Pantothenate kinase isoforms as collateral lethality targets in - - PowerPoint PPT Presentation

Pantothenate kinase isoforms as collateral lethality targets in Glioblastoma Multiforme

NAME OF STUDENT On-Topic Candidacy Exam Adviser: Dr. XXXXXX

PANK1 PTEN Coenzyme A Biosynthesis

Concomitant l loss o

• f p

passeng nger ge genes es o

• ccur with

genomic c de deletion of tum umor s r supp uppressor g r genes

TCGA Pan-Cancer Dataset

1

=0 =1 =2 >2

DNA Copy Number

Prev evalence o e of PT PTEN-PANK1 homozygous d deletions a across c cancer er t types es

PANK1 homozygous deletions are found in cancers with PTEN homozygous deletions

• Prostate Adenocarcinoma
• Glioblastoma
• Ovarian Cancer
• Melanoma
• Cervical Carcinoma

PTEN PANK1 PANK1

2

3

Pantothenate (vitamin B5) 4-phosphopantothenate 4-phosphopantothenoyl L-cysteine Pantothiene-4’-phosphate Dephospho-CoA Coenzyme A (CoASH, CoA)

PANK

OH N H O HO O OH O P OH O O N H O HO OH O

N N N NH2 N O O P O O O O P O O OH O N H N H O HS HO O HO P O O

Plasma membrane Intracellular Extracellular Pantothenate

Pantothenate kinases s ar are r rate l limit itin ing f fac actors i in coenzyme e A A production

Coenzyme e A A regulates es a a multit itude of

• f e

essentia ial c l cellu llula lar f functio ions

Food intake Vitamin B5 Coenzyme A Regulation Acyl group Carrier TCA Cycle Acetyl CoA CO2 + Energy Lipid Metabolism Fatty Acid

• xidation and

Biosynthesis Protein acetylation

4

PANK ANK i isoforms h have e distinct c cellular l localization and nd tissue s e spec ecific distribution

• PANK1
• PANK1⍺ (nuclear; Liver and kidneys)
• PANK1β (cytosolic; Liver)
• PANK2 (mitochondrial; neuronal tissue)
• PANK3 (cytosolic; all tissue types)
• PANK4 (catalytically inactive)

Alfonso-Pecchio A et al (2012) PLOS One, 7: 11, Figure 2 5

Cell e essen entiality o

• f P

PANK ANK i is ev eviden enced b by PANK ANK knockout phen enotype e in n lower er o

• rganisms
• E. coli, S. cerevisiae and D. melanogaster only have one PANK isoform
• PANK knockout is lethal in all unicellular organisms

6

Organismal e essentiality o

• f PANK

ANK is ev eviden enced b by PANK ANK isoform k knockout p phen enotype pes i in mi mice

• Individual PANK paralog knockout mice are viable
• Pank1 -/- : Hepatic fatty acid oxidation and gluconeogenesis impaired in

fasted state

• Pank2 -/-: Pantothenate Kinase Associated Neurodegeneration in human;

retinal degeneration and impaired spermatogenesis in mice

• Pank3 -/- : No known phenotype
• Double knockout mice are embryonic or post natal lethal
• Pank1-/-Pank3-/- and Pank2-/-Pank3-/- dko mice are embryonic lethal
• Pank1-/-Pank2-/- dko are late post-natal lethal

Conclusion: At least two PANK isoforms are required for organism viability

7

8 PANK2/3 Tumors with PANK1-/- Normal tissue Vitamin B5 / Pantothenate Co Enzyme A PANK2/3 Co Enzyme A

PANK1

PANK1 PANKK2/3 shPanK2/3 Co Enzyme A PANK1 PANK2/3 Cell death PANK1 Vitamin B5 / Pantothenate Vitamin B5 / Pantothenate shPanK2/3 Vitamin B5 / Pantothenate

Targeting the redundant isoforms of PANK in tumors with PANK1 homozygous deletion can selectively kill cancer cells.

