RESEARCH ARTICLE
Excitatory and Inhibitory Innervation of the Mouse Orofacial Motor Nuclei: A Stereological Study
Macarena Faunes,1,2 Alejandro Onate-Ponce,~ 1,2 Sara Fern andez-Collemann,1,2 and Pablo Henny1,2* 1Laboratorio de Neuroanatom ıa, Departamento de Anatom ıa Normal, Escuela de Medicina, Pontificia Universidad Catolica de Chile, Santiago, Chile 2Centro Interdisciplinario de Neurociencias, Pontificia Universidad Catolica de Chile, Santiago, Chile
ABSTRACT 3,000, 600, and 1,700 cells, respectively. Neurons in the trigeminal (Mo5), facial (Mo7), ambiguus VGluT11,VGluT21, and VIAAT1 varicosities respec- (Amb), and hypoglossal (Mo12) motor nuclei innervate tively represent: 28%, 41%, and 31% in Mo5; 2%, 49%, jaw, facial, pharynx/larynx/esophagus, and tongue and 49% in Mo7; 12%, 42%, and 46% in Amb; and 4%, muscles, respectively. They are essential for movements 54%, and 42% in Mo12. The Mo5 jaw-closing subdivision subserving feeding, exploration of the environment, and shows the highest VGluT11 innervation. Noticeably, the social communication. These neurons are largely con- VGluT21 and VIAAT1 varicosity density in Mo7 is 5-fold trolled by sensory afferents and premotor neurons of the higher than in Mo5 and 10-fold higher than in Amb and reticular formation, where central pattern generator cir- Mo12. The high density of terminals in Mo7 likely cuits controlling orofacial movements are located. To reflects the convergence and integration of numerous provide a description of the orofacial nuclei of the adult inputs to motoneurons subserving the wide range of mouse and to ascertain the influence of excitatory and complex behaviors to which this nucleus contributes. inhibitory afferents upon them, we used stereology to Also, somatic versus neuropil location of varicosities sug- estimate the number of motoneurons as well as of vari- gests that most of these afferents are integrated in the cosities immunopositive for glutamate (VGluT11, dendritic trees of Mo7 neurons. J. Comp. Neurol. VGluT21) and GABA/glycine (known as VIAAT1 or 524:738–758, 2016. VGAT1) vesicular transporters in the Mo5, Mo7, Amb, and Mo12. Mo5, Mo7, Amb, and Mo12 contain 1,000, VC 2015 Wiley Periodicals, Inc.
INDEXING TERMS: AB_1966444; AB_2079751; AB_2254574; AB_2315551; AB_2340593; AB_2340396; facial nucleus; hypoglossal nucleus; mouse brainstem; nucleus ambiguus; trigeminal motor nucleus; vesicular transporters; VGAT; VGluT1; VGluT2; VIAAT
Mammalian orofacial muscles are involved in vital involved in the rhythmic exploratory behavior-related non-locomotive behaviors including suckling, mastica- movements whisking and sniffing (Deschenes^ et al., tion, swallowing, whisking, facial expression, and the 2012), ear wiggling, and other movements such as production of vocalizations, among others. In humans, blinking. These various patterns of activity are con- they are also involved in the production of speech. Jaw, trolled by brainstem premotor networks and influenced facial, pharynx/larynx/esophagus, and tongue muscles are innervated by motoneurons located in the trigeminal (Mo5), facial (Mo7), ambiguus (Amb), and hypoglossal Grant sponsor: FONDECYT; Grant number: 11100433; Grant sponsor: Anillo; Grant number: ACT-1109, CONICYT, Chile. (Mo12) brainstem motor nuclei, respectively. Current address for Macarena Faunes: Sensory and Motor Systems Group, Department of Anatomy with Radiology, Faculty of Medical and Feeding-related movements, such as mastication, Health Sciences, University of Auckland, Private Bag 92019, Grafton, suckling and licking, involve the rhythmic activity of 1023, Auckland, New Zealand. Mo5 in coordination with Mo7 and Mo12 (Nakamura *CORRESPONDENCE TO: Pablo Henny, Laboratorio de Neuroanatom ıa, Departamento de Anatom ıa Normal, Escuela de Medicina, Pontificia and Katakura, 1995; Travers et al., 1997). These move- Universidad Catolica de Chile. Casa central, Alameda 340, Comuna de ments are often followed by swallowing, where Amb Santiago, Santiago, Chile. E-mail: [email protected] motoneurons take part as well (Jean, 2001). Also, differ- Received April 23, 2015; Revised June 27, 2015; Accepted July 21, 2015. ent subsets of cells within the facial motor nucleus are DOI 10.1002/cne.23862 Published online October 8, 2015 in Wiley Online Library VC 2015 Wiley Periodicals, Inc. (wileyonlinelibrary.com)
738 The Journal of Comparative Neurology | Research in Systems Neuroscience 524:738–758 (2016) VGluT1, VGluT2, and VIAAT in orofacial nuclei
by sensory feedback (Holstege et al., 1977; Buttner-€ approved by the Ethics Committees of the School of Ennever and Holstege, 1985; Chen et al., 2001; Chen Medicine of the Pontificia Universidad Catolica de Chile, and Travers, 2003; Galvin et al., 2004; Grillner et al., and of the Comision Nacional de Investigacion Cient ıf- 2005; Landers and Philip Zeigler, 2006; Nistri et al., ica y Tecnologica CONICYT, both of which conform to 2006; Kinzeler and Travers, 2008; Bianchi and Ges- the guidelines of US National Institutes of Health (NIH) treau, 2009; Moore et al., 2013). Excitatory and inhibi- on animal procedures. Animals were anesthetized with tory neurons located in the midbrain, pontine, and gaseous isoflurane followed by intraperitoneal keta- medullary reticular formation constitute the main pre- mine/xylazine (150 mg/kg and 10 mg/kg, respec- motor afferents to these nuclei, along with primary exci- tively), and then intracardially perfused with phosphate- tatory afferents and various second-order sensory buffered saline (PBS) and 4% paraformaldehyde using a brainstem nuclei (Travers and Norgren, 1983; Nakamura peristaltic pump (Masterflex 7518-00). The brains were and Katakura, 1995; Travers, 2004; Grillner et al., extracted and postfixed for 8 hours in 4% paraformalde- 2005; Takatoh et al., 2013; Stanek Iv et al., 2014). It is hyde, cryoprotected in 30% sucrose, and then cut in expected that the variety of functions and activity pat- 40-lm-thick sections (one of the brains was cut in the terns displayed by the neurons in the different orofacial sagittal plane and the other five were cut in the coronal motor nuclei will be reflected in the type and organiza- plane) using a microtome (Reichert-Jung Hn40, Depew, tion of their afferents. NY) equipped with a freezing stage (Hacker Instruments Previous quantitative analyses on the overall glutama- 220–240v, Fairfield, NJ). Sections from each brain were tergic, GABAergic, and glycinergic innervation of oromo- collected in four series (160 lm between sections in tor neurons have only been done in trigeminal each series) and stored in PBS. motoneurons of cats (Bae et al., 1999; Shigenaga et al., 2005; Shigenaga et al., 2007) and rats (Saha Immunohistochemistry et al., 1991; Yang et al., 1997; Bae et al., 2002). No Antibodies characterization quantitative analyses have been done on the innerva- Primary antibodies used in this study are listed in Table tion of other orofacial motoneurons to allow for com- 1. The rabbit polyclonal anti-VGluT1 antibody (Mab parisons across nuclei. Furthermore, studies on the Technologies, Cat. no. VGT1-3, AB_2315551) was innervation of orofacial motoneurons of the mouse, per- raised against a peptide of the rat VGluT1 (amino acids haps the most important animal model for neuroscience 543–560), which is highly conserved in the mouse nowadays, are still lacking. VGluT1 sequence. This antibody was characterized by With the aim of providing a quantitative description Villalba et al. (2006) in rat and monkey tissue. Western of the main afferents received by each of the orofacial blot analysis showed that anti-VGluT1 labels bands at motor nuclei, we assessed their glutamatergic and 60 kDa, the predicted molecular weight of VGluT1 GABA/glycinergic innervation in the mouse. We used (Villalba et al., 2006). Furthermore, immunostaining is an unbiased stereological technique, the optical fractio- abolished when the antibody is preadsorbed with the nator (West et al., 1991), to estimate the number of peptide (Villalba et al., 2006). The rabbit polyclonal cells and varicosities immunopositive for the vesicular anti-VGluT2 antibody (Synaptic Systems, Goettingen, Germany, Cat. no. 135 403, AB_2254574) was raised transporters for glutamate (VGluT1 and VGluT2) and the against a Strep-TagVR fusion protein of rat VgluT2 vesicular transporter for GABA and glycine (vesicular (amino acids 510–582), and it was characterized by inhibitory amino acid transporter, VIAAT; also known as Hrabovszky et al. (2004) in rat tissue. Immunostaining vesicular GABA transporter, VGAT) in each of these is abolished when the antibody is preadsorbed with the nuclei and their main subdivisions. Finally, we estimated immunization antigen (Hrabovszky, 2004). In addition, the distribution of varicosities over somatic compart- double immunofluorescence assays show that this anti- ments for each vesicular transporter and compared this body and a guinea pig anti VGluT2 produce identical to the overall pattern of innervation within and across labeling (Hrabovszky, 2004). The rabbit anti-VIAAT anti- nuclei. body (VGAT, Synaptic Systems, Cat. no. 131 013, RRID: AB_1966444) was raised against a Strep-TagVR fusion MATERIALS AND METHODS protein of rat VIAAT (amino acids 2–115), and it labels Animals and tissue preparation a band at 55 kDa on western blots of rat hippocam- Six male adult (2 months old) C57BL/6 mice, pus, the predicted molecular weight of VGAT (Stensrud obtained from the animal facility of the Faculty of Bio- et al., 2013). The goat anti-ChAT (Millipore, Bedford, logical Sciences at the Pontificia Universidad Catolica MA, Cat. no. AB144P, AB_2079751) antibody was de Chile, were used in this study. All procedures were raised against the human placental enzyme, and it has
The Journal of Comparative Neurology | Research in Systems Neuroscience 739 M. Faunes et al.
