Low-intensity transcranial focused ultrasound suppresses pain by modulating pain-processing brain circuits

Abstract

There is an urgent and unmet clinical need to develop nonpharmacological interventions for chronic pain management because of the critical side effects of opioids. Low-intensity transcranial focused ultrasound (tFUS) is an emerging noninvasive neuromodulation technology with high spatial specificity and deep brain penetration. Here, we developed a tightly focused 128-element ultrasound transducer to specifically target small mouse brains using dynamic focus steering. We demonstrate that tFUS stimulation at pain-processing brain circuits can significantly alter pain-associated behaviors in mouse models in vivo. Our findings indicate that a single-session focused ultrasound stimulation to the primary somatosensory cortex (S1) significantly attenuates heat pain sensitivity in wild-type mice and modulates heat and mechanical hyperalgesia in a humanized mouse model of chronic pain in sickle cell disease. Results further revealed a sustained behavioral change associated with heat hypersensitivity by targeting deeper cortical structures (eg, insula) and multisession focused ultrasound stimulation to S1 and insula. Analyses of brain electrical rhythms through electroencephalography demonstrated a significant change in noxious heat hypersensitivity-related and chronic hyperalgesia-associated neural signals after focused ultrasound treatment. Validation of efficacy was carried out through control experiments, tuning ultrasound parameters, adjusting interexperiment intervals, and investigating effects on age, sex, and genotype in a head-fixed awake model. Importantly, tFUS was found to be safe, causing no adverse effects on motor function or the brain's neuropathology. In conclusion, the validated proof-of-principle experimental evidence demonstrates the translational potential of novel focused ultrasound neuromodulation for next-generation pain treatment without adverse effects.

Graphical abstract

Figure 1.

Treating chronic and stimulus-evoked hyperalgesia via tFUS stimulation to specific pain processing brain circuits. The conceptual diagram illustrates the hypothesis of treating chronic and stimulus-evoked hyperalgesia through tFUS stimulation in a humanized mouse model of chronic pain in SCD (HbSS-BERK). The modulation of pain-associated behaviors, such as amelioration or deterioration, is proposed to be achieved by selectively stimulating specific pain-processing brain circuits. This is facilitated using a 128-element highly focused random array ultrasound transducer capable of a dynamic focus steering in HbSS-BERK mice. The diagram captures the idea of targeted neuromodulation to influence pain responses in the context of SCD.

Figure 2.

Experimental design. (A) Schematic of behavioral experimental setup comprises various behavioral tests, tFUS stimulation with multiple control groups, and video assessment with subsequent data analysis. Nocifensive reactions to heat, cold, and mechanical stimuli, as well as motor performance, were recorded and evaluated to investigate the effects of tFUS stimulation at different brain locations, including S1HL, insula, and thalamus. (B) Study design for EEG measurements measurements were designed to examine both intrinsic brain activity and heat stimulus–evoked brain activity along with tFUS stimulation. (C) Control experiments were conducted to determine the specificity of the modulatory effect to tFUS stimulation at a particular brain circuit, the effects of negative control (no tFUS stimulation, maintaining experimental procedures) and sham treatment (applying tFUS to a control brain structure near the targeted brain location) were examined. (D) tFUS parameters involved customizing the 128-element random array transducer to steer ultrasound focus. Targeted brain structures were subjected to tFUS with specific parameters, including a fundamental frequency of 1.5 MHz, TBD of 200 μs, PRF of 40 Hz and 3 kHz, ultrasound duration (UD) of 100 and 400 milliseconds, intersonication interval (ISoI) of 2 and 4 second, and total sonication time of 10 minutes, 20 minutes, and 1 hour.

Figure 3.

Short-term modulation of pain-related behaviors in wild-type mice and female HbSS-BERK mice. (A) Experimental design for the single-session treatment with behavioral study. (B-C) Short-term inhibition of heat pain sensitivity in wild-type mice. The impact of tFUS on heat pain-associated behaviors was assessed in both (B) male (n = 8) and (C) female (n = 8) wild-type mice. Single-session tFUS with a PRF of 40 Hz at left S1HL resulted in a significant increase in latency difference between contralateral (right) and ipsilateral (left) hind paw compared with baseline and sham-treated mice. ns, not significant, ∗∗P < .01 using t test with Wilcoxon matched–pairs signed rank test; ^#P < .05 using t test with Mann-Whitney test. (D-F) Short-term modulation of thermal and mechanical hyperalgesia in female HbSS-BERK mice. Quantification of averaged Δ hPWL from hot-plate test, Δ hPWF from cold-plate test, and Δ hPWT from von Frey test was conducted before and after tFUS with control groups (negative control and sham treatment) in HbSS-BERK mice. (D) Single-session tFUS with a PRF of 40 Hz at left S1HL (n = 10) led to a significant increase in the averaged Δ hPWL from hot-plate test compared with baseline, sham treatment (n = 10), and negative control (n = 10). Control groups did not show apparent treatment effect compared with pretreatment baseline. A negative value of Δ hPWL was observed with a PRF of 3 kHz at left S1HL (n = 10), which was significantly different from the baseline and response with the PRF of 40 Hz. (E) Single-session tFUS with PRF of 40 Hz at left S1HL (n = 10) and insula (n = 10) resulted in a significant change in the Δ hPWF from cold-plate test at 8 minutes after treatment compared to the sham treatment and baseline. (F) tFUS with PRF of 40 Hz at left S1HL (n = 10) and insula (n = 10) led to a prominent elevation of Δ hPWT from von Frey test compared with sham-treated (n = 10) and negative control mice (n = 10). A lower withdrawal threshold of the contralateral (right) hind paw was observed with tFUS at PRF of 3 kHz at left S1HL (n = 10) compared with PRF of 40 Hz at S1HL and insula. ∗P < .05, ∗∗P < .01 using t test with Wilcoxon matched–pairs signed rank test; ^#P < .05, ^##P < .01, ^###P < .001, ^####P < .0001 using t test with Mann-Whitney test. ns, not significant.

