Combining Genomics To Identify the Pathways of Post-Transcriptional

Subscriber access provided by Eastern Michigan University | Bruce T. Halle Library

Article

Combining genomics to identify the pathways of posttranscriptional non-genotoxic signaling and energy homeostasis in livers of rats treated with the PXR agonist, PCN Hirohisa Nagahori, Kenji Nakamura, Kayo Sumida, Shingo Ito, and Sumio Ohtsuki J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.7b00364 • Publication Date (Web): 21 Aug 2017 Downloaded from http://pubs.acs.org on August 22, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

1

Combining genomics to identify the pathways of

2

post-transcriptional non-genotoxic signaling and

3

energy homeostasis in livers of rats treated with the

4

PXR agonist, PCN

5

Hirohisa Nagahori, *,†Kenji Nakamura,‡ Kayo Sumida,† Shingo Ito,‡,§ and Sumio Ohtsuki‡,§

6

†

7

3-chome, Konohana-ku, Osaka 554-8558, Japan

8

‡

9

Kumamoto University, 5-1 Oe-honmachi, Chuo-ku, Kumamoto 862-0973, Japan.

Environmental Health Science Laboratory, Sumitomo Chemical Co., Ltd., 1-98, Kasugadenaka

Department of Pharmaceutical Microbiology, Graduate School of Pharmaceutical Sciences,

10

§

11

5-1 Oe-honmachi, Chuo-ku, Kumamoto 862-0973, Japan.

Department of Pharmaceutical Microbiology, Faculty of Life Sciences, Kumamoto University,

12 13

KEYWORDS. PXR, gene expression, subcellular fractions, protein kinases, carcinogenesis,

14

energy metabolism, apoptosis

15

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 44

1

ABSTRACT. Transcriptomic, proteomic, phosphoproteomic, and metabolomic analyses were

2

combined to determine the role of pregnane X receptor (PXR) in non-genotoxic signaling and

3

energy homeostasis in liver after rats were repeatedly orally dosed with the PXR agonist,

4

pregnenolone carbonitrile (PCN), for 7 days. Analyses of mRNAs and proteins in the

5

supernatant, membrane, and cytosolic fractions of enlarged liver homogenates showed diverse

6

expression profiles. Gene set enrichment analysis showed that the synchronous increase in

7

mRNAs and proteins involved in chemical carcinogenesis and the response to drug was possibly

8

mediated by the PXR pathway and proteasome core complex assembly was possibly mediated by

9

the Nrf2 pathway. In addition, levels of proteins in the endoplasmic reticulum lumen and

10

involved in the acute-phase response showed specific increase with no change in mRNA level,

11

and composed of the mitochondrial inner membrane showed specific decrease. The analysis of

12

phosphorylated peptides of Poly(A) RNA binding proteins showed a decrease in phosphorylation

13

possibly by casein kinase 2, which may be related to the regulation of protein expression.

14

Proteins involved in insulin signaling pathways showed an increase in phosphorylation possibly

15

by protein kinase A and those involved in apoptosis showed a decrease. Metabolomic analysis

16

suggested the activation of the pentose phosphate and anaerobic glycolysis pathways, and the

17

increase of amino acid and fatty acid levels, as occurs in the Warburg effect. In conclusion, the

18

results of combined analyses suggest the PXR’s effects are due to transcriptional and post-

19

transcriptional regulation with alteration of non-genotoxic signaling pathways and energy

20

homeostasis.

21 22

INTRODUCTION


2

Page 3 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1


The elucidation of non-genotoxic carcinogenesis has furthered cancer research by identifying

2

the mechanism of tumor initiation, promotion, and progression, but the mechanism is

3

complicated. Non-genotoxic carcinogens act indirectly by disrupting cellular structures,

4

disrupting cell proliferation, or increasing the risk of genetic error (1-3), while genotoxic

5

carcinogens act by direct covalent binding to DNA. Currently, new non-genotoxic and genotoxic

6

carcinogens are detected in 2-year rodent carcinogenicity tests, but no simple in vitro test for

7

detecting non-genotoxic carcinogens has yet been developed. Therefore, a better understanding

8

of the mechanism of non-genotoxic carcinogenesis may reduce the number of tests and also help

9

medical research discover ways to prevent tissue-specific tumor formation.

