THE ANATOMICAL RECORD 295:141–149 (2012)
Reconsideration of the Autonomic Cranial Ganglia: An Immunohistochemical Study of Mid-Term Human Fetuses HIROMICHI KIYOKAWA,1 YUKIO KATORI,1,2 KWANG HO CHO,3* GEN MURAKAMI,4 TETSUAKI KAWASE,1 AND BAIK HWAN CHO5 1 Department of Otorhinolaryngology, Tohoku University School of Medicine, Sendai, Japan 2 Division of Otorhinolaryngology, Sendai Municipal Hospital, Sendai, Japan 3 Department of Neurology, Wonkwang University School of Medicine, Jeonbuk Regional Cardiocerebrovascular Disease Center, Institute of Wonkwang Medical Science, Iksan, Korea 4 Division of Internal Medicine, Iwamizawa Koujin-kai Hospital, Iwamizawa, Japan 5 Department of Surgery, Chonbuk National University School of Medicine, Jeonju, Korea
ABSTRACT The cranial parasympathetic ganglia have been reported to paradoxically contain the sympathetic nerve marker, tyrosine hydroxylase (TH), in addition to neurons expressing parasympathetic markers such as vasoactive intestinal peptide (VIP) and neuronal nitric oxide synthase (nNOS). However, the distribution of these molecules in the cranial ganglia of human fetuses has not yet been examined. Using parafﬁn sections from 10 mid-term human fetuses (12–15 weeks), we performed immunohistochemistry for TH, VIP, and nNOS in the parasympathetic ciliary, pterygopalatine, otic, and submandibular ganglia, and for comparison, the sensory inferior vagal ganglion. The ciliary and submandibular ganglia contained abundant TH-positive neurons. In the former, TH-positive neurons were much more numerous than nNOS-positive neurons, whereas in the latter, nNOS immunoreactivity was extremely strong. No or a few cells in the pterygopalatine, otic, and inferior vagal ganglia expressed TH. Ciliary TH neurons appeared to compensate for classically described sympathetic ﬁbers arising from the superior cervical ganglion, whereas in the submandibular ganglion, nNOS-positive neurons as well as TH neurons might innervate the lingual artery in addition to the salivary glands. Signiﬁcant individual variations in the density of all these markers suggested differences in sensitivity to medicine affecting autonomic nerve function. Consequently, in the human cranial autonomic ganglia, it appears that there is no simple dichotomy between sympathetic and parasympathetic C 2011 Wiley Periodicals, Inc. function. Anat Rec, 295:141–149, 2012. V
Key words: tyrosine hydroxylase; neuronal nitric oxide synthase; submandibular ganglion; ciliary ganglion; pterygopalatine ganglion; otic ganglion; human fetus
Contract grant sponsor: Wonkwang University in 2009. *Correspondence to: Kwang Ho Cho, Department of Neurology, Wonkwang University School of Medicine, Jeonbuk Regional Cardiocerebrovascular Disease Center, Institute of Wonkwang Medical Science, 344-2, Sinyong-dong, Iksan, Jeonbuk, 570-711, Republic of Korea. E-mail: [email protected]
, [email protected]
C 2011 WILEY PERIODICALS, INC. V
Received 29 June 2011; Accepted 4 September 2011 DOI 10.1002/ar.21516 Published online 18 November 2011 in Wiley Online Library (wileyonlinelibrary.com).
KIYOKAWA ET AL.
