6927wileyonlinelibrary.com therapy requires medicine specifi cally formulated and delivered to each patient. However, personalized drug delivery is still a very substantial challenge. We and others have shown that surface modifi cation of nanoparticles by libraries of small organic molecules can markedly alter the uptake in different types of cells, and that the modulation of uptake by surface chemistry can be modeled and predicted quantitatively. [4][5][6] We hypothesized that dual targeting nanoconstructs, i.e., one ligand targets a common receptor overexpressed in a type of cancer and another ligand targets a receptor related only to an individual's genetic background, may get us closer to the goal of personalized delivery. Biochemical studies of diverse markers on cells are complex and potentially time-consuming, and high throughput materials science such as nanocombinatorial chemistry [ 7 ] can provide rapid results that are synergistic to these traditional and more detailed biochemical studies. In particular, nanoparticles, increasingly used to deliver therapies to cells or in disease diagnostics, can also play an important role in understanding how to selectively target cells. Cell uptake of nanoparticles can depend on a number of factors, such as size, shape, surface chemistry and charge, surface coatings, etc. [ 8 ] We have developed a library of nanoparticles that display a diverse array of surface chemistries and have used these to probe the interaction of functionalized nanoparticles with cells, proteins, and enzymes. [ 4,6,9 ] To tackle the complex issue of understanding how to specifi cally target multiple markers on cells we have devised experiments employing a bespoke array of surface functionalized nanoparticles, where the well-known folate targeting moiety (folic acid, FA) is combined with a combinatorial mixture of different surface chemistries previously shown to be taken up selectively by different types of cells. [ 10 ] Herein, we describe how these data may be used to generate quantitative numerical models that predict the uptake of nanoparticles by several types of cells derived from common tumors, identify potential synergistic interactions of folate with other types of surface chemistry in cell-specifi c targeting, and describe how different surface chemistries relate to the specifi c uptake. These are data-driven machine-learning models that provide quantitative predictions of nanoparticle uptake in complex environments, albeit with some sacrifi ce of mechanistic insight.
Results and DiscussionIt is clear from inspection of Figure 1 and Table 1 that the surface chemistry has a marked effect on the degree of uptake
Robust Prediction of Personalized Cell Recognition from a Cancer Population by a Dual Targeting Nanoparticle LibraryTu C. Le , Bing Yan , and David A. Winkler * Nanomaterials are used increasingly in diagnostics and therapeutics, particularly for malignancies. Effi cient targeting of nanoparticles to specifi c cells is an important requirement for the development of successful...