Parabrachial Calca neurons drive nociplasticity

Abstract

Pain that persists beyond the time required for tissue healing and pain that arises in the absence of tissue injury, collectively referred to as nociplastic pain, are poorly understood phenomena mediated by plasticity within the central nervous system. The parabrachial nucleus (PBN) is a hub that relays aversive sensory information and appears to play a role in nociplasticity. Here, by preventing PBN Calca neurons from releasing neurotransmitters, we demonstrate that activation of Calca neurons is necessary for the manifestation and maintenance of chronic pain. Additionally, by directly stimulating Calca neurons, we demonstrate that Calca neuron activity is sufficient to drive nociplasticity. Aversive stimuli of multiple sensory modalities, such as exposure to nitroglycerin, cisplatin, or lithium chloride, can drive nociplasticity in a Calca-neuron-dependent manner. Aversive events drive nociplasticity in Calca neurons in the form of increased activity and excitability; however, neuroplasticity also appears to occur in downstream circuitry.

Keywords: CP: Cell biology; CP: Neuroscience; Calca; allodynia; calcitonin gene-related peptide (CGRP); calcium imaging; cisplatin; mechanical sensitivity (Von Frey assay); neuroplasticity; nitroglycerin.

Figure 1.. Neuropathic pain activates parabrachial Calca neurons and drives persistent changes in their excitability

(A) Representative images from RNAscope in situ hybridization on tissue slices collected 3 or 30 days post sham or pSNL surgery with probes targeting Calca (green), Fos (red), and Cck (white). Scale bars, 100 μm. A dotted line marks the superior cerebellar peduncle (SCP); anterior-posterior bregma level = −5.1. (B) pSNL increased Fos mRNA in Calca neurons in the middle, but not the rostral, PBN at 3 dps. (C) pSNL increased Fos mRNA in non-Calca neurons in the middle, but not the rostral, PBN at 3 dps. (D) pSNL did not change the number of Calca-positive neurons in the rostral or middle PBN at 3 dps. (E) pSNL did not change the number of Cck-positive cells in the rostral or middle PBN at 3 dps. (B–E) Rostral sham, n = 5; middle sham, n = 14; rostral pSNL, n = 11; middle pSNL, n = 10. (F) pSNL did not drive an increase in the expression of Fos mRNA in Calca neurons at 30 dps. (G) pSNL did not drive an increase in the expression of Fos mRNA in non-Calca neurons at 30 dps. (H) pSNL did not change the number of Calca-positive neurons in the rostral or middle PBN at 30 dps. (I) pSNL decreased the number of Cck-positive neurons in the rostral, but not the middle, PBN at 30 dps. (F–I) Rostral sham, n = 5; middle sham, n = 5; rostral pSNL, n = 5; and middle pSNL, n = 7. (B–I) Significance was tested by ANOVA with multiple comparisons. *p < 0.05, ***p < 0.001. Error bars, SEM. (J) Representative traces showing regularly firing Calca neurons 3 and 30 days post sham or pSNL surgery. (K) pSNL 3 days prior to electrophysiology did not change the number of spikes elicited by current injection in the regularly firing population. Sham, n = 3 animals, 7 neurons; pSNL treated, n = 3 animals, 13 neurons. (L) pSNL 30 days prior to electrophysiology increased the number of spikes elicited by current injection in the regularly firing population. Sham, n = 4 animals, 9 neurons; pSNL treated, n = 4 animals, 9 neurons. (K and L) 3 dps sham and pSNL biological replicates, n = 3. 3 dps sham technical replicates, n = 7 and pSNL technical replicates, n = 13. 30 dps sham and pSNL biological replicates, n = 4. 30 dps sham technical replicates, n = 9 and pSNL technical replicates, n = 9. Significance was measured by ANOVA. ****p < 0.0001. Error bars, SEM. (M) Representative traces showing late-firing Calca neurons 3 and 30 days post sham or pSNL surgery. (N) pSNL 3 days prior to electrophysiology did not change the number of spikes elicited by current injection in the late-firing population. Sham, n = 3 animals, 15 neurons; pSNL treated, n = 3 animals, 13 neurons. (O) pSNL 30 days prior to electrophysiology increased the number of spikes elicited by current injection in the late-firing population. Sham, n = 4 animals, 20 neurons; pSNL treated, n = 4 animals, 9 neurons. (N and O) 3 dps sham and pSNL biological replicates, n = 3. 3 dps sham technical replicates, n = 15 and pSNL technical replicates, n = 13. 30 dps sham and pSNL biological replicates, n = 4. 30 dps sham technical replicates, n = 20 and pSNL technical replicates, n = 9. Significance was measured by ANOVA. ****p < 0.0001. Error bars, SEM.

