Effect of an alkyl spacer on the morphology and internalization of MUC1 aptamer‐naphthalimide amphiphiles for targeting and imaging triple negative breast cancer cells

Abstract Despite decades of research, there are few targeted treatment options available for triple negative breast cancer (TNBC), leaving chemotherapy, and radiation treatment regimes with poor response and high toxicity. Herein aptamer‐amphiphiles were synthesized which selectively bind to the mucin‐1 (MUC1) glycoprotein that is overexpressed in TNBC cells. These amphiphiles have a fluorescent tail (1,8‐naphthalimide or 4‐nitro‐1,8‐naphthalimide) which enables self‐assembly of the amphiphiles and allows for easy visualization without the requirement for further conjugation of a fluorophore. Interestingly, the length of the alkyl spacer (C4 or C12) between the aptamer and tail was shown to influence the morphology of the self‐assembled structure, and thus its ability to internalize into the TNBC cells. While both the MUC1 aptamer‐C4‐napthalimide spherical micelles and the MUC1 aptamer‐C12‐napthalimide long cylindrical micelles showed internalization into MDA‐MB‐468 TNBC cells but not the noncancerous MCF‐10A breast cells, the cylindrical micelles showed greatly enhanced internalization into the MDA‐MB‐468 cells. Similar patterns of enhanced binding and internalization were observed between the MUC1 aptamer‐C12‐napthalimide cylindrical micelles and SUM159 and MDA‐MB‐231 TNBC cells. The MUC1 aptamer cylindrical micelles were not toxic to the cells, and when used to deliver doxorubicin to the TNBC cells, were shown to be as cytotoxic as free doxorubicin. Moreover, a pharmacokinetic study in mice showed a prolonged systemic circulation time of the MUC1 aptamer cylindrical micelles. There was a 4.6‐fold increase in the elimination half‐life of the aptamer cylindrical micelles, and their clearance decreased 10‐fold compared to the MUC1 aptamer spherical micelles. Thus, the MUC1 aptamer‐C12‐napthalimide nanofibers represent a promising vehicle that could be used for easy visualization and targeted delivery of therapeutic loads to TNBC cells.


