Photoswitching Near-Infrared Fluorescence from Polymer

Subscriber access provided by DALHOUSIE UNIV

Article

Photoswitching Near-Infrared Fluorescence from Polymer Nanoparticles Catapults Signals over the Region of Noises and Interferences for Enhanced Sensitivity Jie Wang, Yanlin Lv, Wei Wan, Xuefei Wang, Alexander D.Q. Li, and Zhiyuan Tian ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.5b10837 • Publication Date (Web): 27 Jan 2016 Downloaded from http://pubs.acs.org on January 31, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Photoswitching Near-Infrared Fluorescence from Polymer Nanoparticles Catapults Signals over the Region of Noises and Interferences for Enhanced Sensitivity Jie Wang,† §Yanlin Lv, †§ Wei Wan,‡ Xuefei Wang, † Alexander D. Q. Li, ‡ Zhiyuan Tian*,†

†School

of Chemistry and Chemical Engineering, University of Chinese Academy of Sciences,

Beijing 100049, P. R. China, E-mail: [email protected] ‡Department

of Chemistry and Center for Materials Research, Washington State University

Pullman, WA 99164, United States. KEYWORDS: Polymer nanoparticles, Photoswitching, Near-infrared fluorescence, Förster resonance energy transfer, Cellular imaging ABSTRACT: As a very sensitive technique, photoswitchable fluorescence not only gains ultrasensitivity, but also imparts many novel and unexpected applications. Applications of nearinfrared (NIR) fluorescence have demonstrated low background noises, high tissue-penetrating ability and ability of reducing photodamage to live cells. Because of these desired features, NIR fluorescent dyes have been the premium among fluorescent dyes and probes with photoswitchable NIR-fluorescence are even more desirable for enhanced signal quality in the

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

emerging optical imaging modalities, but rarely used because they are extremely challenging to design and construct. Using a spiropyran derivative functioning as both a photoswitch and a fluorophore to launch its periodically modulated red fluorescence excitation energy into a NIR acceptor,

we fabricated

core-shell

polymer

nanoparticles

exhibiting photoswitchable

fluorescence signal within the biological window (∼700-1000 nm) with a peak maximum of 776 nm. Live cells synthesize new molecules constantly, including fluorescent molecules, and also endocytose exogenous particles, including fluorescent particles. Upon excitation at different wavelength, these fluorescent species bring about background noises and interferences nearly covering the whole visible region and therefore render many intracellular targets unaddressable. The oscillating NIR-fluorescence signal with on-to-off ratio up to 76 that the polymer nanoparticles display is beyond the typical background noises and interferences, thus producing superior sharpness, reliability, and signal-to-noise ratios in cellular imaging. Taking these salient features, we anticipate this type of nanoparticles will be useful for in vivo imaging of biological tissue and other complex specimens where two-photon activation and excitation are used in combination with NIR-fluorescence photoswitching.

INTRODUCTION Since the applications of fluorescent dyes in biological systems, the near-infrared (NIR) dyes were identified to have superior advantages over traditional visible fluorescent dyes.1-3 The reasons are several-fold. First, cell autofluorescence in the NIR region is weak and consequently NIR dyes potentially offer highly sensitive detections in complex biological systems.4,5 Second, when compared to UV or visible light, the NIR light has deeper penetration of biological tissues, thus more suitable for in vivo fluorescence imaging. Finally, the NIR photons have much less energy when compared with visible, especially UV counterparts, and thus cause much less


