2.1.

Synthesis and Characterization of the Eu-Luminescence-Labeled Nanoagent

The Eu-luminescent complex, $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$, was synthesized as previously reported.^67,70 Briefly, the benzene solutions of $\mathrm{Eu}{(\mathrm{fod})}_3·3{\mathrm{H}}_2\mathrm{O}$ (51.9 mg, 0.05 mM) and dpbt (20.8 mg, 0.05 mM) were mixed under argon atmosphere and stirred for 3 h. The mixture was then evaporated to remove the solvents and further dried in vacuum to obtain the $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ complexes. Poly(methyl methacrylate-co-methacrylic acid) (PMMA-MAA) was used as the matrix material for the fabrication of the Eu-luminescence-labeled polymeric nanoparticles. See Fig. 1 for the chemical structures of the Eu-complex and the polymer. The $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$-doped polymeric nanoparticles in a colloidal solution were prepared by rapid mixing water with a solution of the Eu-complex and the polymer in acetone under reduced pressure in a microchannel mixer, which was followed by removing acetone via evaporation.⁶⁸ Hereafter, the $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$-doped PMMA-MAA nanoparticle will be referred to as Eu@PMMA-MAA [see the inset of Fig. 2(a) for the nanoparticle configuration].

Fig. 1

Chemical structures of $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ [fod: 6,6,7,7,8,8,8-heptafluoro-2,2-dimethyl-3,5-octanedione; dpbt: 2-($N,N$-diethylanilino-4-yl)-4,6-bis(3,5-dimethylprrazol-1-yl)-1,3,5 triazine] and PMMA-MAA [poly(methyl methacrylate-co-methacrylic acid)]. Product number of PMMA-MAA was 376914 (Sigma-Aldrich LLC.).

Fig. 2

DLS and spectroscopic characterization of the Eu@PMMA-MAA nanoagent. (a) The size distribution and average hydrodynamic diameter ($d_{\mathrm{av}}$) measured by DLS. Inset shows the schematic configuration of the nanoagent. (b) UV–Vis absorption (black) and luminescence (red) spectra of a colloidal solution of Eu@PMMA-MAA. Luminescence excitation wavelength was 400 nm.

The average hydrodynamic diameter of the Eu@PMMA-MAA colloidal nanoparticles ($d_{\mathrm{av}}$) was 26.9 nm as determined by dynamics light scattering [DLS; Fig. 2(a)] at 25°C (Zeta potential and particle size analyzer; Zeta plus, Brookhaven Instruments). Figure 2(b) shows the UV–Vis absorption and the luminescence spectra of the colloidal solution of Eu@PMMA-MAA recorded, respectively, on a UV–Vis absorption spectrometer (UV3600PLUS; Shimadzu, Japan) and a fluorescence spectrophotometer (FLS-980; Edinburgh Instruments Ltd., United Kingdom). The action cross section of $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ complexes, defined as the product of the TPA cross section (${\delta }_{\mathrm{TPA}}$) and the Eu-luminescence quantum efficiency at 613 nm (${\mathrm{\varphi }}_{\mathrm{L}}$), ${\delta }_{\mathrm{TPA}}·{\mathrm{\varphi }}_{\mathrm{L}}$, serves as a metric of the two-photon sensitization efficiency. With the colloidal solution of Eu@PMMA-MAA nanoparticles as the sample and a methanol solution of rhodamine B as the reference (${\delta }_{\mathrm{TPA}}^r=152\mathrm{GM}$, ${\mathrm{\varphi }}_L^r=0.7$ $1\mathrm{GM}={10}^{-50}{\mathrm{cm}}^4·\mathrm{s}·{\mathrm{photon}}^{-1}$),⁷¹ we determined the ${\delta }_{\mathrm{TPA}}$ of the PMMA-MAA embedded $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ complexes following the procedures described in Ref. 64. The cell viability test of the Eu@PMMA-MAA nanoagent was performed following the procedures described in Ref. 68. The data analysis was conducted by Student’s $t$-test. The sampling number of each data point was six, and the difference with $P<0.05$ was considered to be statistically significant.

2.2.