In the human head and neck, the ciliary, pterygopalatine, otic, and submandibular ganglia (CG, PG, OG, SG) are known as the parasympathetic ganglia of which efferent ﬁbers play secretomotor functions, while the nodosa or inferior vagal ganglion (IVG) as one of sensory ganglia (O’Rahilly and Muller, 1996; Berkovitz, 2005). However, the IVG cells contain the same neurotransmitters or proteins as seen in the parasympathetic ganglia (Alm et al., 1995). Tyrosine hydroxylase (TH) is a rate-limiting enzyme of catecholamine synthesis in sympathetic ganglion cells, whereas vasoactive intestinal peptide (VIP) and neuronal nitric oxide synthase (nNOS) are well known markers of parasympathetic neurons. Peptide histidine isoleucine (PHI) is also a major marker of parasympathetic nerves, and is colocalized with VIP (Itoh et al., 1983). Nitric oxide synthase (NOS) is a key enzyme in the production of nitric oxide, a molecule that directly regulates vasorelaxation and blood supply. In humans, nNOS reactivity has been investigated, particularly in the pelvic nerves (e.g., Costello et al., 2011; Hisasue et al., 2010), but the previous descriptions were limited for the human adult cranial ganglia (Tajti et al., 1999; Uddman et al., 1999; May et al., 2004; Rusu and Pop, 2010). To our knowledge, no reports are available for immunohistochemical information of human fetal cranial ganglia. Several reports have documented the paradoxical presence of these markers in the human autonomic nervous system, including VIP-positive neurons in the prevertebral sympathetic ganglia (Roudenok et al., 1999) and TH-positive neurons in the CG (Kirch et al., 1995; Thakker et al., 2008). Likewise, in some ganglia of the human pelvic plexus, TH-positive neurons co-exist with VIP/PHI-positive ones, and both are intermingled without any clear demarcation (Takenaka et al., 2005; Imai et al., 2006). Such colocalization of sympathetic and parasympathetic markers in an autonomic ganglion has also been reported in experimental animals, and the distribution of these molecules appears to vary among animal species, e.g., some SG cells in the rat express TH, whereas those in the monkey do not (Ng et al., 1995). However, various research groups have reported different ﬁndings: for example, with regard to the CG, May et al. (2004) found no TH-positive neurons in humans aged 12–95 years, whereas VIP or nNOS was expressed in almost all neurons, and Tan et al. (1995) found that 35.3% of neurons showed moderate or strong expression of TH in the monkey, whereas only 6.6% did so in the cat. Early establishment of the ganglion cell phenotype is known to occur during prenatal development (Black, 1982; Levitt et al., 1985). However, plasticity of neurotransmitter function in some central and peripheral neuronal populations has been suggested (LeDouarin, 1978; Teitelman et al., 1985). In this context, a comprehensive immunohistochemical study of human fetal cranial ganglia would seem to provide the best understanding of sympathetic/parasympathetic status in the human cranial autonomic nerves. The aim of this preliminary study, therefore, was to clarify the distribution of immunoreactivity for TH, VIP/PHI, and nNOS in the parasympathetic CG, PG, OG, and SG and, for comparison, the sensory IVG in mid-term human fetuses.
MATERIALS AND METHODS The study was performed in accordance with the provisions of the Declaration of Helsinki 1995 (as revised in Edinburgh 2000). We examined 10 mid-term fetuses, comprising ﬁve fetuses at 12 weeks of gestation and ﬁve fetuses at 15 weeks. With the agreement of the families concerned, these specimens had been donated to the Department of Anatomy of Chonbuk National University, Korea, and their use for research had been approved by the University Ethics Committee. All the fetuses had been obtained by induced abortion. After abortion, each mother had been personally informed by an obstetrician about the possibility of fetal donation for research: no attempt had been made to actively encourage donation. Because of randomization of the specimen numbering, it was not possible to trace any of the families concerned. Because of miscarriages and ectopic pregnancies, we could not rule out the possibility that the specimens contained pathology. However, no pathology was evident in the developing umbilical vessels, liver, intestine, adrenal, or kidney of the specimens examined. The donated fetuses were ﬁxed with neutral 10% w/w formalin solution for more than 3 months. Before dehydration, the whole head and neck of each specimen were decalciﬁed using EDTA (0.5 mol/L, pH, 7.5). The time for decalciﬁcation depended on sizes of the specimens (1–3 days at 4 C). Horizontal or frontal parafﬁn sections of the whole head and neck were cut at a thickness of 5 micron, at intervals of 50 micron. Most of the sections were stained with hematoxylin and eosin (HE), while some (5–6 per each stage fetus) were subjected to immunohistochemical staining. The primary antibodies used for immunohistochemistry were (1) rabbit polyclonal anti-human TH (1:100 dilution; Chemicon, Temecula, CA), (2) rabbit polyclonal anti-human VIP (1:100 dilution; Yanaihara Institute, Fujinomiya, Japan), (3) rabbit polyclonal antihuman PHI (1:100 dilution; Yanaihara Institute, Fujinomiya, Japan), and (4) rabbit polyclonal anti-human nNOS (Cell Signaling Technology, Beverly, MA). The secondary antibody was labeled with horseradish peroxidase (HRP), and antigen–antibody reactions were detected by the HRP-catalyzed reaction with diaminobenzidine. Counterstaining with hematoxylin was performed on the same samples. The negative control without a ﬁrst antibody was set up for each of the specimens (an insert in each of Figs. 1–5). The counting and measurement of positive cells were performed using one of good sections under 20 or 40 objective (Olympus CX21, Tokyo, Japan).
