imaging technology a. cameron varano virginia polytechnic ... › bitstream › handle › 10919 ›...

REVEALING MOLECULAR ADVERSARIES OF HUMAN HEALTH USING ADVANCED

IMAGING TECHNOLOGY

A. Cameron Varano

Dissertation submitted to the faculty ofVirginia Polytechnic Institute & State Universityin partial fulfillment of the requirements for the degree of

Doctor of Philosophy In

Translational Biology, Medicine, and Health

Deborah Kelly Lisa Guay-Woodford

Steven Poelzing Zhi Sheng

November 27, 2018 Roanoke, Virginia

Keywords: cryo-electron microcopy, SiN microchips, rotavirus, liquid cell imaging, BRCA1, breast cancer

Copyright © 2018 A. Cameron Varano

Revealing Molecular Adversaries of Human Health Using Advanced Imaging Technology

A. Cameron Varano

ABSTRACT

Single particle electron microscopy (EM) allows us to examine the molecular world and gain

insights into protein structures implicated in human disease. Visualizing the 3D architecture of

the macromolecules can inform drug design and preventative care. While X-ray crystallography

and NMR are able to resolve atomic structures, the methodology is better suited for smaller

structures with limited flexibility. Single particle EM allows us analyze larger structures that

have inherent flexibility.

Protein structures can broadly be categorized as symmetry or asymmetric. There are

computational advantages when analyzing symmetrical structures. Specifically, structural

information can be extrapolated from fewer vantage points. Thus, symmetrical macromolecules

are an advantageous for pioneering new methodologies in single particle EM. Rotavirus double

layered particles (DLPs) are large macromolecular complexes that display icosahedral symmetry.

Previous studies have led to a high resolution structure of transcriptionally inactive rotavirus

frozen in time. However, to more fully understand rotavirus we need to examine the structure

under transcriptionally active conditions. To expand our understanding, we first evaluated these

viral assemblies using cryo-EM under active and inactive conditions. We found both internal and

external structural differences. Based on these findings we sought to further our understanding of

these nano-machines by developing a liquid cell environment to evaluate their dynamics over

time. Our research not only developed a new methodology to evaluate active particles over time,

we also found that the mobility of the DLPs were directly correlated to the level of

transcriptional activity.

When analyzing asymmetrical and flexible protein complexes previous studies have

utilized methodologies to limit the proteins’ conformational variability. While this does allow for

a higher resolution structure, it limits our understanding to a specific orientation and

compromises the biological insights. BRCA1 is an asymmetric protein containing a large flexible

region and is important in the prevention of breast cancer. We utilize silicon nitride microchips

with integrated wells and decorated with a lipid monolayer to capture and image BRCA1

complexes. This imaging platform minimizes heterogeneity and ensures the sample quality while

not biasing confirmation. Thus, allowing for high resolution cryo-EM imaging of flexible native

proteins. We were able to examine BRCA1 complexes from cells at both the primary and

metastatic sites. Our ability to visualize these proteins in their native form provide insights into

the variability of BRCA1 in disease progression. We found that BRCA1 complexes isolated from

metastatic cells have additional density in the C-terminal domain. Our data suggests this density

it due an interaction with p53.

Overall, our methodologies highlight the power of single particle EM for studying protein

complexes. Furthermore, our findings emphasize the importance of examining protein complexes

in their native state.

Revealing Molecular Adversaries of Human Health Using Advanced Imaging Technology

A. Cameron Varano

GENERAL AUDIENCE ABSTRACT

Single particle electron microscopy (EM) allows us to examine the molecular world and gain

insights into protein structures implicated in human disease. Visualizing the 3D architecture of

macromolecules can inform drug design and preventative care. While X-ray crystallography and

NMR are able to resolve atomic structures, the methodology is better suited for smaller

structures with limited flexibility. Single particle EM allow us analyze larger structures that have

inherent flexibility.

Protein structures can broadly be categorized as symmetry or asymmetric. There are

computational advantages when analyzing symmetrical structures. Specifically, structural

information can be extrapolated from fewer vantage points. Thus, symmetrical macromolecules

are an advantageous for pioneering new methodologies in single particle EM. Rotavirus double

layered particles (DLPs) are large, highly symmetrical macromolecular complexes that represent

an ideal model system for developing technology. Previous studies have led to a high resolution

structure of inactive rotavirus DLP frozen in time. However, to more fully understand rotavirus

we need to examine the structure under active conditions. To expand our understanding, we first

evaluated these viral assemblies using cryo-EM under active and inactive conditions. We found

structural differences. Based on these findings we sought to further our understanding of these

nano-machines by developing a liquid cell environment to evaluate their dynamics over time.

Our new methodology revealed new insights into the mobility of the DLPs.

When analyzing asymmetrical and flexible protein complexes previous studies have

utilized methodologies to limit the proteins’ movement. While this does allow for a higher

resolution structure, it limits our understanding to a specific orientation and compromises the

biological insights. BRCA1 is a highly flexible asymmetric protein implicated in the

development of breast cancer. We utilize specialized microchips to capture and image BRCA1

complexes. This imaging platform ensures sample quality and allows for high resolution cryo-

EM imaging of flexible native proteins. We were able to examine BRCA1 complexes from cells

at both the primary and metastatic sites. Our ability to visualize these proteins in their native

form provide insights into the variability of BRCA1 in disease progression. Our data found that

BRCA1 complexes isolated from metastatic cells are structurally different than those at the

primary site.

Overall, our methodologies highlight the power of single particle EM for studying protein

complexes. Furthermore, our findings emphasize the importance of examining protein complexes

in their native state.

vi

To my family and friends.

vii

Acknowledgements

I would like to express my sincere gratitude to my advisor, Dr. Deb Kelly, for her continuous and

enthusiastic support of my Ph.D. research. Her support in my development as a researcher and

member of the greater scientific community has filled my graduate years with growth and

continual exploration. Her encouragement of mid-day runs has kept my graduate years sane.

