Transcriptomic analysis of SEL1L and HRD1 knockout cell lines reveals multifaceted roles of SEL1L beyond the ER quality control

doi:10.21203/rs.3.rs-4484044/v1

Download PDF

Article

Transcriptomic analysis of SEL1L and HRD1 knockout cell lines reveals multifaceted roles of SEL1L beyond the ER quality control

https://doi.org/10.21203/rs.3.rs-4484044/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The endoplasmic reticulum (ER) orchestrates major cellular processes, including protein synthesis, folding, assembly and degradation, to maintain cellular proteostasis. Central to these processes are highly stringent quality control machineries like the ER-associated protein degradation (ERAD). Key players in ERAD include HRD1 and SEL1L, which target misfolded proteins for ubiquitination and facilitate their retro-translocation to the cytosol. Bi-allelic loss-of-function of HRD1 and SEL1L is considered lethal, with hypomorphic variants linked to human diseases, including neurodevelopmental disorders. Despite their well-known roles, a comprehensive transcriptomic characterization of their bi-allelic loss has been lacking.

In this study, we employed CRISPR/Cas9 to generate bi-allelic HRD1-KO and SEL1L-KO HEK293 cell models. Through differential gene expression analysis and co-expression network construction, we identified hub genes and novel regulatory networks. HRD1-KO cells displayed enrichment solely in ER-related processes, suggesting its specific role in ER protein quality control. Conversely, SEL1L-KO cells exhibited a broader impact, affecting mitochondrial function, ERAD-ribosomal quality control interactions, ER-Golgi transport, and Wnt signaling pathway.

These results highlight the distinct roles of HRD1 and SEL1L in ERAD. By unraveling their whole transcriptome impact, our study sheds light on their potential involvement in diverse cellular processes, potentially enhancing our understanding of their cellular processes and disease mechanisms.

ERAD

SEL1L

HRD1

Transcriptomics

WGCNA

hub genes

The endoplasmic reticulum (ER) is a dynamic organelle playing major roles in cellular proteostasis including the orchestration of secretory proteins synthesis, folding, assembly and degradation (Nam and Jeon, 2019). While the ribosomes are attached to the ER, they translate coding mRNAs into polypeptides, and target secretory proteins into the ER lumen for folding and assembly. This process is assisted and controlled by membrane-bound and soluble molecular chaperones, which facilitate their folding and prevent aggregation (Beissinger and Buchner, 1998). In addition to chaperoning processes, ER-targeted proteins often undergo post-translational modifications such as protein glycosylation and GPI anchoring influencing their folding, trafficking, stability and function (Ellgaard et al., 2016). Crucially, the ER harbors a highly stringent quality control systems at the entry point to the secretory pathway. Proteins failing to fold or assemble correctly for any reason usually accumulate in the ER and consequently activate the different arms of the unfolded protein response (UPR). This includes the activation of protein degradation via various processes such as the ER-associated degradation (ERAD). ERAD is a multi-step process that involves the recognition and subsequent retrotranslocation of misfolded and orphaned proteins back into the cytosol for ubiquitination and proteasomal degradation, or to the lysosomes for lysosomal associated degradation. Malfunctioning of this quality control mechanism has been implicated in numerous human diseases. The processes associated with ERAD involve various protein and proteins complexes including the SEL1L-HRD1 retrotranslocation-ubiquitination complex (Qi et al., 2017).

Hydroxymethyl glutaryl-coenzyme A reductase degradation protein 1 (HRD1), a mammalian homolog of yeast Hrd1p, is an ER membrane bound protein. It plays a pivotal role in the ER quality control mechanism by ubiquitinating misfolded proteins, thereby tagging them for proteasomal degradation. Dysregulation in HRD1 gene has been linked to multiple diseases, including cancers, autoimmune diseases, and neurological disorders (Yang et al., 2014; Omura et al., 2018; Liu et al., 2020a; Fan et al., 2021; Badawi et al., 2023; Wang et al., 2023b). Intriguingly, germline deletion of the HRD1 gene in mice leads to in-utero lethality and is associated with arthropathy (Amano et al., 2003). On the other hand, SEL1L, an E3 ubiquitin ligase adaptor or scaffold protein, interacts with HRD1 in the ER membrane, facilitating the retrotranslocation of misfolded proteins for degradation. SEL1L and HRD1 form a dynamic complex that interacts with various ERAD components. Deletion of SEL1L in mice also results in embryonic lethality and pancreatic atrophy confirming its essential developmental functions (Sun et al., 2014). Dysregulation of SEL1L expression has been implicated in various diseases, including neurodevelopmental disorders, fibrosis (Podolsky et al., 2024), and cancers (Yue et al., 2017). Additionally, recent studies have proposed the involvement of SEL1L hypomorphic genetic variants in various diseases, mainly neurodevelopmental disorders (Wang et al., 2024; Weis et al., 2024). Several studies have investigated tissue-specific SEL1L and HRD1 animal KO models (Bhattacharya et al., 2018; Yang et al., 2018; Chen et al., 2020; Yao et al., 2022) highlighting their role in several processes and diseases other than the ERAD, including mitochondrial functions (Zhou et al., 2020), metabolism (Bhattacharya et al., 2018), pancreatic beta cells failure (Wu et al., 2020), B-cells and T-cells immunity (Kong et al., 2016; Xu et al., 2016; Yang et al., 2018) and involvement in neurodelopmental disorders (Weis et al., 2024).