10

Pregnenolone carbonitrile (PCN) is a non-genotoxic carcinogen that acts through the pregnane

11

X receptor (PXR) to cause liver hyperplasia and hypertrophy with cytoplasmic vacuolization and

12

a decrease of periodic acid-Schiff-positive material (4). First identified in 1998, PXR is a nuclear

13

receptor and primarily expressed in the liver and intestine (5). The nuclear receptors, PXR,

14

constitutive androstane receptor (CAR), aryl hydrocarbon receptor (AhR), and peroxisome

15

proliferator-activated receptor α (PPARα) cause liver enlargement and carcinogenesis, and can

16

be activated in a ligand-dependent fashion by numerous structurally diverse xenobiotics,

17

including steroids, pesticides, and drugs (6). So elucidating the role of nuclear receptors in non-

18

genotoxic carcinogenesis should improve understanding of the mechanism of non-genotoxic

19

carcinogenesis.

20

The mechanism of non-genotoxic carcinogenesis mediated by nuclear receptors is established

21

by identifying a drug’s mode of action (MOA) as key and associative events (7). The key events

22

are 1, activation of a nuclear receptor; 2, altered gene expression; 3, increased cell proliferation;

23

4, clonal expansion leading to altered foci; 5, formation of liver adenomas/carcinomas, with


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 44

1

associative events including cytochrome P450 (CYP) induction, liver hypertrophy, and inhibition

2

of apoptosis. PXR plays a central role in transcriptional regulation of drug metabolizing enzymes

3

(8). In addition, recent research revealed several new roles for PXR in inflammation, bone

4

homeostasis, vitamin D metabolism, lipid homeostasis, energy homeostasis, anti-apoptosis, and

5

cancer (9). However, the involvement of these new roles in carcinogenesis remains to be fully

6

elucidated, so establishing the molecular basis for the connection between each event and the

7

stage of tumor formation is important.

8 9

In addition to liver weight change, clinical pathology, and histological pathology analyses, mRNA microarray analysis is one of the tools used to identify MOA mechanisms (6). A previous

10

investigation of PCN-treated rats identified changes in intercellular metabolism, transport of

11

essential small molecules, cell cycle kinetics, and redox balance after a single oral administration

12

(10). The PXR regulates gene networks even more extensively than it activates drug metabolism

13

genes. Therefore, we combined transcriptomic analysis with mass spectrometry (MS)-based

14

proteomics, and metabolomics to monitor the components of biological processes, including

15

mRNAs, proteins, nucleotides, and small molecules (11, 12).

16

The purpose of the present study was to identify the non-genotoxic signaling pathways

17

involved in hepatocarcinogenesis in rats after repeated oral administration of PCN for 7 days

18

using a combination of transcriptomic, proteomic, phosphoproteomic, and metabolomic analyses.

19 20

MATERIALS AND METHODS

21

Animals


4

Page 5 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

Four-week-old Wistar rats (4 groups of 3 males each) were purchased from Japan Laboratory

2

Animals, Inc. (Tokyo, Japan). The animals were kept in aluminum cages for an acclimatization

3

period of 7 days, received either 0 (vehicle control of corn oil), 25, 50, or 100 mg/kg/day of PCN

4

(Sigma, Lot No. 031M1245V, purity of >97%) by gavage for 7 days, and were given free access

5

to pelleted diet (CRF-1, Oriental Yeast Co., Ltd., Tokyo, Japan) and filtered water. All animals

6

were observed daily for mortality, behavioral changes, and signs of toxicity until sacrifice. The

7

animal protocol was approved by the Environmental Health Science Laboratory Institutional

8

Animal Care and Use Committee of Sumitomo Chemical Co., Ltd. and animals were maintained

9

in accordance with the Guide for Animal Care and Use of Sumitomo Chemical Co., Ltd.

