Biological nanoscale fluorescent probes: From structure and performance to bioimaging

In recent years, nanomaterials have attracted lots of attention from researchers due to their unique properties. Nanometer fluorescent materials, such as organic dyes, semiconductor quantum dots (QDs), metal nanoclusters (MNCs), carbon dots (CDs), etc., are widely used in biological imaging due to their high sensitivity, short response time, and excellent accuracy. Nanometer fluorescent probes can not only perform in vitro imaging of organisms but also achieve in vivo imaging. This provides medical staff with great convenience in cancer treatment. Combined with contemporary medical methods, faster and more effective treatment of cancer is achievable. This article explains the response mechanism of threenanometer fluorescent probes: the principle of induced electron transfer (PET), the principle of fluorescence resonance energy transfer (FRET), and the principle of intramolecular charge transfer (ICT), showing the semiconductor QDs, precious MNCs, and CDs. The excellent performance of the three kinds of nano fluorescent materials in biological imaging is highlighted, and the application of these three kinds of nano fluorescent probes in targeted biological imaging is also introduced. Nanometer fluorescent materials will show their significance in the field of biomedicine.


Introduction
For centuries, cancer has been a global problem that has plagued human health. It has become the leading cause of morbidity and mortality in developed countries. The development and research of low-toxicity, high-targeting, and high-selectivity drugs have become a hotspot for many scientists [1,2]. Nanoparticles have a unique size and shape-dependent physical and chemical properties [3,4]. They can be combined with targeting ligands to enter the body, target specific tumor cells, and realize cell imaging in vivo [5]. The development of nanotechnology can mostly help surgeons delineate the edges of tumors, identify residual tumor cells and micrometastasis, and determine whether the cancer is entirely removed [6].
Nanomaterials usually refer to materials with one dimension or more within the range of 0.1-100 nm in the three-dimensional scale of tiny particles [7,8]. Fluorescent nanomaterials include organic fluorescent dyes [9], fluorescent proteins [10], QDs, MNCs, CDs, etc. Organic fluorescent dyes are the most common nano fluorescent probe materials. Among them, rhodamine fluorescent dyes normally have a high quantum yield. They are widely used in biological applications, such as biomolecular labeling and detection, cancer diagnosis, and so on [11]. However, organic dyes and fluorescent proteins have poor fluorescence stability, weak photobleaching resistance, and short fluorescence lifetime, so they are greatly restricted in actual use [12][13][14]. Emerging nano fluorescent materials such as semiconductor QDs, precious MNCs, and CDs are not only used in biological imaging and targeting tumor cell applications due to their excellent physical and chemical properties, but also in drug and gene delivery, biosensors, medical treatment, photocatalysis, electrocatalysis, etc. have been widely used [1,[15][16][17].
This article mainly introduces three materials: semiconductor QDs, precious MNCs, and CDs as nanometer fluorescent probes. First, we introduce the response mechanism of nano fluorescence probes. Then we briefly explain the basic concepts and related properties of the three materials and introduced each material in detail. Finally, we show the specific application of three kinds of nano fluorescent probes in bioimaging and tumor cell targeting.

Design principle of nanometer fluorescent probe
Fluorescent probes are usually composed of three parts, namely recognition genes, linkers, and signal reporter genes. The recognition gene can react with the analyte, and the signal reporter gene (also called fluorophore) can convert the chemical environment change caused by the reaction of the recognition gene with the analyte into an output model that can be easily observed. The linker, as its name implies, connects the recognition gene and the signal reporter gene, becoming the hub of molecular recognition. The analyte reacts with the recognition gene. This reaction is transmitted to the signal reporter gene through the linker. The reaction is reflected by the fluorescent signal change of the signal reporter gene. This path is the response mechanism of the fluorescent probe [18]. There are generally three design mechanisms for the fluorescence recognition process: the principle of photoinduced electron transfer (PET), the principle of fluorescence resonance energy transfer (FRET), and the principle of intramolecular charge transfer (ICT). These design principles are shown in Figure 1. Below we introduce these design mechanisms in detail.

