Enhanced affinity ligands

Abstract

The present invention relates to ligands, nanocrystal complexed with the ligands and their use for bio-imaging.

Claims

1. A ligand having colloidal stability properties and having the following formula II: ##STR00036## wherein R.sub.A has the formula -L.sub.A-M.sub.A: ##STR00037## wherein m is an integer ranging from 1 to 5, and p is an integer ranging from 1 to 6; R.sub.B has the formula -L.sub.B-M.sub.B: ##STR00038## wherein: q is an integer ranging from 1 to 5, s is an integer ranging from 1 to 5, and R.sub.12, R.sub.13, R.sub.14 and R.sub.15 are each independently H, or a group selected from an alkyl, alkenyl, aryl, hydroxyl, halogen, alkoxy and carboxylate; and each of x and y is independently a positive integer, R.sub.1, R.sub.2, R.sub.3, R.sub.4, R.sub.5, and R.sub.6 can be independently H, or a group selected from an alkyl, alkenyl, aryl, hydroxyl, halogen, alkoxy, or carboxylate, wherein the ligand has a molecular weight from about 1 000 g/mol to about 50 000 g/mol.

2. The ligand according to claim 1, wherein x+y is ranging from 5 to 500.

3. The ligand according to claim 1, being functionalized with at least one molecular probe and/or targeting group.

4. A nanocrystal which is complexed with at least one ligand according to claim 1.

5. The nanocrystal according to claim 4, wherein said nanocrystal is a 0D, 1D, or 2D nanocrystal.

6. The nanocrystal according to claim 4, wherein said nanocrystal is a nanosheet, a nanorod, a nanoplatelet, a nanoplate, a nanoprism, a nanowall, a nanodisk, a nanoparticle, a nanowire, a nanopowder, a nanotube, a nanotetrapod, a nanoribbon, a nanobelt, a nanoneedle, a nanocube, a nanoball, a nanocoil, a nanocone , a nanopiller, a nanoflower , or a quantum dot.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1: L2-QDs in pH 2, 3, 5, 7, 9, 11, 13 and saturated NaCl aqueous solutions (from left to right) under a 302-nm UV light (>12 months).

(2) FIG. 2: Principle of the dilution experiment to test L2-QD colloidal stability.

(3) FIG. 3: Influence of the capping ligand on QD colloidal stability (exchange conditions: 65? C./overnight; dilution experiment conditions: 0.3 ?M in 20 mM aqueous NaCl).

(4) FIG. 4: Influence of the exchange (L2/QD) temperature on colloidal stability of L2-QDs (dilution experiment conditions: 0.3 ?M in 20 mM aqueous NaCl).

(5) FIG. 5: Principle of the L2 vs L1 competition experiment.

(6) FIG. 6: Monitoring of fluorescein absorbance remaining on L2-fluorescein-QDs (squares) and accumulated in the surrounding solution (circles), during the exchange experiment between competing ligand L1 and adsorbed (capping) ligands L2-fluorescein (A) or L1-fluorescein (B).

(7) FIG. 7: Fluorescence microscopy images of HeLa cells, after incubating with L2-QDs (A) or L1-QDs (B) and washing, compared to cellular autofluorescence (C). Scale bars 5 ?m.

(8) FIG. 8: Fluorescence microscopy images of HeLa cells, 54 h after electroporation (Scale bars 5 ?m) and corresponding aggregation quantification. (A) L1-QDs tend to form aggregates (bright spots and heaps); (B) L2-QDs are still individual and move around freely within the cell; (C) Quantification of QDs' aggregation using Sd/M index for different surface chemistries.

(9) FIG. 9: Specificity of bioconjugated QDs. L2-biotin-QDs specifically bind to streptavidin-beads, while L2-streptavidin-QDs specifically bind to biotin-beads.

(10) FIG. 10: Ellman's dosage standard curve.

(11) FIG. 11: Ligand exchange L2/QDs.

(12) FIG. 12: Determination of the number of functionalizable amines per L2-fluorescein-QDs.

(13) FIG. 13: Determination of L2-fluorescein desorption rate constant.

(14) FIG. 14: Ligand exchange L2-PEG-NH.sub.2/QDs and bioconjugation with biotin or streptavidin.

(15) FIG. 15: Control of L2-PEG-NH.sub.2-QD bioconjugation with streptavidin (not to scale).

EXAMPLES

(16) The present invention is further illustrated by the following examples.

(17) Ligand design and synthesis. To answer the need of increasing ligand affinity for the nanocrystal surface, we directed our efforts towards ligands exhibiting several attachments points and turned logically to polymerization. We thus synthesized a hydrophilic polymer, L2, derived from the small molecule L1 presented above and resulting from a two-step process (Scheme 2). The first step consisted in the radical random copolymerization of two methacrylamides: one containing the precursor of the dithiol anchoring function (monomer A, obtained from the peptidic coupling between thioctic acid and N-(3-aminopropyl)methacrylamide), the other including the sulfobetaine group (monomer B, commercially available). Due to mismatching monomer solubilities, the solvent had to be adjusted to THF/water 1/1 (v/v) and the A/B ratio was optimized to a maximum of 20/80 to provide a water-soluble polymer. Note that the amounts of initiating (2,2-azobis(2-methylpropionamidine)dihydrochloride, V50) and terminating (3-mercaptopropionic acid, MPA) agents were both chosen equal to 10% molar equivalents relative to the total amount of monomers, in order to prevent the formation of too long chains, thus keeping a small ligand size. The reaction led thereby to bipolymer b-L2 (for bridged-L2) in 70% yield.

(18) ##STR00021##

(19) Then, in a second step, b-L2 was treated by sodium borohydride in water to reduce its disulfide bridge and free the anchoring thiol groups, giving the corresponding polymer L2 (initial A/B ratio=20/80).

(20) Another polymer, L3, was synthesized this way, using an initial A/B molar ratio equal to 10/90.