RESULTS We found the CG along or in the anterior side of the orbital muscle (a smooth muscle band at the inferoposterior margin of the orbit); the PG in the immediately posterior side of the orbital muscle; the OG in the immediately lateral side of the developing pterygoid process; the SG at the medial turn of the lingual nerve in the deep side of the submandibular gland; and the IVG in the lateral and inferior side of the superior cervical ganglion. These landmarks for identiﬁcation had corresponded to the subjects (orbital muscle, pterygoid process) or key structures (lingual nerve, submandibular
RECONSIDERATION OF THE AUTONOMIC CRANIAL GANGLIA
TABLE 1. Summary of the present observations
Specimen 1 (15 weeks; Fig. 1) Specimen 2 (15 weeks; Fig. 2) Specimen 3 (15 weeks) Specimen 4 (15 weeks) Specimen 5 (15 weeks) Specimen 6 (12 weeks; Fig. 3) Specimen 7 (12 weeks; Fig. 4) Specimen 8 (12 weeks; Fig. 5) Specimen 9 (12 weeks) Specimen 10 (12 weeks)
TH VIP/PHI TH nNOS TH VIP/PHI TH nNOS TH, nNOS TH nNOS TH nNOS TH nNOS TH TH, nNOS
þþþ þþþþþ þ þþ þþþþ þþþ ND þþ þ þþ ND þþ ND
þ þþþþþ þþ þþþþ ND ND þþ ND þþ þ ND ND
þ ND þ þþ ND þ ND þ ND þþ þ ND ND
þþþ þþþ ND ND þþþ þþþ þþþ þþ ND þ þþ ND þþþþ þþ þþþþ þþ ND
þ þþþþþ ND ND ND ND ND þ ND ND þ ND þ ND
CG, ciliary ganglion; PG, pterygopalatine ganglion; OG, otic ganglion; SG, submandibular ganglion; IVG, inferior vagal ganglion. ND: no deﬁnite ﬁnding because of failure of the immunostaining. Evaluation of the percentage of positive cells: none (), few (þ), moderate (þþ), many (þþþ), almost half (þþþþ), and more than half (þþþþþ).
gland, superior cervical ganglion, digastric muscle) in our recent studies (Miyake et al., 2010; Katori et al., 2011a,b,c; Osanai et al., 2011).
Individual Difference in Immunoreactivity
nNOS reactivity was not seen in the CG even though positive staining was present in the other ganglia of the same specimens (Figs. 2B, 3B). We observed relatively small nNOS-positive cells out side of the CG (Fig. 4A). The CG contained abundant TH-positive cells. However, the proportion of the positive cells per section varied among specimens: a few (Fig. 2A) or many in numbers (Fig. 1A). Because semiserial sections had been prepared at 50-micron intervals, we were unable to identify whether any cells were double-positive for both sympathetic and parasympathetic markers. TH-positive nerve ﬁbers consistently (8/8) ran through or passed near the CG (Figs. 2A, 3A, 5A). The oculomotor and abducens nerves contained a few TH-positive ﬁbers, but the optic nerve did not.
At any ganglion examined, both ganglion cells and nerve ﬁbers expressed TH reactivity in 8 of the 10 specimens (four each at 12 and 15 weeks), whereas in the other two, no positive cells were seen even in the cervical sympathetic trunk ganglia (Table 1). The TH-positive nerve ﬁbers were evident at multiple sites in the head and neck, including close to and within the cranial ganglia (for details, see below), making them a good indicator of whether immunohistochemical preparation had been good or poor in any particular section. Positive staining for VIP or PHI was seen in 2 of the 10 specimens (both at 15 weeks) and for nNOS in 4 of the 10 (one at 15 weeks, and three at 12 weeks), while the other specimens showed no reactivity (Table 1). VIP or PHI immunostaining was accompanied by marked nonspeciﬁc positive staining, especially in striated muscle. In ﬁve of the eight specimens with TH-positive nerve ﬁbers, VIP or nNOS was also expressed in most or some cells in the cranial ganglia (three at 12 weeks; two at 15 weeks): VIP reactivity was identiﬁed in only one fetus at 15 weeks, and nNOS reactivity in another 4. Figures 1– 5 show all of the ﬁve specimens in which results for both sympathetic and parasympathetic markers were available. However, even in these ﬁve specimens, some ganglia were not found because they had been damaged or washed away during the histological procedure (Table 1). Irrespective of TH, VIP/PHI or nNOS, those immunoreactive ganglion cells took a similar size, but nNOSpositve cells in the SG tended to be small.
