The precise stimulation and imaging of neuronal microstructures are crucial for comprehending the intricate dynamics of neural cells and uncovering the mechanisms underlying neural cell pathology. By achieving precise control and visualization at the subcellular level, this research endeavor holds immense significance in advancing our understanding of neural cell function and dysfunction. In recent years, optical stimulation has gained substantial recognition as a potent tool for investigating the regulation of neural systems, making notable contributions to neuroscience research. This innovative approach has brought significant advancements in our understanding of how neural systems function and interact, opening up new avenues for further exploration and discovery [1–3]. Infrared neural stimulation (INS) is a technique that involves directly applying near-infrared light to neural tissue. This process effectively modulates the firing of action potentials in neurons, resulting in the regulation of brain function [4–6]. INS operates through a photothermal mechanism where electromagnetic energy is converted into thermal energy. This change in temperature induces transmembrane capacitive currents within cells and activates thermosensitive ion channels. These processes collectively influence neuronal activity, unraveling the underlying mechanisms behind the modulatory effects of INS on neuronal activity [7,8]. As a result, optical stimulation has become a versatile and flexible method for modulating neurons. However, wide-field illumination exhibits a broad spectrum of light activation and lacks the ability to selectively activate specific neurons, thereby limiting its utility in studying intricate neural circuitry. In contrast, precise optical stimulation techniques facilitate targeted modulation of specific neurons while minimizing impact on neighboring cells, enabling a more comprehensive and nuanced exploration of neural circuits and their functional roles [9].

To achieve subcellular level precision in implementing the INS, it is necessary to have both control over near-infrared light and visualization of target sites. While various methods like fluorescence microscopy, confocal microscopy, and multiphoton microscopy have been utilized for visualizing cellular structures [10–12], simultaneous control of external fields is also required for precise stimulation of the targets. Microsphere-assisted techniques, emerging as a strong enhancement in conventional optical microscopy, offer the capabilities of real-time and high-resolution imaging [13–18]. More importantly, the presence of microspheres introduces unique optical effects including directional antennas [19–21], whispering gallery mode [22–24], and photonic nanojets (PNJs) [25–27] that enable localization and enhancement of light–matter interactions. Particularly, PNJs have been employed to generate optical trapping forces and achieve high-resolution imaging in cells [28–32], making them a promising candidate for precise stimulation and imaging of neural cells. In this study, we propose the utilization of photonic nanojets (PNJs) for optical stimulation and imaging of neural cells. By employing transparent resin microspheres as microlenses, incident light is effectively focused into subwavelength PNJs that precisely stimulate specific neural cells. Simultaneously, PNJs enhance fluorescence imaging of neural lysosomes, mitochondria, and actin filaments within the cells, enabling real-time monitoring and analysis of fluorescence signals from subcellular structures during the stimulation process.

To achieve a strong lensing effect and high biocompatibility, transparent resin microspheres (diameter: 10 μm) were used as microlenses in this study (Fig. 1). As schematically described, the incident light diverges and illuminates the neural cells over a considerably large range without the presence of microlenses [Fig. 1(a)]. However, by positioning the microlenses close to the neural cells [Fig. 1(b)], the incident light is efficiently focused onto the target area, specifically the neuronal synapse, enabling precise optical stimulation. The lensing effect of these microlenses was demonstrated using polystyrene (PS) nanoparticles suspended in solution through observing Tyndall scattering [33] [Fig. 1(c)], confirming that they effectively focus incident light into PNJs. Based on this lensing capability, our experiments were designed around a scanning optical tweezing (SOT) system integrated with an inverted optical microscope setup [Figs. 1(d)–1(f)] (see detailed description of the experimental setup in Appendix A). By generating PNJs using these microlenses, near-field information from samples can be adequately collected and form virtual images obtained via an optical microscope observation [Fig. 1(d)]. With the assistance of the SOT system that utilizes a 1064 nm Gaussian beam as the trapping light, the microlenses can be trapped and moved along a predesigned path of the scanning optical trap [Fig. 1(e)]. For example, when a circular path is set for the optical trap, the microlenses can be trapped and then move along orbit trajectories [Fig. 1(f)]. This approach enables targeted stimulation and imaging at desired locations.

