An Autoantigen Profile from Jurkat T-Lymphoblasts Provides a Molecular Guide for Investigating Autoimmune Sequelae of COVID-19

In order to understand autoimmune phenomena contributing to the pathophysiology of COVID-19 and post-COVID syndrome, we have been profiling autoantigens (autoAgs) from various cell types. Although cells share numerous autoAgs, each cell type gives rise to unique COVID-altered autoAg candidates, which may explain the wide range of symptoms experienced by patients with autoimmune sequelae of SARS-CoV-2 infection. Based on the unifying property of affinity between autoantigens (autoAgs) and the glycosaminoglycan dermatan sulfate (DS), this paper reports 140 candidate autoAgs identified from proteome extracts of human Jurkat T-cells, of which at least 105 (75%) are known targets of autoantibodies. Comparison with currently available multi-omic COVID-19 data shows that 125 (89%) of DS-affinity proteins are altered at protein and/or RNA levels in SARS-CoV-2-infected cells or patients, with at least 94 being known autoAgs in a wide spectrum of autoimmune diseases and cancer. Protein alterations by ubiquitination and phosphorylation in the viral infection are major contributors of autoAgs. The autoAg protein network is significantly associated with cellular response to stress, apoptosis, RNA metabolism, mRNA processing and translation, protein folding and processing, chromosome organization, cell cycle, and muscle contraction. The autoAgs include clusters of histones, CCT/TriC chaperonin, DNA replication licensing factors, proteasome and ribosome proteins, heat shock proteins, serine/arginine-rich splicing factors, 14-3-3 proteins, and cytoskeletal proteins. AutoAgs such as LCP1 and NACA that are altered in the T cells of COVID patients may provide insight into T-cell responses in the viral infection and merit further study. The autoantigen-ome from this study contributes to a comprehensive molecular map for investigating acute, subacute, and chronic autoimmune disorders caused by SARS-CoV-2.


Introduction
The COVID-19 pandemic has been devastating. After initial recovery from acute SARS-CoV-2 infection, many people continue to suffer from lingering health problems (so called "long COVID" or post-COVID syndrome), such as fatigue, shortness of breath, joint pain, chest pain, muscle pain, loss of smell or taste, and other neurological problems. Although the underlying causes are far from clear, autoimmune effects are likely important contributors to chronic post-COVID disorders. To understand how SARS-CoV-2 infection may induce autoimmune responses, we are establishing a comprehensive COVID autoantigen atlas, i.e., all possible endogenous autoantigens (autoAgs) that may be rendered immunogenic by the viral infection. Because different tissues or cells may give rise to distinct pools of autoAgs, we have been profiling autoAgs from multiple human tissues and cell types, including human lung fibroblast HFL1 cells, human lung epithelial-like A549 cells, and B-lymphoblast HS-Sultan cells [1][2][3][4]. In this study, we report an autoantigen-ome identified from human Jurkat T-lymphoblast cells.
AutoAgs are the targets of autoantibodies (autoAbs) and T-cell autoimmune responses. Typically, selfmolecules are naturally tolerated by the immune system and do not provoke autoimmune responses.
However, certain self-molecules transform into autoAgs and become targets of autoimmune attacks. Thus far, hundreds of autoAgs with seemingly no obvious structural or functional commonality have been identified across various autoimmune diseases and cancers. Our studies have demonstrated that autoAgs do, in fact, share common properties. AutoAgs are commonly released by apoptotic cells, and we found that the glycosaminoglycan dermatan sulfate (DS) has peculiar affinity to apoptotic cells and their autoAgs [5,7,8]. DS and autoAgs can form affinity complexes and cooperatively stimulate autoreactive B1 cells and autoantibody production [5,7,8]. Based on autoAg-DS affinity, we have identified several hundred autoAgs from various cells and tissues [1][2][3][9][10][11].
Proteins eluted with 1.0 M NaCl possess the strongest DS-affinity and, strikingly, 10/13 (90.9%) are known autoAgs ( Table 1), indicating that increasing affinity to DS increases the propensity of a protein to be an autoAg, consistent with our prior findings [1-3, 5, 6, 9-11]. These include histones (H4, H2B types 1-a and 1-b, H2A type 1-a), 60S ribosomal proteins (P0, P2, L6, L7), ACTC1 (skeletal muscle actin), and C1QBP, and PABPC3 (polyadenylate-binding protein 3). Histones and ribosomal P proteins are hallmark autoAgs used in routine clinical tests of autoimmune diseases. Histone autoAbs are nearly always present in druginduced systemic lupus erythematosus, and ribosomal P autoantibodies are tested for to aid in the differential diagnosis of lupus patients with neuropsychiatric symptoms. C1QBP has been repeatedly identified as a putative autoAg in several of our prior studies [1,2,9,10] and was recently confirmed as an autoAg in the neurodegenerative disorder primary open-angle glaucoma [24]. Poly(A)-binding proteins bind the poly(A) tail of messenger RNAs and control mRNA stability and translation initiation. Although PABPC3 has not yet been reported as an autoAg, its paralog PABPC1 has been found to be an autoAg [25].
Protein-protein interaction network analysis by STRING [26] reveals that the DS-affinity autoantigen-ome is highly connected (Fig. 1). There are 787 interactions at high confidence level (vs. 284 expected; enrichment p-value <1.0e-16). These DS-affinity proteins are enriched in several clusters and significantly associated with the cell cycle, protein folding, chromosome organization, RNA splicing, translation, and muscle contraction (Fig. 1). There are 36 DS-affinity proteins associated with the cell cycle, particularly the G2/M checkpoints (26 proteins), the G2/M DNA damage checkpoint, and the G1/S and G2/M transitions.
Pathway and process enrichment analyses by Metascape [27] also reveal that proteins of the DS-affinity autoantigen-ome are significantly associated with cellular response to stress, protein folding, and protein localization to organelles ( Fig. 2A). In addition, they are associated with kinase maturation complex 1, spliceosome, HSF1 activation (activates gene expression in response to a variety of stresses), protein processing in the endoplasmic reticulum, VEGFA-VEGFR2 signaling (major pathway that activates angiogenesis), apoptosis-induced DNA fragmentation, and 17S U2 snRNP.