TPE-TR Bioimaging Apparatus

The instrumentation details of TPE-TR imaging have been reported elsewhere.⁶⁹ Herein, we provide a brief description (Fig. 3). As shown in Fig. 3(a), the TPE-induced autofluorescence of endogenous fluorophores could be rejected by detecting the Eu-luminescence after the decay of autofluorescence. This was realized by setting an appropriate delay time for autofluorescence rejection ($\delta t$), which was longer than the autofluorescence duration ($<100\mathrm{ns}$) but much shorter than the lifetime of Eu-luminescence (0.84 ms). As shown in Fig. 3(b), the laser beam was shaped with aperture A in a way so that the beam after focusing by CL formed a focal line with homogeneous power distribution.⁶⁹ The focal line was scanned over a horizontal plane by regulating the X-scanning unit, while the imaging optics and the intensified charge coupled device (ICCD) for image sensing remained static. The Z-height of imaging plane was adjustable by vertically moving the animal stage.

Fig. 3

(a) Principle of autofluorescence rejection. $\delta t$ is a time delay comparable to the duration of tissue autofluorescence, which is set prior to the activation of an ICCD for image sensing. $\mathrm{\Delta }{t}_{\mathrm{G}}$ is the gate width of the ICCD detector, which is long enough to accommodate the persistence of Eu-luminescence. See text for details. (b) Optical layout of TPE-TR bioimaging. The femtosecond laser source is an oscillator-regenerative amplifier system (solstice; Spectra-Physics). The X-scanning unit consists of a one-dimensional translational stage (ILS-100LM; Newport) mounted with a concave lens ($L_1$, $f=-50\mathrm{mm}$) for beam expanding, an aperture (A) for beam shaping, a convex lens ($L_2$, $f=350\mathrm{mm}$) for beam collimating, a mirror (M1) for beam steering, and a cylindrical lens (CL, $f=400\mathrm{mm}$) for line focusing. Rectangular M2 directs laser to dichroic mirror, DM, which projects the focused beam as a line onto the objective plane. Eu-luminescence is collected by a lens assembly (LA, $f=260\mathrm{mm}$, 1:5). After passing through optical filters (F: FF01-715/SP and FF02-615/20; Semrock), the Eu-luminescence is directed by mirror M3 to the input optics ($f=105\mathrm{mm}$, 1:2.5; Nikkor, Japan) of the ICCD detector (iStar DH334T; Andor). See Ref. 69 for more details.

With delicate timing management among pulsed excitation, mechanical X-scanning, and image sensing, the push-broom type, line-by-line imaging mode was established [cf. Fig. 4(a)]. The average illuminating power of femtosecond laser with a repetition rate of 500 Hz was set at 350 mW. The time delay for autofluorescence rejection was $\delta t=500\mathrm{ns}$, and the gate width for acquiring a line image was $\mathrm{\Delta }{t}_G=1.8\mathrm{ms}$. The frame time was adjustable between 1 to 4 s depending on the repetition rate of pulsed excitation, and it was 4 s in this work. The TPE-TR apparatus has a lateral resolution of $60\mu \mathrm{m}$ in air, and an axial resolution of $\sim 1.8\mathrm{mm}$ in air or 4.6 to 6.6 mm in an intralipid phantom depending on the imaging depth.⁶⁹

Fig. 4

(a) Illustration of the postural status of a mouse subjected to TPE-TR bioimaging: The focal plane (upper panel) and the FOV (lower panel). (b) Bright-field image and (c) its corresponding TPE-TR image recorded for a mouse model 5 min after caudal vein injection of colloidal Eu@PMMA-MAA nanoagent. (d) Cross-sectional photograph showing the liver tissues of a mouse dissected $\sim 2\mathrm{mm}$ apart from the xiphoid to the thoracoabdominal side. The scale in red represents 1 mm per major division.

2.3.

