Keywords

13.1 Introduction

For in vivo fluorescence imaging, visible-emitting fluorescent probes cannot be applied because of the strong absorption and scattering of visible light by intrinsic chromophores, organelles, and cytoskeleton in cells [1, 2]. Most of the visible-emitting fluorescent probes have been used for in vitro imaging and sensing of biomolecules and organelles using conventional fluorescence microscopes [3, 4]. Compared with visible light (400–700 nm), NIR light (700–1400 nm) allows deeper penetration with reduced absorption and scattering in living tissues [5]. In addition, tissue autofluorescence in the NIR region is much lower than the autofluorescence in the visible region [5]. Thus, NIR fluorescence imaging is widely used for non-invasive visualization of deep tissues in living system [6,7,8,9].

So far, conventional NIR region ranging from 700 to 900 nm (1st NIR window) has been used for in vivo imaging [5]. In this NIR region, many kinds of fluorescent probes such as Cy7, ICG, iRFP, and CdSeTe quantum dots (QDs) are commercially available (Fig. 13.1). Recently, NIR fluorescence imaging in the wavelengths of 1000–1400 nm (2nd NIR window) [10,11,12,13,14] has been attracted for clearer deep-tissue imaging at the whole-body level (Fig. 13.2). As the tissue autofluorescence and scattering significantly decrease beyond 1000 nm, 2nd NIR fluorescence imaging offers better spatiotemporal resolution in the deep-tissue imaging [10,11,12]. Unfortunately, compared with conventional NIR fluorescent probes in the 1st NIR window, NIR fluorescent probes that can be used in the 2nd NIR window are very limited.

Fig. 13.1
figure 1

Representative nanoprobes that can be used for bioimaging at the wavelengths of visible and near infrared regions

Full size image
Fig. 13.2
figure 2

First and second NIR windows in biological tissues. Top: These plots of effective attenuation coefficient (on a log scale) versus wavelength show that absorption and scattering from oxygenated blood, deoxygenated blood, skin, and fatty tissue is lowest in either the first (pink shaded area) or 2nd (grey) NIR window. Bottom: Sensitivity curves for typical cameras based on silicon (Si), indium gallium arsenide (InGaAs) or mercury cadmium telluride (HgCdTe) sensors. Reproduced from ref. [10] with permission from Springer-Nature

Full size image

During the past 5 years, several types of NIR fluorescent probes such as single-walled carbon nanotubes (SWNTs) [15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32], PbS QDs [33,34,35,36,37,38,39,40,41,42,43], Ag2S QDs [44,45,46,47,48,49,50,51,52,53,54,55,56,57], and rare earth-doped nanoparticles [58,59,60,61] have been developed for in vivo imaging in the 2nd NIR window. Recently, organic dye-based NIR nanoprobes with low toxicities have attracted much attention for deep-tissue imaging in the 2nd NIR window [62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91]. In this chapter, we focus on the synthesis, optical properties, and applications of NIR fluorescent nanoprobes for non-invasive brain imaging in the 2nd NIR window.

13.2 Optical Property of Brain Tissue

In non-invasive fluorescence imaging of brain, autofluorescence, absorption, and scattering by scalp and skull significantly affect the signal-to-background ratios of the fluorescence images. Autofluorescence and absorption result mainly from the intrinsic chromophores such as nicotinamide adenine dinucleotide phosphate (NADP) and flavin in intracellular compartments [92,93,94]. To get clear NIR fluorescence images of mouse brains, excitation wavelengths are very important to reduce the absorption and autofluorescence by tissues. The detection wavelengths for fluorescence emission are also important to get reduced scattering images.

The absorption spectrum of a mouse brain shows that tissue absorption at the NIR region from 700 to 1400 nm is very low (Fig. 13.3a). The strong absorption at the visible region less than 600 nm is attributed to the absorption by intrinsic chromophores such as flavin and hemoglobin. The intense absorption at around 1500 nm is due to the absorption by water molecules in the brain tissue. Autofluorescence of the mouse brain strongly depends on the wavelength of excitation (Fig. 13.3b). The intensity of autofluorescence over 1000 nm is very low compared with that of the autofluorescence at 520 and 720 nm. Figure 13.4a shows autofluorescence images (at 520, 720, 1100, 1300, and 1500 nm) of brain tissues. It should be noted that the autofluorescence intensity decreases with increasing the emission wavelength, indicating intrinsic chromophores in tissues are less excited at longer wavelengths. Furthermore, the tissue scattering of NIR fluorescence decreases with increasing the emission wavelength (Fig. 13.4b). These optical properties in the 2nd NIR window allow clearer deep-tissue imaging of brain with high signal-to-background ratios, compared with the imaging performed at the visible and 1st NIR regions.