(21) Characterization of the products by gel permeation chromatography in water (Viscotek GPCmax, triple detection) confirmed the expected low number average molecular weights (5,000<M.sub.n<8,000 g.Math.mol.sup.?1, i.e. ?15-25 monomers per chain, for all synthesized polymers) and showed polydispersity indexes (PDI=M.sub.w/M.sub.n) below 3. The dosage of the dithiol functions using Ellman's method allowed, in turn, an estimation of the effective A/B ratio in the polymeric samples: 7/93 for L2 and 2/98 for L3. These A/B ratios are quite different from the initial ratios, probably due to differences in reactivity between the two monomers.

(22) The bipolymers were stored under their oxidized bridged form (b-L2, b-L3) and their reduction was carried out just before any ligand exchange procedure.

(23) Ligand exchange. The synthesized polymers were then dissolved in water and tested in a classical biphasic cap exchange with CdSe/CdS/ZnS core/shell QDs solubilized in chloroform. Surprisingly, the nanoparticles were extracted from the organic phase, but did not get into water, remaining at the interface. We attributed this observation to the poor solubility of the polymer in chloroform: it implies a low partition coefficient between the two solvents and a difficult phase transfer of the QDs. To overcome this problem, we moved to a two-step process. A first monophasic exchange using a small monodentate hydrophilic ligand, namely MPA, on as-synthesized QDs in basic chloroform, led to the precipitation of MPA-QDs and allowed the solubilization of the QDs into water at room temperature. Then, a second step, still monophasic, consisted in the removal of the weak intermediate ligand MPA from the QDs and its replacement by our polymer in an aqueous medium at 65? C. overnight.

(24) After purification of polymer L2-capped QDs by ultrafiltration then ultracentrifugation in an aqueous sucrose gradient, the physical properties of the nanoparticles were characterized by fluorescence spectroscopy and DLS. The nanocrystals showed a slightly reduced quantum yield (in the range of 30-40%, compared to 50-60% before cap exchange) and a suitable small size (15-18 nm as hydrodynamic diameter, for a core/shell diameter of 6-7 nm).

(25) L2-QDs were then subjected to a series of stability experiments to confirm the multiplication of anchoring and hydrophilic functions has a significant influence on the QD colloidal stability as well as on the ligand stability onto the nanoparticles itself.

(26) Stability versus pH and salinity. First, saline L2-QDs aqueous solutions (?0.6 ?M in 3 M aqueous NaCl) proved to be very stable over a large pH range, more precisely from pH 3 to 13, during several months and at 4? C. (>10 months, FIG. 1). The charges of the pending sulfobetaine groups of polymer L2, responsible for its hydrophilicity and for the colloidal stability of the corresponding QDs in water, are indeed insensitive to pH. Moreover, they could stand high salinity conditions, up to a saturated aqueous NaCl solution (?6 M), indefinitely. These properties are explained by the zwitterionic nature of the sulfobetaine groups whose charges are screened by ionic species, so that the attractive interaction between zwitterions is masked and the solvation of the corresponding polymer increases with the ionic strength. This specific feature constitutes a great advantage, compared to poly(ethylene glycol)-coated QDs, which can hardly stand highly saline media. In addition, this proves solubilizing zwitterionic QDs in highly saline or even saturated solutions can be a very convenient long-time storage method, even if the samples are diluted.

(27) Such a test is very common in the literature discussing QD surface chemistry, since stable samples are evidences of a good surface passivation. But it is only a pre-requisite and it is far from being enough to demonstrate QD stability in the conditions of a bio-imaging experiment, which are not only highly diluted, but also less saline and much warmer (37? C.), i. e. much less favorable than those used in this experiment.

(28) Stability versus dilution. Hence the affinity of the ligands for the QDs was tested through dilution experiments: QD samples were diluted to concentrations being in the range of 0.3 ?M in 20 mM aqueous NaCl and left at room temperature. In such conditions, the adsorption/desorption equilibrium existing between the capping ligand and the surface of QDs is shifted towards ligand desorption, so that ligand concentration in solution is homogenized (moderation principle). QDs lose their colloidal stability consequently and begin to aggregate (FIG. 2).

(29) Aggregation kinetics can therefore be monitored by the measurement of QDs' remaining absorbance in solution. Practically, the samples were centrifuged (16,000 g, 5 min) to remove aggregates and the absorbance at 350 nm of the corresponding supernatant was measured, this procedure being repeated over time. The influence of two parameters was underlined by our results: the type of capping ligand (L1, L2, L3) and the temperature of the second exchange step (polymer/QD) in water.

(30) To counter the desorption process, QD-ligand affinity must be as strong as possible. FIG. 3 clearly shows that an increasing number of anchoring functions per ligand prevents the aggregation of the nanoparticles (absorbance decrease with time slows down in the series L1, L3, L2), thus improving QD colloidal stability. More accurately, in the case of our polymers, we believe that the amount of dithiol functions per ligand bound to the QDs must be much greater than suggested by Ellman's assay. As the PDI are higher than 1 and the monomer distributions can vary throughout the polymeric chains, a sorting of the ligands can occur during the exchange step, to the benefit of the chains including the largest number of dithiols, which are therefore more stable. Note that the absorbance of the QDs coated with polymer L2 remained above 75% of the initial absorbance, even after a 20-day experiment under diluted conditions. This demonstrates a remarkable improvement of the colloidal stability over time, compared to L3-QDs, which lost their stability after 15 days and especially, compared to bidentate-ligand-L1-coated QDs, which precipitated in less than 2 days.

(31) Another relevant point is the role of the temperature of the aqueous final ligand exchange step. As shown by FIG. 4 in the case of the exchange reaction with L2, heating is required to produce nanoparticles stable against dilution. This finds its explanation in the competition which takes place during the exchange reaction, between the original and the polymeric ligands on one hand, and among the polymer molecules themselves on the other hand. An elevated temperature is needed to displace native ligands and replace them by the new ones, so that the QDs acquire a good colloidal stability; it can also be necessary to allow the longest chains, which are less mobile but enriched in linking functions, to bind to QDs in a thermodynamically controlled process.

