Antigen structure affects cellular routing through DC-SIGN

Abstract

Dendritic cell (DC) lectins mediate the recognition, uptake, and processing of antigens, but they can also be coopted by pathogens for infection. These distinct activities depend upon the routing of antigens within the cell. Antigens directed to endosomal compartments are degraded, and the peptides are presented on major histocompatibility complex class II molecules, thereby promoting immunity. Alternatively, HIV-1 can avoid degradation, as virus engagement with C-type lectin receptors (CLRs), such as DC-SIGN (DC-specific ICAM-3-grabbing nonintegrin) results in trafficking to surface-accessible invaginated pockets. This process appears to enable infection of T cells in trans We sought to explore whether antigen fate upon CLR-mediated internalization was affected by antigen physical properties. To this end, we employed the ring-opening metathesis polymerization to generate glycopolymers that each display multiple copies of mannoside ligand for DC-SIGN, yet differ in length and size. The rate and extent of glycopolymer internalization depended upon polymer structure-longer polymers were internalized more rapidly and more efficiently than were shorter polymers. The trafficking, however, did not differ, and both short and longer polymers colocalized with transferrin-labeled early endosomes. To explore how DC-SIGN directs larger particles, such as pathogens, we induced aggregation of the polymers to access particulate antigens. Strikingly, these particulate antigens were diverted to the invaginated pockets that harbor HIV-1. Thus, antigen structure has a dramatic effect on DC-SIGN-mediated uptake and trafficking. These findings have consequences for the design of synthetic vaccines. Additionally, the results suggest strategies for targeting DC reservoirs that harbor viral pathogens.

Keywords: C-type lectin; HIV; antigen; endocytosis; polymer.

Fig. 1.

Glycopolymer probes of DC-SIGN endocytosis. A panel of glycopolymers 1–4 of defined length were synthesized bearing an aryl mannoside ligand (red) for DC-SIGN as well as Alexa Fluor 488 (green) for visualization.

Fig. 2.

Comparison of fluorophore-labeled glycopolymer uptake in DC-SIGN–positive (Raji/DC-SIGN) and DC-SIGN–negative (Raji) cells. (A) Cells were treated with glycopolymers 1–4 (40 μM in mannoside ligand for 1 and 2; 10 μM in mannoside ligand for 3 and 4) for 30 min at 37 °C, and internalization was visualized via confocal microscopy. Higher concentrations of glycopolymers 1 and 2 were required to visualize internalization. (Scale bars, 10 μm.) (B) Cells were treated with 1–4 (0.5, 5, 50, and 150 μM in mannoside ligand) for 40 min and placed on ice, and flow cytometry was performed. The resulting fluorescence (arbitrary units) was normalized to account for the average number of fluorophores appended to each polymer.

Fig. 3.

Trafficking of soluble and particulate glycopolymers. (A) TEM of soluble and particulate 275mer polymers. (B and D) Soluble (B) or particulate (D) glycopolymers (green) were added (mannose concentration of 10 μM) to moDCs at 37 °C for 30 min, and their trafficking to transferrin-labeled early endosomes (red) was monitored via confocal microscopy. (C and E) Polymer/transferrin colocalization in B and D, respectively, was assessed at the indicated time points. Pearson’s coefficient was obtained by the Colocalization Threshold plugin in ImageJ using n > 50 cells per treatment. Error bars represent the 95% confidence interval of the mean for representative data from at least two independent experiments. P values represent colocalization of the 10mer compared with each polymer at indicated time point. *P < 0.001. (F) Polymer/Rab5, Rab7, or LAMP-2 colocalization was assessed in Raji/DC-SIGN cells either transfected with mRFP or Rab 5/7 mRFP or stained for LAMP-2 after treatment with polymer for 30 min at 37 °C. [Scale bars: 100 nm (A) and 5 μm (B, D, and F).]

Fig. 4.

Comparison of trafficking of particulate and soluble antigens to HIV. (A) Soluble and particulate fluorophore-labeled glycopolymers (green; 10 μM mannose) were added to Raji/DC-SIGN cells at 37 °C for 30 min. Fluorescence quenching by Trypan blue (2-min exposure) was monitored via confocal microscopy. Arrowheads indicate intracellular fluorescence quenched by Trypan blue. (Scale bars, 10 μm.) (B) Polymer fluorescence before and after Trypan blue treatment was measured in ImageJ (normalized to untreated). Error bars represent the SEM of at least two experiments for n > 50 cells. (C and D) Polymers or fluorescently labeled HIV-1 VLPs (green) were added at 37 °C to Raji/DC-SIGN cells transfected with CD81-mCherry (C) or moDCs stained for endogenous CD81 (D). Colocalization with CD81 was monitored by confocal microscopy. Arrowheads indicate regions of colocalization. (E) Quantitation of colocalization in moDCs in D, utilizing the Mander’s coefficient of antigen fluorescence overlapping with CD81. Error bars represent the SEM for at least two experiments for n > 35 cells. ns, not significant. (F) Polymer (green) and HIV (red) were coincubated in moDCs to assess colocalization. The plasma membrane was stained with wheat germ agglutinin (blue). (G) The overlap of polymer and HIV VLP fluorescence intensity (from F) was compared over the region indicated by a white line. (Scale bars, 5 μm.) *P < 0.01; ***P < 0.0001. A.U., arbitrary units.