2

Page 3 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


photodamage as UV or visible light does. As a result, using NIR for biological and biomedical imaging is a preferred choice.6-13 Most efforts using NIR fluorescence for biological and biomedical imaging are based on the organic dyes with extended π-conjugation features, such as cyanine and phthalocyanine derivatives.14-17 However, aggregation-induced fluorescence selfquenching of the dyes in aqueous milieu owing to their high hydrophobicity and serious crosstalk between the incoming (excitation) light and the outgoing (emission) fluorescence signals arising from the small Stokes shift of the dyes generally hinder their potential of biological and biomedical applications.18, 19 Recently, photoactivation- and photoswitching-based techniques have revolutionized fluorescence imaging and generated unprecedented methodologies in fluorescence imaging that promises either super resolution or ultrahigh sensitivity. For example, fluorescencephotoswitching-based super-resolution imaging dramatically enables visualizing subcellular features with spatial resolution defying the long-standing diffraction limit.20-22 Separately, frequency domain imaging (FDI) can detect molecules that are literally invisible using conventional time-domain integration methods.23 Similarly, many other emerging innovative technologies are being developed to solve important problems that were previously thought difficult, if not impossible, to overcome.24,25 Although improved data-processing algorithms and optical equipment are also needed, fluorescent probes with optimized photoswitching performance continue to be the key-enabling factor in the future development. In terms of the abovementioned advantages NIR-fluorescence, probes with photoswitchable NIR-fluorescence will be highly desirable, particularly for enhanced signal quality in the emerging optical imaging modalities aiming for elucidating signals in biological samples and other complex specimens.26-28 In addition to fluorescence in NIR wavelength region for targeting signal free from common


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 29

fluorescence interferences or noises, high contrast of the photoswitching fluorescence signal is also critical for enhanced detection sensitivity and imaging resolution. However, constructing probes with photoswitchable NIR-fluorescence is still challenging because of the inherent photophysical limitations of photoswitches that are prevalently used in design of photoswitchable fluorescent probes. Specifically, photoswitches developed so far display electronic absorption generally residing in the visible spectrum and are therefore unable to function as energy acceptor or dynamic quencher to modulate the donor’s fluorescence in NIR region via energy transfer process because of energetic uphill. Thus, for photoswitchable fluorescent probes based on Förster resonance energy transfer (FRET) mechanism, using photoswitch as energy acceptor inherently restricts fluorescence photoswitching in the visible region.29-43 For single-chromophore-based photoswitchable probes interconverting between two fluorescent states with noticeable disparity in π-conjugation extents, fluorescence photoswitching is also restricted in the visible region because even the state with relative large π-conjugation merely enables fluorescence emission in the deep-red region at the most. In our initial molecular design, the photoswitch is used as either an acceptor or a donor via FRET mechanism.43 Specifically, we constructed single-chromophore-based photoswitchable fluorescent probes capable of selectively emitting green fluorescence (λmax = 530 nm) and deep-red fluorescence (λmax = 665 nm) upon optical modulation.44,45 More importantly, this type of photoswitchable probe at its red fluorescent state displays relative long lifetime and high quantum yield, which proclaims such state as a potential energy donor/pump for activating the fluorescence of energy acceptor in the NIR region via a energetic downhill process. In this work, we designed and constructed probes possessing photoswitchable NIRfluorescence by transferring excitation energy in the yellow region into the NIR region. This is


4

Page 5 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the first experimental realization that molecular photoswitch can function as energy donor and modulate NIR fluorescence on-and-off, which circumvents the requirement of spectral overlap of absorption band of photoswitches with the emission band of fluorophore and therefore enables photoswitchable fluorescence signals in the quiet NIR region. Other salient features of the resulting probes are their photoswitchable fluorescence signal with the full width at half maximum (FWHM) 93% efficiency. The scattering and reflection from the 375-nm laser were cleaned up by the long-pass filter (Semrock, BLP01488R-25) at the filter cube, while the long-wavelength 561-nm laser was blocked by a notch filter (Thorlabs, NF561-18). The red channel was further purified using a NIR long-pass filter that cuts off at 735 nm that allows NIR fluorescence to pass. The NIR fluorescence image was collected by a highly sensitive Electron-Multiplying Charge Coupling Device (EMCCD) camera (ANDOR Ixon+). The 375-nm laser was switched on for 300 ms to photochemically convert MSP to methoxy merocyanine (MMC), thus turning on NIR fluorescence. Then the 561-nm imaging laser and CCD camera were turned on simultaneously for 100 ms to capture the bright