Fig. 13.3
figure 3

(a) Absorption spectrum of a mouse brain. (b) Autofluorescence of a mouse brain. The fluorescence at 520 and 720 nm was obtained by excitation at 482 and 670 nm, respectively. The fluorescence at 1110, 1300, and 1500 nm was obtained by excitation at 785 nm. Adapted from ref. [41]

Full size image
Fig. 13.4
figure 4

(a) Autofluorescence from brain tissues of mice. The autofluorescence at 520 and 720 nm was measured by excitation at 482 and 670 nm, respectively. The autofluorescence over 1000 nm was measured by excitation at 785 nm. (b) Fluorescence images of 15 μm beads containing five types of QDs (520, 720, 1100, 1300, and 1500 nm emission) though brain tissues of mice. Thickness of the brain tissues was 100 μm. The intensity of fluorescence emission of each QDs was adjusted to the similar level. Upper images show raw data, and lower images show normalized data. Reproduced from ref. [36] with permission from Royal Society of Chemistry

Full size image

13.3 NIR Nanoprobes for In Vivo Fluorescence Imaging

13.3.1 Nanomaterial-Based NIR Nanoprobes

SWNTs in the 2nd NIR window the first-reported fluorescent nanoprobes by Dai group for intravital imaging in mice [24]. SWNTs are cylindrical nanotubes (hundred nanometer in length) consisting of graphene layers, and they have a broad emission in the 2nd NIR region [15]. Although raw SWNTs are insoluble to water and their fluorescence quantum yields are very low (<1%), surface functionalization results in water-dispersible and bright SWNTs [24]. So far, several groups have demonstrated the capability of surface-functionalized SWNTs as NIR fluorescent probes for non-invasive imaging of organs, lymph nodes, tumors, and cerebral vessels in mice [25,26,27,28,29,30,31,32]. For biomedical applications, SWNTs have serious problems on their cytotoxicity and difficulty in exclusion from the body [15, 95, 96].

Nanoparticle-based NIR nanoprobes such as Ag2S QDs, PbS QDs, rare earth-doped nanoparticles are alternative NIR nanoprobes for intravital imaging in the 2nd NIR window (Fig. 13.5). NIR-emitting Ag2S QDs [44,45,46,47,48,49,50,51,52,53,54,55,56,57], and PbS QDs [33,34,35,36,37,38,39,40,41,42,43] are semiconductor nanocrystals that have unique optical properties, such as size-dependent tunable emissions, narrow emission bands, high quantum yields, and high resistance to photobleaching. In 2010, Wang group first synthesized NIR-emitting Ag2S QDs for bioimaging by thermal decomposition of (C2H5)2NCS2Ag [44]. By optimization of the reaction conditions for preparing QDs, they succeeded in the synthesis of emission tunable Ag2S QDs with a high quantum yield (ca. 20%) in the wavelength region from 900 to 1200 nm [52]. Furthermore, they demonstrated the capability of the Ag2S QDs for tumor imaging, in situ tracking of transplanted stem cells, and imaging-guided precise operation of glioma [46, 54,55,56]. In 2013, our group reported a facile method for preparing water-dispersible PbS QDs, which have tunable emissions from 1000–1400 nm with high quantum yields of 6–12% [33]. As the water-dispersible PbS QDs are capped with glutathione (a natural tripeptide), functionalized QDs with biomolecules such as antibody and peptide are easily prepared. We demonstrated the utility of the PbS QDs for non-invasive visualization of lymph nodes as well as breast tumors in living mice. Furthermore, we synthesized highly bright PbS/CdS QDs with a core/shell structure that emit from 1000 to 1500 nm, and we applied these QDs to non-invasive imaging of cerebral blood vessels in mice [34]. In this work, we found that the signal-to-background ratio in the NIR fluorescence imaging of brain tissue at 1300 nm can be improved 76 times compared to the NIR imaging at 720 nm. In 2013, Moghe group reported rare earth (Er3+, Ho3+, Tm3+, and Pr3+)-doped nanoparticles as in vivo shortwave infrared reports for intravital imaging in mice [59]. Other groups also reported rare-earth nanoparticles for organ imaging and cancer early detection [58, 60, 61]. However, biomedical applications of these nanoparticle-based NIR nanoprobes including SWNTs are very limited due to their dose-dependent toxicity [95, 96].

