Aggregation-induced emission luminogen for in vivo three-photon fluorescence lifetime microscopic imaging

Huwei Ni*¶, Zicong Xu*¶, Dongyu Li*, Ming Chen, Ben Zhong Tang and Jun Qian* *State Key Laboratory of Modern Optical Instrumentation Centre for Optical and Electromagnetic Research College of Optical Science and Engineering Zhejiang University, Hangzhou 310058, P. R. China College of Chemistry and Materials Science Jinan University, Guangzhou 510632, P. R. China Department of Chemistry Hong Kong Branch of Chinese National Engineering Research Center for Tissue Restoration and Reconstruction Division of Life Science, State Key Laboratory of Molecular Neuroscience Institute for Advanced Study, Institute of Molecular Functional Materials The Hong Kong University of Science and Technology Clear Water Bay, Kowloon, Hong Kong, P. R. China qianjun@zju.edu.cn


Introduction
Brain imaging has widely arisen people's attention since the last century. In most of the imaging techniques, optical imaging is one of the most favourable methods for its high resolution, no invasion or radiation and°exible combination with other imaging techniques. 1 However, the excitation wavelength of the common°uorescence imaging usually locates in the visible spectral region, which can only reach the super¯cial surface of the biological tissues. Since the brain tissue is a high scattering medium, 2 near-infrared (NIR) light can get much larger penetration depth than visible light, because of its better anti-scattering ability in biological tissues. 3,4 Multi-photon°uorescence microscopic imaging provides a powerful bioimaging approach, which utilizes the NIR light as excitation. In order to achieve a deep penetration, traditional two-photon°uorescence microscopy (2PFM) adopts femtosecond (fs) lasers in NIR-I region (700-900 nm). 5,6 As fs lasers have developed a lot recently, it is becoming more and more convenient to obtain NIR lasers with longer wavelength and higher power density. Meanwhile, three-photon°u orescence microscopy (3PFM) with 900-1700 nm (NIR-II region) fs excitation has been quickly developed. [7][8][9] The 3PFM imaging can penetrate deeper in the brain tissue than 2PFM because excitation lasers in the NIR-II region have less photon scattering. 10 Besides, the three-photon°u orescence (3PF) with higher order nonlinearity brings up higher spatial resolution, larger signal to background ratio, and better optical sectioning ability. 11 Thus, the 3PFM imaging can obtain highcontrast image in the deep tissue.
Although 3PFM has superior imaging performance, the vast majority of endogenous°uorophores have small three-photon absorption (3PA) cross-sections, which cannot be excited e®ectively. Thus, various types of°uorescent probes are adopted to improve the contrast in the 3PFM imaging. 12 However, the metal contained nanoparticles, such as quantum dots, upconversion nanoparticles and gold nanorods, are limited by the excretion di±culty and potential toxicity. The common organic dyes are endowed with high biological compatibility, but limited by small 3PA cross-sections and aggregation-caused quenching property. 13 Hence, aggregation-induced emission luminogens (AIEgens) can be considered as alternatives for in vivo 3PFM imaging. As a type of organic dyes, AIEgens have a large 3PA cross-section and enhanced°uorescence when encapsulated in nanoparticles. 14 When performing the deep-tissue 3PFM imaging, there are still some limitations, such as the small amounts of probes in the targeted samples and the weakened°uorescence intensity at a large depth. Thus, the images of these areas are lack of enough contrast. This could be improved by combining the 3PFM imaging with the°uorescence lifetime imaging microscopy (FLIM), which is suitable for noninvasive study of intracellular processes, 15,16 micro°uidic systems, 17 remote sensing, 18,19 lipid order problems in physical chemistry, 20 temperature sensing 21 and clinical medicine. 22 FLIM can provide a more sensitive and precise image based on the weak signals, compared with the tra-ditional°uorescence intensity imaging. 23 One reason is that the°uorescence lifetime of probe keeps stable at di®erent amounts or under varying intensity of the excitation laser. 24,25 Another reason is that the°uorescence lifetime of each pixel is obtained with time-correlated single photon counting (TCSPC) technique, thus with a signi¯cantly increasing signal-noise ratio. 26 In this paper, DCDPP-2TPA, a novel AIEgen, was adopted as°uorescent probe. The three-photon absorption cross-section of the AIEgen, which has a deep-red°uorescence emission, was proved to be large. The absorption spectrum and the 3PF spectrum of AIEgen nanoparticles were both measured and recorded. The 3PF lifetimes of AIEgen nanoparticles in water and AIEgen molecular in di®erent solvents (THF, chloroform and toluene) were measured. Besides, a 3PF lifetime imaging system based on the TCSPC technique was set up to conduct the further imaging. By injection with AIEgen nanoparticles, brain vasculature of one skull-removed mouse was imaged and the maximum depth reached about 600 m. Even reaching the depth of 600 m, tiny capillary vessels (diameter ¼ 1:9 m) could be distinguished. In addition, the 3PF lifetimes of AIEgen nanoparticles at some representative depths were in accord with the previous results. Finally, a vivid 3D reconstruction was further obtained and presented a wealth of lifetime information. In the future, the combination strategy of 3PFM imaging and FLIM could be further applied in the brain functional imaging and provide a larger detection depth.