CENT NTRAL HY RAL HYPOTHESIS IS

Speci cific c Aim m 1

To determine if PANK activity is cell essential and identify the paralog redundant with PANK1 in cancer cells

9

Hypothesis: Based on the cytosolic co-localization of PANK1 β and PANK3, I hypothesize that PANK3 is redundant with PANK1 β and therefore compensates for loss of PANK1.

To determine if PANK activity is cell essential and identify the paralog redundant with PANK1 in cancer cells

• Identification and generation of cancer cell lines with PANK1 homozygous deletion
• Test the effect of a pan-PANK inhibitor and antisense oligonucleotides against PANK

isoforms

• CRISPR KO of PANK isoforms alone or in the context of PANK1 homozygous deletion
• Generation of PANK1 isogenic rescued cell lines
• Constitutive and inducible knock down of PANK2 and PANK3 in PANK1 deleted and

intact cells

• Cell viability assays

Speci cific A c Aim 1

10

AVANA C ANA CRISPR d R data s set et i identifies P PANK ANK isoforms as non-essen ential ge genes es

11

PANK1 (53354) PANK3 (79646) PANK2 (80025) PANK4 (55229)

Selec ection o

• f in v

vitro model el w with PANK ANK1 homozygous d s deletion

Analyzed based on data from : Klijn, C et al. 2015, Nature Biotechnology 33: 306–312, Supplemental Data 1

Background 12 N=676 25 50 PANK1 TPI (Loading control)

Gener eration o

• f PANK

ANK1 CRISPR R KO cancer c cell l lines es

PANK1 TPI 50 27 Gastric Breast Sarcoma PANK1 TPI

50 37 25

PANK1 TPI

50 37 25

Conclusion: PANK1 is dispensable in cancer cells

13

Gener eration o

• f PANK

ANK1 isogen enic rescued ed c cell l lines

14 PANK1 TPI (Loading control)

Expected Western Blot result

PANK1 over expression

pCMV- PANK1

Lentiviral vectors psPAX2 , pMD2.G

+

G59 cells

Puromycin selection

Transfection

293T cells G59 PANK1 OE cells

Gener eration o

• f He

HeLa La PANK ANK2 and nd PANK ANK3 CRISPR K R KO line

15

Anti-PANK2 and anti-PANK3 antibody development in progress

Expected outcome: PANK isoforms are individually dispensable

16

PANK isoforms protein level in HeLa cells

PANK1 PANK2 TPI (Loading control) PANK3

PANK ANK1 homozygously deleted ed 537 37-MEL cel cells respond to to PANK ANK3 inhib ibit itio ion

17

SK-MEL5 PANK1 +/+ SK-MEL-28 PANK1 +/+

3.15 6.25 12.5 25

ASO (µM) 537-MEL PANK1 -/-

Day 4

Compound 7 7, a a Pan-PAN ANK i inhibitor has modest s t selecti tivity ty for P PANK3 K3

Sharma et al J Med Chem. 2015 Feb 12; 58(3): 1563–1568. PANK Isoforms IC50 PANK1 β 70 ± 1.1 nM PANK2 92 ± 2.0 nM PANK3 25 ± 1.8 nM

18

Compound 7 7 shows s selec ective e toxicity t to P PANK ANK1 1 deleted ed me melanoma t tumo mor-spheres es

0 uM 3.12 uM 6.25uM 12.5 uM 25uM Compound 7 PANK inhibitor 537-MEL PANK1 -/- SK-MEL5 PANK1 +/+ SK-MEL-28 PANK1 +/+ 50uM

Day 6

shRN RNA m A mediated k knockdown of PANK ANK2 2 and PANK ANK3 in n G5 G59 c cells ls

pLKo- shPANK 2/3

2nd generation lentiviral packaging system psPAX2 , pMD2.G + G59 and G59 PANK1 OE cells ( Puromycin selection) Identify shRNA clones with KD>60% To clone into inducible pTRIPZ vector 10 individual short hairpins