TABLE 1. Primary Antibodies Used in This Study
Description of Source, host species, Cat. #, Concentration Antigen immunogen Clone or Lot#, RRID used VGluT1 (vesicular glutamate Peptide (amino acids 543–560, Mab Technologies, raised in Rabbit antisera transporter 1) ATHSTVQPPRPPPPVRDY, NCBI rabbit, Cat# VGT1-3, diluted 1:1000 accession NP_446311) of the rat AB_2315551 VGluT1 coupled to KLH by the addition of an N-terminal cysteine VGluT2 (vesicular glutamate Strep-TagVR fusion protein of rat VGluT2 Synaptic Systems, raised in 1 lg/ml transporter2) (amino acids 510 – 582, ETSEEKCG rabbit, Cat# 135 403, FIHEDELDEETGDITQNYINYGTTKSY AB_2254574 GATSQENGGWPNGWEKKEEFVQES AQDAYSYKDRDDYS, NCBI accession NP_445879) VGAT/VIAAT Strep-TagVR fusion protein of rat VIAAT Synaptic Systems, raised in 2 lg/ml (vesicular GABA (amino acids 2-115, ATLLRSKLTN rabbit, Cat# 131 013, transporter/vesicular VATSVSNKSQAKVSGMFARMGFQA AB_1966444 inhibitory aminoacid ATDEEAVGFAHCDDLDFEHRQGLQ transporter) MDILKSEGEPCGDEGAEPPVEGDI HYQRGGAPLPPSGSKDQAVGAG GEFGGHDKPK, NCBI accession NP_113970) ChAT (choline Human placental enzyme Millipore, raised in goat, Cat# 5 lg/ml acetyltransferase) AB144P, AB_2079751 been characterized with western blot in mouse brain was incubated overnight at room temperature in one of lysate by the vendor and in rat brain lysate by Stensrud the three antibodies for vesicular transporters: anti- et al. (2013). In both cases it is reported to label a sin- VGluT1 (1:1,000, Rb, Mab Technologies, Cat. no. VGT1- gle band at 68–70 kDa. 3), anti-VGluT2 (1:1,000, Rb, Synaptic Systems, Cat. no. Immunostaining with these antibodies across the mouse 135 403), anti-VIAAT (1:500, Rb, Synaptic Systems, Cat. central nervous system (not shown) showed a distribution no. 131 103), or the antibody against ChAT (1:1,000, Gt, consistent with previous reports (VGluT1 and VGluT2: Fre- Millipore, Cat. no. AB144P) diluted in blocking solution. meau et al., 2001, 2004; Kaneko and Fujiyama, 2002; The sections were then washed in PBS and incubated Kaneko et al., 2002; Doig et al., 2010; Michalski et al., in biotinylated secondary antibodies; donkey antira- 2013; VIAAT: McIntire et al., 1997; Sagne et al., 1997; bbit antibody (1:1,000 Jackson ImmunoResearch, Chaudhry et al., 1998; Michalski et al., 2013; Zhang et al., AB_2340593, for immunohistochemistry against vesicu- 2013; ChAT: Ichikawa et al., 1997). lar transporters) or biotinylated donkey antigoat antibody (1:1,000 Jackson ImmunoResearch, AB_2340396, for Immunostaining immunohistochemistry against ChAT) for 2.5 hours at Series of sections from four of the coronal sectioned room temperature, washed again in PBS, and incubated brains were Nissl-stained for quantification of motoneur- in ABC-Elite reagent (1:200, Vector Laboratories, Burlin- ons or prepared for quantification of immunopositive game, CA; prepared according to instructions from the varicosity numbers with immunohistochemistry for vesic- manufacturer) for 1 hour at room temperature. The sec- ular transporters VGluT1, VGluT2, and VIAAT and Nissl tions were washed and preincubated for 5 minutes in counterstaining. The sections from the remaining two PBS with diaminobenzidine (DAB) 0.25 mg/mL and brains (one coronal sectioned and one sagittal sectioned) 0.05% NiSO4 solution, incubated for 3 minutes in the were immunostained against choline acetyltransferase same solution with added 0.015% H2O2, and then trans- (ChAT) to neurochemically identify motoneurons. ferred to PBS and washed. All series of sections were For immunohistochemistry, the sections were incu- mounted on gelatin-coated slides and left to dry. bated in 50% methanol and 0.3% H2O2 for 15 minutes at For Nissl counterstaining, slides were first incubated room temperature to quench endogenous peroxidase in chloroform for 45 minutes. The sections were then activity, then washed in PBS and incubated for 1 hour at rehydrated by incubation in a descending series of room temperature in 3% normal donkey serum (NDS, ethanol; 100%, 90%, 70%, 50% (3 minutes each) and Jackson ImmunoResearch, West Grove, PA) in PBS Triton distilled water (4 minutes). They were then immersed Tx-100 0.3% (blocking solution). Each series of sections in a Nissl solution (0.1% cresyl violet acetate) for
740 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
TABLE 2. Cell Number, Nucleus Volume, and Cell Density in Mo5, Mo7, Amb, and Mo12 (n 5 4)
Cell number Volume (mm3) Density (cells/mm3) Nucleus Subdivision Average SEM Average SEM Average SEM Mo5 DL5 868 107 0.142 0.011 6,097 580 VM5 85 12 0.016 0.002 5,432 1,047 P5 — — 0.050 0.004 — Total 952 110 0.159 0.012 5,984 575 Mo7 DL7 304 28 0.035 0.003 8,759 467 L7 901 75 0.092 0.008 9,966 75 DI7 621 32 0.051 0.002 12,087 394 VI7 231 34 0.019 0.002 12,252 2,150 DM7 421 14 0.041 0.004 10,383 677 VM7 407 39 0.035 0.004 11,537 515 S7 69 14 0.008 0.002 9,596 963 P7 — — 0.026 0.003 — Total 2,957 159 0.282 0.014 10,562 796 Amb rAmb 370 66 0.015 0.001 25, 012 3,667 cAmb 203 36 0.013 0.001 15,772 1,623 Total 588 98 0.028 0.002 20,962 2,689 Mo12 DL12 733 56 0.059 0.001 12,490 1,127 VM12 947 90 0.059 0.004 16,370 2,322 Total 1666 132 0.117 0.004 14,364 1,519
30 seconds to 2 minutes to obtain clear staining within each disector box was 80 3 80 lm over Mo5 and Mo7, nuclei, and then washed in distilled water for 30 sec- and 65 3 65 lm over Amb and Mo12. The grids used onds. If needed, contrast was enhanced using glacial to distribute the counting boxes were equal to the box acetic acid. Sections were then dehydrated in an area to maximize sampling. One out of four serial sec- ascending series of alcohol (50%, 70%, 90%, and 100%) tions (160 lm between sections) was analyzed for the and placed in xylene for at least 30 minutes before cov- entire anteroposterior extent of each nucleus. These erslipping with a xylene-based mounting medium (Entel- parameters yielded Gundersen coefficient of error (CE) lan, Merck, Darmstadt, Germany). ranging between 0.04 and 0.30 for all nuclei and subdi- visions, with overall CE of 0.08, 0.04, 0.11, and 0.06 for Estimation of cell and immunopositive Mo5, Mo7, Amb, and Mo12, respectively (Table 2), varicosity numbers which lie within values acceptable for stereological anal- Estimation of cell number and nucleus volume yses (Guti errez-Ib anez~ et al., 2012; Henny et al., 2012, The Nissl-stained serial sections (160 lm between 2014; Faunes et al., 2013). All the counting was per- sections) were examined using an Olympus BXS1 micro- formed by one researcher. scope equipped with an x-y-z motor stage. The Mo5, Mo7, Mo12, and Amb nuclei were then contoured using Estimation of immunopositive varicosities Stereo Investigator software (MBF Bioscience, Williston, number and proportions VT). The nucleus volume estimations were obtained Contours for the nuclei were obtained from the Nissl according to the Cavalieri principle, and cell number and VGluT1 immunostained series and confirmed with estimates were generated with the optical fractionator equivalent sections exhibiting ChAT immunoreactivity, method (West et al., 1991), using the optical fractiona- and then used for nuclei localization in adjacent series. tor workflow in Stereo Investigator. Briefly, using an oil- Total varicosity number estimates were generated immersion lens (Olympus 60X, numerical aperture 1.42), using the optical fractionator workflow in Stereo Investi- motor neuron nucleoli that fell within the disector boxes, gator. The size of both the counting frames and the that were systematically distributed over the entire vol- grids used to distribute the sampling sites were ume of each nucleus, were counted following stereologi- adjusted to reach a C.E <0.10 for each nucleus, while cal rules to avoid counting bias for cell size, shape, CE for individual subdivisions was between 0.03 and orientation, or distribution. The height of the disector 0.13. The disector box area ranged between 10 3 10 box was 5 lm, with a guard zone of 1 lm. The average lm and 50 3 50 lm, and the grid area between 25 3 section thickness was 6.25 6 0.25 lm. The area of 25 lm and 70 3 70 lm for grids. The disector height
The Journal of Comparative Neurology | Research in Systems Neuroscience 741 M. Faunes et al.
Figure 1. Depiction of the counting process of immunopositive varicosities using the optical fractionator. Brightfield photomicrographs taken at different z depths in a given region of the dorsolateral Mo5 (DL5) and stained for VGluT1 (A–K). The disector box is formed by the counting frame (red/green squares) at different z-depths. Varicosities were counted (green arrows) if located within the permitted regions of the disector box, i.e., inside the box or partially outside but intersecting any of the green borders and/or the top plane of the box. Varicosities intersecting any of the forbidden borders of the disector, i.e., the red borders and the bottom plane (depicted in yellow in K) were not counted (red arrows). In this example, as in several cases in this study, the disector box dimensions were: 10 3 10 3 3.0 lm. Please note that the z-depth scanning process of varicosities was continuous through the z-axis, as carried out online in the micro- scope unit and does not involve taking photomicrographs at different z-depths. Scale bar 5 5 lm.
