Postdoc @Marsonlab @GladstoneInst @UCSF | PhD in Virology with @jkagan1 @harvardmed | Interest in Mechanistic Immunology, Cell Biology, Synthetic Biology
- Tweets 412
- Following 245
- Followers 352
- Likes 670
KO of A20 drove a proinflammatory state resistant to suppressive repolarization and enhanced cytotoxicity in HER2-CAR macrophages. With TNFAIP3 surfacing as a brake on antitumor function now in both T cells and macrophages, it may be an attractive target for cell therapies.
1
142
Our screens converged on the gene TNFAIP3, which codes for the protein A20, as a master regulator of macrophage inflammatory polarization.
1
88
To make pooled screening possible we developed SLICeVLP, sgRNA delivery by lentivirus with Cas9 delivery by eVLPs. We used it for pooled loss-of-function screens and Perturb-seq in primary human macrophages, identifying regulators of inflammatory signaling.
1
110
This platform supports knockout, base editing, and epigenetic silencing, and paired with AAV donor delivery it enables site-specific integration of large sequences by homology-directed repair, even in these post-mitotic cells.
1
126
@HyuncheolJung and I adapted these VLP-based methods to deliver CRISPR ribonucleoproteins to primary human monocytes, macrophages, and dendritic cells at high efficiency while preserving both viability and innate immune responsiveness.
1
140
Big 'Thank you' to @SavvidesLab @DanielleMClancy for writing this thoughtful spotlight @ImmunityCP on our recent work on the 'The pros and con(formation)s of IL-18 activation'
doi.org/10.1016/j.immuni.202…
cell.com/immunity/abstract/S…
nature.com/articles/s41586-0…
nature.com/articles/s41586-0…
1
11
27
3,678
17/ When comparing human caspase-4 to mouse caspase-11, we found that critical residues in the interfaces (e.g. K356 and K293) are not conserved between the two homologs. Introduction of human-specific mutations at these sites greatly enhanced pro-IL-18 cleavage by caspase-11.
1
1
312
16/ We found that caspase-bound pro-IL-18 adopts a conformation that is distinct from mature IL-18: the pro-domain inserts itself into the mature domain to form active- and exosite. Cleavage may then trigger a structural change to mediate release of mature IL-18 from caspase-4
1
1
325
14/ We confirmed by site-directed mutagenesis that both interfaces need to be intact for caspase-4 to bind and cleave pro-IL-18 efficiently (both in vitro and in cells).
1
1
308
13/ The active site interface on the other hand is dominated by electrostatic interactions, which perfectly orient the the pro-domain of pro-IL-18 and the tetrapeptide motif in in the catalytic center of caspase-4 for cleavage.
1
1
315
12/ The exosite involves a conserved residue in caspase-4 (W267), described by Feng Shao and @tsanxiao as critical for cleavage of GSDMD, indicating conservation in recognition mode between these two prominent caspase substrates.
cell.com/cell/pdf/S0092-8674…
cell.com/immunity/pdfExtende…
1
1
354
11/ With their help, we determined a cryo-EM structure of caspase-4 in complex with pro-IL-18, the first structure of a caspase bound to an IL-1 cytokine! The structure revealed two distinct binding interfaces: One close to the caspase active site and a distal protease exosite.
1
2
345
8/ The ability of caspase-4 to efficiently cleave pro-IL-18 allows human (but not murine!) macrophages to release IL-18 (but not IL-1β) in response cytosolic LPS challenge in an inflammasome-independent manner.
1
2
358
7/ Intriguingly, this behavior was species-specific, as murine pro-IL-18 was efficiently cleaved by mouse or human caspase-1, but not by human caspase-4 or mouse caspase-11 (the mouse homolog of caspase-4).
1
1
377
6/ Indeed, caspase-4 (and caspase-5, a paralog of caspase-4) is as efficient at cleaving pro-IL-18 as caspase-1! This contrasts what we see when we use pro-IL-1β as a substrate, which shows a high preference for cleavage by caspase-1.
