Based on studies with the labeled radioligands, [³H]propylbenzilylcholine mustard ([³H]PrBCM), [³H]N-methylscopolamine ([³H]NMS) and [³H]quinuclidinyl benzilate ([³H]QNB), a high density of muscarinic binding sites has been demonstrated in rat, bovine and chick retinas, whereas relatively few binding sites have been detected in frog and salamander retinas [37,38,39]. Autoradiographic experiments in embryo and adult chicken retinas revealed specific mAChR binding sites in the inner synaptic retinal layer [40]. In 1985, Polans et al. found a high density of muscarinic binding sites in the inner plexiform layer (IPL) and the outer plexiform layer (OPL) of the salamander retina [41]. In the same year, Hutchins and Hollyfield presented evidence for a population of mAChRs in the human retina, apparently expressed in the IPL, by using the irreversible ligand, [³H]PrBCM [42].

Based on autoradiographic experiments, it has been suggested in 1988 that mAChR subtype number and distribution change during retinal development [43]. Later, experiments in the ferret retina suggested that the subtypes, number and distribution of mAChRs changes during retinal synaptogenesis [44]. In the study, mAChR-like immunoreactivity was found at amacrine–amacrine cell contacts by electron microscopy and immunohistochemical techniques [44]. Townes-Anderson and Vogt found that mAChRs in the salamander retina are located on amacrine/ganglion, bipolar, and horizontal cells [45]. In 1988, Moroi-Fetters found that stimulation of muscarinic receptors by the subtype-preferring M₁ receptor antagonist, pirenzepine, in the rat retina causes phosphoinositide hydrolysis, which indicated that these receptors appear to be of the M₁ subtype [46].

In 1989, all five mAChR subtypes (M₁, M₂, M₃, M₄ and M₅) were identified [47,48]. Molecular cloning techniques provided a new molecular basis to characterize expression, location and physiological function of all five mAChRs [49]. In 1997, McKinnon et al. examined regulation of mAChR expression in the chicken embryonic retina by using immunoblot, immunoprecipitation and solution hybridization analyses [50]. The authors reported that the M₄ receptor is the main subtype expressed at an early stage in embryonic development, while M₂ and M₃ receptor expression increases at a later stage [50].

However, the precise anatomical location of mAChRs in the retina remained unknown at that time. One year later, Fischer et al. used purified and subtype-specific antibodies directed against M₂, M₃ and M₄ subtypes to detect receptor immunoreactivity in the retina. The study revealed that in the chick retina the M₂ receptor was expressed in amacrine and ganglion cells, the M₃ receptor was expressed in many bipolar cells and small subsets of amacrine cells and the M₄ receptor was found in amacrine and ganglion cells [51]. In an in vitro and in vivo study, Belmonte et al. demonstrated that retinal Müller glial cells can secrete a factor called MARIA (muscarinic acetylcholine receptor-inducing activity) that can regulate M₂ expression in vitro and in vivo [52]. Another in vitro study demonstrated the presence of the M₁ receptor in cultured human retinal pigment epithelium (RPE) at both the mRNA and the protein level [53]. The expression of M₁ receptor mRNA was also observed in the guinea pig retina, and immunohistochemical findings revealed that the M₁ receptor was expressed in all layers of the retina [54]. Strang et al. used RT-PCR, Western blot analysis and immunohistochemistry to identify the expression and distribution of mAChR subtypes in the rabbit retina [55]. The authors detected mRNA expression for all five mAChR subtypes in the whole neural retina by RT-PCR and Western blotting, and they confirmed that all five mAChR subtypes were expressed by subpopulations of bipolar, amacrine, and ganglion cells by immunohistochemical analyses [55]. According to a study by Gericke et al. in 2011, only mRNA for the M₃ receptor was detected in murine retinal arterioles [56]. In contrast, mRNA for all five mAChR subtypes was detected in ophthalmic arteries, but mRNA levels for the odd-numbered subtypes, M₁, M₃ and M₅, were higher than those for the even-numbered subtypes, M₂ and M₄ [57]. Figure 1 shows the distribution of mAChRs within the retina.

