PolyBio Fall 2024 Symposium

Now on Youtube:

https://www.youtube.com/watch?v=nzwH556GsuU

 

Seems like all the Polybio projects from phase 1 are still ongoing (for example the numerous "microclot projects") or negative results aren't published or preprints aren't published, does anybody know?

 

Many of the projects discussed were with a low or very low number of samples. Those types of projects don't necessarily get published but are used to help determine what to do next or apply for larger grants. Just my two cents.

After the live presentation I was left feeling a bit lost. Virus signatures were found in a lot of different types of samples from all over the body. If that's the case, we are still none the wiser on what tissues, fluids, and cell types to focus future research on to determine the disease mechanism.

 

I'm going through watching the symposium a second time much more carefully and taking notes, and I'll try to post them when I'm done. But Morgane Bomsel's talk at 50:47 seems so important that I wanted to post the notes for that now.

---------------------------------------------------------

Main takeaways: In about 2/3 of participants with long COVID, but none who recovered from COVID, megakaryocytes (platelet-producing cells), as well as platelets, were found to be infected with live, infectious SARS-CoV-2. People with LC have greater numbers of megakaryocytes. Platelets in LC more easily form microclots. Spike in the plasma was found in some LC participants.

---------------------------------------------------------

Screenshot of list of papers showing viral components persisting in long COVID.


Bomsel's group previously found that megakaryocytes in severe acute COVID-19 were infected with SARS-CoV-2, and some of the platelets they produced also had the virus, and the virus was capable of infecting other cells. Most non-survivors had infected platelets. Only some survivors had infected platelets, but of these, some (one? all?) had infected platelets several years after initial infection, and had symptoms of long COVID. Thus they hypothesized that megakaryocytes act as a reservoir of SARS-CoV-2 contributing to symptoms of long COVID.


So far have sampled blood from over 50 people with long COVID (symptoms longer than 6 months, median sampling date 16 months after infection). [Edit: she didn't say how many controls they've tested, but it's at least 10, based on the abstract with earlier data linked below.] Also sampled blood from COVID recovered individuals with no history of long COVID.


In about 2/3 of people with LC, their CD41+ megakaryocytes contained spike and viral positive RNA. None of the COVID recovered participants' megakaryocytes had these viral components. Double stranded viral RNA was also found in LC megakaryocytes, but not recovered controls, indicating ongoing replication.


Also found that there were more megakaryocytes in LC than controls, which has also been shown in acute COVID.


Found that platelets in most people with LC also had spike, positive RNA and negative RNA, indicating replication. Tested infectious ability of these platelets, and they were able to cause infection of other cells, again from about 2/3 of people with long COVID.


Found that platelets from LC participants could form microclots more readily, but these microclots were not driven by spike protein, because no spike was found outside of the platelets in these microclots.


About 17-36% of people with LC had detectable spike in the plasma.


Also mentioned that they previously found that serotonin and related metabolites are decreased in LC (in platelets?), likely due to TPH1 deficiency. Potentially contributes to neuronal dysfunction.

They are continuing to validate their findings, as well as attempting to correlate these markers with clinical/symptomatic data.

---------------------------------------------------------

They have previously presented some of this data in a poster at a conference, linked here:

 

The stuff on megakaryocytes being infected would be interesting if it pans out. But the quoted statement doesn't make sense and makes me suspicious. Microclots are protein clumps that are smaller than platelets as far as I can see, not platelet material. And nobody has any evidence for microclots being able to produce symptoms.

 

It is possible I didn't fully understand, so I'll transcribe the exact words she used:

"We then found that platelets from long COVID patients indeed form microclots, unlike blood from COVID recovered or healthy donor. And interestingly, in our case, these platelet aggregates do not seem to be induced by spike, because we can't find spike outside of the platelets in these aggregates."

Maybe a different definition of microclots from what has been used, but still some grouping of platelets? There was an image included of the aggregates.

 

I was going to post all the notes together in one post for the whole symposium when I finished, but since there are a max of 10 images per post, I figured I should split it up, so might as well post as I complete them.

