Skip to main content

Transforming the understanding
and treatment of mental illnesses.

Macrophage Infection by HIV: Implications for Pathogenesis and Cure: Day Two

Watch on YouTube.


DR. KHALILI:  Welcome to the second day of the NIMH-sponsored conference on macrophage infection by HIV and the strategies for elimination.  My name is Kamel Khalili, and I will be moderator of this session for this morning.

Before we get to the talks, I want to cover some housekeeping issues, which, as you see here, all participants will be muted in listen only -- in the listen-only mode, and the cameras will be turned off.  Please submit your question by Q&A answer box any time during the presentation.  If you have any technical difficulties hearing or viewing the webinar, please note this in the Q&A box, and our technical personnel will take care of that.  You can also send an email to  The most important thing here is the speakers, so please stay on the 15 minutes presentation time that was arranged by the organizers.

So with that notion, I would like to introduce you to our speakers.  We have a great lineup of these six talks, and the first talks will be presented by Morgane Bomsel from Institut Cochin, and then the second talk is by James Termini from University of Miami.  The third one is by Rebecca Peters, University of Miami, and then the next one is by Haitao Hu from the University of Texas Medical Branch.  And then next is Howard Gendelman from the University of Nebraska Medical Center, and final talk will be delivered by Rafal Kaminski from Temple University.  Thank you very much.

DR. BOMSEL:  Good morning to all of you.  Thank you for inviting me to this great meeting.  Today I will tell you about two things:  how HIV can infect mucosal macrophages and establish latency in vitro, and how replication-competent HIV reservoir form in macrophages of HIV-infected individual under suppressive cART.

HIV is mainly transmitted at genital mucosa, and what we have tried to do in the lab for the last past year is to understand how circumcision can protect HIV entry in the male genital by 50 percent.  And using ex vivo explant infection as well as reconstruction infection -- mucosal reconstruction infection, we have shown that HIV can penetrate by two pathway to penetrate the penile tissue, first through the foreskin where HIV target Langerhans cells trans-infect CD4 T cell.  Second, HIV penetrates the urethra, which will be one topic of today, and target directly macrophages, but not T cell.  Importantly, HIV-infected cell present in all fluid vectorizing the infection, such as semen or cervical vaginal secretion, mediate efficient penile entry where cell-free viruses is not efficient.

We recently asked whether HIV reservoir form in mucosal macrophage following infection.  Therefore, we reconstructed urethral mucosa made of fibroblast macrophages overlaid by urethral epithelial cell, and after 17 days, the reconstruction polarized was inserted into microscopic device and inoculated in situ with gag-GFP-HIV-infected T cell producing GFP viruses or with cell-free viruses.  When the reconstruction was observed by spinning disc live imaging, as you can see here, T cell adhere to the surface of the epithelium, forming virology synapse able to shed burst of viral particle that will penetrate inside the tissue.

Next, by observing live imaging recording was -- we could measure that the time required for the infected T cell to form a productive synapse with epithelial, which lasts approximately a half an hour before the T cell is able to shed virus for another hour and leave its site of (inaudible) as you just show.  What happens to the virus produced at this viral synapse?  We could follow the virus through the epithelial, as you can see here, and the virus translocate by transcytosis before to infect macrophage.  We found HIV DNA positive in the storm of the reconstruction.  This macrophage also contained HIV protein concentrate in a circular structure reminiscent of virus-containing compartment that are the hallmark of HIV-infected macrophages.  Furthermore, we found that macrophage infection establishes below the site where the viral synapse form both in video and by electron microscopy, as you can see here.

We next quantified the virus produced by this infected reconstruction.  And as you can see here, the virus is produced for approximately two weeks before to stop its production, suggesting such a macrophage until latency.  Furthermore, these macrophages could be reactivated to produce viral particle upon stimulation by LPS that target specifically TLR4 on macrophages, suggest -- as you can see here in pink, suggesting that following mucosal entry, HIV not only infects macrophages, but also establishes a latent replication competent infection, which is the definition of an HIV reservoir.

We next asked whether replication-competent HIV reservoir could form in tissue macrophages in vivo.  I will not repeat the nice introduction Rebecca Veenhuis made yesterday on the crucial role of HIV reservoir to -- as a barrier to eradication but will more concentrate on the differences between T cell and macrophage infection which will help to understand why mucosal tissue macrophages are ideal candidates to form HIV reservoir.

Indeed, macrophage productive infection occurs in nonproliferating T cell -- in nonproliferating cells, like T cell.  Macrophage-accumulating HIV viruses, in intracellular structure I alluded to before, namely virus-containing compartment where you see viral particle.  Macrophage are resistant to HIV cytopathy like T cell.  HIV-infected macrophages naturally stop producing viruses several weeks after infection, at least in vitro, unlike T cell which require cART.  Macrophage can be tissue-resident long-lived cells that can self-renew as nicely presented yesterday by Filip.  HIV macrophage are inefficiently killed by CTL unlike T cell.  And finally, macrophage sustain HIV replication in humanized mice model lacking CD4 T cell as shown by the group of Victor Garcia.  With this model, we could also show that HIV rebounds upon ART interruption.

So in our -- for our study, we used urethral tissues from HIV-infected cART-treated individual.  We sustained and detected the undetectable blood viral load undergoing transgender surgery.  We prepared cell suspension from this tissue and could observe HIV-integrated DNA by nested PCR in the total whole cell suspension.  When the cell was sorted, integrated DNA were only found in macrophages, not on T cell, which was confirmed by Cheo Eugenin using HIV DNA fish.  HIV knife DNA was only present in IB-1 positive macrophages and not in T cell.

Next and most importantly, to define two HIV reservoir, we had to evaluate their capacity to resume HIV production upon stimulation of the cell suspension, what we did using either LPS to target macrophages or using PHA to target CD4 T cell.  And as you can see here, the viral production being quantified by quantitative viral outgrowth assay only stimulation of macrophage both in the epithelium and in the stroma resulted in the production of replication-competent virus from the cell suspension.  Furthermore, when the cell suspension was depleted in macrophages in blue, the viral signal was dramatically decreased almost to insignificant signal.  So these results indicate that this macrophage containing HIV DNA that had been shown for several years by other group were indeed replication competent and, thus, bona fide reservoir.

Next we found HIV RNA in section of tissue both in the epithelium and in the stroma only if it is CD68 macrophages, not on CD3 T cell.  The same for HIV contain p24, p55 that reside only in CD68 macrophages, not in the blue CD3 T cell.  And when we zoom on this macrophage containing the HIV protein, they reside in structure resembling virus-containing compartment, not only by -- detected not only by immunofluorescence, but also at the EM level, and you can see here, virus in a compartment.  And of note, this VCCs in tissue are much smaller than those observed -- found in vitro infection of macrophages.

Then we asked which is a kind of polarization that harbor this tissue macrophage.  And as you can see here, HIV cART -- in cART sequester patient, macrophage express IL-1 receptor, both IL-4 and CD206, but not CD163, which is a polarization state intermediate between M1 and M2, and which is in agreement to what Filip state yesterday about the lack of defined polarization, M1/M2, in vivo.  And interestingly, this set of intermediate macrophage is largely enriched in tissue from infected patient as compared -- also that do exist in an infected tissue.  Finally, we could show that this intermediate macrophage subset are the cells that contain the HIV reservoir as it contained HIV p24, both by morphology and quantitatively by flow cytometry.

So our take-home message is that HIV target mucosal macrophages upon its mucosal entry at least in vitro.  Then that viral reservoir in macrophage can establish very early after HIV infection.  I think that for this audience, this polarized mucosal reconstruction present a valid model to study HIV macrophage reservoir dynamics as well as its regulation by the tissue environment.  Third or fourth, tissue macrophage host replication-competent HIV reservoir in HIV cART-suppressed patient.

And finally, macrophage and T cell reservoir differ in dynamics and nature.  And this this has a clinical relevance as most strategies to cure reservoirs are targeted to eliminate latently-infected cell and block-and-lock do not eradicate latently-infected macrophages.  So HIV cure strategy should consider and adapt to macrophage HIV reservoir.  For instance, here we can propose a TLR agonist when we are studying, TLR agonist that might help in reactivating macrophage reservoir as a first step of shock-and-kill strategy.

So a remaining question, we study the dynamics VCCs and whether macrophage are true or in active latency.  Study also whether similar subset from HIV reservoir are in different tissue.  When we study HIV reservoir, the most interesting part for us is the origin of this HIV reservoir macrophage where there are long-lived versus inflammatory monocyte and, finally, tied to design strategy to eliminate these macrophages.

So I want to acknowledge Yonatan Ganor and Fernando Real, two talented scientists in my lab, as well as many collaborators, especially Cheo Eugenin and Marila Gennaro, and all patient and the funders.  And I thank you for your attention and will now present James Termini for the next talk.

DR. TERMINI:  Okay.  So first, I want to take a chance to thank the -- thank the organizers for inviting me to speak today.  My name is James Termini.  I'm from the University of Miami, and today I'm going to speak to you on the -- Modifying the Effector Functions of Delivered -- AAV-Delivered Anti-HIV Monoclonal Antibodies.  And it's a little bit of a change, so I hope you enjoy.

So AAV, or adeno-associated virus, is really an ideal vector for delivery of broadly-neutralizing antibodies as a treatment for HIV.  I don't have time to go into all the specifics of AAV, but the takeaway points are that it provides for long-term delivery of potent broadly-neutralizing antibodies.  We followed animals for over six years that continuously expressed high levels of antibody, and it's really along the lines of that whole one shot and you're good for life strategy as opposed to other methods of passive transfusion of broadly-neutralizing antibodies which require re-administration every three or four weeks.

And the best part about this technology is that this is a method of treatment that's completely independent of a recipient's immune system.  Like traditional vaccines that need to elicit these broadly-neutralizing antibodies in patients which can take years and years, whatever antibody We encode in the AAV vector will be expressed at high levels long-term.  And probably the most important AAV has been considered safe and it's already being used in the clinic for gene therapy.

So this is some data generated by the Desrosiers Lab.  This is a macaque study here looking at six individual macaques that were given AAV1 intramuscularly to deliver 5L7, which is an anti-SIB antibody.  And you can see that the levels of delivery of antibody can vary anywhere from 20 to 60 to even two to 300 micrograms per mL.  And these antibodies can last a very, very long time, or they can -- it continuously expresses for quite a long time.  This is one of our animals we refer to as our star performer.  It's 84-05.  And it -- we followed this animal for over six years, and it's made between two to 300 micrograms per mL of 5L7 for over six years now.  What's more amazing is that this animal has been IV challenged with a 10x infectious dose of SIV mAB239 six different times and still remains uninfected.

I also wanted to highlight one other animal affectionately referred to as the Miami monkey.  This is a study that the Desrosiers Lab published in Immunity back in 2019.  This is a really amazing animal.  This animal was infected with SHIV-AD8, and 86 weeks of infection was administered the AAV1 express, 10-1074, and 3BNC117.  And you can see the viral load just dropped to undetectable and for 340 weeks.  This this this animal really nicely highlights what I'm going to be talking to you about today because even at these late timepoints, if you were to take a lymph node and perform a viral outgrowth assay, you can still recover virus, although very little, from this animals.  So there are still reservoirs even after such a long-term expression of potent broadly-neutralizing antibodies.

And so one of the things that I focus on is enhancing these antibodies, enhancing the effector function to more appropriately target and destroy the viral reservoir.  And so one of the approaches that I took to try to enhance this activity is by turning to fucose.  So human IgG, the FC H2 portion contains single N-linked glycans depicted here.  Now this N-linked glycan has one little, tiny optional alpha-1-linked -- alpha-1,6-linked fucose.  And what's amazing about this is that antibodies that lack fucose at disposition -- oops, I apologize -- showed really enhanced affinity for FCMR3A on T cells, and this in leads to enhanced ADCC.  So this is not a new approach.  These antibodies are currently being treated -- well, non-fucose later versions are currently being used to treat cancer.  And so this is just a crystal structure here, and you can see this is the N-glycan.  And as I said, this alpha-1,6-linked fucose is an optional residue.  This is kind of a top-down view.  You can see these sugars.  And once again, this is the region that interacts with the Fc gamma RIIIA.

So to highlight this, this is just some data from a publication that was published in 2019 working with rituximab, and they were looking at the ability of rituximab, so anti-CD22, to eliminate CD20 positive cells.  And you can see here from the normal fucosylated versions, the antibody, versus the a-fucsolyated, you see a drastic enhancement of cytotoxicity or cell killing, I mean, going from almost 20 percent at .1 microgram to over 60 percent, which is pretty impressive.

This approach has been looked at before in anti-HIV antibodies.  For instance, this is some data from Moldt et al. published in the Journal of Virology in 2012.  So when you're comparing a-fucosylated B12 to fucosylated, you do see a very nice enhancement of viral inhibition at lower antibody levels even.  You do see an enhancement of ADCC around tenfold as well as you see enhanced NK-cell degranulation as you would expect.

So I started playing around with this technology on some of my more modern kind of broadly-neutralizing antibodies.  So I started by making a FUT8 eight knockout cell line using CRISPR CAS9, and using an AAL probe to stain for the alpha-1,6-linked fucose, you can see antibodies produced in a 293 T cell line have plenty of fucose whereas that is produced in my knockout cell line, we do not see fucose present.  So these are just some standard variations in LS, allow LALA mutant to knock out ADCC activity and then just a LALA/LS combination.  Now, what's interesting is that even with the removal of fucose -- with the removal of fucose, you see identical trimer binding here, and you see just about identical viral neutralization.  However, when you look at the ADCC activities, you know, 10-1074 here, when you're comparing the red circles of the wild-type 10-1074 to the blue squares, which are the a-fucosylated version, you see greater than a tenfold enhancement of ADCC.

What's even more impressive about this technology is when you look at the LALA mutant, so LALA is supposed to abrogate just about all ADCC activity.  Now, when you make LALA in -- without fucose, you actually can see here in the -- these purple squares, the ADCC activity is actually higher than that of even the wild-type 10-1074.  And it just really kind of highlights how powerful just the removal of one simple little alpha-1,6-linked fucose residue is.  I've looked at this in 3BNC117 as well, and you can see that when there is no fucose present on 3BNC117, we have identical trimer binding.  We have identical viral neutralization.  However, once again, comparing the red to the blue, you can see around an eight- to tenfold enhancement of ADCC when you just remove who goes from 3BNC117.

So you might be asking why I'm interested in this approach seeing as it's an approach that typically is used to modify antibodies produced in vitro and I work with AAV.  So in AAV, it transduces a cell and produces my broadly-neutralizing antibodies in vivo.  So I set out to see if I could modify the AAV vector to recapitulate this in an in vivo setting, and so I -- excuse me -- I designed an AAV vector to express shRNAs against fucosyltransferase 8.  Now, this is the glycosyltransferase that add the alpha-1,6-linked fucose.  Now, by inhibiting FUT8, we hoped that the fucosyl content on antibody would be drastically reduced and that hopefully this would boost ADCC activity as well as the antiviral activity of these antibodies, you know, essentially hoping target and destroy viral reservoirs.  So we cloned these AAV vectors in a way that they contain multiple Pol III promoters, each of them independently controlling an shRNA for FUT8.

The construct design looks like this.  So this is actually the Poly A tail downstream of the antibody, and as I said, we -- each of these constructs contained multiple shRNAs, so different constructs either contained two or three shRNAs, and we even had one that just contained one.  And with varying spacer lengths, these actually turned out to be quite important.  This is multiple rounds of testing.  But these constructs were cloned downstream of the antibody and of the Poly A tail in this AAV vector, and it's just upstream of the 3 Prime ITR.

So when you make antibody -- for instance, this is for 4L6 -- with these constructs, we noticed a drastic decrease in the amount of glucose present on the secreted antibody constructs.  For instance, Constructs 5, and 6, and 7, the band is very, very faint compared to the wild-type 4L6 itself.  When you look by real-time PCR, you see about an 80-percent knockdown in the mRNA levels of fucosyltransferase 8.  And most importantly, when you're looking at constructs like 2, 5, and 6, when you actually perform an ADCD assay and compare that to antibody produced in a FUT8 knockout cell line, as you can see here in purple, the ADCC activity is almost identical to that of the AAV construct produced antibody.

Now, in collaboration with Guanping Gao at UMass, we made Constructs 6 and 7 as AAV 2 vectors, and we produced this antibody by AAV transduction and then purification of the antibody.  And that was compared to wild-type-produce 3BNC117 or FUT8 knockout cell line produced.  And what's interesting is Construct 6 here, the levels of ADCC activity are almost identical to that of the FUT8 knock cell line produced, so once again suggesting to us that if, you know, AAV transduces a cell, it makes antibody with very little to no fucose content.

Now, the amazing thing about this technology is you actually can combine it with other approaches to enhance ADCC activity.  One of the famous methods is by just doing FC mutations to enhance the binding to FCGR3A.  This is a very famous del mutant here.  Now, when combined with fucose knockdown, we saw a further enhancement of ADCC activity from around 70- to 80-fold depending on if you're looking at 10-1074 or 3BMC117, which was really exciting to us.  And this can even be further enhanced looking at asymmetrical FC mutations and generating by specific antibodies.  So Fc gamma IIIA is a little asymmetrical, so the mutations on one heavy chain should be a little bit different than the other.  And when you make it as a by-specific antibody, we're seeing a ADCC enhancement over 100-, 110-fold.  To note, even a few of these should be performing even better.  We actually believe we've maxed out our ADCC assay.  So just, once again, you know, fucose removal can be combined with other approaches, and it really does produce an amazing enhancement of ADCC.  And we think that this could really impact the viral reservoir, the residual virus that are left around in macaques.

So in summary, AAV vectors can be engineered to express shRNAs, and they can glycoengineer the antibodies of transduced cells.  These antibodies display enhanced ADCC, and this ADCC can be further enhanced by adding the Fc mutations.  So these glycoengineered AAV vectors, as well as in combination with Fc mutations, are currently being evaluated for their ability to target and destroy viral reservoirs even in macrophages.

So I just wanted to thank everybody, especially the Desrosiers Lab, my mentor, Ron Desrosiers, Danielle Gill, who's my technician in the lab.  She's amazing.  And just also Chema and Sebastian.  They are the ones that did all the beautiful work that you saw in the macaques previously.  And of course, Guangping Gao for all of his help making AAV vectors, and then, of course, the funding support.  So thank you.

MS. PETERS:  Wonderful talk so far.  My name is Rebecca Peters, and I am a graduate student in the lab of Dr. Mario Stevenson at the University of Miami as well as James Termini.  And today I'll be covering some of the ongoing work I've been doing with persistent HIV infection in monocyte-derived macrophages, or MDMs, as well as introducing a possible method of suppression for aqueous in HIV in this system.

So as we are probably all more or less aware, HIV is an incurable disease except in a few very rare circumstances involving bone marrow transplantation.  The current path forward for the infected is a lifetime of antiretroviral therapy.  We know that HIV persists in the body and physiological reservoirs, and we know that the most implicated and studied of these overall is the memory CD4 positive T cell, which is why that cell type is the specific target of many reservoir clearance strategies currently under investigation, including kick and kill and block and lock strategies.

At this this particular meeting, though, I think we're all especially conscious of the fact that myeloid cells, like macrophage, can also support an active HIV infection in vitro for at least some period of time and that the macrophages do not necessarily have the same vulnerabilities as a cell type that CD4 positive T cells do, for example, to cytotoxic CD* positive T cell killing, as an example.  And for reasons like this, research into clearance strategies for myeloid cells is, of course, necessary, although it does trail far behind research into T cell-focused strategies.

So with that in mind, the original focus of our research began with dynamics of HIV infection and MDMs because we have consistent evidence from experiments within the students in lab that HIV infection can persist for many weeks in these cells, and also that the virus eventually tends to fall into an overall state of quiescence, which can be stimulated to produce fresh HIV again, not unlike latency reactivation in T cells.

So experimentally speaking, that quiescence can be seen here as HIV titer tapers off over time.  This is a graph of extracellular HIV RNA copies per microliter where the construct I'm using has a vial envelope to confer macrophage tropism in an LAI backbone.  The diagram here explains how I derive and treat my MDMs over time as well as when I collect the timepoints detailed in the graph.  So in this longitudinal experiment, you can see that I'm looking at the HIV production of MDMs maintained at three different concentrations of human serum to start with the dotted line being 10 percent, which is considered typical, the solid line being 0.5 percent, and the dashed line being 0.1 percent human serum.

I've included tenofovir controls for all these.  For those unfamiliar, tenofovir serves as an infection control and helps account for any carryover plasmid, just to orient you.  And what we can see, first of all, here is that HIV can remain productive in MDMs for a long time.  I do have data from some donors where this persisted for at least 20 weeks, and even here we can see that HIV production does continue for upwards of 12 weeks at 10 percent serum.  And another thing we can see is that simple reduction in serum does change the dynamic of the infection somewhat.  We can see a big titer of HIV RNA is not as high, and we can also see that the -- that there's a decline to what we might consider a quiescent state, and it comes on more rapidly at a lower serum concentration.  Fairly straightforward hopefully.

The next question then would be whether, as in T cells, the use of latency-reversing agents, or LRAs, can lead to an increase in HIV production from these infected MDMs, or in similar -- simpler terms, whether the HIV can be reactivated in this system.  Now, the data I have here comes from supernatant HIV RNA groups maintained on 0.5 percent serum.  And what I've done is stimulate the infected MDMs with either the immunomodulator LPS or the HDAC inhibitor SAHA, both of which are relatively common LRAs from the literature for T cells, with some prior evidence within the Stevenson Lab that they can produce similar results in MDMs as well.

As reflected in this figure, I've consistently seen that both of these LRAs associate with increased levels of HIV supernatant RNA in MDM populations where the production of HIV supernatant RNA had fallen under 1,000 copies per microliter, in this case, at 10 weeks, which -- this is a stage we might call quiescence.  So, therefore, this is telling us that HIV in these MDMs does have the capacity to reactivate when stimulated.

So next, this entire situation caused us to further question the transcriptional activity of the quiescent HIV within ourselves, and we then wondered if manipulation of cellular pathways known to transcriptionally regulate HIV to some degree in T cells would also impact the latency or at least quiescence we were seeing in MDMs.  To investigate this, we added sustained NF-kappa B antagonism to the infected MDMs at two weeks either in the form of resveratrol, which will be shortened to "RV," or caffeic acid, which will be shortened to "CA" for our purposes, and tenofovir for the future is going to be shortened as "TMV," which I already talked about.

So as we can see here, although the degree may not seem large, NF-kappa B antagonism did and does consistently result in a slightly quicker descent into quiescence for the HIV, and certainly this translates to a more tractable experimental setup as well.  More importantly, though, we found that the use of NF-kappa B antagonism seemed to cause something intriguing.  So when LRAs were again used to stimulate MDMs hosting quiescent-stage HIV, reactivation did not occur in either group that had been exposed to NF-kappa B antagonism, as we can see here, where again LPS and SAHA were used to reactivate and the measured output is extracellular HIV RNA from the supernatant.

