key: cord-0732112-32344io6
authors: Farley, Scotland E.; Kyle, Jennifer E.; Leier, Hans C.; Bramer, Lisa M.; Weinstein, Jules; Bates, Timothy A.; Lee, Joon-Yong; Metz, Thomas O.; Schultz, Carsten; Tafesse, Fikadu G.
title: A global lipid map reveals host dependency factors conserved across SARS-CoV-2 variants
date: 2022-02-15
journal: bioRxiv
DOI: 10.1101/2022.02.14.480430
sha: b2fe65c4439e46d7e7e2a8ac0f6c58fffcc7b35c
doc_id: 732112
cord_uid: 32344io6

A comprehensive understanding of host dependency factors for SARS-CoV-2 remains elusive. We mapped alterations in host lipids following SARS-CoV-2 infection using nontargeted lipidomics. We found that SARS-CoV-2 rewires host lipid metabolism, altering 409 lipid species up to 64-fold relative to controls. We correlated these changes with viral protein activity by transfecting human cells with each viral protein and performing lipidomics. We found that lipid droplet plasticity is a key feature of infection and that viral propagation can be blocked by small-molecule glycerolipid biosynthesis inhibitors. We found that this inhibition was effective against the main variants of concern (alpha, beta, gamma, and delta), indicating that glycerolipid biosynthesis is a conserved host dependency factor that supports this evolving virus.

Main Text: SARS-CoV-2 interacts with host membranes at every stage of its life cycle. It di-27 rectly crosses the plasma membrane to enter the cell, replicates inside host-derived membrane 28 compartments, acquires its envelope from the host, and traffics through the Golgi and lysosome 29 to exit the cell. All viruses, by their nature, are wholly dependent on host pathways to meet their 30 metabolic, structural, and trafficking needs, and to be effective, they must modulate these host 31 pathways in some way. One dramatic example of this is the way in which SARS-CoV-2 re-engi-32 neers the host internal membranes into double-membraned vesicles (DMVs) and regions of con- 33 voluted membrane (CM) to facilitate its replication (1, 2). This general pattern of membrane re-34 arrangements is common among (+)-stranded RNA viruses (3-5), although the specific structures 35 vary by species. In flaviviruses such as Zika virus (6) and dengue virus (7), these large-scale 36 membrane alterations are accompanied by vast and varied changes at the molecular lipid level. 37

There are many preliminary lines of evidence suggesting that manipulation of host lipids may be 38 a fundamental feature of SARS-CoV-2 infection. Several lipids and lipid-associated proteins 39 have been identified as biomarkers of infection, including VLDL and HDL particles (8), steroid 40 hormones and various apolipoproteins (9), while both elevated triacylglycerol (TAG) (10) and 41 polyunsaturated free fatty acids (11) have been implicated as markers of severe disease out-42 comes. Furthermore, metabolic disorders such as obesity, diabetes, and hypertension have been 43 described as key risk factors among patients who develop severe disease (12). These observa-44

tions indicate systemic changes in lipid metabolism at an organismal level, but it is still unknown 45 how the virus alters the host lipid metabolism at a cellular level, and how these changes support 46

the viral life cycle. 47

We hypothesized that SARS-CoV-2 would reprogram host lipid biosynthesis, and that the virus 48 would depend on specific host metabolic pathways to survive and replicate effectively. To obtain 49 a comprehensive understanding of how SARS-CoV-2 remodels the cellular lipid composition, 50

we performed a detailed lipid survey of both infected cells and cells ectopically expressing indi-51

vidual SARS-CoV-2 proteins, assessing changes in host lipid composition as a result of infection 52

and as a result of the activity of specific viral proteins. Based on our initial results, we examined 53 lipid droplet flux during infection, and further interrogated the requirements for specific host li-54 pids using small-molecule inhibitors of glycerolipid biosynthesis in multiple strains of SARS-55

