The daily fluctuations of many physiologic and behavioral parameters are differentially influenced by either central or peripheral clocks in mammals. Since substance P (SP) oscillates in some brain tissues and plays an indispensable role in modulating inflammatory pain at the spinal level, we speculated that SP mediates circadian nociception transmission at the spinal level.


In the present study behavioral observation, real-time polymerase chain reaction, luciferase assay, chromatin immunoprecipitation, and immunohistochemistry stain methods were used to investigate the role of SP in the spinal circadian nociception transmission and its regulation mechanism.


Our results showed that under transcriptional regulation of BMAL1:CLOCK heterodimers, SP's coding gene Tac1 expression oscillates in dorsal root ganglion (n = 36), but not in the spinal dorsal horn. Further, the expression of SP cycled in the spinal dorsal horn, and this rhythmicity was potentially determined by circadian expression of Tac1 in dorsal root ganglion. Furthermore, the variation of SP expression induced by formalin was fluctuated in a similar rhythm to behavioral nociceptive response induced by formalin (n = 48); and the nociceptive behavioral circadian rhythm could be abolished through blockade of the SP-Neurokinin 1 receptor pathway (n = 70). Lastly, the variations of spinal SP expression and behavioral nociceptive response were in step, and both were changed by the deletion mutation of clock gene.


We conclude that spinal SP probably plays a pivotal role in modulating circadian inflammatory pain and suggest that peripheral circadian-regulated signaling is potentially an essential pathway for circadian nociceptive transmission.

  • Molecular controls for circadian rhythms have been primarily examined in the suprachiasmatic nucleus, which appears to control peripheral tissue “clocks”

  • Although diurnal variations in pain have been observed, their causes are incompletely understood

  • In mice, substance P production in the dorsal root ganglion and expression in the spinal cord varies in a diurnal nature, accompanied by a diurnal variation in the response to an acute inflammatory stimulus

  • Peripheral clocks may partially explain the diurnal variations in pain

IN mammals, the internal molecular circadian system, with interacting positive and negative feedback loops, not only generates its own oscillation but also regulates the clock-controlled genes to maintain the daily rhythm of many physiologic and behavioral events. The suprachiasmatic nuclei (SCN) has been thought to synchronize and/or maintain the circadian rhythms through its connection with the internal timekeeping system of the passive peripheral clock.1,2However, a rapid growing body of evidence demonstrates that diverse peripheral tissues possess their own clocks and can sustain rhythmicity even in the absence of the SCN,3,,5indicating the molecular clock machinery functions in specific peripheral tissues.

Circadian variations of pain sensation have been documented in human patients with different clinical syndromes6,,8and in animal models for the studies of pain.9,10However, the daily fluctuations in different aspects of a kind of pain sensitivity and physiology have different phases, suggesting that there may be multiple molecules and regulatory pathways modulating different kinds of circadian pain, but up to now the detailed molecular mechanism remains unclear.

One important pain-related modulator, substance P (SP), a tachykinin neuropeptide, is mainly synthesized in small primary afferent sensory neurons of dorsal root ganglion (DRG) cells, and sequentially modulates nociceptive transmission in the spinal level.11,12Previous reports indicated that spinal SP plays an indispensable role in the modulation of inflammatory pain induced by formalin hind paw injection,11,13,14and the behavioral reactions change in diurnal rhythm in mice.15,16It is worthy to note that SP oscillates in many brain regions17,,20except for the mouse SCN,21and it plays a critical role in the photic entrainment of circadian system in rats,21,22suggesting that there could be an association between SP and circadian timing system outside the mouse SCN. We then hypothesized that at the spinal level, molecular genetic clock machinery produces SP oscillation, thereby modulating circadian nociceptive signal transmission. In the present study, we focus on how and to what extent; formalin-induced circadian inflammatory nociception is mediated by SP through transcriptional regulated oscillation in periphery.

General Procedures

All procedures were performed on a total of 384 male C57 BL/6 and Per2Brdm1  mutant mice, 8–10 weeks of age. Per2Brdm1  is a mouse line in which the clock gene Per2  is deleted. The phenotype of the deletion mutation is consistent with a loss-of-function mutation. Animal protocols were approved by the Committee of Animal Use for Research and Education of the Institute of Psychology of Chinese Academy of Sciences (Beijing, China; No. A09030). In general, mice were housed in a standard specific pathogen-free animal facility with room temperature of 23 ± 1°C and given ad libitum  access to food and water. Before each experiment, mice were kept for at least 10 days in a 12-h light/12-h dark cycle, with light on at 6:00 AM (referred to as zeitgeber time-point 0, ZT0) and light off at 6:00 PM (referred to as ZT12).

Behavioral Observation in Formalin Test

After habituation to the testing room for 30 min, the mice received a subcutaneous injection of formalin solution (5%, 10 μl) into the plantar surface of left hind paw. In this part, the mice were divided into four groups: formalin-treated or formalin-untreated of wild-type mice; formalin-treated or formalin-untreated of Per2Brdm1  mutant mice. Mice in each group randomly divided into six time-point subgroups evenly distributed in 24 h (n = 8 for each subgroup at each time-point). The accumulated time spent licking and lifting the injected paw was recorded during two phases: an acute phase (0–10 min after injection) and a tonic phase (10–60 min after injection). Night vision equipment (NV Tracker 1X24 Goggles; Yukon Advanced Optics Worldwide, Inc., Lithuania, Belorussia) was used to observe pain behavior in darkness. All of the data were collected in a blinded fashion.

