Quantum dots (QDs), one of the new nanomaterials, have high fluorescence intensity, photostability, low photobleaching, and simultaneous excitation of particles with different colors by one single wavelength compared to traditional organic dyes.1, 2 In addition, they have the same surface properties, which allow a similar approach for conjugating biomolecules to QDs with different colors; thus, it is possible to monitor biological processes with different labeling molecules over the long term.3, 4
However, even though quantum dots have unique optical properties, as fluorescent probes in biological labeling, their cytotoxicity is of great concern. Derfus 5 found the release of free ions from the CdSe-core QDs because of surface oxidation, and cell death was correlated with the accumulative of free cadmium ions. Lovric 6 found that naked CdTe QDs could damage the plasma membrane, mitochondria, and nucleus and lead to the release of cytochrome c from mitochondria. Moreover, these QDs could induce cells apoptosis. But when the CdS shell and ZnS shell were packed on the core of the CdTe, there was noncytotoxicity to K562 cells for incubation at concentration,7 indicating that as fluorescent probes, QDs need to be prevented from releasing toxic elements in biological applications, whose key problem is to modify the protective agents to the QDs’ surface.
QDs that have a core/shell structure are the most versatile in biology. The cores of this kind of QD are well packaged by ZnS. Usually, this kind of QD is synthesized in hydrophobic organic solvents, as solubilization of these QDs is essential for biological applications. However, the photoluminescence quantum yield (PL QY) of these water-soluble QDs is lower after solubilization, and photostability also declines—all these influence the applications of QDs in various fields.8 CdTe QDs, a recent arrival on the scene, are directly synthesized in a water-phase system and the synthesis method is easy. This kind of QD has also been applied to many areas of biology.9, 10, 11, 12, 13 But CdTe QDs have low photostability and a wide distribution of particle sizes, and their PL QY are not as high as that of QDs.14, 15, 16 Thus, researchers have made great efforts to synthesize QDs with high PL QY and low cytotoxicity for biological applications.
The applications of QDs in cell biological and biomedical imaging are based on the special activity of the conjugated biomolecules. Covalent bonding is the normal way of conjugation bimolecular to QDs.17 Conjugation of denatured protein to the surface of QDs is another new method to prepare the nanocompounds.18, 19 Transferrin (Tf) is an important -globulin and the serum iron transport protein, consisting of a single polypeptide chain of 670 to 700 amino acids containing two structural signatures at the N-terminal and C-terminal and 19 disulphide bridges.20 The relative molecular weight is . It is highly specific to transferrin receptors (TfR) and does not cross-react with other related proteins. It can deliver and adjust the balance of iron in the biological body and is one of the necessary factors in cell growth and propagation. As TfR is overexpressed on the surface of tumor cells, identification and diagnosis of tumors can be achieved using tagged Tf.20
In this study, four nanoparticles were synthesized with , CdTe QDs, and transferrin by covalent bonding and protein denaturation. Their optic capabilities were detected and cytotoxicity in HeLa cells was evaluated with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) viability assay. It was found that QDs capped with denatured transferrin (dTf) present the least toxicity and superior photoluminescence. Preparation of this probe is very simple, and coupling agents are not involved in the program of synthesis, so purification processes are unnecessary. This then is a simple way to prepare nanoparticles with high PL QY and low cytotoxicity.
Materials and Methods
QDs and CdTe QDs were synthesized according to previous reports.16, 21 Sodium dodecyl sulfate (SDS) and (96%) were purchased from Sinopharm Chemical Reagent Co. 3-mercaptopropionic acid, sodium thioglycolate, N-hydrocylsulfo-succinimide (Sulfo-NHS), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), MTT, and transferrin were purchased from Sigma-Aldrich Fine Chemicals (St. Louis, Missouri). All other chemicals and materials used in the experiments were of analytical grade, and water was deionized.
