Neuronal Nitric Oxide Synthase and the Autonomic Innervation of the Mouse Lacrimal Gland

Abstract

purpose. To determine the expression patterns of the vesicular acetylcholine transporter (VAChT), tyrosine hydroxylase (TH) and neuronal nitric oxide synthase (nNOS) in the pterygopalatine ganglion (PPG) and the exorbital lacrimal gland of normal mice.

methods. Mouse PPG and lacrimal glands were processed for single- and double-labeled indirect immunofluorescence studies. Slides were examined with conventional fluorescence microscopy and confocal laser scanning microscopy.

results. All the somata in the PPG expressed both VAChT and nNOS immunoreactivity (IR). The postganglionic axons within the ganglion showed less VAChT-immunoreactive intensity than that seen in the somata, whereas nNOS IR was almost undetectable. In the lacrimal gland, nNOS-positive nerve bundles and fibers were observed to be associated with tear-collecting ducts, blood vessels, and acini. Some nNOS-positive punctate elements appeared to be distributed among acini. Many nerve fibers were VAChT immunoreactive and a small number of fibers were TH immunoreactive in the gland. Most of the VAChT-positive fibers and some of the TH-positive nerves displayed nNOS IR.

conclusions. The expression of nNOS in cells of the PPG and in lacrimal gland nerves suggests that NO may play a role in modulating tear production. The site of action may include the PPG, ducts, blood vessels, acini, nerve fibers, and myoepithelial cells within the gland. NO may modulate parasympathetic and/or sympathetic synaptic transmission or by acting directly on lacrimal gland components. The interaction between NO-ergic and the conventional autonomic input illustrates the complexity of the innervation pattern of the mouse lacrimal gland.

The mammalian lacrimal gland has dense parasympathetic innervation ^{1 2 3} and, in some species, sympathetic innervation. ^{4 5 6 7} Most of the parasympathetic nerve fibers that innervate the lacrimal gland arise from neurons in the pterygopalatine ganglion (PPG). ^{1 2 3} The PPG is the parasympathetic ganglion that receives synaptic input from preganglionic parasympathetic neurons in the brain stem. Postganglionic nerve fibers from the PPG innervate the nasal and palatine mucosae, the extra- and intracranial vasculature, and the lacrimal gland. ⁸  

Nitric oxide (NO) is a diffusible neurotransmitter that mediates a variety of physiological functions. NO is synthesized by nitric oxide synthase (NOS), which has three isoforms: neuronal (nNOS), epithelial (eNOS), and inducible (iNOS). ⁹ Previous studies have reported that nNOS immunoreactivity (IR) is expressed in the PPG and nasal mucosae of rats ⁹ and humans. ¹⁰ Because the lacrimal gland receives dense parasympathetic and some sympathetic innervation in some species, we speculated that nNOS may also be present in the lacrimal gland and that NO might play a role in the neural control of its function. In light of the various physiological functions that NO mediates in the central and peripheral nervous systems and its potential role in the regulation of lacrimal gland secretion, we set out to investigate the expression patterns of nNOS in the normal mouse PPG and lacrimal gland and its relationship to autonomic innervation. 

Choline acetyltransferase (ChAT) and acetylcholinesterase (AChE) are the enzymes that are responsible for the synthesis and breakdown of acetylcholine (ACh), and both ChAT and AChE have been used in attempts to visualize the distribution of cholinergic fibers in tissues. However, there are reports that AChE, as revealed by conventional histochemical and immunohistochemical detection methods, is not restricted to cholinergic neurons ¹¹ and that the synthesis of ChAT is not necessarily restricted to cholinergic neurons. ¹² These reports suggest that neither AChE nor ChAT is an unambiguous marker for cholinergic neurons. 

The vesicular acetylcholine transporter (VAChT) is a proton-dependent transporter that is responsible for packaging ACh into synaptic vesicles, and the immunohistochemical detection of VAChT has emerged as a reliable method for the detection of cholinergic neurons. ¹³ Using a direct double-labeling method in rat, Arvidsson et al. ¹³ found that VAChT and ChAT colocalized in neurons and suggested VAChT as a novel and reliable marker in the autonomic nervous system for cholinergic neurons, which innervate organs such as the salivary and lacrimal glands. Other reports showed that VAChT IR is more sensitive than most ChAT antibodies for the detection of cholinergic terminals ¹⁴ and provides a clearer signal than ChAT labeling. ^{14 15} All these results suggest that VAChT is a highly specific marker for cholinergic somata and nerve fibers. Moreover, although VAChT is present in both the neurons’ somata and terminals, it is concentrated in the terminals and thus better reflects the distribution of cholinergic synapses. ¹⁵  

