Targeting Colon Cancer Cells with Enzyme-Triggered Casein-Gated Release of Cargo from Mesoporous Silica-Based Nanoparticles

Abstract

Colorectal cancer (CRC) is one of the most widely diagnosed cancers worldwide. Despite notable improvements in therapeutic strategies available to CRC patients, late stages of CRC have a higher incidence rate of drug resistance, which is associated with a higher mortality rate. The development of therapeutic strategies that use nanoparticles as a drug delivery system has become one of the most promising potential approaches for cancer therapy. Previous studies have shown that a natural plant alkaloid, veratridine (VTD), suppresses colon cancer cell migration and invasion, two essential factors in tumor metastasis, through activation of the gene that encodes the tumor-suppressor protein UBXN2A. The goal of this study is to develop a nanoassembly to selectively deliver VTD to cancer cells and release it on demand while leaving normal cells intact. We packaged the targeted therapy anticancer molecule VTD inside mesoporous silica nanoparticles (MSNs) impermeable to the blood-brain barrier (BBB) and with selective affinity to CRC cells and sealed the VTD-loaded nanoparticles with an enzymatically cleavable protein. The particles will deliver and release VTD only at the targeted colorectal tumor sites. Since the enzyme MMP-7 protease is dominantly secreted by CRC cells, the release triggered by the enzymes will increase VTD concentration at tumor cells, enhancing the efficiency of the new therapy. We have proven the selective affinity of two types of VTD-carrying particles to CRC cells and enzyme- or acid-triggered VTD release. Negatively surface-charged MSNs showed significant affinity toward positively charged cancer cells but not negatively charged normal fibroblast colon cells, making VTD-MSNs a promising anticancer drug with minimal side effects.

Graphical Abstract

Colorectal cancer (CRC) is one of the most commonly diagnosed malignancies and the second leading cause of cancer death in the United States in both men and women, resulting in approximately 50,000 deaths annually.¹ Metastasis, a characteristic feature of CRC tumors, is dominantly responsible for the high mortality rate in CRC patients.² A 5-year survival rate of 91% for localized tumors in the intestine can be attributed to the success of surgical resection, but this rate drops to 12% for patients diagnosed with a metastatic form of the disease because of a lack of effective therapies. For four decades, the treatment of metastatic CRC underwent minimal development, and patient outcomes remain dismal. Thus, there is an urgent need for new therapies for patients with metastatic CRC.

Research and clinical studies conducted during the past 10 years have increased our understanding of the involved pathways and their downstream target proteins during the formation and progression of CRC tumors. Findings have demonstrated that several genetic and epigenetic changes contribute to the activation of tumorigenic pathways in CRC. Subsequently, these overactivated pathways determine patient prognosis and survival. While chemotherapy is still an important part of the treatment strategy for CRC patients, recent findings have opened a new platform for more effective therapies in CRC patients using a targeted therapy strategy. These drugs function differently from standard chemotherapeutic agents by targeting specific signaling pathways activated in CRC tumors in a stage-dependent manner.³ Targeted therapies are able to inhibit tumor growth and suppress cancer cell migration and invasion while normal cells dominantly remain intact. Several targeted therapies have already been approved by the FDA for CRC patients that are capable of treating advanced colorectal cancer.³ In the past decade, the development of nanoparticle technology has added a new angle for targeted therapy in cancer patients.⁴ A nanoparticle encapsulated with drugs can be designed to possess particular affinity toward cancer cells based on the unique characteristics of those cells. The so-called smart nanoparticles can release an effective dosage of small molecules next to tumor tissues while the surrounding normal organ tissues are exposed to a lower concentration of drugs and remain functional. We have already shown that a natural plant alkaloid, veratridine (VTD), can suppress tumor growth and induce cell death in colon cancer cells.⁵ However, it has been reported that the accumulation of veratridine in an animal model can induce neurotoxicity since VTD also binds to and activates sodium receptors.^6,7 We hypothesized that encapsulation of VTD in nanoparticles with a high affinity to cancer cells is an effective method for elevating the anticancer effectiveness of VTD while significantly lowering the permeabilization of VTD through the blood-brain barrier (BBB).

Human metastatic colon cancer cells notably overexpress matrix metalloproteinase-7 (MMP-7), a protease required to mediate cell invasion and tumor formation.^8-14 The activated form of MMP-7 is dominantly present in tumor masses but absent in normal tissues.¹² MMP-7 is able to digest a large set of proteins located in the extracellular matrix, including casein. MMP-7 has two domains; the first, a pro-domain, is cleaved upon activation of an enzyme.

The second domain of the zinc-binding site is catalytically active. MMP-7 has an approximately 6-fold greater expression in tumor masses versus normal cells.¹⁵ High expression levels of MMP-7 are associated with poor prognosis in CRC patients,¹⁶ which makes the protease an excellent target for enzyme-triggered gated drug delivery by nanoparticles.¹⁷ Enhanced release of MMP-7 by CRC cells enables the MMP-7-triggered gated-drug nanoparticles to release their drug cargo next to the tumor mass, elevating drug effectiveness. The MMP-7-triggered gated-drug nanoparticles can therefore be particularly effective against the metastatic form of CRC tumors with high concentrations of MMP-7 in their microenvironment¹⁸ while generating fewer side effects and lower drug resistance.¹⁹ Finally, the inability of mesoporous silica nanoparticles (MSNs) to penetrate the BBB makes them a promising carrier to reduce the neural toxicity of a potential anticancer drug.²⁰

Here we report the synthesis of a nanoassembly for targeting colon cancer cells by the gated release of an anticancer drug. An enzyme overproduced by the cancer cells cleaves a gatekeeping component of the nanoassembly and releases the drug (Scheme 1). The nanoassembly was successfully tested for the trypsin-triggered release of both a model drug, malachite green (MG), and an anticancer molecule, VTD.⁵ The drug is sealed inside the core of mesoporous silica with a gatekeeping MMP-7 substrate casein¹² conjugated with the core. In addition to selective drug release, we have shown that MSNs selectively associate with colon cancer cells versus normal colon fibroblasts, which will ensure selective drug delivery as well. Incubation of colon cancer cells and normal colon fibroblasts revealed a significant association of MSNs with colon cancer cells along with partial internalization of MSNs, while normal colon fibroblasts showed no affinity to MSNs and zero internalization.

Scheme 1. Schematic Illustration of the Synthesis of a Mesoporous Silica-Based Nanoassembly for Targeting Colon Cancer Cells by Gated Release of an Anti-Cancer Drug, Veratridine (VTD), Which Is Triggered by an Enzyme Overproduced by Colon Cancer Cells (Matrix Metalloproteinase-7; MMP-7) That Cleaves a Gatekeeping Component of the Nanoassembly^a.