10 11

Sample Collection and Preparation of Liver Fractions

12

After a 7-day dosing period, the rats were sacrificed by exsanguination under anesthesia with

13

isoflurane, and the liver and kidneys were dissected out and weighed to determine the extent

14

their enlargement. The supernatant, membrane, and cytosolic fractions of the liver homogenate

15

were prepared as follows. The livers of the control and 100 mg/kg/day groups were homogenized

16

individually in a 3-fold volume of ice-cold 50 mM Tris-HCl buffer (pH 7.4) with 154 mM

17

potassium chloride and PhosSTOP (Roche Diagnostics, Mannheim, Germany). The homogenates

18

were centrifuged at 10,000 g, and the supernatants were collected as supernatant fractions. A part

19

of each supernatant fraction was ultra-centrifuged at 105,000 × g to obtain the cytosolic fraction.

20

The membrane fractions were prepared from the pellet after washing in 100 mM potassium

21

phosphate buffer (pH 7.4) with PhosSTOP. Each fraction was diluted with the buffer to 1 mg/mL


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

of protein after determining protein concentration using the DC protein assay (Bio-Rad

2

Laboratories, Inc., Hercules, CA) with albumin as a standard.

Page 6 of 44

3 4

Sample Preparation for Transcriptomic Analysis

5

The total RNA in liver was prepared as previously described (13). Briefly, total RNA was

6

extracted from liver tissue using the mirVana™ miRNA Isolation Kit (Ambion, Austin, TX).

7

Total RNA was quantified by UV analysis at 260 nm and 280 nm using an Ultraviolet

8

spectrometer (NanoDrop 2000, Thermo Fisher Scientific, Wilmington, DE). The total RNA

9

solution was stored at –80ºC.

10 11

Transcriptomic Analysis

12

The transcriptomic analysis was conducted as described previously (13). The global gene

13

expression profiles were determined in livers from three animals in the control groups and

14

highest-dose group (PCN [100 mg/kg/day]). Gene expression analysis was conducted on each

15

total RNA sample using a SurePrint G3 Rat GE 8×60K Microarray together with a Low Input

16

Quick Amp Labeling Kit, a Gene Expression Hybridization Kit, and a Gene Expression Wash

17

Pack (Agilent Technologies, Wilmington, DE). The procedure was basically conducted

18

following the manufacturer's protocol (version 6.5), and Feature Extraction (version 10.7.3.1)

19

was used to generate tab-delimited files containing data on the relative level of expression of a

20

transcript (gProcessedSignal = Signal) and whether the transcript is reliably detected or not

21

(gIsWellAboveBG = Detection Call; present [1], absent [0]) in the image. The gene expression


6

Page 7 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1

microarray data can be downloaded from the National Center for Biotechnology Information

2

Gene Expression Omnibus (accession no. GSE95475).

3 4

Method of Peptide Digestion for Proteomic Analysis

5

The peptides in liver were extracted and derivatized according to a previously published

6

method (14). The supernatant, membrane, and cytosolic fractions of homogenates were heated at

7

95°C for 5 min after addition of an equal volume of the denaturing solution (final concentration

8

of 12 mM sodium deoxycholate, 12 mM sodium lauryl sulfate, 100 mM Tris-HCl, pH 9.0),

9

followed by sonication in ice for 20 min.

10

Then, the samples were reduced with 10 mM dithiothreitol at room temperature for 1 h,

11

alkylated with 55 mM iodoacetamide in the dark at room temperature for 1 h, diluted 5–fold

12

(v/v) with 50 mM ammonium bicarbonate (16 µg/µL), digested with lysyl endopeptidase at an

13

enzyme/substrate ratio of 1:100 at room temperature for 3 h, digested with sequence-grade

14

modified trypsin at an enzyme/substrate ratio of 1:100 at 37°C for 16 h, vortexed after adding an

15

equal volume of ethyl acetate and 1/200 volume of trifluoracetic acid (TFA, final 0.5%),

16

centrifuged at 15,000 × g for 2 min at room temperature to remove the ethyl acetate phase,

17

allowed to dry to evaporate the solvent completely, dissolved in 5% acetonitrile with 0.1% TFA,

18

and demineralized using a combination of GC (Graphite carbon) and styrene divinylbenzene

