Transcytosis Assay for Transport of Glycosphingolipids across MDCK-II Cells

Absorption and secretion of peptide and protein cargoes across single-cell thick mucosal and endothelial barriers occurs by active endocytic and vesicular trafficking that connects one side of the epithelial or endothelial cell (the lumen) with the other (the serosa or blood). Assays that assess this pathway must robustly control for non-specific and passive solute flux through weak or damaged intercellular junctions that seal the epithelial or endothelial cells together. Here we describe an in vitro cell culture Transwell assay for transcytosis of therapeutic peptides linked covalently to various species of the glycosphingolipid GM1. We recently used this assay to develop technology that harnesses endogenous mechanism of lipid sorting across epithelial cell barriers to enable oral delivery of peptide and protein therapeutics.

Transcytosis is an active (ATP-driven) process. In some cases, transport of large molecules across tight epithelial and endothelial barriers can occur by passive diffusion around the cell through intercellular tight junctions (Fung et al., 2018). But aside from this small level of paracellular leakiness, physiologic meaningful paracellular transport of large molecules around cells occurs only when intercellular tight junctions are dismantled-usually under pathologic conditions. Non-specific and low levels of transcellular transport for large solutes may also occur by fluid-phase endocytosis and miss-sorting of cargo into transcytotic vesicles rather than into the late endosome/lysosome pathway, which receives the bulk of cargo internalized by non-receptor mediated mechanisms. Thus, any assay for transcytosis across healthy epithelial or endothelial barriers in vitro or in vivo needs to control for these confounding non-specific pathways.
We have recently discovered that the endogenous sorting of glycosphingolipids across epithelial barriers can be harnessed for oral delivery of therapeutic peptides in vivo (Garcia-Castillo et al., 2018). Here we describe the in vitro assay using cultures polarized epithelial cells grown on transwell filters that we used to develop the technology that co-opts these endogenous mechanisms of lipid sorting for this purpose. The test compounds used to harness glycolipid sorting are modified to contain a biotin residue to allow for biochemical capture and a fluorophore for detection. Transwell inserts are composed of an inner chamber made with a permeable polycarbonate membrane support where the cells are seeded ( Figure  1). After 3-7 days the epithelial cells assemble into a single cell thick monolayer with sealed tight junctions among cells. As such, the inner chamber of the transwell is exposed to the apical membrane of the epithelial monolayer and models the lumenal surface and can hold 200 μl of solution. The outer chamber-the chamber below the permeable support-is exposed to the basolateral surface of the monolayer and models the serosal surface of the epithelial barrier. This chamber holds 1 ml of basolateral solution. The assembly of the monolayer into a "tight" epithelial barrier with sealed functional tight junctions is routinely measured by passive resistance to small ion transport (TER-Transepithelial Resistance) measured by standard direct current electrophysiology or using the alternate current based machine EVOM. Test compounds are added to the apical chamber, and the assay is allowed to proceed to allow for transport across the barrier, where the basolateral chamber is then sampled for the transcytosed analyte using streptavidin-coated beads ( Figure 2).
Mid-picomolar sensitivity can be achieved with this assay. Our basic approach can be used to measure transport by other molecules that enter the transcytotic pathway-such as for IgG that binds to the Fcγ-receptor FcRn that traffics in both directions across polarized epithelial monolayers (Nelms et al., 2017). The strength of our assay is the ability to directly, quantitatively, and sensitively measure the amount of compound transcytosed. Non-specific paracellular leak between cells versus real transport through cells is controlled in several ways, such as by using an analyte that cannot engage the transcytotic pathway (in our case the reporter peptide lacking fusion to the glycolipid carrier, or by fusion to a glycolipid with ceramide structure that cannot enter the transcytotic pathway), or by inhibiting endocytosis or transcytosis via a 4 °C temperature block, by chemical inhibition of endocytosis using Dyngo, and by siRNA knockdown of genes responsible for transcytosis as described

Materials and Reagents
Notes:

a.
Material and reagents are stored as per the manufacturer's recommendation.

b.
Fluorescent peptides and glycosphingolipid-peptide fusions are used as described in Garcia-Castillo et al., 2018. The assay requires the presence of both a fluorophore and a biotin covalently attached to compound you want to measure (see Figure 3). Peptide fusions containing biotin and fluorophore attached to ganglioside GM1 were synthesized in-house using specialized methods described in (Garcia-Castillo et al., 2018). Precursor peptides were synthesized by New England Peptide (Gardner, MA, USA), and gangliosides were purchased from Dr. Sandro Sonnino (U. Milan, Italy). However, in theory, any macromolecule (peptide, protein, nucleic acid or chemical) containing both a fluorophore and biotin will work in this assay. An example compound is shown below. Here, lower-cased amino acids are depicting D-isomers. Since glycine (shown as upper-cased G) does not have chirality, the peptide essentially contains no L-amino acids that can be susceptible to degradation in vivo.

