HIV Latency Approaches

Daria Hazuda

March 2014

Why ARV treatment doesn’t cure HIV:

2

Activated CD4+ T-cell

HIV

Activated CD4+ T-cell

Apoptosis

HIV

HIV HIV

HIV

Fate of Latently Infected Cells

HIV

Activated CD4+ T-cell

HIV

Resting Memory CD4+ T-cell

HIV

Myeloid cell

Minority of cells become latently

infected

Majority of cells die or are eliminate

Establishment of Latency

HIV PD1

HIV

PD1 and other inhibitory molecules contribute to

T-cell dysfunction

Immune dysfunction reduces the clearance of infected cells

Exhausted T-cell

Dysfunctional B-cell

Immune dysfunction Prevents Clearance of latently infected cells

HIV Particle

• A latent reservoir of long-lived T-cells with integrated HIV DNA is seeded early during acute infection

• Viral persistence results in immune dysfunction

“Flush” and “Kill” Strategy for HIV Eradication

“

HIV genome Memory

CD4+ T cell

DNA

HIV particles

HIV proteins HIV RNA

Dying infected cell Uninfected cell

Antiretroviral therapy

FLUSH Kill

Title Box HIV DNA transcription prevented by restricted

access to needed host enzymes, chromatin remodeling and transcriptional interference

HDAC inhibitors (eg TSA, SAHA) Methyltransferase inhibitors NF-kB activators (Prostratin, PMA) Akt/HEXIM-1 modulators, (HMBA)

nuc-0 nuc-1

HIV TSS

P-TEFb

Latency at the HIV Promoter: Epigenetic Silencing and Transcription

HDAC EZH2

CDK9 HEXIM1

CyclinT1

7S RNA

+

NFAT NFκB

Chromatin Structure

HAT

IκB

Transcription Regulation

Phosphorylation

+

Reactivation Step 1: Relaxation of chromatin

Reactivation Step 2: Stimulate transcription

Oncology HDACIs Being Explored for HIV Latency

Vorinostat (SAHA – Merck): Pan HDACI, oral, AMES positive Latency clinical POC (Margolis, Lewin)

Increase in HIV RNA in resting T-cells Increase in viremia (some pts), no effect vDNA

Ames(+) = potential safety liabilities?

Panobinostat (Novartis): Pan-HDACI; potent/reversible, oral Latency clinical POC (Rasmussen; Lichterfeld CROI ‘14)

Increase in HIV RNA in resting cells Increase in viremia and decrease in vDNA (some pts)

Romidepsin (Gilead): Pan HDACI, potent/irreversible, IV, DDIs Clinical POC in progress

HN

ONH

OOH

L-001079038

Zolinza (Vorinostat)

NH

OOH

HN

HN MeL-001306168

Panobinostat (LBH-589)Phase III - Novartis

HN

HN

HN

O

NH

Me

O

O

OMe

Me

O

O Me

Me

SS

L-001302455Romidepsin

(Istodax)

Can HDACIs be optimized for HIV latency?

Potential Approaches to Improve Efficacy and Therapeutic Window of HDACIs

►Optimize HDAC activity for HIV latency – Increase potency and selectivity profile – PK, understand dose/response relationship

• Differs between reversible and irreversible binders?

► Identify other mechanisms to be used in combination – Enhance overall effect and/or – Reduce HDACI dose

Criteria SAHA Next gen

HDACi Isozyme

Selectivity HDAC3 required 1, 2, 3, 6, 8 1, 2, 3

HIV Latency in Jurkat T-cell

Model EC50 (nM)

Is the cellular activity

≤ SAHA? 1000nM 200 nM

PK profile To be determined in PK/PD expts

very high clearance

Can be tuned

Next Generation HDAC inhibitors with Improved Selectivity/PK Profile for HIV Latency

8

Next gen HDACIs: Historical chemical matter Structure-guided design + modeling Screening

Potential Approaches to Improve Efficacy and Therapeutic Window of HDACIs

►Optimize HDAC activity for HIV latency – Increase potency and selectivity profile – PK, understand dose/response relationship

• Differs between reversible and irreversible binders?

