CRISPR/Cas9-Mediated Whole Genomic Wide Knockout Screening Identies Specic Genes Related to Mineral Absorption Involving in PM2.5 Liver Toxicity

Background: PM 2.5 , also known as ne particles, refers to particulate matter with a dynamic diameter of ≦ 2.5 µm in air pollutants, carrying toxic substances (e.g. heavy metals or minerals), which can pass through the alveolar epithelium and enter the bloodstream and tissues, can have more serious health problems, such as non-alcoholic fatty liver and hepatocellular carcinoma. However, the underlying mechanisms for toxic effect of PM 2.5 are poorly understood. Results: Here, we subjected L02 cells to expose of PM 2.5 and performed a pooled genome ‐ wide CRISPR (clustered regularly interspaced short palindromic repeats)/Cas9 (CRISPR-associated protein 9) loss of function screen to discover new potential targets for PM 2.5 . Enrichr and KEGG were employed to identify candidate genes which might involve in PM 2.5 toxicity. We found that ATP1A2(ATPase Na+/K+ transporting subunit alpha 2), MT1M(metallothionein 1M), SLC6A19(solute carrier family 6 member 19) and TRPV6(transient receptor potential cation channel subfamily V member 6) genes contributed to PM 2.5 toxicity involving in mineral absorption pathway. Moreover, we demonstrated that these candidate genes deciency increased the cell viability and attenuated apoptosis in cells explored to PM 2.5 . Conclusions: ATP1A2, MT1M, SLC6A19 and TRPV6 contribute to PM 2.5 toxicity and may be potential therapeutic targets for PM 2.5 related diseases.


Background
Particulate Matter (PM) is a mixture of solid and liquid particles that are suspended in the air. PM 2.5 describes ne inhalable particles, with diameters that are generally 2.5 micrometers and smaller. PM 2.5 have small diameters, however large surface areas, may therefore be capable of carrying various toxic stuffs (such as heavy metals or minerals: Ca, Cu, K, Mg, Mo, Ps, Pt, Se, and Zn), passing through the ltration of nose hair, reaching the end of the respiratory tract with air ow and accumulate there by diffusion or active transport [1], distributing to other organs through blood, damaging other parts of the body [2]. Many respiratory diseases [1], cardiovascular diseases [3], neurodegenerative diseases [4] and hepatic disease [5] were reported to be related to PM 2.5 . PM 2.5 exposure leads to NAFLD which is known as a silent disease and present in around 20-30% of people [6]. What's more, people living in environments with higher PM 2.5 concentrations have a higher incidence rate ratio of hepatocellular carcinoma (HCC) and patients with HCC has a higher mortality rate [7,8]. Some studies discovered the mechanism of PM 2.5 induced liver disease. Oxidative stress and in ammatory in liver cells aggravated by PM 2.5 contributed to hepatic injury by altering normal lipid metabolism [9], farnesoid X receptor(FXR) [10], ROS/PINK1/Parking signal pathway and NADPH oxidase [11,12] may mediate the effects of PM 2.5 exposure on promoting hepatic brosis. However, the material basis and detailed mechanism of PM 2.5 toxicity are still unknown. A comprehensive study is still necessary.
CRISPR (clustered regularly interspaced short palindromic repeats)/ Cas9 (CRISPR-associated protein 9) is an RNA-guided DNA endonuclease and can be easily programmed to target new sites by altering its guide RNA sequence [13,14]. CRISPR/Cas9 is a powerful gene editing tool that edits DNA in a precise, directed manner. This technology to enable thousands of genes to be modi ed and their function assessed in a single experiment. This cutting-edge tool can help identify and validate novel drug targets or study the underlying human disease [15]. CRISPR/Cas9 has been used to study the pathogenesis of NAFLD and is also becoming a promising genetic engineering approach for liver cancer [16,17]. To explore the effect of PM 2.5 exposure, we rstly applied genome-scale CRISPR/Cas9 knockout (GeCKO) screening technology to discover new regulators involved in PM 2.5 resistance in human L02 cell line.
Moreover, some genes were validated as PM 2.5 resistance genes, indicating that they were important contributors and potential therapeutic targets for toxic effect of PM 2.5 .