Figure 3.

Short-term modulation of pain-related behaviors in wild-type mice and female HbSS-BERK mice. (A) Experimental design for the single-session treatment with behavioral study. (B-C) Short-term inhibition of heat pain sensitivity in wild-type mice. The impact of tFUS on heat pain-associated behaviors was assessed in both (B) male (n = 8) and (C) female (n = 8) wild-type mice. Single-session tFUS with a PRF of 40 Hz at left S1HL resulted in a significant increase in latency difference between contralateral (right) and ipsilateral (left) hind paw compared with baseline and sham-treated mice. ns, not significant, ∗∗P < .01 using t test with Wilcoxon matched–pairs signed rank test; ^#P < .05 using t test with Mann-Whitney test. (D-F) Short-term modulation of thermal and mechanical hyperalgesia in female HbSS-BERK mice. Quantification of averaged Δ hPWL from hot-plate test, Δ hPWF from cold-plate test, and Δ hPWT from von Frey test was conducted before and after tFUS with control groups (negative control and sham treatment) in HbSS-BERK mice. (D) Single-session tFUS with a PRF of 40 Hz at left S1HL (n = 10) led to a significant increase in the averaged Δ hPWL from hot-plate test compared with baseline, sham treatment (n = 10), and negative control (n = 10). Control groups did not show apparent treatment effect compared with pretreatment baseline. A negative value of Δ hPWL was observed with a PRF of 3 kHz at left S1HL (n = 10), which was significantly different from the baseline and response with the PRF of 40 Hz. (E) Single-session tFUS with PRF of 40 Hz at left S1HL (n = 10) and insula (n = 10) resulted in a significant change in the Δ hPWF from cold-plate test at 8 minutes after treatment compared to the sham treatment and baseline. (F) tFUS with PRF of 40 Hz at left S1HL (n = 10) and insula (n = 10) led to a prominent elevation of Δ hPWT from von Frey test compared with sham-treated (n = 10) and negative control mice (n = 10). A lower withdrawal threshold of the contralateral (right) hind paw was observed with tFUS at PRF of 3 kHz at left S1HL (n = 10) compared with PRF of 40 Hz at S1HL and insula. ∗P < .05, ∗∗P < .01 using t test with Wilcoxon matched–pairs signed rank test; ^#P < .05, ^##P < .01, ^###P < .001, ^####P < .0001 using t test with Mann-Whitney test. ns, not significant.

Figure 4.

Sustained modulatory effect on heat hyperalgesia with the single- and multisession tFUS stimulation in female HbSS-BERK mice. (A) Experimental design for the single-session treatment with behavioral study. (B) The averaged Δ hPWL stimulated by single-session tFUS with a PRF of 40 Hz at left insula (n = 10) was significantly increased up to 30 minutes in HbSS-BERK mice compared with the pre-tFUS baseline, as well as values from sham (n = 10) and negative control groups (n = 10). Single-session tFUS with a PRF of 3 kHz (n = 10), sham treatment, and negative control did not result in remarkable changes in Δ hPWL compared with the baseline. ∗P < .05, ∗∗P < .01 using 1-way analysis of variance (ANOVA) with Friedman test; ^#P < .05, ^##P < .01, ^###P < .001 using 1-wayANOVA with Kruskal-Wallis test. (C) Targeting the thalamus with tFUS using a PRF of 40 Hz (n = 10) and 3 kHz (n = 10) did not result in a prominent modulatory effect compared with the prestimulation baseline, as well as sham-treated mice (n = 10) and negative control mice (n = 10), except for 2 measurements (right panel: ^#P = .0104, 8 minutes after tFUS with the PRF of 40 Hz and sham treatment; ^#P = .0277, 60 minutes after tFUS with the PRF of 40 Hz and negative control). ns, not significant, ^#P < .05 using 1-way ANOVA with Kruskal-Wallis test. (D) Sustained amelioration of heat hyperalgesia was examined in older female HbSS-BERK mice (10 to 15-month-old) by measuring latency to noxious heat stimuli compared with prestimulation baseline upon 14-day tFUS stimulation. (E) After multisession tFUS stimulation with a PRF of 40 Hz to S1HL (n = 11) or insula (n = 10) for 14 days, the averaged changes of Δ hPWL were significantly increased for at least 2 hours compared with the baseline values, as well as sham-treated behavioral change (n = 4). The measurement was conducted at 2 hours after tFUS or sham treatment, indicating that the effect of multisession tFUS stimulation is considered to persist for at least 2 hours. ns, not significant, ∗P < .05, ∗∗∗P < .001 using t test with Wilcoxon 1-tailed signed rank test; ^#P < .05, ^##P < .01 using t test with 1-tailed Mann-Whitney test. ns, not significant.