I also thank my dissertation committee: Dr. Lisa Guay-Woodford, Dr. Steven Poelzing and Dr.

Zhi Sheng, for their guidance and insights.

My sincere thanks go to Brian Gilmore and Yanping Liang, who not only taught me many

biochemical techniques, but spent time evaluating and analyzing data with me. For their

willingness to step away from their own work to help me I will always be grateful. I promise to

pay it forward.

Finally, I would like to thank my family and friends. To my family for supporting me in my

endeavors, sending texts and care packages for encouragement and understanding my absence at

holidays. To my friends for balancing out the stress of graduate school with intermural sports and

Sunday night get-togethers.

viii

Table of Contents

Chapter 1-Introduction ........................................................................................................ 1

References ....................................................................................................................... 5

Chapter 2- A Non-Symmetric Reconstruction Technique for Transcriptionally Active

Viral Assemblies ................................................................................................................. 6

Abstract ............................................................................................................................ 7

Introduction .................................................................................................................... 8

Materials and Methods .................................................................................................. 11

DLP Preparation ......................................................................................................... 11

Affinity-capture experiments ..................................................................................... 11

Electron Microscopy .................................................................................................. 12

Image processing and 3D refinement ......................................................................... 12

Results and Discussion .................................................................................................. 13

Images of transcriptionally-active DLPs reveal non-symmetric features .................. 13

3D classification routines can sort DLPs on the basis of structural variability.......... 16

3D reconstructions reveal a continuum of features that represent integrated protein-

RNA density ............................................................................................................... 18

Acknowledgement ......................................................................................................... 21

Author Contribution ...................................................................................................... 22

References ..................................................................................................................... 24

Chapter 3- Visualizing virus particle mobility in liquid at the nanoscale......................... 25

Abstract .......................................................................................................................... 26

ix


DLP preparation and activation.................................................................................. 26

Electron microscopy ................................................................................................... 27

Cinematography ......................................................................................................... 27

3D reconstructions...................................................................................................... 28


Acknowledgement ......................................................................................................... 40


References .................................................................................................................... 41

Chapter 4- Preparation of Disease-related Protein Assemblies for Single Particle Electron

Microscopy ....................................................................................................................... 42

Abstract .......................................................................................................................... 43

Introduction ................................................................................................................... 43

Materials ........................................................................................................................ 44

Cytoplasmic and Nuclear Extraction.......................................................................... 45

Nickel Chromatography Purification ......................................................................... 45

Carbon-coated TEM Grids ......................................................................................... 46

Preparation of uranyl formate heavy metal stain (1% w/v) ....................................... 46

Methods ......................................................................................................................... 47

Cytoplasmic and Nuclear Extraction Procedures ....................................................... 47

Nickel Chromatography Purification ......................................................................... 47

EM Specimen Preparation .......................................................................................... 51

Glow Discharge Procedure for Continuous Carbon Grids ......................................... 51

x

Preparation of Negatively Stained EM Specimens .................................................... 53

TEM Image Collection ............................................................................................... 56

Data Analysis and Representative CTD-FPC Results................................................ 56

Notes .............................................................................................................................. 57

Acknowledgements ....................................................................................................... 58


References ..................................................................................................................... 59

Chapter 5- Cryo-EM-on-a-chip: custom-designed substrates for the 3D analysis of

macromolecules ................................................................................................................ 60

Abstract .......................................................................................................................... 61

Introduction ................................................................................................................... 61


Simian rotavirus DLPs ............................................................................................... 63

Isolation of BRCA1 and p53 protein assemblies ....................................................... 63

SDS-PAGE and immunoblotting ............................................................................... 64

EM specimen preparation and data collection ........................................................... 64

Image Processing and movie production ................................................................... 65


The development of custom-design EM substrates.................................................... 66

Case study #1-Active virus assemblies ...................................................................... 70

Case study #2- Native protein assemblies isolated from breast cancer cells ............. 72

Structural analysis of a novel p53 assembly isolate from glioblastoma cells ............ 73

Conclusions ................................................................................................................... 76

xi



References ..................................................................................................................... 78

Chapter 6- Cryo SiN Platform Reveal BRCA1 Complexes in Metastatic Cancer ........... 80

Abstract .......................................................................................................................... 81

Introduction ................................................................................................................... 81


Authentication of cells, cell culture and BRCA1 enrichment procedures ................. 86

Coomassie blue staining and immunoblot analysis.................................................... 88

Co-immunoprecipitation analysis .............................................................................. 88

EM specimen preparation and imaging ...................................................................... 89

Image processing ........................................................................................................ 91

Movie production ....................................................................................................... 91

Results ........................................................................................................................... 92

BRCA1-BARD1 complexes bind to phosphorylated p53.......................................... 92

Imaging wild-type BRCA1 in metastatic disease ...................................................... 93

Structural spectrum of BRCA1-BARD1 complexes .................................................. 94

Discussion ...................................................................................................................... 95



References ..................................................................................................................... 99

Chapter 7- Conclusions ................................................................................................... 101

xii

List of Figures

Figure 2.1: Schematic to indicate viral processes that can occur during host-cell infection

and initial structural insights of active RV assemblies ....................................................... 9

Figure 2.2: EM images of actively transcribing DLPs show variable features among the

individual viral assemblies ................................................................................................ 15

Figure 2.3: Icosahedral reconstructions of active DLPs reveal internal structural features

........................................................................................................................................... 17

Figure 2.4: Non-symmetric reconstructions of active DLPs reveal structural variability in

a continuous manner throughout the density maps ........................................................... 19

Figure 3.1: Microfluidic system for in situ TEM imaging of virus particles .................... 30

Figure 3.2: Tethering system captures DLPs .................................................................... 31

Figure 3.3 DLPs imaged over time ................................................................................... 33

Figure 3.4: Visualizing and quantifying DLP mobility in liquid ...................................... 35

