3. Discussion
PNX is a newly identified peptide that, in the rat brain, is highly expressed in the hypothalamus, medial division of the central amygdaloid nucleus, the spinal trigeminal tract of the medulla and the spinocerebellar tract [
11]. Moreover, enzyme immunoassays also detected a high level of PNX (>4.5 ng/g of tissue) in the spinal cords of rats [
5]. Immunohistochemical studies in rat also revealed the presence of PNX in trigeminal ganglia, as well as in ganglia comprising neurons projecting to the superficial layers of the spinal cord, such as the DRGs and nodose ganglia [
5]. These projections may belong to Aδ-terminals—thin, myelinated fibers capable of responding to both thermal and mechanical stimuli—or to C-fibers, which are unmyelinated, slow-conducting axons that can, due to their high activation threshold, detect painful stimuli. Both of these types of NFs are assigned to nociceptors as they respond to destructive, thermal, mechanical or chemical stimuli [
12]. Furthermore, PNX
+ afferent NFs targeting the spinal cord were revealed in mice and pigs and were present in all spinal cord segments at equal levels [
6,
13], whereas PNX was observed mainly in NFs located in the superficial DH layer, in laminae I and II [
5,
6]. It is also worth noting that PNX-IR NFs participate in skin innervation in rodents: subcutaneous injections of the fluorescent retrograde tracer Fluorogold labeled a population of DRG cells, some of which also contained PNX, thus raising the possibility that PNX released from these terminals could affect other sensory NF/sensory signaling pattern from the skin to the spinal cord [
13].
In addition, the presence of “true” PNX was confirmed by the use of mass spectrometry (not only by showing that the major peak corresponds to the peptide, but also by verifying its presence using the fragmentation spectra of its individual amino acids) in the extracts of DRG cells (present study) as well as in the spinal cord (both rat [
5] and mouse [
13]).
The available literature implicates PNX in pain transmission, and it has been found that the intrathecal injection of amidated PNX suppresses visceral pain; however, it did not affect thermal pain sensation [
5]. Moreover, PNX’s contribution to sensory modulation is also supported by its participation in inducing the itching effect, as revealed in a study by Cowan and colleagues [
13].
To date, the presence of PNX in DRG neurons has only been studied in rodents [
5,
13]. Thus, for the first time, our study demonstrated the presence of PNX in afferent neurons of porcine DRGs. We have found that PNX
+ neurons accounted for a relatively large subpopulation (approximately 20% of all cells) in all DRGs of each segment of the spinal cord. Furthermore, considering the differences in the size (diameter) of the PNX-containing cells, it was demonstrated that the vast majority of PNX-IR neurons were Sl-sized neurons (average diameter up to 40 μm). M-sized cells (40–70 μm) were distinctly less numerous, while Lg PNX
+ cells (diameter > 71 μm) were only occasionally observed. This finding appears to be inconsistent with the previous results described in rodents, where many of the PNX
+ neurons belonged to the M-sized DRG perikarya [
5]. As there is a consensus that the size of the sensory cell is associated with different “physiological functional classes”, the data presented above may suggest a different involvement of PNX-positive cells in sensory modalities in rodents and the domestic pig. However, this hypothesis requires further verification.
The present study, for the first time, also focused on the immunohistochemical characteristics of PNX
+ DRG neurons. We examined PNX colocalization patterns with several neurotransmitters/their markers that were previously revealed in DRG neurons, with particular attention to those biologically active molecules that were previously described in Sl- and/or M-diameter afferent cells: CGRP, SP, nNOS, GAL, CRT, PACAP, CART and SOM [
14,
15].
Previous studies revealed that, CGRP, together with SP (see below), appears to be the “canonical” transmission molecule of afferent neurons in all mammalian species studied so far (human, horse, dromedary camel, pig, cat, rat and mouse) [
16,
17,
18,
19,
20]. According to the available data, the proportion of sensory neurons containing CGRP (or expressing mRNA encoding this peptide [
21,
22]) ranging, on average, between 50 and 60% of all DRG cells, regardless of the species studied. Furthermore, their axonal processes constitute the largest population of nerve fibers forming not only individual laminae of the dorsal horns of the spinal cord (especially laminae I-III, V and X), but also Lissauer’s tract [
17]. Interestingly, this pattern of intraspinal distribution of CGRP-positive central projections of DRG cells persists in all spinal cord neuromeres, which is most prominent in the thoracic ones [
16].