Fig. 13.5
figure 5

Schematic illustration of typical nanomaterial-based NIR nanoprobes, (a) Polyethylene glycol (PEG)-capped SWNT [24], (b) PEG-capped Ag2S QDs [45], (c) glutathione-capped PbS QDs [33], and (d) rare earth-doped nanoparticle [59]

Full size image

13.3.2 Organic Dye-Based NIR Nanoprobes

Recently, organic dye-based NIR nanoprobes emitting over 1000 nm have been developed as next-generation NIR nanoprobes for intravital bioimaging. Although a variety of NIR organic nanoprobes emitting in the 1st NIR window are commercially available, there are a very limited number of NIR organic nanoprobes that emit over 1000 nm. During a few years, several types of NIR-emitting organic dyes beyond 1000 nm have been reported for bioimaging. Organic dye-based NIR nanoprobes are classified into three types (Fig. 13.6a). Frist type is NIR-dye nanoparticles, where NIR dyes are incorporated into micelles or amphiphilic polymers [62, 63]. Second type is NIR-dye complex, where NIR dyes are conjugated to proteins such as fetal bovine serum [64]. Third type is solely, water-dispersible NIR dyes [65, 67]. To date, these types of organic dye-based NIR nanoprobes have appeared as probes for bioimaging in the 2nd NIR window (Fig. 13.6b). Compared with nanomaterial-based NIR nanoprobes, organic dye-based nanoprobes have well-defined architectures with rapid metabolism and low toxicity [66, 97].

Fig. 13.6
figure 6

(a) Schematic illustration of three types of organic dye-based NIR fluorophore. (b) Chemical structures of typical NIR dyes emitting over 1000 nm (laser dyes, benzo-bis (1,2,5-thiadiazole) dyes, polymethine dyes, and cyanine dyes). Blue and pink color show the electron acceptor and donor unit, respectively

Full size image

In 2013, Dai group reported laser dye, IR-1061 incorporated nanoparticles as a NIR nanoprobe for in vivo imaging [62]. IR-1061 is a commercially available polymethine dye, which is highly hydrophobic and insoluble to water. Thus, they incorporated IR-1061 to nanoparticles consisting of amphiphilic polymer poly (acrylic acid, PAA) and polyethylene glycol-conjugated phospholipid (DSPE-nPEG). The fluorescence quantum yield of the IR-1061 incorporated nanoparticle was 1.8% [62]. They succeeded in performing whole-body imaging of nude mice after intravenous injection of the IR-1061 incorporated nanoparticles. They found the facile excretion of the IR-1061 nanoparticles from the body. Using the similar strategy, Dai group also reported fluorescent copolymer (poly(benzo[1,2-b:3,4-b’]difuran-alt-fluorothieno-[3,4-b]thiophen, pDA) incorporated nanoparticles as a NIR nanoprobe and performed ultrafast fluorescence imaging (>25 flames/sec) in the 2nd NIR window [63].

In the design of NIR-emitting nanoprobes, the energy band gap is known to be significantly affected by conjugation length as well as donor–acceptor (D–A) charge transfer in π-conjugated molecules [78]. In 2016, Dai, Cheng and Hong reported a new type of benzo-bis(1,2,5-thiadiazole) NIR-emitting dyes with D–A–D charge structures (Fig. 13.6b) [65, 69, 80]. This type of NIR nanoprobe (CH1055-PEG) emits at approximately 1050 nm with a quantum yield of 0.3% in an aqueous solution [65]. They modified the fluorescence brightness of CH1055 by complexation of its sulfonated derivative (CH-4T) with bovine serum to produce 110-fold increase in NIR fluorescence. They succeeded to perform molecular imaging of tumors in mice using affibody-functionalized CH1055 [65], leading to a possible application of the NIR dye to tumor detection in humans. Their works showed that the brightness of D–A–D dyes can be significantly improved by their composites with proteins. To date, several derivatives of benzo-bis(1,2,5-thiadiazole) nanoprobes have been reported for in vivo tumor imaging and image-guided surgery [68, 78].

