1. Barriers to HIV Cure Janet M. Siliciano, PhD
Associate Professor of Medicine
Johns Hopkins University School of Medicine
Baltimore, Maryland
FINAL:
Washington, DC: April 15, 2016
From JM Siliciano, PhD, at Washington, DC: April 15, 2016, IAS-USA.

2. After attending this presentation, participants will be able to:
Learning Objectives
After attending this presentation, participants will be able to:
Describe how the latent reservoir for HIV arises
List current approaches to curing HIV infection
Describe how these approaches will be evaluated clinically

3. How does early ART affect likelihood of cure?
Smaller latent reservoir
More rapid reservoir decay
No effect 1

4. How is the latent reservoir best measured?
Viral outgrowth assay
DNA PCR assays
Plasma HIV RNA
Western blot 1

5. Viral dynamics in patients on ART
Start
ART v +
t1/2 = 1 day
100000
Activated CD4+ T cells †
10000
t1/2 = 14 days
1000
Plasma HIV-1 RNA
(copies/ml)
100 10
Eradication in
2-3 years
In order to understand why curing HIV infection is so difficult despite extremely effective ART therapy I would like to briefly discuss viral dynamics which provides a basis for understanding the possibility of curing the infection.review what happens when patients start on a potent active antiretroviral therapy. In untreated patients, plasma virus levels are in the range of 10,000 to 100,000 virus particles per ml of plasma. When patients start on a potent ART regimen, viremia shows rapid exponential biphasic decay of plasma virus all the way down to the limit of detection (50 or 20 copies of HIV RNA / ml of plasma. This decay can be understood in terms of a simple model of viral dynamics in which uninfected cells interact with free virus, resulting in the generation of infected cells. Importantly, all antiretroviral drugs block new infection of susceptible cells but importantly do not block virus production by cells that already have an integrated provirus. Therefore the exponential decrease in viremia indicates the rapid turnover and decay of two infected cell populations that produce most of the plasma virus, which are activated CD4+ T cells, have a short half life in the productively infected state and don’t live long. The decay curve is actually biphasic, indicating that there is another population of infected cells with a longer half life, about 14 days. Assuming that there were only these two populations of infected cells, David Ho predicted in 1997 that it would be possible to cure the infection in 2-3 years. However, that did not happen. And this is because there is another population of infected cells that can produce virus and have a much longer half-life and this population are latently infected resting memory CD4+ T cells which arise as a consequence of the normal physiology of CD4+ cells. 1
Limit of Detection
(50 copies/ml)
0.1
0.01
0.001
100
200
300
Time on ART (d)

6. Physiology of resting and
activated CD4+ T cells
Naive
Memory
These are latently infected CD4+ T cells, and they arise as a result of the normal physiology of CD4+ T cells. Most of the CD4 cells in the body are in a profoundly quiescent state. These resting cells include naïve T cells which have not yet responded to any foreign antigen as well as memory cells that have previously participated in an immune response. These cells circulate throughout the tissues awaiting encounter with antigen.

7. Establishment of immunologic memory
Naive Ag †
Memory
When that happens antigen, the relevant cell becomes activated and divides, ultimately generated many activated effector cells with the same specificity. At the conclusion of the immune response, many of these activated cells die, but some survive and return to a resting state as memory cells. These cells persist for long periods of time, decades in fact, allowing future responses to the same antigen. Ag †

8. HIV infection of activated and resting CD4+ T cells
Naive
HIV Ag †
HIV Ag
Memory †
In HIV infection, the virus replicates mainly in the activated cells. and these infected cells die quickly, hence the rapid decay rate a. The virus does not replicate well in resting T cells
HIV

9. Establishment of the latent reservoir in resting CD4+ T cells
Naive † † †
HIV Ag † †
Memory †
However, one rare occasions some of the activated T cells can become infected as they are in the process of returning to a resting state. This results in a stably integrated form of the viral genome in a memory T cell, a cell population with a very low decay rate. And what is particularly interesting is that as the cell undergoes this profound change from an activated to a resting state, HIV gene expression is turned off.