The following equipment was purchased from the companies indicated in parentheses: Luminescence spectrometer (LS-55, PerkinElmer, Waltham, Massachusetts ), fiber-optic spectrometer (QE65000, Ocean Optics, Dunedin, Florida ), micro-plate reader (ELX808™, Biotek, Winooski, Vermont ), inverted fluorescence microscope (IX71, Olympus,, Nagano, Japan ), cooled color charge-coupled device (CCD, Pixera Penguin 150CL, San Jose, California ), circular dichroism spectrometer (J-810, Jasco, Tokyo, Japan ), vertical electrophoresis system (DYY-6C, Beijing Liuyi Instrument Factory, China ), Gbox-M Biosens Gel Documentation System (Syngene, Cambridge, England), high-voltage power ( , Shanghai Nuclear Research Institute, China).
Modification of Water-Soluble QDs
-chloroform solution was added by sodium thioglycolate powder in an eppendorf (EP) tube. After stirring for , of distilled water was added to the tube, following by stirring and incubation. The supernatant layer of the solution was dissociated for the next precipitation process with acetone. The whole process was repeated more than three times to remove the free sodium thioglycolate. Last, the precipitate was dissolved in deionized water to get water-soluble QDs.
Preparation of QD Probes
The conjugation process of -Tf included three steps. First, EDC phosphate buffered saline (PBS) solution ( , pH 7.4) was added to water-soluble QDs solution . Then the mixture was added to PBS (pH 7.2), followed by shaking. transferrin PBS solution was added into the mixture, and the solution was stirred for at room temperature. The resulting mixture was filtered, and -Tf was obtained.
NHS ( pH 7.4) PBS solution was added to CdTe solution with 3-mercaptopropionic acid as the stabilizer. After incubation at room temperature, PBS and transferring aqueous solution were added, and the mixture was shaken at room temperature for to allow the conjugation reaction yielding CdTe-Tf.
Transferrin was denatured by chemically treating with . The process was as follows: transferrin and were dissolved in deionized water under stirring. The reaction proceeded at room temperature for . After that, the reagent was incubated in a constant temperature water bath at for until no more was generated. Then the denatured transferrin was obtained.
QDs were precipitated with acetone and redissolved in a measured amount of dTf solution, and the admixture was incubated at in a constant temperature water bath for . After being cooled and diluted in PBS (pH 7.2) to the needed concentration, denatured transferrin was coupled with QDs. The preparation of CdTe-dTf and -dTf were finished.
Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE)
10% resolving gel and 4% stacking gel were used in SDS-PAGE electrophoresis. A aliquot of the sample was mixed with sample buffer, and the mixture was then injected into the gel wells. After running for at , , the gel was fixed in 12% acetum for . The fluorescence of samples was observed before staining in 0.01% Coomassie blue R250. The image of dyed protein was obtained after the gel being discolored.
Procedure of Capillary Electrophoresis
Capillary electrophoresis (CE) analyses were carried out on a home-built system. A capillary was fixed on the detecting platform of an inverted fluorescence microscope, while a mercury lamp was used as excitation source, and the excited fluorescence signal of QDs was collected using a fiber-optic spectrometer. CE experiments were all performed in -long fused-silica capillaries. The effective length (length from injection to the detection window) was . Hydrodynamic injection was performed by siphoning at height differences for at the anode. Water-soluble polymer solutions were used as sieving media. A solution of (pH 9.20) was used as CE separation buffer. The separation was achieved at room temperature. Between runs, the capillary was washed with NaOH, pure water, and running buffer for to ensure the reproducibility.
Circular Dichroism Spectrometer
The circular dichroism (CD) spectrum was detected by a CD spectrometer. The spectrum diameter was . The scanning speed was , and the scanning range was . The time constant was , and the resolution was . Samples were dissolved in PBS (pH 7.2) at . The results were averaged from three scans.
