key: cord-0260506-xyotqw7j
authors: Lee, Rhianna E.; Lewis, Catherine A.; He, Lihua; Gallant, Samuel C.; Mascenik, Teresa M.; Dang, Hong; Morton, Lisa C.; Minges, John T.; Theile, Jonathan W.; Castle, Neil A.; Knowles, Michael R.; Kimple, Adam J.; Randell, Scott H.
title: Robust W1282X-CFTR rescue by a small molecule GSPT1 degrader
date: 2021-09-30
journal: bioRxiv
DOI: 10.1101/2021.09.28.462149
sha: 7c31b22d0e746e17447dc22ced9ee2478807a762
doc_id: 260506
cord_uid: xyotqw7j

With the approval of elexacaftor/tezacaftor/ivacaftor (trade name Trikafta), the vast majority of people with cystic fibrosis (CF) are eligible for CF transmembrane conductance regulator (CFTR) modulator therapy. Remaining individuals have premature termination codons or rare CFTR variants with limited treatment options. Although clinical modulator response can be reliably predicted using primary airway epithelial cells, primary cells carrying rare CFTR variants are scarce. To overcome this obstacle, these cells can be expanded by overexpression of mouse Bmi-1 and human TERT (hTERT). We therefore used this approach to develop two non-CF and three CF (F508del/F508del, F508del/S492F, W1282X/W1282X) nasal cell lines and two W1282X/W1282X bronchial cell lines. Bmi-1/hTERT cell lines recapitulated primary cell morphology and ion transport function. The F508del/F508del and F508del/S492F cell lines robustly responded to Trikafta, which was mirrored in the parent primary cells and the cell donors’ clinical response. CC-90009, a novel cereblon E3 ligase modulator targeting the GSPT1 protein, rescued ~20% of wildtype CFTR function in our panel of W1282X/W1282X cell lines and primary cells. Intriguingly, CC-90009 also diminished epithelial sodium channel function. These studies demonstrate that Bmi-1/hTERT cell lines faithfully mirror primary cell responses to CFTR modulators and illustrate novel therapeutic approaches for the W1282X CFTR variant.

The FDA set the precedent for such a change in 2017 when they expanded the use of ivacaftor to patient populations harboring one of twenty-three relatively rare CFTR variants (https://www.accessdata.fda.gov/scripts/cder/daf/index.cfm?event=overview.process&ApplNo=203188).

Although drug label expansions are common, this instance was particularly groundbreaking because the FDA based their decision purely on in vitro data rather than a clinical trial (9). This new paradigm relies on the fidelity of in vitro systems to accurately predict a clinical response.

Primary CF human bronchial epithelial cells (HBECs) obtained at the time of lung transplantation have served as the gold standard for assessing CFTR rescue in vitro (10) (11) (12) . However, the supply of CF explant lungs is limited, particularly for rare CFTR variants. HBECs can also be obtained by bronchial brushing, but the and F508del/S492F CFTR genotypes and comparing them to the parent primary cells, we assessed the fidelity of CFTR modulator responses and the percent of normal CFTR activity restored. We also compared the in vitro cell culture and in vivo cell donors' clinical responses to Trikafta.

We then used this approach to evaluate a potential modulator therapy for patients homozygous for W1282X, a nonsense mutation that generates a premature termination codon (PTC) in the CFTR transcript,

predisposing CFTR mRNA to nonsense-mediated decay (NMD) and absent or truncated protein. Without a targetable protein, modulator therapies are ineffective, leaving these patients without treatment options. Recent studies illustrate that the cereblon E3 ubiquitin ligase modulator, CC-90009, promotes PTC readthrough of nonsense mutations associated with inherited diseases (18). Using one nasal and two bronchial Bmi-1/hTERT conditions ( Figure 1G ). The Bmi-1/hTERT cell line vastly outgrew its parent nasal cells when cultured in EpiX media, whereas CRC culture promoted robust HNEC expansion regardless of Bmi-1/hTERT expression. The cell growth properties of all other nasal cell lines generated here can be found in the online supplements (Supplemental Figure 1A-D) . Notably, the Bmi-1/hTERT bronchial cell lines could be effectively expanded using BEGM as previously described (15) (Supplemental Figure 1E -F). The cell morphology and electrophysiology of nasal cells cultured in EpiX or by CRC are compared in the online supplements (Supplemental Figure 1G -K).