Figure 3.5: 3D structures of DLPs reveal internal density in liquid and ice specimens ... 37

Figure 4.1: Representative immunoblot ............................................................................ 50

Figure 4.2: Applying sample to EM grid .......................................................................... 52

Figure 4.3: Staining EM grid ............................................................................................ 54

Figure 4.4: Representative structural data and output ...................................................... 55

Figure 5.1: Workflow from cancer cells to protein structures .......................................... 67

Figure 5.2: Preparation of macromolecules for the “Cyro-EM-on-a-chip” technique ..... 69

Figure 5.3: Cryo-SiN specimen transfer and image comparisons .................................... 71

Figure 5.4: Structural and biochemical analysis of p53 assembly isolate from human

cancer cells ........................................................................................................................ 75

Figure 6.1: Schematic of the primary structures of p53, BRCA1, and BARD1 ............... 82

Figure 6.2: Schematic of silicon nitride microchip ........................................................... 85

Figure 6.3: An overview of the project work-flow from cells to biological evaluations and

imaging ............................................................................................................................. 87

Figure 6.4: The structural spectrum of BRCA1 complexes in breast cancer.................... 90

1

Chapter 1

Introduction

Structural biology provides literal insights into the infrastructure of life’s most basic properties.

The field of structural biology was revolutionized by the development of single-particle cryo

electron microscopy (cryo-EM). Prior to the advent of cryo-EM proteins and protein complexes

that were large or flexible were unknowable. While X-ray crystallography and NMR are able to

resolve atomic structures the methodology is better suited for smaller structures with limited

flexibility. However, for decades following the advent of the electron microscope significant

hurdles prevented it from being used to analyze single biological protein complexes. The major

hurdles to its use included: (1) transmission electron microscopy creates 2D projections, (2)

maintaining a stable sample in the high vacuum environment of the microscope, (3) the electrons

caused radiation damage to the sample, (4) the scattering pattern of biological sample makes

distinguishing signal from noise difficult.

These challenges were first overcome by David DeRosier and Aaron Klug in 1968 who

used Fourier-Bessel principles to construct a 3D model from 2D projections (1). In order to

overcome the stated obstacles, they created a dry-cast of the proteins in a heavy-metal salt,

uranyl acetate. This methodology is now referred to a negative stain, since structural information

is gained from the inverse pattern of the scattering. The heavy salts also minimize the damage

caused by radiation. The cost, however, was the limited resolution (~15Å). The limited

resolution was an inherent problem, given the size of the salt crystals and its propensity to flatten

fragile protein structures.

In 1974, Kenneth Taylor and Robert Glaeser showed that samples preserved in a frozen-

hydrated state were able to be withstand not only the vacuum and radiation of the microscope,

2

but the intricate features of the proteins were maintained (2). Thus, the field of cryo-EM was

born. Based on their work a team of researchers lead by Jacques Debochet developed the process

of vitrification in liquid ethane (3). In 1990, Richard Henderson resolved the first atomic

structure of bacteriorhodopsin.

The field of cryo-EM remained heavily reliant on structures which had a high level of

symmetry. Until, 1986, when Joachim Frank and Michael Radermacher developed Random

Conical Tilt, a set of algorithms which compared the relationship between the class averages of

2D projections and 3D alignments (4). These algorithms resulted in their development of

mathematical tools used for transforming particles from 2D micrographs into 3D reconstructions

(5).

In more recent years, the field of single particle cryo-EM has accelerated due to

advancements in image processing. Mathematical tools, such as RELION, have adopted more

sophisticated algorithms based on maximum-likelihood(6). Detectors are now able to digitally

and directly interpret electrons and thus maximize the signal to noise ratio (7). However, one

aspect which has remained relatively unchanged since the work of Debochet is sample

preparation. The reproducibility of sample preparation remains part scientific method and part art

form.

The ideal sample preparation results in vitreous ice formation with an ice-thickness not much

more than ~100nm, but thicker than the protein complex of interest. A sample frozen under the

same conditions can often produce sample grids that have noticeably different ice thickness.

Additionally, the protein particles must be kept away from the destructive air-water interface.

The Brownian motion created during the vitrification process can cause protein complexes to

collide with the air-water interface(8). When proteins come in contact with this interface their

3

unique 3D structure is lost becoming featureless spheres. These particles are referred to AWIPs

(air-water interface particles). AWIPs are not only lost for analysis, but if large enough they

interfere with imaging the complexes of interest. The formation of AWIPs is of particular

concern when imaging native proteins containing flexible regions. While the use of

glutaraldehyde cross-linking can minimize the challenges of protein flexibility (9), it is less than

ideal for imaging native protein complexes. The use of lipid monolayer help alleviate the

formation of AWIPs and optimize ice thickness (10). If the monolayer is further decorated with

immunocapture assemblies, the flexibility of the protein can be further stabilized for both native

and recombinant proteins(11). Despite the addition of monolayers, obtaining ideal ice thickness

has remained inconsistent. When examining native proteins, which inherently are less

concentrated, obtaining reliable ice thickness is critical to resolving the 3D architecture.

To that end, the research presented here provides a methodology which allows native protein

complexes to be examined in both a liquid and frozen environment. By employing traditional

cryo-EM methods, we examined double-layered rotavirus particles. We found the variability of

internal density of the virus to be related to its level of transcriptional activity. However, the

traditional methodologies used in cryo-EM limited the exploration of these findings. We,

therefore, created liquid cell imaging techniques to gain temporal insights. In order to limit the

movement, we utilized a silicon nitride (SiN) microchip with incorporated wells. The resulting

findings not only furthered our understandings of rotavirus structure during transcription, but of

the potential use of SiN microchips for imaging. We then combined the use of SiN microchips

with traditional cryo-EM sample preparation to image native BRCA1 complexes from metastatic

cancer cells. Our data provides insights into structural differences between BRCA1 isolated

from metastatic cancer cells when compared with protein isolated from primary cells. The data

4

suggest the structural differences can be attributed to an interaction with p53. These differences

are not distinguishable on the genomic level, highlighting the importance of examining cancer

associated protein complexes in 3D. Here we begin to construct a structural spectrum of BRCA1

complexes with respect to disease progression.