► Identify other mechanisms to be used in combination – Enhance overall effect and/or – Reduce HDACI dose

nuc-0 nuc-1

HIV TSS

P-TEFb

Latency at the HIV Promoter: Epigenetic Silencing and Transcription

HDAC EZH2

CDK9 HEXIM1

CyclinT1

7S RNA

+

NFAT NFκB

Chromatin Structure

HAT

IκB

Transcription Regulation

Phosphorylation

+

Reactivation Step 1: Relaxation of chromatin

Reactivation Step 2: Stimulate transcription

Novel Ultra High Throughput Screen to Identify New Mechanisms and Combinations

“Other"

HDAC Inhibitors

Farnesyltransferase Inhibitors

66.5%

16.1%

17.4%

Latent Jurkat T-cell model (Jon Karn) Screened 2.9 million compounds in the presence of 250 nM SAHA All hits titrated plus and minus 250 nM SAHA Identified HDACIs plus 2400 unique hits

FTi’s with Differential Binding Modes and Structures

MRK-16 MRK-17

From: Bell, I. J. Med. Chem. 2004, 47, 1869-1878.

X-ray Crystallographic images of different farnesyl-transferase inhibitor binding modalities. (A) Zinc Binding, (B) Exit Groove Blocker

A B

Correlation Between Farnesyl Transferase Inhibition and HIV-Latency Activation

13

• Strong Positive correlation between FTi potency (IC50 enzyme assay) and HIV latency activation in the presence of 250nM SAHA (EC50 in Jurkat T-cell model system)

• Knock-down of FTi-beta subunit leads to activation of HIV LTR in a Jurkat model system

ADD si RNA KD data

Scrambled siRNA

shRNA knockdown of FTβ Subunit

Untreated

3 Days 10 Days

Farn

esyl

-Tra

nsfe

rase

E

nzym

atic

Ass

ay IC

50 (µ

M)

Jurkat T-cell Induction (EC50, µM)

FTIs Increase Emax and Lower SAHA EC50

1

SAHA FTi #1

FTIs Enhance HDACI Effectiveness in Primary T-cells by Stimulating Transcription (J.Karn)

eGFP

Ac

tive

PTEF

-b

0 101 102 103 104 105

R670-A

010

110

210

310

410

5B

530-

A

58.75% 41.25%

0.00% 0.00%

R2: 58.75%57.24% Unstimulated

35.86% Stimulated

NEF

0 101 102 103 104 105

YG610-A

010

110

210

310

410

5R

780-

A0.90% 45.52%

50.90% 2.69%

R6: 0.90%

50.37% Unstimulated

47.21% Stimulated

Cyclin D3 0 101 102 103 104 105

YG610-A

010

110

210

310

410

5R

780-

A

0.72% 0.90%

95.87% 2.51%

R6: 0.72%

95.25% Unstimulated

0.54% Stimulated

Cyclin D3

0 101 102 103 104 105

R670-A

010

110

210

310

410

5B

530-

A

96.39% 3.61%

0.00% 0.00%

R2: 96.39%92.99% Unstimulated

2.62% Stimulated

NEF

0 101 102 103 104 105

YG610-A

010

110

210

310

410

5R

780-

A

0.34% 4.68%

89.75% 5.24%

R6: 0.34%

88.49% Unstimulated

5.46% Stimulated

Cyclin D3

0 101 102 103 104 105

R670-A

010

110

210

310

410

5B

530-

A74.00% 26.00%

0.00% 0.00%

R2: 74.00%71.88% Unstimulated

22.20% Stimulated

NEF

No Stimulation 500nM SAHA 10µM FTi 10µM FTi + 500nM SAHA

0 101 102 103 104 105

YG610-A

010

110

210

310

410

5R

780-

A

1.00% 49.44%

46.70% 2.86%

R6: 1.00%

46.14% Unstimulated

51.62% Stimulated

Cyclin D3

0 101 102 103 104 105

R670-A

010

110

210

310

410

5B

530-

A

23.54% 76.46%

0.00% 0.00%

R2: 23.54%22.37% Unstimulated

67.32% Stimulated

NEF

Validation of FTIs as Inducers of Latent HIV Expression: Summary of data to date

► Farnesyl Transferase (FT) Inhibitors from different structural classes induce latent HIV expression

► Effect on latent HIV expression is reproduced by siRNA knockdown

► Positive correlation between FT enzymatic activity and HIV latency activation in a Jurkat T-cell model system

► FTIs synergize with SAHA in a primary T-cell model and in latently infected memory T-cells isolated from HIV infected patients

► FTIs may increase active pTEFb levels in resting cells

16

Per

cent

Act

ivat

ion

(Nor

mal

ized

) to

SA

HA

)

MRK 23 (HDACi) Prostratin TNFα No Enhancer

Log Compound (M)

FTIs Synergize with other HIV Latency Activation Mechanisms

Identifying Compounds which Synergize with other Established Mechanisms

Jurkat HIV Latency T-cell Model (Luc

reporter)