12:40 Nicholas Huot–The Role of THEMIS in Modulating NK Cell Activity during COVID

Main takeaways: Downregulation of a molecule called THEMIS in NK cells could be involved in viral persistence, possibly related to a lowered ability of NK cells to detect or kill infected cells.

------------------------------------------

One year ago, they found persisting SARS-CoV-2 in alveolar macrophages, 18 months after infection in monkeys (study thread). IFN-gamma was able to reduce viral persistence in these macrophages. IFN-gamma also increased MHC-E expression on surface of the macrophages.

MHC-E is recognized by an inhibitory receptor called NKG2a on NK cells. MHC-E reduces NK cell degranulation activity. Also, NK cells were shown to be more inhibited in monkeys with the highest viral persistence.

New data: Investigating potential of a molecule called THEMIS (Thymocyte-expressed molecule involved in selection) to reduce NK cell inhibition. THEMIS is involved in T cell selection (removing T cells that might cause autoimmunity). Modulates T-cell receptor strength by binding to a molecule called SHP-1. SHP-1 is also involved in NK cell inhibition, so they hypothesized that THEMIS is involved in inhibition of NK cells leading to viral persistence. Found that THEMIS is downregulated in NK cells of chronically infected monkeys in both blood and bone marrow.



They isolated NK cells and tested responses to different cytokines and other chemicals. For B-cells or T-cells, none of the cytokines had much effect on THEMIS expression. In NK cells, IL-12, IL-18, and IL-1b all decreased THEMIS expression. These cytokines are all expressed by macrophages, so they think macrophages may be involved.

Pilot study in humans using CORSER 5 cohort. In people who had a SARS-CoV-2 infection, one month after infection, their NK cells expressed higher levels of NKG2a and THEMIS. The levels are closer to normal at 6 months.

 

27:16 David Price–Infectious, Immune and Microbiome Signals in the Long COVID Lung

Main takeaways: In long COVID: SARS-CoV-2 specific T cells more inhibited - potentially indicates ongoing exposure. Lower SARS-CoV-2 neutralizing antibody activity - potentially allows virus to proliferate. Expansive protein upregulation in various systems. Complement system more activated. New, ongoing study is looking at lung tissue and fluid.

-----------------------------------------------------------

Previous study (link): Looked at peripheral blood in long COVID and healthy controls. No major differences in frequencies of immune system cells, including SARS-CoV-2 specific T cells. But on T cells specific to some non-spike SARS-CoV-2 proteins, there was upregulation of a few receptors, particularly the inhibitory receptors TIM-3 and PD-1. Price says this is known to be an indicator that there is repeated exposure to an antigen.


Those with long COVID had significantly lower neutralizing antibody activity against SARS-CoV-2, but no differences in total SARS-CoV-2 antibodies or antibody-dependent NK cell activation/cytotoxicity. Potentially enables viral persistence.


Proteomics: Large numbers of proteins involved in a variety of body systems are upregulated in long COVID.


New since last symposium: found markers associated with breathlessness, including TNF (inflammatory marker) and thromboxine a2 (platelet activation marker).

Complement system more activated in long COVID. Able to predict long COVID vs controls with area under the curve of nearly 0.9.


The findings suggested that there may be ongoing tissue damage caused by viral reservoirs driving long COVID. Thus the current study is looking at tissue from the initial site of infection (lung) in people with long COVID and respiratory symptoms (n=15), and healthy controls (n=15). Will include transbronchial biopsies and bronchoalveolar lavage fluid. They hope to link clinical data with experimental data. Currently still awaiting most of the data from the collaborators, but lots of data expected to be available by next symposium in 6 months.

 

39:03 Resia Pretorius–Microclots, Biofilms and NETs: optimizing methods for biomarker identification

Honestly, this one was very confusing so I probably missed a lot of relevant details, so watch it if you are interested in microclots.

---------------------------------------

Goal: Identify biomarkers that may indicate specific long COVID symptoms. Promising candidate: microclots. These also entrap inflammatory markers, which allows easier study of inflammation.