Tyrosine hydroxylase (TH) is the rate-limiting enzyme responsible for the synthesis of dopamine and has been used as the marker for the sympathetic innervation of the lacrimal gland. Therefore, we have used antisera to nNOS, VAChT, and TH to examine the distribution of these markers in the PPG and lacrimal gland of the mouse. Our results have confirmed that all the neurons in the PPG are positive for VAChT as well as nNOS. Most of the nerve fibers in the glands were positive for VAChT, with many also showing nNOS IR. Some TH-positive fibers were seen in the gland and a few of them were also positive for nNOS. These data suggest that NO could play a role in the regulation of lacrimal gland secretion. 

Part of this article has been presented in abstract form. 

Materials and Methods

Tissue Preparation

Five Swiss Webster (SW) and five C57 female mice were obtained from commercial vendors (Taconic Farms, Germantown, NY, or Charles River, Wilmington, MA). All animals were kept in a 12-hour light–dark cycle and maintained in an accredited animal facility with freely available food and water. They were managed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. To obtain tissues, the animals were killed with an overdose of halothane and decapitated around midday. The lacrimal glands and PPG were removed and placed in freshly prepared 4% buffered paraformaldehyde. After 3 to 4 hours of fixation at 4°C, the tissue was placed in 0.1 M phosphate buffer containing 30% sucrose at pH 7.4 for at least 12 hours at 4°C. The lacrimal gland and ganglia were then placed in optimal cutting temperature (OCT) embedding medium (Sakura Finetek USA, Torrance, CA), serially sectioned at 10 μm with a cryostat (Leica, Deerfield, IL), and collected on slides (Superfrost Plus; VWR Scientific, West Chester, PA). The sections were dried and then stored at −20°C until used. 

Immunohistochemistry

The antibodies used were polyclonal rabbit anti-rat NOS1 (R-20; Santa Cruz Biotechnology, Santa Cruz, CA), at dilutions of 1:400 for lacrimal gland sections and 1:200 for ganglion; and goat anti-VAChT polyclonal antibody (Chemicon International, Temecula, CA) at dilutions of 1:2000 for lacrimal gland and 1:1000 for ganglion. The VAChT antibody has been well characterized and has been shown to colocalize with ChAT in both central and peripheral nervous systems, including lacrimal gland. ^{13 14} The sheep anti-TH polyclonal antibody (Chemicon) was used at a dilution of 1:200. Sections were incubated in primary antibody diluted with 0.1 M sodium phosphate buffer (PBS) overnight or for 48 hours for TH. For control samples, primary antibodies were omitted. Secondary antibodies used were fluorescein isothiocyanate (FITC)–conjugated donkey anti-rabbit and anti-sheep IgG, and Texas red–conjugated donkey anti-goat and anti-sheep IgG (Jackson ImmunoResearch, West Grove, PA), all at a dilution of 1:200. The secondary antibodies were applied for 1 hour. The slides were then washed with three changes of PBS and one change of 4 mM sodium carbonate (pH 10.0), coverslipped, and examined with a conventional fluorescence microscope (Eclipse E800M; Nikon, Melville, NY). The images were captured with a digital camera (Spot; Diagnostics Instruments, Sterling Heights, MI), and analyzed on a desktop computer with image analysis software (PhotoShop; Adobe Systems, Mountain View, CA). Additional images were also obtained with a confocal laser scanning microscope (TCS SP; Leica). At least 21 sections were examined per gland to investigate all lobes within the gland. 

Results

Pterygopalatine Ganglion

In the PPG, all the neurons appeared to display various intensities of nNOS IR. The neurons, which were interspersed among nerve fibers that do not synapse within the PPG, were of various sizes (20–40 μm) and were clearly differentiated from the nerve fibers (Fig. 1A) . nNOS IR appeared to be distributed relatively evenly within the neuronal somata, except that labeling was excluded from the nucleus (Fig. 1B) . However, nNOS IR was weak or undetectable in the axons of these cells in contrast to the easily observed fibers in the lacrimal gland, which is one of the targets of the neurons in the PPG. 

All the neurons in the PPG also displayed VAChT IR. The somata were of different sizes and, except for the nucleus, were filled by the labeling (Fig. 1C) . VAChT IR was also detectable in presynaptic terminals that resembled large baskets. Axons were clearly visible as well but were less intensely labeled. The labeling of both the cell bodies and the axons stood out clearly from the rest of the nerve fibers that course through the ganglion without synapsing. nNOS and VAChT appeared to be colocalized in all the neuronal somata in the PPG. 

