By Clemens Wülfing, Fenja Amrei Schuran, Inga Hölge, Julia Urban, Jasmin Oehlmann and Hauke Simon Günther
Nearly all body tissues are connected to the central nervous system through the axonal terminals of neurons. But with regard to lymphoid organs, this direct innervation, its extent and significance, as well as the role of afferent and efferent pathways remains a matter of debate [1-5]. Previous studies indicate that many lymphoid organs receive innervation distinct from the known autonomic fibers reaching blood vessels or smooth muscles, and that those axons can have very close contacts to leukocytes [6, 7]. But regarding the composition and fine structure of this parenchymal innervation our knowledge is still limited. Autonomic efferent fibers reaching primary and secondary lymphoid organs are widely accepted for the sympathetic nervous system but remain in question for the parasympathetic nervous system division [3, 8-10].
Looking at afferent fibers, the field is getting much smaller. Besides somatic fibers innervating the skin, which are in close contact to Langerhans cells, autonomic afferent fibers have been probed only in some lymphoid organs, having their origin either in the dorsal root ganglia or the vagal nodose ganglion. These findings, however, are based on only a few approaches [4, 11-13]. Thus, little is known about the extent and magnitude of these sensory pathways or their role in neuronal processing, and the integration of the immunological information either locally or in the central nervous system. One step to further unravel that innervation we took in 2015 , and afterwards we evolved the new concept of neurally hard-wired cells as a form of synaptic communication.
I. New morphology findings
Up to now, we limited our investigations to the cervical lymph nodes. Now, we examined lymph nodes at many other locations including the skin and mucosa draining lymph nodes. Additionally we investigated most other secondary lymphoid organs like Peyer’s patches, nasopharynx associated lymphatic tissue (NALT), bronchus associated lymphatic tissue (BALT), the spleen and even skin but also the primary lymphoid organs thymus and bone marrow. All results seem to fit together like a big puzzle .
Figure 1: Schematic drawing of a neurally hard-wired immune cell (wIC). In green the arriving neurite is shown, which encloses only the cell body by omitting membranous extensions. Note that the neurally hard-wired immune cells show both lymphocyte, but also myeloid shape.
I. 1. wAPC is one type of many wIC
In 2015, we described the neurally hard-wired antigen presenting cells (wAPC) in the cervical lymph nodes of rodents and humans. These cells were reached and densely enclosed by single neurite terminals particularly in the T-cell zone.
Now extending our investigations to other lymphoid organs, we detected other immune cells with a similar innervation morphology but without antigen presenting capacity. Therefore, we consequently named those cells neurally hard-wired immune cells (wIC). It is very likely that wAPC are only one possible manifestation of wIC, as antigen presenting cells are also only one group of all leukocytes. However, wAPC like wIC could be regularly observed in all other secondary lymphoid organs (Figures 1 and 3) and in the thymic medulla. Particularly, wAPC were always present in the T-cell zone of all different types of lymph nodes, like also in all other secondary lymphoid organs like Peyer’s patches, NALT and BALT, whereas wIC did not have that regular shaped appearance. Lastly, the similar presence of wIC and wAPC in the dermis was an intriguing fact, as skin is not consistently defined as a lymphoid organ in the literature [16, 17]Only in the spleen were we not able to find innervated immune cells, with one exception, where a small lymphocyte seemed to be closely associated to neural structures.
Figure 2: Schematic drawing of a neural nexus (NN). In green the high density of neurites are illustrated, located in the parenchyma below the entrance site. “X” is a wildcard for the antigen entrance site, the name depends on the type of secondary lymphoid organ (see text for details).
I. 2. Neural nexus at the gates
As shown previously in cervical lymph nodes, we verified the appearance of a high density of neurites in the subsinoidal layer in all other types of lymph nodes investigated. Moreover, it turned out that this phenomenon also occurred at other antigen entrances sites, which have a gate character, like the border between the medullary sinus and cords. Most notably was the fact that the same high density of neurites could be observed at similar gates in all other secondary lymphoid organs. This includes the subepithelial dome of Peyer’s patches, the subsinoidal layer beyond the marginal sinus of the spleen and at the outer margins of BALT and NALT. This regular phenomenon we referred to as “neural nexus” (NN) [Figures 2 and 3]. Another finding was the identification of a zone with lower cell density right below the outer margin of BALT, appearing like a subcapsular sinus in lymph nodes. Interestingly, and matching the other results – the neural nexus in BALT was located exactly beyond that sinus structure. Additionally, the skin also showed similar neural structures in the dermis resembling the morphology of a neural nexus. However, it remains unclear if these neurites in the dermis are likewise part of the systematic innervation of lymphoid organs, as it is well known that the dermis contains many free nerve endings for somatic sensory purposes.