CoV-2. 56 57

We performed global lipidomic profiling of HEK293T cells overexpressing the ACE2 protein, 59

which we infected with SARS-CoV-2 or mock-infected for 24 hours (Fig 1A) . Each condition 60 was repeated in biological quintuplicate. Total cellular lipids were extracted following the 61 method of and analyzed by liquid chromatograph electrospray ionization tan-62 dem mass spectrometry (LC-ESI-MS/MS). The abundances of the identified lipids were normal-63

ized by comparison to internal standards for quantitative analysis. In total, we identified 514 64 unique lipids spanning the glycerolipid, phospholipid, sphingolipid, and acylcarnitine categories 65

(Supplementary Data 1). Of these, 409 (79.6%) were statistically altered between SARS-CoV-2 66 and mock infection (Benjamini-Hochberg adjusted p < 0.05, analysis of variance [ANOVA] 67 test), changing between 2-and 64-fold in response to infection. Principal component analysis 68 (PCA) of these observations confirmed that infection status accounted for most of the changes 69 ( Fig 1B) , with the five infected samples and the five mock samples falling into two distinct 70 groups. 71

We then examined how these changes in host lipid composition broke down based on class and 72 acyl chain. Glycerolipids and phospholipids showed the largest and most significant changes 73 ( Fig 1C and Fig S1) , with increasing triacylglycerol (TAG) and decreasing cardiolipin (CL) be-74

ing the most altered. Examining the nature of the individual lipid species that changed in more 75 detail (Fig 1C) , we observed that the TAG species change based on their fatty acid composition. 76 TAG species that bear polyunsaturated fatty acid (PUFA) chains were increased an average of 8-77 fold more than saturated or monounsaturated species. This trend was also observed in phospho-78 lipids: saturated phospholipids (phosphatidylcholine, PC; phosphatidylethanolamine, PE; phos-79 phatidylglycerol, PG; phosphatidylinositol PI) almost universally decreased, while many polyun-80 saturated species increased, notably P-PC (phosphatidylcholine, plasmalogen-linked) by 2.7-fold 81

, PC by 1.5-fold, and PG by 1.7-fold. Other notable phenotypes included a decrease in lysolipids 82 (Lyso-PE by 5.1-fold and Lyso-PC by 3.1 fold), a decrease in CL by 5.9-fold, and an increase in 83 ceramide (Cer) by 1.8-fold. 84 85

The genome of SARS-CoV-2 encodes 29 individual proteins (Fig 2A) . Some of these proteins 87 have been directly studied for SARS-CoV-2, but the roles of most of them must be extrapolated 88 by comparison with the proteins of SARS-CoV, which are better studied. With two substantial datasets of virus-induced lipid changes, we sought to link the changes ob-130  served in live virus infection to the action of specific viral proteins. First, we performed unsuper-131 vised clustering of the normalized lipid species observed in the protein-transfected dataset by t-132 SNE ( Fig 3A) . While most phospholipids did not cluster substantially, TAG, in particular, 133 formed distinct clusters, and, in an echo of the live virus phenotype, saturated species and poly-134 unsaturated species clustered separately. Of note, two other molecular features of infection 135

-Cer and CLalso sorted into distinct clusters. 136

The dominant features of lipid remodeling in live virus infection were an increase in TAG, Cer, 137

and phospholipids bearing polyunsaturated fatty acyl chains; and a decrease in lysolipids, DAG, 138 CL, and saturated phospholipids ( Fig 3B) . In order to assess each viral protein for its ability to 139 produce these changes, the average fold change for each of these classes was calculated for each 140 condition ( Fig 3C) . Once again we saw that the virus has multiple proteins that influence remod-141 eling of the lipid environment of its host cells, suggesting a distinct role for each viral protein.