Cell Culture Conditions and Treatments
The cytotoxicity of QDs was evaluated by MTT [3-(4, 5-dimethylthiazol -2-yl)-2, 5-diphenyltetrazolium bromide] viability assay. HeLa cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% fetal bovine serum, streptomycin, and penicillin at in a humidified atmosphere with 5% . For MTT assays, cells were dispensed in 96-well plates at a density of 1000 cells/well. After incubating for , the cells were washed with fresh medium. The treated cells were added by different concentrations (from ) of the nanoparticles and incubated for , , , , and , respectively. At different incubation time, MTT stock solution was added to each well, and cells were cultured for another at in the dark. The media was then removed, and dimethyl sulfoxide (DMSO) was put in each well. Plates were shaken gently for while the cells were lysed with DMSO. The absorbance at was measured with a micro-plate reader. In this study, the data were from three or four independent experiments, with the same treatment repeated in triplicate. The cell viability was normalized to 100% for the control well containing no QDs and the same treatments.
For imaging, the cells were treated as described earlier and exposed to QD probes for defined time intervals. After being incubated for the different time periods (2, 12, 24, 48, and ), the cell morphology and metabolic activity were detected by an IX71 inverted fluorescence microscope. Before and after the QD probes being washed out, fluorescence and transparent images were taken, and the spectrum was recorded from four parallel wells for each concentration of the probes.
Results and Discussion
Characterization of QD Probes
The photofluorescence (PL) spectra of these QDs probes and original QDs were measured. As shown in Fig. 1 , compared to the original , CdTe QDs [Figs. 1 and 1, curve a, respectively], the dTf-coated and CdTe QDs (Figs. 1 and 1, curve b, respectively) showed a higher PL QY and PL peak positions shifted to blue. But the Tf conjugated, , and CdTe QDs displayed contrary results. Their PL QY was lower and peak positions shifted to red compared with the original QDs (Figs. 1 and 1, curve c, respectively). The quantum yields were measured by the optically dilute measurement method using rhodamine (whose PL QY is assumed to be 95% in ethanol) as in previous reports.22 PL QYs were 26%, 33%, and 18%, respectively, for QDs, -dTf, and -Tf, while those of CdTe QDs, CdTe-dTf, and CdTe-Tf were 21%, 35%, and 16%, respectively. All these illuminated that dTf directly conjugated to the surface of QDs. Moreover, the PL QYs of dTf-modified QDs increased, while that of Tf-modified QDs decreased.
Proof of Conjugation by SDS-PAGE
SDS-PAGE was carried out on QDs coated with dTf to prove the conjugate formation. The concentration of QDs was kept stable, while the concentration of dTf increased. Figure 2 shows the electrophoresis results for the different concentration ratios of dTf-coated QDs (wells I, II, III), pure dTf (well IV), pure Tf (well V), heated QDs (well VI), and original QDs (well VII). The evidence displayed that the color of bands “a” in wells I, II, III became deeper with increased concentration of dTf. At the same time, the luminescent image of the gel is shown in Fig. 2, where the bands “a” of dTf-coated showed strong luminescence in well II and III; these were possibly due to the change of charge of protein after denatured transferrin coating, resistance was augmented, and the protein band shifted from position b (Fig. 2, well IV) to band a. By comparing the two bands in well II and well III in Fig. 2, it is found that band b of dTf also appeared in well III, suggested that there was excessive dTf in the ratio of 6:6 (dTf: ). However, it is puzzling that well I had faint luminescence as well as the same concentration with well II and well III. The reason was due to the sample being heated in the preparation. The same treatment of was carried out, and the results showed that the original QDs had a wide band with strong fluorescence (Fig. 2, well VII), while after heating, the fluorescence intensity of QDs was enormously decreased (Fig. 2, well VI). This was confirmed in the photoluminescence spectra as well. The fluorescence intensity of QDs under heat sharply dropped (Fig. 1, curve d). In our previous work, the photoluminescence of water-soluble QDs presented sensitive temperature dependence.16, 23 The fluorescence intensity of QDs has irreversible decreased, which is the reason for faint light in well I, Fig. 1. Thus, it was probable that in well I, dTf could not package QDs absolutely because of the small ratio. Therefore, the fluorescence intensity was greatly reduced after heating. With the increase of dTf, QDs were coated completely and hold back the effect of heat, so it exhibited high fluorescence. The result was the same as Fig. 1 (curve b). Thus, it is easy to determine why bright bands were in well II and III.