Results
A genome-wide CRISPR/Cas9-mediated screen to identify PM 2.5 resistance genes The rapid emergence of CRISPR/Cas9-mediated genome editing technologies has now allowed us to discover additional modulators of PM 2.5 . We introduced a GeCKO library containing 123,411 sgRNAs targeting 19,050 human genes into lentiviral vectors and generated a pool of cells, in which every targeted gene theoretically carried a loss-of-function mutation [18]. After treating the pooled cells with PM 2.5 exposure insult, viable cells were collected for analysis of the inserted sgRNAs in genome (Fig. 1A). The sgRNAs ampli ed from the surviving cells' genomes were determined by next-generation sequencing (NGS), and the candidate genes were ranked depending on the number of unique sgRNAs versus NGS reads. As shown in the scatter diagram of sgRNA number and corresponding sequencing reads of genes, the detected genes were well-distributed in every sgRNA (Fig. 1B). Enrichment of multiple sgRNAs was found in our GeCKO screening, suggesting that loss of the cognate genes conferred resistance to PM 2.5 insult. Of the 19,050 genes tested, 614 top-rank genes with high number (4-6) of unique sgRNAs were involved in PM 2.5 toxic effect (Additional le 1). We had also discovered that the de ciency of some interesting genes related to mineral absorption, ATP1A2, ATP1B2, MT1M, SLC6A19 and TRPV6, might involve the protection of PM 2.5 toxic.
KEGG pathway and GO analysis of PM2.5 resistance genes To understand the roles of these candidate genes in apoptosis and their association with PM 2.5 resistance, Enrichr (https://maayanlab.cloud/Enrichr/) [19] was used for KEGG pathway and GO enrichment analysis of top-rank genes with high number (4-6) of unique sgRNAs, and the top 10 signi cant pathways and functions were selected in accordance with p values. KEGG analysis indicated that some genes (ATP1A2, SLC6A19, MT1M, TRPV6 and ATP1B2) were belong to mineral absorption pathway (p = 0.019602244) (Fig. 2, Additional le 2). ATP1A2, ATP1B2, MT1M, SLC6A19 and TRPV6 are all responsible for absorption of metals and minerals(cadmium, zinc, copper and calcium ions,etc.), which means knocking down ATP1A2, ATP1B2, MT1M, SLC6A19 or TRPV6 may keep L02 cells from transporting metals and minerals in PM2.5. What's more, in ammatory mediator regulation of TRP channels (ADCY9, ASIC5, CALML5, CYP4A22, CALML4, IL1RAP, PIK3R1,PLCG1) was the other pathway affected by PM2.5 in our study which indicated that oxidative stress and in ammatory mediates PM2.5-induced cell damage, and it was well known that PM 2.5 could cause oxidative stress and in ammatory in liver cells [9].
The GO analysis demonstrated that PM 2.5 resistance genes were primarily involved in GO biological process: biomineral tissue development(p = 0.0014), skeletal system development(p = 0.0018), inorganic anion transmembrane transport(p = 0.0033), GO molecular function: inorganic anion transmembrane transporter activity(p = 0.0033), inhibitory extracellular ligand-gated ion channel activity(p = 0.0099), chloride channel activity(p = 0.01), transmitter-gated ion channel activity(p = 0.017), and GO cellular component: tertiary granule membrane (p = 0.0069), speci c granule membrane(p = 0.06), bicellular tight junction (p = 0.069)etc. (Fig. 2, Additional le 3). It was obvious that minerals played an important role in biomineral tissue development and skeletal system development [20,21]. We found that other GO functions may also be indirectly related to transcellular transport of ion and minerals (or metals). Thus, it can be seen that PM 2.5 resistance genes may be responsible for mineral toxicity and mineral absorption.
On the other hand, oxidative stress and in ammation are a consequence of the absorption of metal ions in PM 2.5 . In our study, oxidoreductase activity(GO:0016709) and T cell receptor binding (GO:0042608) were found in the GO analysis, which might correspond to PM 2.5 -induced oxidative stress and in ammation [22,23].
Taken together, our GeCKO screen had identi ed the PM 2.5 resistance genes in L02 cells that were involved in mineral absorption pathway and in ammation-related pathway. GO analysis also revealed that PM 2.5 resistance genes involved the physiological functions of minerals and in ammation, oxidative stress. We were surprised to nd that mineral absorption de ciency might be involved in PM 2.5 exposure protection.
Effect of the mineral absorption related genes from GeCKO screen on PM 2.5 -induced apoptosis To validate the top-ranking genes from the GeCKO screen, we next knocked down the expression of the genes by expressing siRNA sequences targeting mineral absorption related genes. The ratios of cell viability and apoptosis of the during PM 2.5 exposure were compared to that of the control cells that transfected an siRNA control. We observed a signi cant increase in the ratio of cell viability during PM 2.5 exposure insult following knockdown of ATP1A2, MT1M, SLC6A19 or TRPV6 ( Figure.3). Meanwhile, knocking down ATP1A2, ATP1B2, MT1M, SLC6A19 or TRPV6 could decrease in the ratio of apoptosis during PM 2.5 exposure (Figure. 4). The ndings for mineral absorption related genes were consistent with the integrative analysis during the genome-wide CRISPR/Cas9-mediated screen, which suggested that these genes might sensitize cells to apoptosis following PM 2.5 exposure. Discussion PM 2.5 particles are widely believed to be associated with NAFLD [5] and HCC ,etc [7]. As PM 2.5 exposure induced the activation of apoptotic pathway, there is no doubt that apoptosis is associated with the formation of these diseases and is one of the important pathological features [24,25]. CRISPR screening facilitates discovery of key genes or genetic sequences that elicit a speci c function or phenotype for a cell type, which is conducted to identify gene involving in PM 2.5 resistance. Of the 19,050 genes tested, our GeCKO screen identi ed 614 genes with high number (4-6) of unique sgRNAs that may be involved in PM 2.5 toxic effect. The precent study demonstrates that PM 2.5 resistance genes were involved in minerals physiological function and mineral absorption, etc., which may be important for liver cell apoptosis and damage.