Figure 2.. Parabrachial Calca neurons are necessary for the induction of neuropathic pain

(A) Bilateral injections of AAV1-SYN-DIO-YFP or AAVDJ-Ef1a-DIO-GFP:TeTx into the PBN of Calca^Cre/+ mice. Representative images show expression of YFP and TeTx. Scale bars, 100 μm. A dotted line marks the SCP; anterior-posterior bregma level = −5.1. (B) TeTx expression in PBN Calca neurons prevents the development of pSNL-driven allodynia, measured by von Frey assay. Calca^Cre/+:YFP, n = 5 and Calca^Cre/+: TeTx, n = 5. (C) Bilateral injections of AAV1-SYN-DIO-mCherry or AAV1-CBA-DIO-hM4Di:mCherry into the PBN of Calca^Cre/+ mice. Representative images show expression of mCherry and hM4Di:mCherry. Scale bars, 100 μm. A dotted line marks the SCP; anterior-posterior bregma level = −5.1. (D) hM4Di/CNO inhibition of PBN Calca neurons ameliorates pSNL-driven allodynia at 3 dps, measured by von Frey assay. Calca^Cre/+:mCherry, n = 5 and Calca^Cre/+: hM4Di, n = 5. (E) hM4Di/CNO inhibition of PBN Calca neurons did not ameliorate pSNL-driven allodynia at 30 dps, measured by von Frey assay. Calca^Cre/+:mCherry, n = 5 and Calca^Cre/+:hM4Di, n = 5. (F) Bilateral injections of AAV1-SYN-DIO-YFP or AAVDJ-Ef1a-DIO-GFP:TeTx into the PBN of Calca^Cre/+ mice 14 days after pSNL ameliorated established allodynia, measured by von Frey assay. Calca^Cre/+:YFP, n = 6 and Calca^Cre/+:TeTx, n = 4. (G) pSNL produced allodynia in Calca-null mice, measured by von Frey assay. (B and D–G) Significance was tested by ANOVA with multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Error bars, SEM.

Figure 3.. Activation of parabrachial Calca neurons is sufficient to drive nociplasticity

(A) Bilateral injections of AAV1-Ef1a-DIO-mCherry or AAV1-hSyn-DIO-hM3Dq:mCherry into the PBN of Calca^cre/+ mice. Representative images show expression of mCherry and hM3Dq. Scale bars, 100 μm. A dotted line marks the SCP; anterior-posterior bregma level = −5.1. (B) 3 days of CNO injection (1 mg/kg, i.p.) resulted in persistent allodynia, measured by von Frey assay. Calca^cre/+:mCherry, n = 5 and Calca^cre/+:hM3Dq, n = 7. (C) Behavior timeline for von Frey (VF), Hargreave’s (HG), and hot plate (HP) assays before and after 5 consecutive days of CNO injection. Calca^Cre/+:mCherry, n = 6 and Calca^cre/+:hM3Dq, n = 6. (D) 5 days of CNO injection decreased the paw withdrawal threshold, measured by VF assay, which persisted after the last CNO injection. (E) 5 days of CNO injection decreased the paw withdrawal latency, measured by HG assay, which persisted after the last CNO injection. (F) 5 days of CNO injection increased nocifensive behaviors, measured by HP assay, which persisted after the last CNO injection. (G) 3 days of CNO injection (1 mg/kg, i.p.) resulted in persistent allodynia in Calca-null mice, measured by VF assay. Calca^Cre/Cre:mCherry, n = 6 and Calca^Cre/Cre:hM3Dq, n = 6. (B and D–G) Significance was tested by ANOVA with multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Error bars, SEM. (H) Bilateral injections of AAV1-SYN-DIO-hM3Dq:mCherry into the PBN of Calca^cre/+ mice followed by 3 days of CNO or saline injection and 48 h of no stimulation prior to electrophysiology. (I) Representative traces showing regularly and late-firing Calca neurons. (J) 3 days of CNO injection (i.p.) prior to electrophysiology resulted in an increase in the number of spikes elicited by current injection in the regularly firing population. Saline treated, n = 3 animals, 14 neurons; CNO treated, n = 3 animals, 11 neurons. (K) 3 days of CNO injection (i.p.) prior to electrophysiology resulted in an increase in the number of spikes elicited by current injection in the late-firing population. Saline treated, n = 3 animals, 5 neurons; CNO treated, n = 3 animals, 11 neurons. (J and K) Significance was measured by ANOVA. ***p < 0.001, ****p < 0.0001. Error bars, SEM.

Figure 4.. Nociplastic effect scales with the duration of Calca neuron activation