^aThe gatekeeping component, MMP-7 substrate, will cleave in the presence of colon tumor cells where MMP-7 is notably overexpressed and release the anti-cancer drug at the targeted colon cancer cells. Additionally, the hydroxyapatite particle (HAP) component of the hybrid mesoporous silica-hydroxyapatite nanoassemblies (MSNs/HAPs) will dissolve upon encountering the acidic environment surrounding cancer cells, releasing the anti-cancer drug subsequent to the dissolution of the gatekeeping component. To design a gated drug delivery system, the mesoporous silica nanoparticle (MSN) surface is modified with amino groups followed by carboxylic groups to enable their covalent bonding with the protein’s amino groups, as shown in Scheme 1A. The gatekeeping protein component is electrostatically attached to the MSNs/HAPs, as shown in Scheme 1B. Particles are denoted by the blue color; drug load is denoted by the orange color.

RESULTS AND DISCUSSION

Synthesis and Characterization of MSN and MSN/HAP-Based Nanoassemblies for Gated Drug Release.

The MSNs were synthesized using a sol–gel reaction. In order to design a gated drug delivery system, the MSN surface was modified with amino groups followed by carboxylic groups to enable their covalent bonding with the protein’s amino groups (Scheme 1A). The hybrid mesoporous silica nanoparticles-hydroxyapatite particles (MSNs/HAPs) were synthesized to take advantage of their biodegradable properties and pH responsiveness due to the basicity of the hydroxyapatite (HAP) particle component (Scheme 1B). The synthesized drug delivery materials were characterized by scanning electron microscopy (SEM), transmission electron microscopy (TEM), dynamic light scattering (DLS), powder X-ray diffraction (PXRD), and Fourier transform infrared spectroscopy (FTIR) to determine the particles’ size, shape, structure, and functional groups.

SEm, TEM, EDS, and XPS Analysis.

SEM and TEM images were obtained to observe the particle morphology and mesostructure of synthesized MSNs and MSNs/HAPs. As shown in the SEM image in Figure 1, MSN particles are uniformly spherical (Figure 1A), while the hybrid MSN/HAP is made of both spherical and rod-shaped nanoparticles (Figure 1B). The TEM image of the MSNs confirms uniform spherical morphology and the beehive pore structure arrangement, representing the well-ordered internal mesoporous structure (Figure 1C). SEM images of MSNs after surface modification show that the particles maintained their spherical shape after the surface modifications by amination and carboxylation (Figure S1,A,B). As illustrated in Figure S2,A-C, the average particle diameters of MSNs, amino-functionalized mesoporous silica nanoparticles (MSNs-NH₂), and carboxyl-modified mesoporous silica nanoparticles (MSNs-COOH) are 95, 110, and 112 nm, respectively. The hybrid MSNs/HAPs exhibit lower dispersibility in water than the exceptionally hydrophilic pristine MSNs.

Figure 1.

Morphological characterization of mesoporous silica nanoparticles (MSNs) and mesoporous silica nanoparticle-hydroxyapatite particle (MSN/HAP) samples: scanning electron microscopy (SEM) images of (A) MSNs, SEM operated at 4 kV, and (B) MSNs/HAPs, SEM operated at 2 kV; transmission electron microscopy (TEM) images of MSNs, showing their (C) ordered porous structure and (D) element mapping; and (E) X-ray photoelectron spectroscopy (XPS) analysis of MSNs/HAPs.

To examine the structure of the hybrid MSNs/HAPs comprehensively, element mapping analysis of the particle surface and composition distribution of silicon, calcium, phosphorus, and oxygen of the particles were performed with an Oxford Instruments X-MaxN 50 energy dispersive spectroscopy (EDS) probe, which is attached to the SEM and X-ray photoelectron spectroscopy (XPS) measurements. According to our observations, silicon (Si), calcium (Ca), and phosphorus (P) elements are evenly dispersed in the hybrid MSNs/HAPs nanostructures (Figure 1D,E).

PXRD and FTIR Analysis.

The PXRD patterns of the MSNs and hybrid MSNs/HAPs are presented in Figure 2A-B. Low-angle PXRD spectra illustrate that all types of synthesized silica nanoparticles show a strong diffraction peak in the range of 2.3° to 2.35° (2θ), which is characteristic of the (100) diffraction plane of the unit cell (Figure 2A). Also, there are two more low intensity diffraction peaks in the regions of 4.0° to 5.0° (2θ) due to the (110) and (200) planes.²¹ The identified diffraction peaks illustrate the formation of the well-arranged MSNs with the long-range order hexagonal structures. The intensity of the PXRD spectrum peaks for functionalized particles is slightly lower than for MSNs, but their pattern remains unchanged. The region between 2Θ values of 20–50 has some characteristic diffraction peaks in the wide-angle PXRD spectra of MSNs/HAPs compared with MSNs (Figure 2B). The MSNs/HAPs show characteristic diffraction peaks at 2Θ values of 25.9, 28.1, 29.01, 32.01, 32.96, 34.1, 40, 46.4, and 53, which are attributed to the (002), (102), (210), (211), (112), (300), (310), (222), and (213) diffraction planes, respectively. The identified diffraction peaks confirm the incorporation of hexagonal hydroxyapatite (JCPDS 09-0432) with the MSNs, which is consistent with the work of Hao et al.²²

Figure 2.

(A) Low-angle powder X-ray diffraction (PXRD) patterns of mesoporous silica nanoparticles (MSNs) and mesoporous silica nanoparticles-hydroxyapatite particles (MSNs/HAPs). (a) MSNs, (b) amino-functionalized mesoporous silica nanoparticles (MSNs-NH₂), (c) carboxyl-modified mesoporous silica nanoparticles (MSNs-COOH), and (d) MSNs/HAPs. (B) Wide-angle PXRD patterns of (a) MSNs and (b) MSNs/HAPs. (C) Fourier transform infrared spectroscopy (FTIR) spectra of (a) MSNs, (b) MSNs-NH₂, and (c) MSNs-COOH, and (d) MSNs/HAPs. (D) Dzeta potential measurements for MSNs, MSNs-NH₂, MSNs-COOH, MSNs carrying fluorescein isothiocyanate (MSNs-FITC), and MSNs carrying rhodamine-B isothiocyanate (MSNs-RBITC) at pH 7.4.

