Abstract
Many cellular processes are governed by protein–protein interactions
that require tight spatial and temporal regulation. Accordingly, it is
necessary to understand the dynamics of these interactions to fully
comprehend and elucidate cellular processes and pathological disease
states. To map de novo protein–protein interactions with time
resolution at an organelle-wide scale, we developed a quantitative mass
spectrometry method, time-resolved interactome profiling (TRIP). We
apply TRIP to elucidate aberrant protein interaction dynamics that lead
to the protein misfolding disease congenital hypothyroidism. We
deconvolute altered temporal interactions of the thyroid hormone
precursor thyroglobulin with pathways implicated in hypothyroidism
pathophysiology, such as Hsp70-/90-assisted folding, disulfide/redox
processing, and N-glycosylation. Functional siRNA screening identified
VCP and TEX264 as key protein degradation components whose inhibition
selectively rescues mutant prohormone secretion. Ultimately, our
results provide novel insight into the temporal coordination of protein
homeostasis, and our TRIP method should find broad applications in
investigating protein-folding diseases and cellular processes.
Keywords: Temporal Proteomics, Thyroglobulin, Hypothyroidism,
Proteostasis, Bioorthogonal Protein Labeling
Subject terms: Biotechnology & Synthetic Biology, Proteomics
Synopsis
graphic file with name 44320_2024_58_Figa_HTML.jpg
A new temporal proteomics technique reveals the protein interaction
dynamics of folding and secretion pathways. The method clarifies how
mutant forms of the pro-hormone thyroglobulin result in congenital
hypothyroidism by elucidating the defects in protein quality control.
* TRIP combines pulsed unnatural amino acid labeling,
biotin-streptavidin enrichment, and TMTpro-multiplexed proteomics
to study time-resolved protein interactions.
* Secretion-defective thyroglobulin mutants exhibit prolonged
interaction dynamics with chaperone, disulfide exchange, and
degradation pathway interactions.
* Depletion of degradation factors VCP and TEX264 rescues mutant
prohormone secretion.
* VCP inhibition shifts temporal protein interactions to rescue
mutant thyroglobulin secretion.
__________________________________________________________________
A new temporal proteomics technique reveals the protein interaction
dynamics of folding and secretion pathways. The method clarifies how
mutant forms of the pro-hormone thyroglobulin result in congenital
hypothyroidism by elucidating the defects in protein quality control.
graphic file with name 44320_2024_58_Figb_HTML.jpg
Introduction
Protein–protein interactions drive functional diversity within cells
and are often closely connected to the observed phenotypes for cellular
processes and disease states (Bludau and Aebersold, [35]2020). Protein
homeostasis (proteostasis) is a critical cellular process that relies
on tightly regulated protein interactions. The proteostasis network
(PN), consisting of protein-folding chaperones, trafficking, and
degradation components, maintains the integrity of the proteome by
ensuring the appropriate trafficking and localization of properly
folded proteins while recognizing misfolded, potentially detrimental
states and routing them for degradation (Karagöz et al, [36]2019;
Needham et al, [37]2019; Behnke et al, [38]2016; Pohl and Dikic,
[39]2019). The concerted action of hundreds of proteostasis factors is
referred to as protein quality control (PQC). Perturbations to the PN
through genetic, age-related, or environmental factors manifest in
several disease states, including amyloidosis, neurodegeneration,
cancer, and others (Taldone et al, [40]2020; McDonald et al, [41]2022;
Wright et al, [42]2021; Kuo et al, [43]2021; Marinko et al, [44]2021).
Identifying and quantifying protein–protein interactions has been
critical for comprehending the pathogenesis of these disease states.
Approaches include yeast two-hybrid systems, co-immunoprecipitation
coupled with western blot analysis, the Luminescence-based Mammalian
IntERactome (LUMIER) assay, as well as affinity purification—mass
spectrometry (AP-MS) (Taipale et al, [45]2012, [46]2014; Piette et al,
[47]2021; Rizzolo et al, [48]2017; Rizzolo and Houry, [49]2019; Wright
and Plate, [50]2021). These methods have been powerful for mapping
steady-state proteostasis interactions to disease states, yet most lack
the ability to measure interaction dynamics over time. Proximity
labeling mass spectrometry (BioID and APEX-MS) has had limited use to
spatiotemporally resolve protein–protein interactions only following
protein maturation, as synchronization of newly synthesized protein
populations is challenging to achieve (Lobingier et al, [51]2017; Perez
Verdaguer et al, [52]2022). Unnatural amino acid incorporation methods,
such as biorthogonal non-canonical amino acid tagging (BONCAT), pulsed
azidohomoalanine (PALM), or heavy isotope labeled azidohomoalanine
(HILAQ) can identify newly synthesized proteins but have not focused on
a single endogenously expressed protein or group of proteins in the
context of disease (Dieterich et al, [53]2006; Bagert et al, [54]2014;
Ma et al, [55]2018; McClatchy et al, [56]2015; Ma et al, [57]2017;
Howden et al, [58]2013; van Bergen et al, [59]2022).
Similar DNA and RNA labeling approaches are used to temporally resolve
nascent DNA–protein and RNA–protein interactions in cell culture and
whole organisms. Isolation of proteins on nascent DNA (iPOND) has
revealed the timing of DNA–protein interactions during replication and
chromatin assembly (Cortez, [60]2017; Sirbu et al, [61]2011; Munden et
al, [62]2022). Thiouracil cross-linking mass spectrometry (TUX-MS) and
viral cross-linking and solid-phase purification (VIR-CLASP) are used
to study the timing of RNA–protein interactions during viral infection
(Kim et al, [63]2020; Phillips et al, [64]2016). In contrast, no
methods were previously available to identify de novo protein–protein
interactions of newly synthesized proteins in a client-specific manner
with temporal resolution, which has motivated our efforts here.
In earlier work, we mapped the interactome of the secreted thyroid
prohormone thyroglobulin (Tg), comparing the WT protein to
secretion-defective mutations implicated in congenital hypothyroidism
(CH) (Wright et al, [65]2021). Tg is a heavily post-translationally
modified 330 kDa prohormone that is necessary to produce
triiodothyronine (T3) and thyroxine (T4) thyroid-specific hormones
(Citterio et al, [66]2019; Coscia et al, [67]2020). Tg biogenesis
relies extensively on distinct interactions with the PN to facilitate
folding and eventual secretion. Our previous results identified
topological changes in Tg–PN interactions among CH-associated Tg
mutants compared to WT. Nonetheless, the lack of temporal resolution
precludes more mechanistic discernment of these changes in PN
interactions. While some changes may simply correlate with disease
pathogenesis, others may be directly responsible for the aberrant PQC
and secretion defect of the mutant Tg variants.
To address this shortcoming, we developed time-resolved interactome
profiling (TRIP) to capture and quantify interactions between Tg and
interacting partners throughout the life cycle of the protein. We found
that Tg mutants are characterized by both discrete changes with select
PN components and broad temporal alterations across Hsp70/90,
N-glycosylation, and disulfide/redox-processing pathways. Moreover, we
find that these perturbations are correlated with alterations in
interactions with degradation components. We coupled our TRIP method
with functional siRNA screening and uncovered that VCP (p97) and TEX264
are two key regulators of Tg processing. VCP and TEX264 inhibition or
silencing in thyroid cells rescued the secretion of mutant Tg,
representing—to our knowledge—the first restorative approach based upon
proteostasis modulation to increase mutant Tg secretion.
Results
TRIP temporally resolves Tg interactions with PQC components
To develop the time-resolved interactome profiling method, we
envisioned a two-stage enrichment strategy utilizing epitope-tagged
immunoprecipitation coupled with pulsed biorthogonal unnatural amino
acid labeling and functionalization (Fig. [68]1A). Cells are
pulse-labeled with homopropargylglycine (Hpg) to synchronize newly
synthesized protein populations. After the Hpg pulse, samples are
collected across timepoints throughout a chase period (Fig. [69]1A, Box
1) (Kiick et al, [70]2001; Beatty et al, [71]2006). The Hpg alkyne
incorporated into the newly synthesized population of protein is then
conjugated to biotin using copper-catalyzed azide-alkyne cycloaddition
(CuAAC) (Fig. [72]1A, Box 2). Subsequently, the first stage of the
enrichment strategy globally captures the client protein and binding
partners using epitope-tagged immunoprecipitation, followed by elution
(Fig. [73]1A, Box 3). The second enrichment step then utilizes a
biotin–streptavidin pulldown to capture the Hpg pulse-labeled, and
CuAAC-conjugated population, enriching samples into time-resolved
fractions (Fig. [74]1A, Box 4) (Li et al, [75]2020; Thompson et al,
[76]2019).
Figure 1. Time-resolved interactome profiling to identify interactions with
newly synthesized proteins.
[77]Figure 1
[78]Open in a new tab
(A) Key steps necessary for the TRIP workflow. Box 1 shows pulsed
unnatural amino acid labeling with Hpg to incorporate an alkyne
functional group into newly synthesized proteins followed by in situ
cross-linking of protein interactions using DSP. Box 2 shows labeled
protein functionalization with copper-catalyzed azide-alkyne
cycloaddition (CuAAC) click chemistry. Box 3 shows FLAG
co-immunoprecipitation to globally enrich labeled and unlabeled Tg. Box
4 shows subsequent streptavidin-biotin enrichment of the pulse-labeled,
time-resolved Tg fractions. (B–D) Western blot analysis of the
two-stage purification strategy with continuous Hpg labeling. FRT cells
stably expressing WT Tg were continuously labeled with Hpg (200 μM) for
4 h and cross-linked with DSP (0.5 mM) for 10 min to capture transient
proteostasis network interactions ( + Xlink). Lysates were
functionalized with a TAMRA-Azide-PEG-Desthiobiotin probe using CuAAC
and subjected to the dual affinity purification scheme as described.
(B) FLAG IP inputs as described in Box 2 in (A). Top shows a
fluorescence image of TAMRA-labeled proteins, and the bottom shows
immunoblots of Tg (IB: FLAG) and interactors HSP90B1, HSPA5, PDIA4, and
loading control GAPDH. (C) FLAG IP elutions as described in Box 3 in
(A). The top shows a fluorescence image of TAMRA-labeled Tg and total
Tg (IB: FLAG). The bottom shows immunoblots of interactors (HSP90B1,
HSPA5, and PDIA4). Samples were then subjected to biotin pulldown. (D)
Biotin pulldown elutions as described in Box 4 in (A). The top shows a
fluorescence image of TAMRA-labeled Tg and total Tg (IB: FLAG). The
bottom shows immunoblots of interactors (HSP90B1, HSPA5, and
PDIA4). [79]Source data are available online for this figure.
Thyroglobulin was chosen as the model secretory client protein. We
generated isogenic Fischer rat thyroid (FRT) cells that stably
expressed FLAG-tagged Tg (Tg-FT), including WT or mutant variants
(A2234D and C1264R) (Appendix Fig. S[80]1). WT Tg is readily secreted
from the FRT cells, while C1264R Tg shows only minimal residual
secretion (~2% of WT when detected after immunoprecipitation) (Appendix
Fig. S[81]1D). A2234D is fully secretion deficient, consistent with the
prior characterization of these variants (Hishinuma et al, [82]1999;
Pardo et al, [83]2009). We first set out to determine if Tg could
tolerate immunoprecipitation after pulse Hpg labeling,
dithiobis(succinimidyl propionate) (DSP) cross-linking, and CuAAC
conjugation with a trifunctional tetramethylrhodamine
(TAMRA)-Azide-Polyethylene Glycol (PEG)-Desthiobiotin probe. We showed
previously that a C-terminal FLAG-tag is tolerated by Tg and allows
efficient immunoprecipitation, while the DSP crosslinker aids in
capturing transient protein–protein interactions that take place during
Tg processing (Wright et al, [84]2021). We pulse-labeled FRT cells
expressing WT or C1264R Tg for 4 h with Hpg, performed DSP
cross-linking, CuAAC, and tested our two-stage enrichment strategy via
western blot analysis (Fig. [85]1B–D). Pulsed Hpg labeling, DSP
cross-linking, and CuAAC did not significantly affect
immunoprecipitation efficiency and allowed for robust two-stage
enrichment of WT and C1264R Tg-FT with well-validated Tg interactors
HSPA5 (BiP), HSP90B1 (Grp94), and PDIA4 (ERp72) (Fig. [86]1B,C) (Menon
et al, [87]2007; Baryshev et al, [88]2004; Wright et al, [89]2021).