(32) These experiments were repeated on several QD and polymer syntheses batches and proved to be highly reproducible. Nevertheless, some slight variations in stability can occur, but only after very long times of experiment (>1 month).

(33) After these characterizations of L2-QD colloidal stability, we wanted to understand it better and verify to which extent it could be attributed to the affinity of ligand L2 for the surface of the QDs.

(34) Stability versus a competing ligand. To further test the stability of L2 onto QDs and demonstrate the adaptability of the polymeric synthesis to diverse applications, we introduced a monomer including a functional group during the polymerization process. Monomer C, bearing a reactive amine end function, was added to and copolymerized with methacrylamides A and B, in an A/B/C ratio equal to 20/70/10, to give, after disulfide reduction, terpolymer L2-NH.sub.2. As an illustration, this functionalizable polymer could be coupled to a fluorescein carboxylic derivative and then, exchanged with QDs (Scheme 3). Absorbance measurements on these L2-fluorescein-QDs let us estimate a number of ten functionalizable amines per nanoparticle, after subtraction of QD characteristic absorbance.

(35) ##STR00022##

(36) These functionalized QDs proved to be particularly useful in the determination of L2-fluoresceinwhich is a modified L2-type ligandstability at the surface of QDs. To L2-fluorescein-QDs was opposed a 10,000-fold molar excess of competing ligand L1 (FIG. 5). In these conditions, and due to ligand adsorption/desorption dynamic equilibrium, re-adsorption of desorbed L2 was highly disfavored. Desorption of L2-fluorescein could be followed by the measurement of remaining fluorescein absorbance on QDs over time. As this measurement required the separation of the QDs from the surrounding solution (performed by ultrafiltration), cumulative fluorescein absorbance in solution could be monitored in parallel (FIG. 6, A).

(37) The same type of experiment was carried out with L1-fluorescein-QDs versus L1 (FIG. 6, B). In that case, the so-called L1-fluorescein-QDs were QDs covered with a mixture of L1 and a fluorescein-PEG-dithiol bidentate ligand, which were also subjected to a 10,000-fold molar excess of L1. From the modeling of absorption data by an exponential decay, we were able to estimate the corresponding apparent desorption rate constants in the presence of competing L1, namely k.sub.off L2-fluorescein?0.0045 h.sup.?1 and k.sub.off L1-fluorescein?1.5 h.sup.?1. These results demonstrate undoubtedly the overwhelming stability of L2-type ligands, compared to that of L1-type bidentate ligands.

(38) Stability in an intracellular medium. To conclude the study of L2-QD stability, their behavior in an intracellular environment was examined and compared to L1-QDs. First, HeLa cells were incubated with 1 ?M L1- or L2-QD solutions for 6 h. After washing, the fluorescence microscopy image of L2-QD-incubated cells (FIG. 7A) showed no significant difference regarding cellular autofluorescence (FIG. 7C), whereas cells treated with L1-QDs had a slightly higher fluorescence level (FIG. 7B). Even though this level can be considered as low on an absolute scale, this is indicative of a small amount of non-specific adsorption in the case of L1-QDs. As for L2-QDs, they do not present any detectable non-specific adsorption on cell membranes, which represents again a worthwhile improvement.

(39) Intracellular stability was checked, in turn, using electroporation as internalization process: via an electric shock, 1 ?M solutions of L1- or L2-QDs were incorporated in HeLa cells. The monitoring of QD fluorescence showed that, 54 h after incorporation, L1-QDs begin to aggregate (bright spots, FIG. 8A), while L2-QDs remain individualized, mobile and well distributed throughout cell cytoplasm (FIG. 8B). Quantification of QD aggregation was realized using a method developed in our lab. This method consists in estimating the Standard deviation of pixel fluorescence intensities normalized by their Mean (average) value over the whole cytoplasm (Sd/M index). Briefly, highly dispersed QDs yield uniform cytoplasm labeling and low Sd/M, while bright aggregates spots yield high Sd/M. This index makes a clear difference in favor of L2-QDs, not only compared to L1-QDs, but also to other QDs with various surface chemistries, including QDs encapsulated into poly(ethylene glycol)-phospholipid micelles (Micelle-QDs) and Qtracker?, commercial QDs encapsulated into an amphiphilic copolymer (FIG. 8C). L2-QD stability properties over time are thus definitely confirmed and meet all the relevant conditions for an application in long-term bioimaging.

(40) Towards biotargeting. The possibility of L2-QD functionalization has already been demonstrated by the incorporation of a fluorescent dye in a L2-type polymer before the ligand exchange. To further develop the opportunities for functionalization and tend to biotargeting, we designed and synthesized terpolymer L2-PEG-NH.sub.2 bearing a more reachable amine reactive group (longer spacer arm) than in terpolymer L2-NH.sub.2 (Scheme 4). The synthesis of the amine-containing monomer D was achieved using a previously described protocol.

(41) ##STR00023##

(42) The experiment of specific binding relied on the strong non-covalent affinity (K.sub.a?10.sup.14 L.Math.mol.sup.?1) between biotin (a small biomolecule) and streptavidin (a 60 kDa protein). After ligand exchange, L2-PEG-NH.sub.2-capped QDs were bioconjugated, via their amine function, with either biotin or streptavidin, using respectively a classical peptidic coupling based on the reagents DCC (dicyclohexylcarbodiimide) and NHS (N-hydroxysuccinimide), or a thiol/maleimide reaction. Bioconjugated QDs, when reacted with either biotin- or streptavidin-functionalized agarose beads, bound specifically to the beads bearing the complementary biomolecule. No non-specific binding could be pointed out (FIG. 9).