was set to 3.0 lm for VGluT1 and VGluT2 series and These proportions therefore indicate the abundance of 2.5 lm for VIAAT series, following assessment of a given type of varicosity in relation to the sum of most adequate antibody penetration and staining. Average of the glutamatergic innervation (we did not include the section thickness was 8.5 6 0.40 lm. Figure 1 vesicular transporter type 3, VGluT3) and the entire describes the counting procedure within a disector fol- GABAergic/glycinergic innervation (McIntire et al., 1997; lowing stereology rules. Two researchers performed Sagne et al., 1997; Chaudhry et al., 1998; Fremeau et al., these and the perisomatic varicosity estimations (see 2001, 2004; Kaneko and Fujiyama, 2002; Kaneko et al., below). To assess rating differences between them, we 2002). compared their counts for the same boxes (30 boxes, including 10 for each vesicular transporter). Their Estimation of immunopositive varicosities results differed by less than 1% (204 and 206 varicos- number and proportions over somatic ities, with an average difference of 0.067 6 0.94 vari- compartments cosities per counting box). The distribution of the immunopositive varicosities over Relative abundances were obtained by dividing the somatic or nonsomatic compartments was studied by number of varicosities immunopositive for a given vesic- quantifying the number and proportion of varicosities ular transporter by the total number of immunopositive displaying a perisomatic position. This was carried out varicosities for a given nucleus or nucleus subdivision. using the optical fractionator method as described in
742 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
the previous subsection. Given that perisomatic varicos- are presented as mean 6SEM (n 5 4) unless otherwise ities represent a small fraction of the total population, stated. the counting frames used were larger than those for Images in figures were adjusted for brightness, con- the total number of varicosities estimation and the CEs trast, and color balance using Adobe Photoshop CS6 were generally higher (up to 0.3 for whole nuclei and (San Jose, CA), so as to provide a clear and comparable 0.54 for subdivisions). The criterion to consider a vari- set of examples. Panels, lettering, and drawings were cosity as perisomatic was for it to be in clear apposi- prepared using Adobe Illustrator CS6. tion with an unequivocally identified cell body, as stained with Nissl. Due to the fact that sections of cell RESULTS bodies may become difficult to recognize, especially Estimation of cell number and nucleus near their top and bottom edges, it is likely that the volume number of perisomatic varicosities is underestimated. Cell numbers and nucleus volumes for Mo5, Mo7, Yet this criterion should not have any bias towards a Amb, and Mo12 were estimated on Nissl-stained coro- particular vesicular transporter or structure and thus nal sections using the optical fractionator method. The we consider it valid for comparison purposes. nuclei were identified according to the mouse brain To obtain an estimate of somatic innervation, the atlas of Franklin and Paxinos (2008) and with the aid of number of perisomatic varicosities for a given vesicular a ChAT immunostained series as a reference (Fig. 2). transporter was compared to the total number of vari- cosities of the same vesicular transporter in each Trigeminal motor nucleus nucleus. To assess whether varicosities of a given type The mouse Mo5 is a large nucleus that appears round- of vesicular transporter showed a tendency for somatic or oval-shaped in cross section, and is located medial innervation, the relative abundance of perisomatic vari- to the trigeminal principal sensory nucleus (between cosities was obtained by dividing the number of periso- bregma –4.72 mm and –5.34 mm, and lateral 1.2 mm matic varicosities immunopositive for a given vesicular and 1.75 mm; Franklin and Paxinos, 2008). It is subdi- transporter by the total number of perisomatic varicos- vided into two readily distinguishable cell populations: a ities. Values of relative abundance of perisomatic vari- larger, dorsolateral population, mainly innervating the cosities were then compared to the relative abundance jaw-closing muscles (masseter, temporalis, and medial of total varicosities. pterygoid), which we will refer to as dorsolateral trigem- inal motor nucleus (DL5), and a smaller, ventromedial Statistical analysis and image processing population innervating mainly the jaw-opening muscle For statistical analyses of cell number and density anterior digastric, and the mylohyoid, which we will we used repeated measures analysis of variance refer to as ventromedial trigeminal motor nucleus (ANOVA) with nucleus as factor. For statistical analysis (VM5) (Terashima et al., 1994; Fig. 2A). In our analysis -3 of varicosity density (mm ) we first used two-way we also considered the neuropil located medial to DL5 repeated measures ANOVA with vesicular transporter and dorsal to VM5, called the peritrigeminal zone (P5), and nuclei subdivision as factors to test for differences which contains dendrites of trigeminal motoneurons within each nucleus. As Mo5 proved to have different (Lingenhohl and Friauf, 1991; Fig. 2A,E) and thus repre- immunopositive varicosity densities while Mo7, Amb, sents a region where motoneurons receive afferents. and Mo12 subdivisions proved to have no significant The average number of cells estimated in Mo5 is 952 differences, we then considered Mo5 different subdivi- 6 110 and its estimated volume is 0.159 6 0.012 sions and Mo7, Amb, and Mo12 as a whole in two-way mm3 (Table 2). The cell density estimated from these repeated measures ANOVA to compare varicosity den- data is 5,984 6 575 cells/mm3. There are no differen- sity (mm-3), varicosities per cell, and perisomatic to ces in cell density between the two subdivisions of total varicosities ratio across nuclei. The relative abun- Mo5, but DL5 has a much higher cell number than VM5 dances of perisomatic and total number of immunoposi- (paired t-test, P < 0.01). This could be expected con- tive varicosities were compared with two-way repeated sidering that DL5 innervates a much larger muscle measures ANOVA with nucleus and position (periso- group than VM5. Volume and cell number estimations matic/total) as factors for each vesicular transporter. for each subdivision of Mo5 are shown in Table 2. Square root transformation was used in the cases where normal distribution and/or homoscedasticity Facial nucleus assumptions were violated. All post-hoc comparisons The mouse Mo7 is a large nucleus located ventral and were done with the Student-Newman-Keuls test. Data caudal to Mo5 (between bregma –5.8 mm and –6.36 mm,
The Journal of Comparative Neurology | Research in Systems Neuroscience 743 M. Faunes et al.
Figure 2. Cytoarchitectonic and inmunohistological identification of Mo5, Mo7, Amb, and Mo12. Brightfield photomicrographs of Nissl- stained (A–D) and corresponding ChAT-immunostained (E–H) coronal sections of the trigeminal motor nucleus (Mo5; A,E, approximately at bregma –5.2 mm), the facial motor nucleus (Mo7; B,F, approximately at bregma –5.9 mm), nucleus ambiguus (Amb; C,G, approximately at bregma –6.8 mm), and the hypoglossal motor nucleus (Mo12; D,H, approximately at bregma –7.6 mm). Note that motor nuclei and their subdivisions are readily distinguishable using cytoarchitectonic criteria Abbreviations: Ad1, adrenergic cells; DI7, dorsal intermediate sub- nucleus of Mo7; Mo10, dorsal motor nucleus of vagus; DL5, dorsolateral subnucleus of Mo5; DL7, dorsolateral subnucleus of Mo7; DL12, dorsolateral subnucleus of Mo12; DM7, dorsomedial subnucleus of Mo7; Gi, gigantocellular reticular formation; IRt, intermediate reticular formation; L7, lateral subnucleus of Mo7; m5, motor root of the trigeminal nerve; PCRt, parvicellular reticular formation; Pr5, principal sen- sory trigeminal nucleus; P5, peritrigeminal zone; P7, perifacial zone; S7, suprafacial zone; Su5, supratrigeminal zone; VI7, ventral intermedi- ate subnucleus of Mo7; VM5, ventromedial subnucleus of Mo5; VM7, ventromedial subnucleus of Mo7; VM12, ventromedial subnucleus of Mo12. Scale bar 5 200 lm. and lateral 0.75 mm and 1.75 mm; Franklin and Paxi- Table 2), which innervates the muscles around the eyes nos, 2008). It is subdivided into seven cell populations and the forehead (Komiyama et al., 1984; Terashima innervating the superficial muscles of the face. These et al., 1993). The total volume of Mo7 is 0.282 6 are the dorsolateral (DL7), dorsal intermediate (DI7), 0.014 mm3, and the average cell density is 10,562 6 lateral (L7), ventral intermediate (VI7), dorsomedial 796 cells/mm3 (Table 2). Volume and cell number esti- (DM7), ventromedial (VM7), and the suprafacial subdivi- mations for each subdivision of Mo7 are shown in sions (S7, located dorsally and separated to the other Table 2. subdivisions, Fig. 2B). Following a musculotopic organi- zation, the lateral regions of Mo7 innervate the more Nucleus ambiguus rostral muscles, medial regions innervate the more cau- The Amb is a narrow but long column of cells along the dal muscles, ventral regions innervate ventral muscles, rostro/caudal axis located caudal to Mo7 (between and dorsal regions innervate dorsal muscles of the face bregma –6.36 mm to –8.00 mm, and lateral 1.25 mm (Komiyama et al., 1984; Terashima et al., 1993). Ventral and 1.50 mm; Franklin and Paxinos, 2008). Studies in to the VI subdivision there is a ChAT-immunopositive the rat and the rabbit indicate that Amb can be subdi- (ChAT1) neuropil, which we will refer to as the perifa- vided into columns that are distinguishable in sagittal cial zone (P7, Fig. 2B,F), that most likely represents a sections (Bieger and Hopkins, 1987; Kitamura et al., region of afferent control of motoneurons, similar to P5 1993). The rostral compact column contains esophago- in Mo5. motor neurons, the intermediate semicompact column The average number of cells estimated in Mo7 is contains pharyngolaryngomotor neurons, and the caudal 2,957 6 159 (Table 2). L7, which innervates the naso- loose formation contains laryngomotor neurons (Bieger labialis muscle in the upper lip and therefore is involved and Hopkins, 1987; Kitamura et al., 1993). in whisking (Komiyama et al., 1984; Terashima et al., In Nissl- and ChAT-stained sagittal sections of the 1993) is the subdivision with the greatest number of mouse Amb (not shown) at least two subdivisions were cells (901 6 75; Table 2), followed by DI7 (621 6 32; distinguished. This allowed us to subdivide the Amb in