Although there are some mechanisms of PM 2.5 induced liver deceases are reported, such as in ammatory, ROS and FXR, they focus on studying the physiological changes and molecular pathways in the body after PM 2.5 exposure [9,11,26]. As Qingzhao Li ,etc. reported main heavy metals of PM 2.5 were more likely to be distributed in the liver [2]. The cell permeability and transfer of toxic Chemicals(metals and minerals) constituted PM 2.5 might also play an important role in PM 2.5 for liver cell apoptosis and damage, but were less explored. In present study, genes related to mineral absorption in GeCKO screening were con rmed by expressing siRNA sequences. KEGG analysis of 614 genes with high number (4-6) of unique sgRNAs showed that some genes (ATP1A2, SLC6A19, MT1M, TRPV6 and ATP1B2) were belong to mineral absorption pathway. What's more, GO analysis con rmed that genes involved in PM 2.5 toxic effect might produce many biological functions, such as biomineral tissue development, skeletal system development, inorganic anion transmembrane transport(GO Biological Process), inorganic anion transmembrane transporter activity, inhibitory extracellular ligand-gated ion channel activity, transmitter-gated ion channel activity(GO Molecular Function), and tertiary granule membrane (GO Cellular Component), which directly or indirectly related to mineral absorption.
Knocking down ATP1A2, SLC6A19, MT1M and TRPV6 signi cantly increased cell viability ratio and decreased apoptosis. MT1M is an important member of metallothioneins (MTs) family that contains the short peptide of cysteine and sulfur protein with high a nity to heavy metals, including cadmium, zinc, and copper, and play a vital role in metal ion homeostasis and detoxi cation [27]. TRPV6 is the most Ca 2+selective members of the TRP ion channel family, play an important role in intestinal Ca 2+ absorption. Pharmacological interference with TRPV6 in the digestive system is likely to interfere with calcium absorption and bone mineralization as well as with epithelial cell hyperplasia and early stages of malignancies [28]. ATP1A2 encodes the α2 isoform of the Na + ,K + -ATPase's catalytic subunit, an ion channel/ion transporter. SLC6A19 encodes the amino acid transporter B(0)AT1, mediating neutral amino acid transport from the luminal compartment to the intracellular space [29]. In generally, MT1M, ATP1A2, SLC6A19 and TRPV6 relates to mineral absorption. When knocking down these genes by transfection of siRNA, the cell activity increased and apoptosis decreased compared to control under the effect of PM 2.5 .
As these genes are responsible for mineral absorption, loss of them may stop the absorption of minerals or metals into the cell or imbalance of metal ion homeostasis caused by PM 2.5 , which means these genes/proteins are responsible for transporting metals and minerals involved in PM 2.5 toxic effect to cells.
On the other hand, PM 2.5 may induce oxidative stress and in ammation that indirectly mediated PM 2.5 induced apoptosis and diseases [30]. Studies reported oxidative stress and in ammation in liver cells aggravated by PM 2.5 contributed to hepatic injury by altering normal lipid metabolism [9]. In our GeCKO screening, loss of some genes related to in ammation and oxidoreductase activity, such as IL1RAP [31], PIK3R1, PLCG1 [32], CYP26B1, CYP4F11 [26] and MICAL2 [33], might relieve PM 2.5 toxicity [34].
There are some limits in our study that can be improved and explored in the future. First, we focused on the genes of transporters. We did not specify which metals or minerals transported into cells were responsible for the damage of L02 cells and the mechanism of minerals mediated apoptosis. Second, metals or minerals concentration had not been measured in our study. This work will be done in our lab in the future to elucidate the role of metals or minerals transporters in the PM 2.5 toxicity.
In conclusion, metals are main components of PM 2.5 [35] and play an important role in PM 2.5 -induced apoptosis [36], it is reasonable that mineral absorption results apoptosis of cells which induces and deteriorates deceases. In our study, loss of genes involving to mineral absorption protected the L02 cell, which indicated that metals of PM 2.5 could distribute to liver by transporters and induced apoptosis of several deceases [2]. We can inspect that methods of prevention and treatment may be explored on the genes or proteins of mineral absorption in the future. Conclusion PM 2.5 can damage cells, cause damage to tissues and organs, and cause a very negative impact on human health. This genome-wide CRISPR/Cas9-mediated screen found genes related to mineral absorption (ATP1A2, ATP1B2, MT1M and TRPV6) may be involved in this process, which can provide theoretical support for the prevention and treatment of PM 2.5 -induced damage.