| INTRODUCTION
Breast cancer is a leading cause of cancer related mortalities, and TNBC represents approximately 15-20% of all invasive types of breast cancer. 1 TNBC, characterized by its lack of estrogen and progesterone receptors and normal HER2 expression, has proven extremely difficult to treat, even after over a decade of research. 2 Due to the lack of many primary therapeutic targets, there are few targeted therapy options available, leaving only surgery, chemotherapeutics, and radiation therapy as viable treatment options. [2][3][4][5] This route tends to be less effective with more pronounced side effects.
As such, discovery of targeted delivery strategies is critical to improving breast cancer patient outcomes and experiences. 6,7 One potential target is the transmembrane glycoprotein MUC1, which has been the subject of significant clinical research for both antibody-based treatments and vaccines. Not only is MUC1 known to be overexpressed in most carcinomas, including~90% of breast cancer tumors, but it is linked to immune evasion, cancer progression, and metastasis. [8][9][10][11][12][13] Incomplete O-linked glycosylation (underglycosylation) of MUC1 expressed on the surface of epithelial cancer cells results in exposed peptide epitopes and shortened carbohydrate chains, such as the Thomsen-nouvell (Tn) antigen. [14][15][16] Tn is a single Nacetylgalactosamine (GalNAc) added to a serine or threonine of a protein, and represents the first step of an O-linked glycosylation pathway. Thus, underglycosylation of MUC1 on cancer cells can generate glycans truncated at the initial GalNAc (Tn), leaving them exposed for targeting. The MUC1 glycoprotein typically internalizes via clathrin-mediated endocytosis or macropinocytosis and to a lesser extent via caveolae-mediated endocytosis, thus enabling cell internalization of targeted nanoparticles bound to it. [17][18][19][20] In this study, a single-stranded DNA (ssDNA) aptamer was used to target MUC1 which specifically binds to the GalNAc regions of the protein. More importantly, the MUC1 aptamer was shown to bind to MUC1+ cancer cells, but not MUC1− cells or even MUC1+ normal human primary epithelial cells displaying fully glycosylated MUC1. 14 Thus, this aptamer, targeting the underglycosylated MUC1 overexpressed on the surface of cancer cells, presents a promising targeting ligand for the design of a novel delivery system aimed at TNBC.
While aptamers are often conjugated to other drug delivery systems and molecular diagnostic tools such as quantum dots or peptide, polymeric, or Au nanoparticles, the ssDNA itself can be induced to self-assemble into nanostructures via conjugation of a hydrophobic tail. [21][22][23][24][25][26] Direct conjugation of ssDNA aptamers to hydrophobic tails largely generates spherical micelles, however the inclusion of an alkyl spacer between a hydrophobic di-alkyl tail and hydrophilic ssDNA headgroup induced more complex and interesting morphologies. [27][28][29][30] These may in turn be able to impact the ability of a delivery system to associate with cells given the widely reported impact of size and shape on nanoparticle delivery. [31][32][33][34] The hydrophobic tail of the amphiphile promotes self-assembly, and can also serve other functions such as enabling visualization or delivering a therapeutic load. 35,36 In this work, 1,8-naphthalimide (or 4-nitro-1,8-naphthalimide for enhanced fluorescence) was chosen as an example of a hydrophobic fluorophore that can serve as a tail for the design of the MUC1 aptamer-amphiphile, as it has been used in a wide variety of in vitro and in vivo studies for imaging, and shows no cytotoxicity on its own. 37 Altogether, this makes it a suitable choice for the amphiphile tail. Incorporation of a spacer between the hydrophobic tail and hydrophilic headgroup has been shown to play a vital role on the assembly and binding properties of both ssDNAamphiphiles and peptide-amphiphiles. 27,30,38 While a myriad of spacers of varying lengths and types could have been used, in this work C 4 and C 12 spacers were utilized as previous work has demonstrated that alkyl spacers can promote the assembly of ssDNA supramolecular architectures. 29,30 After the synthesis of different MUC1 aptamer-amphiphiles, circular dichroism (CD) spectroscopy was used to evaluate the effect of the tail conjugation on the aptamer's secondary structure, and their assembled structures were identified via cryo-