RNA Isolation and Real-time Polymerase Chain Reaction (PCR)

Bilateral lumbar 3 (L3)–L5DRG and spinal cord (SC) segments were dissected and collected on dry ice, and total RNA was isolated and purified, respectively (6 mg of DRG and 30 mg of SC). The concentration of each individual total RNA sample was standardized as 250 ng/μl. Equal volume of this standardized total RNA from six mice (same time-point) were pooled and used for complementary DNA synthesis. To generate single-strand complementary DNA, 2 μg total pooled RNA was used as the starting template for the first strand complementary DNA synthesis, using the PCR complementary DNA Synthesis Kit (Promega, Madison, WI), according to the manufacturer's instructions.

Real-time PCR was performed using the Bio-Rad Laboratories DNA Engine OPTICON 2 system (Hercules, CA) with SYBR Green detection, and the primers were listed (see Supplemental Digital Content 1, Table S1,, which is a table listing all primers for real-time PCR used in this study). Results were first normalized through the amount of target gene messenger RNA (mRNA) in relation to the amount of reference 18s ribosomal RNA gene. The values at other time-points were calibrated to the value of the time-point with the highest mRNA expression level, which is designated as 1. All data were collected in a blinded fashion.

Chromatin Immunoprecipitation Assay

The eight mice pooled DRG (L3–L5) tissues were homogenized in phosphate buffered saline containing phenylmethanesulfonyl fluoride, and fixed with 1% formaldehyde. The cells were then lysed in 5 ml lysis buffer on ice. After centrifuged, the pellet was resuspended with 1 ml sonication buffer for 20 min and subsequently sonicated 10 times (15 s, 15 s spaced). Then, 20% of the total supernatant fraction was collected as input control. Twenty percent of the total chromatin supernatant was diluted fivefold with dilution buffer. The diluted solutions were shaken in a rotary incubator for 30 min at 4°C with 40 μl salmon sperm DNA/protein A agarose slurry, and followed by centrifugation for 1 min. The resulting supernatants were incubated with 14 μl antibodies overnight at 4°C, then 30 μl salmon sperm DNA/protein A agarose was added and shaken in the rotary incubator for 4 h at 4°C. After centrifugation, the protein A agarose/antibody/histone complex were washed with 1 ml of washing buffers and eluted twice with 250 μl elution buffer. Cross-links were reversed by adding 20 μl of 5 M NaCl to all reactions, and heating at 65°C for 4 h. DNA was extracted and real-time PCR was carried out on extracted DNA samples with five sets of primer probe sets, which covered 1.313 kb of the promoter region of Tac1  (see Supplemental Digital Content 1, Table S2,, which is a table listing all primers for chromatin immunoprecipitation assay used in this study). Results were expressed as C  Tvalues, which were used to determine the amount of BMAL1 binding DNA. ΔC  Tindicated the difference between the number of cycles necessary to detect the PCR products for BMAL1 binding DNA and corresponding reference input DNA. ΔΔC  Twas the difference between the ΔC  Tof the different pooling tissue samples at each time-point of ZT (ZT4, ZT12, and ZT20). Data were expressed as 2−ΔΔC  Tto give an estimate of the amount of BMAL1 binding DNA in the tissue at different time-points relative to the reference input DNA.

Construction of Plasmids

The DNA fragment of Tac1  promoter was obtained by amplification of a 1.538 kb fragment of the 5′ flanking region of Tac1  gene from mouse genomic DNA, using the primers as the following:



This fragment was then directly cloned into Kpn  I-Smal  digested pGL3-Basic luciferase reporter vectors (Promega), resulting in a −1011/+527 bp (relative to the transcriptional initiation site) promoter construct, which was confirmed by sequencing. Plasmid of hBmal1  and hClock  were provided by Dr. Xiao Zhong Peng (Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China23). hClock  was cloned into the Hind  III and Xho  I sites of pcDNA3.1, whereas hBmal1  was cloned into the Bam  H I and Eco  R I sites of pcDNA3.1.

Mutagenesis of E-box

The Class I E-box at −61 bp upstream of the transcriptional start site in the promoter region of the Tac1  gene was mutated from CACGTG to TGAGTG by site-directed mutagenesis using the Muta-DirectTMKit (Beijing SBS Genetech Co., Ltd., Beijing, China), according to the manufacturer's instructions. For a 50 μl reaction, 20 ng plasmid DNA template (constructed with 1.538 kb fragment of the 5′ flanking region of Tac1  gene, as described previously) was mixed with 10 pM forward primers (5′- CGTGGGGAGAGTGTTGAGTG GCTCTACAGGCT −3′) and reverse primers (5′- AGCCTGTAGAGCCACTCA ACACTCTCCCCACG −3′), 2 μl dNTP mixture (each 2.5 mM), 1 μl Muta-direct™ enzyme (Beijing SBS Genetech Co., Ltd., Beijing, China), 5 μl 10× reaction buffer, and 38 μl ddH2O. The reaction was first incubated at 95°C for 30 s, followed by 18 cycles of 95°C for 30 s, 55°C for 1 min, and 72°C for 6.5 min. The PCR product was incubated at 37°C for 3 h with 1 μl (10U/μl) of Mutazyme™ enzyme. Ten microliters of the enzyme-treated sample was put into 50 μl competent cell (DH5 α) and then transformed. The mutated construct was verified by sequencing.