Mice Model and Pharmacokinetics Imaging

Immune-deficient balb/c nude mice (6 to 8 weeks, 20 to 30 g Charles River Laboratories, China) were used in this work, and they were treated in compliance with the Guide for the Care and Use of Laboratory Animal Resources (1996, National Research Council). Before experiment, a healthy mouse was kept in a box attached with a minitype anesthetic ventilator (Matrx VIP3000). After being anaesthetized with isoflurane vapor, the mouse was mounted on the animal stage equipped with a ventilation mask supplying anaesthetic vapor and oxygen and with an exhaust port connected to a cryogenic trap.⁶⁹ The mouse was injected via caudal vein with colloidal solutions of Eu@PMMA-MAA nanoagent ($0.1\mathrm{mg}·{\mathrm{mL}}^{-1}$ and $300\mu \mathrm{L}$), and the real-time images were recorded every 5 min.

2.4.

FOV Orientation and Time Origin of TPE-TR Imaging

A mouse was mounted on the stage at a desired Z-height [cf. Fig. 3(b)] with the help of an auxiliary optoelectrical positioning system, which was based on a laser triangulation scheme for distance sensing.⁷² The xiphoid of the mouse was taken as the referencing point, which was illuminated with a beam of weak continuous-wave laser and viewed with an auxiliary CCD camera. This auxiliary positioning system combined with the regulation of animal stage ensured the precise location of the TPE-TR imaging plane. To examine the superficial TVs, the imaging plane was placed 3 mm beneath the xiphoid [Fig. 4(a)]. As shown by the bright-field image recorded prior to the TPE-TR imaging, a single FOV could cover the thoracoabdominal area [Fig. 4(b)]. Accordingly, the TV network could be visualized clearly with the TPE-TR imaging [Fig. 4(c)]. To facilitate the subsequent kinetics analysis, we define the timing of caudal vein injection (duration, $\sim 2\mathrm{s}$) as a pseudotime origin.

3.1.

Two-Photon Sensitization Efficiency, Photostability, and Cell Toxicity of the Nanoagent

Figure 5(a) plots the action cross section (${\delta }_{\mathrm{TPA}}·{\mathrm{\varphi }}_{\mathrm{L}}$) of $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ complexes doped in the nanoagent against the near-infrared TPE wavelengths. The ${\delta }_{\mathrm{TPA}}·{\mathrm{\varphi }}_{\mathrm{L}}$ is 34.9 GM at 800 nm, which is close to the maximal value of 42.5 at 840 nm. Since a Eu@PMMA-MAA nanoparticle accommodated about 400 $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ molecules, the overall action cross section of the nanoagent was estimated to be $\sim 1.5\times {10}^4\mathrm{GM}$ at 800 nm. Figure 5(b) shows the time evolution profile of the Eu-luminescence from the Eu@PMMA-MAA nanoparticles sequestered in the liver tissues taken from a mouse, which before sacrifice was administered with the colloidal nanoagent via tail-vein injection. The luminescence intensity dropped by $\sim 3\%$ over the time course of 2 h, indicating an excellent photostability of the nanoagent, which is indispensable for the kinetics analysis based on luminescent labeling. Figure 5(c) shows the viability for three different lines of cells including the liver hepatocellular carcinoma (HepG-2), the Hela, and the human embryonic kidney 293 (HEK 293) cells, which were incubated with the colloidal nanoagent at different concentration. It is seen that, after 24 h, the cell viability remained 75% to 90% for different types of cells, indicating the rather low cell toxicity of the nanoagent.⁷³ In addition, no abnormal physiological reaction was observed for the mice during the TPE-TR imaging experiment, proving the negligible biotoxicity of the polymeric nanoagent.

Fig. 5

(a) One-photon (solid line) and two-photon (open circle) luminescence excitation spectra of $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ in PMMA-MAA nanoparticles. (b) Photostability of the nanoparticles sequestered in the liver tissues from a mouse. (c) Viability of HepG-2, Hela, and HEK 293 cells after incubation with the colloidal Eu@PMMA-MAA nanoagent for 24 h.⁶⁸

3.2.