• Found that NETs are entrapped in some microclots.
• Found that spike protein can cause fibrinogen to turn into insoluble microclots. LPS (lipopolysaccharides, produced by bacteria) can as well.
• Biofilm is found in some microclots in some people with LC, which may indicate bacterial involvement.
• Also, Syndecan-1, a marker of endothelial damage, is found in a subset of microclots.
• Microclots shown to be able to combine into macroclots in stroke patients.

 

1:02:57 Johan Van Weyenburg–Blood transcriptomics reveal persistent SARS-CoV-2 RNA as a candidate biomarker, a real world Long COVID cohort

Main takeaways: Performed blood transcriptomics on LC patients (looking at circulating RNA to see what proteins are currently being produced by both human and virus). RNA related to viral proteins, including nucleocapsid, was upregulated. Antisense RNA from SARS-CoV-2 was also increased, which indicates ongoing replication. Also increases in some RNA related to platelets and memory B cells, plus RNA for a potential target for Paxlovid. Performed unblinded trial of Paxlovid. There was symptom improvement in some participants, as well as decreases in RNA for some viral proteins, but no decrease in viral antisense RNA.

---------------------------------------

They are interested in blood transcriptomics to identify biomarkers as it can identify both human and viral RNA. (Thread for previously published article)

48 LC and 12 controls, matched for age, sex, time since infection (average 2 years), vaccine status, and number of comorbidities.

He basically repeated this part from paper: "Several viral RNAs were upregulated: nucleocapsid, ORF7a, ORF3a, Mpro (a nirmatrelvir plus ritonavir [Paxlovid] target), and antisense ORF1ab RNA. Specifically, the upregulation of antisense ORF1ab RNA suggests ongoing viral replication. SARS-CoV-2-related host RNAs (ACE2/TMPRSS2 receptors, DPP4/FURIN proteases) and RNAs prototypical for memory B-cells and platelets4 were also upregulated."


Also IgE RNA upregulated, which may be related to allergies seen in patients.

For antisense RNA (indicates replication) 65% of LC positive vs 25% of controls. For viral load, 60% of LC are higher than a cutoff vs 8% of controls. (I don't fully understand how the cutoff is determined. From paper: "The red horizontal line represents the cutoff for positivity (ten normalised counts for individual transcripts and 50 normalised counts for the viral load)."


Prediction using both FYN (involved in immune system) and viral antisense yields 94% sensitivity and 92% specificity.


Patients with higher anxiety/depression had higher antisense viral RNA and lower immunometabolism (based on combination of immune markers in transcriptome).


Performed study of Paxlovid with 16 people with LC. Participants opted in to Paxlovid treatment. The rest only used anticoagulants (n=49). Using cloud based platform that allows real time access to patient data, including symptoms and adverse events. Primary outcome is % drop in personalized symptoms. Secondary outcomes include transcriptomics markers. About 32% lower symptom score on average at 15 days. (Did not present data from control group here as far as I can tell.)


For secondary outcome of COOP score, which asks about things like fitness, social activities, and feelings, there was a significant decrease (improvement) in Paxlovid group but not control group.


They saw decreases in some viral RNAs, such as nucleocapsid, with Paxlovid treatment, but surprisingly, no change in antisense viral RNA.


They are continuing to follow the participants, and so far have an average of 8 months of data for Paxlovid group and average 12 months for control group.

Included recommendation to consider personalized symptom scores for trials (patients choose most personally relevant symptoms to track), as they may be more sensitive to improvements, and they are faster and easier to complete.

 

1:17:35 Marcus Buggert–LymphPASC Project Proposal: Exploring Lymphoid Immune Activation in Post Acute COVID-19

Main takeaways: Ongoing study trying to see if SARS-CoV-2 persists in gut and tonsils, and seeing if persistence leads to T cell activation in tonsils. Another ongoing study is looking at protein differences in long COVID. Have found pathways correlating with breathlessness, but also found that long COVID is associated with platelet activation, inflammation, cell proliferation, and cell death.

