Nanoparticle Surface Functionalization: How to Improve Biocompatibility and Cellular Internalization

1Nanoparticles (NPs) are ultrafine particles with a size between 10 and 500 nm composed of different organic and/or inorganic materials. These particles have been widely studied because of their unique properties and find suitable application in a great number of biomedical fields like biomolecule detection, vaccines, regenerative medicine, and tissue engineering, gene and drug delivery, cancer therapy, high accuracy diagnosis, and theranostics.

2When NPs are used in biomedical applications two very important characteristics must be considered: toxicity and cellular uptake. In fact, the biocompatibility of NPs is one of the most critical characteristics of nano-platforms to be suitable for biomedical purposes, and the NPs capability to be internalized by target cells, compared to not-target cells, is a very important goal.

3The use of NPs has grown exponentially in the last 15 years especially for cancer treatment, because of their capability to perform high precision tasks, such as the delivery of drugs and imaging contrast agents (CAs) directly to tumor cells, by using a large number of molecular targets. In particular, the accumulation of nanocarriers in cancer cells can occur through two different mechanisms: passive and active targeting. In passive targeting, NPs accumulate in the proximity of the tumor site as result of the altered permeability of tumor blood vessels. This phenomenon, known as the enhanced permeability retention (EPR) effect, permits the passive accumulation of NPs to solid tumors and/or metastatic sites, simply through their particular physical properties including size, shape, and superficial charge. Active targeting exploits the biofunctionalization of the NPs surface by using ligands with a strong affinity and specificity for overexpressed receptors and molecules on the tumor cells or secreted proteins in the tumor microenvironment (TME). Active and passive targeting are phenomena often occur simultaneously, and one does not preclude the other (Figure 1).

4The use of NPs for active targeting of drugs and/or contrast agents is based on the interaction of properly modified NPs surface with a molecular target overexpressed by cells and tissues. The molecules used to modify the NPs surface include small proteins, peptides, antibodies, aptamers, and oligosaccharides. Furthermore, the biochemical modification of the NPs surface with these specific targeting ligands is often necessary to reduce toxicity (as widely applied for silver NPs) and to increase their stability in biological fluids. An example is represented by the use of human albumin, whose presence on the NPs surface reduces toxicity and achieves active targeting at the same time. The chemical modification of the NPs surface is a step-by-step process that requires chemical-physical and biological characterization for each step. Transmission electron microscopy (TEM), scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FTIR), dynamic light scattering (DLS), and ζ-potential analysis are some of the most useful and simple techniques used to study shape, size, chemical composition, and superficial charge of NPs. In contrast, the biological evaluation of NPs requires a more complex analysis with the assessment of numerous variables depending on the experimental model, conditions, and other specific parameters. In conclusion, it is crucial to consider the effects of the NPs surface modification on stability, uptake, and biocompatibility to achieve an efficient nanoplatform for biomedical applications.

5The focus of this review is to summarize the most used techniques to evaluate NPs toxicity and uptake and to describe the most useful strategies to avoid these issues through the modification of the NPs surface.

6NPs can be applied to cancer for two main purposes: diagnosis and therapy. The usage of nanoplatforms for diagnosis can be classified by two different approaches: biomolecule detection and imaging techniques. Biomolecule detection is based on the identification of tumor markers that are often present in a very small amount in biological samples. The possibility to modify NPs by specific ligands with high affinity for these markers can improve the identification and quantification of desired compounds.

7Medical imaging techniques have evolved very quickly over the past 20 years, overall in cancer disease, to identify very little tumor masses for an early diagnosis. This evolution has occurred through the combined use of classical imaging techniques (like MRI, PET and SPECT) with NPs and advancing toward the use of innovative imaging techniques like photoacoustic (PA), surface-enhanced Raman scattering (SERS), and near infrared light up-conversion (NIR-Up). All these techniques are very useful in combination with NPs-mediated delivery in order to accumulate CAs in the tissues of interest.

8For cancer therapy, there are currently numerous agents available depending on the disease stage, location, type of tissue, age, and conditions of the patient. The use of classical chemotherapeutic compounds, radiation, and surgical interventions are still useful and widely used for treating tumors, but the side effects are often a serious concern. Furthermore, the recurrence of cancer due to the incomplete removal of malignant cells is a major reason of poor prognosis. The use of NPs for an early and precise identification of cancer cells can help to avoid delayed diagnosis, and the specific delivery of a limited amounts of drug directly to cancer cells can effectively reduce chemotherapy-related side effects. At the same time, the usage of NPs can be useful to bypass the drug resistance of some tumors, like melanoma. In this scenario, the surface modification of NPs represents an important strategy to successful develop specific and biocompatible nano-platforms for precise and sensitive therapy and/or diagnosis.