Each feature of infection was recapitulated by at least one protein, and different proteins appear 143 to be responsible for different aspects of the live virus lipid phenotype. 144

In particular, TAG increase was recapitulated by six proteins (orf6, nsp13, nsp5, orf9c, nsp1, and 145 nsp11). Cer increase was recapitulated by six as well (nsp6, orf6, nsp5, orf9c, orf3a, and orf7a), 146

and polyunsaturated PC (both ether-and ester-linked) increase was recapitulated by four (orf6, 147 orf9c, orf9b, and E). Of note, orf6 and orf9c recapitulated all three of these distinctive altera-148 tions, and also recapitulated the most individual phenotypes of any protein. fection. TAG is produced through the acylation of DAG by DGAT1 or DGAT2, where it is then 153 sequestered in lipid droplets that can be accessed as a source of fatty acids. TAG breakdown is 154 the result of several lipases that remove an acyl chain to produce DAG ( Fig 4A) lipid droplets and an anti-dsRNA antibody to mark the site of viral replication ( Fig 4B) . We see a an average of 6.7 at 24hpi, to an average of 21.5 at 48hpi. Lipid droplet area increases from zero 171 pixels per droplet at 8hpi, to an average of 177 pixels per droplet at 24hpi, to an average of 400 172 pixels per droplet at 48hpi. However, there does not appear to be any colocalization of the lipid 173 droplets and dsRNA, suggesting that the virus is not using lipid droplets directly as a platform 174

for replication ( Fig 4E) . 175

To further validate these observations, similar experiments were performed in the human epithe-176 lial Caco2 cell line. Here, a slight increase in lipid droplet number was observed, from an aver-177 age of 8 lipid dropletss per cell at 8 and 24hpi, to an average of 15.9 lipid droplets per cell at 178 48hpi, although the increase was not significant. Lipid droplet area, however, did significantly 179 increase throughout the course of infection, to a similar degree as in HEK293T-ACE2 cells, from 180 an average of 136.5 pixels per droplet at 8hpi to an average of 192.5 pixels per droplet at 24hpi 181

to an average of 431.1 pixels per droplet at 48hpi (Fig 4F-G) . Once again, colocalization with 182 dsRNA was not observed (Fig 4H) . inhibition is more effective than inhibiting only one lipase. The importance of DAG production 204

to the virus, perhaps as a precursor to TAG, is indicated by the efficacy of U-73122 ( Fig 5F) , 205

which inhibits phospholipase-C-dependent processes (49). 206

To directly compare the inhibitors of central glycerolipid metabolism, we designed a more de-207 tailed study to test a range of concentrations for each inhibitor. We tested a range of two-fold di-208 lutions of each compound, and in parallel with the focus-forming assay to assess viral replica-209 tion, we performed a resazurin-based cytotoxicity assay to verify that any deficiency in viral pro-210 duction was not due to impaired cell viability ( Fig S4) . The most effective inhibitor by about 211 fifty-fold was GSK2194069 (EC50 = 1.8 nM, 293T-ACE2). GSK2194069 blocks FASN, sug-212 gesting that de novo lipid synthesis is strictly required for viral survival. Orlistat followed in effi-213 cacy (EC50 = 94 nM, 293T-ACE2), highlighting the importance of both fatty acid synthesis and 214

lipolysis to the virus. The other broad-spectrum lipase inhibitor, CAY10499 (EC50 = 157 nM, 215 283T-ACE2) had a similar efficacy to PF04620110 (EC50 = 490 nM, 293T-ACE2). Atglistatin, 216 the most specific lipase inhibitor, became cytotoxic before complete inhibition was achieved, and 217 so an EC50 could not be calculated; certainly it is higher than 10 µM, showing again that the vi-218 rus is not dependent on the activity of one specific lipase, but rather on a certain lipid composi-219 tion. Taken together, these results indicate a profound dependence on host lipid metabolism, and 220

in particular glycerolipid flux. The de novo synthesis of TAG is required, as is the ability to re-221 lease the fatty acids sequestered in this neutral storage lipid through lipolysis. 222 223 Glycerolipid biosynthesis is a conserved host dependency factor for variants of SARS-CoV-224 2 225