Figure 5.

Modulation of heat stimulus–evoked brain activity and chronic pain resting state in female HbSS-BERK mice. (A) The single- and multisession tFUS treatment conditions used in EEG study. (B) Chronic pain quantification in brain rhythms at rest, showing a broadband increase of oscillations from theta to beta frequencies in female HbSS-BERK mice (n = 8) compared with female wild-type mice (n = 8). ^#P < .05 using t test with Mann-Whitney test. (C) Heat pain quantification in brain oscillations from the negative control group. HTHPS but not LTHPS resulted in a significant decrease in theta, alpha, and beta power in the contralateral brain region of S1HL in HbSS-BERK mice (n = 8). Note the temperature categories: baseline (BL), low temperature (LT; 26.2 ± 0.2°C < T ≤ 36.2 ± 0.2°C), and high temperature (HT; 38.9 ± 0.2°C < T ≤ 46.4 ± 0.1°C). ∗P < .05 using t test with Wilcoxon signed rank test. (D) Differences (Δ) of oscillatory power during normalized HTHPS and averaged normalized LTHPS. Post-tFUS with a PRF of 40 Hz, but not 3 kHz, at S1HL induced a significant increase in Δ power from theta to beta frequencies (n = 8), contrasting the heat pain–related EEG pattern. ^#P < .05, ^##P < .01, ^###P < .001 using t test with Mann-Whitney test. (E) Multisession tFUS at S1HL induced a significant decrease in power from theta to beta frequencies (n = 5), showing a remarkable difference from the power of untreated HbSS-BERK mice while being similar to the power of wild-type mice. ^#P < .05, ^##P < .01 using t test with Mann-Whitney test. (F) At 80 seconds after heat stimulus at hind paw, power in alpha and beta frequencies was significantly suppressed by multisession tFUS with a PRF of 40 Hz but not by single-session tFUS with a PRF of 40 Hz (n = 8 for single-session tFUS and n = 4 for multisession tFUS). ^#P < .05 (t test with Mann-Whitney test). ns, not significant.

Figure 6.

tFUS safety evaluation in female HbSS-BERK mice. (A) The single- and multisession tFUS treatment conditions used in safety study. (B-C) The effect of tFUS on the motor coordination and balance was investigated in rotarod testing. After 60 minutes of single-session tFUS stimulation with a PRF of 40 Hz to S1HL (n = 11) or thalamus (n = 11) and after 60 minutes of multisession tFUS with a PRF of 40 Hz to S1HL (n = 4) or insula (n = 6) for 14 days with 1 hour per day, no significant alteration was observed in the averaged latency to fall compared with prestimulation baseline as well as control groups. In addition, there was no significant difference of control groups from their baseline values using t test with Wilcoxon matched–pairs signed rank test and 1-way ANOVA with Kruskal-Wallis test. (D) Double-blind histological analysis using hematoxylin and eosin (H&E) and TUNEL stains showed that multisession tFUS stimulation to the left S1HL or insula had no adverse impact on tissue structural integrity and nuclear morphology within the treated brain region compared with contralateral brain regions and sham treatment. Scale bars, 50 μm. ns, not significant; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.

Figure 6.

tFUS safety evaluation in female HbSS-BERK mice. (A) The single- and multisession tFUS treatment conditions used in safety study. (B-C) The effect of tFUS on the motor coordination and balance was investigated in rotarod testing. After 60 minutes of single-session tFUS stimulation with a PRF of 40 Hz to S1HL (n = 11) or thalamus (n = 11) and after 60 minutes of multisession tFUS with a PRF of 40 Hz to S1HL (n = 4) or insula (n = 6) for 14 days with 1 hour per day, no significant alteration was observed in the averaged latency to fall compared with prestimulation baseline as well as control groups. In addition, there was no significant difference of control groups from their baseline values using t test with Wilcoxon matched–pairs signed rank test and 1-way ANOVA with Kruskal-Wallis test. (D) Double-blind histological analysis using hematoxylin and eosin (H&E) and TUNEL stains showed that multisession tFUS stimulation to the left S1HL or insula had no adverse impact on tissue structural integrity and nuclear morphology within the treated brain region compared with contralateral brain regions and sham treatment. Scale bars, 50 μm. ns, not significant; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.