Nociceptor-like behaviors differ between young and aged female rats after chronic CFA-induced inflammation
Figure 1A shows mechanical paw withdrawal thresholds (PWT) measured throughout a period of 120 days in young and aged female rats after a single unilateral injection of CFA. As expected, CFA triggered mechanical allodynia which peaked between 7 and 14 days after treatment (CFA7-14) at both ages. From that point onwards, there was a partial recovery in the young rats while the older rats remained allodynic (PWT ~ 6g) until CFA21. Strikingly, while young animals reached a full recovery from mechanical allodynia at CFA77, the aged group never fully recovered during our observation period of 120 days. Another difference between the groups is the lack of a contralateral effect in the young and a slight but significant mechanical hypersensitivity in the old that reverts within a week. Finally, the baseline mechanical thresholds for the young and old were 16 and 26 g respectively, and in both cases, the contralateral PWT went up as the rats aged, in agreement with our previous findings [59].
The absolute differences in baseline thresholds required data normalization where 100% represents the initial values. Given that the most prominent changes in PWT happened before CFA49, we decided to shorten the x-axis to reflect this time frame. Thus, the time courses in Fig. 1B depict the percentage variations at each observation time point instead of the absolute PWT values and allow us to compare young versus aged rats rather than the effects of lateralization. The plot on the left corresponds to the side contralateral to treatment and shows a small but significantly reduced mechanical sensitivity in the aged rats at CFA4 and CFA7 compared to young rats (F(1,18) = 7.391, p = 0.0141)). The PWT as % of baseline was significantly smaller compared to the pre-CFA values for older animals at both times (p = 0.0106 and 0.0127, indicated by the letter b). This may indicate the presence of an early systemic effect of inflammation in this age group. The mechanical threshold of the 6-month-old rats showed a decrease relative to the initial baseline at CFA4 (p < 0.0001) and then tended to increase and become statistically higher at CFA49 (P = 0.0325, both indicated by the letter a).
The plot on the right shows significant differences between young and aged rats ipsilaterally in terms of the evolution of the response to chronic cutaneous inflammation. The magnitude of the mechanical hypersensitivity is significantly higher in the older group at CFA4, CFA14, CFA21 and CFA28 (F(1,18) = 7.024, p = 0.0163, indicated by asterisks). Compared to the initial baseline values, we observed significant differences in both ages at all times (indicated by the letters a and b) except for the young at CFA49 (indicated by a dash numeral).
The ipsilateral time courses for young and old suggest different kinetics in the recovery from the effects of inflammation on mechanical sensitivity in the two groups. To test this hypothesis, we used a non-linear regression model to ascertain whether a single curve was a good fit for all data sets. We found that this was the case contralaterally (p = 0.0549; F(DFn, DFd) = 2.364 (4, 172)), but ipsilaterally two different curves were required to fit the data sets (p < 0.0001; F(DFn, DFd) = 9.660 (4, 172)) (Suppl. Figure 2). We then decided to determine at what time after CFA the two age groups reached 50% of baseline PWT recovery. This rendered a striking result: 50% recovery was achieved 21 days after CFA for young and after 33 days for aged rats. Thus, it is evident that for mechanical hypersensitivity younger animals exhibited a much faster initial recovery after the induction of inflammation than aged rats.
Next, we evaluated the response to acetone (cold allodynia) and spontaneous pain in both age groups. Figure 1C shows that the baseline cold sensitivity (indicated by horizontal dashed lines in the plot) agrees with what we reported previously for untreated, healthy rats of 6 and 24 months of age [59]. We only detected differences between the two groups at CFA4 (p = 0.0145) and CFA35 (p = 0.0455), and in general the young were more sensitive to cold than the old. Evidence of significant cold allodynia was present at CFA4 in young rats and at CFA1 for the older group, indicated by the letters a (p = 0.0338) and b (p = 0.0164) respectively. This may indicate a slower onset of painful cold sensitivity in the younger rats.
We made several important discoveries regarding spontaneous pain as assessed by spontaneous foot lifting. First, the intensity of this pain was significantly higher in older than younger rats at CFA1 (p = 0.0023). Second, SFL peaked at CFA1 in the old and CFA4 in the young. Finally, this kind of pain resolved itself in both groups between days 21 and 28. This is the first time that such a clear difference in the time course and intensity of spontaneous pain has been described in groups of differing ages.
These unique behavioral differences between young and aged rats may arise from differential changes in the expression of canonical nociceptive-associated molecules. Here we focused on Nav1.8 and ASIC3 for two main reasons. First, we have recently shown that their expression patterns differ between healthy young and old, geriatric rats [59]. Second, these molecules have been previously implicated in inflammatory pain but have not been examined in detail in the context of prolonged chronic inflammation (see Introduction).