FTIR spectra for all MSNs, MSNs-NH₂, MSNs-COOH, and MSNs/HAPs were obtained to identify the functional groups/surface-incorporated layers after surface modifications (Figure 2C). All spectra exhibited the bands characteristic of mesoporous silica.²² Thus, the strong peak at 1070 cm⁻¹ is attributed to asymmetric Si–O–Si stretching. The vibrational band at 799 cm⁻¹ represents symmetric Si–O stretching. The hydroxyl stretching vibration (O–H) can be identified as the broader band at 3400 cm⁻¹. The band at 450 cm⁻¹ is attributed to the Si–O–Si bending, and the band at 965 cm⁻¹ represents the stretching vibrations of Si–OH groups. Amine functionalization can be detected by new bands Figure 2C,b. Asymmetric bending of NH₂ can be identified by the characteristic band at 1560 cm⁻¹.²³ This confirms the efficient attachment of NH₂ groups to the surface, while lowering the surface concentration of free silanol groups reduces the band intensity at 1070 cm⁻¹. The carboxyl group attached to the silica surface can be proven by the sharp peak at 1656 cm⁻¹, which is attributed to the C═O vibration of the COOH group Figure 2C,c. Thus, FTIR spectroscopy observations have confirmed successful MSN surface functionalization in all steps. As shown in Figure 2C,d, the MSN/HAP FTIR spectrum is similar to the MSN sample. There are notable vibrational bands at 569 and 607 cm⁻¹, representing the Si– O–Ca asymmetric bending mode. The characteristics of the asymmetric Si–O–Si stretching vibration peak at about 1070 cm⁻¹ overlap with the asymmetric stretching mode of vibration of P–O bands of PO₄ tetrahedra.²⁴

DLS and dzeta Potential Measurements.

DLS analysis showed that the hydrodynamic diameter of MSNs, MSNs-NH₂, and MSNs-COOH was approximately 191.6 nm (PDI-0.287), 268.3 nm (PDI-0.387), and 215.5 nm (PDI-0.358), respectively (Figure S2d-f). The higher hydrodynamic diameter of MSNs-NH₂ compared with MSNs and MSNs-COOH suggests better solvation of the protonated sterically available amino groups. As expected, all hydrodynamic diameters exceeded the TEM diameters, as they include the solvation layer around the particles. As shown in Figure 2D, the dzeta potentials for MSNs, MSNs-NH₂, and MSNs-COOH are −26.33 mV, 6.41 mV, and −53.54 mV, respectively. The negative dzeta potential of MSNs (−26.33 mV) is due to the deprotonation of silanol groups on the surface. The positive potential of aminated MSNs proves the introduction of protonated amino groups to the particles’ surface. The carboxy-functionalized particles exhibit the lowest potential due to deprotonated carboxyl groups.

Fluorescence-Labeled MSNs Show a Drastically Higher Affinity to and Internalization Rate in Cancer Cells versus Normal Fibroblast Colon Cells.

In accordance with the literature, there is a significant difference in electrical properties between normal and cancerous cells.²⁵ It has been shown that cancer cell transformation leads to elevated insertion of phospholipids in their plasma membrane at low pH, resulting in a total positive charge in cancer cells.²⁶ In contrast, fibroblast cells have a negative charge.²⁷ To examine whether colon cancer cells and colon normal fibroblast have different affinity to negatively charged MSNs, cancer and fibroblast colon cells from the large intestine²⁸ were treated with MSNs carrying fluorescein isothiocyanate (MSNs-FITC; see Supporting Information). The results shown in Figure 3A,B indicate that HCT-116 colon cancer cells are associated with an enriched portion of MSNs bound to the plasma membrane. Additionally, z-stack images show the internalization of MSNs into cancer cells. However, normal colon fibroblasts treated with MSNs showed no association with or internalization of MSNs (Figure 3C,D). It has been shown that nanoparticles associated with plasma membranes enter into an endocytosis process,²⁹ leading to membrane wrapping of the nanoparticles followed by cellular uptake.³⁰ A set of flow-cytometry analysis revealed a significant internalization of MSNs carrying rhodamine-B isothiocyanate (MSNs-RBITC) into HCT-116 after 48 h of incubation (Figure 3E-H). Selective interaction of MSNs with colon cancer follow-up with internalization into cells highlights that MSNs can turn into an effective drug carrier for VTD in colorectal cancer cells while normal cells remain mostly intact.

Figure 3.

Colon cancer cells show a higher affinity to mesoporous silica nanoparticles (MSNs) than normal colon fibroblasts. MSNs carrying fluorescein isothiocyanate (MSNs-FITC) were prepared in pure dimethyl sulfoxide (DMSO; 5 mg/mL). HCT-116 colon cancer cells and CCD-33Co normal colon fibroblast cells were incubated with MSNs-FITC (60 ng/mL) for 90 min. Unattached MSNs-FITC were removed by PBS. Fixed cells stained with antibeta-tubulin (red) and diamidinos-2-phylindole (DAPI; blue) were subjected to confocal microscopy studies. (A,B) Panel A shows an enriched pool of MSNs-FITC associated with cancer cells (thin arrow). The z-stack images revealed that a portion of MSNs-FITC were internalized into cells (thick arrows) during incubation. Panel B is untreated control cells. (C,D) CCD-33Co fibroblast colon cells received the same treatment and were stained with antibeta-tubulin and DAPI. Panel C indicates that MSNs-FITS particles show no affinity to these normal colon fibroblast cells. Panel D is the CCD-33Co cells with no treatment. These results suggest that the plasma membranes of cancer cells, which possess a positive surface charge, enable the electrostatic interactions with negatively charged MSNs-FITS (14 mV, Figure 2D). Panels E—H show HCT-116 cells that were treated with MSNs carrying rhodamine-B isothiocyanate (MSNs-RBITC), as described in Panel A. Cells treated with MSNs-RBITC (19 mV) were incubated for 4, 24, and 48 h in a 37 °C incubator. Cells were washed with ice-cold PBS, followed by flow-cytometry. Results show significant cellular uptake of MSNs-RBITC after 48 h mediated by endocytosis pathways (n = 3, P < 0.001).

Protein Coupling and Quantification.

Casein and bovine serum albumin (BSA) were used as sealing agents to hold the drug inside the MSN channels. A bicinchoninic acid (BCA) assay was performed to determine the amount of casein or BSA coupled to the MSN channels. While BSA was used as the gatekeeping protein for proof of concept, casein (an MMP-7 substrate) was used as a low-price model of the MMP-7 substrates expressed by cancer cells. The protein conjugation efficiency (ratio of conjugated protein amount [mg] to MSNs and added protein amount for the conjugation [mg] with different MSNs) is presented in Table S1. There was a notable increase in the efficiency of casein conjugation compared with BSA conjugation to the particles. Covalent coupling of casein to MSNs-COOH was most efficient (31.9%). Electrostatic attachment of casein to the MSNs/HAPs occurred with a similar efficiency due to the coordination of carboxylate-groups with the calcium cations of the hydroxyapatite component. As expected, MSNs lacking the hydroxyapatite component exhibited much lower coupling efficiency (9.99%), which is accounted for by mere electrostatic attraction. On the basis of the better coupling efficiency of casein over BSA, we selected casein as the gatekeeping element for the nanoassemblies modeling triggered drug release.