Capillary Electrophoresis Analysis
To choose the optimal ratio of dTf to QDs, capillary electrophoresis was used. First, the pure QDs were measured and an electrophoresis peak appeared at retention time about (Fig. 3 , curve a). After conjugation with the ratio 6:1 (QDs: dTf), two electrophoresis peaks were observed at about and (Fig. 3, curve b). It was speculated that the conjugation of dTf to QDs was not homogeneous in this proportion and the amount of QDs coated by dTf was different. However, at ratio 6:3 (QDs: dTf), only one electrophoresis peak appeared (Fig. 3, curve c). It was confirmed that the electrophoresis peak was caused by the -dTf. When the ratio of QDs to dTf was 6:6, there was still only one peak. Therefore, it was affirmed that dTf completely coated on QDs with ratio 6:3 (QDs: dTf). The results were coherent with the SDS-PAGE. A band of protein emerged in well III (Fig. 2, band b), indicating that the dTf was in excess. The ratio 6:3 of QDs:dTf was the optimal ratio for conjugation of dTf and QDs.
Cytotoxicity of CdTe-dTf, -dTf, CdTe-Tf, -Tf, , and CdTe QDs for the HeLa Cells
Based on the preceding results, to evaluate the effect of the concentration and incubative time on the cytotoxicity, six kinds of QDs were compared at five concentrations in HeLa cells As shown in Fig. 4 , it was found that -dTf were almost nontoxic when they were exposed to HeLa cells for different incubation times. There was almost no change in the morphology of cells incubated with this probe, and the cells were growing well. Cells were seldom seen in suspension. At the same time, the amount of cells increased gradually. At the incubation time of , the density was obviously augmented, indicating that the proliferation of cells has been occurring. The fluorescence imaging showed the contour of HeLa cells too. Simultaneously, QD probes were seen gathered in the cells. The position of the PL peak was detected always at from by spectrometer (Fig. 5 ). The cytotoxicity of -dTf evaluated with the MTT viability assay is shown in Fig. 6 . It was found that this kind of QD probe was almost nontoxic to HeLa cells; even at and incubated for , cell viability was still at 92%.
Compared with -dTf, the cell density decreased slightly after incubation with -Tf for ; meanwhile, the proliferation was slower [Fig. 7 ]. During incubation, a PL peak at was also obtained (data not shown here). The cell viability decreased to 86.7% at , as shown in the MTT viability assay [Fig. 8 ], while the viability of cells declined with prolonged incubation time.
CdTe-dTf of was more highly toxic for HeLa cells than QD probes. Live cells were greatly reduced after incubation, and suspended cells increased. A few fixed cells were observed in wells after washing CdTe-dTf at [Fig. 7]. As shown in Fig. 8, the cytotoxicity of CdTe-dTf was time and concentration dependent. Cell viability was depressed by prolonging time and augmenting dose. 22.5% cells were alive after at concentration [Fig. 8].
