Exosomal cargo including microRNA regulates sensory neuron to macrophage communication after nerve trauma.

Editor's note: This comment is co-authored by Rafael González-Cano and Michael Costigan, Boston Children’s Hospital and Harvard Medical School, Boston, US.

Why does the dorsal root ganglia (DRG) exist? Why not just locate the neuronal cell bodies throughout the nerve itself? After all, this is how neurons are spaced in the CNS along the signaling pathways they serve. Most would answer this question, "Simple―it’s for protection," since the DRG sits in its own bone capsule close to the spinal cord. Undoubtedly that is one, and likely the most important, explanation for the DRG existence. But are there other reasons for the cell bodies to sit so close to each other? Communication has often been suggested as another reason, and possibly the cell bodies achieve this via electrical signaling (Kim et al., 2016). The plethora of cytokines and chemokines produced following nerve injury would also suggest that speedy and effective information transfer is necessary, especially in times of crisis (Austin and Moalem-Taylor, 2010).

One newly recognized way that cells communicate is through miRNAs and in this very well-performed study by Simeoli et al., Marzia Malcangio and colleagues have examined this system in sensory neurons. After peripheral nerve injury, miR-21-5p is upregulated in the cell bodies of most, if not all, damaged neurons, and this miRNA along with others are packaged into exosomal vesicles ready for release. The authors show how this process occurs with some elegant experiments using capsaicin stimulation of cultured DRG neurons. These cells respond to capsaicin treatment by increasing the expression of exosomal protein components, upregulating certain miRNAs, assembling the loaded vesicles and releasing them into the cell culture medium. Once released, the authors were able to collect the free exosomes and assay for the presence of miRNAs. Following capsaicin treatment, miR-21-5p and the other miRNAs were increased at least twofold in this exosomal fraction. To control for the specific action of capsaicin on sensory neurons, these experiments were repeated in TRPV1 KO DRG cultures, and in this case the vesicle components and miRNAs were not upregulated, conclusively proving this miRNA signaling occurs through sensory signaling.

Next, the authors proved the released endosomes are phagocytosed by macrophages. As their name suggests, these dendritic immune cells are all too happy to devour multiple types of cellular and subcellular debris, these morsels classically including viruses, bacteria, or other foreign entities. To prove that endosomes were on the menu, the vesicles were labeled with fluorescent dye and cultured with macrophages, and the subsequent presence of dye within the macrophages proved that these cells did indeed engulf them. Next, the authors fed the macrophages endosomes derived from either capsaicin-treated or control neurons and demonstrated that the capsaicin-treated vesicles resulted in regulation of macrophage effector genes including those that change these macrophages toward an inflammatory M1-like phenotype relative to the more regenerative M2-like cells.

To formally show that miR-21-5p was capable of regulating these macrophage effector genes, cultured macrophages were transfected with synthetic miR-21-5p mimics and compared with scrambled controls. These transfections regulated the macrophage genes, again sending their phenotype toward the inflammatory M1 state, a result demonstrated by qPCR and confirmed with FACS analysis.

The authors then moved in vivo, returning to nerve injury to assay the effects of miR-21-5p on neuropathic mechanical allodynia. Both intrathecal delivery of a miR-21-5p antagomir and conditional deletion of miR-21 specifically in sensory neurons reduced neuropathic mechanical hypersensitivity, and reduced the number of macrophages that migrated into the injured DRG. The in vivo results suggest that the role of exosomal miR-21-5p in the damaged DRG is to attract monocytes from the bloodstream and, once within this tissue, convert them into M1-like cells.

Studies prior to this paper have also shown that miR-21-5p is regulated in DRG neurons by nerve injury, but this work concentrated on the effects of this miRNA in the neurons themselves. Strickland et al. demonstrated that miR-21-5p can increase axonal regeneration in DRG neurons via a mechanism connected to Sprouty2 (Sprty2; Strickland et al., 2011). Expressed in early neuronal development (Chambers et al., 2000), the Sprouty proteins act as regulators of receptor tyrosine kinase signaling (Neben et al., 2017), and, interestingly, results within the study by Simeoli et al. also demonstrate that Sprty2 is regulated in macrophages by miR-21-5p, where it may aid polarization of these cells (Atomura et al., 2016). So, in addition to miR21-5p acting in the damaged neurons that produce it, to regulate genes involved in axonal growth, these neurons actively export this signal to attract monocytes and determine their phenotype once present.

Early after nerve injury it makes sense that the recruited macrophages are proinflammatory as they need to clear away any debris resulting from nerve injury and induce painful tactile hypersensitivity to prevent excessive, potentially damaging limb movement while the nerve starts to regrow. Later, in mice subject to sciatic crush, where the nerve can regenerate efficiently, it would be interesting to see if the miRNA produced by the DRG cells starts to convert the macrophages into an M2-like phenotype. Alternately, in the case of a ligated nerve injury such as in the SNI model, how long the neurons continue to overexpress miR21-5p and other proinflammatory signals would be fascinating to define. Such signals likely maintain non-resolving inflammation within the chronically damaged peripheral nervous system and in doing so support continued neuropathic pain symptoms.

This study is exceptional in the fact that it offers us well-worked-up molecular proof of the communication between differing cell types in the DRG via miRNA signaling. It can be likened to different support services within a town close to a highway communicating with one another following a multiple car pileup. Local communications are necessary in order to coordinate and administer initially the emergency support services, followed by those tailored to clearing the wrecked vehicles, and finally those intended to fix the road surface before returning it to use.