Extraocular muscles involved in convergence are innervated by an additional set of palisade endings that may differ in their excitability: A human study

Karoline Lienbacher 1, Kathrin Sänger 2, Sebastian Strassburger 2, Oliver Ehrt 3, Günther Rudolph 3, Miriam Barnerssoi 2, Anja K E Horn 4

Abstract

Palisade endings are located at the myotendinous junction of extraocular muscles in most mammals. Irrespective of their unclarified function as motor or sensory nerve endings, a specialized role in convergence is proposed, based on their high number in the medial rectus muscle (MR). Further support comes from a study in monkey demonstrating that only the MR and inferior rectus muscle (IR) contain an additional population of palisade endings that express the calcium-binding protein calretinin (CR) in addition to choline acetyltransferase (ChAT). Here we studied, whether CR-positive palisade endings are present in human as well and confined to extraocular muscles most active during convergence. The systematic analysis of all eye muscles of 17 human specimen revealed that only the MR and IR contain an additional population of CR-positive palisade endings and multiple en-grappe endings, which target non-twitch muscle fibers along their whole length. Approximately 80% of all palisade endings in the MR expressed CR. Furthermore, the intrafusal muscle fibers of some muscle spindles in the MR were innervated by CR-positive annulospiral nerve endings that transmit the signals of muscle length changes to the brain. All extraocular muscles contained few thin CR-positive, but ChAT-negative nerve fibers, possibly representing free sensory or autonomic endings arising from the trigeminal ganglion. As in monkey, in the medial periphery of the human oculomotor nucleus ChAT-positive neurons were found to co-express CR. Therefore these neurons most likely represent the cell bodies of CR-positive palisade endings in the MR. Unlike in monkey, these neurons do not lie within a compact cell group, but are more scattered.
In conclusion, the MR and IR in human contain two histochemically different populations of palisade and multiple endings that may contribute to ocular alignment and convergence in a different way.

Keywords
Oculomotor nucleus, Calretinin, Non-twitch muscle fibers, Multiply-innervated muscle fibers, C-group

1 Introduction

Extraocular muscles have a highly complex architecture enabling the whole spectrum of fast and slow eye movements to allow fixation of stationary or moving visual targets and gaze stabilization during own body movements (Leigh and Zee, 2015). Each muscle consists of an orbital and a global layer. Only the global layer inserts directly at the sclera via a tendon. Based on their innervation two main categories of muscle fibers can be distinguished in the global layer: (1) twitch fibers targeted by a single “en-plaque” endplate (singly-innervated fibers, SIF) and (2) non-twitch fibers contacted by multiple “en-grappe” nerve endings (multiply-innervated fibers, MIF) along the whole muscle length (Spencer and Porter, 2006). The myotendinous junction of the global layer contains a third nerve ending type, the palisade endings. They form a cuff around the tip of MIFs and provide numerous synaptic contacts at the tendon and to lesser extent at the muscle fiber (Blumer et al., 2016). Although a proprioceptive eye position signal has been discovered in the primary somatosensory cortex in monkey (Wang et al., 2007) and sensorimotor cortex in human (Balslev et al., 2011), the receptors and exact pathways that transmit these signals are not clear yet. Classical proprioceptors like muscle spindles or Golgi tendon organs are poorly developed in extraocular muscles or even absent in some species. Accordingly, palisade endings were considered as possible proprioceptors for this function in extraocular muscles (Ruskell, 1999; Steinbach, 2000).
However, there are two properties that put a sensory role into question and rather support a motor function of palisade endings. The first is their cholinergic nature (Blumer et al., 2009), and the second is the location of their cell bodies in the peripheral groups of the motonuclei of extraocular muscles in the brainstem (Lienbacher et al., 2011; Zimmermann et al., 2011). Irrespective of their exact nature, a special role of palisade endings in convergence is indicated by their high number in the medial rectus muscle (MR) compared to other extraocular muscles (Blumer et al., 2016). This is supported by the recent discovery of an additional population of calretinin (CR)-immunoreactive palisade endings only present in the medial rectus (MR) and inferior rectus muscle (IR) in monkeys (Lienbacher et al., 2018). CR is a calcium binding protein which may modulate neuron excitability by rapid calcium buffering (Camp and Wijesinghe, 2009).
Here we investigated, whether a similar CR-positive subpopulation of palisade endings is present in human extraocular muscles as well. We further explored the location of their cell bodies within the midbrain by their chemical signature.