Given that our most effective inhibitors all relate in some way to the dynamics of TAG produc-226 tion, we hypothesized that their efficacy is due to the virus's specific requirements for lipid drop-227 lets. We performed microscopy of cells treated with selected inhibitors at 10 µM overnight ( does not support replication: pure accumulation of TAG resulting from the inhibition of lipolysis 240

is as detrimental to infection as preventing its synthesis. 241 SARS-CoV-2 interacts with host lipids at every stage of its life cycle. To rule out the possibility 242 that glycerolipid metabolism is necessary for the initial attachment and endocytosis of the virus, 243

we performed an entry assay using S-pseudotyped lentivirus. For this experiment, lentiviruses 244

were generated that display the SARS-CoV-2 S protein and carry a GFP reporter; lentiviruses 245 coated instead with the VSV G protein were used as a control. Successfully infected cells express 246 GFP, and quantitative microscopy was used to assess infection ( Fig S5B) . 293T-ACE2 cells 247

were treated overnight with selected inhibitors of glycerolipid biosynthesis and then infected 248 with either of these two lentivirus constructs. We did not observe a significant reduction in viral 249 entry in the presence of any of the inhibitors tested, suggesting that the virus depends upon this 250 lipid biosynthetic pathway to facilitate the intracellular stages of its life cycle (Fig 6C) . 251

The continued global transmission of SARS-CoV-2 has led to the emergence of variants of con-252 cern (VOC) that show evidence of increased transmissibility (50)or resistance to prior immunity 253

(51, 52) ( Fig 6D) . The major VOCs include the B. strain, to infect cells that had been pre-treated overnight with 10 µM CAY10499, GSK2194069, 267 PF04620110, and Orlistat, and assessed viral proliferation by focus forming assay. We per-268

formed these experiments in both 293T-ACE2 cells and Caco2 cells. We observed very few dif-269 ferences in efficacy of the compounds among the four strains tested ( Figure 6E and Fig S6) propose here a model for how SARS-CoV-2 uses lipid droplets to support infection ( Figure 6F ). 283 We show that lipid droplet proliferation is a consequence of infection, and that both TAG synthe-284 sis and lipolysis are required for effective replication. The lipid droplet phenotype appears to be 285 part of a profound reprogramming of cellular lipid metabolism which is induced directly by indi-286 vidual viral proteins; strikingly, polyunsaturated lipids are dramatically increased while saturated 287 lipids are decreased, suggesting that viral membrane structures require a particularly high level 288 of fluidity. While lipid droplets do not appear to be parts of the viral replication complex, given 289 the very low levels of colocalization between dsRNA and BODIPY in infected cells, it seems 290 likely that their roles in buffering lipid levels and facilitating membrane plasticity support the 291 ambitious coronaviral membrane rearrangements. 292

Using small-molecule inhibitors of glycerolipid metabolism, we showed that SARS-CoV-2 fun-293 damentally requires host lipid metabolic pathways for its survival and proliferation. Our findings 294

highlight the dynamic and specific involvement of host lipids in infection: SARS-CoV-2 requires 295 both de novo fatty acid and TAG synthesis, and lipolysis, simultaneously promoting lipid synthe-296 sis and providing specific lipids for viral processes. We further showed that these inhibitors work 297 as effectively against the recently emerging SARS-CoV-2 variants of concern as they do against 298 the original WA1 strain, demonstrating the advantage of designing host-targeted therapeutics 299 against a conserved host dependency pathway. 300

Our findings fill an important gap in our understanding of host dependency factors of corona-301 virus infection. Our systematic analysis of the protein-by-protein effect on host lipids reveals a 302 complex network of many individual viral proteins responsible for diverse aspects of host lipid 303

remodeling. Both of our lipidomics datasets are resources for understanding cellular disease pa-304 thology and suggest potential directions for therapeutic discovery, highlighted by the success of 305 several inhibitors of glycerolipid biosynthesis in blocking viral replication. In light of the evolv-306

ing nature of SARS-CoV-2, it is critical that we understand the basic biology of its life cycle in 307 order to illuminate additional avenues for protection and therapy against this global pandemic 308 pathogen, which spreads quickly and mutates with ease. 309 water, and 10 µL of an internal standard cocktail (Avanti EquiSPLASH) was added. Extracts 369

were left for one hour at 4 ˚C, then the layers were separated by centrifugation (3,000xg for 10 370 minutes), and the chloroform layer was moved to a fresh tube. 2 mL fresh chloroform was added 371

to the aqueous layer, mixed, left for one hour at 4 ˚C, separated by centrifugation, and then added 372

to the first chloroform layer. The combined chloroform layers were dried under a stream of nitro-373 gen. These dried extracts were frozen at -80 ˚C and sent to PNNL on dry ice. fragmentation. Lipid identifications were made using previously outlined fragment ions (60). 386