19

(SDB) containing micropipette GL-Tips (GL Sciences, Tokyo, Japan). After washing of

20

micropipettes by centrifugation with 50 µL of 80% acetonitrile solution containing 0.1% TFA

21

and by 50 µL of 5% acetonitrile solution with 0.1% TFA, the sample (50 µg protein) in solution

22

was added to the SDB, centrifuged, and washed with 50 µL of 5% acetonitrile solution with


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 44

1

0.1% TFA. The peptides were eluted from GC + SDB twice with 50 µL of 80% acetonitrile

2

solution containing 0.1% TFA. The collected 100 µL of peptides solution was dried and the dried

3

peptides were dissolved and diluted in 0.1% TFA up to 0.2 µg/µL for LC-MS analysis.

4 5

Preparation of Phosphorylated Peptides for Phosphoproteomic Analysis

6

The phosphorylated proteins were prepared from the supernatant, membrane, and cytosolic

7

fractions of liver homogenate by highly selective phosphopeptide enrichment using aliphatic

8

hydroxy acid modified metal oxide chromatography (HAMMOC) (15). Briefly, the trypsin-

9

treated protein (150 µg protein in 75 µL), as described in the previous section, was mixed with

10

150 µL of 60% acetonitrile solution with 300 mg/mL lactic acid and 0.3% TFA (buffer A),

11

loaded onto a titanium dioxide column (Titansphere Phos-TiO Kit, 3 mg of titanium dioxide

12

beads, 10 µm, GL Sciences, Torrance, CA) and washed into the column with 20 µL of buffer A

13

and 20 µL of 80% acetonitrile solution with 0.4% TFA (buffer B). After centrifugation of the

14

column, the eluent was added to the same column, and the column was centrifuged again and

15

washed 3 times with 20 µL of buffer A and 50 µL of buffer B. The phosphorylated peptide was

16

eluted from the column by addition of 50 µL of 5% ammonia solution followed by 50 µL of 5%

17

pyrrolidine solution. The eluted phosphopeptides were diluted twofold in 20% trichloroacetic

18

acid and subjected to the demineralization procedure described in the previous section. The

19

collected solution was dried and dissolved in 17 µL of 0.1% TFA solution.

20 21

Proteomic and Phosphoproteomic Analyses by SWATH using LC-MS/MS


8

Page 9 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1


The proteomic analysis was conducted as previously reported (14) using nano-LC-ESI-MS/MS.

2

Samples were loaded onto a PepMap 100 precolumn (100-µm internal diameter × 20-mm length,

3

Thermo Fisher Scientific) at 5 µl/min and separated on a PepMap RSLC column (75-µm internal

4

diameter × 250-mm length, Thermo Fisher Scientific) packed with 2-µm C18 beads with 100-Å

5

pores. The flow rate was 300 nl/min and the gradient of solvent B (100% ACN and 0.1% formic

6

acid) during the 107-minute run was as follows: 2% B at 0–3 min, to 25% B at 63 min, to 50% B

7

at 78 min, 98% B at 80–85 min. Sample analysis was performed using a Triple TOF 5600 mass

8

spectrometer controlled by Analyst TF1.6 software (SCIEX, Ontario, Canada). For Information

9

Dependent Acquisition (IDA), precursor ions were scanned from 300 to 1008, and product ions

10

were scanned from 100 to 1600 with an accumulation time of 50 msec for a maximum of 20

11

precursor ions/cycle. For SWATH-MS acquisition, the sequential windowed acquisition of all

12

theoretical fragment ion (SWATH) window for precursor ions was set at m/z 13 Da, including 1

13

Da of overlap, and the mass range was from 100 to 1600 using the same collision energy settings

14

as for IDA. Cycle time was 3.05 sec. IDA data were analyzed using ProteinPilot Version 4.5

15

(SCIEX), and the UniProt rat proteome database was searched. The confidence in protein

16

identification for the dataset was evaluated versus the false discovery rate (FDR) determined in a

17

concomitant search of the UniProt rat proteome database for the reverse sequences. The FDRs

18

were all lower than 1%. The spectral library for peak identification in SWATH-MS data was

