Ever since their discovery in 1981, quantum dots (QDs) have intrigued scientific minds worldwide with their unique physical and optical properties. Quantum dots are nano-sized semiconductor crystals with a broad emission range [1]. Their emission spectra are closely related to their diameter ^{[2][3][4][5][6][7]}[2,3,4,5,6,7]. This specific characteristic of quantum dots allows for easy manipulation of QDs’ size to achieve the desired fluorescence color during production. Quantum dots are also superior to many other fluorescence probes, such as organic dyes, since QDs’ fluorescence is known to be photobleaching-resistant ^{[8][9][10][11][12]}[8,9,10,11,12]. Due to their physical properties, the structure of quantum dots is relatively stable. In some recent studies, QDs have been demonstrated to remain intact under various pH levels [13], high UV exposure [12], and exposure to oxidative conditions [14]. A standard structure of QDs is composed of a core—often heavy metals—encapsulated by a protective shell ^[15][16][15,16]. QDs can also be conjugated with different types of ligands, which often dictate QDs’ interaction with their surrounding environment ^{[17][18][19][20][21]}[17,18,19,20,21]. The addition of a protective shell and surface ligands increases the solubility and strengthens the structure of QDs ^{[9][11][15][22][23][24]}[9,11,15,22,23,24]. The protective shell also greatly reduces the exposure of the heavy metal contents residing in the core of QDs ^[25][26][25,26], considered to be one of the main mechanisms of QDs’ toxicity. Another characteristic of quantum dots is that they can be synthesized using many methods. However, these methods are often categorized into either top-down or bottom-up approaches [27]. In the top-down approach, larger precursors are used to form smaller QDs as products through hydrothermal methods [28], electrochemical methods [29], laser ablation [30], etc. On the other hand, the bottom-up approach uses smaller precursors, such as organic molecules, to build larger QD structures through pyrolysis [31] and heat [32]. Although each approach has its own advantages and disadvantages [33], the variety of methods enables developers to choose the best-suited approach.

Due to the diverse types of available quantum dots, QDs are often categorized by their core composition. One of the most widely used cores for quantum dots is cadmium. Cadmium-based QDs, such as cadmium selenide (CdSe), cadmium telluride (CdTe), and cadmium sulfide (CdS), are of great interest to researchers due to their phenomenally high quantum yield compared to other types of QDs ^[34][35][36][34,35,36]. Thus, cadmium QDs are an excellent candidate for biomedical applications such as trackable drug delivery and bioimaging. However, the potential for cadmium-based QD usage for these biological purposes is currently hindered due to the potential toxicity of their cadmium core ^{[37][38][39][40][41]}[37,38,39,40,41]. Cadmium is well known to cause many major problems for the environment, as well as for human health ^{[42][43][44][45]}[42,43,44,45]. It has been reported that cadmium exposure can lead to serious diseases in adults [46] and cause detrimental developmental impairments in children [47]. The United States Environmental Protection Agency (EPA) has also classified cadmium as a probable carcinogenic agent for humans [48]. As such, the usage of cadmium-based QDs raises concerns for many people due to the possible leakage of cadmium ions. In response, countless studies have been dedicated to studying the impact of QDs on the cellular, molecular, tissue, and even organismal levels ^{[49][50][51][52][53][54][55][56]}[49,50,51,52,53,54,55,56]. In the past decade, many studies have reported the toxicity of cadmium-based QDs, including inducing ROS levels ^[51][57][51,57], triggering apoptosis [57], alternate gene expression profiles ^[58][59][60][58,59,60], damaging the structure and function of mitochondria [61], negative impact on the reproductive system ^[62][63][62,63], causing neurotoxicity [64], and many other undesirable side effects. Due to the countless negative impacts, cadmium QDs are often deemed to be unideal for in vivo biomedical applications. Thus, researchers have shifted their attention to developing other types of QDs that do not contain cadmium.