We found this very interesting because, despite common issues with donor and batch variance in primary monocyte-derived macrophages in general, this was something we've been seeing very consistently.  And it translates across two separate compounds, caffeic acid and resveratrol, caffeic acid being a compound found in beeswax, and resveratrol being a compound found often in red wines.  And both are being used to antagonize NF-kappa B in particular.

Now, at this point, a key question, comment, or concern might be that the manipulations I've been talking about, so the serum reduction to 0.5 percent and, in particular, the addition of NF-kappa B antagonism, is creating severe functional handicaps for the MDMs because, after all, NF-kappa B is a master regulator that directly and indirectly impacts hundreds of genes.  And even apart from that, the -- in some ways we're also a far cry from fully defining what exactly human serum provides MDMs in culture, though, of course, we know it to be necessary and important and best used in concentrations of about 10 percent.  Changing that would almost certainly deprive the cells of important factors, whether we can define them yet in total or not.  So to follow up on these possible concerns, I did a number of assays to check survivorship, viability, and functionality, and I will cover a few of those today.

For one, I tested the ability of the MDMs to perform the major macrophage function of phagocytosis, so that is depicted here.  To start with the different serum levels, so 10 percent, .5 percent, and 0.1 percent human serum, are not included as lines on the graph to the right for reasons of clarity.  But the representative picture still clearly give the idea -- give an idea of the dramatic differences to phagocytic capacity that seem to be clearly caused by reduction in serum with 10 percent performing best clearly and .1 percent serum performing very poorly in comparison.  And for clarity, cytochalasin D in the bottom right corner is a phagocytosis inhibitor which is used as a control.

In contrast, the graph on the right catalogs infected and uninfected MDMs which either have or have not undergone NF-kappa B antagonism.  And overall, we see that the NF-kappa B antagonism does not seem to translate into major changes in phagocytic capacity when using either caffeic acid or Resveratrol as antagonists, though they may trend and a little bit differently from one another.  Likewise, we can see a similar trend when we consider LDH data where we keep in mind that LDH serves as a fairly reliable indicator of cell death.  Through this time course, we can see that regardless of infection status, NF-kappa B antagonism does not result in a statistically significant difference in survivorship over time.  We can see that infected untreated MDMs may have a slightly better survivorship in general than other groups, which could have a virological explanation, but that's not pertinent to whether or not NF-kappa B antagonism results in a great deal of die-off, which we can see here that it does not.

So that's all well and good so far.  We see the effects of the serum reduction, the effects of the antagonism, that sustained antagonism seems to prevent the expected reactivation of the HIV.  We've addressed some concerns about the possible negative effects of these manipulations, but all this begs the question, are the effects of the NF-kappa B antagonism reversible.  In answering this question, I expanded to include a few more LRAs as well.  So I tried reactivating with the immunomodulator TNF-alpha, the PKC agonist PMA, and the mitogen PHA in addition to LPS and SAHA, which I used before.  You can see that pictured here.

Now, this data includes groups where the NF-kappa B antagonism was stopped, or you'll see the word "interrupted" in the figure.  And what we can see is that in comparison to the untreated cluster of groups on the very left which responded variously to the different LRAs, but certainly, again, responded robustly to LPS in SAHA, we see that in comparison, neither the NF-kappa B antagonized groups nor the groups with interrupted NF-kappa B antagonism showed any reactivation, which does lend support to the idea that any possible transcriptional suppression in the NF-kappa B antagonized groups does last after the antagonism is removed.

And when we look more closely at a snapshot of NF-kappa B activation under these conditions, in here what we're looking at is a ratio of phosphorylated and NF-Kappa B, which we consider to be activated over the total amount of NF-kappa B present.  But when we do look at these NF-kappa B levels, we see something interesting as well.  So we see that the NF-kappa B is responding normally to LPS challenge in the untreated group at the left as expected, but that there's also what we might consider a normal and reactive NF-kappa B response in those NF-kappa B antagonized groups where the antagonism has been stopped.  So there's a recovery in these MDMs as far as their NF-kappa B response, yes, and yet the suppression of HIV reactivation continues as we saw in the previous slide.

And when we look at two key cytokines across the same time span even without interruption of antagonism, something else emerges that is fairly striking.  To introduce this, IL-10 and CXCL2 are both cytokines.  They're variously controlled by NF-kappa B, but both of them have some dependence or interrelation with the pathway.  But we do see here that long-term NF-kappa B antagonism has an extremely clear effect on the levels of IL-10 at the time of reactivation, and that this contrasts with what we see in CXCL2 under the same conditions, which is that the levels of CXCLT were largely unaffected.  The pathways governing these two cytokines are interrelated, but further investigation into the ways that they differ could help to shed light on the mechanism of quiescent HIV suppression by NF-kappa B antagonism.  So obviously something is happening here with all of this.

I've established the effect of -- I have established and explored the effects of serum reduction in NF-kappa B antagonism just as a recap, and I've demonstrated that NF-kappa B antagonism appears to suppress typical reactivation in quiescent HIV in MDMs.  I've touched upon possible deleterious effects of the manipulations, confirmed that the suppression of reactivation continues past interruption of NF-kappa B antagonism, and accounted for NF-kappa B and cytokine activity around the time for reactivation.  A final component to this story that may be obvious at this point is to look further into the activity of the NF-kappa B pathway ideally to the level of epigenetics, perhaps to try to confirm or deny that the suppressed effect is the result of perhaps an epigenetic modification, which, if that's true, that could be very difficult to reverse, and it could be grounds for future therapeutic research.  Both of these compounds are safe for human consumption and commonly used as supplements already for various things.

So I'd like to thank and acknowledge the Stevenson Lab of course for the resources, protocols, and backgrounds that have been provided for this continued research, as well as the University of Miami and CFAR for their resources, and the Committee for guidance.  Thank you, everyone here, for your attention.  And next we'll be hearing from Haitao Hu from the University of Texas.  Thank you.

DR. HU:  Cool.  All right.  First of all, thanks for the opportunity to present our work.  So today I'm going to talk about the modulation of BRD4 to induce HIV epigenetic suppression in microglial cells and the myeloid cells.

So in our lab, we have been interested in some host epigenetic mechanisms that regulate HIV transcription and latency, and a long-term goal is to hopefully identify some approaches that can target this mechanism to either disrupt or reinforce HIV latency.  So we particularly -- in recent years, we particularly investigate a protein called BRD4, which is shown here, that is a bromodomain-containing protein that belongs to the BET protein family.

So based on the studies in the cancer field, we know that the function of the BRD4 is actually really diverse and versatile.  But basic function or activity of this protein, it sort of serve as an epigenetic reader, which means it utilizes bromodomain to selectively interact with the acetylated histone of chromatin and to serve as a scaffold platform.  Meanwhile, the protein utilize additional functional domains to recruit or interact with a wide variety of the pattern of proteins as shown here, including the transcription factors, chromatin modifiers, and remodelers, etcetera, to bring it to the gene promoter region to regulate the gene transcription.

So a role of BRD4 in HIV transcriptional regulation was introduced roughly 16 years ago with some divergent results being reported.  Some early studies showed that BRD4 can promote HIV transcription, but more recently, some studies -- there a couple -- a number of them show that BRD4 can actually suppress HIV transcription and promote latency.  So consistent with some of these more recent studies, a Pan-BET inhibitor, which is called JQ1, was identified roughly 11 years ago that later was tested in the HIV field.  It was shown that the compound can inhibit or modulate the BRD4 or the BET proteins non-selectively in a manner so that HIV transcription can be activated.  So a potential or proposed mechanism of JQ1 is that it disrupts the competition of the BRD4 with HIV Tat protein for cellular CDK9 that leads to the enhanced HIV transcription elongation.

So a few years ago, using structure guided drug design and J-Lat base to screen, our collaborative groups identified a new molecule as shown here called ZL0580 and a few analogues that are interestingly distinct from JQ1, but induce HIV suppression.  And here are some of the representative data, and we show that compared to the JQ1.  And our compounds can actually induce HIV suppression in a dose-dependent matter and based on both GFP report of protein expression in the J-Lat cells as well as based on the viral RNA -- 3LTRNA in the cells.  And this is the kinetic data, so.

And we verified that the effect of these compounds on HIV was actually largely mediated by BRD4 because we use the CRISPR/Cas9 to knock out BRD4 in the J-Lat, and we also knock out BRD2 in the J-Lat just as an internal BET protein control.  And what we observed later on was if you -- if you knock out the BRD4 protein expression but not actually the BRD2, you largely abrogated the effects of these compounds in HIV transcription based on both GFP protein as well as the viral RNA.

So, you know, our ongoing studies, we did RNA seq and tried to understand the global impacts of these two compounds in J-Lat cells on transcription.  And what we found was the data looks very consistent with HIV results in that these two compound basically induce sort of like opposing, or I would say distinct transcriptional profiles as you can see summarized here.  For those upregulated in JQ1, they are downregulated by the 0580, and vice versa for those downregulated by JQ1, they're upregulated by these compounds.  And the other thing is if we knock out the BRD4, the effects of these -- our compounds on the cellular gene transcription.  This is the list of the selected genes.  Based on the RNA seq data on PCR, the effect was lost.  So for the conforming, at least it's fully or largely dependent of this BRD4 protein.

So next we tested these compounds of HIV in microglia and myeloid cells.  And through collaboration with Dr. Jonathan Kind's group at Case Western, they kindly provided, as we said, immortalized microglial cell line called here the HC69, which was engineered to contain HIV provirus as well as an EGFP as a reporter to allow the subsequent analysis.  And then we -- using these three online, we find a very similar pattern in a way that, if we look at the JQ1, it promotes the HIV expression in the microglia based on the GFP protein, but our compounds really substantially reduced the GFP expression in these microglial cells.  And in this case, we use the H micromolar in each condition for the compound.  And other than the GFP, we also quantified the viral RNA in these cases, the early multi-spliced viral RNA, as well as the GFP mRNA, and you can see that RNA expression was substantially reduced or suppressed.  This is just a one-time dataset in the 24 hours post-treatment.

So we also observed that the effects of these compounds on HIV transcription in the microglia is actually pretty durable, and we did two system.  One is to activate microglia which were stimulated by TNF-alpha.  We also used the resting microglia just in the absence of any activation.

So for the activated microglia, we followed two treatment models.  One is just the single treatment at day zero, and we also did repeated treatments of the cells with our compound at day zero, three, and seven, and then we monitored the viral gene expression at various timepoint post-treatment.  And we can see the single treatment leads to viral suppression.  We can see that up to day 21 post-treatment, and it looks like this effect gradually lost around day 28.  And in this repeated treatment model, we can further see the extended suppression up to day 41 post-treatment.  And we are still continuing to monitor the effect -- the durable effects of these compounds on suppression in these microglial cells.  And we see a very similar pattern here in the resting condition as well based on -- we can see up to day 14 post-treatment very strong suppression.

And so to explore some potential epigenetic effects of this compound, so we did this experiment.  So the microglial cells were pretreated with this compound, single dose.  And then -- so then we monitored the cells various days, on day seven, 14, and 21.  And at this various time point post-treatment, the cells sales were reactivated with TNF-alpha, tried to  -- tried to activate the latent HIV, and then we look at the viral gene expression.

And what we found here was, based on both early multi-spliced mRNA and the GFP reporter RNA, we can see, compared to the Molt control group, which means there was no pretreatment, the pretreatment of the cells with the compound leads to -- sort of render these cells more resistant to the subsequent TNF-alpha or the latent HIV reactivation based on this viral or the GFP transcription, indicating that the compounds may exert some epigenetic effect on HIV in the cells.  For the sake of time I didn't present all this cellular toxicity data, et cetera, and it's just -- it just ensure -- so the observed data so far, they are not related due to the -- simply due to the cellular toxicity.  And we also measure these compounds, an addition of two monocytic cell lines, including the U1 and OM10.1, and we see a similar or consistent result.  It's suppressive to HIV transcription, both cell line as well.  We didn't do extensive analysis of these two cell lines.  It's just a one-time dataset point.  One timepoint data is 24 hours of treatment.

We did some mechanistic explorations, and as I mentioned earlier, one potential mechanism of action or activity of the BRD4 is actually to compete with the viral Tat for cellular p-TEFb CDK9 that leads to the inhibition of HIV transcription based on some previous models.  So what we did here was we did a co-IP analysis to measure the binding of Tat with CDK9 in both microglial cells and U1 cells, and we did both activated condition as well as the resting condition.  I call it "resting condition" which means there was no stimulation at all.  And we find a very consistent pattern, and that is, compared to the non-treatments on the no-compound control, this compound appeared to really reduce the binding of Tat to CDK9 across all these conditions.  And we observed similar data in the J-Lat cells as well.

And we also looked at the chromatin structure and the DNA accessibility in the HIV LTR region using this assay called high-resolution MNase nucleosomal mapping analysis.  And, again, for the sake of time, I won't go through the details for this assay, but as you see, the strongest signal here, which means that you have reduced DNS accessibility with repressive chromatin structure because they are protected from the DNS digestion.  And we can see, compared to the low-treatment control, the compounds appear to induce a more repressive nucleosomal structure at the LTR region.

So lastly, in the last part, I'm going to use a few minutes to briefly talk about another ongoing study related to these which hasn't been published yet in terms of the BRD4 isoforms and HIV transcription in microglia.  So since the discovery of this molecule a few years ago, we have been trying to really understand why the two molecules appear to target the same protein, or at least the same protein family, but induce totally distinct functional outcomes.  And last year we saw a very interesting paper published by Dr. Charles' group from UT Southwestern.  They demonstrated the opposing function of the BRD4 isoforms in breast cancer.  So the BRD4 has multiple isoforms, including the long isoform and two additional short isoforms, S(a) and S(b).  And basically, they found these isoforms play opposing roles in regulating the tumor gene expression in the breast cancer model.

And then we went back to our data, especially the CRISPR/Cas9, the J-Lat cell, and we find the CRISPR/Cas9 was fairly efficient in knocking out the BDR4 protein expression.  But later on we found an issue that this is the BRD4 long isoform based on the size as well as the antibody we used, which was the BRD4-long isoform-specific antibody that recognized the CDT region of the protein.  Then we repeated this experiment and using a Pan-BRD4 antibody that recognized the bromodomain of the protein that is shared by all these isoforms.  And what we found, our CRISPR/Cas9 wasn't isoform specific.  It knock out BRD4 long isoform as well as the short isoform S(a).  For the S(b), there was also a pretty significant reduction, it wasn't -- even though completely knocked out.

So then we reached out to the group at UT Southwestern and started a collaboration.  They provided us with the isoform-specific siRNA, and it actually worked pretty well.  This is the data for the microglial cells that were transfected with these different isoform-specific siRNA, including the long S(a) and S(b).  So you can see the top panel shows the antibody or shows the BRD4 for long protein.  This is the S(a) protein.  This is the S(b) protein.

And we can see it induced pretty good specific knockdown.  Especially the S(a) was almost completely out, the long isoform substantial reduction.  S(b) compared to S(a), the long isoform, it wasn't that strong, but there was certainly some reduction.  And then we quantified the HIV transcription in these microglial cells following siRNA transfection.  This is the data for the 72 hours post-transfection based on both GFP and early multi-spliced RNA.  And there's an interesting pattern where we see the long S(a) isoform siRNA knocked down, induced in-house to HIV transcription.  And in contrast, if we knock down the S(b), it looks -- it reduced HIV transcription.

So this data remains preliminary and we are trying to repeating everything and extend it to the J-Lat cells as well.  So but the data indicate that these isoforms may play distinct or even opposing roles in HIV transcription as well as in the microglia.

So to quickly summarize, through a structure-guided design, we identified a new small molecule that induce HIV suppression through BRD4, and the suppression appears to be pretty durable.  Preliminary data support that isoforms may have opposing functions in regulating HIV transcription in the microglia.

So I would like to thank the current and the former lab members who have really contributed to this study as well as my collaborators at UTMB and several groups outside of the UTMB as the external collaborators, as well as the funding from NIMH.  So thank you for your attention.  I'll stop here and I'll pass this over to Dr. Howard Gendelman.  Thank you.

DR. GENDELMAN:  Okay.  Thank you very much to the organizers for allowing me to speak today on CRISPR Exonic Disruption.  I'm Dr. Howard Gendelman.  I'm the chair of pharmacology and experimental neuroscience, and I'd like to spend just a moment and talk about the history of the myeloid HIV interactions.

The actual first conference began almost 35 years ago at the Rockefeller Institute, and I think what's very pertinent about that first conference on AIDS or HIV and the macrophage is that the singular events or what led to what we know about the pathobiology of myeloid HIV interactions actually are still true today after 35 years.  We know that the macrophage plays many roles in disease.  It clearly affects by altering the metabolic state of the brain in releasing factors, proinflammatory cytokines and chemokines that alter brain homeostasis, and that there is a multitude of end organ diseases that are affected by this altered activation of myeloid cells as well as the dissemination of the virus within an infected human host and others during the course or during transmission of HIV and AIDS.

SPEAKER:  And, Howard, just let me know when you would like for the slides to advance.

DR. GENDELMAN:  Okay.  You can go to the second slide, the macrophages in HIV over 35 years.

So we can now move to the macrophages in HIV and the 35-year journey, and I'd like to pause and salute Jan Orenstein, who really was the originator or orchestrator of the macrophage myeloid interactions, and shown that HIV can be harbored in mononuclear phagocytes through intravascular accumulation of virions.  We now know it's a transmission of the cytoplasmic membrane, and really laid the foundation of this work that we all enjoy today.  He died just a few weeks ago.

What I will present today is part of a multidisciplinary program that really begins with medicinal and polymer chemistry to synthesize nanoformulations and lipid nanoparticles that can be used to specifically target latent reservoirs of HIV in CD4 T cells as well as mononuclear phagocytes.  We've developed, in conjunction and collaboration with Benson Edagwa in our program, an alteration of the pharmacokinetic parameters of long-acting antiretroviral drugs, and we have made libraries of year-long medicines that can be given after a single injection and target sites of ongoing restrictive viral growth in conjunction with combination therapy.

We've also developed, in conjunction with Bhavesh Kevadiya, who presented yesterday, the notion of nanotheranostics where we can label these antiviral therapies and monitor their distributions in latent reservoirs in lymphoid system, in gut, and bone marrow, and in brain.  And in early collaboration with Kamel Khalili, who's chairing this session today, we've developed the systems to actually investigate not only the perpetration of HIV in these reservoirs, but to focus on excision of latent viral DNA in target cells.

Next slide.

So today we're going to talk about the CRISPR HIV guide RNAs and design and biology.  And there's really two takeaway messages from this.

Next slide.

And the two takeaway messages is how do we design our CRISPR guide RNAs that will detect most HIV proviral DNAs that can be part -- or part of an infection or transmission in an infected human host.  Now, we've monitored and aggressively sought guide RNAs that will look at consensus sequences and that will be overlapping with both structural and regulatory proteins of HIV, looking at both entropy and conservation.  And we located the first exon and the second exon of Tat, made these guide RNAs, tested them and facilitated their transfer in Cas/CRISPR systems to make RNPs, mRNAs, and DNAs in conjunction with our abilities to excise latent HIV within a human -- a human host ultimately, but within experimental animals to date.

Next slide.

So when we look at this, we have to develop a testing system.  This is a complex slide, but I want to call your attention to two panels.  Panel A with our Tat DNA, when we've looked at a number of altered consensus sequences of HIV, tested them in a number of molecular clones of divergent sequences, and then asked the simple question, what is the best combination of guides that we can produce that will ultimately excise in some proportion -- in some significant proportion -- of proviral DNA.  And not only the proportion of proviral DNA, but actually list beyond a single sequence -- in this case NL4-3 -- to look at consensus among many proviral DNA molecular clones.  And then take and on-target cleavage and look at the single-guide RNAs versus the dual-guide RNAs for their intel rate percentage to generate the most and best sequence for optimal delivery and excision of proviral HIV.

Next slide.

So how do we test this system?  Well, really, there's a couple important variables that we want to look at.  One is the model system, in this this case using a latent U1, or ACH2, or LIT-1 system where we can actually infect these latently-infected cells using lentiviral constructs, transfection, electroporation.  And what we'll lead into is our development of lipid nanoparticles which appear to be the most sensitive and specific delivery that will optimize the target take-on, which are lately-infected lymphocytes and monocytes circulating in blood and lymphoid target systems.

In this case, we looked at ACH2 cells that contain a single copy of HIV, developed a lentiviral recombinant construct, and then infected that lentivirus that carry the guide RNAs and the Cas9 sequence at varying multiplicities, and asked the question, can we prevent development or production of progeny virus by measuring RT activity over the course after TNF induction of a latently- and viral-infected cells line.  In this case, we were successful both at a multiplicity of 10 with our TatDE construct and a multiplicity of 1 using this lentiviral guide RNA.  But the question that we had to ask, is this something that we can translate into an affected human host.  Can we build enough lentiviral systems to use this as a carrier for our guide RNAs or would we need to look beyond this carrier modality.

Next slide.

So the question we had to ask first is not only the delivery, but the efficiency.  So we had to do cross-disciplinary work, and in this case -- and in this case we had to affect latent cells by making new recombinants, in this this case, the Tat DE, not the guide RNAs, but the modification within the proviral DNA to excise or modify those fragments to investigate the specificity of our guide RNA systems.  We also had to look at what we excised by sequencing that fragment band or that excised fragment band.  And using the altered recombinant profile DNA, as you can see in D, we were able to attenuate showing the specificity of our guide RNAs, both by infection using our lentiviral constructs and an indicator GFP positive system and through induction by measuring RT activity in the latently-infected cell that we can induce with TNF or PMA.

Next slide, please.

So the question that we had to pose now that we had a very sensitive and specific guide RNA.  We can show excision through sequence analysis.  We can actually delete through molecular recombinations of deletion of that specific area of the proviral DNA within the subgenomic fragment of HIV.  So we knew we had a sensitive and specific guide that actually was more efficient than other fragments and subgenomic fragments that we had developed in parallel.  So how do we deliver it?  What do we know about CRISPR biology, and is it all in the delivery to find those latently-infected cells for ultimate excision of the proviral DNA?

Next slide.

So what we've done for the first time in delivery within our Tat DNA plasmids, to start off in making lipid nanoparticles that were significant in their orchestration, development, and composition.  And this is something that we had done in parallel for long-acting antiretroviral therapies and developing of long-acting prodrugs.  We had to look at the PEG composition, zwitterionic lipids, the cationic lipids, the composition of cholesterol, and the ratios of each.  Why is this so important?  Because the lipophilicity could maximize entry into HI latently-infected cells.  So the CRISPR system now was integrated and encased in a depot within the lipid nanoparticle with our Tat DNA constructs, in conjunction with our Cas9 endonuclease.