9The surface modification of nanoparticles is a powerful methodology to fix or attenuate issues related to NPs toxicity and uptake, since both phenomena are closely related to the NPs surface composition.

10The surface functionalization of NPs involves a process that aims to improve and/or add properties useful for the use of NPs in medical applications. Different types of nanomaterials have characteristic chemical properties and functional groups exposed on their surface to be used in the first steps of functionalization. Generally, the first phase of the surface modification is based on the use of homo- or hetero-bifunctional cross linkers to the aim to add an organic functional group (R-NH2, R-COOH, etc.), useful to bind biological molecules. For silica NPs, the most used linkers are aminosilanes that introduce an amino group on the NPs surface for the next bio-conjugation. Noble metals, like gold, can be functionalized by using crosslinkers with -SH or -NH2 groups able to react with the metal and to produce a covalent bond. These bi-functional linkers, such as thio-carboxylic acids, have at the other end functional groups to use for binding ligands. Metal oxides can be easily modified by using a ligand exchange strategy based on the substitution of the original surfaces with functional groups such as diol, amine, carboxylic acid, and thiol useful for the next steps. The carbon-based nanomaterials contain a significant fraction of sp2 hybridized carbon atoms that can be exploited to generate functional groups. Through the oxidation it’s possible to generate -COOH, -OH, and -C = O on the NPs surface; through halogenation, it’s possible to obtained halogenated carbon that can be further modified, for example by reaction with the amine group; through cycloaddition it’s possible to insert different type of functional groups. Table 1 summarizes types of nanomaterials, their chemical groups and/or composition, and the suitable compounds or processes that can be used for surface modification by using crosslinkers.

11The modification of the NPs surface can be achieved using two different approaches: non-covalent and covalent conjugation. The non-covalent strategy is based on a large number of weak interactions (electrostatic, ionic, van der Walls and hydrophobic interactions, absorption, hydrogen bonds) and it is specially used with metallic and silica NPs. Non-covalent bonds have the advantage of being relatively simple and do not affect the structure of the used molecules and their interaction with biological targets. Conversely, non-covalent modifications can be easily influenced by different variables, such as pH and ionic strength.

12The covalent bond strategy can be obtained by using many alternative approaches, depending on the composition of the NPs. Moreover, this strategy allows modifications at several levels using sequential functionalization. This methodology can be exploited to achieve structures with multiple functions, such as diagnosis and therapy to implement the theranostic approach. Usually, the covalent bond of ligands to the NPs surface can be achieved using various linker molecules. An example is PEG, that can be synthesized with specific functional groups at the ends and used as homobifunctional or heterobifunctional linkers to perform a wide range of functionalization processes. showed how the production of specific heterobifunctional PEG molecules is still an active research field and how this molecule can be very useful to design efficient nano-platforms for medical applications. Thanks to its polymeric nature, PEG can be also used as a spacer for high molecular weight molecules in order to space them across the surface of the NPs and to reduce steric hindrance of bonded ligands, allowing bioconjugation at high density. Generally, non-covalent interactions are used to load nanoparticles with molecules that must be released in target cells, such as drugs or RNAi, while covalent bonds are employed to bind ligands useful to achieve targeting and/or to reduce he toxicity of NPs. Recently, the use of sensitive bonds, such as pH-sensitivity or heat-sensitivity, to develop nanoplatforms for a controlled drug-release has been widely explored. In particular, the tumor microenvironment is very acidic when compared to the normal microenvironment. pH-sensitive nano-platforms can been designed for controlled drug release specifically triggered by the acidity of the tumor environment. In a recent work, a cationic polymer PBAE pH-sensitive was used to cover liposome NPs loaded with doxorubicin. Furthermore, the NPs surface was modified with hyaluronic acid (HA) to perform active targeting via CD44. In vivo experiments confirmed the results obtained in vitro, showing that the DOX-loaded NPs inhibited the growth of tumor more efficiently compared to free drug and also reduced side-effects. Another example of pH-sensitive NPs is reported by. In this work, mesoporous silicon nanoparticles were functionalized with amine conjugated phenylboronic acid linked to dopamine with a pH-sensitive covalent bond. In a weakly acidic environment, as like the tumor environment, the nanoparticles released dopamine owing to the hydrolysis of boronic-ester bond between the two molecules. This intelligent release resulted in an inhibition of vascular endothelial cell migration and tubule formation.