(43) Conclusion

(44) Whatever the extreme conditions opposed to our new polymeric ligand (pH, salinity, dilution or adsorption competition), the different experiments reported herein confirmed the outclassing stability of multidentate polyzwitterion L2 at the surface of CdSe/CdS/ZnS QDs, compared to bidentate monozwitterion L1. Moreover, this work led usfor the first time to our knowledgeto display some quantitative information about ligand desorption rates, and to better argue about the improved ligand anchoring the poly(dithiol)s are responsible for, with respect to single dithiol. Finally, L2 excellent properties resulted also in an increased colloidal and intracellular stability of the corresponding L2-coated quantum dots, making this polymer a remarkable ligand for long-term live-cell imaging experiments based on fluorescent nanocrystals.

(45) Additional Results

(46) Materials and Instrumentation

(47) Streptavidin was purchased from Biospa; APMA.HCl (N-(3-aminopropyl)methacrylamide hydrochloride) was purchased from Tebu-bio; SPP (3-sulfopropyldimethyl-3-methacrylamidopropylammonium inner salt; 3-[3-methacrylamidopropyl-(dimethyl)ammonio]propane-1-sulfonate), from Raschig GmbH (Ralu?Mer SPP); sulfo-1c-SPDP (sulfosuccinimidyl 6-(3-[2-pyridyldithio]-propionamido)hexanoate), sulfo-SMCC (sulfosuccinimidyl-4-(N-maleimidomethyl)cyclohexane-1-carboxylate) and DTT (dithiothreitol), from Pierce; all other chemicals used in this study (including functionalized agarose beads) were purchased from Sigma-Aldrich. All of these purchased chemicals were used without further purification unless otherwise specified. Dry THF was obtained from distillation on sodium/benzophenone ketyl. Chromatography on silica was carried out on Kieselgel 60 (230-240 mesh, Merck) and analytical TLC was performed on Merck precoated silica gel (60 F.sub.254); chemicals were visualized by heating with a solution of 5-7% phosphomolybdic acid in ethanol. .sup.1H NMR spectrum was recorded on a Bruker Avance DPX 400 spectrometer at 400.13 MHz. Chemical shifts (?) are expressed in ppm and coupling constant (J) in hertz. Absorption measurements were carried out with a Cary 5E UV-vis-NIR spectrophotometer (Varian). Fluorescence measurements were acquired using a Fluoromax-3? fluorimeter (Jobin Yvon, Horiba). Dynamic light scattering measurements (DLS) were performed on a CGS-3 goniometer system equipped with a HeNe laser illumination at 633 nm (Malvern) and an ALV 5000/EPP correlator (ALV).

(48) Polymeric ligands Syntheses

Synthesis of monomer A (5-(1,2-dithiolan-3-yl)-N-(3-methacrylamidopropyl)-pentanamide, Scheme 5)

(49) ##STR00024##

(50) To a suspension of APMA.HCl (2 g, 11.2 mmol) in dichloromethane (20 mL) was added triethylamine (2.5 mL, 17.9 mmol). Methanol (2 mL) was introduced to obtain complete solubilization. A solution of LA (2.76 g, 13.4 mmol) in dichloromethane (5 mL) was then added, followed by NHS (1.58 g, 13.8 mmol) in one portion. The reaction mixture was cooled down to 0? C. with an ice bath and a solution of DCC (dicyclohexyl carbodiimide) (3.00 g, 14.4 mmol) in dichloromethane (10 mL) was injected dropwise. The medium was warmed up to room temperature and further stirred overnight. A pale yellow solution containing a white precipitate was obtained. The solution was washed by a 0.1 M aqueous HCl solution (2?50 mL), deionized water (1?50 mL) and a 0.2 M aqueous NaOH solution (2?50 mL). The organic phase was separated, dried over MgSO.sub.4, filtrated and concentrated under reduced pressure. The crude residue was purified by chromatography on silica (eluent: hexane/ethyl acetate 1/4, then hexane/acetone 1/1) to give A (2.88 g, 8.71 mmol, 78%) as a pale yellow solid. R.sub.f=0.37 (hexane/acetone 1/1); .sup.1H NMR (CDCl.sub.3, 400 MHz): ? 7.03 (sl, 1H); 6.87 (sl, 1H); 5.72 (s, 1H); 5.29 (s, 1H); 3.53-3.39 (m, 1H); 3.29-3.20 (m, 4H); 3.14-3.01 (m, 2H); 2.43-2.35 (m, 1H); 2.18 (t, J=8.0 Hz, 2H); 1.92 (s, 3H); 1.88-1.80 (m, 1H); 1.68-1.55 (m, 6H); 1.48-1.33 (m, 2H).

Synthesis of polymer b-L2 (poly(5-(1,2-dithiolan-3-yl)-N-(3-methacryl-amidopropyl)pentanamide-co-3-[(3-methacrylamidopropyl)dimethylammonio]-propane-1-sulfonate), Scheme 6)

(51) ##STR00025##

(52) To a solution of B (SPP, 1.17 g, 4 mmol, 4 equiv.) in deionized water (20 mL) was added a solution of A (331 mg, 1 mmol, 1 equiv.) in THF (20 mL). A solution of V50 (2,2-azobis(2-amidinopropane) hydrochloride) (130 mg, 0.5 mmol, 0.5 equiv.) in deionized water (2 mL) was further added in one portion. The pale yellow mixture was stirred and degassed by argon bubbling for 40 min. MPA (42 ?L, 0.5 mmol, 0.5 equiv.) was injected into the reaction medium, which was stirred overnight at 60? C. under argon atmosphere. THF was evaporated under reduced pressure; the residual solution was extracted with 20 mL of dichloromethane and the aqueous phase was separated. A 9-fold excess of ethanol was poured into the latter phase to precipitate the polymer, which was separated by centrifugation (50-mL centrifuge tubes, 2,800 g, 10 min) and further dried overnight under vacuum in the presence of P.sub.2O.sub.5 as a desiccant. The polymer was obtained as an off-white solid (1.05 g, 70%).