Autofluorescence Suppression and Imaging Depth

It is seen in Fig. 6 that, when the delay time for autofluorescence rejection was set as $\delta t=0\mathrm{ns}$, the view field was overwhelmed by the autofluorescence. This implies that, under femtosecond laser excitation, the TPE-induced fluorescence from endogenous fluorophores may interfere seriously with the Eu-luminescence images. However, the autofluorescence was drastically suppressed by setting a $\delta t$ of 5 to 20 ns. Further prolonging $\delta t$ could completely reject the biological autofluorescence, e.g., setting $\delta t$ at 50 to 500 ns essentially did not change the image brightness. The negligible decrease in Eu-luminescence intensity in $\delta t=50$ to 500 ns is due to the long persistence of Eu-luminescence (lifetime, 0.84 ms) and the much shorter lived autofluorescence (lifetime, a few tens of nanoseconds). Besides the autofluorescence rejection, the TPE-TR bioimaging is nearly free from the effect of laser heating, which is benefited from the low dose of laser exposure, the spatially decentralized energy distribution, and the effective thermal dissipation in a living body.⁶⁹ The heating free performance is crucial for deriving the pharmacokinetics information under normal physiological conditions.

Fig. 6

Bright-field image and corresponding TPE-TR images recorded with the indicated delay time for autofluorescence rejection [$\delta t$, cf. Fig. 3(a)]. The mouse model was injected via caudal vein with a colloidal solution of Eu@PMMA-MAA (0.1 mg/mL, $300\mu \mathrm{L}$), which was sacrificed 10 min after the injection. The scale bar of 6 mm in the bright-field image applies to the TPE-TR images.

In the case of living tissues, the imaging depth and spatial resolution are dependent on the light wavelength, the absorption, and scattering properties of tissues and the brightness of the labeled objects.^43,47 As described below, we have estimated the TPE-TR imaging depth based on the anatomical structure and the model of subcutaneous tissue imaging. First, we measured the anatomical depth of the liver in proximity to the xiphoid [Fig. 4(d)]. In viewing the fact that the liver sequestrated with the nanoagent in a high concentration was distinctly visualized by TPE-TR imaging (vide infra), we roughly estimated the imaging depth to be slightly larger than 1 mm. To be more precise, we further prepared a bulky PMMA-MAA particle (projection section, $0.9\times 1.9{\mathrm{mm}}^2$), which was doped with $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$ complexes at a concentration of $0.01\mathrm{mg}·{\mathrm{mL}}^{-1}$. This concentration was one-tenth of that of the colloidal solution used for the caudal vein injection, so as to mimic the sequestration of nanoparticles in subcutaneous tissues. We then covered the bulky particle with individual pieces of mouse lash meat of different thicknesses and accordingly recorded the transdermal TPE-TR images [cf. Fig. 7(a)]. It is seen that the original morphology of the labeled particle (control) was retained at an imaging depth of $D=0.9\mathrm{mm}$. However, it became blurry yet distinctly recognizable with increasing $D$ up to 2.3 mm. Interestingly, the imaging quality does not alter appreciably upon varying the subcutaneous tissue thickness from 1.3 to 2.3 mm [Fig. 7(b)].

Fig. 7

(a) Transdermal TPE-TR images of a $\mathrm{Eu}{(\mathrm{fod})}_3\mathrm{dpbt}$-doped PMMA-MAA particle (projection section, $0.9\times 1.9{\mathrm{mm}}^2$ covered with individual pieces of mouse lash meat of known thickness ($D$). Control: the TPE-TR image of the uncovered particle. The scale bar of 1 mm in control applies to the other TPE-TR images. (b) Semilogarithmic plot of the total Eu-luminescence intensity against tissue thickness.

3.3.