Transcriptional Assay

293T cells (human embryonic kidney 293T cell line, Cell Bank of Type Culture Collection of Chinese Academy of Sciences, Shanghai Institute of Cell Biology, Chinese Academy of Sciences, Shanghai, China) were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin (HyClone, Beijing, China). One day before transfection, plated cells were in 1 ml of growth medium without antibiotics so that cells would be 90–95% confluent at the time of transfection (0.5–2 × 105cells/well for a 24-well plate). The cells were transfected with 1.6 μg (total) of plasmids using LipofectamineTM2000 (Invitrogen, Paisley, United Kingdom), according to the manufacturer's instructions. Cell extracts were prepared 24 h after transfection by a lysis buffer. Twenty microliters of the extract were taken for a luciferase assay using a luminometer (Victor3 V Multilabel Counter model 1420; PerkinElmer, Waltham, MA), as described by the manufacturer (Promega). For statistical analysis, a paired Student t  test was applied.

Immunohistochemistry Stain

The timing of the SP-immunoreactivity measurements in the formalin-treated animals was the time-window 40–45 min after 10 μl 5% formalin injection or formalin-untreated mice at the same time. As our previous report,24all animals were anesthetized and then perfused through the ascending aorta with 0.9% (w/v) saline followed by 100 ml of 4% (w/v) paraformaldehyde in 0.1M phosphate buffer (pH 7.4). After perfusion, the L4and L5SC of wild-type mice (naïve, n = 6; formalin-treated, n = 4) and Per2Brdm1  mutant mice (formalin-untreated, n = 4; formalin-treated, n = 4) were removed and saturated overnight at 4°C. Frozen sections (30 μm-thick) were cut and the sections were incubated free floating at 4°C in a monoclonal rat-anti-SP (1:500; MAB356; Chemicon, Billerica, MA) for 48 h, followed by biotinylated rabbit-anti-rat IgG (1:200; Vector, Burlingame, CA) for 6 h at room temperature. Finally, the sections were incubated with CyTM3-conjugated Streptavidin (1:1000; Jackson Immunoresearch, Newmarket, United Kingdom) for 4 h at room temperature. In addition, the sections have incubated with fluorescein Griffonia simplicifolia lectin (isolectin B4, IB4) (1:200; Vector) for 4 h. To characterize primary antibody specificity, we employed a negative control, replacing the primary antibody with antibody dilution, and a positive control, using the antibody with cells known to contain SP (see Supplemental Digital Content 2, Figure S1,, which is a figure showing the negative and positive controls for SP-immunoreactivity).

In dorsal rhizotomy experiment, four mice for each time-point were employed. After anaesthetization, one side of the SC of mice was exposed by laminectomy at the L6segment region. L3–L5right dorsal roots were transected. Six days later, the animals were perfused at ZT4 or ZT20. The SCs were removed for immunohistochemistry fluorescent stain.

Quantitative Analysis

The sections were observed with a confocal laser-scanning microscope (FV1000; Olympus, Tokyo, Japan). Adapted from Ranson's report,25the area measurement of SP-immunoreactivity terminal profile (10–12 nonadjacent sections per mouse) in the superficial layer (but not cell body in the deep layer) was analyzed at the injection side in the formalin-treated mice or ipsilateral side in naïve mice by using the software (FV10-ASW, Olympus). The data were collected in a double-blinded fashion. To determine the labeling area for each time-point exactly on the same coronal sections, because the size of the labeling area can be influenced by the exact position of the section in the antero-posterior axis, the total areas of SC were analyzed for homogeneity of variance using ANOVA. The analysis results revealed that there were no statistically significant of-treatment effects, time-of-day effects, or the interactions of treatment-time.

Intrathecal Administration

A mechanically modified polyethylene-10 tubing (OD = 0.28 mm, ID = 0.07 mm) was inserted from the thoracic 2 (T2) level of the SC and the terminal reached at L5level for intrathecal administration. At least 5 days after surgical operation the mice without dyskinesia were used in behavioral tests. In brief, the data from 104 mice were used in statistical analysis (antagonist treatment: 38 mice, vehicle treatment: 32 mice, and dose-dependent experiment: 34 mice). To test the dose-dependent effects, 10 min before formalin injection, different doses of the antagonist, N -acetyl-l-tryptophan-3, 5-bis(trifluoromethyl)benzyl ester (L-732,138) (dissolved in 70% dimethyl sulfoxide, Sigma-Aldrich, St. Louis, MO) were intrathecally administered with a volume of 5 μl. Furthermore, 100 nM of L-732,138 or vehicle was used to test the time-dependent variation of nociceptive behavioral response.

Data Analysis

All values are reported as mean ± SE. Data analysis was performed with the software package SPSS 13.0 for Windows (SPSS, Inc., Chicago, IL). The statistical significances were determined using one-way ANOVA followed by Tukey post hoc  test (the statistically significant main effects for the one-way ANOVA are followed by post hoc  testing that adjusts the P  values of the individual post hoc  tests), paired Student t  test, independent Student t  test, or Pearson's correlation analysis; P < 0.05 was considered significant.

Time effects were analyzed with one-way ANOVA. Then, cosine regression analysis of each time-series was performed using SPSS 13.0 statistics software and MATLAB. Each data set was fitted to a general cosine equation model:

A  and B  are predictors for this function,26,27 T  is the period (24 h or 12 h in the present study), and M  is the MESOR (midline estimating statistic of rhythm). A R2value and a P  value for the rejection of the zero-amplitude assumption were determined for each component in the cosine model separately and overall, with rhythm detection considered statistically significant if P < 0.05 for any period tested.