Indium-based QDs were among the earliest developed non-cadmium QDs. The recent literature suggests that indium-based QDs are less toxic compared to cadmium QDs. Several studies have found that compared to cadmium-based QDs, indium-based QDs have less impact on cell viability, induce less DNA damage, and are overall not toxic [65]. However, contradictory results have also been reported. Davenport et al., reported a similar toxic effect on cell viability between cadmium selenide/zinc sulfide quantum dots (CdSe/ZnS QDs) and indium phosphide/zinc sulfide quantum dots (InP/ZnS QDs) [66]. Cullen et al., found that InP/ZnS QDs reduced the endpoint optical density of the budding yeast Saccharomyces cerevisiae, while CdSe/ZnS QDs only caused a prolonged lag phase but did not greatly impact the final optical density [58]. Therefore, more studies need to be conducted before indium-based QDs could be used as a safe alternative. In addition to indium-based QDs, other types of QDs such as copper-based QDs [67], silver-based QDs [68], and many others have also been developed. However, these quantum dots have also been shown to have a negative impact on cells. For example, with ternary copper indium disulfide/zinc sulfide quantum dots (CulnS₂/ZnS QDs), a high dose induced a minor inflammatory response in the lymph nodes of mice [69]. Silver quantum dots have been reported to exert toxicity by interfering with plants’ photosynthesis processes [70]. Thus, collectively, these studies suggest that each type of QDs has its own toxicity mechanism that needs to be further investigated. Another alternative to cadmium quantum dots is doped quantum dots. Among these, manganese-doped zinc sulfide dots (Mn: ZnS d-dots) are considered to have the most potential due to their low toxicity [71]. Furthermore, Mn-doped QDs have been shown to possess both fluorescence and magnetic properties, making them an ideal candidate for multimodal imaging [72].

2. Ligands

In addition to using alternative core types, modifications to other QD components, such as the surface ligand, have also been explored to maximize QDs’ biocompatibility and efficacy for in vivo applications. Surface ligands are essential for the interaction between QDs and the surrounding environment [17]. Therefore, understanding the behavior of QDs with different conjugated ligands is the first step toward picking the right ligand type for potential clinical usage. In the pharmaceutical field, one of the key factors for an agent to be considered for biological applications is solubility ^[73][75]. Thus, numerous strategies have been developed to improve the solubility of chemical treatments ^[74][75][76][76,77,78] and drug delivery carriers ^[77][78][79][79,80,81]. For QDs, solubility issues can be avoided by coating the nanocrystal with hydrophilic surface ligands. For example, Ghani et al., replaced the original hydrophobic tri-octyl phosphine oxide (TOPO) ligands on CdSe/ZnS QDs with different types of water-soluble bisphosphonate (BIP)-based ligands, such as ethylene diphosphonate (EDP), imido diphosphonate (IDP), and methylene diphosphonate (MDP). The results showed that the exchanged ligand improved the water solubility and dispersal of CdSe/ZnS QDs [22]. Furthermore, EDP-, IDP-, and MDP-conjugated CdSe/ZnS QDs were shown to have lower toxicity effects compared to CdSe/ZnS QDs conjugated with TOPO ligands. Additionally, the study found that EDP and MDP were significantly taken up by IGROV-1 ovarian cancer cells compared to IDP-conjugated CdSe/ZnS QDs [22]. The difference in the cellular uptake of QDs with different target ligands showed that certain ligands are more suitable than others for biological applications. It is also worth noting that the ligand-exchange step could be skipped if QDs are synthesized in an aqueous phase [71], making it easier to conjugate the desired ligands on QDs. In another study, Al-Hajaj et al., found that equally sized CdSe(CdZnS) QDs with different conjugated ligands had distinct modes of entry and were taken up in different quantities. Their data revealed that negatively charged CdSe(CdZnS) QDs-CA (QDs conjugated with cysteamine ligands) entered HEK293 human kidney cells and HepG2 human liver cells at a much higher levels compared to positively charged CdSe(CdZnS) QDs, such as CdSe(CdZnS) QDs-CYS (QDs conjugated with cysteine), CdSe(CdZnS) QDs-MPA (mercaptopropionic acid), and CdSe(CdZnS) QDs-DHLA (dihydrolipoic acid) [21]. Ligand types also had an impact on the rate of CdSe(CdZnS) QDs elimination. CdSe(CdZnS) QDs-CA clearance from HepG2 and HEK293 cells took place in the first 3 h post-treatment, while around 80% of CdSe(CdZnS) QDs-CYS was still retained in HepG2 cells after 6 h post-treatment [21]. In addition to affecting the rate of QDs’ cellular uptake, surface ligand types also influence the interaction of QDs with biological components. In 2023, Yu et al., investigated the interaction and impact of two different ligands on CdSe/ZnS QDs—glutathione (GSH-QDs) and dihydrolipoic acid (DHLA QDs)—on the alpha chymotrypsin (ChT) enzyme. Their data revealed that GSH-QDs weakly inhibited ChT’s catalytic activity, while DHLA-QDs greatly inhibited ChT activity. Both DHLA-QDs and GSH-QDs were able to bind to ChT at a 1-to-8 ratio. However, DHLA-QDs have a greater affinity for ChT compared