The LC-MS/MS raw data files were analyzed using LIQUID (60), and then all identifications 387 were manually validated by examining the fragmentation spectra for diagnostic and fragment 388 ions corresponding to lipid acyl chains. Identifications were further validated by examining the 389 precursor ion isotopic profile and mass measurement error, extracted ion chromatogram, and re-390 tention time for each identified lipid species. To facilitate quantification of lipids, a reference da-391 tabase for lipids identified from the MS/MS data was created, and features from each analysis 392

were then aligned to the reference database based on their m/z, and retention time using MZmine 393 2 (61). Aligned features were manually verified, and peak apex-intensity values were reported 394 for statistical analysis. 395 396

Lipidomics -QC, normalization, and statistical comparison methods 397

Lipidomics data were collected in positive and negative ionization mode and analyzed using R. 398

Each ionization mode datasets was normalized using an IS specific to the respective ionization 399 mode. We required that an IS be quantified for every sample to be considered for normalization 400 purposes. Further, normalization factors should not be related to the biological groups being 401 compared to avoid the potential introduction of bias into the data. Single concentration screen for replication inhibition (all strains of SARS-CoV-2) 443

The highest concentration for each inhibitor that did not cause cytotoxicity was selected for this 444 assay. For most described inhibitors 10 µM was used, except for MAF (100 µM), and remdesivir 445

(2 µM). Each cell line (Caco2 or 293T-ACE2) was seeded in 96-well plates at a density of 446 10,000 cells per well and treated overnight with each inhibitor prior to infection with SARS-447

CoV-2 with an MOI of 0.1. The infection was continued for 48 hours. To quantify viral produc-448 tion, focus-forming assays were performed on the supernatants, described in detail below. 449 450

Pseudovirus lentivirus production 451 293T cells were seeded at 2 million cells/dish in 6cm TC-treated dishes. The following day, cells 452

were transfected as described above with lentivirus packaging plasmids, SARS-CoV-2 S plas-453 mid, and IzGreen reporter plasmid (66). After transfection, cells were incubated at 37 ˚C for 60 454

hours. Viral media was harvested, filtered with a 0.45 µm filter, then frozen before use. Virus 455 transduction capability was then determined by fluorescence using a BZ-X700 all-in-one fluores-456 cent microscope (Keyence), and a 1:16 dilution of viral stocks was found to be optimal for neu-457 tralization assays. 458 459

Pseudovirus entry assay 460

Neutralization protocol was based on previously reported experiments with the SARS-CoV-2 S 461 pseudotyped lentivirus (66). 293T-ACE2 cells were seeded on tissue-culture-treated, poly-lysine 462 treated 96-well plates at a density of 10,000 cells per well. Cells were allowed to grow overnight 463 at 37 ˚C, and then treated with selected inhibitors as described above for live virus infection. Lz-464

Green SARS-CoV-2 S pseudotyped lentivirus was added to 293T-ACE2 cells treated with 5 465 µg/mL polybrene and incubated for 48 hours before imaging. Cells were fixed with 4 % PFA for 466 1 hour at room temperature, incubated with DAPI for 10 minutes at room temperature, and im-467 aged with BZ-X700 all-in-one fluorescent microscope (Keyence). Estimated area of DAPI and 468 GFP fluorescent pixels was calculated with built-in BZ-X software (Keyence). There were five 469 biological replicates for each condition, and the biggest outlier was removed from analysis due to 470 inherent variability in the assay. 471 472