The PG contained a few TH-positive ganglion cells in 2 of the specimens (Figs. 1C, 5A; 2/8). In the other six specimens, TH-positive neurons were absent but TH-positive nerve ﬁbers passed through the PG and exited to the palatine nerves (Figs. 1C, 2C, 3C). In contrast to diffuse and strong reactivity for VIP (Fig. 1D), weak nNOS reactivity was seen in a moderate number of the ganglion cells (Figs. 2D, 3D, 4B), although diffuse nonspeciﬁc staining was also evident. The immunoreactivity of the OG was similar to that of the PG with the nNOS reactivity being evident in a small or moderate number of ganglion cells (Figs. 2F, 3F, 4C). In two specimens (12 and 15 weeks), the OG carried a few cells with weak TH reactivity, although the nerve ﬁbers showed stronger positivity near the OG (Figs. 2E, 5D). The mandibular nerve contained no or only a few TH-positive ﬁbers.
Ciliary Ganglion or CG
Submandibular Ganglion or SG
Reactivity for the parasympathetic marker VIP was seen in most or some of the CG cells (Fig. 1B). However,
The SG contained both TH-positive and nNOS (or VIP)positive ganglion cells (Figs. 1EF, 3GH, 5BC; Table 1): the
Pterygopalatine Ganglion (PG) and Otic Ganglion (OG)
KIYOKAWA ET AL.
Fig. 1. Immunohistochemstry of TH and VIP using a 15-week fetus. Specimen No.1 in Table 1. Horizontal sections. The anterior and medial directions are indicated by arrow with ‘‘ant’’ or ‘‘med,’’ respectively. Panel A (TH) and panel B (VIP) display the near sections including the ciliary ganglion (star). Panel C (TH) and panel D (VIP) show the near sections including the pterygopalatine ganglion (star). Arrows in panel C indicate candidates of TH-positive neurons. An insert in panel D is the negative control without the ﬁrst antibody. Panel E (TH) and panel F (VIP) show the near sections including the submandibular ganglion (star). Panel G (TH) exhibits the otic ganglion and TH-positive
ﬁbers (arrows) along the internal carotid artery (ICA). Panel H (TH) and panel I (VIP) show the near sections including the inferior vagal ganglion (star) and the superior cervical ganglion (SCG). Arrows in panel H indicate candidates of TH-positive neurons. Panels A–F, H and I were prepared at the same magniﬁcation (scale bar in panel A), while panel G at the lower magniﬁcation. III, oculomotor nerve; V2, maxillary nerve; V3, mandibular nerve; VI, abducens nerve; LN, lingual nerve; LR, lateral rectus muscle; MH, mylohyoideus muscle; SMG, submandibular gland.
proportion was similar in two specimens at 15 weeks, but in specimens at 12 weeks, TH-positive cells showed a proportion smaller than nNOS-positive cells. Notably, in the head and neck, nNOS reactivity was strongest in and around the SG: in addition to ganglion cells, nerve ﬁbers in and around the SG were also weakly positive (Figs. 3H, 4D). However, we were not able to trace the ﬁbers to any likely target. VIP reactivity was also found in some parts of the submandibular gland (Fig. 1F).
Inferior Vagal Ganglion or IVG The IVG contained abundant VIP/PHI-positive neurons (Fig. 1i) and a few TH-positive ganglion cells (Figs. 1H, 5E). Unexpectedly, we found no nNOS reactivity in the IVG (Figs. 3J, 4E). Near the IVG, the cervical sympathetic trunk as well as nerve ﬁbers around the carotid and facial arteries showed strong TH expression (Figs. 1H, 3i, 5E). Some of the latter ﬁbers ran superiorly into the cranial cavity along the internal carotid
RECONSIDERATION OF THE AUTONOMIC CRANIAL GANGLIA
Fig. 2. Immunohistochemstry of TH and nNOS using a 15-week fetus. Specimen No.2 in Table 1. Frontal sections. The anterior and superior directions are indicated by arrow with ‘‘ant’’ or ‘‘med,’’ respectively. The left-hand side column displays TH immunostaning, while the right-hand side nNOS. Panels A and B, the ciliary ganglion; panels C and D, the pterygopalatine ganglion; panels E and F, otic
ganglion and the negative control without the ﬁrst antibody (insert of panel F). Arrows (Arrowheads) indicate candidates of TH-positive neurons (nerve ﬁbers). All panels were prepared at the same magniﬁcation (scale bar in panel A). II, optic nerve. We failed to stain the submandibular and inferior vagal ganglia of this specimen (Table 1).
artery (Figs. 1G, 5D). In addition, the superior cervical ganglion of the sympathetic nerve trunk contained VIPpositive cells, although most of the reactivity appeared to be nonspeciﬁc (Fig. 1i). Consequently, the CG and SG contained TH-positive neurons with various proportions, whereas nNOS immunoreactivity was extremely strong in the SG. Strangely, we found no nNOS reactivity in the IVG. Signiﬁcant individual variation in the expression of all of these molecules was evident.
