Fluolab【JACS】突破深层组织限制：可调控 131 nm 光谱范围的有机力致发光纳米平台
通讯作者： Wenliang Wang, Huiliang Wang
文章概要
引言
光遗传学技术虽然彻底改变了神经科学研究，但光线在生物组织中的散射和吸收限制了其进入深层脑区的能力，传统方法往往需要植入侵入性光纤。为了解决这一痛点，研究团队开发了一种基于超声驱动的有机力致发光（Mechanoluminescence, ML）纳米颗粒。这种技术利用超声波的深层渗透能力，在组织内部原位激发光线，从而实现非侵入性的精准神经调制。目前已有的力致发光系统大多存在发光颜色单一、强度低且对超声响应机制理解不足的问题。本研究通过引入能量转移机制和系统性的声敏剂筛选，成功克服了这些挑战，为深层组织光遗传学提供了全新的工具箱。
Scheme 1. Schematic Illustration of the Design of a Multicolor Mechanoluminescent Nanoparticle System (Exemplified by L012-RhB NPs) and Its Application in Sono-Optogenetic Activation of Neurons Expressing ChRmine,
主要实验及结论
研究人员首先构建了一个基于 FRET（福斯特共振能量转移） 机制的多色发光平台。他们以蓝光发射的 L012 纳米颗粒作为能量供体，通过单步或级联能量转移，将多种荧光受体分子嵌入体系中。实验结果显示，通过精确调节供体与受体的摩尔比例，该平台实现了从蓝色（461 nm）到红色（592 nm）的连续波长调节，覆盖了约 131 nm 的光谱范围。这种灵活性使得该平台能够匹配目前主流的所有光遗传学执行器。在稳定性测试中，该体系表现出极佳的抗淬灭能力，其持续发光时间比已报道的同类系统延长了 2.5 到 5.4 倍。
Figure 1. Chemiluminescence (CL) characterization and optimization of multicolor L012-based nanoparticles. (a) Construction schematic of multicolor nanoparticles employing highly efficient single and cascade FRET, wherein L012 acts as donor and FL, EY, PB, or RhB serve as acceptors for achieving progressively longer wavelengths, created via BioRender.com. (b-c) Optical images and intensity quantification of L012 NPs and L012-FL NPs under H2O2 (mean ± s.e.m., n = 3 independent samples). (d) CL spectra of varying FL concentrations in L012-FL NPs. (e–g) Optical images of L012-EY (e), L012-PB (f), and L012-RhB (g) nanoparticles at different concentrations. (h) Concentration-dependent CL intensities of L012-EY, L012-PB and L012-RhB NPs (mean ± s.e.m., n = 3 independent samples). (i–k) CL spectra of L012-FL NPs incorporating increasing concentrations of EY (i), PB (j), and RhB (k), respectively. (l) CL spectra displaying five distinct emission colors. (m) Size distribution profiles of these nanoparticles via DLS. (n) Compositional and physicochemical characterization of the multicolor nanoparticle system (mean ± s.e.m., n = 3 independent samples).
Figure 2. Sono-ML characterization of multicolor L012-based nanoparticles. (a) Schematic representation of the multicolor Sono-ML emission mechanism for L012-based nanoparticles under FUS stimulation, created via BioRender.com. (b–f) The dynamic intensity profiles of L012, L012-FL, L012-EY, L012-PB and L012-RhB NPs emissions under intermittent ultrasound irradiation (1.5 MHz, 1.55 MPa, 100 ms on, 900 ms off). (g) True-color emission optical images of these nanoparticles under the same ultrasound irradiation. (h) Comparative ML intensities of all nanoparticles under standardized conditions (1.5 MHz, pulse 100 ms on, 900 ms off, 1 Hz, 1.55 MPa; mean ± s.e.m., n = 10 independent samples). (i, j) ML intensities of the multicolor nanoparticle platforms at various ultrasound frequencies (pulse 100 on 900 ms off for 1 Hz, 50 on 450 ms off for 2 Hz, 25 on 225 ms off for 4 Hz and 12.5 on 112.5 ms off for 8 Hz) and pulse intervals (pulse 100 on 900 ms off for 100 ms light emission, 300 on 700 ms off for 300 ms light emission, 500 on 500 ms off for 500 ms light emission and 900 on 100 ms off for 900 ms light emission) (Both trials i and j were performed at 1.5 MHz and 1.55 MPa; mean ± s.e.m., n = 10 independent samples). (k) ML intensities of all nanoparticles at increasing ultrasound peak pressures (mean ± s.e.m., n = 10 independent samples). (l) Effective ML durations for all nanoparticle variants under cyclic FUS stimulation (1.5 MHz, pulse 1 s on, 1 s off, 1.55 MPa). (m) Comparative ML duration analysis of Lipo@IR780/L012, HOF@L012, and β-CD@IR780/L012 (1.5 MHz, pulse 1 s on, 1 s off, 1.55 MPa; mean ± s.e.m., n = 3 independent samples).