Measurement of compound EC50 473

Compounds from the single concentration screen that showed efficacy against SARS-CoV-2 rep-474 lication were tested to measure compound EC50. The cell line of interest (293T-ACE2 or Caco2) 475 was seeded in 96-well plates at a density of 10,000 cells per well, and treated overnight with 2-476 fold dilutions of each compound, starting from 50 µM for Atglistatin, PF04620110, 477 GSK2194069, and CAY10499, and starting at 1 µM for Orlistat. Each condition was tested in 478 quadruplicate. The next day cells were infected as described above, and the infection was contin-479 ued for 48 hours, and then the supernatants were used in a focus forming assay, as described be-480 low. 481 482

Focus forming assay 483

Vero E6 cells were seeded in a 96-well plate at a density of 20,000 cells per well. CoV-2 serum, diluted 1:5,000 in perm buffer) for either 2hr room temperature or overnight at 4 496 ˚C. Antibody was removed and plates were washed 3 x 5 minutes with 200 µL/well PBST (0.1% 497 tween in PBS). Plates were incubated with 50 µL secondary antibody (anti-llama HRP, goat IgG) 498

for either 2hr room temperature or overnight at 4 ˚C. Antibody was removed and plates were 499

washed 3 x 5 minutes with 200 µL/well PBST. Plates were stained with 50 µL/well TrueBlue pe-500 roxidase substrate for 30 minutes. Foci were imaged on an ImmunoSpot S6 Macro ELISPOT im-501 ager, and then counted using the Viridot R package (67). 502 503

Quantification and Statistical Analysis 504 EC50 values were calculated using the Hill equation in the R software package. Unless otherwise 505 stated, P values are from one-way ANOVA tests without adjustments for multiple comparisons, 506

with P < 0.05 considered statistically significant. 507 508

Data and Code Availability: The raw lipidomics datasets generated during this study have been 509 deposited and will be available at ftp: 

Representative images of HEK293T-ACE2 cells treated with each indicated inhibitor (10 µM) or vehicle (DMSO), infected with SARS-CoV-2 (MOI = 1), and stained to visualize lipid droplets (BODIPY 493/503), and dsRNA. Images are representative of three independent experiments. (B) Quantification of lipid droplet numbers in (A). Data are mean ± SE; * p < 0.05, ** p < 0.01, *** < 0.001, one-way ANOVA (C) GFP fluorescence resulting from an infection with lentivirus pseudotyped with either SARS-CoV-2 Spike protein or VSV G protein.

Estimated area of DAPI and GFP fluorescent pixels was calculated with built-in BZ-X software (Keyence) and GFP fluorescence was normalized to the DAPI signal for each condition. There were five biological replicates for each condition, and the biggest outlier was removed from analysis due to inherent variability in the assay. Data are mean ± SD. (D) A model for neutral lipid flux during SARS-CoV-2 infection. All SARS-CoV-2 genomes enter the cytosol (shown here by direct fusion with the plasma membrane; endosomal entry has also been reported, especially in Vero E6 cells, but is thought to be less physiologically relevant to human infection). Viral proteins are expressed by host metabolic machinery; orf6, orf9c, nsp1, nsp5, and nsp13 all directly induce TAG formation via DGAT1, which is inhibited by PF04620110. Lipid droplets proliferate following infection, and are also sources for raw lipid material released by lipolysis, which is inhibited by CAY10499 and orlistat. These raw materials may be sources of energy, signaling mediators, and lipids for the creation of viral replication complexes. Assembled virions are trafficked through the Golgi and released from the cell by lysosomal exocytosis. (E) Unrooted phylogenetic tree of SARS-CoV-2 variants of concern, generated by Nextstrain (68,69) , an open-source repository of pathogen genomic data. (F) Inhibition of the original strain and four variants of concern of SARS-CoV-2 in 293T-ACE2 cells by four inhibitors of glycerolipid biosynthesis, each at 10 µM overnight prior to an 48hour infection. Data are from three independent experiments; data are mean ± SE.