13Several methods have been used to evaluate the internalization of nanoparticles in a specific cellular context that can be distinguished as label-free and label-based techniques. Among the label free methods, the most commonly used are transmission electron microscopy (TEM), scansion electron microscopy (SEM), and Raman microscopy. These procedures, unlike label-based ones, have the advantage of not requiring the use of fluorophores that could affect the NPs size or chemical properties. Furthermore, using labeled-NPs complicates the discrimination of internalized NPs or those attached to the cell membrane. Conversely, TEM and SEM imaging, despite offering high-resolution down to the cellular organelle scale, are relatively expensive and time-consuming. Furthermore, TEM and SEM, unlike Raman microscopy, are destructive imaging approaches.

14TEM analysis is widely used to study nanoparticle uptake and cellular localization. showed a better affinity of the epoxy-coated SiO2 nanoparticles for the cell membrane, compared to bare NPs, in A549 cells and human monocytes. Furthermore, TEM allowed to identify the nanoparticles in endosomes. used TEM analysis to evaluate the uptake of bare and APTES-coated iron oxide nanoparticles in PC3 (prostate cancer epithelial cell) and BPH1 (benign prostate hyperplastic epithelia cell) cell lines. Results showed that PC3 cells internalized coated particles with higher efficiency than BPH1 cells. used SEM analysis to correlate the size of gold nanoparticles with their cellular localization in HEp-2 cells. Results showed that the presence of nanoparticles in the cytosol and nucleus was simply based on their sizes, regardless of the incubation time.

15Among label-free techniques, Raman microscopy presents several advantages useful to study nanoparticles uptake. Compared to SEM and TEM analysis, Raman microscopy requires minimum sample preparation and allows in vitro and in vivo cellular imaging. This methodology was used by to evaluate the internalization kinetics and intracellular localization of diatomite-based nanoparticles in a lung epidermoid carcinoma cell line. The results indicated the presence of NPs up to 72 h, without damage to cell viability or morphology. Raman microscopy has also been used to study the uptake of metal-based nanoparticles. Using this technique, explored the cellular uptake of iron oxide nanoparticles in breast cancer cells. The NPs were totally internalized in cells, displaying a cytoplasmic localization with a direction toward the nucleus after 24 h of incubation.

16Label-based techniques include methods that exploit a fluorescent signal emitted by the nanoparticles. This signal can be attributed to an intrinsic property of the NPs or to the use of a fluorophore added to the nanoparticles. This modification can be obtained adding the fluorescent tag inside the NPs structure during synthesis or binding the tag to the NPs surface.

17The use of this approach to study NPs cellular uptake has several advantages like the ease of use, no need or simple specimen preparation, possibility to analyze live-cells and to perform time-lapse acquisitions. However, fluorescence methods do not allow a quantitative analysis, but only comparison between different experimental conditions (semi-quantitative analysis). This is due to the fact that the signal is not absolute because its intensity depends on the excitation source, the number of fluorophores per NP, the quantum yield of fluorophores or the NP itself, and the sensitivity of the detector.

18The most used label-based techniques to study NPs cellular uptake are confocal fluorescent microscopy (CFM) and flow cytometry (FC).

19Owing to very high resolution, contrast and penetration depth, CFM allows the detection of very low amounts of NPs and to localize them to cellular compartments. used CFM to study the uptake of various types of NPs [liposomes, mesoporous silica NPs, poly(lactide-co-glycolide) NPs, and nanohydrogels] in myoblasts with a high proliferative rate and in myotubes characterized by a low proliferative rate, to correlate proliferative rate with NPs uptake. The results showed that there was a lower uptake in myotubes compared to myoblasts and the inability to penetrate in nucleus by all analyzed NPs in both cell types. evaluated melanin-silica hybrid nanoparticles uptake in two pancreatic cancer cell lines where NPs appeared as punctate small vesicles, indicating an endocytic mechanism of internalization.

20The use of flow cytometry to study NPs uptake is widely used, despite the impossibility of discriminating NPs attached on cellular surface from the internalized ones., using NPs labeled with the fluorescent dye fluorescein isothiocyanate (FITC), evaluated TiO2 NPs uptake in NIH/3T3 and A549 cell lines. The authors observed that NIH/3T3 cells internalized the TiO2 NPs more efficiently than A549 cells, confirming the importance of the cell type in the uptake behavior. Flow cytometry can also be used to investigate the internalization pathway of the nanoparticles. investigated the uptake pathway of bare and IL13-modified nanoparticles using several specific endocytosis inhibitors and flow cytometry. Their results showed that bare NPs were internalized by macropinocytosis while IL13-modified NPs were internalized by a clathrin-mediated pathway, typical of a receptor-mediated uptake.

