Letter pubs.acs.org/NanoLett
Cancer-Microenvironment-Sensitive Activatable Quantum Dot Probe in the Second Near-Infrared Window Sanghwa Jeong,† Jaejung Song,‡ Wonseok Lee,† Yeon Mi Ryu,§ Yebin Jung,† Sang-Yeob Kim,§,# Kangwook Kim,†,¶ Seong Cheol Hong,† Seung Jae Myung,§,#,⊥ and Sungjee Kim*,†,‡
Downloaded via KAOHSIUNG MEDICAL UNIV on November 9, 2018 at 14:32:26 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
Department of Chemistry and ‡School of Interdisciplinary Bioscience and Bioengineering, Pohang University of Science and Technology (POSTECH), San 31, Hyojadong, Nam-gu, Pohang 37673, South Korea § Asan Institute for Life Sciences, Asan Medical Center, 88 Olympic-ro, 43-gil, Songpa-gu, Seoul, 05505, Republic of Korea # Department of Convergence Medicine, University of Ulsan College of Medicine, 88 Olympic-ro, 43-gil, Songpa-gu, Seoul, 05505, Republic of Korea ¶ Department of Civil and Environmental Engineering, Korea Army Academy at Young-cheon, Yeongcheon-si, Gyeongsangbuk-do South Korea ⊥ Department of Gastroenterology, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-ro, 43-gil, Songpa-gu, Seoul, 05505, Republic of Korea S Supporting Information *
ABSTRACT: Recent technological advances have expanded fluorescence (FL) imaging into the second near-infrared region (NIR-II; wavelength = 1000−1700 nm), providing high spatial resolution through deep tissues. However, bright and compact fluorophores are rare in this region, and sophisticated control over NIR-II probes has not been fully achieved yet. Herein, we report an enzyme-activatable NIR-II probe that exhibits FL upon matrix metalloprotease activity in tumor microenvironment. Bright and stable PbS/CdS/ZnS core/shell/shell quantum dots (QDs) were synthesized as a model NIR-II fluorophore, and activatable modulators were attached to exploit photoexcited electron transfer (PET) quenching. The quasi type-II QD band alignment allowed rapid and effective FL modulations with the compact surface ligand modulator that contains methylene blue PET quencher. The modulator was optimized to afford full enzyme accessibility and high activation signal surge upon the enzyme activity. Using a colon cancer mouse model, the probe demonstrated selective FL activation at tumor sites with 3-fold signal enhancement in 10 min. Optical phantom experiments confirmed the advantages of the NIR-II probe over conventional dyes in the first near-infrared region. KEYWORDS: Second near-infrared window, quantum dot, activatable probe, cancer microenvironment, molecular imaging vibrational modes that typically makes organic fluorophores unreachable to NIR-II. QDs are also compact in size and allow multiplexed imaging due to their narrow emissions;10 these features cannot be achieved using organic macromolecules (polymer dots)11 or carbon nanostructures (e.g., SWCNs).6 PbSe,12 PbS,13 CdAs,14 Ag2S,15 and Ag2Se16 QDs can emit at the NIR-II region; however, to date FL-activatable QD probes have not been reported for NIR-II. To properly exploit the NIR-II window, the development of NIR-II probes that can provide specific, sensitive signals is critical. Herein, we report an activatable QD probe in NIR-II. Activatable probes, which modulate signals at the event of interest,17 are useful because the signal is less sensitive to the probe distribution, thereby
O
ptical imaging can provide information on many biological systems and affords multiplexing, high sensitivity, noniodization, fast acquisition (real-time imaging), and high spatial resolution.1,2 However, it is mainly limited to systems that can be imaged with small penetration depths (e.g., cells), because of the diffusiveness of light due to scattering. The second near-infrared region (NIR-II; wavelength = 1000− 1700 nm) has recently emerged as a new optical window that can significantly reduce photon scattering and increase imaging depth.3−5 Single-walled carbon nanotubes (SWCNs) has been used for NIR-II fluorescence (FL) imaging such as vascular imaging,6 cerebral imaging,7 nitric oxide sensing,8 and intraoperative surgical tumor area guidance.9 However, SWCNs exhibit rather low photoluminescence quantum yields (PL QYs) and limited tunability at the emission wavelength. Semiconductor nanocrystals, quantum dots (QDs), can be good candidates at the NIR-II because they are free from © 2017 American Chemical Society
Received: October 11, 2016 Revised: January 9, 2017 Published: January 26, 2017 1378
DOI: 10.1021/acs.nanolett.6b04261 Nano Lett. 2017, 17, 1378−1386
Letter
Nano Letters
Figure 1. (a) Absorption and (b) fluorescence spectra of PbS/CdS/ZnS QDs with PbS and PbS/CdS QDs. (c) Time-dependent FL intensity of water-soluble PbS/7CdS QD and PbS/CdS/ZnS QD (NIRQD) under 910 nm, 200 mW/cm2 laser. (d) TEM images of PbS, PbS/CdS, and PbS/ CdS/ZnS QDs. (e) STEM image of PbS/CdS/ZnS QDs. Two-dimensional EDS mapping of the same position in (e) for (f) Cd and (g) Zn.