nerve-associated marker molecules using parafﬁnembedded fetal specimens because of the signiﬁcant variation in the level of reactivity or the proportion of positive cells. In fact our present results may have reﬂected individual postmortem and/or in utero conditions. Taguchi et al. (1999) reported that adult cadavers of individuals with a history of diabetes showed weak reactivity for cholinesterase in the pelvic nerve plexus. Moreover, previous studies of human autonomic nerve ganglia were based on immunostaining using ‘‘cryostat’’ sections: e.g., Kirch et al. (1995) and Thakker et al. (2008) have demonstrated TH-positive cells in the adult human CG, and using materials obtained from elderly individuals, Tajti et al. (1999) and Uddman et al. (1999) have found VIP-positive neurons in the superior cervical ganglion or the OG and PG, respectively. However, immunohistochemistry using cryostat sections of human fetal materials are very difﬁcult
DISCUSSION This study demonstrated immunoreactivity for TH, VIP/PHI, and nNOS in all of the cranial parasympathetic ganglia in human fetuses. First it seems appropriate to discuss the difﬁculty in immunostaining of these
KIYOKAWA ET AL.
Fig. 4. Immunohistochemstry of nNOS using a 12-week fetus. Specimen No.7 in Table 1. Horizontal sections. The anterior and medial directions are indicated by arrow with ‘‘ant’’ or ‘‘med,’’ respectively. Panel A, the ciliary ganglion (stars). The nNOS-positive cells (arrows in panel A) at the anterior margin of the ganglion are bit smaller than usual ganglion cells; panel B, the pterygopalatine ganglion. Diffuse and weak reactivity appeared to be non-speciﬁc; panel C, otic ganglion and the negative control without the ﬁrst antibody (insert of panel C); panel D, the submandibular ganglion; panel E, the inferior vagal ganglion (negative). All panels were prepared at the same magniﬁcation (scale bar in panel C). II, optic nerve; V2, maxillary nerve; V3, mandibular nerve; SMG, submandibular gland. Fig. 3. Immunohistochemstry of TH and nNOS using a 12-week fetus. Specimen No.6 in Table 1. Horizontal sections. The anterior and medial directions are indicated by arrow with ‘‘ant’’ or ‘‘med,’’ respectively. The left-hand side column displays TH immunostaning, while the right-hand side nNOS. Panels A and B, the ciliary ganglion (star); panels C and D, the pterygopalatine ganglion (star) and the negative control without the ﬁrst antibody (insert of panel D). The nerve of the pterygoid canal (NPC) contains abundant TH-positive ﬁbers (panel C); panels E and F, otic ganglion (star); panels G and H, the submandibular ganglion (star); panels I and J, the inferior vagal ganglion (star) and the superior cervical ganglion (SCG). Arrowheads in panel A indicate candidates of TH-positive ﬁbers. Arrows (or Arrowheads) in panel G indicate candidates of TH-positive neurons (nerve ﬁbers). Panels A–H were prepared at the same magniﬁcation (scale bar in panel G), while panels I and J at the lower magniﬁcation. II, optic nerve; III, oculomotor nerve; V2, maxillary nerve; V3, mandibular nerve; SMG, submandibular gland.
according to our experience (unpublished data). Nevertheless, in the present data, we were able to compare the results obtained from several ganglia in a single specimen because of uniformity of the preparation conditions. Even in fetal parafﬁn sections, TH immunohistochemistry was usually (8/10) conducted successfully and even nerve
ﬁbers were stained. However, in the present series, it was difﬁcult to account for the lack of nNOS reactivity in the IVG although it is sensory. In the ﬁnal paragraph, we will return to the issue of individual variation in immunohistochemical expression. In contrast to the PG, OG and IVG, the CG and SG contained abundant TH-positive cells. According to Crouse and Cucinotta (1965), SG cells show a progressive reduction in number from 17 weeks until full term. Thus, we cannot rule out a possibility that, in the CG and SG, TH-positive neurons are eliminated in the late fetal stage. Why do the CG and SG express TH, in contrast to the other three ganglia? Rohrer and colleagues (Ernsberger et al., 1989; Saadat et al., 1989) have found a ‘‘ciliary neurotrophic factor’’ in embryonic chick and rat that inhibits proliferation and differentiation of sympathetic neurons to control the levels of choline acetyltransferase and TH. The density of this factor (a 22.5kDa molecule) is highest in the eyeball (Adler et al., 1979), but it can also be extracted from peripheral nerves (Ernsberger et al., 1989). Using in situ hybridization as well as immunohistochemistry, Tyrrell et al.