Characterization of the TVs as Visualized by In Vivo TPE-TR Imaging

Noninvasive and intravital TPE-TR images of three mice were recorded 5 min after the caudal vein injection of colloidal Eu@PMMA-MAA nanoagent [Figs. 8(a)–8(c), right column]. These images were obtained by placing the scanning plane 3 mm beneath the xiphoids of the mice. Despite the individual difference, each TV network consists of a pair of axial TVs (ATVs) along with the tributary and lateral TVs (LTVs) between the ATV pair. From the cross-sectional distribution profiles of Eu-luminescence intensity [cf. Fig. 8(b), inset], the ATV and LTV diameters, respectively, are determined to be 180 to $460\mu \mathrm{m}$ and 130 to $350\mu \mathrm{m}$ for the mice. These dimensions are in agreement with the TV diameters characterized by the use of anatomy-assisted optical microscopy.^24,54 From the axillary to the femoral side, the ATVs become slightly thinner, whereas the spatial distribution of LTV thickness is heterogeneous. In Fig. 8, it is seen by comparing the TPE-TR images that some of the LTVs are interconnected, but the amount and patterns of interconnection are different among the mice. Here, we note that the LTVs are associated with collateral and more tenuous veins, which were not resolved with the present spatial resolution. The polymeric nanoagent in such higher-order veins, including venules and blood capillaries, was in substantially lower concentration compared to that in the principal LTVs, and hence appeared as a weak luminescent background.

Fig. 8

(a)–(c) Bright-field images (left column) of three individual mice and the corresponding TPE-TR images (right column) recorded 5 min after caudal vein injection of colloidal Eu@PMMA-MAA nanoagent. The mouse bodies were aligned to render a similar orientation of the TV networks. Insets in the TPE-TR image of (b) show the representative cross-sectional distribution of Eu-luminescence intensity across the LTVs indicated by the red arrows. The solid lines in red were obtained by fitting the intensity profiles to a Gaussian model function, and the full width at half maximum was taken as the vessel diameter.

For a healthy mouse, it takes about a minute for the tail-vein administered nanoagent to be circulated into the ATVs as demonstrated by a recent study using real-time NIR-II optical bioimaging.⁵⁶ In this work, the TPE-TR scanning was started 2 to 5 min after tail-vein injection, a time when the concentration of nanoagent in TV bloods increased approximately to the maximum. Afterward, the blood clearance process began, which led to a systematic decay of the luminescent contrast of the TPE-TR images (vide infra). With the general geometric features of TV network in mind, we will come to examine the transport and clearance processes of the polymeric nanoagent.

3.4.

Noninvasive and Real-Time TPE-TR Imaging of the TV Networks

To investigate the transport and clearance processes of the nanoagent in the peripheral TV blood streams, we placed the scanning plane 3 mm below the xiphoid of a mouse model. Figure 9 shows the real-time TPE-TR images at representative timing. The TPE-TR image at 2.5 min renders a distinct TV network along with a weak and broad Eu-luminescent background, which originated from the sparsely distributed nanoagent in the higher-order venules and blood capillaries. In addition, the femoral artery of the left hind limb was also visualized ($a\text{-}1$) despite optical defocusing. As time elapsed, the image of TV network became considerably dimmer at 35 min and eventually became unrecognizable after 60 min. In parallel, the Eu-luminescence in the region immediately beneath the xiphoid, $a\text{-}2$, as well as in the abdominal region, $a\text{-}3$, became gradually stronger. Note that the Eu-luminescent area $a\text{-}2$ was the reminiscence of nanoagent accumulation in the liver, because the upper layer of the liver was sensed by TPE-TR imaging with the present imaging depth [cf. Figs. 4(d) and 7]. The high capacity of liver sequestration is to be confirmed by the TPE-TR imaging analysis of dissected organs (vide infra). On the other hand, the relatively weak and long-lasting Eu-luminescence in area $a\text{-}3$ is presumably attributed to the retention of nanoagent in the milky spots, which are submillimeter lymphoid tissues consisting of macrophages and other lymphocytes on the omentum majus.⁷⁴

Fig. 9

Bright-field image of a living mouse, and the corresponding real-time TPE-TR images at the indicated time after caudal vein injection of colloidal Eu@PMMA-MAA nanoagent (see Video 1 for more details of spatiotemporal evolution). The outlines a-1, a-2, and a-3 in irregular shapes in the 2.5-min image are the ROIs indicating the respective domains of femoral artery, liver, and abdomen as visualized by TPE-TR imaging. The scanning plane was placed 3 mm beneath the xiphoid. The spatial bar in the bright-field image holds for the TPE-TR images (Video 1, MPEG, 4.9 MB [URL: https://doi.org/10.1117/1.JBO.24.6.066009.1]).

