The recycling and transcytotic pathways for IgG transport by FcRn are distinct and display an inherent polarity

Abstract

The Fc receptor FcRn traffics immunoglobulin G (IgG) in both directions across polarized epithelial cells that line mucosal surfaces, contributing to host defense. We show that FcRn traffics IgG from either apical or basolateral membranes into the recycling endosome (RE), after which the actin motor myosin Vb and the GTPase Rab25 regulate a sorting step that specifies transcytosis without affecting recycling. Another regulatory component of the RE, Rab11a, is dispensable for transcytosis, but regulates recycling to the basolateral membrane only. None of these proteins affect FcRn trafficking away from lysosomes. Thus, FcRn transcytotic and recycling sorting steps are distinct. These results are consistent with a single structurally and functionally heterogeneous RE compartment that traffics FcRn to both cell surfaces while discriminating between recycling and transcytosis pathways polarized in their direction of transport.

Figure 1.

The fusion protein FcRn-EGFP transports IgG normally in polarized MDCK cells. (a) FcRn-EGFP–mediated transcytosis of NIP-IgG (shaded bars) or mutant NIP-IgG-IHH (open bars) across MDCK cell monolayers measured by NIP-specific ELISA, as described in Materials and methods. Results shown are mean ± SD (n = 4; ANOVA of IgG vs. IHH, P < 0.0002). (b) FcRn-EGFP mediates endocytosis of Alexa Fluor 647–IgG but not Alexa Fluor 647–IgG–IHH from either apical and basolateral surfaces of polarized MDCK cell monolayers, as assessed by confocal microscopy. Intensity scale bars shown demonstrate that images between IgG and IHH were equally contrasted. Bar, 5 µm. (c) IgG is present in FcRn-containing compartments. Cells were incubated with Alexa Fluor 647–IgG as described in panel b and imaged by 3D confocal microscopy. Compartments occupied by both proteins were quantified by first identifying FcRn-EGFP– and IgG-containing objects using an automatic threshold-finding algorithm, and analyzing the colocalization between FcRn-EGFP and IgG using the Manders colocalization coefficient (termed M_{target protein}, see Materials and methods). The distribution of colocalization M_{target protein} for IgG (Target protein) in each FcRn-identified object (Compartment) is shown (data were collected from 40–50 cells of three independent experiments). The horizontal bars indicate a mean M_{target protein} of 0.95 and 0.98 for apically applied IgG (ap IgG) and basolaterally applied IgG (bs IgG), respectively. (d) FcRn-EGFP recycles NIP-IgG (shaded bars) or NIP-IgG-IHH (open bars) after endocytosis from either the apical or basolateral surfaces. In brief, as described in Materials and methods, FcRn-EGFP cell monolayers were incubated with NIP-IgG or the IHH mutant at acidic pH to allow uptake from either cell surface for 1 h. IgG bound at the plasma membrane was removed by washing cells at neutral pH at 4°C. The monolayers were then returned to 37°C and chased for an additional hour in fresh buffer lacking IgG. Results shown are mean ± SD (n = 4, ANOVA for IgG vs. IHH, P < 0.03). (e) Time course of FcRn-EGFP–mediated NIP-IgG basolateral-to-apical transcytosis, recycling, and cell-associated IgG. Individual monolayers were loaded with NIP-IgG or the IHH mutant for 1 h, and surface IgG was stripped at 4°C. The monolayers were then returned to fresh buffer at 37°C and chased for the indicated times to measure recycling (closed triangles), transcytosis (squares), and total IgG cell-associated content (diamonds). Recycling of NIP-IgG-IHH to the basolateral membrane is also shown as control (open triangles). A representative experiment where each point represents the mean of three separate measurements is shown. Error bars indicate the SD of three independent experiments. (f) FcRn is mostly excluded from the lysosomal compartment. Polarized cells were fixed and immunostained for LAMP1. The distribution of colocalization M_{target protein} for FcRn (Target Protein) in each LAMP1-identified object (Compartment) is shown. The horizontal bar indicates mean M_{target protein} = 0.02.

Figure 2.

FcRn transports IgG into the common endosome. Compartments from 15–40 FcRn-EGFP polarized cells from 3D confocal images were analyzed for content of indicated markers as described in Fig. 1 and the Materials and methods section. Representative images are shown in panels a, c, and e. Panels b, d, and f show quantification of each analysis, with the number of objects studied indicated. Horizontal bars in the graphs indicate mean M_{target protein}. (a and b) Compartments masked for FcRn-EGFP or Tf-R show a fraction of endosomes containing both molecules. (c and d) Compartments masked for FcRn-EGFP show a fraction containing IgG internalized from both apical and basolateral membranes. Alexa Fluor 568–IgG and Alexa Fluor 647–IgG were applied to apical or basolateral membranes for 1 h, respectively. (e and f) Compartments masked for FcRn-EGFP show a fraction containing IgG internalized from the apical membrane and Tf internalized from the basolateral membrane. Alexa Fluor 568–IgG and Alexa Fluor 647–Tf were applied to apical or basolateral membranes for 1 h, respectively. In a, c, and e, the panels to the right show enlarged views of the boxed portions on the left. Circles highlight the same object in each panel. Bars, 10 µm.