Nasal cell morphology and function is contingent upon in vitro culture conditions and media composition (24, 25) . To optimize the differentiation of our Bmi-1/hTERT nasal cell model, we differentiated P15 UNCNN2T cells in UNC air-liquid interface (ALI) (26) , Vertex ALI (27) , and Pneumacult ALI (STEMCELL Technologies) media (Supplemental Figure 2A -E). The bioelectric properties were dramatically different between the three media conditions, with basal short-circuit current (Isc) averaging 7.1 ± 11.9 μA/cm 2 , 9.6 ± 1.6 μA/cm 2 , and 27.7 ± 7.9

μA/cm 2 in UNC ALI, Vertex ALI, and Pneumacult ALI, respectively (mean ± SD). The apical addition of amiloride inhibited Isc by approximately 3%, 97%, and 29% in UNC ALI, Vertex ALI, and Pneumacult ALI, respectively. In vivo electrophysiology, as measured by nasal potential difference, is characterized by partial amiloride-sensitive Na+ transport (~50% basal Isc) (28). Thus, in vivo nasal electrophysiology is not perfectly modeled by any of the Figure 3G -J). The greatest CFTR response was again seen with Pneumacult ALI.

Thus, Pneumacult ALI differentiation medium was used for the remainder of the study for all nasal and bronchial

Bmi-1/hTERT cell lines unless otherwise indicated.

Bmi-1/hTERT nasal cell lines model primary HNEC morphology and function. Hematoxylin and eosin (H&E) and Alcian blue-periodic acid-Schiff (AB-PAS) staining revealed a pseudostratified mucociliary epithelium in P6 and P15 UNCNN2T cells that was morphologically similar to parent HNECs at P2 (Figure 2A ). These results were confirmed by whole-mount immunostaining, which illustrated the presence of both MUC5ACproducing goblet cells (green) and α-tubulin + ciliated cells (white) ( Figure 2B -D). From these data, we concluded that Bmi-1/hTERT nasal cell lines at mid-and late-passages produce a well-differentiated mucociliary epithelium that reflects primary cell morphology. Measurements with a 24-channel transepithelial current clamp amplifier (TECC-24) device demonstrated that parent HNEC electrophysiology was also recapitulated by the UNCNN2T cell line at mid-and late-passage, but with a significant increase in CFTR activity in P5 versus parent cells ( Figure   2E -G). From this, we concluded that Bmi-1/hTERT nasal cell lines model primary HNEC morphology and function for at least 15 passages. Representative histology and whole-mount immunostaining of all other nasal cell lines are shown in the online supplement (Supplemental Figure 4) . . We also tested a recently described drug combination (29) which combines the CFTR corrector VX-809 with two novel small molecule correctors, 3151 and 4172 (each at 5 µM). Here, we call this treatment "triple corrector combination," or "3C." UNCX4T cells were treated with VX-445/VX-661, 3C, or a vehicle control (DMSO) and were assayed for CFTR function using a TECC-24 device ( Figure 3A , C). VX-445/VX-661 and 3C rescued 23.0 ± 1.4% and 19.5 ± 4.8% of wildtype CFTR function, respectively (mean ± SD), determined by dividing the UNCX4T forskolin (FSK) response by the average response of non-CF nasal cell lines (46.5 ± 5.0 µA/cm 2 ) ( Figure 2F ; Supplemental Figure 1H ). These data align with preclinical studies of VX-445/VX-661 and clinical observations of Trikafta in F508del homozygous populations (30). VX-445/VX-661 and 3C treatment also significantly rescued CFTR ion transport in the paired parent nasal cells compared with a DMSO control ( Figure 3B , C), rescuing 27.3 ± 5.1% and 26.7 ± 3.9% of wildtype CFTR function, respectively (mean ± SD) ( Figure 3C ). These data suggest that the novel CFTR corrector combination, 3C, might be as effective as Trikafta in rescuing F508del-CFTR and could represent a therapeutic candidate for those who cannot tolerate or do not respond to Trikafta treatment.

