Global analysis of lysine acetylation in the brain cortex of K18‐hACE2 mice infected with SARS‐CoV‐2

Severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) has infected hundreds of millions of people all over the world and thus threatens human life. Clinical evidence shows that SARS‐CoV‐2 infection can cause several neurological consequences, but the existing antiviral drugs and vaccines have failed to stop its spread. Therefore, an understanding of the response to SARS‐CoV‐2 infection of hosts is vital to find a resultful therapy. Here, we employed a K18‐hACE2 mouse infection model and LC‐MS/MS to systematically evaluate the acetylomes of brain cortexes in the presence and absence of SARS‐CoV‐2 infection. Using a label‐free strategy, 3829 lysine acetylation (Kac) sites in 1735 histone and nonhistone proteins were identified. Bioinformatics analyses indicated that SARS‐CoV‐2 infection might lead to neurological consequences via acetylation or deacetylation of important proteins. According to a previous study, we found 26 SARS‐CoV‐2 proteins interacted with 61 differentially expressed acetylated proteins with high confidence and identified one acetylated SARS‐CoV‐2 protein nucleocapsid phosphoprotein. We greatly expanded the known set of acetylated proteins and provide the first report of the brain cortex acetylome in this model and thus a theoretical basis for future research on the pathological mechanisms and therapies of neurological consequences after SARS‐CoV‐2 infection.


INTRODUCTION
The coronavirus disease 2019 (COVID-19) is a long-term global health emergency resulting from the recently identified severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Patients with SARS-CoV-2 mainly exhibit respiratory symptoms [1], but SARS-CoV-2 can also produce and increase the risk of neurological symptoms, such as taste and smell disorders, encephalitis, meningitis, Parkinson's disease (PD), and Alzheimer's disease (AD) [2][3][4]. Understanding these neurological consequences is vital to the diagnosis, therapy, and prognosis of COVID-19 patients, and the K18-hACE2 mouse is a useful model to study the pathogenesis and evaluate interventions for COVID-19 [5].
Protein lysine acetylation (Kac) modification is related to many key cellular processes, such as gene transcription, protein folding, signal transduction, autophagy, and metabolism, and is closely associated with neurological diseases such as AD [6,7]. The acetylomes of various organs and cell lines of humans, mice, rats, and even some plants have been reported [7][8][9][10][11][12][13][14][15][16][17][18][19][20], but the acetylome of COVID-19 patients or in any animal models of COVID-19, including K18-hACE2 mice, have not been analyzed. To explore the relationship between neurological consequences of COVID-19 and acetylation in the brain cortex, we performed the first investigation of acetylome in the cortex of K18-hACE2 mice infected or uninfected with SARS-CoV-2 and studied the differences between the SARS-CoV-2 infection group and the control group.
We detected the largest amount of acetylated proteins in the mouse brain cortex in any study to date, which greatly expands the existing body of acetylome data. In addition, this work lays a foundation for subsequent research on COVID-19 pathogenesis and treatment.

Identification of lysine-acetylated peptides, proteins, and sites
To globally determine the host protein acetylation upon SARS-CoV-2 infection, we assessed the brain cortexes from three normal K18-hACE2 mice and three infected K18-hACE2 mice (control and SARS-CoV-2 infection group, n = 3 per group) and executed label-free quantification acetylome analysis ( Figure 1A). The viral nucleic acid load test results ( Figure S1) showed that the three infected mouse brain cortexes had the same degree of infection. We detected 29,309 lysine acetylated (Kac) peptides (4242 unique Kac peptides with 3829 Kac sites) in 1735 proteins ( Figure 1B), and this number exceeded the numbers of acetylated proteins previously detected in brain tissues of mice [11,12,17]. For further analysis, we selected 3530 quantifiable Class I Kac sites in 1482 proteins (quantifiable proteins) that possessed intensities in more than 50% biological replicates in one group. Each quartile was divided into a class group based on a localization probability from 0 to 1 (localization probability of Class I > 0.75) [21]. In the volcano plot, 201 proteins with upregulated acetylation and 439 proteins with downregulated acetylation were found ( Figure 1F and Table   S1). The validation of the MS data quality indicates the reliability of our data (S19, Figures S2, S3 and S4).