Preparation of DCDPP-2TPA nanoparticles
The AIE nanoparticles were synthesized in the following way. An aliquot of 5 mL chloroform solution containing 2.2 mg DCDPP-2TPA and 26.4 mg F-127 was added into a°ask¯rst and the°ask was put into the ultrasonic cleaner for 5 min to ensure even mixing. The°ask was then placed in a rotary evaporator to evaporate all liquid. Until the chloroform was completely removed, 2.2 mL DI water was used to dissolve the nanoparticles in the°ask.
After that, the°ask containing aqueous dispersion of AIEgen nanoparticles was put into the ultrasonic cleaner again and sonicated for another 5 min. At last, a clear aqueous dispersion of DCDPP-2TPA nanoparticles was gathered.

Characterizations of DCDPP-2TPA nanoparticles
The size distribution of AIE nanoparticles synthesized by us was measured via dynamic light scattering (DLS) using a particle size analyzer (90 Plus, Brookhaven Instruments Co., USA). The zeta potential of nanoparticles was recorded with the same instrument. The morphology of nanoparticles was observed via a transmission electron microscope (TEM, JEM-1200EX, JEOL, Japan). Both extinction and transmission spectra (300-1800 nm) were measured using an UV-Vis-NIR scanning spectrophotometer (Agilent Cary 5000).

Cell viability analysis
The cytotoxicity of DCDPP-2TPA nanoparticles was evaluated via the instruction of cell counting kit-8 (CCK-8, Beyotime). About 20,000 cells/well in a 100 L suspension were incubated in 96-well plates for 24 h. A 100 L HBSS (Gibco) containing nanoparticles of di®erent concentrations was then added into each well. After incubated for 1 h, the culture medium was removed and the cell well was washed three times with HBSS. An aliquot of 100 L culture medium containing CCK-8 (10%) was added into each well and incubated for another 4 h. Finally, the absorbance at 450 nm and 625 nm was measured with a microplate reader (Thermo, USA).

Animal preparation
All of the animal experiments and housing procedures were conducted strictly according to the requirements of the Institutional Ethical Committee of Animal Experimentation of Zhejiang University.
2.6. The one-photon and three-photon°u orescence measuring system The fs laser beam with a center wavelength of 1550 nm was divided into two perpendicular directions by a polarization beam splitter (PBS), adopted to excite the 3PF and trigger the time synchronization of TCSPC system, respectively (Scheme 1). A quarter waveplate was put before the PBS in order to adjust the power of laser beam in the two divided directions. The stronger laser beam was introduced into a scanning microscope (FV1200&BX61, Olympus) and re°ected by an 800 nm short-pass dichroic mirror (DC) to excite AIEgen nanoparticles. The focus was adjusted by controlling the Z-axis of the water-immersed objective (XLPLN25XWMP2, Olympus, 25Â, working distance ¼ 2:0 mm, NA ¼ 1:05) on computer. The 3PF signals collected by the water-immersed objective passed through the DC and were captured by a photomultiplier tube (PMT, HPM-100-50, Becker&Hickl GmbH). The weaker laser beam was utilized as synchronization signals via a photodetector (PD) in the TCSPC system (Becker&Hickl GmbH, SPC-150). The 3PF lifetime imaging system was used to measure°uorescence lifetimes of AIEgen nanoparticles in water and AIEgen molecular in di®erent solvents (THF, chloroform and toluene). Furthermore, the system was adopted to conduct the three-photon°uorescence lifetime imaging microscopy (3PFLIM) on mouse. A healthy ICR mouse underwent craniotomy surgery and the open window of the skull was protected by a clean glass cover. Before imaging, the mouse under anesthetic was intravenously injected with 200 L AIEgen nanoparticles in 1Â PBS (1 mg/mL). During the imaging, the mouse was immobilized on the stage.
shown in Fig. 1(b). The size distribution and TEM image of these AIEgen nanoparticles ( Fig. 1(c)) indicated an average size of approximately 40 nm, which means that these nanoparticles were suitable for circulation in blood vessels in the further bioimaging. 28 The zeta potential of DCDPP-2TPA nanoparticles was about -13:45 AE 0:65 mV. In addition, DCDPP-2TPA nanoparticles showed negligible toxicity toward cells (Fig. 1(d)).
The°uorescence spectra with one-photon and three-photon excited were almost the same. The central wavelength of the°uorescence spectra was 645 nm.