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

NIR fluorescence image. To switch the MMC to MSP, the 561-nm yellow laser was turned on for additional 220 ms. Then again, the 561-nm imaging laser and CCD camera were turned on simultaneously for 100 ms to capture the dark NIR fluorescence image. The above laser and camera pulse sequences complete a periodic cycle that comprises of both the bright and dark states of the NIR fluorescence. This cycle was typically repeated 25 times to enhance the signal to noise ratio and image quality. Cell Growth and Delivery of NPs. HeLa cells were cultured in the glass-bottom 35-mm MatTek’s culture dishes (MatTek Corp.) filled with Dulbecco's modified eagle medium (DMEM) at 37 °C in a CO2 incubator. At 70-80% confluence, the old DMEM was aspirated out and NPs were delivered into cells by adding 200 µL of the NPs suspension in 1.00 mL DMEM to each culture dish. Then the cell-NPs system was further incubated for 2 hours before the medium containing NPs was removed. The cells with internalized polymer NPs were replenished with fresh DMEM, and typically cultured overnight. Prior to imaging, the live cells transporting NIRphotoswitching fluorescent NPs were washed with PBS buffer three times and finally replenished with phenol-free DMEM for imaging purposes. The MatTek’s culture dishes containing live cells were sealed under the CO2 enriched atmosphere although no new CO2 enriched gas was delivered during imaging. The HeLa cells with internalized polymer NPs in MatTek’s culture dishes were placed in a constant-temperature-controlled device custom-made for the Olympus X81 sample stage. This device was kept the temperature at 37 ˚C throughout the imaging measurements. Photoswitching NIR fluorescence Imaging of NPs in Living Cells. For cell imaging, the same laser system and optics as described in Microscope Design and Measurements were used with the exception documented below. The lasers and CCD camera pulse sequences are similar to


8

Page 9 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


those of imaging NPs on glass with slight adjustments for optimization of the biological samples. By default, the photoswitchable NIR fluorescent NPs are stable in NIR dark state and thus were imaged with the 561-nm for 100 ms synchronized with the CCD camera. Then the photoswitching 375-nm laser was turned on for 300 ms to convert the MSP to MMC and thus turn on the NIR fluorescence. Immediately, the 561-nm imaging laser was switched on together with the CCD camera for 100 ms. To switch the NPs back to the dark-state, the 561-nm yellow laser was turned on for additional 220 ms, thus finally to complete the cycle. Most experiments collected 25 repeated of the above cycle to enhance signal-to-noise ratio and image quality. RESULTS AND DISCUSSION Preparation of photoswitchable NIR fluorescent probes. As illustrated in Scheme 1, the target photoswitchable NIR fluorescent probes were constructed via a radical-initiated microemulsion polymerization approach by incorporating the photochromic components, MSP, and the NIR fluorescent dye, NIR775, into the hydrophobic core of the core-shell polymer NPs with hydrophilic shell. Specifically, MSP moiety was covalently incorporated in the polystyrene (PST) skeletons that were cross-linked with DVB during the polymerization process and the NIR775 dyes were also incorporated into the cross-linked PST matrix and therefore reside in the hydrophobic cavities of the NPs owing to their highly hydrophobic characteristics. Owing to such unique structural feature, MSP and NIR775 components are expected to be more or less evenly incorporated throughout the core part of the NPs with minimized self-quenching. Such core-shell polymer nanostructures have been experimentally validated, confirming that the hydrophobic dyes reside in the hydrophobic cores of the NPs.48 Owing to such core-shell structure, in aqueous milieu the photoactive component MSP and NIR775 are protected in the hydrophobic core of the NPs and their fluorescence is unlikely to be quenched by components of


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

biological milieu that cannot permeate through the hydrophilic shell. Additionally, the hydrophobic PST matrix in the core part of the NPs surrounds the photoactive MSP and NIR775 components, providing another protective layer for them. These as-prepared NPs are easily suspended in water due to their hydrophilic shell, which makes them suitable for live-cell experiment. On the other hand, because the polymer chains are cross-linked, individual particles remain insoluble and thus retain their identities.