744 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
the coronal series of sections into a compact rostral and Mo12 motoneurons, we next estimated the total Amb (rAmb, Fig. 2C) and a semicompact caudal Amb number of VGluT1, VGluT2, and VIAAT immunopositive (cAmb), each comprising about half the rostrocaudal varicosities (VGluT11, VGluT21, and VIAAT1, respec- extent of the nucleus. rAmb and cAmb presumably cor- tively) in these nuclei and their subdivisions. Examples respond to esophagomotor and pharyngolaryngomotor of immunopositive varicosities in different orofacial neurons, respectively. motor nuclei are shown in Figure 3. The average cell number estimated for Amb is 588 6 In all nuclei except for Mo5 (see below) the density 98, and the nucleus volume is 0.028 6 0.002 mm3 of VGluT11 varicosities is lower than the densities of (Table 2). The cell density estimated from these data is VGluT21 and VIAAT varicosities (two-way repeated 20,962 6 2,689 cells/mm3. There is a significant differ- measures ANOVA, P < 0.05). ence in cell density between rAmb and cAmb (25,012 6 In Mo5, the total numbers of VGluT11, VGluT21, 3,667 cells/mm3 and 15,772 6 1,623 cells/mm3 and VIAAT1 varicosities are 221,409 6 25,186 (28%), respectively, paired t-test P < 0.05). Indeed, rAmb has a 323,275 6 65,548 (41%), and 245,105 6 19,813 higher cell density than Mo5, Mo7, and Mo12 (one-way (31%), respectively (Table 3). The density of VGluT11 repeated measures ANOVA, P < 0.01). This difference, varicosities is higher in DL5 than in VM5 and P5 3 as well as the smaller size of Amb motoneurons (cell (1,670,964 6 259,467/mm versus 310,082 6 3 3 size was not quantified in this study) is readily apparent 40,907/mm and 467,819 6 44,267/mm ; two-way in Nissl- and ChAT-stained sections (see Fig. 2). repeated measures ANOVA, P < 0.01, Table 3), in line with the study by Pang et al. (2009) in the rat. The den- sity of VGluT21 varicosities is also higher in DL5 than Hypoglossal nucleus 3 The Mo12 is located ventral to the vagal motor nucleus in VM5 and P5 (2,293,334 6 533,713/mm versus 1,082,565 6 206,427/mm3 and 847,296 6 144,935/ and has a rostrocaudal extent of about 1 mm (between 3 bregma –7.00 mm to –8.00 mm) and a mediolateral mm ; two-way repeated measures ANOVA, P < 0.01, Table 3). extent of 0.5 mm from the midline (Franklin and Paxinos, In Mo7, the total numbers of VGluT11, VGluT21, 2008). It can be subdivided into dorsolateral and ventro- and VIAAT1 varicosities are 108,349 6 7,253 (2%), medial cell populations (DL12 and VM12, respectively, 2,248,283 6 281,299 (49%), and 2,259,788 6 Fig. 2D). Motoneurons in DL12 send their axons through 155,224 (49%), respectively (Table 3). There are no dif- the lateral branch of the 12th (or hypoglossal) nerve and ferences between Mo7 subdivisions for any of the mainly innervate muscles involved in tongue retrusion, vesicular transporters (two-way repeated measures while those in VM12 send their axons though the medial ANOVA). branch of the hypoglossal nerve and innervate tongue pro- In Amb, the total numbers of VGluT11, VGluT21, trusors (Krammer et al., 1979; Uemura-Sumi et al., 1988). and VIAAT1 varicosities are 5,108 6 802 (12%), The average cell number of Mo12 is 1,666 6 132 18,313 6 3,774 (42%), and 20,180 6 2,736 (46%), cells, the nucleus volume is 0.117 6 0.004 mm3, 3 respectively (Table 3), and there are no significant dif- and the cell density is 14,364 6 1,519 cells/mm ferences between rAmb and cAmb. (Table 2). There are no significant differences in cell In Mo12, the total numbers of VGluT11, VGluT21, number or density between the two subdivisions. and VIAAT1 varicosities are 7,806 6 1,038 (4%), 117,178 6 24,129 (54%), and 92,726 6 23,976 (42%), Estimation of the number and relative respectively (Table 3). There are no differences between abundance of immunopositive varicosities Mo12 subdivisions for any of the vesicular transporters VGluT1 and VGluT2 proteins are expressed by differ- (two-way repeated measures ANOVA, P > 0.05). ent glutamatergic neuronal populations and represent Overall across nuclei, the VGluT11 varicosity den- specific markers of a glutamatergic phenotype. Within sity is significantly higher in DL5 than in Mo7, Amb, neurons, they are expressed almost exclusively in axon and Mo12 (Fig. 4A, two-way repeated measures terminals (Fremeau et al., 2001, 2004; Kaneko and ANOVA, P < 0.01). DL5 also has a higher density of Fujiyama, 2002; Kaneko et al., 2002). Likewise, the VGluT21 and VIAAT1 varicosities than Amb and VIAAT protein is expressed by GABAergic, glycinergic, Mo12 (Fig. 4B,C, two-way repeated measures ANOVA, and GABA/glycinergic populations, mainly in axon ter- P < 0.05). On the other hand, the VGluT21 and minals, and represents an unequivocal marker for inhib- VIAAT1 varicosity densities are significantly higher in itory neurons (McIntire et al., 1997; Sagne et al., 1997; Mo7 than in Mo5, Amb, and Mo12 (Fig. 4B,C, around Chaudhry et al., 1998). To assess the glutamatergic 5-fold higher than Mo5, and 10-fold higher than and GABA/glycinergic innervation of Mo5, Mo7, Amb, Amb and Mo12; two-way repeated measures ANOVA,
The Journal of Comparative Neurology | Research in Systems Neuroscience 745 M. Faunes et al.
Figure 3. VGluT1, VGluT2, and VIAAT innervation of the orofacial motor nuclei. High-magnification brightfield photomicrographs of Nissl and VGluT1 (A,D,G,J,M,P,S), VGluT2 (B,E,H,K,N,Q,T), and VIAAT (C,F,I,L,O,R,U) immunostained sections of Mo5, Mo7, Amb, and Mo12. Note the high VGluT1 labeling in DL5 (A) and the high VGluT2 (H,K) and VIAAT (I,L) labeling in the subdivisions of Mo7. Arrowheads indi- cate examples of immunopositive varicosities. Amb, nucleus ambiguus; DL5, dorsolateral subdivision of Mo5; DL12, dorsolateral subdivi- sion of Mo12; L7, lateral subdivision of Mo7; VM5, ventromedial subdivision of Mo5; VM7, ventromedial subdivision of Mo7; VM12, ventromedial subdivision of Mo12. Scale bar 5 5 lm.
746 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
Figure 3. Continued
P < 0.01). Similarly, the number of VGluT11 varicosities perisomatic immunopositive varicosities are listed in per cell is significantly larger in DL5 than in Mo7, Amb, Table 4. and Mo12, and the number of VGluT21 and of VIAAT1 In Mo5, the relative abundances of perisomatically varicosities per cell is significantly larger in Mo7 than in located VGluT11 varicosities (21%, Table 4) is signifi- all the other orofacial motor nuclei (data not shown, cantly lower than the overall proportion of VGluT11 two-way repeated measures ANOVA, P < 0.05). varicosities (28%, Table 3) (two-way repeated measures ANOVA, P < 0.01), suggesting a slight tendency of VGluT11 varicosities in this nucleus to localize in the Estimation of the number and relative neuropil (most likely innervating dendrites) instead of abundance of immunopositive varicosities cell bodies. No differences between perisomatic and on somatic compartments total relative abundances for VGluT21 or VIAAT1 vari- In Figure 5 we show examples of varicosities cosities were found in this nucleus, nor for any of VGluT11, VGluT21, and VIAAT1 considered as the vesicular transporters in Mo7, Amb, and Mo12 perisomatic in this study. The numbers and densities of (Tables 3, 4).
The Journal of Comparative Neurology | Research in Systems Neuroscience 747 M. Faunes et al.
TABLE 3. Immunopositive Varicosities Number, Relative Abundances, and Densities in Mo5, Mo7, Amb, and Mo12 (n 5 4)
VGluT1 Immunopositive varicosities number Density Relative Average SEM abundance Average SEM Mo5 DL5 193,575 23,633 0.30 1,670,964 259,467 VM5 4,279 689 0.13 310,082 40,907 P5 23,555 2,437 0.21 467,819 44,267 Total 221,409 25,186 0.28 1,230,073 155,552 Mo7 DL7 9,339 2,032 0.02 292,332 41,583 L7 34,645 1,909 0.03 456,049 41,715 DI7 22,176 1,889 0.03 511,834 49,269 VI7 8,490 705 0.03 539,475 37,436 DM7 6,744 384 0.01 201,892 27,672 VM7 5,825 1,837 0.01 179,931 47,385 S7 3,934 1,336 0.03 558,442 28,711 P7 17,195 1,649 0.04 664,296 125,200 Total 108,349 7,253 0.02 407,807 42,861 Amb rAmb 2,420 806 0.11 222,581 86,118 cAmb 2,689 538 0.13 207,613 35,882 Total 5,108 802 0.12 210,354 35,132 Mo12 DL12 4,122 456 0.03 77,126 3,630 VM12 3,684 743 0.04 69,155 13,766 Total 7,806 1,038 0.04 72,804 7,450 VGluT2 Immunopositive Density varicosities number Relative Average SEM abundance Average SEM
Mo5 DL5 265,674 61,791 0.41 2,293,334 533,713 VM5 14,938 2,640 0.47 1,082,565 206,427 P5 42,663 7,363 0.37 847,296 144,935 Total 323,275 65,548 0.41 1,796,005 319,442 Mo7 DL7 249,262 32,850 0.47 7,802,258 821,779 L7 648,148 95,008 0.47 8,531,843 917,021 DI7 354,141 35,616 0.49 8,173,657 774,527 VI7 138,192 21,862 0.50 8,780,631 599,145 DM7 285,130 41,611 0.51 8,535,917 323,079 VM7 272,979 26,325 0.50 8,432,188 315,796 S7 61,616 24,346 0.52 8,746,508 531,916 P7 238,814 34,639 0.49 9,225,999 532,046 Total 2,248,283 281,299 0.49 8,462,132 655,252 Amb rAmb 8,369 3,604 0.37 805,811 243,287 cAmb 9,945 1,670 0.47 792,837 154,803 Total 18,313 3,774 0.42 822,956 199,253 Mo12 DL12 62,444 12,008 0.51 1,057,368 174,395 VM12 54,734 12,305 0.57 947,447 157,365 Total 117,178 24,129 0.54 1,003,207 164,784 VIAAT Immunopositive Density varicosities number Relative Average SEM abundance Average SEM
Mo5 DL5 184,209 18,749 0.29 1,590,118 184,652 VM5 12,581 514 0.40 911,729 102,987 P5 48,316 4,127 0.42 959,568 89,092 Total 245,105 19,813 0.31 1,361,724 147,803
748 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
TABLE 3. Continued
VIAAT Immunopositive Density varicosities number Relative Average SEM abundance Average SEM Mo7 DL7 277,208 32,413 0.52 8,677,026 561,926 L7 690,905 51,838 0.50 9,094,667 809,565 DI7 345,752 39,871 0.48 7,980,027 918,179 VI7 130,443 12,595 0.47 8,288,226 656,345 DM7 266,031 20,255 0.48 7,964,136 801,753 VM7 264,177 35,178 0.49 8,160,301 754,592 S7 51,868 16,047 0.44 7,362,844 1,083,785 P7 229,389 7,052 0.47 8,861,898 711,904 Total 2,259,788 155,224 0.49 8,505,437 728,308 Amb rAmb 11,830 1,968 0.52 1,108,236 228,396 cAmb 8,350 955 0.40 727,513 105,966 Total 20,180 2,736 0.46 860,212 167,744 Mo12 DL12 54,877 17,790 0.45 883,739 225,696 VM12 37,850 6,706 0.39 667,724 132,178 Total 92,726 23,976 0.42 780,440 173,847
Figure 4. Densities of immunopositive varicosities in the orofacial motor nuclei. VGluT11 (A), VGluT21 (B), and VIAAT1 (C). Dots repre- sent average and error bars represent the standard deviation (n 5 4). Asterisks represent significant differences with the rest of the nuclei/subdivisions (two-way repeated measures ANOVA, see text).