2 Methods

Human extraocular muscles were obtained 24–48h after death from 17 bodies donated to the Institute of Anatomy of the Ludwig-Maximilians University with written consent of the donor at lifetime. Human brainstem tissue (9 control cases; 24–72h after death) was obtained from the Reference Center for Neurodegenerative Disorders of the Ludwig-Maximilians University. All procedures were approved by the Local Research Ethics Committees, and the study is carried out in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. The age of the donors ranged from 60 to 96years.
After removal of the eye from the orbit all eye muscles were dissected from the bulbus leaving the myotendinous junction intact. The muscles were fixed in 100% methanol or in 4% paraformaldehyde for at least 3 days and were then transferred through increasing concentrations of sucrose for frozen sectioning. The extraocular muscles were cut flat from the orbital to the global layer at 20μm thickness using a cryostat (Thermo Scientific Microm HM 560; Fisher Scientific, Germany) and thawmounted onto glass slides (Superfrost Plus; Thermo Scientific, Germany). Sections containing the global layer with the myotendinous junction were processed for simultaneous detection of either synaptosomal associated protein 25 (SNAP-25) to visualize the complete innervation or choline acetyltransferase (ChAT) and CR with immunofluorescence methods as described previously (Lienbacher et al., 2018).
The brain tissue was immersed either in 4% paraformaldehyde in 0.1M phosphate buffer, pH 7.4, or in 10% formalin for 7 days. Blocks of the midbrain were embedded in paraffin, and serial sections of 7 and 10μm thickness from each case were cut. After deparaffination and antigen retrieval, neighboring 7μm thick sections were stained for either CR or ChAT using immunoperoxidase protocols as described before (Che Ngwa et al., 2014).

3 Results

3.1 Identification of calretinin-positive endings in the extraocular muscles

After a systematic analysis we found that all palisade endings in all extraocular muscles expressed the cholinergic marker ChAT, but only the MR and IR contained a subset of palisade endings that expressed CR-immunoreactivity in addition (Fig. 1A–D). The analysis of 7 human MR revealed that 50–80% of identified palisade endings in the distal myotendinous junction expressed CR. Further, the vast majority of en-grappe endings in the MR and IR expressed only ChATimmunoreactivity (Fig. 1G and H), but a consistent portion contains CR as well (Fig. 1E and F).
Fluorescence photograph of nerve endings in the extraocular muscles stained for calretinin (CR) and choline acetyltransferase (ChAT) or synaptosomal nerveassociated protein 25 (SNAP-25). CR-positive palisade endings of the MR (A and B; arrows) contain ChAT as well (C and D; arrows). Only few en-grappe endings in the MR and IR express CR-immunoreactivity (E and F; arrows), but not in other eye muscles as the inferior oblique (IO) (G and H; arrows); but en-grappe endings are always ChAT-positive (H; arrows). Some muscle spindles in MR and IR have CR-positive annulospiral endings (K and L), but other spindles lack CR (I and J). CR-positive spiral endings were only found in the MR (closed arrow). Note the single CR-positive fiber traveling within the nerve fiber bundle stained for SNAP-25 entering the muscle (M and N, open arrow). ChAT-positive en-plaque endings did not express CR (O and P). Some thin CR-positive nerve fibers lacking ChAT may represent autonomic or sensory fibers (Q and R). Scale bar: 20μm.
CR-positive palisade and multiple endings had no preferential location, but they were intermingled with CR-negative counterparts. With SNAP-25immunostaining we could identify few muscle spindles by annulospiral nerve endings around the intrafusal muscle fibers in all extraocular muscles (Fig. 1I–L). Only in the MR and IR some muscle spindle afferents were present expressing CR-immunoreactivity (Fig. 1K and L).
In addition, only in the MR some CR-positive spiral endings were observed (Fig. 1M and N). In all extraocular muscles, the en-plaque endings targeting fast twitch muscle fibers in the global layer expressed only ChAT, but not CR (Fig. 1O and P). Furthermore, few thin CR-positive nerve fibers meandering between the muscle fibers were observed in all extraocular muscles, that did not express ChAT-immunoreactivity (Fig. 1Q and R).