Measurement of three-photon°u orescence lifetime
DCDPP-2TPA was directly dissolved in tetrahydrofuran, chloroform and toluene without encapsulation. The 3PF lifetimes of DCDPP-2TPA in these three solutions were measured, which were 570 ps in tetrahydrofuran, 960 ps in chloroform and 3366 ps in toluene ( Fig. 2(a)). Compared with them, 3PF lifetime of DCDPP-2TPA nanoparticles in water was around 5215 ps, relatively longer since the aggregated state of DCDPP-2TPA molecules in nanoparticles was rather compact. 34 Three-photon°u orescence intensity image (in grayscale) and three-photon°uorescence lifetime image (quadrate color map) of DCDPP-2TPA nanoparticles in water (contained in a glass capillary tube) were both recorded ( Fig. 2(b)). The distribution of 3PF lifetime within the color map region (in Fig. 2(b)) was shown in Fig. 2(c), and the central lifetime was measured to be 5.2 ns.

In vivo three-photon°uorescence lifetime microscopic imaging
In vivo 3PFLIM was further conducted, and DCDPP-2TPA nanoparticles were used as the three-photon°uorescent probes to stain the brain vasculature of mice. An eight-week-old male mouse with craniotomy was anesthetized¯rst and then intravenously injected with 200 L aqueous dispersion of DCDPP-2TPA nanoparticles (1 mg/mL, PBS, 1Â). The mouse was imaged by the aforementioned 3PF lifetime microscopic system under fs laser (with a center wavelength of 1550 nm) excitation, and the°uorescence lifetime images of brain vasculature at di®erent depths were obtained. Cerebrovascular 3PFLIM images were recorded at every 20 m and the penetration depth of blood vessels reached about 600 m. 3D reconstructions of brain vasculature were further built with the software of Imaris (Fig. 3). 3PFLIM images of brain vasculature at di®erent depths were recorded, and some speci¯c images (0, 200, 400, 600 m) were shown in Figs. 4(a), 4(d), 4(g) and 4(j). In order to con¯rm whether the°u orescence signals came from the DCDPP-2TPA nanoparticles and whether the°uorescence signals in vivo have changed, the 3PF lifetimes of certain spots of the blood vessels in these images were analyzed, which were 4086 ps at 0 m, 4656 ps at 200 m, 4646 ps at 400 m and 6470 ps at 600 m (Figs. 4(b), 4(e), 4(h) and 4(k)). Although DCDPP-2TPA nanoparticles distributed at various depths in the mouse brain, their°uorescence lifetimes were still stable, ranging from 4 ns to 7 ns, which were within the 3PF lifetime distribution of DCDPP-2TPA nanoparticles in vitro (Fig. 2(c)). This fact illustrated that DCDPP-2TPA nanoparticles could be a stable probe to be utilized in in vivo 3PFLIM and keep the same aggregated state both in vivo and in vitro. Afterwards, to evaluate the spatial resolution of 3PF lifetime images, a line was drawn across capillary vessels in each image at di®erent depths, and the full width at half-maximum (FWHM) of the intensity curve was calculated by Gaussian¯tting (Figs. 4(c), 4(f), 4(i) and 4(l)). At the depth of 600 m, tiny capillary vessels (diameter ¼ 1:9 m) could still be distinguished clearly.

Conclusions
In summary, aqueous dispersion of AIEgen (DCDPP-2TPA) nanoparticles was synthesized and characterized. Moreover, three-photon°uorescence lifetime imaging microscopy (FLIM) was realized by combining time-correlated single photon counting (TCSPC) technique and three-photon°uorescence microscopic system. Three-photon°uorescence lifetimes of AIEgen in di®erent solvents and AIEgen nanoparticles in water were obtained basing on this setup. Afterwards, with the fs laser (with a center wavelength of 1550 nm) excitation and AIEgen nanoparticles, the vivid 3D reconstruction of threephoton FLIM cerebrovascular imaging of a skullopened mouse was obtained by the 3PFLIM, and the imaging depth reached 600 m. 3PF lifetimes of nanoparticles in the living mouse were the same as that measured in vitro, showing that DCDPP-2TPA nanoparticles had the same aggregated state both in vivo and in vitro. Furthermore, tiny capillary vessels as small as 1.9 m could be distinguished clearly on the bottom of the image (600 m). Three-photon FLIM can achieve in vivo 4D (3D + lifetime) brain structural and functional imaging, which increases lifetime information while ensuring high resolution and deep penetration.

Con°ict of Interest
The authors declared that there was no con°icts of interest in this work.