Driven by BMAL1:CLOCK Heterodimers, SP-encoding Gene (Tac1) Diurnal Oscillated in DRG

To test our initial hypothesis of the existence of functionalized clock molecular machinery in peripheral tissue-modulating circadian nociceptive transmission, we examined expression profiles of circadian clock genes and SP-encoding gene Tac1  by real-time PCR in DRG and SC of mice entrained to a 24-h light-dark cycle (12:12). As expected, in DRG the Tac1  gene and the clock genes (Bmal1 , Clock , Npas2 , Per1 , Per2 , Rev-erb α) showed robust circadian expression (fig. 1A and 1B, and see Supplemental Digital Content 1, Table S3,, which is a table listing rhythm parameters for mRNA expression of genes in DRG and SC). Transcriptions of the positive factors (Bmal1 , Clock , Npas2 ) oscillated in antiphase to those of the negative factors (Per1 , Per2 , Rev-erb α) (fig. 1C). This pattern is consistent with the previous reports in other peripheral tissues.28,29In addition, in the SC both positive regulators (Bmal1  and Clock ) and Tac1  were expressed in weak (Bmal1 , P = 0.03, M = 0.68, amplitude, or AMP = 0.3) or no rhythmic oscillation (see Supplemental Digital Content 2, Figure S2,, which is a figure showing the mRNA expression level of circadian genes and Tac1  in SC and Supplemental Digital Content 1, Table S3, As a control, although N -methyl-D-aspartate receptors (NAs) have been demonstrated to regulate many aspects of pain transmission,30,31we did not detect circadian oscillation in the expression of NR1 , NR2A , NR2B , and NR2C  subtypes in DRG and the SC (fig. 1D, and see Supplemental Digital Content 2, Figure S3,, which is a figure showing the mRNA expression level of NRs  in SC). By contrast, their expression curves were likely to pulse in a 12-h rhythm.

Next was questioned how the molecular clock governs Tac1  gene expression in DRG. In general, rhythmic transcription can be driven by CLOCK:BMAL1 heterodimers or Rev-erb α binding to the transcriptional enhancers, such as E boxes or Rev-erb α monomer-binding sites in the promoter region of the clock-controlled genes.1Analysis of the promoter region of the mouse Tac1  gene revealed that within the 1.6 kb region upstream of the transcriptional start site, there was one class I (CACGTG) and five class II (CANNTG) putative E-box motifs but no Rev-erb α monomer-binding sites consensus motif (fig. 2A). To examine whether CLOCK:BMAL1 heterodimers drive Tac1  gene transcription through these E-box enhancers, we analyzed DNA-binding activity in each E-box by using chromatin immunoprecipitation assay. The results displayed that in both DRG and SC, the class I E-box exhibited obvious binding activity for BMAL1 (fig. 2B1, and see Supplemental Digital Content 2, Figure S2, In contrast, the class II E-box elements didn't show measurable BMAL1-binding ability. Furthermore, this BMAL1-binding activity of the class I E-box displayed a significant time-effect (one-way ANOVA followed by Tukey post hoc  test, F = 34.69, P < 0.01) from ZT4 to ZT20 (fig. 2B2). Using a luciferase reporter assay in 293T cells, we monitored the promoter activity of the Tac1  gene in a 1.6 kb region 5′ to the start site of transcription. The results showed that only when coexpressed, but not expressed alone, BMAL1 and CLOCK up-regulated luciferase activity by 15-fold (paired Student t  test, P < 0.001, fig. 2C), and the CLOCK:BMAL1-dependent activation was abolished when the class I E-box was mutated.

Oscillation of SP in SC Originated from Tac1  Gene Circadian Expression in DRG

Next, we wanted to determine whether SP expression oscillated in SC. Immunohistochemistry fluorescent stain results revealed that in lumbar SC sections, SP-immunoreactivity was mainly concentrated in primary afferent terminals of the superficial laminae (lamina I and the outer part of lamina II), and its expression level oscillated with a trough at ZT4 and a peak at ZT20 in mice (figs. 3A1 and 3A2, and see Supplemental Digital Content 1, Table S4,, which is a table listing the rhythm parameters for SP expression in SC). By contrast, IB4-immunoreactivity was distributed in the inner part of lamina II with a steady fashion (fig. 3B). Further, once L3–L5primary afferent filaments were unilaterally rhizotomized, SP-immunoreactivity expression in the superficial laminae completely diminished on the dorsal rhizotomy side at ZT4 (trough) and ZT20 (peak), but remained intact on the uncut side (fig. 3C). These results are consistent with previously published work32in confirming that SP in the superficial layer of dorsal horn were mainly transported from primary afferents.

Formalin-induced Diurnal Variation of Nociceptive Behavioral Response

To identify the diurnal pattern of formalin-induced nociceptive reactions in mice, formalin-induced nociceptive behavioral responses were recorded at the six time-points in behavioral observation. The results displayed that there was a weak fluctuation (M = 171.02, AMP = 34.16) of the nociceptive reactions in the acute phase (fig. 4and also see Supplemental Digital Content 1, Table S5,, which is a table listing the rhythm parameters for formalin-induced nociceptive behavioral response, with P = 0.05, one-way ANOVA followed by Tukey post hoc  test; P = 0.001, M = 171.02, AMP = 34.16, R2= 0.89, cosine regression analysis). Differently, in the tonic phase, the nociceptive response showed a robust pattern in diurnal oscillation with a peak at ZT4 and a trough at ZT20 (P = 0.017, one-way ANOVA followed by Tukey post hoc  test; P = 0.005, M = 452.86, AMP = 53.36, R2= 0.83, cosine regression analysis).