RECONSIDERATION OF THE AUTONOMIC CRANIAL GANGLIA
Fig. 5. Immunohistochemstry of TH and nNOS using a 12-week fetus. Specimen No.8 in Table 1. Frontal sections. The anterior, superior, and medial directions are indicated by arrow with ‘‘ant,’’ ‘‘sup,’’ or ‘‘med,’’ respectively. Panel A (TH), the ciliary ganglion (star), and the pterygopalatine ganglion (asterisk); panel B (TH); and panel C (nNOS), the submandibular ganglion (star); panel D (TH), otic ganglion (star); panel E (TH), the inferior vagal ganglion (star), and the superior cervical ganglion (SCG). An insert in panel C is the negative control without the ﬁrst antibody (a near section of panel C). Arrowheads in panel A indicate candidates of TH-positive ﬁbers. Arrows in panels A, D, and E
indicate candidates of TH-positive neurons. Panels A, B, D, and E were prepared at the same magniﬁcation (scale bar in panel A), while panel C at the higher magniﬁcation. II, optic nerve; III, oculomotor nerve; V2, maxillary nerve; V3, mandibular nerve; VI, abducens nerve; DP, digastric muscle posterior belly; ECA, external carotid artery; GG, genioglossus muscle; ICA, internal carotid artery; IO, infra-orbital nerve; LR, lateral rectus muscle; MC, Meckel’s cartilage; PTT, pharyngotympanic tube; RC, Reichert’s cartilage; SMG, submandibular gland.
(1992) demonstrated that ciliary TH-positive neurons were increased in number in rats that had been sympathectomized at birth. Therefore, they are likely to compensate for the major sympathetic nerves projecting from the superior cervical ganglion. On the other hand, we have found no reports of previous studies focusing on the mechanisms of development of submandibular TH-positive neurons. Ibanez et al. (2010) have reported the coexistence of adrenergic and cholinergic neurons in the human aberrant laryngeal ganglion, near the SG. Such colocalization is very similar to that in the human pelvic ganglion plexus, where the ratio of adrenergic to cholinergic neurons is approximately 6:4 (Takenaka et al., 2005; Imai et al., 2006), and may occur at the oral and anal ends of the alimentary canal, through mechanisms that are still unexplained. Henrich et al (2003) reported dense nNOS-positive nerves supply to the guinea pig lingual artery. The SG may issue nNOS positive nerves to the tongue although, according to our experiences, commercially obtained
antibodies for nNOS are not available for identiﬁcation of nerve ﬁbers in human body. Rusu et al. (2011) demonstrated a difference in discrimination between the human mast cell and ganglion cell in immunohistochemistry of nNOS. Actually, we also observed mast cell-like nNOS-positive cells near a CG (Fig. 4A). However, the abundant positive cells in the SG were unlikely to be the mast cell. Dichotomy between sympathetic and parasympathetic ganglia in the human body as well as in experimental animals seems more complex than previously envisaged. Uemura et al. (1987) appear to have provided the ﬁrst report of ciliary TH-positive neurons in the monkey, dog, cat and rat. Although human anatomy was not included, Hardebo et al. (1992) concluded that (1) combined localization of dopamine-beta-hydroxylase and neuropeptide Y or (2) the presence of TH was not an exclusive marker of peripheral adrenergic neurons, and discussed whether the presence of enzymes and peptides represents a ‘‘remnant’’ of different expression during ontogenesis, or
KIYOKAWA ET AL.
target-speciﬁc functions in the adult. Another factor that runs counter to the classical concept of the sympathetic nerve system is found in nerves innervating striated muscle and muscle-feeding arteries (see also the paragraph above). Lindh et al. (1989) reported, in cats, that ‘‘sympathetic’’ trunk ganglia contained clusters of VIP/ PHI-positive neurons that innervated blood vessels in striated muscles. Likewise, in guinea pig, Anderson et al. (1996) considered VIP-positive nerve ﬁbers to represent the non-noradrenergic ‘‘sympathetic’’ system innervating the feeding arteries of muscles. Consequently, the classical concept of the sympathetic and parasympathetic nervous systems does not seem to correspond to the distribution of marker molecules such as TH, VIP/PHI and nNOS. Finally, it is necessary to consider again the issue of individual variation. Although the numbers of TH-positive cells in the CG varied considerably, nNOS reactivity also varied in the PG and OG. At the beginning of this discussion, we hypothesized differences in postmortem and/or in utero conditions between specimens. Although such technical issues are certainly relevant, we cannot rule out the possibility that so-called parasympathetic ganglia actually display individual variations in the proportions of neuron subpopulations. The signiﬁcant individual variations would likely result in individual differences in sensitivity or response to receptor blockers or stimulators of autonomic nerve function in routine medicine.