Enzyme-Triggered In Vitro MG Release.

First, we identified the minimum concentration of casein able to provide substantial retention of MG inside MSNs-COOH by performing a series of experiments at 0%, 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, and 0.8% w/v (w, the mass of casein in g; v, the volume of solvent in mL) of casein (Figure S5). Without any casein, nearly 85% of loaded MG was released from MG/MSNs-COOH within 24 h, which demonstrated no significant MG encapsulation inside the pore channels without a pore-blocking capping agent. As the concentration of casein increased, more and more MG remained inside the nanoparticles, reaching 76% retention at 0.6% w/v (Figure 4A). A higher concentration of casein led to only a slight increase in MG retention. Therefore, 0.6% was determined to be the optimum concentration of casein to efficiently hold the MG inside the MG/CAS-MSNs-COOH.

Figure 4.

In vitro controlled release of a drug using mesoporous silica nanoparticles (MSNs) and hybrid mesoporous silica nanoparticles-hydroxyapatite particles (MSNs/HAPs). (A) Effect of varying w/v% of casein as a carboxyl-modified mesoporous silica nanoparticle (MSN-COOH) capping layer on the entrapping percentage of malachite green (MG). (B) Effect of varying w/v% of trypsin as enzymatic media to release entrapped MG by cleaving the casein layer that capped the MSN-COOH pore channels. (C) Cumulative release (%) profiles of MG from MG/CAS-MSNs-COOH under enzymatic and acidic conditions; red and black curves represent MG release profiles in the presence of the 0.10 w/v% trypsin mixture and the absence of any enzymes from casein-capped MSNs-COOH, respectively, at pH 7.4, RT. Blue curve represents MG release profile in the absence of any enzymes from casein-capped MSNs-COOH at pH 5.0, RT. (D) Cumulative release (%) profiles of MG from MG/CAS-MSNs/HAPs under enzymatic and acidic conditions; blue and black curves represent MG release profiles in the presence of the 0.25 w/v% trypsin mixture and the absence of any enzymes from casein-capped MSNs/HAPs, respectively, at pH 7.4, RT. Red curve represents MG release profile in the absence of any enzymes from casein-capped MSNs/HAPs at pH 5.0, RT. (E) Cumulative release (%) profiles of VTD from VTD/CAS-MSNs-COOH under enzymatic conditions; blue and red curves represent VTD release profiles in the presence of the 0.25 w/v% trypsin mixture and the absence of any enzymes from casein-capped MSNs-COOH, respectively, at pH 7.4, RT. Black curve represents VTD release from MSNs-COOH (without any casein) at pH 7.4, RT. (F) Cumulative release (%) profiles of VTD from VTD/CAS-MSNs/HAP under enzymatic and acidic conditions; blue and red curves represent VTD release profiles in the presence of the 0.25 w/v% trypsin mixture and the absence of any enzymes from casein-capped MSNs/HAPs, respectively, at pH 7.4, RT. Pink curve represents VTD release profile in the absence of any enzymes from casein-capped MSNs/HAPs at pH 5.0, RT. Black curve represents VTD release from MSNs/HAPs (without any casein) at pH 7.4, RT. Each data point represents the mean of the experiments performed in triplicate along with their corresponding standard deviation.

Similarly, we found that the same concentration of casein was optimal for the MSNs/HAPs, which provided ~66% retention of MG for 24 h inside the MG/CAS-MSNs/HAPs. With the optimized concentration of casein (0.6 w/v% casein) in hand, trypsin-triggered drug release profiles of MG/CAS-MSNs-COOH and MG/CAS-MSNs/HAPs were investigated.

Next, we determined the minimum concentration of trypsin that triggers substantial MG release from the MG/CAS-MSNs-COOH as well as its threshold triggering concentration by conducting enzyme-responsive drug-releasing experiments at 0%, 0.005%, 0.0075%, 0.01%, 0.05%, 0.1%, 0.2%, and 0.25% w/v% concentrations of trypsin (Figure S6). As expected, a faster release was observed at higher concentrations of trypsin. Since there was a similar MG accumulated release (%) for 24 h with no trypsin and with 0.005 w/v% trypsin, and some release was observed at higher concentrations, 0.005 w/v% was determined as the threshold triggering concentration of trypsin. On the other end, no notable enhancement in the MG cumulative release (%) from MG/CAS-MSNs-COOH was achieved by increasing trypsin concentration over 0.1 w/v% (Figure 4B). Thus, 0.1 w/v% concentration was identified as the optimum trypsin concentration that triggers nearly 84% release of MG from the MG/CAS-MSNs-COOH at pH 7.4, 25 °C, for 24 h (Figure 4C). In the control experiment, only ~20 w/v% of MG leaked from the nanoassembly under the same conditions with no trypsin.

For the mesoporous silica/hydroxyapatite MG/CAS-MSNs/ HAPs (0.6 w/v% casein), the determined optimum concentration of trypsin (0.25 w/v%) was able to release ~95% of MG for 70 h at pH 7.4, 25 °C. Without trypsin, only ~18% MG was released from the system under the same conditions (Figure 4D). Comparing the time-release profiles from the two types of nanoassemblies (MG/CAS-MSNs-COOH and MG/CAS-MSNs/HAPs) shows that the addition of the hydroxyapatite component to the material slightly retards the enzyme-triggered MG release that requires a higher concentration of trypsin, which can be attributed to the affinity of the phosphoprotein casein to hydroxyapatite. However, the overall MG release was more efficient and complete compared with the control experiment baseline.

Enzyme-Responsive In Vitro VTD Release.

After completing triggered drug delivery studies with the model drug element MG, we used VTD, an anticancer drug molecule, to replace the model drug MG for site-directed smart drug delivery design via MSNs-COOH and hybrid MSNs/HAPs. Without any casein, 86.13% of loaded VTD was released from VTD-MSNs-COOH within 48 h, which demonstrated no significant VTD encapsulation inside the pore channels without a pore-blocking capping agent. The optimum concentration of casein (0.6% w/v) efficiently holds 80.6% of the loaded VTD inside the MSNs within 48 h. 0.25 w/v% trypsin was able to trigger 85.45% release of VTD from the VTD/CAS-MSNs-COOH at pH 7.4, 25 °C, for 48 h (Figure 4E). Hybrid VTD-MSNs/HAPs released 82.89% of the loaded VTD without any casein attachment within 50 h. Casein attachment (0.6% w/v) encapsulated 63% of the loaded VTD inside the hybrid MSNs/HAPs for 50 h, representing significant VTD encapsulation inside the pore channels with a pore-blocking capping agent. For VTD/CAS-MSNs/HAPs (0.6 w/v% casein), the determined optimum concentration of trypsin (0.25 w/v%) was able to release ~77% of VTD for 50 h at pH 7.4, 25 °C (Figure 4F).

pH-Triggered In Vitro MG Release.

