Submitted to

���1������1��

DOI: 10.1002/adfm.((please add manuscript number)) Combined optical and MR bioimaging using rare-earth ions doped NaYF4 nanocrystals

Rajiv Kumar, Marcin Nyk, Tymish Y. Ohulchanskyy, Christopher A. Flask and Paras N. Prasad*

[*] Prof. Paras N. Prasad, Dr. R. Kumar, Dr. M. Nyk, Dr. T. Y. Ohulchanskyy

Institute of Lasers, Photonics and Biophotonics, Department of Chemistry State University of New York, Buffalo, NY 14260, USA E-mail: pnprasad@acsu.buffalo.edu Dr. C.A. Flask Case Center for Imaging Research, Case Western Reserve University Cleveland, OH, USA

[**] This work was supported by grants from the National Institute of Health (R01CA119358-01 and RO1CA104492) and the John R. Oishei Foundation. We thank Dr. I. Roy and Dr. D. Sukumaran for useful discussions. Supporting Information is available online from Wiley InterScience.

Keywords: Upconversion nanophosphors (UCNPs), downconversion nanophosphors (DCNPs), mutimodal nanoprobes, optical imaging, magnetic resonance imaging

ABSTRACT

We report novel nanoprobes for combined optical and magnetic resonance (MR) bioimaging.

Fluoride (NaYF4) nanocrystals (20-30 nm size) co-doped with the rare earth ions Gd3+ and

Er3+/Yb3+/Eu3+ have been synthesized and dispersed in water. An efficient up- and

downconverted photoluminescence (PL) from the rare-earth ions (Er3+ and Yb3+ or Eu3+)

doped into fluoride nanomatrix allows optical imaging modality for the nanoprobes.

Upconversion nanophosphors (UCNPs) show nearly quadratic dependence of the

photoluminescence intensity on the excitation light power, confirming two-photon induced

process and allowing two-photon imaging with UCNPs with low power continuous wave

(CW) laser diode due to the sequential nature of the two-photon process. Furthermore, we

have modified both UCNPs and downconversion nanophosphors (DCNPs) with

Submitted to

���2������2��

biorecognition biomolecules, such as anti-claudin-4 and anti-mesothelin, and showed in vitro

targeted delivery to cancer cells using confocal microscopy. The possibility to use nanoprobes

for optical imaging in vivo was also demonstrated. We have also shown that Gd3+ co-doped

within the nanophosphors imparts strong T1 (Spin-lattice relaxation time) and T2 (spin-spin

relaxation time) for high contrast MR Imaging. Thus, the nanoprobes, based on the fluoride

nanophosphors doped with rare earth ions have been shown to provide the dual modality of

optical and magnetic resonance imaging.

1. Introduction

Multimodal bioimaging is a new frontier in biology and medicine [1-3], which combines

more than one imaging modalities such as optical [4], nuclear [5, 6], ultrasound [7, 8], and

magnetic resonance imaging (MRI) [9-10] modalities. Optical imaging provides the highest

sensitivity and spatial resolution for in vitro imaging, is inexpensive, robust and portable to be

used for in vivo imaging, but still lacks the full capability to obtain anatomical and

physiological details in vivo. At the same time, optical imaging is the only technique among

all the medical imaging techniques which can provide cellular or molecular level information

with almost single molecule sensitivity [3, 11]. On the other hand, magnetic resonance imaging

provides an excellent spatial resolution and depth for in vivo imaging, providing exceptional

anatomic information, but suffers from limited sensitivity and lacks resolution for imaging at

the cellular level. Combination of MRI and optical imaging can lead to the development of

new approaches that will bridge gaps in resolution and depth of imaging between these two

modalities. Potential benefits of combined optical and magnetic resonance imaging, along

with recent advances in engineering of the nanoscale materials, have already stimulated a

development of the hybrid magnetic fluorescent nanoprobes for imaging in vitro and in

vivo.[12-16] Use of nanoparticles as molecular specific reporters offers an increased signal over

Submitted to

���3������3��

traditional probes as well as easier conjugation chemistries.[17-19] Modifying the surface of

these multimodal nanoparticles with targeting agents can improve contrast and provide

information about specific biomarkers in both modalities. [20-22]

Highly dispersible nanocrystals of rare-earth compounds, such as oxides [23], phosphates [24],

fluorides [25], vanadates [26], and sulfides [27], have recently become a new focus of research

due to their unique optical properties. The most interesting feature of lanthanide luminescence

spectra are their sharp spectral lines much resembling the spectra found in the case of free

ions. The lanthanide ions also demonstrate long (submillisecond or millisecond range)

photoluminescence (PL) lifetime, [28] thus providing capability of time-gated detection for

autofluorescence suppression. [29] Rare-earth fluoride compounds, AREF4 (A=alkali; RE=rare

earth), have attracted interest as a matrix for lanthanide ions in different luminescence

applications. In particular, NaYF4 is considered among the most efficient host material for

upconversion phosphors, which exhibits visible emission upon infrared (IR) excitation when

doped with Yb3+ and Er3+/Tm3+ ions.[30] Synthesis of colloidal hexagonal nanocrystals of

NaYF4 doped with various upconversion RE3+ ions has been extensively reported [31-33].

Incorporation of luminescent RE3+ ions into NaYF4 nanocrystals for downconversion

luminescence has been also shown [34], when the emission wavelength is red shifted relatively

to the excitation wavelength. Chatterjee et al [35] have recently demonstrated use of

upconversion nanophosphors (NaYF4: Er3+, Yb3+) for in vitro imaging of cancer cells and in

vivo imaging in tissues, claiming such advantages of these nanoprobes as absence of photo-

damage to living organisms, very low auto-fluorescence, high detection sensitivity, and high

light penetration depth in biological tissues.