Scheme 1. Photochromic MSP was functionalized with a styrene moiety and copolymerized with other monomers into the cross-linked hydrophobic cavities of core-shell polymer NPs. The NIR fluorescent dye, NIR775, was also incorporated into the hydrophobic core of the NPs during the polymerization process due to its hydrophobic nature. Optically switching MSP between its open- and closed-states enables NIR fluorescence on-and-off photoswitching. Transmission electron microscope (TEM) and dynamic light scattering (DLS) techniques were used to acquire the shape and size information of the polymer NPs. Figure 1A displays the TEM image of a representative batch of polymer NPs, which reveals their regular spherical shape with an average diameter of ∼50 nm. DLS measurement of the same NPs sample reports an average hydrodynamic radius of ∼60 nm and a relatively narrow polydispersity. Such


10

Page 11 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


discrepancy in the obtained diameter results is probably attributable to the hydrodynamic swelling effect as the DLS measurement was carried out with the core-shell NPs carrying hydrophilic shell suspended in an aqueous environment while the TEM measurement was made with dry NPs under high vacuum conditions. The zeta potential of the aqueous dispersion containing the as-prepared polymer NPs is approximately -42 mV, suggesting good stability of the NPs dispersion sample. Specifically, their particle size and zeta potential were well maintained and no aggregation was found over one month. Additionally, no appreciable leaching or loss of dyes was observed for NPs dispersed in PBS buffer (pH 7.2) and acidic aqueous milieu (pH 3.5) up to 72 hours, as shown in Figure S-5, suggesting the NIR775 dyes were tightly trapped within the hydrophobic core and did not suffer from dye leaching problem under our experimental conditions.

Figure 1. (A) Representative TEM image of a version of polymer NPs exhibiting spherical particles with average diameter of ∼50 nm. (B) DLS measurement of the same sample reports an average hydrodynamic radius of ∼ 60 nm. Ensemble NIR-fluorescence photoswitching of the polymer NPs. The NIR-fluorescence photoswitching of the polymer NPs is based on the activation/deactivation of FRET process


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

between the optically interconvertible MSP and MMC system, the energy donor, to the NIR fluorescent dye, the energy acceptor. Typically, MSP cannot transfer its excitation energy to its nearby NIR775 dye because MSP’s green fluorescence has poor overlap with the optical absorption of the NIR775 (Figure S-1 in Supporting Information (SI)). When MSP is photoswitched to MMC by 375-nm light, however, the resulting MMC has a strong absorption band centered at 575 nm and emits red fluorescence centered at 665 nm;44,45 this red-emission band overlaps appreciably with the absorption band of the NIR775 dye, thus transferring the excitation energy to NIR fluorescence (Figure 2). The net effect is that exciting at 575-nm yellow light produces NIR fluorescence at 776 nm, a whopping 200 nm shift or 559 meV difference in energy. No currently known dyes have such a large Stoke shift, and therefore the photoswitched NIR fluorescence based on the strategy in Figure 2 is free from common fluorescence interferences or noises generally appearing from 570 nm to 730 nm due to relatively small Stoke shifts. Typically, a 561-nm laser was used to excite the NIR fluorescence; this excitation line is just off the MMC absorption peak (575 nm) but provides good efficiency (Figure 2A). For traditional fluorescence techniques, such an excitation generally generates fluorescence signals as well as interferences simultaneously in the region from 570 nm to 730 nm, here after called region of noises and interference.4, 5 Through FRET process within NPs, our fluorescence signals appear in the NIR region or region of detection (Figure 2B). Because the NP prevent potential NIR interfering dyes or molecules from receiving the excitation energy stored in MMC, the NIR fluorescence is almost solely from the NIR775 dyes residing in the hydrophobic core of the core-shell polymer NPs.48 Fluorescence in the region of noises and interference can be easily filtered away via a long-pass filter to yield nearly interference-free NIR fluorescence.