Figure 3: Systematic pattern of neural structures in different secondary lymphoid organs, exemplary show the neural nexus (NN) and neurally hard-wired immune cells (wIC). The neural nexus is always located at antigen entrance sites: The subepithelial dome in Peyer’s patches; the layer below the marginal sinus in the spleen; the subsinoidal layer below the subcapsular sinus in lymph nodes or below a comparable “sinus-like” layer at the border of BALT. The wIC, which also include neurally hard-wired antigen presenting cells (wAPC) are always located in the T-cell zone and other parenchymal areas of lymphoid organs. Only in the spleen, the appearance of wIC is questionable. Lymph nodes, Peyer’s patches, spleen and BALT of Sprague–Dawley rats stained with monoclonal anti-neurofilament (green) and DAPI (blue).
I. 3. Neural thymic epithelial cell at the thymus gate
In the thymus, the cortico-medullary junction is the perivascular zone between the thymic cortex and medulla. This zone likewise can be seen as a gate, as its vessels are the main entrance site for thymocytes and other immune cells. Of note, here we observed another new phenomenon, namely, the thymic epithelial cells also showing associated neurofilament expression. But deviating to the finding of wAPC or wIC – this neurofilament signal not only exhibited all the cell bodies, but also followed most of the fine membranous extensions. Unfortunately, with the method of immunohistochemical staining, the exact location of the neurofilament signal could not be visualized. Two scenarios are possible: either it was located in the cytoplasm of the thymic epithelial cells themselves, or it was closely associated with their cell membrane, being part of an axonal terminal, as could be outlined for the wAPC or wIC (see Figure 4). Therefore, we propose that the thymic epithelial cells showing neurofilament signals at the corticomedullary junction should be referred to as “neural thymic epithelial cells” (nTEC).
Figure 4: Schematic drawing of neural thymic epithelial cells (nTEC). Two scenarios are possible: A) nTECs are densely covered by terminals from innervating neurites. B) nTEC express neurofilaments themselves.
II. Location and distribution of the structures
As summarized in Tables 1 and 2 (see end of text), we have seen the new neural structures with similar cytoskeletal components at two different locations in all examined primary and secondary lymphoid organs. First, they appeared as a dense network of neurites always at antigen entrance sites, which have a gate function (neural nexus and nTEC). Second, they appeared as abundant innervation of single immune cells in the T-cell zone (wIC and wAPC). Additionally, these two phenomena could be observed redundantly in both possible lymph drainage pathways. In case of the skin draining pathway, neural nexus like wIC and wAPC appeared in the dermis like also subsequently in the draining lymph nodes. In case of the mucosa draining pathway, those neural structures appeared in Peyer’s patches like also subsequently in their draining lymph nodes.
Figure 5: Tracing results of vagal afferent fibers: Fluoro-Ruby tracer-signal in superficial cervical lymph nodes of Sprague–Dawley rats 60 min after injection in the ganglion nodosum. The morphology of the neurite approaching the immune cells and enclosing only their cell bodies in a very close manner resembled the morphology seen with neurofilament staining. A: wIC in the T-cell zone, with higher magnifications in C, D, E and co-staining with anti-neurofilament in F. B: The neural nexus below the subcapsular sinus.
III. Possible physiologic and functional meaning
Thus, with all the new findings described above, we would like to expand the concept of neurally hard-wired cells as a form of synaptic communication evolved on this platform in 2016. Subsequently, we will integrate it into the larger context of possible topographic information of how the central nervous system receives information from the immune system.
III. 1. The neuro-immune synapse as an afferent connection
If/when the nervous system should get detailed information about immunological processes it is useful to “speak” and/or “listen” to single immune cells. But how to achieve a separation of such a communication within an environment full of myriads of communication factors from many different sources? One hypothesis might be a densely enclosing contact – like the one we observed in the case of wIC. Therefore, we propose to talk about the concept of a neuro-immune synapse (NI-synapse), a term used very scarcely until now . In neurobiology, the term “synapse”, historically, requires evidence for vesicular transmission. Yet, so far, it has not been shown, except for the known sympathetic innervation of lymphoid organs. Nevertheless the term “synapse” is now well established for intrinsic immunological synapses between immune cells, formed by diverse adhesion molecules and without any vesicles released . Consequently, because of the very dense enclosing contact of neurite terminals to single immune cells, we strongly suggest wIC and wAPC to be a synaptic contact with yet unknown adhesion molecules.