(53) Synthesis of Polymer b-L3. b-L3 was synthesized in the same manner as b-L2, with an A/B molar ratio equal to 10/90 (A: 166 mg, 0.5 mmol; B: 1.28 g, 4.4 mmol; V50: 130 mg, 0.5 mmol; MPA: 42 ?L, 0.5 mmol).

(54) Ellman's Dosage (Scheme 7)

(55) ##STR00026##

(56) Dithiol groups of the different polymers were quantified using DHLA (dihydrolipoic acid) as a standard.

(57) A sodium phosphate buffer solution (0.1 M, pH=8) was prepared by dissolving sodium hydrogen phosphate (3.3 g, 23.3 mmol), sodium dihydrogen phosphate (0.2 g, 1.7 mmol) and EDTA (93 mg, 0.3 mmol) in water (250 mL).

(58) Lipoic acid (31 mg, 0.15 mmol) was dissolved in the sodium phosphate solution (5 mL) and the solution was cooled down to 0? C. (ice bath). NaBH.sub.4 (60 mg, 1.6 mmol, 10 equiv.) was added and the mixture was stirred at 0? C. for 30 min. Sulfuric acid (1.5 M, 3 mL) was added and the final volume was adjusted to 50 mL, using the sodium phosphate buffer solution (42 mL), to give a 3-mM solution of DHLA called thereafter Standard DHLA solution. A set of DHLA standards from 0 to 0.5 mM was prepared from this solution and from the sodium phosphate buffer (see Table 1).

(59) Unknown samples were prepared by dissolving L2 (100 mg) or L3 (200 mg) in sodium phosphate buffer (5 mL). The solutions were cooled down to 0? C., NaBH.sub.4 was added (30 mg) and the mixtures were stirred at 0? C. for 30 min. Sulfuric acid was added to each solution (1.5 M, 1.5 mL), then sodium phosphate buffer (3.5 mL). 2 mL of these solutions were diluted to a final volume of 10 mL to obtain the unknown sample solutions.

(60) Ellman's reagent solution was prepared by dissolving DTNB (5,5-dithio-bis-(2-nitrobenzoic acid)) (4 mg) in sodium phosphate buffer (1 mL). A set of test tubes was prepared, each containing Ellman's reagent solution (50 ?L) and sodium phosphate buffer (2.5 mL).

(61) Each DHLA standard or unknown sample (250 ?L) was added to separate test tubes.

(62) Solutions were mixed, incubated at room temperature for 15 min, and their absorbance at 412 nm was measured. DHLA standards were used to generate a standard curve (FIG. 10) that allowed the determination of unknown concentrations. Results are summarized in Table 1.

(63) TABLE-US-00001 TABLE 1 Ellman's dosage: standard curve generation and determination of unknown sample concentrations. V.sub.Standard DHLA sol. [dithiol] Absorbance [dithiol].sub.exp Sample V.sub.sodium phosphate buffer (mL) (mM) at 412 nm (mM) Standard 1 4 0 0 0.000 Standard 2 3.917 0.083 0.0625 0.142 Standard 3 3.833 0.167 0.125 0.285 Standard 4 3.667 0.333 0.25 0.572 Standard 5 3.333 0.667 0.5 1.158 L2 1.045 0.453 L3 0.693 0.300

(64) Estimation of the real A/B ratio in polymers L2 and L3: Average number of monomer A/polymeric chain:

(65) $.Math.N_A.Math.=\frac{{\left[\mathrm{dithoil}\right]}_{\exp }}{{\left[\mathrm{polymer}\right]}_{\mathrm{sample}}}$ Average number of monomers/polymeric chain:

(66) $.Math.N_A+N_B.Math.?\frac{M_n\left(\mathrm{polymer}\right)}{M_A}\left(M_A?M_B?300g.{\mathrm{mol}}^{-1}\right).Math.\{\begin{array}{c}A\text{/}B?7\text{/}93\mathrm{in}L2\\ A\text{/}B?2\text{/}98\mathrm{in}L3\end{array}$
CdSe/CdS/ZnS QDs Synthesis

(67) 600-nm-emitting CdSe/CdS/ZnS QDs were synthesized using slight modifications of previously published procedures. CdSe cores were synthesized by reaction of trioctylphosphine selenide and cadmium oleate in octadecene, oleylamine and trioctylphosphine oxide. Three monolayers of CdS shell, followed by two monolayers of ZnS, were grown using cadmium oleate, zinc oleate and sulfur diluted in octadecene following the SILAR (Successive Ionic Layer Adsorption and Reaction) procedure.

(68) Ligand Exchange L2/QDs: Standard Procedure (Scheme 8 and FIG. 11)

(69) ##STR00027##

(70) CdSe/CdS/ZnS core/shell QDs in hexane (4 nmol) were precipitated with ethanol and centrifuged (16,000 g, 10 min). The supernatant was removed, the QDs were redispersed in hexane (0.2 mL) and the procedure was repeated once. The QDs were then taken up in chloroform (1 mL). MPA (100 ?L, 1.1 mmol) was dissolved in a freshly prepared solution of TMAOH.5H.sub.2O (400 mg, 2.2 mmol) in chloroform (2 mL), using a sonicating bath. 1 mL of the basic organic phase was added to the QD colloidal dispersion. The mixture was stirred, then left at room temperature. After 15-30 min typically, MPA-QDs aggregated. The suspension was centrifuged to remove the basic organic supernatant (16,000 g, 5 min) and the nanoparticles were washed twice with chloroform (brief stirring and centrifugation at 16,000 g, 5 min). MPA-QDs were taken up in a 10-mM sodium tetraborate buffer (2 mL, pH=9).

(71) A solution of b-L2 (40 mg) in deionized water (2 mL) was cooled down to 0? C. with an ice bath and NaBH.sub.4 (10 mg) was added in one portion. The solution was warmed up slowly to room temperature and stirred for 30 min. The aqueous solution of MPA-QDs was added (2 mL) and the mixture was stirred vigorously at room temperature. After 30 min, the vial containing the aqueous mixture was sealed and stored without stirring at 65? C. overnight to complete cap exchange. The L2-QD aqueous solution was cooled down to room temperature and excess free solubilized ligands and reagents were removed by two rounds of membrane ultrafiltration at 16,000 g using a Sartorius Vivaspin? 500 ?L disposable filter (cutoff 30 kDa) in 20 mM aqueous NaCl.