Unlike the total densities of VGluT21 and VIAAT1 DISCUSSION varicosities, the density of perisomatic VGluT21 and This study provides estimations of volume and cell VIAAT1 varicosities are not larger in Mo7 than in any number for the orofacial motor nuclei Mo5, Mo7, Amb, other nucleus. This suggests that the large difference in and Mo12. The type and organization of afferent inputs VGluT21 and VIAAT1 varicosities between Mo7 and to these motor nuclei was also examined by estimating the other orofacial motor nuclei is due to a higher the number of total and perisomatically located amount of dendrite-located varicosities in this nucleus. VGluT11, VGluT21, and VIAAT1 varicosities. According In turn, the total density of VGluT11, VGluT21, and VIAAT1 perisomatic varicosities is larger in DL5 than in to our stereological estimations, mouse orofacial motor Mo7, Amb, and Mo12 (two-way repeated measures nuclei Mo5, Mo7, Amb, and Mo12 contain 1,000, ANOVA, P < 0.05). The proportions of perisomatic to 3,000, 600, and 1700 cells, respectively. Cell den- total varicosities are lowest in Mo12 and Mo7 and high- sity in Amb is much higher than in Mo5, Mo7, and est in Amb and Mo5 (for VGluT11, VGluT21, and Mo12. These nuclei also exhibit different patterns of VIAAT1; 0.04, 0.05, and 0.07 for Mo5; 0.02, 0.01, and glutamatergic and GABA/glycinergic innervation. Mo5 0.01 for Mo7; 0.07, 0.08, and 0.09 for Amb, and 0.01, receives a significantly higher VGluT11 innervation, pre- 0.01, and 0.01 for Mo12, respectively). sumably originating from the trigeminal mesencephalic
The Journal of Comparative Neurology | Research in Systems Neuroscience 749 M. Faunes et al.
Figure 5. Perisomatic VGluT1, VGluT2, and VIAAT innervation of the orofacial motor nuclei. High-magnification brightfield photomicrographs of Nissl and VGluT1 (A,D,G,J), VGluT2 (B,E,H,K), and VIAAT (C,F,I,L) immunostained sections of Mo5 (A–C), Mo7 (D–F), Amb (G–I), and Mo12 (J–L). Solid arrowheads indicate examples of perisomatic immunopositive varicosities. Scale bar 5 5 lm.
750 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
TABLE 4. Perisomatic Immunopositive Varicosities Number, Relative Abundances, and Densities in Mo5, Mo7, Amb, and Mo12 (n 5 4)
VGluT1 Perisomatic immunopositive varicosities number Density Relative Nucleus Subdivision Average SEM abundance Average SEM Mo5 DL5 7,053 765 0.22 60,882 4,481 VM5 159 43 0.06 11,512 2,269 Total 7,212 771 0.21 40,066 2,701 Mo7 DL7 165 24 0.04 5,152 830 L7 575 106 0.04 7,569 958 DI7 371 25 0.08 8,553 453 VI7 199 73 0.11 12,634 2,500 DM7 113 16 0.03 3,368 799 VM7 128 41 0.03 3,961 1,145 S7 61 22 0.08 8,646 1,291 Total 1,611 118 0.05 6,062 166 Amb rAmb 228 50 0.11 18,398 3,949 cAmb 123 43 0.09 9,135 2,279 Total 351 81 0.10 16,375 3,812 Mo12 DL12 128 17 0.03 2,507 519 VM12 144 11 0.04 2,504 700 Total 250 40 0.03 2,503 605
VGluT2 Perisomatic immunopositive varicosities number Density Relative Average SEM abundance Average SEM Mo5 DL5 12,301 3,633 0.39 106,185 29,337 VM5 1,427 344 0.50 103,383 19,976 Total 13,728 3,971 0.40 76,266 20,289 Mo7 DL7 1,822 546 0.49 57,033 15,341 L7 6,345 1,306 0.46 83,518 13,329 DI7 2,282 471 0.50 52,673 9,724 VI7 1,012 399 0.55 64,307 20,089 DM7 2,013 581 0.51 60,273 11,006 VM7 1,976 284 0.44 61,051 6,093 S7 404 135 0.54 57,318 7,356 Total 15,855 3,597 0.48 59,674 9,826 Amb rAmb 674 228 0.32 66,676 14,793 cAmb 757 166 0.54 64,697 22,146 Total 1,431 348 0.41 66,305 15,904 Mo12 DL12 2,196 548 0.51 36,850 8,346 VM12 1,881 804 0.49 28,615 10,098 Total 4,077 1,268 0.50 33,044 8,630 VIAAT Perisomatic immunopositive varicosities number Density Relative Average SEM abundance Average SEM Mo5 DL5 12,402 852 0.39 107,054 12,105 VM5 1,266 265 0.44 91,725 26,049 Total 13,667 776 0.39 75,932 7,260
The Journal of Comparative Neurology | Research in Systems Neuroscience 751 M. Faunes et al.
TABLE 4. Continued
VIAAT Perisomatic immunopositive varicosities number Density Relative Average SEM abundance Average SEM Mo7 DL7 1,726 107 0.46 54,037 3,011 L7 6,793 1,135 0.50 89,419 10,108 DI7 1,951 198 0.42 45,038 2,858 VI7 624 161 0.34 39,670 9,886 DM7 1,811 329 0.46 54,206 5,385 VM7 2,421 365 0.53 74,798 14,254 S7 287 77 0.38 40,679 27,004 Total 15,578 1,310 0.47 58,632 2,311 Amb rAmb 1,216 114 0.57 101,163 15,974 cAmb 520 111 0.37 45,650 9,941 Total 1,737 100 0.49 72,671 8,905 Mo12 DL12 1,950 416 0.46 31,684 4,691 VM12 1,798 660 0.47 29,925 9,544 Total 3,747 1056 0.46 30,786 6,794 nucleus afferents carrying proprioceptive activity from loss starts before 6 months. However, he did not find jaw-closing muscles, associated with the role of Mo5 in any age-related cell loss in Mo12 (Sturrock, 1990b). mastication (Pang et al., 2009). It also receives a higher The differences could also be due to the use of differ- VGluT21 and VIAAT1 innervation than Amb and Mo12. ent quantitative methods employed by the two studies Mo7, in turn, receives a much higher density of (design-based in this study, versus Abercrombie’s VGluT21 and VIAAT1 innervation than Mo5, Amb, and model-based in Sturrock’s studies; West et al., 1991). Mo12. To the best of our knowledge, this is the first Nevertheless, Nimchinsky et al. (2000), using stereol- report of quantitative differences in terminal varicosities ogy, estimated roughly the same number of Mo12 cells among orofacial motor nuclei. A summary of the main as the present study, but a 50% higher number of results of this study is provided in Figure 6. Mo7 cells.
Cell number and density Neurochemical and quantitative description In Mo5, the jaw-closing motoneurons of DL5 are of varicosities much more numerous than the jaw-opening cells of Differing patterns of innervation between Mo5, Mo7, VM5 (about 10-fold). This is as expected when compar- Amb, and Mo12 were found following estimations of ing the size of the muscles respectively innervated by the number of total and perisomatic VGluT11, them (McLoon and Andrade, 2013). The subpopulation VGluT21, and VIAAT1 varicosities. Mo5, specifically its L7 is the largest of Mo7, as would be expected due to DL5 (jaw-closing) subdivision, shows a conspicuously the importance of vibrissal muscle control in rodents. high proportion of VGluT11 varicosities when compared Finally, average cell density is significantly larger in the to the other orofacial motor nuclei, and a higher density presumably esophagomotor neurons rAmb than in of VGluT21 and VIAAT1 varicosities than Amb and cAmb, Mo5, Mo7, and Mo12. Mo12. In turn, Mo7 exhibits a much higher density of Our estimated cell numbers differ from those VGluT21 and VIAAT1 varicosities than Mo5, Amb, and obtained by Sturrock (1987, 1988, 1990a,b), who Mo12 (Table 3, Fig. 6). estimated the number of trigeminal, facial, ambiguus, Previous quantitative reports on excitatory and inhibi- and hypoglossal motoneurons to be 50%, 40%, tory innervation in rat or cat cranial motor nuclei have 25%, and 35% below ours. This discrepancy could addressed GABA, glycine, glutamic acid decarboxylase relate to differences in the age of the adult animals (GAD), and glutamate immunoreactivity in Mo5 neurons used (2 months old in this study, versus 6 months and (Saha et al., 1991; Yang et al., 1997; Bae et al., 1999, older in the aforementioned study). In fact, Sturrock 2002; Shigenaga et al., 2005, 2007). In the present shows age-related neuron loss in all three branchio- study we used immunoreactivity against vesicular trans- meric nuclei, i.e., Mo5, Mo7, and Amb (Sturrock, 1987, porters instead of neurotransmitters or synthetic 1988, 1990a), and thus it is possible that motor neuron enzymes. Unlike the expression of GABA, glycine, GAD,
752 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
Figure 6. Relative densities of motor neurons and of VGluT11, VGluT21, and VIAAT1 varicosities in the orofacial motor nuclei of the mouse. Schematic representation of coronal sections of the mouse orofacial motor depicting the relative values of motor neuron and immunopositive varicosities densities, based on the data provided in Tables 2 and 3. Note the high density of cell bodies in the Amb nucleus, the high density of VGluT21 and VIAAT1 in Mo7, and the high VGluT11 innervation in DL5. Amb, nucleus ambiguus; DL5, dorso- lateral subnucleus of the Mo5; Mo5, trigeminal motor nucleus; Mo7, facial nucleus; Mo12, hypoglossal nucleus; VM5, ventromedial subnu- cleus of the Mo5.
and glutamate, the expression of vesicular transporters Mo5 innervation is specific to presynaptic terminals and type of neuro- Mo5 neurons receive projections arising from various transmitter utilized (McIntire et al., 1997; Sagne et al., regions of the brainstem, including the mesencephalic 1997; Chaudhry et al., 1998; Fremeau et al., 2001, trigeminal nucleus, the Edinger-Westphal nucleus, the 2004; Kaneko and Fujiyama, 2002; Kaneko et al., principal and spinal sensory trigeminal nuclei, the 2002). Thus, we consider it a trustworthy marker supratrigeminal region, the raphe nucleus, and the par- for neurotransmitter liberation sites. The use of immu- vicellular, intermediate and gigantocellular medullary, nohistochemistry for VIAAT does not allow a distinction and pontine reticular formation (Kidokoro et al., 1968; to be made between GABAergic and glycinergic varicos- Holstege, 1983; Travers and Norgren, 1983; Vornov and ities. However, previous studies in rat or cat using elec- Sutin, 1983; Li et al., 1993; Mogoseanu et al., 1993; tron microscopy suggest that there is a significant Kamogawa et al., 1994; Travers et al., 2005; Paik et al., proportion of inhibitory synapses containing both neuro- 2009). While the general pattern of innervation of DL5 transmitters (Saha et al., 1991; Yang et al., 1997; Bae and VM5 is overall similar, in some of the premotor et al., 1999, 2002; Shigenaga et al., 2005, 2007). Fur- nuclei the cells of origin of the fibers reaching each ther studies will be needed to clarify differences subdivision are to some extent spatially segregated (Li between, GABAergic, glycinergic, and GABAergic/glyci- et al., 1995). Furthermore, some specific projections nergic subpopulations in mice. target preferentially DL5 (e.g., the trigeminal
The Journal of Comparative Neurology | Research in Systems Neuroscience 753 M. Faunes et al.