Despite the well-established roles of HRD1 and SEL1L in the ERAD machinery, a comprehensive transcriptomic characterization of their KO cells is lacking in the literature. To delve deeply into the effect of these ERAD components deficiency on the transcriptome and the potential mechanisms behind their associated diseases, CRISPR/Cas9 gene editing tools were used in human embryonic kidney cells (HEK293) to generate HRD1-KO and SEL1L-KO cellular models (Kizhakkedath et al., 2018; Gariballa et al., 2022). For this manuscript, we have employed various techniques to identify differentially expressed genes, construct co-expression networks, and pinpoint hub genes providing novel insights into the regulatory networks and functional significance of HRD1 and SEL1L.

1.1. Summary of Transcriptome sequencing, mapping, and gene expression level data

To investigate the role of HRD1 and SEL1L, HEK293 cells were used to generate these ERAD components knockout models using CRISPR/Cas9 genome editing tool. Gene deletion was confirmed using Sanger sequencing. Biallelic deletion is also confirmed by diminished protein expression of HRD1 and SEL1L by western blotting analysis in comparison to their parental HEK293 cells performed previously by Gariballa et al, and Kizhakkedath et al, respectively (Kizhakkedath et al., 2018; Gariballa et al., 2022).

RNA Sequencing results show that the total raw reads were distributed between 39,431,088 and 47,377,380 reads across all samples. the percentage of the reads that have been mapped to reference genome has ranged between 88% and 96%. The majority of these reads were aligned to the exonic regions with percentages ranging between 90% and 94.4%. Only few reads, ranging between 1.07% and 1.97% were mapped to intergenic regions, presumably due to weak annotations to the reference genome. Additionally, GC distribution across the samples has ranged between 49.1% and 50.18%. Gene expression level is calculated using a negative binomial generalized linear model in Deseq2 R package.

1.2. Transcriptome profiling of HRD1 KO HEK293 cells

Principal component analysis (PCA) of HRD1-KO and control HEK293 cells has generated two biological clusters based on the expression data (Fig. 1A), where HRD1-KO form a separate cluster compared to control HEK293 cells. This variability in the data is explained by the first two components (82%). As shown in Fig. 1B, knocking out HRD1 in HEK293 cells has induced a significant change in expression of 1839/17205 genes, including 994 upregulated and 789 downregulated genes with fold change > 1.5 and statistical significance of adjusted p-value < 0.05.

Gene ontology enrichment analysis of all differentially expressed genes (DEGs) show that a variety of biological processes have been significantly enriched (Fig. 1C). The highly enriched processes can be categorized into three major categories mainly related to response to unfolded protein, ER stress and transport and organelle organization including ER to Golgi vesicle mediated transport and Golgi organization. Moreover, KEGG pathway enrichment analysis of the 1839 genes has yielded only 2 KEGG pathways of adjusted p-value < 0.05, hsa04141 (Protein processing in endoplasmic reticulum) and hsa01240 (biosynthesis of cofactors). Digging deeper into the genes involved in the protein processing in ER pathway, we find that out of the 46 differentially expressed genes and enriched in this pathway, only three were downregulated, displayed in blue boxes in Fig. 2 and all the remaining genes that are displayed in red boxes were upregulated in response to HRD1 knockout in HEK293 cells. These dysregulated genes were involved in various processes including chaperoning processes, like the luminal chaperones Hsp40 and BiP, and the lectin-like chaperones including calnexin (CNX) and calreticulin (CRT). Furthermore, in addition to UPR-related genes like ATF6, PERK and XBP, ERAD related components including SEL1L, HERP, Derlin and EDEM have been dysregulated. Moreover, the only genes that were downregulated were proapoptotic genes (Bax/Bak) and cytoplasmic components of the ubiquitin ligase complex (FBP and RBX1).

1.3. Characterization of SEL1L knockout HEK293 cells

Similarly, the first two principal components explain 90% of the variation in SEL1L knockout data, where SEL1L and control HEK293 cells form separate clusters along the first principal component. Considering adjusted p-value of 0.05, we get 3728 differentially expressed genes. In an attempt to reduce this number and get more specific dysregulated genes in response to SEL1L knockout, we have set the adjusted p-value to 0.01. As a result, the transcriptome profiles of SEL1L KO cells have revealed 3052/16979 differentially expressed genes, where 1694 were upregulated and 1358 were down regulated using adjusted p-value of 0.01 and fold change of 1.5 as a cutoff criterion. To validate RNA-seq results, several upregulated and downregulated DEGs, with varying fold change (NSUN7, CTSZ, HERPUD1, DAG1, HRD1, SFRP2 and PXMP4) were selected, and their mRNA expression level was assessed by real-time quantitative PCR assay (Fig. 3).

Among SEL1L-KO DEGs, as shown in Fig. 4A, 910 (of which 536 are upregulated) genes were also dysregulated in response to HRD1 knockout, while 2142 were unique to SEL1L KO cells. Categorization of the commonly regulated genes through enrichment bioinformatic analysis identified 12 biological processes that were mainly related to ER response to unfolded proteins and intracellular vesicle-mediated protein transport mainly including the ER and the Golgi. Moreover, enrichment analysis of the uniquely regulated genes by SEL1L knockout identified 159 enriched processes in which they were categorized into major bigger processes including aerobic respiration, nucleotides biosynthesis and metabolism, protein ubiquitination and signal transduction including Wnt signaling pathway (Fig. 4B).