10. NFκB sites in the HIV LTR
U 3 R
U 5
Modulatory region
Enhancer
Core
Cell DNA
This is the promoter of HIV. Several groups have shown that HIV gene expression is heavily dependent upon the host transcription factor NFkB which is excluded from the nucleus in resting cells. Thus HIV gene expression is automatically extinguished as the cells return to a resting state
AP1
NFAT1
USF1
Ets1
LEF
NFB
NFAT
Sp1
TBP
LBP1
Nabel G, et al. Nature. 1987;326:
Tong-Starksen SE, et al. PNAS. 1987;84:6845:6849.
Bohnlein E, et al. Cell. 1988;53:
Duh EJ, et al. PNAS. 1989;86:

11. Reactivation of latent HIV
Naive † † Ag †
HIV † †
Memory †
The end result is a stably integrated but transcriptionally silent form of the viral genome in a long lived memory T cell. This is a perfect mechanism for viral persistence - it allows the virus to persist essentially as pure information, unaffected by immune responses or antiretroviral drugs. If the cell becomes activated in the future, it can begin to produce virus again. In 1995, our group developed assays to detect these cells in vivo and found that they are present in everybody with HIV infection. Ag †

12. An assay for latently infected cells
ml blood
Negative
control
1.6x102
5x106
106
2x105
4x104
8x103
Purified resting
CD4+ T cells
1/1,000,000
PHA + irradiated
allogeneic PBMC
d2: add CD4+ lymphoblasts from HIV- donors
d7: add CD4+ lymphoblasts from HIV- donors
Describe assay. So this was the assay that we used to detect the presence and persistence of latently infected CD4 cells in patients
p24 Ag
Chun et al., Nature, 1997
Finzi et al., Science, 1997

13. Slow decay of latently infected
CD4+ T cells
10000
Time to eradication
> 73.4 years
0.0001
0.001
0.01
0.1 1 10
100
1000 2 3 4 5 6 7
Time on ART (years)
(per 106 cells)
Frequency
Finzi et al., Nature Med., 1999
Siliciano et al., Nature Med., 2003 -
As you can see from the results of this longitudinal study with many patients who were doing well on ART with undetectable plasma virus but yet we were able to measure latently infected cells with replication competent virus. As you can see, the problem is that this population of infected cells has in extremely long half life measured in years, not days. The latent reservoir is the major barrier to eradication and no matter how long patients have been on treatment, viremia rebounds 2 weeks treatment is stopped.

14. Slow decay of the reservoir
t1/2 = 44 months
Siliciano et al.,
Nature Med., 2003
The problem is that this population of infected cells has in extremely long half life measured in years, not days. The latent reservoir is the major barrier to eradication and no matter how long patients have been on treatment, viremia rebounds 2 weeks treatment is stopped. This study was published in 2003 but last year , Dr. Margolis and his colleagues published the results of their study in which they also measured the decay of latently infected resting CD4 cells isolated from suppressed patients. Many of these patients were on the newer ART regimens but as you can see, the results are essentially identical to the earlier study. The frequency of infected cells is about the same, 1 in a million with a 2 log range and essentially no decay. The half-life is almost identical in these 2 studies. We calculated a half-life of 44 months and Margolis calculated a half life of 43 months. So, even though the newer antiretroviral drugs are better with regard to convenience and tolerability, this has not changed the fundamental problem of persistence of latently infected cells.
t1/2 = 43 months
Crook et al, JID 2015

15. Latency results from infection of memory precursor cellsAg
These cells arise by infection in a relatively narrow time window and what is just an unfortunate accident of the patterns of gene expression after T cell activation. However, latently infected cells are present in all infected individuals and are the major barrier to curing infection
Deng et al., submitted