mesencephalic nucleus and the supratrigeminal region, GABA/glycinergic innervation representing respectively Li et al., 1995) and others target preferentially VM5 the 70% and 30% over all DL5 and the 60% and 40% (e.g., the medullary gigantocellular reticular nucleus, Li near the soma of DL5 cells. Furthermore, studies on et al., 1995; and the pontine reticular formation dorsal the innervation of the rat Mo5 neurons suggest that to the superior olive, Holstege, 1983). GABA/glycinergic boutons represent 53% apposing Tract tracing, immunohistochemical, in situ hybridiza- Mo5 neurons (Bae et al., 2002). This contrast could be tion, and physiological studies have shown that the vast due to differences between the design of the present majority of the VGluT11 varicosities in Mo5 arise from work and those of the single-cell studies. While our the mesencephalic trigeminal nucleus, as also demon- sampling was systematic, allowing every part of the strated by transection of the trigeminal motor root, nuclei under study to have an equal chance of being although some could originate from VGluT11 neurons sampled, the single-cell level studies were designed to in the reticular formation ventral to Mo5, which have always sample the soma and primary dendrites of the been described in the rat (Pang et al., 2009). VGluT11 neuron under study and therefore they better represent varicosities are preferentially located in the DL5 (Pang the innervation of these specific compartments. It is et al., 2009; present study), which is consistent with also very likely that the differences in the markers the known monosynaptic connection between muscle (vesicular transporters versus neurotransmitters) and in spindle afferents and jaw-closing motoneurons media- the animal model systems (mouse versus cat or rat) ting a jaw-closing reflex (Appenteng et al., 1978; Lin- contribute to these discrepancies. genhohl and Friauf, 1991; Luo and Li, 1991; Yabuta et al., 1996). According to our results, in the mouse Mo7 innervation these sensory afferents would represent 30% of all Neurons in Mo7 receive projections from various nuclei the terminals innervating the DL5 motor neuronal pool. in the midbrain, pons, and medulla; such as the supe- This proportion is in turn the highest for VGluT11 vari- rior colliculus, the parabrachial complex, the supratrige- cosities among the orofacial motor nuclei. minal region, the principal and spinal sensory trigeminal VGluT21 varicosities in Mo5 arise from various cen- nuclei, and the medullary reticular formation. Some of tral structures, such as the supratrigeminal region, the these afferents differ between orofacial muscle- parvicellular and intermediate reticular formation, and, projecting subdivisions and ear and eye muscle- to a lesser extent, the oralis and interpolaris subdivi- projecting subdivisions (Travers and Norgren, 1983; Li sions of the spinal trigeminal nucleus (Travers et al., et al., 1993, 1997; Pinganaud et al., 1999; Hattox 2005; Pang et al., 2009). The density of these varicos- et al., 2002; Morcuende et al., 2002; Takatoh et al., ities is higher in DL5 than in VM5; even though this dif- 2013). ference is not as dramatic as for VGluT11 varicosities VGluT11 varicosities in Mo7 represent only 2% of (see also Pang et al., 2009). They represent 41% and 47% of the varicosities we identified in DL5 and VM5, all the terminals studied in our analysis, and presum- respectively. ably originate from the mesencephalic trigeminal GABAergic and glycinergic cells projecting to Mo5 nucleus, which contains the somata of primary afferents are located in the supratrigeminal region, the parvicellu- (Pang et al., 2009). lar, intermediate and gigantocellular reticular formation, Most of the excitatory input to Mo7 is VGluT21. Ret- and the principal trigeminal sensory nucleus, as shown rograde tracing combined with immunohistochemistry in rats and cats (Kidokoro et al., 1968; Kamogawa and in situ hybridization has shown that glutamatergic et al., 1994; Li et al., 1996; Travers et al., 2005; Paik VGluT2 expressing facial premotor neurons are located et al., 2009). In the mouse, these afferents represent in the superior colliculus, the spinal trigeminal nuclei, 29% and 40% of all the glutamate, GABA, and glyciner- and many regions of the brainstem reticular formation, gic innervation of DL5 and VM5 neurons, respectively. including cells relaying input from the vibrissal motor Electron microscopy studies of the innervation of cat cortex (Takatoh et al., 2013; Matthews et al., 2014). jaw-opening Mo5-labeled neurons at the single-cell level According to our results, these afferents constitute suggest that glutamatergic and GABA/glycinergic syn- 50% of the glutamate, GABA, and glycinergic innerva- apses contacting intermediate and distal dendrites rep- tion of Mo7 neurons. resent 53% and 44% of all boutons, respectively, Most facial GAD1 and glycine1 premotor neurons while in proximity to the soma and primary dendrites are located in the pontine reticular formation, the these proportions shift to 37% and 60% (Bae et al., supratrigeminal region, the medullary reticular forma- 1999; Shigenaga et al., 2005, 2007). These percen- tion, and the spinal trigeminal sensory nuclei, as stud- tages contrast with our finding of glutamatergic and ied in the rat (Li et al., 1997). These afferents thus
754 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
constitute 50% of the glutamate, GABA, and glyciner- tion of Mo12. As in the case of Mo5 (Pang et al., gic innervation of Mo7 motoneurons. 2009), these varicosities could represent the proprio- The VGluT21 and VIAAT1 density of varicosities ceptive jaw afferents from the mesencephalic trigeminal found in Mo7 exceeds by far that found in Mo5, Amb, nucleus (Zhang et al., 2001). and Mo12. Mo7 plays a key role in a variety of behaviors The VGluT21 varicosities represent 54% of the glu- including complex exploratory whisking and sniffing, tamate, GABA, and glycinergic input to Mo12. Many which constitute an important part of the animal sensory VGluT21 premotor neurons projecting to Mo12 are perception and sophisticated behaviors (Desch^enes located in the medullary reticular formation, particularly et al., 2012). On the other hand, Amb participates in in the intermediate reticular formation (Travers et al., rhythmic and to some extent stereotyped movements 2005). A high proportion of these neurons project also involved in swallowing and in the production of vocaliza- to Mo5, thus providing a substrate for the coordinated tions, which are not as developed in mice as in humans activation of these nuclei during feeding-related behav- or songbirds (Arriaga et al., 2012), and Mo5 mainly in iors (Travers et al., 2005). mastication and other feeding behaviors. The multiple The GABA and glycinergic hypoglossal premotor neu- connections associated with the wide variety of Mo7 rons are located mainly in the pontine and medial reticular functions (Takatoh et al., 2013) may account for its formation, the Kolliker-Fuse nucleus, the spinal trigeminal higher varicosity density. Furthermore, Mo7 along with nuclei, and the intermediate zone of the cervical spinal Mo12 exhibit the lowest proportion of perisomatically cord (Li et al., 1997). These account for 42% of the gluta- located varicosities of the orofacial motor nuclei, sug- mate, GABA, and glycinergic input to Mo12. gesting a higher degree of afferent activity integration at As mentioned above, Mo12 is the only somatomotor the dendritic level in these nuclei. (non-branchiomotor) nucleus examined in this study. Furthermore, according to the work of Sturrock, it is Amb innervation the only orofacial motor nucleus that does not display The afferent connections of Amb motoneurons have age-related cell loss (Sturrock, 1987, 1988, 1990a,b). been less studied than those of the other orofacial This nucleus is also spared of cell loss in the amyotro- motor nuclei, likely due to inherent technical difficulties phic lateral sclerosis (ALS) mouse model with a G86R related to the small size and location of the nucleus. mutation in superoxide dismutase-1 (Nimchinsky et al., However, retrograde tracing studies in the rat suggest 2000). Our results suggest that these remarkable differ- that it receives projections from the nucleus of the soli- ences cannot be associated in a straight-forward man- tary tract, the parabrachial complex, the spinal trigemi- ner with differences in the pattern of VGluT11, nal nucleus, and the medullary reticular formation VGluT21, or VIAAT1 innervation. (Travers and Norgren, 1983; Nunez-Abades et al., 1990; Bao et al., 1995; Cunningham and Sawchenko, 2000; CONCLUSION Hayakawa et al., 2000; de Sousa Buck et al., 2001), particularly in the nucleus retroambiguus, which also By describing key aspects of the organization of exci- controls respiratory motoneurons and is involved in the tatory and inhibitory premotor neuron terminals on the production of vocalizations (Holstege, 1989, 2014). The mouse Mo5, Mo7, Mo12, and Amb, this study lays the neurochemical phenotypes of each of these afferents groundwork for further studies defining the identity and have not yet been identified in detail, although it is influence of the networks that regulate activity of orofa- clear that glutamate and GABA are major neurotrans- cial motor nuclei. By looking at the overall perisomatic mitters in this nucleus (Mueller et al., 2004; Sun et al., and neuropil localization of terminals, this study will 2008, 2010; Suzuki et al., 2010). also provide a comparative framework for future studies looking at structural correlates of function (Henny Mo12 innervation et al., 2012), as well as of neurodegeneration in motor The innervation of Mo12 originates from pontine and diseases such as ALS. In this regard, it is interesting to medullary structures including the parabrachial com- consider that one of the hypotheses for early degenera- plex, the supratrigeminal region, the spinal trigeminal tion of motoneurons in ALS, which include neurons of nuclei, the nucleus of the solitary tract, and the medul- Mo5, Mo7, Amb, and Mo12, is neuroexcitotoxicity, a lary reticular formation (Borke et al., 1983; Travers and condition dependent on the ratio of excitatory versus Norgren, 1983; Takada et al., 1984; Li et al., 1993, inhibitory afferent activity (Kanning et al., 2010; Quin- 1997; Travers, 2004; Travers et al., 2005). lan, 2011). Future comparative studies between vulnera- The VGluT11 varicosities represent only the 4% of ble versus nonvulnerable motoneurons (i.e., oculomotor all the of the glutamate, GABA, and glycinergic innerva- neurons) may shed light on this hypothesis (Medina et al.,
The Journal of Comparative Neurology | Research in Systems Neuroscience 755 M. Faunes et al.