| RESULTS AND DISCUSSION
The MUC1 aptamer-amphiphiles were synthesized as described in Scheme 1. Successful synthesis of the tail-spacer molecules was verified by 1 H NMR (Figures S1 and S2), and liquid chromatography-mass spectrometry was used for the verification of the masses of the MUC1 aptamer-amphiphiles. All initial experiments were performed with the MC4N and MC12N aptameramphiphiles, and the ssDNA in all nanoparticles of this study is the aptamer sequence, which is typical for aptamer micelles. [39][40][41][42] Conjugation of the spacer and aptamer did not inhibit the fluorescence  (Figure 1(a)), which showed that the MUC1 aptamer, MC4N and MC12N aptamer-amphiphiles all had a peak maximum at 275 nm, and a minimum at~244 nm, consistent with a stem-loop secondary structure that has been observed before for the MUC1 aptamer. 30 Therefore, conjugation of the MUC1 aptamer to the tail-spacer molecules did not affect the secondary structure of MUC1 aptamer. Cryo-TEM was used to characterize the self-assembled structures formed by the two different MUC1 aptamer-amphiphiles. As can be seen in Figure 1(b), the MC4N aptamer-amphiphiles self-assembled into spherical micelles with 11 ± 2 nm in diameter (n = 100), while the MC12N aptamer-amphiphiles formed micrometer long cylindrical micelles with a diameter of 13 ± 3 nm (n = 60) ( Figure 1(c)). The packing parameter, that is commonly used to describe the shapes of selfassembled molecules, is defined as the (cross-sectional area of the hydrophobic tail)/(equilibrium area per molecule at the aggregate surface). In our case, the cross-sectional area of the hydrophobic tail-spacer, regardless of the alkyl spacer length, is defined by the cross-sectional area of the 1,8-naphthalimide, while the equilibrium area occupied by each amphiphile at the aggregate surface is influenced by the steric and electrostatic repulsions present between the aptamer headgroups. 43 The packing parameter would therefore be identical for the MC4N and MC12N aptamer-amphiphiles. However, it has also been shown theoretically that for the same value of the packing parameter, the sphere-to-rod transition parameter increases with increasing tail length, thus making the formation of cylindrical micelles more favorable for MC12N aptamer-amphiphiles with the longer C 12 hydrophobic spacer than the MC4N aptamer-amphiphiles with the shorter C 4 spacer. 43 While both self-assembled nanoparticles should have the inherent targeting ability of the MUC1 aptamer, this distinct difference in structure enables the probing of the impact of the size and shape on their binding and internalization with the TNBC cells.  Our work demonstrates that the overall shape and size of the self-assembled structures dramatically impacts cell association and internalization behavior in the presence of a targeting aptamer. There is evidence in the literature regarding the effect of physical properties such as size, shape, and stiffness on cell uptake. 47 For example, gold and some polymeric rod-like or cylindrical nanoparticles showed reduced cellular uptake compared to spheres, while targeted rod-like polystyrene nanoparticles coated with antibodies exhibited higher specific cell uptake than their spherical counterparts. 20,32 Mathematical modeling showed that elongated particles coated with ligands can form more substantial contacts with their targeting receptors, compared to spheres, due to the engagement of multiple ligand-receptor interactions. 48 Overall, the data suggest that the self-assembled MUC1 aptamer nanofibers internalize specifically into different TNBC cells with great affinity and can be seen without additional staining due to the presence of the naphthalimide tail. Thus, the MC12N aptamer nanofibers were used further to examine the ability of these nanoparticles to deliver a therapeutic load, such as DOX, to the TNBC cells. DOX has been shown to intercalate into the double-stranded regions of stemloops, thus forming physical complexes with aptamers through noncovalent intercalations. 49  This work lends further credence to the significant impact size and shape can have on targeted delivery and the growing body of evidence that with the appropriate morphology, one can achieve increased specific targeting and internalization into cancer cells, as well as longer circulation times. Thus, the MUC1 aptamer nanoparticles designed in this study present a promising platform for targeted delivery.

| CONCLUSIONS
Given the difficulty of treating TNBC, the development of a targeted nanoparticle delivery system would be highly impactful. Such a system would enable enhanced efficacy and greatly reduced off-target effects for a variety of treatment options including gene therapy and traditional chemotherapeutics. In this work, a MUC1 aptamer-amphiphile was designed which provides targeting, an entry mechanism into the MUC1-expressing TNBC cells, and visualization capabilities due to its naphthalimide tail. The use of a MUC1 aptamer could also be used with a wide variety of cancers, as nearly all carcinomas overexpress MUC1. Two different amphiphile constructs were generated to examine the impact of the spacer length on self-assembly and cell association. Cryo-TEM images demonstrated that the short C 4 alkyl spacer led to the production of spherical micelles, while the C 12 alkyl spacer led to micrometer long cylindrical micelles. The targeting was highly effective, with neither construct showing association with the MCF-10A cells, while both vehicles showing significant internalization into the TNBC cells. That said, the cell internalization of the MC12N nanofibers was greatly enhanced over that of the MC4N spherical micelles, suggesting that the shape of the aptamer nanoparticle plays an enormous role in its ability to interact and internalize with the cancer cell. In addition, it was shown that the MUC1 aptamer nanofibers could be used effectively for the delivery of a chemotherapeutic, such as DOX, and had a more favorable pharmacokinetic profile than the MUC1 spherical micelles, thus making them a promising targeted drug delivery system.

SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of this article.