Herein, we comment on the spatiotemporal evolution of the image patterns in the $a\text{-}2$ region (Fig. 9). From 2.5 to 20 min, the asymmetrical and diffusive luminescence image in $a\text{-}2$ was slightly upper shifted and transformed from vertical into horizontal orientation. Such spatiotemporal evolution is most likely due to the redistribution of the initially captured nanoagent among different lobes of the liver. We also would like to comment on the axial resolution of TPE-TR imaging. We have recorded, for the same mouse used for collecting the TPE-TR images as shown in Fig. 9, a complementary set of TPE-TR images by focusing on the femoral artery of the left hind limb. This was done in the same TPE-TR experiment by alternatively placing the scanning plane at 3 mm and at 10 mm, i.e., by targeting on the TV network and the femoral artery alternatively. The TPE-TR images focused on the femoral artery are shown in Fig. 10. It is seen that, with a 7-mm displacement of the focusing depth, the TV network visualized clearly in Fig. 9 became nearly invisible, implying that the axial resolution is substantially better than 7 mm. For the peripheral TVs at a submillimeter depth, the lateral resolution is $\sim 80\mu \mathrm{m}$ as determined from the cross-sectional distribution of Eu-luminescence intensity of the finest LTVs (cf. Figs. 8 and 9). However, as mentioned in Sec. 3.2, the lateral resolution deteriorates considerably at an imaging depth exceeding 1 mm. To improve the imaging depth and the associated spatial resolution, the adoption of NIR-II luminescence labels and/or confocal imaging geometry would be helpful. At the present stage, our TPE-TR imaging modality with a 1 to 2 mm imaging depth is potentially useful for the detection of subcutaneous tumor models and peripheral vasculatures as OCT does.^21,52

Fig. 10

Bright-field image of a living mouse, and corresponding in vivo real-time TPE-TR images at the indicated timing after caudal vein injection of colloidal Eu@PMMA-MAA nanoagent. The outlines $a\text{-}1$, $a\text{-}2$, and $a\text{-}3$ in irregular shapes in the 2.5-min image are the ROIs indicating the respective domains of femoral artery, liver, and abdomen as visualized by TPE-TR imaging. The scanning plane was placed 10 mm beneath the xiphoid. The scale bar of 6 mm in the bright-field image holds also for the TPE-TR images. This set of images is complementary to that in Fig. 9, i.e., they were recorded for the same mouse by alternatively changing the imaging depth.

The aforementioned transport and clearance dynamics in $a\text{-}1-a\text{-}3$ regions are more clearly illustrated by the Eu-luminescence evolution profiles shown in Fig. 11(a). It is seen that the nanoagent concentration in both $a\text{-}2$ and $a\text{-}3$ domains increase in 2.5 to 10 min (cf. Fig. 9). This rising phase reflects the process of nanoagent uptake, which is followed by a clearance phase as indicated by the subsequent luminescence decay. The luminescence fluctuation in 20 to 80 min of the $a\text{-}2$ kinetics was due to the change in the postural and/or the physiological conditions of the mouse as corroborated by Video 1. On the other hand, the $a\text{-}1$ kinetics in Fig. 11(a) shows that the nanoagent concentration in the femoral artery decreased monotonically to a background level.

Fig. 11

(a) Time evolution profiles of Eu-luminescence in the ROIs of $a\text{-}1-a\text{-}3$ in Fig. 9. (b) Spatial distribution of the Eu-luminescence decay time constant (clearance time constant, in min) along with the apparent clearance efficiency in parenthesis [amplitude ratio defined by Eq. (2), in %]. These values are derived from the Eu-luminescent kinetics of the ATVs and LTVs by fitting them to a monoexponential decay function, Eq. (1). Inset shows a pair of representative TV kinetics along with fitting curves (red solid). (c) Normalized Eu-luminescent kinetics of the ATVs and LTVs as indicated by the irregular ROIs adjacent to the numerals as depicted in (b).