to GSH-QDs, suggesting that the binding affinity of DHLA-QDs with ChT is one of the key factors for the inhibition of ChT activity. The different ligands also resulted in different binding mechanisms between the QDs and ChT, as DHLA-QDs bound to ChT via hydrophobic interactions, while GSH-QDs bound to ChT through hydrogen bonding and van der Waals forces [20]. Recently, some studies have reported that various types of QDs could also interact with common proteins, such as bovine serum albumin (BSA) ^[80][81][82][82,83,84]. It has been shown that in the presence of CdSe QDs, BSA could bind and form corona protein complexes ^[83][85]. The formation of corona protein complexes in the serum could lead to several issues, such as protein structure alteration, protein aggregation, and protein denaturation. In humans, a similar protein called human serum albumin, which is abundant in human serum, has also been reported to form bind to CdSe/ZnS QDs ^[84][86]. Thus, it is possible that the same phenomenon could also take place when QDs are used for in vivo applications in humans. The interaction between QDs with different ligands and other biological materials is consistent with prior reports, where ligands played an essential role in the binding of proteins to nanoclusters ^[85][87]. Thus, future research should investigate ligand-dependent interactions between QDs and proteins to develop a modified ligand with minimal unwanted QDs–protein interactions. Collectively, the studies above highlight the importance of choosing the right surface ligands for QDs. Depending on the aim of the application, ligand choice can enhance the safety and efficacy of QDs by altering the interaction between QDs and biological components.

3. QDs as a Labeling Agent

Due to their unique characteristics, QDs are vastly useful as a fluorescence label. Previously, Q-tracker 565 and Q-tracker 655 from the Quantum Dots Corporation (Hayward, CA, USA) were shown to effectively label several hematological cell lines (KG-1, HL-60, and SUDHL-16), as well as cells derived from the bone marrow and umbilical cords, by residing in the intracellular space of cells. The labeling of these cells was shown to last from one to two weeks post-incubation. Furthermore, these QDs were found to remain in cells through four cell division cycles, with decreasing QD fluorescence after each cycle. In cells such as HL-60 cells and umbilical-cord-derived CD34+ cells, QDs were also shown to be retained through cell differentiation ^[86][88]. These findings showed that QDs could reside intracellularly and label different hematological cell lines for an extended period of time, providing evidence that QDs could be a useful tool for hematological cell imaging. However, the same study also revealed that intracellular labeling by QDs was seen in all tested cell lines, hinting that the labeling by these QDs is not selective. As such, the same group of researchers attempted to target QDs to specific cells by conjugating QDs with streptavidin (QDs-SA). To target cells that specifically express CD33 on the cell surface, QDs-SA were incubated with biotinylated anti-CD33 antibodies prior to cell treatment. The results showed that QDs-SA incubated with biotinylated anti-CD33 selectively bind to cells that express CD33 (HL-60) ^[86][88]. These results indicate that when QDs are not conjugated with a selective ligand, they can be randomly internalized by multiple hematological cell lines, thus providing evidence that the presence of targeting ligands is essential for the specific binding. Apart from whole-cell labeling, QDs have also been used to label specific organelles’ structures. Traditionally, organic dyes are commonly used as fluoroprobes to visualize different structures of cells. However, problems such as short lifetime and weak signal intensity limit their efficiency in cell imaging. As QDs are well known to have photobleaching-resistant fluorescence, they could also be used as superior labeling agents. In one study, streptavidin-conjugated CdSe/ZnS QDs were used to label the actin cytoskeleton of SK-BR-3 cancer cells. The authors found that QDs-streptavidin were able to clearly label the biotinylated F-actin structure of cancer cells. Furthermore, CdSe/ZnS QDs-streptavidin were also able to label other cell structures, such as the nucleus of SK-BR-3 cancer cells pre-incubated with nuclear antigens and biotinylated anti-human IgG ^[87][89]. In the most recent 2023 study, neutravidin-conjugated CuInS₂/ZnS (CIS/ZnS) QDs and neutravidin-conjugated CdSe/ZnS QDs were used as F-actin labeling agents for super-resolution imaging. Although improvements in labeling density are still needed, both QD types effectively labeled F-actin and significantly improved the resolution compared to conventional fluorescence imaging ^[88][90]. In addition to fixed structural imaging, QDs could also be used as probes to study real-time cellular processes. Hatakeyama et al., used Qdot (QD655) in combination with HaloTag technology to study the dynamics of the cytosolic myosin motor protein. In this study, QDs were conjugated with a HaloTag ligand and electroporated into cells, where HaloTag ligand-QDs found the protein of interest (myosin) that was fused with the HaloTag protein (Figure 1). In this way, QDs were able to indirectly bind