Here, we report the synthesis of a new generation of the functionalized fluoride

nanophosphors of size <50 nm, which form stable aqueous dispersion that can be used for

bimodal imaging: optical and MR imaging. Down- and upconverted luminescence from the

Submitted to

���4������4��

rare-earths ions (Eu3+, Er3+ and Yb3+) doped into nanosized fluoride matrix allows optical

imaging modality for the nanoprobes and co-doping with Gd3+ provides high contrast MRI

modality. These nanoparticles are devoid of the drawbacks inherent to other optical imaging

probes, namely the broad emission and non-photostability. Furthermore, upconversion

nanophosphors demonstrate nearly quadratic dependence of the PL intensity on the excitation

power, which provides benefits of conventional two-photon induced imaging, such as

intrinsic 3D localization of two-photon excitation and low photobleaching and photodamage

of the sample outside the excitation volume. [3, 36, 37] Use of UCNPs does not require high

photon intensity, which is necessary for two-photon excitation of organic dyes or quantum

dots used for two-photon microscopy. Because of existence of metastable intermediate energy

level in nanophosphors, an upconversion process involves the sequential absorption of two

photons, as opposed to simultaneous one in case of two-photon absorption, when intermediate

level is virtual. This makes upconversion process much more efficient in a comparison to two-

photon absorption [38] and allows excitation and imaging of unconversion nanophosphors

using low-power and continuous wave (CW) laser diodes, which are inexpensive and ready

available, in contrast to femtosecond pulsed lasers needed for simultaneous two-photon

excitation.

Keeping in mind that gadolinium (Gd3+) is a paramagnetic relaxation agent extensively used

in MRI, [39] we have co-doped the fluoride nanophosphors also with Gd3+ ions, thus

introducing MRI modality. We have shown that relaxivity, which is the measure of an MRI

agent’s ability to shorten the inherent water-proton relaxation times and the quantitative

measure of the MRI contrast agent efficacy, [40] is higher for the Gd3+ doped nanophosphors

than for the gadolinium chelate compounds currently used in conventional MRI. [37] We

believe that use of such contrast agents can boost MRI signal and improve contrast at a lower

dosage. It is worth also noting that most Gd3+-based contrast agents currently used in clinical

Submitted to

���5������5��

medicine are fluid agents that are distributed extracellularly and excreted exclusively via the

kidneys. [40] Because of its toxicity, the hydrated Gd3+ ion is sequestered by chelation to

reduce potential toxic effects. However, even for chelated Gd3+-complexes, the release of the

metal ion in vivo during metabolic processes and the subsequent toxicity are pertinent issues.

[41] Gd3+ ions doped into rigid matrix of fluoride nanocrystal, as fabricated here, can avoid this

release issue. In addition, the dual imaging modality of the nanoprobes, based on the fluoride

nanophosphors doped with rare earth ions for optical and magnetic resonance imaging can

provide complementary anatomic, functional, and molecular information. We have also

demonstrated active targeting of pancreatic cancer cells by using anti-claudin 4 and anti-

mesothelin as targeting biomolecules conjugated to the nanophosphors. To the best of our

knowledge, the synthesis and biomedical application of water-soluble hybrid NaYF4:RE3+,

Gd3+ nanoparticles, combining both optical and MR imaging modalities, has not been

reported before.

2. Results and discussion

NaYF4 nanophosphors co-doped with various rare earth ions were synthesized with a size

distribution within range of 25-30 nm as shown by the TEM image in Figure 1, A. The size of

the water dispersible nanophosphors remains practically the same after the ligand exchange

process, as can be seen in Figure 1, B. Dispersions of these nanophosphors demonstrate good

colloidal stability, optical transparency and produce intense photoluminescence (Figure 2).

Submitted to

���6������6��

Figure 1. Representative TEM images of NaYF4 nanophosphors doped RE3+ ions.

Samples for TEM were prepared from chloroform (A) and water (B) dispersions.

Figure 2. Colloidal dispersions of the NaYF4 nanophosphors co-doped with Gd3+ and RE3+

ions demonstrate good stability, optical transparency and efficient photoluminescence. A, B-

UCNPs (NaYF4:10%Yb3+,2%Er3+,10%Gd3+) in CHCl3; C, D-DCNPs

(NaYF4:10%Eu3+,10%Gd3+ ) in CHCl3. PL was excited with 975 nm (B) and 365 nm (D).

To investigate how different ratios of rare earth co-dopants influence on the optical/MRI

properties of the nanophosphors, we have synthesized up-conversion NaYF4 nanophosphors

co-doped with three ratios of the lanthanides: 18 % of Yb3+ and 2 % of Gd3+; 15 % of Yb3+

and 5 % of Gd3+; 10 % of Yb3+ and 10 % of Gd3+. The percentage of Er3+ remained the same

for all samples, 2%. The results for the steady-state and time-resolved absorption and

A B

A B C D

Submitted to

���7������7��

photoluminescence spectroscopy for upconversion nanophosphors (NaYF4, co-doped with

Gd3+, Er3+ and Yb3+) are shown in Figure 3. The absorption spectra of the dispersions show

the presence of Yb3+ ions doped into nanocrystal matrix by peak in the range of ~970 nm. The

presence of Er3+ ions is indicated by the absorption peaks at around 520 and 650 nm. [42] As

one can see, the water dispersions of the nanophosphors demonstrate higher level of light

scattering, apparently caused by the attachment of the ligands and resulted in an appearance of

significant “background” in absorption spectra. It is worth noting that we have not observed

any colloidal instability for the nanophosphors in water over a period of at least one month.

The Dynamic Light Scattering (DLS) experiment on the aqueous dispersions of

nanophosphors has also indicated absence of aggregation of the nanoparticles (see Supporting

information). The PL emission spectra of the nanophosphors show multiple sharp spectral

lines, characteristic of luminescent nanophosphors. Yb3+ excited with 975 nm transfers energy

to Er3+, which results in the characteristic Er3+ emission with peaks at ~ 520, 538, 550 nm

(“green” bands) and ~ 649, 653, 667 nm (“red” bands). [43] The “green” PL is a radiative

deactivation of the 2H11/2 and 4S3/2 energy levels of the Er3+, whereas “red” PL corresponds to

4F9/2→4I15/2 transition. [32, 43] both levels are populated through 2F5/2 level of Yb3+, thus

upconverting energy (Figure 1 in Supporting information).