为了进一步优化发光效率，团队深入探讨了超声触发活性氧（ROS）产生的底层逻辑。他们系统筛选了 15 种不同结构的有机声敏剂，涵盖了卟啉、花菁和氧杂蒽等类别。通过结合实验数据与理论模型，研究者发现声敏剂的电子能隙（S1-S0）与单线态氧的产生效率呈正相关，而 HOMO-LUMO 能隙则与羟基自由基的生成相关。基于这一发现，他们建立了性能预测模型，并为不同波长的发光需求匹配了最优组合，例如使用 PPa 优化蓝光强度，使用 EY 优化红光表现，发光强度较常规条件提升了约 2 倍。
Figure 3. Screening of sonosensitizers for ultrasound-induced ROS generation and theoretical correlation modeling. (a) The small-molecule organic sonosensitizers investigated for ROS generation are categorized into three structural classes: porphyrin, cyanine, and xanthene derivatives. (b) Reaction scheme illustrating the detection of 1O2 using DPBF, whose characteristic absorption peak at 420 nm decreases proportionally with increasing 1O2 concentration. (c) Reaction mechanism of •OH detection using MB, whose 680 nm absorption peak decreases upon reaction with •OH. (d) UV–vis spectra showing no significant change in DPBF absorption in the presence of ICG without ultrasound stimulation (1.5 MHz, 1.55 MPa, pulse 500 ms on, 500 ms off). (e) Time-dependent UV–vis spectra demonstrating DPBF decomposition and 1O2 generation by ICG under ultrasound stimulation. (f) Quantitative analysis of DPBF decomposition induced by ultrasound in the presence of ICG compared to controls (mean ± s.e.m., n = 3 per group). (g) UV–vis spectra indicating negligible MB decomposition in the absence of ultrasound. (h) Time-dependent UV–vis spectra demonstrating MB degradation by •OH produced from ICG under ultrasound stimulation. (i) Quantitative analysis of MB decomposition with and without ultrasound irradiation in the presence of ICG (mean ± s.e.m., n = 3 per group). (j) Comparative analysis of 1O2 generation efficiency for 15 sonosensitizers under identical ultrasound parameters. (k) Comparative analysis of •OH generation efficiency for 15 sonosensitizers under identical ultrasound parameters. (l) Schematic Jablonski diagram illustrating ROS generation mechanisms of sonosensitizers activated by ultrasound. (m) Correlation plot demonstrating the positive relationship between relative 1O2 generation efficiency and S1–S0 energy gap (mean ± s.e.m., n = 3 independent samples). (n) Correlation plot showing the negative relationship between relative •OH generation efficiency and HOMO–LUMO energy gap (mean ± s.e.m., n = 3 independent samples).
Figure 4. Comparative Sono-ML screening of 15 sonosensitizers. (a) Comparative blue Sono-ML intensities of L012 NPs paired with 15 sonosensitizers under FUS stimulation (1.5 MHz, 1.55 MPa, pulse 100 ms on, 900 ms off; mean ± s.e.m., n = 5 independent samples). (b) Comparative green Sono-ML intensities of L012-FL NPs combined with 15 sonosensitizers under identical ultrasound conditions (mean ± s.e.m., n = 5 independent samples). (c) Comparative red Sono-ML intensities of L012-RhB NPs in combination with 15 sonosensitizers under the same ultrasound parameters (mean ± s.e.m., n = 5 independent samples). (d) Representative Sono-ML intensity dynamics of L012 NPs combined with IR780 or PPa sonosensitizers under ultrasound stimulation (1.5 MHz, 1.55 MPa, pulse 100 ms on, 900 ms off). (e) Representative Sono-ML intensity dynamics of L012-FL NPs combined with IR780 or PPa sonosensitizers under the same ultrasound parameters. (f) Representative Sono-ML intensity dynamics of L012-RhB NPs combined with IR780 or EY sonosensitizers under identical ultrasound stimulation conditions. (g) Positive correlation analysis illustrating the relationship between relative 1O2 generation efficiency and blue Sono-ML intensity (mean ± s.e.m., n = 5 independent samples).