ACKNOWLEDGEMENTS The authors are grateful to Ms. Shizue Mochizuki (Tohoku University Hospital) for her meticulous efforts to prepare sections.
LITERATURE CITED Adler R, Landa KB, Manthorpe M, Varon S. 1979. Cholinergic neurotrophic factors: intraocular distribution of trophic activity for ciliary neurons. Science 204:1434–1436. Alm P, Uvelius B, Ekstro¨m J, Holmqvist B, Larsson B, Andersson KE. 1995. Nitric oxide synthase-containing neurons in rat parasympathetic, sympathetic and sensory gamglia: a comparative study. Histochem J 27:819–831. Anderson RL, Gibbins IL, Morris JL. 1996. Non-noradrenergic sympathetic neurons project to extramuscular feed arteries and proximal intramuscular arteries of skeletal muscles in guinea-pig hindlimbs. J Autonomic Nerv Sys 61:51–60. Black IB. 1982. Stages of neurotransmitter development in autonomic neurons. Science 215:1198–1204. Berkovitz BKB. 2005. Neck. In: Standring S, editor. Gray’s anatomy, 39th ed. Edinburgh: Elsevier Churchill Livingstone. Chapter 31, p 557. Costello AJ, Dowdle BW, Namdarian B, Perdersen J, Murphy DG. 2011. Immunohistochemical study of the cavernous nerves in the periprostatic region. Br J Urol Int 107:1210–1215 Crouse GS, Cucinotta AJ. 1965. Progressive neuronal differentiation in the submandibular ganglia of a series of human fetuses. J Comp Neurol 125:259–271. Ernsberger U, Sendtner M, Rohrer H. 1989. Proliferation and differentiation of embryonic chick sympathetic neurons: effects of ciliary neurotrophic factor. Neuron 2:1275–1284. Hardebo JE, Suzuki N, Ekblad E, Owman C. 1992. Vasoactive intestinal polypeptide and acetycholine coexist with neuropeptide Y, dopaminebeta-hydroxylase, tyrosine hydroxylase, substance P or calcitonin gene-related peptide in neuronal subpopulations in cranial parasympathetic ganglia of rat. Cell Tissue Res 267:291–300.
Henrich M, Haberberger RV, Hempelmann G, Kummer W. 2003. Quantitative immunohistochemical investigation of the intrinsic vasodilator innervations of the guinea pig lingual nerve. Auton Neurosci 103:72–82. Hisasue S, Hashimoto K, Kobayashi K, Takeuchi M, Kyoda Y, Sato S, Masumori N, Tsukamoto T. 2010. Baseline erectile function alters the cavernous nerve quantity and distribution around the prostate. J Urol 184:2062–2067. Ibanez M, Valderrama-Canales FJ, Maranillo E, Vazquez T, Pascual-Font A, McHanwell S, Sanudo J. 2010. Human laryngeal ganglia contain both sympathetic and parasympathetic cell types. Clin Anat 23:673–682. Imai K, Furuya K, Kinugasa Y, Kawada M, Omote K, Namiki A, Uchiyama E, Murakami G. 2006. Human pelvic extramural ganglion cells: a semiquantitative and immunohistochemical study. Surg Radiol Anat 28:596–605. Itoh N, Obata K, Yanaihara N, Okamoto H. 1983. Human preprovasoactive intestinal peptide contains a novel PHI-27-like peptide, PHM-27. Nature 304:547–549 Katori Y, Kim JH, Rodriguez-Vazquez JF, Kawase T, Murakami G, Cho BH. 2011a. Early fetal development of the intermediate tendon of the digastricus and omohyoideus muscles: a critical difference in histogenesis. Clin Anat 24:843–852. Katori Y, Rodriguez-Vazquez JF, Kawase T, Murakami G, Cho BH, Abe S. 2011b. Early fetal development of hard tissue pulleys for the human obliquus superior and tensor veli palatini muscles. Ann Anat 193:127–133. Katori Y, Takeuchi H, Rodrı´guez-Va´zquez JF, Kitano H, Murakami G, Kawase T. 2011c. Fetal development of the human epiglottis revisited: appearance of GFAP-positive mesenchymal cells and ﬁbrous connections with other laryngeal and lingual structures. Ann Anat 193:149–155. Kirch W, Neuhuber W, Tamm ER. 1995. Immunohiostochemical localization of neuropeptides in the human ciliary ganglion. Brain Res 681:229–234. LeDouarin NM. 1978. Adrenergic