5

References

1. D. J. De Rosier, A. Klug, Reconstruction of three dimensional structures from electron

micrographs. Nature 217, 130-134 (1968).

2. K. A. Taylor, R. M. Glaeser, Electron diffraction of frozen, hydrated protein crystals. Science

186, 1036-1037 (1974).

3. J. Dubochet, Lepault, R., Freeman, R.. Berriman, J.A., Homo, J.C., Electron microscopy of

frozen water and aqueous solutions. Journal of microscopy 128, 219-237 (1982).

4. J. Frank, Three-dimensional electron microscopy of macromolecular assemblies : visualization of biological molecules in their native state. (Oxford University Press, Oxford ; New York, ed. 2nd,

2006), pp. xiv, 410 p.

5. J. Frank et al., SPIDER and WEB: processing and visualization of images in 3D electron

microscopy and related fields. J Struct Biol 116, 190-199 (1996).

6. S. H. Scheres, RELION: implementation of a Bayesian approach to cryo-EM structure

determination. J Struct Biol 180, 519-530 (2012).

7. A. R. Faruqi, R. Henderson, Electronic detectors for electron microscopy. Curr Opin Struct Biol

17, 549-555 (2007).

8. K. A. Taylor, R. M. Glaeser, Retrospective on the early development of cryoelectron microscopy

of macromolecules and a prospective on opportunities for the future. J Struct Biol 163, 214-223

(2008).

9. B. Kastner et al., GraFix: sample preparation for single-particle electron cryomicroscopy. Nat Methods 5, 53-55 (2008).

10. D. W. Taylor, D. F. Kelly, A. Cheng, K. A. Taylor, On the freezing and identification of lipid

monolayer 2-D arrays for cryoelectron microscopy. J Struct Biol 160, 305-312 (2007).

11. B. L. Gilmore et al., Preparation of Tunable Microchips to Visualize Native Protein Complexes

for Single-Particle Electron Microscopy. Methods Mol Biol 1764, 45-58 (2018).

6

Chapter 2

A Non-Symmetric Reconstruction Technique for Transcriptionally-

Active Viral Assemblies

Amina Rahimi1*, A. Cameron Varano1,2*, Andrew C. Demmert1,3, Linda A.

Melanson4, Sarah M. McDonald1,3,5, and Deborah F. Kelly1,3,6**

1. Virginia Tech Carilion Research Institute, Roanoke, VA 24016

2. Graduate Program in Translational Biology, Medicine, and Health, Virginia Tech,

Blacksburg, VA 24061, USA.

3. Virginia Tech Carilion School of Medicine, Roanoke, VA 24016

4. Life Sciences Division, Gatan, Inc., Pleasanton, CA 94588

5. Department of Biomedical Sciences and Pathology, Virginia-Maryland College of

Veterinary Medicine, Virginia Tech, Blacksburg, VA 24061

6. Department of Biological Sciences, Virginia Tech, Blacksburg, VA 24061

*These authors contributed equally to the work.

**Correspondence to: D.F.K. ([email protected]); Tel: +1-540-526-2031; Fax +1-540-985-3373.

mailto:[email protected]

7

Abstract

The molecular mechanisms by which RNA viruses coordinate their transcriptional activities are

not fully understood. For rotavirus, an important pediatric gastroenteric pathogen, transcription

occurs within a double-layered particle that encloses the viral genome. To date, there remains

very little structural information available for actively-transcribing rotavirus double-layered

particles, which could provide new insights for antiviral development. To improve our vision of

these viral assemblies, we developed a new combinatorial strategy that utilizes currently

available high-resolution image processing tools. First, we employed a 3D classification routine

that allowed us to sort transcriptionally-active rotavirus assemblies on the basis of their internal

density. Next, we implemented an additional 3D refinement procedure using the most active

class of DLPs. For comparison, the refined structures were computed in parallel by (1) enforcing

icosahedral symmetry, and by (2) using no symmetry operators. Comparing the resulting

structures, we were able to visualize the continuum that exists between viral capsid proteins and

the viral RNA for the first time.

8

Introduction

RNA viruses are ubiquitous in nature and represent some of the most severe pathogens known to

mankind (e.g., influenza A virus, Ebola virus, poliovirus, etc.). Rotaviruses (RVs) are non-

enveloped, eleven-segmented, double-stranded RNA viruses that can cause life-threatening

gastroenteritis in children. RVs express their genes as single-stranded, messenger RNA (mRNA)

molecules early in infection through the process of transcription [1]. RV transcription occurs

within intact, subviral double-layered particles (DLPs) within the cytoplasm of the host cell [2].

Transcriptionally competent DLPs (Figure 2.1A) are formed when the outermost VP7-VP4 layer

of the mature, triple-layered RV virions (TLP) are shed during cell entry [3]. DLPs primarily

contain viral genomic RNA that is encapsulated by a protective protein layer (VP6). Also found

within DLPs are core shell proteins (VP2), viral polymerases (VP1), and mRNA capping

proteins (VP3) that coordinate the production of mature mRNA transcripts [4]. In the cytoplasm,

viral transcripts are recognized by host cell ribosomes and act as functional templates for protein

synthesis (Figure 2.1A) [5]. Given the importance of RV transcripts during the viral replication

cycle, it is expected that abrogation of their synthesis by targeted antivirals will effectively

prevent RV-induced disease. Still, the molecular mechanisms by which RV DLPs are able to

function as such exquisite mRNA-producing nanomachines are not fully understood,

consequently hampering the design of pharmacological inhibitors.

