Recent Advances in the Use of Cadmiumfree Quantum Dots for Bioimaging

Introduction

Quantum dots (QDs) are nanometer-sized (2–10 nm) semiconducting crystalline materials that exhibit size-dependent optoelectronic properties. Most notably, the wavelength of their photoluminescence (PL) emission tracks directly with the size of the material as synthesized. In the last few decades, QDs have garnered tremendous attention as nanoparticle (NP) platforms for various applications in electronic devices, bioimaging, biosensing, and drug delivery due to their unique features. Among the many outstanding optical properties of QDs, their large absorption coefficients, narrow emission bands, large stoke shifts, high PL quantum yield (QY), and excellent photostability are highly advantageous for biological imaging.¹ For example, a sample can be labeled with different sized QDs and excited with a single excitation source far removed to the ultraviolet portion of the spectrum to produce distinct bright emission colors in the visible, making QDs particularly suitable for multiplexed imaging. QDs can be coated with different hydrophilic surface ligands with functional chemical groups (e.g., carboxylic acid, amine) to impart colloidal stability in aqueous environments and to facilitate conjugation to biologicals (peptides, polymers, and ligands), dyes, and drugs.

QDs for biological applications typically consist of a core/shell structure capped with hydrophilic ligands. To date, QD systems based on CdSe/ZnS structures have been the primary focus of researchers due to ease of synthesis and post-synthesis surface modification. Typically, the core of these QDs are composed of elements from groups II-VI (e.g., CdTe and CdSe), with a shell composed of ZnS. The shell protects against degradation of the core and maintains QD PL while also serving as an anchor point for ligand attachment. However, despite numerous studies demonstrating the stability and the largely innocuous nature of Cd-based QD systems in vitro and in vivo,^2–4 concerns regarding the long-term persistence of Cd-containing QDs and associated heavy metal-induced toxicity have spurred interest in the development of alternative QD configurations. Over the last decade, many research groups have focused on developing fabrication methods to synthesize different sized Cd-free QDs with various core/shell configurations and surface coatings. Here, the primary focus has been developing biocompatible Cd-free QDs with optoelectronic and surface properties comparable to existing Cd-based QDs. Among the various non-Cd QDs developed, indium phosphide (InP), copper indium sulfide (CuInS₂), and graphene QDs have been highly studied for biosensing and bioimaging applications. Although many properties of these non-Cd QDs are comparable to the Cd-based QDs, their surface chemistry and colloidal and optical stability in physiological environments have lagged behind that of Cd QDs. This review provides a snapshot of the current state of the art for the design, synthesis, and surface modification strategies for non-Cd QDs for use in bioimaging and sensing applications. We focus our discussion on the progress made in the last 5 years and highlight the critical challenges to overcome to advance this field further.

Structural Constructions of Cd-free QDs

Like Cd-based QDs, Cd-free QDs also have a size range of 2–10 nm and are comprised of core/shell architectures. However, Cd-free QD cores are typically comprised of elements from groups III–V (e.g., InP) (Figure 1). These cores, particularly InP, are structurally more robust and stable due to covalent bonds in their matrix.^5,6 However, the PL intensity and QY of these cores have generally been lower than Cd-containing core QDs. Several studies have used shell coatings with a broader bandgap material to overcome this challenge, such as ZnS and/or ZnSe layers. These shell coatings 1) reduce interactions between the exciton and the outer surface of the core nanocrystal to minimize surface defects and quenching, 2) facilitate better control of emission wavelength, PL lifetime, and QY, and 3) enhance chemical reactivity for ligand exchange. Based on the core/shell combination and band alignment of the valence and conduction bands of the constituent materials, these QDs can be divided into three types (i.e., type I, type II, and quasi type II), described elsewhere.⁷ In general, type I core/shell QDs possess a conduction band of the shell at higher energy than that of core, whereas the valence band of the shell is at lower energy relative to the core. Cd-free type I core/shell QDs, such as InP/ZnS, CuInS₂/ZnS, AgInS₂/ZnS, and ZnSe/ZnS, have been widely developed for various biological applications (Figure 1).⁸ In the following section, we discuss some of these examples and highlight their contribution to developing Cd-free QDs for bioimaging and sensing.