3.2 Location of putative cell bodies of calretinin-positive palisade endings

Based on the identification of a large population of CR-positive palisade endings in the MR, the oculomotor nucleus (nIII) region was analyzed for the presence of neurons expressing ChAT- and CR-immunoreactivity in neighboring thin paraffin sections. ChAT-immunostaining revealed typical multipolar motoneurons within the cytoarchitectural boundaries of the nIII and the Perlia nucleus (NP) at the midline at more rostral planes. These correspond to SIF motoneurons innervating twitch muscle fibers (Horn et al., 2008). Smaller, more oval shaped ChAT-positive neurons were scattered in the periphery and at the midline between both nIII. The close analysis of adjacent thin 7μm sections immunostained either for ChAT or CR revealed double-labeled neurons at sections through midplanes of nIII (Fig. 2A and B). Except few neurons within the nIII boundaries, neurons coexpressing ChAT and CR were mainly found in the periphery of the medial aspect of nIII. These double-labeled neurons did not form a compact cell group, but lie rather scattered between both nIII and further rostrally between nIII and NP (Fig. 2C and D). Based on their location and coexpression of ChAT and CR we consider these neurons as the cell bodies of CR-positive palisade endings and/or multiple endings of the MR and IR in correspondence to monkey (Lienbacher et al., 2018).
Taken together, we identified subpopulations of CR-positive palisade and multiple en-grappe endings, of spiral and annulospiral endings only in the MR and IR. The ChAT- and CR-positive neurons at the medial nIII border most likely represent the cell bodies of CR- and ChAT-positive palisade and/or multiple endings in MR and IR.
Brightfield photographs of adjacent thin transverse sections through the oculomotor nucleus (nIII) in human stained for ChAT (A) or CR (B). The arrows indicate two ChAT-positive putative cell bodies of palisade endings (A) that express CR (B). Plot of ChAT-positive neurons (gray dots) and ChAT- and CR-positive neurons (red dots) in two transverse sections through the oculomotor nucleus: a caudal plane (C) and a more rostral plane (D). Double-labeled neurons are primarily found at the medial border of nIII. The rectangle in (C) indicates the area of the photograph in (A) and (B). CCN: central caudal nucleus; EWcp: (non-preganglionic) central projecting neurons of the Edinger-Westphal nucleus, NP: nucleus of Perlia.

4 Discussion

The present study demonstrates in human that a consistent portion of palisade endings and en-grappe endings only in the MR and IR contain the calcium-binding protein CR, which is absent from en-plaque endings in all extraocular muscles. Thus, these findings are in line with recent work in monkey showing that more than 80% of palisade endings in the distal myotendinous junction of MR and 20–40% in IR contain CR (Lienbacher et al., 2018). The analysis of sections rather than whole-mount preparations did not allow to determine, whether CR-positive palisade endings and en-grappe endings form anatomical entities, which arise from a common neuron (Lienbacher and Horn, 2012). Whole-mount preparations of extraocular muscles in cat suggested that palisade endings are continuous with axons that supply the respective muscle fiber via multiple endings (Zimmermann et al., 2013). A developmental study in cat further supports the view that palisade endings are extensions of axons targeting the non-twitch muscle fibers via multiple endings (Blumer et al., 2017). However, by way of comparison the low number of CR-positive versus CR-negative multiple endings and the high number of CR-positive versus CR-negative palisade endings indicates a different morphology of both systems, which has to be investigated in future studies.