21To study NPs biocompatibility, there are several aspects, summarized in Figure 2 that can be evaluated: cell viability (Figure 2A), cytotoxicity (Figure 2B), proliferation (Figure 2C), apoptosis/necrosis (Figure 2D), cellular morphology alteration, oxidative stress (Figure 2E), inflammatory response, and hemotoxicity (Figure 2F). Using different methodologies exploiting commercial kits and/or protocols adapted to the type of analysis it is possible to assess these aspects. Some of the most useful markers/pathways are summarized in Table 2.

22The physiochemical properties of NPs, like shape, size, charge, material and surface chemical groups, influence their toxicity and uptake efficiency. Some of these, such as surface charge and chemical groups, can be easily modified by surface modification.

23The functionalization of the NPs surface to enhance biocompatibility can be exploited by using different molecules; among these PEG is one of the most used for in vitro and in vivo applications. showed how the addition of PEG on the surface of Fe3O4 and SiO2 NPs increased the biocompatibility when PEGylated NPs were used at high concentrations (200 μg/mL) with murine neural stem cells, unlike bare Fe3O4&SiO2 that caused a viability reduction of about 50% already at a dose of 20 μg/mL. This study is a clear example of the high convenient properties of PEG to increase NPs biocompatibility. Furthermore, the presence of PEG on the NPs surface also improves the hemocompatibility, as observed with chromium-doped zinc gallate and diatomite based NPs.

24To reduce the toxicity of NPs, dextran is a widely used to modify nanoparticles. Dextran is a complex branched polysaccharide usually exploited to modified iron-oxide NPs surface. showed how the addition of dextran to iron-oxide NPs increased biocompatibility in zebrafish larvae; in particular, the treatment with dextran-coated NPs did not determine any significant mortality or changes in the hatching rate of the larvae. The toxicity of dextran modified iron-oxide NPs was also investigated by on Jurkat cells. The authors observed low toxicity and small effects on membrane integrity up to 72 h of incubation. The use of dextran with iron-oxide nanoparticles to improve biocompatibility was also studied using primary cells. demonstrated that dextran-NPs had no significant effects on cell viability and apoptosis on human primary monocytes cells.

25A further oligosaccharide generally used to enhance NPs biocompatibility is chitosan, as reported by and. observed a decrease of toxicity in three different cell lines when chitosan-NPs were compared to iron-oxide NPs. used chitosan to modify silver NPs. Results showed that chitosan-coated silver nanoparticles had higher biocompatibility when compared with silver nanoparticles without surface modification in human fibroblast cells.

26A more complex NPs surface modification as way to enhance biocompatibility, stability and dispersity was been reported by. The study demonstrated that encapsulation of mesoporous silica NPs using a lipid bilayer is a useful way to improve biocompatibility and hemocompatibility.

27Besides the increase of biocompatibility, NPs surface modification is a very important tool for the modulation of the body’s immune response against the particles. NPs, after the injection in blood flow, interact with a lot of aspecific proteins like opsonin, complement proteins, immunoglobulins, fibronectin, and apolipoproteins (protein corona) and this interaction can modify NPs behavior and can trigger an immune response. The NPs surface modification can be exploited to promote their escape from the immune system and to increase their half-life in blood by reducing the clearance due to macrophages of the mononuclear phagocyte system (MPS). Conversely, in the next-generation vaccines, the NPs surface modification can trigger the immune response toward a specific antigen.

28Escape from the immune system can be obtained through the modification of NPs surface with different types of molecules that make NPs “invisible” to the immune system cells. For this purpose, the most used compounds are hydrophilic polymers that bind water molecules producing a shield on NPs surface. This layer of water reduces the interaction with opsonin and/or macrophages receptors. This strategy can increase NPs circulation time and can reduce the clearance.

29It is well known that the cellular uptake of NPs is influenced by the physicochemical properties of NPs, such as its composition, size, shape, surface charge, surface functionalization, and surface hydrophobicity/hydrophilicity. When nanoparticles interact with the constituents of the plasma membrane, they are mainly taken up by cells by endocytosis, which is commonly classified in phagocytosis (macrophages) and pinocytosis (all cellular types). The latter can be distinguished in clathrin-dependent endocytosis, caveolae-dependent endocytosis, macropinocytosis, and clathrin, and caveolae-independent endocytosis. The modification of the physicochemical properties of NPs surface can be exploited to enhance cellular uptake.