Correlation between Formalin-induced Variations of SP Expression Level and Oscillation of Nociceptive Behavioral Response

To further investigate the relationship between the nociceptive behavioral response and the formalin-induced SP expression in spinal dorsal horn, we measured SP-immunoreactivity expression in the SC after formalin injection. As shown in figures 3A1 and 3A2, at each time-point formalin induced up-regulation of SP-immunoreactivity expression to a certain level ([6.17 ± 0.25]×104μm2∼[6.59 ± 0.19]× 104μm2). The cosine regression analysis showed that the level of SP-immunoreactivity expression did not cycle across a day after formalin injection (fig. 3A2, and see Supplemental Digital Content 1, Table S4, Subsequently, by using the temporal point-by-point analysis to compare the difference between the formalin-treated curves and the control curves, we detected the pattern of formalin-induced SP expression variations in the superficial laminae of the SC. As shown in figure 5A, the formalin-induced elevation of SP-immunoreactivity expression flattened the circadian oscillation (see Supplemental Digital Content 1, Table S4,, and it is worth to note that the pattern closely matched the diurnal fluctuation in inflammatory nociceptive behaviors induced by formalin (Pearson's correlation coefficient, r = 0.97, P < 0.01; fig. 5B).

Abolishment of Circadian Nociceptive Feature by Blockade of SP-neurokinin-1 (NK1) Receptor Pathway

To further verify whether SP is involved in the circadian feature of nociceptive behavior, the SP–NK1 receptor pathway was blocked by intrathecal administration of a nonpeptide NK1 receptor antagonist L-732,138 before subcutaneous formalin injection. Behavioral observation results indicated that administration of L-732,138 attenuated inflammatory nociceptive responses in a dose-dependent manner (fig. 6A). We then chose the dosage of 100 nM of L-732,138 valuate the efficiency of SP–NK1 receptor pathway on the circadian feature of nociceptive behavioral response. As shown in figure 6B and table S5, although nociceptive behavioral response were maintained in a certain degree for noxious stimuli (max: 209.29 ± 20.87 s at ZT16, min: 185.20 ± 11.85 s at ZT8), their circadian variations were abolished (F = 0.14, P = 0.982, one-way ANOVA followed by Tukey post hoc  test). In addition, NK1 receptor does not show any circadian expression either in SC or in DRG (see Supplemental Digital Content 1, Table S3,, and Supplemental Digital Content 2, Figure S4,, which is a figure showing the mRNA expression level of NK1  in DRG and SC).

Deletion Mutation of Clock Gene Altered SP Expression and Nociceptive Behavior Response

To evaluate the regulation effects of clock genes on spinal SP and nociceptive behavioral response, Per2Brdm1  mutant mice were used in behavioral observation and immunohistochemistry stain experiments. In Per2Brdm1  mutant mice, the acute response to formalin nociceptive stimulation did not show a statistically significant circadian fluctuation; however, the tonic response showed marked circadian oscillation (fig. 7A, and also see Supplemental Digital Content 1, Table S5, Notably, this oscillation in Per2Brdm1  mutant had an opposite phase to the one in the wild-type with a peak at ZT16 and a trough at ZT8 (see fig. 4B and fig. 7A2).

We then tested the 24 h profiles of SP expression in primary afferent terminals in the superficial laminae of the spinal dorsal horn in Per2Brdm1  mutant mice. In the absence of a nociceptive formalin injection, the 24-h cosine regression analysis showed that SP expression oscillated with a trough at ZT14 and a peak at ZT2 (fig. 7B, and also see Supplemental Digital Content 1, Table S4, Furthermore, after formalin injection the curve of SP expression level oscillated with a peak at ZT0 and a trough at ZT12. Same as in wild-type mice, we detected the pattern of formalin-induced SP expression variations in the superficial laminae of the SC. The results indicated that the variation pattern was in an oscillated manner. Pearson's correlation analysis further showed a significant fit between the variation pattern of SP expression and the behavioral circadian fluctuation induced by formalin injection in Per2Brdm1  mutant mice (Pearson's correlation coefficient, r = 0.88, P < 0.05; fig. 7C).

It used to be generally thought that circadian clock in the SCN is a master oscillator to synchronize and initiate passive peripheral clocks in peripheral tissues. However, a recent study in the functioning of the synthetic mammalian circadian system33suggests that circadian machinery could perform in the cell with an autonomous manner in the absence of SCN clock. Other studies demonstrated that the peripheral food-entrained clock could overcome the influence of the SCN clock to shift the phase of mouse locomotor activity, further suggesting that some circadian behavior could be controlled by peripheral clock.3,4We have now shown that in DRG there is a circadian positive and negative feedback loop and cyclic expression of the nociceptive related gene, Tac1 . And further, the Tac1  circadian expression in DRG is controlled by the transcriptional regulation of CLOCK:BMAL1 heterodimers through circadian binding on the upstream class I E-box element of the Tac1  promoter. This evidence indicates a crucial relationship between machinery of the ganglionic circadian positive and negative feedback loop and generation of rhythm of Tac1  expression. Spinal SP oscillated in a diurnal rhythm and circadian feature was vanished when SP–NK1 receptor pathway was blocked, demonstrating that spinal SP is potentially a crucial role in spinal circadian nociceptive transmission. Furthermore, the unilateral disruption of primary afferent suggests that SP in the dorsal horn mainly originated from the DRG, combining with the diurnal fluctuation of Tac1 . It is reasonable to speculate that the circadian feature of nociceptive transmission in the SC was potentially controlled by the clock in DRG, rather than by the SCN central clock. Additionally, a previous study used retrograde and anterograde tracing to show that there is no direct neuronal projection between SCN and the superficial layer of the dorsal horn, although there is a bineuronal connection from SCN to the intermediolateral cell column in the SC.34 