9

Figure 2.1: Schematic to indicate viral processes that can occur during host-cell infection

and initial structural insights of active RV assemblies. (A) The infectious RV virion sheds its

outer most protein layer (VP4-VP7) as it enters the host cell forming a double-layered particle

(DLP). VP1 and VP3 proteins coordinate the production of mature mRNA transcripts of RV

genome segments. The mRNA transcripts are then extruded into the cytoplasm of the host cell.

Host ribosomes are responsible for translating the viral mRNAs into viral proteins. (B) Cryo-EM

3D reconstructions of DLPs (Class 1 - 3) having different levels of internal density. The diameter

of each reconstruction is 80 nm. (C) Representative images of negatively stain DLPs that

comprise similar Class 1, 2, or 3 reconstructions. The transcriptional activities of the particles

(none, moderate, or high) were interpreted based upon the amount of surrounding mRNA. White

arrows point to single-stranded mRNAs. These results suggest that Class 3 particles, which are

robustly transcribing, have the most ordered internal density. The width of each panel is 200 nm.

Figures in panels (B) and (C) were adapted from Kam et al., 2014 [10].

10

Three-dimensional (3D)structures of non-transcribing RV DLPs have been determined

to high resolution (3.8 Å) using cryo-electron microscopy (cryo-EM) or X-ray crystallography

[6,7]. Not only have these structures revealed atomic-level details of VP6 and VP2, theyhave

also provided insight into the location of viral polymerase complexes (comprised of VP1 and

VP3) within the particle [8]. Specifically, in the maps of the inactive DLP, a hub of density is

detected just below each 5-fold axis, which is partially attributed to VP1/VP3 complexes.

Unfortunately, the only available 3D structures of actively transcribing DLPs were determined

>15 years ago at lower resolution (25 Å) [5,9]. These active DLP structures show mRNA exiting

the DLP through VP2-VP6, but they do not reveal any detailed features of the particle.

Recent cryo-EM structural assessments performed by our laboratories revealed the

unexpected finding that transcriptionally-active DLPs exhibited varying degrees of disorder in

their external VP6 capsids [10]. Additionally, the DLPs with the most disordered external

density contained more internally ordered features within the 3D structure. Upon inspecting the

individual DLP assemblies that represented the active structures, we also found greater quantities

of mRNA surrounded the particles (Figure 2.1B &C, Class 3). Complimentary to this result,

particles exhibiting externally ordered capsids observed in the same cryo-EM images correlated

with very little internally ordered density; and the particles that constituted this structure were

surrounded by very little mRNA in the EM images (Figure 2.1B &C, Class 1). Yet another 3D

structure exhibited some external disorder and some internally ordered features, having a modest

degree of mRNA in the area surrounding the particles (Figure 2.1B &C, Class 2) [10]. Here, we

build upon these intriguing findings to gain a deeper understanding of protein and genomic RNA

arrangements within transcriptionally-active DLPs. Specifically, we address the following

questions: 1) in computing 3D structures of active DLPs, how does the use of icosahedral

11

symmetry during the refinement procedures influence the resulting density map; and 2) what

structural information can we gain about active viral assemblies without assuming icosahedral

symmetry during refinement procedures?

Materials and Methods

DLP preparation

Rotavirus (strain SA11-4F) DLPs were prepared as described in our previous work [10].

Transcription activation reactions (25-μl each) were performed in eppendorf tubes and each

mixture contained the following components: 1 μg DLPs prepared in 100 mM Tris-HCl pH 7.5,

6 mM MgAc, 4 mM DTT, 2 mM each of ATP, GTP, CTP, UTP, and 1μl RNasin (Promega

Corp., Madison, WI). The reactions mixtures were incubated for ~30 minutes at 37°C. After the

incubation period, aliquots of the transcription reactions (3-μl each) were used for Affinity

Capture experiments.

Affinity-capture experiments

EM Affinity grids containing Nickel-nitrilotriacetic acid (Ni-NTA) coatings were prepared using

holey carbon grids (C-flat - 2/1 grids; Protochips, Inc.) as previously described [10]. The Ni-

NTA functionalized layers were comprised of 25% Ni-NTA lipids and 75% 1,2-dilauryl-

phosphatidylcholine (DLPC) filler lipids (Avanti Polar Lipids). Protein adaptors were

sequentially added to the Ni-NTA coated grids. We first added aliquots (3-μl) of His-tagged

Protein A (0.01 mg/ml) (Abcam) in buffer solution containing 50 mM HEPES, pH 7.5, 150 mM

NaCl, 10 mM MgCl2 and 10 mM CaCl2. The Protein A solution was incubated for 1 minute on

each grid and the excess solution was blotted away with filter paper. Next, 3-μl aliquots of VP6-

12

specific guinea pig polyclonal antisera (#53963) (0.01 mg/ml) contained in the same HEPES

buffer solution were added to Protein A-decorated grids. After a 1-minute incubation step, the

excess solution was gently removed using a Hamilton syringe. Finally, the transcriptionally

active DLPs (2-μl aliquots of 0.1 mg/ml) were added to the antibody-decorated grids for a 2-

minute incubation. Frozen-hydrated specimens were prepared by plunge-freezing the grids into

liquid ethane slurry using a Gatan Cryoplunge™ 3 equipped with GentleBlot capabilities (Gatan,

Inc.) while employing a one-sided blotting routine for approximately 8 seconds.

Electron Microscopy

Frozen-hydrated specimens were transferred to a Gatan 626 cryoholder (Gatan, Inc.) and

maintained under liquid nitrogen until transferred into the TEM. Specimens were imaged using a

FEI Tecnai Spirit BioTwin TEM (FEI, Co., Hillsboro, OR) equipped with a LaB6 filament and

operated at an acceleration voltage of 120 kV under low-dose conditions (~5 electrons / Å2).

Images of transcriptionally-active DLPs were recorded on a FEI Eagle 2k HS CCD camera with

a pixel size of 30-μm using a defocus range from -1.0 μm to -3.0 μm at a nominal magnification

of 60,000× for a final sampling of 5 Å / pixel at the specimen level.