Furthermore, the C-terminal FLAG-tag and Hpg labeling are necessary for
this two-stage enrichment strategy, and DSP cross-linking is necessary
to capture these interactions after stringent wash steps (Fig. [90]1D;
Appendix Fig. S[91]2).
Next, we investigated whether TRIP could temporally resolve
interactions with these PN components. We pulse-labeled WT Tg-FRT cells
with Hpg for 1 h, followed by a 3 h chase in regular media capturing
timepoints in 30-min intervals and analyzing via western blot or TMTpro
LC-MS/MS (Fig. [92]2A). Our previous study indicated that ~70% of WT
Tg-FT was secreted after 4 h, while ~30% of A2234D and 15% of C1264R
was degraded after the same time period (Wright et al, [93]2021).
Therefore, we reasoned that a 3-h chase period would be enough time to
capture the majority of Tg interactions throughout processing,
secretion, cellular retention, and degradation, while still being able
to capture an appreciable amount of sample for analysis. For WT-Tg,
interactions with HSPA5 peaked within the first 30 min of the chase
period and rapidly declined, in line with previous observations, but
PDIA4 interactions were not detectable by western blot analysis
(Fig. [94]2B) (Menon et al, [95]2007; Kim and Arvan, [96]1995). To test
the ability of TRIP to distinguish temporal differences in PQC
interactions across mutant Tg variants, we performed TRIP on FRT cells
expressing C1264R Tg, a known patient mutation implicated in CH
(Hishinuma et al, [97]1999; Kanou et al, [98]2007). For C1264R,
interactions with HSPA5 were highly abundant at the 0 h timepoint and
remained mostly steady throughout the first 1.5 h (Fig. [99]2C). A
similar temporal profile was also observed for HSP90B1. In addition,
interactions with PDIA4 were detectable for C1264R and were found to
gradually increase throughout the first 1.5 h of the chase period,
before rapidly declining (Fig. [100]2C). We noticed similar temporal
profiles for PDIA4 and HSPA5 to our western blot analysis, when
measured via TMTpro LC-MS/MS as further outlined below
(Fig. [101]2D,E). In particular, the HSPA5 WT-Tg interaction declined
within the first hours, yet for C1264R Tg, the HSPA5 interactions
remained mostly steady over the 3-h chase period (Fig. [102]2E).
Figure 2. TRIP temporally resolves Tg interactions with protein quality
control components.
[103]Figure 2
[104]Open in a new tab
(A) Workflow for TRIP protocol utilizing western blot or mass
spectrometric analysis of time-resolved interactomes. (1) Cells are
pulse-labeled with Hpg (200 μM final concentration) for 1 h, chased in
regular media for specified timepoints, and cross-linked with DSP
(0.5 mM) for 10 min to capture transient proteoastasis network
interactions; (2) lysates are functionalized with a
TAMRA-Azide-PEG-Desthiobiotin probe using copper CuAAC Click reaction;
(3) lysates undergo the first stage of the enrichment strategy where
Tg-FT is globally captured and enriched using immunoprecipitation; (4)
eluted Tg-FT populations from the global immunoprecipitation undergo
biotin–streptavidin pulldown to capture the pulse Hpg-labeled, and
CuAAC-conjugated population of Tg-FT, enriching samples into
time-resolved fractions; (5) time-resolved fraction may then undergo
western blot analysis or (6) quantitative liquid chromatography–tandem
mass spectrometry (LC-MS/MS) analysis with tandem mass tag (TMTpro)
multiplexing for analysis. The (−) Hpg control channel is used to
identify enriched interactors and a (−) Biotin pulldown channel to act
as a booster (or carrier). (B, C) TRIP western blot analysis of WT
Tg-FT (B) and C1264R Tg-FT (C). Samples were processed as described
above in (A). FLAG IP Input panel shows immunoblot of Tg (IB: FLAG),
and interactors HSP90B1, HSPA5, PDIA4 and loading control GAPDH. Biotin
pulldown elution panel shows a fluorescence image of TAMRA-labeled Tg
and immunoblot of Tg interactors HSP90B1 and HSPA5. Validated Tg
interactors show higher and delayed enrichment with the misfolded
C1264R Tg mutant in (C) compared to WT in (B). PDIA4 interactions were
not detectable by western blot analysis for biotin pulldown elution
with WT Tg. (D, E) Plots showing the scaled enrichment of select Tg
interactors HSPA5 (E) and PDIA4 (D) compared across constructs. Samples
were processed as described above in (A) and analyzed by mass
spectrometry. The solid line corresponds to mean and shading represents
the SEM (N = 5 for WT Tg; N = 6 for A2234D and C1264R Tg). Data in
Dataset [105]EV4. [106]Source data are available online for this
figure.
These data highlight the utility of TRIP to not only identify changes
in protein interactions over time but also monitor how these
interactions differ for a given protein of interest across disease
states. Moreover, these data corroborate our previous findings that
steady-state interactions with HSPA5, HSP90B1, and PDIA4 are elevated
for C1264R Tg (Wright et al, [107]2021).
TRIP identifies altered temporal dynamics associated with Tg processing
We benchmarked the utility of our TRIP approach to temporally resolve
previously identified and novel interactors, as the Tg interactome has
not been fully characterized in native tissue. We focused on A2234D and
C1264R Tg as they present with distinct defects in Tg processing, and
mutations are localized in separate structural domains (Kanou et al,
[108]2007; Hishinuma et al, [109]1999). Following the Hpg pulse-chase
labeling scheme and dual affinity purification, the time-resolved Tg
fractions were trypsin/Lys-C digested and labeled individually with
isobaric TMTpro tags. Subsequently, two sets of TRIP time course
samples (0, 0.5, 1, 1.5, 2, and 3 h) could be pooled using the 16plex
TMTpro and analyzed by LC-MS/MS (Fig. [110]2A). In total, five
biological replicates were analyzed for WT and six biological
replicates were analyzed for A2234D and C1264R (Dataset EV[111]8).
Aside from the experimental samples, we utilized a (–) biotin pulldown
booster (or carrier) channel with cells that were pulse-labeled with
Hpg, underwent CuAAC functionalization, and immunoprecipitation, but
did not undergo biotin–streptavidin enrichment (Fig. [112]2A). This
booster sample acted to (1) aid in increased peptide/protein
identification—compared to the much lower abundant chase samples; and
(2) benchmark Tg interactors in FRT cells compared to our previously
published dataset (Tsai et al, [113]2020; Petelski et al, [114]2021).
When comparing the booster channel to our (–)Hpg negative control
samples, most previously identified Tg interactors were strongly and
significantly enriched (Appendix Fig. S[115]3). Our dataset in this
study showed appreciable overlap between our previous results in
HEK293T cells identifying 75 of the previous 171 Tg interactors and
identifying 198 new interactors (Fig. [116]3A). Several ribosomal and
proteasomal subunits, trafficking factors, and lysosomal components
were not identified in our previous dataset (Appendix Fig. S[117]3;
Dataset EV[118]2). We then took our list of previously identified
interactors and PQC components found to be enriched in the (–) biotin
pulldown samples compared to (–) Hpg and carried these proteins forward
to time-resolved analysis utilizing the Hpg-chase samples (Dataset
EV[119]6).
Figure 3. TRIP identifies altered temporal dynamics associated with Tg
processing.
[120]Figure 3
[121]Open in a new tab
(A) Venn diagram showing the overlap in proteostasis components
identified as Tg interactors here compared to our previous dataset
(Wright et al, [122]2021). (−) Biotin pulldown vs (−) Hpg samples were
used to identify Tg interactors in FRT cells. Data available in Dataset
EV[123]3. (B, C) Plot showing the relative pathway enrichment for
Hsp70-/90-assisted folding & chaperoning interactors (B) and
disulfide/redox-processing interactors (C) with WT, A2234D, or C1264R
Tg constructs. For this and all subsequent panels, the average log2
fold change enrichment value across timepoints for a given interactor
were used to scale data. Positive enrichment was scaled from 0 to 1.
Lines represent the median scaled enrichment for the group of
interactors and shades correspond to the first and third quartile
cutoff. All source data for this and subsequent panels can be found in
Dataset EV[124]4. (D, E) Heatmap showing the relative enrichment for
individual Hsp70-/90-assisted folding & chaperoning interactors (D) and
disulfide/redox-processing interactors (E) with WT, A2234D, or C1264R
Tg constructs. Relative enrichment scaled as described above in (C).
(F) Unbiased k-means clustering of TRIP profiles to determine
co-regulated groups of interactors. Aggregate time profiles for the
most prominent clusters are shown on the left (WT) and the right
(C1264R). The line corresponds to the mean scaled log2 fold enrichment,
and the shading represents the 25–75% quartile range within each
cluster. The Sankey plot in the center shows the shift of interactors
between clusters from WT to C1264R. (G–I) Heatmap showing the relative
enrichment for select glycan processing (G), proteasomal degradation
(H) and autophagy interactors (I) with WT, A2234D, or C1264R Tg. (J)
Plot showing the relative pathway enrichment for proteasomal
degradation interactors. (K) Plot comparing the relative enrichment of
VCP throughout the time course for WT, C1264R, and A2234D Tg. The TRIP
data (left) is contrasted to aggregate (steady-state) interactomics
data (right) (Wright et al, [125]2021). TRIP resolves dynamic VCP
interaction changes with mutant Tg, while these changes are muted in
the aggregate data. On the right, data are represented as mean ± SEM
(N = 12 biological replicates for C1264R, and 6 biological replicates
for A2234D). On the left, a solid line corresponds to the mean, and
shading represents the SEM (N = 5 for WT Tg; N = 6 biological
replicates for A2234D and C1264R Tg).
To map the time-resolved Tg–PN interactome changes, we considered the
Hpg-chase samples (0–3 h) and compared the enrichment of Tg interactors
to the (–) Hpg control. The enrichments were normalized to Tg protein
levels to account for gradual changes due to Tg secretion or
degradation, and positive enrichment values were scaled from 0 to 1. We
organized the interactors according to distinct PN pathways known to
influence Tg processing (Appendix Fig. S[126]4; Datasets EV[127]3 and
EV[128]4).
To benchmark the TRIP methodology, we chose to monitor a set of
well-validated Tg interactors and compare the time-resolved PN
interactome changes to our previously published steady-state
interactomics dataset (Wright et al, [129]2021). Previously, we found
that CALR (Calreticulin), CANX (Calnexin), ERP29 (PDIA9), ERP44, and
P4HB (PDIA1) interactions with mutants A2234D or C1264R Tg exhibited
little to no change when compared to WT under steady-state conditions
(Fig. [130]EV1A). However, in our TRIP dataset we were able to uncover
distinct temporal changes in engagement that were previously masked
within the steady-state data. Our time-resolved data deconvolutes these
aggregate measurements, revealing prolonged CALR, ERP29, and P4HB
engagements for both A2234D and C1264R Tg mutants compared to WT
(Fig. [131]EV4B–F). We found that these measurements for key
interactors and PN pathways exhibited robust reproducibility, as
exemplified by the standard error of the mean for the TRIP data
(Fig. [132]EV1B–I; Appendix Fig. S[133]4B).