19

generated from the IDA data of same samples. The results files were imported into the PeakView

20

SWATH Processing Micro App (SCIEX, Concord, ON, Canada). The conditions of SWATH

21

data analysis were as follows: all peptides per protein, 5 transitions per peptide, peptide

22

confidence threshold of 95%, false discovery rate threshold of 1% with exclusion of shared

23

peptides, 2.0 min extracted-ion chromatogram (XIC) window, and 0.040 Da XIC width. Raw


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

data files of LC-MS/MS analysis have been deposited in jPOST (http://jpostdb.org, jPOST ID:

2

JPST000258/PXD006305).

Page 10 of 44

3 4

Preparation of Extracts for Metabolomic Analysis

5

The metabolites in liver were extracted and derivatized according to a previously published

6

procedure (13). Briefly, metabolomic analysis was conducted on liver samples from four animals

7

in each of the control and PCN groups.

8 9

Liver samples (ca. 10 mg) were weighed, homogenized twice in 50 volumes of methanol/water/chloroform (2.5:1:1, v/v/v) using a single-bead TissueLyser mill (Qiagen, Haan,

10

Germany) set at 25 Hz for 4 min, and centrifuged at 20,400 × g for 5 min at 4°C. After their

11

transfer to clean tubes, the resultant supernatants (300 µl) were mixed with 4 µl of 2.5-mM 2-

12

isopropylmalic acid (Sigma-Aldrich, MO) dissolved in pyridine (used as an internal standard)

13

and 133 µl of water before being centrifuged at 20,400 × g for 5 min at 4°C. Exactly 200 µl of

14

the supernatants were transferred to clean tubes and dried under reduced pressure before

15

oximation and silylation. For oximation, 30 µl of 20-mg/ml methoxyamine hydrochloride

16

(Sigma-Aldrich) dissolved in pyridine was added to each dried sample. The mixture was

17

incubated for 90 min at 30°C, silylated by adding 30 µl of N-methyl-N-trimethylsilyl-

18

trifluoroacetamide (GL Sciences), incubated for 30 min at 40°C, and analyzed by gas

19

chromatography (GC)/MS.

20 21

Targeted Metabolomic Analysis by GC-MS/MS


10

Page 11 of 44

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1


The metabolomic analysis was conducted as previously reported (13). All analyses were

2

carried out using a GCMS-TQ8030 system (Shimadzu, Kyoto, Japan) equipped with a

3

FactorFour capillary column DB-5 (length, 30 m; internal diameter, 0.25 mm; film thickness,

4

1.00 µm) (Agilent Technologies, Inc.) under the following conditions: injection temperature

5

280°C, and helium gas flow rate through the column 39 cm/sec, with column oven temperature

6

maintained at 100°C for 4 min, rising 10°C /min to 320°C where it remained for 11 min. A

7

sample of 1.0 µl was injected using a split mode (split ratio 1:50). Ions were generated at the

8

most suitable electron impact energy for each metabolite, and the ion spectra were recorded over

9

the mass range m/z 70–1000 Da.

10

The chromatogram acquisition, detection of mass spectral peaks, and their waveform

11

processing were performed using GC-MS solution software Version 4.11 (Shimadzu, Kyoto,

12

Japan). The identification of metabolites was performed employing an in-house mass spectral

13

library constructed from the results of standard analysis, GC/MS Metabolite Mass Spectral

14

Database (Release 1.01), and an NIST Mass Spectral Library (NIST 11) search. To perform

15

semi-quantitative assessment, the peak intensity of each quantified ion was calculated using the

16

peak intensity of 2-isopropylmalic acid as an internal standard. The ratio of each individual

17

treatment group value to the control group average was calculated.

18 19

Data analysis

20

All the presented data were mean ± SD for N=3 animals, and a two-tailed Welch’s t-test was

21

applied. For all gene products (mRNAs in the liver and proteins in the supernatant, membrane, or

22

cytosolic fractions of liver homogenates) compared between the highest dose and control groups


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 44

1

excluding those with low signal (

Combining Genomics To Identify the Pathways of Post-Transcriptional

Recommend Documents