Our behavioral evaluations showed that BTZ was able to induce in mice a dose-related neuropathy characterized by the presence of allodynia and hyperalgesia. Preclinical data have demonstrated that antineoplastic drugs can activate both innate and adaptive immune responses as well as peripheral and central neuronal accessory cells like satellite cells, Schwann cells, astrocytes, and microglia [
4,
28]. In particular, chemotherapeutics can cross the blood-nerve barrier, accumulating in dorsal root ganglia and in the peripheral nerves and exerting a toxic action with consequent immune cell infiltration and activation [
8]. Consistently, our biochemical and immunofluorescence results indicate that after 14 days of BTZ, in the presence of an hyperalgesic and allodynic state, PNS stations are characterized by increased levels of macrophage activation markers, i.e., CD68 and TLR4, and by the presence of a pro-inflammatory cytokine profile due to high levels of the investigated pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) and low levels of the anti-inflammatory cytokine IL-10 which is particularly evident in DRG. Moreover, our electron microscopy evaluations demonstrated that DRG of BTZ mice are characterized by ultrastructural abnormalities like the presence of some partially detached satellite cell sheaths and some swollen mitochondria in neurons and satellite cells. At this time point, we only observed an increase of GFAP in the spinal cord without measuring any other biochemical alteration. Our results showing an increase of CD68 marker and of TLR4 expression in the peripheral nervous tissues are consistent with the recent literature that suggests the importance of macrophage infiltration and activation in PNS in CIPN development [
8]. For example, it has been recently demonstrated that the intravenous immunoglobulin administration was able to reduce or prevent BTZ-induced heat and mechanical allodynia in rat by decreasing or preventing M1 macrophage infiltration [
10] in the peripheral nerves. In our study, by increasing the BTZ cumulative dose (day 28), we observed a further lowering of the response thresholds to mechanical and thermal stimuli in BTZ-treated animals. The increased hypersensitivity correlates with a more severe structural damage in DRG and with the appearance of a more pronounced neuroinflammatory condition also evident at the spinal cord level. Interestingly, after 28 days of BTZ treatment, we also observe an overexpression of PK system in all the tissues involved in pain transmission (sciatic nerve, DRG, spinal cord). Immunofluorescence data suggest that in PNS at the higher BTZ dose, CD68 + cells co-express PK2. Therefore, we can suppose that infiltrating activated macrophages may represent an important source of PK2 in DRG and sciatic nerve even if it appears evident that other cell types like satellite cells and neurons may contribute to PK2 increase. In our paradigm, the activation of the PK system in BIPN is delayed in comparison with painful symptoms and the precocious neuroinflammation. This later PK activation was somehow surprising, since in previous work from ours [
13,
14] and other groups, PK system [
20] activation well correlated with the development of hypersensitivity. Here, we demonstrated that in BIPN, this chemokine family has however an important role in sustaining, maintaining, and worsening hypersensitivity, neuroinflammation, and structural damage of DRG. In fact, chronic treatment with PC1, even if it was started in the presence of an established hypersensitivity, was able to counteract the further decrease of mechanical and thermal thresholds, to preserve against the neurotoxic damage to DRG and to reverse the established neuroinflammation, rebalancing pro- and anti-inflammatory cytokines in sciatic nerve and DRG. We can assume that during BTZ treatment, infiltrated and resident-activated immune cells, in association with satellite cells and Schwann cells, produce pro-inflammatory cytokines leading to a further recruitment of immune cells into damaged nervous tissues. These infiltrating macrophages not only produce PK2 but also express PK-Rs receptors [
26]; hence, PK2 can act in autocrine or paracrine way sustaining a neuroinflammatory loop that exacerbate the neuronal damage and sustain a progressive glial activation at the spinal cord level [
20]. A possible signaling pathway could be the one suggested by the group of Qu et al. [
28]. Authors demonstrated that STAT3 signaling plays a crucial role in PK2 regulation and that phosphorylated STAT3 can directly bind to the Pk2 promoter. Furthermore, a recent study [
