Understanding How Cationic Polymers' Properties Inform Toxic or Immunogenic Responses via Parametric Analysis

Abstract

Cationic polymers are widely used materials in diverse biotechnologies. Subtle variations in these polymers' properties can change them from exceptional delivery agents to toxic inflammatory hazards. Conventional screening strategies optimize for function in a specific application rather than observing how underlying polymer-cell interactions emerge from polymers' properties. An alternative approach is to map basic underlying responses, such as immunogenicity or toxicity, as a function of basic physicochemical parameters to inform the design of materials for a breadth of applications. To demonstrate the potential of this approach, we synthesized 107 polymers varied in charge, hydrophobicity, and molecular weight. We then screened this library for cytotoxic behavior and immunogenic responses to map how these physicochemical properties inform polymer-cell interactions. We identify three compositional regions of interest and use confocal microscopy to uncover the mechanisms behind the observed responses. Finally, immunogenic activity is confirmed in vivo. Highly cationic polymers disrupted the cellular plasma membrane to induce a toxic phenotype, while high molecular weight, hydrophobic polymers were uptaken by active transport to induce NLRP3 inflammasome activation, an immunogenic phenotype. Tertiary amine- and triethylene glycol-containing polymers did not invoke immunogenic or toxic responses. The framework described herein allows for the systematic characterization of new cationic materials with different physicochemical properties for applications ranging from drug and gene delivery to antimicrobial coatings and tissue scaffolds.

Figure 1

High throughput synthesis of 107 polymers. Statistical copolymers were prepared via RAFT polymerization using 50–100% of the amine-containing monomers (R₁ = BocAEMA or DMAEMA) and 0–50% of the hydrophobicity-modifying monomers (R₂ = TEGMA or BMA) at five different molecular weights (M_n = 7.5–60 kg/mol) to map a broad domain space of physicochemical properties.

Figure 2

Results and analysis of high throughput immunological and toxicity screening. (A) IL-1β and (B) LDH screens at high (100 μg/mL), medium (25 μg/mL), and low (6.25 μg/mL) concentrations for each of the 107 polymer library entries.

Figure 3

Identification of trends in the high throughput data set. (A) A principal component analysis was conducted on the high throughput screen to identify structure–property relationships. The first two principal component axes (PCA-1 and PCA-2) are plotted. Three regions of interest are identified as shown. (B, C) Validation of IL-1β secretion results in BMDCs from (B) wild-type and (C) NLRP3-deficient mice. Polymers were incubated with LPS-primed BMDCs at the indicated concentrations for 5 h, and IL-1β was analyzed via ELISA. A two-way ANOVA with Dunnett’s multiple comparison’s test was used to determine statistical significance relative to the LPS only treatment (no label = not significant).

Figure 4

Imaging the rupture of cells treated with representative polymers reveals distinct modes of rupture and death. Cellular swelling was induced by the inflammasome-activating 60 kg/mol DMAEMA-BMA copolymer but not the toxic 15 kg/mol AEMA-BMA copolymer when treated with three different cell lines. This phenomenon was found to persist independently of NLRP3 using NLRP3-KO cells. Time lapse videos can be found in Movies S1–S3 and Figures S13, S15, and S16. For the image denoted with “&”, t = 40 min as all cells died and imaging was stopped. The scale bar is representative of all images.

Figure 5

Analysis of cationic polymer localization in the cell for 8 representative polymers. Treatment of THP-1 cells with AF488-labeled polymers was employed to determine cellular localization of polymers at 15, 30, and 60 min after treatment with cells. Cells were LPS primed, stained with the indicated dyes, treated with 100 μg/mL polymers for the indicated times, and then washed and immediately imaged using a confocal microscope (40× oil lens). The scale bar is representative of all images.

Figure 6

Probing the role of active transport in inflammasome activation induced by representative polymers. THP-1 ASC-GFP cells were treated with (A) the indicated polymers or (B) the indicated polymers + cytochalasin D, and cell death and ASC speck formation were evaluated as a function of time and plotted as shown. Time lapse videos can be found in Movies S4–S5 and Figures S18–S21. (C) Models of cell–polymer interactions for each of the three classes of polymers identified in this study are depicted.

Figure 7

In vivo model of immunotoxicity induced by representative polymers. (A) Experimental paradigm for immunotoxicity study and survival after injection. (B) Body weight was monitored 1, 3, and 6 h after injection for mice treated with each polymer. (C–F) Serum cytokines assayed 6 h after injection of polymers was determined via multiplex cytokine panel. (G) Spleen leukocyte composition (% of live CD45⁺ cells) averaged for mice treated with each polymer. For (B), statistics were conducted using one-way ANOVA with Tukey’s multiple comparisons test relative to PBS. For (C–F), an analogous test was performed, where all of the multiple comparisons were queried.