DS-affinity autoantigen-ome related to COVID-19
To find out how many of the DS-affinity autoAgs identified from Jurkat T-cells are affected by in SARS-CoV-2 infection, we searched for them in a multi-omic COVID database compiled by Coronascape .
Among the COVID-altered DS-affinity proteins, 17 are up-regulated only, 35 are down-regulated only, and 71 are altered (up or down depending on study conditions) at protein and/or RNA levels in SARS-CoV-2 infected cells. The COVID database was assembled from different cell and patient tissue types by multiple research laboratories using different technologies, including proteomics, phosphoproteomics, ubiquitinomics, and bulk and single-cell RNA sequencing.
Six DS-affinity proteins are found in the interactomes of SARS-CoV-2 viral proteins, i.e., these host proteins interact directly or indirectly with the viral proteins [29,40,44]. Specifically, HSPA5 (GRP78/BiP) interacts with Nsp2 and Nsp4, HYOU1 interacts with Orf8, PRKCSH and MAPRE1 interact with Orf3, and BZW2 interacts with the viral M protein. HSPA5/BiP (binding immunoglobulin protein) has been consistently identified by DS-affinity in our previous studies, and we have also recently reported that DS-BiP association plays important roles in regulating precursor autoreactive B1 cells [8]. HYOU1 (hypoxia upregulated protein 1) was also found overexpressed at protein level in the urine of COVID-19 patients and . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint up-regulated at mRNA level in B cells from 4 patients out of a cohort of 7 hospitalized COVID-19 patients [33,48]. HYOU1 belongs to the heat shock protein 70 family, accumulates in the endoplasmic reticulum under hypoxic conditions, and has been shown to be up-regulated in tumors. PRKCSH (glucosidase 2 subunit beta) is an N-linked glycan processing enzyme in the endoplasmic reticulum, and mutations of this gene have been associated with the autosomal dominant polycystic liver disease. MAPRE1 (microtubuleassociated protein RP/EB family member 1) binds the plus-end of microtubules and regulates microtubule cytoskeleton dynamics. BZW2 (basic leucine zipper and W2 domain 2) may be involved in neuronal differentiation and is associated with congenital hypomyelinating neuropathy.
Similar to the 140 DS-affinity protein autoantigen-ome, the 125 COVID-altered DS-affinity proteins are most significantly associated with RNA metabolism and protein folding (Fig. 2B). In addition, they are associated with establishment of protein localization to organelles, kinase maturation complex 1, emerin complex 24, DNA conformation change, spliceosome, cellular response to heat stress, smooth muscle contraction, VEGFA-VESFR2 signaling pathway, prothymosin alpha C5 complex, regulation of protein dephosphorylation, and telomerase pathway (Fig. 2B). Protein-protein interaction network analysis also confirms that the COVID-altered DS-affinity protein network is strongly associated with mRNA processing, translation, chromosome organization, protein processing in the endoplasmic reticulum, CCT/TriC chaperonin, and apoptosis ( Fig. 3).
CALM1 has recently been identified as one of the autoAgs in multisystem inflammatory syndrome in children from SARS-CoV-2 infection [13]. Six 14-3-3 proteins are identified, all of which are autoAgs.

