The sympathetic nervous response in inflammation.

So far, we introduced a new model of neuroimmune regulation specified in arthritis. All these elaborate mechanistic and structural adaptations during inflammation need to serve some purpose, however, otherwise they would not have been positively selected during evolution. In recent hypothetical modeling, a framework was developed that tries to explain the underlying meaning. An activated immune system needs a significant amount of energy above that required for the normal non-inflamed state [ 93 ]. The activation of the SNS and the HPA axis at the beginning of inflammation helps to provide enough energy, because activation of these axes mobilizes energy-rich fuels mainly by increasing lipolysis, glycogenolysis, muscle protein breakdown, and gluconeogenesis (Figure  4 ). At the beginning of an inflammatory innate immune response, the SNS but also HPA axis support inflammation by non-specific means; for example, mobilization of leukocytes [ 22 , 26 ], increasing blood pressure and heart rate, increasing lymph flow [ 21 ], plasma extravasation [ 94 ], antigen uptake and presentation [ 37 ] (Figure  4 ). In this initial phase of inflammation, SNS activity also ‘programs’ adaptive immune cells via β2-AR - for example, B cells to produce increased amounts of antibodies and T cells to produce more or less cytokines dependent on the context of activation [ 95 ]. This mainly proinflammatory action takes place on a systemic level in secondary lymphoid organs like the spleen and lymph nodes, where immune cells are programmed and then released to attack the intruder. At the local site of inflammation, however, SNS activity contributes primarily to anti-inflammatory mechanisms, mainly by direct influence of neurotransmitters on immune cells [ 4 ]. Besides the local promotion of regulatory B cells (see above), also macrophages stimulated via the β2-ARs acquire an anti-inflammatory M2 phenotype [ 96 ] and β2-AR stimulation also inhibits TNF production [ 97 ] (Figure  2 ). On the other hand, stimuli via α-ARs are proinflammatory: for example, α2-AR stimulation increases reactive oxygen species in macrophages [ 98 ]. Therefore, the net outcome of stimulating ARs on immune cells strongly depends on the receptor engaged and, therefore, on the receptor expression pattern (which might change during the course of inflammation [ 4 , 45 ]) and neurotransmitter concentration, because NE binds preferentially to α-ARs, only binding to β-ARs at high concentrations (for example, [ 99 ]). However, why do some immune cells, like macrophages, express both α-ARs and β-ARs, which will counteract each other in terms of immunoregulation? One possible explanation is that, due to the different binding affinities of NE to these AR subtypes, this system can be used as a distance detector to the source of catecholamines. In this respect, repulsion of sympathetic nerve fibers from inflamed tissue makes sense, since it is not favorable to inhibit the immune response (high concentrations of catecholamines preferentially stimulate anti-inflammatory β-ARs) before the antigen is cleared (Figure  4 ). Therefore, this distance detector system (simultaneous expression of α-AR and β-AR on immune cells) provides a means for the body to define sites of permitted inflammation (low SNS fiber density, low catecholamine concentration) and, on the other hand, prevent uncontrolled spreading of inflammation by preventing bystander activation (high SNS fiber density, high catecholamine concentration) (Figure  2 ). To get an impression of the contribution of SNS to local anti-inflammatory mechanisms, the eye is a good example. The eye is known as an exceptional immune-privileged site, dominated by anti-inflammatory mechanisms. It has been shown that sympathetic denervation of the eye leads to a decrease in anti-inflammatory molecules, like tumor growth factor-β, which results in a complete loss of the immune-privileged status [ 100 ]. Therefore, repulsion of SNS fibers from inflamed tissue is an effective means to increase local inflammation (Figures  2 and 4 ). This has been positively selected during evolution to clear invading microbes but not to serve chronic autoimmune inflammation. We hypothesize that catecholamine-producing cells start to play a role in the later inflammatory phase, possibly as a compensatory mechanism for the local loss of SNS fibers. These TH-positive cells can be anti-inflammatory as described above. One might argue that it is easier to just shut down SNS activity at the systemic level than to repel nerve fibers from local inflamed tissue, but SNS activity stays high during many chronic inflammatory conditions (for example, [ 101 ]). Concerning the energetic aspect discussed above, this is beneficial in terms of providing enough energy to feed the activated immune system on a systemic level. In contrast to the SNS activity, which is still high in chronic inflammation, HPA axis activity is relatively reduced, not down to normal, but to a level without immunosuppression, to not disturb the local immune response (Figures  3 and 4 ). Overall, the system takes on an ‘inflammation configuration’, including repulsion of sympathetic nerve fibers from local inflamed tissue to create an area of permitted inflammation, high SNS activity on a systemic level, and reduced HPA activity without local immunosuppression, but provision of energy-rich fuels is still maintained and important (Figure  4 ). These processes are positively selected during evolution to serve short-term acute inflammation [ 93 , 102 ]. If these processes persist for too long, they cause harm because the body is in a constant state of catabolism and volume overload. Known disease sequelae in chronic inflammatory conditions can be explained by this constant activation of the SNS and HPA axis and the resultant catabolic state, like cachexia, high blood pressure, insulin resistance, and so on [ 93 , 102 ].