The presence of the acid-soluble hydroxyapatite component in the MG/CAS-MSN/HAP material allowed us to take advantage of an additional triggering factor of the drug release: lowering pH from 7.4 to 5.0, which mimics a common elevated acidity in cancerous tissue due to increased metabolism.³¹ We followed drug release profiles at pH 5.0 and 7.4 for both MG/CAS-MSNs-COOH (Figure 4C) and MG/CAS-MSNs/HAPs (Figure 4D). For the material lacking the hydroxyapatite component, lowering pH from 7.4 to 5.0 did not substantially affect drug release for 24 h in the absence of trypsin (triangles and squares, Figure 4C). Conversely, lowering pH from 7.4 to 5.0 did trigger MG release from the hydroxyapatite-containing MG/CAS-MSN/HAP nanoassembly in the absence of trypsin. The cumulative release of MG over 70 h drastically increased from ~18% to ~88% (Figure 4D, circles and squares). We attribute this effect to the degradation of the hydroxyapatite component on the hybrid MSNs/HAPs at elevated acidity. This assumption was confirmed by SEM with energy-dispersive X-ray analysis (EDX) imaging, which has shown that the hybrid MSN/HAP nanostructures disintegrated upon elevated acidity (pH = 5), and the hydroxyapatite components almost completely dissolved within the first 6 h (Figures S7 and S8).

pH-Triggered In Vitro VTD Release.

pH-responsive drug delivery studies of MG using both MSNs and hybrid MSNs/HAPs identified that MSNs without the hydroxyapatite component did not produce any MG delivery kinetics upon pH changes from pH 7 to pH 5. Thus, we have replaced the model drug MG for controlled drug delivery via only MSNs/HAPs. In the absence of the enzyme agent, VTD/CAS-MSNs/HAPs (0.6 w/v% casein) only released 27.54% of the loaded VTD for 50 h at pH 7.4, 25 °C. The cumulative release of VTD over 50 h significantly increased from 27.54% to 69.94% upon pH changes from pH 7.4 to pH 5.0 for VTD/CAS-MSNs/HAPs (Figure 4F, circles and downside triangles).

CONCLUSIONS

Covalent attachment of casein to MSNs by amide linkage efficiently trapped inside the particles both the model drug MG and the anticancer drug VTD. Enzymatic cleavage of the casein gatekeeping element with trypsin efficiently released both compounds from the particles, proving the concept of the enzyme-triggered release of VTD sealed inside a drug delivery assembly by an MMP-7 substrate.

The MSN-based nanoassembly for the controlled release of VTD is optimized by selecting casein over BSA as the gatekeeping element and by optimization of the surface density of casein and concentration of the enzyme triggering release.

Electrostatic attachment of casein to the hybrid MSNs/HAPs enables them to trap the model drug MG inside of themselves. Either enzymatic cleavage of the casein gatekeeping element with trypsin or degrading the hydroxyapatite component at a lower pH efficiently released the drug from the particles. This points at the potential of MSNs/HAPs as an alternative to their MSN counterpart for the controlled release of VTD to take additional advantage of the more acidic environment surrounding tumor tissues. This research is currently underway.

The MSNs demonstrated a much higher affinity to CRC cells than to healthy colon fibroblasts. Coupled with the significant internalization of the particles by CRC versus healthy cells, this finding highlights the potential of our nanoassembly for targeted drug delivery, which is a powerful addition to controlled drug release. We have already shown that VTD-dependent expression of UBXN2A leads to the suppression of mortalin protein, an oncoprotein, overexpressed in CRC.⁵ A set of in vitro and in vivo experiments have been designed to examine the apoptotic as well as antimigration and invasion effects of VTD-MSNs. We expect that selective delivery of VTD by nanoparticles will suppress tumor cell migration and invasion because primary and metastatic cancer cells release MMP7 at different levels. Before moving the cancer treatment to a mouse model, we will replace the model MMP-7 substrate casein with the commercially available substrates released by cancer cells and expand the in vitro studies to epithelial cancer cells. In addition to affecting CRC,³² mortalin plays tumorigenic roles in diverse tumor types, including breast cancer,³³ hepatocellular carcinoma,³⁴ and ovarian cancer.³⁵ This current study has the potential to open the door for the application of the targeted delivery and controlled release of VTD beyond CRC.

METHODS

Materials.

Cetyltrimethylammonium bromide (CTAB), tetraethyl orthosilicate (TEOS, 99.98%), 3-aminopropyltriethoxysilane (APTES), fluorescein isothiocyanate (FITC), and eugenol (4-allyl-2-methoxyphenol) were purchased from Sigma-Aldrich. BSA, casein, trypsin solution, absolute ethanol (EtOH), succinic anhydride, 1-ethyl-3-(3-(dimethylamino)-propyl)carbodiimide hydrochloride (EDC), N-hydroxysulfosuccinimide sodium salt (sulfo-NHS), N,N-dimethylformamide (DMF), 2-(N-morpholino)ethanesulfonic acid (MES buffer), and glacial acetic acid were purchased from Fisher Scientific. The reagents used in this study are of analytical grade and used as received.

Instrumentation.

SEM (SIGMA FE-SEM, ZEISS) and TEM were used to characterize the morphology and size of the prepared particles. A powder X-ray diffractometer (Ultima IV, Rigaku) was used to analyze the phase distribution and structure of the prepared powder materials. FTIR spectroscopy (IFS Equinox 55 Spectrometer, Bruker) was carried out to probe a functionalized particle surface. Dzeta potential values and hydrodynamic diameters of the particles were obtained using the Malvern Zetasizer ZS nano series. The drug concentration in solution was monitored by UV spectrometry using a UV-500 UNICAM spectrometer and Shimadzu high-performance liquid chromatography (HPLC) system. Oxford Instruments X-MaxN 50 EDS probes attached to FE-SEM and XPS (K-Alpha, Thermo Fisher) were utilized for EDX elemental distribution and surface element analysis for the particles, respectively. Centrifuging was performed with an Eppendorf 5810R centrifuge. Confocal microscopy studies were conducted using an Axiovert 200 instrument from ZEISS.

SEM.

A powder sample (5 mg) was dispersed in ethanol (10 mL) using an ultrasonication probe. One drop of the suspension was deposited on a reflective face of the silicon wafer and dried at 60 °C for 30 min to obtain a sample specimen for SEM analysis. Particle size and shape were observed via SIGMA FE-SEM (ZEISS) with an accelerating voltage of 2 kV and 4 kV.