2.2
Lysine acetylation motif characterization, subcellular localization, and pathway enrichment analysis of acetylated proteins in the brain cortex After that, we evaluated 3623 sequences surrounding the identified quantifiable Class I Kac sites and selected the five highest-scored motifs ( Figure 1G). The results suggested a preference for histidine (H) or asparagine (N) at the +1 position relative to acetylation, whereas there was a predilection for glycine (G), leucine (L), and isoleucine (I) was found next to acetylation. Accordingly, a motif analysis determined that the KacH and KacN motifs were the most conservative sequences of the Kac site. A comparison of these motifs with the acetylated sequence motifs of other mouse tissues or cell lines [11,[13][14][15][16][17][18][19] and species [7][8][9]12] indicated that KacH and KacN motifs are both found in the mouse hippocampus [17] and that the KacH motif is also found in the testis and hepatocytes of mice, which suggested that different preferences for motifs of Kac sites in different species and tissues.
We subsequently conducted a subcellular localization analysis of quantifiable proteins, which were mostly in the cytoplasm, membrane, nucleus, and mitochondrion ( Figure 1H). Because many acetylated proteins belonged to multiple compartments, there were considerable overlaps among these major cellular components ( Figure 1I). These results were different from the findings obtained in other species and tissues [7,9,13,17,18], showing the differences in the subcellular localization of acetylated proteins among different species and cell types.
To understand the metabolic and human disease pathways of acetylated proteins in the brain cortex, a Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was conducted for quantifiable acetylated proteins. We found that pathways of a variety of neurological disorders, such as PD, amyotrophic lateral sclerosis (ALS), Huntington's disease (HD), prion disease, and AD were especially prominent for acetylated proteins ( Figure 1J and Table S4). HD and PD pathways frequently appeared in other acetylation-related studies for different species or tissues [7-9, 11, 13, 14, 16], suggesting that acetylation-regulated pathways were relatively conserved, although many pathways were different. Additionally, the Kac motif, subcellular localization, and KEGG pathway enrichment analyses of the overall, control, and SARS-CoV-2 infection groups showed little difference (Figures S6, S7, and Tables S5 and S6).

Subcellular localization and functional categories of proteins with upregulated and downregulated acetylation reacting to SARS-CoV-2 infection
To gain a deeper understanding of proteins with upregulated and downregulated acetylation during SARS-CoV-2 infection, a series of F I G U R E 1 Overview of acetylome analysis in the brain cortex. (A) Workflow for the K18-hACE2 mouse brain cortex acetylated proteome analysis study. (B) Identified spectra, peptides, proteins, and lysine acetylation (Kac) sites for the acetylome. The control and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection were used to analyze the label-free acetylome (each group has three biological replicates.). Venn diagram of (C) identified acetylated proteins, (D) identified unique acetylated peptides, and (E) identified Kac sites in the two groups.  ) with a confidence score ≥0.9. The "common proteins" represent Kac proteins in all five pathways. Red circles represent proteins with upregulated acetylation, blue circles represent proteins with downregulated acetylation, and green circles represent proteins that have both upregulated and downregulated acetylation. If a protein was present in two or more pathways in KEGG analysis, only the pathway with the lower p value was included in the PPI. those with downregulated acetylation were also concentrated in carbon metabolism, systemic lupus erythematosus, and HD [12]. Overall, we could speculate that the pathways involving the acetylated proteins showing differential abundance were somewhat conserved in the brain tissues of infected mice, which might help study acetylation in tissues infected with other pathogens.
In  nonstructural proteins (nsp), and nine accessory proteins [22]. The N protein has an N-terminal (NTD) and a C-terminal (CTD) domain, which is located before, between, and after flexible and intrinsically disordered regions (IDRs). We found only one acetylation site, K346, in the CTD of the N protein ( Figures 3A, B). The CTD binds to RNA and connects the RNA to the envelope [23,24]. Therefore, we hypothesized that the acetylation of the CTD may facilitate the combination of CTD and RNA, and the assembly of the SARS-CoV-2 RNA into ribonucleocapsid complexes and viral particles, which promotes viral replication in vivo.