The light focusing property of the microlenses was simulated using 2D finite-element-method (FEM) calculations (COMSOL Multiphysics 5.6). In the simulation, a plane wave with a wavelength of 1064 nm was used as the incident light, and the refractive indices of the microlens (diameter: 10 μm) and its surrounding medium (water) were set as 1.68 and 1.33, respectively. The simulated energy density distributions show that when a microlens is placed at the optical axis, it focuses the incident light into a PNJ with a full width at half maximum (FWHM) of 650 nm [Figs. 2(a) and 2(b)]. The FWHM was utilized to represent the spot size of the photonic nanojet. Therefore, with the assistance of microlenses, the FWHM compression is 3.7 times. To evaluate its imaging performance, we applied these microlenses for imaging self-assembled PS nanoparticles (NPs) with a diameter of 80 nm (see Appendix A for detailed description on self-assembling NPs). In the experiment, a water-immersion objective (magnification: 60×; NA: 1.0) with an inverted optical microscope was utilized for imaging the self-assembled PS NPs. The suspended microlens was precisely positioned onto the surface of the PS NPs using the SOT (see Appendix A for the preparation of particle suspension). During illumination, the evanescent near-field wave, which carries high spatial frequency information from the sample, is effectively coupled to the microlens and subsequently transformed into a propagating wave for far-field observation [34]. With the presence of the microlens, the outlines of the PS NPs became distinguishable within the field of view [Fig. 2(c)]. By measuring the intensity variation along a transverse cross section through the center of two light spots (NP1 and NP2) emitted by the PS NPs, we obtained a distance of 667 nm between two adjacent peaks [Fig. 2(d)]. Taking inspiration from the highly effective numerical aperture provided by microlenses [35], we utilized them to enhance weak signal detection in single quantum-dot-coated NPs. In absence of a microlens, only a very faint green fluorescence signal (emission wavelength: 512–558 nm) was detected at the NPs’ position [Fig. 2(e1)]. By contrast, with assistance from a microlens, a significantly enhanced fluorescent signal (3.2-fold increase) was observed [Fig. 2(e2)]. This phenomenon primarily relies on the focusing properties and light collection capability of the microlenses, which modulate the optical fields of the excitation and emission light, respectively [18,25,36,37]. The fluorescence enhancement by the microlens was also evaluated through FEM simulations to analyze the energy density distributions. In the simulation, a point source emitter with a wavelength of 532 nm was placed in direct contact with the surface of the microlens. The emission patterns for emitters, both with and without the microlens, were obtained by simulating the field distributions above the emitter [Fig. 2(f)]. It is evident that when equipped with a microlens, the emission becomes highly directional (with a reduced divergence angle from 140° to approximately 62°), effectively redirecting it towards the objective (collection angle: 97.5°) and thereby enhancing the signal collecting.

To evaluate the performance of the microlens in biological imaging, we conducted imaging of various subcellular structures in living neural cells. The utilized living cells were commercially available neural cells (see detailed cell culture description in Appendix A). With the optical traps from the SOT, the deployment of the microlens is confined within a specific region inside the living cells. The trapped microlens can be manipulated flexibly, enabling real-time imaging and detection of fluorescent signals within the region of interest. As shown in Fig. 3(a), a living neural cell marked by a fluorescent probe for intracellular Ca2+ distribution (green, emission wavelength: 512–558 nm) was excited by blue light (465–495 nm) (see detailed cell staining description in Appendix A). In absence of the microlens assistance, it is not possible to resolve a targeted axon through conventional microscope imaging even at optimal focal plane settings. However, when the microlens was trapped and subsequently moved towards the targeted axon using SOT manipulation, efficient enhancement of fluorescent signals occurred, and clear visualization of the axon structure was achieved. The normalized intensity distribution was obtained along the line of observation through the axon in the absence and presence of microlenses [Fig. 3(b)]. The curve of the case with the microlens exhibits two distinct small peaks, indicating a significant enhancement in the resolving capability of the imaging system due to the inclusion of microlenses. Moreover, the microlens can be rapidly moved to enable scanning imaging, thereby facilitating a wider field of view. With the assistance of the SOT, the microlens was moved along the axon of the neural cell, followed by image reconstruction through signal stitching [Fig. 3(c)]. Under fluorescence mode, the fluorescence of the axon is weak without the microlens [Fig. 3(d)]. When the microlens moves along the axon of a neural cell, the fluorescent signals were efficiently enhanced [Fig. 3(e)]. This approach is feasible in overcoming limitations such as a small field of view and inherent spherical aberrations associated with microlenses while achieving high-throughput scanning imaging and signal detection in vitro. In addition to axons, this enhancement in imaging facilitated by the microlens is also applicable for lysosomes [Fig. 3(f)], mitochondria [Fig. 3(g)], and actin filaments [Fig. 3(h)] within neural cells under fluorescence modes (see a detailed description of cell staining in Appendix A). By manipulating the microlens towards targeted structures, a significant enhancement was observed in fluorescent signals resulting in clear and visible examination details.