Bioimaging and Sensing with InP and CuInS2 QDs

InP/ZnS has emerged as the most popular type of Cd-free QD for use in bioimaging.⁹ Wu et al. conjugated InP/ZnS QDs to an anti-vascular endothelial growth factor receptor-2 (VEGFR-2) monoclonal antibody and loaded them with miR-92A miRNA to inhibit proliferation of cancerous myeloid cells.¹⁰ In this way, they produced a bifunctional InP nanocomposite (IMAN) for targeted imaging and therapy. In K562 leukemia cell models examined using near-infrared (NIR) imaging, the IMAN fluoresced and localized to the plasma membrane before entering the cells (Figure 2Ai). The researchers also demonstrated that greater doses of IMAN incubated with K562 cells for 36 hours led to more efficient cell killing than InP QDs or miR inhibitor alone (Figure 2Aii). In vivo, the use of the VEGFR-2 had a significant effect on the specificity of tumor localization. K562 tumorbearing nude mouse models were given tail vein injections of phosphate-buffered saline (PBS), 10 mg/kg InP QDs, or 10 mg/ kg IMAN and tracked using NIR imaging (Figure 2Aiii). Notably, while InP QDs were diffusely located within the abdomen, only IMAN localized to the tumor site. Further, the group showed that a consistent intravenous treatment of 10 mg/kg IMAN over 20 days led to significantly higher tumor volume reduction within the mouse model than the free miR 92a inhibitor (Figure 2Aiv).

Similarly, CuInS₂-based QDs have also emerged as another platform for developing heavy metal-free QDs for bioimaging. Yang et al. used 2-[bis[2-[carboxymethyl-[2-oxo-2-(2 sulfanylethylamino) ethyl]amino]ethyl]amino]acetic acid (DTDTPA)-modified CuInS₂/ZnS QDs chelated with gadolinium ions (for MRI) for a bimodal imaging nanoparticle (QDs@ DTDTPA-Gd NPs) capable of use with NIR fluorescence or magnetic resonance imaging (MRI) (Figure 2Bi).¹¹ DTDTPA was chosen due to its water solubility and ability to chelate gadolinium while being able to attach to the shell of the QD. Within a HeLa-tumor bearing nude mouse model, increasing magnetic resonance signal intensity was observed within the tumor over 48 hours, subsequent to an intravenous dose of 20 mg/kg QDs@DTDTPA-Gd NPs (Figure 2Bii). A similar increase in fluorescence intensity from the QD was also observed (Figure 2Biii). Encouragingly, organs harvested from sacrificed mice after 48 hours showed that QDs@DTDTPA-Gd NPs localized primarily to the liver, tumor, and spleen via passive targeting accumulation (Figures 2Biv, 2Bv).

Another research team applied CuInS₂/ZnS QDs embedded within a glycol-chitosan matrix (GCM). After coating with 11-mercaptoundecanoic acid (MUA) to prime for use in aqueous solutions, conjugation of the composite to RGD peptides for improved tumor targeting to produce cRGDyk-GCM-QDs (Figure 3Ai).¹² Examination of in vitro cell viability for GCM-QDs and MUA-QDs using an MTT assay found no significant loss in viability. In vivo, the researchers used MRI fluorescence imaging to observe localization of intravenously injected cRGDyk-GCMQDs at the site of the subcutaneous RR1022 xenograft tumor within nude mice (Figure 3Aii).

Bioimaging and Sensing with Other Types of Cd-free QDs

Literature reports other Cd-free QD configurations, such as an InP core. For example, Zhang et al., formulated an oil-soluble InP/ZnSe/ZnS core/multi-shell QD. This InP/ZnSe/ZnS QD is unique in its ability to emit at two distinct wavelengths with peak emissions in the visible and near-infrared (NIR) (Figures 3Bi, ii).¹³ The surface of the InP/ZnSe/ZnS QDs was modified with a poly(acrylic acid)-octylamine amphiphilic (PAA) polymer grafted with an Arg-Gly-Asp (RGD) peptide for water solubility and tumor localization. Within a Bcl-7402 tumor-bearing mouse model, the PAA polymer/RGD-modified QDs were visually located at the tumor site using NIR fluorescence within one hour after intravenous injection of a 1 mg/mL solution of QD (Figure 3Biii). The control MCF-7 tumor-bearing mouse model, which lacks the protein targeted by RGD, did not exhibit targeted localization to the tumor site. The case of the Bcl-7402 mouse model treated with QDs lacking the RGD modification obtained similar results.

Although used less commonly than InP/ZnS and CuInS₂/ZnS QDs, the adaptation of other types of Cd-free QDs found use in physiological applications. Zn-doped AgInS₂ (AIZS) QDs were utilized by Zang et al. to create biocompatible AIZSgraphene oxide (AIZS-GO) nanocomposites.¹⁴ To produce these, oleylamine-modified GO was assembled to AIZS QDs using thermal decomposition and mini emulsion (Figure 3Ci). Incubation of the AIZS-GO nanocomposites occurred in SK-BR-3 breast cancer cells, and proliferation assay (MTT) accessed the cell viability. At 0.8 μg/mL, the cells retained 84% viability, suggesting acceptable biocompatibility and low cytotoxicity of the AIZS-GO nanocomposites (Figure 3Cii). The group further demonstrated a brief in vivo proof of ability to optically image and distinguish the AIZS-GO nanoparticles within a SK-BR-3 cancer cell-bearing mouse model (Figure 3Ciii). QDs based on ZnSe core structures are another Cd-free alternative. Zhou et al. used manganese-doped ZnSe QDs as a dual MRI/fluorescent imaging probe.¹⁵ These Mn-ZnSe QDs were loaded into mesoporous silica particles to create MSN@QDs (Figure 3Di) that were further decorated with a tumor-targeting transferrin ligand to make TRF MSN@QDs. The group observed increased cellular uptake with the transferrin ligand in an in vitro 4T1 cell model as measured by cellular fluorescence intensity (Figure 3Dii). In a nude mouse, subcutaneous injections of single-QDs, MSN@QDs, and TRF-MSN@QDs were given, and the fluorescence was examined (Figure 3Diii). While single- QDs did not register above background fluorescence levels, MSN@QDs and TRF-MSN@QDs showed significant fluorescence intensity, suggesting that the enriching the QDs aided in bioimaging capability.