Submitted to

���8������8��

500 600 700 800 900 1000

CHCl3 suspensions

Aqueous suspensions

Wavelength (nm)

18 % Yb3+, 2 % Er3+,2 % Gd3+

15 % Yb3+, 2 % Er3+,5 % Gd3+ 10 % Yb3+, 2 % Er3+,10 % Gd3+

Abso

rban

ce

A

500 550 600 650 7000

10

20

30

40

50

Emis

sion

inte

nsity

(a.u

.)

Wavelength (nm)

18 % Yb3+,2 % Er3+,2 % Gd3+

15 % Yb3+,2 % Er3+,5 % Gd3+

10 % Yb3+,2 % Er3+,10 % Gd3+

B

8 16 32 64 128

128256512

1024204840968192

163843276865536

I~P2

Emis

sion

inte

nsity

(a.u

.)

Excitation power (mW)

at 538 nmat 667 nm

C 0.0 0.5 1.0 1.5

0.01

0.1

1

4

3

21

Emis

sion

inte

nsity

(a.u

.)

Time (ms) D

Figure 3. Optical properties of UCNPs (NaYF4 : x %Yb3+, 2%Er3+, y % Gd3+ ; x =18, 15, 5;

y = 2; 5; 15). A - Absorption spectra of chloroform and aqueous dispersions; B – PL spectra

of chloroform and aqueous dispersions, excitation with 975 nm; C – power dependence of PL

intensity at peak positions (538 and 667 nm) for NaYF4 : 10 %Yb3+, 2%Er3+, 10 % Gd3+ in

water; D – PL decays for water dispersion of the UCNPs. Green (1) and red (3) shows

NaYF4 : 10 %Yb3+, 2%Er3+, 10 % Gd3+, blue (2) and orange(4) shows NaYF4 : 18 %Yb3+,

2%Er3+, 2 % Gd3+. A shorter decay was observed at 538 nm, a longer decay at 667 nm. Black

line shows Instrument Responce Function (IRF).

Submitted to

���9������9��

It is worth noting that the ratio of the intensity of the “red” band to that of the “green” band

is noticeably different for aqueous and chloroform dispersions of the nanophosphors; red

emission is more intense in water dispersion (Figure 2A in Supporting information). This

variable green-to-red ratio (GRR) has been investigated and reported by several groups. [30, 43,

44] Its value is considered to depend on the excitation density, crystallographic phase and any

additional impurities which can increase the rate of multiphonon relaxation between the

green-emitting 2H11/2 and 4S3/2 levels and the red-emitting 4F9/2 level. [30, 43] The presence of

Gd3+ ions in our case can be an additional factor decreasing value of GRR for UCNPs. At the

same time, in nanophosphors a larger amount of Er3+ ions is located closer to the surface,

when compared to bulk material. This is why surface ligands used to disperse nanocrystals in

water can significantly increase the rate of the 2H11/2, 4S3/2 →4F9/2 transition. On the other

hand, higher light scattering in water nanophosphor dispersion results in a lower excitation

density for the exciting light and caused a decrease in the rate of the 2F5/2→4S3/2 process in

comparison to that of 2F5/2→4F9/2. A similar change in GRR has been also seen in the PL

spectra of the aqueous dispersion of the UCNPs for the different power of excitation (Figure

2B in Supporting information).

The photoluminescence lifetime for nanophosphors is very different from the fluorescence

lifetime characteristic for conventional fluorescence imaging probes, which is typically in the

nanosecond range. For UCNPs, it is in the submillisecond range (Figure 3, D). The lifetime

does not noticeably depend on the percentage of Er3+ and Yb3+, but is significantly different

for the red and the green bands, indicating a higher rate for the deactivation of the 4S3/2 level

(Figure 1 in supporting information). The different percentage of Gd3+ does not noticeably

influence the rate of the PL decays.

The power dependence for both the red and the green PL bands is close to quadratic (Figure

3, C), thus unambiguously confirming a two-photon induced process. It is important to note

Submitted to

���10������10��

that two-photon excitation in two-photon fluorescence laser scanning microscopy and other

fluorescence imaging techniques is an essential tool providing a number of advantages over

single photon excitation. The use of near-infrared light for excitation allows a larger

penetration depth in biological tissues and results in lower phototoxicity and reduced

background due to the relatively low two-photon absorption cross-section of most

biomolecules responsible for autofluorescence in biological systems. The quadratic

dependence of the excitation efficiency on the power of excitation leads to an intrinsic 3D

localization of two-photon excitation and also results in low photobleaching and photodamage

of the sample outside the focal region. [3, 36, 37] Relatively low two-photon excitation cross-

sections of the nanoprobes require the use of a high-repetition femtosecond (fs) pulsed laser

for two-photon induced imaging as well as the use of an objective lenses with a high

numerical apertures to focus the laser beam.[37] On the other hand, the use of UCNPs does not

require high photon flux and allows excitation and imaging of unconverting nanophosphors

using low-power and continuous wave (CW) laser diode, which is inexpensive and readily

available, when compared to femtosecond pulsed lasers.

To show possibility of single photon induced luminescence optical imaging with Gd3+

doped nanophosphors, we have also synthesized downconversion nanophosphors co-doped

with Eu3+, as described above. DCNPs co-doped with Eu3+ and Gd3+ shows specific and

distinct spectral lines of Eu3+ in the PL excitation and emission spectra (Figure 4, A).