While analyzing this large gene set has provided a broad overview of the enriched biological processes, we sought of analyzing smaller subsets of genes for increased power and further a more focused biological insight. Thus, we have investigated co-varying genes and their associated biological processes, where we aimed to apply weighted gene co-expression analysis (WGCNA) on the 3052 differentially expressed genes. Firstly, sample clustering has been performed using these genes’ expression data to make sure no outliers are present in the data. As shown in Fig. 5A, no outlier sample was detected, where control HEK293 cells and SEL1L-KO cells form different clusters. WGCNA is an unsupervised analysis tool that is based on scale free topology criteria. To meet the latter, a soft threshold (power) of 6 was set (Fig. 5B). Densely interconnected genes were then hierarchically clustered by modules coded by different colors. Using a signed network construction, seven different modules/clusters of positively correlated genes shown in Fig. 5C were generated. The size of the modules ranged from 200 to 650 genes. The expression profiles of the different modules, summarized by the cluster’s eigengene, are represented in Fig. 6 by their corresponding module eigengenes and heatmaps. WGCNA had successfully clustered co-expressed genes, where 3 generated modules (yellow, green, and brown) had decreased expression profile in response to SEL1L knockout compared to control HEK293 cells. Genes of the yellow and green modules were enriched in aerobic respiration and nucleotides biosynthesis and metabolism. However, the 525 genes clustered together in the brown module were interestingly overrepresented in biological processes related to the ribosome and ribosomal RNA (rRNA) related processes, which was masked when analyzing all DEGs together. Furthermore, positively regulated genes were clustered together into 4 clusters (blue, black, turquoise, and red) based on their fold change profiles. Blue and black modules genes were enriched in processes mainly related to vesicle-mediated processes and endoplasmic reticulum stress, respectively. Whereas red module genes (294 genes) were enriched in signal transduction related processes mainly the Wnt signaling pathway. Additionally, 622 genes were assigned to the turquoise module displaying enrichment in general systemic biological processes like urogenital and renal systems development and cell-substrate adhesion.

In an attempt to identify hub genes that might be responsible for these variations, WGCNA modules served as input for STRING software to construct protein-protein interaction (PPI) networks. Subsequently, we have used Molecular Complex Detection (MCODE) plug-in in Cytoscape to identify densely connected complexes/genes within STRING PPI networks. MCODE analyzes nodes (representing genes/proteins) and edges (representing the interaction between genes/proteins) of the network of interest. It employs graph-theoretical algorithms to identify the highly connected nodes and applies a scoring system which helps highlight the most significant and biologically relevant entities within the network. As a result, based on the highest score in the studied modules, we have identified a hub gene (seed) for each module, NDUFB4, ATP5F1C, NACA, ZFYVE1, VCP, SEC23B, and CCNE2. These hub genes were then validated statistically using their module membership based on WGCNA co-expression networks. Firstly, a module eigengene is calculated for each module. For each gene in the module, we have calculated its Pearson correlation with the eigengene, named module membership. The module membership measure (correlation coefficient) provides an idea about the degree of association between the gene’s expression profile and the module’s overall expression pattern. To be considered significant, we have set a threshold of 0.9, where genes having a coefficient above this threshold are considered to have a strong module membership and likely to play important role within the module. Interestingly, all biologically identified genes had a module membership of minimum 0.94 as shown in Table 1, except for the turquoise module gene ZFYVE1, its module eigengene is 0.821.

Table 1

Module hub genes and their corresponding MCODE score and WGCNA module membership.
Module	Hub gene	MCODE Score	WGCNA module membership
Yellow	NDUFB4	12	0.949
Green	ATP5F1C	12.4	0.995
Brown	NACA	14	0.991
Turquoise	ZFYVE1	5	0.821
Black	VCP	7	0.988
Blue	SEC23B	9	0.984
Red	CCNE2	3	0.985

The transcriptomic profiling performed in this study has revealed significant insights into the molecular mechanism caused by the transcriptomic alterations of HRD1 and SEL1L knockout in HEK293 cellular model, shedding light onto their roles in cellular processes other than the ER quality control. While HRD1 knockout cells displayed enrichment solely in ER-related processes, SEL1L knockouts have exhibited a broader impact that triggers more extensive cellular response in the studied model.

The absence of enriched processes and pathways outside the ER highlights the specificity and the crucial role of HRD1 in the ER quality control, specifically the ERAD machinery. These findings are consistent with the known role of HRD1 in the ERAD. However, several studies have reported an important role of HRD1 in immune regulation, including T cell immunity and B cells development and survival (Kong et al., 2016; Xu et al., 2016; Xu and Fang, 2020). While the exact mechanism behind these roles is still being investigated, it is largely thought to be ERAD related, where HRD1 targets proteins involved in T and B cells development like p27kip1 (Xu et al., 2016), pre-BCR (Yang et al., 2018), and CD95/Fas (Kong et al., 2016) for degradation. Knowing that deficiency in HRD1 is embryonic lethal in mouse model (Amano et al., 2003), it is not associated clinically with any human phenotype. Moreover to investigate the pathophysiological role of HRD1, studies have performed tissue targeted deletion in mammalian models, specifically mice (Xu et al., 2016; Wei et al., 2018; Yang et al., 2018; Chen et al., 2020). Few are the studies that have assessed the transcriptional changes in response to HRD1 deficiency. In an HRD1 knockout vascular smooth muscle cellular model, Wang et al, showed that HRD1 deficiency led to upregulation of genes related to ER stress response, prolifiration and migration and downregulation of contractile-related genes (Wang et al., 2023a). Additionally, in a recent proteomic study that aimed to finding substrates of SEL1L-HRD1 complex, Wei et al, performed RNA sequencing on HRD1 deficient HEK293T cells and BAT cells but did not perform a detailed transcriptomic analysis. In this study, authors showed that HRD1 induced a significant diffetential expression in 280 and 377 genes for p value of 0.05 but didn’t specify a fold change selection criteria in HEK293T and BAT, respectively (Wei et al., 2024).

