General trends in the calnexin-dependent expression and pharmacological rescue of clinical CFTR variants
- The James Tarpo Jr. and Margaret Tarpo Department of Chemistry, Purdue University, United States
- Department of Chemistry, Vanderbilt University, United States
- Program in Chemical and Physical Biology, Vanderbilt University, United States
- Department of Pediatrics, Emory University School of Medicine, United States
- Department of Chemistry, Indiana University, United States
- Institute for Drug Discovery, Leipzig University, Germany
- Department of Biological Sciences, Vanderbilt University, United States
Figures
Figure 1 with 4 supplements
Surface immunostaining profiles of cystic fibrosis (CF) variant libraries.
Flow cytometry was used to characterize the distribution of CFTR surface immunostaining intensities among recombinant CANX KO and parental HEK293T cells expressing a pool of 232 CF variants. (A) A histogram depicts the distribution of CFTR surface immunostaining intensities among recombinant CANX KO (orange) and HEK293T (blue) cells treated with vehicle (DMSO). The mean fluorescence intensities of parental HEK293T cells stably expressing WT or ΔF508 CFTR are shown for reference. (B) A histogram depicts the distribution of CFTR surface immunostaining intensities among recombinant parental HEK293T cells expressing the CF variant library treated with vehicle (light blue) or with 3 μM VX-661 + 3 μM VX-445 (dark blue). The mean fluorescence intensities of parental HEK293T cells stably expressing WT or ΔF508 CFTR are shown for reference. (C) A histogram depicts the distribution of CFTR surface immunostaining intensities among recombinant CANX KO cells expressing the CF variant library treated with vehicle (orange) or with 3 μM VX-661 + 3 μM VX-445 (red). The mean fluorescence intensities of CANX KO cells stably expressing WT or ΔF508 CFTR are shown for reference.
-
Figure 1—source data 1
PDF containing the original western blot with band indicated for CRISPR KO validation.
- https://cdn.elifesciences.org/articles/107180/elife-107180-fig1-data1-v1.zip
-
Figure 1—source data 2
Original image of CANX KO validation western blot.
- https://cdn.elifesciences.org/articles/107180/elife-107180-fig1-data2-v1.zip
Figure 1—figure supplement 1
Comparison of deep mutational scanning measurements derived from first and second generation cystic fibrosis (CF) variant libraries.
A subset of surface immunostaining intensities for 129 CF variants derived from deep mutational scans using the library of 232 variants described herein is plotted against the corresponding 129 values that were previously reported from a deep mutational scan of the first generation library described in McKee et al., 2023. Values represent the average of three biological replicates, and error bars represent the standard deviation. A linear best fit is shown for reference (Pearson’s R2 = 0.9975).
Figure 1—figure supplement 2
Validation of Cas9-mediated CANX knockout cells.
(A) Western blot compares the levels of CANX protein expression across a series of clones isolated following transfection with a Cas9 RNP loaded with anti-CANX guide RNA. The lanes containing lysate from the parental cell line and the knockout line used for the studies detailed herein are indicated. An anti-cyclophilin B stain loading control is shown for reference. (B) The alignment of a Sanger sequencing read to the CANX gene sequence demonstrated that the chosen knockout clone indicated in panel A contains a deletion within exon 8.
Figure 1—figure supplement 3
Surface and internal expression of N-glycan knockouts in parental and CANX KO backgrounds.
A series of HA-tagged CFTR variants was transiently expressed in parental HEK293T (open) and CANX KO (patterned) HEK 293T cells. Surface CFTR on intact cells was first immunostained with a Dylight-550-conjugated anti-HA antibody, then fixed and permeabilized prior to immunostaining of the intracellular CFTR with an Alexa Fluor 647-conjugated anti-HA antibody. Relative surface and intracellular immunostaining intensities were then measured by flow cytometry. A bar graph depicts the relative surface immunostaining intensity (white) and relative intracellular immunostaining intensity (gray) CFTR immunostaining for each CFTR variant. Values represent the average of three biological replicates, and error bars represent the standard deviations. These data show that both WT and ΔF508 CFTR exhibit impaired expression in the absence of CANX. This expression defect persists for single mutants of either glycosylation site (N894 or N900). However, the CANX dependence of expression is lost for the double mutant that lacks N-linked glycosylation sites. Consistent with the findings in Rosser et al., 2008, these trends validate that CFTR exhibits lower expression in CANX knockout cells. Our results also suggest this expression defect depends upon its recognition of both N-linked glycans.