Similarly to the UCNPs, downconversion ones also demonstrate substantially delayed PL

emission; its lifetime is in millisecond range (~10 ms), as seen in Figure 4, B. This highly

characteristic feature of Eu3+ photoluminescence provides one more option which can be

potentially used for time-resolved photoluminescence imaging. [29, 45]

Submitted to

���11������11��

450 500 600 650 700 7500

1

2

3

4

0.0

0.2

0.4

0.6

0.8

1.0

PL e

xcita

tion

inte

nsity

(a.u

.)

Wavelength (nm)

NaYF4:10%Eu3+, 10%Gd3+

414

523

462

532

588611

648

698

A

PL e

mis

sion

inte

nsity

(a.u

.)

Figure 4. Optical properties of DCNPs (NaYF4 : 10% Eu3+, 10%Gd3+) in water. A – PL

excitation and emission spectra; PL was excited with 532 nm and excitation spectrum was

obtained for emission at 611 nm. B. PL decay for DCNPs (NaYF4 : 10% Eu3+, 10%Yb3+).

Confocal microscopy was used to study the selective uptake of the up- and downconversion

nanophosphors by pancreatic cancer cells in vitro, using pre-equated amount of nanoparticles.

As shown in the Figure 5, there is almost no uptake of the nanophosphors capped with

mercaptopropionic acid only (A, C), whereas when the same nanophosphors were conjugated

with targeting biomolecules such as anti-claudin 4 (B) and anti-mesothelin (D) , the uptake of

the nanoparticles was considerably enhanced. As can be seen in the Figure 5, nanophosphors

mostly label the cell membrane. This enhanced labeling efficiency is a result of receptor-

mediated uptake of these bioconjugated nanoparticles with their corresponding receptors

which are known to be overexpressed on the surface of these cells. [19] The cell viability assay

of the nanoparticles showed no considerable cellular toxicity over a period of 48 hrs (see

Supporting figure S3).

0 10 20 30 40

0.01

0.1

1

C

PL

inte

nsity

(a.u

.)

Time (ms)

peak at 590 nm Instrument Response Function

B 0 10 20 30 40

0.01

0.1

1

C

PL

inte

nsity

(a.u

.)

Time (ms)

peak at 590 nm Instrument Response Function

B

Submitted to

���12������12��

Figure 5. Confocal images of Panc 1 cells treated with UCNPs (NaYF4: 2% Er3+, 10%Yb3+,

10%Gd3+.; A,B) and DCNPs (NaYF4 : 10% Eu3+, 10%Gd3+.; C,D). A and C shows non-

bioconjugated nanophosphors, B and D shows cells targeted with nanophosphors conjugated

with anti-claudin 4 and anti-mesothelin, respectively.

We have also checked whether nanophosphors can be imaged using Maestro GNIR FLEX

system, an Instrument used typically for whole body optical imaging in small animals. Figure

6 presents images of nanoparticles in aqueous dispersions acquired with Maestro system. As

one can see, spectrally unmixed images clearly demonstrate a feasibility to image and

spectrally distinguish the characteristically emitting nanophosphors (shown as orange and

Submitted to

���13������13��

red), using the Maestro imaging software. This experiment underlines the potential of

extending the application of these nanophosphors for in vivo bioimaging as well.

Figure 6. Spectrally unmixed Maestro images of aqueous dispersions of DCNPs (NaYF4 :

10% Eu3+, 10%Gd3+) (A) and UCNPs (NaYF4: 2%Er3+, 10%Yb3+, 10%Gd3+) (B).

We have also examined the paramagnetic behavior of the Gd3+ doped fluoride nanocrystals

and their ability to be used as contrast agents in MRI. Figure 7 shows the effect of different

concentration of Gd3+ doped UCNPs on the relaxivity (R1=1/T1 and R2=1/T2) of the water

protons at 400MHz. The studies performed on aqueous solutions of different Gd3+ doped

nanophosphors confirmed that T1 is inversely proportional to the concentration of the co-

doped paramagnetic agent, as shown previously for commercial MRI contrast agents. [40, 46]

2 4 6 8 10

1

2

3

4

5

6

7 as measured linear fit

y=0.63x-0.41

1/T1

(Sec

-1)

Gd3+ concentration (mM)2 4 6 8 10

150

200

250

300

350

400 as measured linear fit

y=28.5x+115.78

1/T2

(Sec

-1)

Gd3+ concentration (mM)

Submitted to

���14������14��

Figure 7. R1 and R2 determination for the different gadolinium concentrations doped in

nanophosphors in salt free aqueous solution. Data were acquired at 400 MHz and 25°C.

In order to determine whether these gadolinium doped nanophosphors could be used as

contrast agents in MR imaging, their contrast effect in solution was tested using MRI.

Different concentrations of suspensions of nanophosphors, containing 0-10mM of Gd3+, as

well as pure water for background signal, were placed in the centrifuge tubes. Representative

T1-weighted (TR/TE=500ms,10ms) and T2-weighted (TR/TE=10000ms,70ms) MRI images

are shown in Figure 8. The MR images of the nanophosphor suspension clearly showed the

negative enhancement effect on T2-weighted sequences, a result very similar to that obtained

for the superparamagnetic iron oxide nanocrystals used as contrast agents in MRI. [47] As a

result, the tubes containing the Gd3+ doped nanophosphors appeared dark and invisible in the

T2-weighted MR image (Figure 8). On the other hand, the tubes that contained water and void

nanophosphors remained brighter and visible.