16. Residual viremia Plasma HIV RNA (copies/ml) Time on HAART (days/years)
Start Therapy
HAART
100000 Ag
10000
1000
Plasma HIV RNA
(copies/ml)
100
Limit of
Detection
(50 copies/ml) 10 1
0.1
1 copy/ml
0.01
0.001
The existence of a stable latent reservoir suggests that there should be at least one more phase in this decay curve of viremia. You can image that every day a small number of these latently infected cells become activated and begin to produce virus. This virus will not be able to infect additional cells because of the drugs but could in priniciple be detected in the plasma with a more senstive assay. In 1999, Roger Pomerantz showed that what HAART really does is reduce viremia to a new steady state that is below the limit of detection of clinical assays. Typically it is about 1 copy/ml of plasma.
There has been concern that this residual viremia represents some level of ongoing replication not fully suppressed by HAART, and that this residual revirmia by contribute to the remarkable stability of the reservoir by reseeding it. We have advanced the alternative hypothesis that residual viremia simple represents release of virus from stable reservoirs. To test this hypothesis, we have carried out direct sequence analysis of this trace level of viremia, which is difficult to do because the average level is about 1 virus particle/ml of plasma. We demonstrated that this residual viremia is drug sensitive, archival in character, and non-evolving. These results, particularly the stunning lack of viral evolution, are consistent with the idea that residual viremia largely represents release of virus from stable reservoirs rather than continuous ongoing cycles of replication.
100
200
200
years
Time on HAART (days/years)
Hermankova et al, JAMA, 2001
Persaud et al, J Virol, 2003
Kieffer et al, J Infect Dis, Nettles et al, JAMA, 2005
Bailey et al, J Virol, 2006
Brennan et al, J Virol, 2009
Sensitive to current regimen
Archival
Non-evolving

17. Residual viremia Plasma HIV RNA (copies/ml) Time on HAART (days/years)
Start Therapy
Add 4th drug
HAART
100000 Ag
10000
1000
Plasma HIV RNA
(copies/ml)
100
Limit of
Detection
(50 copies/ml) 10 1
0.1
1 copy/ml
0.01
0.001
The existence of a stable latent reservoir suggests that there should be at least one more phase in this decay curve of viremia. You can image that every day a small number of these latently infected cells become activated and begin to produce virus. This virus will not be able to infect additional cells because of the drugs but could in priniciple be detected in the plasma with a more senstive assay. In 1999, Roger Pomerantz showed that what HAART really does is reduce viremia to a new steady state that is below the limit of detection of clinical assays. Typically it is about 1 copy/ml of plasma.
There has been concern that this residual viremia represents some level of ongoing replication not fully suppressed by HAART, and that this residual revirmia by contribute to the remarkable stability of the reservoir by reseeding it. We have advanced the alternative hypothesis that residual viremia simple represents release of virus from stable reservoirs. To test this hypothesis, we have carried out direct sequence analysis of this trace level of viremia, which is difficult to do because the average level is about 1 virus particle/ml of plasma. We demonstrated that this residual viremia is drug sensitive, archival in character, and non-evolving. These results, particularly the stunning lack of viral evolution, are consistent with the idea that residual viremia largely represents release of virus from stable reservoirs rather than continuous ongoing cycles of replication.
100
200
200
years
Time on HAART (days/years)
Residual viremia cannot be reduced by treatment intensification
Dinoso et al, PNAS, 2009
Many later raltegravir intensification studies

18. The first cure
Host immune system, including latently infected cells, largely eliminated by condition regimen (chemo + irradiation and by graft vs host disease.
Donor cells protected from HIV infection due to absence of CCR5
What about Mr. Brown’s entire immune system …

19. “Boston Patient B” Plasma HIV RNA (copies/ml)
Below limit
of detection
10,000,000
Matched
allogeneic
HSCT
Stop
ART
1,000000
100,000
10,000
TDF
FTC
RAL
Plasma HIV RNA
(copies/ml)
1000
100
Recently Henrich et al reported two transplant cases in which CCR5 wild type donors were used and the donor cells were protected from infection simply by continuing antiretroviral therapy throughout the tranplant period and for a few years thereafter. Then treatment was stopped and what happened was really interesting. The normal rebound at 2 weeks did not occur. Viremia remained undetectable for several months and then suddenly shot to a million copies/ml, presumably reflecting the activation of one of a small number of latently infected cells remaining in the patient. 10 1
-42
-30 2 4 6 8
Time after Rx interruption (months)
Henrich et al, JID, 2013