在功能性验证阶段，研究团队将该纳米平台应用于离体原代神经元的调制。实验选择了三种具有代表性的光敏蛋白：ChR2（蓝光激活）、eOPN3（绿光抑制）和 ChRmine（红光激活）。在微型化聚焦超声（FUS）的驱动下，纳米颗粒发射出的高强度脉冲光成功诱发了神经元的同步放电或有效抑制。统计数据显示，其放电诱导概率高达 97%，而抑制组的放电概率则显著降至 17%。整个过程展现了极高的时空分辨率和生物兼容性，证明了该系统在模拟生理电活动方面的强大潜力。
- Figure 5. In vitro sono-optogenetics using mechanoluminescent nanoparticles in primary neurons expressing ChR2, eOPN3, and ChRmine. (a) Schematic illustration of the sono-optogenetic setup on a MEA system, showing: (i) experimental overview; (ii) ChR2 activation by L012 NPs under FUS; (iii) eOPN3 activation by L012-FL NPs; and (iv) ChRmine activation by L012-RhB NPs, created via BioRender.com. (b) Overlaid spectra of the CL emission from L012 NPs and the action profile of ChR2, demonstrating spectral compatibility. (53) Below: representative Sono-ML trace used in MEA testing. (c) MEA recordings of ChR2-expressing neurons under four experimental conditions: FUS(−)/L012 NPs(−); FUS(+)/L012 NPs(−); FUS(−)/L012 NPs(+) and FUS(+)/L012 NPs(+). All recordings were conducted under standardized ultrasound conditions (1.5 MHz, 1.55 MPa, 100 ms on/900 ms off; stimulation from 10 to 40 s). (d) Quantitative analysis of MEA signal changes across conditions and spike probability of ChR2-expressing neurons (n = 3 per group, two-way ANOVA and multiple comparisons test). (e) Overlaid spectra of L012-FL NPs and eOPN3 opsin action spectrum, confirming spectral overlap. (54) Below: Sono-ML pattern used for the MEA experiment. (f) MEA recordings of eOPN3-expressing neurons under continuous electrical stimulation across four experimental conditions: FUS(−)/L012-FL NPs(−), FUS(+)/L012-FL NPs(−), FUS(−)/L012-FL NPs(+), and FUS(+)/L012-FL NPs(+). In the FUS(+)/L012-FL NPs(+) group, continuous Sono-ML irradiation was applied from 10 to 40 s. (g) Statistical analysis of MEA signal changes and spike probability of eOPN3-expressing neurons (n = 3 per group, two-way ANOVA and multiple comparisons test). (h) Overlaid spectra of L012-RhB NPs and ChRmine opsin. (19) Below: corresponding Sono-ML trace used in the MEA test. (i) MEA recordings of ChRmine-expressing neurons under the same four experimental conditions and ultrasound parameters as in (c). (j) Quantitative analysis of MEA signal variations and spike probability of ChRmine-expressing neurons across all tested groups (n = 3 per group, two-way ANOVA and multiple comparisons test). All plots show mean ± s.e.m. unless otherwise mentioned. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001; ns, not significant.
总结及展望
这项研究通过创新的分子设计和机制解析，成功开发了一套光谱可调、发光持久且高效的有机力致发光纳米平台。该平台不仅解决了深层组织光传递的难题，还通过声敏剂筛选法则为未来高性能力致发光材料的设计提供了理论指导。研究团队指出，这种非侵入性的声-光转换策略未来不仅限于神经科学领域，在生物成像、精准基因编辑以及临床治疗等深层组织应用场景中都具有极广阔的推广价值。接下来，团队将重点开展体内生物分布及长期安全性评估，推动该技术向临床转化迈进。