In 2017, Sletten group reported flavylium polymethine nanoprobes for near- and shortwave-infrared imaging [79]. They synthesized a new series of polymethine dyes with dimethylamino flavylium heterocycles and found that a flavylium dye (Flav 7) emitting at ca. 1050 nm is 13 times brighter than IR-26 (quantum yield: 0.05). They achieved whole-body imaging of nude mice by intravenous injection of Flav 7 micelles consisting of mPEG-DSPE lipids, suggesting the possible translation of polymethine nanoprobes to optical diagnostics in NIR region over 1000 nm. In 2018, Zhang group reported the synthesis and application of a cyanine dye (FD-1080) that emit at around 1100 nm for deep-tissue high-resolution dynamic bioimaging [67]. The excitation wavelength (1064 nm) of this NIR probe was longer than that (650–980 nm) of previous reported NIR probe. Thus, this NIR probe allowed deeper tissue imaging due to the high penetration of excitation light (1064 nm). The quantum yield of PD-1080 was 0.31% in an aqueous solution and could be increased to 5.94% after combining with fetal bovine serum. This dye is the first-reported NIR nanoprobe that can be excited at the wavelength longer than 1000 nm.

More recently, several groups have found that a commercially available dye, indocyanine green (ICG) with an emission peak of 830 nm can be used to NIR fluorescence imaging in the 2nd NIR window [72, 73, 83, 85]. ICG is the only NIR nanoprobe that is approved by the Food and Drug Administration (FDA) for clinical use in humans. Although the NIR emission of ICG over 1000 nm is very week, its emissions in blood and vasculatures are clearly detected by an InGaAs camera. In 2018, Bawendi and Bruns group showed that ICG can be used as a NIR nanoprobe for in vivo fluorescence imaging over 1000 nm, including intravital microscopy, non-invasive real-time imaging in blood and lymph vessels, imaging of hepatobiliary clearance, and molecular targeted in vivo imaging [73]. In the same year, Annapragada group reported that ICG-incorporated liposomes show higher contrast to noise ratios compared to free ICG in the 2nd NIR window, allowing visualization of hind limb and intracranial vasculatures [72, 83]. Sun and Chen group reported that the NIR emissions of ICG including IRDye800 and IR-12N3 have the potential to accelerate clinical translation of NIR fluorescence imaging in the 2nd NIR window [85]. While no FDA-approved NIR organic nanoprobes with an emission peak over 1000 nm exist, the emission of ICG may give rapid translation of longer NIR fluorescence to humans in clinical applications.

13.4 NIR Fluorescence Detection System for Brain Imaging

In most of the commercially available in vivo imaging systems, conventional NIR wavelenths ranging from 700 to 900 nm (1st NIR optical window) are used for deep-tissue imaging. This is because the conventional NIR photodetectors (silicon CCD camera) are sensitive in the 1st NIR region, and 1st NIR-emitting probes (e.g., Indocyanine green, Cy 7, and CdSeTe QDs) are commercially available. Although 1st NIR fluorescence imaging is useful for the non-invasive visualization of organs and tissues, its spatial resolution is not enough to observe cellular dynamics. As tissue autofluorescence and scattering significantly decrease with increasing the excitation/emission wavelength, fluorescence imaging in the 2nd NIR region should be very useful to get better spatiotemporal resolution in deep-tissue imaging [10]. However, there are no commercially available imaging systems with high spatiotemporal resolution in the 2nd NIR window.

Our 2nd NIR microscope imaging system is based on the Macro Zoom System with zoom function from 0.63× to 6.3× (Fig. 13.7). Optical system is optimized for VIS, 1st NIR, and 2nd NIR fluorescence imaging. Solid-state lasers for 645, 785, and 978 nm excitation, and emission filters of 1100 ± 25 nm, 1300 ± 25 nm, and 1500 ± 25 nm are equipped to the optical system. A Xe lamp was used as the excitation light source at 482 nm for VIS imaging. A Si EM camera (iXon3, Andor) is used for VIS and 1st NIR fluorescence imaging, and an InGaAs CMOS camera (C10633-34; Hamamatsu photonics) is used for 2nd NIR fluorescence imaging.