Methods
Lentiviral production of the sgRNA library.
Lentiviral transduction of the sgRNA library.
L02 cells were purchased from ATCC, cultured at 37℃, 5% CO 2 in the DMEM medium containing 10% fetal bovine serum (Invitrogen). 3x10 8 L02 cells were infected with the GeCKO library at a multiplicity of infection (MOI) of 0.1 aiming for ensure that most cells receive only 1 viral construct with high probability in full DMEM supplemented with 10% fetal bovine serum, 4 mM l-glutamine, and 10 µg/ml penicillin and streptomycin in the presence of 10 µg/ml of Polybrene. 48 hr after infection, the medium was removed and fresh DMEM was added to the cells containing 1 ug/mL puromycin for 7 days culture. PM 2.5 resistance gene screen and DNA sequencing Cells were exposed to PM 2.5 for 48 hours and then returned to 95% air, 5% CO2, and glucose-containing medium for recovery 6h. The surviving cells were collected. The genomic DNA from surviving cells was isolated using the Blood & Cell Culture DNA Midi Kit (Quiagen, Hilden, Germany) and stored at -20°C. PCR was performed in two steps and resulting amplicons from the second PCR were sequenced using a HiSeq 2500 (Illumina) as described by Dr. Feng Zhang [14]. The raw sequencing data were processed and analyzed using customized CRISPR-Cas9 library screen pipelines. Brie y, sequencing reads were rst de-multiplexed by using the barcode in the reverse primer, and processed by Cutadapt to remove sequences from beginning to sgRNA priming site primers. Trimmed reads were used to map sgRNA sequences to pooled GeCKO v2 libraries A and B. Read counts of sgRNA for each sample were quanti ed by MAGeCK v5.6.0. Count data were ltered and normalized, and essential sgRNA and genes were ranked by MAGeCK.
Cell culture, RNA interference Oligo RNA was purchased from GenePharma (Shanghai, China) siRNA and control siRNA transfection was performed with Lipofectamine 2000 (Invitrogen) according to the instructions of the manufacturer, 50 pmol siRNA/well was transfected with 1 μL of Lipofectamine 2000 in the 24 wells for 48h, after RNA transfection, the cells were exposed to PM 2.5 for 48 hours and then returned to 95% air, 5% CO 2 , and glucose-containing medium for different recovery times to induce cell apoptosis.

MTT assay
Cell viability was determined by 3, (4,5-diamethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay. Thus, 30 μL of the MTT solution was added to each well and incubated at 37 °C for 3 h. After 3 h, the wells were aspirated, and the plates were left to dry overnight. The next day, 50 μL of dimethyl sulfoxide (DMSO) was added to each well to solubilize the formazan crystals. The plates were then put on a shaker for 1 h and read spectrophotometrically at 570 nm in a plate reader. The data was then analyzed and is represented as percent cell viability.

Apoptosis assay
Cells were washed with PBS and resuspended in binding buffer before Annexin-V-FITC and propidium iodide (PI) double staining, according to the manufacturers′ instructions (BD Biosciences Pharmingen, USA). Brie y, L02 cells were collected and washed twice with PBS, followed by addition of 500 μL binding buffer, 5 μL Annexin V-FITC and 10 μL PI then were added to each group and cultures were incubated at 37℃ for 10 min in dark. The apoptosis was analyzed by ow cytometry and Cell QuestPro software (BD Biosciences).

Statistics
The data is presented as mean ± SD. The signi cance of differences between the groups was determined by paired Student's t-test and/or one-way ANOVA by the GraphPad Prism 6 software, with 0.05 as the level of signi cance.