Lacrimal Gland

Many nerve fibers in the lacrimal gland displayed nNOS IR, and there were smaller bundles of fibers with numerous varicosities that could be observed around tear-collecting ducts, blood vessels, and acini. 

Large bundles of nerve fibers were labeled, mostly in the interlobular areas, running along with large tear-collecting ducts and blood vessels. Smaller nerve bundles and fibers were observed to branch from the larger ones. Some nerve fibers were observed to be closely associated with collecting ducts, and some appeared to encircle the outer surface of ducts (Fig. 2A) . Some punctate nNOS IR was also observed in association with acinar and ductal cells. Some nerve fibers with prominent varicosities appeared to be in close proximity to the acini (Figs. 2A 2B 2E) . Although quantification was not performed in the present study, examination of serial sections indicated that approximately 80% to 90% of the acini exhibited either punctate nNOS-immunoreactive staining or were in close contact with nNOS-positive nerve fibers. 

Dense VAChT IR was found in the lacrimal gland. Numerous varicosities were observed among acini, presumably representing synaptic vesicles containing ACh (Fig. 2C) . These VAChT-positive varicosities were close to the basolateral membrane of the acini. They formed a loose network around the acini, and each nerve fiber appeared to be in close association with more than one acinar cell. 

Some nerve fibers were also TH immunoreactive. However, compared with the dense VAChT-positive varicosities in the gland, many fewer TH-positive varicosities were observed. Most of the TH IR was seen in association with the blood vessels, which are known to be innervated by sympathetic neurons (Fig. 2D) . Some of the labeled nerves were observed to course among acini, and nerve fibers with many varicosities were occasionally seen in close association with acini (Fig. 2F) . 

nNOS IR was present in many but not all the fibers that displayed VAChT IR (Figs. 2B 2C) . Also, some nNOS-positive fibers that did not display VAChT IR were observed. However, in most of the fibers nNOS IR appeared to colocalize with VAChT IR. nNOS IR was detectable in some, but not all, of the sparse TH-positive fibers that coursed among the acini. However, no nNOS IR was observed to colocalize with TH IR in the inner layer of blood vessel walls. 

Discussion

The PPG is a parasympathetic relay center and receives its inputs from the preganglionic parasympathetic neurons in the brain stem through the greater superficial petrosal nerve. These preganglionic fibers then synapse with the neurons in the ganglion and the postganglionic axons project to target organs, such as the lacrimal gland and nasal mucosae. 

The present study demonstrated that virtually all the neurons in the mouse PPG were VAChT immunoreactive. To our knowledge, this is the first description of VAChT IR in the neurons of the PPG in mouse. Our results are consistent with those of other investigators who used ChAT or AChE as the cholinergic marker and found that virtually all the PPG neurons in rat, ¹⁶ chicken, ¹⁷ and pigeon ¹⁸ display ChAT or AChE IR. Also, the present study demonstrated that most of the neurons in the PPG were nNOS immunoreactive. This is in agreement with other reports in monkey, ¹⁹ human, ^{9 19} mouse, ²⁰ rat, ¹⁰ pigeon, ¹⁸ and cat. ²¹ The neuronal profiles in the mouse PPG varied in size and labeling intensity for both VAChT and nNOS. However, there was no apparent systematic relationship between the neurons’ somata size and the observed labeling intensity. This result is consistent with the observations in the human PPG. ⁹  

An interesting observation in the present study was the apparent difference in the labeling patterns of VAChT and nNOS in the portion of the axons within PPG (Figs. 1B 1C) . In the case of VAChT IR, even though the labeling in the somata was more intense than that in axons, the axons also displayed clear VAChT IR. However, nNOS IR was virtually undetectable in the axons, even though the somata were clearly nNOS positive. This observation is in contrast to reports in pigeon ¹⁸ and rat, ¹⁰ which have demonstrated that the postganglionic axons in the PPG are also nNOS immunoreactive. Because most of the parasympathetic nerves in the lacrimal gland originate from the PPG and many of them were immunoreactive to nNOS, and because all the neurons in the PPG showed both VAChT and nNOS IR, it would be reasonable to expect nNOS to be detectable in the axons that extend from the neurons to the lacrimal gland. However, it is possible that NO synthesized in the postganglionic somata serves to modulate synaptic transmission within the ganglion and that in the terminals in the gland it modulates secretion. That nNOS was undetectable in the portions of the axons within the PPG may indicate that nNOS, which is being transported to the terminal, is present in quantities below the detection threshold for the methods used. 