4.1 Calretinin-expression in nerve endings in extraocular muscles

The functional significance of CR-expression in the different nerve endings is not clear. Other than a general role in the maintenance of the homeostasis of the intracellular calcium ion concentration as a buffer (Schwaller, 2009), CR is proposed to play a role in the modulation of neuronal excitability (Camp and Wijesinghe, 2009). Unlike the calcium-binding protein parvalbumin, which is found in fast-firing neurons (Morris et al., 1999), no clear association of CR-expression with a functional neuron type is known. The selective presence of CR in premotor saccadic burstneurons for upward eye movements, but its lack in down-burst neurons, indicates that the CR-expression does not depend on the firing characteristics themselves (Adamczyk et al., 2015; Zeeh et al., 2013). The CR-expression in a subpopulation of palisade endings adds a further example, which may indicate that CR modulates more subtle properties of a given functional cell group. The fact that the majority of palisade endings in the MR and a robust number in the IR are CR-positive suggests a specialized role of these endings in both muscles. The MR and IR are primarily involved in convergence, often in synergy for viewing in the lower field of gaze during reading or working with the hands (Mok et al., 1992). One hypothesis is that all extraocular muscles contain a basic set of palisade endings, which do not express CR, and are involved in the fine alignment of the eye position during conjugate eye movements. The CR-positive palisade endings in MR and IR may represent a more excitable type of endings for sustained discharge used for the fine adjustment of eye position during convergence, when focusing on near targets (Lienbacher et al., 2018).
CR has been found in many sensory endings including afferent nerve fibers of the dental pulp (Ichikawa et al., 1995) and rapidly adapting mechanoreceptors, such as muscle spindle afferents in the hindlimb of chicken (Duc et al., 1994). In the present study we also found CR-positive annulospiral endings in a subset of muscle spindles in the MR. As other muscle spindle afferents these CR-positive afferents most probably derive from trigeminal ganglion cells. This is indicated by tract-tracing injections into the extraocular muscles that revealed a robust population of neurons in the trigeminal ganglion with different histochemical properties (Fackelmann et al., 2008) including CR (Lienbacher, own observations). Similarly, a subset of simple spiral endings only in the MR was found to express CR. Based on their structural characteristics, such as postjunctional folds, spiral endings in extraocular muscles are considered as motor nerve endings (Ruskell, 1984), which originate from the motonuclei (Lienbacher et al., 2011). Again, the functional significance of the CR-expression in a subset of muscle spindle afferents (sensory) and simple spiral endings (motor) only in the MR is unclear. A more general function may be provided by the CR-positive fine nerve fibers found in all extraocular muscles. The lack of ChAT-immunoreactivity in these fibers indicate a non-motor function. Whether these are free sensory nerve endings arising from the trigeminal ganglion or autonomic fibers of the sympathetic system remains to be studied with co-staining for tyrosine hydroxylase (Neuhuber and Schr€odl, 2011).

4.2 Location of CR-positive putative cell bodies of palisade endings in human

The location of neurons expressing CR- and ChAT-immunoreactivity at the medial aspect of the nIII found in the present study corresponds to the location of putative MIF motoneurons previously described in humans. These were identified by ChATexpression, but lack of the specialized condensed extracellular matrix in form of perineuronal nets (Horn et al., 2008). In monkey the cell bodies of MR and IR palisade and multiple endings are located in the rather circumscribed C-group at the dorsomedial border of nIII, where MR and IR neurons form two separate groups (Tang et al., 2015). Accordingly, CR-positive neurons were found intermingled with CR-negative ones within the respective MR and IR populations in the C-group of monkey (Lienbacher et al., 2018). Unlike in monkey, a compact C-group has not been identified in human so far (Horn et al., 2008). This is in line with the scattered distribution of CR- and ChAT-positive neurons throughout the periphery of nIII. This suggests a more widespread distribution of the cell bodies of palisade and multiple endings of MR and IR in human compared to monkey, but requires further investigation with different markers.