Although our results suggest that clock machinery in the DRG probably regulates circadian expression of SP via  regulation of Tac1  transcription, the indirect stimulatory role of the central circadian mechanism in modulating SP oscillation cannot be completely eliminated. This is because of two reasons: first, although the circadian clocks operative in peripheral cell types are as robust as those residing in SCN neurons, in intact animals the phase coherence between peripheral oscillators must be established by daily signals generated by the SCN master clock.1Second, SCN could regulate the rhythm in peripheral organs through the autonomic nervous system and the hormone secretory system, and these systems can also indirectly influence nociceptive primary afferents.5,35,36 

As a member of the tachykinin family, SP is considered to be a pain-related neurotransmitter from primary sensory afferent fibers.37It is distributed in numerous regions in the central nervous system but is highly concentrated in the most superficial regions of the dorsal horn.38,39Once peripheral tissue is injured by prolonged noxious stimulation, SP and excitatory amino acids (glutamate or aspartate) are released from the primary afferent terminal, binding to the neuronal cell membrane NK1 and NRs, respectively, and cell excitation then occurs with calcium ion influx into neuronal cells by means of membrane-dependent events.40Recent works on inflammatory nociceptive defects (attenuation of licking behavior) after formalin injection in both Tac1  knockout mice11,14and the African naked mole-rat (heterocephalus glaber ), a species naturally lacking SP,13supports the hypothesis that spinal SP plays an indispensable role in modulating inflammatory pain. Furthermore, several studies have demonstrated that the increase of SP in the dorsal horn follows the nociceptive stimulus induced by formalin injection,41,42and the long-lasting conveyance of nociceptive input also increases SP-encoding mRNA expression in DRG.43It is worthy to note that these previous results merely suggested the qualitative correlation between the increment of SP and the intensity of inflammatory noxious stimuli in a single time-window. Differentially, our study presented the diurnal fluctuation of SP variations under same exogenous stimuli, and this fluctuation is well matched by the oscillation of behavioral response, these demonstrating that the circadian nociceptive transmission is potentially because of the regulation of DRG clock on Tac1  gene, and further the activity of the peripheral clock is autonomic and independent on the intensity of exogenous stimuli. The changes in SP expression and circadian behavioral nociceptive response in clock gene-deleted mice supplied an evidence for the effects of clock genes on spinal nociceptive transmission.

The oscillation of Tac1  mRNA hints that the fluctuation of synthesis of SP in DRG is one reason for the mediation for circadian nociceptive transmission. However, the formalin-induced SP-like immunoreactivity increase in the dorsal horn is blocked by intrathecal naloxone treatment;44on the other hand, the mediation of SP in the sorting of δ-opioid receptors into large dense-core vesicles is essential for modulation of the sensitivity of nociceptive afferents to opioid.45This suggests that the rhythmic increment of formalin-induced SP expression in the dorsal horn is also probably resulted by the demand amount of potentially rhythmic sorting of δ-opioid receptors, which is not proved, and further correlates with a potential circadian opioid-mediated spinal analgesia. Furthermore, previous studies have indicated the release of SP into the superficial dorsal horn following nociceptive activation of the hind paw and the efficacy of a systemically administered NK-1 receptor antagonist in blocking SP-induced facilitation of a spinal nociceptive reflex46,,48, which is coordinated with the present results that with strong inhibition of the SP pathway's nociceptive-transmitting efficiency in the superficial dorsal horn, the circadian feature of inflammatory pain was completely abolished. Therefore, we also cannot exclude the probability of the circadian feature of formalin-induced release of SP in the superficial dorsal horn.

The nociceptive pathways of persistent inflammatory pain are distinguished from other categories of pain, such as physiologic pain, which has acute traits49,,52, and neuropathic pain, which is characterized by nerve tissue damage53,,55. Further, the different oscillating characteristics between hot pain10and formalin-induced inflammatory pain also suggest that there could be various regulatory molecules and pathways accounting for the divergences of circadian features in different pain sensitivities. Therefore, although our results indicate that SP is specifically involved in formalin-induced circadian inflammatory pain, the possibility of participation of other molecules cannot be excluded in other types of pain transmission. However, the present results of the circadian feature of acute inflammatory, which is mediated by spinal SP and controlled by peripheral clock, combining with the clinical reports6,,8, urges us to rethink the therapeutic treatment in clinic, especially in perioperative medicine/pain management or trauma medicine in which inflammation is involved. Perhaps we have to reconsider the management of acute pain, and the time of the day the patient is exposed to acute pain. Also, future drugs focused on SP–NK1 receptor pathway should be paid more attention to, and maybe the complex function of analgesic drugs and the drugs for local circadian regulation will be considered in the future.

In summary, the present study demonstrated that SP in the spinal dorsal horn is potentially a major factor in modulating the circadian inflammatory nociceptive response, driven by autonomic peripheral circadian signaling originating from DRG. Our results provide new insight into understanding the molecular mechanism of other types of circadian pain behaviors, and could potentially lead to developing chrono-based means for clinical pain management.