If we consider the options that the NI-synapse can “speak” (efferent connection) or “hear” (afferent connection) our preliminary results would suggest the latter. Tracing experiments with Fluoro-Ruby injected into the vagal nodose ganglion resembled the wIC and wAPC morphology and type of innervation of single immune cells seen with neurofilament staining (Figure 5).
Additionally, beside early works supporting our findings  other facts are relevant: for efferent purposes it will be more useful to influence whole microdomains of lymphoid organs, because of the complex interplay not only between hematopoietic cells, but also between them and the surrounding stromal cells, e.g. by building chemokine gradients. This also applies to the regulation of blood flow, as efferent signaling to specialized microdomains will be essential to distribute leukocytes to the right place in time, a known task for the sympathetic nervous system [3, 21].
III. 2. The neuro-immune synapse and the immunological homunculus
If we take a look at the greater picture and the scientific literature, an interesting concept emerged a decade ago – the immunological homunculus, which received many but quite different responses [4, 22]. Indeed, this is comprehensible, as the idea of ugly dwarfs developed in the late Middle Ages, in the context of alchemical theories, as miniature formed humans, is perhaps not suitable here. On the other hand, ugly dwarfs are famous concepts, not only in the Goethe’s Faust, and besides the use in the philosophy of perception and the mind, but also when looking at the cortical homunculus of our brain. The latter one is known as a distorted representation of the body, where specific areas of the brain are dedicated to the processing of sensory or motor functions in specific parts of the body. So why should it be ignored in case of the immune system? The concept of the immunological homunculus likewise proposes specific cortical but also subcortical representations in the brain, where different sensory information out of the immune system is processed in specific areas according to information qualities. After neuronal integration, equally organized efferent answers can also be assigned to specific areas of the brain (Figure 6).
Figure 6: Schematic illustration of our recent findings of neural structures in secondary lymphoid organs put in context with the concepts of NI-synapse and immunological homunculus. Neural structures are colored green in all drawings as we suppose them to build up one coherent sensory system. The circle in the brain with three colored parts should represent the immunological homunculus, which gets topographically mapped information from different secondary lymphoid organs. Looking at the cellular level, in all lymphoid organs, single immune cells could be found densely surrounded by neural structures (wIC), probably forming neuro-immune synapses. Likewise in all lymphoid organs a high density of neurites always in gate areas could be observed as the neural nexus phenomenon (NN). The two dashed circles represent two different innervation locations of wIC and NN (see text). LN (red): lymph nodes, GALT (brown): gut associated lymphatic tissue, BALT (yellow): bronchus associated lymphatic tissue, NALT (yellow): nasopharynx associated lymphatic tissue, CNS: central nervous system, NN: neural nexus, wIC: neurally hard-wired immune cell.
The systematic innervation patterns we observed in all lymphoid organs examined and the evolving concept of a NI-synapse seem to argue for the existence of such an immunological homunculus: Neural nexus, nTEC, wIC and wAPC as systematic neural structures (always at gates or in the T-cell zone of primary and secondary lymphoid organs) may represent the neuroanatomical and morphological correlate of topographic information for the CNS. And the NI-synapse of the wIC and wAPC may be one manifestation of the communication channels with their molecular details still to be unraveled (Figure 6).
For the sake of completeness we have to consider that the concept of the immunological homunculus has been developed earlier in another context [23, 24]. Cohen used the term for describing a possible self-image the immune system builds up from self-antigens to differentiate self and non-self. With the recent identification of extrathymic AIRE expressing cells, permanently presenting self-antigens in secondary lymphoid organs for peripheral tolerance , we have another candidate for the NI-synapse.
So what type of immunological homunculus is the right one, or “which dwarf is working in the brain”? Is it the immunological self-image build of self-antigens or is it the cortical representation of body areas dedicated to the processing of specific immune related sensory and motor functions? Maybe both dwarfs exist side by side using neural nexus, wIC and nTEC for the neuro-immune communication purposes.
