Glaucoma is one of the leading causes of blindness in the world. Over 67 million people are affected by glaucoma (1, 2). Elevated intraocular pressure (IOP) is the major risk factor for glaucoma. Primary open-angle glaucoma (POAG) is now defined as a progressive disease of retinal ganglion cells characterized by structural change in the optic disk and by typical, slowly progressive loss of function. The most common form of the group of glaucomatous diseases is POAG with about 80%. Ocular hypertension (OHT) is characterized by elevated IOP without optic nerve degeneration. One to 2% of patients with OHT per year convert to POAG.

The pathogenesis of the optic nerve damage in POAG is complex and associated with increased IOP, neurotoxicity and apoptosis (3, 4), changes of the extracellular matrix (5, 6), activation of glia cells (7), an interrupted transport of neutrophins (8), oxidative stress (9, 10), and hypoxia due to ocular and systemic vascular dysregulation (9). There are also reports of glaucoma-associated degenerative processes of the central visual pathways in the brain (11). Thus, new targets for therapeutic intervention, such as improving ocular blood flow and direct neuroprotection of retinal ganglion cells, are under investigation.

Since elevated IOP is the primary risk factor for the development and progression of glaucoma, the lowering of the IOP is the primary goal of all treatment strategies (12). Many studies have shown that IOP reduction can slow the progression of glaucoma (13–17) and delay or even prevent the onset of retinal ganglion cell loss. Lowering the IOP by 1 mmHg prevents progressive visual field loss by 10% (18). Conservative [e.g., β2-adrenergic receptor (β2AR) blocker] and surgical therapeutic options are offered to achieve IOP lowering. The β2AR blocker Timolol reduces the IOP in humans (19) by suppressing the rate of aqueous humor formation (20, 21) by 30–50% (20–26). The relative potency of the β2-adrenergic blocker Timolol is higher than that of the β1-adrenergic receptor blocker Betaxolol (27), indicating the prominent role of β2ARs in the formation of aqueous humor and/or regulation of its flow.

Trabecular meshwork and ciliary body express β2ARs (28, 29), regulating the aqueous humor system in the eye and consequently influencing IOP. Thus, β2AR plays a key role in the regulation of both aqueous humor production and outflow. In addition, β2AR is also expressed in human optic nerve and in microvessels (30, 31). β2ARs are members of the G protein-coupled receptors (GPCR) family, comprised of more than 600 genes (32, 33). They have seven membrane spanning domains, three intra- and extracellular loops, an extracellular N-terminus, and an intracellular C-terminal tail. Most of the GPCR form homodimers upon ligand activation. It has been shown that functional autoantibodies (AAbs), directed against GPCR, are associated with various human diseases. The first AAb against a GPCR was described for the β2AR in patients with allergic asthma in 1980 (34). Several groups identified agonistic AAb against the β1-adrenergic receptor in Chagas’ disease (35, 36), dilated cardiomyopathy (37–39), ischemic cardiomyopathy (40), and myocarditis (41). β2- and α1-AAb were seen in Alzheimer’s and vascular dementia (42). β1-, β2-, and α1-AAb were shown to be increased in patients with preeclampsia (43). α1-AAb were also detected in primary and malignant hypertension (44).

We hypothesize that AAb against GPCR might be involved in the pathogenesis of glaucomatous disease by influencing the dynamics of aqueous humor.

To study the hypothesis that circulating immunoglobulins of patients with glaucoma activate GPCRs, isolated rat cardiomyocytes were employed as target cell. Frequency modulation of spontaneously beating neonatal cardiomyocytes is highly efficient to investigate such targets (1–3, 37, 45). The rat cardiomyocytes contain a spectrum of relevant GPCRs, e.g., β1-, β2-, α-ARs, and Angiotensin-1 receptor, which display a high homology to their corresponding human receptors. The homology of the amino acids between the human and the rat second extracellular loop is 88%; however, the homology in the estimated binding site of the AAbs in this extracellular loop is 100%. The cells were exposed to IgG prepared from controls or patients with glaucoma and OHT. To verify the specificity of the effects, the heart cells were treated with receptor-specific agonists in the presence or absence of receptor-specific antagonists and compared with the AAb-mediated responses. In screening studies using respective agonists and blocking agents an interaction of the glaucoma-associated IgG with the β1AR, α-AR, or with AT1-receptor system has been excluded (not shown). Based on these findings, the potential importance of β2AR-directed antibodies associated with glaucoma was analyzed in more detail. These findings were the basis for the following experiments. Investigations of the target loop as well as the target epitope of the GPCR and the IgG isotype of the AAb were added. Surface plasmon resonance analysis confirmed the receptor specificity of the agonistic β2AR AAb.