4.3 Role in convergence

Although the exact role of palisade endings is still not understood there is increasing evidence for a specialized role in convergence. Only the eye muscles of frontal-eyed species including human are consistently equipped with palisade endings, but not the eye muscles of all lateral-eyed species (Blumer et al., 2016). The largest number of palisade endings is present in MR (Blumer et al., 2016), and those are the first to develop postnatally in cat (Blumer et al., 2017). The discovery of an additional possibly more excitable CR-positive subpopulation of palisade and multiple endings only in MR and IR in monkey (Lienbacher et al., 2018) and here in man, provides further specializations in these muscles, which may contribute to convergence. Recent work in monkey revealed that along with the preganglionic neurons of the lens and pupil the C-group receives a specific input from the central mesencephalic reticular formation possible providing vergence in the near response (Bohlen et al., 2016; May et al., 2016). Up to date no attempts have been made to study the inputs specifically to CR-positive neurons within the C-group.
In conclusion there are two populations of palisade endings in the MR and IR that may contribute in a different way to fine adjustment of eye position in conjugate and vergence eye movements, respectively. In theory, the malfunction of one or both types of palisade endings, either Choline as proprioceptive organs or effectors, could cause congenital strabismus or nystagmus (Bui Quoc and Milleret, 2014), also suggested from modeling studies (Optican and Zee, 1984). Some evidence comes from morphological studies demonstrating subtle ultrastructural alterations in nerve endings at the myotendinous junction in the specimen of strabismic patients, whereas those in the muscle belly appeared normal (Domenici-Lombardo et al., 1992).

