Age-dependent sensitization of cutaneous nociceptors during developmental inflammation

Abstract

Background: It is well-documented that neonates can experience pain after injury. However, the contribution of individual populations of sensory neurons to neonatal pain is not clearly understood. Here we characterized the functional response properties and neurochemical phenotypes of single primary afferents after injection of carrageenan into the hairy hindpaw skin using a neonatal ex vivo recording preparation.

Results: During normal development, we found that individual afferent response properties are generally unaltered. However, at the time period in which some sensory neurons switch their neurotrophic factor responsiveness, we observe a functional switch in slowly conducting, broad spiking fibers ("C"-fiber nociceptors) from mechanically sensitive and thermally insensitive (CM) to polymodal (CPM). Cutaneous inflammation induced prior to this switch (postnatal day 7) specifically altered mechanical and heat responsiveness, and heat thresholds in fast conducting, broad spiking ("A"-fiber) afferents. Furthermore, hairy skin inflammation at P7 transiently delayed the functional shift from CM to CPM. Conversely, induction of cutaneous inflammation after the functional switch (at P14) caused an increase in mechanical and thermal responsiveness exclusively in the CM and CPM neurons. Immunocytochemical analysis showed that inflammation at either time point induced TRPV1 expression in normally non-TRPV1 expressing CPMs. Realtime PCR and western blotting analyses revealed that specific receptors/channels involved in sensory transduction were differentially altered in the DRGs depending on whether inflammation was induced prior to or after the functional changes in afferent prevalence.

Conclusion: These data suggest that the mechanisms of neonatal pain development may be generated by different afferent subtypes and receptors/channels in an age-related manner.

Figure 1

Percent of total slowly conducting, broad spiking afferents. A: Examples of action potentials generated from a fast conducting, broad spiking afferent (“A”-fiber) and a slowly conducting, broad spiking afferent (“C”-fiber) during development. From the onset of the electrical stimulus (arrows), the initiation of the “A”-fiber action potential was approximately twice as fast as the “C”-fiber potential. B: The percentage of mechanically sensitive, but thermally insensitive “C”-fibers (CM) out of all C-fibers recorded were significantly reduced beginning at postnatal day 10 (P10). C: Conversely, the percentage of mechanically sensitive and heat sensitive (and sometimes cold/cooling responsive) “C”-fiber afferents (CPM) was significantly increased beginning at P10. *p value < 0.05, χ².

Figure 2

Response properties of fast conducting, broad spiking afferents after cutaneous inflammation initiated at P7. A: No changes in mechanical thresholds were found in the fast conducting, broad spiking afferents that were mechanically sensitive (AM) or mechanically and heat sensitive (APM) at any time point after inflammation of the hairy skin initiated at postnatal day 7 (P7). B: However, the heat thresholds of the APM neurons were found to be significantly reduced one and three days after P7 inflammation. Heat thresholds returned to naïve levels by 7d post hairy skin inflammation. C: The mean peak instantaneous frequencies to mechanical deformation of the skin were also found to be significantly increased one day after P7 inflammation in the AM/APM neurons. D: In addition, the mean peak instantaneous frequencies to heat stimuli were enhanced in APMs one day after cutaneous inflammation initiated at P7. *p value <0.05, mean ± SEM; One-way ANOVA with Tukey’s post hoc test for thresholds and Kruskal-Wallis and Dunn’s post hoc test for instantaneous frequencies. See text for animal and afferent numbers.

Figure 3

Response properties of mechanically sensitive, thermally insensitive (CM) and polymodal (CPM) C-fibers after P14 inflammation. A: No changes were observed in the mechanical thresholds of CPM neurons at any time point after inflammation of the hairy skin at P14. B: No differences in heat threshold in the CPM fibers were detected after P14 inflammation. C: However, there was an increase in the mean peak instantaneous frequencies to mechanical stimulation of the skin in the CPM neurons one day after P14 inflammation. This effect was transient as firing returned to naïve levels by the three day time point. D: Similar results were found in terms of the mean peak instantaneous frequencies to heat stimuli in the CPM fibers. However, the enhanced firing to heat was maintained at the three day time point and resolved by day seven. E: Although there was not a statistically significant reduction in mechanical thresholds in the CM neurons, there was a trend towards reduced mechanical thresholds in these afferents one day after injection of carrageenan into the hairy skin at P14. F: The CM fibers however did display increased mean peak instantaneous frequencies to mechanical stimulation of the skin one day after P14 inflammation. *p value <0.05; ^#p value < 0.1, mean ± SEM; One-way ANOVA with Tukey’s post hoc test for thresholds and Kruskal-Wallis with Dunn’s post hoc tests for mean peak instantaneous frequencies. See text for animal and afferent numbers.