In order to draw a bow between the molecular findings and clinical data, correlation analyses of the patients’ clinical parameter with the β2AR AAb were done. Finally, a pilot proof-of-principal study was performed in four glaucoma patients using an extracorporal immunoadsorption (IA) for removal of AABs against β2AR and other immunoglobulins G in order to monitor a potentially transient or permanent change of IOP.

In the present study, we describe formerly unknown circulating AAb in patients with glaucoma (50). We screened patient sera for antibodies that activate a GPCR expressed by cardiomyocytes and detected AAb against the β2AR employing specific inhibitory peptides. These agonistic AAb induced a dose-dependent stimulation and their effects were blocked by the selective antagonist ICI118.551, and neutralized by peptides corresponding to the second extracellular loop of the β2AR (AS 181–192).

Autoantibody can be detected by several methods: (1) enzyme-linked immunosorbent assay detects the patient’s antigen bound AAb by a specific (to human IgG) animal sourced and enzyme (peroxidase or alkalic phosphatase) labeled antibody; (2) fluorescence (microscopy) detects AAb as (1), but the detection of the AAb is fluorescence labeled (e.g., fluorescein isothiocyanate, FITC); (3) radioimmunoassay detects AAb as (1), however, the detection of the AAb is radio labeled (e.g., iodine); (4) bioassay (with spontaneous beating neonatal rat cardiomyocytes) measures the increase in the beating rate, induced by added IgG (fraction from human serum) due to the binding of AAb to the corresponding receptor (here: β2-AR). The detection methods (1)–(4) measure the AAb indirectly, however method (4) additionally directly. As the detected signal extent of the bioassay is proportional to the AAb bound to the surface fixed target, we decided to choose the cardiomyocyte bioassay for the detection of the AAb. Next to the detection of the AAbs, this method enables functional analyses.

The AAb found in glaucoma patients are of the IgG3 subclass. The antibodies that trigger effector functions, and that are most likely to be involved in immunoregulatory activity, are IgG3 and IgG1 (51). In addition, it is known that IgG3 is involved in antibody-dependent cellular cytotoxicity (52). Complement activation is most effective with IgG3 and, to a lesser extent, with IgG1, IgG2, and IgG4. Usually, complement binding and activation leads to destruction of the target structure. The plasma concentration of IgG3 is low, and its half-life is shorter than that of any of the other IgG-subclasses. Further investigations are necessary to elucidate whether complement activation and cell death are involved in the pathogenesis of glaucomatous diseases.

Recent studies support the hypothesis that glaucoma-associated neurodegenerative processes are partially caused by immune-mediated mechanisms (53–56). These studies report complex AAb repertoires in patients with glaucoma which were directed against retinal and optical nerve antigens. Such autoantibody reactivities were shown in optical nerve extracts from bovine eyes. The later findings are of interest in respect to ocular pressure-dependent and independent autoimmune responses. With regard to AAb-mediated immunoreactions, these pathomechanisms seem to be driven by inflammation.

The β2AR is one of the best investigated GPCR. The modulation of cardiac contractility is also mediated by β2AR, and its activation might have chronic effects on cellular metabolism, on cell growth, or on excitability (57). This receptor seems to be involved in allergic asthma, several cardiomyopathies, such as Chagas’ disease, and cardiac electrical disturbances (58–60). However, whereas the AAb in the case of Chagas disease were functionally active in cardiomyocytes, β2-AAb from asthmatic patients react in the opposite way. Respective loop analyses explained these findings; in the Chagas-related cardiac disorders the loop II was identified to be the target of AAbs, whereas in the asthmatic disease the loop III was recognized. Thus, it was concluded that the various extracellular loops of the β-AR are functionally linked to distinct antibody-mediated responses.