Figure EV1. TRIP data for individual interactors.
[134]Figure EV1
[135]Open in a new tab
(A) Aggregate (steady-state) interactomics data comparing the
enrichment of Tg interactors for mutant Tg to WT Tg (data from Wright
et al, [136]2021). Interactions are mostly unchanged for mutant Tg
relative to WT Tg. Data is represented as mean ± SEM (N = 12 biological
replicates for C1264R, and 6 biological replicates for A2234D). (B–F)
Plots comparing the relative enrichment of interactors CALR (B), CANX
(C), ERP29 (D), ERP44 (E), and P4HB (F) throughout the TRIP time course
for WT, A2234D, and C1264R Tg. TRIP data can resolve dynamic
interaction changes for several mutant Tg interactors, while these
changes are muted in the aggregate data. Solid line corresponds to mean
and shading represents the SEM (N = 5 for WT Tg; N = 6 biological
replicates for A2234D and C1264R Tg). (G–I) Plots comparing the
relative enrichment of individual disulfide/redox-processing
interactors throughout the TRIP time course for WT (G), A2234D (H), and
C1234R (I). Individual protein disulfide isomerases exhibit distinct
peak times when interactions reach maximum, thereby revealing an order
to their engagement. For instance, PDIA3, PDIA4, PDIA6, and P4HB peak
at 0 h for WT Tg, while TXNDC12 peaks later at 1 h. Moreover, the exact
temporal sequence of PDI engagements is shifted for A2234D (H) and
C1234R Tg (I). Solid line corresponds to mean and shading represents
the SEM (N = 5 for WT Tg; N = 6 biological replicates for A2234D and
C1264R Tg). Data available in Dataset EV[137]4.
Figure EV4. Pulse-chase analysis of WT and C1264R Tg with pharmacological VCP
inhibition.
[138]Figure EV4
[139]Open in a new tab
Autoradiographs and quantifications of pulse-chase analysis of C1264R
Tg-FT (A, B) and WT Tg-FT (C, D) in FRT cells with ML-240 treatment.
Cells were pre-treated with ML-240 or DMSO for 15 min prior to pulse
labeling with EasyTag ^35S Protein Labeling Mix (Perkin Elmer,
NEG772007MC) for 30 min and chased for 4 h with DMSO or ML-240
treatment, collecting samples at 0-, 2-, and 4-h time points.
Autoradiographs from a representative experiment are shown in (A, C).
Quantification is shown in (C, D). Data is normalized to the timepoint
of maximum Tg recovery (C1264R) or 0 h (WT) and represented as
mean ± SEM. Statistical testing was performed using an unpaired
Student’s t test with Welch’s correction with P values as indicated.
N = 5–6 biological replicates as shown.
Next, we monitored temporal changes more broadly across proteostasis
network pathways. We found that both A2234D and C1264R exhibited
prolonged interactions with components of Hsp70/90 and
disulfide/redox-processing pathways (Fig. [140]3B–E). Particularly,
A2234D and C1264R showed increased interactions with HSPA5, HSP90B1,
HYOU1 (Grp170), DNAJC3 (ERdj6), and SDF2L1 throughout the chase period
(Fig. [141]3D). Conversely, WT interactions peaked at the 0 h timepoint
and consistently tapered off for many of these components
(Fig. [142]3B,D). Similarly, divergent temporal interactions were
observed in the case of disulfide/redox-processing components. PDIA3
(ERp57), PDIA4 (ERp72), PDIA6 (ERp5), and ERP29 have all been heavily
implicated in Tg processing (Fig. [143]3E) (di Jeso et al, [144]2005;
Menon et al, [145]2007; di Jeso et al, [146]2014; Baryshev et al,
[147]2006). While WT interactions with these components showed similar
trends as Hsp70/90 chaperoning components— peaking at the 0 h timepoint
and consistently decreasing—prolonged mutant interactions peaked later
at 1–1.5 h for both A2234D and C1264R (Fig. [148]3C,E). Furthermore,
individual protein disulfide isomerase interactions peaked at distinct
times, thereby revealing an order to their engagement
(Fig. [149]EV1G–I).
To assess temporal interaction changes in an unbiased fashion and
identify protein groups exhibiting comparative behavior, we carried out
k-means clustering of the temporal profiles for WT and C1264R. This
analysis revealed a large divergence in the interaction profiles. For
WT Tg, only one cluster exhibited steadily decreasing interactions
(cluster 4), while others increased with time, or showed peaks at
intermediate timepoints (Figs. [150]3F and [151]EV2A). On the other
hand, C1264R largely exhibited clusters with decreasing interactions
over time (Figs. [152]3F and [153]EV2B). Cluster 2 for WT with bimodal
interactions at early and late timepoints contains many Hsp70/90
chaperoning components. For C1264R Tg, many Hsp70/90 chaperoning
components and disulfide/redox-processing components are instead part
of cluster 2’, which exhibited an initial rise in interactions strength
before plateauing (Figs. [154]3F and [155]EV2A,B). In addition, we
carried out k-means clustering of the combined WT and C1264R time
series, which revealed similar clusters (Fig. [156]EV2C). This
divergent temporal engagement between WT Tg and the destabilized C1264R
mutant is aligned with the patterns observed in the manual grouping
(Fig. [157]3B,C), highlighting that the unbiased temporal clustering
can reveal broader patterns in the reorganization of the proteostasis
dynamics.
Figure EV2. Unbiased clustering of TRIP data.
[158]Figure EV2
[159]Open in a new tab
(A, B) Unbiased k-means clustering of TRIP profiles for WT (A) and
C1264R (B) to determine co-regulated groups of interactors. k-means
clustering was carried out by using the k-means function from the
tslearn python package with the data being normalized using the scaler
mean variance function. This analysis resulted in 7 distinct clusters
for WT and 6 clusters for C1264R. The line corresponds to the mean
scaled log2 fold enrichment and the shading represents the 25–75 %
quarter range within each cluster. (C) Heatmap showing unbiased k-means
clustering of the combined WT and C1264R Tg TRIP profiles. Only
interactors identified in both datasets were included. (D) Heatmap for
interactors in C1264R Cluster 3 (from B), which displayed the strongest
interactions at the initial 0 h timepoint. The scaled log2 fold change
enrichment for individual interactors is shown, and the individual
interactors are grouped by pathway. Several interactors related to
autophagy (brown) and glycan processing (purple), including the
glycoprotein folding sensor UGGT1, are present in this cluster.
TRIP highlights link between glycan processing and ER-phagy pathways
One area of particular interest was the crosstalk and correlation
between interactions with glycan-processing components and degradation
pathways. The link between the glycosylation state of ER clients and
ER-associated degradation (ERAD) is well established, whereas more
recently defined autophagy at the ER (ER-phagy) or ER to
lysosome-associated degradation (ERLAD) represents alternative
degradation mechanisms for ER clients (Christianson et al, [160]2008,
[161]2012; Fregno et al, [162]2021; Chiritoiu et al, [163]2020; Fregno
et al, [164]2018). Previously, we showed that A2234D and C1264R differ
in interactions with N-glycosylation components, particularly the
oligosaccharyltransferase (OST) complex. Efficient A2234D degradation
required both STT3A and STT3B isoforms of the OST, which mediate
co-translational or post-translational N-glycosylation, respectively
(Kelleher et al, [165]2003; Cherepanova and Gilmore, [166]2016). TRIP
revealed differential interactions with glycosylation components that
may lead to altered degradation dynamics. Many glycan-processing
enzymes, lectin chaperones, and several subunits of the OST complex
were identified (Fig. [167]3G; Appendix Fig. S[168]3A). While STT3A
interactions across all constructs showed similar temporal profiles, we
observed prolonged interactions for lectin chaperones CALR and CANX
with mutant Tg (Fig. [169]EV1B,C). The most striking difference
observed was with the “gatekeeping” glycosyltransferase UGGT1
(Fig. [170]3G). This protein regulates glycoprotein folding through the
CANX/CALR cycle by re-glycosylating ER clients, and thus triggering
reengagement with the lectin chaperone cycle (Lamriben et al,
[171]2016). UGGT1 interactions with WT remain moderate from 0.5 to
1.5 h and peak at 3 h, while interactions for A2234D and C1264R peaked
earlier and were more pronounced throughout the chase period. These
differential interactions with UGGT1 may suggest changes in the
monitoring of the Tg folded state. Moreover, CANX has been directly
linked to emerging mechanisms of ERLAD for glycosylated ER clients
(Forrester et al, [172]2019; Fregno et al, [173]2018).
Proteasomal and autophagic degradation pathways exhibited broad
differences in interaction dynamics for WT and mutant Tg
(Fig. [174]3H–J). The ERAD-associated lectin OS9 (Erlec2), and ATPase
VCP both peaked at the 3 h chase timepoint for WT Tg (Fig. [175]3H,K).
Conversely, A2234D exhibited more prolonged VCP interactions, and OS9
interactions peaked at the 2 h timepoint (Fig. [176]3H,K). For C1264R,
we observed much stronger and prolonged VCP interactions, as well as
additional interactions with the ERAD-associated mannosidase EDEM3 and
E3 ubiquitin ligase adaptor SEL1L (Hrd3) (Christianson et al,
[177]2008, [178]2012) (Fig. [179]3H). Most notably, our previous
aggregate steady-state data showed no significant difference for VCP
interactions between WT and mutant Tg, yet our TRIP workflow was able
to resolve these temporal dynamics (Fig. [180]3K) (Wright et al,
[181]2021), further highlighting the utility of this novel TRIP
methodology.
Our original dataset identified several lysosomal components, yet it
was unclear how Tg might be delivered to the lysosome. Recent work has
identified several selective ER-phagy receptors, and highlighted
ER-phagy mechanisms for the clearance of mutant prohormones and other
destabilized clients from the ER (Chen et al, [182]2021; Cunningham et
al, [183]2019). It was intriguing then to identify several lysosomal
and autophagy-related components and observe differential temporal
profiles across WT and C1264R Tg constructs (Fig. [184]3I; Appendix
Fig. S[185]4). For A2234D, interactions with these components were more
sparse. The most intriguing observation was the enrichment of three
different ER-phagy receptors, ATL3 (Atlastin-3), CCPG1 (Cpr8), and RTN3
(Reticulon 3) between WT and C1264R, along with the RTN3 adaptor
protein PGRMC1 (Fig. [186]3I) (Chen et al, [187]2019; Liang et al,
[188]2018; Smith et al, [189]2018; Grumati et al, [190]2017; Chen et
al, [191]2021). CCPG1 and RTN3 were found to specifically interact with
C1264R, with RTN3 interactions peaking at 0 h and then decreasing,
while CCPG1 interactions peaked later (Fig. [192]3I). In the C1264R
k-means clustered profiles, autophagy interactions largely group
together in the same cluster, showing stronger interactions at earlier
timepoints. In the same cluster are glycosylation components (UGGT1,
STT3B, and MLEC), further supporting a possible coordination for C1264R
Tg between lectin-dependent protein quality control and targeting to
autophagy (Fig. [193]EV2B,D).
siRNA screening discovers key regulators of construct-specific Tg processing
We developed an RNA interference screening platform to investigate
whether the temporal interaction changes discovered by TRIP are
functionally important for Tg PQC. Moreover, we were interested in
identifying factors whose modulation may act to rescue mutant Tg
secretion. HEK293 cells were engineered to stably express
nanoluciferase-tagged Tg constructs (Tg-NLuc) and screened against 167
Tg interactors and related PN components (see Dataset EV[194]5 for the
list of genes). The NLuc tag allowed us to monitor changes to both
intracellular Tg abundance in the cell lysates, and Tg levels in the
conditioned media to assess secretion rates in a 96-well format
(Fig. [195]4A). Importantly, the NLuc tag did not alter the secretion
of WT Tg, and CH-associated mutants maintained the same secretion
deficiency (Appendix Fig. S[196]5) (England et al, [197]2016).