We can see our characterization both by transmission, electron microscopy, and analysis of the particle.  So we were able to optimize the size, the shape, the stability of what we were producing, and the lipophilicity so we can maximize entry into cells.  And when we first tested this, either by electroporation or transduction, or transfection, we were able to once again use the lipid nanoparticles as we did with the lentiviral constructs, as we did with clear transfection efficiency, and show a basis and a significant dose response in preventing the production of progeny RT activity, showing in G the excision of the anticipated fragment that was ultimately sequenced, and then showing where these guide RNAs were developed within the locus of Tat DE in H.

Next slide.

But we had to move this further.  We had to move this further.  Why?  Because we were still struggling with making the bridge, the ultimate bridge, between the transduction, transfection, and the lentiviral constructs in something that we can ultimately move to larger animals and then to humans.  And what we had to do is develop these lipid nanoparticles so we can use them as a natural entry.

So let me take a step back.  What are we trying to do here?  Ultimately we're trying to increase the efficiency of the entry of the lipid nanoparticles into target cells.  How are we doing that?  Using our prior experiences with formulation efficiency and our long-acting antiretroviral drugs, we were able to devise a composition of the lipid nanoparticles based on the cholesterol, and the ionizable lipids, and the DSPC, and the composition to affect ultimate improvements in lipophilicity for entry through the cell membrane and in conjunction using more established microfluidic technology.

So within this lipid nanoparticle, we had the guide RNA and the Cas9, but before we can use this in any efficacious assay, we had to test how well these particles were entering.  And rather than use the CRISPR system, we used Firefly and GFP as our indicator modality.  And we were able to show, as you can see in C, D, and E, using J-Lat and U1 and controls in the LRU, or by flow cytometry, or by induction, that we were able to induce, after exposure of these lipid nanoparticles, to the GFP or the flu constructs the levels of entry of these lipid nanoparticle recombinant DNAs and messenger RNAs almost to the level of 100 percent of these latently-infected cells cultivated in vitro and in G.  And the lipid treated and the untreated, you can actually see the efficiency and specificity of the numbers or percentage of GFP-positive cells.

Next slide, please.

But the ultimate delivery of the system really had to evolve in looking both by transfection efficiency and the induction of what we were seeing vis-à-vis the proviral DNA.  So in these experiments, what we wanted to recapitulate here is a natural use of our CRISPR/Cas guide RNA, Cas9, efficiency of excision of the proviral DNA in a biologically-relevant system, and first testing this -- first testing this using transfection and electroporation, and ultimately as part of a natural exposure.

Three things that we saw that are important.  Number one is that the use of these lipid nanoparticles using R and P electroporated cells were able to demonstrate that these cells remain viable, number one.  Number two --

DR. KHALILI:  Howard, I'm sorry, your time is up.  Can you wrap up?

DR. GENDELMAN:  I have one more slide and then I'm finished.  And then the excision band showing 100 percent elimination of the proviral DNA.

Next slide.

So these are the last two slides demonstrating that we were able to get these guide RNAs into the nucleus --

Next slide.

-- by looking at trafficking.

Next slide.

And the last slide before the summary is demonstrating through natural infection or natural exposure, greater than 92 percent efficiency of excision in E and F and lack prevention of induction of the J-Lat integrated provirus DNA cells.  In summary -- in summary --

Next slide.

-- what I really wanted to put together is a statement by Henry Ford, is that when we look at these multi-multiple modalities of bringing people together in collaboration to perform technologies that maybe we're not so familiar with, that we can do things more than we think we can do, not by ourselves, by our working in a greater collaboration.

And the next slide and the final slide is that this is a city, it's a group of important people, and I mentioned -- unfortunately I couldn't show my slides, but also built on the work and the collaboration of Kamel Khalili that we've extended consequently, and the work on the chemistry with Benson Edagwa, the theranostics of Bhavesh Kevadiya, and a group of teams in humanized mice and monkeys in Sid Byrareddy and Larisa Poluektova and Santhi Gorantia, who've all worked together to move, I think, the part and progression to a finality.

I'd like to turn this work over now to Rafal Kaminski who will deliver the talk in this series.

DR. KAMINSKI:  So, again, good morning, everyone.  Thank you very much, Organizers, for this opportunity.  My name is Rafal Kaminski.  I am assistant professor at Temple University in Philadelphia, and I would like to share with you some of my data from ALCAM project when I used CRISPR technology to target outcome in myeloid cells.

So let me introduce the ALCAM.  "ALCAM" stands for activated leukocyte cell adhesion molecule.  Another name, it is CD166.  This is type I transmembrane glycoprotein from immunoglobulin family, and basically it is responsible for two types -- for cell-to-cell adhesion.  And we have two types of these adhesion events:  homotypic ALCAM-ALCAM or heterotypic ALCAM-CD6.  CD6 is expressed mostly on T cells, so this interaction will take place between lymphocyte -- T lymphocytes and outcome-expressing cells.

And ALCAM is pretty -- it's pretty ubiquitously expressed in almost every tissue.  But what is important, it is induced on activated immunocells, lymphocytes, and myeloid cells, and also inactivated endothelium.  And if we -- it comes to endothelium, there is part of ALCAM which is constitutively expressed there because it is building these non-canonical junctions between endothelial cells.  So it is important for maintenance of endothelial barriers.

It comes to T cells, there are some publications showing that outcome is important for T cell aggregation, and we have both interactions there with ALCAM-ALCAM and ALCAM-CD6 interactions.  And also it was shown to be involved in stabilizing with immunological signup during antigen presenting, so here we have interaction between T cell and myeloid cells, dendritic or macrophage.  The most likely outcome would be involved also in interactions between the myeloid cells.

So if we put these two information together, the presence of ALCAM in the junctions, endothelial junctions, and overexpression of ALCAM in activated immunocells, so no wonder that the outcome is important for this extravasation of activated immunocells through the endothelium to the tissue.  And it was nicely demonstrated for lymphocytes of B cells and T cells, also for monocytes, and recently also for activated dendritic cells.  And this outcome is important for this later stages of research process, actual transmigration of diapedeses.  So it seems like if you have a higher level of outcome, it is easier for the cells to pass through these endothelial junctions.

And to put HIV in the picture, I wanted to highlight two publications, so one from Dr. Berman group when they show that in HIV-infected patients, the monocytes, they have higher level of outcome expression compared to uninfected controls, and especially with intermediate CD14/CD16 monocytes, which have highly migratory phenotype.  They have high level of outcome, and by blocking this outcome with antibody, you can block the transmigration of the cells to the blood-brain barrier.  And another publication from Dr. Walker's group from MIT.  They identify ALCAM as host-dependency factors for T cells and showed critical role of ALCAM in aggregation of T cells and cell-to-cell virus transmission.  And this most likely, again, would be made for interactions between myeloid cells and myeloid and T cells.  So our R21 was based on this information.

So why targeting -- so targeting ALCAM should affect two very important virus dissemination processes.  One of them is this transendothelial migration of HIV-infected immunocells, leukocytes, and another one is with cell-to-cell virus transmission, and this should be important for transmission from myeloids to other myeloids, myeloid to T cell, T cell to myeloid cell, and between the T cells.  And there is also this effect on this -- stabilizing this immunological sign up, so most likely, this will also affect antigen-driven proliferation of latently-infected T cells.  So by targeting ALCAM should disrupt these processes and should -- that should result in reduction of the number of infected cells and the dissemination to the tissues.  And that's a result that should lead to -- limit the size and tissue distribution of viral reservoir.

So to target ALCAM, I used CRISPR/Cas9 technology.  We pioneered and developed this multiplex CRISPR approach when we have two guide RNAs targeting our target sequence and cleavage-adverse target sites will lead to excision of this intervening sequence and big deletion.  So I designed two guide RNAs, which one -- Exon 1 and Interim 1 of ALCAM gene.  In Interim 1, we have the START codon and signaling peptide, so successful deletion of this region would lead to removal of START codon and complete knockout of the ALCAM expression in treated cells.

And this excision can be easily demonstrated by PCR genotyping, as you can see here, for monocytic cell 99527.  So in control clones, we can amplify this Exon 1/Interim 1 region, and this outcome knockout clones receive truncated amplicon, which is the one will be 1,185 base by deletion without this ALCAM START codon.  And similar cell line I developed for endothelial hcMEC/D3 cell line.  This was confirmed by sequence.  And then, of course, the next step is to check the expression of ALCAM.  So here is RNA expression and cell surface expression.  The you can see the entity, I'll say, have much higher level of ALCAM compared to this e9527 monocytic cell line.  And these red dots are the knockout cells, so it was completely knocked out in these monastic cells and also drastically reduced in the endothelial cells and was confirmed by immunolabeling flow cytometry.  So these are control clones.  These are knockout clones.

So we move to the functional assays, so, first, the simple adhesion assay.  So monocytes were labeled fluorescently with CSFE and plated on monolayer of endothelial cells, and then after 30 minutes we checked how many monocytes stayed attached to endothelium, and the results you can see here.  So when we use wild-type endothelium, we can see the decrease in adhesion of ALCAM knockout monocytes compared to controls.  When we use this ALCAM knockout endothelium, we could see already reduction even using wild-type monocytes, and there was no additional reduction using these ALCAM knockout monocytes.

And this was also confirmed by another assay, transendothelial resistance assay.  So here with collaboration with Dr. Slava Rom from Temple.  So as you can see here, upon plating with monocytes, the transendothelial resistance drops.  This red line shows wild-type monocytic cells, and then eventually it recovers to the Basal level.  If you use these ALCAM knockout cells, the drop is much smaller and the cells and the resistance recovers very quickly within four hours, so with also smaller -- lower attachment of this ALCAM knockout cell to endothelium.

So when I move to this primary monocytes, so here, as a delivery we use a AAV6 system, so we use monocytes from three different donors.  Here, the expression of CRISPR/Cas9 components were verified by reverse transcription pictures.  So you can see Cas9 mRNA and guide RNAs expression in these AAV6-treated cells, which led to the decrease in ALCAM mRNA and surface expression of ALCAM.  And again, we noticed reduction of adhesion with CRISPR-treated -- CRISPR-outcome treated monocytes to endothelium.  And also we perform transmigration also using mCP1 psychomotor attractant.  And again, we saw great reduction of transmigrationability of these primary AAV6 CRISPR-treated cells comparing to control.

So when -- we wanted to add HIV to the picture, so here we changed the strategy, and I used these immunoclone protein complexes, electroporation with RNP-CRISPR as a delivery method.  So primary monocytes are isolated day one, and they are immediately electroplated with this control RNP-CRISPR containing only Cas9 or this experimental one containing Cas9 plus ALCAM guide RNAs.  Next day the cells are infected with CCR5-tropic GFP-reported virus, and after two days we verified the CRISPR editing and performed some functional assays.  So after these two days, our infection level was around six to eight percent by GFP reported.

And so again, here you can see this PCR genotyping verifying successful excision of ALCAM Exon 1 in these CRISPR-treated cells.  This is data from six donors, so here we have -- first two are uninfected and then infected for other six donors.  And this led to this decrease of ALCAM -- surface ALCAM expression, which you can see here, for uninfected and infected cells.  And this was the reason for lower transmigration activity of that cells, so it was not so significant for uninfected cells, but it was clearly significant for infected cells.  And here you can also appreciate because it shows that HIV-infected monocytes, they transmigrate more when uninfected.

So to demonstrate this critical role of ALCAM in cell-to-cell virus transmission, we performed with co-culture assays.  So we have infected donor cells and uninfected acceptor cells, so we mix them in a ratio one-to-one using these different conditions and let them grow for six days.  So at the end, we have 30 to 50 percent infection-positive cells -- GFP-positive cells.  So as you can see here, if we use donor cells and control cells -- donor cells, control-infected cells, and accepter cells, control-uninfected cells, the infection after six days was around 50 percent.  If we -- if our acceptor cells didn't have ALCAM when -- there was a little bit drop of infection.  If we used outcome-negative donor cells, it went down farther.  And finally, if both donor and acceptor cells were treated with CRISPR and ALCAM RNPs, we have the biggest drop.  So this indicates that this ALCAM expression on these myeloid cells is important for the virus spread between uninfected and infected cells, and that the -- we combined those two assays, so transmigration plus this co-culture assay.

So here on the bottom chamber, we played with monocyte-derived macrophages which were stained for red, so we can see one here.  And in the top well, you have these endothelial cells and HIV-infected cells are green, so they were protected with control antibody or anti-ALCAM antibody.  They were let -- we let them transmigrate for 16 hours when the top chamber was removed and cultures were left for another six days.  And these double positive red GFP-positive cells were scored by flow cytometry, and the result is here.  So you can see the cell results from three donors.  These are control IG treated and anti-ALCAM treated.  So you can see that we saw the reduction of viral infection spread in all three donors, and two of them was around 40 percent.  In one, very little, just four percent.  So there is donor-related variability.

So at the end, of course we would like to CRISPR and knockout of ALCAM to be present only in HIV-infected cells.  So just I wanted to show you we developed this HIV expression-dependent CRISPR system.  So we screened different mutants of HIV promoter, long terminal repeats, and identified with minus-80 plus 66 as a minimal one, which we still thought inducible and have some promoter activity, and we used this to drive Cas9 expression.  So if you look at uninfected versus infected primary monocytes, you can see that, upon infection, Cas9 expression is induced.  This is without expression cells.  And you can also see clearly that excision of Exon 1 of ALCAM is induced in these HIV-infected cells, again, showing this inducibility and Tat dependency of this CRISPR expression and CRISPR activity in the cells.

So to summarize, we are showing the disruption of ALCAM in primary monocytes resulting in reduction of the adhesion of transmigration abilities, and also reduce this myeloid cell-to-myeloid cell infection spread in coculture assays.  And we also developed this HIV expression-dependent CRISPR system.  So next step would be validation of these findings in animal models

And here I would like to acknowledge my collaborators from Temple, Dr. Burdo and Dr. Rom.  And also from -- moving forward with humanized mice model, I would like to acknowledge my future collaborators, Dr. Prasanta Dash and Santhi Gorantia from UNMC.  And I hope that one day we will move also to monkey model and we start -- maybe we'll get this pilot grant with Dr. Andrew MacLean.  Of course I want to thank NIMH for the funding and thank my students who were rotating in my lab, especially, Francesca and Konrad, who are with me for one year.  And thank you very much, and I will take questions.  Thank you.  So I turn it over to Dr. Khalili, our moderator.

DR. KHALILI:  Thank you very much.  At this point, I would like to invite all the speakers basically come to the camera so we can start the question and answer and start answering questions.  Questions for Dr. Morgan was that, "Have you tried other LRA to reactivate HIV macrophage?"

DR. BOMSEL:  Yes, indeed.  We have tried interferon, and it didn't work, and other classical T cell RNAs that didn't reduce any viruses.

DR. KHALILI:  Okay.  Another question comes to you is about the -- about the "How do you distinguish if this structure in the macrophage are VCC but not endosomes, and how do you specifically identify that compartment is VCC?"

DR. BOMSEL:  So what has -- one marker is to check for the presence of tetraspanins, such as CD63 or CD81 or CD9, and indeed we have found the co-localization with CD63.  So it looks like bona fide VCCs.

DR. KHALILI:  Okay.  For doctor -- for James is, "How well do you -- how well do the antibodies generate the brain and all the -- all the -- all the specific modification that can enhance crossing the blood-brain barrier?"

DR. TERMINI:  Yeah, that's a very good question, and I don't know exactly the answer to that.  I do know that -- so a colleague of mine in the Desrosiers Labs is actually looking at the administration of AAV, you know, in the neural tract itself to get expression and elimination in the -- of course across the blood-brain barrier.  But I can't -- I can't speak to how well these AAV-delivered antibodies will cross.  I don't know that.

DR. KHALILI:  Very good.  Jim, one more question from Dr. Brown --

DR. TERMINI:  Mm-hmm.

DR. KHALILI:  -- is that, "Is the long-lived retention of a high antibody related to the presence of viral antigen?"

DR. TERMINI:  So what happens on the AAV transduce of the cell is it actually stays more episomally in the -- in the cell itself, and muscle cells are very, very long lived and will continuously express.  So to my knowledge, the only real genes that are being expressed are the antibody within then my construct itself, so that's the only thing really between the ITR that's being delivered.  So it's not really like a lot of other viruses that are going to have other viral genes for survival and, you know, wreak havoc.

DR. KHALILI:  So the -- one more question from you -- for --


DR. KHALILI:  -- before I move to the next speaker is that the -- "what are the effect of shutting down fructose during the glycosylation in proteins on -- other than antibodies, basically the --

(Cross talking.)

DR. KHALILI:  -- of target.

DR. TERMINI:  Yeah.  Yeah.  Understood, yeah.  And, you know, this is something that I worried about a lot in the beginning as well.  I've made a lot of other glycosylation-modified cell lines where, like, I knocked glycosylation completely, and I thought, oh, this is definitely going to be, you know, just detrimental to life.  And the funny thing is out of all the cell lines and of all the transductions I've done, it seems like the cells are perfectly fine.  Is it affecting alpha-1,6-linked fucose on other proteins  Definitely.


DR. TERMINI:  Is it -- is it adding some other type of -- you know, is it adding some other type of sugars, or is it just -- or is it just the proteins work fine without the fucose?  That I can't speak to, but it -- the cells themselves seem fine and it doesn't look to be toxic in any way.

DR. KHALILI:  Okay.  All right.  Rebecca, I have a question for you.  "Have you looked at the macrophages that are treated with NF-kappa B antagonist and done reactive to see if they contain integrated DNAs?"

MS. PETERS:  Well, I haven't looked for integrated DNA specifically.  What I have looked at is total cell-associated DNA -- HIV DNA in the cellular populations, and I have found that there is a bit of a reduction in that level of cell-associated DNA when the -- when either NF-kappa B inhibitors are used, either Resveratrol or caffeic acid.

DR. KHALILI:  So the -- what about the impact of that on the other cells -- cellular genes, NF-kappa B containing cellular genes?

MS. PETERS:  I'm sorry?

DR. KHALILI:  What is the impact of the -- of the antagonist on the NF-kappa B cellular genes?  Do you see any of target or does it have effect on the cellular gene expression?

MS. PETERS:  Gene expression isn't something that I have looked at specifically.  The mechanism of action for each of the two antagonists is different.  For caffeic acid, it's not actually understood at this point what that mechanism is.  For Resveratrol, it's known that it does inhibit the IKB alpha phosphorylation, which inhibits or prevents translocation to the nucleus.  But as far as actually mapping those, I haven't looked into that.

DR. KHALILI:  Okay.  All right.  Thank you.  Dr. Hu, I have a couple question for you.  One of them is that, first of all, it was very nice talk.  I enjoyed that personally.  But the question is that you're focusing on the BRD4, and then -- through the epigenetic.  And again, goes back to the point that the -- what impact this kind of manipulation may have on the cellular gene with respect to the toxicity.  You mentioned that you have done a study on toxicity, but you didn't show any data.  Can you elaborate on that, what kind of -- in the long term, what kind of impact it may have on congenic pathway?

DR. HU:  So that's a really good question.  So there are two levels of answers to that.  First of all, like, many inhibitors or modulators target a host pathway, and induce some, you know, some other effects in terms of the cellular gene expression of target.  So we did RNA seq, and we see that it does induce a global impact on the transcriptomics, and it induce an opposing profile compared to the JQ1, which is true.  And second, in terms of the -- in terms of the toxicity, we did a lot on the cellular levels and the microglia and the J-Lat use different assays, but mostly use the flow-based live and dead cells staining.  So based on the data we have so far, the compounds, it's CC50 -- in the J-Lat cells, it's around like 40 to 50 micromolar.  And, you know, you immortalize the microglial cell lines, it's more than 100 micromolar, so it's not too bad.

We also -- so far we have done extensively some in vivo analysis for in vivo toxicity of these compounds in mouse, and it's interesting that so far we did -- we did -- we did two models, both the acute toxicity model and a chronic toxicity model.  In the mouse, at least so far, we didn't see a very obvious toxic effect using the dose we have tested in the mouse.  But I still -- I want to highlight it is likely that this type of manipulation of the epigenetic pathway likely have some long-term effect, which we have probably not known yet.

DR. KHALILI:  Yes.  Thank you very much.  Rebecca, I have one more question here from you.  I'm sorry passed earlier -- I didn't bring it up earlier.  That's coming from Dr. Maggirwar.  The question is about, "Which NF-kappa B subunit were blocked by the inhibitors in your assays?"

MS. PETERS:  Well, as I said, the mechanism for certainly caffeic acid, it's not known what that mechanism is.  And either of these inhibitors, both of them affect NF-kappa B, but I wouldn't say that they're the most specific.  They're certainly non-toxic.  But so caffeic acid, it would never -- we would not be able to answer that question at this point just based on what we understand of it.


MS. PETERS:  And for resveratrol, again, it effects the IKB alpha phosphorylation, and so that prevents translocation to the nucleus, which prevents everything downstream.  So as far as subunit, that's, you know --

DR. KHALILI:  Okay.  So the most concern was whether or not it's on C rail or B rail, so I don't think that you have an answer for that.

MS. PETERS:  There's no -- yeah.

DR. KHALILI:  All right.  Dr. Gendelman, there are a number of questions here for you that -- and then one of the question, which is very important is, that, "You have using this double basically targeted multiplex guide RNA for targeting of the -- for the region within the HIV, which is between the Tat -- covering Tat region.  How do you know that Tat is inactivated on these cells?"

DR. GENDELMAN:  Okay.  So I -- a couple -- a couple ways to show that, one --

DR. KHALILI:  Which one did you use?

DR. GENDELMAN:  I used the Tat -- the first and the second exons of Tat and I excised both.  The guide RNAs that we made we call Tat D and E were made from both of those exons.  That's first thing.  Second thing is we actually did a double conjunction -- what we call double conjunction where we've actually made Tat mutant proviruses as -- just to show as a -- as both a negative control.  And using that Tat negative proviruses, actually looked at the biology or the virology of that.  It was in one of my earlier slides.

The second thing that we did is several different assays, okay, in terms of inactivation of the HIV provirus, and as well as the virus itself.  First thing that we did is we took the lipid nanoparticles that were carrying the Tat D and E, and we exposed them to the monocytes.  These are primary monocytes.  And after days of -- consecutive days -- five, 10, 15 days -- we challenged those cells with HIV, and we were able to show inactivation vis-à-vis there was no evident viral replication as measured by RT activity and p24 from those cells.  That's number two.

Number three is we took the U1 and the ACH2 cells that contained one or more integrated provirus DNA that are restricted in terms of viral production.  And we challenged them with the lipid nanoparticles in replicated experiments, and then exposed them for 24 and 48 hours with either TNF or PMA.  In our controls, the controls demonstrated robust HIV production, both RT and p24 as well as the digital droplet PCR for HIV.  And then we took the supernatant from those inoculated treated cells -- viral-inoculated treated cells -- and we did coculture analysis either by cell or by virus to see if we could cover progeny virus, and we were not able to do it.