Both the DCNPs and UCNPs also showed a pronounced contrast on the T1 weighted image

sequences. It can be seen in Figure 8 that samples with 10mM of Gd3+ (for both DCNPs and

UCNPs) showed the maximum contrast whereas for the samples with 2 and 5 mM of Gd3+

showed a slightly less contrast, which was still very high in comparison to plain water and

void nanophosphors. The direct proportionality of the Gd3+ concentration on the contrast

enhancement can be readily seen by the linearity of the plots of R1 (1/T1) and R2 (1/T2) as a

function of gadolinium concentration. We have determined the specific relaxivities r1, r2

from these plots, the values are: r1=0.14 mM-1s-1 , r2=8.7 mM-1s-1 . As reported earlier, [48, 49]

most commonly used commercial contrast agent, Gd-DTPA, has a r1=3.7 mM-1s-1 and r2=5.8

mM-1s-1,. It is worth also noting that these values are reported per Gd3+ ion. At the same time,

Submitted to

���15������15��

single nanophosphor probe can contain many Gd3+ ions, thus providing much higher signal

per single probe, in comparison with single molecule based probes such as Gd-DTPA.

Figure 8. T1 and T2 weighted images of NaYF4 doped with different concentrations of

Gd+3 and Er+3/Yb3+ (UCNPs) and Eu+3 (DCNPs). NaYF4 with 0mM Gd+3 and water were

used as control.

The T1/T2 relaxation plots for all of the imaging agents as well as a deiononized water

sample as control are shown in Figure 9. Estimated T1/T2 values from the least squared error

regression of the curves are shown in Table 1.

Table 1: T1 and T2 values as estimated from the MRI images.

Control UCNPs DCNPs Sample

0mM Gd3+ Water 2mM Gd3+ 5mM Gd3+ 10mM Gd3+ 10mM Gd3+

T1 (msec) 2183 3140 1417 1007 536 579

T2 (msec) 40.9 1330 22.3 16.5 8.8 12.4

Submitted to

���16������16��

0 2 4 6 8 10

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

y=0.14x+0.41

as measured linear fit

R1

(1/m

s)

Gd3+ concentration (mM)0 2 4 6 8 10

20

40

60

80

100

120

y=8.7x+23.9

as measured linear fit

R2

(1/m

s)

Gd3+ concentration (mM)

Figure 9 Plots of R1 (1/T1) and R2 (1/T2) as a function of gadolinium concentration for

Gd3+ doped nanophosphors as measured from 9.4T MRI.

The Gd3+ (10mM) doped in DCNPs has little effect on both T1 and T2 in comparison with

the same concentration of the Gd3+ (10mM) doped in UCNPs. However, this difference may

be due to a variation in the sample preparation or differences in the relaxivity provided by the

different lanthanides (Eu vs Er); This needs to be determined. In any case, the pronounced

contrast enhancement for both T1 and T2 weighted image sequences as well as high

relaxivities for both DCNPs and UCNPs suggests the potential use of these nanoparticles as

excellent contrast agents in MRI.

3. Conclusions

In conclusions, we have synthesized in the organic phase NaYF4 nanocrystals of 20-30 nm

size, co-doped with various rare-earths ions (Eu3+, Er3+, Yb3+, Gd3+), and subsequently

dispersed them in aqueopus phase. These nanoparticles are colloidally stable and easily

detectable by both MR and optical imaging methods. Their excellent photoluminescence

properties make them suitable probes for in vitro and in vivo optical bioimaging. We have

functionalized these nanoparticles by conjugation with tumor specific antibodies and shown

targeted imaging of living cancer cells in vitro, without any sign of cellular toxicity. Superior

Submitted to

���17������17��

magnetic contrast of the Gd3+ doped nanophosphors also underscores the potential of these

nanoparticles as robust MR imaging probes. In summary, owing to the combined presence of

efficient optical and MR imaging probes, along with the facility of site-specific delivery,

makes these non-toxic nanoprobes ideal candidates as ‘new generation’ of diagnostic probes

for routine clinical application. In vivo experiments using small animals bearing experimental

tumors are underway.

4. Experimental

Synthesis of NaYF4 nanocrystals doped with rare earth elements: The synthetic strategy

follows a similar protocol as described earlier with modifications. [31,50] Fixed amounts of

commercial Gd2O3 , Er2O3, Yb2O3 and Y2O3 were mixed and solubilized in hot ∼50%

concentrated trifluoroacetic acid and slowly evaporated to dryness in vacuum oven at 70°C.

The different molar compositions of the feed external solution were [Er/(Gd+Er+Yb+Y)] =

0.02, [Yb/(Gd+Er+Yb+Y)] = 0.18; 0.15; 0.10, and [Gd/(Gd+Er+Yb+Y)] = 0.02; 0.05; 0.10,

respectively. Next, the obtained trifluoroacetate precursor salt was added to the three necked-

flask with octadecane (30 ml), oleic acid (30 ml) and 4.5 mmol of sodium trifluoroacetate.

The molar ratio of Na(CF3COO) to RE(CF3COO) was kept 1.8 to form a pure nanocrystalline

matrix of α-NaREF4. The solution was heated to 110°C under vacuum with stirring for ∼30

min to remove water and oxygen during each 10 minutes time the flask was purged with dry

nitrogen. The yellow solution was then heated to 300°C under nitrogen and kept under

vigorous stirring for 1h. After synthesis, the mixture was cooled to room temperature and

precipitated by acetone using ultrasonic bath and collected by centrifuge at 11,000 rpm for 30

min and then washed with ethanol. Finally, the entire amount of NC was directly dispersed in

chloroform (4 ml). For downconversion luminescence NaYF4: Eu3+,Gd3+ nanophosphors

were synthesized using the same procedure as that used for the synthesis of NaYF4:Er3+, Yb3+,

Gd3+ co-doped sample, except that Er2O3 was replaced with EuCl3. The different molar

Submitted to

���18������18��

compositions of the feed external solution was [Eu/(Gd+Eu+Y)] = 0.1, [Gd/(Gd+Eu+Y)] = 0.1,

respectively.