1996; Kanning et al., 2010). Finally, changes in excitatory Bianchi AL, Gestreau C. 2009. The brainstem respiratory net- versus inhibitory innervation of these nuclei may also work: an overview of a half century of research. Respir Physiol Neurobiol 168:4–12. account for the brainstem motor deficits associated with Bieger D, Hopkins DA. 1987. Viscerotopic representation of the other motor diseases such as Parkinson’s disease (Plow- upper alimentary tract in the medulla oblongata in the rat: man-Prine et al., 2009) or bruxism (Bader and Lavigne, the nucleus ambiguus. J Comp Neurol 262:546–562. 2000), in which the functionality of these motor neuronal Borke RC, Nau ME, Ringler RL Jr. 1983. Brain stem afferents of hypoglossal neurons in the rat. Brain Res 269:47–55. groups is disrupted. Buttner-Ennever€ J, Holstege G. 1985. Anatomy of premotor centers in the reticular formation controlling oculomotor, ACKNOWLEDGMENTS skeletomotor and autonomic motor systems. Prog Brain We thank the Centro de Investigaciones M edicas (CIM) Res 64:89–98. Chaudhry FA, Reimer RJ, Bellocchio EE, Danbolt NC, Osen KK, and Direccion de Investigacion (DIEMUC) for the use of Edwards RH, Storm-Mathisen J. 1998. The vesicular their Microscopy Unit, funded by MECESUP PUC0815 GABA transporter, VGAT, localizes to synaptic vesicles in grant for scientific equipment. This work was also sup- sets of glycinergic as well as GABAergic neurons. J Neurosci 18:9733–9750. ported by an Interdepartamental Grant awarded to PH and Chen Z, Travers JB. 2003. Inactivation of amino acid receptors Dr. Pablo Fuentealba (Departamento de Psiquiatr ıa, UC). in medullary reticular formation modulates and sup- We thank Paulina Merino for technical support, and Dr presses ingestion and rejection responses in the awake Natalie Doig, Prof. Jorge Mpodozis, and Dr. Cristi an rat. Am J Physiol Regul Integr Comp Physiol 285:R68–83. Chen Z, Travers SP, Travers JB. 2001. Muscimol infusions in Guti errez-Ib anez~ for critical reading of the article. the brain stem reticular formation reversibly block inges- tion in the awake rat. Am J Physiol Regul Integr Comp CONFLICT OF INTEREST Physiol 280:R1085–1094. Cunningham ET Jr, Sawchenko PE. 2000. Dorsal medullary The authors declare no conflicts of interest. pathways subserving oromotor reflexes in the rat: impli- cations for the central neural control of swallowing. J Comp Neurol 417:448–466. ROLE OF AUTHORS de Sousa Buck H, Caous CA, Lindsey CJ. 2001. Projections of MF designed and supervised the study, carried out the paratrigeminal nucleus to the ambiguus, rostroven- quantification of varicosities and statistical analysis, pre- trolateral and lateral reticular nuclei, and the solitary tract. Auton Neurosci 87:187–200. pared figures and tables, and wrote the article. AO carried Desch^enes M, Moore J, Kleinfeld D. 2012. Sniffing and whisk- out the quantification of varicosities in all nuclei and car- ing in rodents. Curr Opin Neurobiol 22:243–250. ried out statistical analysis. SF carried out the quantifica- Doig NM, Moss J, Bolam JP. 2010. Cortical and thalamic innervation of direct and indirect pathway medium-sized tion of cell bodies and prepared figures. PH designed and spiny neurons in mouse striatum. J Neurosci 30:14610– supervised the study, prepared figures, and wrote the arti- 14618. cle. All authors read and commented on the article. Faunes M, Fernandez S, Gutierrez-Ibanez C, Iwaniuk AN, Wylie DR, Mpodozis J, Karten HJ, Marin G. 2013. Laminar seg- regation of GABAergic neurons in the avian nucleus LITERATURE CITED isthmi pars magnocellularis: a retrograde tracer and Appenteng K, O’Donovan MJ, Somjen G, Stephens JA, Taylor comparative study. J Comp Neurol 521:1727–1742. A. 1978. The projection of jaw elevator muscle spindle Franklin KBJ, Paxinos G. 2008. The mouse brain in stereotaxic afferents to fifth nerve motoneurones in the cat. coordinates. San Diego: Elsevier Academic Press. J Physiol 279:409–423. Fremeau RT Jr, Troyer MD, Pahner I, Nygaard GO, Tran CH, Arriaga G, Zhou EP, Jarvis ED. 2012. Of mice, birds, and men: Reimer RJ, Bellocchio EE, Fortin D, Storm-Mathisen J, the mouse ultrasonic song system has some features Edwards RH. 2001. The expression of vesicular gluta- similar to humans and song-learning birds. PLoS One 7: mate transporters defines two classes of excitatory syn- e46610. apse. Neuron 31:247–260. Bader G, Lavigne G. 2000. Sleep bruxism; an overview of an Fremeau RT Jr, Voglmaier S, Seal RP, Edwards RH. 2004. oromandibular sleep movement disorder. Sleep Med Rev VGLUTs define subsets of excitatory neurons and suggest 4:27–43. novel roles for glutamate. Trends Neurosci 27:98–103. Bae YC, Nakamura T, Ihn HJ, Choi MH, Yoshida A, Moritani M, Galvin KE, King CT, King MS. 2004. Stimulation of specific regions Honma S, Shigenaga Y. 1999. Distribution pattern of of the parabrachial nucleus elicits ingestive oromotor behav- inhibitory and excitatory synapses in the dendritic tree of iors in conscious rats. Behav Neurosci 118:163–172. single masseter alpha-motoneurons in the cat. J Comp Grillner S, Hellgren J, Menard A, Saitoh K, Wikstrom MA. Neurol 414:454–468. 2005. Mechanisms for selection of basic motor pro- Bae YC, Choi BJ, Lee MG, Lee HJ, Park KP, Zhang LF, Honma grams—roles for the striatum and pallidum. Trends Neu- S, Fukami H, Yoshida A, Ottersen OP, Shigenaga Y. rosci 28:364–370. 2002. Quantitative ultrastructural analysis of glycine- and Guti errez-Ib anez~ C, Iwaniuk AN, Lisney TJ, Faunes M, Marin gamma-aminobutyric acid-immunoreactive terminals on GJ, Wylie DR. 2012. Functional implications of species trigeminal alpha- and gamma-motoneuron somata in the differences in the size and morphology of the isthmo rat. J Comp Neurol 442:308–319. optic nucleus (ION) in birds. PLoS One 7:e37816. Bao X, Wiedner EB, Altschuler SM. 1995. Transsynaptic local- Hattox AM, Priest CA, Keller A. 2002. Functional circuitry ization of pharyngeal premotor neurons in rat. Brain Res involved in the regulation of whisker movements. J Comp 696:246–249. Neurol 442:266–276.
756 The Journal of Comparative Neurology | Research in Systems Neuroscience VGluT1, VGluT2, and VIAAT in orofacial nuclei
Hayakawa T, Takanaga A, Maeda S, Ito H, Seki M. 2000. Landers M, Philip Zeigler H. 2006. Development of rodent Monosynaptic inputs from the nucleus tractus solitarii to whisking: trigeminal input and central pattern generation. the laryngeal motoneurons in the nucleus ambiguus of Somatosens Mot Res 23:1–10. the rat. Anat Embryol 202:411–420. Li YQ, Takada M, Mizuno N. 1993. Identification of premotor Henny P, Brown MT, Northrop A, Faunes M, Ungless MA, interneurons which project bilaterally to the trigeminal Magill PJ, Bolam JP. 2012. Structural correlates of heter- motor, facial or hypoglossal nuclei: a fluorescent retrograde ogeneous in vivo activity of midbrain dopaminergic neu- double-labeling study in the rat. Brain Res 611:160–164. rons. Nat Neurosci 15:613–619. Li YQ, Takada M, Kaneko T, Mizuno N. 1995. Premotor neu- Henny P, Brown MT, Micklem BR, Magill PJ, Bolam JP. 2014. rons for trigeminal motor nucleus neurons innervating Stereological and ultrastructural quantification of the the jaw-closing and jaw-opening muscles: differential dis- afferent synaptome of individual neurons. Brain Struct tribution in the lower brainstem of the rat. J Comp Neu- Funct 219:631–640. rol 356:563–579 Holstege G. 1989. Anatomical study of the final common path- Li YQ, Takada M, Kaneko T, Mizuno N. 1996. GABAergic and way for vocalization in the cat. J Comp Neurol 284:242– glycinergic neurons projecting to the trigeminal motor 252. nucleus: a double labeling study in the rat. J Comp Neu- Holstege G. 2014. The periaqueductal gray controls brainstem rol 373:498–510. emotional motor systems including respiration. Prog Li YQ, Takada M, Kaneko T, Mizuno N. 1997. Distribution of Brain Res 209:379–405. GABAergic and glycinergic premotor neurons projecting Holstege G, Kuypers HG, Dekker JJ. 1977. The organization of to the facial and hypoglossal nuclei in the rat. J Comp the bulbar fibre connections to the trigeminal, facial and Neurol 378:283–294. hypoglossal motor nuclei II. An autoradiographic tracing Lingenhohl K, Friauf E. 1991. Sensory neurons and motoneur- study in cat. Brain 100:265–286. ons of the jaw-closing reflex pathway in rats: a combined Holstege G, Graveland G, Bijker-Biemond C, Schuddeboom I. morphological and physiological study using the intracel- 1983. Location of motoneurons innervating soft palate, lular horseradish peroxidase technique. Exp Brain Res pharynx and upper esophagus. Anatomical evidence for 83:385–396. a possible swallowing center in the pontine reticular for- Luo PF, Li JS. 1991. Monosynaptic connections between neu- mation. Brain Behav Evol 23:47–62. rons of trigeminal mesencephalic nucleus and jaw- Hrabovszky E, Turi GF, Kallo I, Liposits Z. 2004. Expression of closing motoneurons in the rat: an intracellular horserad- vesicular glutamate transporter-2 in gonadotropin- ish peroxidase labelling study. Brain Res 559:267–275. releasing hormone neurons of the adult male rat. Endo- Matthews DW, Desch^enes M, Furuta T, Moore JD, Wang F, crinology 145:4018–4021. Karten HJ, Kleinfeld D. 2014. Feedback in the brainstem: Ichikawa T, Ajiki K, Matsuura J, Misawa H. 1997. Localization an excitatory disynaptic pathway for control of whisking. of two cholinergic markers, choline acetyltransferase and J Comp Neurol 523:921–942. vesicular acetylcholine transporter in the central nervous McIntire SL, Reimer RJ, Schuske K, Edwards RH, Jorgensen system of the rat: in situ hybridization histochemistry and EM. 1997. Identification and characterization of the immunohistochemistry. J Chem Neuroanat 13:23–39. vesicular GABA transporter. Nature 389:870–876. Jean A. 2001. Brain stem control of swallowing: neuronal net- McLoon LK, Andrade F. 2013. Craniofacial muscles: a new work and cellular mechanisms. Physiol Rev 81:929–969. framework for understanding the effector side of cranio- Kamogawa H, Manabe K, Kondo M, Naito K. 1994. Supra- and facial muscle control. Heidelberg, Germany: Springer. juxtatrigeminal inhibitory premotor neurons with bifurcat- Medina L, Figueredo-Cardenas G, Rothstein JD, Reiner A. ing axons projecting to masseter motoneurons on both 1996. Differential abundance of glutamate transporter sides. Brain Res 639:85–92. subtypes in amyotrophic lateral sclerosis (ALS)-vulnera- Kaneko T, Fujiyama F. 2002. Complementary distribution of ble versus ALS-resistant brain stem motor cell groups. vesicular glutamate transporters in the central nervous Exp Neurol 142:287–295. system. Neurosci Res 42:243–250. Michalski D, Hartig W, Krugel K, Edwards RH, Boddener M, Kaneko T, Fujiyama F, Hioki H. 2002. Immunohistochemical Bohme L, Pannicke T, Reichenbach A, Grosche A. 2013. localization of candidates for vesicular glutamate trans- Region-specific expression of vesicular glutamate and porters in the rat brain. J Comp Neurol 444:39–62. GABA transporters under various ischaemic conditions Kanning KC, Kaplan A, Henderson CE. 2010. Motor neuron in mouse forebrain and retina. Neuroscience 231:328–344. diversity in development and disease. Annu Rev Neurosci Mogoseanu D, Smith AD, Bolam JP. 1993. Monosynaptic 33:409–440. innervation of trigeminal motor neurones involved in Kidokoro Y, Kubota K, Shuto S, Sumino R. 1968. Possible mastication by neurones of the parvicellular reticular for- interneurons responsible for reflex inhibition of moto- mation. J Comp Neurol 336:53–65. neurons of jaw-closing muscles from the inferior dental Moore JD, Deschenes M, Furuta T, Huber D, Smear MC, nerve. J Neurophysiol 31:709–716. Demers M, Kleinfeld D. 2013. Hierarchy of orofacial Kinzeler NR, Travers SP. 2008. Licking and gaping elicited by rhythms revealed through whisking and breathing. Nature microstimulation of the nucleus of the solitary tract. Am 497:205–210. J Physiol Regul Integr Comp Physiol 295:R436–R448. Morcuende S, Delgado-Garcia JM, Ugolini G. 2002. Neuronal Kitamura S, Nagase Y, Chen K, Shigenaga Y. 1993. Nucleus premotor networks involved in eyelid responses: retrograde ambiguus of the rabbit: cytoarchitectural subdivision and transneuronal tracing with rabies virus from the orbicularis myotopical and neurotopic representations. Anat Rec oculi muscle in the rat. J Neurosci 22:8808–8818. 237:109–123. Mueller PJ, Buckwalter JB, Clifford PS. 2004. Tracheal tone Komiyama M, Shibata H, Suzuki T. 1984. Somatotopic repre- and the role of ionotropic glutamate receptors in the sentation of facial muscles within the facial nucleus of nucleus ambiguus. Brain Res 1021:54–62. the mouse. Brain Behav Evol 24:144–151. Nakamura Y, Katakura N. 1995. Generation of masticatory Krammer EB, Rath T, Lischka MF. 1979. Somatotopic organi- rhythm in the brainstem. Neurosci Res 23:1–19. zation of the hypoglossal nucleus: a HRP study in the Nimchinsky EA, Young WG, Yeung G, Shah RA, Gordon JW, rat. Brain Res 170:533–537. Bloom FE, Morrison JH, Hof PR. 2000. Differential
The Journal of Comparative Neurology | Research in Systems Neuroscience 757 M. Faunes et al.