a.
Aspirate MDCK-II growth media from the stock cell culture flask used for passaging.

b.
For a T-75 flask of MDCK-II cells, add 5-10 ml of warmed, sterile PBS. Gently tilt/rotate flask to wash the cells. Remove PBS.

c.
For a T-75 flask, add 3 ml of 0.25% Trypsin-EDTA. Tilt/rotate flask to make sure cell surface is evenly coated.

d.
Return to a 37 °C incubator with 5% CO 2 under humidified conditions for 5-10 min.

e.
Observe under a light microscope with phase contrast. If cells are properly detached, they will appear round and floating in suspension.

f.
Collect cell suspension in conical tube and pipet cells 6-8 times up and down. No clumps should be observed. Cell suspension should be a homogeneous mixture.

2.
Add MDCK-II cells at a density of 200,000 cells in 200 μl in complete growth media (see Recipes) to the apical side of a Transwell® insert (12-well 0.4 μm pore size Transwell®−65 mm). In addition, 1 ml complete growth media is added to the basolateral chamber.

b.
Each condition is tested as biological duplicates (i.e., 2 Transwell® inserts per condition). Experiments must always include untreated and reporter-peptide controls.

3.
Incubate the plate for 2-3 days to allow for polarization at 37 °C with 5% CO 2 under humidified conditions.

1.
Centrifuge the tube containing lyophilized compound at 13,000 rpm (~13,800 × g) for 3 min to pellet the material to the bottom of tube.

a.
Sonication is done in a bath sonicator (VWR, AquaSonic, model 50T

b.
At this point, this stock solution will be used to make the final dilution into "Apical" media for adding onto cells.

4.
To prepare this Apical test solution, dilute the appropriate volume of stock into Apical Solution (see Recipes) to reach a final concentration of 0.1 μM.
Note: The goal is to have a final 1:1 ratio of lipid to dfBSA.

1.
Check electrical resistance of MDCK-II Transwell® inserts after 2-3 days using EVOM Epithelial Voltohmmeter to measure integrity of tight junctions.
Note: Acceptable electrical resistances of MDCK-II monolayers used for transcytosis experiments is in the 250Ω-300Ω range.

4.
Allow cells to equilibrate for 15 min in a 37 °C/5% CO 2 cell culture incubator.

7.
To quantify transcytosed peptide and lipid-peptide fusions, collect basolateral media in pre-labeled Eppendorf tubes and proceed to streptavidin pull-down assay.

8.
To calculate the apparent permeability coefficient (PAPP, cm/sec), apical chamber solution is also collected in pre-labeled Eppendorf tubes. A standard curve ranging from 0 nM to 200 nM is used to interpolate concentration of peptide or lipid-peptide fusion in apical chamber after 3 h.

Wash Streptavidin magnetic beads 3 times with TBS-T.
Notes:

a.
Beads are never vortexed. They are brought into solution in stock tube by gentle inverting.

b.
Ten microliter beads are needed per sample.
Example: For 12 samples, place 120 μl Streptavidin beads in an Eppendorf tube and wash 3 times each with 1 ml TBS-T using magnetic rack.

3.
Incubate basolateral samples with streptavidin beads overnight at 4 °C with headover rotation covered in foil.

1.
Collect the beads using a magnetic rack and wash 3 times with TBS-T.
Note: Each wash is done with 1 ml TBS-T and beads mixed by inverting.

2.
Bound peptide or lipid-peptide fusions are eluted from beads by addition of 220 μl elution buffer (See Recipes) and boiling. Notes:

a.
Invert the beads to ensure that they are in solution.

b.
After beads are in solution, boil for 2 min at 65 °C in a heat block.

3.
After boiling, collect the beads using a magnetic rack.

5.
Fluorescence is read on a TECAN Spark microplate reader for Alexa-488 channel and against a standard curve for each compound.

c.
A basolateral standard curve for each compound being tested is made in elution buffer ranging from 0 pM to 1,000 pM.
i. Make 1 μM solutions of peptide and lipid-peptide stocks in elution buffer.

Data analysis
GraphPad Prism software is used to calculate the concentration of lipid-peptide fusion. As detailed above, a standard curve for each test compound is used to relate fluorescence intensity to a known concentration of peptide or lipid-peptide fusion. Absorbance at 495 nm is initially used to calculate stock concentrations. For detailed instructions on graphing and interpolating standard curves, the reader is referred to the GraphPad software website.
To calculate the apparent permeability coefficient (PAPP, cm/sec) a standard curve ranging from 0 nM to 200 nM is used to interpolate concentration of peptide or lipid-peptide solution remaining in the apical chamber after a 3 h continuous incubation.