Silencing of NAPA (α-SNAP) and LMAN1 (Ergic53) were found to increase
WT Tg-NLuc lysate abundance but had no effects on the two mutants
(Fig. [198]4B; Appendix Fig. S[199]6A; Dataset EV[200]5). NAPA is a
member of the Soluble N-ethylmaleimide-sensitive factor Attachment
Protein (SNAP) family and plays a critical role in vesicle fusion and
docking, while LMAN1 is a mannose-specific lectin that functions as a
glycoprotein cargo receptor for ER-to-Golgi trafficking (Song et al,
[201]2017; Zhao et al, [202]2007; Marinko et al, [203]2021). For mutant
Tg-NLuc constructs we found CTSZ (cathepsin Z) silencing decreased
A2234D Tg-NLuc lysate abundance, while GUSB (β-glucoronidase) silencing
increased C1264R lysate abundance (Fig. [204]4B).
Figure 4. siRNA screening finds regulators of construct-specific Tg
processing.
[205]Figure 4
[206]Open in a new tab
(A) siRNA screening workflow utilizing NLuc-tagged Tg to monitor lysate
and media abundance. Approximately 36 h after transfection with 25 nM
siRNAs cells were replenished with fresh media and Tg-NLuc abundance in
lysate and media was measured after 4 h using the nano-glo luciferase
assay system. (B) Violin plots showing the relative Tg-NLuc abundance
changes in lysate with siRNA knockdown of select genes. Tg-NLuc
abundance in lysate was measured 4 h after replenishing with fresh
media using the nano-glo luciferase assay system. Hits labeled within
the plot were defined as changes in Tg-NLuc abundance by greater than
3σ. N = 2 biological replicates for WT and A2234D Tg. N = 3 biological
replicates for C1264R Tg. (C) Violin plots showing the relative Tg-NLuc
abundance changes in media with siRNA knockdown of select genes. Sample
processing and cutoff criteria to identify hits are described above in
(B). N = 2 for WT and A2234D Tg. N = 3 for C1264R Tg. Full data for (B,
C) is available in Appendix Fig. S[207]6 and Dataset
EV[208]5. [209]Source data are available online for this figure.
Remarkably, we identified six genes whose silencing rescued mutant
Tg-NLuc secretion in a construct-specific manner. NAPA silencing
increased secretion of A2234D Tg-NLuc (Fig. [210]4C). This contrast to
the reduction in WT secretion with NAPA silencing may suggest an
alternative role for NAPA in regulating mutant Tg processing as other
proteins involved in vesicular fusion and trafficking have been
implicated in ER-phagy (Cui et al, [211]2019; Liang et al, [212]2020).
Silencing of P3H1 (Lepre1), an ER-resident prolyl hydroxylase,
increased C1264R Tg-NLuc secretion but not WT nor A2234D (Fig. [213]4C)
(Vranka et al, [214]2004). Silencing of several protein degradation
genes robustly increased mutant Tg secretion: VCP, HERPUD1, TEX264 and
RTN3. VCP silencing increased both A2234D and C1264R Tg-NLuc secretion
(Fig. [215]4C). VCP is associated with ERAD but also aids in several
diverse cellular functions including the interplay between proteasomal
and autophagic degradation (Hill et al, [216]2021; Christianson et al,
[217]2008, [218]2012). VCP silencing exclusively affecting mutant Tg
corroborates our TRIP dataset, and suggests a more prominent role for
VCP in mutant Tg PQC compared to WT. VCP interactions were sparse for
WT Tg, while they remained more steady throughout the chase period for
the mutants (Fig. [219]3H,K). HERPUD1, TEX264 and RTN3 silencing
selectively increased C1264R secretion, but did not alter WT nor A2234D
secretion (Fig. [220]4C). HERPUD1 is a ubiquitin-like protein and
associates with VCP during ERAD (Christianson et al, [221]2012; Needham
et al, [222]2019; Okuda-Shimizu and Hendershot, [223]2007). The
ER-phagy receptors RTN3 and TEX264 localize to subdomains of the ER to
facilitate degradation of specific ER clients and organellular regions
(Chino et al, [224]2019; Fielden et al, [225]2022; An et al, [226]2019;
Grumati et al, [227]2017; Chen et al, [228]2021; Cunningham et al,
[229]2019). Unfortunately, TEX264 was not identified in our TRIP data,
but RTN3 was found to specifically interact with only C1264R
(Fig. [230]3I; Appendix Fig. S[231]4). Of note, while the ER-phagy
receptor CCPG1 was identified in our mass spectrometry dataset, siRNA
silencing of CCPG1 did not significantly alter Tg-NLuc abundance in
lysate or media, nor did silencing of SEC62 or RETREG1 (FAM134B), two
additional ER-phagy receptors found to regulate ER dynamics (Appendix
Fig. S[232]6) (Fumagalli et al, [233]2016; Bhaskara et al, [234]2019).
This is the first study to broadly investigate the functional
implications of Tg interactors and other PQC network components on Tg
processing. Coupling these data with our TRIP methodology helped to
deconvolute PQC dynamics associated with Tg and identify pathways
implicated in the aberrant secretion of CH-associated Tg mutations. The
discovery of several protein degradation components as hits for
rescuing mutant Tg secretion may suggest that the blockage of
degradation pathways can broadly rescue the secretion of A2234D and
C1264R mutant Tg, a phenomenon similarly found for destabilized CFTR
implicated in the protein-folding disease cystic fibrosis (Vij et al,
[235]2006; Pankow et al, [236]2015; McDonald et al, [237]2022).
Trafficking and degradation factors selectively regulate Tg processing in
thyroid cells
We examined the hits from the initial siRNA screen in FRT cells stably
expressing Tg constructs to test whether their silencing exhibited
similar phenotypes in thyroid-specific tissue. Thyroid tissue must
synthesize and fold a large amount of Tg as it is the main protein
produced and can make up more than 50% of all protein components within
the thyroid gland (di Jeso and Arvan, [238]2016). Silencing of NAPA led
to a ~50% increase in WT-Tg lysate abundance, while NAPA and LMAN1
silencing both led to marginal decreases in WT-Tg secretion after 4 h,
consistent with the results in HEK293 cells (Figs. [239]5A,B and
[240]EV3A,B). Using ^35S pulse-chase analysis, we confirmed that NAPA
silencing significantly increased lysate retention by 15% over 4 h and
decreased secretion by 18% (Fig. [241]EV3H). To understand if these
results were directly attributable to NAPA function, we performed
complementation experiments where FRT cells treated with NAPA siRNAs
were cotransfected with a human NAPA plasmid. WT-Tg lysate abundance
decreased when NAPA expression was complemented, confirming that the
observed retention phenotype could be attributed to NAPA silencing
(Fig. [242]EV3I). These results established that NAPA acts as a
pro-secretion factor for WT Tg.
Figure 5. Trafficking and degradation factors regulate Tg processing in FRT
cells.
[243]Figure 5
[244]Open in a new tab
(A, B) Western blot analysis (A) and quantification (B) of WT Tg-FT
secretion from FRT cells transfected with select siRNAs hits from the
initial screening dataset. The red asterisk denotes a non-specific
background band within the western blot. Cells were transfected with
25 nM siRNAs for 36 h, media exchanged and conditioned for 4 h, Tg-FT
was immunoprecipitated from lysate and media samples, and Tg-FT amounts
were analyzed via immunoblotting. Data are represented as mean ± SEM
from N = 6 biological replicates. (C, D) Western blot analysis (C) and
quantification (D) of C1264R Tg-FT secretion from FRT cells transfected
with select siRNA hits from the initial screening dataset. Red asterisk
denotes a non-specific background band within the western blot. Cells
were transfected with 25 nM siRNAs for 36 h, media exchanged and
conditioned for 8 h, Tg-FT was immunoprecipitated from lysate and media
samples, and Tg-FT amounts were analyzed via immunoblotting. All
statistical testing performed using an unpaired Student’s t test with
Welch’s correction with P values as indicated. Data are represented as
mean ± SEM from N = 5 biological replicates (one RTN3 sample excluded
due to sample handling error). [245]Source data are available online
for this figure.
Figure EV3. Validation of siRNA screening hits in FRT cells.
[246]Figure EV3
[247]Open in a new tab
(A–G) Relative expression of knockdown targets in engineered WT Tg-FT
FRT cells with and without siRNA silencing measured by qRT-PCR. Data
was first normalized to a GAPDH loading control followed by
normalization to median expression of non-targeting transfected samples
and represented as mean ± SEM. (A) NAPA (α-SNAP), (B) LMAN1, (C) VCP,
(D) RTN3, (E) TEX264, (F) HERPUD1, (G) LEPRE11 (P3H1). Statistical
testing was performed using an unpaired Student’s t test with Welch’s
correction with P values as indicated. N = 3–7 biological replicates as
shown. Primers for detection are described in Dataset EV[248]7. (H)
Pulse-chase analysis of WT Tg-FT in FRT cells with NAPA (α-SNAP) siRNA
knockdown. Approximately 36 h after transfection with 25 nM siRNAs
cells were pulse-labeled with EasyTag ^35S Protein Labeling Mix (Perkin
Elmer, NEG772007MC) for 30 min and chased for 8 h, collecting samples
at 0-, 4-, and 8-h time points. Data is normalized to timepoint of
maximum Tg recovery and represented as mean ± SEM. % Degradation is
defined as
[MATH: 1−Tgtlysate<
/mi>+Tgtmedi
ax100 :MATH]
. Where
[MATH:
Tgt
lysate :MATH]
is the fraction of Tg-FT detected in the lysate at a given timepoint n,
and
[MATH:
Tgt
media
:MATH]
is the fraction of Tg-FT detected in the media at a given timepoint n.
Statistical testing performed using an unpaired Student’s t test with
Welch’s correction with P values as indicated. N = 6 biological
replicates. (I) Complementation of NAPA knockdown partially reverses
WT-Tg retention. FRT cells stably expressing WT Tg were cotransfected
with NAPA siRNA and siRNA resistant NAPA expression plasmid. Cells were
harvested 40 h post transfection and lysates were analyzed by Western
blot to monitor changes in WT-Tg amounts. Quantification (mean ± SEM)
is shown on the right (N = 3 biological replicates). (J, K) Individual
siRNA knockdown of VCP (J) and TEX265 (K) in C1264R Tg-FRT cells to
confirm the increase in Tg secretion. Cells were transfected with 25 nM
siRNAs for 36 h, media exchanged and conditioned for 8 h, Tg-FT was
immunoprecipitated from media samples, and Tg-FT amounts were analyzed
via immunoblotting. Multiple individual siRNAs recapitulated the
increase in C1264R secretion. Data is represented as mean ± SEM for
N = 10 (J) or 7 (K) biological replicates.
In C1264R Tg-FRT cells, RTN3, TEX264, and HERPUD1 silencing marginally,
yet significantly decreased C1264R lysate abundance (Figs. [249]5C,D
and [250]EV3C-G). Moreover, VCP and TEX264 silencing significantly
increased C1264R secretion by threefold and twofold, respectively,
after 8 h, in line with our results in HEK293 cells (Fig. [251]5C,D).
In contrast, silencing of RTN3 significantly decreased C1264R Tg
secretion by fivefold, in opposition to the increased secretion
observed in HEK293 cells. Several individual VCP and TEX264 siRNAs were
able to partially recapitulate these increased secretion phenotypes on
C1264R Tg-FT, confirming that the effect is mediated by the respective
gene silencing (Fig. [252]EV3J,K).