PANK1 PANK3

Expected Result

PANK1 PANK2 TPI TPI 20

Ex Examin inin ing t the e effect of

• f l

los

• ss of
• f PANK ac

activ ivit ity

• n
• n c

can ancer c cell v ll viab abilit ility

G59 shPANK2/3 pCMV PANK1 cells G59 shPANK2/3 cells

• Cell viability Assay with Crystal violet

staining/Cell-titre-glo assay

• Colony formation Assay
• Cell growth Analysis by IncuCyte
• FACS with annexin V-PE and 7-AAD

to detect apoptosis

21

Expect cted R Result: P : PANK3 inhibition will select ctively y imp mpede PANK1 nul null tum umor ce cell growth in n vitro

• + -

+ - + Dox

ishPANK3 ishPANK2 G59 PANK1 OE ishScr

• + -

+ - + Dox

ishPANK3 ishPANK2

G59 PANK1 -/-

ishScr ***

• Slow rate of cell growth in G59

shPANK3

• Increased annexin V-PE and and

reduced 7-AAD

22

AIM1- Pitfal alls ls/Alternative Approac

• ach
• Individual isoform may be essential and CRISPR KO may not be feasible.
• Recently released AVANA dependency map identifies PANK isoforms as non essential
• shRNA mediated knock down of PANK may lead to a complete cell death
• Modulation of dox concentration to measure the effect of acute loss of PANK activity.
• If shRNA mediated knock down is not informative in identification of the redundant isoform
• Generate inducible CRISPR KO of PANK isoform.
• If PANK activity is dispensable in cancer cells, it will contradict my hypothesis, but also suggest that cancer

cells can survive without Co-enzyme A.

• Unlikely, because all other downstream enzymes in the CoA biosynthesis pathway are essential based on the dependency score,

suggesting that this pathway is essential.

23

Hypothesis: I hypothesize that the loss of PANK activity will deplete CoA level from the cells which can impact critical metabolic and transcriptional profile of the cells.

Spe pecifi fic Ai Aim 2

24

To determine the biochemical consequences of PANK ablation

To determine the biochemical consequences of PANK ablation

• Profile small molecule metabolites by metabolomics
• Determine the effects of protein acetylation
• Identify key signaling pathways by RPPA
• Transcriptomics following PANK ablation in PANK1 homozygously deleted

and PANK1 intact cells.

Specific Ai Aim 2 2

Co Co-enzyme A A pl plays c critical r role i in n a mul ultitude o

• f

f bi biochemical r reactions

AcetylCoA is required for all protein acetylation reaction!

26

Met etabolomics to s to iden enti tify k key ey m met etabolites a es alter tered ed b by co-en enzyme A e A dep epleti tion

G59 shPANK3 cells G59 PANK1 OE shPANK3 ther PANK1 intact Glioma Cells: D423, D502, LN319

Extraction of polar metabolites with 80% methanol at -80 C

Mass spectrometry (LC/MS) ASARA MS core at BIDMC Analysis of metabolites

27

Expected Outcome:

• Massive disruption of Fatty Acid

Biosynthesis and oxidation pathway

• Decreased metabolites in the CoA

synthesis pathway

• Decrease in OCR and increase in

glycolytic intermediates, as well as ECAR

Status of metabolites in PANK1 deleted G59 cells

28

Assess essmen ent o t of global changes i es in protei ein a acety etylati tion a and g gene e e expressi ession i in G59 cells

Preparation of tumor/cell lysates in the RPPA buffer and samples submitted to the proteomics core at MD Anderson RNA extraction from the cells or tumor and submitted to the MD Anderson Core for an unbiased RNA seq analysis

29

Global Acetylation levels detected by western blot G59 shPANK3 cells G59 PANK1 OE shPANK3 cells Other PANK1 intact Glioma Cells: D423, LN319, D502

Expected O Outcome

A global decrease in protein acetylation

TPI

Ac-lys

Expected changes

• Changes in protein stability and localization
• Increase in p53/caspase mediated apoptosis
• Increased sensitivity to DNA damage
• Disruption in protein synthesis machinery in

ribosomes

• Decreased structural integrity of tubulins

30

RNA seq and RPPA

AIM2- Pitfal alls ls/Alternative Approac

• ach
• If there is no considerable effect in co-enzyme A levels or the subsequent effects on metabolites as well as

acetylated proteins, this could suggest two possibilities.