Photoinduced electron transfer (PET)
In PET, there is a linker between the receptor and the fluorophore. The linker connects the fluorophore and the receptor, while also separating the two during fluorophore excitation. Electrons are transferred from the highest occupied molecular orbital (HOMO) of the acceptor (or donor) to the semi-occupied molecular orbital (SOMO) of the fluorophore (or acceptor), resulting in fluorescence quenching. However, if the non-bonded electron pair of the donor is combined with the analyte, the redox potential of the donor increases, and the HOMO energy is lower than the fluorophore, which will cause the fluorescence to be switched on, and that limits the PET process [19,20].

Fluorescence resonance energy transfer (FRET)
FRET refers to the process in which the excitation energy of a donor transfers to the acceptor when the emission spectrum of the donor overlaps with the excitation spectrum of the acceptor, causing the acceptor to emit fluorescence. This is a non-radiation intermolecular dipole effect. The key factors that determine the fluorescence resonance energy transfer are mainly two points: (1) the emission spectrum of the donor and the excitation spectrum of the acceptor effectively overlap; (2) the distance between the donor and the acceptor needs to be maintained between 1-10 nm. The third one is the relative orientation of donor emission dipole moment and acceptor absorption dipole moment [21].

Intramolecular charge transfer (ICT)
In ICT system, the fluorophore and the acceptor are directly connected to a single entity. This single entity can provide electrons as an electron donor or accept electrons as an electron acceptor, so it has two functions. The addition of analytes will cause the intensity of the dipole on one side of a single entity to change, thereby affecting the overall intensity change and spectral shift. When the analyte interacts with the acceptor end, the electron-withdrawing property of the acceptor increases, and the emission spectrum shifts red; when the analyte interacts with the donor segment, the electron-donating properties of the donor decrease, and the emission spectrum shifts blue. Compared with the PET mechanism, there is no spectral shift during fluorescence quenching in the PET process, and a significant spectral change occurs in the ICT mechanism [19,20].

Semiconductor quantum dots (QDs)
Semiconductor QDs are novel fluorescent probes for biological and medical applications, usually composed of nanoscale II-VI and III-V elements. QDs are a kind of nanoparticles with quantum domain limiting effect, whose size is usually between 1-10 nm and nearly spherical [22][23][24].
Due to the development of quantum confinement, semiconductor QDs have unique optical properties. Compared with organic dyes, QDs have many advantages in fluorescence properties [25]. Firstly, QDs have a broad and continuous excitation spectrum and a narrow and symmetrical fluorescence emission spectrum ( Figure 2) [26]. The size of QDs determines the wavelength of fluorescence emission. By changing the size and chemical composition of QDs, the fluorescence emission spectrum of QDs can be adjusted from near-ultraviolet to near-infrared, with the wavelength range of 400-2000 nm [27]. Second, compared with organic dyes, QDs have a more massive stokes shift, which means that the difference between the peak of the excitation wavelength of QDs and the height of the emission wavelength is massive. In this way, the overlap of the excitation spectrum with the emission spectrum and its influence on the fluorescence detection in low signal detection can be avoided [28,29]. Third, QDs have high fluorescence intensity and good stability. Unlike conventional organic fluorescent dyes, which are easily bleached by light, QDs are 100 to 1000 times more resistant to light bleaching than organic dyes and fluorescent proteins [25,26]. In addition, the fluorescence yield of QDs is high, and the fluorescence life generated by QDs is long, and their excited state life lasts 20-50 ns. This enables the QDs to obtain fluorescence signals without background interference during use [22,30]. QDs are used in bioimaging, drug delivery, biosensor, cell tracking, and other fields due to their excellent optical properties [31].
In 1985, Brus first discovered QDs in colloids: ZnS and CdS particles [32]. Then in 1998, Alivisatos and Nie made the first successful use of semiconductor QDs as fluorescent probes in biology [33,34]. Since the synthesized QDs contain heavy metal elements (such as cadmium), they cannot be directly used in biological applications, so they need to be treated with water solubility [33]. Nie et al. used ZnS to encapsulate CdSe and then added a thioglycolic acid solution ( Figure 3). On one hand, the sulfhydryl group in thioglycolic acid can be combined with Zn atoms. On the other hand, the carboxylic acid group of thioglycolic acid makes QDs water-soluble. Studies have shown that the optical properties of the obtained QDs have not been changed [34]. At present, there are three main types of methods for synthesizing QDs, namely physical method, chemical method, and biological method. Material synthesis methods include laser physical vapor deposition and laser irradiation of large particles. However, QDs prepared by this method have low yield and poor stability. Therefore, many researchers use chemical processes to synthesize QDs [35]. Chemical synthesis method is mainly divided into an organic synthesis method and an aqueous synthesis method. QDs obtained by this method have good monodispersity and optical properties [30].