Figure 1—figure supplement 4
Relative surface immunostaining histograms of parental (A) and CANX KO (B) cell lines with approximate binning gates indicated as dashed lines.
Gates were redrawn for each experiment, ensuring that each bin contained approximately 25% of cellular events. Actual surface immunostaining histograms and precise gates can be found in the Mendeley directory for each experiment.
Figure 2 with 1 supplement
Influence of calnexin on cystic fibrosis (CF) variant plasma membrane expression (PME).
The difference PME in CANX KO cells relative to the parental HEK293T cells is shown for 232 CF variants across three biological replicates. (A) The change in expression for each variant is plotted against its position within the CFTR sequence. The boundaries of the six CFTR domains highlighted for reference. (B) A box and whisker plot depicts the statistical distributions for the change in variant PME across variants within each subdomain. The upper and lower edges of the box reflect the 75th and 25th percentile values, and the upper and lower whiskers reflect the 90th and 10th percentile values, respectively. The lines within each box represent the median value and the squares within the box represent the average. ** denotes p < 0.01 for a Mann–Whitney U-test. (C) Values for the change in PME for each variant are projected onto their corresponding residues within a structural model of CFTR (5UAK).
Figure 2—figure supplement 1
Plasma membrane expression of cystic fibrosis (CF) variants in parental HEK293T cells and CANX knockout HEK293T cells.
Deep mutational scanning measurements of the surface immunostaining intensities of 232 CF variants in parental HEK 293T cells are plotted against the corresponding surface immunostaining intensities in CANX knockout cells. Values represent the average of three biological replicates, and error bars represent the standard deviations. A line of best fit is shown for reference (m = 0.26).
Figure 3
Influence of calnexin on the sensitivity of cystic fibrosis (CF) variants to VX-445.
Deep mutational scanning was used to quantitatively compare the effect of 3 μM VX-445 on the plasma membrane expression (PME) of 232 CF variants across three biological replicated in CANX KO cells relative to the corresponding parental HEK293T cell line. (A) A scatter plot depicts the change in the VX-445 response in CANX KO cells relative to parental cells against the corresponding PME in parental HEK293T cells for each CF variant. (B) A scatter plot depicts the change in the VX-445 response in CANX KO cells relative to parental cells against the position of the mutated residue for each CF variant. (C) The change in the VX-445 response in CANX KO cells relative to parental cells for each CF mutant is projected on to the structure of VX-445-bound ΔF508 CFTR (PDB 8EIG). The structure of VX-445 is shown in yellow for reference. Negative values indicate a weaker response and positive values indicate a greater response to VX-445. Transmembrane helices 10 and 11 and ICL4 are depicted in green.
Figure 4
VX-445-mediated suppression of conformational defects in NBD2.
Structural modeling was used to compare the conformational states of the apo and VX-445-bound active structures of WT CFTR and six rare cystic fibrosis (CF) variants (P5L, G85E, V232D, L1077P, W1098R, and N1303K). (A) The average Rosetta energy scores (± SEM) for the 100 lowest scoring models of the VX-445-bound state are plotted against those of the apo CFTR variant models. A reference line corresponding to no stabilization is shown for reference. All variants except for the non-responsive G85E fall below the line, which confirms VX-445 enhances stability. (B) The total ΔRMSD of the active conformation of NBD2 is shown for variants bound to VX-445. Red bars show increasing deviations from the native NBD2 conformation in the mutant models and blue bars how much VX-445 suppresses these conformational defects in NBD2. (C) Maps of the change in root mean squared deviation (RMSD) between G85E modeled with and without VX-445 show which structural regions are stabilized by VX-445. Structurally variable regions in the ensemble shown in red, while areas adopting a more ordered conformation are shown in blue. VX-445 appears to primarily suppress conformational defects within the NBD2 region of this variant. (D) Maps of the change in RMSD between L1077P modeled with and without VX-445 show which structural regions are stabilized by VX-445. Structurally variable regions in the ensemble are shown in red, while areas adopting a more ordered conformation are shown in blue. VX-445 appears to primarily suppress conformational defects within the NBD2 region of this variant. (E) Maps of the change in RMSD between N1303K modeled with and without VX-445 show that few structural regions are stabilized by VX-445 for N1303K, which responds poorly to VX-445 in vitro. Statistical significances were calculated using a non-parametric Wilcoxon signed-rank test compared to zero to determine if distributions changes were significantly different from zero, and p-values were depicted by *<0.05, **<0.01, and ****<0.0001. Error bars indicate the standard error of the mean.