DR. KHALILI:  Okay.  Thank you.  Howard, thank you very much.  I think the question was whether or not you actually showed that the Tat is not present in the -- in the cell.

DR. GENDELMAN:  That was the last part --

DR. KHALILI:  -- doing that directly you were looking at -- the other important question is -- came up is, "How do you assess the off target in your system?

DR. GENDELMAN:  Yeah, off targets candidates were, so we took the top candidates.  Now, the off target -- let me just finish the last part of that.  The last part of that is we direct sequencing to show that the Tat was gone.  So we have the --

DR. KHALILI:  So how did you assess for the -- how did you assess for off target?  Can you answer that?

DR. GENDELMAN:  Yeah, the off target, we took the top guide RNA candidates, not only the CAT DNA, but all the candidates that we wanted to move forward, and we evaluated them in CRISPR off -- and CRISPR finder, the genome Cas finder tools.  This is a standard protocol that we've used all along.  And the sites were selected based on the highest-predicted off-target toxicity sites.  So we looked at that by sequential analysis with -- vis-à-vis the gene, and the testing in these sites were selected for our primary screening, and then we enlarged beyond that because we had a number of guide RNAs that we were screening, and then -- and then moved it beyond that for the Tat DNA.  We were not able to show any putative off-target genomic events, but that's how we screened.

DR. KHALILI:  Okay.  So you did not -- those are the good first step, but it doesn't really tell you that -- whether or not there is -- off target exists or not.  And then my other question is that in some of your J-Lat, you had the size of the fragment.  It's about 500 or something.  You said excised fragment.  What does that excised fragment mean?

DR. GENDELMAN:  Yes, this excised fragment is the -- is a space or the predicted fragment from what we -- what we had done vis-a-vis the Cas9.  And what we've done in all our --

DR. KHALILI:  No, this excise fragment is the fragment --

DR. GENDELMAN:  Well, let me finish.

DR. KHALILI:  -- which is removed from the -- form the viral DNA?  Is that what you're calling excised fragment?

DR. GENDELMAN:  We did both, what we removed and what was left, but we sequenced both, and the excised fragment is what we removed.  We sequenced that, what was left, and we sequenced that as well.  So everything was substantiated by sequence analysis, by SANG or sequence.

DR. KHALILI:  Okay.  Did you see any reintegration of the remaining of the viral DNA or excised fragment in the host genome?

DR. GENDELMAN:  Right.  That is something we're actually doing now, and the reason that we hadn't done that to start with is we wanted to screen many of the fragments of the guide RNAs first to see our sensitivity specificity and our delivery screening system.


DR. GENDELMAN:  But now we're doing that now.

DR. KHALILI:  All right.  The important question is -- that came also from Dr. Morgane Bomsel, that question is about -- again, go back to the how can you target the specifically-infected cells in vivo.  So whatever you showed was in vitro cell culture system.  How are you going to target in vivo system by --

DR. GENDELMAN:  Right.  So we're -- those experiments are actually ongoing.  And I think if you followed the last slide is what we're -- really have developed and done much of in the long-acting antiretrovirals is establishing targeting lipid nanoparticles.  So we have developed the means to place linkers from specific antibodies conjugated to our lipid nanoparticles that will target both lymphocytes, and depending on in vivo, exactly what we will be targeting, monocytes as well.  We're working with Dan Peer out of Tel Aviv in Israel to help us in part of that, and part of that we're developing on our own.

So these are all targeted lipid nanoparticles that -- and in conjunction with that, we're looking at the optimal formulation of the lipids itself.  So the lipophilicity gives us a big advantage in terms of entry into CD4 positive T cells.  So the composition --

DR. KHALILI:  I think you want to target also myeloid cells.  After all, this is a myeloid conference, right?

DR. GENDELMAN:  Yeah, I think I know something about macrophages.  It may be a long ago.


(Cross talking.)

DR. GENDELMAN:  I worked with macrophages before you were born, 40 years ago.  Maybe not before you were born, but before most of you were born.

DR. KHALILI:  Thank you very much.  Is there any other question for any of the participant, because right now there's not much -- nothing left here.

(No response.)

SPEAKER:  Kamel, I think there's a few questions left over in the Q&A, not in the moderator chat, but in the Q&A.

DR. KHALILI:  Okay.  Okay.  The one question is that, "Have you tried other LRA to reactivate HIV in the macrophage?"  This is -- I don't -- I don't know which -- addressed to which speaker, but is there -- is the LPS the strongest one for HIV reactivation?

DR. BOMSEL:  I guess it was for me, but I already answered.

DR. KHALILI:  Okay.  So I don't really see any more questions here.

DR. HU:  I see some additional here in the Q&A.

DR. KHALILI:  Well, for some -- they don't appear on my thing.

DR. HU:  I see one is for me.

DR. KHALILI:  Let me see here.

SPEAKER:  Haito, then maybe you can read off --

DR. KHALILI:  Yeah, maybe you can bring it up.

DR. HU:  Oh, okay.  Yeah, I can do that.  So one question for me:  "Have you seen other bromodomain proteins like a BRD2 in also inducing HIV suppression in microglia?"  Yes, I -- that's a really good question.  So far, we have not looked at other bromodomain-containing proteins in microglial cells yet, but we did look at BRD2 knockout in the J-Lat cells as I see in the second or third slide.  Compared to BRD4, we didn't see a strong effect if we knock out BRD2, at least in the J-Lat.  So yeah, I think it will be interesting next two steps to look at other bromodomain-containing proteins in microglia in regulating HIV transcription as well, and maybe likely also the -- some additional isotypes, yes.

DR. KHALILI:  All right.  Thank you very much.  Is there any other questions?  I really don't see anymore questions on my screen.

DR. TERMINI:  I see one for me actually.  "Do you think these antibody mutations would enhance ADCC elimination of macrophages?"  And so the answer to that is, yes, I do, and that's actually an active collaboration that we're working with to actually evaluate that.


DR. KAMINSKI:  If I may, I also see one question for myself.  So I got a question from Caroline about, "Is outcome found on HIV itself or other extracellular vesicles, and if so, could it contribute to its spread?"  So that is very interesting question, and I don't know.  I didn't find any literature about it.  But this clustering of ALCAM on the membrane is associated with interaction with tetraspanin CD9.  So most likely  -- and as you know, tetraspanins are markers of extracellular vesicles, so that is very likely.  But ALCAM is present on this, and, yeah, this is very interesting, but nobody -- to my knowledge, nobody check if there is -- if it affects spread of HIV.  And actually, we were discussing about this with my collaborator, Dr. Kihansi, and we were trying -- we will do it in the future.  So that's very good question.  Thank you.

DR. BOMSEL:  I see also a question from Rahm Gummuluru.  "What facilitates HIV transitories across mucosal epithelium?  Is this dependent or independent mechanism?"  And second question, "Which HIV-infected T cells themselves penetrate the epithelium?"  So it's quite good question.  A synapse formation is indeed dependent on envelope and cells -- T cells do not adhere to the -- to the epithelium as we have shown with the system.   We have also shown years ago that synapse formation depends -- and HIV production at the synapse depends on RGD-dependent integrins, so I guess it would seem here.

And concerning the second question, no, we couldn't find in this -- in no system, by the way.  We used for years HIV translocation across epithelium.  I remember some papers showing that if you activate the simple epithelium reconstruction, monocyte -- infected monocyte can penetrate.  But, no, it's not (inaudible).

DR. KHALILI:  All right.  Thank you very much.  So I guess we have ahead of -- let me see what's -- so what's next?  I guess we take time off, right?

DR. TERMINI:  Yeah, 10-minute break.

DR. HU:  Ten-minute break, yeah.

DR. KHALILI:  Ten-minute break.

SPEAKER:  Okay.  Thank you, everyone, and we'll return again at 12:15 to start the next session.

DR. TERMINI:  All right.  Thank you.


DR. SPUDICH:  Good afternoon, everyone.  Welcome back to this really important meeting focusing on macrophage biology and HIV infection.  My name is Serena Spudich, and I'm from Yale University, and I'm here today to moderate the Session 5, which is entitled "CNS Comorbidities in the Era of ART:  Involvement of Myeloid Reservoirs."  And I'll just start by thanking the organizers, and NIMH, and the Ragon Institute for hosting this really important meeting.

And I know that we heard from Dr. Gendelman that there's been a long history of interest in myeloid cells and macrophages in HIV infection, but I think that it's really a unique opportunity to hear from so many investigators with diverse experience and diverse expertise all related to these cell types and their role in HIV infection as well as other infections SARS-CoV-2 in this concentrated two days of science.  And I've really enjoyed many of the talks and heard from a lot of speakers that I didn't previously know, so I think this has been a fantastic meeting so far.  One of the special opportunities in focusing on myeloid cells and macrophages is the opportunity to focus on tissues rather than circulating or lymph node C4 cells, and there have been a number of talks about tissues already, including genital tissues as well as CNS tissues.

But this session is composed of three talks that will be focusing primarily on potential CNS-relevant myeloid reservoirs or myeloid infection.  So there are three speakers in the session:  Dr Hisashi Akiyama from Boston University, Dr. Lishomwa Ndhlovu from Weill Cornell Medicine, and Dr. Joan Berman from Albert Einstein Collect of Medicine.  We're looking forward to these talks, and we will follow this by a brief discussion session and then another 10-minute break, so thanks very much.  Dr. Akiyama, do you want to go ahead?

DR. AKIYAMA:  Thank you very much for the kind introduction.  Good afternoon, everyone.  My name is Hisashi Akiyama from Boston University.  I'd like to thank the organizer and NIH to give me such a great opportunity to present my data here.  And today I'm going to talk about innate responses in HIV-1-infected microglia.

So as all of know that HIV-1 individuals on combinational antiviral therapy still have higher risks for HIV-associated non-AIDS complications -- HANA -- which includes cardiovascular atherosclerosis and neurocognitive disorders and so on.  But this data shows that the percentage of neurocognitive disorder either in the pre-cART era or cART era.  And you can appreciate, that severe form of dysfunction, like HIV-associated dementia, has been dramatically decreased due to the introduction of combination therapy.  However, still 50 percent of people suffer from some sort of neural cognitive dysfunctions in the era of cART.  So we are wondering what contributes to HANA in those cART-suppressed HIV-positive individuals.

Several laboratory groups have shown the presence of HIV or SIV RNA tissues in HIV-positive or SIV-positive individuals on suppressive cART.  For example, Ko and colleagues and Lamers and colleagues have demonstrated the presence of viral RNA by RNA scope in the brain tissue from HIV individual on successful therapy.  Likewise, (inaudible) lab has shown a quantitative presence of viral RNA in the brain in -- from the SIV-infected macaques on suppressive therapy.  So it is clear that HIV or SIV infection persists in tissue reservoirs such as CNS, which is postulated to drive chronic immune activation.

So what are those HIV- or SIV-positive cells?  As we all know now, these are microglia and perivascular macrophages.  I'm going to focus on microglia today.  In my career, CNS-resident yolk-sac-derived macrophages and are one of the long-lived HIV-1 reservoir in the brain due to its share of renewable capacity.  And as you can see here, the CD68-positive macrophages are also positive for HIV viral DNA in this picture.  So, yeah, HIV -- microglia are found to be HIV positive in patients on therapy, and microglia can drive chronic immune activation, leading neuronal damage.  But mechanisms underlying this HIV persistence and chronic immune activation in HIV-1-infected microglia in macrophages still remained unclear.

So we wanted to investigate the mechanism.  So we have previously shown that HIV-1 infection -- HIV infection in macrophages are induced from proinflammatory cytokine responses.  We used monocyte-derived macrophages, MDM, and we derived them from monocytes in the presence of Human AB serum and M-CSF, and infected them with HIV-1 replication competent virus, and we harvested them for measurement.  As you can see here, this is one of the flow cytometry analysis that in HIV-infected p24-positive macrophages, CD169, also known as Siglec-1, which is an ISG and a marker for immune activation in myeloid cells, such as dendritic cells and macrophages, was highly upregulated.  And as you can notice that also the bystander in uninfected population also upregulated CD169, suggesting that there was a secretion of type I interferon into the culture supernatant.  We have also detected from Ag cytokines such as IP10 in the culture supernatant, so we decided to investigate the molecular mechanisms of this HIV-1 into similar activation.

So we wanted to address what kind of viral proteins are involved in this immune activation.  So we used HIV mutants which lacked expression of either Gag-pol, or envelope, or VPU, Vif, Vpr, or Nef.  But none of them -- when we transfused them into macrophages, none of them had impact on the expression of CD169, suggesting that these proteins are not responsible for this immune activation in macrophages.

So we got interested in viral RNA, so we used a Rev mutant which, as you know, that Rev is important to export by HIV-1 intron-containing RNA into the -- into the cytosol.  So when it expressed this Rev mutant, which cannot do this job, so which is deficient for intron-containing RNA export, even though we could establish a similar level of infection because GFP is expressed in the left region.  So this GFA expression is Rev independent, so there's no difference in the -- in the infection rate.  But this Rev mutant failed to induce or upregulate CD169 in macrophages, suggesting that the presence of intron-containing RNA in the cytosol is important for this immune activation in macrophages.

Now, we knew that RNA is important.  Next we wonder if, you know, what kind of RNA sensing molecules are involved in.  So we decided to knock down MAVS, which is a mitochondrial protein which plays an important role in transducing signal sensing of viral RNA in the cytosol and to induce type I interferon response.  So when we knocked the MAVS expression by shRNA macrophages, HIV-1 infection no longer induced upregulation of CD169, suggesting that this MAVS-dependent pathway is involved in the sensing and also secretion of type I interferon -- type I interferons into the culture supernatant.

So in summary, so HIV-1 RNA, in particular, in intron-containing RNA in the cytosol which is exported by this CRM1-dependent and Rev-dependent pathway, is the trigger of innate immune responses in macrophages.  And MAVS is required for this immune activation, which leads to the secretion of low level of type I interferon and proinflammatory cytokines.  We didn't show the data, but CRM1 inhibitor like KPT, can block  -- which block nuclear export of HIV containing -- HIV-1 intron-containing RNA, which inhibited innate immune responses in macrophages.

It should be noted that current cART regimen cannot prevent either HIV-1 RNA transcription or nuclear export of intron-containing RNA.  It's highly plausible that HIV-1 intron-containing RNA-driven immune activation may persist in HIV-positive individuals on successful therapy.  So based on these findings in macrophages, we have hypothesized that persistent de novo expression of HIV-1 intron-containing RNA induces innate immune activation in microglia, tissue-resident macrophages, which contributes to neuroinflammation and neurotoxicity.

To study this, there was a problem which is access to primary human CNS-resident cells because that is very limited.  That's why we decided to do take advantage of two in vitro cellular platforms.  One is monocyte-derived microglia, and the other -- and the other one is IPSC-derived microglia.  For the sake of time, I'm going to focus on IPCA-derived microglia hiMG in this talk.

There are multiple protocols available to generate hi-microglia from IPSC, but we chose one which was published by Takata and colleagues in (inaudible) Lab, which was published several years ago.  So basically, we generated a yolk sac-derived primitive macrophage hiMAC from IPSC line.  Then at the same time, in parallel, we generated a hiNeuron from the same IPSC cell line by co-culturing them for additional two weeks.  Macrophages were differentiated in microglia hiMG.  So we characterized them -- characterize them by messenger RNA analysis, and we found microglia-specific RNA genes, such as TM119 and P2RY12, was highly upregulated in the co-culture.  We confirmed the expression of macrophage markers, such as IBA-1 and TM-1 -- sorry -- microglia or macrophage marker IBA-1 and TMEM119 in the co-culture, and we clearly see the two distinct population by flow cytometry.  One is microglia, which is P2RY12 positive, and the other one -- population is neuron, which is tubulin beta 3-positive.

Then we infected the co-culture with HIV-1 replication-competent virus.  Then we measured intracellular p24 in the microglial population.  Then it turned out that microglia are highly susceptible to HIV infection.  Sometimes we got like 40 percent infection in the microglia population, which was completely blocked by efavirenz, which blocks RT process or raltegravir which blocks integration.  And KPD, once again is a CRM1 inhibitor that blocks nuclear export of intron-containing RNA into the cytosol, including the messenger RNA for gag, so KPD block the expression or p24.  And interestingly, we have observed the secretion of type I -- sorry -- proinflammatory cytokines, IP-10, in the co-culture upon infection of microglia with HIV-1, which was completely blocked by efavirenz, raltegravir, and KPT, which is a CRM1 inhibitor, suggesting that intron-containing RNA in the cytosol is the key to trigger these innate responses in microglia.

We also measured microglia viability by flow cytometry and found that microglia infection with HIV led to the decrease -- reduction in live cells, and also we measured the functions of my microglia.  And one of the important homeostatic functions of microglia is clear lines of debris, such as fibrillar form amyloid-beta.  So we measured uptake of (inaudible) labeled fibrillar amyloid-beta by infected or uninfected microglia.  And we found that infected microglia, which is p24 positive, had reduced capacity to update for data or the fibrillar amyloid-beta, suggesting that -- suggesting that HIV-1 infection microglia leads to -- leads to the microglia damage as well as functional impairment.

In conclusion, IPSC-derived microglia hiMGs are useful tools to study HIV-1 infection of CNS region in microglia, and microglia are highly susceptible to HIV-1 infection, which is in agreement with the previous finding using primary human microglia.  And, importantly, expression of HIV-1 intron-containing RNA in microglia results in innate immune activation and impairment of homeostatic function, and which may contribute to chronic neuroinflammation in HIV-positive individuals on ART.

So in the next couple slides I want to introduce some preliminary ongoing work.  So further data I have shown was from the 2D culture of neuron and microglia, but we tried to start this 3D tri-culture model just to better mimic the complex interaction of CNS-resident cells, and also in order to investigate the role of astrocytes, which is another key player in the neural statis in neuronal inflammation in addition to microglia.  So we derived microglia, neurons, and astrocytes, those three key players, from IPSC and assembled it into this ball-like structure.  We call it a spheroid.

So here's an example of immunosuppressant stating of the spheroid.  So we infected microglia with HIV-1 expressing GFP, so you can see some GFP-positive cells, together with astrocytes, G5-positive, in neurons which is TSG-1 or tubulin beta 3-positive.  Now we can see this nice assembly of those three key players into the spheroid.  Now, we would like to -- using this 3D model, we tried to investigate the role of HIV-1 infection in microglia, particularly the role of intron-containing RNA in microglia activation in neuropathogenesis.

So with that, I'd like to thank all of the members in Rahm Gummuluru's lab as well as our collaborators, Gustavo Mostoslavsky and Christine Cheng, and fundings, and for your attention.  I'm now happy to take any questions at the end of the session, and I now pass it over to Dr. Lishomva Ndhlovu.  Thank you.

DR. NDHLOVU:  Okay.  Great.  I want to thank you, Dr. Serena Spudich, for leading the session.  I want to thank Kiera Johnson -- Kiera Clayton -- sorry -- and Jeymohan Joseph from NIMH from the Ragon Institute for helping put this session together.

So I've talked about myeloid HIV reservoirs from establishment to aberrant cognitive trajectories.  I'm professor of immunology in medicine and neuroscience at Weill Cornell Medicine, and our program has been focused on understanding the role of monocytes and other non-T cell reservoirs in HIV.  And I show this slide really to just try to highlight some of the interest we have had as we think about the two cases of HIV cure and just pose the question as to what role monocytes may have played in the -- in these two successful cases of remission.

I think the "London patient," it's interesting to know that he was on a T cell-depleting antibody therapy, anti-CD52, prior to his transplant and irradiation and stem cell transplantation.  And it's notable to know that targeted CD52 does seem to target other cells, including dendritic cells, NK cells, and myeloid cells.  It's also notable that the "Berlin patient" on the CD33 depleting antibody mAb, which at the time was not authorized in the U.S., but it was actually being used by him in Germany, and that has the potential to impact the myeloid reservoirs.

So we're particulary interested, as we think about latent reservoirs, we're particularly interested in studying the role of peripheral monocytes, and I think there's been some really exciting discussions about macrophages and microglia, and just pose the discussion about how we think monocytes may be relevant as a potential transitory cell that may enter from the blood into the periphery.  And this really nice review Frontiers sort of outlines some of the mechanisms by which infected monocytes may actually transverse virus into microglia in the brain, and thereby seeding virus in this compartment.

It's also notable that this can also happen in the lymphatic system, and we do know that it's a lymphatic system linking the CNS to the brain, and potentially that could be another route of entry for infected myeloid cells.  It's important to note that they could be different subsets of monocytes that are more predelicted to retain virus that may also seed.  We know these are short-lived cells, so the question always comes as to the role they may play in the bone marrow as a potential source of seeding of cells that are actually infected into the periphery and eventually into the tissues.

So there are a number of assays that have been presented at the meeting and presented at the previous meeting in Miami two years ago on potential uses of viral outgrowth assays in this department, and this has also been shown in primates and at this meeting today.  So the sensitivity and accuracy of measuring myeloid reservoirs in the periphery in this compartment is of continued interest and may be relevant if you think about curative strategies.

I'm going to give you a little background about what I'll talk about, two findings we've had in the past few years.  We do know that a subgroup of individuals with HIV, despite initiating ART very early, soon after acute infection, we already begin to see evidence of regional brain atrophy and potentially impaired cognition.  This is always notable to mention that this is always in the subgroup of people, and there are certainly strong indicators that the myeloid -- the monocyte macrophage activation status are really strong predictors of indicated of abnormalities in the brain.  And this work by Carpenter back in 2020 has shown these effects happening after two years of sustained and antiretroviral therapy.  Even though they were initiated soon after infection, we do see impact on the brain.  And the mechanism by which these may be occurring could be driven in part by the myeloid compartment.  These are through local damage of inflammatory mediators or direct damage to the neurons.

So what we tried to do in this part of the talk is really tried to determine that the establishment of HIV infection in the monocyte occurs very early during acute infection, whether this could be linked in any changes in cognitive performance, even though we are able to rapidly initiate and antiretroviral therapy needs in these individuals and sustain that suppression for long periods of time.

To do that, we've taken advantage of our HIV -- acute HIV cohort in Bangkok, Thailand where to date we've screened over 400,000 individuals -- 400,000 samples out of 682 individuals, and many of them -- actually most of them actually were able to initiate antiretroviral therapy.  And then these individuals, mostly male and MSMs, were 81 percent of the potential circulating virus in that region.  We're able to capture individuals within 19 days of infection, and we have a number of individuals of different Fiebig stages.  But what's exciting with this cohort of longitudinal follow-up was we were able to do a number of CNS-related measures, including lumbar punctures, and leukapheresis, and brain imaging data as well.  And we have a very robust repository, and this it was a very large collaboration to put this operation together.