20. The Mississippi baby >2 years
1,000000
ART discontinued
These delayed rebound cases prove that HIV can persist in a latent form for years and then begin to replicate
Below limit
of detection
AZT
3TC
LPV/r
100,000
10,000
Plasma HIV RNA
(copies/ml)
>2 years
1000
100
Recently, our Hopkins colleague Debbie Persaud has reported a second possible cure case, an infant who acquired HIV in utero from an infected mother who had no prenatal care. At birth, the viral load was 20,000 and an astute pediatrician started an antiretroviral therapy 30 hours after birth. Viremia fell to below the limit of detection and remained there However, at 18 months, therapy was stopped against medical advice. Remarkably, the viral load remained undetectable. Viral outgrowth assays done by Dr. Persaud have been negative. We think what happened in this case was that therapy was started before the latent reservoir in resting memory T cells could be established because 10 10 20 30 40 50
Months after Birth
Persaud D et al., NEJM 2013

21. Approaches to HIV cure Gene Rx † † Prevent reactivation † †
TCR pathway agonists †
Prevent reactivation
LRAs
(HDACi) † †
We have explored these issues using our primary cells model. When latency is reversed through T cell activation, the infected cells die quickly. But when latency is reversed with a histone deacytlase inhibitor that does not induce T cell activation, the cells don’t die. They just sit there and produce virus. †
Induce elite control
Shock and kill

22. Other approaches to HIV cure
Gene Editing Strategies used in Cure Research: target integrated provirus with engineered nucleases (ZFN,TALENS,) or CRISPR/Cas9 x x
Problems
Various gene editing strategies are being tested both in vitro and in clinical trials. These strategies involve nucleases which act on specific DNA sequences to create mutations. They can be targeted to the HIV provirus to render it unable to make infectious viral particles. The main problem with these approaches is that you have to deliver the nucleases into every infected cell in vivo. There is currently no way to do this. In addition, there are off target effects.
No way to deliver enzymes into every infected cell in vivo
Off target effects

23. Other approaches to HIV cure
Gene Rx – ZFN targeting CCR5 gene in patient CD4+ T cells or HSC. Reinfuse engineered, HIV-resistant cells back into patients x x
Problems
These nucleases can also be used to disrupt the CCR5 gene in CD4+ T cells or hematopoietic stem cells from patients. These HIV-resistant cells are then reinfused into the patient. The problem with this approach is what to do about all of the non-engineered cells that are still HIV susceptible. In the case of the Berlin patient, which received a transplant with resistant cells, all of the HIV susceptible host T cells, including latently infected cells, were eliminated by the conditioning regimen and/or by graft vs host disease which is a consequence of an allogeneic bone marrow transplant. Autologous cells can be readily engineered to knock out CCR5, but without graft vs host disease, the viral reservoir will not be eliminated.
HIV can still replicate in non-engineered cells. (In Berlin patient, CCR5+ host cells eliminated by conditioning regimen and graft vs. host effects)

24. Fundamental approach to HIV cure
How do we identify latency reversing agents?
TCR pathway agonists
LRAs
(HDACi) †
Will cells be eliminated following reversal of latency?
We have explored these issues using our primary cells model. When latency is reversed through T cell activation, the infected cells die quickly. But when latency is reversed with a histone deacytlase inhibitor that does not induce T cell activation, the cells don’t die. They just sit there and produce virus.
How do we measure the reservoir in eradication trials?