30PEG, usually tested to enhance biocompatibility, can also be explored to increase NPs uptake. The PEG monomer is about 0.35 nm in length with a molecular weight of about 2 kDa, and can be synthesized at different lengths. This molecule reduces aggregation and increases NPs stability in biological fluids. Furthermore, PEGylation of NPs determines a decrease of the interaction with not-specific proteins resulting in a “stealth” effect able to enhance PEG-NPs circulation time and to reduce phagocytosis. This increased stability of PEG-NPs is directly related to higher cellular uptake if compared with bare NPs that can aggregate in biological environment and/or to be phagocytized by immune system cells.

31As reported by, the use of PEG can effectively change the behavior of NPs in relation to its delivery in tissues. The addition of PEG of different sizes (2.5, 5, and 10 kDa) on poly(lactic-co-glycolic) acid (PLGA) on the NPs surface can enhance particle diffusion by up to 10-fold into the cervical mucus and the binding to mucin proteins, depending on the PEG density and size. Furthermore, the conjugation with PEG results in a shift in the charge of PLGA NPs, from −45 to + 8 mV, with the NPs size increase related to the efficiency of PEG coverage. showed that NPs properties can be affected by the length of PEG molecule and by the surface functionalization density (PEG molecules/nm2). In particular, when longer PEG molecules and a high density were used to functionalize gold NPs, a reduction in the non-specific protein adsorption was observed. However, this kind of modification resulted in a decrease in the uptake of NPs in all cell lines tested.

32Other molecules used to modify NPs surface involve changes in the superficial charge. Generally, these consist of amino ending molecules (R-NH2) that at physiological pH (about 7.4) are positively charged. demonstrated how the surface charge of silica nanoparticles (SiO2) affects the cellular uptake of NPs in the HaCaT cell line and in skins explants. Negatively charged SiO2 NPs showed lower uptake levels compared to positive 3-aminopropyl-trimethoxysiliane (APS) modified NPs. Conversely, reported that polystyrene-modified NPs with -NH2 groups on the surface enhanced toxicity, related to the high reactivity of the amino groups. The higher uptake and toxicity of positively charged NPs, compared to negatively charged NPs, has also been demonstrated by in an intestinal epithelial cell model.

33Furthermore, the superficial charge of NPs can be selectively modified to modulate the cellular uptake. Zwitterionic ligands, like carboxybetaines and sulfobetaines, show a variety of the positively and negatively charged groups allowing the modulation of charge densities to optimize solubility and to avoid the interactions of the protein corona, making NPs highly stable in biological fluids. This modification reduces non-targeted uptake and opsonization of modified-NPs, and increases their accumulation in target tissues (e.g. cancer cells). An example of the zwitterionic ligands useful to enhance NPs uptake is reported by. In this work, silica coated CdSe/CdS quantum dots were biofunctionalized with PEG and with sulfobetaines in order to evaluate the impact of these modifications on the NPs cellular uptake. The authors showed the differences in the cellular internalization and, in particular, they revealed that the uptake of the zwitterion-modified NPs happened more easily compared to PEG-modified NPs in HeLa cells.

34In gold-NPs were modified with pyranine (a negatively charged dye) that reduced the NPs uptake. This behavior can be reversed through the addition of a positively charged molecular cage that neutralizes the negative charge of gold NPs and allows cellular uptake. The key role of the surface charge of NPs in cellular uptake efficiency was also demonstrated by, where the surface of fluorophore-conjugated polystyrene nanoparticles (f-PLNPs) was modified with different types of functional groups (acetyl, zwitterionic, carboxyl) and molecules (guanidinium, polyethylene glycol, sulfonic acid). The nanoparticles were incubated with THP-1 cells (phagocytic) or A549 cells (non-phagocytic). Results showed a NPs superficial charge-dependent uptake by both cell lines; in particular, an increasing trend in internalization was observed in positively charged modified-NPs and this correlation was stronger in the THP-1 cells compared to A549 cells.

35Another strategy widely explored to increase NPs uptake is based on the use of cell penetrating peptides (CPPs). These molecules are composed of a specific aminoacidic sequence, usually polycationic or amphipathic structures, that enhance NPs uptake. used a CPP derived from protamine to modified lipid-based NPs for the efficient delivery of siRNA. This modification enhanced internalization of siRNA in B16F10 cells. exploited the penetrating capability of Tat peptide to functionalize PLGA NPs surface. They observed that while bare NPs did not enter HeLa cells, when Tat peptide was added (Tat-NPs), cellular uptake was detected.