The authors thank Charles Randy Gallistel, Ph.D. (Professor, Department of Psychology, Rutgers University, New Brunswick, New Jersey), and Xiaoxi Zhuang, Ph.D. (Professor, Department of Neurobiology, Pharmacology and Physiology, University of Chicago, Chicago, Illinois), for constructive comments during the preparation of this manuscript. The authors also thank Xiao Zhong Peng, Ph.D. (Professor, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China), for the generous gift of plasmids.

Reppert SM, Weaver DR: Coordination of circadian timing in mammals. Nature 2002; 418:935–41
Van Gelder RN, Herzog ED, Schwartz WJ, Taghert PH: Circadian rhythms: In the loop at last. Science 2003; 300:1534–5
Fuller PM, Lu J, Saper CB: Differential rescue of light- and food-entrainable circadian rhythms. Science 2008; 320:1074–7
Hara R, Wan K, Wakamatsu H, Aida R, Moriya T, Akiyama M, Shibata S: Restricted feeding entrains liver clock without participation of the suprachiasmatic nucleus. Genes Cells 2001; 6:269–78
Son GH, Chung S, Choe HK, Kim HD, Baik SM, Lee H, Lee HW, Choi S, Sun W, Kim H, Cho S, Lee KH, Kim K: Adrenal peripheral clock controls the autonomous circadian rhythm of glucocorticoid by causing rhythmic steroid production. Proc Natl Acad Sci U S A 2008; 105:20970–5
Cutolo M, Seriolo B, Craviotto C, Pizzorni C, Sulli A: Circadian rhythms in RA. Ann Rheum Dis 2003; 62:593–6
Rigas B, Torosis J, McDougall CJ, Vener KJ, Spiro HM: The circadian rhythm of biliary colic. J Clin Gastroenterol 1990; 12:409–14
Solomon GD: Circadian rhythms and migraine. Cleve Clin J Med 1992; 59:326–9
Christina A, Merlin N, Vijaya C, Jayaprakash S, Murugesh N: Daily rhythm of nociception in rats. J Circadian Rhythms 2004; 2:2
Crockett RS, Bornschein RL, Smith RP: Diurnal variation in response to thermal stimulation: Mouse-hotplate test. Physiol Behav 1977; 18:193–6
Cao YQ, Mantyh PW, Carlson EJ, Gillespie AM, Epstein CJ, Basbaum AI: Primary afferent tachykinins are required to experience moderate to intense pain. Nature 1998; 392:390–4
DeVane CL: Substance P: A new era, a new role. Pharmacotherapy 2001; 21:1061–9
Park TJ, Lu Y, Jüttner R, Smith ES, Hu J, Brand A, Wetzel C, Milenkovic N, Erdmann B, Heppenstall PA, Laurito CE, Wilson SP, Lewin GR: Selective inflammatory pain insensitivity in the African naked mole-rat (Heterocephalus glaber). PLoS Biol 2008; 6:e13
Zimmer A, Zimmer AM, Baffi J, Usdin T, Reynolds K, König M, Palkovits M, Mezey E: Hypoalgesia in mice with a targeted deletion of the tachykinin 1 gene. Proc Natl Acad Sci U S A 1998; 95:2630–5
Perissin L, Boccalon S, Scaggiante B, Petrelli L, Ortolani F, Porro CA: Diurnal changes of tonic nociceptive responses in mice: Evidence for a proalgesic role of melatonin. Pain 2004; 110:250–8
Perissin L, Facchin P, Porro CA: Diurnal variations in tonic pain reactions in mice. Life Sci 2000; 67:1477–88
Kerdelhue B, Palkovits M, Karteszi M, Reinberg A: Circadian variations in substance P, luliberin (LH-RH) and thyroliberin (TRH) contents in hypothalamic and extrahypothalamic brain nuclei of adult male rats. Brain Res 1981; 206:405–13
Shigeyoshi Y, Maebayashi Y, Okamura H: Co-localization of preprosomatostatin mRNA and preprotachykinin A mRNA in neurons of the rat suprachiasmatic nucleus. Brain Res Mol Brain Res 1997; 48:159–63
Weber M, Lauterburg T, Tobler I, Burgunder JM: Circadian patterns of neurotransmitter related gene expression in motor regions of the rat brain. Neurosci Lett 2004; 358:17–20
Labrecque G, Vanier MC: Biological rhythms in pain and in the effects of opioid analgesics. Pharmacol Ther 1995; 68:129–47
Kim YI, Kim SH, Kim DY, Lee HW, Shin HC, Chung JM, Han HC, Na HS, Hong SK: Electrophysiological evidence for the role of substance P in retinohypothalamic transmission in the rat. Neurosci Lett 1999; 274:99–102
Kim DY, Kang HC, Shin HC, Lee KJ, Yoon YW, Han HC, Na HS, Hong SK, Kim YI: Substance p plays a critical role in photic resetting of the circadian pacemaker in the rat hypothalamus. J Neurosci 2001; 21:4026–31
Li R, Yue J, Zhang Y, Zhou L, Hao W, Yuan J, Qiang B, Ding JM, Peng X, Cao JM: CLOCK/BMAL1 regulates human nocturnin transcription through binding to the E-box of nocturnin promoter. Mol Cell Biochem 2008; 317:169–77
Chen T, Hui R, Wang XL, Zhang T, Dong YX, Li YQ: Origins of endomorphin-immunoreactive fibers and terminals in different columns of the periaqueductal gray in the rat. J Comp Neurol 2008; 509:72–87
Ranson RN, Priestley DJ, Santer RM, Watson AH: Changes in the substance P-containing innervation of the lumbosacral spinal cord in male Wistar rats as a consequence of ageing. Brain Res 2005; 1036:139–44
Nelson W, Tong YL, Lee JK, Halberg F: Methods for cosinor-rhythmometry. Chronobiologia 1979; 6:305–23