Next, we assessed our F508del/S492F compound heterozygous cell line, UNCX3T, for its response to VX-445/VX-661, 3C, or a DMSO control ( Figure 3D , F). We found that UNCX3T responded well to VX-445/VX-661 treatment, recapitulating 22.3 ± 0.7% of wildtype CFTR function (mean ± SD) ( Figure 3F ). However, 3C treatment was less effective, producing 14.6 ± 1.5% of wildtype function (mean ± SD). These findings were recapitulated in the UNCX3T parent cells ( Figure 3E , F). Because 3C treatment produced less CFTR functional response in UNCX3T (one copy of F508del) than in UNCX4T (two copies of F508del), we posit that this modulator combination does not rescue S492F-CFTR as effectively as F508del-CFTR. Even so, the response to 3C falls well within the therapeutic window (i.e., 10% of wildtype CFTR function over baseline) and could serve as an alternative therapy. From these studies, we concluded that Bmi-1/hTERT nasal cell lines generated from CF donors accurately predict the primary cell response to FDA-approved and novel CFTR modulators. W1282X-CFTR is rescued by CC-90009, a novel cereblon modulator that promotes PTC readthrough.

Current therapies are not effective in rescuing the truncated protein generated by the W1282X-CFTR variant.

One proposed treatment strategy is to promote ribosomal readthrough of the PTC to generate full-length protein literature, we suspect that the primary mechanism for W1282X-CFTR rescue is enhanced ribosomal readthrough (18). CFTR protein level is low in primary airway epithelial cells and made even lower by NMD in cells carrying PTC variants, making it challenging to assess. Importantly, we did not observe CC-90009 to increase CFTR activity in F508del-or wildtype-CFTR primary HBECs; rather, it was slightly reduced (Supplemental Figure 6 ).

This supports the notion that CC-90009 specifically acts to promote PTC readthrough rather than by non-specific effects on cellular ion transport. We measured CFTR mRNA by qRT-PCR and found a significant increase in response to CC-90009 treatment ( Figure 5E ). This finding is consistent with the published observation that these too may prove therapeutic. Increasing airway hydration through ENaC inhibition has been proposed previously as a treatment strategy for CF (44). Future mechanistic work will be necessary to discern the link between CC-90009 and ENaC and may uncover novel pathways for ENaC ubiquitination.

In summary, we have shown that treatment with CC-90009 increases CFTR function and decreases

ENaC expression in W1282X patient-derived cell lines and primary airway epithelial cells. GSPT1 degradation by CC-90009 represents a new therapeutic strategy for treating nonsense mutations in the CFTR transcript.

The extension of CFTR modulators to 90% of people with CF represents a tremendous victory. However, developing targeted therapies for the remaining 10% will be challenging. This challenge is heightened by the limited availability of rare-genotype primary cells. For this reason, the field has turned to cell lines for drug discovery efforts. Though primary HBECs are the gold standard of CF disease modeling, they can only be expanded for ~10 population doublings using conventional growth methods (10) or ~25 doublings using newer CRC culture methods (12, 49) . HNECs also exhibit extended growth capacity using CRC technology (21-23). However, with extensive population doublings comes squamous transformation and a reduction in ion transport (12) . Here, we describe the development of Bmi-1/hTERT cell lines derived from primary airway epithelial cells that can be expanded for at least 30 population doublings (15 passages) while recapitulating primary cell morphology and CFTR function. The ability to amply expand and assay CFTR function in patient-derived cells represents a major advancement.