Metal nanoclusters (MNCs)
Fluorescent MNCs are likely to replace organic dyes or QDs [17]. Noble MNCs are typically smaller than 2 nm and consist of several or dozens of atoms [36]. Similar to semiconductor QDs, emission wavelength of noble MNCs can be adjusted from UV-Visible light to near-infrared light, and a single light source can excite the emission wavelength of multiple colors. Emission energy of MNCs can be regulated by determining the number of constituent atoms and ligands. Similarly, noble MNCs have high fluorescence intensity and broad stokes shift [17,37]. However, unlike semiconductor QDs, noble MNCs exhibit low toxicity and ultrafine size [38]. Large nanoparticles can quickly cause kidney barrier effect and cause serious side effects in the body. Noble MNCs can be effectively removed in the body due to their small size [37]. Noble metal nanoparticles have a strong surface plasmon resonance effect due to the electron excitation of the continuous conductive band. In contrast, noble MNCs, due to their small size close to the Fermi wavelength of electrons, will lose their surface plasmon resonance effect, showing discrete electronic structure and molecular properties [17,39,40]. It is important to note that colloid and cluster are often used interactively by researchers. Generally speaking, metal nanoparticles larger than 10 nm are called colloid, while metal nanoparticles smaller than 2 nm are called MNCs [41]. Noble MNCs also have some properties related to size, such as HOMO-LUMO transition, oxidizing reduction, magnetism, and handshape (as is shown in Figure 4) [37,40,41]. Besides, noble MNCs also have an essential property, which can be used as a new type of nano-catalysts in the fine chemical industry, medicine, food additives, and other fields [42]. The ultra-fine size and low toxicity of noble MNCs make their applications in biological imaging and fluorescence labeling very attractive [43].
In the 1960s, Malatesta firstly reported the synthesis of Au NCs using phosphine as a ligand. However, the structure of the Au NCs was still unknown [44]. After continuous research, Malatesta deduced the crystal structure of the Au NCs [45]. The precious MNCs were prepared by the template method. They had good biocompatibility and were quickly modified to adapt to specific applications. Therefore, the template method has become the primary method to prepare precious MNCs [46]. At present, proteins, DNA, polymers, dendrimers, and small sulfhydryl molecules are used to prepare noble MNCs.