(A) Bilateral injections of AAV1-Ef1a-DIO-mCherry or AAV1-hSyn-DIO-hM3Dq:mCherry into the PBN of Calca^cre/+ mice. Representative images show expression of mCherry and hM3Dq. Scale bars, 100 μm. A dotted line marks the SCP; anterior-posterior bregma level = −5.1. (B) 1 day of CNO injection (1 mg/kg, i.p.) produced allodynia, measured by VF assay. Calca^cre/+:mCherry, n = 4 and Calca^cre/+:hM3Dq, n = 7. (C) Bilateral injections of AAV1-Ef1a-DIO-mCherry or AAV1-Ef1a-DIO-ChR2:mCherry into the PBN of Calca^cre/+ mice. Representative images show expression of mCherry and ChR2. Scale bars, 100 μm. A dotted line marks the SCP; anterior-posterior bregma level = −5.1. (D) 20 min of 473-nm photostimulation (20 Hz, 2 s on, 2 s off) resulted in allodynia, measured by VF assay. Calca^cre/+:mCherry, n = 6 and Calca^cre/+:ChR2, n = 7. (E) 5 min of 473-nm photostimulation (20 Hz, 2 s on, 2 s off) resulted in allodynia, measured by VF assay. Calca^cre/+:mCherry, n = 4 and Calca^cre/+:ChR2, n = 5. (F) 5 days of 5-min, 473-nm photostimulation (20 Hz, 2 s on, 2 s off) resulted in allodynia, measured by VF assay. Calca^cre/+:mCherry, n = 4 and Calca^cre/+:ChR2, n = 4. (B and D–F) Significance was tested by ANOVA with multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Error bars, SEM.

Figure 5.. Chronic exposure to all aversive stimuli tested drives nociplasticity regardless of sensory modality

(A) 5 days of NTG exposure (10 mg/kg, i.p.) produced allodynia, measured by VF assay. Vehicle, n = 4 and NTG, n = 4. (B) 3 days of LiCl exposure (0.2 M, 15 mL/kg, i.p.) produced allodynia, measured by VF assay. Vehicle, n = 6 and LiCl, n = 8. (C) 3 days of robobug chase (10 min) produced allodynia, measured by VF assay. Vehicle, n = 5 and robobug, n = 5. (D) Bilateral injections of AAV1-hSYN-DIO-YFP or AAVDJ-Ef1a-DIO-GFP:TeTx into the PBN of Calca^Cre/+ mice. Representative images show expression of YFP and TeTx. Scale bars, 100 μm. A dotted line marks the SCP; anterior-posterior bregma level = −5.1. (E) TeTx expression in PBN Calca neurons prevented the development of NTG-driven allodynia, measured by VF assay. Calca^Cre/+:YFP, n = 6 and Calca^Cre/+:TeTx, n = 5. (F) TeTx expression in PBN Calca neurons prevented the development of cisplatin-driven allodynia, measured by VF assay. Calca^Cre/+:YFP, n = 6 and Calca^Cre/+:TeTx, n = 6. (A–C, E, and F) Significance was tested by ANOVA with multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Error bars, SEM.

Figure 6.. Calca neuron activity during the development of mechanical allodynia

(A) Schematic of Ca²⁺ imaging during NTG-induced mechanical allodynia. (B) Multiday tracking of Calca neuron fluorescence activity. Top: field of view (FOV) of a representative animal for 4 days. Bottom: representative neural activities throughout 4 days of 3 neurons marked by arrows in FOVs. Scale bar, 50 μm. (C) NTG injection increased unstimulated neuronal activity. Elevated neuronal activity persisted 24 and 48 h post injection. Individual neurons are aligned across days in the heatmap. (D) Average calcium transient area under the curve increased following NTG injection. This increase in fluorescence activity remained elevated 24 and 48 h post injection. (E) NTG injection increased the number of times of 8 applications that mice responded to a 0.4-g VF filament. Error bars indicate mean ± SEM. (F) Representative traces of filament-evoked neural activities. (G) The number of Calca neurons responsive to application of a 0.4-g VF filament increased after NTG injection. The increase in responsive neurons persisted 24 and 48 h after NTG injection. (H) The percentage of Calca neurons responsive to application of a 0.4-g VF filament increased from 24% after vehicle injection to 49.5% after NTG injection. The percent of 0.4-g VF filament-responsive neurons remained elevated, at 40.5%, 24 and 48 h after NTG injection. (I) The majority of neurons unresponsive to 0.4-g VF filament application following vehicle injection (i.p.) became responsive to the 0.4-g filament following NTG injection (i.p.). (J) About half of the neurons responsive to 0.4-g VF filament application following vehicle injection (i.p.) became unresponsive to the 0.4-g filament following NTG injection (i.p.). (A–J) n = 4 animals, 79 neurons. Significance was tested by ANOVA with multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Error bars, SEM.

Figure 7.. Intrathecal NPY does not reverse hM3Dq/CNO-driven allodynia

(A) pSNL or bilateral injections of AAV1-Ef1a-DIO-mCherry or AAV1-hSyn-DIO-hM3Dq:mCherry into the PBN of Calca^cre/+ mice. Representative images show expression of mCherry and hM3Dq. Scale bars, 100 μm. A dotted line marks the SCP; anterior-posterior bregma level = −5.1. (B) Intrathecal (i.t.) injection of NPY^Leu,Pro into pSNL animals reversed pSNL-driven allodynia. (C) I.t. injection of NPY^Leu,Pro into Calca^Cre:hM3Dq animals treated with CNO for 3 days did not reverse allodynia. (D) I.t. injection of NPY^Leu,Pro into control Calca^Cre:mCherry animals treated with CNO for 3 days did not affect the paw withdrawal threshold. (B–D) Significance was tested by ANOVA with multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Error bars, SEM.