Based on the expression studies, all five mAChR subtypes have been detected in the retina with the individual subtypes showing an overlapping expression pattern.

Peralta et al. and Bonner et al. first cloned and sequenced human mAChRs, which are encoded by the CHRM1 to CHRM5 genes [58,59]. The genes give rise to the five subtypes, M₁, M₂, M₃, M₄ and M₅ [3]. The mAChR family belongs to the superfamily of seven-transmembrane receptors, which mediates cellular signal transduction pathways via G-proteins [60]. The M₁, M₃ and M₅ subtypes, which efficiently couple to Gq/11 subtype G proteins, can mobilize phosphoinositides to generate inositol 1, 4, 5-triphosphate (IP3) and 1, 2-diacylglycerol (DAG) via activation of phosphoinositide-specific phospholipase Cβ (PLCβ), leading to an increase in intracellular cytosolic calcium (Ca²⁺) levels and protein kinase C (PKC) activity [61]. This may help to stimulate nitric oxide (NO) production, since neuronal nitric oxide synthase (nNOS) is calcium/calmodulin-dependent [62]. The M₂ and M₄ receptors preferentially couple to pertussis toxin-sensitive Gi and Go proteins, causing an inhibition of the cAMP-dependent pathway via suppression of adenylyl cyclase [63]. Furthermore, both Gq/11-, Gi/o-coupled with mAChRs may exert effects through activation of small GTPases, such as Rho and Ras, and downstream effectors, such as phosphoinositide-3 kinases and mitogen-activated protein kinases [60].

Muscarinic receptors play an important role in the development of the retina and in processing visual information [64]. For example, mAChRs regulate the function of bipolar cells in the ON/OFF channel, and the input and output of amacrine, bipolar, ganglion and horizontal cells [52,64]. Additionally, Jardon et al. suggested that a cholinergic loop of amacrine cells could be involved in the inhibitory pathway from the ON channel to the OFF channel in the frog retina carrying “light on” and “light off” information from the retina to the brain [65,66]. Muscarinic cholinergic transmission exerts a substantial contribution in the retina [67]. In an in vitro study, the muscarinic antagonist, QNB, enhanced the amplitude of the electroretinogram (ERG) b-wave (a measure of ON bipolar cell activation), and induced moderate vasoconstriction in the cat retina [67].

Although mAChRs regulate many important cellular signaling pathways, it is difficult to assign specific functional roles for individual mAChR subtypes in the retina. Jositsch et al. tested the specificity of mAChRs antibodies under different conditions in immunohistochemical labelling on tissue sections by analyzing specimens from respective gene-deficient mice and wild-type mice [68]. The data indicated that immunohistochemical detection of mAChR subtypes in tissue sections is limited to the M₂ subtype [68]. It has been reported that cells frequently co-express more than one mAChR subtype that increases the difficulty of assigning a functional response to a single receptor subtype [3]. The lack of highly selective pharmacological ligands and antibodies for individual mAChR subtypes has hampered conclusions regarding the physiological role of individual muscarinic receptor subtypes. The development of genetically modified mice devoid of M₁- to M₅ receptors helped to circumvent the problem of assigning a specific function to an individual mAChR subtype [69,70,71,72,73,74,75]. M₁-M₅ receptor knockout mouse models (M₁-M₅R−/−) have also been used in studies of ocular tissues.