----------------------------------------------------

Discussing study LymphPASC, which explores viral persistence in GI tract and tonsils, and tries to see if this contributes to immune activation and symptoms.

For GI tract portion, looking at 15 LC and 15 age/sex matched recovered controls. 6 biopsies per participant throughout GI tract. Using single cell RNA sequencing on immune and non-immune cells to identify virus. Trying to identify specific cells that harbor viral RNA, and eventually viral proteins. Also eventually will analyze transcriptional differences in these cells.

They have previously seen that in general, killer T cells in the tissues, and particularly in the gut, are less cytotoxic, so they are seeing if SARS-CoV-2 specific killer T cells in the gut are less cytotoxic in the long COVID group. The data seems to support this so far. Less cytotoxicity may make it easier for virus to survive in the gut.

Also looking at changes in microbiome and integrity of gut epithelial barrier.

So far have collected tissue from 23 out of 30 participants. Analysis will begin soon, within the next month or so.

For tonsils: studies have shown increased T cell activation in tonsils, as well as persistent SARS-CoV-2 in tonsils in children. Looking at tissue from 60 participants prior infection, 30 prior vaccination, and 40 neither.

If they detect virus, they will try to detect localized T cell activation, which might be easier to detect close to the source of the virus, as opposed to in the blood. They think T cell activation may cause symptoms.

Different study: seeing which LC symptoms associate with systemic protein alterations. Along with David Price, looking at 150 combined LC and controls. Breathlessness found to associate with markers from multiple systems and pathways. (Mentioned in David Price talk as well.) Many of these proteins have previously been linked to collapsed lung, shortness of breath, lymphedema, etc.


Also found many pathways elevated in LC in general, including pathways involved in chronic inflammation, cell death and cell growth (often linked to cancer), cell migration and cytoskeletal dynamics (linked to wound healing and fibrosis), and most interestingly pathways linked to platelet activation.

 

Glad it's helpful! Definitely might take longer than I thought to get through all of these though, my brain was barely functioning last night after doing a few.

 

1:30:09 Michela Locci–Lymph Node Immune Responses in Long COVID

Main takeaways: This group has found differences in germinal center B cells (GC is involved in selecting and proliferating the best B cells) in long COVID. As opposed to controls, where proportion of B cells specific to SARS-CoV-2 in the germinal centers decreased over time after infection, in long COVID they increased, but still reached much lower levels. Numbers of total and SARS-CoV-2 specific B cells in the germinal centers were lower in LC. There were also fewer SARS-CoV-2 specific antibodies in the body, and lowered ability to neutralize the virus with antibodies. In the transcriptome of the B cells, they found some differences between groups, including differences related to interferon.

----------------------------------------------------------------------------------------

Locci's lab is interested in germinal centers, which are structures in lymphoid organs, such as lymph nodes and the spleen, which are important in B cell development. After an infection, the B cells most effective at binding to an antigen such as SARS-CoV-2 are selected here to be used in the immune response. The highest quality B cells can then differentiate into plasma cells, which can release antibodies to fight an infection. Germinal center B cells can also differentiate into memory B cells, which stick around after the infection to detect if it comes back, and if it does, they differentiate into plasma cells to mount a fast response. (Going by how I understood what she said and a quick skim of Wikipedia, might not be 100% accurate.)

Prior studies have suggested differences in B cell responses in long COVID.


This group hypothesizes that there is altered germinal center activity in LC, possibly releasing B cells which don't work as well, for example by not binding as well to the virus.

They have previously studied germinal center responses to vaccination by giving a vaccine in the arm, then using a needle (using process called fine needle aspiration, or FNA) to extract some cells from the nearby lymph node under the arm.

The current study is instead looking at the lymph nodes in the neck, because these are more involved in infections in the throat, where SARS-CoV-2 is often at high levels.

The study is testing 16 convalescent (COVID recovered) and 23 with long COVID, including many with fatigue and cognitive dysfunction. They are mainly looking at two time points: T1 is anywhere from 3 weeks to 6 months post-infection, and T2 is anywhere from 6 months to around a year post-infection.