This review is based on a systematic search of the PubMed database using the search terms ‘sympathetic nervous system’, ‘peripheral nervous system’, ‘nerve fiber’, ‘neuroimmun*’, ‘norepinephrine’, ‘arthritis’, ‘collagen induced arthritis’, ‘rheumatoid arthritis’, ‘autoimmune diseases’, ‘autoimmunity’. Articles (including abstracts) published in English or German up to March 2014 were considered. All retrieved articles were screened for eligibility based on title, abstract, and full content.

From the current conceptual and experimental knowledge, certain hypotheses can be derived about potential clinical and therapeutic approaches that might improve clinical practice. Clinical data applying the current knowledge specifically on sympathetic regulation of inflammation is scarce. However, one promising approach that underscores the importance of sympathetic downstream signaling in anti-inflammation is the inhibition of phosphodiesterase (PDE)4, an enzyme that degrades cAMP. Increasing cAMP by inhibiting this enzyme shows promising results in psoriatic arthritis, which led to the approval of the PDE inhibitor apremilast for this disease entity [ 103 ]. PDE inhibitors are also currently being tested for several other clinical entities; for example, psoriasis, rheumatoid arthritis, and Behcet’s syndrome [ 103 ]. Taking into consideration that a general increase in cAMP might also support detrimental effects as discussed above, it is noteworthy that PDE4 is the predominant PDE isoform expressed in immune cells [ 104 ]. However, whether increasing cAMP by pharmacologic PDE inhibition will support disease sequelae is not clear at the moment and further research is needed. Right now, neuroimmunology in the sense presented in this review is on the verge of clinical translation. In terms of sympathetic control of inflammatory arthritis a possible approach is to follow the success seen in animal models and put effort into developing novel cellular therapies; for example, after induction of TH in certain immune cells or treatment of B cells with sympathetic stimuli to increase their regulatory potential. On the other hand, the systemic permanent overactivation of the SNS as discussed above could also be a potential target for intervention; for example, by psychological or pharmacological means. However, clinical data are missing at the moment and further research is warranted. For this research an approach to support local activation of sympathetic mechanisms, like increasing cAMP in immune cells (for example, PDE4 inhibition) but on the other hand decreasing systemic SNS activation to prevent disease sequelae, needs to be the focus.

Inflammation causes increased activity of the SNS with release of NE and co-transmitters in lymphoid organs and inflamed local sites. Immune cells carry receptors (for example, ARs) to detect and process signals from the SNS. The reaction of the immune cell to neurotransmitters is variable depending on the context of receptor engagement (activation state of the cell, expression pattern of neurotransmitter receptors, microenvironment, cytokine milieu, and distance from the catecholamine source (concentration)). On a systemic level, the signals from the SNS are proinflammatory in the initial phase of inflammation, whereas anti-inflammatory effects are dominant in the late or chronic phases of an inflammatory response, at least in collagen-induced arthritis. Upon initiating an inflammatory process, the body adopts an ‘inflammatory configuration’ with increased systemic SNS and HPA axis activity. This reaction can be interpreted as an ‘energy appeal reaction’ resulting in the provision of enough energy-rich fuels, like glucose and free fatty acids, to fulfill the needs of an activated immune system. If inflammation becomes chronic, as in chronic inflammatory illness, the system changes into a ‘chronic inflammatory condition’ that is characterized by 1) still increased systemic activity of the SNS, 2) still increased activity of the HPA axis but without immunosuppression (glucocorticoid receptor desensitization and inadequacy), and 3) local repulsion of SNS fibers from inflamed tissue, including lymphoid organs, to create zones of permitted inflammation. The immune response is more or less uncoupled from central regulation to avoid the anti-inflammatory influence of the brain. All mechanisms ensure an optimal fight against an antigen. These adaptations are evolutionarily positively selected to clear the antigen, usually an intruding microbe. However, if a ‘chronic inflammatory configuration’ persists, as in autoimmunity, the effects are detrimental because of the persistently increased SNS activity, HPA activity, and the resultant chronic catabolic state. This leads to known comorbidities in chronic inflammatory disease, like cachexia, high blood pressure, insulin resistance, and increased cardiovascular mortality. The challenge is now to translate this conceptual knowledge into clinical benefit.

The sympathetic nervous system (SNS) is an integrative system that reacts to dangerous situations, and activation of the SNS is part of the classical ‘fight and flight’ response. This is common knowledge. However, the SNS is not active just in these extreme situations, but is part of constant regulatory machinery that keeps body functions in a steady-state equilibrium. Of course, the SNS is not alone in performing these tasks but is interwoven into complex regulatory circuits. Therefore, it is not possible to analyze the action of the SNS in inflammation without considering the other important players, like the hypothalamic-pituitary-adrenal (HPA) axis, and the sensory nervous system and vagal nervous system (VNS). For a detailed description of the functional anatomy of the autonomous (SNS and VNS) and sensory nervous system, as well as the HPA axis, we refer the reader to respective standard textbooks of physiology since this is established and common knowledge and a detailed description would go beyond the scope of this review. In the first part of this review we focus on important highlights concerning the SNS and inflammation. In the second part, the standalone facts will be integrated to try to understand the deeper meaning of this regulatory machinery in inflammatory disease. As an example, we refer to findings concerning neuroendocrine immune regulation in arthritis.