Figure 5 with 1 supplement
Influence of calnexin on the sensitivity of cystic fibrosis (CF) variants to VX-445 + VX-661.
Deep mutational scanning was used to quantitatively compare the effect of 3 μM VX-445 + 3 μM VX-661 on the plasma membrane expression (PME) of 232 CF variants across three biological replicates in CANX KO cells relative to the corresponding parental HEK293T cell line. (A) A scatter plot depicts the change in the VX-445 + VX-661 response in CANX KO cells relative to parental cells against the corresponding PME in parental HEK293T cells for each CF variant. (B) A scatter plot depicts the change in the VX-445 + VX-661 response in CANX KO cells relative to parental cells against the position of the mutated residue for each CF variant. (C) The change in the VX-445 + VX-661 response in CANX KO cells relative to parental cells for each CF mutant is projected onto the WT CFTR structure (PDB 5UAK). Negative values indicate a weaker response and positive values indicate a greater response to VX-445 + VX-661.
Figure 5—figure supplement 1
Influence of correctors on the plasma membrane expression in parental and CANX knockout cells.
Deep mutational scanning measurements of the plasma membrane expression of cystic fibrosis (CF) variants in CANX knockout cells in the presence of (A) 3 μM VX-445, (B) 3 μM VX-661, or (C) 3 μM VX-445 + 3 μM VX-661 are plotted against the corresponding variant measurements under identical conditions in the parental HEK293T cell line. Values represent the average of three biological replicates, and error bars represent the standard deviation. Trends are generally linear, and there are few variants that stray from the trend, which suggests CF variants that respond in one cell line have a similar response in the other.
Figure 6 with 1 supplement
Interactome of cystic fibrosis (CF) variants in CANX KO cells.
Proteomic mass spectrometry was used to compare the protein–protein interactions formed by CF variants in the context of CANX KO cells relative to the parental cell line. (A) A cartoon depicts the workflow for multiplexed affinity purification-mass spectrometry (AP-MS) experiments. (B) Violin plots depict the log2 fold change in protein abundances of various classes of interactors that associate with CF variants in CANX KO cells relative to their corresponding abundance in the parental cell line. Each data point represents individual protein interactors that are grouped according to their pathways. (C–E) Violin plots depict the log2 fold change in protein abundances of various classes of interactors that associate with ΔF508 in CANX KO cells treated with various correctors relative to their corresponding abundance in the parental cell line under identical treatment conditions. Each data point represents individual protein interactors that are grouped according to their pathways. Statistical significance is denoted as follows: Red asterisks indicate a significant deviation from zero, assessed using a one-sample t-test and Wilcoxon test against a hypothetical value of 0. Black asterisks denote significant differences from wild-type (WT) levels, determined using a repeated measures one-way ANOVA with Geisser–Greenhouse correction (*p < 0.05, **p < 0.01, ***p < 0.001, ****p <0.0001).
Figure 6—figure supplement 1
Comparative interactome profiling of cystic fibrosis (CF) variants in parental and CANX knockout HEK293T cells.
Proteomic mass spectrometry was used to compare the protein–protein interactions formed by CF variants in the context of CANX knockout cells relative to the parental cell line. (A) Heat maps depict the log2 fold change in protein abundances of various classes of interactors that associate with each indicated CF variant in CANX KO cells relative to their corresponding abundance in the parental HEK293T cell line. (B) Heat maps depict the log2 fold change in protein abundances of various classes of interactors that associate with ΔF508 in CANX KO cells treated with 3 μM of the indicated corrector(s) relative to their corresponding abundance in the parental HEK293T cell line under identical conditions. Protein interactors that are grouped according to their pathways.
Figure 7 with 2 supplements
Influence of calnexin on the functional rescue of cystic fibrosis (CF) variants.