Mutant Tg is selectively enriched with the ER-phagy receptor TEX264
Intrigued by the finding that TEX264 silencing increased C1264R Tg
secretion without affecting WT Tg, we asked whether TEX264 exhibited
differential interactions with Tg constructs. HEK293 Tg-NLuc cells were
transfected with either a fluorescent GFP control, or C-terminal
FLAG-tagged ER-phagy receptors, followed by FLAG co-IP. The NLuc assay
and western blotting were then used to monitor Tg enrichment
(Fig. [253]6A). The negative control SEC62 did not yield any
appreciable enrichment of Tg compared to GFP control (Fig. [254]6B),
consistent with SEC62 not impacting Tg constructs in the siRNA screen.
Conversely, co-IP of TEX264 resulted in the enrichment of all Tg
variants compared to GFP control when monitored by NLuc assay, with
C1264R and A2234D being significantly enriched. C1264R Tg exhibited a
threefold increased interaction compared to WT Tg and almost a twofold
increase compared to A2234D Tg (Fig. [255]6B). This increase in C1264R
enrichment was also observable by western blot analysis (Fig. [256]6C).
In addition, we monitored Tg enrichment with ER-phagy receptors CCPG1
and RTN3 via Western blot as both were found to be C1264R Tg
interactors within our TRIP dataset. RTN3L is found to be the only RTN3
isoform involved in ER turnover via ER-phagy (Grumati et al,
[257]2017). WT and C1264R Tg-NLuc were modestly enriched with RTN3L
compared to control samples expressing GFP. Conversely, we found that
all Tg variants exhibited modest interactions with CCPG1 compared to
control samples expressing GFP, although less than with TEX264
(Appendix Fig. S[258]7).
Figure 6. Mutant Tg is selectively enriched with the ER-phagy receptor
TEX264.
Figure 6
[259]Open in a new tab
(A) Schematic for western blot and NLuc analysis for identifying
TEX264-Tg interactions. HEK293T cells stably expressed WT or mutant
Tg-NLuc were transfected with FLAG-tagged ER-phagy receptors TEX264,
SEC62, or GFP control. After FLAG Co-IP, samples were analyzed using
the using the nano-glo luciferase assay system or immunoblot analysis
to monitor Tg enrichment. (B) Luminescence data from FLAG-Co-IP and
nano-glo luciferase analysis. Tg is selectively enriched with TEX264
compared to GFP fluorescent control, or SEC62. Mutant Tg exhibits
higher enrichment compared to WT. Data represented as mean ± SEM.
Statistical testing was performed using a one-way ANOVA with post hoc
Tukey’s multiple testing corrections with adjusted P values as
indicated. N = 8 biological replicates. (C) Western blot analysis of
FLAG Co-IP samples. Top panel shows IP inputs and bottom panel IP
elutions with Tg (IB: Tg), ER-phagy receptors (IB: FLAG and IB: TEX264)
and loading control Tubulin. Tg is selectively enriched with TEX264
compared to GFP fluorescent control, or SEC62, with C1264R Tg
exhibiting higher enrichment compared to WT. [260]Source data are
available online for this figure.
Together, these data suggest that TEX264, CCPG1, or RTN3L engage with
Tg during processing, and CH-associated Tg mutants may be selectively
targeted to TEX264. Furthermore, ER-phagy may be considered as a
degradative pathway in Tg processing, as other studies have mainly
focused on Tg degradation through ERAD (Tokunaga et al, [261]2000;
Menon et al, [262]2007).
Pharmacological VCP inhibition selectively rescues C1264R Tg secretion
Considering the promising finding that silencing of degradation factors
by siRNA rescued C1264R secretion, we sought to investigate whether
pharmacological modulation of select Tg processing components could
similarly improve Tg secretion. There are no selective inhibitors
currently available for TEX264, but several for VCP. We monitored
C1264R Tg secretion from FRT cells in the presence of VCP inhibitors
ML-240, CB-5083, and NMS-873. ML-240 and CB-5083 are ATP-competitive
inhibitors that preferentially target the D2 domain of VCP subunits,
whereas NMS-873 is a non-ATP-competitive allosteric inhibitor that
binds at the D1-D2 interface of VCP subunits (Chou et al, [263]2013,
[264]2014; Anderson et al, [265]2015; le Moigne et al, [266]2017; Tang
et al, [267]2019). ML-240 and NMS-873 have been shown to decrease both
proteasomal degradation and autophagy, in line with VCP playing a role
in both processes (Chou et al, [268]2013, [269]2014; Her et al,
[270]2016). Conversely, while CB-5083 is known to decrease proteasomal
degradation it has been shown to increase autophagy. (Anderson et al,
[271]2015; le Moigne et al, [272]2017; Tang et al, [273]2019). We found
that treatment with ML-240 was able to significantly increase C1264R Tg
secretion without altering C1264R Tg lysate abundance, corroborating
the siRNA silencing data (Fig. [274]7A,B). To investigate whether this
rescue of C1264R Tg secretion with ML-240 treatment was specific to
C1264R Tg, we also monitored WT-Tg abundance. In contrast, we observed
that ML-240 significantly reduced WT-Tg abundance in both lysate and
media 4 and 8 h after treatments (Fig. [275]7C,D).
Figure 7. Secretion rescue of C1264R Tg is associated with temporal
remodeling of the Tg interactome.
[276]Figure 7
[277]Open in a new tab
(A, B) Western blots analysis (A) and quantification (B) of C1264R
Tg-FT FRT cells treated with VCP inhibitors ML-240 (10 μM), CB-5083
(5 μM), NMS-873 (10 μM), or vehicle (0.1% DMSO) for 8 h. Lysate and
media samples were subjected to immunoprecipitation and analyzed via
immunoblotting. Data are normalized to the median C1264R Tg-FT
abundance of DMSO-treated samples. Data represented as mean ± SEM.
Statistical testing was performed using an unpaired Student’s t test
with Welch’s correction with the exact P value indicated. N = 6
biological replicates. (C, D) Western blots analysis (C) and
quantification (D) of WT Tg-FT FRT cells treated with ML-240 for 4 and
8 h. Lysate and media samples were processed and analyzed as described
above. Data are normalized to the median WT Tg-FT abundance of
DMSO-treated samples. Data representation and statistical testing as in
(B). Data are represented as mean ± SEM from N = 6 biological
replicates. (E) ^35S Pulse-chase analysis of C1264R Tg-FT FRT cells
with ML-240 treatment. Cells were pre-treated with ML-240 or DMSO for
15 min prior to pulse labeling with ^35S for 30 min and chased for 4 h
with DMSO or ML-240 treatment. Data are represented as mean ± SEM. Full
data panel available in Fig. [278]EV4A,B. Statistical testing was
performed using an unpaired Student’s t test with Welch’s correction
with the exact P value indicated. N = 6 biological replicates. (F) ^35S
Pulse-chase analysis of WT Tg-FT FRT cells with ML-240 (10 μM)
treatment. Samples were processed and analyzed as described in (E).
Data are represented as mean ± SEM. Full data panel available in
Fig. [279]EV4C,D. Statistical testing was performed using an unpaired
Student’s t test with Welch’s correction with the exact P value
indicated. N = 6 biological replicates. (G–N) TRIP analysis showing the
relative enrichment for select Tg interactors across proteostasis
pathways between untreated and ML-240 (10 μM) treated C1264R Tg. Cells
were treated with ML-240 (10 μM) and analyzed utilizing the TRIP
workflow (Fig. [280]2). The average log2 fold change enrichment values
across timepoints for a given interactor are used to scale data.
Positive enrichment values were scaled from 0 to 1. The enrichment is
shown for Glycan Processing (G, H), Hsp70/90-Assisted Folding (I, J),
Disulfide/Redox Processing (K, L), and Proteasomal Degradation (M, N).
Aggregate pathway enrichment is displayed in (G), (I), (J) and (M),
where lines represent the median scaled enrichment for the group of
interactors and shades correspond to the first and third quartile
cutoff. Heatmaps with individual enrichments of interactors is shown in
(H), (J), (L), and (N). Data are available in Dataset
EV[281]4. [282]Source data are available online for this figure.
We hypothesized that ML-240 treatment may differentially regulate Tg
degradation and processing in a construct-specific manner. Hence, we
turned to a ^35S pulse-chase assay to fully characterize Tg degradation
and processing dynamics and found that lysate abundance of C1264R Tg
was not significantly changed at the 4 h timepoint with ML-240
treatment compared to DMSO (Figs. [283]7E and [284]EV4A,B). This
paralleled the previous results with VCP silencing and ML-240 treatment
under steady-state conditions in FRT cells. Impressively, C1264R Tg
secretion was increased tenfold at the 4 h timepoint with ML-240
treatment compared to DMSO with no significant change in C1264R Tg
degradation in the presence of ML-240 (Figs. [285]7E and [286]EV4A,B).
Conversely, for WT-Tg ML-240 treatment led to a significant increase in
lysate accumulation at 4 h compared to DMSO (Figs. [287]7F
and [288]EV4C,D). This increase in lysate abundance correlated with a
significant decrease in WT-Tg secretion from 67% to 4% in the presence
of ML-240 compared to DMSO, without altering WT degradation
(Figs. [289]7F and [290]EV4C,D).
To understand whether this rescue in secretion was uniquely linked to
VCP inhibition or could be more broadly attributed to blocking Tg
degradation, we tested the proteasomal inhibitor bortezomib, and
lysosomal inhibitor bafilomycin. Bafilomycin increased WT-Tg lysate
abundance, and bortezomib significantly increased A2234D lysate
abundance, consistent with a role of these degradation processes in Tg
PQC (Fig. [291]EV5A). When monitoring Tg-NLuc media abundance, neither
bafilomycin nor bortezomib significantly altered WT, A2234D, or C1264R
abundance, confirming that general inhibition of proteasomal or
lysosomal degradation does not rescue mutant Tg secretion
(Fig. [292]EV5B).
Figure EV5. TRIP of C1264R Tg-FT FRT cells with pharmacological VCP
inhibition.
[293]Figure EV5
[294]Open in a new tab
(A, B) Effect of pharmacologic inhibition of protein degradation in Tg
retention and secretion. HEK293T cells stably expressed Tg-NanoLuc
variants were treated with proteasomal degradation inhibitor Bortezomib
(10 µM) or lysosomal inhibitor Bafilomycin A1 (10 µM) for 8 h. Prior to
treatment, the media was exchanged to condition secreted Tg. Tg lysate
amounts (A) or media amounts (B) were then measured by the nano-glo
luciferase assay system. Data is normalized to DMSO condition and
represented as mean ± SEM. Statistical testing was performed using an
unpaired Student’s t test with Welch’s correction with P values as
indicated. N = 6–8 biological replicates as shown. (C, D) Assessment of
UPR upregulation with ML-240 treatment. (C) Western blot analysis to
assess phospho-eIF2α levels in of C1264R Tg-FT FRT cells treated with
VCP inhibitor ML-240 (10 µM), tunicamycin (1 µg/mL), or vehicle (0.1%
DMSO) for 2 h. Lysate samples were analyzed via immunoblotting. Data is
normalized to the mean C1264R Tg-FT abundance of tunicamycin-treated
samples. Data represented as mean ± SEM from N = 2 biological
replicates. (D) Activation of UPR markers monitored via qPCR in C1264R
Tg-FT FRT cells treated with ML-240 (10 μM) for 3 h. HSPA5 and ASNS
expression remained unchanged in C2164R Tg-FT FRT cells but led to a
significant decrease in WT Tg-FT FRT cells. Only DNAJB9 showed a
significant increase in transcript levels for both C1264R and WT Tg-FT
FRT cells. This suggest that ML-240 dependent rescue of C1264R Tg is
not due to global remodeling of the ER proteostasis network via UPR
activation. Data was first normalized to a GAPDH loading control
followed by normalization to median expression of DMSO-treated samples
and represented as mean ± SEM. Statistical testing performed using an
unpaired Student’s t test with Welch’s correction with P values as
indicated. N = 6 biological replicates. Primers for detection are
described in Dataset EV[295]7. (E) Viability analysis using Propidium
iodide control samples, WT & C164R Tg-FT FRT cells with ML-240
treatment. Cells were treated with DMSO or ML-240 (10 μM) for 4 h,
harvested, and stained with propidium iodide (1 μg/mL). FRT cells
permeabilized with 0.2% Triton were used as a positive staining
control. Unstained, non-permeabilized FRT cells were used as a negative
staining control.