compared with that in the first near-infrared region (NIR-I; 650−950 nm). PbS/CdS/ZnS core/shell/shell QDs were synthesized and exploited as bright NIR-II emitting fluorophores with PL QY > 10% in aqueous media. We chose PbS QDs as the NIR-II QD emitter due to their brightness (2 orders of magnitude brighter than SWCNs) and relatively well-known postsynthetic treatments (e.g., overcoating and cation exchange) and surface modifications.22,23 The PbS/CdS/ZnS QDs were synthesized by three steps [see Supporting Information (SI)]: (1) synthesis of PbS cores, (2) partial cation exchange to PbS/CdS (core/ shell) QDs, and (3) ZnS shell overcoating. The synthesized PbS cores were typically 7.5 nm in size, the cation-exchange resulted in PbS cores 3.3 nm in radius and the CdS shell 2.1 nm in thickness (∼7 monolayers). The thick CdS shell is important as it provides thermal stability over 200 °C, which is necessary for the subsequent ZnS overcoating. Overcoating with the thin ZnS shell (∼1 monolayer) typically doubled the PL QY, giving the PbS/CdS/ZnS QDs a PL QY of ∼50% in organic phase. The PbS cores typically showed an absorption peak around 1780 nm, and the cation-exchanged PbS/CdS QDs produced a featureless, blue-shifted spectrum (Figure 1a). The CdS shell
significantly reducing the background. Cancer microenvironment-specific activatable probes have been reported in visible and NIR-I regions using tumor-related enzymes such as MMP and cathepsin families.18−20 We designed a protease-activatable NIR-II QD (PA-NIRQD) probe that is activated by the activity of matrix metalloproteinase (MMP), which is a hallmark of the cancer microenvironment.21 Bright and stable PbS/CdS/ZnS (core/shell/shell) QDs that emit at 1200 nm were synthesized, and activatable FL modulators were conjugated to the surface ligands. The modulators contain methylene blue (MB), which quenches FL via photoexcited electron transfer (PET), and a protease-cleavable peptide sequence. The NIR-II QDs were quenched by PET from the QDs to MB when the QDs were in close proximity to MB, and FL was activated upon modulator cleavage, resulting in separation between the emitter and quencher. The peptide sequence was optimized to impart flexibility, facilitating enzyme access and guaranteeing effective quenching upon activation. As a proof-of-concept, we conducted ex vivo imaging of colon cancer in a small mouse model to demonstrate the activatable probe. Selective PL activation at tumor sites was observed with up to 3-fold signal enhancement in 10 min. Imaging in the NIR-II region was also 1379
DOI: 10.1021/acs.nanolett.6b04261 Nano Lett. 2017, 17, 1378−1386
Letter
Nano Letters
Figure 2. (a) Schematic of photoexcited electron transfer (PET) from PbS/CdS/ZnS (Core/Shell/Shell) QDs to MB. (b) Schematic of the surface modification of a QD for the NIR-II activatable probe [PEG = poly(ethylene glycol)]. (c) Displacement of the FL quencher, methylene blue, from a QD by the enzymatic cleavage of MMP-2.