Carbon dots (CDs)
In 2004, Xu et al. inadvertently isolated a kind of fluorescent nanoparticles when preparing carbon nanotubes by arc discharge method [47]. Later, in 2006, the Sun research group used the laser ablation method to prepare the fluorescent nanoparticles and named them CDs [48]. CDs are carbon nanoparticles with a particle size of less than 10 nm [49]. CDs are also referred to as carbon quantum dots (CQDS) in some literature. CDs have excellent photoluminescence properties. They show high fluorescence quantum yield, anti-swelling, and anti-bleaching photoluminescence decay, and long concentration range. They could be adjusted from deep ultraviolet to near infrared [50][51][52][53]. Besides, compared with QDs, CDs have good biocompatibility and low toxicity, which can almost be ignored.
Moreover, at a high concentration of CDs, the damage of CDs to organisms is shallow [53]. There are functional groups such as the hydroxyl group, carboxyl group, and carbonyl group on the surface of CDs. These functional groups cause CDs have good water solubility and provide the possibility of functionalization of CDs, which improves the effectiveness of CDs in practical application [51,54]. Due to the high price and large size of QDs, which affect the observation in vivo, QDs are subject to many limitations in practical application [27]. However, CDs with small size, low production costs, and low prices can be produced on a large scale, which is considered as a superior substitute for semiconductor QDs [54].
Synthesis methods of CDs are generally divided into two types: top-down and bottom-up. CDs obtained by the top-down approach tend to have high crystallinity and complete structure [51,55]. However, CDs with amorphous carbon core, large surface functional groups, and doping sites may be obtained by the bottom-up method ( Figure 5) [51,55]. Therefore, according to the different carbon cores, CDs are generally divided into graphene quantum dots (GQDs), carbon nanodots (CNDs), and polymer dots (PDs). GQDs are made up of one or more layers of graphene with edges connected by chemical groups and with a degree of crystallinity. CNDs are generally spherical carbon nanoparticles with no lattice. PDs are usually polymerized, assembled, or cross-linked by a linear nonconjugated polymer [56,57]. The photoluminescence mechanism of CDs has long been controversial among researchers. Three theories have been discussed the most, namely quantum confinement theory, molecular state theory, and surface state theory [52,58,59]. The difference of the synthesis method, prepared CDs structure, and performance will be different, so it is important to understand the preparation method of CDs [58].

QDs in biological targeting and biological imaging
Traditional fluorescent materials (such as organic dyes, fluorescent proteins, etc.) are prone to photobleaching, and the signal intensity is low. Multiple light sources are required to excite different fluorophores and other defects, which are subject to many restrictions in actual use [14,60,61]. QDs have vigorous fluorescence intensity and stability and have massive stokes shifts, photobleaching resistance, and controllable optical properties. Therefore, in biomedical applications such as bioimaging and targeting labels, QDs are very attractive to researchers [62]. In general, QDs can enter cells four ways: (a) passive absorption through endocytosis; (b) binding or modification with biopolymers, and then internalization through interaction with cell membranes and endocytosis; (c) direct physical manipulation of the cells (such as microinjection); (d) use of several uptake modes at the same time in combination [63].
In the past ten years, research on the biological imaging of QDs has significantly been developed [30]. Generally speaking, directly prepared QDs have poor biocompatibility, have specific toxicity, and are prone to non-specific adsorption and aggregation during use in the biological environment. Therefore, they need to be modified in biomedicine (usually using natural coupling [64] or encapsulation [65]). Then they can possess good water solubility, biocompatibility, and low cytotoxicity during use [14,66,67]. Gao et al. used amphiphilic triblock copolymers to improve QDs and used them for in vivo imaging in mice and targeted imaging of human prostate cancer cells. Passive targeted imaging is achieved through the penetration and retention of QDs at the tumor site. At the same time, QDs can also be combined with antibodies, and the targeted effect of antibodies and antigens can be used to achieve actively targeted imaging. However, the speed and efficiency of active targeting are significantly higher than the rate and efficiency of passive targeting (Figures 6) [28].
QDs are also used in multi-mode imaging [62,68], multicolor imaging [60,69], and multiphoton imaging [70]. Among them, multicolor imaging can identify individual malignant tumors in a complex biological tissue environment. This application provides detailed molecular information and morphological features in clinical medical diagnosis [60]. Multi-mode imaging combines QDs imaging technology with other imaging technologies (such as nuclear magnetic resonance imaging, computed tomography, etc.) [30]. Shibu et al. used nuclear magnetic resonance imaging technology and near-infrared fluorescence imaging technology to track the fate of modified QDs in mice, and found that kidneys can effectively excrete QDs within 48 h [68]. QDs have more excellent resolution under two-photon excitation and also have higher temperature sensitivity. Highquality thermal gradient images can be recorded in the fluid, and the internal temperature changes of individual cells can also be measured [70].