Finally, we monitored the activation of the unfolded protein response
(UPR) in the presence of ML-240 in FRT cells expressing C1264R Tg-FT.
Phosphorylation of eIF2α, an activation marker for the PERK arm of the
UPR, was induced within 2 h of ML-240 treatment (Fig. [296]EV5C). We
further investigated the induction of UPR targets via qRT-PCR. HSPA5
and ASNS transcripts, markers of ATF6 and PERK UPR activation,
respectively, remained unchanged or slightly decreased after 3 h
treatment with ML-240 in C1264R Tg cells (Fig. [297]EV5D). Only DNAJB9,
a marker of the IRE1 arm of the UPR, showed a significant increase in
both WT-Tg and C2164R Tg-FRT cells (Fig. [298]EV5D). Moreover, ML-240
did not significantly alter cell viability after 3 h, as measured by
propidium iodide staining (Fig. [299]EV5E). Overall, these results
highlight that the short ML-240 treatment induces early UPR markers,
but the selective rescue of C1264R Tg secretion via ML-240 treatment is
unlikely the results of global remodeling of the ER PN due to UPR
activation.
Secretion rescue of C1264R Tg is associated with temporal remodeling of the
interactome
After identifying that VCP inhibition via ML-240 rescued C1264R Tg
secretion, we sought to use TRIP to capture PQC interaction changes
that correlated with increased secretion. TRIP was carried out in FRT
cells expressing C1264R Tg in the presence of ML-240 to monitor
temporal changes in PN interactions (Appendix Fig. S[300]8; Dataset
EV[301]4). Both glycan-processing and Hsp70-/90-chaperoning pathways
exhibited broad decreases in C1264R Tg interactions in ML-240 treated
samples compared to untreated samples (Fig. [302]7G,H). Particularly,
CALR, CANX and UGGT1 interactions tapered off more rapidly within 0.5 h
compared to untreated C1264R. In contrast, interactions with key
Hsp70-/90-chaperoning components remained relatively steady throughout
the chase period before peaking at the 3 h timepoint (Fig. [303]7I,J).
Conversely, C1264R interactions with chaperones HSPA5 and HSP90B1, and
co-chaperones DNAJB11, DNAJC10, and DNAJC3 decreased and mimicked the
WT-Tg temporal profile (Figs. [304]3D and [305]7J).
Interactions with disulfide/redox-processing components exhibited
milder but marked declines at intermediate timepoints with ML-240
treatment compared to untreated samples (Fig. [306]7K,L; Appendix Fig.
S[307]8). PDIA4 interactions remained much lower before peaking at the
3 h timepoint (Fig. [308]7L). Conversely, ERP29 and PDIA3 interactions
remained largely unchanged in the presence of ML-240 (Fig. [309]7L).
The most striking interaction changes occurred with proteasomal
degradation components, which remained steady until 1.5 h, but then
abruptly declined with ML-240 treatment at later timepoints
(Fig. [310]7M,N). This decline tracks with changes to the
glycan-processing machinery, highlighting that the coordination between
N-glycosylation and diverting Tg away from ERAD may be a key to the
rescue mechanism.
Discussion
The timing of protein–protein interactions implicated in cellular
processes and pathogenic states remains pivotal to our understanding of
disease mechanisms. Nonetheless, the temporal measurement of these
interactions has remained difficult and has proven to be a bottleneck
in elucidating the coordination of complex biological pathways such as
the PN. Here, we developed an approach to temporally resolve
protein–protein interactions implicated in PQC. Furthermore, we
complemented these data with a functional genomic screen to further
characterize and investigate the implications of these protein–protein
interactions on Tg processing. This combination of our novel TRIP
approach coupled with functional screening deconvoluted previously
established PQC mechanisms for Tg processing while also providing new
paradigms within PQC pathways that are critical for the secretion of
the prohormone.
TRIP has allowed the identification and resolution of unique temporal
changes in Tg interactions with glycan-processing components including
CANX, CALR, and UGGT1, while contrasting WT to mutant Tg variants.
These changes subsequently correlate with altered interactions with
ERAD components EDEM3, OS9, SEL1L, and VCP. Moreover, we identified a
broader scope of Tg interactors in thyroid tissue, including ER-phagy
receptors ATL3, CCPG1, RTN3, and the RTN3 adaptor protein PGRMC1.
Identification of these receptors establishes a direct link from Tg
processing to ER-phagy or ERLAD degradation mechanisms. In addition,
glycan-dependent and independent mechanisms have been established for
the degradation of ERAD-resistant ER clients (He et al, [311]2021;
Fregno et al, [312]2021). This overlap in degradation pathways may be
similarly at play in the case of Tg biogenesis and processing,
especially with transient disulfide-linked aggregation taking place
during nascent Tg folding and mutants forming difficult-to-degrade
aggregates (Kim et al, [313]1993; di Jeso et al, [314]2005; Menon et
al, [315]2007; di Jeso et al, [316]2014).
An important finding was that TRIP was capable of resolving subtle
temporal interaction changes, for example, with CALR, ERP29, ERP44,
P4HB, and VCP, which were otherwise masked in steady-state
interactomics data (Wright et al, [317]2021). Nonetheless, there are
some limitations within our TRIP methodology. TRIP relies on a
two-stage purification strategy which increases sample handling and
limits the amount of protein that can be subsequently enriched. Both
add inherent variability to the workflow. Furthermore, pulsed labeling
with unnatural amino acids has been shown to slow portions of protein
translation (Kiick et al, [318]2001; Dieterich et al, [319]2006; Bagert
et al, [320]2014; Lang and Chin, [321]2014). To address this, we
utilized a labeling time of 1 h which allows us to generate a large
enough labeled population of Tg-FT for TRIP analysis, but some early
interactions are likely missed within the TRIP workflow. In the case of
mutant Tg, performing the TRIP analysis for much longer chase periods
(6–8 h) may provide insightful details to the iterative binding process
of PN components that is thought to facilitate protein retention within
the secretory pathway. In addition, within our dataset, we noticed that
the temporal profiles for ribosomal and proteasomal subunits,
trafficking and lysosomal components are inherently difficult to
measure across experiments. To efficiently measure these components
temporally the TRIP methodology will require further optimization.
The functional implications of protein–protein interactions can be
difficult to deduce, especially in the case of PQC mechanisms
containing several layers of redundancy across stress response
pathways, paralogs, and multiple unique proteins sharing similar
functions (Wright and Plate, [322]2021; Bludau and Aebersold,
[323]2020; Karagöz et al, [324]2019; Braakman and Hebert, [325]2013).
This led us to establish a siRNA screening platform to complement our
TRIP data and broadly investigate the functional implications of PQC
components. With this assay, we found the trafficking factor NAPA was
heavily implicated in WT-Tg secretion. Most strikingly, we found that
VCP and TEX264 were implicated in C1264R processing, and siRNA
silencing of either led to rescue of C1264R secretion. Rescue of mutant
Tg trafficking from ER to Golgi, but not secretion, with
low-temperature correction had been documented previously (Kim et al,
[326]1996). Yet, this is the first study, to our knowledge, that
identified a restorative approach to mutant Tg secretion. To expound
upon this, our findings that pharmacological VCP inhibition selectively
rescues C1264R secretion compared to WT, and mutant Tg was selectively
enriched with TEX264 further corroborated the TRIP data establishing
that Tg mutants undergo differential interactions with degradation
pathways compared to WT Tg. Similar phenomena have been observed for
the partitioning and differential interactions of other prohormone
proteins, such as proinsulin and proopiomelanocortin, and
proteasome-resistant polymers of alpha1-antitrypsin Z-variant. RTN3 is
implicated in these differential interactions and is shown to be
selective for these prohormones compared to other ER-phagy receptors
(Chen et al, [327]2021; Cunningham et al, [328]2019; Fregno et al,
[329]2018). While A2234D and C1264R Tg were preferentially enriched
with TEX264 compared to WT, it remains unclear what other accessory
proteins may be necessary for the recognition of TEX264 clients (Chino
et al, [330]2019; An et al, [331]2019). Furthermore, TEX264 function in
both protein degradation and DNA damage repair further complicates
siRNA-based investigations (Fielden et al, [332]2022). Further
investigation is needed to fully elucidate (1) if Tg degradation takes
place via ER-phagy and (2) by which mechanisms this targeting is
mediated.
As we discovered that pharmacological VCP inhibition with ML-240 can
rescue C1264R Tg secretion yet is detrimental for WT-Tg processing, it
is unclear whether VCP may exhibit distinct functions for WT and mutant
Tg PQC. Finally, as ML-240 is shown to block both the proteasomal and
autophagic functions of VCP it is unclear which of these pathways may
be playing a role in the rescue of C1264R, or detrimental WT processing
(Chou et al, [333]2013, [334]2014).
We used our TRIP method to monitor the changes in interactions
associated with the rescue of C1264R Tg with pharmacological VCP
inhibition. We found that this rescue is correlated with broad temporal
changes in interactions across glycan processing, Hsp70-/90-chaperoning
and proteasomal degradation pathways, while exhibiting more discrete
changes with select disulfide/redox-processing components. Mapping
these temporal changes in response to pharmacological VCP inhibition
and C1264R rescue highlights the capabilities of TRIP to not only
resolve protein–protein interactions across disease states but also
identify compensatory mechanisms that may take place with drug
treatment or other modulating techniques like gene inhibition or
activation used to study or treat disease states. Consequently, TRIP
should find broad applicability for delineating the proteostasis
deficiencies that give rise to diverse protein misfolding diseases and
elucidating other cellular interactome dynamics.
Methods
Reagents and tools table
Reagent/resource Reference or source Identifier or catalog number
Antibodies
KDEL Enzo Life Sciences ADI-SPA-827
M2-FLAG Sigma-Aldrich F1804
PDIA4 Protein Tech 14712-1-AP
TG Proteintech 21714-1-AP
GAPDH GeneTex GTX627408
Tubulin-Rhodamine Bio-Rad 12004165
Goat anti-mouse Star-bright700 Bio-Rad 12004158
Goat anti-rabbit IRDye800 LI-COR 926-32211
G1 Anti-DYKDDDDK affinity resin GenScript L00432
High-Capacity Streptavidin agarose resin Pierce 20357
Chemical compounds
NMS-873 Cayman Chemical 17674
CB-5083 Cayman Chemical 19311
ML-240 Cayman Chemical 17373
Dithiobis(succinimidyl propionate) (DSP) Thermo Scientific 22585
2-(4-((Bis((1-(tert-butyl)-1H-1,2,3-triazol-4-yl)methyl)amino)methyl)-1
H-1,2,3-triazol-1-yl)acetic acid (BTTAA) Click Chemistry Tools 1236
Carboxytetramethylrhodamine (TAMRA)-Azide-Polyethylene Glycol
(PEG)-Desthiobiotin BroadPharm BP-22475
DharmaFECT1 Transfection Reagent Dharmacon T-2001
Lipofectamine 3000 Invitrogen L3000
Commercial assays
Nano-Glo Luciferase Assay System Promega N1110
Cell lines
Human HEK293 Flp-In Thermo Fisher [335]R75007
Fischer Rat Thyroid (FRT) Flp-In (Sabusap et al, [336]2016) N/A
Recombinant DNA
Tg-FLAG - pcDNA3.1 + /C-(K)-DYK Genscript OHu20241
pcDNA5/FRT (Sabusap et al, [337]2016) N/A
pOG44 Thermo Fisher V600520
pcDNA5/FRT-Tg-FLAG This study N/A
pcDNA5/FRT-Tg-NLuc This study N/A
pMRX-INU-TEX264-FLAG Addgene 128258
pMRX-INU-SEC62-FLAG Addgene 128263
Oligonucleotides
PCR primers This study Dataset EV[338]7
Software
Bio-Rad Image Lab Bio-Rad
[339]https://www.bio-rad.com/en-us/product/image-lab-software
Prism 9 GraphPad
[340]https://www.graphpad.com/scientific-software/prism/
Proteome Discoverer 2.4 Thermo Fisher [341]https://www.thermofisher.com
CFX Maestro Bio-Rad
[342]https://www.bio-rad.com/en-us/product/cfx-maestro-software-for-cfx
-real-time-pcr-instruments
FlowJo BD Biosciences
[343]https://www.flowjo.com/solutions/flowjo/downloads
[344]Open in a new tab
All unique/stable reagents generated, including plasmids and cell
lines, are available from the corresponding author
(lars.plate@vanderbilt.edu) with a complete Materials Transfer
Agreement.