For example, Barathi et al. studied the role of each of the mAChR subtypes in the development of myopia by using M₁-M₅R−/− mice [76]. The authors found that M₂ receptors play a crucial role in myopia development by hindering scleral fibroblast cell proliferation and further scleral remodeling [76]. Based on studies in M₃R−/− mice, Gericke et al. showed that the M₃ receptor is responsible for mediating cholinergic responses in retinal arterioles and the ophthalmic artery [56,57,77]. Laspas examined the amount of cells in the retinal ganglion cell (RGC) layer and the amount of axons in the optic nerve in 5-month-old M₁R−/− and wild-type mice and found no significant difference between both groups [78]. More recently, the same laboratory conducted experiments in 5- and 15-month-old M₁R-M₅R−/− mice to examine whether one of the mAChRs and age have an influence on neuron survival in the retina [30]. Based on these studies, the M₁ receptor was found to be critical for RGC survival in the aging mouse retina [30]. These examples show how genetically modified mice may help to better understand the physiological roles of individual muscarinic receptor subtypes in the retina.

According to their differential coupling to intracellular signaling cascades, the mAChR subtypes have been divided into two subfamilies, the “M₁-like” mAChR subfamily and the “M₂-like” mAChR subfamily [79]. The odd-numbered subtypes, M₁, M₃, and M₅, belong to the “M₁-like” family, which couple to the Gq protein and activate phospholipase C (PLC)-dependent signaling pathways. In neuronal tissue, activation of this signaling cascade increases neuronal excitability through activation of nonspecific cation channels, release of Ca²⁺ from intracellular stores, or inhibition of Ca²⁺-activated K⁺ channels [80]. On the contrary, members of the “M₂-like” family, the even-numbered M₂ and M₄ receptor subtypes, are generally linked to inhibition of adenylyl cyclase activity [79]. Activation of M₂ and M₄ receptors decreases neuronal activity via activation of a subset of K⁺ channels, the inhibition of Ca²⁺ channels, or the inhibition of the Ca²⁺ priming of K⁺ channels [55,81,82]. This suggests that the different subtypes may subserve different functions in the retina.

A plethora of studies demonstrated that the M₁ muscarinic receptor subtype may be crucial for neuron survival in the retina [30]. For example, M₁ receptor activation protected retinal neurons from glutamate-induced cytotoxicity [83]. It has been proposed that activation of M₁ receptors reduces Ca²⁺ influx into the cell and the expression of Bcl-2 and Caspase-3 [62,83,84]. This effect can be blocked by the M₁-preferring muscarinic receptor antagonist, pirenzepine [84]. Additionally, activation of M₁ receptors significantly increases the survival of RGCs in vitro [85]. Pereira et al. indicated that the mechanism involved in M₁ receptor activity and the survival of RGCs is by the release of polypeptides and activation of insulin receptor kinase receptors [85]. An in vitro study demonstrated that activation of mAChRs effectively protects against hypoxia-induced apoptosis in RGCs via modulation of the hypoxia-inducible factor 1-alpha (HIF-1α) pathway [86]. Based on these in vitro data, Laspas et al. conducted a study in 5-month-old M₁R−/− mice and age-matched wild-type mice to test the neuroprotective role of the M₁ receptor in vivo. However, the authors found no differences in the number of retinal neurons and the amount of optic nerve axons between M₁R−/− and wild-type mice [30]. In a more recent study in 2019, Laspas et al. examined the potential role of all five muscarinic receptor subtypes on neuroprotection in the RGC layer in congenic mAChR−/− mice of different age categories [30]. Intriguingly, the authors observed that the lack of the M₁ receptor was associated with a reduced RGC density in aged mice. Aged M₁R−/− mice also displayed elevated ROS levels in the RGC layer and increased retinal mRNA expression for the prooxidative NADPH oxidase 2 (NOX2) and reduced mRNA levels for the antioxidative enzymes, superoxide dismutase 1 (SOD1), hemeoxygenase-1 (HO-1) and anti-glutathione peroxidase 1 (GPx1) [30]. The findings provided the first direct evidence that the lack of the M₁ receptor leads to accelerated RGC loss in mice via changing in the oxidative/antioxidative balance in favor of oxidative in the retina [30]. L-satropane was reported to be effective in preventing retinal neuron damage, which may be attributed to decreasing cell apoptosis and amyloid-β (Aβ) production via activation of M₁ receptor subtype [87]. Moreover, in an in vivo study, the M₁ muscarinic receptor was reported to exert protective effects on RGCs via activation of insulin growth factor 1 (IGF-1) and insulin growth factor 1 receptor (IGF-1R) [88]. Additionally, activation of PKC delta was suggested to regulate neurotrophin levels by M₁ muscarinic receptor activation ultimately leading to an increase in RGCs’ survival in vitro in the retina [89,90]. In 2021, an in vivo study in rats showed that huperzine A lowers intraocular pressure via the M₃ receptor and exerts neuroprotective effects in the retina by increasing endogenous ACh levels and activating M₁ receptors and their downstream AKT/MAPK signaling pathways [31].