36Due to the importance of enhancing the uptake of NPs, research into alternative molecules that bind on the NPs surface is very active. reported the use of phosphatidylcholine (PC) modified with different alkyl chain lengths (from C12 to C18), to increase lipid-PLGA hybrid NPs internalization. Even if PC had little effect on NPs stability and physicochemical properties, results showed an enhanced cellular uptake of hybrid nanoparticles in HepG2 or A549 cells. Furthermore, the PC-modified NPs uptake increased proportionately to the length of PC alkyl chain.

37In recent years, surface modification of NPs was mainly aimed at active targeting and cellular uptake by exploiting the specific interactions of NPs surface ligands with a wide range of receptors overexpressed in cancer cells. The molecules useful to perform NPs active uptake include several main categories, among these the most used are antibodies, small peptides, proteins, aptamers, carbohydrates, and small molecules (Figure 3). During the conjugation process, usually performed by covalent interactions, these molecules are linked to the NPs surface in order to preserve their ability to bind the target receptors.

38Monoclonal antibodies (mAbs) used to perform active uptake of NPs are extensively studied due to their high stability, specificity and binding ability (Figure 3A). reported a very efficient uptake of PLGA NPs modified with an anti-HER2 mAb. Uptake evaluation in HER2 positive (MDA-MB-453) and negative (MCF-7 and BT-20) cells demonstrated higher internalization levels in a HER2 positive cell line, compared to the negative ones. Another example related to the use of targeted uptake with the anti-HER2 antibody was reported by. In this work, boron-containing gold NPs were bioconjugated with an anti-HER2 antibody and their uptake was studied both in vitro and in vivo. Results showed an elevated selective uptake in N87 human gastric cancer cells of mAb-gold NPs when compared to bare NPs. In vivo experiments indicated accumulation of mAb-gold NPs in tumor sites and a reduction of non-specific uptake at others anatomical districts. Another receptor widely used to perform active uptake in cancer cells is the epidermal growth factor receptor (EGFR). As reported by, the functionalization of PLGA NPs surface by using a therapeutic mAb against EGFR (Cetuximab) enhanced the NPs uptake in an in vivo model. A further example of mAb modified NPs to improve cellular internalization was reported by, where PLGA nanoparticles were functionalized with a mAb against heparan sulfate proteoglycan 2 (HSPG2) protein, a surface receptor highly expressed in triple negative breast cancer (TNBC) that binds growth factors such as VEGF-A and FGF-2, acting as “co-receptor.” Even in this study the presence of mAb on the NPs surface increased specific cellular uptake in Luciferase-expressing MDA-MB-231-LM2 cells.

39Antibodies are very useful molecules to enhance specific NPs uptake, but they have high molecular weight (about 150 kDa), and this could be an issue in bioconjugation process, above all with smaller NPs (<10 nm). The use of the antibody antigen-binding fragments (Fabs) to perform NPs active uptake is a viable alternative, as reported by (Figure 3B). To target breast cancer cells, the PEG-PLGA NPs were loaded with paclitaxel and everolimus and the NPs were coated with anti-HER2 and anti-EGFR Fabs. Results showed higher uptake levels in HER2 and EGFR positive cell lines (SKBR3) compared to negative or low EGFR expressing cell lines (MCF-7 and MDA-MB- 436).

40A viable alternative to the use of antibodies and Fabs to modify NPs surface is represented by peptides that bind with high affinity to specific receptors (Figure 3C). Indeed, specific peptides can be obtained by the screening of phage libraries and by the isolation of binding sequences from proteins using three-dimensional (3D) structural analysis. A small peptide (GE11) with high affinity for EGFR was used by to modify nanomicelle containing evodiamine (GE11-Evo-NPs) in order to enhance drug uptake into cancer cells. Results obtained in vivo showed that GE11-Evo-NPs allow higher evodiamine concentration in tumor after intravenous administration, compared to the drug alone. Furthermore, the accumulation of GE11-Evo-NPs in tumor masses in vivo was higher than Evo-NPs. Another study where a peptide was used to enhance active NPs accumulation is reported by. In this study, hybrid chitosan/poly(N-isopropylacrylamide) NPs functionalized with K237 peptide (that binds KDR/Flk-1 receptor) were described. The modified NPs showed higher uptake in the MDA-MB-231 cell line overexpressing KDR/Flk-1 compared to K237 free NPs and to L929 cells that do not express the KDR/fLK-1 receptor. The use of peptides was also explored to perform a personalized therapy approach. showed that the surface functionalization of diatomite-based NPs with an idiotype-specific peptide (Id-peptide), which recognizes hypervariable region of immunoglobulin B-cell receptor, enhanced Id-NPs uptake in specific myeloma cells of threefold compared to nonspecific cells. Furthermore, same results were obtained if a random control peptide was used.