Image processing and 3D refinement

Images of transcriptionally-active DLPs were recorded as described above and individual

particles were selected from the images using automated routines in the PARTICLE software

package (http://www.image-analysis.net/EM). Prior to particle selection, the raw images were

inverted, normalized and CTF-corrected using standard routines in PARTICLE. The final image

stack containing 2769 particles was exported in MRC format for 3D reconstruction and

13

refinement calculations in the RELION software package [11]. Within the RELION package, we

used a reference map for the DLP structure [7] that was available from the website of the

Grigorieff laboratory. Implementing the RELION program, 3D classification routines identified

3 distinct classes while enforcing icosahedral symmetry through 25 cycles of refinement. The

particles that comprised the 3D structure and displayed the most disordered exterior exhibited

well-defined density within the internal particle. These particles were then subjected to 10

subsequent rounds of refinement while enforcing either C1 or I1 symmetry, and while dividing

the particles into two equal halves. The resulting 3D structures were resolved to ~15 Å as

independently verified using the RMEASURE program [12]. Overall parameters for the global

reconstruction routines included a pixel size of 5 Å, a reference model low-pass filtered to 50 Å,

and a regularization parameter of T = 4 over an angular search space of 7.5 degrees.

Results and Discussion

Images of transcriptionally-active DLPs reveal non-symmetric features

Improving our understanding of protein-RNA interactions can provide important new insights to

dissect infection at the molecular level. The dynamic manner in which these interactions occur

remains unclear, especially in the case of viral mRNA production. To expand upon our previous

findings of transcriptionally-active RV DLPs, we implemented a new computing protocol to

better visualize the dynamic nature of these complexes. We collected images of frozen-hydrated

transcriptionally-active DLPs that were tethered to C-flat holey carbon Affinity Grids

(Protochips, Inc.). These grids were decorated with adaptor molecules including IgG antibodies

against the VP6 capsid protein (please see Methods section). EM images of the DLP specimens

14

were acquired using a FEI Tecnai Spirit BioTwin TEM (FEI, Co., Hillsboro, OR) equipped with

a LaB6 filament and operated at an acceleration voltage of 120 kV under low-dose conditions (~5

electrons / Å2).

The raw images (Figure 2.2A) revealed DLP assemblies having a diameter between 80 -

90 nm, some of which were associated with RNA strands, consistent with our previous

observations [10,13]. A number of particles in the images displayed strongly symmetric features

although these attributes varied among the population. To further understand the structural

variation in the DLPs, we processed the images for downstream reconstruction procedures.

Using the PARTICLE software package (http://www.image-analysis.net/EM), we inverted and

normalized each image (Figure 2.2B) then implemented standard routines for CTF correction

(Figure 2.2C). Individual DLPs were selected from the corrected images using the automated

processes implemented in PARTICLE as previously described [10]. The resulting image stack

containing 2769 individual particle images was imported into the RELION software package for

3D reconstruction and refinement routines.

15

Figure 2.2: EM images of actively transcribing DLPs show variable features among the

individual viral assemblies. (A) Representative EM image of transcriptionally active DLPs in

ice. The image contrast was inverted and normalized (B) then CTF-corrected (C) using standard

procedures in the PARTICLE software package, providing refined images for 3D reconstruction.

Insets highlight individual DLPs with surrounding mRNA. Scale bar is 100 nm.

16

3D classification routines can sort DLPs on the basis of structural variability

The EM images clearly indicated that particle variability existed within the active DLP

population. Hence, computing a variety of 3D structures may better represent the inherent

mobility present among the diverse population of complexes. We used a standard 3D

classification routine within the RELION software package to test for the number of statistically

significant 3D structures present in our stack of DLP images. Based on Bayesian inference and

utilizing a filtered starting model (50-Å) of the DLP structure [7], RELION identified three

distinct structures present in our images after 25 cycles of refinement. This observation is

consistent with our previous structures (Figure 2.1B) that were computed by applying

icosahedral symmetry operators (I1 space group) during the 3D classification process [10].

Again, we noted that the structures having distinctly ordered internal features correlated with the

greatest degree of proximal mRNA.

Overall, the 3D classification routine reproducibly distinguished various structures

among the active particles. However, considering the fact that not all DLPs displayed perfect

symmetry in the images, we investigated whether applying the icosahedral symmetry operator to

the active DLPs obscured information about protein-RNA arrangements. We performed

additional refinement calculations using only the particles contained within the presumed most

active class (Figure 2.1, Class 3). These DLPs contained the strongest internal features and had

greater quantities of nearby mRNA strands in the images. We calculated new 3D structures of

these assemblies using the RELION software package, with and without the use of symmetry

operators (I1 vs C1 space groups). In each case, we used the same reconstruction parameters,

varying only the symmetry operators during the calculations. We implemented 10 refinement

cycles in RELION while dividing the data into two halves for direct comparison.

17

Figure 2.3: Icosahedral reconstructions of active DLPs reveal internal structural features.

Particles that comprised the Class 3 structure shown in Figure 2.1B were refined using the

RELION software package while enforcing icosahedral symmetry. The images were equally

divided into two halves (blue and yellow) during the refinement procedure. Contoured sections

are shown at 10-nm intervals through the density ending at the midsection of the particle. 3D

reconstructions are ~80 nm in diameter. Scale bar is 20 nm.

18

3D reconstructions reveal a continuum of features that represent integrated protein-RNA

density

Upon examining the resulting reconstructions generated from applying icosahedral symmetry,

we visualized strong internal density along the 5-fold axis in contoured sections through the EM

structures (Figure 2.3). Protein subunits arranged distinctively along the 5-fold axis have been

previously identified as VP1 and VP3 in non-transcribing DLPs. Our new symmetrized density

maps resolved to ~15 Å-resolution also revealed strong globular density at this position and is

consistent with the work of Lawton and colleagues [5,9]. By examining through the sections of

the density map, we also visualized features that collectively defined the protein-RNA

contribution to the particle integrity. This information was missing from the original analyses of

non-transcribing DLPs, which were focused on defining high-resolution features in the VP6

protein capsid [7].