25. Current status of LRA trials
Numerous LRAs identified in studies with transformed cell lines and primary T cell model systems
Few shown to work ex vivo with cells from patients
TCR pathway agonists
LRAs
(HDACi) †
In clinical trials, no reduction in the reservoir yet demonstrated
In clinical trials, evidence for increases in cell-associated HIV RNA (Archin et al.)
We have explored these issues using our primary cells model. When latency is reversed through T cell activation, the infected cells die quickly. But when latency is reversed with a histone deacytlase inhibitor that does not induce T cell activation, the cells don’t die. They just sit there and produce virus.
Some evidence for slight transient increases in plasma HIV RNA after LRA treatment (romidepsin, panobinostat, TLR7 agonist)

26. Assay for reversal of latency using patient resting CD4+ T cells
500 x106 resting
CD4+ T cells
Positive control
+Test compound
TCR
agonist
LRA
we are now using leukapheresis so that we can get hundreds of millions of resting CD4 cells from patients on HAART. The cells are plated in limiting dilution and treated with test compounds. The next step is normally the addition of CD4+ T lymphoblasts from a normal donor to expand any viruses that are released. However, in this situation, there is a problem because these lymphoblasts can themselves cause activation of resting cells either through allogeneic effects or cytokine production. This result in a high background. So instead, we used MOLT4-cells transfected with CCR5. These cells don’t cause any backstimulation and give support viral outgrowth as well as primary lymphoblasts. In collaboration with Dave Margolos, we are making assays of this kind available to the AIDS research community through an NIH sponsored program within the Martin Delaney CARE collaboratory. We hope this will help to clear up …./
Measure intracellular HIV RNA
and virion release
Shan et al, J Virol, 2013
Laird et al, PLOS Pathogens, 2013
5 x 106 cells/well
Bullen et al., Nature Med, 2014

27. Induction of HIV RNAs by LRAs
Total RNA
polyA primers
18 hrs
Bullen et al, Nat Med 2014

28. Induction of HIV RNAs by combinations of LRAs
Single LRA
+ Bryostatin-1
+ Disulfiram
% of PMA/ionomycin
Laird et al, in preparation

29. Fundamental approach to HIV cure
How do we identify latency reversing agents?
TCR pathway agonists
LRAs
(HDACi) †
Will cells be eliminated following reversal of latency?
We have explored these issues using our primary cells model. When latency is reversed through T cell activation, the infected cells die quickly. But when latency is reversed with a histone deacytlase inhibitor that does not induce T cell activation, the cells don’t die. They just sit there and produce virus.
How do we measure the reservoir in eradication trials?

30. Fate of infected CD4 cells after latency reversal in vivo is unknown
TCR pathway agonists
HDACi 6 4 7
Days after reactivation 3 2 5
120
100 60 80 40 20
Residual GFP+ cells
(%) †
αCD3+αCD28
We have explored these issues using our primary cells model. When latency is reversed through T cell activation, the infected cells die quickly. But when latency is reversed with a histone deacytlase inhibitor that does not induce T cell activation, the cells don’t die. They just sit there and produce virus.
HDACi
Shan et al, Immunity, 2012

31. CTL killing of latently infected cells treated with SAHA
Normal donor 1
Normal donor 2
Normal donor 3 1 8
Elite suppressor 1
Elite suppressor 2
Elite suppressor 3
Surviving infected cells (%) 6 4
If you do this with cells from normal donors, there is not HIV specific killing. If you it with cells from elite supressor, who generally have strong HIV-specific CTL responses, you see the disappearance of infected cells. 2 2 4 6 8
Time of coculture (days)
E:T = 1:1
Shan et al, Immunity, 2012

32. CTL killing of latently infected cells treated with SAHA
Normal donor 1
Normal donor 2
Normal donor 3 1 8
Elite suppressor 1
Elite suppressor 2
Elite suppressor 3
Surviving infected cells (%) 6
HAART patient 1
HAART patient 2
HAART patient 3
HAART patient 4
HAART patient 5
HAART patient 6 4
Now the critical issue is what happens in patients on HAART, because eradication will be tried in these patients. We found that in most patients, killing was weak, reflecting qualitative and quantitative defects in the HIV-specific CTL response. Importantly, these defects could be reversed by in vitro stimulation of the CD8 cells with gag peptides. 2 2 4 6 8
Time of coculture (days)
E:T = 1:1
Shan et al, Immunity, 2012

