Engineering adjuvants for predictable immunity

Insight from Steven Reed

A designer approach to adjuvant development may be required to improve upon natural immunity, to reliably induce protective immunity to pathogens against which we have no effective vaccines, and to develop effective therapeutic vaccines for cancer and other types of disease for which current therapies are inadequate.

The quest to design biological structures for adjuvant use began many years ago, exemplified by the work of Edgar Ribi to dissect the adjuvant activity from LPS while eliminating the toxic components. More recent work has defined the precise structure–function aspects of adjuvant binding to toll-like receptors (TLRs) using synthetic chemical entities. However, most adjuvants in current use, including those categorized loosely as “alum,” are not homogeneous and their composition is not well characterized. The effects of chemical composition, charge, particle size, shape, and particle association (e.g., encapsulation and internalization versus surface binding) all influence immunological responses, but precise, controlled studies with defined compositions and controlled variables, using innate responses ex vivo to predict adaptive responses in vivo, are relatively few.

Schematic of a layered double hydroxide showing the positively charged cation layers (red/blue circles) and negatively charged anion layers (green circles) sandwiched between two layers of water molecules.

The study by Williams et al. in this issue uses immune response readouts from cultured human and mouse dendritic cells to mathematically model the relationship between adjuvant materials and the immune responses they induce, adding a rational approach to the concept that “immunity can be determined purely by chemistry.” The authors used physically and chemically defined structures that resemble alum (layered double hydroxides [LDH]; see figure). The modeling predicted that varying three key physicochemical properties of LDHs (the ionic radius, the c-parameter [the interlayer spacing within the LDH], and the zeta potential [the magnitude of the electrical charge of the layer around the LDH particle]; see figure) would influence innate and adaptive responses in a predictable manner. These predictions were tested and verified in mouse immunization studies and ex vivo studies on human macrophages.

While it may still be some way off, this novel mathematical approach to predicting the immune responses induced by candidate crystalline adjuvants offers the exciting possibility of rational design of superior adjuvants.