Cellular Structures: Studying Physical Mechanisms of Cell Organization on Micron Length Scales

Top

00:00:06.22 My name is Mike Rosen and today
00:00:09.01 I'm going to tell you about a problem I see
00:00:10.27 as one of the really great problems today
00:00:12.28 in biology. And that is how eukaryotic cells are
00:00:15.28 organized on micron length scales.
00:00:17.23 Now I'll begin by telling you that eukaryotic cells
00:00:21.16 are organized across a huge range of length scales.
00:00:24.25 And the organization of individual atoms and protein
00:00:27.12 molecules to the assembly of those proteins into larger
00:00:30.13 and larger complexes, to the organization of those
00:00:33.29 complexes into distinct subcellular compartments,
00:00:36.19 out to membrane bound organelles in cells, out to
00:00:40.19 structures that consist of the entire length of the cell.
00:00:43.17 What I'll propose to you is that we understand the
00:00:47.04 organization on the shortest and longest length scales
00:00:49.28 quite well. So on the shortest scales, we understand a great
00:00:53.10 deal about how macromolecules function through
00:00:57.07 decades of structural biology and biophysical studies.
00:00:59.26 On the longer length scales, we understand through
00:01:03.15 cell biological studies that many of the membrane bound
00:01:06.25 compartments of the cell, for example, the nucleus,
00:01:09.08 the endoplasmic reticulum, golgi apparatus, the mitochondria,
00:01:12.22 all of those are very well understood in terms of their structures
00:01:16.09 and functions. However, there's an intermediate length
00:01:19.21 scale. Roughly 100 nm to 1 micron, on which we understand
00:01:26.03 the organization much, much, much more poorly.
00:01:28.15 Because there are in cells many structures of this
00:01:32.16 size that are not bound by membranes, that consist of
00:01:36.06 tens of thousands to hundreds of thousands of
00:01:39.05 individual macromolecules that are involved in many distinct
00:01:42.16 areas of biology. Now I'm just showing a short of panel of
00:01:47.00 those, a representative panel of those structures here.
00:01:50.11 So, for example, there are some structures, the JUNQ
00:01:54.10 and the IPOD structures that are involved in protein
00:01:56.22 quality control. There are structures in the nucleus called
00:01:59.25 PML nuclear bodies that are involved in DNA repair.
00:02:03.16 There are puncta of various metabolic enzymes that
00:02:07.09 have been observed in yeast and higher eukaryotes
00:02:10.09 that are believed to be involved in controlling metabolism.
00:02:12.28 There are puncta of numerous signaling molecules,
00:02:16.02 such as the disheveled protein here, that are involved
00:02:19.14 somehow in mediating signaling activities. There are RNA
00:02:22.23 granules of various flavors, for example, these stress
00:02:25.10 granules which are involved in translational control.
00:02:28.01 There are also numerous structures that exist in
00:02:31.22 the domain of the nucleus that are involved in organizing
00:02:34.25 of chromatin, which are believed to play an important role
00:02:37.20 in gene regulation. So these micron sized membrane-less
00:02:42.02 structures are found really throughout all of eukaryotic
00:02:45.17 biology. But we understand very little about their functions.
00:02:50.16 We understand a lot about the individual molecules that
00:02:54.02 control -- that are contained within these structures,
00:02:56.24 but we don't understand why those molecules, or how
00:03:00.10 those molecules are brought together into these micron
00:03:02.24 sized structures to impart new functionality to cells.
00:03:06.26 It turns out that there are also a number of potentially
00:03:11.00 analogous structures that are two-dimensional, that is
00:03:13.15 clusters of proteins on membranes. And those again
00:03:17.00 are important in very larges swathes of biology, and I've
00:03:20.09 just listed a few here. The immunological synapse, focal
00:03:24.24 adhesions, invadopodia structures; the initiation of
00:03:28.04 invadopodia structures are involved in T-cell activation
00:03:31.09 cell migration, tumor metastasis. And again, like the
00:03:35.08 three-dimensional equivalences, we know very little about
00:03:37.28 the nature of these structures and their functions in cells.
00:03:41.22 So, there are a series I think of fairly discrete questions
00:03:48.01 that we would really like to know about these structures.
00:03:51.05 To fill in the gap between macromolecular understanding
00:03:54.08 and membrane-bound structure understanding. So
00:03:58.21 first, what is the physical nature of the structure?
00:04:01.11 What are they? What are the molecules that compose them?
00:04:04.19 How are those held together to give a discrete structure?
00:04:07.20 What are their biochemical and cellular functions
00:04:11.00 of these cellular structures? And sort of to corollary
00:04:15.17 questions, how do the physical properties dictate the
00:04:19.18 biochemical and cellular functions? And finally, how
00:04:22.09 can the physical properties, and thus functions,
00:04:25.08 be controlled by cells to organize various biological processes?
00:04:29.24 So, these I think are fairly straightforward set of questions
00:04:34.25 but they're all very difficult questions. Let me spend a few
00:04:37.19 minutes explaining to you some of the challenges and