12

Page 13 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Photoswitchable MMC absorbs at 575 nm (black curve) and emits at 665 nm (red curve); this emission overlaps with the absorption of NIR775 dye (blue curve) (A). Photoswitching of the photochrome using UV and visible light turns the absorption at 575 nm on-and-off, which in turn switches the NIR fluorescence on-and-off while exciting the particles with a 561-nm laser (B). This NIR fluorescence can be turned on-and-off depending on whether the photochrome in the NPs is in the ring-closed spiropyran form MSP or ring-open merocyanine form MMC. In the MSP configuration, 561-nm excitation cannot efficiently excite the NIR dye and thus the NPs remain in the dark state with only residue NIR signals (Figure 2B). In the MMC form, however, 561-nm excitation turns on strong NIR fluorescence ranging from 760 nm to 800 nm and


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

switches the NPs to the bright state. Owing to the core-shell structural feature, the fluorophores (MMC and NIR 775 dye) confined within the hydrophobic cavities with protective shell are isolated from nonradiative decay pathways or electron-transfer pathways originating from collisions with solution components. Additionally, both MMC moiety and NIR 775 molecule reside in the hydrophobic cavity, which enables efficient proximity for FRET and good quantum yield of φ = ∼0.28 for MMC and φ = ∼0.10 for NIR 775, respectively. These factors, together with the spectral overlap between the absorption band with the emission band of MMC component, contributed to efficient FRET from MMC to NIR dye with efficiency up to ∼83%. More importantly, the efficient FRET enabled the NIR fluorescence photoswitching with on-tooff ratio up to 67, a strong contrast that is advantageous in super-resolution imaging20-22, frequency-domain imaging (FDI)23, and high-contrast bioimaging43,49, 50. By reversibly switching MSP to MMC via alternating illumination of UV and visible light, reversible NIR-fluorescence on-and-off photoswitching was achieved. NIR-fluorescence photoswitching of the polymer NPs at single particle level. The 50nm NPs were imaged in the NIR using alternating 561-nm and 375-nm excitations. The dark NPs were photoswitched to the bright NPs by 300-ms illumination of a 375-nm laser while the NIR images were collected in 100 ms under a 561-nm laser excitation using a NIR long pass filter that cuts off at 735 nm (OD≥4). The imaging excitation plus an additional 220-ms 561-nm illumination photoswitched the NIR NPs to the dark state, thus completing one photoswitching cycle. This NIR dark-bright cycle was repeated 25 times with one dark frame and one bright frame in each cycle. The summation of all 50 frames yields the total integrated intensity in time domain (Figure 3A) and Fourier transform produces the corresponding FDI image at the


14

Page 15 of 29

TDI

A

FDI

B 1200

300 800 200 400 100 0 0 dark

bright

dark

bright

dark

C 30 20

bright

dark

bright

dark

10

bright

0

NIR Fl. Int. (cps)

6

D

40

(130,98)

E

(119,98)

30

4

20 2 10 00

10

20

30

40

50

0

0

10

Time (100-ms bin) 200

NIR Fl. Int. (cps)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


20

30

40

50

Time (100-ms bin)

F

FD

100 0 0 800 600

100

200

300

400

G

500

TD

400 200 0 0

100

200

300

400

500

Distance (in pixels)

Figure 3. Single-particle imaging of photoswitchable NIR fluorescent NPs was illustrated in both time (A) and frequency domain (B). The 50-frame time-integrated image (A) and its corresponding frequency domain image (B) are nearly identical, indicating most NPs are photoswitchable and emit modulated bright-dark NIR fluorescence. In panel (C), individual frame (60×60 pixels) acquired as a function of time shows bright-and-dark oscillation mostly in diffraction-limit spots. Plot of NIR fluorescence intensity at the peak of single particles reveal the same on-and-off modulation pattern for well-behaved photoswitching behavior (D). Plot of the background spot only 11 pixel away from the signal in (D) reported no regular periodicity at all (E). NIR fluorescence line profiles passing through the same point in time (A) and frequency (B) domains were depicted in (F) and (G), respectively.