In this study, we directly compare the functional response of patient-derived Bmi-1/hTERT cell lines and the primary cells from which they were developed. Across five CF donors (three nasal and two bronchial) carrying a range of CFTR mutations and using both FDA-approved and novel CFTR modulators, we found that patientderived cell lines accurately predicted primary cell CFTR rescue. We created two nasal epithelial cell lines from donors carrying the F508del allele. These cell lines served as a positive control for effective modulator response (i.e., Trikafta) and provided proof-of-concept that patient-derived cell lines are predictive of clinical response.

One of the two F508del-carrying cell lines, UNCX3T, was created from a compound heterozygous donor with genotype F508del/S492F. S492F encodes a rare mutation in the nucleotide-binding domain 1 (NBD1) of CFTR degradation alone with no apparent aminoglycoside synergy. This difference could be due to cell-type specific differences in GSPT1 expression or the unique biology of CFTR. We previously reported a robust non-linear relationship between CFTR function and the proportion of CFTR-expressing cells (68) . Accordingly, even low levels of PTC readthrough could account for the ~20% of wildtype CFTR function recovered by CC-90009 treatment.

A weakness of this study is the inability to detect CFTR protein expression in PTC-carrying primary airway epithelial cells. Even in non-CF cells, CFTR protein abundance is low and CFTR is barely detectable after immunoprecipitation with a CFTR antibody. Studies in progress are evaluating more sensitive methods to detect changes in CFTR protein expression. Despite current CFTR protein evidence, the lack of therapeutic effect on F508del-and wildtype-CFTR (Supplemental Figure 6 ) strongly suggests that CC-90009's activity is specific to PTC-relevant pathways.

Partial suppression of NMD by ribosomal readthrough with aminoglycosides (69, 70) or with CC-90009 (18) is a well-established phenomenon. Likewise, we found that CC-90009 increased the level of CFTR mRNA expression ( Figure 5E ). However, a far greater increase was seen after treatment with Smg1i (Supplemental Figure 5D ) with no corresponding increase in CFTR ion transport (Supplemental Figure 5A-B) . Thus, we posit that the functional rescue of W1282X-CFTR is likely not explained by NMD suppression alone. Follow-up proteomic studies are required to definitively confirm PTC readthrough as the primary mechanism of W1282X- We speculate that the decreased level of β-and γ-ENaC may destabilize the heterotrimer and cause the loss of α-ENaC protein. Future work to understand the connection between CC-90009 and ENaC will be necessary.

One concern of therapeutically degrading GSPT1 is the possibility of normal termination codon (NTC) readthrough. However, NTCs are thought to be more highly regulated than PTCs since basal levels of readthrough are ~10 times more likely at a PTC (60). Further, studies of aminoglycoside-and Ataluren-induced readthrough demonstrate no adverse effect on translation termination at NTCs (53, 77). While these studies set the precedent that readthrough therapies are safe, future work will be required to determine whether GSPT1 degradation affects translation termination at NTCs.

In this study, we describe a pipeline for developing patient-derived airway epithelial cell lines to model rare CFTR variants. Bmi-1/hTERT nasal and bronchial cell lines are highly representative of primary airway cells and predictive of clinical response to CFTR modulators. We present the novel finding that GSPT1 degradation by CC-90009 promotes robust functional rescue of W1282X-CFTR. Follow-up studies to understand the mechanism of rescue will be vital to develop this new therapeutic strategy for treating people with CF currently ineligible for modulator therapy.