Braga et al. first analyzed the levels of M₃ receptors in retinal cell cultures treated with 50 ng/mL phorbol 12-myristate 13-acetate (PMA, a PKC activator) for 48 h. PMA induced a marked increase in M₃ receptor levels [91]. Based on pharmacological studies employing muscarinic subtype-preferring antagonists, Borda et al. observed that carbachol can stimulate NOS activity and increase the expression of nNOS and iNOS mRNA in the rat retina via activation of M₁/M₃ receptor subtypes [92,93].

The expression of nicotinic AChR (nAChR: α3, α4, α6, α7, β2 and β4 nAChR subunits) and/or mAChR by amacrine and ganglion cells has been described in retinas of Rhesus monkeys and rabbits [94,95,96]. Retinal nAChRs mediate visual processing and may have effects on refractive development and ocular neovascularization [94]. In the retina, there was an overlap in the expression patterns of M₁, M₄ and M₅ muscarinic receptors with those of non-α7 and α7 nAChRs in presumptive amacrine and ganglion cells [97]. Strang et al. suggest that the determining the role of mAChRs in retinal processing is complicated by the concomitant expression of nAChRs by the same cells [55]. A study in α7 nAChR−/− mice demonstrated that M₂ and M₄ mAChR subtype transcripts were significantly upregulated in the RGC layer [98].

There are not many studies on the functional role of M₂ receptors in the retina. Several pieces of evidence suggest that activation of M₂ and M₄ receptors is involved in visual processing [55]. The M₂ receptor has been reported to increase Ca²⁺ influx, exclusively due to Ca²⁺ mobilization from intracellular stores [99]. The M₂ muscarinic receptor was shown to inhibit adenylyl cyclase activity and to activate inwardly rectifying potassium (K⁺) channels [100]. However, these responses can be rapidly attenuated by receptor desensitization [100,101]. Antal et al. found that activation of M₂ receptors regulates feedforward inhibition following activation of RGC synapses in a manner that is strongly dependent on the number of activated RGCs [102]. Cimini et al. indicated that the production of NO in response to M₂ muscarinic receptor activation may lead to an increase in cGMP, which can modulate the mutual interactions of acetylcholine-glycine-gamma-aminobutyric acid (GABA) in the inner retina [103].

The M₄ receptor exerts a direct inhibitory control on dopamine D₁-like receptor signaling [104]. In a rat glaucoma model, Almasieh et al. demonstrated that activation of M₁ and M₄ receptors promotes RGCs’ survival [15]. Moreover, the partially selective M₁/M₄ muscarinic antagonist, pirenzepine, was reported to be successful in preventing myopia progression in animal models [105]. The M₄-selective antagonist, himbacine, could also prevent myopia in chicken by daily intravitreal injections [106].

All these studies provide evidence that individual mAChRs exert specific functional roles in the retina, which offers new therapeutic perspectives for mAChR ligands. A variety of studies reported associations of endothelial NOS (eNOS) genetic polymorphism or adrenergic receptor gene polymorphisms with retinal diseases [107,108]. Additionally, mAChR subtype gene polymorphisms have been reported [109]. Unfortunately, studies on mAChR genetic polymorphisms with respect to retinal diseases have not been reported so far [109]. Such studies would be appreciated to shed some light on the role of mAChR in the development of specific retinal diseases.