To quantify the spatiotemporal evolution of the nanoagent in the TV network, both ATVs and LTVs were divided into the segmental regions of interest (ROIs) as shown in Fig. 11(b), and the corresponding Eu-luminescence kinetics curves are shown in Fig. 11(c). Note that some of the luminescence kinetics in Fig. 11(c) as derived from the proximity of $a\text{-}2$ and $a\text{-}3$ regions exhibit higher background, which are due to apparent high retention efficiency in these regions. The time evolution profile of each local ROI could be well accounted for by a monoexponential decay function:

Figure 11(b) shows the spatial distribution of the clearance time constant along with the corresponding apparent clearance efficiency in parenthesis. It turns out that the ATVs and the LTVs exhibit a similar clearance time constant, 25 to 29 min, which is marginally longer than the clearance time constant of the femoral artery ($\sim 25\min$). In addition, the apparent clearance efficiency, ${\eta }_{\mathrm{c}}=80\%$ to 90%, is also similar for the entire TV network. Furthermore, the wide spread Eu-luminescence background exhibited a clearance time constant of $\sim 30\min$ and an apparent clearance efficiency of $\sim 80\%$ as derived from the Eu-luminescence kinetics in the ROIs in proximity of ATVs and LTVs (data not shown). Therefore, the TV network and the wide spread luminescent background all showed a rather high clearance efficiency. On the other hand, analysis of the $a\text{-}2$ and the $a\text{-}3$ kinetics in Fig. 11(a) derived the clearance time constant (apparent retention efficiency) of $\sim 100\min$ (32%) for the liver domain ($a\text{-}2$), and of $\sim 50\min$ (38%) for the abdomen domain ($a\text{-}3$). The relatively long clearance time constant and high retention rate of the liver domain ($a\text{-}2$) are to be attributed to the liver sequestration effect, whereas similar kinetics behavior of the abdominal domain ($a\text{-}3$) are presumably due to the retention of nanoagent in the milky spots on the omentum majus.⁷⁴

3.5.

Distribution of the Polymeric Nanoagent in the Internal Organs

The blood flow in the principal TVs and their tributary vessels plays an important role in the transport, clearance, and residence of polymeric nanoagents. To examine the distribution of the nanoagent in the internal organs accompanied the TV-blood clearance, the same mouse used for the in vivo pharmacokinetics analysis was sacrificed immediately after the TPE-TR imaging, and six different organs were accordingly dissected and subjected once again to TPE-TR imaging. The in vitro TPE-TR images are shown in Fig. 12(a). Judging from the Eu-luminescence brightness, the liver obviously exhibits the highest capacity of nanoagent deposition.

Fig. 12

(a) TPE-TR images of the internal organs dissected from the same mouse as that used for the intravital TPE-TR imaging (cf. Fig. 9). The image brightness was adjusted by the indicated multiplication factors. The scale bar of 6 mm of the heart image applies to the other images. Broken lines outline the corresponding bright-field images. (b) Comparison of the overall Eu-luminescence intensity for the organs shown in (a).

Figure 12(b) compares the total Eu-luminescence intensity of the internal organs displayed in Fig. 12(a). It is seen that the capacity of nanoagent sequestration of liver is $\sim 15\text{-}\mathrm{folds}$ of the spleen or the lung, $\sim 140\text{-}\mathrm{folds}$ of the kidney, and $\sim 550\text{-}\mathrm{folds}$ of the heart. Besides the absolute volumes, the vasculature characteristics are considered to be mainly responsible for the different retention capacities of the organs. The rather high uptake capacity of the liver, as well as the spleen despite a smaller volume, can be attributed to the abundant artery and venous vessels, and especially to the sinusoids at their vascular endings.^75,76 The lung and the kidney are both rich in blood capillaries, and the difference in the retention capacities is likely attributed to the difference in their bulkiness. The much lower uptake capacity of the heart is obviously due to the much faster blood flow in the cardiac vessels and the relatively less tributary vasculatures. In contrast to the internal organs, no polymeric nanoagent was detected in the brain tissues, which is obviously due to the effective blood–brain barrier. Taken together, our results prove that the liver is the major organ for the nanoagent deposition during the 2-h circulation in blood vasculatures.