A molecular 516 pore spans the double membrane of the coronavirus replication organelle

SARS-CoV-2 struc-518 ture and replication characterized by in situ cryo-electron tomography

Composition and three-dimen-520 sional architecture of the dengue virus replication and assembly sites

+) RNA virus replication compartments: a safe home for (most) viral replication

Ultrastructural Characteriza-524 tion of Zika Virus Replication Factories

A global lipid map defines a net-526 work essential for Zika virus replication

The role of lipids in 528 the inception, maintenance and complications of dengue virus infection

Proteomic and Metabolomic Characterization

Low HDL and high triglycerides 535 predict COVID-19 severity

High plasma concentration of 537 non-esterified polyunsaturated fatty acids is a specific feature of severe COVID-19 pneumonia

Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area

A rapid method of total lipid extraction and purification

A unifying 545 structural and functional model of the coronavirus replication organelle: Tracking down RNA synthesis

Ultrastructural Characteriza-548 tion of SARS Coronavirus

Membrane remodeling by SARS-CoV-2 -double-enveloped viral replication

Apoptosis-induced changes in mitochondrial lipids

Lipid droplet functions beyond energy storage

The lipid droplet is an important 596 organelle for hepatitis C virus production

Hepatitis C virus core protein, lipid droplets and steatosis

Dengue Virus-Induced Autophagy Regulates Lipid Metabolism. Cell Host & Mi-600 crobe

BODIPY 493/503 Staining of Neutral Lipid Droplets for Microscopy and Quantifica-602 tion by Flow Cytometry

Analysis of Cell Viability by the alamarBlue Assay. Cold Spring Harb 604

Growth, detection, quantification, and inactivation 606 of SARS-CoV-2

A human fatty acid 608 synthase inhibitor binds beta-ketoacyl reductase in the keto-substrate site

Inhibition of pancreatic lipasein vitroby the covalent inhibitor tetrahy-610 drolipstatin

Orlistat Is a Novel Inhibitor of Fatty Acid Synthase with 612

Discovery of PF-04620110, a Potent

Bioavailable Inhibitor of DGAT-1

Hormone-sensitive lipase: control of intracellular tri-(di-)acylglycerol and choles-616 terol ester hydrolysis

CAY10499, a novel monoglyceride lipase inhibitor evidenced by an 618 expeditious MGL assay

Development of small-620 molecule inhibitors targeting adipose triglyceride lipase

Receptor-Coupled Phospholipse-C-Dependent Processes in Human Platelets and Polymorphonuclear Neutrophils

reinfection, and transmissibility associated with the SARS-CoV-2 variant B.1.1.7: an ecological study. Lancet Pub-626 lic Health

Escape of SARS-CoV-2 501Y.V2 from 628 neutralization by convalescent plasma

Increased resistance of SARS-CoV-2 vari-630 ant P.1 to antibody neutralization

Fast-spreading U.K. virus variant raises alarms

Emergence and rapid 633 spread of a new severe acute respiratory syndrome-related coronavirus 2(SARS-CoV-2) ilneage with multiple spike 634 mutations in South Africa

Genomics and epidemiology 636 of the P.1 SARS-CoV-2 lineage in Manaus

Reduced neutralization of 638 SARS-CoV-2 B.1.617 by vaccine and convalescent serum

Nextstrain: real-time tracking of 640 pathogen evolution

TreeTime: Maximum-likelihood phylodynamic analysis

CoV-2 protein interaction map reveals targets for drug repurposing

LIQUID: an-open source software for identifying lipids in LC-MS/MS-based lipidomics data

MZmine 2: Modular framework for processing

ualizing, and anlyzing mass spectrometry-based molecular profile data

Controlling the False Discovery Rate: A Practical an Powerful Approach to 665

Multiple Testing

Profiler 3.0: Next-generation image processing for biology

671 Fiji: an open-source platform for biological-image analysis

Analysis of Cell Viability by the alamarBlue Assay

Protocol and Reagents 676 for Pseudotyping Lentiviral Particles with SARS-CoV-2 Spike Protein for Neutralization Assays

Viridot: An 680 automated virus plaque (immunofocus) counter for the measurement of serological neutralizing responses with ap-681 plication to dengue virus

Nextstrain: real-time tracking of pathogen evolution

TreeTime: Maximum-likelihood phylodynamic analysis

Dr. Fikadu Tafesse (tafesse@ohsu.edu). 706