33. CTL escape variants dominate in the latent reservoir of chronic patients1
Acute Pt10 5
A*02:01 G K 9 E V 9 S L 9 T V 9 E I 8 G L Y 9 D L 9 F K 1 1
Chronic Pt 18 5
A*02:01 W F 9 S L 9 T V 9 T L 9 H A 9 P Y 9 V I 9 F K 1 1 5
Acute Pt 12
A*03:01 K K 9 R K 9 S V 9 T L 9 H A 9 G L 9
Frequency of variants (%) 1
Chronic Pt 39 5
A*03:01 K K 9 R K 9 L Y 9 S V 9 T L 9 H A 9 G L 9 1 5
Acute Pt07
B*58:01 R Y 1 1 V L 8 T W 1 Y L 9 Q W 9 1
Chronic Pt12 5
B*57:01 L Y 9 I W 9 K F 1 1 T W 1 Q W 9 CT L
epitopes in HI V
-1 Gag
Documented Escape
Diminished Response
Mutation T
ype Not Determined
Deng et al, Nature, 2015

34. Fundamental approach to HIV cure
TCR pathway agonists
LRAs
(HDACi) †
How do we measure the reservoir in eradication trials?
How do we identify latency reversing agents?
Will cells be eliminated following reversal of latency?
We have explored these issues using our primary cells model. When latency is reversed through T cell activation, the infected cells die quickly. But when latency is reversed with a histone deacytlase inhibitor that does not induce T cell activation, the cells don’t die. They just sit there and produce virus.

35. An assay for latently infected cells
ml blood
Negative
control
1.6x102
5x106
106
2x105
4x104
8x103
Purified resting
CD4+ T cells
1/1,000,000
PHA + irradiated
allogeneic PBMC
d2: add CD4+ lymphoblasts from HIV- donors
d7: add CD4+ lymphoblasts from HIV- donors
The viral outgrowth assay originally used to define the reservoir requiring 2-3 week of culture work in a BLS3 facility and there is great interest in simpler methods. Of course you can detect the viral genome with PCR
p24 Ag
Chun et al., Nature, 1997
Finzi et al., Science, 1997

36. Assays for latent HIV Viral outgrowth assay (VOA) DNA PCR
TCR
agonist
Viral outgrowth
assay (VOA)
DNA PCR
PCR for proviral DNA
Measure intracellular
HIV RNA
Induction of HIV RNA
TCR
agonist
we are now using leukapheresis so that we can get hundreds of millions of resting CD4 cells from patients on HAART. The cells are plated in limiting dilution and treated with test compounds. The next step is normally the addition of CD4+ T lymphoblasts from a normal donor to expand any viruses that are released. However, in this situation, there is a problem because these lymphoblasts can themselves cause activation of resting cells either through allogeneic effects or cytokine production. This result in a high background. So instead, we used MOLT4-cells transfected with CCR5. These cells don’t cause any backstimulation and give support viral outgrowth as well as primary lymphoblasts. In collaboration with Dave Margolos, we are making assays of this kind available to the AIDS research community through an NIH sponsored program within the Martin Delaney CARE collaboratory. We hope this will help to clear up …./
Measure virion release
Induction of virion production
TCR
agonist

37. Viral outgrowth vs PCR assays
Total HIV
DNA
2 LTR
circles
Residual
viremia
Assay
Total HIV DNA
Integrated HIV DNA
Cell/tissue
Resting CD4
PBMC
Resting CD4
PBMC
Resting CD4
Rectal CD4
PBMC
Plasma
10,000 1
0,000
Infected cell frequency
(per 106)
1,000
rho = 0.19
p = 0.31
rho = 0.07
p = 0.71 1 ,
300x
Plasmas HIV RNA (copies/ml)
100 1 10 1
We recently compared the viral outgrowth assay with several state of the art PCR assays and basic result was that PCR assays give infected cell frequencies that are on average 300 fold higher than, and poorly correlated with the viral outgrowth assay. 1 1
r = 0.38
p = 0.28
r = 0.70
p < 0.01
r = 0.41
p = 0.13
r = 0.05
p = 0.86
0.1
0.1
Cohort
Chronic
Acute
Chronic
Acute
Chronic
Acute
Chronic
Acute
Chronic
Acute
Chronic
Acute
Chronic
Acute
Chronic
Acute
Eriksson et al, PLOS Pathogens, 2013