Cytotoxicity of Compounds +/- EC20 of known HIV Activators

N=3

~2000 HIV Latency uHTS hits (non-HDACi) Complete Dose Response,

+/- EC20 of SAHA, HDACi (1,2,3), TNFα, JQ1, HMBA, Prostratin

Complete Dose Response of uHTS hits +/- EC20 of known HIV

Activators

Re-test of compounds with decreased EC50 and/or increased Emax in the presence of known HIV activators

N=3

Synergy Profile of uHTS Hits of Unknown Mechanism

EC50 Synergy Emax Synergy EC50 and Emax Synergy

HMBA 0 0 0

JQ1 12% 72% 7%

PKC 15% 81% 11%

HDACi (1,2,3) 14% 78% 10%

TNFa 13% 80% 9%

19

Some Key Questions in HIV Latency Drug Discovery ►Can we improve the efficacy/safety of

latency activating agents ►Can we increase the extent and spectrum

(breadth) of activation through combinations ►Will the same agents/combinations work in

all cells ►If we increase HIV expression will it lead to

cell death or will another modality be required to eliminate these cells

►How much expression is “enough”

Jurkat 2C4 Cells (Luciferase) Compounds Synergistic

Jurkat Population Cells (eGFP) No Synergy Detected

Differential Activity of FTIs in Two Different Jurkat T-cell Models

Some Key Questions in HIV Latency Drug Discovery ►Can we improve the efficacy/safety of

latency activating agents ►Can we increase the extent and spectrum

(breadth) of activation through combinations ►Will the same agents/combinations work in

all cells ►If we increase HIV expression will it lead to

cell death or will another modality be required to eliminate these cells

►How much expression is “enough”

gp120 is Expressed on Latently Infected Jurkats After Stimulation w/ TNF or SAHA

TNF alpha

SAHA

% p

ositi

ve GFP +

gp120 +

GFP + gp120 +

% p

ositi

ve

Untreated cells: 2G12

TNF treated: 2G12

ng/ml

nM *KARN cells were treated with TNF or SAHA for 24 hrs. GFP and gp120 expression were determined at by imaging and flow cytometry.

23

“

Approaches for the “Kill”

DNA

HDACi with Optimal Selectivity/PK

+ Synergistic Agent

FLUSH

• Ab-mediated Killing • IMR/Resurrection of

immune response +/- • Therapeutic Vaccine

KILL

HIV Antibody Conjugates can Eliminate HIV-infected Cells in vivo Platform mAb MOA POC

(Model) Pros Cons

Bispecific mAb (BsAb)

αgp120/αCD3

Redirected T-Cell killing (CD3)

Yes (Rhesus Macaque)

Highly specific Does not require high antigen expression/ internalization Clinical precedence for MOA

Uncertainty around cis/trans effect with CD3 expression Clinically significant immune activation?

Antibody Drug Conjugate (ADC)

αgp120 Intracellular cytotoxin delivery (Doxorubicin, PE, ricin)

Yes (BLT Mouse)

Highly specific Intracellular toxicity Clinical precedence

Requires internalization May require high antigen expression

Antibody Radioisotope Conjugate

αgp41 Localized radiation induced damage (Bi213 )

Yes (BLT Mouse)

Highly specific Does not require high antigen expression or internalization Clinical precedence

Manufacturing and regulatory hurdles Short t1/2 of isotope and commercial impact

25

Why ARV treatment doesn’t cure HIV:

26

Activated CD4+ T-cell

HIV

Activated CD4+ T-cell

Apoptosis

HIV

HIV HIV

HIV

Fate of Latently Infected Cells

HIV

Activated CD4+ T-cell

HIV

Resting Memory CD4+ T-cell

HIV

Myeloid cell

Minority of cells become latently

infected

Majority of cells die or are eliminate

Establishment of Latency

HIV PD1

HIV

PD1 and other inhibitory molecules contribute to

T-cell dysfunction

Immune dysfunction reduces the clearance of infected cells

Exhausted T-cell

Dysfunctional B-cell

Immune dysfunction Prevents Clearance of latently infected cells

HIV Particle

• A latent reservoir of long-lived T-cells with integrated HIV DNA is seeded early during acute infection

• Viral persistence results in immune dysfunction

The Latent Reservoir Established Early in Infection….may be Maintained by

Multiple Mechanisms

HIV/CMV persistence

Immune activation/ Inflammation

Immune dysfunction

Homeostatic proliferation

Addressing Latency Through Multiple Interventions…

HIV/CMV persistence

Immune activation/ Inflammation

Immune dysfunction

Homeostatic proliferation

Therapy intensification

Anti- inflammatory?

Immune reconstitution?

Shock and Kill

Acknowledgements

The Merck HIV

Team

Many collaborators..