Primary cell isolation and tissue culture. Primary HNECs were obtained by nasal curettage from two non-CF and three CF donors, yielding 2.1 x 10 6 cells on average. Demographics and CFTR genotypes are summarized in Table I . Freshly obtained samples were transferred into Lactated Ringer's Solution. Tissue dissociation was performed by treating with dithiothreitol (DTT, 0.5 mg/mL; Sigma-Aldrich, D0632) and DNase for each donor. Data were analyzed using a linear mixed-effects model with the donor as a random effect factor. E) Relative CFTR and ENaC mRNA levels by qRT-PCR in UNCX2T and parent cells. N=4-6. Data were analyzed using an ordinary linear model. F) Western blot for α-ENaC in UNCCF9T parent cells pretreated with escalating doses of CC-90009 for 24 hours. Protein expression normalized to GAPDH and relative to the DMSO control quantified below. Cells were grown in Vertex ALI media to increase the level of ENaC expression to be detectable by western blot. All data presented as mean ± SD. ** = p<0.01; *** = p<0.001. 

Cystic fibrosis genetics: from molecular understanding to clinical application

Primary epithelial cell models for cystic fibrosis research

Rescue of DeltaF508-CFTR trafficking and gating in human cystic fibrosis airway primary cultures by small molecules

Pharmacological rescue of conditionally reprogrammed cystic fibrosis bronchial epithelial cells

Brushed nasal epithelial cells are a surrogate for bronchial epithelial CFTR studies

Feasibility of nasal epithelial brushing for the study of airway epithelial functions in CF infants

Different culture conditions affect drug transporter gene expression, ultrastructure, and permeability of primary human nasal epithelial cells

Optimization of human nasal epithelium primary culture conditions for optimal proton oligopeptide and organic cation transporters expression in vitro

Well-differentiated human airway epithelial cell cultures

Use of primary cultures of human bronchial epithelial cells isolated from cystic fibrosis patients for the pre-clinical testing of CFTR modulators

Conditionally reprogrammed cells represent a stem-like state of adult epithelial cells

A pilot study of the effect of gentamicin on nasal potential difference measurements in cystic fibrosis patients carrying stop mutations

Gentamicin-induced correction of CFTR function in patients with cystic fibrosis and CFTR stop mutations

Advances in therapeutic use of a drug-stimulated translational readthrough of premature termination codons

Elucidating the mechanism of action of CC-90009, a novel cereblon E3 ligase modulator, in AML via genome-wide CRISPR screen

CC-90009, a novel cereblon E3 ligase modulator, targets GSPT1 for degradation to induce potent tumoricidal activity against acute myeloid leukemia (AML)

Clinical activity of CC-90009, a cereblon E3 ligase modulator and first-in-class GSPT1

degrader, as a single agent in patients with relapsed or refractory acute myeloid leukemia (R/R AML): First results from a phase I dose-finding study

Pharmacodynamic responses to CC-90009, a novel cereblon E3 ligase modulator, in a phase I dose-escalation study in relapsed or refractory acute myeloid leukemia (R/R AML)

A novel cereblon E3 ligase modulator eradicates acute myeloid leukemia stem cells through degradation of translation termination factor GSPT1

Assessing human airway epithelial progenitor cells for cystic fibrosis cell therapy

Sense from nonsense: therapies for premature stop codon diseases

Suppression of premature termination codons as a therapeutic approach

Divergent effects of translation termination factor eRF3A and nonsense-mediated mRNA decay factor UPF1 on the expression of uORF carrying mRNAs and ribosome protein genes

Central role for ENaC in development of hypertension

Regulation of blood pressure, the epithelial sodium channel (ENaC), and other key renal sodium transporters by chronic insulin infusion in rats

Defective regulation of the epithelial Na+ channel by Nedd4 in Liddle's syndrome

Liddle's syndrome: heritable human hypertension caused by mutations in the beta subunit of the epithelial sodium channel

Mutations in subunits of the epithelial sodium channel cause salt wasting with hyperkalaemic acidosis, pseudohypoaldosteronism type 1

PTC124 targets genetic disorders caused by nonsense mutations

The authors thank the cell donors; Dr. Snyder