Fig. 13.7
figure 7

(a, b) Up-right fluorescence microscope system for in vivo imaging in VIS, 1st NIR, and 2nd NIR region (400–1400 nm). (c) Sensitivity curves for typical cameras based on silicon (Si), indium gallium arsenide (InGaAs) sensors. Si and InGaAs cameras are sensitive within the 1st and 2nd NIR windows, respectively

Full size image

Dai group used a high-resolution microscopic system for NIR imaging of brain imaging cerebral vessels (Fig. 13.8) [31]. High-magnification intravital imaging of cerebral vessels was carried out in epifluorescence mode with an 808-nm diode laser (RMPC lasers, 160 mW) as the excitation source and two objective lenses (4× and 10×) for microscopic imaging. The mouse with scalp hair removed was intravenously injected with a solution of SWNTs and placed in a home-made stereotactic platform fixed on a motorized 3D-translational stage that allowed for the digital position adjustment and readout of the mouse relative to the objective. The emitted fluorescence was filtered through a 1000-nm long-pass filter, a 1300-nm long-pass filter, and a 1400-nm short-pass filter to ensure only photons in the 1300–1400 nm.

Fig. 13.8
figure 8

A schematic of NIR fluorescence imaging for non-invasive through-scalp and through-skull brain vascular imaging. NIR-IIa emission: 1300–1400 nm. Reproduce from ref. [31] with permission from Springer-Nature

Full size image

13.5 Non-invasive Brain Imaging Using NIR Nanoprobes

13.5.1 Cerebral Blood Vessels

13.5.1.1 SWNT Probes

To date, mouse brain imaging has largely relied on magnetic resonance (MR), X-ray computed tomography (CT), and positron emission tomography (PET). However, these imaging modalities have limited spatial resolution and long scanning times. During the past 5 years, a number of reports on non-invasive brain imaging of mice in the 2nd NIR window have appeared using NIR nanoprobes such as SWNTs [31, 98], QDs [34, 41, 99,100,101,102], rare earth-doped nanomaterials [104, 105], and organic dyes [65,66,67,68,69,70,71,72,73,74,75,76,77, 104,105,106,107]. In 2014, Dai et al. first reported 2nd NIR fluorescence imaging of a mouse brain by using (SWNTs) (Fig. 13.9) [31]. They performed through-scalp and through-skull fluorescence imaging of mouse cerebral vasculatures without craniotomy, utilizing the intrinsic photoluminescence of SWNTs in the 1.3–1.4 nm NIR window. They found that reduced photon scattering in the NIR region allowed fluorescence imaging to a depth of >2 mm in mouse brain with sub-10 μm resolution. In this fluorescence imaging, they achieved dynamic NIR fluorescence imaging (5.3 frames/sec) of cerebral blood perfusion.

Fig. 13.9
figure 9

NIR fluorescence imaging of mouse brain vasculatures with SWNT–IRDye800 in different NIR subregions. (a), A C57Bl/6 mouse head with the hair removed. (b–d), Fluorescence images of the same mouse head in the NIR-I, NIR-II, and NIR-IIa regions. In (d), the inferior cerebral vein, superior sagittal sinus, and transverse sinus are labeled 1, 2, and 3, respectively. Reproduced from ref. [31] with permission from Springer-Nature

Full size image

13.5.1.2 QD Probes

In 2014, our group first reported the non-invasive fluorescence angiography of a mouse head using PbS/CdS QDs in the 2nd NIR window [34]. To date, several types of QDs including PbS QDs, Ag2S QDs, and InAs QDs have been used for brain imaging in the 2nd NIR window [100,101,102,103]. Figure 13.10 shows the fluorescence angiography for a mouse head using bovine serum albumin (BSA)-conjugated VIS, 1st NIR, and 2nd NIR-emitting QDs [34]. The fluorescence images of mouse cerebral vessels were measured by using band-path filters (525, 720, and 1300 nm) after injection of each QDs in a mouse tail vein. Autofluorescence of the mouse body dramatically decreased in the angiography of 2nd NIR images compared with that of the VIS and 1st NIR images. Blood vessels showed a clearer image in the 2nd NIR region due to the higher penetration and lower scattering of the 2nd NIR light in the tissue. The spatial resolution of the fluorescence image of the blood vessels was significantly improved by increasing the imaging wavelength, which also increased the signal-to-background ratio of the 2nd NIR fluorescence images compared with VIS or 1st NIR fluorescence images.