AutoAgs from altered phosphorylation and ubiquitination
Thirty-eight of the 125 COVID-affected DS-affinity proteins have phosphorylation changes in SARS-CoV-2 infection (Fig. 4). Their molecular functions include histone binding (6 proteins), RNA binding (10 proteins), helicase activity (5 proteins), ATP binding (12 proteins), DNA binding (14 proteins), and hydrolase activity (11 proteins). These COVID-altered phosphoproteins are significantly associated with gene expression, chromosome organization, and mRNA metabolism. Chromosome-associated proteins are particularly . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. Ubiquitination is typically the "kiss of death" modification that marks proteins destined for degradation by the proteasome, although ubiquitination may also modulate protein interaction and activity. Intriguingly, we identified UBA1 (ubiquitin-like modifier-activating enzyme 1), which catalyzes the first step in ubiquitin conjugation and plays a central role in ubiquitination, as a ubiquitination-altered DSaffinity autoAg, which is consistent with our previous studies [1,2].

DS-affinity proteins altered in T cells of COVID-19 patients
Because Jurkat cells were established from human T-cell lymphoblastic leukemia, we searched for DSaffinity proteins that were altered in T cells of seven COVID-19 patients [33].  (Table 1). NACA, ACTG1, and SRSF5, which were down-regulated at the mRNA level, are also known autoAgs. EEF1B2 (or EEF1B, elongation factor 1-beta) . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint has not been identified as an autoAg, although other similar elongation factors such as EEF1A and EF2 are known autoAgs (see references in Table 1).
Among up-regulated proteins, LCP1 was up-regulated in CD4+ T cells of 2 patients (out of 4 patients with available data) and in CD8+ T cells of 2 patients (out of 5 patients with available data), with one of the patients having LCP1 up in both CD4+ and CD8+ T cells. Up-regulation of heat shock proteins, particularly HSPA5 and HSP90AA1, was detected in CD4+ T cells of 2 patients and CD8+ T cells of 1 patient. MCM4 upregulation was detected in CD8+ T cells of 3 out of 6 patients. Among down-regulated proteins, NACA was detected in CD4+ T cells of 1 patient and CD8+ T cells of all 3 patients whose data were available. EEF1B2 was down in CD4+ T cells of 3 patients (out of 5 with available data) and down in CD8+ T cells of 2 out of 3 patients. ACTG1 down-regulation was detected in CD4+ T cells of 2 patients and CD8+ T cells of 1 patient.

SRSF5 was down in CD4+ T cells of 3 out of 5 patients.
Among these T-cell-altered proteins, LCP1 and NACA are perhaps most interesting. LCP1 (plastin-2, an actin binding protein) has been found to play a significant role in T cell activation in response to costimulation through TCR/CD3 and regulates the stability of the immune synapse of naïve and effector T cells [50]. NACA (nascent polypeptide-associated complex subunit alpha) binds to newly synthesized polypeptide chains as they emerge from the ribosome, blocks their interaction with the signal recognition particle, and prevents inappropriate targeting of non-secretory polypeptides to the endoplasmic reticulum. NACA is an IgE autoAg in atopic dermatitis patients with chronic skin manifestations [51]. The significance of these T-cell proteins in COVID-19 and autoimmunity merits further study.

Conclusion
In order to establish a comprehensive COVID-19 autoantigen-ome, we have been profiling autoAgs from different cell and tissue types. Compared to other cells we have examined, Jurkat T-cells contain relatively fewer DS-affinity autoAgs than HFL1 lung fibroblasts, A549 lung epithelial cells, HS-Sultan B-lymphoblasts, and HEp-2 fibroblasts. Although cells share numerous autoAgs, each cell type gives rise to unique COVIDaltered autoAg candidates, which may explain the wide range of symptoms experienced by patients with autoimmune sequelae of SARS-CoV-2 infection. We believe that our effort of discovering autoAgs across different cell types provides a comprehensive and valuable autoAg database for better understanding of autoimmune diseases and post-COVID-19 health problems.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint

Jurkat T-cell culture
The human T lymphoblast Jurkat cell line was obtained from the ATCC (Manassas, VA) and cultured in complete RPMI-1640 medium. The growth medium was supplemented with 10% fetal bovine serum and a penicillin-streptomycin-glutamine mixture (Thermo Fisher). The cells were grown at 37 °C in a CO2 incubator.