1
1
435
5/ Starting off, we characterized the ability of human caspase-4 to cleave cytokines of the IL-1 family in vitro. We found (in line with recent work by Guy Salvesen and others like @taabaman) that out of all the cytokines tested, pro-IL-18 is the only high efficiency substrate.
1
1
430
Taken together, we show that GSDMD becomes oxidized during pyroptosis in a RagA-dependent manner. This establishes GSDMD as a redox-sensitive protein and implicates cellular redox state as a key determinant of GSDMD activities.
And all of this is mediated by a single cysteine.
1
2
252
Intriguingly, a single cysteine (the common culprit C192) was necessary and sufficient for ROS-mediated enhancement of GSDMD pore formation and pyroptosis. When we mutated this one site to a non-oxidizable amino acid, exogenous ROS did no longer lead to a boost in cell death.
1
1
241
We next used a sulfenic acid-specific pulldown strategy to evaluate relative cysteine oxidation within NT-GSDMD. Here we found that NT-GSDMD is much less oxidized in RagA KO cells compared to WT cells. Again, this defect could be rescued by treatment with ROS-inducing agents.
1
2
258
In a great collaboration with mass spec master mind @HaopengXiao from @echouchani lab, we first confirmed that several cysteines (including C192) in NT-GSDMD become oxidized during pyroptosis.
1
3
279
All of these ROS stimulators were also able to overcome defects in GSDMD oligomerization, pore formation and pyroptosis in RagA KO cells (to varying degree)!
ROS induction also boosted GSDMD-mediated pyroptosis in WT cells, supporting the idea of more ROS=more GSDMD pores!
1
1
2
381
To test this hypothesis, we assembled a panel of known ROS inducing chemicals, including common PAMPs like bacterial LPS and fungal beta-glucan and mitochondrial poisons like rotenone. As expected all of these chemicals rescued ROS defects in RagAKO cells.
1
1
293
Interestingly, these pyroptosis defects correlated with decreased cellular reactive oxygen species (ROS) levels, both at a basal state and upon dox-induction of NT-GSDMD.
1
1
285
In macrophages deficient for the Ragulator-Rag components RagA or RagC, NT-GSDMD was unable to form oligomeric membrane pores and cause cell death.
1
1
293
Using a synthetic pyroptosis system, where the pore-forming N-terminus of GSDMD (NT-GSDMD) is induced by doxycycline (Dox), we had identified the Ragulator-Rag complex as a positive regulator of GSDMD oligomerization.
1
350
How are phagosomal bacteria sensed by cytosolic pattern-recognition receptors to stimulate an innate immune response?
Find the answer in this expert review by my @jkagan1 lab mate @StephRaglandPhD on the 'exodus' of PAMPs into the cytosol.
onlinelibrary.wiley.com/doi/…
4
21
Special thank you also to the editors at Science Immunology @SciImmunology for considering our paper for the cover and Science Visuals for creating this beautiful piece of cover art!
(12/12)
1
5
To conclude, we were able to group all inflammatory caspases into three distinct classes (see below) based on their ability to sense LPS and/or process pro-IL-1β, highlighting the functional diversity and plasticity in inflammatory caspase functions found in nature.
(10/12)
1
4
Lastly, my favourite part: We screened a range of mammalian caspase-4 homologs and found that most carnivoran caspase-1/4 hybrids (e.g. the ones found in cats and other felines, but not canines) operated in a similar manner to our redesigned human caspase-4.
(9/12)
1
1
4
Comparative analysis allowed us to identify the exact residues responsible for this switch in cleavage specificity: A conserved set of charged amino acids (blue), located proximal to the active center. These are distinct from the hydrophobic exosite for GSDMD (yellow).
(6/12)
1
3
Using cell-based assays in engineered murine macrophages and canine primary cells, supplemented with biochemical data we concluded that canine caspase-1/4 hybrids, despite their sequence homology to human caspase-4, are actually functional homologs of caspase-1.
(4/12)
1
3