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

photoswitching frequency of 1.19 Hz (Figure 3B). None-switching particles have no intensity in frequency domain and these results confirm that such NIR NPs have good photoswitching behavior. The NIR dark-and-bright pattern was depicted in Figure 3C and the fluorescence intensity trajectory of a single particle is plotted in Figure 3D. Both Figures confirm the photoswitching of the NIR fluorescence. In contrast to the well-behaved fluorescence on-and-off switching that the NPs aqueous dispersion sample exhibits (Figure S-5), the fluorescence intensity trajectory of a single particle (Figure 3D) displays noticeable damping at the “On” state accompanying with elevated level at the “Off” state. Such discrepancy in repeatability of fluorescence switching between the ensemble and single particle measurements, particularly the elevated level at the “Off” state in the latter, likely means that the time allocation for activation, imaging, and deactivation, which was determined with the consideration of balancing the quality of the image and the temporal resolution, deviates from that enabling optimum photoswitching performance. It is also noted that despite such deviation from the optimum switching repeatability, high on/off switching ratio still persisted as shown in Figure 3D, which eventually enables high detection sensitivity as demonstrated in the next section. To demonstrate that the NIR fluorescence oscillation is indeed originated from the NPs, Figure 3E plots a random background pixel, whose intensity trajectory has nothing but random noises. Previously, photoswitching MSP-MMC deep-red NPs (665 nm) has demonstrated that FDI yields signal-tonoise (S/N) ratio about two orders of magnitude over the integrated time-domain imaging (TDI).23 Figure 3F and 3G present the fluorescence intensity profiles of these NIR NPs in frequency domain and time domain, respectively. The S/N ratio for the FDI is 35.7 while the TDI is 55.7. What is the key underpinning effect that enables the TDI to be comparable with FDI?


16

Page 17 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The answer is that FRET catapults the signals from the region of noises and interference to the quiet NIR region.

A

1000

B

300 200

600

100

TDI

200

C

FDI

0

D

Bright field

DAPI

Figure 4. Live HeLa cells were incubated with 50-nm NPs; very weak NIR fluorescence near the cell’s nucleus was detected in TDI (A). However, FDI revealed that these signals carried no photoswitching frequency, thus not from photoswitchable NPs (B). Imaging the same area under bright field (C) and DAPI fluorescence dark field (D) clearly shows that this cell had no measurable photoswitchable NPs endocytosed even though the NPs were relatively small. NIR-fluorescence photoswitching of the polymer NPs in cells. When the 50-nm NPs were incubated with live HeLa cells, no measurable amount of NIR fluorescence from these particles were unambiguously identified in FDI (Figure 4B). Many cells were examined and the results were similar, indicating low efficiency of cellular uptake. Counter-intuitively, polymer NPs with relative large diameter, ∼100 nm, readily went into live cells of the same origin (Figure 5A-F). Again, both TDI and FDI yield nearly identical pattern (Figure 5B and 5C) with the former has slightly higher noises similar to Figure 4. Fluorescence imaging of the DAPI dye (Figure 5D) reveals the nucleus of the cell and the NIR NPs are converging to the perinuclear area as evidenced by the overlay of the DAPI and FDI image (Figure 5E) as well as DAPI, FDI, and bright field images (Figure 5F). These results coupled with data in Figure 4 suggest that NPs did not just penetrate the cells; rather the cells facilitated the transportation of the NPs from