Changes in the relative mRNA levels of Nav1.8 and ASIC3 in the DRG of young versus aged rats at CFA1 and CFA4
To ascertain possible expression changes of Nav1.8 and ASIC3 mRNA levels, we ran a qPCR analysis on DRGs at 2 times after CFA injection (Fig. 2). In young rats, we found significantly elevated Nav1.8 mRNA ipsilaterally at CFA4 compared to normal rats of the same age (p = 0.0480, indicated by a) and also at that time compared against the contralateral side (p = 0.0451). In aged animals, we observed significantly lower mRNA levels contralaterally compared with normal (p = 0.0065, indicated by b) and also elevated mRNA ipsilaterally at CFA1 compared to contralateral (p = 0.0050) (Fig. 2A).
For ASIC3 mRNA, young DRG showed higher levels ipsilaterally at CFA1 compared to normal (p = 0.0123, letter a) and also compared to the contralateral side (p = 0.0049). DRG from older rats exhibited elevated levels of ASIC3 mRNA ipsilaterally, compared to normal at CFA1 (p = 0.0309) and CFA4 (p = 0.0480) both indicated by b. ASIC3 levels were also significantly increased between ipsilateral and contralateral sides at both times (CFA1, p = 0.0123 and CFA4, p = 0.0451) (Fig. 2B).
These data showed that Nav1.8 and ASIC3 mRNA levels change at different times after CFA suggesting that these molecules may play different roles in the response of both age groups to cutaneous inflammation. However, this approach did not allow us to evaluate possible changes occurring within neuronal subpopulations in the DRG. Thus, we decided to conduct a full quantitative immunohistochemical analysis of the expression of both proteins over a period that extends to day 42 after CFA. This interval encompasses both the acute and chronic inflammatory phases [67] as well as the behavioral recovery shown in Fig. 1.
Changes in the expression of Nav1.8 after CFA in DRG neurons of young versus aged rats
Figure 3A shows representative images of Nav1.8 staining from 1, to 42 days after a single intradermal injection of CFA (ipsilateral side) versus the expression in the untreated contralateral side. We observe time-dependent bilateral variations in Nav1.8 expression after CFA treatment in some but not all neurons.
The bar scatter plot in Fig. 3B shows the quantitative analysis of the relative levels of staining for each age group over the duration of treatment. At CFA1 we found a significant contralateral decrease in aged rats (A) compared to young animals (Y) (p = 0.0173) and a significant increment in Nav1.8 ipsilaterally only in the aged rats (p = 0.0020). At CFA4 Nav1.8 levels dropped both contra and ipsilaterally in the aged rats (p = 0.0473 and p = 0.0197, respectively) compared to younger rats. At the same time, we observed an increment ipsilaterally in the young rats (p = 0.0197) compared to contra. We did not observe any significant changes between young and aged DRG at CFA7.
Interestingly, at CFA14 the levels of Nav1.8 went up ipsilaterally in old rats versus young (p = 0.0002). From this time onwards the only change we noticed was a significant drop in the aged only contralaterally at CFA28 (p = 0.0036). When compared to normal (untreated rats, N) we observed significantly higher Nav1.8 at CFA4 ipsilaterally in the young animals (p = 0.0498) and a similar pattern at CFA14 in the aged (p = 0.0017).
Because neuronal size and conduction velocity correlate within the DRG[47] we performed a detailed analysis of Nav1.8 expression in neurons classed by size (Suppl. Figure 3). For small, C fiber-like neurons, at CFA1 we observed a significantly lower contralateral expression in aged vs. young rats (p = 0.0042) and for the aged only a significant increment ipsilaterally (p = 0.0086). At CFA4 we found a significantly higher % intensity of Nav1.8 in the young vs. aged rats (p = 0.0240). In agreement with the overall pattern described in Fig. 3B, we saw no significant differences at CFA7. This changed at CFA14: We found significantly higher levels of Nav1.8 in aged rats both contra- and ipsilaterally (p = 0.0478 and 0.01). 14 days later, this trend reverted and the younger rats had higher levels of the channel at both sides, suggesting systemic changes (p = 0.0161 for contra and p = 0.0309 for ipsi). The expression of Nav1.8 seems to become stable and on par for each group at CFA42 (Suppl. Figure 3A).
In contrast to the various ups and downs of % intensities observed in small neurons, medium-sized cells only exhibited significant alterations at CFA1, with higher levels ipsilaterally in the aged rats (p = 0.0009) and at CFA28, with lower levels in the same age group but contralaterally (p = 0.0161). We found no other significant differences (Suppl. Figure 3B).