For transferring the nanocrystals from chloroform solution to aqueous phase, we used the

ligand exchange method using 3-mercaptopropionic acid which results in water dispersible

carboxyl terminated nanoparticles. In a typical protocol 2ml of chloroform dispersion of

nanocrystals was further diluted with 3ml of chloroform and incubated with 3-

mercaptopropionic acid (5 ml). After the solution was stirred overnight, HPLC grade water (5

ml) was added to it and further stirred for 30min. The turbid solution obtained was

centrifuged and the pellet was redispersed in HPLC grade water (5 ml) by sonication. Finally

the aqueous dispersion of the nanocrystals was filtered with a 0.45µm syringe filter and stored

at 4oC for future use.

Conjugation of –COOH terminated nanocrystals with anti-claudin 4: For in vitro cellular

targeting studies, the carboxyl terminated aqueous dispersion of nanocrytals was conjugated

to different biomolecules like anti-claudin 4 and transferrin by using a typical EDC chemistry.

1ml stock solution of –COOH terminated nanocrystals was mixed with 0.1M EDC solution

(25 µl) and gently stirred for 30 minutes. Next, anti-claudin 4 (5 µl of 0.5 mg/ml) was added

into this mixture and incubated at room temperature for 2 hrs to allow the free amino groups

of the antibody to covalently bond to the carboxyl groups on the nanoparticles. Human

pancreatic cancer cell line Panc 1 was cultured in Dulbecco minimum essential media (MEM-

α) with 10% fetal bovine serum (FBS), 1% penicillin, and 1% amphotericin B. The day before

nanoparticles treatment, cells were seeded in 35 mm culture dishes. On the treatment day, the

cells, at a confluency of 70-80 % in serum-supplemented media, were treated with the

nanophosphors at a specific concentration (1 mg/ml media) for two hours at 37°C.

Submitted to

���19������19��

Optical spectroscopy: Optical absorption, photoluminescence steady-state and time-

resolved spectroscopies were used to characterize the spectral properties of the luminescent

nanomaterials. UV-vis absorption spectra were acquired using a Shimadzu UV-3600

spectrophotometer. Photoluminescence (PL) excitation/emission spectra were recorded on a

Fluorolog-3 spectrofluorimeter (Jobin Yvon, Longjumeau, France). Fiber coupled laser diode

(Qphotonics) emitting at 975 nm was used as the excitation source in case of UCNPs.

Fluorolog-3 spectrofluorimeter was also engaged in lifetime measurements. In this case

Infinium oscilloscope (Hewlett-Packard) coupled to the PMT of the spectrofluorimeter was

used to acquire PL decays. A second harmonic 532 nm beam from nanosecond pulsed

Nd:YAG laser (Lotis TII, Belarus) operating at 20 Hz was used to excite DCNPs. Qphotonics

laser diode operating in pulsed mode (pulse width of 200 µs) at 200 Hz was used as the PL

excitation source in case of UCNPs.

In vitro Confocal Microscopy and Optical Imaging: Confocal fluorescence laser scanning

microscopy was performed using a confocal microscope (MRC-1024, Bio-Rad, Richmond,

CA), which was attached to an upright microscope (Nikon model Eclipse E800). A water

immersion objective lens (Nikon, Fluor-60X, NA 1.0) was used for cell imaging. As source of

excitation, we have used the diode-pumped solid state laser (Millenia, Spectra-Physics)

providing 532 nm line for excitation of DCNPs or single mode laser diode (Qphotonics)

operating at 975 nm to image UCNPs. Excitation light was coupled to upper port of the

microscope through single mode optical fiber.

Received: ((will be filled in by the editorial staff)) Revised: ((will be filled in by the editorial staff))

Published online: ((will be filled in by the editorial staff))

Submitted to

���20������20��

[1] S. A. Wickline, G. M. Lanza, J. Cell. Biochem. 2002, 39, 90.

[2] J. R. Heath, M. E. Phelps, L. Hood, Molecular Imaging and Biology 2003, 5, 312.

[3] P. N. Prasad, Introduction to Biophotonics. Wiley-Interscience, New York, 2003.

[4] M. Bruchez Jr., M. Moronne, P. Gin, S. Weiss, A. P. Alivisatos, Science 1998, 281,

2013.

[5] H. R. Herschman, Science 2003, 302, 605.

[6] R. Y. Tsien, Nat. Rev. Mol. Cell Biol. 2003, 4, SS16.

[7] G. M. Lanza, S. A. Wickline, Progress in Cardiovascular Diseases 2001, 41, 13.

[8] G. M. Lanza, S. A. Wickline, Current problems in cardiology 2003, 28, 625.

[9] E. M. B. Haacke, M. L. Thompson, R. Venkatesan, Magnetic Resonance Imaging:

Physical Principles and Sequence Design, John Wiley, 1999.

[10] D. M. McDonald, P. L. Choyke, Nat. Med. 2003, 9, 713.

[11] J. Michaelis, C. Hettich, J. Mlynek, V. Sandoghdar, Nat. Med. 2000, 405, 586.

[12] C. Wu, J. Hong, X. Guo, C. Huang, J. Lai, J. Zheng, J. Chen, X. Mu, Y. Zhao, Chem.

Comm. 2008, 6, 750.

[13] M. F. Kircher, U. Mahmood, R. S. King, R. Weissleder, L. Josephson, Cancer Res.

2003, 63, 8122.

[14] S. Santra, H. S. Yang, P. H. Holloway, J. T. Stanley, R. A. Mericle, J. Am. Chem. Soc.

2005, 127, 1656.

Submitted to

���21������21��

[15] L. Josephson, M. F. Kircher, U. Mahmood, Y. Tang, R. Weissleder, Bioconjug. Chem.

2002, 13, 554.

[16] Y. S. Lin, S. H. Wu, Y. Hung, Y. H. Chou, C. Chang, M. L. Lin, C. P. Tsai, C. Y.

Mou, Chem. Mat. 2006, 18, 5170.

[17] M. E. Akerman, W. C. Chan, P. Laakkonen, S. N. Bhatia, E. Ruoslahti, Proc. Natl.

Acad. Sci. USA 2002, 99, 12617.