The functional properties of CF variants are compared in various cells that feature endogenous CANX expression or deficient CANX expression under various experimental conditions. (A) Flow cytometry was used to monitor CFTR-mediated quenching of a halide-sensitive YFP (hYFP) relative to an mKate expression marker upon activation of stably expressed CFTR variants with 25 μM forskolin (Fsk) and 50 μM genistein in parental HEK293T and CANX KO HEK293T cells. Representative flow cytometry measurements of single-cell hYFP: mKate ratios are plotted over time following CFTR activation for recombinant cells stably expressing a series of CF variants. Fitted curves for single exponential fit decay are shown for reference. Bar graphs depict the fitted half-lives of the hYFP quenching reactions among cells expressing each indicated CF variant in parental HEK293T cells (CANX+, open bars) and CANX KO cells (CANX−, dashed bars) treated with (B) DMSO or with (C) 3 μM VX-661, 3 μM VX-445, and 10 μM VX-770. Values represent the mean ± SEM (n = 3). Fischer rat thyroid (FRT) monolayers transiently expressing (D) ΔF508 or (E) WT CFTR were cultured 5 days on permeable transwells and chronically treated with CFTR modulators (72 hr, 5 μM), vehicle control (DMSO), and/or various siRNA (72 hr, 100 nM). Bar plots depict average short-circuit currents (ΔISC) following acute application of 10 μM Fsk, 5 μM VX-770, and/or 10 μM of the CFTR-specific inhibitor-172 (Inh172). (F) A bar graph depicts the total functional activation of ΔF508 CFTR that is achieved under various conditions upon stimulation with Fsk and VX-770 relative to WT stimulated with Fsk. Values represent mean ± SEM (n = 4). Asterisks annotate statistical evaluation of Fsk ± VX-770 values compared to non-specific (NS, red) or CANX-specific (blue) siRNA. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001 (two-way ANOVA). Currents from non-treated cells (NT) are shown for reference.
Figure 7—figure supplement 1
Impact of a CFTR-specific inhibitor on observed hYFP quenching kinetics in recombinant cells expressing cystic fibrosis (CF) variants.
The functional properties of CF variants are compared in various cells that feature endogenous CANX expression or deficient CANX expression under various experimental conditions. Bar graphs depict the fitted half-lives of the hYFP quenching reactions among cells expressing each indicated CF variant in parental HEK293T cells treated with either vehicle (open bars) or with 10 µM of the CFTR-specific inhibitor-172 (Inh172). A slowing of quenching was observed in the context of recombinant cell lines expressing each variant except T1036N. Though it is unclear why inhibition was not observed in this cell line, it is possible that this variant may have limited affinity for the inhibitor. Nevertheless, this cell line exhibits an intermediate quenching rate that lies between wild-type and ΔF508, which suggests the observed quenching is unlikely to be non-specific.
Figure 7—figure supplement 2
Validation of siRNA-mediated CANX knockdowns in Fischer rat thyroid (FRT) cells.
Quantitative reverse-transcriptase PCR was used to measure the relative abundance of the CFTR and CANX transcripts in FRT cells following transfection with either a non-specific or CANX-specific siRNA. ΔCT values for the target transcripts were normalized relative to the corresponding values of actin B transcript. Statistical significance was calculated using a two way T-Test (n=4) assuming unequal variance where * indicates a p-value of <0.05. Error bars indicate standard error of the mean.