differentiation of cells of the cholinergic ciliary and Remark ganglia in avian embryo following in vivo transplantation. Proc Natl Acad Sci USA 75:2030– 2034. Levitt P, Maxwell CD, Pintar JE. 1985. Speciﬁc cellular expression of monoamine oxisase B during early stages of quail embryogenesis. Dev Biol 75:148–167. Lindh B, Lundberg JM, Ho¨kfelt T. 1989. NPY-, galanin-, VIP/PHI-, CGRP-, and substance P-immunoreactive neuronal subpopulations in the cat autonomic and sensory ganglia and their projections. Cell Tissue Res 256:259–273. May CA, Neuhuber W, Lutjen-Drecoll E. 2004. Immunohistochemical classiﬁcation and functional morphology of human choroidal ganglion cells. Invest Ophthalmol Vis Sci 45:361–367. Miyake N, Hayashi S, Cho BH, Kawase T, Murakami G, Fujimiya M, Kitano H. 2010. Fetal anatomy of the human carotid sheath and structures in and around it. Anat Rec 293:438–445. Ng YK, Wong WC, Ling EA. 1995. A study of the structure and function of the submandibular ganglion. Ann Acad Med Singapore 24:793–801. O’Rahilly R, Muller S. 1996. Human embryology and teratology, 2nd ed. New York: Wiley-Liss. p 396–401. Osanai H, Abe S, Rodrı´guez-Va´zquez JF, Verdugo-Lo´pez S, Murakami G, Ohguro H. 2011. Human orbital muscle: a new point of view from the fetal development of extraocular connective tissue. Invest Ophthalmol Vis Sci 52:1501–1506. Roudenok V, Kuhnel W, Rogov Y, Nerovnja A. 1999. Develpmental changes in vasoactive intestinal polypeptide immunoreactivity in the human paravertebral ganglia. Ann Anat 181:561–565. Rusu MC, Pop F. 2010. The anatomy of the sympathetic pathway through the pterygopalatine fossa in humans. Ann Anat 192: 17–22. Rusu MC, Pop F, Bos cu AL, Jianu AM, Dermengiu D, Curca GC, Hostiuc S. 2011. Anatomical and immunohistochemical considerations on the microinnervation of trachea in humans. Ann Anat 193:13–22.
RECONSIDERATION OF THE AUTONOMIC CRANIAL GANGLIA Saadat S, Sendtner M, Rohrer H. 1989. Ciliary neurotropic factor induces cholinergic differentiation of rat sympathetic neurons in culture. J Cell Biol 108:1807–1816. Taguchi K, Tsukamoto T, Murakami G. 1999. Anatomical studies of the autonomic nervous system in the human pelvis by the wholemount staining method: left-right communicating nerves between bilateral pelvic plexuses. J Urol 161:320–325. Tajti J, Mo¨ller S, Uddman R, Bodi I, Edvinsson L. 1999. The human superior cervical ganglion: neuropeptides and peptode receptors. Neurosci Lett 263:121–124. Takenaka A, Kawada T, Murakami G, Hisasue S, Tsukamoto T, Fujisawa M. 2005. Interindividual variations in distribution of extramural ganglion cells in the male pelvis. Eur J Urol 48:46–52. Tan CK, Zhang YL, Wong WC. 1995. A light- and electron microscopic study of tyrosine hydroxylase-like immunoreactivity in the ciliary ganglion of monkey and cat. Histol Histopathol 10:27–34.
Teitelman G, Joh TH, Grayson L, Park DH, Reis DJ, Iacovitti L. 1985. Cholinergic neurons of the chick ciliary ganglia express adrenergic traits in vivo and invitro. J Neurosci 5:29–39. Thakker MM, Huang J, Possin DE, Ahmadi AJ, Mudumbai R, Orcutt JC, Tarbet KJ, Sires BS. 2008. Human orbital sympathetic pathways. Ophthal Plast Reconstr Surg 24:360–366. Tyrrell S, Siegel RE, Landis SC. 1992. Tyrosine hydroxylase and neuropeptide Y are increased in ciliary ganglia of sympathectomized rats. Neuroscience 47:985–998. Uddman R, Tajti J, Mo¨ller S, Sundler F, Edvinsson L. 1999. Neuronal messengers and peptide receptors in the human sphenopalatine and otic ganglia. Brain Res 826:193–199. Uemura Y, Sugimoto T, Nomura S, Nagatsu I, Mizuno N. 1987. Tyrosine hydroxylase-like inmmunoreactivity and catecholamine ﬂuorescence in ciliary ganglion neurons. Brain Res 416:200– 203.