Methods and protocols
Plasmid production and antibodies
Tg-FLAG in pcDNA3.1 + /C-(K)-DYK plasmid was purchased from Genscript
(Clone ID OHu20241). The Tg-FLAG gene was then amplified and assembled
with an empty pcDNA5/FRT expression vector using a HiFi DNA assembly
kit (New England BioLabs, E2621). This plasmid then underwent
site-directed mutagenesis to produce pcDNA5-C1264R-Tg-FLAG, and
pcDNA5-A2234D-Tg-FLAG plasmids.
An oligonucleotide fragment encoding Nanoluciferase (NLuc) was ordered
from Genewiz. To generate pcDNA5/FRT-Tg-NLuc plasmids the Tg gene was
amplified from the pcDNA5/FRT-Tg-FLAG plasmid and assembled with the
NLuc fragment using a HiFi DNA assembly kit (New England BioLabs). To
generate the respective mutant construct plasmids for A2234D and C1264R
Tg, site-directed mutagenesis was performed using a Q5 polymerase (New
England BioLabs). All oligonucleotide sequences used for site-directed
mutagenesis can be found in Dataset EV[345]7.
Cell line engineering
FRT cells were cultured in Ham’s F12, Coon’s Modification (F12) media
(Sigma, cat. No. F6636) supplemented with 5% fetal bovine serum (FBS),
and 1% penicillin (10,000 U)/streptomycin (10,000 μg/mL). All cell
lines were tested monthly to ensure they were free of mycoplasma
contamination. To generate FRT flp-Tg-FT cells, 5 × 10^5 cells cultured
for 1 day were cotransfected with 2.25 μg of flp recombinase pOG44
plasmid and 0.25 μg of FT-Tg pcDNA5 plasmid using Lipofectamine 3000.
Cells were then cultured in media containing Hygromycin B (100 μg/mL)
to select site-specific recombinants. Resistant clonal lines were
sorted into single-cell colonies using flow cytometry and screened for
FT-Tg expression (Appendix Fig. S[346]1).
To generate HEK293 flp-Tg-NLuc cells, 5 × 10^5 cells cultured for 1 day
were cotransfected with 2.25 μg of flp recombinase pOG44 plasmid and
0.25 μg of NLuc-Tg pcDNA5 plasmid using Lipofectamine 3000. Cells were
then cultured in media containing Hygromycin B (100 μg/mL) to select
site-specific recombinants. Polyclonal lines were screened for Tg-NLuc
expression and furimazine substrate turnover (Appendix Fig. EV[347]5).
Time-resolved interactome profiling
Fully confluent 15-cm tissue culture plates of FRT cells were used.
Cells were washed with phosphate-buffered saline (PBS) and depleted of
methionine by incubating with methionine-free Dulbecco’s Modified Eagle
Medium (DMEM) supplemented with 5% dialyzed fetal bovine serum (FBS),
1% l-glutamine (2 mM final concentration), 1% l-cysteine (200 μM final
concentration), and 1% penicillin (10,000 U)/streptomycin
(10,000 μg/mL) for 30 min. Cells were then pulse-labeled with
Hpg-enriched DMEM supplemented with 1% Hpg (200 μM final
concentration), 5% dialyzed FBS, 1% l-glutamine (200 mM final
concentration), 1% l-cysteine (200 μM), and 1% penicillin (10,000
U)/streptomycin (10,000 μg/mL) for 1 h. After pulse labeling, cells
were washed with F12 media containing tenfold excess methionine (2 mM
final concentration). Cells were then cultured in normal F12 media
supplemented with 5% FBS and chased for the specified timepoints. Cells
were harvested by washing with PBS and then cross-linked with 0.5 mM
DSP in PBS at 37 °C for 10 min. Cross-linking was quenched by the
addition of 100 mM Tris pH 7.5 at 37 °C for 5 min. Lysates were
prepared by lysing in Radioimmunoprecipitation assay (RIPA) buffer
(50 mM Tris pH 7.5, 150 mM NaCl, 0.1% SDS, 1% Triton X-100, 0.5%
deoxycholate) with protease inhibitor cocktail (Roche, 4693159001).
Protein concentration was normalized to 1 mg/mL using a BCA assay
(Thermo Scientific, 23225), and cell lysates underwent click reactions
with the addition of 0.8 mM copper sulfate (diluted from a 20 mM stock
in water), 1.6 mM BTTAA (diluted from a 40 mM stock in water), 5 mM
sodium ascorbate (diluted from a 100 mM stock in water), and 100 μM
TAMRA-Azide-PEG-Desthiobiotin ligand (diluted from a 5 mM stock in
DMSO). Samples were allowed to react at 37 °C for 1 h while shaking at
750 rpm. Cell lysates were then precleared on 4B sepharose beads
(Sigma, 4B200) at 4 °C for 1 h while rocking. Precleared lysates were
immunoprecipitated with G1 Anti-DYKDDDDK affinity resin overnight at
4 °C while rocking. The resin was washed four times with RIPA buffer,
and proteins were eluted twice in 100 μL immunoprecipitation elution
buffer (2% SDS in PBS) by heating at 95 °C for 5 min. Eluted samples
from FLAG immunoprecipitations were then diluted with PBS to reduce the
final SDS concentration to 0.2%. The solutions then underwent
streptavidin enrichment with high-capacity streptavidin agarose resin
(Pierce, 20359) for 2 h at room temperature while rotating. The resin
was then washed with 1 mL each of 1% SDS, 4 M Urea, 1 M sodium
chloride, followed by a final wash with 1% SDS (all wash buffers
dissolved in PBS). The resin was frozen overnight at −80 °C and samples
were then eluted twice with 100 μL biotin elution buffer (50 mM Biotin
in 1% SDS in PBS) by heating at 37 °C and shaking at 750 rpm for 1 h.
Eluted streptavidin enrichment samples were precipitated in
methanol/chloroform, washed three times with methanol, and air-dried.
Protein pellets were then resuspended in 3 μL 1% Rapigest SF Surfactant
(Waters, 186002122) followed by the addition of 10 μL of 50 mM HEPES pH
8.0, and 34.5 μL of water. Samples were reduced with 5 mM
tris(2-carboxyethyl)phosphine (TCEP) (Sigma, 75259) at room temperature
for 30 min and alkylated with 10 mM iodoacetimide (Sigma, I6125) in the
dark at room temperature for 30 min. Proteins were digested with 0.5 μg
of trypsin/Lys-C protease mix (Pierce, A40007) by incubating for 16-18
h at 37 °C and shaking at 750 rpm. Peptides were reacted with TMTpro
16plex reagents (Thermo Fisher, 44520) in 40% v/v acetonitrile and
incubated for 1 h at room temperature. Reactions were quenched by the
addition of ammonium bicarbonate (0.4% w/v final concentration) and
incubated for 1 h at room temperature. TMT-labeled samples were then
pooled and acidified with 5% formic acid (Fisher, A117, v/v). Samples
were concentrated using a speedvac and resuspended in buffer A (97%
water, 2.9% acetonitrile, and 0.1% formic acid, v/v/v). Cleaved
Rapigest SF surfactant was removed by centrifugation for 30 min at
21,100 × g.
For TRIP analysis coupled with ML-240 treatment, C1264R Tg-FRT cells
were processed as described above with ML-240 (10 μM) supplemented in
Hpg pulse media and throughout the chase period.
Liquid chromatography–tandem mass spectrometry
MudPIT microcolumns were prepared as previously described (Fonslow et
al, [348]2012). Peptide samples were directly loaded onto the columns
using a high-pressure chamber. Samples were then desalted for 30 min
with buffer A (97% water, 2.9% acetonitrile, 0.1% formic acid v/v/v).
LC-MS/MS analysis was performed using an Exploris480 (Thermo Fisher)
mass spectrometer equipped with an Ultimate3000 RSLCnano system (Thermo
Fisher). MudPIT experiments were performed with 10 μL sequential
injections of 0, 10, 30, 60, and 100% buffer C (500 mM ammonium acetate
in buffer A), followed by a final injection of 90% buffer C with 10%
buffer B (99.9% acetonitrile, 0.1% formic acid v/v) and each step
followed by a 140 min gradient from 4 to 80% B with a flow rate of
500 nL/minute on a 20 cm fused silica microcapillary column (ID 100 μm)
ending with a laser-pulled tip filled with Aqua C18, 3 μm, 125 Å resin
(Phenomenex). Electrospray ionization (ESI) was performed directly from
the analytical column by applying a voltage of 2.2 kV with an inlet
capillary temperature of 275 °C. Data-dependent acquisition of mass
spectra was carried out by performing a full scan from 400 to 1600 m/z
at a resolution of 120,000. Top-speed data acquisition was used for
acquiring MS/MS spectra using a cycle time of 3 s, with a normalized
collision energy of 32, 0.4 m/z isolation window, automatic maximum
injection time, and 100% normalized AGC target, at a resolution of
45,000 and a defined first mass (m/z) starting at 110. Peptide
identification and TMT-based protein quantification was carried out
using Proteome Discoverer 2.4. MS/MS spectra were extracted from Thermo
Xcalibur.raw file format and searched using SEQUEST against a Uniprot
rat proteome database supplemented with the human thyroglobulin gene
(accessed 03/2014 and containing 28,860 entries). The database was
curated to remove redundant protein and splice-isoforms. Searches were
carried out using a decoy database of reversed peptide sequences and
the following parameters: 20 ppm peptide precursor tolerance, 0.02-Da
fragment mass tolerance, minimum peptide length of 6 amino acids,
trypsin cleavage with a maximum of two missed cleavages, dynamic
methionine modification of +15.995 Da (oxidation), dynamic protein
N-terminus +42.011 Da (acetylation), −131.040 (methionine loss),
−89.030 (methionine loss + acetylation), static cysteine modification
of +57.0215 Da (carbamidomethylation), and static peptide N-terminal
and lysine modifications of +304.2071 Da (TMTpro 16plex).
Immunoblotting and SDS-PAGE
Cell lysates were prepared by lysing in RIPA buffer with protease
inhibitor cocktail (Roche), and protein concentrations were normalized
using a BCA assay (Thermo Scientific). Lysates were then denatured with
1× Laemmli buffer + 100 mM DTT and heated at 95 °C for 5 min before
being separated by SDS-PAGE. Samples were transferred onto
poly-vinylidene difluoride (PVDF) membranes (Millipore, IPFL00010) for
immunoblotting and blocked using 5% nonfat dry milk dissolved in
tris-buffered saline with 0.1% Tween-20 (Fisher, BP337-100) (TBS-T).