The acetylation of the SARS-CoV-2 N protein, and PPI networks among SARS-CoV-2 viral proteins and mice proteins with changed acetylation
According to a previous study about PPI among SARS-CoV-2 and human proteins [23], we matched differentially expressed acetylated proteins of mice to humans by the blast and constructed networks among SARS-CoV-2 proteins and mice proteins in this study ( Figure 3C and Table S14). We found 61 differentially expressed acetylated proteins interacted with 26 viral proteins. Some acetylated proteins are involved in neurological diseases pathways, mostly in mitochondria, such as voltage-dependent anion-selective channel protein1-3 (Vdac1-3), cytochrome b-c1 complex subunit 7 (Uqcrb), ATP synthase subunit (Atp5b, Atp5f1, Atp5o), and so on. The N protein interacts with heterogeneous nuclear ribonucleoprotein A3 (Hnrnpa3), which is shown in the pathway of ALS, The N protein also interacts with histone H1 (Hist1h1e, Hist1h1b) called linker histone, which is vital for various diseases, such as cancers, AD, and viral infection [25].

Replication-transcription complex (RTC) performs all RNA synthesis
and is composed of nsp7, nsp8, nsp9, nsp12, and nsp13 [24]. According to the UniProt protein annotations, these nsps interact with a series of acetylated proteins that play roles in ATP synthesis, endoplasmic reticulum translation, composition of cytoskeleton, cell proliferation, and cell migration. In summary, these viral proteins interact with a variety of acetylated proteins, which may be drug targets to treat SARS-CoV-2 infection.

DISCUSSION AND CONCLUSION
SARS-CoV-2 infection may increase the risk or potentiate the severity of some neurodegenerative disorders [3,26]. However, the reliability of this assumption and the exact molecular mechanisms of these neurodegenerative disorders remain to be researched. In this study, the acetylomes of brain cortexes for K18-hACE2 mice with and without infection were detected and we identified 3829 acetylated sites in 1735 proteins. Compared with the number of acetylated proteins found in other mouse tissues and cell lines in previous studies [13-16, 18, 19], we found 473 newly identified acetylated proteins in our study ( Figures S5A, B, and Table S2). KEGG analysis manifested that these proteins strongly took part in glutamatergic synapse and synaptic vesicle cycle ( Figure S5C and Table S3) and were different from acetylated proteins in other tissues [13][14][15][16]18]. The results manifested that the types of acetylated proteins differed among different tissues and participated in specific pathways.
The microtubule-associated protein tau (Mapt) with upregulated acetylation played a vital role in the PD and AD pathways, particularly in the latter (note that tau is attributed to both PD and AD in Figure 2E, Table S13). AD is the most common cause of demen- and K586 acetylation were identified for the first humans and mice [17,27,28], which corresponded to K190 and K294 in human tau protein.
People with PD exhibit motor dysfunction and features of this disease are dopaminergic neurons in the substantia nigra dense region and Lewy bodies (mainly consisting of α-synuclein) [29]. The N-terminal acetylation of α-synuclein (Snca), which interacts with tau and has an upregulated Kac site, as shown in Figure 2E, can affect the stability, protein levels, and neuronal toxicity of α-synuclein [30,31]. However, other acetylated sites in α-synuclein have rarely been reported. We and HBO1, and MOF) [6,32]. It is important to note that the selfacetylation of KATs is common and seems to function differently for different KATs. For example, the histone acetyltransferases p300 and CBP can self-acetylate at various lysine residues and the autoacetylation at specific sites can lead to increased basal activity [32,33].
Moreover, lysine deacetylases (KDACs), reversing acetylation, can be classified into two major families: Zn 2+ -dependent deacetylases and NAD + -dependent sirtuin deacetylases [6]. In this study, we also  (Table   S15). In contrast, only one acetylated site in KAT (K13 in the histone acetyltransferase p300) was upregulated. This may be the main reason for the reduction in the number of acetylated sites we identified after SARS-CoV-2 infection ( Figure 1B).
We note that our study still has some limitations. Due to the shortage of COVID-19 patients' tissues, we used brain cortexes from transgenic K18-hACE2 mice, but the acetylomes of mice infected with SARS-CoV-2 are somewhat different from those of humans, which reduces the reliability of our results. Besides, the underlying mechanism of how acetylation affects neurological consequences after SARS-CoV-2 infection was not investigated here. We would study the mechanism and find effective treatments for the neurological consequences of COVID-19 (Supporting method).
In conclusion, this study provides comprehensive acetylome data of the brain cortex of the K18-hACE2 mouse model after SARS-CoV-

CONFLICT OF INTEREST STATEMENT
The authors have declared no conflict of interest.