Critical Considerations, Toxicity, and Future Outlook

Recent significant advancements in developing non-Cd QDs for bioimaging and sensing show promise for alternative NP materials for biological applications. Still, challenges remain in terms of the long-term persistence of the core elemental components (e.g., In, Cu, Se). Just as with Cd-containing QDs, the potential leaching of metal ions from within the QD core is influenced by the physicochemical properties of the QDs, including size, shape, and surface ligand functionalization. These properties primarily determine the fate of the QDs’ biodistribution, biodegradation, clearance, and elimination when administered in vivo. Currently, the primary technical challenge facing non-Cd QD development is focused on improved colloidal stability, biocompatibility, and targeting. Here, lessons learned from Cd QDs guide the engineering approach, but our current knowledge about the chemistry and synthetic parameters of non-Cd QDs impedes progress.

It is still challenging to make non-Cd QDs that rival the narrow PL bandwidth, high QY, and ease of spectral tuning (UV-NIR) available in the Cd QD series. It is an even further challenge to maintain these properties when the QDs are transferred to water. To date, the highest quality Cd QDs for use in biological applications are core/shell structures where a wider-band gap inorganic shell encapsulates the core and ideally prevents leaching of potentially toxic metal ions. ZnS shells are commonly used on Cd QDs to create a non-toxic outer surface, improve their PL, and increase their chemical and photostability. Additionally, biocompatible hydrophilic ligands and bioconjugation strategies have been optimized for ZnS coated QDs. Thus, non-Cd QDs should be amenable to incorporating a ZnS shell to transfer existing technology and ease their transition for replacing Cd-based QDs.

In particular, researchers have focused heavily on InP QDs as a less toxic alternative to Cd-based QDs. However, refinement of their properties via chemical synthesis is complicated by their covalent nature, available phosphorus precursors, sensitivity towards oxidation, and reaction mechanisms (nucleation and growth).¹⁶ Bare InP QDs are notorious for being prone to oxidation while at the same time exhibiting poor QYs. It is, therefore, essential to overcoat them with passivating inorganic shells, such as ZnS, to enhance their PL efficiency and stability. Despite significant advances in synthesis and shell overcoat technology, it is still challenging to tune emission wavelengths and simultaneously yield bright samples with narrow line widths. Significant research effort is being made to refine the synthesis of InP QDs, understand their surface chemistry, and optimize their inorganic surface passivation. CuInS₂ QDs face many of these same challenges. Indeed, the origin of emission from CuInS₂ QDs is still being elucidated, making it even more challenging to achieve narrow PL from these QDs.¹⁷ The formation of robust core/shell structures (e.g., CuInS₂/ZnS) is complicated by the rich surface chemistry of CuInS₂ QDs coupled with the influence of precursors and reaction conditions.¹⁸ However, the successful formation of CuInS₂ QDs with stable ZnS shells enables more rapid testing of surface modifications, enhancing the potential for interfacing these QDs with biological systems.

Conclusion

Over time, the use of heavy metal-free QDs within biological applications has become more popular due to their notably reduced cytotoxicity. However, the challenge remains to develop a Cd-free QD with a narrow PL and concomitant high QY. While the core/shell model has been effective thus far when using heavy metal-free QDs in limited in vivo bioimaging work, these QDs still lag behind heavy metal-incorporating core/shell QDs in PL efficiency and QY. To combat this shortcoming, we expect to see further research on QD synthesis and use in vitro. We also expect to see further combinatorial approaches, including core/multi-shell structures and additional modifications post-QD synthesis. Looking forward, we anticipate approaches that utilize the QD’s facilitation of bound ligands at its surface to enhance QD targeting ability for bioimaging use in vivo. Such technologies must leave the QD’s imaging capabilities uncompromised while reliably directing the QD through the body to the desired imaging site. Finally, while the number of in vivo studies applying Cd-free QDs to bioimaging remains low compared to other approaches currently in the field of nanoscience, we believe more will emerge within the next several years.

Acknowledgments

The authors acknowledge the NRL Base Funding Program and the NRL Institute for Nanoscience for financial support. K.E.R. is a Ph.D. candidate in the Fischell Department of Bioengineering, University of Maryland College Park.