In addition to being involved in the conduction of sensory and pain stimuli in a wide variety of mammals, including human [
17,
18,
23,
24,
25], CGRP is also known to be a very potent vasodilator [
26] and a deficiency in CGRP release is related to a lack of a vasodilation reflex [
27]; thus, blockade of CGRP secretion exerts an analgesic effect in people suffering from migraine [
28].
However, CGRP, along with SP, released from the terminals of DRG neurons in response to noxious mechanical, chemical or thermal stimuli [
29] may also contribute to the development and maintenance of neurogenic inflammation: CGRP acts as an extremely potent vasodilator [
30] while SP, acting on neurokinin 1 (NK1) receptors, evokes increased vascular permeability [
31]. Therefore, it can be concluded that the intensified secretion of these peptides may contribute to the formation of edema, increased blood flow and the influx of inflammatory cells at the site of inflammation.
The above-mentioned close interdependence and cooperation of CGRP and SP in numerous regulatory loops seem to result from the fact that both substances coexist very often in the same sensor cell [
14,
17,
20] from which they are released together and synergistically interact with target cells, supporting each other. For example, as revealed by Biella and co-workers [
32], the excitatory effects of SP arising from spontaneous and noxious activity were significantly enhanced by CGRP in rats. Although the mechanism by which CGRP may potentiate the effects of SP is not fully clear, there is evidence that CGRP may delay the enzymatic degradation of SP [
33,
34]. CGRP has also been reported to increase the release of SP [
35] as well as the excitatory amino acid transmitters glutamate and aspartate [
36] from central terminals of DRG neurons, possibly leading to the strengthening of synaptic connections in the spinal dorsal horn, as well as increasing the effectiveness of these substances on the peripheral targets (for details, see [
37]).
The second largest population of PNX
+ DRG cells observed in this study were SP-containing neurons, accounting for approximately 89% of all PNX
+ afferent cells. This observation seems to be in stark contrast to the data obtained in studies of the co-occurrence of PNX and SP in mouse DRG cells, where both neurotransmitters were present in separate cell subpopulations [
5]. Therefore, it seems reasonable to hypothesize that PNX is probably involved in regulating different functions of DRG sensory cells in both species.
As shortly mentioned above, SP, like CGRP, is primarily considered a transmitter of sensory and pain stimuli, the presence of which has been shown in DRG cells in every species studied so far, including humans [
38,
39,
40,
41]. So far, studies on the presence and distribution patterns of SP
+ cells in the DRGs of the domestic pig have shown the presence of cells containing this tachykinin, especially in subpopulations of Sl- and M-sized neurons. Moreover, there is irrefutable evidence that SP occurs in both the somatic as well as visceral subclasses of these perikarya [
14,
42]. This is consistent with the data obtained in other species studied so far, including humans [
21], guinea pigs [
39], mice [
5] and rats [
43,
44], with regard to both the number of chemically encoded afferent cells and their classification into individual size classes.
While the co-existence of PNX with CGRP in the studied porcine DRG cells was shown in a very high percentage of cells (approximately 96% of all PNX+ cells), only slightly more than half of the CGRP+ cells (56%) also contained PNX. Moreover, a very similar picture emerged from the studies on the co-occurrence pattern of PNX and SP performed in this study: almost 90% (exactly 88.5%) of all PNX+ cells contained SP, and approximately 50% of all SP+ cells were also PNX-positive. The co-localization of both peptides was observed mainly in Sl- and M-sized ganglion cells.