• a. CoA pool from mitochondria can move to the cytosol when PANK1/PANK3 activity is eliminated.

a. CoA is highly charged and therefore cannot cross mitochondrial membrane

• b. Alternatively, it is also possible that there are other sources of CoA that have not been studied and identified before.

Both RNA-seq and RPPA will be informative in addressing this possibility.

N N N NH2 N O O P O O O O P O O OH O N H N H O HS HO O HO P O O

31

Hypothesis: PANK3 inhibition will selectively kill intracranial PANK1 deleted glioma tumors in vivo but not its isogenic rescued tumors .

Specific Ai Aim 3

32

To validate PANK paralog inhibition as targeted therapy for PANK1-homozygously deleted tumors in vivo

• Generate xenografted tumors with PANK1 homozygously deleted and intact

cells in mice, with an inducible CRISPR or shRNA against PANK isoforms

• Test the efficacy of the PANK inhibitor in vivo
• Monitor tumor growth by IVIS and T2 MRI
• Identify target engagement markers with the metabolomics results and test the
• ligonucleotides for anti neoplastic activity

Spe pecifi fic Ai Aim 3 3

To validate PANK paralog inhibition as targeted therapy for PANK1-homozygously deleted tumors in vivo

G59 shPANK3 G59- pCMV PANK1 shPANK3 (PANK1 rescued) N=5 Intracranial injection Confirmation

• f tumor

establishment by T2 weighed MRI and IVIS and doxycycline administration Week 1 Weekly Tumor monitoring Tumor regression: Follow mice for 2 months Euthanize at morbidity Post Mortem Metabolomics and Histopathologic al Analysis N=5

Orthot

• top
• pic tumor imp

mplantation i in nude mi mice

34

Expected R Result lt: P PANK3 i 3 inhib ibit ition

• n l

leads t to PANK1 n 1 null ll t tumor

• r regression
• n.

Histopathological Analysis: Expected Outcome

• Reduced Ki67+ in G59 shPANK3

(+dox) cells

• Increased Cleaved Caspase 3
• Reduction in phenotype

characteristic of GBM tumors such as pseudopallisading cells, necrosis, hemorrhage G59 shPANK3 G59- pCMV PANK1 shPANK3 (PANK1 rescued)

+DOX +DOX

35

AIM3- Pitfalls/Alternative Approach

• pHAGE-CMV promoter might get methylated in mice, which can reduce PANK1 expression.
• I will alleviate this problem by using alternative plasmids such as pWPXL or pWPT which contains the EF-1 alpha promoter, and can

allow constitutive expression of the transgene.

• If there is a complete reduction of tumor growth, I will administer doxycycline water to mice at a lower

dosage so that loss of PANK2/3 will be acute.

• If no discernable differences are present between the PANK1 deleted and PANK1 reconstituted G59 cells due

to a compromise in knock down efficiency by the shRNA,

• I will intra-tumorally inject PANK2/3 ASOs alone or in combination with the pan PANK inhibitor in the mice and assess the effect.

Alternatively, we can use the inducible CRISPR against PANK2/3 in G59 cells and determine the effect of PANK ablation in tumors in vivo.

36

• Hypothesis: Targeting the redundant isoforms of PANK in tumors with

PANK1 homozygous deletion can selectively kill cancer cells.

• AIM1: To determine the cell essentiality of PANK isoforms and identify the

isoform redundant with PANK1 in cancer cells

• AIM2: To determine the biochemical consequences of PANK elimination
• AIM3: To validate PANK paralog inhibition as a targeted therapy in PANK1

homozygously deleted tumors in vivo.

Summa mmary

37