In the large neuron subpopulation, overall % intensities for Nav1.8 were lowest, as this group represents heavily myelinated A-fibre-like neurons that are mostly non-nociceptive[25] (Suppl. Figure 3C). At CFA1 we described significantly elevated Nav1.8 ipsilaterally in the aged rats (p = 0.0012) and at CFA14, we found the lowest levels of Nav1.8 in the aged group – incidentally this was lower than the ipsilateral expression for the younger rats (p = 0.0132).
Changes in the expression of ASIC3 after CFA in DRG neurons of young versus aged rats
Figure 4A depicts representative images of ASIC3 at 1 to 42 days after a single intradermal injection of CFA (treated side) versus the untreated contralateral side. The staining patterns showed both similarities and differences compared to Nav1.8 staining described in the previous section.
Figure 4B shows the quantitative analysis of the relative % intensities of ASIC3 immunostaining for each age group and the duration of treatment. At CFA1, young DRG neurons showed a significant ipsilateral increment compared to the contralateral side (p = 0.047). Notably, we found no significant differences between ipsilateral and contralateral expression of ASIC3 at any time after CFA1 in the DRG neurons from young rats. In sharp contrast, we observed significant ipsilateral increments in the expression at all times examined (from CFA1 to CFA42) in the neurons from older rats (p = 0.0007 for CFA1, p = 0.0150 for CFA4, p = 0.020 for CFA7, p = 0.0087 for CFA14, p = 0.0014 for CFA28 and p = 0.048 for CFA42). We found several statistically significant differences between ipsilateral levels between both ages at all times except CFA1. In particular, at CFA4 (p = 0.0031), CFA7 (p = 0.0042), CFA14 (p = 0.0002), CFA28 (p = 0.0014) and CFA42 (p = 0.0001). Compared with normal (N) we observed a significantly higher ASIC3 expression at CFA1 ipsilaterally in the young animals (p = 0.0197). In the aged rats, we found significantly elevated ASIC3 ipsilaterally relative to normal at CFA1 (p < 0.0001), CFA4 (p = 0.0097), CFA7 (p = 0.0003) and CFA14 (p = 0.0114).
Thus, a pattern emerges that shows significantly higher than normal levels of ASIC3 at CFA1 for young rats (indicated by a) followed by a sustained drop from that time onwards. Meanwhile in the old rats the protein levels go up at CFA1 and remain elevated until CFA14 compared to normal (indicated by b), and up to CFA42 relative to contralateral expression levels.
Suppl. Figure 4 presents the data arising from the detailed analysis of ASIC3 expression in DRG neurons classed by size. In the small C fiber-like subpopulation at CFA1, we found a significant elevation ipsilaterally in the aged DRGs (p = 0.0001). From CFA4 up to CFA42 we observed a similar pattern: significantly elevated ipsilateral ASIC3 compared to contralateral in the aged (p = 0.0477, 0.0023, 0.0087, 0.0017 and 0.0374 for CFA4, 7, 14, 28 and 42 respectively), and also significantly higher levels of the receptor in aged versus young rats ipsilaterally (p = 0.0224, 0.0291, 0.0036, 0.0329 and 0.0014 for CFA4, 7, 14, 28 and 42 respectively) (Suppl. Figure 4A).
Suppl. Figure 4B shows the changes in ASIC3 expression in medium-sized DRG neurons. At CFA1, we encountered raised ASIC3 ipsilaterally in young and older rats compared to contralateral (p = 0.0291 and 0.0031, respectively). At CFA4, 7 and 14 we found significantly elevated levels of the receptor ipsilaterally in the aged group compared to its contralateral side (p = 0.0309, 0.0473 and 0.0132, respectively) and also compared to the ipsilateral younger sides (p = 0.0065, 0.0197 and 0.0001, respectively). Unexpectedly, and only at CFA14, we observed a significantly higher % intensity contralaterally in the young rats (p = 0.0290). At CFA28 and 42, the only statistically significant differences occurred between the old versus young ipsilateral sides (p = 0.0172 and 0.0003). However, ipsilateral levels tended to be higher than the corresponding contralateral sides in the aged rats, albeit data dispersion prevented it from achieving statistical significance.