Thus, based on the comparison of the relative numbers of DRG cells containing the individual transmitters, there is no mathematical possibility that all three peptides do not co-occur in at least approximately 80–85% of all PNX
+ cells. This suggestion is further, although indirectly, supported by the observation of the high degree of co-occurrence of PNX with SP and/or CGRP in the axonal fibers of DRG cells forming the dorsal horn laminae of the porcine spinal cord [
6].
Moreover, considering the degree of co-localization of PNX and nNOS in DRG cells (approximately 50% of all PNX+ cells; see below), it seems that neurons containing all four transmitting substances must also be present in at least 30% of all the porcine DRG cells. Unfortunately, at present, insufficient knowledge of the possible functions of PNX in peripheral sensory pathways prevents us from suggesting its physiological significance, either in the case of hypothetical PNX+/CGRP+/SP+ cells or in the population of cells additionally co-expressing nNOS.
nNOS is a marker of the nitrergic subpopulation of afferent cells that are involved in the regulation of numerous physiological processes in the peripheral and central nervous systems. Nitric oxide (NO), produced by nNOS, is an important neurotransmitter in the periphery, mainly in the control of, among others, neuronal plasticity [
45], neurogenesis [
46], neuroprotection [
47] as well as modulation of nociceptive transmission in neuropathic pain [
48] and it is believed that inhibition of nNOS may be a valuable strategy in the treatment of not only neuropathic pain [
49] but also several other illnesses [
50].
In the present study, PNX
+/nNOS-containing neurons formed the third largest (approximately 50%) population among all afferent cells capable of synthesizing and releasing PNX as their transmitter; the vast majority of these neurons were Sl- and M-sized cells. This is well in line with the data obtained in other species (humans, rats, mice, sheep, dromedary camels [
20,
51,
52,
53]) where nitrergic afferent neurons were found in an average of 40–50% of all DRG cells, with the highest percentage in the subpopulation of Sl- and M-sized neurons. It should be emphasized that the high degree of nNOS co-localization with SP and CGRP, reported by Russo and colleagues [
53], additionally supports the above hypothesis, suggesting the existence of a PNX
+/SP
+/CGRP
+/NOS
+ population of DRG cells.
The co-existence of PNX and GAL was found exclusively in porcine DRGs in approximately one-fifth (20.7%) of Sl-sized neurons. As neither the immunoreactivity to GAL alone nor the co-occurrence of PNX and GAL immunoreactivities was observed in M- or Lg-sized cells, this strongly suggests the involvement of cells containing these two peptides in the regulation of nociceptive activities in the domestic pig. This suggestion corresponds well with the data from studies in other species, especially rats: GAL was also found in Sl-sized DRG cells in rodents (in a few cells [
54]) and humans (approximately 12% [
22]); moreover, it has also been observed that their number increases dramatically after peripheral nerve damage [
55,
56]. Numerous studies indicate that GAL may reduce the perception and transmission of pain by increasing the pain threshold as a result of increasing K
+ conductance or decreasing Ca
2+ conductance in the brain, spinal cord and peripheral neurons [
57,
58]. Moreover, in behavioral studies, it was found that reducing the effect of GAL leads to an increase in neuropathic pain-like behaviors and, on the contrary, the upregulation of GAL leads to a decrease in neuropathic pain in animal models [
56].
In contrast to its co-occurrence pattern with CGRP, SP or GAL, PNX co-localized with the other studied substances (neuropeptides: PACAP, CART and SOM; calcium-binding protein: CRT) in only a relatively few porcine DRG cells.
In the case of the co-occurrence of PNX and PACAP, a significantly lower degree of colocalization was observed in this study (about 7% of all PNX
+ cells simultaneously contained PACAP); at the same time, however, it should be emphasized that the vast majority of neurons chemically coded by the presence of both peptides belonged to the population of small, probably nociceptive, neurons. These observations agree fairly well with the data obtained in rodents, where PACAP
+ cells were part of a Sl-diameter neuronal population [
59,
60]. It should be noted, however, that the number of PACAP
+ cells in rat DRGs was slightly higher, reaching approximately 10–17.5% of all sensory nerve cells [
61], suggesting interspecies differences.