38. Non-induced proviruses
Resting
CD4+ T cells
Negative
control
1.6x102
5x106
106
2x105
4x104
8x103
full length, single genome analysis
Non-induced proviruses
PHA + irradiated
allogeneic PBMC
Are they inducible?
d2: add CD4+ lymphoblasts from HIV- donors
d7: add CD4+ lymphoblasts from HIV- donors
This means that there are many proviruses that are not detected in the culture assay. We call these non-induced proviruses. We don’t know that they are non-inducible. To find out Ya-Chi Ho, a very talented HHMI international fellow, has used full length single genome analysis to characterize a large number of non-induced proviruses from wells that are negative for viral outgrowth.
p24 Ag
Ho et al, Cell, 2013

39. Non-induced proviral clones (n=213)H O 2
TGG
TAG
Trp
Stop
Hypermutated
32.4%
Results from the analysis of 213 non-induced proviral clones from 8 different patients are summarized here About a third lethally hypermutated by another host restriction factor APOBEC3G which causes GA hypermutation.
Ho et al, Cell, 2013

40. 32.4% of non-induced proviruses have lethal GA hypermutation
HXB2
Pt 09 clone 31E05
Pt 09 clone 31E11
Pt 20 clone 36D12
Pt 20 clone 33C03
Pt 20 clone 33C09
Pt 20 clone 33G10
ATG  ATA
MI
start codon mutation
HXB2
Pt 09 clone 31E05
Pt 09 clone 31E11
Pt 20 clone 36D12
Pt 20 clone 33C03
Pt 20 clone 33C09
Pt 20 clone 33G10
Here is the start codon of gag and gag pol ORFs. It is followed by a glycine codon, generais fting an APOBEC3G consensus sequence. So the start codon is frequently mutated to Ile. And whenever you have a Trp, mutation of either one or both of the Gs to A give a stop codon. So these proviruses have stop codons in most open reading frames and are dead for replication but are detected by standard PCR. Similarly,
TGG  TAA, TAG, TGA
Tryptophan  stop codon
nonsense mutation
Ho et al, Cell, 2013

41. 45.5% of non-induced proviruses have large internal deletions2
TGG
TAG
Trp
Stop
Large internal
deletion
45.5%
Hypermutated
32.4%
Almost half of the non-induced proviruses have large internal deletions which arise during reverse trancription due to an intermoleclar template switching event between short regions of repeated sequence. This deletes one of the repeats and the intervening sequence. The proviruses are missing a large fraction of the genome and are dead for replication. But could be detected by PCR depending on the primers.
Ho et al, Cell, 2013

42. Deletions and hypermutation
Pt. KB7
Pt. KB6
Pt. KB5
First did this with chronically treated patients to have unbiased screen in patients.
Dark blue are intact regions. Lighter blue is hyper mutated sequence. White is mapped deletion. Dark gray is non-sequenced deletion. Other gray tones are estimates of min and max based on primer locations. As you can see, highly defective, even more than the previous study.
Question leading in. …….Do deletions accumulate slowly over time, or is reverse transcription only successful in making an intact template a fraction of the time?
Pt. KB3
Pt. KB8
Bruner et al, submitted