In agreement with our observations taking into account the whole DRG (Fig. 4B), we also found that at all times after CFA injection, the large DRG-aged neurons showed up-regulation of ASIC3 ipsilaterally compared to contralateral (p = 0.0025, 0.0114, 0.0486, 0.0001, 0.0291 and 0.0422 for CFA1, 4, 7, 14, 28 and 42 respectively). This was also the case for young rats at CFA1, the only time we observed a significant increment ipsilaterally (p = 0.0114). With the sole exception of CFA1 and CFA14, at all other times we noticed a statistically significant elevation in ASIC3 when comparing old versus young ipsilateral levels of the receptor (p = 0.0475, 0.0018, 0.0049 and 0.0329 for CFA4, 7, 28 and 42) (Suppl. Figure 4C).
The emerging patterns of age-related differential behaviors (Fig. 1) combined with the time-dependent changes in expression of Nav1.8 and ASIC3 after CFA (Figs. 2–4) suggest that a pharmacological intervention timed to coincide with the peak expression of a nociceptive marker might be a more rational strategy to treat chronic inflammatory pain in different age groups.
Thus, we expected that inhibition of Nav1.8 may be equally effective in young and aged rats whereas antagonism of ASIC3 may be more beneficial in aged rats (as they have higher levels of ASIC3 after chronic CFA inflammation). Bearing this in mind, we decided to test our hypothesis and our findings are described in the next section.
Effects of selective inhibition of Nav1.8 and ASIC3 on behavioral parameters in young and aged rat models of long-term inflammation
Figure 5A shows the effects of A803467 (a selective inhibitor of Nav1.8) and APETx2 (a high-affinity blocker of ASIC3) delivered in vivo by minipump relative to the vehicle over 28 days. APETx2 is the most well-studied and well-known inhibitor of ASIC3 channels; however, at generally high concentrations this toxin was shown to inhibit in vitro DRG Nav1.8 currents with an IC50 of 2.6 µM [11; 80]. In our experiments, we used APETx2 at a concentration of 0.22 µM, ~ 12 times lower than this IC50, and ~ 3 times higher than the IC50 for ASIC3 in the same neurons. Under these conditions and given a reported Hill coefficient close to 1, we expect the toxin effect to be dominated by inhibition of ASIC3 with negligible contributions from Nav1.8 effects.
We conducted this set of experiments over 28 days primarily because in that time frame we observed the resolution of spontaneous pain in both age groups. Additionally, this time interval was long enough to ascertain whether our planned pharmacological interventions altered significantly the evolution of mechanical algesia.
In young rats, we found that both drugs attenuate mechanical allodynia at CFA1 and reverse it completely by CFA21. The treatment with A803467 also alleviated significantly the mechanical-induced pain at CFA4 and led to a slight mechanical hyposensitivity by CFA21 (p = 0.0004; F(DFn, DFd) = 3.338 (12, 102)). The comparison between the two treatments demonstrated non-inferiority for either of them at this age (Fig. 5A left). In aged rats (Fig. 5A right) APETx2 exhibited a small alleviating effect at CFA1. Neither drug was significantly effective at CFA4. This changed from CFA7 onwards, when both drugs first alleviated and then resolved the mechanical pain present (p < 0.0001; F(DFn, DFd) = 4.725 (12, 120)). Interestingly, at CFA1 (p = 0.0007) and CFA14 (p = 0.0498), we found that the treatment with APETx2 was significantly better than A803467.
In Fig. 5B we standardized the raw data from Fig. 5A and plotted the resulting % of baseline PWT against time. This allowed us to evaluate the kinetics of recovery from mechanical hypersensitivity with the treatments. We found that in young rats, A803467 and APETx2 both similarly accelerated the recovery, with these animals surpassing 50% of baseline PWT by day 4 after CFA, whereas in the absence of treatment, the crossover time was at 18 days (left plot). By contrast, in the aged rats, the group that received APETx2 achieved recovery earlier than the rats treated with A803467 (CFA7 vs. CFA14). Thus, antagonizing either ASIC3 or Nav1.8 resulted in a significant shift to shorter times in the recovery from mechanical allodynia compared with untreated rats (see also Fig. 1B). Note that as a whole the pattern of pharmacological response to both drugs evaluated with the % of baseline PWT closely resembles those present in Fig. 5A.
Next, we examined the effect of the drugs on the responses to acetone (Fig. 5C). The plot on the left shows that in young females both compounds lowered the cold sensitivity at CFA1, 4 and 7, an effect that was also present at CFA14 only for A803467 (p < 0.0042; F(DFn, DFd) = 2.604 (12, 114)). Aged rats responded to both treatments only at CFA7 and to APETx2 at CFA21 (right plot) (p = 0.6806; F(DFn, DFd) = 0.7694 (12, 120)). Neither drug showed statistically significant superiority of effect on cold allodynia for either age group.