[18] J. Qian, K. T. Yong, I. Roy, T. Y. Ohulchanskyy, E. J. Bergey, H. H. Lee, K. M.

Tramposch, S. He, A. Maitra, P. N. Prasad, J. Phys. Chem. B 2007, 111, 6969.

[19] R. Kumar, I. Roy, T. Y. Ohulchanskyy, L. N. Goswami, A. C. Bonoiu, E. J. Bergey, .

M. Tramposch, A. Maitra, P. N. Prasad, ACS Nano 2008, 2, 449.

[20] X. Michalet, F. F. Pinaud, L. A. Bentolila, J. M. Tsay, S. Doose, J. J. Li, G.

Sundaresan, A. M. Wu, S. S. Gambhir, S. Weiss, Science 2005, 307, 538.

[21] A. M. Morawski, P. M. Winter, X. Yu, R.W. Fuhrhop, M. J. Scott, F. Hockett, J. D.

Robertson, P. J. Gaffney, G. M. Lanza, S. A. Wickline, Magn. Reson. Med. 2004, 52, 1255.

[22] D. Artemov, N. Mori, B. Okollie, Z. M. Bhujwalla, Magn. Reson. Med. 2003, 49, 403.

[23] Y. C. Cao, J. Am. Chem. Soc. 2004, 126, 7456.

[24] K. Ko�mpe, H. Borchert, J. Storz, A. Lobo, S. Adam, T. Mo�ller, M. Haase, Angew.

Chem. – Int. Edition 2003, 42, 5513.

[25] J. W. Stouwdam, F. C. J. M. Van Veggel, Nano Letters 2002, 2, 733.

[26] K. Riwotzki, M. Haase, J. Phy. Chem. B 1998, 102, 10129.

Submitted to

���22������22��

[27] T. Mirkovic, M. A. Hines, P. Sreekumari Nair, G. D. Scholes, Chem. Mat. 2005, 17,

3451.

[28] X. Y. Chen, E. Ma, G. K. Liu, J. Phy. Chem. C 2007, 111, 10404.

[29] K. Hanaoka, K. Kikuchi, S. Kobayashi, T. Nagano, J. Am. Chem. Soc. 2007, 129,

13502.

[30] K. W. Kramer, D. Biner, G. Frei, H. U. Gudel, M. P. Hehlen, S. R. Luthi, Chem. Mat.

2004, 16, 1244.

[31] Y. W. Zhang, X. Sun, R. Si, L. P. You, C. H. Yan, J. Am. Chem. Soc. 2005, 127, 3260.

[32] F. Wang, X. Liu, J. Am. Chem. Soc. 2008, 130, 5642.

[33] J.-C. Boyer, F. Vetrone, L. A. Cuccia, J. A. Capobianco, J. Am. Chem. Soc. 2006, 128,

7444.

[34] X. Liang, X. Wang, J. Zhuang, Q. Peng, Y. D. Li, Inorg. Chem. 2007, 46, 6050.

[35] D. K. Chatteriee, A. J. Rufalhah, Y. Zhang, Biomaterials 2008, 29, 937.

[36] W. Denk, J. H. Strickler, W. W. Webb, Science 1990, 248, 73.

[37] R. A. Niesner, V. Andresen, M. Gunzer, Immunological Rev. 2008, 221, 7.

[38] S. Heer, K. Kompe, H. U. Gudel, M. Haase, Adv. Mat. 2004, 16, 2102.

[39] G. R. Choppin, K. M. Schaab, Inorg. Chim. Acta 1996, 252, 299.

[40] P. Caravan, J. J. Ellison, T. J. McMurry, R. B. Lauffer, Chem. Rev. 1999, 99, 2293.

[41] J. S. Mann, J. Computer Assisted Tomography 1993, 17, S19.

Submitted to

���23������23��

[42] P. Salas, C. Angeles-Chavez, J. A. Montoya, E. De la Rosa, L. A. Diaz-Torres, H.

Desirena, A. Martinez, M. A. Romero-Romo, J. Morales, Optical Materials 2005, 27, 1295.

[43] H. Schafer, P. Ptacek, K. Kompe, M. Haase, Chem. Mat. 2007, 19, 1396.

[44] J. F. Suyver, J. Grimm, K. W. Kramer, H. U. Gudel, J. Luminescence 2005, 114, 53.

[45] A. Picot, A. D'Aleo, P. L. Baldeck, A. Grichine, A. Duperray, C. Andraud, O. Maury,

J. Am. Chem. Soc. 2008, 130, 1532.

[46] G. Strich, P. L. Hagan, K. H. Gerber, R. A. Stutsky, Radiology 1985, 154, 723.

[47] C. Corot, P. Robert, J. M. Idee, M. Port, Adv. Drug Del. Rev. 2006, 58, 1471.

[48] K. M. Donahue, D. Burstein, W. J. Manning, M. L. Gray, Magn. Reson. Med. 1994,

32, 66.

[49] R.B. Lauffer, Chem. Rev. 1987, 87, 901.

[50] O. Ehlert, R. Thomann, M. Darbandi, T. Nann, ACS Nano 2008, 2, 120-124.

Submitted to

���24������24��

The table of contents Multimodal NaYF4 nanophosphors doped with rare earth ions are synthesized and

characterized. The nanophosphors exhibit good optical and magnetic properties which can be

exploited in vitro and in vivo for optical fluorescence and MR imaging.

