Characterization of the HIV-1 Transcription Profile after Romidepsin Therapy in vivoSara Moron-Lopez1,2, Peggy Kim2, Ole S. Søgaard3, Martin Tolstrup3, Joseph K. Wong1,2, Steven A. Yukl1,2

1. University of California San Francisco, San Francisco, CA, USA; 2. San Francisco VA Medical Center, San Francisco, CA, USA; 3. Aarhus University Hospital, Aarhus, Denmark

Introduction

MethodsStudy design and samples

Results

Conclusions••

•

Acknowledgments

References

5’LTR GAG

POL

ENV

TAT

REV 3’LTR

VPR VPU

VIF NEF

RU3 U5TAR

“TAR”

“Read-through”

“longLTR”

“Tat-Rev”

“PolyA”

Assaytargets

HIV-1genome

Blocks to initiation- Integration into transcriptionally- silent regions- Epigenetic modifications- Lack of host initiation factors- Lack of viral factors (e.g. Tat)- Transcriptional interference

Blocks to elongation- Lack of host elongation factors- Presence of inhibitory host fac-tors- Nucleosomes- Lack of viral factors (e.g. Tat)

Blocks to completion- Lack of host distal elongationfactors- Lack of 3’ end processing (e.g.polyadenylation)

Blocks to post-transcriptional modifications- Lack of or aberrant splicing- Lack of nuclear export- RNA interference

AAAAAAAAHIV-1transcripts AAAAAAAA

AAAAAAAA

ReadthroughShort

Singly-splicedMultiply-spliced

“Read-through”- Suggests transcriptional interfe-rence

“longLTR”- Proximal elongation

“PolyA”- Completion of transcription- Surrogate for HIV protein

“Tat-Rev”- Completion of splicing- Potential to overcome blocks toinitiation, elongation, and export- Surrogate for productive infection“TAR”

- Found in all HIV transcripts- Initiation of transcription- >2-fold excess over longLTRsuggests inhibition of elongation

Blocks to HIV transcription and HIV transcription profile

Figure 2. Blocks to HIV transcription and assays to investigate the HIV transcription profile. Diagram of the blocks that regulate HIV transcription and assays used to characterize the HIV transcription profile.

Time (days)-21 0 7 14 21 77 84 91

105

11211

916

117

518

229

5

Vacc-4x + rhuGM-CSF Romidepsin ATI (N=16)

Samples (cryopreserved PBMC)N=9

N=17

RNA/DNA TriReagent extraction

DNARNAddPCR

RTpolyAtailing

RT

TAR Read-through LongLTR PolyA Tat-Rev LongLTR TERT

Figure 1. Study design and sample processing. Cryopreserved peripheral blood mononuclear cells (PBMCs) were available from nine participants at baseline (day -21) and 4h after the second (day 112+4h) and the third (day 119+4h) romidepsin infusions. Eight of the nine participants underwent an ATI. Peripheral blood mononuclear cells were pelleted and nucleic acids were extracted using TRI Reagent. Cell-associated HIV transcripts (read-through, total initiated [TAR], 5’ elongated [R-U5-pre-Gag; “longLTR”], polyadenylated [“polyA”], and multiply-spliced [Tat-Rev]), total HIV DNA (longLTR), and the cellular gene TERT were quantified in duplicate using ddPCR [5].

Romidepsin increases read-through, total, elongated, and polyadenylated but not multiply-spliced transcripts

Contact:Steven A. Yukl

steven.yukl@ucsf.edu 4150 Clement Street

San Francisco, CA, 94121+1(415)221-4810

-21 112+4h 119+4h

10 0

10 1

10 2

10 3

10 4

Read-through

RNA

copi

es/1

06 PBM

C

p=0.0195

Time (days)

10 0

10 1

10 2

10 3

10 4

10 5

Total (TAR)p=0.0078

-21 112+4h 119+4hTime (days)

10 1

10 2

10 3

10 4

10 5

Elongated (longLTR)

p=0.0039

p=0.0039

-21 112+4h 119+4hTime (days)

10 0

10 1

10 2

10 3

10 4

Polyadenylated (PolyA)

p=0.0273

-21 112+4h 119+4hTime (days)

10 -1

10 0

10 1

10 2

Multiply-spliced (Tat-Rev)

-21 112+4h 119+4hTime (days)

Romidepsin increases elongation but not completion or multiple-splicing

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Transcriptional interference