conjugation. For example, reliable QD-distance-dependent quenching could not be obtained using AuNRs and SWCNs as quenchers for NIR-II QDs. Polymer nanoparticles (e.g., conjugated polymer dots) in aqueous media typically have amphiphilic structures with the energy acceptors hidden within the cores,11 making them inefficient distance-dependent quenchers. Thus, we sought a small PET-based quenching molecule for our NIRQD activatable probe. PETs with QDs have been studied for bioprobe28 and optoelectronic applications,29 however mostly at visible range. For example, Medintz et al. reported QD−dopamine conjugates as an intracellular pH sensor, where the pH-dependent population of quinone (as a result of dopamine oxidation) determined the QD quenching by PET.30 NIRQD has quasi-type-II electronic characteristics, making it well-suited for a PET quencher; electrons are facilely delocalized in the CdS shells and holes are confined in the PbS core (Figures 2a and S3).31 This configuration is advantageous for photoexcited electrons to transfer to the quenching molecules on QD surface. Holes remaining in the QD should effectively quench the QD PL. The ZnS shell needs to be thin so that it does not hinder PET and affords brightness and stability. MB was chosen as the PETbased quencher for NIRQD because the LUMO lies 0.35 eV below the conduction edge of PbS core in NIRQD, thus photoinduced electrons in NIRQD could be transferred readily to LUMO level of MB in energetically favorable pathway. PET from bare PbS QD to MB is efficient and fast (lifetime = 0.4 ps),32 and photoexcited electrons in NIRQD should easily spread to access the MB PET quencher. The photoexcited electron in quasi-type-II PbS/CdS QD can spread over entire QD and reach the LUMO level of MB on the surface. No spectral overlap exists between QD emission and MB absorption (Figure S4). PA-NIRQDs were developed by conjugating NIRQD with activatable modulators (AcMs), which have an MB quencher and a protease-cleavable peptide sequence (Figure 2b). MMP was selected as a representative target in the tumor microenvironment; MMP families participate in the early state of tumorigenesis and are increasingly expressed during tumor
had a zinc blende crystal structure, as confirmed by X-ray diffraction (Figure S1). Both PbS/CdS QD and PbS/CdS/ZnS QD showed the PL at 1200 nm (Figure 1b). Transmission electron microscopy (TEM) images confirmed that the PbS cores and PbS/CdS QDs were of the same size, which increased by 0.3 nm after ZnS overcoating (Figure 1d). The distributions of Cd and Zn within each type of QD (Figure 1e− g) were mapped using two-dimensional energy dispersive spectroscopy (EDS), and the EDS elemental analysis indicated a Pb/Cd/Zn ratio of 4:80:16. The ZnS shell thickness estimated from elemental analysis was 0.36 nm, which agrees with the TEM measurements. Hereafter, the PbS/CdS/ZnS QDs are denoted as NIRQD and compared to (i) PbS/CdS core/thick shell (seven monolayers) QDs without ZnS coating (noted as PbS/7CdS QD), (ii) PbS/CdS core/thin shell (two monolayers) QDs (noted as PbS/2CdS QD), and (iii) 4.1 nm PbS QDs that emit at 1200 nm. NIRQD and the three other QDs were ligand-exchanged with dihydrogen lipoic acids (DHLAs) and transferred into the aqueous layer (see SI for details). When introduced into water, the PL of NIRQD was slightly diminished, whereas those of PbS/7CdS QD, PbS/ 2CdS QD, and bare PbS QD decreased significantly to 20%, 3%, and 1% of the initial PL QYs, respectively (Figure S2). The ZnS shell was essential for maintaining a high PL QY and photostability. Under 910 nm (200 mW/cm2) laser exposure, NIRQD retained 82% of its initial PL, whereas that of PbS/ 7CdS decreased to 55% in 60 min (Figure 1c). The photostability of NIRQD was comparable to that of NIR-II emitting SWCNs and nearly an order of magnitude higher than that of indocyanine green.9 The activatable fluorescent probes developed in the visible and NIR-I regions were typically modulated via energy transfers;24−27 however, energy transfer-based PL modulation is difficult in NIR-II probes due to the scarcity of small quenching molecules that absorb strongly in the NIR-II region. Although gold nanorods (AuNRs), SWCNs, and conjugated polymer nanoparticles are candidates for such quenchers, their large sizes make it difficult to modulate the fluorophore using external stimuli, retain colloidal stability, and control 1380
DOI: 10.1021/acs.nanolett.6b04261 Nano Lett. 2017, 17, 1378−1386
Letter
Nano Letters
Figure 3. (a) Absorption spectra of NIRQD and PA-NIRQD conjugated with AcM. (b) Relative PL QY of PA-NIRQD as a function of the molar ratio of AcM to NIRQD. (c) Time-dependent FL activation of PA-NIRQD coincubated with the different concentration of MMP-2. (MMP-In = global MMP inhibitor). (d) Cell viability test for cytotoxicity of PA-NIRQD in HeLa cell by cell counting kit-8 assay.