Now what we did here is really follow over several years, and this is through funding from NIMH, to follow 30 individuals over a two-year period where they had serial cognitive assessments performed imaging but also had CSF and blood samples taken, including leukapheresis at multiple timepoints from the time they were diagnosed in acute HIV, but also soon after infection.  And what we did was we followed them up over a 144-week period, and I'm going to show you data in these early timepoints here where we were able carry out a number of these assessments in these individuals.

What we were able to do was carry out cell-associated -- with leukapheresis, we were able to identify a large volume of cells and really detection.  The goal here was to try to see if we could detect the replicating virus from the myeloid compartment throughout this period, but also this was done through high monocyte cell sorting through high-sensitivity cell sorting to increase the accuracy and the purity of the cells.  Four cognitive tests.  This was not by all means and extensive battery, but really trying to capture cognitive status based on norms within the region.  And then we also had an opportunity to measure a couple of cellular biomarkers, essentially linked to myeloid activation, and then eventually developed a QPCR method to try to begin to measure viral quantification within the cell types.  And what we discovered in these early pilot studies was we really needed a large number of cells -- in excess of 40 million -- for us to be able to reliably quantify virus in this particular cell compartment.  Most assays typically have been designed to study CD4 T cells, and these were our primary and secondary outcomes for these assessments.

What I will tell you is that we were able to identify -- on the left in the circles are the monocytes, and on the right are the CD4 cells.  We did see that there was a variation in individuals with detectable cell-associated HIV RNA prior to ART, and this is within a -- over a two-year period we were able to see that.  Most of it actually was suppressed to undetectable levels based on our assays, but there were some individuals that were retained, potentially ones with high levels of detectable RNA at higher levels.  But essentially pretty much everybody with the CD4 cells, we were able to detect consistently persistence of HIV post-ART in these individuals.

And what we tried to do was to see if there were any association with some of those four cognitive domains.  We did see that Color Trails performance scores tended to be a bit lower in those individuals with the detectable virus in the myeloid compartment versus those without -- with undetectable virus.  As you can see, their performance scores, they're lower.  They tended to perform worse on these tests, and higher performance suggests less -- a much better performance on those tests.  So just to let you know we're really looking at these detectable versus undetectable groups at baseline between those assessments.

And then sort of -- you know, so where we are, we're really noticing these subdomain performances.  We did not see this in the other subdomains and we looked at these segregations in the cohort.  What was interesting, though, is over time, because we had cognitive assessments over the 144 weeks -- this is a collaboration with the University of Missouri St. Louis -- we were able to show that the trajectories of change in cognitive performance over the time was actually segregating quite differently between the two groups over that period.  The most striking dataset really was shown within the Color Trails Making A Test performance studies, those with detectable virus in gray.  There was already segregation at baseline, and whether there's some predetermined factors that will allow for viral persistence is something that we wanted to study.  But these are all baseline values, and we can see that their trajectories are certainly segregated and in some cases converge, but other times remain segregated throughout the course of 144 weeks.

Overall, the performance of these individuals was actually about what -- were actually normal.  Subgroups were performing under -- just a small percentage of individuals did not perform, were slightly impaired, but overall the performance segregation was quite evident based on the detectability of viruses at baseline in these individuals.  We also looked at the soluble biomarkers.  The only two that really popped out as being a sort of mildly significant is really the levels of neopterin in IL-6, and the plasma at baseline really were beginning to segregate into these two groups based on their detectable RNA in their myeloid compartment.

So really just to summarize this first part of the talk, we do see active HIV persistence in the myeloid cells.  The monocytes are the substantive donors early in acute infection.  It is substantially reduced by ART, and we do see improved performance overall globally in three out of the four domains.  We note that the detectable baseline levels of cell-associated monocytes was associated with reduced changes in performance, and principally, these are representing psychomotor speed performance, which has been known to be impaired in HIV, particularly over two years, even in this early stage of infection, despite very early initiation of ART and continuous suppression over two years.  So we do note that those with detectable virus had delayed favorable cognitive trajectories, and we think this could be an important early predictor of brain health diagnosed during acute stage of infection.

And this really underlines that, even though we aren't producing ART in the subgroup of individuals, we may need to think about additional interventions to preserve brain health in these individuals and acute infection.  We're excited with our cohorts that we're able to follow them quite long term, in some cases, over six to seven years now, and really interested to know what would happen to them over time, which really speaks to the beauty of a cohort is allowing us to follow up beyond these timelines I've given to see whether there's any long-term unfavorable phenotypes that we might be able to uncover in these two segregated groups.  And there's a lot of work further to be done.

I think I'm okay with time, but I wanted to just finally end is to talk about some studies we're doing to try to better understand the dynamics of that establishment in that compartment by using health epigenetic studies to try to see the evolution of acute infection and the establishment of virus in this study, but also to see the relationship with cognitive or clinical outcomes in this group.

To do this work, this is work that was done at our lab with collaboration with Michael Cawley, was to capture these individuals at the early stage of infection, 22 individuals, cell sort both CD4s and T cell monocytes who perform that DNA methylation array to try to profile the methylation status of these two cell types.

And just because of time, I'm just going to show you some of the data.  What was very intriguing is that we are able to see, compared to control individuals, quite different profiles of the myeloid cells.  The epigenetic changes do seem to occur very early when we compare these with uninfected controls.  But more interestingly, when you compare sort of the changes that we see between the myeloid and the CD4 T cells, we're seeing a much greater change in the monocytes compared to the T cells, which was actually quite surprising in the differential methylation status.  In fact, what was interesting is there's really a much greater change in CPG sites that would were differentially methylated in the myeloid compartment versus the CD4 T cells with a few overlapping sites between the two cell types.

What was very interesting is that when we -- when you look at these individuals from the time they get on antiretroviral therapy, it was striking that you do see the demethylation occurring in particular genes, and this is one particular gene using H2 pre-ART across the different stages, really unchanged after that initiation of ART.  So ART really doesn't seem to restore any of these epigenetic changes that are induced by HIV, and we see this across the board, with the exception of the interferon genes.  They are -- they are clearly impacted by ART, but by and large, we don't really see much lasting changes.

What was interesting with this particular site -- methylation site within CH2 is it encodes the protein that's maintaining transcription repression.  And if you look at this chronic infection, you can see that post-ART, we actually don't see much increase in methylation in this particular gene, and this is the "Berlin patient" showing you that.  To some degree, whether this is natural human methylation status or whether this is sort of a setting where we've actually cleared virus, whether that is restored is something of interest as we think about the effects of curative trials.  But this knockdown of this particular gene has been shown to impact HIV latency, and it may be part of this multi-enzyme complex that may be suppressing transcription genetically.

So these are really just a snapshot of some of the things we're now looking at in identifying sites for the genome and the host that could be relevant for persistence of virus in some cells, but not in others, in different individuals.  So just building on that work, some work by Rob Paul and Phil Chen who just recently got his PhD -- we're excited about that -- who's shown that in this particular cohort, we actually are able to see that monitor the trajectories over time in these individuals in a larger cohort.  But because we have all this baseline datasets -- clinical data, but also logical data -- the logical database line, we threw in our methylation datasets into gene a learning machine learning model to see what would be the best predictors of these outcomes.  And it was very evident that the -- certain sites within the genome of the -- epigenome of the monocytes were really the strongest -- incredibly strong predictors of change over time in acute infection, again, speaking to the value of the myeloid compartment versus all other compartments, but also other neurological parameters that one would consider relevant as we try to predict CNS outcomes in this population.

We have actually taken this dataset even further.  It's shown that in this cohort, we were actually able to identify individuals with pretty good favorable phenotypes.  We followed these people.  This is just a group that really just do very well.  They have no serious coding events, as you can see in this table, and, again, what are the strongest predictors?  Again, we're finding, again, a number of sites within the methylation profile of these monocytes, just incredibly strong predictors of these outcomes over 96 weeks.

So really, in summary, we now know that HIV rapidly disrupts the host immune epigenetic landscape.  The monocytes appear to be more impacted than CD4 T cells.  This is not mitigated by ART with less than a one-percent differentially-methylated loci being changed.  There's a durable epigenetic memory that may be shaped by HIV in the cell type.  We know these individuals' epigenetic states really may determine the clinical course of long-term trajectories, particularly performance in cognitive testing that we have seen.

So we're really excited with these datasets, and we're now expanding these signatures to follow these individuals further and identify individuals that may be relevant to follow up for curative interventions, but also studies that we've already ongoing looking at viral rebound during this stage of infection.

This work has now been published, and this really couldn't be done without a huge team of individuals.  I can speak to the participants in the RV254 study, but also all the members that participate in the study, but also in the funding agencies that support all these activities, IHR, which is now recently as part of the clinical studies in Bangkok.  This is the lab that we've been supported by the NIMH and NINDS in some of the work that we're doing.

And I want to end by just letting you know that there is a biotypes meeting that's coming up next week, so feel free to either take a picture of this and please attend.  I think I'll be an exciting opportunity for us to better understand the CNS complications of people living with HIV and what biotypes we might be able to think about.  So I'm happy to take questions at the end and thank you for your time.

I'm now going to pass on to Joan Berman from Albert Einstein.

DR. BERMAN:  Thank you, and I'm absolutely delighted to speak at this meeting, and I'd like to thank the organizers for this exciting program.  My work will dovetail really well with those just presented in the past two sessions, so that's even more exciting.

So I'm going to talk to you about viral seeding by mature monocytes and potential therapies to reduce do CNS viral reservoirs in the ART era.  We've already had quite a bit of a conversation about HIV infection and neurocognitive impairment.  Just to repeat that virus seeding in the CNS occurs early after peripheral infection, and that 15 to 40 percent of people living with HIV still have some form of cognitive impairment despite ART and regardless of viral load, and that there are no therapeutic interventions at the moment for people with HIV on ART.

So the mechanisms of CNS HIV infection have been covered quite extensively in this talk, so I will just briefly say that infected monocytes will cross the blood-brain barrier in response to cytokines or chemokines, differentiate into long-lived macrophages, and infect other cell types, specifically microglia or low-level astrocytes.  These cells will also become activated even if not infected.  There'll be a low-level viral environment setup that will recruit additional cells in reseeding reservoirs, and ultimately neurotoxic hosts and viral vectors will damage neurons leaving to impairment in a large percentage of people.  CCL2 is the most potent monocyte chemoattractant, and it is significantly elevated in the CSF and CNS of people with HIV even on ART.  CCR2 is the only receptor for CCL2 human monocytes.

There's a specific subset of monocytes that we and many other labs have shown are sent to an HIV neuropathogenesis, and that's a CD14 and CD16 monocytes, CD14 being the LPS co-receptor and CD16 the FC gamma 3 receptor.  In this flow cytometry plot, the CD14 on y-axis and CD16 on the x-axis.  You can see in people who are healthy the number of 14/16 positive monocytes is very low, five to 12 percent, but in a person with HIV, it can get as high as 40 percent.  This is really important because these cells are highly susceptible to HIV infection compared to other monocyte subsets, and they preferentially cross the blood-brain barrier.

So not only can these cells cross the blood-brain barrier, a former student in my lab, Dr. Deanna Williams also showed that when PBMC were taken from people with HIV and put in our blood brain barrier model for transmigration to CCDL2, the monocytes from people without cognitive impairment transmigrated more than other monocyte subsets, but those from people with impairment increased their transmigration even more compared to baseline and compared to CCL to the people from the normal cognition.

But these monocytes from people with HIV as well as HIV-infected cultures are heterogeneously infected.  Some harbor HIV and others are exposed to HIV viral proteins and inflammatory mediators.  So some, we call them HIV-positive meaning that they carry the virus, and others are exposed.  They've been exposed to all these mediators, et cetera, but they do not carry virus.  So we call one HIV-plus and the other HIV-exposed, which leads us to the goals of this study, which are to characterize the entry of HIV-positive CD14 and CD16 monocytes into the CNS and identify interventional strategies to reduce infection and reseeding of viral reservoirs in the context of ART, and, therefore, neurocognitive impairment, two different ways:  using PBMC obtained from people with HIV and ART and also monocytes matured HIV infected and treated with ART in vitro.

So part one is to characterize transmigration of HIV harboring mature monocytes using PBMC from people with HIV on ART.  This is a schematic of our blood brain barrier.  We co-culture human brain microvascular EC and human astrocytes on a -- on a tissue culture insert that as three micron pores.  The astrocytes put the processes through the membrane and seal the barrier.  We then put PBMC or monocytes or other cell types we see on the top of the barrier, which is the peripheral side, and chemokine on the bottom, which is The CMS side.  We allow the cells to transmigrate, collect them post-transmigration, analyze them by flow cytometry and by a variety of molecular techniques.

So the experimental setup is in collaboration with Susan Moore Gallo -- Dr. Susan Moore Gallo, who is the director of the Manhattan HIV Brain Bank.  She has a very well-established and well-characterized cohort of people living with HIV.  We obtained blood.  We also did neuropsychological testing and imaging, but I'm only going to talk about the transmigration studies today because we're blinded to the other data.

So people with HIV will come in, and we will obtain their blood, their the PBMC.  We'll take some of those PBMC and quantify the 14/16 positive monocytes as well as the CD3 T cells by flow cytometry.  We'll also do DDPCR and DNA/RNA scope for the presence of HIV.  So the first part will show you the quantification by flow cytometry, and here you can see that in response to CCL2, the mature monocytes transmigrate a great percentage compared to the percentage input of other cell types when compared to the baseline control.  T cells also migrate, but not nearly to the same extent.  This is when we normalize to one to the baseline, and this is a 3.5-fold increase of mature monocytes and a 1.2-fold increase of T cells.

So the next part is how do we quantify HIV DNA.  What cells pre- and post-transmigration are carrying the virus, and are they enriched post-transmigration?  So first I'll talk about the ddTCPR studies.  When we quantify whether there's an enrichment for cells harboring HIV, is we do quantitative DDTCPR on the pre-transmigration of the cells prior to their transmigration.  We get the number of HIV gag copies of DNA per 10-to-the-6 PMBC, and then the same for post-transmigration.  We divide the post transmigration copy number by the pre-transmigration copy number to get an enrichment factor.  If the enrichment factor is greater than one, it means that the cells harboring HIV translate -- transmigrate more than those that are just expose.  If it's equal to one, there's no enrichment, and if it's less than one, it means that, in fact, the cells that are exposed, but not harboring, HIV have a selective advantage across the barrier.

And here are the interesting and exciting data.  Pre-transmigration, the PBMC and a million cells have 31 copies of HIV gag DNA compared to 645 post, which is a 20-fold increase.  So these are PMBC.  We didn't know whether the monocytes or T cells both, who was harboring the virus, who was transmigrating.  So in collaboration with DR. Eliseo Eugenin and Maribel Donoso at the University of Texas, we did DNA/RNA scope.  Again, we quantified an enrichment factor, this time looking at HIV Nef DNA pre- and post-transmigration, Nef DNA and gag/pol RNA pre- and post-, and Nef DNA, gag/pol mRNA, and HIV p24 protein pre- and post-.  And what the data very clearly indicate is that that monocytes transmigrate across the blood-brain barriers.  The ones that are carrying DNA have a 2.5-fold enrichment.  Those that have both DNA and gag/pol RNA, meaning active viral transcription, had an 11.4-fold increase.  And those with three active viral translation had a 6.5 increase.  This was really exciting data because these individuals have been on suppressive ART for a minimum of two years and many for up to 10 years.  And so even with such low-level infection, we could still detect the fact that those carrying virus entered the CNS selectively.

So the next part is to characterize the mechanisms of transmigration that mediate HIV-positive mature monocyte entry into the CNS using monocytes matured, infected with HIV, and treated with ART and cultured to identify targets to block the establishment and reseeding of viral reservoirs, and to reduce and eliminate cognitive impairment.

So just as a reminder, PBMC from healthy individuals have very low levels of CD14/CD16 cells, and we need a lot of these to be able to study because these are the cells that are preferentially infected with HIV among monocyte subsets.  So we developed a culture method in which we take leukopaks from the blood bank, isolate PBMC, select the antibody for the monocytes, and then we culture the monocytes non-adherently so they don't differentiate into macrophages for two or three days in the presence of the MCSF to mature as many cells as possible into those 14/16-postive cells.

This is a picture of pre-.  This is the PBMC, the number of 14/16s that are in this sample, and then after culture, 80 percent of them are the mature monocytes with enough cells to study.  We can then infect some of these cells with HIV ADA, and then after infection, treat them with ART, Tenofovir, and Emtricitabine, which the backbone of all the patients in our study, have as part of their ART therapy and test them in our transmigration assays and in our molecular studies.

So first, we'll show that ART-treated HIV-positive mature monocytes preferentially transmigrate across the blood-brain barrier to CCL2 -- this is our model system -- and that all cells, both infected and exposed, transmigrate significantly to CCL2.  When we look at their HIV DNA, there's a selective advantage yet again of the cells carrying virus compared to the pre-transmigration, and this is a 2.54-fold increase.

So the question is can we block this transmigration.  Can we block specifically the HIV-positive cells?  So you've already been given a really nice introduction to junctional proteins.  So briefly, our lab showed that CCL2, the chemokine receptor for CCL2, and JAMA-A and ALCAM junction of proteins that are essential to mature monocyte transmigration across the blood-brain barrier are increased on the mature monocytes in people with HIV, even in the presence of that.  So we set up a strategy to determine whether we can block specifically the cells carrying HIV.

So on top of the insert we have media.  This is CCL2 on the bottom.  Then we also treat the cells with Cenicriviroc, CCL2/CCL5 inhibitor, anti-JAM-A or anti-ALCAM.  We'll allow them to transmigrate to CCL2 in our controls in IgG1.  After 24 hours of transmigration, we collect the cells and we quantify the number of cells of transmigrated to flow cytometry data as both infected and exposed cells, and then we collect the cells also analyze them by DDPCR.  The data are really compelling.

So first, CCR2 may be a potential target to reduce or eliminate transmigration of ART-treated HIV harboring CD14- and CD15-positive monocytes.  There's an increase to CCL2 of these cells for transmigration, and treatment with cenicriviroc on the top of the blood-brain barrier model where the monocytes are put greatly reduces the number of cells that transmigrate.  But it's important to note that there are still cells going, and when we quantify by DDPCR that DNA copies the enrichment factor, the post-transmigration is 2.5-fold enriched, and with CBC, it's below baseline.  So several -- most of the cells that are infected are reduced in transmigration as compared to those that have just been exposed.

GMA is even more exciting.  Here you can see that it reduces the number of cells that transmigrate.  The IGG control has more effect.  Again, a 2.4-fold increase post-transmigration, but look at this.  Anti-JAM-A antibodies absolutely eliminated any detectable HIV-harboring cells in seven of the nine samples tested, independent samples, and the IgG had no effect.  Similarly, ALCAM also blocked, did not reduce the number of cells quite as much, but it really reduce those harboring HIV.

So what does all this tell us?  So the importance of the findings are that even with fully-suppressed viral replication, mature monocytes carrying virus still selectively enter the CNS, can replenish reservoirs, facilitate chronic neuroinflammation, and contribute to neuronal damage and cognitive impairment.  And these data also indicate strategies that target scope -- target host proteins to limit transmigration, specifically the HIV harboring mature monocytes across the blood-brain barrier, contributing to eradication of CNS viral reservoirs and of neurocognitive disorders.

It's really important to say that the vast majority of this work was done by an outstanding M.D./Ph.D. student in the lab as well as another terrific graduate student, Rosiris, who is no longer there, and Mike Veenstra, Dr. Dionna Williams, who at Hopkins started a lot of these project.  And other people are continuing it, like great collaborators, Dr. Susan Morgello, Dr. Eugenin, Dr. Donoso.  And Dr. Clements and Gama and Veenhuis Dr Susan taught us how to DDPCR and participated in some of the early studies.  These are our neurocognitive testers, et cetera.

So I want to say also that this cohort was non-initiated ART very, very early.  It was -- the vast majority of these individuals were not given ART until their CD4 counts fell below 350 as per protocol, but yet our data still dovetail really well with the presentation that Lish gave earlier.  And so there's a lot more work to do, but all these studies and all these talks have underscored the importance of the infected monocyte in HIV neuropathogenesis and reservoirs.  Thank you, and I'd be delighted to take questions when this session starts.

DR. SPUDICH:  Thank you so much.  I'd like to ask Dr. Berman to leave her camera on, Dr. Ndhlovu and Dr. Akiyama to please turn on your cameras.  And first of all, I want to tell you this was absolutely a fabulous session.  Those were beautiful talks.  Couldn't have been better linked and I think have really already answered some questions for me, and have -- also you've made our job easier by stimulating a lot of questions already in the Q&A.  So we'll be taking some questions -- we have 15 minutes or so -- and then we'll have a 30-minute actually break after this session.

So I'm going to start out by a question for Dr. Akiyama, and by the way, there are many congratulations also in the Q&A on these great talks, but I'm not going to repeat those.  "Dr. Akiyama, can you please comment on the contribution of astrocytes to the HIV-associated inflammatory response in your spheroid model?"

DR. AKIYAMA:  Yeah, that's a great question.  So we have done some preliminary work on astrocytes from IPSC just, you know, phenotype them and infect them with HIV.  And we found little infection in low-level astrocytes.  So in the 3D culture, we expect that astrocytes is not going to be the main producer as an HIV, you know infected cells, but we expect that, for example, exposure of viral particles astrocyte by, you know, activated astrocytes or infected microglia, loss of cytokines that might also influence the activation status of astrocytes.  And altogether, they -- you know, astrocyte may play an important role to either -- how to say -- amplify the neurotoxic features induced by infection with HIV.

DR. SPUDICH:  Yeah.  I mean, I think a really important part of your model is the idea that you have these different cell types all integrated rather than a single cell type in an organoid, but this spheroid and trying to reproduce what's happening in the brain.  I think it's really a fantastic opportunity.

DR. AKIYAMA:  Mm-hmm.

DR. SPUDICH:  Next question is for Lish.  So the statement is, "Your data suggests that macrophages are more sensitive to ART compared with T cells.  Is this known to be the case?  Do you know how many of the macrophages still harbored HIV despite reduction to below detectable limits by ART?"

DR. NDLOVU:  Yeah.  So I guess the question was whether macrophages were more sensitive to ART.  I think what we were showing is that actually epigenetically, that both the CD4s and the monocytes actually we're not impacted by antiretroviral therapy.  We did not see any dramatic shift in the methylation  -- differential methylation state before ART and after ART.  So one would think that antiretroviral therapy would have a more potent effect, but we did not see changes there.  I guess that's sort of how the data was presented.

The second question is how much virus.  So we were looking at monocyte RNA, so cell-associated RNA.  We did see pretty much most subjects by this PCR method were able to show detectable RNA in the myeloid compartment, certainly less than what we see in the CD4 compartment, but after ART, less than -- I think about 40 percent did show detectable RNA.