The presence of NOS in the preganglionic ²² and postganglionic PPG neurons suggests that NO may play a role in neurotransmission within the PPG. Retrograde labeling studies of the preganglionic parasympathetic neurons projecting to the PPG in rabbit demonstrated that although all the retrogradely labeled neurons in the brain stem displayed ChAT IR, only 75% were NOS positive, although the NOS isoform was not specified. ²² Our observations further demonstrated nNOS in the postganglionic neurons within PPG colocalized with VAChT in all the neurons’ somata. Studies in Torpedo synaptosomes indicated that NO may decrease ACh release, ²³ and other reports have shown that NO can act as a neuromodulator in modulating synaptic transmission at several types of synapses. ^{24 25 26 27} There are thus two potential sources of NO that may play a role in modulating cholinergic transmission within the PPG: the preganglionic neuronal terminals and the postganglionic somata. 

Although there are reports of NOS distribution in the PPG and some of its targets, such as nasal mucosae, ^{10 20} sinus mucosa, ²⁸ and cranial blood vessels, ²¹ to our knowledge, this is the first report of nNOS IR in postganglionic neurons of the PPG and nerve fibers in the lacrimal gland. 

It has been established that the mammalian lacrimal gland is densely innervated by the parasympathetic nervous system. ^{1 2 29} Our observations, based on VAChT as a parasympathetic marker, confirmed that the mouse lacrimal gland was densely innervated by VAChT-positive nerve fibers. The small nerve bundles and fibers that were VAChT immunoreactive appeared to be found only at the base of the acinar cells. They were never seen to project into the acini or between acinar cells. These observations were consistent with findings in cat. ⁴ The VAChT-positive nerve fibers included large and small nerve bundles that coursed along the tear-collecting ducts and apparent nerve fibers that formed a dense network. Every acinar cell seemed to be in close proximity to a VAChT-positive nerve fiber. 

In contrast to the dense distribution of parasympathetic nerves in the lacrimal gland, only sparse TH IR nerve fibers were detectable in the gland among the secretory acini, although more TH IR was present in the inner layer of interlobular blood vessels. The sympathetic innervation of the lacrimal gland varies among species, and there are some discrepancies between the observations of different investigators. For example, there are reports that every acinar cell is in close proximity to an adrenergic fiber in cat, ⁴ dog, ^{5 7} human, and monkey. ⁶ However, there are reports of findings in other studies in humans ³⁰ and in monkeys ³¹ that are inconsistent with these findings. In rat ^{7 32} and mouse, ⁷ only very sparse TH-positive nerve fibers have been described, and most of them were in association with the blood vessels, whereas only a few were found among acini. In guinea pig, some researchers reported rich adrenergic innervation in lacrimal glands, ³² whereas others described a virtual absence in the gland except along the blood vessels. ⁷ Our observations appeared to be in agreement with a previous report in the mouse. ⁷  

Many of the VAChT-positive, and some of the TH-positive, nerve fibers in the lacrimal gland also appeared to exhibit nNOS IR. Although the role of NO in the lacrimal gland is not known, the colocalizations of nNOS with VAChT and with TH, in addition to the highly diffusive nature of NO, support the notion that there may be some interaction between NO and other autonomic transmitter systems within the lacrimal gland. For example, as discussed with respect to the PPG, NO could exert its effects by influencing ACh transmission. 

The observation of TH-immunoreactive nerve fibers coursing among the acini suggest that sympathetic nerves may play a direct role in modulating tear secretion, in addition to regulating blood flow within the lacrimal gland. The presence of nNOS IR in some of the TH-immunoreactive fibers raises the possibility that there may be some interaction between these two systems as well. 

There are five targets that could be influenced by NO at the lacrimal gland level: ducts, blood vessels, acini, nerve fibers, and myoepithelial cells. Many of the nNOS-positive nerve bundles ran alongside the ducts in the interlobular areas and some of the smaller nerve fibers and their varicosities appeared to be in close association with the ducts (Fig. 2A) . Physiologically, NO is a potent vasoactive agent ^{33 34 35} and can readily diffuse across membranes and mediate vasodilation within vascular smooth muscle. ^{36 37 38} Thus, the blood vessels in the lacrimal gland, which are primarily innervated by sympathetic nerves and supplemented by some parasympathetic ones, ³⁹ might be modulated by NO. This is significant, because in the cat and rabbit, vasodilation in the lacrimal gland correlated with tear flow. ^{40 41} Other studies have also demonstrated that the blood flow in the submandibular gland of cats could influence secretory function. ^{42 43}  

Ultrastructural studies have demonstrated that nerve terminals in the monkey lacrimal gland are in close proximity to acinar cells. ⁴⁴ In the present study, numerous parasympathetic and fewer sympathetic fibers were observed among the acini, and many of the parasympathetic and some sympathetic fibers exhibited nNOS labeling. It is possible, therefore, that NO may influence tear production from the acini, which are the functional units of the lacrimal gland. Based on the fact that most nerve fibers have nNOS and that NO is a highly diffusive molecule and can readily spread to adjacent areas, it is reasonable to speculate that NO could spread and exert its influence on most or perhaps all the acini. 