Upon examining the 3D density maps calculated from the same images of active DLPs,

but without imposing icosahedral symmetry, we found very different results (Figure 2.4A). The

overall structures appeared to be strongly assymetric, but showed a continuum of density

throughout the maps. In comparing the same cross-sections of the non-symmetrized

reconstructions with those of the icosahedral structures, we found that the inherently disordered

C1 structures did not show distinct 5-fold channels, nor did they provide a clear view of

individually distinct subunits. However, when we consider the fact that the active DLP

assemblies that comprised the reconstructions were similarly lacking in strong symmetry

elements, it seems reasonable that the more dynamic reconstructions may also provide a realistic

view of the most active DLPs structures.

19

Figure 2.4: Non-symmetric reconstructions of active DLPs reveal structural variability in a

continuous manner throughout the density maps. (A) Particles that comprised the Class 3

structure shown in Figure 2.1B were refined using the RELION software package in the C1

space group. The images were equally divided into two halves (blue and yellow) during the

refinement procedure. Contoured sections are shown at 10-nm intervals through the density

ending at the midsection of the particle. 3D reconstructions are ~80 nm in diameter. (B) As an

internal control, we calculated EM reconstructions using the C1 space group for inactive DLPs.

The resulting structures were highly symmetric as expected, considering the strong symmetry

elements that were visible within the individual particles that constituted this population. Scale

bar is 20 nm.

20

As a control experiment, additional 3D structures were calculated in the C1 space group

for non-transcribing particles contained within the same image stack (Figure 2.4B). For the

reconstructions of non-transcribing assemblies that were calculated without imposing icosahedral

operators, we still observed strong symmetry elements within the resulting structures as expected

for the inactive assemblies. As an external control, we also examined cryo-EM images of non-

transcribing DLPs prepared in buffer solution but lacking the required elements to induce

transcriptional activity. EM structures computed in the C1 space group for these inactive

assemblies indicated a statistically single population having features consistent with the

reconstructions shown in Figure 2.4B.

In comparing our results to the high-resolution structures of the transcriptionally-active

subviral particle of cypovirus, a member of the same viral family as RV (i.e., Reoviridae), we

found an essential commonality – a series of conformational changes must occur in the viral

capsid during mRNA synthesis, capping, and egress [14]. It remains unknown whether this

process is being driven by the structural mobility of the internal RNA genome or by fluid protein

rearrangements in response to the changing RNA landscape during transcription.

Collectively, the experimental results presented here complement the multiscale

molecular dynamics simulations of Miao and Ortoleva [15]. Their studies showed that structural

transitions in viral capsids begin locally then propagate across the continuum of the outer protein

layer. Their findings also support the idea of nonsymmetric intermediate states among ordered,

icosahedral viral assemblies. In particular, they suggested that the use of symmetry elements

alone does not reveal the full regime of viral structural transitions [15].

Additional work by Vlijmen and Karplus [16] demonstrated the existence of structural

variability among viral transition states using normal mode analysis while applying icosahedral

21

symmetry. Although both native and altered states of viral capsids displayed symmetrical

elements, the transitional pathway to achieve these variable conformations was not ordered. In

fact, less than 2% of viral capsid motions were considered to be symmetrical in nature, indicating

that asymmetrical transition states are more likely to dominate atomic fluctuations [16].

Finally, studies performed by Brooks et al. [17,18] described the need to examine viral

capsid reconstructions with and without enforcing symmetry elements. Their reasoning for these

considerations draws from the fact that for part of the viral lifespan, protein capsids can assume

an icosahedral symmetric state. However, nonsymmetric intermediates must also exist due to the

elastic properties of the capsid constituents and changes in morphology permitted by buckling

transition theory [17,18].

Based on these theoretical frameworks, and considering the continuum of density

visualized in the more disordered structures presented here, it seems feasible that the movements

within the DLP cores are influenced by RNA-induced mobility in cooperation with internal

protein subunits. As such, the transitional movements in the VP6 capsid are likely to be

asymmetric due to the natural fluctuations needed to accommodate RNA synthesis and extrusion.

We expect similar movements may occur in other pathogenic RNA viruses, and that our new

strategies can be broadly utilized to provide new insights for structure-based drug design.

Overall, we anticipate future work involving in situ TEM analysis of transcriptionally-active

viral assemblies within a liquid environment may reveal additional information to delineate real-

time mechanisms within these exquisite nanomachines.

Acknowledgements

The authors acknowledge the following funding sources that supported this work, 1R21AI113402-

01 NIAID/NIH (D.F.K., S.M.M) and 1R01AI116815-01 NIAID/NIH (D.F.K., S.M.M).

22

Author Contribution

D.F.K. conceived this project and supervised the completion of all experiments. A.R., A.C.V.

and D.F.K. prepared the manuscript. A.C.V., A.R., and A.C.D. performed the experiments.

L.A.M. provided technical support. S.M. provided reagents and technical support.

23

References

1. McDonald SM (2013) RNA synthetic mechanisms employed by diverse families of RNA

viruses. Wiley Interdiscip Rev RNA 4: 351-367.

2. Jayaram H, Estes MK, Prasad BV (2004) Emerging themes in rotavirus cell entry, genome

organization, transcription and replication. Virus Res 101: 67-81.

3. Trask SD, McDonald SM, Patton JT (2012) Structural insights into the coupling of virion

assembly and rotavirus replication. Nat Rev Microbiol 10: 165-177.

4. Prasad BV, Chiu W. (1994) Structure of rotavirus. Curr Top Microbiol Immunol 185: 9-

29.