Noble MNCs in biotargeting and bioimaging
Although QDs have been successfully applied to cancer targeting and cell imaging in living animals, the mechanism of biodegradation in animals remains unclear due to the toxicity of QDs because of their heavy metal components [71,72]. In addition, the size of QDs is relatively large. Although this allows QDs to have extremely high resolution, it also reduces imaging efficiency and sensitivity, thus limiting their application in clinical imaging [73]. Compared with QDs, noble MNCs not only have the advantages of QDs in biological imaging but also have the characteristics of small size, low toxicity, and good biocompatibility. Therefore, they become more attractive fluorescent probes in biological imaging [72,74,75]. Wu et al. were the first group to use precious MNCs in near-infrared fluorescence imaging. They found that noble MNCs avoided uptake of the reticuloendothelial system due to their small size.
Thereby, they passively targeted Hela tumors in tumorbearing mice through the retention effect (EPR) [73]. Among the precious MNCs, the properties of Au NCs have been studied the most extensively [46,76]. However, due to the small size of Au clusters, the yield of Au NCs, synthesized by traditional methods, is low. Fluorescence intensity is not enough to achieve biological imaging. Therefore, Sun et al. used iron oxidase to assemble two Au NCs. The fluorescence intensity of the cluster is enhanced, and the fluorescence emission is adjustable. This kind of fluorescent probe can be used for whole-body imaging, especially kidneys [77]. Researchers not only use the fluorescence intensity of precious MNCs for biological imaging [78] but also use their fluorescence lifetime for imaging. Shang et al. developed the Fluorescence Lifetime Imaging (FLIM) technology. Through life-gating, cell imaging can be achieved without autofluorescence. FLIM images can not only show the uptake of MNCs by cells but also provide information on the environmental changes of NCs [79].
Due to the small size of precious MNCs, a suitable template preparation (such as glutathione) is chosen to make them quickly cleared by the kidneys and are therefore discharged out of the body after biological imaging [80,81]. While targeting tumor cells, precious MNCs can also provide sufficient information on cancer cells by detecting other substances in the organism (such as a high concentration of reactive oxygen species [82], PH value [83], etc.). Like fluorescent probe imaging with QDs, noble MNCs probe in vivo imaging can also be used with other imaging methods (such as computed tomography, magnetic resonance imaging technology) to achieve multi-mode imaging and provide more tumor information [84], as is shown in Figure 7.

CDs in biotargeting and bioimaging
CDs are used as an excellent fluorescent bio-imaging agent due to their broad absorption spectrum, narrow emission spectrum, adjustable photoluminescence, good chemical stability, and light stability, and photobleaching resistance [85,86]. Compared with QDs and noble MNCs, CDs have the advantages of low toxicity, good biocompatibility, low cost, facile synthesis, easy surface modification, and large-scale preparation [86][87][88][89][90]. The physical and chemical properties of CDs can be controlled by the reactants, reaction products, and reaction by-products [86]. Passivating CDs by using different organic substances and placing the prepared CDs in an oxidized state can improve the fluorescence quantum yield of the CDs, thereby obtaining higherquality biological imaging [91].
There are various functional groups on the surface of the CDs. These functional groups not only cause CDs to have good biocompatibility but also provide the possibility of surface modification of the CDs [92]. Hua et al. combined the surface functional groups of CDs with protoporphyrin IX (PpIX) to obtain multifunctional fluorescent CQDs. This kind of CDs can not only perform high-quality nucleolus imaging of fixed cells but it can cause that cells are imaged with high-quality nucleoli ( Figure 8) [88]. Due to poor biocompatibility and high toxicity, quantum dots are rarely used for drugs and gene delivery. CDs can realize the dual application of non-viral gene vectors and biological imaging probes at the same time, so that genetic material can be transferred to specific cells of patients to achieve gene therapy [93,94]. Zhou et al. modified the surface of CDs to make it have a positive surface charge, so that the prepared CD can be used as a nanoprobe for biological imaging, and as an efficient nonviral gene carrier to achieve drug and gene transfer. They also found that CDs have high transfection efficiency and cannot enter the nucleus during the transfection process. They speculate that the high transfection efficiency of CDs comes from the strong aggregation of macromolecular pDNA, which prevents its enzymatic hydrolysis during transportation. On the other hand, small size of the prepared CDs and their positive surface charge comtribute to high transfection rate [94].
Some researchers are interested in the development of carbon points emitting red fluorescence and their application in near-infrared biological imaging [95][96][97]. Li et al. attached electron acceptor groups (rich in sulfoxide/carbonyl groups) to the outer layer and edges of the CDs. These groups lead to increased surface oxidation and discontinuous energy levels, thereby enhancing the infrared absorption band and near-infrared fluorescence emission of CDs. They then successfully applied the prepared CDs to the imaging of the stomach of live mice [95].