to myosin and act as a probe to study its intracellular movements and interactions. Using this technology, the authors were able to observe myosin’s movement along the actin filament ^[89][91]. In a more recent study by Zhang et al., the combination of QDs and HaloTag proteins was used in a similar manner to study the movement of proteins selectively expressed on the surface of mammalian cells through mammalian display technology. In this case, the HaloTag protein was chosen as the displayed surface protein. QDs conjugated with HaloTag ligands (HTL-QDs) were added to act as a probe to track the movement of the displayed HaloTag protein throughout the experiment. Around 30 min after the temperature shift, the displayed HaloTag protein tagged with HTL-QDs moved from the membrane to the cytoplasm. This indicates that the membrane-displayed protein is not always on the cell membrane but is capable of reentering the cells upon temperature changes. The internalized HLT-QDs were transported by membrane-bound vehicles and located near the nucleus but never entered it. The HLT-QDs signals were eventually either detected in the lysosome or recycled back to the membrane surface ^[90][92]. Thus, quantum dots are an excellent fluorescence probe for the labeling and imaging of cellular processes. In addition to using QDs to study the intracellular dynamics of cells, some researchers have also used QDs to study the communications between cells. Extracellular vesicles (EVs) are secreted by cells such as cancer cells and carry information that helps the communication between cells to promote cancer’s proliferation and invasion ^[91][92][93][93,94,95]. Due to their associations with cancer cells, researchers have recently become interested in using extracellular vesicles like exosomes as cancer biomarkers ^[94][96]. To achieve this, it is essential to understand the behavior and interaction of EVs with other components. Therefore, fluorescence probes such as QDs have been used to label and track the dynamics of EVs in recent years. In 2020, Zhang et al., visualized EVs by conjugating them with fluorescence QDs (QDs-EV). The conjugation of EVs with QDs yielded high-quality fixed and live imaging. Furthermore, the interaction between QDs-EVs and microglial BV-2 cells was also detected around 1 h post-co-incubation ^[95][97]. In another study, gold carbon dots (GCDs) were used as a nanoprobe to study exosomes secreted by a human breast cancer cell line (SKBR3). This was achieved by using anti-HER2 as an adaptor for GQDs to target HER2 receptors on SKBR3-derived exosomes. By tracking the QDs-labeled exosomes, Jiang et al., were able to observe the uptake of exosomes by HeLa cells. These QDs-labeled exosomes were eventually detected in the lysosome of HeLa cells ^[96][98]. QDs showed vast potential in the labeling of EVs and their behavior in vitro. However, it is unclear whether the same efficacy could be achieved in vivo. In the human body, there are multiple cell types that can secret EVs, giving rise to a diverse range of EVs in the tumor microenvironment ^[97][99]. As such, targeting a specific type for in vivo labeling of EVs is exceptionally challenging. However, if successful, much insight could be gained through the observation of heterogeneous EV interactions through multicolor imaging. QDs have also been used to study the behavior and movement of cancer cells in the body. For instance, Voura et al., used QDs to label cancer cells and studied cancer’s distribution to other organs. In this study, CdSe/ZnS QD-labeled B16F10 cancer cells were injected through the tail vein, and their distribution was tracked over time. CdSe/ZnS QD-labeled cells were found mostly in the lungs. However, a small population of cells was also able to arrive at other organs through the lungs’ capillary networks, with low tumor formation frequency. The same study used CdSe/ZnS QDs with different emission ranges to perform multicolor imaging. Two populations of B16F10 cells were labeled with two different types of CdSe/ZnS QDs (510 green and 570 red) and tracked throughout the body. They found that these two populations tended to colonize and form tumors at the same location in the lungs ^[98][100]. This result provides a hint that tumor formation in these areas is not random ^[98][100]. The use of QDs to study the migration pathways of cancer cells has indeed provided insights into their metastasis behavior. However, a few limitations exist. First, it is unclear whether QD labeling alters any structures, processes, or behaviors of cancer cells. To overcome this issue, the trafficking path of QDs in cancer cells, as well as the specific interactions with cellular components, should be investigated in detail. Second, cancer cells were labeled in vitro before being injected back into mice via the tail vein. Thus, the conditions were different from the extravasation and metastasis of cancer cells in vivo. As such, injected cancer cells may behave differently from cancer cells existing in native biological conditions. In addition, previous studies have shown that nanoparticles, including quantum dots, could be retained in the liver and kidneys for an extended amount of time. Therefore, the interactions of QDs with vital organs, as well as the QDs’ clearance mechanisms, need to be investigated.