We described systematic neural structures with a similar appearance in all investigated primary and secondary lymphoid organs. These neural structures are redundantly present at all sites along the antigen recognition, processing and presentation pathway particularly at important sites of first antigen contact. All these findings seem to substantiate our evolving concept of 2016to a possible neuro-immune synapse with afferent character. Moreover they could be integrated in the concept of the immunological homunculus as systematic neural structures for some kind of topographical information. Certainly, the more precise features and mechanisms behind the NI-synapses, immunological homunculi and the self-images of the immune system warrant further studies and more detailed investigation.
Table 1: Overview of the new morphology findings and their distribution. Symbols & abbreviations: +) structure is apparent, –) structure is not apparent, +/-) structures could not be clearly assigned, N.D. no data because of preparation methods.
Table 2: Overview of the new morphology findings and their location in the lymphoid tissue. The green text describes the exact location of the structures in the lymphoid tissue. Symbols & abbreviations: –) structure is not apparent, +/-) structures could not be clearly assigned, N.D. no data because of preparation methods.
Clemens Wülfing, Fenja Amrei Schuran, Inga Hölge, Julia Urban, Jasmin Oehlmann, Hauke Simon Günther– Group for Interdisciplinary Neurobiology and Immunology, Biozentrum Grindel, Institute of Zoology, University of Hamburg, Martin-Luther-King-Platz 3, 20146 Hamburg, Germany: Corresponding Author: Clemens Wülfing, email: firstname.lastname@example.org
List of Non-Standard Abbreviations
AIRE, autoimmune regulator (transcription factor for expressing self-antigens); APC, antigen presenting cell; BALT, bronchus associated lymphatic tissue; CNS, central nervous system; GALT; gut associated lymphatic tissue; LN, lymph node; NALT; nasopharynx associated lymphatic tissue; NI-synapse, neuro-immune synapse; NN; neural nexus (new finding); nTEC; neural thymic epithelial cell (new finding); wAPC; neurally hard-wired antigen presenting cell (one type of wIC); wIC, neurally hard-wired immune cell (new finding – lymphocyte like myeloid shape).
1 Chavan SS, Ma P, Chiu IM. Neuro-immune interactions in inflammation and host defense: Implications for transplantation. Am J Transplant 2018; 18: 556-63.
2 Chesne J, Cardoso V, Veiga-Fernandes H. Neuro-immune regulation of mucosal physiology. Mucosal Immunol 2019; 12: 10-20.
3 Dantzer R. Neuroimmune Interactions: From the Brain to the Immune System and Vice Versa. Physiol Rev 2018; 98: 477-504.
4 Pavlov VA, Chavan SS, Tracey KJ. Molecular and Functional Neuroscience in Immunity. Annu Rev Immunol 2018; 36: 783-812.
5 Veiga-Fernandes H, Mucida D. Neuro-Immune Interactions at Barrier Surfaces. Cell 2016; 165: 801-11.
6 Felten DL, Felten SY, Carlson SL, Olschowka JA, Livnat S. Noradrenergic and peptidergic innervation of lymphoid tissue. J Immunol 1985; 135: 755s-65s.
7 Novotny GE, Kliche KO. Innervation of lymph nodes: a combined silver impregnation and electron-microscopic study. Acta anatomica 1986; 127: 243-8.
8 Bellinger DL, Lorton D. Autonomic regulation of cellular immune function. Autonomic neuroscience : basic & clinical 2014; 182: 15-41.
9 Elenkov IJ, Wilder RL, Chrousos GP, Vizi ES. The sympathetic nerve–an integrative interface between two supersystems: the brain and the immune system. Pharmacological reviews 2000; 52: 595-638.
10 Janig W. Sympathetic nervous system and inflammation: a conceptual view. Autonomic neuroscience : basic & clinical 2014; 182: 4-14.
11 Goehler LE, Gaykema RP, Hansen MK, Anderson K, Maier SF, Watkins LR. Vagal immune-to-brain communication: a visceral chemosensory pathway. Autonomic neuroscience : basic & clinical 2000; 85: 49-59.
12 Quan N. In-depth conversation: spectrum and kinetics of neuroimmune afferent pathways. Brain Behav Immun 2014; 40: 1-8.
13 Weihe E, Nohr D, Michel S, Muller S, Zentel HJ, Fink T, Krekel J. Molecular anatomy of the neuro-immune connection. The International journal of neuroscience 1991; 59: 1-23.
14 Wulfing C, Gunther HS. Dendritic cells and macrophages neurally hard-wired in the lymph node. Sci Rep 2015; 5: 16866.