progression along the adenoma-to-carcinoma sequence.21 The NIRQD surfaces were codecorated with DHLAs and zwitterionic ligands using previously reported procedures,33 which showed slightly negative surface charge (zeta potential ≈ −10 mV). The zwitterionic ligands, which contain quarternary amine and sulfonate, were used to enhance the colloidal stability and to suppress the nonspecific adsorptions.33 The zwitterionic QD surface can keep the probe small and compact, which can promise fast and deep permeations for the probe. The amine end of AcM was carbodiimide coupled with the carboxylic acid in DHLA (Figure 2b; see SI for details). The hydrodynamic size of PA-NIRQDs increased slightly from 10.0 to 11.7 nm after AcM conjugation, confirming the absence of agglomeration during conjugation (Figure S5). The AcM chain consisted of three parts from the amine end to the other end: a short poly(ethylene glycol) (PEG) chain, a peptide sequence of GGPLGVRGGC with an MB maleimide conjugated to the cysteine, and a DDDD poly aspartic acid sequence (Figure 2, see SI for details). The peptide sequence contained a proteasecleavable specific sequence of PLGVR (cleavage site between G and V) that can be recognized and cleaved by specific proteases such as MMPs, which are endopeptidases that break the peptide bonds of nonterminal amino acids. This provided flexibility in the design of AcM, where the cleavable site is sandwiched between two other parts. The part bearing MB is designed to leave the QD surface upon the enzyme-induced cleavage of the peptide sequence (Figure 2c). AcM contains two specifically designed units sandwiching the MMP-cleavable peptide sequence. The first, a short PEG chain (PEG8, a linear tetramer of miniPEG; 8-amino-3,6-dioxaoctanoic acid), was introduced between the QD and the cleavage site to endow spatial flexibility and facilitate the access of approaching enzymes. The second, the poly aspartic acid peptide sequence (DDDD; tetramer of aspartic acid), was attached at the terminus so that the MB containing the leaving group bore a negative charge, facilitating the departure of the leaving group
from the negatively charged QD surface via electrostatic repulsion. The number of AcMs conjugated per QD was varied from 8 to 64, and the PLQYs were monitored; 8 AcMs per QD quenched the PL to 41%, while 40 AcM conjugations reduced the PL to 26% of the initial value (Figure 3b). The control AcM without MB quencher, NH2-PEG8-GGPLGVRGGCDDDD, was conjugated to NIRQD and showed no quenching effect (Figure S6). In subsequent experiments, 40 AcMs were conjugated to PA-NIRQD. After the conjugation, the samples were rigorously purified by repeated dialysis to eliminate free unconjugated AcMs. The absorption spectrum of PA-NIRQD confirmed the coexistence of NIRQD and MB (Figure 3a). The enzyme-induced FL activation of PA-NIRQD was tested by coincubating PA-NIRQD with different concentrations of MMP-2, a representative of the MMP family that shows high activity for the PLGVR sequence cleavage.19,20 Enzyme concentration-dependent FL activation was observed for PANIRQD. At [MMP-2] = 30 μg/mL, the FL of PA-NIRQD increased by over 200% within 10 min and 250% within 60 min (Figure 3c). At 10 μg/mL MMP-2, the increase in FL was 130% 5 min after the enzyme addition. When no MMP-2 was added (0 μg/mL), PA-NIRQD FL remained unchanged throughout the experiment. In a second control, MMP-In (N-isobutyl-N-(4-methoxyphenylsulfonyl)glycyl hydroxamic acid, a global MMP inhibitor) was coadded with 10 μg/mL MMP-2.34 The coaddition of MMP-In completely blocked the FL activation of PA-NIRQD; the FL was similar to that when no enzyme was added. PA-NIRQD cytotoxicity was examined for MMP-2 concentrations of up to 1 μM and times of up to 24 h using three cell types (HeLa, HT29, and CCD841; Figures 3d and S7). No significant cytotoxicity was observed, and the cytotoxicity levels were lower or comparable to those of other NIR-II probes.35,36 We have also performed biotoxicity studies using small animals. Extensive blood biochemical analysis including hematological parameters and typical biochemical 1381
DOI: 10.1021/acs.nanolett.6b04261 Nano Lett. 2017, 17, 1378−1386
Letter
Nano Letters
Figure 4. (a) Time-dependent FL activation of PA-NIRQD and PA-NIRQD-noPEG in the presence of 20 μg/mL MMP-2 enzyme. (b) Gel electrophoresis assay of AcM, AcM-DD, and AcM-noAsp. (c) Relative PL QY of PA-NIRQD, PA-NIRQD-DD, and PA-NIRQD-noAsp. (d) Timedependent FL activation of PA-NIRQD, PA-NIRQD-DD, and PA-NIRQD-noAsp with an AcM/QD ratio of 12.