While SEL1L serves as an adaptor protein for HRD1, aiding in the recognition and targeting of misfolded proteins in the ER, our research proposes that SEL1L possesses a multifaceted role distinct from its involvement in ERAD. Following an analytical pipeline that includes unsupervised identification of differentially expressed genes, construction of co-expression networks, enrichment analysis, and identification of hub genes, we were able to identify various major regulatory genes, including NDUFB4, ATP5F1C, NACA, VCP, SEC23B, CCNE2. These genes represent key players in different cellular processes, indicating the impact of SEL1L deficiency on various cellular functions. Among the identified hub genes, NDUFB4 and ATP5F1C are components of the mitochondrial respiratory chain and ATP synthase complex, respectively. Dysregulation of these genes suggests a potential role of SEL1L in mitochondrial function and cellular energetics. Interestingly, in a brown adipocyte cellular model, Zhou et al, showed that SEL1L regulates mitochondrial dynamics and ER-mitochondria contacts (Zhou et al., 2020). Additionally, SEL1L deficient HepG2 cells displayed an impaired mitochondrial morphology associated with reduced oxidative phosphorylation and mitochondrial membrane potential (Liu et al., 2020b). Disruption of mitochondrial dynamics caused by SEL1L deficiency was also linked to reduction in T-cell metabolism (Correa et al., 2022; Correa-Medero et al., 2024). Although no direct interaction has been proposed between SEL1L and the mitochondrial proteins NDUFB4 and ATP5F1C, the proposed mechanism behind the mitochondrial dysfunction caused by SEL1L is via the ER-mitochondrial contact sites known as MAMs.

Interestingly, NACA, also known as nascent polypeptide-associated complex subunit alpha, is identified as a hub gene in the brown WGCNA module, whose genes were enriched mainly in ribosome biogenesis and cytoplasmic translation. NACA physiologically is involved in ribososme binding during translation, it influences protein folding through its interaction with nascently synthesized proteins (Schroeder et al., 2022). NACA is also found to play a major role in the translocation of nascent polypeptide into the ER (Zhang et al., 2012). While there are no studies that invetsigate possible SEL1L interaction with NACA, some proposes an evidence of ERAD and ribosomal quality control cross talk while translating membrane proteins (Phillips and Miller, 2020). Furthermore, SEC23B is a component of the COPII vesicle coat complex involved in ER-to-Golgi trafficking (Yehia et al., 2021), indicating a potential role for SEL1L in regulating ER-Golgi transport pathways, noting that SEL1L is highly expressed in protein secreting cells like B cells and pancreatic cells (Biunno et al., 2006). Additionally, our study proposes the ER related changes to be mediated largely thorugh VCP, or valosin-containing protein. Several studies have shown that VCP is a major component of the SEL1L/HRD1 complex, however, Lin et al, have recently shown that VCP directly interacts with SEL1L (Lin et al., 2024). Finally, cyclin E2 (CCNE2), a cyclin involved in cell cycle regulation, suggests potential alterations in cell cycle dynamics including proliferation in SEL1L-deficient cells. Some studies have reported the implication of SEL1L in various signaling pathways including the ERK pathway (Diaferia et al., 2013), notch signaling pathway (Liu et al., 2021), and TGF-beta signaling pathway (Shrestha et al., 2020). Moreover, our analysis proposes the regulation of Wnt pathway by SEL1L. While no studies have reported potential interaction between SEL1L and Wnt pathway, a SEL1L liver deficient mouse model displays increased Wnt5a aggregates in the ER associated with moderate elevation of cyclin D1 (CCND1) (Bhattacharya et al., 2022). Interestingly, in the current analysis, we suggest an interaction with CCNE2, which was shown to phosphorylate β-catenin, the major substrate for Wnt pathway (Shah and Kazi, 2022).

Overall, our study provides valuable insights into the transcriptomic changes associated with SEL1L and HRD1 deficiencies, shedding light on their roles in cellular processes beyond the ERAD. In conclusion, our findings underscore the specificity and critical role of HRD1 in ER quality control and mainly ERAD. Conversely, we evidenced the multifunctional role possessed by SEL1L, where we highlight novel roles implicated in various organelles and processes including the ribosome, ER-Golgi vesicle transport, and cell signaling and proliferation. These findings contribute to a comprehensive understanding of SEL1L and HRD1 functions and mechanisms and pave the way for future studies elucidating their roles in their implicated pathophysiology. Moreover, further investigations into the functional consequences of these transcriptomic changes will deepen the understanding of SEL1L and HRD1 biology and their implicated pathology.