Ratio

Read

-thr

ough

/TAR

-21 112+4h 119+4hTime (days)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Elongation

Ratio

long

LTR

/TAR

p=0.0156p=0.0156

-21 112+4h 119+4hTime (days)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Multiple splicing

Ratio

Tat-

Rev/

poly

A

-21 112+4h 119+4hTime (days)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Completion

Ratio

poly

A/lo

ngLT

R

p=0.0156

-21 112+4h 119+4hTime (days)

HIV DNA and elongated transcripts predict time to viral rebound

A HIV DNA correlations

Time

Time Tim

eto

VL>50cp/ml

toVL>1,000cp/m

l

tosu

ppression

*

**

**

day -21

day 112+4h

day 119+4h

-1.0

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1.0

21

23

24

25

26

27

28

29

34

0 10 20 30 40 50

0

500

1000

1500day -21

Days to VL>50cp/ml

long

LTR

HIV

DN

A/10

6 PBM

C

Rho= -0.8144p=0.0180

0 10 20 30 40 50

day 112+4h

Days to VL>50cp/ml

Rho= -0.8795p=0.0062

0

500

1000

1500

0 10 20 30 40 50

0

500

1000

day 119+4h

Days to VL>50cp/ml

Rho= -0.9102p=0.0030

B

R ead-thro

ughTAR

longLTR

PolyA

Tat-Rev

R ead-thro

ughTAR

longLTRPolyA

Tat-Rev

day -21

day 112+4h

day 119+4h

Increase(d119+4h to d-21)

HIV RNA correlations

**

Time to VL>50cp/ml Time to VL>1,000cp/ml

0 20 40 60

0

5000

10000

15000

day 112+4h

Days to VL>1000cp/ml

long

LTR

HIV

RNA

copi

es/

106

PBM

C

Rho= -0.7785p=0.0295

600 20 40

0

2000

4000

6000

8000

day 119+4h

Days to VL>1000cp/ml

Rho= -0.7665p=0.0323

C

R ead-thro

ughTAR

longLTR

PolyA

TatRev

day -21

day 112+4h

day 119+4h

**

** *

* * *

HIV DNA vs transcripts correlations

0 500 1000 1500

0

200

400

600

800

1000day -21

longLTR HIV DNA copies/106 PBMC

Read

-thr

ough

HIV

RNA

copi

es/

106

PBM

C

Rho= 0.7833p=0.0172

0 500 1000 1500

0

500

1000

1500

2000day 112+4h

Rho= 0.8954p=0.0022

longLTR HIV DNA copies/106 PBMC

0 200 400 600 800 1000

0

200

400

600

800

1000day 119+4h

Rho= 0.8667p=0.0045

longLTR HIV DNA copies/106 PBMC

0 500 1000 1500

0

1000

2000

3000

4000

5000day -21

long

LTR

HIV

RNA

copi

es/

106

PBM

C

Rho= 0.7833p=0.0172

longLTR HIV DNA copies/106 PBMC

0 500 1000 1500

0

10

20

30

40

50day -21

Tat-

Rev

HIV

RNA

copi

es/

106

PBM

C

Rho= 0.8000p=0.0138

longLTR HIV DNA copies/106 PBMC

0 200 400 600 800 1000

0

20

40

60

80day 119+4h

Rho= 0.7667p=0.0214

longLTR HIV DNA copies/106 PBMC

Detection of different TAR sequences after romidepsin infusion

day -21 day 112+4h day 119+4h Neg Ctl

Romidepsin infusions #ID25

cell-associated HIV RNA TAR amplification Figure 3. HIV transcription profile before and after romidepsin therapy.Dynamics of each HIV transcript per million PBMC at baseline (day 0) and 4 hours after the second (day 112+4h) and the third (day 119+4h) romidepsin infusions. Each color represents a different individual from the REDUC part B study. Determinations below the limit of quantification (LOQ) are represented as solid dots.

Figure 4. Blocks to HIV transcription before and after romidepsin therapy.Changes in: transcriptional interference (read-through/total transcripts), elongation (elongated/total transcripts), completion (polyadenylated/elongated transcripts), and multiple-splicing (multiply-spliced/polyadenylated transcripts)at baseline (day 0) and 4 hours after the second (day 112+4h) and the third (day 119+4h) romidepsin infusions.Each color represents a different individual from the REDUC part B study.