Fig. 13.10
figure 10

Fluorescence angiography of a mouse head. The images show its Vis (520 nm), 1st NIR (720 nm), and 2nd NIR (1300 nm) fluorescence angiographies, where excitation wavelengths are 488 nm, 670 nm, and 785 nm, respectively. CdSe/ZnS QDs, CdSeTe/CdS QDs, PbS/CdS QDs were used for fluorescence imaging at 525, 720, and 1300 nm, respectively. Reproduced from ref. [34]. Copyright (2014) with permission of Royal Society of Chemistry

Full size image

The 2nd NIR fluorescence imaging shows deeper penetration with lower scattering compared with the VIS and 1st NIR fluorescence imaging. Although NIR light over 1000 nm can penetrate across the skin and scalp of the mouse brain, it is difficult to determine the precise value of brain imaging depth in living mice. Figure 13.11 shows the imaging depth for an isolated mouse brain. Visualization depth for the cerebral blood vessels was evaluated by measuring z-stacked images for the isolated brain. In our NIR imaging system, maximum depth for the visualization of the fine structure of cerebral blood vessels was determined to be ca. 1.6 mm. For a brain of living mouse, we can perform NIR fluorescence imaging of cerebral blood vessels with high spatial resolution. Figure 13.12 shows non-invasive NIR fluorescence images of cerebral blood vessels of a nude mouse after injection of PbS QDs via a tail vein. Immediately after injection of the QDs, strong NIR fluorescence signals were detected and clear images of the cerebral blood vessels could be taken. Although the intensity of the NIR fluorescence signals was gradually decreased, the NIR fluorescence images of the cerebral blood vessels could be taken for ca. 5 minutes post-injection of the PbS QDs. In this brain imaging, we could observe small blood capillaries with a spatial resolution of ca. 10 μm.

Fig. 13.11
figure 11

(a) Bright-field image of a mouse brain perfused by PbS QDs. (b) Raw fluorescence images and sectioning images at the depth of 0.2 mm, 0.8 mm, and 1.6 mm from the surface. The sectioning image was obtained from a raw image minus its previous and next image. Red circles with dotted lines show cerebral blood vessels appearance after sectioning. Scale bar: 1 mm. Adapted from ref. [41]

Full size image
Fig. 13.12
figure 12

NIR fluorescence images of cerebral blood vessels in a nude mouse. NIR fluorescence images were taken 1 s, 20 s, 40 s, and 60 s post-injection of PbS QDs (200 μL, 2 μM) through a tail vein. Excitation: 670 nm, Emission >1000 nm. Laser power; 25 mW/cm2. Exposure time: 100 ms. Scale bar: 1 mm. Adapted from ref. [12]. Copyright (2018) The Electrochemical Society

Full size image

13.5.1.3 Rare-Earth Nanoprobes

Rare earth (Er3+)-doped nanoprobes can also be used for brain imaging in the NIR window beyond 1500 nm [104, 105]. In general, with increasing the emission wavelength in fluorescence imaging, tissue scattering of the emission light is decreased. Thus, fluorescence imaging beyond 1500 nm would give clearer images of the brain compared with the imaging wavelength in the 2nd NIR region. Zhong et al. reported Er/Ce co-doped NaYbF4 nanocrystals for in vivo fluorescence imaging in the NIR region between 1500 and 1700 nm [104]. This Er/Ce co-doped nanoprobe shows bright emission at 1550 nm under 980 nm excitation. In this probe, Ce doping suppresses the upconversion pathway while boosting down conversion by ninefold to produce bright 1550 nm emission. The authors reported that the quantum yield of this rare-earth nanoprobe was 0.27–2.73% with a highest value among reported down conversion rare earth-doped nanomaterials, leading to fast in vivo cerebrovascular imaging with a 20 ms exposure time in the NIR region between 1500 and 1700 nm (Fig. 13.13).