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 29

extra cellular matrix into the perinuclear area of the cells. It has been demonstrated in other group that the manner that the nanoparticles were internalized by cells, and their subsequent intracellular fate, strongly depend on particle size and specifically, in potassium-depleted cells, the smaller particles displayed relatively low cell uptake efficiency as compared to that the larger particles displayed.51 One plausible explanation that larger particles have a higher efficiency to cellular uptake herein is that larger particles contact more area on the cell membrane and thus enhancing the probability of triggering membrane receptors for endocytosis. Fluorescence profiles passing through the same vertical line (at x = 97/300 pixels) are displayed in Figure 5G and 5H. The elevated baseline in the time domain data (pixel 50-220) indicates that residue NIR interferences are present like diffused clouds. In this aspect, the FDI is superior with much flat baseline and some unfocused NPs at different positions along the z-axis. The TDI S/N ratio is 17.4, whereas the FDI S/N ratio is 31.3. Again these values are comparable because of the catapulting effect generated by the FRET from MMC to NIR775 dye. Reliably, when compared with a random point (Figure 5I), the NIR photoswitching NPs can be unambiguously identified because not only they have high NIR intensity, but also their intensities oscillate in sync with the photoswitching frequency of 1.18 Hz (Figure 5J). A

B

TDI 70k

C

FDI 18k

50k 12k 30k 6k 10k 0

BF

D

E

DAPI

F

DAPI+FDI


DAPI+FDI+BF

18

Page 19 of 29

9k

G

TD

NIR Fl. Int. (cps)

NIR Fl. Int. (cps)

40k 30k 20k 10k

X=97

0

50

150

H

3k

X=97

0

250

50

I

50 30 10 0

(224,173) 10

20

150

250


NIR Fl. Int. (cps)

70

FD

6k


NIR Fl. Int. (cps)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


J

1200

(110,258)

800 400 0

30

40

50

0

Time (100-ms bin)

10

20

30

40

50

Time (100-ms bin)

Figure 5. After incubation with live HeLa cells (A), the 100-nm larger NPs were endocytosed by the cell and transported to the perinuclear region as detected in TDI (B) and FDI (C). The nucleus was stained by DAPI (D) and the overlays of DAPI and FDI (E) or DAPI, FDI and bright field images (F) confirmed photoswitchable NPs near the cell’s nucleus. NIR fluorescence line profiles passing through a nanoparticle in time (A) and frequency domain (B) were depicted in (G) and (H), respectively. Plot of the background intensity shows random noises (I), whereas plot of NIR fluorescence intensity at the peak of single particles reveals on-and-off modulation pattern synchronized with the photoswitching lasers (J). FRET mechanism has been widely applied to modulate the donor fluorescence for constructing photoswitching systems. Typically, photochromic components with one of their alternating states displaying intense absorption in visible region were used as the dynamic quenchers (acceptors) to quench (modulate) the donor fluorescence, thus enabling on/off or dualcolor fluorescence photoswitching.29, 43 The spectral overlap of such dynamic acceptor (quencher) with the fluorescence emitter plays critical role in determining FRET efficiency and therefore the photoswitching contrast. Owing to the requirement of spectral match and energetic downhill in FRET, such photoswitched donor fluorescence either locates in the visible region30-42 or displays sufficiently broad emission band spanning over visible to NIR region46. For the former, extremely large visible fluorescence photoswitching on/off contrast can be achieved on the basis


19


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

of efficient spectral overlap between the emission band of visible fluorophore and the absorption band of photoswitch (the low-bandgap quencher). For the latter, however, an intrinsic trade-off between the position of photoswitching fluorescence signal and the spectral match exists. Specifically, such system demands the visible absorption band of the photoswitch (the quencher) to overlap with the NIR emission band thus a broad emission band with band edge extending to the visible region is prerequisite. Consequently, system with such trade-off balanced generally displays broad photoswitchable fluorescence spanning over visible to NIR region and limited switching on/off contrast. Using MMC as actuator for catapulting the excitation energy into NIR acceptor with a whopping shift of 200 nm, NPs developed herein displayed photoswitchable fluorescence signals in the quiet NIR region with

Photoswitching Near-Infrared Fluorescence from Polymer

Recommend Documents