DR. SPUDICH:  So thanks, Lish.  So this is for Dr. Berman.  "Are CD4-positive/CD-16-positive referred to intermediate subtypes of monocytes, and are these subsets and monocytes more susceptible to HIV infection than classical and non-classical monocytes?"

DR. BERMAN:  Sorry.  Yeah, CD14/CD16-positive monocytes are the most susceptible subset of monocytes to HIV infection.  Whelan Pulliam and lots of lots of people, Harvard, etcetera, have shown that.  On neuroclassical, there was intermediate, which are the 14/16s that I call mature, and then there's the non-classical which are not affected either.

DR. SPUDICH:  So another question for Dr. -- Hisashi.  So "have you tested HIV latency in your 2D microglia culture," or, I think, the steroids or the 2d microglia culture, and "can you reactivate the HIV with some latency reactivating agents?  In your 3D culture, what cells can be infected by HIV," and then the question about microglia or astrocytes.  And that may be a question of stimulating your next set of experiments, but it would be interesting to hear if you have some ideas.

DR. AKIYAMA:  Right.  So the -- for the 2D microglia culture with neuron, we finished experiment about, like, two weeks post-infection, so I think I learned that two weeks is not enough to establish, like, a latency or quiescent infection in microglia or macrophages.  So yeah, we plan to extend the culture period up to, like, maybe four weeks or even longer to see -- you know, test those reactivation.  It will be very interesting.  Also for the 3D culture, as I briefly said, that we had preliminary work doing, like, astrocyte infection, but, you know, they are not productively infected that much.  So but still, so we expect that -- those 3D culture mcl will be the main infected cells.  Yeah, that's we think, but we're going to find out.

DR. SPUDICH:  Another question for Dr. Ndhlovu, and I think this is talking about probably the Thai cohort.  "With one of the parameters that differentiate groups with development of some cognitive impairment being viral RNA and monocytes, are critical cells expressing virus or latently infected?"

DR. NDLOVU:  Yes.  If I understand that, are these newly infected or cells expressing virus in the myeloid compartment pre-ART.  It's a great question.  Difficult to answer.  You know, I think they are -- we're certainly interested in looking at the viral products in the -- in the blood and seeing whether those might be driving a number -- some of these segregated outcomes that we're seeing.  So a difficult question to ask.  We just don't have the data to confirm those delineations.

DR. SPUDICH:  Well, and perhaps I can clarify.  Maybe the question was is there active viral replication, or is this -- you know, you're detecting RNA.

DR. NDLOVU:  Correct.

DR. SPUDICH:  So do you think it's just transcripts -- intracellular transcripts that are sort of there by chance?

DR. NDLOVU:  Yeah.  So we've actually developed -- yeah, we've actually developed a quantitative assay using TZM-bl cells, and we've tested this in the Thai cohort samples at baseline and actually after ART.  And we do see sufficient data there is viral outgrowths.  So whether those cells can now further replicate using the TZMBL assay, we do believe that they are potentially replication competent.

DR. SPUDICH:  So another question for Dr. Berman.  Actually a couple here.  I think these maybe go together.  So one is, "Can you comment on the permeability of free virus trafficking across the blood-brain barrier and its contribution to the HIV reservoir?"  And also as a quick one, "How specific is cenicriviroc for CCR2?"

DR. BERMAN:  It's not specific to see CCR2.  It's specific to CCR2 and CCR5.  But in the 24 hours that we have it added and since its cells are already on ART to start with, we don't think it's making a significant contribution.  With regard to cell pre-virus, we've tried this extensively.  We put cell pre-virus, Luciferase virus, which often is less effective, you know, on the -- and put it on the top of the co-culture and macrophages on the bottom, T cells on the bottom, you know, any type of cell types that would get infected.  And we have seen no infection even when we treat the barrier with some immunomodulators.

So we don't think cell-free virus is playing a significant role, although there is literature suggesting that the endothelial cells may take it up and spit it out.  And it could be that the time frame in which we did our assays did not detect that, but I don't think it's a significant contribution, at least for our model.

DR. SPUDICH:  So going back to Dr. Akiyama.  "Have you examined interferon signaling pathways in your model?

DR. AKIYAMA:  Yeah, that's interesting thing.  So we haven't done -- had a detailed mechanistic analysis using IPS-derived microglia, but when we use the macrophage as a model, so type I interferon pathway plays a major role.  So, for example, if we inhibit a type I interferon in the supernatant by using, like, neutralizing reagent, we don't see lots of different interferon responses.  So we are planning to do -- like, block or investigating those interferon pathway in this 3D culture model, whether or not, you know, type I interferon, for example, plays an important role in neuropathogenesis in the future.  But yeah, it's a great question.  Yeah, nice suggestion.

DR. SPUDICH:  New directions, right?


DR. SPUDICH:  And question for Dr. Ndhlovu, and I think you're typing in a response, but let's answer it verbally.

DR. NDLOVU:  Okay.

DR. SPUDICH:  "Have you looked for an association with cognitive trajectories in HIV DNA in monocytes?

DR. NDLOVU:  Yes.  So this acute cohort, the strongest association with HIV RNA at baseline, we have not seen a robust signal with the HIV DNA levels, and we have both atoned with integrated levels.

DR. SPUDICH:  Okay.  Another question for Dr. Berman.  "Would these targets also affect transmigration of T cells, essentially CTLs, which could eliminate HIV-infected cells."  Interesting question.

DR. BERMAN:  Very interesting.  T cells don't have JAM-A, and the antibody that we're using for ALCAM -- does not target the region that T cells express.  T cells are missing a region of ALCAM compared to monocytes, so we deliberately chose that antibody.  It's a great question.  I will comment that it's really important, we don't want to eliminate all cells from going.  We don't want to get into the reactivation of other infectious diseases, so it's a very important question.  We want to be able to let there be baseline trafficking and learning surveillance.

DR. SPUDICH:  Absolutely.  Another question for you, Dr. Berman, "And this could potentially be accomplished through your collaboration with Dr. Morgello with the Manhattan Brain Bank.  Have you looked at virus DNA and CD-positive, CD16 monocytes by in situ hybridization in the brains of people who died HAND?"

DR. BERMAN:  We are in the process of addressing that question, and we're also going to correlate the transmigration of infected cells from this population with cognitive impairment imaging.  And so a lot of those questions will be addressed once we break the code.

DR. SPUDICH:  Okay.  There's another question for Dr. Akiyama, and maybe this will be our final question before we wrap up before the break.  "You mentioned that p24 positive microglia are not doing phagocytosis well enough.  Wonder if p24-positive microglia represent UNIKA cell population.  Is p24 supposed to be depleted in infected cells?"

DR. AKIYAMA:  Right.  We gated on -- you know, when we did, like, psychometric analysis, we gated on live cells.  So it is possible that those cells are, you know, either -- it can be for a sub-population which results in, like, but it impairs functions, but also it's possible that these are cells that can die.  But we stopped the experiment in -- on day six of infection, so it's possible.  Both ways are possible, yes.

DR. SPUDICH:  Maybe I'll end with one final question, which is sort of a challenging one for Dr. Ndhlovu.  "Current models suggests monocytes enter the circulation for a day or so before migrating to tissues.  If monocytes have a short half-life, there should be continuous CD1 virus.  How might that seeding occur under ART?"

DR. NDLOVU:  That's a fantastic question.  In fact, we've had -- that's been a conundrum for us for some time.  Clearly the bone marrow hypothesis where perhaps there's continuous seeding by compartment is one that is of interest to us.  We've also recently shown, and I think somebody spoke about this earlier, was the link with platelets.  We've actually shown persistence of virus.  It's not replicating, so it's sort of maybe as a carrier for virus in the setting of ART that may go back and forth between the myeloid, but in the peripheral monocytes.  They're almost hitched at -- they're limbed together which is seen as by electron microscopy.  And then third, I think the discussion about monocytes sort of having a very short half-life, it's -- you know, we have to take that with a pinch of salt.  I'm always trying to bring the question as to, you know, if they are transitioning into other cell types and different issues, you know, is that considered the life span a little bit longer.  I think -- I think that could be brought into question, and we should think about other ways by which myeloid persistence might be driving trafficking into different tissue compartments.  But good questions.

DR. SPUDICH:  So yeah.  I mean, wonderful kind of big picture question which I think is relevant to everyone's talks.  So this was an absolutely fabulous session.  I really, really want to thank you all for putting so much thought and, you know, work into pairing these talks, and to be active participation of the audience and all of the questions that we got.  I think that I'll end by thanking you all and announcing that we now have a 30-minute break.  I will be able to let my dog outside so she'll stop whining.  All of you heard her.  And we really look forward to the afternoon sessions which will restart at 1:50 p.m.  So thanks again to the terrific speakers, and we'll look forward to further discussions after the break.  Thank you.

DR. BERMAN:  And thank you, Serena.


DR. JOSEPH:  Okay.  Welcome back from lunch.  Coming up is the second session from awardees of the "Role of Myeloid Cells in Persistence and Eradication of HIV Reservoirs from the Brain RFA."  I just wanted to mention that 16 applications were funded from this RFA.  You heard from half of them yesterday, so today we'll hear from the rest of the RFA awardees.  So I just wanted to briefly mention that the research goals and objectives of this FOA are to contribute to the knowledge and understanding about how myeloid and microglial cell populations contribute to HIV persistence and/or viral rebound.  Mechanistic studies involved in the establishment, maintenance, and resurgence of myeloid reservoirs in relationship to effect and timing of ART and strategies to target myeloid reservoirs were also encouraged.

Now, I would like to introduce my co-chair, Dr. Amanda Brown from Johns Hopkins University.  She's also a recipient of an award from this RFA.  She will be giving the first talk, but before she does that, she will introduce the other speakers.  Over to you, Amanda.

DR. BROWN:  Thank you so much, Dr. Joseph.  It's a pleasure to be here, and as you say, it's been a fantastic meeting.  I would like to introduce our speakers for this session.  After me there will be Antoine Chaillon from the University of California-San Diego, followed by Sanjay Maggirwar from George Washington University, then Michal Toborek from the University of Miami, followed by Melissa Churchill from the Royal Melbourne Institute of Technology in Australia, and last will be Guochun Jiang from the University of North Carolina-Chapel Hill.  Thank you.

Hello.  My name is Amanda Brown.  I'm an associate professor here at Johns Hopkins University in the Department of Neurology and Neuroscience.  Today I'm speaking on, "Toward Understanding the Role of Adult Human Microglia in the Ongoing Persistence of Adult HIV and its Associated Neuropsychiatric Comorbidities."  I'm going to first tell you a little bit about the gap that we're addressing in this new study, some of the advances of my colleagues that have allowed us to embark on this innovative research, and give you some idea of the aims and the support of data that supports this direction.

So we really want to get a deeper mechanistic understanding of the role of adult human microglia in the ongoing persistence of HIV in the central nervous system, and we believe and know that this is likely a barrier to a cure.  And in both persistence and neuropsychiatric comorbidities, neural information plays a role, and both inflammation in the periphery as well as in the brain is important.

We know from early studies that HIV -- a collection of studies, human as well as animal, that HIV gets into the brain very early.  Microglia is one of the longest-lived in the brain, and so we know that and believe that HIV can be a significant reservoir at this site.  However, it's been difficult to really get at the mechanistic knowledge of this host-pathogen interaction because of this difficulty of studying the brain.  So most of our knowledge comes from these beautiful fate-mapping studies using mice, and we know that microglial that come from the yolk sac are the ones that are early -- come early during embryonic differentiation, and later as the fetus develops the fetal liver, and then the bone marrow also make contributions to the early myeloid progenitors that go into the brain.  But their transcriptional signatures differ, and so we know that there's this heterogeneity.

And over the last five years, we've really learned a lot about how these cells are specialized at the level of tissues.  They can be self-renewed, but in cases of injury, you can get infiltration of these progenitors from the bone marrow.  And so in the last five years, there have been these advances in understanding how monocytes derived from bone marrow can be differentiated into microglial-like cells given the right signals.

The other advanced is these humanized mice that, in fact, are proof of principle that blood-derived myeloid cells going to the brain can differentiate into nice, healthy microglia in the presence of human IL-34.  And so our first aim will be to characterize the extent, and breadth, and the fidelity of human in vitro-derived microglial cells to go to the brain and populate and serve as reservoirs for HIV.

And we've been able to do this.  We have 14-day-old cultures here -- shown here, and you see the diversity of phenotype as well as these processes on our microglia.  We can keep them in culture for months, and so we know that these cells then can be used for studies of viral reservoir formation.  They express the receptors that we expect, and additionally they are phagocytically active, being able to phagocytize amyloid beta.

And so some of the quantitative measures, will -- we will assess the complexity arborization of our microglia.  We will compare and contrast both the in vivo and the in vitro transcriptomes to what is available also in the field and what has been submitted through other studies and so we can understand how representative these microglia are of human macrophages.  And we'll use the exquisite techniques to assess HIV, RNA, and DNA in these cells.  Additionally, in Aim 2, we will have an in vitro model in which we will assess the hierarchy of proinflammatory signaling.  And this is based on our recent studies and actually a body of work showing that osteopontin is actually a regulator of proinflammatory signaling.  And we have an in vitro study using human primary-derived macrophages, and we found that when -- in the context of viral infection, when HIV is inhibited or -- sorry  -- when we inhibit osteopontin expression, we see upregulation of proinflammatory signals, like TNF alpha and interferon gamma.

And we'll be in Aim 2 defining the kinetics of proinflammatory signaling.  We will be assessing the role of biologics sex because we know that microglial function can vary in a sex-dependent manner.  We will also quantify the expression of several of the mediators of the TNF alpha pathway, both a soluble and membrane-bound receptors, and we will be able to dissect the contribution of HIV infection versus osteopontin through the use of selective inhibition of osteopontin.

So I want to thank Dr. Jeymohan Joseph, Dianne Rausch, Maureen Goodenow for supporting this work over the years, and for the opportunity to present and to have this fantastic workshop.  That's the end of my talk, and we will go on to the next one.  Thank you.

DR. CHAILLON:  Hi.  First, I want to thank again the NIH for funding our proposal and for this opportunity to present work entitled, "Brain Myeloid Cells are Sources of HIV-Associated Damage and Viral Dispersal."  As most of you know, persons with HIV continue to have HIV-associated neurological dysfunction even during ART treatment, and there are no complementary therapy that can lessen HIV-associated damage in the CNS.  And prime suspect of these damages are brain myeloid cells, which harbor most HIV in the CNS.

Our proposed solution is to collect analyze tissue across the CNS of persons with HIV enrolled in the unique Last Gift cohort, an ongoing rapid autopsy study consisting of altruistic, terminally-ill person with HIV.  We will examine in our proposal the role of brain myeloid cell in HIV persistence and its impact on local inflammatory-induced damage, and their role as a source of viruses that can egress from the CNS to re-seed peripheral organs.

Overall hypothesis is brain myeloid cells contribute to HIV persistence in the CNS with local heterogeneity and serve as a source of HIV brain damage and viral dispersal when treatment is interrupted.  We also hypothesize that these mechanisms are associated with ARV and opioid level.  Cohort study is designed is as follow.  We proposed cohort of 20 participants from the Last Gift cohort.  Five of them decided to voluntarily stop their treatment.  We'll collect brain tissues across the CNS with the location listed here, and we'll additionally collect ileum, spleen, blood, CSF, and other tissue as part of the Last Gift cohort.

The proposed experiment includes ddPCR and HIV full-length single genome sequencing to map the CNS reservoir and to develop phylodynamic model to regulate viral dispersal within the CNS and across the blood-brain barrier.  We also propose to perform HIV integration at sequencing to evaluate the role of clonal expansion in HIV persistence, and perform single cell RNA-seq to evaluate how inflammation and differential expression of inflammasome genes in the CNS relates to HIV reservoir characteristics and to local blood and opioid level.  Finally, we propose ex vivo functional assay to determine the state of post-translation inflammasome activation of bone marrow cells.

Briefly, in Aim 1, we map CNS HIV reservoir in relation to BMC distribution, drug and opioid level, the main hypothesis being that CNS area with higher density of microglial cells, lower drug level, and higher opioid level will have higher level of intact HIV DNA and RNA and higher level of HIV activity.  We already generate some preliminary at the tissue level, and here I'm just showing you the ddPCR in two participant, and we clearly showed heterogeneity of HIV DNA and RNA level across the brain in these two participant.

We were also able to generate single genome sequencing that are in this participant and in just genome from one participant, including sequencing from the CNS.  And you can find more information in the recent Journal of Clinical Investigation paper here.  And two, we propose to determine the role of brain myeloid cell as a source of HIV dispersal, the hypothesis being that the brain myeloid cell carrying HIV can migrate within the CNS, and that HIV is able to egress from the brain toward the periphery when ART is interrupted in relation to local drug and alveolar.

Using env sequencing from bulk tissue and/or validated by Bayesian phylodynamic model, we were able to show that it can migrate within the CNS.  Like for example, in Aim 1, it is participant from (inaudible) on the left throughout the Basal ganglia on the right, but also across the blood-brain barrier.  A wonderful collaborator from UNC, Dr. Jiang, was also able to get from rhesus macaque more than one million cell/gram of brain tissue consisting mostly of 90 percent of microglia but also up to 10 percent of macrophages.

In Aim 3, we propose to evaluate the size and activity of brain myeloid cell reservoir in relation to local inflammation, and the hypothesis being that there  would be relationship between inflammasome expression pathway and HIV reservoir characteristics.  Here I'm just showing you the timeline for the next five years.

Finally, I wanted to present you the wonderful team of expert from UCSD and UNC, including Dr. Davey Smith, Dr. Sarah Gianella, Dr. Guochun Yang, Dr. Hoffman, Dr. Nadia Beliakova-Bethell, and Ronald Ellis.  I want to thank you for attention and thank you all the team, and participants.

DR. MAGGIRWAR:  Thank you so much for inviting me for this talk today.  I'm representing my student collaborators, particularly Natalia Soriano-Sarabia and Mirko Paiardini from Emory.  The goals of our study are to investigate whether platelets harbor infectious virus particles during acute phase of infection, thereby facilitating viral dissemination, promoting seeding and persistence of HIV reservoir in the CNS.

Since platelets do express multiple HIV attachment factors, we speculated whether platelets can serve as a viral host.  So my student, Sydney Simpson, took the plasma samples from viremic patients -- they were treatment live -- and isolated platelets and then imaging cytometry.  As you can see here, two representative samples shows that platelets do contain gp120 inside them.  The super resolution microscopy also detected gp120 inside the platelets.  This was further confirmed by transmission electron microscopy.  In this you can see that intact platelet particles can be found in the open canicular system of platelets.

So as part of this analysis, so we screened multiple samples.  These are matching samples derived from HIV-infected subjects prior to or three months after ART initiation.  As you can see here, nearly five percent of platelets were GFP positive before ART, and after ART this percentage of platelets were dropped to nearly .5 percent.  So the percentage of gp120-positive platelets were strongly correlative to the individual viral load rather than the treatment itself.

Activated platelets are known to form high number of platelet monocyte complexes in HIV-infected subjects.  So typically, if you look at these complexes, you'll find two to three platelets bound to a single monocyte.  This interaction between these two cells is very dense and particularly receptor mediated.  There are two types of complexes formed in the circulation of these subjects:  type I complex where we have distinct platelets attached to the monocyte, and type II complexes where platelets are lost, but some of their components are still expressed on their surface.

So at this point, we asked this question, whether platelets render monocytes permissive for HIV infection.  And so for this experiment, what we did was we took monocytes alone and exposed them to the HIV.  At the same time, we also incubated monocytes with platelets and then infected them with HIV.  And then the HIV replication was followed by p24 (inaudible).  As you can see here, the levels of p24 for -- were sharply increased when -- in monocyte culture that was pre-incubated with platelets, suggesting then that platelets promote HIV replication in monocytes.  These same cultures were followed up for extended period of time in presence of M-CSF for their survival.  In this experiment, you can see that there's a distinct replication state state of latency, and the reactivation of virus can be detected.

We did the next experiment in which we asked whether HIV delivered by infected platelets to the monocyte also can assume latency, and similar experiment was carried out.  In this case, we did see the active replication until day 14, and then the sharp drop in HIV replication, suggesting the latency.  This is a true latency because there is a reactivation by a prior study which is very significant.  So altogether this data suggests that platelets augment HIV infection and latency in monocytes.

So finally, we ask this question, whether pharmacologic inhibition of platelet activation can prevent migration of HIV-infected monocytes into the CNS.  To do this, we took mice, they were evaluated, pre-treated with clopidogrel, and then we IP-injected HIV-infected monocytes in them and followed their migration to the CNS.  As you can see here, the administration of Clopidogrel, it's also known as Plavix.  Administration of this drug sharply reduced the migration of HIV-infected monocytes in this case, and these are the total monocytes, migration of these into the CLS, suggesting that the platelets have multiple roles.  And based on that, we propose that this conceptual model, that platelets promote viral latency.  It promotes the seeding and persistence of HIV reservoirs in the CNS by three steps.

First, the transmission of HIV from latently-active infected platelets to the monocyte happens.  In the second, interaction of activated platelets with monocyte render Want to say it's more promising for HIV infection.  And then platelets facilitate the migration of these HIV-infected monocytes into the CNS where they differentiate into macrophages, and then also help spread the virus to the microbial cells.  So it should be noted that it is challenging to investigate the range of viral transmission dissemination by platelets to uncommon host, and how it may contribute to the heterogeneity of viral resistance in the CLS and beyond CNS.  Also the roles of platelets in HIV infection are dominated by pathologic inflammation, but evident, but not exclusive consequences in hemostasis.

Finally, I am very grateful to my students, my collaborators, and the NIH for funding.  These are the major drivers of the work that I presented here today Thank you so much.

DR. TOBOREK:  Hello.  My name is Michal Toborek.  I am professor in the Department of Biochemistry and Molecular Biology at the University of Miami School of Medicine.  Dr. Mario Stevenson is co-investigator on the grant.

So the main interest of my laboratory is on the blood-brain barrier and specifically the involvement of the blood-brain barrier in HIV infection.  And one of the most mysterious cell type of the blood-brain barrier and probably the entire brain are parasites.  Those are neural cells which have brain epithelial cells in approximately 90 to 100 percent.  Okay.  They are specifically covering endothelial cells on the brain capillaries, not on the larger vessels, okay?  Larger vessels are covered by smooth muscle cells.  As I mentioned, parasites are mysterious cells.  They hey are of different origin in the brain.  However, recent evidence shoes shows that substantial subpopulation of brain parasites actually originates from myeloid progenitors.

So our central hypothesis is short and sweet, and it's just that the brain parasites confirm HIV reservoirs in the CNS.  To study this, we have three specific aims.  Aim 1 is determine whether brain parasites can harbor latent HIV infection in HIV-suppressed individuals.  However, you know, to address latent infection of brain parasites, we first need to have evidence that parasites can be actively infected.