TEM.

Particle morphologies were observed via TEM with an accelerating voltage of 200 kV. Monodispersed solution from MSNs (1 mg/mL) was diluted 20 times with Nanopure water. A drop of the resultant solution was deposited on copper-coated TEM grids and air-dried for 2 days in a vacuum desiccator.

PXRD.

The low-angle and wide-angle PXRD spectra for MSNs and hybrid MSNs/HAPs were obtained from a Rigaku Ultima IV diffractometer using Cu Kα radiation (λ = 1.54 A°, 40 kV, 44 mA). The stepwise increase for small-angle was 0.02° s-1 over the range of 0.5° to 5°.

DLS and Zeta Potential.

A Malvern Zetasizer Nano ZS was used to conduct the hydrodynamic diameter and zeta potential measurements for MSNs. The MSNs (1 mg) were dispersed in 10 mL of PBS solution (pH = 7.4, 0.1M) by sonicating for 10 min using an ultrasonication probe. A 2 mL portion of 0.1 mg/mL MSNs suspension was transferred into a low volume cuvette (PLASTIBRAND, semimicro, PMMA, l = 1 cm). Hydrodynamic diameter measurements were performed three times (12 runs for each analysis). An MSN suspension (0.1 mg/mL in PBS, pH 7.4) was transferred to a disposable capillary zeta cell to measure zeta potential. Zeta potentials were measured 5 times for each sample, and each measurement consisted of 10 runs. The presented zeta potentials values are the averages of those measurements. The conductivity of the measured suspension was always above 0.8 mS/cm for PBS.

HPLC Analysis of VTD.

The quantitative determination of VTD was performed by a Shimadzu HPLC system equipped with an LC-20AD VP pump, an SPD-20A variable wavelength UV/vis detector, a DGU-20A degasser model (Shimadzu), and a reverse phase C-18 column (150 mm × 4.6 mm I.D.; particle size 3 μm; Econosphere).

A mixture of methanol and 0.1 M ammonium acetate in a ratio of 60:40 was used as the mobile phase. The filtered mobile phase was pumped at a flow rate of 0.25 mL/min. The column temperature was maintained at 25 °C. The detection wavelength for the eluent was set at 220 nm, and data were acquired, stored, and analyzed with the software lab solution LC/GC (Shimadzu). A standard curve was constructed for VTD in the range of 1.5–150 μg/mL using 150 μ/mL VTD as the external standard. A good linear relationship was observed between the concentration of VTD and the ratio of the peak area of VTD with a high correlation coefficient (r = 0.9999) (Figure S5). The retention time of VTD using the present HPLC method was found to be 18.83 min. The standard curve constructed as described above was used for estimating VTD in the drug-releasing samples of nanostructures.

Synthesis of MSNs.

Synthesis was performed by the detergent-templated method.¹⁷ Specifically, CTAB (2.00 g, 5.48 mmol) was dissolved in 960 mL of Nanopure water. To this solution, 7 mL of 2.00 M NaOH in Nanopure water was added. The solution was stirred at 80 °C, and TEOS (11.477 mL, 0.0514 mol) was then added dropwise (0.5 mL/min). The mixture was stirred for another 2 h at 80 °C to render a white precipitate, which was collected by centrifugation (11000 rpm for 8 min) and washed with absolute ethanol (20 mL × 3) and then with Nanopure water until the supernatant pH value reached 7.00. After each washing step, particles were separated by centrifugation, transferred to a crucible, dried at 60 °C overnight, crushed into a fine powder, calcined at 550 °C in air for 2 h, cooled down to room temperature, and stored in a closed container. Yield was 1.95 g.

Synthesis of FITC-MSNs.

The particles were prepared according to a known procedure.³⁶ First, a solution of FITC (1 mg, 2.56 μmol) and APTES (14 μL, 60.09 μmol) in 15 mL of anhydrous DMF was prepared. Separately, CTAB (2.0 g) was dissolved in 960.0 mL of Nanopure water, and 7 mL of 2.00 M NaOH in Nanopure water was added to the CTAB solution. To this solution, TEOS (11.477 mL, 0.514 mmol) was added dropwise (0.5 mL/min) at 80 °C at stirring. Next, the FITC/APTES/DMF solution was added dropwise (1 mL/min) to the surfactant solution. The mixture was stirred for 2 h at 80 °C to render an orange precipitate, which was separated by centrifugation (11000 rpm for 8 min), washed with methanol three times, washed next with Nanopure water three times, and then dried under vacuum for 12 h (70 cm Hg) at 80 °C. The dried particles (1.9 g) were refluxed for 18 h in a solution of 1.0 mL of 37.4% aqueous HCl and 100.0 mL of methanol to remove the surfactant template (CTAB). The solids were washed with methanol three times (15 mL × 3) and then with Nanopure water three times, separated by centrifugation, and dried at 80 °C in a vacuum oven (70 cm Hg) for 24 h to remove the solvent molecules from the pores of the MSNs. A yellowish-color powder of MSNs-FITC was produced (1.85 g).

Synthesis of MSNs-NH₂ (Amination of MSNs).

MSNs (1.00 g) were dispersed in 30 mL of absolute ethanol by ultrasonication for 5 min using an ultrasonication probe (model FS-250N, 100 W) and aminated by the known procedure.³⁷ After sonication, 1 mL of APTES and 0.6 mL of glacial acetic acid were added to the solution. The mixture was slowly stirred for 24 h. After 24 h, aminated particles were separated by centrifugation. Separated particles were washed with absolute ethanol three times (10 mL × 3). After each washing step, particles were separated by centrifugation and dried at 80 °C in a vacuum oven (70 cm Hg) for 24 h. The collected yield was 0.96 g of MSNs-NH₂.

Synthesis of MSNs-COOH (Carboxylation of MSNs).

Carboxylation of MSNs was performed by the known procedure.³⁸ Succinic anhydride (25 mg) was dissolved in 25 mL of DMF to prepare a 1 mg/mL succinic anhydride solution. Aminated MCM-41 particles (200 mg) were mixed with the succinic anhydride solution and gently stirred for 24 h. Particles were separated by centrifugation (11000 rpm for 8 min) and washed with DMF three times (10 mL × 3). The separated product was dried at 50 °C in a vacuum oven (70 cm Hg) for 24 h. MSNs-COOH were collected (195 mg).

Synthesis of Hybrid MSNs/HAPs.