Figure 3.

Overexpression of MyoVb tail blocks FcRn-dependent IgG transcytosis but not recycling. All experiments were performed in MDCK cells stably expressing a tetracycline-inducible EGFP–MyoVb tail and FcRn-HA. (a) Doxycycline induces expression of EGFP–MyoVb tail in nonpolarized (lanes 1 and 3) and polarized cells (lanes 2 and 4), as measured by SDS-PAGE and immunoblotting using monoclonal antibodies against EGFP. The bottom shows corresponding actin immunoblotting. **, cross-reacting protein band insensitive to doxycycline treatment. Numbers to the left indicate kD. (b) Doxycycline-induced expression of EGFP–MyoVb tail collapses Rab11-containing endosomes into a dense apical structure consistent with the ARE. The induced structure does not contain Tf-R. Polarized cells treated with doxycycline were fixed and immunostained for Rab11 and Tf-R. Bar, 10 µm. (c) Doxycycline-induced expression of EGFP–MyoVb tail inhibits transcytosis of NIP-IgG (shaded bars) in both directions as measured by ELISA; results are mean ± SD (n = 5). NIP-IgG-IHH (open bars) was used as control for nonspecific transport. *, P < 0.001. (d) Doxycycline-induced expression of EGFP–MyoVb tail has no detectable effect on recycling of NIP-IgG (shaded bars) to either apical or basolateral cell surfaces as measured by ELISA; results are mean ± SD (n = 3). NIP-IgG-IHH (open bars) is used as a control for nonspecific transport. (e) Doxycycline-induced expression of EGFP–MyoVb tail has no detectable effect on endocytosis of Nip-IgG (shaded bars) from either apical or basolateral membranes. Cells kept on ice (time 0) or incubated with NIP-IgG-IHH (open bars) provide control for nonspecific transport. A representative study with three independent measurements for each point is shown. Error bars indicate the SD of three independent experiments. (f) Doxycycline-induced expression of EGFP–MyoVb tail does not divert FcRn to the lysosomal compartment. Polarized cells treated or not treated with doxycycline were fixed and immunostained for FcRn (HA staining) and LAMP1. The distribution of colocalization of FcRn with LAMP1 was quantified using the Manders coefficient as described (see main text and Materials and methods). Horizontal bars in the graphs indicate that mean M_{target protein} equals 0.04 and 0.08 for +Dox and −Dox, respectively.

Figure 4.

FcRn transports IgG through the MyoVb tail–induced compartment. (a) FcRn partially localizes with the EGFP–MyoVb tail, as assessed in fixed cells stained for FcRn-HA. (b) Time course studies show that IgGs entering cells from apical or basolateral surfaces eventually accumulate in the EGFP–MyoVb tail–induced compartment. Alexa Fluor 568–IgG and Alexa Fluor 647–IgG were added at the apical and basolateral reservoirs, respectively; cells were incubated for 5, 15, 30, or 60 min, then fixed and imaged by confocal microscopy. Objects were defined based on the intensity of the EGFP–MyoVb tail; then, the colocalization of apically applied IgG (ap IgG), basolaterally applied IgG (bs IgG), or both IgGs was measured and quantified as described (see Materials and methods). In the micrographs, the panels on the right show enlarged views of the boxed portions on the left. The horizontal bars in the graphs indicate mean M_{target protein}. Bars, 10 µm.

Figure 5.

Rab25 regulates a sorting step that specifies transcytosis, whereas Rab11a regulates recycling in the basolateral pathway only. All experiments use doxycycline-induced shRNA against Rab25 (left panels) or Rab11a (right panels) in polarized FcRn-EGFP cells. (a) shRNA against Rab25 suppresses gene expression as assessed by RT-PCR. (b) shRNA against Rab11a suppresses gene expression as assessed by SDS-PAGE and immunoblotting. Numbers on the left indicate kD. (c and d) Gene suppression of Rab25 (c) or Rab11a (d) does not alter trafficking of FcRn–IgG complexes into the common RE. Alexa Fluor 568–IgG was applied apically and Alexa Fluor 647–Tf was applied basolaterally for 60 min; they were then fixed and imaged by confocal microscopy. Scatter plots show the degree of colocalization of IgG with Tf. The horizontal bars in the graphs indicate mean M_{target protein}. (e and f) Gene suppression of Rab25 (e) but not Rab11a (f) inhibits transcytosis of NIP-IgG (shaded bars) in both directions across the cell. NIP-IgG-IHH is used as control. Results are mean ± SD. (e, n = 6; f, n = 4). (g and h) Gene suppression of Rab11a inhibits NIP-IgG (shaded bars) recycling in the basolateral direction only. NIP-IgG-IHH (open bars) is used as control. Results are mean ± SD (n = 4). *, P < 0.001. (i and j) Doxycycline-induced silencing of Rab25 or Rab11a does not divert FcRn to the lysosomal compartment. Polarized cells treated or not treated with doxycycline were fixed and immunostained with LAMP1. The degree of colocalization of FcRn with LAMP1 was estimated using the Manders coefficient as described (see Materials and methods). Bar indicates mean M_{target protein}.