Several years ago, the effect of monoclonal antibodies directed against the second extracellular loop of the beta2 adrenoceptor was investigated. Interestingly, similar results were detected as seen with the AAbs prepared from the sera of glaucoma patients (61). The agonist-like effect of this mAB (6H8) was blocked by the specific β2-adrenergic antagonist ICI118.551 and a peptide HWYRATHQEAINCYANETC corresponding to the second extracellular loop of the β2-adrenoceptor. The β1-adrenoceptor antagonist bisoprolol was without influence. The AAb were directed against loop II and recognized the overlapping epitopes AINCYAN and ANETCCD. Considering that the peptide corresponding to the second extracellular loop and the short overlapping peptides cannot form the correct conformation of the extracellular loop, the anti-β2 receptor antibodies were pretreated with the peptides corresponding to the first, second or third extracellular loop for 1 h. Under these conditions, only the peptide that corresponds to the second extracellular loop was able to neutralize the AAb activity. The peptides of the first and third extracellular loops were without effect. Similar results we observed also for the short overlapping peptides with an amino acid sequence of 5–6 amino acids. In these experiments, the sequence (AINCYANETCCD) was able to neutralize the AAb action. Therefore, it is assumed that this sequence represents the epitope of the AAb on the second extracellular loop. Plasmon resonance and affinity purification confirmed the direct interaction of the AAb with loop II of the β2AR. The agonistic β2AR antibodies described here are good mechanistic candidates that might induce and maintain an elevated IOP as a critical step in the development of glaucomatous diseases.

In this study, we present first in vivo data of four patients with POAG that underwent IA in a pilot proof-of-principal study: glaucoma therapy is considered to be effective when IOP reduction is 20% from wash out baseline (62). Even not all patients in a study group must show a substantial IOP decrease of 20%, in cases when representative parts of the patients display even more than 20% IOP reduction. Thus, an IOP decrease could be observed in 3 of 4 patients after IA: patients 2 and 3 showed an IOP decrease of 27.8% (OD)/32.1% (OS) and 11% (OD)/46.3% (OS) using equal number of topical antiglaucomatous eye drops. An IOP decrease of 2.9% (OD) and 6.2% (OS) could be observed 12 months after IA in patient 4. As it is known that each single millimeter of mercury increases glaucoma progression rate about 12–13% (18) even the IOP decrease in this patient is important and enables winning sighted lifetime. The IOP of patient 4 is considered to be even lower after IA as the number of local antiglaucomatous eye drops decreased from 4 to 0, thus, measured IOP values under no local therapy can be seen as a lowered IOP after IA with even higher IOP under quadruple therapy before IA. Summarizing, immunoadsorption seemed to lower IOP or even number of antiglaucomatous therapy. Possibly, the decrease of IOP could even be larger, if β2AR AAbs are blocked specifically. Because IA is a non-specific method for adsorbing AAbs and other IgGs, several other AAb were removed in addition to the agonistic β2AR AAb. As known for other autoimmune diseases, AAb relapse. This effect was also seen in our patients. Detection of β2AR AAbs went along with an IOP increase (patients 2 and 4). The delay in the decrease of IOP might be possibly due to a re-sensitization of the β2AR in the absence of stimulating AAbs. In addition, the therapy was well tolerated and no adverse effects were observed.

The most important target of glaucoma treatment is elevated IOP. A large body of evidence has established the importance of the reduction of the IOP in the medical management of glaucomatous diseases (13–17, 63). The pathogenic damage of the optic nerve in POAG is not well understood. It is supposed to be associated with increased IOP, glutamate toxicity (64), interrupted transport of neutrophins (8), apoptosis (4, 65), extracellular matrix changes (5, 6), and hypoxia due to ocular and systemic vascular dysregulation (66).

The aqueous humor dynamic plays the major role in the regulation of IOP. Four factors maintain the steady-state IOP of the healthy eye. These are flow of aqueous humor, resistance to outflow, episcleral venous pressure (67), and uveoscleral outflow (68). The IOP reflects the balance between in- and outflow of aqueous humor. A major strategy in medical treatment of glaucoma is the reduction of inflow and, thereby, normalization of IOP. Understanding the mechanisms and regulation of inflow is, thus, of undoubted clinical relevance. Several mechanisms underlying increased inflow have been identified, however, the integration and regulation of these mechanisms is still elusive (69). The balance of in- and outflow of aqueous humor is modulated by the local β2AR activity and may be chronically influenced by agonistic AAbs.