Okay.  So parasites do contain -- do express HIV receptors such as CD4, CCR5, and CXC4, and we have evidence that impacted parasites can produce -- actively produce p24 in here in the cell culture media and also expression by immunocytochemistry.  We also have evidence from HIV-infected mice.  It's very nice co-localization of PDGF receptor beta, which is market of parasite with p24-infected mice, but not so obviously non-infected mice, and co-localization of splice tray with NG2 right here.  NG2 is also market of parasites.

We had very exciting human evidence that parasite infection actually occurs in humans.  So from HIV-infected brains, we isolate micro vessels, and then we demonstrated this very nice co-localization between p24 and PDGF receptor beta.  So this was the first human evidence that parasites can be infected with HIV, and then Dr. King's group independently confirmed -- fully confirmed this result also by co-localization of PDGF receptive beta and HIV p24.  So Aim 1 is ongoing.  We are using a cohort of human brain samples with history of achieving viral suppression, and we're studying the viral -- the presence of viral DNA and viral RNA in brain parasites in those human samples.

Aim 2 is to evaluate transcriptional signatures of latently-infected human brain parasites.  So our preliminary data showed -- actually confirmed the hypothesis that HIV -- that parasites can be latently infected by HIV.  We have this very nice integration of viral DNA with host being parasite DNA during the course of infection, and then we have reactivated virus from these latent phases of HIV infection.

So in Aim 2, we are taking advantage of HIV GKO, a duo-color HIV reporter which can distinguish between uninfected, latent, and actively-infected cells by having fluorescent markers under the control of different promoters.  So make -- to make the long story short, our active -- actively-infected parasites are staying in green, the latently-infected parasites are staying red, and during the course of infection we have this very nice switch from active to latent infection.

We can isolate the pools of -- the different pools of parasites -- actively infected, latently infected -- in the cells in between which we are switching from active from latent infection.  Then we are running transcriptomic studies, and we already identified some of the signatures of actively-infected parasites which are -- is full of parasites, and latently-infected parasites.  We have lots of gene candidates to study and also associated signaling pathways.

In Aim 3, we are looking to evaluate functional outcomes associated with HIV infection brain parasites.  So we are looking at certain interactions between parasites which are acutely or latent infected with other cells of neurovascular unit, primarily endothelial cells because we have these very interesting structures between parasites and endothelial cells.  They are called peg-socket contacts which, like basement membrane, allow the interaction between GAP junction mediators, inflammatory mediators, and trophic factors, you know, crossing from one cell type to another without interference from basement membrane.

The main conclusion of the talk is that parasites account for HIV infection and may be involved in HIV reservoirs in the CNS.  I would like to thank the laboratory, everybody working hard on this project, and acknowledge the great support from NIMH.  Thank you so much for your attention.

DR. CHURCHILL:  My name is Melissa Churchill, and today I'll present an outline of our study entitled, "The Characterization of Intact and Defective HIV Reservoirs in Myeloid Cells in the Brain."  The principal investigators are myself and Dr. Thomas Angelovich from RMIT University and Dr. Michael Roche from the Peter Doherty Institute.

The overall goal of our study is to comprehensively and systematically describe the HIV reservoir in the CNS.  We hypothesize that a subset of myeloid cells in the CNS in our treated subjects will contain intact replication-competent genomes with the potential to be reactivated to produce HIV.  Aim 1 of our study is to quantify and characterize the intact and defective myeloid reservoir.  We use droplet digital PCR for pol to quantify the frequency and size of the reservoir, intact proviral DNA assay to determine the frequency, and size, and nature of the intact and defective reservoir, DNA sequencing to assess compartmentalization of this reservoir, and laser capture and microdissection to phenotypically assess the reservoir.

In Aim 2, we'll characterize the functional capacity of the intact and defective proviruses in the reservoir.  We use comprehensive sequence analysis.  We will clone the intact and defective proviruses.  We'll assess the replication competence and tropism of these proviruses.  And we'll -- finally, we'll assess the transcriptional and translational competence of these proviruses.

To carry out these studies, we've curated a cohort of brain and peripheral tissues from people who died with HIV.  All of these samples were generously provided by the NNTC.  The cohort consists of six uninfected individuals, 18 individuals characterized as viremic and died with detectable HIV viral loads, and 12 carefully selected individuals who were virally suppressed at the time of death.  These individuals were chosen because they had undetectable viral loads in both CNS and plasma for a minimum of two years prior to death.  These individuals also had relatively healthy T cell counts and a median time of viral suppression of 6.2 years.

To determine if a reservoir existed in these individuals, an initial screen was carried out using a droplet digital PCR for the presence of pol.  As you can see, there was no significant difference in the frequency or the size of the reservoir between the viremic and the virally-suppressed people with HIV.  To further characterize these genomes, we used the intact proviral DNA assay, or the IPDA.  The IPDA is basically a multiplexed droplet digital PCR reaction using probes targeting both the side region and the env region of the genome.  In this droplet digital PCR plot schematic, env fluorescence is -- signal is shown on the x-axis and psi-fluorescent signal is shown on the y-axis.  Each dot on the plot represents an individual droplet containing a single PCR reaction.  Droplets present in the top right-hand quadrant in orange represent droplets that contain both env and psi signal.  As these droplets contain both primer sets, they're defined as containing intact proviral genomes.  Conversely, the blue and green droplets contain only psi or env-positive signal and are, therefore, defined as containing defective proviruses, either three prime deleted or five Prime deleted.

Cecile Khana's group showed that the IPD efficiently identifies greater than 90 percent of deleted proviruses, and importantly, it can detect intact proviral genomes at a rate that correlates with the quantitative viral outgrowth assay.  When we apply this assay to a viremic patient, you can see that intact proviral genomes can be seen in orange along with five prime deleted and three prime deleted genomes.  A similar result is also found for the virally-suppressed patients.

This figure shows the results of the IPDA for eight patients, four viremic and full virally-suppressed patients.  The x-axis is the type of genome we were detecting, and the y-axis is the provirus pertaining to the six cells analyzed.  You can see that there's no significant difference in the number of proviruses between the viremic and the virally-suppressed patients, but importantly, intact HIV proviruses are present in the CNS despite viral suppression.

If we then look at the relative proportion of intact proviruses, firstly we see that the majority of proviruses in the CNS are defective with either three prime or five prime deletions as you can see indicated by the blue and the green bars, and approximately 10 to 20 percent of the proviral genomes are intact, and these are indicated by the orange bars.  Secondly, the composition of the reservoir is similar between ART-suppressed and viremic people with HIV.  We've examined the sequences of some of these proviruses, and what we've found is that the CNS is likely to be a phylogenetically distinct reservoir.  The sequences compartmentalize between the lymphoid and the CNS compartments for both the virally-suppressed and the viremic people with HIV.

So in summary, to date we've shown that the intact and defective proviral genomes can be detected in the CNS of people with HIV despite viral suppression.  We've had a number of key challenges.  Firstly, the isolation of non-sheared genomic DNA from frozen CNS tissue is certainly a challenge, and it's critical for all of the experiments that we intend to carry out.  Access to sufficient amounts and numbers of CNS tissue has also been challenging.

I'd like to acknowledge the people involved in this study, first, the Neuroinfections Group that I lead at RMIT University, in particular, Tom Angelovich and Catherine Cochrane; secondly, the Peter Doherty Institute:  Michael Roche, Sharon Lewin, and Matt Gartner; our collaborators, Jake Estes and Bruce Brew.  Our funding our funding bodies, the Melbourne HIV Cure Consortium, NHMRC, and, of course, the National Institute of Drug Abuse at the NIH.  None of this work is possible without the generous provision of tissue samples from the National NeuroAIDS Tissue Consortium or the NNTC.  Thank you.

DR. JIANG:  First of all, we thank NIH for funding our current research, and also we thank the committee for inviting us for this presentation.

So the goals of this study is, first, to determine whether long-lived microglia serve as stable reservoirs in the brain.  These are being investigated with the brain tissue from ART-suppressed animals and tissue "Last Gift" cohort, which will be collected immediately after the patient death.  And second, we want to exploit the mechanisms of HIV persistence in the brain.  For the long term, we want to develop strategy to eradicate HIV reservoir in the brain.

In the last few years, we have developed a protocol to isolate brain cell subsets from different part of the brain, including cortex, hippocampus, Basal ganglia, and with this cell, we want to determine brain reservoirs by measuring reservoir size, pursuing latency in mechanisms, and development of the cure strategies.

This are some of the example that we collected in last few years.  We first purified brain mono-cell with CD-relevant antibody for the sorting.  With that, we discovered that more than 99 percent of CD cell are positive.  And further, we found that more than 93 percent of CRR, CD45 are low.  And while less than five percent of cell express high level of the CD45, including the majority of these cell are microglial cell, and less than five percent of the cell are macrophage.

During this process, we analyze with high-sensitive RT-qPCR analysis, and we show that those purified brain myeloid cell are free of T cell.  And then we started to isolate microglial cells using a similar protocol from ART-suppressed animals, and we found that, as you see here, that those microglial cell express highest -- high specific -- highly-specific microglia cell marker called TMEL19 where more than 99 percent of cell express (inaudible) microglia markers.  And we found that during in vitro culture, those cell can proliferate very well.  Meanwhile, cell viability are pretty much close to 100 percent of reliability.

And then we decide to determine the frequency of SIV reservoirs in this isolated brain microglia, and we found that, first of all, this SIV retroviral DNA co-localized with CD68 protein marker.  Second, we determined that the frequency of total SIV proviral DNA -- of the integrated virus DNA ranged from 100 to 1,000 copy in the microglial cell.  And lastly, we discovered during in vitro culture from passage zero to passage 9, SIV proviral DNA keep constant.  So those preliminary data indicate that this microglia ex vivo model could be a very good, you know, tool for us to study HIV latency, including mechanism of HIV latency establishment in the microglia and possible care strategy, and it can be developed with this model.

For example, for in vitro, we found that SIV RNA can be reactivated in this isolated microglial cell by several latency-reverse agent, including HDAC inhibitor, SAHA, histone methyltransferase inhibitor, GSK343, and (inaudible) including TNF-alpha and AZD582.  As a representative picture, we show here that SRA RNA can be induced in this microglia cell, which expressed as CD68 marker after reactivation with ACD582.

And lastly, we developed our growth assay that a co-culture of the microglial cell with CEM174 T cell, and we found that the frequency of this regular size ranged from 10 to 30 units within the microglial cell.  And we did realize that there's some challenge ahead of this study.  First, there may be some challenge isolate and grow this brain cell from the tissue of these people living with HIV.  Second, the levels of some brain cell subsets may be relatively low, such as CD4 pol cell.  And lastly, the frequency of the HR reservoirs may be null in a brain in which one protein, such as p24 or p27, could be very hard to measure by QVOA.

So this project has been collaborated with many lab in UNC Cure Center and across this country.  Initially, we collect the tissue from Dr. Huanbin Xu's lab in Tulane National Primate Center, and Dr. Dennis Hartigan-O'Connor's lab in UC Davis.  And currently, we have obtained a few tissue sample from Dr. Ann Chahroudi's lab in Emory University.  Meanwhile, we are starting to collaborate with Sara Gianella's lab in UC San Diego for "Last Gift" cohort.  And most recently, we started collaborating with Dr. Victor Garcia in UNC Lineberger Comprehensive Cancer Center, and Dr. Wenhui Hu's lab in Temple University to try to develop a new gene editing system that specifically targeting HIV reservoir in the microglia.  Thank you very much for listening.

DR. BROWN:  Fantastic.  That is the end of the recorded sessions.  I'd like to invite all of the speakers to please turn their cameras on, and we will begin the question and answer session.  And those in the audience, please feel free to continue to type in any questions that you may have.

So I'll start.  There is a question about -- to me about major sex differences.  "What are the major sex differences in microglial function?  Could you elaborate on this and which particular sex difference do you expect will play a role in HIV infection?"  Well, I have to admit this is a new study, so we haven't dived deeply as yet into this.  But we do know that during development, microglia play an important role in the masculinization during development.  But moreover, we know the role of sex hormones, for example, in driving differences in inflammatory processes, either protection or exacerbating inflammation, and so those do differ between females and males, so it may be somewhat difficult in our -- the experimental model.

I have thought about this, right, because the donor cells will be from a particular sex, but then our recipients will be one sex or other.  But this also follows up on work -- our most recent studies where we inhibited osteopontin expression, and because of experimental limitations, we were only able to assess female mice.  And then we saw in this study that the information was exacerbated when osteopontin expression was down modulated.  And so we will -- we'll be interested to look at differences between males and females.  At least based on our analyses from our immunohistochemical studies, you know, we didn't see a striking difference in males, for example.

Okay.  I'm going to go to the next question, and this is for Dr. Maggirwar.  "How does HIV permeate into platelets?"

DR. MAGGIRWAR:  Oh, thanks for asking this question because I deliberately skipped those slides in my talk.  So platelets do take up HIV naturally.  Basically the platelets kind of don't really use the standard steps for HIV replication and infection.  So the steps like fusion, or the encoding of the capsids, or RT, or integration, none of that happens in platelets.  What they do is they naturally soak HIV from the surrounding, and it's not in the -- in the platelets.  It is more like in the folds of membranes that is called open canalicular system.  It is inside the platelet but away from the cytosol.  And typically, platelet activation is linked with this uptake of HIV.  If you block platelet activation, you block the interaction with HIV.  And if you look at the clinical data, so we showed that the number of HIV-positive platelets correlates with the viral load but not so much with the CD4 T cell count.  So kind of it's unrelated to the immune responses.  It's more Or at the platelet level.  Thank you.

DR. BROWN:  Okay.  The next one is for Michal.  "Very interesting talk.  Can you comment if infected parasites contribute to the inflammatory responses in the brain?"

DR. TOBOREK:  Yes, they do.  So I'm sorry.  I'm just checking the question.  Well, so we don't have in vivo evidence from humans, but we have in vitro evidence from cells.  And parasite-infected cells definitely produce inflammatory responses.  Interleukin 6 was the cytokine which was produced in the highest extent by infected parasites.  And then when we took the supernatant-conditioned media from infected parasites and we put them into endothelial cell cultures, we could disrupt integrity of the endothelium.  So I would say that they definitely contribute to neuroinflammatory responses in the brain.

DR. BROWN:  Thank you.  The next one is for Melissa.  "How do you confirm that the intact virus you detect in brain tissue samples by IPDA originates from myeloid cells in the brain?  What is the probability that a large fraction of them you detect are from infiltrating CD4-positive T cells?"

DR. CHURCHILL:  So that's something we've considered, and some of the -- some of the ways that we're looking at that is we've tried to stain the T cells for virus, and we haven't seen any infected T cells in the brain.  But secondly, we've used laser capture microdissection to isolate microglia, and we found that the -- most of the virus that we've seen comes from the microglia or the macrophage population.

DR. BROWN:  Could you expand on that because LcnD has been with us for a while.

DR. CHURCHILL:  Oh sorry.

DR. BROWN:  And I was wondering if there's some -- you know, been some innovations because that was a criticism in the past.

DR. CHURCHILL:  Well, we're only -- we're only -- we're only looking at small fragments of the virus.  We're just trying to detect the virus, so we look at V3, and that's very -- that's actually very efficient.

DR. BROWN:  Okay.  Okay.  The next question, again, for Dr. Churchill, "Do you have PBMC data for intact effective proviruses viruses for the cohorts?  Any correlations with what is seen in the CNS?"

DR. CHURCHILL:  So what we do have is for a subset of the patients, we have a the PBMC data or lymphoid tissue data.  And what we see is that there's a very similar patent of intact to deleted viruses, but there's certainly a larger reservoir in the lymphoid tissue than in the PBMCs -- than in the CNS.

DR. BROWN:  And maybe why don't we just stay with this?  So you mentioned your cohort, that you've been very critical or focused on those that are suppressed, right, selecting those patients.  Can you compare/contrast a little bit what -- how your studies going forward differ from what we have learned in the past about compartmentalization because you didn't see any difference in the level of intact proviruses viruses between the treated and untreated -- treated and treated -- untreated groups.

DR. CHURCHILL:  That's correct.  We didn't see any difference.  I'm not quite sure what you're asking.  Sorry.

DR. BROWN:  Well, does it relate to the antiretroviral therapy because it suggests that --

DR. CHURCHILL:  Most likely.  Yes, most likely.  Yep.  We haven't seen any difference at all in the level of pol or in the level of -- or the ratio of intact to defective proviruses viruses between viremic and aviremic.  We haven't looked at encephalitic patients.  We've just used viremic patients.

DR. BROWN:  Thank you.  Okay.  This is for Michal, Dr. Toborek.  "We and others have previously shown that the transcriptional latency largely does not exist infected T cells or myeloid cells.  This was shown by short TAR non-coding RNA in infected latent T cells or myeloid cells.  Is there Basal transcription in latent, and this with cART treatment, in infected platelets or parasites, that is, the presence of short TAR RNA?"  So do you see short TAR RNA in your parasites?

DR. TOBOREK:  We didn't look at them.  We know that parasites, like some other myeloid cells, they maintain basic -- very low HIV replication.  So the latency in parasites and microglia, they never go into complete dermis, okay?  There is always -- in our hands there is always some simplification going on.  It's different between myeloid cells compared to T cells which can stop HIV replication completely, but in our cells, we have very low -- very low p24 production.  It never goes into zero like with macrophages, for example.  The same -- the same for myeloid cells.

DR. JOSEPH:  Thank you.  For Dr. Jiang:  "You mentioned that isolated microglia were proliferating well.  How long can you grow microglia in culture?  Are those resistant -- the latency-reversing agents toxic to microglia or neurons?  And could you comment" --

DR. JIANG:  First of all --

DR. BROWN:  Yeah, go ahead.

DR. JIANG:  Sorry.  First of all, this cell can be accrued outside at least, like, two weeks.  We have -- after that we freeze that in the (inaudible), and then we definitely can regrow them.  We still don't know exactly how long they can grow.  The maximum time, we don't know.  Second, the latency-reversal agent that we use in this study now is not toxic to the cell, but we know some of the (inaudible) agent is definitely toxic to either microglia cell or the neuron.  We haven't shown here yet.  And then for the last one, we do not know.  We do not know for some of the RNA.  And we know that RNA have some, you know, issue can efficiently, you know, across the pb, but for this latency-reversing agent, we don't know.  Most of them never tested in the brain yet.

DR. BROWN:  Mm-hmm.  And so your microglia -- just going to that proliferation again, do they -- over time do they go to senescence?  Do they go to a more resting state?

DR. JIANG:  We don't -- we haven't test whether they are in a resting state yet, but they are pervading.  I don't know.  When you are talking about resting for the microglia, I'm not totally understand what that -- what that mean.

DR. BROWN:  Right, because in the brain, right, in the nervous system, unless there's some injury or some replacement --

DR. JIANG:  Yeah.

DR. BROWN:  -- not proliferating.

DR. JIANG:  Yeah.  I mean, that's a sum up.  I saw there's a lot of people ask this kind of question, you know.  You know, we don't know whether actually reservoirs rely on this.  We don't know that.  Second, particularly for this study, the goal is that we try to develop ex vivo, too, to address the critical question that whether, you know, latency can be established in microglia and whether microglia cell are the stable reservoir.  Actually, you know, we already detect HIV DNA/RNA in either our suppressed animal or our suppressed patient in the brain tissue.  We haven't shown here even an RNA or DNA.  We don't claim to see that.

DR. BROWN:  Excellent.  Thank you.  Dr. Churchill, "What mechanisms drove the compartmentalization you observed between the periphery and the CNS/HIV/DNA populations between the study groups?"

DR. CHURCHILL:  Goodness.  That's a question that we've been trying to answer for 30 years.  I seriously don't think that we have an answer for it yet.  So don't know.  Art, immune pressures, I don't know.

DR. BROWN:  Yeah.  Now we have the glymphatics.

DR. CHURCHILL:  That's right.

DR. BROWN:  We have the SCOLE bone marrow.

DR. CHURCHILL:  I'm hoping someone else will answer that for us.

DR. BROWN:  Okay.  Dr. Toborek.  Let's see.  You didn't answer.  "Are there differences in differentiation states of parasites between mice and humans that may influence the permeability of the blood-brain barrier and perhaps the dynamics of HIV-infected cell populations between these models?"  So they're different in differentiation states between mice and humans.

DR. TOBOREK:  I'm not aware of differences in differentiation of parasites.  The same with differentiation of neurons or astrocytes.  I think the cells will differentiate the same way and they would behave the same way in mouse brains or human brains.  So I'm not -- I'm not aware of any differentiation.  This being said, you know, the tools are primarily based on the mice, and then so when we do -- when we do ECO HIV Infection Model, we use mice.  When we do in vitro model, we use human parasites infected by HID.  And when you -- when we take the human tissue, obviously we have human brains.

We didn't observe any differences between those different models, so when we have infections out in in in vitro, we have infection in HIV-infected brain, and the same week with EcoHIV in mice.  So I don't think there are any differences, and, you know, taking into consideration different brains, we see very similar responses.

DR. BROWN:  Because you mentioned parasites, right, have some type of plasticity.  But whatever your culture conditions are, their phenotype and I guess genotype stays constant.

DR. TOBOREK:  Yes.  Yes, this is what I would say, yeah.  I don't see differentiated -- differences in differentiation of parasites.

DR. BROWN:  Okay.  Great.  Let me just see this question here.  I guess another attendee.  "The parasites you're using are primary human brain parasites," so you said yes.


DR. BROWN:  Mm-hmm.  "Are those cells dividing often, and what is their half-life of these cells?  Do they express CCL2 during HIV infection?"

DR. TOBOREK:  They do express CCL2, they express nicely chemokines, and they respond to HIV infection by producing a spectrum of inflammatory mediators, including chemokines, including CCL2.  They proliferate nicely, so we get them from a company.  They proliferate nicely.  The proliferation rate would be like typical for cell culture.  We don't see any differences in proliferation between parasites and astrocytes and epithelial cells.  You know, the density when we use for splitting cells approximately the same, and then it takes three, four days to generate confluent cultures.  Now, with the primary cells, like with all primary cells, we cannot go forever with passages, so we stop approximately seven, eight passage of parasite, and then we take a new batch of this.  So we -- they're not -- you know, they are not immortalized.  The lifespan would be limited to seven, eight passages.

The good thing is that when we do parasites from different batches, and we already work with parasites for many years, so we -- by definition we have different batches of cells from the company.  The responses are very similar, you know.  So from -- it's very constant from experiment to experiment, from year to year, and from investigator to investigator.

DR. BROWN:  Thank you.  Okay.  We have another series of questions here.  Let me find it here.  Oh sorry.  Forgot about Dr. Chaillon.  "The data supporting the viral egress from the brain to the periphery is sketchy for two reasons.  Highly-mutated viral quasi-species originated from the infected cells in the CNS are hard to track down in the periphery, and two, the minuscule contribution of brain-derived HIV to the overall viremia originated from T cells/macrophages in the periphery.  How will you overcome these difficulties in your studies?"