1. Generation of HEK293 cells deficient of HRD1 and SEL1L using CRISPR/Cas9-directed genome editing

HRD1 and SEL1L knockout HEK293 cells were generated as described previously (Kizhakkedath et al., 2018; Gariballa et al., 2022). Briefly, HRD1 KN2.0 non homology mediated CRISPR/cas9 kits (KN407132, Origene inc.) and the GeneArt® CRISPR Nuclease Vector with orange fluorescent protein (OFP) Reporter (Life Technologies) were used to generate HRD1 and SEL1L knockout cell lines, respectively. For HRD1, two different gRNA vectors embedded in pCas-Guide CRISPR vector were used to cut the genome at the target site. Cut site is then repaired through the integration of a linear donor DNA containing a selection marker (puromycin) and reporter gene (GFP) (LoxP-EF1A-tGFP-P2A-Puro-LoxP). Moreover, two pairs of gRNAs targeting SEL1L were designed and integrated into OFP/Cas9 plasmids. HEK293 cells seeded into 6 well-plates were transfected with gRNAs vector and the linear DNA donor using FuGENE HD transfection reagent (Promega). HRD1 KO transfected cells were split 5 times to dilute out non-integrated donor DNA. Remaining cells were then treated with 1µg/ml puromycin (Sigma) predetermined by a kill curve for 7 days to select positive knockout clones. Moreover, OFP positive SEL1L clones were isolated using fluorescence-activated cell sorting (FACS) (BD FACSCanto II, BD Biosciences) (Kizhakkedath et al., 2018). Single clones were then seeded into 96 well plates, allowed to grow and then transferred into bigger plates for validation. Gene knockout has been confirmed at the DNA level using Sanger sequencing, and at the protein levels using immunoblotting (Kizhakkedath et al., 2018; Gariballa et al., 2022).

2. RNA extraction, quality checks, library preparation

Cells were pelleted and stored at -80 °C and sent for Novogene for sequencing. RNA was extracted, quantified using Nanodrop and sample integrity and purity were checked using a Nanodrop and Agilent bioanalyzer. Only samples with RNA integrity number (RIN) > 9 have been processed for library preparation. Messenger RNA (mRNA) was purified from total RNA using poly-T oligo-attached magnetic beads. After fragmentation, the first strand cDNA was synthesized using random hexamer primers, followed by the second strand cDNA synthesis using either dUTP for directional library or dTTP for non-directional library. cDNA Library was then checked with Qubit and PCR for quantification and bioanalyzer for size distribution detection.

3. RNA sequencing

Sequencing of the samples was performed using NovaSeq platform using 150nt and paired end reads. Image data files generated NovaSeq are transformed into sequenced reads CASAVA base recognition (Base Calling). Raw data are stored in FASTQ format files, which contain sequences of reads and corresponding base quality. Reads were then tested for sequencing error rate, GC content distribution and then filtered to obtain clean reads for downstream analysis.

4. Real time Polymerase Chain Reaction

RNA of wild type and knockout cell lines seeded into 6-well plates were extracted using TRIzol™ reagent solution (Invitrogen) following manufacturer’s protocol. First strand cDNA was generated from total RNA using GoScript Reverse Transcriptase System (Promega). mRNA expression profiles were measured using SYBR Green Real-Time qPCR Master Mix (Thermo Fisher Scientific) by the Quantstudio™ 7 Flex Real Time PCR System in a 96-well block (Applied Biosystems™). Sequence of target genes primers used are listed in Table 2. RT-PCR results were analyzed using GraphPad Prism (version 10.0.3).

Table 2

Primers list for genes selected for real-time PCR.
Gene	Forward primer	Reverse primer
NSUN7	TGCCTCCAGTCTTTATGCCA	AACTGCTCTTTGGCCCCTAT
CTSZ	CCCCATCGTTTAAGGCCATG	AACTCTCAGGAACACTCGCA
HERPUD1	ACCTTTCTTCCCCTGGATGG	AACAAAAGCCCCTGATGCAG
DAG1	CCTGACCTAGCACACACTGA	TGCCTGCGAGTTCATTCATG
HRD1	CACACCTTCCCACTCTTTGC	TGGAAAATGTGGTTGCAGGG
SFRP2	TTCCCCAAGCACACTCCTAG	TACAAGATTCGGGTGGGCTT
PXMP4	CCCTCTGTCTTTCTGCCTGA	AGATGGGGAGGGTGTTTGAG

5. Differential gene expression analysis

Differential gene expression analysis between wild type and knockout samples was performed using the R package DESeq2 (version 1.38.3). Raw read counts from RNA-seq were used as input and DESeq2 workflow was applied to estimate the size factors, dispersion parameters, and fold changes for each gene (Love et al., 2014; Badawi et al., 2019). Low count genes (less than 10 in at least one sample) were filtered using independent filtering method to adjust for multiple testing. We considered genes with an adjusted p-value less than 0.05 and 0.01 and absolute log2 fold change greater than 0.58 as differentially expressed for HRD1 and SEL1L analysis, respectively. DESeq2 results function was used to extract the list of differentially expressed genes and their corresponding statistical significance. Analyzed data were visualized using by principal component analysis (PCA) plot and volcano plot to identify potential outliers in the samples and gene expression patterns.

6. Enrichment analysis

Gene ontology (GO) analysis of the differentially expressed genes was performed using the R package clusterProfiler (version 4.6.2). We used the org.Hs.eg.db (version 3.16.0) package to map the gene symbols to Entrez IDs and obtain the GO annotations for human genes. Hypergeometric test was used to calculate the p-values and the Benjamini-Hochberg method to adjust for multiple testing. We considered GO terms with an adjusted p-value less than 0.05. We used the enrichplot functions to generate plots to visualize the top enriched GO terms for the biological processes. We also performed KEGG pathway analysis using the same package and parameters.

7. Weighted gene co-expression analysis

Deseq2 differentially expressed genes were normalized using varianceStabilizingTransformation function. Normalized matrix was used as an input for weighted gene co-expression analysis (WGCNA). To achieve a scale-free topology criterion for network construction, a soft threshold (beta), also known as power was set to 6. Based on pairwise correlations between genes, an adjacency matrix was constructed. This matrix is then transformed into a topological overlap matrix (TOM). Hierarchical clustering using flashClust (version 1.01-2) function was then performed and cutreeDynamic function of WGCNA package (version 1.72-5). Genes of similar expression profiles are clustered together in clusters refered as modules.