Conclusions and outlook
In summary, nano fluorescent probes have great potential in biomedicine (Table 1). They can quickly target tumor cells and image biological cells. This provides excellent opportunities for medical personnel in cancer treatment. While targeting tumor cells, nanometer fluorescent probes can be combined with drugs and genes to accurately transport therapeutic drugs or expressed genes to tumor cells, thereby achieving the effect of cancer treatment. Through cell imaging in organisms, it is possible to understand relevant information of tumor cells better and help medical personnel to treat cancer better. Semiconductor QDs, precious MNCs, CDs, and other nano fluorescent materials all have adjustable emission spectra, good fluorescence, chemical stability, and vigorous fluorescence intensity will become the leading research direction of nanometer fluorescent probes.
Compared with the three materials, QDs and MNCs were developed earlier as nanoscale fluorescent probes, and the synthesized QDs have certain toxicity due to the presence of heavy metal elements. Therefore, in the process of using quantum dots in vivo, it is necessary to modify them, and improve the biocompatibility of quantum dots to reduce toxicity by means of biological coupling or ligand exchange. Compared to the other two materials, the size of the QDs is larger, which reduces the efficiency and sensitivity of imaging, which makes it more restricted in actual use. The size of MNCs is small, and the presence of the protective agent provides MNCs with good biocompatibility and low cytotoxicity. It is very noteworthy that the fluorescence properties of precious metal nanoclusters and precious metal nanoparticles are slightly different, which is easy to be confused in some literature. Noble metal nanoparticles have a surface plasmon resonance effect due to their large size, so their fluorescence is weak. However, the size of the noble metal nanoclusters is close to the Fermi wavelength of electrons, and the surface plasmon resonance effect disappears, showing strong fluorescence properties. CDs were discovered relatively late and have been extensively studied in the past decade. They have been used in various fields such as biomedicine, photocatalysis, electrocatalysis, and biosensors. Compared with QDs, CDs are smaller in size, have good biocompatibility, and are less toxic; compared with MNCs, CDs have a wider source of raw materials and are easily obtained from ordinary biomass materials. In addition, CDs have low cost, can be mass-produced, simple synthesis, and easy surface modification. CDs are likely to become excellent green fluorescent materials. Due to the late discovery of CDs, there are many synthesis methods, so the photoluminescence mechanism of CDs has not been fully understood. Currently, quantum confinement theory, molecular state theory and surface state theory are more common. Researchers continue to explore the photoluminescence mechanism of CDs. In addition, for CDs, there is a need to develop green synthesis methods. To achieve this goal, one can start from two aspects: raw materials and the synthesis process of CDs. Raw materials of CDs can be selected from renewable natural fuels, such as organisms, biological wastes and protein products to replace common reaction precursors such as graphene and carbon nanotubes. Synthesis methods of CDs are mainly divided into two types: top-down method and bottom-up method. Relatively speaking, top-down method requires more stringent conditions, and since raw materials are more expensive, this method is not suitable for large-scale production. Therefore, bottom-up method has become the mainstream method for CDs synthesis. It can be considered to meet the requirement of synthetic CDs under the condition of reducing energy use. As a relatively young nano fluorescent material, CDs still have a lot of potential for researchers to develop.