29] demonstrated that phosphorylated STAT3 levels were significantly increased after BTZ administration and that STAT3 activation in DRG contributes to BIPN. On the basis of these data, we can speculate that the activation of STAT3, consequent to pro-inflammatory cytokine increase in the peripheral nervous stations [
30], could be one of the mechanisms involved in the PK2 upregulation. The effect of PC1 may be in part related to its ability to reduce macrophage activation and infiltration in PNS and to prevent PK system upregulation that plays a crucial role in prompting spinal cord neuroinflammation. In addition, as also supported by the acute antiallodinic effect of PC1, PKs can also act on PK-Rs expressed by neurons and glial cells enhancing pain pathway transmission [
17] which also occurs through TRPV1 sensitization [
31,
32]. Our results also confirm the importance of astrocytes in CIPN [
33]. In fact, in our experiments, GFAP is the only marker that we find precociously activated in the spinal cord. It was recently suggested that the presence of mechanical hypersensitivity due to BTZ treatment correlated to an upregulation of GFAP [
34,
35], and more recently, Salvemini’s group described that the development of BIPN is lost when S1PR1 (sphingosine-1-phosphate receptor 1) is deleted in astrocytes, suggesting a central role of astrocytes in sustaining CNS sensitization [
33]. Interestingly, by blocking the activation of the PK system with PC1, we prevent a further neuroinflammatory condition in the spinal cord. In BTZ + PC1-treated animals, in fact, we did not detect any increase of glial activation markers CD68 and TLR4 that are indeed significantly enhanced in BIPN animals after 28 days of BTZ treatment. As already observed in other experimental models, IL-1 and IL-10 appear to be the main cytokines modulated in the spinal cord in the presence of a neuropathic state and the treatment with PC1 is able to prevent the IL-1/IL-10 unbalance. We can therefore hypothesize that in the spinal cord, there is an early activation of astrocytes that is independent from the PK system. Astrocytes start to produce PK2, as demonstrated by the colocalization of PK2 and GFAP signals in the immunofluorescence experiments and confirming what we already observed in the CCI model [
14]. PK2 promotes microglia activation and cytokine alterations that may participate in central sensitization; antagonizing the PK system prevents this later activation. The precocious astrocyte activation is not completely reverted by the PK antagonism and may be responsible for the only partial anti- hyperalgesic effect observed in PC1 mice. We must however underline that in PC1-treated mice, we did not observe a BTZ dose-related increase of GFAP signal, suggesting that blocking PK2 may be useful to control astrogliosis. Moreover, our data support the well-known flow of neuroimmune activation from the periphery to the central nervous system [
36‐
38] at the basis of the development of pathological pain and underline the role of the prokineticin system in this sensitization process.
Finally, the data reported in this study could have translational implications. First of all, considering that BIPN develops in about 1/3 of BTZ-treated patients, PC1 may be administered only when the symptoms have appeared allowing patients continue the chemotherapeutic treatment. In addition, we show that PC1 has a protective role in a two-cycle BTZ schedule: in fact, in the second cycle, the allodynic effect promoted by BTZ is less evident in mice previously treated with PC1 if compared to that observed in BTZ-only re-treated mice. Furthermore, the second PC1 treatment completely normalizes the mechanical thresholds. Considering that patients often undergo multiple chemotherapeutic cycles, the protective role exerted by PC1 on a second chemotherapeutic cycle could be clinically relevant in order to slow down the re-appearance of the side effects. We plan to deeply investigate the reason behind this protective role of PC1 in future studies. At the moment, we can only speculate that the protective effect exerted by PC1 could be due to its ability to counteract neuroinflammation and more likely to its protective role on DRG ultrastructure. It can be hypothesized that chronic PC1 treatment may induce long-lasting modification in the PK system or enhance protective mechanisms that may be important in a second BTZ cycle, but further experiments are needed to sustain this possibility.