Primary antibodies were incubated either at room temperature for 2 h,
or overnight at 4 °C. Membranes were then washed three times with TBS-T
and incubated with secondary antibody in 5% nonfat dry milk dissolved
in TBS-T either at room temperature for 1 h or overnight at 4 °C.
Membranes were washed three times with TBS-T and then imaged using a
ChemiDoc MP Imaging System (Bio-Rad). Primary antibodies were acquired
from commercial sources and used at the indicated dilutions in
immunoblotting buffer (5% bovine serum albumin (BSA) in Tris-buffered
saline pH 7.5, 0.1% Tween-20, and 0.1% sodium azide): KDEL (1:1000), M2
anti-FLAG (1:1000), PDIA4 (1:1000), thyroglobulin (1:1000).
Tubulin-Rhodamine primary antibody was obtained from commercial sources
and used at 1:10000 dilution in 5% milk in Tris-buffered saline pH 7.5,
0.1% Tween-20 (TBS-T). Secondary antibodies were obtained from
commercial sources and used at the indicated dilutions in 5% milk in
TBS-T: Goat anti-mouse Star-bright700 (1:10,000), Goat anti-rabbit
IRDye800 (1:10,000).
qRT-PCR
RNA was prepared from cell pellets using the Quick-RNA miniprep kit
(Zymo Research). cDNA was synthesized from 500 ng total cellular RNA
using random hexamer primer (IDT), oligo-dT primer (IDT), and M-MLV
reverse transcriptase (Promega). qPCR analysis was performed using iTaq
Universal SYBR Green Supermix (Bio-Rad) combined with primers for genes
of interest and reactions were run in 96-well plates on a Bio-Rad CFX
qPCR instrument. Data analysis was then carried out in CFX Maestro
(Bio-Rad). All oligonucleotide sequences used for qRT-PCR can be found
in Dataset EV[349]7.
siRNA screening assay
Proteins previously identified as Tg interactors, and other key
proteostasis network components were selected and targeted using an
siGENOME SMARTpool siRNA library (Dharmacon) (Wright et al, [350]2021).
HEK293 cells stably expressing Tg-NLuc constructs were seeded into
96-well plates at 2.5 × 10^4 cells/well and transfected using
DharmaFECT1 following the DharmaFECT1 protocol (Dharmacon) with a 25 nM
siRNA concentration. Approximately 36 h after transfection, cells were
washed with PBS and replenished with fresh DMEM media. After 4 h,
Tg-NLuc abundance in the lysate and media were measured using the
nano-glo luciferase assay system according to the manufacturer protocol
(Promega). Four controls were included for the experiments, including a
non-targeting siRNA control, a siGLO fluorescent control to monitor
transfection efficiency, a vehicle control containing transfection
reagents but lacking any siRNAs, and a lethal TOX control. Data was
median normalized across individual 96-well plates (Chung et al,
[351]2008). Data represents two independent experiments for WT-NLuc and
A2234D-NLuc, and three independent experiments for C1264R NLuc. Cutoff
criteria for hits were set to those genes that increased or decreased
Tg-NLuc abundance in lysate or media by 3σ.
FRT siRNA validation studies
For siRNA silencing follow-up experiments, Tg-FRT cells were seeded
into six-well dishes at 6.0 × 10^5 cells/well transfected using
DharmaFECT1 following the DharmaFECT protocol (Horizon) with a 25 nM
siRNA concentration. For NAPA complementation experiments, siRNA
(25 nM) and corresponding plasmid (0.83 µg per six-well dish) were
cotransfected with DharmaFECT Duo (Horizon) Approximately 36 h after
transfection, cells were washed with PBS and harvested for qRT-PCR for
western blotting. For qRT-PCR siRNA target transcript levels were
normalized to a GAPDH loading control to monitor siRNA silencing
efficiency. For immunoblot analysis ~36 h after transfection cells were
washed with PBS and replenished with fresh DMEM media. After 4 h in the
case of WT Tg and 8 h in the case of C1264R or A2234D Tg, cells and
media samples were harvested. Cells were lysed with 1 mL of RIPA with
protease inhibitor cocktail (Roche), and lysate and media samples were
subjected to immunoprecipitation with G1 Anti-DYKDDDDK affinity resin
overnight at 4 °C while rocking. After three washes with RIPA buffer,
protein samples were eluted with 3× Laemmli buffer with 100 mM DTT
heating at 95 °C for 5 min. Immunoblot quantification was performed
using Image Lab Software (Bio-Rad).
^35S pulse-chase assay
Confluent six-well dishes of FRT cells (approximately 1 × 10^6/well)
were metabolically labeled in DMEM depleted of methionine and cysteine
and supplemented with EasyTag ^35S Protein Labeling Mix (Perkin Elmer,
NEG772007MC), glutamine, penicillin/streptomycin, and 10% dialyzed FBS
at 37 °C for 30 min. Afterward, cells were washed twice with F12 media
containing 10× methionine and cysteine, followed by a burn-off period
of 10 min in normal F12 media. Cells were then chased for the
respective time periods with normal F12 media, lysed with 500 μL of
RIPA buffer with protease inhibitor cocktail (Roche) and 10 mM DTT.
Cell lysates were diluted with 500 μL of RIPA buffer with protease
inhibitor cocktail (Roche) and subjected to immunoprecipitation with G1
anti-DYKDDDDK affinity resin overnight at 4 °C. After three washes with
RIPA buffer, protein samples were eluted with 3× Laemmli buffer with
100 mM DTT heating at 95 °C for 5 min. Eluted samples were then
separated by SDS-PAGE, and gels were dried and exposed on a storage
phosphor screen. Radioactive band intensity was then measured using a
Typhoon Trio Imager (GE Healthcare) and quantified by densitometry in
Image Lab (Bio-Rad).
For pulse-chase analysis coupled with ML-240 treatment, cells were
pre-treated with vehicle (0.1% DMSO) or ML-240 (10 μM) for 15 min prior
to ^35S pulse labeling. Vehicle (0.1% DMSO) or ML-240 (10 μM) treatment
was then maintained throughout the pulse labeling and chase period.
VCP pharmacological inhibition studies
For the follow-up experiment with VCP inhibitors, confluent six-well
dishes of C1264R Tg-FT FRT cells were washed with PBS and fresh F12
media supplemented with ML-240 (10 μM), CB-5083 (5 μM), NMS-873
(10 μM), or vehicle (0.1% DMSO) was added and incubated for 8 h. For WT
Tg-FT cells, confluent six-well dishes were similarly used, washed with
PBS, and fresh F12 media supplemented with ML-240 (10 μM) or vehicle
(0.1% DMSO) was added and incubated for 4 or 8 h. Cells were lysed with
1 mL of RIPA with protease inhibitor cocktail (Roche), and lysate and
media samples were subjected to immunoprecipitation with G1
Anti-DYKDDDDK affinity resin overnight at 4 °C while rocking. After
three washes with RIPA buffer, protein samples were eluted with 3×
Laemmli buffer with 100 mM DTT heating at 95 °C for 5 min. Immunoblot
quantification was performed using Image Lab Software (Bio-Rad).
Viability with ML-240 was monitored using propidium iodide staining.
Briefly, cells were treated with either ML-240 (10 μM) or vehicle (0.1%
DMSO) for 4 h before being harvested and stained with propidium iodide
(1 μg/mL) at room temperature for 15 min in the dark. Cells
permeabilized with 0.2% Triton were used as a positive staining
control. Staining intensity of cells was then analyzed by flow
cytometry and data analysis was carried out in FlowJo (BD Biosciences).
TEX264 & Tg-NLuc co-immunoprecipitation studies
HEK293 flp-Tg-NLuc cells were cultured at 1.0 × 10^5 cells/well in
12-well tissue culture dishes for 1 day and transfected with either a
fluorescent control, C-terminal FLAG-tag TEX264, or C-terminal FLAG-tag
SEC62 using a calcium phosphate method. Confluent plates were harvested
by lysing with 300 μL of TNI buffer (50 mM Tris pH 7.5, 150 mM NaCl,
0.5% IGEPAL CA-630 (Sigma-Aldrich)) and protease inhibitor (Roche).
Lysates were sonicated for 10 min at room temperature and normalized
using a BCA Assay (Thermo Scientific). Cell lysates were then
precleared on 4B sepharose beads (Sigma, 4B200) at 4 °C for 1 h while
rocking, then immunoprecipitated with G1 Anti-DYKDDDDK affinity resin
overnight at 4 °C while rocking. The resin was washed four times with
TNI buffer and resuspended in 250 μL of TNI buffer. In total, 50 μL
aliquots were taken and measured using the nano-glo luciferase assay
system according to the manufacturer protocol (Promega). For
differential enrichment, statistical analysis was performed in Prism 9
(GraphPad) using a one-way ANOVA with post hoc Tukey’s multiple testing
corrections.
For western blot analysis HEK293 flp-Tg-NLuc (1.0 × 10^6 cells/dish in
10-cm tissue culture dishes) were cultured for 1 day and transfected
with either a fluorescent control, TEX264, or SEC62 using a calcium
phosphate method. Cells were collected and lysed using 300 μL of TNI
buffer with protease inhibitor (Roche) sonicated at room temperature
for 10 min. Lysates were normalized, precleared, and immunoprecpitated
as described above. Proteins were eluted using 6× Laemmli buffer +
100 mM DTT, separated by SDS-PAGE and transferred to PVDF membrane
(Millipore) for immunoblotting.
Mass spectrometry interactomics and TMT quantification data analysis
To identify Tg interactors (−) biotin pulldown vs (−) Hpg samples were
processed using the DEP pipeline (Zhang et al, [352]2018). Enriched
proteins were determined based on those with a log2 fold change of 2σ
and Benjamini–Hochberg adjusted P value (false discovery rate) of 0.05.
For pathway enrichment analysis of identified proteins, EnrichR was
used and GO Cellular Component and Molecular Function 2018 terms were
used to differentiate secretory pathway associated proteins from
background (Chen et al, [353]2013). For time-resolved analysis, data
were processed in R with custom scripts. Briefly, TMT abundances across
chase samples were normalized to Tg TMT abundance as described
previously and compared to (-) Hpg samples for enrichment analysis
(Wright et al, [354]2021). For relative enrichment analysis, the means
of log2 interaction differences were scaled to values from 0 to 1,
where a value of 1 represented the timepoint at which the enrichment
reached the maximum, and 0 represented the background intensity in the
(−) Hpg channel. Negative log2 enrichment values were set to 0 as the
enrichment fell below the background.
Inconsistencies in the quantification of Tg bait protein were observed
for replicate 5 of WT Tg Hpg-chase samples, likely due to sample loss
during the enrichment. Hence, this replicate was excluded from further
time-resolved analysis. All analysis scripts are available as described
in “Data availability”.
Supplementary information
[355]Appendix^ (4.4MB, pdf)
[356]Peer Review File^ (764.3KB, pdf)
[357]Dataset EV1^ (1.3MB, xlsx)
[358]Dataset EV2^ (97.2KB, xlsx)
[359]Dataset EV3^ (257.8KB, xlsx)
[360]Dataset EV4^ (117.8KB, xlsx)
[361]Dataset EV5^ (85.6KB, xlsx)
[362]Dataset EV6^ (21.2KB, xlsx)
[363]Dataset EV7^ (11.2KB, xlsx)
[364]Dataset EV8^ (11.9KB, xlsx)
[365]Source data Fig. 1^ (15.5MB, zip)
[366]Source data Fig. 2^ (9MB, zip)
[367]Source data Fig. 4^ (90.9KB, zip)
[368]Source data Fig. 5^ (83.1MB, zip)
[369]Source data Fig. 6^ (1,023.7KB, zip)
[370]Source data Fig. 7^ (155.1MB, zip)
[371]Expanded View Figures^ (921.2KB, pdf)
Acknowledgements