Protein disaggregation by molecular chaperones

The problem with protein aggregation

Protein mis-folding and aggregation are deep constraints on the growth, development and life of all biological organisms.  Work on a variety of human diseases has shown that the incorrect folding and/or aggregation of important cellular proteins can be involved in serious pathologies.  Examples include cystic fibrosis, thalassemias, alpha1-antitrypsin deficiency, and a variety of amyloid neuropathies such as Alzheimer’s, Huntington’s and Parkinson’s disease.  Misfolded and aggregated proteins can even generate infectious particles that can, in the case of prion proteins, propagate their own formation upon spreading from cell to cell. Protein misfolding and aggregation also present serious issues for the large-scale expression of medically and industrially important proteins.


The role of molecular chaperones in protein disaggregation

The competition between productive protein folding and aggregation led to the early evolution of specialized molecular machines designed to deal with this problem. These machines are referred to as molecular chaperones. Many years of effort have uncovered some of the mechanisms employed by molecular chaperones, which can be broadly divided into three basic types: (1) passive blockage of aggregation, (2) direct facilitation of protein folding and (3) structural disassembly and disaggregation.  Some molecular chaperones are highly specialized for one of these functions, while other molecular chaperones appear to be capable of all three actions.  Of these general mechanisms, the disassembly and disaggregation of protein structures is the least well understood.  Our goal is to develop a detailed mechanistic understanding of how molecular chaperones recognize and dismantle protein aggregates.


The difficulty of stydying protein aggregation

The investigation of protein aggregation and disaggregation by molecular chaperones presents a serious technical challenge.  In general, an aggregating protein can rapidly populate a heterogeneous mix of assembled states that span several orders of magnitude in size.  The precise aggregation or disaggregation pathway followed can be an exquisitely sensitive function of environment such that shear forces, fractionation, or solvent conditions experienced during measurement can dramatically perturb the population distribution under study. 

While a variety of methods have been used to study protein aggregation, the inherent heterogenity of an aggregating protein sample and the likely importance of rare nucleating events and structural transitions, implies that a single-particle method capable of characterizing complex population distributions would be a powerful complement to other approaches.  One of the most versatile general strategies for observing single particles and detecting rare molecular events involves the optical fluorescence labeling of a target sample. Fluorescence labeling has proven to be a highly sensitive method for the single-particle detection of protein aggregates in a variety of assays.    


A single molecule approach

In order to study protein aggregation and disaggregation by molecular chaperones, we will employ both established single molecule methods, as well as develop novel techniques.  One new method we have developed in collaboration with Dr. Jason Puchalla and Dr. Robert Austin at Princeton University, is an intensity-based, single-particle method capable of resolving complex species distributions, while minimally perturbing a sample under study (Puchalla, J. et al.(2008), PNAS, 105: 14400-14405).  This technique, which we refer to as Burst Analysis Spectroscopy (BAS) permits the direct quantitation of the species distribution of fluorescently labeled protein assemblies at low concentration. 


Fluorescence Burst Analysis Spectroscopy (BAS)

BAS requires flowing fluorescently labeled particles through an excitation volume and measuring the photon bursts recorded by a sensitive detector.  Sample movement is accomplished by rapid rotation of the sample using a motorized open-loop stage or through continuous sample flow in a microfluidic device.  In a typical measurement, each labeled particle that transits through the excitation volume produces a fluorescent burst whose peak intensity value is determined, then stored for subsequent analysis. The sample scan rate is chosen to be much faster than the particle transit rate due to free diffusion but much slower than the fluorescent excitation/de-excitation rate.  For a typical confocal excitation profile under these conditions, the peak intensity of each burst occurs just as the particle crosses the central plane of the excitation volume where the laser intensity is highest.

A histogram of the excitation intensities corresponding to the central plane of the excitation volume is equivalent to a noiseless model of the burst amplitude histogram of a single, uniformly bright species transiting this volume.  Because there is a much greater probability of a particle crossing the beam at a region of low excitation intensity, the raw burst data will contain a large number of small-amplitude events.  To compensate, the collected burst data can be logarithmically binned in amplitude.  For a typical combined Gaussian–Lorentzian (GL) excitation model, logarithmic binning for a single, homogeneous species yields a histogram that is well approximated by a power law.  Importantly, additional noninteracting species of differing intrinsic brightnesses lead to simple and predictable offsets in the histogram up 
to the maximum burst amplitude of each new species. 


Extracting population distributions

In general, a particle distribution is likely to be composed of multiple subspecies each possessing a different intrinsic brightness. To recover the species number distribution, one needs to differentiate dim species passing through the center of the excitation profile from bright species passing through the edge of the profile. This problem can be solved by constructing a corrected burst histogram by using information about the most highly fluorescent species detected in the course of a measurement.  If all particles cross the excitation beam profile at random and uncorrelated positions, the most intense burst detected can, in principle, come only from the most highly fluorescent objects passing through the center of the excitation volume. The average number of bursts contributing to each of the lower intensity bins from the passage of the most fluorescent species through regions of lower excitation intensity can then be calculated.  After the contribution to each histogram intensity bin from the most fluorescent species is removed, any bursts that remain in the second brightest bin can come only from a second, but slightly dimmer, fluorescent species. This procedure can then be repeated.  Thus, each intensity bin can be corrected for the average contribution coming from each object of greater fluorescence intensity to yield a corrected histogram that estimates the number distribution.  By applying this approach in a time-resolved fashion, it is possible to track the population-resolved dynamics, in free solution, of complex assembly and disassembly events, like protein aggregation.










How do molecular chaperone networks disassemble protein aggregates?

Interestingly, different organisms organize their protein disaggregation machinery in different ways, with most bacteria, many unicellular eukaryotes and plants employing a core bi-chaperone network composed of an Hsp70 system coupled to an Hsp100 system.  Exactly how these two molecular chaperone systems cooperate to dismantle protein aggregates is not well understood.  In order to examine this problem, will employ the disaggregating chaperones of E. coli, consisting of the DnaKJ-GrpE Hsp70 system and the Hsp100 homolog ClpB, as a model.  While we will employ a variety of approaches in our work, we are particularly interested in studying this problem using minimally perturbing, free-solution single molecule and single particle fluorescence methods.  Our goal to develop a detailed mechanistic understanding of how this essential molecular chaperone network extracts and refolds proteins from aggregates.


Department of Biochemistry and Biophysics                           Texas A&M University              2017