(72) L2-QDs were purified by ultracentrifugation at 268,000 g for 25 min in a 10%-40% sucrose gradient in 20 mM aqueous NaCl. The QD band was collected, residual sucrose was removed by four rounds of ultrafiltration (Vivaspin? 500 ?L, cutoff 30 kDa, 16,000 g, 10 min) and L2-QDs were further washed by five rounds of ultrafiltration with a 20 mM NaCl aqueous solution (Vivaspin? 500 ?L, cutoff 30 kDa, 16,000 g, 10 min). L2-QDs were finally taken up in 20 mM aqueous NaCl.

(73) Stability Vs pH and Salinity

(74) L2-QDs (12 ?L, 16 ?M in 20 mM aqueous NaCl) were added to solutions of different pH prepared from HCl or NaOH solutions (300 ?L, pH from 1 to 13) containing NaCl (50 mg), or to a saturated aqueous solution of NaCl. L2-QDs colloidal solutions were kept at 4? C. over months and stability (i. e. possible aggregation) was controlled by centrifugation (16,000 g, 5 min).

(75) Stability Vs Dilution (FIG. 2)

(76) The concentrations of the different QD solutions in 20 mM aqueous NaCl were determined by the measurement of their absorbance at 350 nm and the corresponding samples were diluted to 0.3 ?M by addition of a 20 mM aqueous NaCl solution. The volume of each sample was in the range of 1 mL. Before each measurement, diluted QD samples were centrifuged at 16,000 g for 5 min. Aggregated QDs fell down at the bottom of the centrifuge tube and the absorbance at 350 nm of the supernatant was then measured. This solution was eventually recovered and left at room temperature until next measurement.

(77) QD samples diluted in a 1 M aqueous NaCl solution were washed beforehand by three rounds of ultrafiltration with a 1 M NaCl aqueous solution (Vivaspin? 500 ?L, cutoff 30 kDa, 16,000 g, 10 min).

(78) Ligand Competition Experiments

(79) Competition L2 Fluorescein vs L1

(80) Synthesis of b-L1 . See the supporting information in a previous publication from our lab (Scheme 9)

(81) ##STR00028## Synthesis of Polymer b-L2-Fluorescein Synthesis of Fluorescein-NHS (Scheme 10)

(82) 5(6)-carboxyfluorescein (200 mg, 0.53 mmol, 1 equiv.) was dissolved in DMF (2 mL). NHS (61 mg, 0.53 mmol, 1 equiv.) was added in one portion, then a solution of DCC (110 mg, 0.53 mmol, 1 equiv.) in DMF (0.4 mL), in one portion. The reaction mixture was protected from light and stirred overnight at room temperature. A white precipitate formed, which was filtered and the filtrate was diluted with DMF to obtain a final volume of 5 mL. The reaction was supposed to be quantitative and the solution was stored in the dark at 4? C. until use.

(83) ##STR00029## Synthesis of polymer b-L2-NH.sub.2 (poly(N-(3-aminopropyl)methacrylamide-co-5-(1,2-dithiolan-3-yl)-N-(3-methacrylamidopropyl)pentanamide-co-3-[N,N,N-(3-methacrylamidopropyl)-dimethyl-ammonio]propane-1-sulfonate, Scheme 11)

(84) APMA.HCl (89 mg, 0.5 mmol, 1 equiv.), then triethylamine (140 ?L, 1 mmol, 2 equiv.) and B (1.03 g, 3.5 mmol, 7 equiv.) were dissolved in deionized water (20 mL). To this solution was added a solution of A (331 mg, 1 mmol, 2 equiv.) in THF (20 mL). A solution of V50 (130 mg, 0.5 mmol, 1 equiv.) in deionized water (2 mL) was further added in one portion. The pale yellow mixture was stirred and degassed by argon bubbling for 40 min. MPA (42 ?L, 0.5 mmol, 1 equiv.) was injected into the reaction medium, which was stirred overnight at 60? C. under argon atmosphere. THF was evaporated under reduced pressure; the residual solution was extracted with 20 mL of dichloromethane and the aqueous phase was separated. A 9-fold excess of ethanol was poured into the latter phase to precipitate the polymer, which was separated by centrifugation (2,800 g, 10 min) and further dried overnight under vacuum in the presence of P.sub.2O.sub.5 as a desiccant. The polymer was obtained as an off-white solid (0.69 g, 48%).

(85) ##STR00030## Functionalization of b-L2-NH.sub.2 by a Fluorescein Dye (Scheme 11)

(86) To a solution of b-L2-NH.sub.2 (40 mg) in aqueous NaHCO.sub.3 (1.25 mL, 0.2 M, pH=9) was added a solution of fluorescein-NHS in DMF (750 ?L, 0.106 M, ?10 equiv.). The reaction mixture was stirred 2 h at room temperature. The labeled polymer b-L2-fluorescein was purified by several rounds of ultrafiltration (Vivaspin? 500 ?L, cutoff 3 kDa, 16,000 g) until the filtrate was not fluorescent anymore. The resulting residue was employed for the ligand exchange without further purification. Ligand Exchange (Scheme 12)

(87) CdSe/CdS/ZnS core/shell QDs (4 nmol) were exchanged with 20 mg of b-L2-fluorescein, treated beforehand by NaBH.sub.4 (10 mg) for 30 min to afford L2-fluorescein, according to the standard two-step process described above for ligand exchange with L2.