Figure 5. Association between HIV DNA and HIV transcripts before and after romidepsin infusion, and between HIV levels and time to rebound (after ATI) or time to subsequent suppression (after restarting ART). (A) Spearman correlations between the total HIV DNA per million PBMCs and time to rebound after ATI (measured as daysto VL>50 copies/ml of plasma and days to VL>1,000 copies/ml of plasma), and time to suppression (quantified as days toVL<50 copies/ml of plasma after ART reinitiation);(B) Spearman correlations between the different HIV transcripts per million PBMCs and time to viral rebound (measuredas days to VL>50 copies/ml of plasma and days to VL>1,000 copies/ml of plasma); and(C) Spearman correlations between total HIV DNA per million PBMCs and the different HIV transcripts per million PBMCs.Each color represents a different individual from the REDUC part B study.

Figure 6. Droplet digital PCR (ddPCR) plot showing the detection of initiated HIV transcripts (TAR RNA) with different amplitudes (likely different sequences) after romidepsin infusion. One dimension ddPCR plot corresponding to the amplification of HIV RNA TAR sequences from participant #ID25 in channel 1 (FAM).The plot represents the droplet fluorescence amplitude detected at baseline (day 0) and 4 hours after the second (day 112+4h) and the third (day 119+4h) romidepsin infusions.Red square highlights the cloud of droplets with higher fluorescence amplitude, detected after romidepsin therapy.

Antiretroviral therapy (ART) cannot eliminate the HIV genomes integrated in latently infected cells, which are a major barrier to cure HIV [1-3].

One strategy to eradicate HIV consists of reactivating viral transcription with latency-reversing agents (LRAs), such as histone deacetylase inhibitors (HDACi).

A recent clinical trial, REDUC part B, analyzed the administration of the therapeutic HIV vaccine Vacc-4x and rhuGM-CSF as local adjuvant, in combination with the HDACi romidepsin. This approach showed an increase in unspliced cell-associated HIV RNA and residual plasma viremia after romidepsin infusions, along with a reduction in total HIV DNA [4]. However, the mechanism by which romidepsin reverses HIV latency in vivo remains unclear.

AIM: To characterize the HIV transcription profile before and after romidepsin therapy in available samples from the REDUC part B study.

1. After romidepsin infusions, we observed: Reactivation of transcriptionally silent proviruses (Fig. 6), An increase in HIV transcriptional initiation and especially elongation, but not completion or multiple splicing (Fig. 3-4), An inverse correlation between time to rebound after ATI and levels of both total HIV DNA and elongated HIV RNA (Fig. 5).

2. Romidepsin may play a role in strategies to reverse latency, but newapproaches are needed to increase HIV transcriptional completion andmultiple splicing, which are likely necessary for productive infection andimmune recognition/killing of HIV-infected cells.

3. Therapies that increase HIV transcription but do not lead to killing ofinfected cells may actually shorten time to rebound after ATI.

This work was supported by the National Institute of Allergy and Infectious Diseases at the National Institutes of Health (1R01AI132128, R56AI116342, R33AI116218, R56AI091573, U19AI096109), the National Institute of Diabetes and Digestive and Kidney Diseases at the National Institutes of Health (1R01DK108349), the U.S. Department of Veternas Affairs (IK2 CX000520, I01 BX000192), and the American Foundation for AIDS Research (amfAR) Institute for HIV Cure Research (109301).

1. Chun TW et al. Presence of an inducible HIV-1 latent reservoir during highly active antiretroviral therapy. PNAS (1997).2. Wong JK et al. Recovery of replication-competent HIV despite prolonged suppression of plasma viremia. Science (1997).3. Finzi D et al. Identification of a reservoir for HIV-1 in patients on highly active antiretroviral therapy. Science (1997).4. Leth S et al. Combined effect of Vacc-4x, recombinant human granulocyte macrophage colony-stimulating factor vaccination, andromidepsin on the HIV-1 reservoir (REDUC): a single-arm, phase 1B/2A trial. Lancet HIV (2016).5. Yukl S et al. HIV latency in isolated patient CD4+T cells may be due to blocks in HIV transcriptional elongation, completion, and splicing.STM (2018).

Limitations1. The parent study had sequential study interventions (vaccination and then romidepsin), and we did nothave access to samples between vaccination and romidepsin.2. Samples were available from only 9 of 17 trial participants, of whom only 1 had an increase in viral loadafter romidepsin.3. The presence of non-B subtypes may have affected HIV levels and detection frequencies.

#397