parameters was performed using mice that had received intravenous injections of 100 μL 0.1 nmol PA-NIRQD PBS solution. One hundred microliters of PBS buffers were used for control injections. The amount of PA-NIRQDs introduced to the mice was the same as the amount used for ex vivo imaging experiments, which simulates the case when the entire amount of PA-NIRQDs applied to the colorectal tissues was completely absorbed into the blood circulation system of the mice. For over a 7 day period, no noticeable difference was found for the weight and behaviors of the QD-treated mice when compared to the control group (Figure S8). After 7 days, the mice were sacrificed for blood collection. Biotoxicity was investigated for red blood cell, white blood cell, platelet, hemoglobin, aspartate aminotransferase, alanine aminotransferase, creatinine, and blood urea nitrogen. They were all remained within the ranges observed for the control group and did not suggest any acute toxicity or damage to renal or hepatic function (Figure S9). To investigate the role of the PEG8 part of AcM, a control AcM without the PEG8 component (AcM-noPEG) was prepared and conjugated to NIRQD to obtain the control PA-NIRQD-noPEG. We have observed no meaningful difference in the quenching efficiencies between AcM and AcMnoPEG to NIRQDs presumably due to the structural flexibility of the PEG8 component. For an example, identical conjugations of 40 AcMs or 40 AcM-noPEGs to NIRQDs showed only 5% difference in the PL QYs (Figure S10). The FL activation experiment with 20 μg/mL MMP-2 was then repeated using PA-NIRQD and PA-NIRQD-noPEG (Figure 4a). Unlike PA-NIRQD, which showed typical FL activation upon enzyme addition, PA-NIRQD-noPEG showed no FL activation. The experiment was repeated using free AcMnoPEGs instead of PA-NIRQD-noPEGs. Interestingly, the free AcM-noPEGs were completely cleaved by the enzymes, as confirmed by MALDI-TOF mass spectroscopy (Figure S11). Enzyme cleavage was only deactivated when the cleavage site was conjugated to QD with no PEG spacer. The enzyme
activity was critically suppressed when the enzyme recognition site was close to the QD surface, presumably due to steric hindrance, preventing enzyme access. Without the PEG8 spacer between the QD surface and the enzyme recognition site, the enzyme had to be significantly structurally deformed to access the recognition site, thereby destroying the enzymatic activity. These results indicate that PEG8 is critical for the enzymeinduced FL activation of PA-NIRQD; this design concept should be applicable to other enzyme-based nanosystems. To evaluate the role of the poly aspartic acid (DDDD) sequence in AcM, we prepared two control AcMs: one with DD instead of DDDD (AcM-DD) and the other with no poly aspartic acid unit (AcM-noAsp). The charge states of AcM, AcM-DD, and AcM-noAsp were confirmed negative, zero (or zwitterionic), and positive, respectively, by gel electrophoresis (Figure 4b). All three AcMs had MB and showed blue bands upon electrophoresis. The three AcMs were conjugated to NIRQDs at varying conjugation ratios, and the resulting FL quenching behaviors were similar (Figure 4c), indicating that the charge state of AcMs affects neither conjugation nor FL quenching efficiency. The enzyme-induced FL activation experiment was repeated using the NIRQD conjugated to the three AcMs (Figure 4d). When conjugated to the AcM with DDDD, PA-NIRQD showed higher FL activation compared to PA-NIRQD-DD and PA-NIRQD-noAsp as early as 10 min after enzyme addition (see SI for details). For PA-NIRQDnoAsp, the FL signal decreased slightly after 10 min, presumably due to the readsorption of the MB-containing leaving groups via electrostatic attraction to the QD surface. This clearly demonstrates that control over the interaction between the fluorophore and quencher is essential in an effective activatable probe. Nanoparticle-based fluorophores often have multiple charges on the surface; thus, the charge state of the quencher (or the modulator) is an important factor to consider. In the case of PA-NIRQD, AcM is electrostatically repulsed from the QD and thus stretched out, which may 1382
DOI: 10.1021/acs.nanolett.6b04261 Nano Lett. 2017, 17, 1378−1386
Letter
Nano Letters
Figure 5. (a) Schematic of ex vivo mouse colon cancer NIR-II FL imaging by spraying with an activatable probe. Colon tissue is resected from an AOM/DSS-treated tumor-bearing mouse and cut longitudinally, and PA-NIRQD is sprayed evenly on the tissue surface. (b) Ex vivo FL imaging of colon cancer by spraying with the activatable probe, PA-NIRQD: color image of the intact mouse colon cancer model and time-dependent FL image after proble spraying. Areas T1−T4 and N are the selected tumor regions and normal colon tissue, respectively. (c) Time-dependent FL signals of the selected region after probe spraying. (d) Ex vivo FL imaging of a normal healthy colon by spraying with PA-NIRQD: color image of the intact mouse colon model and time-dependent FL image after spraying with the probe. Areas T1-T4 and N are the selected tumor regions and normal colon tissue, respectively. (e) Time-dependent FL signals of selected regions A1−A5 after probe spraying. (f) Ex vivo FL imaging of colon cancer by spraying with NIRQDs: color image of intact mouse colon cancer model and time-dependent FL image after probe spraying. Areas T1−T5 and N indicate the selected tumor regions and normal colon tissue, respectively. (g) Time-dependent FL signals of the selected regions after probe spraying.