They found that in convalescent participants a high proportion of germinal center B cells were specific to SARS-CoV-2 at T1, and these are nearly gone at T2. In contrast, in those with long COVID, the proportion of such B cells was much lower at T1 compared to the healthy group, and instead of decreasing at T2, the proportion of these cells increased. Even after increasing, the level is still much lower than the peak in controls.


They also looked at the levels of SARS-CoV-2 antibodies over time. They found that these antibody levels were slightly lower soon after infection in LC, and the levels decreased faster over time than in recovered participants.


They looked at the ability of antibodies to neutralize (neutralizing antibodies attach to the virus to interfere with its function) SARS-CoV-2 in vitro (outside the body), and found a reduction of these types of antibodies at both time points compared to controls.


And they found a correlation in controls, where the larger the germinal center response (proportion of B cells specific to SARS-CoV-2 in GC), the more neutralizing antibodies were detected. But no correlation between these two metrics was found in LC.


Also looked at clonal size of total germinal center B cells and GC B cells specific to the virus (someone correct me if I'm wrong, but I think this basically means numbers of these cells in the germinal center). Found that for both measurements, and at both time points, the clonal size for LC was always lower. They think this is related to the mechanism of B cell selection, so they are digging more into this currently.


Mentioned that transcriptomics of the memory B cells (looking at RNA in the cell; highlights which proteins are being made) showed some differences, including related to interferon.

They plan to continue to test whether long COVID B cells have less ability to bind to the virus, and whether altered B cell responses are associated with viral persistence.

 

1:41:42 Akiko Iwasaki–Probing the Role of Endogenous and Latent Viruses in Long COVID

Main Takeaways: Iwasaki's lab is studying latent viruses' (including EBV and endogenous retroviruses) potential involvement in various diseases. They have found that endogenous retrovirus expression is upregulated in lupus, and antibodies against these ERVs from lupus patients are more likely to induce phagocytosis by neutrophils, so they are continuing to study whether these or similar antibody differences can lead to disease. They are also involved in a study of Truvada and maraviroc in long COVID. They are used against HIV, and Truvada has previously shown inhibitory activity action in EBV and ERVs.

-------------------------------------

Iwasaki's lab is exploring the possibility that acute infection causes reactivation of latent viruses, such as herpesviruses like EBV, or endogenous retroviruses.

About 95% of adults carry latent EBV. Multiple prior studies have found a link between long COVID and EBV reactivation.


Retroviruses are viruses that irreversably integrate into the DNA of a cell. One example of a retrovirus is HIV. If a retrovirus infects a germline cell (these produce sperm or egg cells), they will persist into future generations, and are then called endogenous retroviruses (ERV, or HERV for human endogenous retroviruses). About 1-8% of the human genome consists of ERVs. The majority of these ERVs no longer do anything due to mutations over time, but some have been linked to diseases, including MS. (https://en.wikipedia.org/wiki/Endogenous_retrovirus)

A few years ago they developed tools to identify RNA being transcribed from ERV DNA, as well as to identify antibodies that target ERV antigens.


They looked at people with systemic lupus erythematosus (SLE or lupus), and found increased expression of multiple ERVs, particularly an ERV called K-10. They tested the antibody response of SLE and healthy controls to this ERV in vitro but did not find much difference in the antibody response.


But then they examined one specific function of antibodies, which is marking a pathogen for phagocytosis by a neutrophil (antibody-dependent neutrophil phagocytosis, or ADNP). Phagocytosis was increased when using antibodies from SLE bound to this ERV, compared to controls. There was also increased formation of neutrophil extracellular traps in response to SLE antibodies bound to the ERV.


They thus hypothesize that the antibodies against ERVs in SLE being more capable of inducing phagocytosis may be involved in the pathogenesis of lupus. They are continuing to investigate whether antibodies against EBV or ERVs can lead to disease when bound to these antigens.