For the synthesis of hybrid MSNs/HAPs, the known procedure²² was slightly changed to introduce Ca²⁺ and PO₄³⁻ with CaCl₂ and Na₂HPO₄ as their respective sources. In a typical procedure, to the solution of CTAB (2.00 g, 5.48 mmol) under gentle stirring 960 mL of Nanopure water and 7 mL of NaOH 2.00 M in Nanopure water was added, which produced a solution with pH 11. Next, Na₂HPO₄(1.275 g) was added to the solution, and the mixture was stirred at 80 °C for 1 h. Then, TEOS (11.48 mL) was added dropwise (0.5 mL/min) to the above solution under vigorous stirring, followed by a one-portion addition of 20 mL of an aqueous solution containing 2.20 g of CaCl₂. The mixture was stirred continuously for another 4 h to render a white precipitate that was collected by centrifugation, washed three times with Nanopure water, and then washed three times with absolute ethanol (15 mL × 3). Finally, collected particles were dried at 60 °C in a vacuum oven (70 cm Hg) overnight, crushed into a fine powder, calcined at 550 °C in a furnace for 2 h to remove the CTAB, and cooled to room temperature. The MSN/HAP material (2.15 g) was obtained.

Immunofluorescence Assay.

Human HCT-116 and CCD-33Co cell lines provided by the ATCC (American Type Culture Collection; Manassas, VA, USA) were cultured per ATCC instructions. HCT-116 is a primary colorectal carcinoma cell line, and CCD-33Co is a noncancer human colon fibroblast cell line originating from the large intestine. Cells received fresh media 1 day before treatment. MSNs conjugated to FITC or RBITC were dispersed in pure DMSO (5 mg/mL) followed by 3 min of sonication (30 s sonication at 25% amplitude 5 times with short breaks on ice). The MSNs-RBITC used for internalization experiments showed much better solubility in dimethyl sulfoxide (DMSO) than in MSNs-FITC in DMSO by roughly an order of magnitude. We used 10 s sonication at 25% amplitude 3 times with short breaks on ice between sonications. After dilution in corresponding media (60 ng/mL), MSNs-FITC were subjected to another 30 s of sonication using the same ultrasound amplitude described above. Cells seeded in 6-well plates on coverslips received media containing dissolved MSNs-FITC (confocal microscopy studies) or MSNs-RBITC (time-point internalization experiments) and incubated at 37 °C (5% CO₂). Control wells received only media. After 90 min of incubation, unattached MSNs-FITC were removed by washing 2 times at room temperature with PBS without calcium and magnesium. Cells were fixed and stained with antibeta-tubulin and DPAI as previously described.³⁹ Stained cells were subjected to confocal microscopy studies. To confirm accurate images of the association and internalization of MSNs-FITC with treated cells, we used z-stack confocal images to capture a three-dimensional picture of cells and associated nanoparticles. All conducted cellular and molecular experiments have been approved by the Institutional Biosafety Committee of The University of South Dakota under protocol number IBC-19-06.

In Vitro Cellular Uptake by Flow Cytometry.

We used a method previously described by Xu et al.⁴⁰ Briefly, HCT-116 cells were seeded on a 6-well plate and incubated (5% CO₂ at 37 °C). After 48 h, cells were treated with MSNs-RBITC (final concentration: 60 ng/mL) incubated at 37 °C under 5% CO₂ for 4, 24, and 48 h. We used MSNs-RBITC instead of FITC since HCT-116 cancer cells have a green autofluorescence. Unattached MSNs were thoroughly washed with ice-cold PBS two times. Then the cells were detached with Accutase cell detachment solution (BD Biosciences, USA) followed by the addition of culture medium to neutralize the Accutase. After centrifugation (400 g for 3 min), the supernatant was discarded, and one mL PBS (1×) was added to resuspend cells for flow cytometry using a BD Accuri C6 flow cytometer equipped with a 488 nm argon laser. AFL2 bandpass emission (585/40) was used to measure cells with internalized rhodamine. The fluorescence of MSNs-RBITC in 10,000 events were collected in triplicate. The cells untreated with MSNs were used as the control. The flow cytometry data were analyzed using FlowJo software followed by statistical analysis using Prism-GraphPad software. The one-way Anova and Tukey’s multiple comparison test was applied to compare the three time-points. Values of p < 0.05 were considered significantly different.

MG Loading of MSNs-COOH and MSNs/HAPs.

A dispersion of MSNs-COOH or MSNs/HAPs (5 mg) in 1.5 mL MG solution (0.05 mg/mL in 0.1 M PBS buffer, pH 7.4) was placed on a rotary mixer for 24 h at room temperature. Then it was centrifuged (11000 rpm, 8 min) to remove unencapsulated MG. The separated particles were washed with a 1.5 mL PBS buffer (pH 7.4, 0.1 M) three times to remove the surface-adsorbed MG. All washing decantants were collected, and the amount of nonloaded MG was determined by UV/vis absorbance at λ_max 618 nm with the aid of a calibration graph of known concentrations of MG solutions (Figure S5). The drug encapsulation efficiency (DEE) and drug loading efficiency (DLE) of MG for particles were calculated using the following equations:

$$\begin{gathered} \text{DEE}=\frac{\text{fed drug}(\mu \mathrm{g})-\text{drug in the supernatant}(\mu \mathrm{g})}{\text{fed drug}(\mu \mathrm{g})}\times 100 \\ \text{DLE}=\frac{\text{fed drug}(\mu \mathrm{g})-\text{drug in the supernatant}(\mu \mathrm{g})}{\text{mass of the dried nanoparticles}(\mu \mathrm{g})}\times 100 \end{gathered}$$

The obtained MG-loaded particles from MSNs-COOH and MSNs/HAPs will be referred to as MG/MSNs-COOH and MG/MSNs/HAPs, respectively.

VTD Loading of MSNs-COOH and MSNs/HAPs.

The MSNs-COOH or MSNs/HAPs (5 mg) were dispersed in 1.5 mL of a VTD solution (0.01 mg/mL in 0.1 M PBS buffer, pH 7.4). Similar steps as in the MG loading procedure were followed to load VTD into the particles. The obtained VTD-loaded particles from MSNs-COOH and MSNs/HAPs will be referred to as VTD/MSNs-COOH and VTD/MSNs/HAPs, respectively. The supernatant was collected, and the amount of unloaded VTD was determined using the HPLC method. Twenty microliters of collected supernatant was injected and analyzed by a Shimadzu HPLC system equipped with a LC-20AD pump, SPD-20A UV/vis detector, and DGU-20A degasser model. An Econosphere C-18 column (4.6 mm × 150 mm × 3 μm) was used to carry out separations with 60% methanol and 40% water (containing 0.1 M ammonium acetate) in mobile phase at a uniform flow rate of 0.25 mL per minute. The detection wavelength was set at 220 nm. The operating temperature of the column was maintained at 25 °C by the column-oven.