8. Protein-protein interaction network

Genes of these modules are used as an input for the freely available STRING database (version 12.0) (Franceschini et al., 2013). Generated networks were then sent to Cytoscape (version 3.10.2) (Shannon et al., 2003) for visualization. Highly connected proteins were identified using Molecular Complex Detection (MCODE) clustering algorithm, a plug-in in cytoscape, based on STRING biological interactions (Bader and Hogue, 2003).

Acknowledgements

None

Data availability

RNA-seq data generated and analyzed during the current study are available in the SRA repository under the accession number PRJNA1113371 (https://dataview.ncbi.nlm.nih.gov/object/PRJNA1113371?reviewer=bpl8ptrl00enhojkmich6e8rcl).

Funding

This work was supported by ASPIRE, the technology program management pillar of Abu Dhabi’s Advanced Technology Research Council (ATRC), via individual grant AARE19-086 and the ASPIRE Precision Medicine Research Institute grant VRI-20-10.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Ethical statement

The manuscript doesn’t include data or description of human patients or animal experiments and therefore does not require ethical committee approval at this institution.

Authors Contribution

SB, NG, PK, generated the cell lines and analyzed the data. SB generated the first draft of the manuscript. BRA conceptualized the project, supervised the overall direction and edited all versions of the manuscript. All authors read and approved the final manuscript.