41The use of natural proteins that interact with cancer cell receptors has been widely investigated, as reported in several recent studies (Figure 3D). used transferrin (Tf) to functionalize the surface of doxorubicin-loaded PLGA NPs. The interaction between transferrin and Tf receptors (Tf-R), highly expressed in cancer cells, was exploited to enhance uptake and drug release. Results showed that Tf-PLGA@DOX NPs greatly reduced viability of human epithelial cervical cancer cells HeLa (Tf-R positive), when compared to immortalized HaCaT keratinocytes with low Tf-R expression. Another protein used to perform NPs active delivery is human serum albumin (HSA) that interacts with tumor-associated protein SPARC. modified hybrid melanin-silica-silver nanoparticles (MelaSil_Ag) surface using HSA to enhance cellular uptake in breast cancer cells. Results showed that MelaSil_Ag-HSA NPs were mostly internalized by SPARC positive cell line (HS578T) compared to SPARC negative cells (MCF10a).

42Furthermore, other molecules are currently being explored to modify NPs surface and to perform active uptake. Among these, aptamers represent a useful molecule since are cheap and easy to synthetize (Figure 3E). Aptamers are short nucleic acid sequences (dsDNA, ssDNA or RNA) with specific 3D structure that can bind a molecular target in cancer cells with high specificity and affinity. reported the conjugation of the aptamer S-MUC-1 (able to bind MUC1 protein) on paclitaxel-loaded protein nanoparticles. Incubation of aptamer-modified and bare NPs with MCF-7 cells (that overexpress MUC-1 protein) showed higher uptake levels of aptamer-functionalized NPs when compared to the bare NPs. Further examples of aptamers as ligand for active uptake are reported by and. modified the surface of mesoporous silica nanoparticles (MSNs) with Sgc8 aptamer able to bind protein tyrosine kinase-7 (PTK-7) on human acute T lymphocyte leukemia cells. Sgc8-MSNs NPs were also loaded with DOX, in order to enhance drug uptake in leukemia cells. The results obtained with two different cell lines, CCRF-CEM with high expression of PTK-7 and Ramos cells with no expression of PTK-7, showed a high uptake level of Sgc8-MSNs in CCRF-CEM cells compared to Ramos cells. Furthermore, no differences in NPs uptake either cell lines was observed using bare MSNs. used an aptamer designed to bind CD133 receptor on osteosarcoma cells in order to enhance the uptake of lipid NPs loaded with all-trans retinoic acid (ATRA). Results showed high internalization of aptamer-modified NPs in CD133 positive cells compared to the negative ones, confirming the CD133 mediated targeting.

43Exploitation of simple molecules to enhance active uptake also includes the use of carbohydrates (Figure 3F). Among these, one of the most studied is hyaluronic acid (HA). This molecule enhances the NPs uptake through its interaction with CD44 protein. recently demonstrated that HA can be easily used to biofunctionalized carbon dots loaded with doxorubicin to enhance drug uptake in CD44 overexpressing cells (4T1). Results obtained by competitive assay performed with free HA showed that the HA-modified NPs were internalized through the binding of HA-modified NPs with the CD44 receptor. Furthermore, in vivo experiments showed that the enhanced accumulation into tumor tissue was confirmed. The use of carbohydrates was also explored for targeted gene-delivery in macrophages by. PLGA-PEG NPs decorated with various carbohydrates (mannose, galactose and dextran) and loaded with eGFP (enhanced Green Fluorescent Protein) mRNA and GFP (Green Fluorescent Protein) plasmid DNA (used as reporters) was developed. The results showed that carbohydrates on the NPs surface, especially mannose and dextran, improved the active uptake of carbohydrate-modified-NPs in the Raw 264.7 murine macrophage cell line, that express specific receptors for mannose and dextran.

