Our lab motto is "from one, many":
a reflection on the number of roles
actin can perform in the cell.
We strive to uncover novel connections in cell biology, starting from the basis of the actin cytoskeleton. Actin is an extremely abundant protein (about 200 mM in the cytosol of most mammalian cells) and can assemble into filaments extremely rapidly, almost anywhere. Some very prominent cellular structures are actin-based, such as stress fibers, lamellipodia and filopodia. The prevalence of these structures, however, somewhat overshadows the more subtle populations of filaments that are clearly serving important functions. Two examples of intense current interest in the lab are: endoplasmic reticulum-associated actin, and actin that polymerizes around mitochondria that are under stress. We are interested in the mechanisms that trigger this actin polymerization, and the functional consequences of this polymerization. On a different subject, we are also interested in actin-based structures called “tumor-microtubes” that emanate from pancreatic cancer cells. Each of the lab’s interests is discussed below.
Endoplasmic reticulum-associated actin
Traditionally, the ER has not been thought of as a hotbed of actin. In the last few years, however, we and others have found that an ER-associated isoform of the formin INF2 can rapidly assemble actin filaments in response to a number of stimuli (see the video to the left). This “actin burst” has two immediate consequences that we know of: to stimulate transfer of calcium between ER and mitochondria (Chakrabarti et al (2018) J. Cell Biol); and to stimulate oligomerization of the dynamin GTPase Drp1, and its recruitment to mitochondria (Ji et al (2017) J. Cell Biol.). These events lead to mitochondrial division. Mitochondrial division is a key regulatory mechanism, controlling the activity these important cellular energy generators. In addition, the mitochondrial division machinery is important for two homeostatic processes: mitophagy and apoptosis. Defects in mitochondrial division are linked to multiple neurodegenerative diseases. Mutations in INF2 itself are linked to a neuropathy (Charcot-Marie-Tooth disease) and to a kidney disease (focal and segmental glomerulosclerosis). Our focus is to define three things in this process:
The activation mechanism of INF2. We now have exciting results showing that activation involves de-acetylation of actin! See Mu A et al, Nature Cell Biology (2019) and Mu A et al, PNAS (2020) for details. The regulatory mechanisms of actin acetylation/deacetylation on INF2 both biochemically and in cells are of very active interest to us.
The mechanisms by which INF2-mediated actin filaments stimulate mitochondrial calcium up-take, Drp1 recruitment, and mitochondrial division. We think that key events are occurring at ER-mitochondrial contact sites, but other effects might involve more wide-spread actin filaments. We know that myosin II is necessary for both mitochondrial calcium transfer and Drp1 recruitment, but we do not know the architecture of these filaments. This work will require both innovative cellular imaging and detailed biochemical studies.
Actin polymerization around stressed mitochondria
In addition to INF2-mediated actin filaments, a very different population of actin has very different effects on mitochondria. When mitochondria are stressed, they often get "depolarized." In other words, they lose their membrane potential, due to loss of electron transport chain activity which provides the proton gradient across the inner mitochondrial membrane (IMM). A very early response to mitochondrial depolarization is assembly of a dense actin “cloud” around the mitochondrion, which we call “ADIA” (Acute Damage-Induced Actin). ADIA is transient, peaking within 5 min and being gone by 10 min. We have shown that ADIA is independent of INF2, but depends upon another actin polymerization factor, Arp2/3 complex (Fung et al J. Cell Sci. (2019)). At a slightly later time, depolarized mitochondria undergo extensive shape changes, through remodeling of the IMM. We were surprised to find that ADIA does not induced these shape changes, but actually inhibits them. We are presently trying to determine the function of ADIA, with the working hypothesis that the cloud actually temporarily inhibits mitophagy, allowing the mitochondria a chance to recover. Intriguingly, actin polymerizes at least twice more at later stages of mitophagy: at 1-2 hrs after depolarization (PDIA, or Prolonged Damage-Induced Actin) and several hours after depolarization (AAA, or Autophagy-Associated Actin). While all of these polymerization events are through Arp2/3 complex, they are activated by different up-stream factors and clearly serve very different purposes. While we currently focus on ADIA, the ensemble of these responses is very intriguing.
In all of our research, we use a combination of cell-free biochemistry, live-cell microscopy, and other techniques such as fractionation of cellular extracts. We are trying to push our biochemical studies more toward reconstitution now, using full-length proteins under their proper regulation. These studies have lead to some striking insights that will be the subject of publications yet-to-come. Eventually, I hope, these will lead to cell-free reconstitution of ER-associated and mitochondrially-associated actin polymerization in the future.
Our live-cell studies are focused on establishing well-controlled, reproducible systems that can be quantified and used as reliable bio-assays to test biochemical predictions (“cellular biochemistry”). These experiments generate some striking images, but that is not our focus. In the era of the sound bites and tweets and “look at me!” science, we strive to take the long view. Of course, we’re not averse to the occasional tweet as well!
Learn more about the basics of actin biochemistry!
Tumor microtubes in pancreatic cancer
The laboratory has a long-standing interest in filopodia: thin finger-like protrusions from the plasma membrane. That interest has morphed into a fascination to amazing structures that go by several names: tunneling nanotubes, cytonemes, or tumor microtubes (TMTs). We somewhat stumbled across these when we saw them in abundance in a patient-derived pancreatic cancer cell line (Latario et al Mol. Biol. Cell (2020)). A number of things intrigue us about the cytoskeletal composition of TMTs. In particular, the actin filaments seem to be near the membrane, not running down the center of the TMT. That space is occupied by microtubules and intermediate filaments in abundance. So, TMTs clearly are no simply very big filopodia. Perhaps, they resemble more a neuronal axon or dendrite, as has been suggested for cytonemes. Of course, the function of TMTs is of huge interest. Some say that they might transmit cellular contents between cells, such as organelles like mitochondria, RNAs, or even bulk cytosol. We know that bulk cytosol is not occurring for our TMTs, but do not know about transfer of other components. We think the TMTs make adherens junction-type connections between cells.