Amano, T., Yamasaki, S., Yagishita, N., Tsuchimochi, K., Shin, H., Kawahara, K., et al. (2003). Synoviolin/Hrd1, an E3 ubiquitin ligase, as a novel pathogenic factor for arthropathy. Genes Dev 17, 2436–2449. doi: 10.1101/gad.1096603
Badawi, S., Mohamed, F. E., Varghese, D. S., and Ali, B. R. (2023). Genetic disruption of mammalian endoplasmic reticulum-associated protein degradation: Human phenotypes and animal and cellular disease models. Traffic 24, 312–333. doi: 10.1111/tra.12902
Badawi, S., Paccalet, A., Harhous, Z., Pillot, B., Augeul, L., Van Coppenolle, F., et al. (2019). A Dynamic Transcriptional Analysis Reveals IL-6 Axis as a Prominent Mediator of Surgical Acute Response in Non-ischemic Mouse Heart. Front Physiol 10, 1370. doi: 10.3389/fphys.2019.01370
Bader, G. D., and Hogue, C. W. (2003). An automated method for finding molecular complexes in large protein interaction networks. BMC Bioinformatics 4, 2. doi: 10.1186/1471-2105-4-2
Beissinger, M., and Buchner, J. (1998). How chaperones fold proteins. Biol Chem 379, 245–259.
Bhattacharya, A., Sun, S., Wang, H., Liu, M., Long, Q., Yin, L., et al. (2018). Hepatic Sel1L-Hrd1 ER-associated degradation (ERAD) manages FGF21 levels and systemic metabolism via CREBH. EMBO J 37, e99277. doi: 10.15252/embj.201899277
Bhattacharya, A., Wei, J., Song, W., Gao, B., Tian, C., Wu, S. A., et al. (2022). SEL1L-HRD1 ER-associated degradation suppresses hepatocyte hyperproliferation and liver cancer. iScience 25, 105183. doi: 10.1016/j.isci.2022.105183
Biunno, I., Cattaneo, M., Orlandi, R., Canton, C., Biagiotti, L., Ferrero, S., et al. (2006). SEL1L a multifaceted protein playing a role in tumor progression. Journal of Cellular Physiology 208, 23–38. doi: 10.1002/jcp.20574
Chen, L., Wei, J., Zhu, H., Pan, H., and Fang, D. (2020). Energy supplementation rescues growth restriction and female infertility of mice with hepatic HRD1 ablation. Am J Transl Res 12, 2018–2027.
Correa, L. O., Lee, R., Dils, A., Vijendra, A. I., and Carty, S. A. (2022). Sel1L, an Endoplasmic Reticulum Associated Degradation Adaptor, Regulates CD8+ Tcell Fate. Blood 140, 5508–5509. doi: 10.1182/blood-2022-170207
Correa-Medero, L. O., Jankowski, S. E., Hong, H. S., Armas, N. D., Vijendra, A. I., Reynolds, M. B., et al. (2024). ER-associated degradation adapter Sel1L is required for CD8+ T cell function and memory formation following acute viral infection. Cell Reports 43. doi: 10.1016/j.celrep.2024.114156
Diaferia, G. R., Cirulli, V., and Biunno, I. (2013). SEL1L Regulates Adhesion, Proliferation and Secretion of Insulin by Affecting Integrin Signaling. PLoS One 8, e79458. doi: 10.1371/journal.pone.0079458
Ellgaard, L., McCaul, N., Chatsisvili, A., and Braakman, I. (2016). Co- and Post-Translational Protein Folding in the ER. Traffic 17, 615–638. doi: 10.1111/tra.12392
Fan, Y., Wang, J., Jin, W., Sun, Y., Xu, Y., Wang, Y., et al. (2021). CircNR3C2 promotes HRD1-mediated tumor-suppressive effect via sponging miR-513a-3p in triple-negative breast cancer. Molecular Cancer 20, 25. doi: 10.1186/s12943-021-01321-x
Franceschini, A., Szklarczyk, D., Frankild, S., Kuhn, M., Simonovic, M., Roth, A., et al. (2013). STRING v9.1: protein-protein interaction networks, with increased coverage and integration. Nucleic Acids Res 41, D808–D815. doi: 10.1093/nar/gks1094
Gariballa, N., Kizhakkedath, P., Akawi, N., John, A., and Ali, B. R. (2022). Endoglin Wild Type and Variants Associated With Hereditary Hemorrhagic Telangiectasia Type 1 Undergo Distinct Cellular Degradation Pathways. Front Mol Biosci 9, 828199. doi: 10.3389/fmolb.2022.828199
Kizhakkedath, P., John, A., Al-Gazali, L., and Ali, B. R. (2018). Degradation routes of trafficking-defective VLDLR mutants associated with Dysequilibrium syndrome. Scientific Reports 8, 1583. doi: 10.1038/s41598-017-19053-8
Kong, S., Yang, Y., Xu, Y., Wang, Y., Zhang, Y., Melo-Cardenas, J., et al. (2016). Endoplasmic reticulum-resident E3 ubiquitin ligase Hrd1 controls B-cell immunity through degradation of the death receptor CD95/Fas. Proc Natl Acad Sci U S A 113, 10394–10399. doi: 10.1073/pnas.1606742113
Lin, L. L., Wang, H. H., Pederson, B., Wei, X., Torres, M., Lu, Y., et al. (2024). SEL1L-HRD1 interaction is required to form a functional HRD1 ERAD complex. Nat Commun 15, 1440. doi: 10.1038/s41467-024-45633-0
Liu, L., Yu, L., Zeng, C., Long, H., Duan, G., Yin, G., et al. (2020a). E3 Ubiquitin Ligase HRD1 Promotes Lung Tumorigenesis by Promoting Sirtuin 2 Ubiquitination and Degradation. Mol Cell Biol 40, e00257-19. doi: 10.1128/MCB.00257-19
Liu, Q., Yang, X., Long, G., Hu, Y., Gu, Z., Boisclair, Y. R., et al. (2020b). ERAD deficiency promotes mitochondrial dysfunction and transcriptional rewiring in human hepatic cells. J Biol Chem 295, 16743–16753. doi: 10.1074/jbc.RA120.013987
Liu, X., Yu, J., Xu, L., Umphred-Wilson, K., Peng, F., Ding, Y., et al. (2021). Notch-induced endoplasmic reticulum-associated degradation governs mouse thymocyte β−selection. eLife 10, e69975. doi: 10.7554/eLife.69975
Love, M. I., Huber, W., and Anders, S. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology 15, 550. doi: 10.1186/s13059-014-0550-8
Nam, S. M., and Jeon, Y. J. (2019). Proteostasis In The Endoplasmic Reticulum: Road to Cure. Cancers (Basel) 11, 1793. doi: 10.3390/cancers11111793
Omura, T., Matsuda, H., Nomura, L., Imai, S., Denda, M., Nakagawa, S., et al. (2018). Ubiquitin ligase HMG-CoA reductase degradation 1 (HRD1) prevents cell death in a cellular model of Parkinson’s disease. Biochemical and Biophysical Research Communications 506, 516–521. doi: 10.1016/j.bbrc.2018.10.094
Phillips, B. P., and Miller, E. A. (2020). Ribosome-associated quality control of membrane proteins at the endoplasmic reticulum. J Cell Sci 133, jcs251983. doi: 10.1242/jcs.251983
Podolsky, M. J., Kheyfets, B., Pandey, M., Beigh, A. H., Yang, C. D., Lizama, C. O., et al. (2024). Genome-wide screens identify SEL1L as an intracellular rheostat controlling collagen turnover. Nat Commun 15, 1531. doi: 10.1038/s41467-024-45817-8