References

Adamczyk, C., Strupp, M., Jahn, K., Horn, A.K.E., 2015. Calretinin as a marker for premotor neurons involved in upgaze in human brainstem. Front. Neuroanat. 9, 153.
Balslev, D., Albert, N.B., Miall, C., 2011. Eye muscle proprioception is represented bilaterally in the sensorimotor cortex. Hum. Brain Mapp. 32, 624–631.
Blumer, R., Konakci, K.Z., Pomikal, C., Wieczorek, G., Lukas, J.R., Streicher, J., 2009. Palisade endings: cholinergic sensory organs or effector organs? Invest. Ophthalmol. Vis. Sci. 50, 1176–1186.
Blumer, R., Maurer-Gesek, B., Gesslbauer, B., Blumer, M., Pechriggl, E., Davis-Lo´pez de Carrizosa, M.A., Horn, A.K.E., May, P.J., Streicher, J., de la Cruz, R.R., Pastor, A´ .M., 2016. Palisade endings are a constant feature in the extraocular muscles of frontal-eyed, but not lateral-eyed, animals. Invest. Ophthalmol. Vis. Sci. 57, 320–331.
Blumer, R., Streicher, J., Davis-Lo´pez de Carrizosa, M.A., de la Cruz, R.R., Pastor, A.M., 2017. Palisade endings of extraocular muscles develop postnatally following different time courses. Invest. Ophthalmol. Vis. Sci. 58, 5105–5121.
Bohlen, M.O., Warren, S., May, P.J., 2016. A central mesencephalic reticular formation projection to the supraoculomotor area in macaque monkeys. Brain Struct. Funct. 221, 2209–2229.
Bui Quoc, E., Milleret, C., 2014. Origins of strabismus and loss of binocular vision. Front. Integr. Neurosci. 8, 71.
Camp, A.J., Wijesinghe, R., 2009. Calretinin: modulator of neuronal excitability. Int. J. Biochem. Cell Biol. 41, 2118–2121.
Che Ngwa, E., Zeeh, C., Messoudi, A., B€uttner-Ennever, J.A., Horn, A.K.E., 2014. Delineation of motoneuron subgroups supplying individual eye muscles in the human oculomotor nucleus. Front. Neuroanat. 8, 2.
Domenici-Lombardo, L., Corsi, M., Mencucci, R., Scrivanti, M., Faussone-Pellegrini, M.S., Salvi, G., 1992. Extraocular muscles in congenital strabismus: muscle fiber and nerve ending ultrastructure according to different regions. Ophthalmologica 205, 29–39.
Duc, C., Barakat-Walter, I., Droz, B., 1994. Innervation of putative rapidly adapting mechanoreceptors by calbindin- and calretinin-immunoreactive primary sensory neurons in the rat. Eur. J. Neurosci. 6, 264–271.
Fackelmann, K., Nouriani, A., Horn, A.K.E., B€uttner-Ennever, J.A., 2008. Histochemical characterisation of trigeminal neurons that innervate monkey extraocular muscles. Prog. Brain Res. 171, 17–20.
Horn, A.K.E., Eberhorn, A., H€artig, W., Ardelenanu, P., Messoudi, A., B€uttner-Ennever, J.A., 2008. Perioculomotor cell groups in monkey and man defined by their histochemical and functional properties: reappraisal of the Edinger-Westphal nucleus. J. Comp. Neurol. 507, 1317–1335.
Ichikawa, H., Deguchi, T., Nakago, T., Jacobowitz, D.M., Sugimoto, T., 1995. Parvalbuminand calretinin-immunoreactive trigeminal neurons innervating the rat molar tooth pulp. Brain Res. 679, 205–211.
Leigh, R.J., Zee, D.S., 2015. The Neurology of Eye Movements, fifth ed. Oxford University Press, New York.
Lienbacher, K., Horn, A.K.E., 2012. Palisade endings and proprioception in extraocular muscles: a comparison with skeletal muscles. Biol. Cybern. 106, 643–655.
Lienbacher, K., Mustari, M., Ying, H.S., B€uttner-Ennever, J.A., Horn, A.K.E., 2011. Do palisade endings in extraocular muscles arise from neurons in the motor nuclei? Invest. Ophthalmol. Vis. Sci. 52, 2510–2519.
Lienbacher, K., Ono, S., Fleuriet, J., Mustari, M., Horn, A.K.E., 2018. A subset of palisade endings only in the medial and inferior rectus muscle in monkey contain calretinin. Invest. Ophthalmol. Vis. Sci. 59, 2944–2954.
May, P.J., Warren, S., Bohlen, M.O., Barnerssoi, M., Horn, A.K.E., 2016. A central mesencephalic reticular formation projection to the Edinger–Westphal nuclei. Brain Struct. Funct. 221, 4073–4089.
Mok, D., Ro, A., Cadera, W., Crawford, J.D., Vilis, T., 1992. Rotation of Listing’s plane during vergence. Vision Res. 32, 2055–2064.
Morris, N.P., Harris, S.J., Henderson, Z., 1999. Parvalbumin-immunoreactive, fast-spiking neurons in the medial septum/diagonal band complex of the rat: intracellular recordings in vitro. Neuroscience 92, 589–600.
Neuhuber, W., Schr€odl, F., 2011. Autonomic control of the eye and the iris. Auton. Neurosci. 165, 67–79.
Optican, L., Zee, D.S., 1984. A hypothetical explanation of congenital nystagmus. Biol. Cybern. 50, 119–134.
Ruskell, G.L., 1984. Spiral nerve endings in human extraocular muscles terminate in motor end plates. J. Anat. 139, 33–43.
Ruskell, G.L., 1999. Extraocular muscle proprioceptors and proprioception. Prog. Retin. Eye Res. 18, 269–291.
Schwaller, B., 2009. The continuing disappearance of “pure” Ca2+ buffers. Cell. Mol. Life Sci. 66, 275–300.
Spencer, R.F., Porter, J.D., 2006. Biological organization of the extraocular muscles. Prog. Brain Res. 151, 43–80.
Steinbach, M.J., 2000. The palisade ending: an afferent source for eye position information in humans. In: Lennerstrand, G., Ygge, J., Laurent, T. (Eds.), Advances in Strabismus Research: Basic and Clinical Aspects. Portland Press, London, pp. 33–42.
Tang, X., B€uttner-Ennever, J.A., Mustari, M.J., Horn, A.K.E., 2015. Internal organization of medial rectus and inferior rectus muscle neurons in the C group of the oculomotor nucleus in monkey. J. Comp. Neurol. 523, 1809–1823.
Wang, X., Zhang, M., Cohen, I.S., Goldberg, M.E., 2007. The proprioceptive representation of eye position in monkey primary somatosensory cortex. Nat. Neurosci. 10, 640–646.
Zeeh, C., Hess, B.J., Horn, A.K.E., 2013. Calretinin inputs are confined to motoneurons for upward eye movements in monkey. J. Comp. Neurol. 521, 3154–3166.
Zimmermann, L., May, P.J., Pastor, A.M., Streicher, J., Blumer, R., 2011. Evidence that the extraocular motor nuclei innervate monkey palisade endings. Neurosci. Lett. 489, 89–93.
Zimmermann, L., Morado-Diaz, C.J., Davis-Lopez de Carrizosa, M.A., de la Cruz, R.R., May, P.J., Streicher, J., Pastor, A.M., Blumer, R., 2013. Axons giving rise to the palisade endings of feline extraocular muscles display motor features. J. Neurosci. 33, 2784–2793.