantly in relationship to autoimmunity to myelin associated proteins. The eye is a common target in MS, but compared with the CNS, there is actually a paucity of myelin related antigens in the eye. The middle box depicts the intermediate uvea, which is the commonest site of the ocular MS manifestations. It is argued that a true understanding of MS will only evolve when the site-specific factors that lead to such patterns of disease are discerned.

Indeed, a non-autoimmune component for MS is being increasingly supported from histopathology and molecular imaging that suggest a primary toxic or degenerative basis and that inflammatory responses are secondary [71, 72]. Therefore, an evaluation of brain-specific factors such as degenerative or toxic changes that lead to innate immune activation could play an important role in understanding MS pathogenesis.


Some immunologists have argued that the extreme complexity of the immune system functioning may go unmeasured and is something that will be difficult to measure in the future [73]. Whilst that may be the case, ultimately for the eye, the prevention of blindness, without ever really fully understanding the complexities of the immune system, is perfectly acceptable. Herein, we have argued that understanding the interplay between innate and adaptive immune responses based on an appreciation of ocular physiology provides a novel basis for understanding immunity in a relevant disease model context. The foundations for our approach were laid by Mechnikov and Ehrlich more than a century ago. Whilst their studies seemed contradictory and were not reconciled at the time, they are actually closely intertwined and together hold the key to elucidate the mechanisms of eye disease and inflammatory disease in general. Providing we can ascertain the principal immune mechanism leading to tissue inflammation and whether aberrations reside in adaptive or innate immunity and by extension the tissue where innate immune activation takes place, then a scheme for the prevention of visual loss can be developed. The onus to unravel the Mechnikov’s thorns or organ-specific factors contributing to disease is not just the realm of immunology but also medicine.