Also discussed an ongoing trial of Truvada and maraviroc in long COVID. These drugs are currently used against HIV. Truvada has shown some efficacy against EBV as well. It is currently being studied for preventing HERV activation in ALS, and EBV reactivation in MS. Maraviroc "will block inflammatory cell migration to tissues."

Enrolling 30 participants for each drug and 30 for placebo (n=90 total) for daily low dose treatment for 3 months. Measuring patient reported outcomes, as well as changes in immune, transcriptional, and viral signatures.

 

1:52:30 John Wherry–T cells as biosensors in Long COVID

Main takeaways: Wherry's lab is experimenting with looking at levels of T cells specific to certain antigens, such as SARS-CoV-2, as sensors which likely imply presence of these antigens. They are looking at these levels in acute, recovered, and long COVID cohorts. They found SARS-CoV-2 spike T cell activation in at least 25-40% of people with long COVID, but only around 5-15% of recovered participants. T cells specific to the latent viruses EBV and VZV are also more commonly activated in long COVID than in recovered controls. They also confirmed earlier data that these latent viruses are reactivated in acute COVID. They did not find differences between any groups for flu or CMV T cell activation. They are continuing to examine these antigen specific T cells more closely, including through single cell transcriptomics.

---------------------------------------

Wherry's lab is interested in probing the immune system to test the relationship of persistent or reactivated viruses to long COVID. This talk is specifically about using T cells specific to certain antigens to act as "biosensors" to show if there is persistent virus or increased levels of latent viruses.

The exact methodology is above my pay grade, but in essence they create molecules which are combinations of HLA proteins as carriers bound to pieces of the virus, and they see if any T cells bind to these molecules. They tested this technology after vaccination as well as after a breakthrough infection of SARS-CoV-2. The area in the colored portion of each square represents T cells specific to that part of the SARS-CoV-2 virus. After vaccination (first row) there were T cells specific to spike protein, but none of the other three viral proteins. After an infection, there are T cells for all four parts of the virus.


They can also track markers on these antigen specific T cells over time that provide more information, such as markers related to activation (CD38) and proliferation (Ki67). The first row of the following chart shows the proliferation marker with an increase from about days 5-10 after infection, and the activation marker increasing earlier and staying high longer.


They also created these tracking molecules for other viruses, including EBV, influenza, VZV (chickenpox/shingles), and CMV.


The lab has access to three long COVID cohorts for running these tests, each of which includes an acute, recovered, and long COVID group.


I'm not sure if these images are examples for three individual participants or for the groups as a whole, but they show that in acute COVID, 88% of spike specific T cells are activated, nearly 0 are activated in the recovered chart, and 27% are activated in long COVID.


This chart shows data from the full cohort, showing that about 25% of people with long COVID have spike-specific T cell activation, while only 5% of recovered participants have activation, which suggests that there is antigen still present in at least a portion of people with long COVID.


Again I think these are from individuals, but they show that EBV specific T cells are substantially activated in acute COVID, which agrees with previous findings of COVID causing EBV reactivation.


Looking at the full groups again for other viruses, they found that more people with long COVID vs recovered had T cell activation for EBV (16% vs 6%) and VZV (13% vs 4%), but there was no difference in activation of T cells specific to CMV or flu. This suggests that there is some reactivation of the first two viruses.


They are developing a technology called Barcode Enabled Antigen Mapping (BEAM) for looking at these T cells in more detail, including examining their RNA and surface phenotypes. They hope to have data from this experiment in the next few weeks to begin analysis.

There is not a good correlation between circulating spike and activated spike-specific T cells, and they are trying to figure out the reason.

 

I think it is just a question of them having no idea what they are talking about. Microclots are clots not platelet clumps. The level of incompetence in this sort of work takes some believing but that is how it is.

 

I am afraid that none of this stuff looks meaningful to me. So many things you can measure in immunology are just not close enough to what you need to know to be any help. It is a bit like trying to work out who the murderer was in a French detective novel when you don't know French. You can count how often names come up but you will never know who did it.