Casein and BSA were used as capping agents to hold the drug inside the MSN-COOH or MSN/HAP channels.

Synthesis of Covalently Protein-Capped Drug-Loaded MSNs-COOH.

Activation and covalent coupling of carboxylated particles was performed by the known procedure.³⁸ VTD/MSNs-COOH or MG/MSNs-COOH (5 mg) were mixed thoroughly with 1 mL of MES buffer (0.1 M, pH 6.0). Then, 24 μL of 250 mM EDC in Nanopure water and 240 μL of 250 mM sulfo-NHS (in 0.1 M MES buffer, pH 6.0) was quickly added to the 1 mL of drug/MSNs-COOH. The mixture was vortexed and incubated by mixing on a rotary wheel for 30 min at room temperature. Then it was centrifuged to pellet the particles. The supernatant was decanted, and the particles were gently washed with 1 mL MES buffer (pH 6.0) to remove excess EDC and sulfo-NHS. The collected pellet was redispersed in 400 μL of PBS buffer (0.1 M, pH = 7.4), followed by the addition of casein (1 mL of 0.1 M PBS buffer, pH 7.4) solution at given concentrations (0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, and 0.8%, w/v). The suspension was mixed for 5 h on a rotary wheel at room temperature. Finally, casein-capped drug-loaded particles (VTD/CAS-MSNs-COOH or MG/CAS-MSNs-COOH) were centrifuged and rinsed with 1 mL of PBS buffer (pH 7.4) twice gently to remove the noncoupled casein. The combined supernatants were analyzed by a BCA assay.

The same procedure was used for the covalent attachment of BSA to drug-loaded MSNs-COOH.

Synthesis of Electrostatically Casein-Capped Drug-Loaded MSNs/HAPs.

Next, casein was electrostatically coupled to the MG/MSNs/HAPs. A casein solution of a given concentration in 1 mL of 0.1 M PBS buffer (pH 7.4) was mixed with 400 μL of a suspension containing 5 mg of MG/MSNs/HAPs. The suspension was mixed for 5 h on a rotary wheel (24 rpm) at room temperature. Casein-coupled particles (MG/CAS-MSNs/HAPs) were centrifuged out (11000 rpm for 8 min) and rinsed twice with 1 mL of PBS buffer (pH 7.4) to remove excess casein. The combined supernatants were analyzed for protein content by BCA assay.

Enzyme-Triggered In Vitro MG Release.

The MG release studies were performed by collecting the time profiles of the MG concentration in solution without (control series) and in the presence of trypsin (trypsin series).

The control experiment series: MG/CAS-MSNs-COOH (5 mg) was dispersed in 1.5 mL of 0.1 M PBS buffer (pH 7.4) at room temperature and placed in a rotary wheel for the duration of the experiment. At the specified time points (sampling times), the suspension was centrifuged to pellet the particles. A 300 μL portion of the supernatant was taken to measure MG release through absorbance at λ = 617 nm. Then, 300 μL of PBS buffer was added to maintain the supernatant volume, the pelleted particles were resuspended, and the rotation was continued until the next sampling point.

The trypsin experiment series: MG/CAS-MSNs-COOH (5 mg) was dispersed in 1.1 mL of 0.1 M PBS buffer solution (pH 7.4) followed by 0.4 mL of trypsin solution (0.1 M PBS) at a specified concentration, which was added at room temperature and placed in a rotary wheel for the duration of the experiment. The sampling of MG concentrations was performed as in the control experiment series. The time when the trypsin mixture was added to the suspension of casein-capped drug-loaded particles was set as zero. The MG release profile was observed over 24 h. The same procedure was followed for MG/CAS-MSNs/HAPs.

A series of trypsin-concentrated solutions (400 μL) in 0.1 M PBS buffer (pH 7.4) was used to determine the optimum trypsin concentration that releases the maximum amount of drug from the protein-coupled drug-loaded materials.

Enzyme-Triggered In Vitro VTD Release.

The VTD release studies were performed in solution without (control series) and in the presence of 0.25% w/v trypsin (trypsin series).

The control experiment series: VTD/CAS-MSNs-COOH (5 mg) was dispersed in 1.5 mL of 0.1 M PBS buffer (pH 7.4) at room temperature and placed in a rotary wheel for the duration of the experiment. At the specified time points (sampling times), the suspension was centrifuged to pellet the particles. A 300 μL portion of the supernatant was taken to measure VTD release by the Shimadzu HPLC system. Then, 300 μL of 0.1 M PBS buffer was added to maintain the supernatant volume, the pelleted particles were resuspended, and the rotation was continued until the next sampling point.

The trypsin series: VTD/CAS-MSNs-COOH (5 mg) were dispersed in 1.1 mL of 0.1 M PBS buffer solution (pH 7.4) followed by 0.4 mL of 0.25% w/v trypsin solution, which was added at room temperature and placed in a rotary wheel for the duration of the experiment. Sampling of VTD concentrations was performed as in the control experiment series. The time when the trypsin mixture was added to the suspension of casein-capped VTD-loaded particles was set as zero. The VTD release profile was observed over 48 h.

pH-Triggered In Vitro MG Release.

pH-triggered MG release studies were performed from both casein-capped MSNs-COOH and MSNs/HAPs at two pH values (7.4 and 5.0) by the following procedure. MG/CAS-MSNs-COOH or MG/CAS-MSNs/HAPs (5 mg) was dispersed in 1.5 mL of each of two different pH 0.1 M PBS solutions (7.4 and 5.0) at room temperature. Sampling of MG concentrations was performed according to the protocol of the enzyme-triggered drug release described earlier in this paper. All drug release experiments were repeated three times for each condition to produce three replicates, and the cumulative percentage of the released drug was determined by the equation in the section immediately following.

pH-Triggered In Vitro VTD Release.

pH-triggered VTD release studies were performed from only casein-capped MSNs/HAPs at two pH values (7.4 and 5.0) by the same procedure described for pH-responsive model drug release experiments.

All drug release experiments were repeated three times for each condition to produce three replicates, and the cumulative percentage of the released drug was determined by the following equation:

$$\begin{gathered} \text{cumulative percentage release} \\ =\frac{\text{drug released at time}t+\Sigma \text{withdrawn drug before time}t}{\text{total drug loaded into nanoparticles}}\times 100 \end{gathered}$$

Statistical Analysis.

All statistical values presented in Figure 3 for flow-cytometry experiments were analyzed with the software GraphPad Prism 9 and utilized the one-way ANOVA. A p value of ≤0.05 was used to compare mean values and signify a statistically significant result.

ABBREVIATIONS

CRC

colorectal cancer

VTD

veratridine

MG

malachite green

MSNs

mesoporous silica nanoparticles

HAPs

hydroxyapatite particles