(88) ##STR00031## Number of Functionalizable Amines Per L2-Fluorescein-QDs

(89) The absorbance at 500 nm of fluorescein-NHS was measured in 0.2 M aqueous NaHCO.sub.3 and let us determine the corresponding molar extinction coefficient: 249 =53,990 L.Math.mol.sup.?1.Math.cm.sup.?1. The absorbance of L2-fluorescein-QDs was measured in the same conditions from 350 to 700 nm. The spectrum was then deconvoluted to separate QD and fluorescein absorbances, as exemplified in FIG. 12. QD absorbance at 350 nm led us to QD concentration; fluorescein absorbance at 500 nm and the corresponding c, to dye concentration. The concentration ratio gave a number of 10 dyes per QD, that is, 10 functionalizable amines per QD. Ligand Competition and Measurement of L2-Fluorescein Desorption Rate Constant

(90) A solution of b-L1 (100 mg) in 20 mM aqueous NaCl (1 mL) was cooled down to 0? C. with an ice bath and NaBH.sub.4 (20 mg) was added in one portion. The solution was warmed up slowly to room temperature and stirred for 30 min. To 120 ?L of this L1 solution were added 100 ?L of 1 M aqueous HCl, then 480 ?L of 1 M aqueous NaHCO.sub.3.

(91) L2-fluorescein-QDs (19 ?L, 32.3 ?M, 0.6 nmol in 0.2 M aqueous NaHCO.sub.3, pH=9) were added to the resulting mixture and the absorbance of the solution was measured from 400 nm to 700 nm. This solution, containing L2-fluorescein-QDs and the competing ligand L1, was transferred to a Vivaspin? 500 ?L disposable filter (cutoff 30 kDa) and left at room temperature.

(92) Before each next measurement, L2-fluorescein-QDs were separated via ultrafiltration (16,000 g, 10 min) and the surrounding solution was recovered. The absorbance of this surrounding solution was measured and cumulated with previous measurement(s). A freshly prepared 700-?L solution of L1 was added to the QDs and the absorbance of this QD solution was, in turn, measured. Each spectrum of L2-fluorescein-QDs was the deconvoluted to obtain the remaining fluorescein absorbance on QDs.

(93) Measurements were normalized at t=? and t=0 respectively, to give the evolution of remaining fluorescein on L2-fluorescein-QDs and the evolution of cumulative fluorescein in solution surrounding L2-fluorescein-QDs, as a function of time. Exponential fittings of experimental data led to the determination L2-fluorescein desorption rate constant, as illustrated in FIG. 13.

(94) Competition L1-Fluorescein vs L1 Synthesis of b-L1. See Scheme 9 and the supporting information in a previous publication from our lab. Synthesis of b-L1-Fluorescein. b-L1-fluorescein is a LA-PEG-NH.sub.2 molecule labeled with fluorescein via fluorescein-NHS (see above for the synthesis of fluorescein-NHS, Scheme 10). LA-PEG-NH.sub.2 was synthesized in a five-step process already described and starting from a PEG600 polymer (Scheme 13).

(95) ##STR00032## Ligand Exchange

(96) CdSe/CdS/ZnS core/shell QDs in hexane (4 nmol) were precipitated with ethanol and centrifuged (16,000 g, 10 min). The supernatant was removed, the QDs were redispersed in hexane (0.2 mL) and the procedure was repeated once. The QDs were then taken up in chloroform (1 mL). b-L1 (100 mg, 1.6 ?mol) in deionized water (1 mL) was treated by NaBH.sub.4 (20 mg) for 30 min, L1 solution was added to the QDs in chloroform and the biphasic mixture was heated at 65? C. overnight. The aqueous phase was separated and concentrated by ultrafiltration (Vivaspin? 500 ?L, cutoff 30 kDa, 16,000 g). L1-QDs were washed with 20 mM aqueous NaCl via three rounds of ultrafiltration (Vivaspin? 500 ?L, cutoff 30 kDa, 16,000 g) and taken up in 20 mM aqueous NaCl (200 ?L). A solution of b-L1-fluorescein, treated beforehand with NaBH.sub.4 to give L1-fluorescein (200 ?L, 25 mM in 20 mM aqueous NaCl, ?10% mol relative to QD-coating-L1), was added to L1-QDs and the mixture was heated at 65? C. overnight. Typical treatment (concentration of the sample, washings with aqueous 20 mM NaCl using ultrafiltration) and purification (ultracentrifugation in a 10%-40% sucrose gradient in 20 mM aqueous NaCl, concentration and washings with 0.2 M aqueous NaHCO.sub.3 using ultrafiltration) afforded L1-fluorescein-QDs in 0.2 M aqueous NaHCO.sub.3. Ligand Competition and Measurement of L1-Fluorescein Desorption Rate Constant

(97) A solution of competing ligand L1 was prepared as reported above for the measure of L2-fluorescein desorption rate constant. L1-fluorescein-QDs (52 ?L, 11.4 ?M, 0.6 nmol in 0.2 M aqueous NaHCO.sub.3, pH=9) were added to the L1-containing mixture (10 mM) and the absorbance of the solution was measured from 400 nm to 700 nm. This solution, containing L1-fluorescein-QDs and the competing ligand L1, was transferred to a Vivaspin? 500 ?L disposable filter (cutoff 30 kDa) and left at room temperature. Before each next measurement, L1-fluorescein-QDs were separated via ultrafiltration (16,000 g, 10 min) and the surrounding solution was recovered. The absorbance of this surrounding solution was measured and cumulated with previous measurement(s).

(98) The time scale of this experiment being far shorter than the determination of L2-fluorescein desorption rate constant, it was not necessary to add systematically a freshly prepared solution of L1 to the QDs before QD absorbance measurement. The initial solution of L1 could be re-used as prepared, except for the last measurement.

(99) Data treatment was performed according to the procedure detailed for L2-fluorescein desorption rate constant in FIG. 13.

(100) Stability in an Intracellular Medium

(101) Cell Culture

(102) HeLa cells were grown in DMEM medium supplemented with 10% FBS and 1% antibiotics.