Although little is known about the functions of myoepithelial cells, it has been suggested that they could be involved in the release of secretory product by contraction. However, others have proposed that these cells’ only role is to maintain the contour of the glandular end pieces, serving as the exoskeleton of the acini. ⁴⁵ It has been shown that these cells are attached to the acini in a lacy arrangement, that they contain a network of α-smooth muscle actin that is similar to that of smooth muscle, ^{46 47} and that M₃ muscarinic receptors are expressed on their surfaces. ⁴⁷ In that NO has been reported to influence cholinergic neurotransmission ²³ and muscular tone, ^{36 37 38} it can be speculated that NO may also exert some influence on the myoepithelial cells. The exact mechanism and to what extent these cells are involved are unknown. 

Functional studies in the salivary glands, which have some similarities to lacrimal glands anatomically and functionally, have shown that NO can modulate protein, electrolytes, and water production in salivary secretion. ^{48 49 50 51 52} NO may influence tear secretion through various mechanisms, including control of cGMP production, ^{36 37 38 53} modulation of ion channels, ^{54 55} activation of intracellular signaling pathways, ⁵⁶ and regulation of Na,K- ATPase. ⁵⁷  

It must be emphasized that it is still unclear what role NO plays in the lacrimal gland, although research on salivary glands has shown that NO plays a role in salivary secretion. Another point that should be made is that because our observations were made in the course of a study of Sjögren’s syndrome, we used only female mice, and the findings reported herein may not be applicable to male mice. 

In summary, nNOS was expressed in both the neurons in the mouse PPG and nerve fibers in the lacrimal gland. In the PPG, nNOS was present in all the VAChT-positive neurons, and many of the parasympathetic and some sympathetic nerves in the lacrimal gland colocalized with the nNOS-positive nerves. These observations support the notion that NO may play a role in modulating tear production through various mechanisms. These observations illustrate the complexity of both the innervation pattern of the mouse lacrimal gland and of the mechanisms by which autonomic input may influence the secretory process. 

 Supported by Grants EY0940607 (BW), P30 EY03039 (KTK), and EY07845 (KTK) from the National Eye Institute.

 Submitted for publication March 2, 2001; revised June 18, 2001; accepted July 18, 2001.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked“ advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

VAChT and nNOS labeling in the PPG. (A) VAChT visualized with FITC was observed in the somata, with presynaptic terminals shown as large puncta (arrow) and VAChT IR in the postganglionic somata appeared uniformly throughout the cytoplasm with some small puncta visible (arrowhead). (B) nNOS-FITC was expressed in all the somata, but virtually no nNOS IR could be detected in the axons. (C) VAChT-positive Texas red was detected in both somata and axons, although the intensity in axons appeared weaker than in the somata (arrows). In the somata, nNOS IR and VAChT IR was colocalized. (B) and (C) are from the confocal microscope. Scale bar, 30 μm.

Figure 2.

 

nNOS, VAChT, and TH IR in the lacrimal gland. (A) nNOS-immunoreactive nerve bundles (arrow) and fibers, were observed in the gland. Some nerve fibers with varicosities were in close proximity to acini. Immunoreactive fibers were observed to encircle the ducts (arrowhead). Some punctate nNOS labeling was found to be associated with both acinar and ductal cells. (B, C) Confocal microscope images double labeled for nNOS and VAChT of lacrimal gland. (B) Some nNOS punctate labeling was associated with acinar cells. Arrow: nerve bundle; arrowhead: several varicosities. (C) VAChT IR was observed in many of the nNOS-immunoreactive fibers and varicosities. (D) TH immunoreactive nerve fibers were observed to be in association with blood vessels. (E, F) Nerve fibers that were nNOS immunoreactive (E) also displayed TH IR (F). Although these labels appeared to be colocalized, the possibility cannot be ruled out that the labeling represents two different fibers. Scale bar, 30 μm.