It is worth noting that both PACAP and GAL are dramatically up-regulated in damaged DRG cells (up to 75% in the rat [
60]), especially in Sl-sized cells (probably nociceptive perikarya), and act as neuroprotective and/or pro-regenerative peptides in this scenario [
62,
63,
64]. Unfortunately, so far, there are no data on the possible functions of PNX in these regulatory mechanisms, and therefore the question of the possible existence of a subpopulation of PNX
+/GAL
+/PACAP
+ cells, as well as their presumed targets and possible regulatory functions remains open and requires further, more detailed research.
In contrast to the patterns of PNX co-localization with the other biologically active agents discussed above, where the vast majority of observed PNX
+ cells belonged almost exclusively to Sl- or M-diameter cells, PNX and CART co-localization was observed in the pig in all three neuronal size classes, with the M-sized PNX
+/CART
+ cells being the most numerous. While the data on the presence of CART in porcine DRG cells collected in this study appear to be in full agreement with the results obtained by Zacharko-Siembida and colleagues [
65] in the lumbar porcine DRGs (in both studies, the relative number of CART+ cells oscillated around 5% of all DRG cells), a slightly more numerous (approximately 10%) population was observed in rodents [
66]. Remarkably, Zacharko-Siembida and colleagues found that the vast majority (approximately 70%) of porcine CART
+ DRG cells simultaneously contain CGRP. This allows for the assumption that some of the sensory cells may also contain, in addition to PNX
+/CGRP
+/SP
+/NOS
+, CART, which additionally suggests the need for in-depth research on both the chemistry and the functional significance of these sensory cells.
The last of the neuropeptides whose pattern of co-occurrence with PNX was the subject of interest in this work was SOM, a substance present in both the pig and the rat [
67] in a small (up to 10%) subpopulation of the DRG cells. It has been shown that a small fraction (about 6% of all PNX
+ cells) also contains SOM, which, in the light of the available literature data (both PNX and SOM are implicated as transmitters subserving the itch conduction pathway [
13,
68,
69]) the most recent data, see [
70], suggests that these cells are representatives of pruriceptors. However, further studies are needed to unveil the exact relevance of PNX in this neural modality.
Intracellular calcium-binding proteins may play an important role in the nociceptive excitability of neurons [
71,
72], as well as in proprioception [
73], as evidenced by the presence of calbindin (CB) in excitatory interneurons within the superficial DH [
74]. Taking this into account, we decided to analyze the co-occurrence of CRT (protein closely related to the above-mentioned CB) with PNX in porcine primary afferent cells. It was found that PNX coexisted with CRT in a relatively small population of DRG neurons (approximately 10.5% of all PNX-containing cells), due to the fact that CRT was expressed mainly in Lg-sized neurons, and the vast majority of PNX
+ neurons were small-sized cells. These data are in line with reports emerging from studies concerning the CB or CRT expression patterns in rats and fowl [
15,
73,
75], which showed that the vast majority of CB- or CRT-positive neurons were M to Lg in size [
73].
Limitations of the Study
The most important limitation of the present study was the inability to perform simultaneous colocalization studies of a larger number of neurotransmitters in the same nerve cell, due to the unavailability of appropriate primary and/or secondary antibodies. Similarly, the authors are aware that the value of this work would be higher if it were possible to demonstrate expression patterns of membrane/cytosolic receptor subtypes specific to the studied neurotransmitters; unfortunately, as in the case of studies based on multiple colocalizations in a single afferent cell, the unavailability of antibodies recognizing the mentioned receptors was an obstacle. The last limitation was the inability to link the observed PNX+ cell subtypes with their target tissues within the assumptions of this study: an analysis of the presence of PNX+ cells in the DRGs of so many spinal cord neuromeres would require the injection of a retrograde neuronal tracer into individual internal organs in such a large number that the planned experiment was simply becoming impossible. On the other hand, the use of an anterograde tracing technique was excluded due to the need to inject the tracer directly into the studied DRG, which would certainly cause changes in the expression pattern(s) of the tested substances.