Author response image 1
Tables
Key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Antibody | Anti-Hemagglutinin (Mouse Monoclonal) | Invitrogen | Cat# 26183-D550; RRID:AB_2533052 | 2–2.2.14, DyLight 550 conjugate, mouse monoclonal, 1:100 |
| Antibody | Anti-Hemagglutinin (Mouse Monoclonal) | Invitrogen | Cat# 740005MP647; RRID:AB_3666184 | 2–2.2.14, Alexa Fluor 647 conjugate, mouse monoclonal, 1:100 |
| Antibody | Human CFTR C-Terminus Antibody | Bio-Techne | Cat# MAB25031; RRID:AB_2260673 | Mouse Monoclonal, 6 mg antibody/ml of beads |
| Strain, strain background (Escherichia coli) | NEB10β | New England Biolabs | Cat# C3020K | Electrocompetent |
| Strain, strain background (Escherichia coli) | NEBDH5α | New England Biolabs | Cat# C2987H | |
| Chemical compound, drug | iodoacetamide | Sigma-Aldrich | Cat# I6125 | |
| Commercial assay, kit | Fugene 6 | Promega | Cat# E2691 | |
| Peptide, recombinant protein | 2X Kappa HiFi HotStart Ready Mix | Roche Diagnostics | Cat# KK2602 | |
| Chemical compound, drug | VX-661 | APExBIO | Cat# A2664 | |
| Chemical compound, drug | VX-445 | Selleckchem | Cat# S8851 | |
| Chemical compound, drug | VX-770 | Selleckchem | Cat# S1144 | |
| Chemical compound, drug | Forskolin | Selleckchem | Cat# S2449 | |
| Chemical compound, drug | Genistein | Selleckchem | Cat# S1342 | |
| Chemical compound, drug | Amiloride | Sigma-Aldrich | Cat# A7410 | |
| Chemical compound, drug | CFTRinh-172 | Selleckchem | Cat# S7139 | |
| Chemical compound, drug | Roche cOmplete Protease Inhibitor Cocktail | Roche Diagnostics | Cat# 11697498001 | |
| Commercial assay or kit | Lipofectamine CRISPRMAX Cas9 Transfection Reagent | Invitrogen | Cat# CMAX00008 | |
| Commercial assay or kit | ZymoPURE Plasmid Miniprep Kit | Zymo Research | Cat# D4211 | |
| Commercial assay or kit | Zymopure Endotoxin-Free Midiprep Kit | Zymo Research | Cat# D4200 | |
| Commercial assay or kit | ZR-96 DNA Clean-Up Kit (Shallow Well) | Zymo Research | Cat# D4018 | |
| Commercial assay or kit | Select-a-Size DNA Clean & Concentrator | Zymo Research | Cat# D4080 | |
| Commercial assay or kit | DNeasy Blood and Tissue Kits for DNA Isolation | QIAGEN | Cat# 69506 | |
| Commercial assay or kit | Detergent-Compatible Bradford Assay | Pierce Biotechnology | Cat# 23246 | |
| Transfected construct (H. sapiens) | Tet-Bxb1-BFP HEK293T | Laboratory of Doug Fowler | Clone 37 described in Matreyek et al., 2017 | |
| Cell line (H. sapiens) | HEK293T for DMS CANX KO | This paper | Clone 37 described in Matreyek et al., 2017 | |
| Cell line (H. sapiens) | HEK293T | ATCC | Cat# CRL-11268; RRID:CVCL_0063 | |
| Cell line (R. norvegicus) | Fischer rat thyroid (FRT) cells | Michael Welsh | ||
| Sequence-based reagent | Custom CFTR primer pool | Agilent Technologies Inc | See Supplementary file 3. Supplemental Primer List Related to Variant and Deep Sequencing Library Production | |
| Sequence-based reagent | CANX siRNA | QIAGEN | #SI02666335 | |
| Sequence-based reagent | NS siRNA | QIAGEN | #1027310 | |
| Software, algorithm | Origin 2019 | OriginLab | RRID:SCR_014212 | |
| Software, algorithm | Image Studio V. 5.2 | LI-COR Biosciences | RRID:SCR_015795 | |
| Software, algorithm | Structure Refinement from Cryo-EM maps | Wang et al. | ||
| Software, algorithm | RosettaLigand | Lemmon & Meiler | ||
| Software, algorithm | BCL ConformerGenerator | Mendenhall et al. | ||
| Software, algorithm | RosettaCM | Song et al. | ||
| Software, algorithm | RosettaMembrane | Barth et al. | ||
| Software, algorithm | OCTOPUS | Viklund & Elofsson | ||
| Software, algorithm | Proteome Discoverer 2.4 | Thermo Fisher | RRID:SCR_014477 |
Additional files
-
MDAR checklist
- https://cdn.elifesciences.org/articles/107180/elife-107180-mdarchecklist1-v1.pdf
-
Supplementary file 1
CFTR variant read counts.
- https://cdn.elifesciences.org/articles/107180/elife-107180-supp1-v1.xlsx
-
Supplementary file 2
CFTR expression table.
- https://cdn.elifesciences.org/articles/107180/elife-107180-supp2-v1.xlsx
-
Supplementary file 3
CFTR primer list.
- https://cdn.elifesciences.org/articles/107180/elife-107180-supp3-v1.xlsx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
General trends in the calnexin-dependent expression and pharmacological rescue of clinical CFTR variants
eLife 14:RP107180.
https://doi.org/10.7554/eLife.107180.3