43. Non-induced proviruses
Non-induced proviral clones (n=213)
Nonsense mutations/
INDELS 3.8%
11.7% Intact
genome N H O 2
TGG
TAG
Trp
Stop
Deletion in ψ/
MSD site 6.5%
Large internal
deletion
45.5%
Hypermutated
32.4%
…..However, 12% of the non-induced proviruses appear completely intact at the primary sequence level. To determine whether these viruses can replicaiton Ya-Chi painfully reconstructed them using the direct sequening result and gene synthesis to create a near full length genome that is inserted into a plasmid carrying a reference HIV provirus NL43. A small region of NL43 sequence is corrected to pateint sequence so that after transfection into 293 cells, virus productionm and infeciton of primary CD4 clls, we get a provurs that is 100% pateint derived. We can then ask how well this virus replicates
What is particulary
Ho et al, Cell, 2013

44. Replication capacity of intact non-induced proviruses
0.01
0.1 1 10
100
1000
0.01
0.1 1 10
100
1000
NL4-3
Rep-Comp
Patient 10
Patient 16
p24 (ng/ml)
Intact non-
induced 7
0.01
0.1 1 10
100
1000 7
0.01
0.1 1 10
100
1000
Relative to
Patient 17
Patient 20 1 3 5 1 3 5
Time post infection (days)

45. Size of latent reservoir
HIV DNA
62 fold
VOA
Intact
Scale=100/106
Ho et al, Cell, 2013

46. Can intact non-induced proviruses
be induced?
ml blood
Negative
control
1.6x102
5x106
106
2x105
4x104
8x103
Recover cells from
negative wells
Purified resting
CD4+ T cells
PHA + irradiated
allogeneic PBMC
p24 Ag
d2: add CD4+ lymphoblasts from HIV- donors
d7: add CD4+ lymphoblasts from HIV- donors
Ho et al, Cell, 2013
The viral outgrowth assay originally used to define the reservoir requiring 2-3 week of culture work in a BLS3 facility and there is great interest in simpler methods. Of course you can detect the viral genome with PCR

47. Can intact non-induced proviruses
be induced?
Resting CD4+ T cells
PHA+ allo PBMC + -
47%
53%
PHA+ allo PBMC + -
39%
61%
PHA+ allo PBMC + -
39%
61%
Ho et al, Cell, 2013
Nina Hosmane

48. Take home points There is no clinical assay for the latent reservoir
DNA PCR assays widely used for reservoir analysis mainly defect grossly defective proviruses
The quantitative viral outgrowth assay remains the best available assay for the latent reservoir, but better assays are urgently needed.
Other approaches: transient blips following LRA administration, time to rebound after ART interruption

49. Predicting time to rebound after reservoir reductions
Hill et al, PNAS 2014

50. Time to rebound Log reduction in latent reservoir Time to rebound
Miss.
baby 6
Boston
pt. A
Boston
pt. B 5
Berlin pt. 4
Chun et al.
Log reduction in
latent reservoir 3 2 1
1 wk
1 mo
3 mo
1 yr
10 yr
Lifetime
Time to rebound
Hill et al, PNAS 2014

51. What will cure look like?
1,000,000
Therapeutic vaccination
100,000
cART
cLRAs
10,000
Plasma HIV RNA (copies/ml)
1000
100
(weeks) (years)
Time Post Infection

52. Korin Bullen Ya-Chi Ho Robert Siliciano Greg Laird Liang Shan Kai Deng
I would like to thank several outstanding graduate students in our group, Ya-Chi Ho, Korin Bullen, Liang Shan, Kai Deng and Greg Laird, who did most of the experimental work I showed, and also all of our collaborators in the assay comparison study, particularly Steve Deeks. That study was supported by amFAR and by the Martin Delaney Collaboratories and by HHMI. Thank you and I will be happy to take any questions.

53. Thanks Collaborators Funding Steve Deeks Richard Flavell Dave Margolis
Joel Gallant
Joe Cofrancesco
Doug Richman
Martin Nowak
Matt Strain
Sarah Palmer
Una O’Doherty
Steve Yukl
John Mellors
Funding
NIH: Martin Delaney Collaboratories
CARE and DARE
Howard Hughes Medical Institute
Foundation for AIDS Research
(amfAR): ARCHE
Johns Hopkins Center for AIDS
Research
Bill and Melinda Gates Foundation