Fig. 13.13
figure 13

Fast in vivo brain imaging with Er-doped nanoprobes in the 2nd NIR region. (a) Color photograph of a C57Bl/6 mouse preceding NIR fluorescence imaging. (b–d) Time-course NIR brain fluorescence images (exposure time: 20 ms) showing the perfusion of Er-doped nanoprobes into various cerebral vessels. Reproduced from ref. [104] with permission from Springer-Nature

Full size image

13.5.1.4 Organic Dye Nanoprobes

Very recently, much attention has been paid to organic dye-based NIR nanoprobes for in vivo imaging in the 2nd NIR window due to their low toxicities and rapid clearance from the body [62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77, 106, 107]. The safety of NIR fluorescent probes is crucial for the application of 2nd NIR fluorescence imaging to biomedical and clinical fields. The fluorescence brightness of organic dye-based NIR nanoprobes is lower than that of nanomaterial-based NIR nanoprobes because of the smaller extinction coefficients of organic dyes. To overcome the low fluorescence brightness, several efforts have been made to increase the fluorescence brightness of organic dye-based NIR probes.

Wan et al. reported a bright organic NIR nanoprobe (p-FE) for three-dimensional imaging of cerebral vasculatures [66]. They encapsulated an organic NIR dye (FE) in the hydrophobic interior of an amphiphilic polymer, poly (styrene-co-chloromethyl styrene)-graft-poly(ethylene glycol) (PS-g-PEG), to produce a bright and biocompatible NIR nanoprobe (size: 12 nm) that can be used for the fluorescence imaging in the 2nd NIR window. The fluorescence quantum yield of p-FE in aqueous environment was ca. 16.5%. With this bright organic NIR nanoprobe, non-invasive ultrafast in vivo NIR imaging of cerebral blood vessels with a short exposure time of 2–5 ms was achieved (Fig. 13.14). In addition, the bright organic NIR nanoprobe enabled three-dimensional NIR fluorescence imaging of cerebral blood vessels using a confocal imaging system.

Fig. 13.14
figure 14

(a) Photo and wide-field NIR-II epi-fluorescence imaging of the brain in a mouse injected with p-FE (808 nm excitation, emission >1200 nm) with an exposure time of 5 ms. (b) Ex vivo confocal imaging of brain in a mouse injected with p-FE (785 nm excitation, emission >1100 nm, laser power ~30 mW, PMT voltage ~500 V). Reproduced from ref. [66] with permission from Springer-Nature

Full size image

13.5.2 Brain Tumors

NIR fluorescence imaging is useful for the non-invasive visualization of brain tumors as well as cerebral blood vessels in living mice. In 2016, Antaris et al. reported a small NIR dye (CH1055) for mouse brain imaging in the 2nd NIR window [65]. They synthesized a small molecule (CH1055, 0.97 kDa) and PEGylated CH1055 (8.9 kDa) and showed the capability of these NIR organic nanoprobes for orthotopic glioblastoma brain tumor imaging. They used an orthotopic glioblastoma brain tumor-bearing mouse by implanting U87MG cells in the mouse brain at a depth of 4 mm with the left hemisphere (Fig. 13.15a, b). Once the brain tumor reached a diameter of 2–3 mm, they intravenously injected PEGylated CH1055 (100 μg) to the mouse and observed NIR fluorescence from the tumor at periodic time points over next 3 days. They observed that 6 h post-injection, the tumor was clearly visible during high magnification of NIR fluorescence (>1200 nm) imaging (Fig. 13.15c). After 24 h, the tumor was clearly visible with a tumor-to-normal tissue ration of 4.25 when using a whole-body imaging system (Fig. 13.15d). In this study, the accumulation of PEGylated CH1055 to a brain tumor was attributed to passive tumor uptake via the enhanced permeation and enhanced (EPR) effect [108]. Recently, Tian et al. reported the fluorescence imaging of a glioblastoma brain tumor by using active tumor uptake of RGD peptide-conjugated NIR dye (IR-BEMC6P@RGD) [76]. They detected strong tumor fluorescence (>1300 nm) through scalp/skull at 12 h post-injection.