Keyword: Nanocrystals, Magnetic Materials, Optically Active Materials Marcin Nyk, Tymish Y. Ohulchanskyy, Christopher A. Flask, Paras N. Prasad*

Title: Combined optical and MR bioimaging using rare-earth ions doped NaYF4 nanocrystals ToC figure

Column Title: R. Kumar et al./Multimodal NaYF4 nanophosphors

Submitted to

���25������25��

Supporting Information

Combined optical and MR bioimaging using rare-

earth ions doped NaYF4 nanocrystals

Rajiv Kumar, Marcin Nyk, Tymish Y. Ohulchanskyy, Christopher A. Flask and

Paras N. Prasad*

[*] Prof. Paras N. Prasad, Dr. R. Kumar, Dr. M. Nyk, Dr. T. Y. Ohulchanskyy Institute of Lasers, Photonics and Biophotonics, Department of Chemistry State University of New York, Buffalo, NY 14260, USA E-mail: pnprasad@acsu.buffalo.edu Dr. C.A. Flask Case Center for Imaging Research, Case Western Reserve University Cleveland, OH, USA

Submitted to

���26������26��

Materials

The synthesis of NaYF4 was carried out using standard oxygen/water-free procedures and

commercially available reagents. Rare-earth (RE) oxides (99.99%, Aldrich), oleic acid (90%.

Alfa Aesar), 1-octadecane (90%), trifluoroacetic acid (99%, Sigma), sodium trifluoroacetate

(98%, Sigma), 3-mercaptopropionic acid (99+% Aldrich), absolute ethanol, chloroform

(≥99.8% Sigma), acetone and HPLC water (Sigma) were used as received. Anti-claudin 4

and anti-mesothelin were purchased from Invitrogen.

Transmission Electron Microscopy

Transmission Electron Microscopy (TEM) was used to determine particle size and size

distribution. Chloroform or aqueous dispersions of the nanoparticles were dried on formvar-

coated copper grids (obtained from Electron Microscopy Sciences Inc.) and visualized using a

JEOL JEM 2020 electron microscope.

Cell viability assay

The Panc 1 cells were dispensed into a 96-well flat-bottom microtiter plate (Nunc)

(10,000 cells/well) and allowed to attach overnight using MEM-α medium with 10% FBS.

Cell viability was assessed by the Cell Titer-Glo™ luminescent cell viability assay (Promega

Corporation, Madison, WI). This assay is a homogenous method to determine the number of

viable cells in culture based on quantity of the ATP present, which signals the presence of

metabolically active cells. [1] In brief, after 4hrs and 48hrs treatment with anti-claudin 4

conjugated nanophosphors, the medium was changed and the Cell Titer-Glo reagent was

added to the cells, mixed for 2 min and allowed to incubate for 10 min at room temperature.

Luminescence was measured using a microplate luminometer (Bio- Tek Synergy HT

microplate reader) and data were expressed as a percentage of the control. Tests were

Submitted to

���27������27��

performed in quadruplicate. Each point represents the mean ±SD (bars) of replicates from one

representative experiment.

Leaching studies of Gd3+ from the Nanophosphors matrix

For checking the leaking of the gadolinium ions from NaYF4 matrix, different samples kept

at room temperature for different time durations were spin filtered at 11000 rpm for 20 min

using Nanosep 100K OMEGA spin filters. The filter membrane holds almost 100% of the

nanophosphors as evident by the absence of any emission from the lower aqueous filrate. The

filtrate obtained after spin filtration was incubated with 0.05mM Arsenazo III and analyzed

for any Gd3+ presence by monitoring the absorption band of arsenazo-Gd3+ complex at 658

nm[2]. Water and GdCl3 were used as negative and positive controls respectively (Supporting

figure S5).

Paramagnetic behavior of the gadolinium ions doped in fluoride nanocrystals and

magnetic resonance imaging with these nanocrystals

The effects of paramagnetic behavior of gadolinium ions doped in both upconversion and

downconsion nanophosphors on proton relaxation times were measured in aqueous solutions

on a Varian INOVA spectrometer at 400 MHz, utilizing a standard, inversion recovery

sequence. For this time relaxation studies with the aqueous dispersion of the gadolinium

doped nanophosphors, different concentrations of gadolinium doped nanocrytals were used.

The samples for measurements were made in 10% D2O (v/v).

Submitted to

���28������28��

Supporting Figure 1. Schematic energy levels of Yb3+ and Er3+ doped in NaYF4 nanocrystals.

500 550 600 650 700

0

1

A

10 % Yb3+,2 % Er3+,10 % Gd3+ in water 10 % Yb3+,2 % Er3+,10 % Gd3+ in CH3Cl

Wavelength (nm)Nor

mal

ized

em

issi

on in

tens

ity (a

.u.)

500 550 600 650 7000

1

B

Wavelength (nm)

Nor

mal

ized

em

issi

on in

tens

ity (a

.u.)

10 % Yb3+,2 % Er3+,10 % Gd3+ in water

Excitation power of 12 mW Excitation power of 89 mW

Supporting Figure 2. Differences in green-to-red ratio (GRR) for chloroform and water dispersions of UCNPs under the same power of excitation (A) and for different power of excitation in water dispersion (B).

Submitted to

���29������29��

0

10

20

30

40

50

60

70

80

90

100

110

Control 0.4mg/ml 0.8mg/ml 1.2mg/ml 1.6mg/ml 2.0mg/ml

Concentration of NPs

% c

ell v

iabi

lity

4 hrs

48 hrs

Supporting Figure S3. Cell viability assay for Panc-1 cell line treated with different concentrations of NPs. Supporting Figure S4: The DLS data showing the hydrodynamic diameter for the aqueous suspension of the nanophosphors.

1 5 50 500 5000 Diameter (nm)

I N T E N S I T Y

Submitted to

���30������30��

Supporting Figure S4: Absorption behavior of Arsenazo III with different filterates obtained after spin filteration of UCNPs aqueous solution to study the leaching of Gd3+ from the NaYF4 matrix. Water and GdCl3 were used as negative and positive controls respectively.

References:

[1]. F. Fan, K. V. Wood, Assay and Drug Development Technologies 2007, 5, 127.

[2]. E. Hvattum, P.T. Normann, G.C. Jamieson, J.-J. Lai and T. Skotland. J. Pharmaceut.

Biomed. Anal. 1995, 13, 927.