Protein extraction
Protein extraction was performed as previously described [6]. In brief, Jurkat cells were lysed with 50 mM phosphate buffer (pH 7.4) containing the Roche Complete Mini protease inhibitor cocktail and then homogenized on ice with a microprobe sonicator until the turbid mixture turned nearly clear with no visible cells left. The homogenate was centrifuged at 10,000 g at 4 °C for 20 min, and the total protein extract in the supernatant was collected. Protein concentration was measured by absorbance at 280 nm using a NanoDrop UV-Vis spectrometer (Thermo Fisher).

DS-affinity fractionation
The total proteomes extracted from Jurkat cells were fractionated in a DS-Sepharose column [6]. About  was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021.

Protein sequencing was performed at the Taplin Biological Mass Spectrometry Facility at Harvard Medical
School. Proteins in gels were digested with sequencing-grade trypsin (Promega) at 4 °C for 45 min. Tryptic peptides were separated in a nano-scale C18 HPLC capillary column and analyzed in an LTQ linear ion-trap mass spectrometer (Thermo Fisher). Peptide sequences and protein identities were assigned by matching the measured fragmentation pattern with proteins or translated nucleotide databases using Sequest. All data were manually inspected. Proteins with ≥2 peptide matches were considered positively identified.

COVID data comparison
DS-affinity proteins were compared with currently available COVID-19 multi-omic data compiled in the Coronascape database (as of 02/22/2021) . These data have been obtained with proteomics, phosphoproteomics, interactome, ubiquitome, and RNA-seq techniques. Up-and down-regulated proteins or gene transcripts were identified by comparing cells infected vs. uninfected by SARS-CoV-2 or COVID-19 patients vs. healthy controls. Similarity searches were conducted to identify DS-affinity proteins that are up-and/or down-regulated in viral infection at any omic level.

Protein network analysis
Protein-protein interactions were analyzed by STRING [26]. Interactions include both direct physical interaction and indirect functional associations, which are derived from genomic context predictions, high-throughput lab experiments, co-expression, automated text mining, and previous knowledge in databases. Each interaction is annotated with a confidence score from 0 to 1, with 1 being the highest, indicating the likelihood of an interaction to be true. Pathways and processes enrichment were analyzed with Metascape [27], which utilizes various ontology sources such as KEGG Pathway, GO Biological Process, Reactome Gene Sets, Canonical Pathways, CORUM, TRRUST, and DiGenBase. All genes in the genome were used as the enrichment background. Terms with a p-value <0.01, a minimum count of 3, and an enrichment factor (ratio between the observed counts and the counts expected by chance) >1.5 were collected and grouped into clusters based on their membership similarities. The most statistically significant term within a cluster was chosen to represent the cluster.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint

Autoantigen literature text mining
Every DS-affinity protein identified in this study was searched for specific autoantibodies reported in the PubMed literature. Search keywords included the MeSH keyword "autoantibodies", the protein name or its gene symbol, or alternative names and symbols. Only proteins for which specific autoantibodies are reported in PubMed-listed journal articles were considered "confirmed" or "known" autoAgs in this study.

Acknowledgements
We are grateful to Dr. Jung-hyun Rho for his technical assistance to the study. We thank Ross Tomaino and the Taplin Mass Spectrometry Facility of Harvard Medical School for expert service with protein sequencing.

Funding Statement
This work was partially supported by Curandis, the US NIH, and a Cycle for Survival Innovation Grant (to MHR). MHR acknowledges NIH/NCI R21 CA251992 and MSKCC Cancer Center Support Grant P30 CA008748. The funding bodies were not involved in the design of the study and the collection, analysis, and interpretation of data.

JYW is the founder and Chief Scientific Officer of Curandis. MHR is a member of the Scientific Advisory
Boards of Trans-Hit, Proscia, and Universal DX, but these companies have no relation to the study.

Authors' contributions
JYW conducted the study and wrote the manuscript. WZ performed some of the experiments. MWR and VBR assisted with the study and manuscript preparation. MHR consulted on the study and edited the manuscript. All authors have approved the manuscript.
. CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint Table 1. Abbreviations . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made

DS-affinity autoantigens from Jurkat T-cells and their alterations in SARS-CoV-2 infection
The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint . CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint CC-BY 4.0 International license available under a was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint (which this version posted July 7, 2021. ; https://doi.org/10.1101/2021.07.05.451199 doi: bioRxiv preprint