Fig. 13.15
figure 15

(a) Color photograph of a nude mouse preceding high-magnification NIR-II imaging, with an outline over suture lines. (b) Graphic representation demonstrating the location of a U87MG orthotopic glioblastoma brain tumor under both the scalp and skull. (c, d) High-magnification NIR-II fluorescence imaging (1200 nm long-pass filter, 800 ms) showing strong tumor fluorescence detectable through both the scalp and skull 6 h post-intravenous injection. Reproduced from ref. [65] with permission from Springer-Nature

Full size image

13.5.3 Cerebrovascular Disorders

NIR fluorescence imaging can be also used for the visualization of the pathophysiological state of blood vessels in sepsis. Thrombosis in cerebral blood vessels is induced by administration of lipopolysaccharide (LPS) to mice (Fig. 13.16a). Heparin is used as an inhibitor of blood coagulation [109,110,111]. The magnified images with the scalp removed (Fig. 13.16b) of blood vessels, showed septic clots (i.e., thrombosis), and the number of clots was increased by administration of LPS. The administration of heparin resulted in the suppression of the number of clots (Fig. 13.16b and c). Figure 13.16d shows the immunohistochemistry for an LPS-administrated brain slice, indicating the formation of clots in cerebral blood vessels. The blood coagulation was quantified by enzyme-linked immunosorbent assays (ELISA) (Fig. 13.16e). Thrombin–antithrombin complex (TAT) is a valid biomarker for disseminated intravascular coagulation [112]. After administration of LPS, averaged TAT values were significantly increased, and the level of TAT was recovered by heparin administration. This study suggests that 2nd NIR fluorescence imaging is useful for the detection of thrombosis in an LPS-injected mouse.

Fig. 13.16
figure 16

(a) Time course of the experimental procedure for LPS and heparin administration. (b) NIR fluorescence images (>1000 nm) of cerebral blood vessels before and after administration of LPS (LPS (−) and LPS (+)) and the image followed by additional administration of heparin (LPS + heparin), with scalp removed. Lower panel shows the magnification of the images shown by red rectangular. Red arrowheads show clots. Scale bars: 1 mm. (c) Statistical analyses of the clots in the cerebral vessels. (d) Immunofluorescence staining of LPS-treated cerebral blood vessels, where anti-fibrinogen antibody (Alexa FLuor 488) was used for staining of fibrinogen. Fibrinogen helps the formation of blood clots. Scale bar: 10 μm. (e) ELISA assays for TAT in blood plasma. Adapted from ref. [41]

Full size image

13.6 Summary and Outlook

In this chapter, we presented recent progress in NIR fluorescent nanoprobes and techniques for brain imaging in the 2nd NIR window. During the past 5 years, a variety of the NIR nanoprobes have been synthesized, and the proof of principle studies on their capabilities for non-invasive brain imaging have been performed. The pioneer work by Dai group using SWNTs has proven the advantages of NIR fluorescence imaging of brain tissues in the 2nd NIR window [31]: deeper penetration, reduced scattering, and low-autofluorescence in deep-tissue imaging. Nanomaterial-based NIR nanoprobes such as Ag2S QDs, PbS QDs, and rare earth-doped nanoparticles have also contributed to prove their superior properties for brain imaging. Although the nanomaterial-based NIR nanoprobes such as SWNTs cannot be applied to clinical fields because of their cytotoxic properties, these NIR nanoprobes should be very useful for the study of cancer cell metastasis, immune/inflammatory response, and stem cell dynamics in the animal level.

For biomedical and clinical applications in humans, the NIR nanoprobes must be rapidly metabolized and excluded from the body. The renal filtration threshold for rapid clearance via urine excretion is known as ca. 5 nm [113]. In this regard, smaller organic dye nanoprobes are suitable as NIR probes for fluorescence imaging in humans. In a few years, researchers have developed several types of NIR-dye nanoprobes that emit over 1000 nm [62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91]. These organic NIR nanoprobes have shown the possible application of NIR intravital imaging to biomedical fields. Notably, recent studies showed that the FDA-approved NIR dye, ICG can be used to perform in vivo NIR imaging over 1000 nm. This finding permits the rapid translocation of 2nd NIR-emitting organic dyes (>1000 nm) in the clinical fields.

In the practical use of the 2nd NIR-emitting organic dyes in the clinical fields, high brightness and safety are necessary as optical contrast agents. At the same time, highly sensitive NIR imaging system with a high spatiotemporal resolution should also be developed for the clinical use such as non-invasive visualization of blood vessels and tumors. At present, except for conventional NIR imaging system (700–900 nm for animals, there is no NIR fluorescence (>1000 nm) imaging system for humans [114]. In the near future, by developing highly sensitive 2nd NIR-emitting organic dyes and intravital imaging system, NIR fluorescence imaging in the 2nd NIR window will be an indispensable tool for non-invasive imaging in biomedical and clinical fields [115].