To demonstrate the ability of PA-NIRQD to respond to the cancer microenvironment, fresh colon tissues were prepared from azoxymethane (AOM)/dextran sulfate sodium (DSS)treated tumor-bearing mice (Figure 5a). The AOM/DSS mouse model has colorectal tumors that express high levels of MMPs.37 Immediately after sacrificing the mice, fresh colon
provide a greater probability of interacting with incoming enzymes. After enzyme-induced cleavage, AcM releases the negatively charged leaving group, which is repulsed from the negatively charged QD surface. This guarantees the effective and rapid FL activation of PA-NIRQD. 1383
DOI: 10.1021/acs.nanolett.6b04261 Nano Lett. 2017, 17, 1378−1386
Letter
Nano Letters
Figure 6. Tissue phantom study of Genhence 750 (NIR-I)-sprayed colon cancer and PA-NIRQD (NIR-II)-sprayed colon cancer. (a) Gray-scaled fluorescence images of a colon cancer model sprayed with NIR-I or NIR-II in 1% intralipid liquid phantom (depth = 0−5 mm). fwhm was calculated along the dotted red lines. (b) fwhm as a function of the depth of the 1% intralipid phantom.
activity and successfully distinguished tumor sites in the colorectal tumor mouse model. Thus, PA-NIRQD has the potential for the image-guided biopsy of cancers. As an activatable probe, PA-NIRQD does not require washing and is less sensitive to probe distribution. As the upregulation of MMP enzymes occurs in a wide range of tumors,39 the PANIRQD probe can also be applied to other types of tumors. In addition, the simple replacement of the cleavable peptide sequence in PA-NIRQD can generate new activatable probes for other protease-related diseases such as atherosclerosis, rheumatoid arthritis, apoptosis, cardiovascular diseases, diabetes, and HIV.40 NIR-II emitting PA-NIRQDs can be particularly useful for whole body imaging which requires deep-tissue penetrations. We are currently pursuing whole body tumor imaging upon intravenous administrations of PANIRQDs into tumor-xenografted mice. We are hopeful that such NIR-II QD imaging can lead to tomographic molecular imaging that provides three-dimensional and quantitative data. We also plan to apply PA-NIRQDs to whole body small animal imaging using patient derived xenografted models for therapeutic efficacy screening. We performed optical phantom experiments to demonstrate the advantages of PA-NIRQD over NIR-I dyes. Intralipid aqueous solution (1 wt %) was used as the liquid phantom to simulate human skin tissue.41 Two colon tissues from the AOM/DSS mouse model were prepared as in the experiments depicted in Figure 5a. One colon tissue was sprayed by a commercially available NIR-I fluorescent dye (Genhence 750, PerkinElmer), and the other was treated with PA-NIRQD as in Figure 5a (see SI for details). The NIR-I dye was not activatable, and the FL peak was at ∼780 nm (Figure S13). The NIR-I dye-treated colon tissue was excited by a 10 mW/cm2, 660 nm light-emitting diode, and the FL image was taken using an NIR-I sensitive Si CCD camera. Without the optical phantom, the tissues treated with both NIR-I dye and NIR-II PA-NIRQD showed similar brightness under the CCD cameras (Figure 6). A thin (