Figure 4

Prevalence of mechanically sensitive, thermally insensitive (CM) and polymodal (CPM) “C”-fibers after inflammation at P7. A: The reduction in the percentage of total CM fibers at P10 in naïve mice was delayed by cutaneous inflammation at P7. The percentage of CM neurons detected three days after P7 inflammation (P10) was not found to be different than the percentage of these cells in naives at P7 or at one day post carrageenan injection (P8). The numbers of CM neurons detected at the three day time point was significantly higher than that detected in time matched naives at P10. By seven days post inflammation (P14), there was no difference in CM fiber prevalence between naives and inflamed mice. B: The normal increase in CPM neuron prevalence at P10 in naïve mice was also delayed by inflammation at P7. The percentage of CPM neurons detected three days after P7 inflammation of the hairy skin was significantly lower than the numbers of CPM neurons found in time matched naives (P10). The increase in CPM neuron prevalence was resolved however by the seven day time point after inflammation (P14). *p value < 0.05 vs P7 and P8 (1d post inflammation); **p value < 0.05 vs time matched naïve controls (P10), χ². See text for animal and afferent totals.

Figure 5

Intracellularly stained and physiologically characterized sensory neurons in naïve mice from the ex vivo preparation. A neurobiotin stained “A”-fiber neuron (arrows) from a P7 mouse (A) that responded to mechanical stimuli and heat stimuli (D) was found to be immunoreactive for both TRPV1 (B; red) and ASIC3 (C; blue). One “C”-fiber neuron (arrows) recovered from a P8 mouse that was intracellularly stained with neurobiotin (E) and responded to mechanical deformation of the skin but not heat stimuli (H) was found to be immunonegative for TRPV1 (F; red) and did not bind isolectin B4 (IB4; G; blue). A polymodal (L) “C”-fiber neuron intracellularly stained with neurobiotin (I) and recovered from a P21 naïve ex vivo preparation (arrows) was found to be TRPV1 immunonegative (J; red), but did bind IB4 (K; blue). Scale bar for images: 40 μm.

Figure 6

Intracellularly stained and physiologically characterized afferents in mice with inflammation recovered after ex vivo recording. A neurobiotin (NB) stained (A) “C”-fiber neuron (arrows) that was recovered from a mouse seven days after P14 cutaneous inflammation was found to be mechanically sensitive but thermally insensitive (D), and was found to contain TRPV1 protein (B; red) but did not bind isolectin B4 (IB4; C; blue). The TRPV1 staining in this particular neuron did not appear to be membrane bound as the labeling pattern for TRPV1 was within the staining pattern for NB (A). One polymodal (H) “C”-fiber (arrows) that was filled with NB (E) was found to contain TRPV1 (F; red) but did not bind IB4 (G; blue). This cell was recovered from a mouse with cutaneous inflammation induced at P14 (three day time point). Finally, another polymodal “C”-fiber (L; arrows) that was intracellularly filled with NB (I) was found to contain both TRPV1 (J; red) and bound IB4 (K; blue). This latter cell was obtained from a mouse with inflammation at P7 at the one day time point. Scale bar for images: 40 μm.

Figure 7

Western blot analysis of whole DRGs in naïve and P7 inflamed mice for GFRα1. A: Quantification of GFRα1 protein in naïve L2/L3 DRGs from P7 through P21 revealed that levels of this co-receptor were significantly elevated beginning at P10 and remained elevated thereafter compared to P7 levels. B: However, after cutaneous inflammation at P7, GFRα1 protein did not display enhanced expression at one day or three days after inflammation; the age in which it normally increased in naïve ganglia (P10). In fact, there appeared to be a slight decrease in the mean expression of GFRα1 protein after P7 inflammation, but this did not reach statistical significance. C: Comparing levels of GFRα1 protein between P7 and P10 (which is also 3d post inflammation) in naïve and inflamed mice shows that inflammation of the hairy hindpaw skin blocked the normal increase in this receptor normally seen in naïve mice. D: Examples of western blots for GFRα1 protein in naïve mice from P7 through P21 or after P7 inflammation at 1d and 3d. Bands are detected around the predicted molecular weights for GFRα1 and GAPDH. *p value <0.05 vs P7 naives. Values for GFRα1 are normalized to GAPDH levels prior to quantification. Normalized mean ± SEM; One-way ANOVA with Tukey’s post hoc test. Presented as percent changes.

Figure 8

Schematic representation of the age-dependent sensitization of cutaneous sensory neurons during developmental inflammation. Peripheral injury (cutaneous inflammation) sustained prior to the observed critical period (P10; bold) of sensory neuron development (at P7; arrow) specifically results in sensitization of A-fibers (AM/APM; left). Inflammation of the skin after this time point however (at P14; arrow) only sensitizes mechanically sensitive, thermally insensitive (CM) and polymodal (CPM) C-fibers to peripheral stimuli (right). We hypothesize that these unique age-related changes in afferent responsiveness may be due to upregulation of specific genes in the DRGs at the different time points (*).