Usually, night-time aqueous flows are approximately half of those observed at daytime in active humans (70–72). The ability of aqueous flow stimulation by epinephrine during night-time shows a low basal β-adrenergic stimulation during the night (23). Patients with POAG form in average 15% more aqueous humor/24 h than control subjects of the same age. This is due to a significantly higher aqueous flow in the night in comparison to control subjects (73). Thus, a β-adrenergic overdrive precludes the downregulation of the aqueous flow during the night. This observation is underlined by the lack of receptor desensitization by agonistic AAb displayed in Figure 3. Usually, β2AR agonists such as clenbuterol desensitize the receptor within 1–2 h (Figure 3B). By contrast, the β2AR AAb do not desensitize the β2AR in vitro for at least 5 h. This chronic β2-adrenergic stimulation may explain the higher aqueous flow and the elevation of the IOP in patients with POAG. Alteration of the circadian rhythm of IOP in POAG can be explained by the presence of AAb. The lack of tachyphylaxia was also observed for agonistic antibodies against anti-β1AR-, anti-α1AR-, anti-AT1-, and anti-muscarinic M2 receptor in several cardiovascular diseases (74, 75).

The AAb might also play a modulatory role in the aqueous outflow. Putney et al. (76) found an increase of cell volume of trabecular meshwork cells via Na⁺-K⁺-Cl⁻ cotransporters in glaucoma leading to a reduced outflow. As β-adrenergic stimulation results in an activation of Na⁺-K⁺-Cl⁻ cotransporters, agonistic AAb might also contribute to reduced outflow facility.

An “organ specificity” for the AAb against the β2AR can be assumed as patients of glaucoma suffered not of other diseases, which can be caused by these AAb (e.g., allergic asthma). It can be hypothesized that the chemical surrounding has to be changed before the autoantibodies can act specifically. Previous data showed that the βAR is not accessible for the AAb under adequate oxygen conditions. However, the βAR seemed to be unmasked, if ischemia occurred (77, 78). This could be a potential reason as glaucoma is hypothesized to be a disorder with local ischemia (79).

The data presented here offer another insight into eye disorders associated with high IOP. The newly identified AAb against the β2AR were exclusively found in the sera of patients with POAG and OHT but neither in those with cataract, nor in controls. Based on these observations, we suggest that these AAbs are not only immunological markers with a high incidence in POAG and OHT, but might also be pathomechanistically relevant β2AR activating molecules. The common, broadly successful therapy with β2AR antagonists further supports the idea that AAb-mediated mechanisms might be of pathogenic importance. In addition to their effect on IOP, the AAb might also have a direct effect on retinal ganglion cells as β-adrenoreceptor antagonists showed neuroprotective properties (80).

In patients with POAG and OHT, however, not in controls, agonistic AAb directed to β2AR were detected. These AAb interacted with the second extracellular loop of β2AR (peptides 181–187 and 186–192), thus functionally different links to distinct antibody-mediated responses are assumed. Data of our clinical pilot proof-of-principal study might suggest that a reduction of total IgG, including AABs against β2AR (IgG3 isotype) by IA, transiently decreases IOP. These findings might indicate a possible role of these AAb in the dynamics of aqueous humor and might support a contribution of adaptive autoimmunity in the etiopathogenesis of POAG and OHT. A specific blocking of AABs against β2AR might be of interest for further studies.

The study followed the tenets of the declaration of Helsinki for research. Written informed consent to use serum samples for research purposes was obtained from each participants of the study. The institutional review board of the University Hospital Erlangen approved the protocols (3483; ClinicalTrials.gov Identifier NCT00494923).

RK, MH, and GW had the idea and planned the study; GW, QF, and SB performed the laboratory work; AJ, JR, and RV planned and performed the clinical trial; HK did the synthetic work; JH did laboratory work; AS and BH performed data acquisition and statistical analysis; RR was involved in the analysis and revised the manuscript, FH was involved in the analysis, BH was responsible for the draft of the manuscript. US-S interpreted results and edited the manuscript. All authors have no financial interest in any topic regarding the present study and approved the final version of the manuscript.

(1) Patent: EP 1832600 A1 (2) RK is working for Fresenius Medical Care GmbH. The reviewer RK declared a shared affiliation, with no collaboration, with one of the authors, HK, to the handling editor.