(103) Non-Specific Adsorption on HeLa Cell Membranes

(104) Cells were incubated at 37? C. with L1- or L2-QDs diluted to 1 ?M in Opti-MEM? for 10 min (FIG. 7). Cells were then rinsed five times (centrifugation) and imaged. The non-specific adsorption on cell membranes of L2-QDs was comparable with L1-QDs and is almost not detectable.

(105) All images were acquired with the same parameters, using a widefield epifluorescence microscope (IX71 Olympus), a 60?1.2 NA water objective and an EM CCD camera (cascade 512B Roper). Excitation and collection of fluorescence of QDs were performed with 425/60 nm and 605/40 band pass filters.

(106) Electroporation of HeLa Cells

(107) L1- or L2-QDs were diluted in DMEM to 1 ?M in a final volume of 100 ?L and mixed to 50?10.sup.4 cells in suspension in a 2 mm electroporation cuvette. The cuvette was subjected to 0.15 kV for a 28 ms pulse using a Gene Pulser (Biorad) electroporator. Cells were rinsed 3 times, and deposed on LabTek in DMEM/F12 medium supplemented with 10% FBS, 1% antibiotics and 1% HEPES. Cells were imaged after 54 h at 37? C., 5% CO.sub.2.

(108) All images were acquired with a widefield epifluorescence microscope (IX71 Olympus) using a 60?1.2 NA water objective, Chroma filters and an EM CCD camera (cascade 512B Roper).

(109) Aggregation quantifications with micelle-QDs and Qtracker?, showed for comparison in FIG. 8C, were performed in the same conditions as those described above. Micelle-QDs were prepared from the encapsulation of CdSe/CdS/ZnS QDs in 100% 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-[methoxypoly(ethylene glycol)-2000]carboxamide (Nova) micelles (Scheme 14). Qtracker? 655 non-targeted QDs were purchased from Invitrogen.

(110) ##STR00033##
Towards Biotargeting

(111) Monomer D synthesis. Monomer D was synthesized from tetra(ethylene glycol), according to a four-step process (Scheme 15), adapted from a protocol initially developed with a PEG600 as starting material.

(112) ##STR00034##

(113) b-L2-PEG-NH.sub.2 synthesis. The synthesis of this terpolymer (Scheme 16) was carried out according to the procedure described for the synthesis of L2-NH.sub.2 (see the Competition experiment section).

(114) ##STR00035##

(115) Ligand exchange L2-PEG-NH.sub.2/QDs. Ligand exchange was performed according to the standard procedure indicated for L2, with 20 mg of b-L2-PEG-NH.sub.2 treated by 10 mg of NaBH.sub.4, for 4 nmol of CdSe/CdS/ZnS QDs (FIG. 14).

(116) L2-PEG-NH.sub.2-QD/biotin coupling (FIG. 14). 0.1 mmol of biotin were mixed with equimolar amounts of DCC and NHS in 10 mL DMF and stirred at room temperature overnight to yield a solution of 10 mM NHS-activated biotin. Then 0.5 nmol of L2-PEG-NH.sub.2-QDs were diluted in 400 ?L sodium bicarbonate buffer (0.2 M, pH=9), and mixed with 50 nmol of NHS-biotin for 30 min. The QDs were then purified with one round of ultrafiltration on Vivaspin 30 kDa (16,000 g, 10 min), one filtration on a NAP-5 column (GE Healthcare) and one final round of ultrafiltration (Vivaspin 30 kDa, 16,000 g, 10 min).

(117) L2-PEG-NH.sub.2-QD/streptavidin coupling (FIG. 14). Typically, QDs capped with L2-PEG-NH.sub.2 (1 nmol) were dispersed in a NaHCO.sub.3 buffer (400 ?L, 0.2 M, pH=9) and mixed with a sulfo-1c-SPDP solution (30 ?L, 10 mg/mL in DMSO) for 15 min. The solution was then concentrated using ultrafiltration (30 kDa MW cutoff, Vivaspin?, Vivascience, 16,000 g, 10 min) and diluted in a NaHCO.sub.3 buffer (400 ?L, 0.2 M). Then a DTT solution (20 ?L, 23 mg/mL in DMSO) was added, and the solution was stirred for 15 min. The DTT was eliminated using one round of ultrafiltration (16,000 g, 10 min), followed by purification on a NAP-5 column and another round of ultrafiltration (16,000 g, 10 min). The resulting concentrated QD solution was diluted in a HEPES buffer (300 ?L, 0.1 M, pH=7).

(118) In another vial, streptavidin (1 mg) diluted in a NaHCO.sub.3 buffer (200 ?L, 0.2 M) was mixed with a sulfo-SMCC solution (2.6 ?L, 10 mg/mL in DMSO) for 20 min, concentrated with one round of ultrafiltration (30 kDa MW cutoff Vivaspin?, 16,000 g, 10 min) and rediluted with a HEPES buffer (300 ?L, 0.1 M).

(119) The resulting freshly prepared QD and streptavidin solutions were mixed together and stirred for 20 min at room temperature. The resulting streptavidin-QD solution was then concentrated using ultrafiltration (16,000 g, 10 min). Excess streptavidin was eliminated using ultracentrifugation on a 10%-40% sucrose gradient (268,000 g, 25 min), and the functionalized streptavidin-QDs were finally reconcentrated using ultrafiltration (16,000 g, 10 min) and redispersed in HEPES buffer.

(120) Control of the Bioconjugation with Functionalized Agarose Beads. The efficiency of the conjugation was checked by mixing, during 15 min, 6 pmol of L2-streptavidin-QDs with biotin-functionalized agarose beads. The beads were washed four times in HEPES by mild centrifugation. The control was performed by mixing L2-streptavidin-QDs with streptavidin-functionalized agarose beads. Only the biotin-functionalized beads appeared fluorescent (FIG. 15).

(121) The bioconjugation of biotin to form L2-biotin-QDs was tested in the same way on streptavidin-functionalized agarose beads, with a control on biotin-functionalized agarose beads.