DR. CHAILLON:  A good question and exciting challenges, but I think we have the approach to overcome these challenges by doing single genome HIV sequencing in these genome sites.  We can data mine the intactness and -- of the viruses that will be feeding Bayesian model.  And we also generating a lot of sequences across the body in various tissue sites from the Last Gift award.  So we should be able to account for these by sources of viral rebound across the body, including the CNS.  So I think we have to the proper assay to look into it, and we are also considering doing match integration site and sequencing to track down sex trafficking within the CNS and across the wire.  These are challenges.

DR. BROWN:  Let me just turn off my microphone.  The next question is for me.  "Do you think there is an interplay between infected-circulating CD4 positive T cells in myeloid reservoirs and inflammation in the CNS?"  So definitely, you know, in the context where there is escape, so that there is not full suppression of HIV replication in CD4-positive T cells, definitely we know that there's this impact in communication of soluble factors between the periphery and the central nervous system.  So it could definitely have an impact in the case where antiretroviral therapy is suboptimal.

And another part of that question is, "In the CSF, we detected" -- and this is from Dr. Spudich.  "In CSF we detected HIV-infected T cells that may be constantly trafficking to the brain even during ART.  Can you examine this contribution in your study?"  That is a terrific suggestion since in our models we -- there will be CD4-positive T cells there available, so we could definitely look at that.  We will be using from Santhi Gorantia the IL-34 mice, but I don't have experience with those mice.  We've used the regular NSG mice, and we did do some studies to try to -- flow cytometry to look for immune subsets in the brain in our prior studies and didn't see a whole lot of T cells.  But perhaps with IL-34 and enhanced, you know, differentiation of the human cells, we could look at that.  Thank you.

This is for Dr. Toborek.  Let's see.  I think this is different.  Okay.  "Can you comment on the potential relationship between parasite infection with HIV and markers of vascular inflammation that have been observed in people living with HIV?  Do parasites that surround capillaries contribute to vascular information?"  We'll stop there.

DR. TOBOREK:  Yeah.  So I think it was similar question before.

DR. BROWN:  Okay.

DR. TOBOREK:  Infected parasites are producing the spectrum and inflammatory mediators, inflammatory cytokine, CCL2, and they definitely would contribute to vascular inflammation and probably also neuroinflammation which will be associated with HIV infection in the brain.

DR. BROWN:  Okay.  Good.  I think Dr. Maggirwar is going to get the last question.  "Did you check the CD4-positive T cell level of the patients you studied?  Indeed, we recently showed that platelets from immunological responders, those who have sustained low CD4 and are virally suppressed, harbor replication-competent HIV in their platelets.  This virus can productively infect macrophages in vitro in a process blocked by the drug that blocks one of FAB."  Okay.  "Did you check for other viral components in your platelets?"

DR. MAGGIRWAR:  That's a very good question, and, Morgane, I'm familiar with your STM paper.  It is very nicely done and extensive work.  So to your question, we did look at the CD4 T cell count before the antiretroviral drug treatment and three months after the ART.  And before the ART, it was somewhere around 400 cells per microliter, whereas with the three months of ART, it led to the recovery of CD4 T cells up to somewhere around 700 cells per microliter.  And so this is very interesting question, though, that -- whether that is linked somehow with the persistence of HIV in those platelets under the suppressive conditions.  And so we don't know the answer for it yet, but together we can find out.

So your next question about the other compartments in the platelets, so there are a lot of advanced granules and other granular compartments in the platelets.  So we did the transmission electron microscopic image of the entire platelets, and where we found HIV particles were only in open canalicular system, which is not really in the platelets.  It's somewhat outside.  So don't know.  Thank you.

DR. BROWN:  Thank you.  This is the end of our session.  I want to thank all of the speakers.  It was an excellent, dynamic discussion section.  I don't know if I turn it back to Dr. Joseph, but we are --

DR. JOSEPH:  Yes.  So we'll have a 10-minute break before the closing final session.  Thank you all, yeah.


DR. CLEMENTS:  Hello, everyone.  That was a terrific session yesterday and today, very exciting things we discussed and talked about.  And we're -- the end session will be to really ask each of the session chairs to give a short overview of what the most important messages were from their session.  And then at the end, Kiera and I will talk about some things that we've -- we saw that perhaps would add to that, points of discussion that we feel that might add to that, unless, Kiera, you want to do that first.  I think we should let the groups go, and then if they cover our -- some of our points, then we'll just use the ones that are left, okay?

DR. CLAYTON:  Okay.  So I will start.  So just a brief overview of Session Number 1.  We focused on macrophages and the immune system during HIV, and we did have one talk on SARS-CoV-2 infection, so thanks to Rahm Gummuluru.  So the key takeaways from this really is that, you know, in the case of HIV, we see differences in HIV's interferon sensitivity that are observed, you know, during the transmission, during the chronic stage, as well as rebound infection, and that, you know, from the first talk, we saw that interferon may actually enhance latency in macrophages, but really, you know, interferon potentially -- or how interferon resistance potentially plays a role for latency in different stages of infection.  And I think that's still an outstanding question.

We also looked at innate immune sensing of viruses and how that can contribute to elevated levels of inflammation.  And of course we saw this with multiple talks for HIV, but of course from Rahm's talk we also saw that for SARS-CoV-2.  And then there was a nice discussion at the end in terms of how this -- the inflammation could potentially contribute to inflammation-related comorbidities, such as cardiovascular disease.  So that was a very nice set of talks to really prop up how macrophages in the immune system are stimulated by HIV and other viruses.

And in terms of future research, you know, we think it should focus not only on defining mechanisms of interferon resistance and the induction of latency, but also the downstream effects that contribute to macrophage-mediated inflammation in the context of different comorbidities.  You know, this may contribute to the development of new therapies to target inflammation in the context of HIV as well as SARS-CoV-2 and potentially other pathogens that infect macrophages.  So this may not be pathogen specific.  It could be, you know, obviously the host response.

We also have to think of other pathogens that infect macrophages and use them as hideouts, such as Ebola virus.  That infects macrophages.  Mycobacterium tuberculosis infects macrophages.  Cytomegalovirus, which is another chronic infection, infects macrophages, and there's a whole -- there's a whole list that we could go through.  So just something to think about.  And the last thing that I wanted to point out was, you know, macrophages don't really exist.  Especially in tissues, they don't really exist in a vacuum.  They constantly interact with different cells of the immune system.  And how the presence of HIV in this long-lived reservoir can affect other cells of the immune system is something that we need to study as well.

So with that, I'll hand it over to Tom Hope, who's going to give us an overview of Session 2.

DR. HOPE:  So our session was, I think, focused on a number of new approaches and tools that are being developed to try to study these reservoirs.  And I think that one -- some of the themes that come out of this is really the need to focus on primary cells, especially some animal models and human tissues.  These studies are more challenging, but they really are getting at the relevant cell types.  The context there is really quite important in terms of the anatomy and physiology and the influence of that, the local environment on these reservoirs, and also sort of -- a little more of a systematic evaluation, I think, of the infections in the CNS.  I think there is a lot of great work out there, but there's a lot more work to be done.

And so overall for, I think, future plans, I think one of the things that came out throughout the meeting was that, by their very nature these myeloid cells are tissue resident and tissue specific.  I liked what Kiera just said about interacting with all their partners, but they're also influenced by their environment.  In each environment, the inflammatory state of that environment, other circumstances of that environment are impacting those cells.  And although it's convenient, I think that we all have this model we use to -- for something we call a macrophage or sometimes a little bit different, called the dendritic cell, to kind of re-evaluate those models and see how much that is like the cells we're really interested in, say, in vivo.  And maybe some of some of us will be happy about that and some of us would be horrified.  So but I think we need to understand that part to try to understand to do these things in vivo and to really get a sense of the relevance.

And I think the other thing is that we need to be less dogmatic maybe a little bit because these ideas like these cells can never divide, that's not true.  It's very much not true.  And it's just like neurons don't divide after we're teenagers.  These things stick, and -- but we should let them go.  So we have to appreciate that these are complex populations that can originate in different places, and they can self-renew and divide.

So all that being said, it is so exciting to have been part of this meeting, to see all the great science, and to see that we are moving forward to solve big problems.  And then I will pass it on to Rebecca.

DR. VEENHUIS:  Thank you.  I'm Rebecca, and I moderated the session with Janice.  The big takeaways from this session, it was a nice overview of a lot of the grants that were funded through the RFA, and it sort of directs where we're going in the future and what lines of inquiry are currently open.

But I think a lot of the key takeaways were that we just -- we need better ways to identify these infected myeloid cells that could potentially contain replication-competent virus, particularly in the brain, and a lot of the methods suggested in the -- in the grants are working towards this avenue.  We also need a better way to understand how viral infection of myeloid cells could potentially induce a permanent latency.  Why do macrophages shut down viral infection and still maintain the virus in their genome that could potentially be reactivated later, if it can be reactivated.  And then also understanding how and why myeloid cells are resistant to HIV-induce cell death.  This was touched upon in several of the talks and how macrophages can get infected and then don't die as a result of the infection, but then potentially go latent.

In general, future research, it seems to be going, at least based on this talk, towards a focus on single-cell methods in the sense of ex vivo single-celled to really understand the cells as soon as they're removed from the tissue.  What do they look like, what can they do, what can't they do, knowing that as soon as they leave the tissue and enter a dish, they do change, so to be very cautiously optimistic with methods like that.  But then also to -- as a way to distinguish them.  So how are they distinct based on their metabolic signatures, biomolecular mechanisms that enable viral shut down, or even the ability to resist HIV-mediated cell death, as I mentioned previously.  With the overarching goal of producing myeloid-specific pharmacological interventions.  But I really like what Tom said in that we need to step away probably from the dogmatic ideas that we've been indoctrinated with since the beginning of HIV and just try things that are not expected because macrophages are not CD4s and they're different, but overall, we seem to be heading more in that direction.

And with that, I would like to turn it over to Dr. Khalili to give Session 4's overview.

DR. KHALILI:  Yes, hi.  Can I have my -- yes.  This session is started with a -- with a terrific talk by Dr. Bomsel, who basically presented to me very convincing data that -- showing that in urethral macrophage and in mucosal macrophage there is a presence of the replication-competent HIV, and then also virus can remain in the latent stage in these -- in these tissues and cells.  And I think that was a very clear set of -- very clean set of studies that, in in my view, nailed down the detection of the virus in the -- in the tissues and the myeloid cells actually.

But then there was a series of studies on the -- looking at the epigenetic and transcription regulation of the virus in the -- in myeloid cells macrophage, and -- which was interesting, in particular, the epigenetic approach that the -- that the BRD4 was -- probably play a role in either activation or suppression of the virus.  Initially it was shown that it might be important for activation virus or later on established by -- studies by Bruce Walker's laboratory that actually it could be separation is involved in the -- in the virus replication in those infected cells.  That was very interesting observation.

And then -- and then role of the AAV in the nonhuman primate that -- showing that how broadly-neutralizing antibody can stay for a long time, and that was basically gene delivery approach, and then keep the virus at bay for a very long period of time, yet the virus is not eliminated.  And then comes Gendelman, who tries to demonstrate that the virus can be inactivated genetically by using CRISPR technology, but the studies were done mostly on the -- almost entirely in the cell culture and then delivery.  The good thing about the studies that I liked was developing of the new delivery system, lipid-based delivery system, but, again, whether or not this lipid deliberate system can target specific cells or uniformly get every cells is a -- is a question that needs to be addressed and then -- and then work done on it.

And then the last talk was about the combination of the -- of the CRISPR technology that basically, although Kaminski presented that only one-half of that -- the other half is something that the -- that is ongoing project, is going on in the different laboratories, including my own laboratory, and showing that can be combined, the CRISPR technology, for elimination or inactivation of the virus on one hand, and also modification of the cellular genes, which in this case, Kaminski demonstrated that ALCAM could target -- one could target where you can basically mitigate the chance for the virus to spread and replicate in other cells.  So that was basically a take over from the --

But with respect to the future strategy, I think if one needs to enter into the genetic and epigenetic, it's -- I think it's a great idea because, after all, when the virus impact the cells, the pace of the infection changes from -- infectious disease become like a genetic disease.  So use of epigenetic and a genetic approach is the right way of doing it, in my view.  And then, but it is a careful -- it should be done very carefully in terms of the -- that this strategy does not impact on the -- on the integrity of the host chromosome because the outcome could be devastating.

And then one of the concerns I have the about the epigenetic studies, it talk whether or not that basically working with the BRD4 can stimulate the autogenic pathways in the -- in the -- in the -- in the patient.  But most of the studies were done in the -- in the cell culture, and so you really don't know that -- the in vivo impact of that over the years where it's going to go.  But overall, I think the strategy of looking at the transcription genetic and epigenetic approach and then nail it -- nailing down that -- the role of the macrophage in harboring the virus was interesting, and they present it very well throughout.

Now, I'm going to pass it to Serena, I guess.

DR. SPUDICH:  Yeah, that's right.  Thank you very much, Dr. Khalili.  So I'm going to be summarizing Session 5.  And first of all, I just wanted to comment that, you know, I think the observations of the chairs of the other session are really thought provoking, and I think some of the issues and sort of thinking outside the dogma.  And this is a group, I think, of investigators who are often being in that space, but I think thinking about that rigorously and making sure that people outside of the people attending this meeting are also interested in looking at things outside the dogma.  And thinking about how as a sort of myeloid-centered field we can communicate that effectively outside this group I think is extremely important.

So I was moderating a session that had three really terrific talks.  I think they were very succinct and clear.  One focused on microglial models using both two-dimensional and three-dimensional models -- a three-dimensional model including also other cell types in the brain and thinking about how HIV infection in microglial cells interacts with neuroinflammation potentially as a substrate to the development of neurologic injury in HIV.  And the other two talks were really focused on trafficking of infected cells into the CNS establishment, and persistence, and maintenance of HIV reservoirs potentially in myeloid compartments.  And these focused on human studies, taking advantage of a unique cohort of people who have acute HIV infection who then start immediate antiretroviral therapy.  And then from Dr. Berman, using an extremely valuable blood-brain barrier model that I think has been extremely revealing to look at sort of chronic infection and trafficking in that setting.

So the key takeaways which were very kindly provided by my speakers are that, even in the setting of suppressed viral replication, mature monocytes seem to be entering the CNS carrying HIV, and we saw some compelling data that there's selective trafficking potentially of infected monocytes versus simply exposed monocytes through the blood-brain barrier model.  And these I think have the potential to sort of lead to neuropathogenesis, including maintenance and replenishment of reservoirs.

We saw interesting data from the microglial model suggesting that HIV intron-containing RNAs, including defective RNAs, induce immune responses in microglia, and it was interesting that it was actually cytosolic virus that seemed to be producing these changes.  And finally that cell-associated RNA levels in human blood monocytes from donors during acute HIV infection seem to be important in determining the trajectory of a variety of aspects of neuropathogenic sort of mechanisms after the initiation of early ART.

So I think a lot of these questions related to the sort of establishment and maintenance of reservoirs within the CNS and neuroinflammation.  And questions that came up, I think, in our session and, I think, are sort of ideas for future research relate to using, for example, emerging HMOX1 tools.  We talked earlier.  Others commented on using maybe single-cell approaches to identify therapeutic targets that may block some of the trafficking and also may reverse abnormalities after HIV.  And then the use, I think, of either brain organoids, the spheroids, and also particularly targeted monocyte and myeloid interventions that may make a difference for long-term outcomes.

So thanks very much, and I think I'm turning it over to Dr. Brown.

DR. BROWN:  Thank you so much, Dr. Spudich.  And from the discussion session, I think, you know, the folks get it in terms of having this R21 format and R01 formats to test some of our dogma.  And so there was a lot of excitement about unexplored CNS or underexplored CNS cells, particularly the parasites, and also the platelet monocyte interactions and their role as reservoirs that have not been really addressed or accessible to ART.  And all of the speakers, everyone is really using the latest, most comprehensive approaches and quantification -- digital droplet, PCR, full-length sequencing, QVOA, IPDA assays -- and also doing functional analyses to really look at what is replication competent, what is defective.

And some of the challenges, you know, that we just talked about, you know, we're using cell lines, but we're pushing the edge in using primary cells to really see and hope that our -- that our findings are translatable to people living with HIV.  So again, the prominent role of humanized mice where we can study viral infection over the long term and really understand, as others have said, is the potential of these different myeloid cells to serve as a latent reservoir and their susceptibility are not to ART.

And again, the quantification, some of the challenges, you get low cell numbers.  But I was impressed with, you know, for example, the fish RNA scope type of approaches that, again, ART will help us get to that level of sensitivity of detection.  But then we also have to keep in mind latency reversal agents and other things to eradicate, to stimulate turnover other reservoir.  What is the impact going to be on the central nervous system?  We really have to be able to control these technologies and things that we use to control inflammation in the host response so that we can have, you know, healing and n regenerative impacts to return the system to homeostasis.

And so future research is really, you know, going to really drive towards understanding the mechanisms for latency in these myeloid cells and continue to make improvements on quantifying viral reservoirs.  And I'll end there turn it back over to the moderators.

DR. CLAYTON:  Okay.  Thank you to all the panelists.  That was -- that was a very nice summary and discussion of potential future research.

So Janice and I are just going to go through a couple points to try and put all this in kind of a bigger picture context here.  So I know we had a few points to start.  I guess I could start with something and then hand it over to Janice.  You know, there's obviously a need to study tissue-resident macrophages as opposed to the monocytes that we get in the blood, so really trying to develop technologies where we can study these tissue-resident macrophages is going to be key, you know, not just looking at brain tissues, but other tissues that are macrophage rich, such as the -- such as the liver and the gut and potentially the skin to get ahold of some Langerhans cells, I think, would be really interesting.  And kind of building off of some of the PET-CT work that was done in the monkeys.  You know, would it be possible to develop immuno-PET to actually detect HIV reservoirs within ART-suppressed individuals.  so I think that would be something really interesting.

And then, you know, there's been a lot of discussion of LRAs, and obviously pushing HIV out from latency.  But when we think about cure-related strategies or any sort of therapeutic intervention, you know, there's the whole prototypical shock and kill, in which case you would push HIV out of latency and then there would be potentially some immune component to eliminate that, but then there's also block and lock to completely silence HIV.  And, you know, I guess we have to weigh the advantages and the disadvantages of each, you know.  Are we going after more of a reservoir elimination type of strategy with the shock and kill, or is this, you know, something we want to limit more of the inflammation and so, you know, block and lock may actually be a better therapy.  And I'd be curious to see what the panelists think of that.

DR. CLEMENTS:  Yeah, I'll pick up from there.  One of the issues I saw, discussed, and sort of talked about in various ways, are the, as you say, the assays that we're using to measure the reservoirs, particularly the myeloid reservoirs.  We're currently using two assays:  QVOA assays and intact proviral DNA assays.  Are they accurately showing us what the true reservoir is?  Are they capturing everything that we would want to know?

I think many of us are using these assays.  We depend on them.  But I think we have to test and push them, too, and make sure that the results we're getting are really reflective of what's in brain tissues and of non-human primates, and the exciting work that was just reported about doing it in frozen brain with Melissa.  So I think there's a lot of potential there, but we can't just sit on our laurels and use those essays without really testing them and understanding whether they're giving us the right answer.

There probably needs to be a focus on designing LRAs and other specific agents for macrophages.  While there are LRAs for T cells, I think there is -- there are very few ways of taking those -- the macrophages and really activating them to produce virus.  As somebody who's been working on macrophages a long time, they have their own mind.  They have a mind of their own, and when you activate them they do -- they do different things in T cells.  So that's another thing that I think we have to really think about and perhaps devise or identify novel ways of reactivating the reservoir and looking at how the macrophage actually manages that reservoir.

I think you took -- you covered the shock and kill which I was going to talk about, too.  I think shock and kill is very interesting.  It hasn't -- I agree with you that shock and kill has not been very impressive, and so I think -- and it would cause, particularly in macrophages, a lot of inflammatory reactions in tissues.  And so actually trying to impact the -- both the T cells and myeloid cells by locking down the latent virus seems to me a very -- very important area that should be focused on.  How do we do that?  Macrophages are normally pretty quiescent unless they are activated, so we might take advantage of that.

So those are the sorts of things that I think were brought to my mind, but altogether it's been a very exciting meeting, and many, many ideas and new ideas, and new faces have been focused on here in this meeting.  So I think it's been very exciting.

DR. CLAYTON:  Excellent.  And there is actually one comment in the Q&A if I could just pose that since we have about a minute left here.


DR. CLAYTON:  So there's a potential question to all.  "One inherent problem with many of our assays when looking for the virus outside the cell may be that we are not really testing for genuine viruses.  Looking for p24 RT-PCR for a certain portion of the virus, RT assays, gag, et cetera, are easy to test but are not a real virus rescue assay if one is interested in the presence of actual virus.  It is possible today to separate the virus away from others using EVs, exosomes, incomplete viruses, et cetera, and then test their effects on CNS injury.  Therefore, what are the chances that most of the pathology in the CNS from infected myeloids, astrocytes, parasites, especially under ART, may from EVs and not directly from virus transfer to other cells?  This, therefore, requires model systems, such as CNS 3D cultures, for long-term or six-months-plus studies."  So also something to consider.

DR. CLEMENTS:  Great.  I'd like to thank all the speakers, too.  They did a great job, and it was really an exciting meeting for two days.  I really enjoyed it.  Does Jeymohan want to say -- so now we are going to hand over the meeting to Dianne.  Dianne, are you available?

DR. RAUSCH:  Yep.  I'm here.  Okay.  Thanks, Janice, and we just want to thank everybody, all the speakers, the moderators, and the meeting organizing committee -- Kiera, Janice, Mario, and Jeymohan -- for such an informative two-day meeting.  We had a -- heard a lot of really interesting science, which really expands our knowledge of the role of macrophages in HIV pathogenesis and cure.  I think the goal of the meeting, as Jeymohan had outlined at the beginning, was clearly met through these presentations and discussions and did a great job of capturing these goals and presenting research to address all these topics.

Many of the questions and discussions that followed the presentations also provided stimulating ideas and considerations for new approaches, and we hope that these dialogues will continue and open new opportunities for expanded collaborations.  And finally, I think the closing session did -- was great, a great summary of the sessions and discussed key gaps and priorities that have been identified.  And we particularly appreciate the closing session with the recommendations to think outside the dogma.  This is great for us as a program to hear, and it's helpful as we think about new ideas for new priorities and initiatives.

And just to remind you, we will be preparing a meeting report to summarize these discussions and potential next steps, and that will be posted.  A recording of this meeting will also be posted and -- in about six weeks at on the NIH website.  And a reminder that the Journal of Leukocyte Biology will be presenting a -- or providing us with the opportunity to do a special issue to address many of the things that were presented at this conference.

So on that, I will close the meeting.  Thank you again, everyone, and I hope you have a great rest of your day.