Qi, L., Tsai, B., and Arvan, P. (2017). New insights into the physiological role of ERAD. Trends Cell Biol 27, 430–440. doi: 10.1016/j.tcb.2016.12.002
Schroeder, A. M., Nielsen, T., Lynott, M., Vogler, G., Colas, A. R., and Bodmer, R. (2022). Nascent polypeptide-Associated Complex and Signal Recognition Particle have cardiac-specific roles in heart development and remodeling. PLOS Genetics 18, e1010448. doi: 10.1371/journal.pgen.1010448
Shah, K., and Kazi, J. U. (2022). Phosphorylation-Dependent Regulation of WNT/Beta-Catenin Signaling. Front. Oncol. 12. doi: 10.3389/fonc.2022.858782
Shannon, P., Markiel, A., Ozier, O., Baliga, N. S., Wang, J. T., Ramage, D., et al. (2003). Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks. Genome Res 13, 2498–2504. doi: 10.1101/gr.1239303
Shrestha, N., Liu, T., Ji, Y., Reinert, R. B., Torres, M., Li, X., et al. (2020). Sel1L-Hrd1 ER-associated degradation maintains β cell identity via TGF-β signaling. J Clin Invest 130, 3499–3510. doi: 10.1172/JCI134874
Sun, S., Shi, G., Han, X., Francisco, A. B., Ji, Y., Mendonça, N., et al. (2014). Sel1L is indispensable for mammalian endoplasmic reticulum-associated degradation, endoplasmic reticulum homeostasis, and survival. Proc Natl Acad Sci U S A 111, E582-591. doi: 10.1073/pnas.1318114111
Wang, H. H., Lin, L. L., Li, Z. J., Wei, X., Askander, O., Cappuccio, G., et al. (2024). Hypomorphic variants of SEL1L-HRD1 ER-associated degradation are associated with neurodevelopmental disorders. J Clin Invest 134, e170054. doi: 10.1172/JCI170054
Wang, L., Ren, Z., Wu, L., Zhang, X., Wang, M., Niu, H., et al. (2023a). HRD1 reduction promotes cholesterol-induced vascular smooth muscle cell phenotypic change via endoplasmic reticulum stress. Mol Cell Biochem. doi: 10.1007/s11010-023-04902-0
Wang, Y., Wang, S., and Zhang, W. (2023b). HRD1 functions as a tumor suppressor in ovarian cancer by facilitating ubiquitination-dependent SLC7A11 degradation. Cell Cycle 22, 1116–1126. doi: 10.1080/15384101.2023.2178102
Wei, J., Yuan, Y., Chen, L., Xu, Y., Zhang, Y., Wang, Y., et al. (2018). ER-associated ubiquitin ligase HRD1 programs liver metabolism by targeting multiple metabolic enzymes. Nat Commun 9, 3659. doi: 10.1038/s41467-018-06091-7
Wei, X., Lu, Y., Lin, L. L., Zhang, C., Chen, X., Wang, S., et al. (2024). Proteomic screens of SEL1L-HRD1 ER-associated degradation substrates reveal its role in glycosylphosphatidylinositol-anchored protein biogenesis. Nat Commun 15, 659. doi: 10.1038/s41467-024-44948-2
Weis, D., Lin, L. L., Wang, H. H., Li, Z. J., Kusikova, K., Ciznar, P., et al. (2024). Biallelic Cys141Tyr variant of SEL1L is associated with neurodevelopmental disorders, agammaglobulinemia, and premature death. J Clin Invest 134, e170882. doi: 10.1172/JCI170882
Wu, T., Zhang, S., Xu, J., Zhang, Y., Sun, T., Shao, Y., et al. (2020). HRD1, an Important Player in Pancreatic β-Cell Failure and Therapeutic Target for Type 2 Diabetic Mice. Diabetes 69, 940–953. doi: 10.2337/db19-1060
Xu, Y., and Fang, D. (2020). Endoplasmic reticulum-associated degradation and beyond: The multitasking roles for HRD1 in immune regulation and autoimmunity. J Autoimmun 109, 102423. doi: 10.1016/j.jaut.2020.102423
Xu, Y., Zhao, F., Qiu, Q., Chen, K., Wei, J., Kong, Q., et al. (2016). The ER membrane-anchored ubiquitin ligase Hrd1 is a positive regulator of T-cell immunity. Nat Commun 7, 12073. doi: 10.1038/ncomms12073
Yang, H., Qiu, Q., Gao, B., Kong, S., Lin, Z., and Fang, D. (2014). Hrd1-mediated BLIMP-1 ubiquitination promotes dendritic cell MHCII expression for CD4 T cell priming during inflammation. Journal of Experimental Medicine 211, 2467–2479. doi: 10.1084/jem.20140283
Yang, Y., Kong, S., Zhang, Y., Melo-Cardenas, J., Gao, B., Zhang, Y., et al. (2018). The endoplasmic reticulum-resident E3 ubiquitin ligase Hrd1 controls a critical checkpoint in B cell development in mice. J Biol Chem 293, 12934–12944. doi: 10.1074/jbc.RA117.001267
Yao, X., Wu, Y., Xiao, T., Zhao, C., Gao, F., Liu, S., et al. (2022). T-cell-specific Sel1L deletion exacerbates EAE by promoting Th1/Th17-cell differentiation. Mol Immunol 149, 13–26. doi: 10.1016/j.molimm.2022.06.001
Yehia, L., Liu, D., Fu, S., Iyer, P., and Eng, C. (2021). Non-canonical role of wild-type SEC23B in the cellular stress response pathway. Cell Death Dis 12, 1–12. doi: 10.1038/s41419-021-03589-9
Yue, Y., Cui, X., Fraass, B. A., Tuli, R., and Liu, Z. (2017). SEL1L Stratifies the Prognosis of Breast Cancer Patients Treated with Radiotherapy. International Journal of Radiation Oncology, Biology, Physics 99, S55. doi: 10.1016/j.ijrobp.2017.06.138
Zhang, Y., Berndt, U., Gölz, H., Tais, A., Oellerer, S., Wölfle, T., et al. (2012). NAC functions as a modulator of SRP during the early steps of protein targeting to the endoplasmic reticulum. Mol Biol Cell 23, 3027–3040. doi: 10.1091/mbc.E12-02-0112
Zhou, Z., Torres, M., Sha, H., Halbrook, C. J., Van den Bergh, F., Reinert, R. B., et al. (2020). Endoplasmic reticulum-associated degradation regulates mitochondrial dynamics in brown adipocytes. Science 368, 54–60. doi: 10.1126/science.aay2494

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Transcriptomic analysis of SEL1L and HRD1 knockout cell lines reveals multifaceted roles of SEL1L beyond the ER quality control

Status:

Version 1

Abstract

Figures

Introduction

Results

1.1. Summary of Transcriptome sequencing, mapping, and gene expression level data

1.2. Transcriptome profiling of HRD1 KO HEK293 cells

1.3. Characterization of SEL1L knockout HEK293 cells

Discussion

Materials and Methods

1. Generation of HEK293 cells deficient of HRD1 and SEL1L using CRISPR/Cas9-directed genome editing

2. RNA extraction, quality checks, library preparation

3. RNA sequencing

4. Real time Polymerase Chain Reaction

5. Differential gene expression analysis

6. Enrichment analysis

7. Weighted gene co-expression analysis

8. Protein-protein interaction network

Declarations

References

Additional Declarations

Status:

Version 1