00:04:40.10 understanding them. And I think a lot of those challenges can
00:04:42.26 be illustrated by comparing a macromolecular assembly,
00:04:47.06 even a very large one like the eukaryotic ribosome, shown
00:04:50.13 here, to let's say a generic cellular punctum, these micron
00:04:54.08 sized membrane-less structures. So one of the first points
00:04:58.01 of comparison that one can make is the size.
00:05:00.05 So, even the eukaryotic ribosome, which is a gigantic
00:05:03.15 macromolecular machine, when you really scale it
00:05:08.21 to the size of one of these cellular puncta, it's miniscule.
00:05:11.23 So this is the size of a 25nm ribosome would be
00:05:16.19 compared to a 1 micron in diameter cellular punctum.
00:05:21.12 Perhaps even more importantly, though. Not only are these
00:05:25.11 much, much smaller, but the size of a macromolecular
00:05:27.27 complex is fixed by the components that compose it.
00:05:31.17 Whereas these cellular puncta, the sizes are highly
00:05:34.11 variable, sometimes they'll be small in cells, sometimes
00:05:37.03 they'll be large. They can transition from large to small
00:05:40.12 in response to different perturbations. Let me take the
00:05:43.13 ribosome back up to an unrealistic size to make my
00:05:46.25 additional comparisons. The composition is also very
00:05:50.29 different, macromolecular complexes are usually of
00:05:54.09 fixed stoichiometry such that there's always one copy
00:05:57.05 of this protein and one copy of that RNA. However,
00:06:00.19 the composition of these cellular puncta is highly
00:06:03.17 variable, molecules go in and out of them, some molecules
00:06:06.22 stay, some molecules go. They also are very different
00:06:11.17 to the level of structure, so macromolecular complexes are
00:06:14.19 stereochemically defined across their entire length. So
00:06:17.16 if I know where this amino acid is at one end, I know
00:06:21.02 the organization of that amino acid at the other end
00:06:23.16 of the molecule. However, these cellular puncta are almost
00:06:26.15 certainly not defined over their entire lengths. In fact,
00:06:29.21 we don't know on which length scales they are stereochemically
00:06:33.03 defined, and over what scale that definition decreases
00:06:36.03 as one moves out. They also are very different in
00:06:40.14 terms of dynamics. Much biochemistry of macromolecular
00:06:43.26 machines is governed by fluctuations on microsecond
00:06:46.29 and millisecond timescales. Whereas in these cellular
00:06:50.09 structures, yes the molecules within them move on those
00:06:53.11 timescales, but the molecules also move within these
00:06:56.15 structures and also between these structures and the
00:06:59.04 surrounding media on timescales that are much longer,
00:07:01.23 seconds to minutes. And frankly, it's really unclear at this
00:07:05.05 point which of those timescales of fluctuations are most
00:07:08.10 important to the function of these structures.
00:07:10.21 And finally, I think one of the most important issues
00:07:14.16 is function. So we know that macromolecular machines
00:07:18.04 function according to the domains that compose them.
00:07:21.24 Whereas, it's really unclear in these cellular structures
00:07:24.29 what functionality can be specifically ascribed to the
00:07:28.16 assembly itself and is distinct from the function of the
00:07:31.19 individual proteins within them. What does it mean to
00:07:34.13 organize them into this much, much larger structure?
00:07:36.22 We simply don't know at this point.
00:07:38.27 And finally, regulation, we know that macromolecular
00:07:43.11 machines are regulated often by alterations between
00:07:46.12 discrete configurations, inactive state and the active
00:07:49.29 state. However, because we don't understand the functions
00:07:53.10 of these cellular structures, it's unclear really what
00:07:56.23 should be controlled. What should we be looking for
00:07:59.02 to understand how function is altered by cellular
00:08:01.21 processes? Now, because of all of these difficulties
00:08:06.23 we really need new conceptual frameworks and new
00:08:10.03 experimental tools to understand these micron-sized
00:08:13.12 membrane-less cellular structures. Now I'm going to tell you
00:08:16.22 over the next few minutes a little bit about some ideas
00:08:20.26 that I and others in the field have had about this.
00:08:23.12 So, there are a couple of potential frameworks, conceptual
00:08:27.05 frameworks that derive from polymer science and soft
00:08:30.11 matter physics that may play some role here. And
00:08:33.22 one of them is this idea of polymerization, because
00:08:36.19 most simply, if you think, how do you make a big structure
00:08:38.24 from a bunch of little structures, the little structures may
00:08:41.14 just polymerize to make something much larger.
00:08:44.21 And when those assembling structures are multivalent,
00:08:48.12 systems can undergo what's called a cell gel phase
00:08:52.21 transition, such that in low fractional binding,
00:08:56.17 one gets distribution of small complexes. However,
00:09:00.02 as fractional binding goes up, there's a very sharp
00:09:02.18 line beyond which the system becomes capable of
00:09:05.19 assembling into huge macromolecular assemblies.
00:09:09.00 And it's possible then, that this soft gel transition
00:09:12.29 or polymerization in general, may contribute to these