44In addition to the molecules already mentioned, there are other small molecules useful to perform NPs active uptake (Figure 3G). used folate (FA) to functionalize chitosan-lipid hybrid NPs in order to increase uptake through the interaction of FA with folate receptor (FR), highly expressed on cancer cells. Results showed higher uptake of FA-conjugated NPs in an ovarian cancer cell line (SK-OV-3) when compared to bare NPs. Similar results were obtained with the 3D spheroid cell model, with a 2.4-fold higher uptake of FA conjugated NPs compared to uptake by the control group. Other simple molecules used to perform NPs active uptake are anisamide (AA) and phenylboronic acid (PBA) able to bind sigma receptors and sialic acid (SA), respectively. These molecules were successfully investigated by and. Ramzy et al., functionalized the surface polymeric NPs loaded with thymoquinone (TQ) by using AA, in order to enhance uptake and drug accumulation in colon cancer cells. Researchers used three colon cancer cell lines, HT-29 (overexpressing sigma receptor), HCT-116, and Caco-2, to evaluate the toxicity of AA-TQ-NPs. Results showed higher toxicity of AA-TQ-NPs in HT-29 cells, compared to the other two cell lines. Furthermore, bare TQ-NPs showed lower toxicity in HT-29 cells compared to AA-functionalized NPs. These results are due to the active uptake of AA-modified NPs after interaction with the sigma receptor. The use of PBA was described by Qian et al. with soy protein-based NPs. The high affinity of PBA for SA, which is overexpressed in tumor cells, was exploited to enhance the uptake of PBA-modified NPs in cell lines with different SA expression. The results showed a high uptake level of PBA-NPs in HepG2 cells (SA positive) when compared to SH-SY5Y cells (SA negative).

45In the last years, the development of nanoparticles functionalization strategies is considerably grown, due to potential applications of NPs in nanomedicine. Surface modification is evolved from the use of simple molecules (PEG), to the aim to decrease toxicity, clearance and immune response, to more specific and complex ligands, in order to increase specificity and efficacy. The evolution of the NPs surface bioconjugation went hand in hand with the identification of specific cellular targets. The possibility to tune the NPs physico-chemical properties in combination with the knowledge of cancer biology makes possible the use of these nano-platforms in biomedical applications, both for therapy and diagnostic, providing a great contribution to the advance of nanomedicine.

46Uptake and biocompatibility are two of the most important features of a usable nanoplatform for medical applications, and overcoming issues related to these two aspects is the first goal in the development of NPs. For this reason, the study of toxicity and cellular uptake is the first step during the biological assessment of NPs.

47The NPs surface modification is a powerful instrument to enhance uptake and biocompatibility, as confirmed by the vast amount of scientific papers focused on this topic. These studies demonstrate that the conjugation of molecules on the NPs surface can effectively enhance biocompatibility both in vivo and in vitro, due to the modification of surface charge and to the inactivation of reactive chemical groups that can affect cellular membrane stability. Moreover, the addition of specific molecules can also enhance NPs passive and active uptake, reducing systemic toxicity in vivo and allowing high precision therapy and/or diagnosis. The binding of molecules on the NP surface can be obtained by covalent and non-covalent approaches. The former is widely used to bind proteins, antibodies, aptamers and peptides exploited to enhance uptake and to perform active targeting, while non-covalent interactions are generally used for loading of drugs and for all molecules that must be released in the cells.

48The protocols for the conjugation of molecules on NPs surface depend on the nanomaterial used and on the available functional groups, for this reason the use of linker molecules or the modification of NPs and/or ligands in order to obtain a stable conjugation could be necessary. Furthermore, some molecules used to functionalize NPs have high molecular weight due to the overall size of the proteins and the use of a spacer (e.g., PEG(n)) could be necessary to stave off the ligand from NP surface. The use of spacers also has an effect on conjugation density. In fact, in order to bind a sufficient amount of molecules to the NPs it is necessary to reduce steric hindrance. As reported in numerous studies, the conjugation density is a very important parameter that can affect the behavior of NPs. Moreover, it is important to note that the conjugation of some molecules, usually used to enhance uptake, can also increase the NPs biocompatibility.

49Due to the high variability of nanomaterials and ligands to modify the NPs surface, it is not possible to follow specific guidelines to functionalize NPs. It is necessary take in account that: (I) the addition of big size molecules could change the size of the NPs, thus influencing uptake; (II) some molecules used for active targeting could change their 3D structure during the functionalization protocol and to lose the ability to bind target molecules; (III) the conjugation with non-covalent bonds could result in an unstable surface modification due to the influence of environmental conditions like pH and ionic strength; (IV) the conjugation density and the orientation of ligands on NPs surface is a key parameter to enhance uptake; (V) the reduction of toxicity could be due to a reduction in NPs uptake; and (VI) positively charged NPs are usually better internalized by cells.

50In conclusion, through the handling of surface characteristics, the nanoparticles can be transformed in smart platforms, containing therapeutic and imaging agents as well as stealth property, delivering drugs to specific tissues and providing controlled release therapy. This targeted and sustained drug delivery decreases the drug related toxicity and the frequency of treatments. Nanoparticles have proven useful in the treatment of cancer, and many other diseases, also providing advancement in diagnostic and theranostic applications.

51GS and AL designed the review. All authors wrote, read and approved the manuscript.

52The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.