For CdTe-Tf, the morphology of HeLa cell labeling with concentration for was absolutely globose, and nearly no fixed cells existed [Fig. 7]. With the MTT viability assay, of CdTe-Tf resulted in the decrease of cell viability up to 60% in , and only 14.2% cells were alive with for incubation [Fig. 8].
and CdTe QDs
The pure and CdTe QDs were used to evaluate the cytotoxicity in HeLa cells. When the QDs were washed out after incubation, no fluorescence in the cells could be detected in the cells. With prolonged incubation time, a few scattered fluorescent groups were detected. However, the fluorescence was visually cluttered and the outline of cells were not visible, while the location of labeling was uncertain too [Fig. 9 ]. The probable reason was that with prolongation of the incubation, a few QDs were adsorbed on the surface of the cells nonspecifically and could not be washed out. Comparing the images of these two QDs, CdTe QDs (Fig. 9) have higher cytotoxicity than core-shell QDs [Fig. 8]. In this case, the proliferation of cells was slow; there were only a few cells with incubation after washing off the CdTe QDs, and nearly all cells were dead after . However, a few proliferous cells were observed when of QDs were exposed to HeLa cells for [Fig. 7]. The result was approved by the MTT viability assay. The and CdTe QDs led to the decrease of cell viability up to 58.8% [Fig. 8] and 3.2% (Fig. 10 ) with incubation and concentration.
These results demonstrated that the order of cytotoxicity of QDs in HeLa cells from high to low was -dTf, -Tf, , CdTe-dTf, CdTe-Tf, and CdTe. MTT viability assay showed the same result. For incubation and concentration, the cell viability was up to 92% with the lowest cytotoxicity -dTf, while down to 3.2% with the highest cytotoxicity CdTe QDs.
The toxicity of QDs depends on various factors that come from both inherent physicochemical properties and environmental conditions. It was reported that QD size, concentration, outer coating functional groups and materials, mechanical stability, the species of cells, and exposure times have been included as determining factors in QD cytotoxicity.5, 11, 24, 25, 26, 27, 28, 29, 30, 31 The QD probes we synthesized exhibited different degrees of cytotoxicity in HeLa cells. -dTf was almost nontoxic and CdTe-Tf was highly cytotoxic. -dTf appeared highly biocompatible, possibly because had inherent core-shell structure and the outer-coated ZnS inhibited the release of Cd ion. On the other hand, dTf directly wrapped on the surface of QDs could be seen as another “shell” and enhanced the stability of QDs. CdSe inside the structure was not easy to disassemble, and dTf provided a highly stable layer for QDs. All these led to reducing the influence from environmental conditions. While the QD probes prepared by coupling agent have another protection layer compared with the original QDs, the cytotoxicity is a little depressed. But if Tf was not coated on the QDs directly, the QD probes were barely affected by the outer environment, and stability was not as good as dTf-coated QDs. As a result, the damage by these probes was bigger. For uncoated QDs, no biomolecule was covered on their surface, and they lacked effective coating; thus, QDs were severely toxic to cells. Furthermore, ion was more easily released from non–core/shell structure CdTe QDs and led to enormous toxicity.
The cell images revealed that -dTf had lower cytotoxicity and could be used to label live cells, indicating that dTf maintained their biological activities after coating onto QDs. CD spectra were employed to verify the result too. This showed that compared with Tf (Fig. 11 , curve c), little change of dTf’s peak and a small blue shift of the negative peak occurred (Fig. 11 curve a). Also there was little influence on the peak after dTf was coated onto QDs (Fig. 11, curve b). These data show that the structure of transferrin remained mostly intact after denaturalization. Partial allostery possibly happened and dTf maintained partial biological activities.
The study of quantum dots has become increasingly popular; but cytotoxicity is one of the key problems for their prevalent application. In this paper, denatured transferrin was adopted to modify the quantum dots to prepare quantum dot probes. The results showed that this special method could reduce the cytotoxicity and increase the optical properties of quantum dot probes over covalent conjugations. The PL QY of this probe was 7% higher than original QDs, and after exposure to a probe for , over 92% of HeLa cells were still alive. This work presents a method for preparation of QD bioprobes with high optical properties and low cytotoxicity and offers profitable help for biological application of QDs.
This work was supported by the National Natural Science Foundation of China (Grant No. 30670553) and the National High Technology Research and Development Program of China (863 Program: 2007AA10Z328). We also thank the Analytical and Testing Center (Huazhong University of Science and Technology) for help in measurement.