vulnerability of oculomotor, facial, and hypoglossal nuclei Sun QJ, Berkowitz RG, Pilowsky PM. 2008. GABA A mediated in G86R superoxide dismutase transgenic mice. J Comp inhibition and post-inspiratory pattern of laryngeal con- Neurol 416:112–125. strictor motoneurons in rat. Respir Physiol Neurobiol Nistri A, Ostroumov K, Sharifullina E, Taccola G. 2006. Tuning 162:41–47. and playing a motor rhythm: how metabotropic gluta- Sun HZ, Zhao SZ, Cui XY, Ai HB. 2010. Effects and mechanisms mate receptors orchestrate generation of motor patterns of L-glutamate microinjected into nucleus ambiguus on gas- in the mammalian central nervous system. J Physiol tric motility in rats. Chin Med J 123:1052–1057. 572(Pt 2):323–334. Suzuki T, Nakazawa K, Shiba K. 2010. Swallow-related inhibition Nunez-Abades PA, Portillo F, Pasaro R. 1990. Characterisation of in laryngeal motoneurons. Neurosci Res 67:327–333. afferent projections to the nucleus ambiguus of the rat by Takada M, Itoh K, Yasui Y, Mitani A, Nomura S, Mizuno N. means of fluorescent double labelling. J Anat 172:1–15. 1984. Distribution of premotor neurons for the hypoglos- Paik SK, Lee HJ, Choi MK, Cho YS, Park MJ, Moritani M, sal nucleus in the cat. Neurosci Lett 52:141–146. Yoshida A, Kim YS, Bae YC. 2009. Ultrastructural analy- Takatoh J, Nelson A, Zhou X, Bolton MM, Ehlers MD, Arenkiel sis of glutamate-, GABA-, and glycine-immunopositive BR, Mooney R, Wang F. 2013. New modules are added boutons from supratrigeminal premotoneurons in the rat to vibrissal premotor circuitry with the emergence of trigeminal motor nucleus. J Neurosci Res 87:1115–1122. exploratory whisking. Neuron 77:346–360. Pang YW, Ge SN, Nakamura KC, Li JL, Xiong KH, T, Mizuno N. Terashima T, Kishimoto Y, Ochiishi T. 1993. Musculotopic 2009. Axon terminals expressing vesicular glutamate organization of the facial nucleus of the reeler mutant transporter VGLUT1 or VGLUT2 within the trigeminal mouse. Brain Res 617:1–9. motor nucleus of the rat: origins and distribution pat- Terashima T, Kishimoto Y, Ochiishi T. 1994. Musculotopic terns. J Comp Neurol 512:595–612. organization in the motor trigeminal nucleus of the reeler Pinganaud G, Bernat I, Buisseret P, Buisseret-Delmas C. 1999. mutant mouse. Brain Res 666:31–42. Trigeminal projections to hypoglossal and facial motor Travers JB. 2004. Oromotor nuclei. In: Paxinos G, editor. The nuclei in the rat. J Comp Neurol 415:91–104. rat nervous system. Amsterdam: Elsevier Academic Press. Plowman-Prine EK, Sapienza CM, Okun MS, Pollock SL, Travers JB, Norgren R. 1983. Afferent projections to the oral Jacobson C, Wu SS, Rosenbek JC. 2009. The relationship motor nuclei in the rat. J Comp Neurol 220:280–298. between quality of life and swallowing in Parkinson’s dis- Travers JB, Dinardo LA, Karimnamazi H. 1997. Motor and pre- ease. Mov Disord 24:1352–1358. motor mechanisms of licking. Neurosci Biobehav Rev 21: Quinlan KA. 2011. Links between electrophysiological and 631–647. molecular pathology of amyotrophic lateral sclerosis. Travers JB, Yoo JE, Chandran R, Herman K, Travers SP. 2005. Integr Comp Biol 51:913–925. Neurotransmitter phenotypes of intermediate zone retic- Sagne C, El Mestikawy S, Isambert MF, Hamon M, Henry JP, ular formation projections to the motor trigeminal and Giros B, Gasnier B. 1997. Cloning of a functional vesicu- hypoglossal nuclei in the rat. J Comp Neurol 488:28–47. lar GABA and glycine transporter by screening of genome Uemura-Sumi M, Itoh M, Mizuno N. 1988. The distribution of databases. FEBS Lett 417:177–183. hypoglossal motoneurons in the dog, rabbit and rat. Anat Saha S, Appenteng K, Batten TF. 1991. Quantitative analysis Embryol 177:389–394. and postsynaptic targets of GABA-immunoreactive bou- Villalba RM, Raju DV, Hall RA, Smith Y. 2006. GABAB tons within the rat trigeminal motor nucleus. Brain Res receptors in the centromedian/parafascicular thalamic 561:128–138. nuclear complex: an ultrastructural analysis of GABABR1 Shigenaga Y, Moritani M, Oh SJ, Park KP, Paik SK, Bae JY, and GABABR2 in the monkey thalamus. J Comp Neurol Kim HN, Ma SK, Park CW, Yoshida A. 2005. The distribu- 496:269–287. tion of inhibitory and excitatory synapses on single, Vornov JJ, Sutin J. 1983. Brainstem projections to the normal reconstructed jaw-opening motoneurons in the cat. Neu- and noradrenergically hyperinnervated trigeminal motor roscience 133:507–518. nucleus. J Comp Neurol 214:198–208. Shigenaga Y, Bae YC, Moritani M, Yoshida A. 2007. Spatial West MJ, Slomianka L, Gundersen HJG. 1991. Unbiased ster- distribution patterns of excitatory and inhibitory synap- eological estimation of the total number of neurons in ses in the dendritic tree differ between jaw-closing and - the subdivisions of the rat hippocampus using the optical opening motoneurons. Arch Oral Biol 52:321–324. fractionator. Anat Rec 231:482–497. Stanek Iv E, Cheng S, Takatoh J, Han B-X, Wang F. 2014. Yabuta NH, Yasuda K, Nagase Y, Yoshida A, Fukunishi Y, Monosynaptic premotor circuit tracing reveals neural Shigenaga Y. 1996. Light microscopic observations of substrates for oro-motor coordination. ELife 3(e02511). the contacts made between two spindle afferent types Stensrud MJ, Chaudhry FA, Leergaard TB, Bjaalie JG, Gundersen and alpha-motoneurons in the cat trigeminal motor V. 2013. Vesicular glutamate transporter-3 in the rodent nucleus. J Comp Neurol 374:436–450. brain: vesicular colocalization with vesicular c-aminobutyric Yang HW, Min MY, Appenteng K, Batten TF. 1997. Glycine- acid transporter. J Comp Neurol 521:3042–3056. immunoreactive terminals in the rat trigeminal motor Sturrock RR. 1987. Changes in the number of neurons in the nucleus: light- and electron-microscopic analysis of their mesencephalic and motor nuclei of the trigeminal nerve relationships with motoneurones and with GABA- in the ageing mouse brain. J Anat 151:15–25. immunoreactive terminals. Brain Res 749:301–319. Sturrock RR. 1988. Loss of neurons from the motor nucleus Zhang J, Luo P, Pendlebury WW. 2001. Light and electron of the facial nerve in the ageing mouse brain. J Anat microscopic observations of a direct projection from 160:189–194. mesencephalic trigeminal nucleus neurons to hypoglossal Sturrock RR. 1990a. A comparison of age-related changes in motoneurons in the rat. Brain Res 917:67–80. neuron number in the dorsal motor nucleus of the vagus Zhang J, Yang LM, Pan XD, Lin N, Chen XC. 2013. Increased and the nucleus ambiguus of the mouse. J Anat 173: vesicular gamma-GABA transporter and decreased phos- 169–176. phorylation of synapsin I in the rostral preoptic area is Sturrock RR. 1990b. Stability of motor neuron and inter- associated with decreased gonadotrophin-releasing hor- neuron number in the hypoglossal nucleus of the ageing mone and c-Fos coexpression in middle-aged female mouse brain. Ann Anat 173:113–116. mice. J Neuroendocrinol 25:753–761.
758 The Journal of Comparative Neurology | Research in Systems Neuroscience