Stolwijk AM, Straatman H, Zielhuis GA: Studying seasonality by using sine and cosine functions in regression analysis. J Epidemiol Community Health 1999; 53:235–8
Liu S, Cai Y, Sothern RB, Guan Y, Chan P: Chronobiological analysis of circadian patterns in transcription of seven key clock genes in six peripheral tissues in mice. Chronobiol Int 2007; 24:793–820
Yamamoto T, Nakahata Y, Soma H, Akashi M, Mamine T, Takumi T: Transcriptional oscillation of canonical clock genes in mouse peripheral tissues. BMC Mol Biol 2004; 5:18
Tan PH, Yang LC, Shih HC, Lan KC, Cheng JT: Gene knockdown with intrathecal siRNA of NMDA receptor NR2B subunit reduces formalin-induced nociception in the rat. Gene Ther 2005; 12:59–66
Watanabe M, Mishina M, Inoue Y: Distinct spatiotemporal distributions of the N-methyl-D-aspartate receptor channel subunit mRNAs in the mouse cervical cord. J Comp Neurol 1994; 345:314–9
Takahashi T, Otsuka M: Regional distribution of substance P in the spinal cord and nerve roots of the cat and the effect of dorsal root section. Brain Res 1975; 87:1–11
Tigges M, Marquez-Lago TT, Stelling J, Fussenegger M: A tunable synthetic mammalian oscillator. Nature 2009; 457:309–12
Vrang N, Mikkelsen JD, Larsen PJ: Direct link from the suprachiasmatic nucleus to hypothalamic neurons projecting to the spinal cord: A combined tracing study using cholera toxin subunit B and Phaseolus vulgaris-leucoagglutinin. Brain Res Bull 1997; 44:671–80
Day M: Sympathetic blocks: The evidence. Pain Pract 2008; 8:98–109
Schlereth T, Birklein F: The sympathetic nervous system and pain. Neuromolecular Med 2008; 10:141–7
Iversen L: Substance P equals pain substance? Nature 1998; 392:334–5
Radhakrishnan V, Henry JL: Electrophysiology of neuropeptides in the sensory spinal cord. Prog Brain Res 1995; 104:175–95
Henry JL: Substance P and inflammatory pain: Potential of substance P antagonists as analgesics. Agents Actions Suppl 1993; 41:75–87
Radhakrishnan V, Henry JL: Electrophysiology of neuropeptides in the sensory spinal cord. Prog Brain Res 1995; 104:175–95
Kantner RM, Kirby ML, Goldstein BD: Increase in substance P in the dorsal horn during a chemogenic nociceptive stimulus. Brain Res 1985; 338:196–9
Zhang RX, Mi ZP, Qiao JT: Changes of spinal substance P, calcitonin gene-related peptide, somatostatin, Met-enkephalin and neurotensin in rats in response to formalin-induced pain. Regul Pept 1994; 51:25–32
McCarson KE, Krause JE: The formalin-induced expression of tachykinin peptide and neurokinin receptor messenger RNAs in rat sensory ganglia and spinal cord is modulated by opiate preadministration. Neuroscience 1995; 64:729–39
McCarson KE, Goldstein BD: Time course of the alteration in dorsal horn substance P levels following formalin: Blockade by naloxone. Pain 1990; 41:95–100
Guan JS, Xu ZZ, Gao H, He SQ, Ma GQ, Sun T, Wang LH, Zhang ZN, Lena I, Kitchen I, Elde R, Zimmer A, He C, Pei G, Bao L, Zhang X: Interaction with vesicle luminal protachykinin regulates surface expression of delta-opioid receptors and opioid analgesia. Cell 2005; 122:619–31
McCarson KE, Goldstein BD: Release of substance P into the superficial dorsal horn following nociceptive activation of the hindpaw of the rat. Brain Res 1991; 568:109–15
De Koninck Y, Henry JL: Substance P-mediated slow excitatory postsynaptic potential elicited in dorsal horn neurons in vivo  by noxious stimulation. Proc Natl Acad Sci U S A 1991; 88:11344–8
Yashpal K, Radhakrishnan V, Coderre TJ, Henry JL: CP-96,345, but not its stereoisomer, CP-96,344, blocks the nociceptive responses to intrathecally administered substance P and to noxious thermal and chemical stimuli in the rat. Neuroscience 1993; 52:1039–47
Le Bars D, Gozariu M, Cadden SW: Animal models of nociception. Pharmacol Rev 2001; 53:597–652
Yeomans DC, Proudfit HK: Characterization of the foot withdrawal response to noxious radiant heat in the rat. Pain 1994; 59:85–94
Yeomans DC, Proudfit HK: Nociceptive responses to high and low rates of noxious cutaneous heating are mediated by different nociceptors in the rat: Electrophysiological evidence. Pain 1996; 68:141–50
Yeomans DC, Pirec V, Proudfit HK: Nociceptive responses to high and low rates of noxious cutaneous heating are mediated by different nociceptors in the rat: Behavioral evidence. Pain 1996; 68:133–40
Bennett GJ, Xie YK: A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man. Pain 1988; 33:87–107
Kim SH, Chung JM: An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 1992; 50:355–63
Seltzer Z, Dubner R, Shir Y: A novel behavioral model of neuropathic pain disorders produced in rats by partial sciatic nerve injury. Pain 1990; 43:205–18