The ability of microlens-generated PNJs to enhance neural activation and imaging in cells was also investigated. Broadly, changes in intracellular Ca2+ were reflected by fluorescence intensity with or without the presence of a microlens, which was also utilized for recording neuronal activity. It is important to note that neural cells exhibit interconnectivity, and in the absence of external stimulation, they maintain an equilibrium between internal and external currents [Fig. 4(a)]. Placing a microlens near the cell membrane results in strong focusing of incident light, leading to a substantial increase in energy density that initiates channel opening in the membrane. This allows cations to enter neural cells, thereby activating the neurons [Fig. 4(b)]. To demonstrate the effect of PNJs on cells, the distribution of intracellular Ca2+ was marked by a Fluo-4 AM probe. Exciting these neural cells with blue light (465–495 nm) causes them to emit green fluorescence as an indication of successful calcium ion expression. When trapping the microlens using a 1064 nm laser beam and moving it towards the target synapse, the fluorescence signal is effectively enhanced [Fig. 4(c)]. To investigate signal modulation in neural cells, the microlens was brought into contact with the axon. The 1064 nm laser (power: 100 mW) was intermittently activated and deactivated every 2 s to stimulate the neural cells, thereby achieving signal regulation. The precise stimulation range acting on the neural cell could approximate the spot size of the photonic nanojet. Remarkably, in the presence of microlenses, the intensity of inward current was significantly higher than that without the microlenses, even when maintaining the same intensity of stimulated light [Fig. 4(e)]. In neural networks, individual neural cells are interconnected and engage in communication with each other. It has been reported that the brain consists of diverse types and functional categories of neural cells, which establish connections to facilitate material exchange and signal transmission. These interconnected neural cells collectively form a highly intricate neural network [38]. To investigate the finer structures of neural cells, we employed microlenses for precise stimulation of synapses between two neural cells. In our experiment, SOT was utilized to manipulate the microlens, accurately positioning it between the synapses that established connections between two neural cells. By adjusting the duration of the stimulation light, effective control over the neural cell’s signal was achieved. With the assistance of the microlens, fluorescence intensity at the synapse connecting these two neural cells was enhanced, enabling clear visualization of the synaptic gap [Fig. 4(d)]. Importantly, synapses with a microlens exhibited higher fluorescence intensity compared to those without a microlens [Fig. 4(f)]. The phenomenon can be attributed to the microlenses’ ability to focus incident light, effectively increasing the convergence of excitation light and resulting in enhanced fluorescence intensity. Moreover, when the highly focused stimulation light interacts with the cell membrane, it induces a reversal of membrane potential, leading to an influx of calcium ions into the cell. This influx activates neural cells and contributes to the observed fluorescence enhancement [39]. These results confirmed that the presence of microlenses enhances both photogenetic stimulation and optical imaging capabilities.

We have proposed an optical method for stimulation and imaging via PNJs, demonstrating the utilization of transparent resin microspheres as microlenses for fluorescence imaging and signal enhancement. The incorporation of microlenses enhances the local optical intensity, enabling precise spatial adjustment of neural activity while minimizing adverse effects on non-target neural cells. Real-time detection of neural cell fluorescence signals during cell stimulation presents a feasible approach towards the development of high-precision optical stimulation combined with optical imaging. The proposed method holds potential in facilitating accurate diagnosis and treatment of neuronal cell-related disorders in biomedical research.