00:09:16.03 micron sized structures. Another idea that's been
00:09:19.21 put forward recently is this notion of phase separations.
00:09:23.00 So I'm sure everyone's familiar if you mix oil and
00:09:25.25 water together and you shake them up, you can get these
00:09:28.28 little oil droplets that will float around in aqueous solution.
00:09:31.28 Well, it turns out, macromolecules, proteins and RNA,
00:09:35.11 in particular, can also do this. And in fact, as a system
00:09:41.13 polymerizes, as proteins polymerize they become less
00:09:44.02 soluble, they become more able to undergo this
00:09:47.18 kind of phase separation to produce concentrated
00:09:50.00 protein droplets, for example, in aqueous solution.
00:09:53.15 So again, it's been proposed that some kind of
00:09:55.23 phase separation may underlie many of these cellular
00:09:59.01 structures. I'm sure though, that with more thought and
00:10:02.19 more experiments, there are other conceptual frameworks
00:10:07.00 that also could be brought to understand these cellular
00:10:10.21 structures. In terms of tools, the tools that are used
00:10:16.21 to study macromolecular structures, x-ray crystallography,
00:10:20.15 NMR, electron microscopy, are really best to describe
00:10:24.04 discrete entities or fluctuations between discrete
00:10:27.23 conformational states. However, as I told you these cellular
00:10:31.22 structures are almost certainly not discrete entities.
00:10:34.02 And so we need new tools that will describe them as
00:10:37.06 distributions of structures, distributions of stoichiometries.
00:10:40.13 And probably also tools that will allow us to assess and
00:10:43.24 perturb the material properties, which may play a role in
00:10:46.09 the functions. And of course we need ways of relating
00:10:49.11 these properties to the individual macromolecular interactions
00:10:52.23 that govern their formation. In terms of function,
00:10:58.15 often in the fields of biophysics and cell biology,
00:11:01.02 biochemistry of purified components is used to understand
00:11:05.23 mechanistically, the functions of individual molecules
00:11:08.28 or individual entities. The problem is that most of these
00:11:13.11 cellular structures, in fact, can't be purified. And so we
00:11:16.17 need then ways of characterizing their biochemical
00:11:18.27 properties in situ, or to reconstitute these structures
00:11:22.11 from their individual parts. And perhaps most perplexing
00:11:26.00 at this point is, we need ways of directly interrogating
00:11:28.26 the functions of the assemblies themselves.
00:11:31.05 Again, as distinct from the components that compose them.
00:11:34.03 And we really need new experimental approaches to
00:11:37.05 address that issue of assembly. And finally, we need new
00:11:41.03 theoretical descriptions of these structures. And it may be
00:11:44.04 possible to take theories that exist already in polymer
00:11:47.09 science, colloid science, material science, and apply
00:11:50.18 them to biology. But an important distinction between
00:11:53.13 biology and these other physical sciences is that biology
00:11:56.07 is almost always held out away from equilibrium by the
00:11:59.21 consumption of energy through ATP hydrolysis. And so
00:12:02.26 one then needs to take these equilibrium studies and
00:12:06.24 apply them to biology by incorporating this idea of
00:12:09.29 systems away from equilibrium. And I know there are studies
00:12:14.04 of active matter in the soft matter physics world that
00:12:17.09 may very well be applicable to biology. So, let me just conclude
00:12:22.00 by coming back to what we would like to know in these systems.
00:12:24.15 What are they? What do they do? How does what
00:12:28.25 they are dictate what they do? And how can biology
00:12:31.29 control all of this to mediate various biological processes?
00:12:35.16 So I think if over say, the next decade, we can understand
00:12:38.19 all of these questions, we'll fill in this interesting and
00:12:44.00 confusing gap between macromolecules on one end
00:12:47.12 and membrane-bound organelles on the other, and
00:12:49.25 give us a thorough across scales understanding of biology.
00:12:54.02 So hopefully I've inspired some of you to think about this
00:12:57.11 problem and hopefully you, over the course of your career,
00:12:59.24 will be able to answer many of these questions.
00:13:01.25 Thank you.

This Talk

Speaker: Michael Rosen

Audience:

Educators of Adv. Undergrad / Grad
Researcher
Educators

Recorded: July 2014

Talk Overview

Michael Rosen argues that we have considerable insight into how cellular structures are formed at the smallest and largest scales (e.g., proteins and organelles, respectively). However, we know little about cell organization of intermediate-sized structures, such as cellular bodies and membrane puncta, which exist at the micron length scale. He considers the need for new tools, approaches, and ways of thinking to investigate the mechanisms of these structures and understand their functionality.

Speaker Bio

Michael Rosen

Michael Rosen is Chair of the Department of Biophysics at University of Texas Southwestern and an Investigator of the Howard Hughes Medical Institute. He studies the assembly and disassembly of actin filaments, a major component of the cytoskeletal network in cells, and focuses on understanding how these filaments organize into higher-order assemblies. Continue Reading