5. Lawton JA, Estes MK, Prasad BV (1997) Three-dimensional visualization of mRNA

release from actively transcribing rotavirus particles. Nat Struct Biol 4: 118-121.

6. McClain B, Settembre E, Temple BRS, Bellamy AR, Harrison, SC (2010) X-ray Crystal

Structure of the Rotavirus Inner Capsid Particle at 3.8 angstrom Resolution. J Mol Biol

397: 587-599.

7. Zhang X, Settembre E, Xu C, Dormitzer PR, Bellamy R, Harrison SC, Grigorieff N. (2008)

Near-atomic resolution using electron cryomicroscopy and single-particle reconstruction.

Proc Natl Acad Sci U S A 105: 1867-1872.

8. Estrozi LF, Settembre EC, Goret G, McClain B, Zhang X, Chen JZ, Grigorieff N, Harrison,

SC (2013) Location of the dsRNA-dependent polymerase, VP1, in rotavirus particles. J

Mol Biol 425: 124-132.

9. Lawton JA, Estes MK, Prasad, BV (1999) Comparative structural analysis of

transcriptionally competent and incompetent rotavirus-antibody complexes. Proc Natl

Acad Sci U S A 96: 5428-5433.

10. Kam JA, Demmert AC, Tanner JR, McDonald SM, Kelly DF (2014) Structural Dynamics

of Viral Nanomachines. Technology 2(1): 1-5.

11. Scheres SH (2012) A Bayesian view on cryo-EM structure determination. J Mol Biol 415:

406-418.

12. Sousa D, Grigorieff N (2007) Ab initio resolution measurement for single particle

structures. J Struct Biol 157: 201-210.

13. Tanner JR, Demmert AC, Dukes MJ, Melanson LA, McDonald SM, Kelly DF (2014)

Cryo-SiN – An Alternative Substrate to Visualize Active Viral Assemblies. J Anal Mol

Tech 1 (1): 1-6.

14. Yang CW, Jia G, Liu HR, Zhang K, Liu GQ, Sun F, Zhu P, Cheng LP (2012) Cryo-EM

structure of a transcribing cypovirus. Proc Natl Acad Sci U S A 109: 6118-6123.

15. Miao Y, Ortoleva PJ (2009) Viral Structural Transition Mechanisms Revealed by

Multiscale Molecular Dynamics/Order Parameter eXtrapolation Simulation. Biopolymers

93 (1): 61-72.

16. van Vlijmen HW, Karplus M (2005) Normal Mode Calculations of Icosahedral Viruses

with Full Dihedral Flexibility by Use of Molecular Symmetry. J Mol Biol 350: 528-542.

24

17. May ER, Brooks III CL (2012) On the Morphology of Viral Capsids: Elastic Properties

and Buckling Transitions. J Phys Chem B 116 (29): 8604-8609.

18. May ER, Feng J, Brooks III CL (2012) Exploring the Symmetry and Mechanism of Virus

Capsid Maturation Via an Ensemble of Pathways. Biophys J 102: 606-612.

25

Chapter 3

Visualizing virus particle mobility in liquid at the nanoscale

A. Cameron Varano1,2, Amina Rahimi1, Madeline J. Dukes3, Steven Poelzing1,

Sarah McDonald1 and Deborah F. Kelly1

1. Virginia Tech Carilion Research Institute, Roanoke, VA 24016



3. Application Science, Protochips, Inc. Raleigh, NC 27606, USA.

26

Abstract

Currently, there remains a critical need to develop real-time imaging resources for life sciences.

Here, we demonstrate the use of high resolution in situ imaging to observe biological complexes

in liquid at the nanoscale. Using a model virus system, we produced the first time-resolved

videos of individual biological complexes moving in solution within an electron microscope.

Materials and Methods DLP preparation and activation

Simian rotavirus (strain SA11-4F) DLPs were purified as previously described (1). Transcription

reactions (25-μl each) were in carried out in eppendorf tubes incubated for ~30 minutes at

37°C.6 Briefly, each mixture contained the following: 1 μg DLPs prepared in 100 mM Tris-HCl

pH 7.5, 6 mM MgAc, 4 mM DTT, 2 mM each of ATP, GTP, CTP, UTP, and 1μl RNasin

(Promega Corp., Madison, WI). Following the 30-minute incubation period, 3-μl aliquots of the

reaction mixtures were applied to antibody-decorated SiN chips used for subsequent

experiments. Enzymatically active DLPs (2-μl aliquots of 0.1 mg ml-1) were added to the

antibody-decorated microchips for a 2-minute incubation. Microchips containing the tethered

DLPs were then loaded into the Poseidon 200 liquid specimen holder (Protochips, Inc.) for in

situ TEM imaging. Antibody-tethered grids for cryo-EM control experiments were produced

using the same procedures, but upon holey carbon grids (C-flat - 2/1 grids; Protochips, Inc.)

rather than microchips. Frozen-hydrated specimens were plunge-frozen into a liquid ethane

slurry using a Gatan Cryoplunge™ 3 equipped with GentleBlot capabilities (Gatan, Inc.) and

employing a one-sided blotting step for approximately 8 seconds.

27

Electron microscopy

All specimens were eamined using a FEI Spirit BioTwin TEM (FEI Company, Hillsboro, OR,

USA) equipped with a LaB6 filament operating at 120kV under low-dose conditions (< 1

electron per Å2). Images were recorded using a FEI Eagle 2k HS CCD camera having a pixel

size of 30 μm. Images of DLPs in liquid and in ice were recorded at a nominal magnification of

60,000 × with a final sampling of ~5 Å per pixel using a defocus range of -1.5 to -3 μm. For

image series acquisition, we collected sequential images at intervals ranging from 0.25 – 1 s-1.

For the image series analyzed here, we selected representative DLPs from the images using the

PARTICLE software package (http://www.image-analysis.net/EM/). For cryo-EM imaging, we

employed the same TEM and imaging parameters.

Cinematography

Intact particles sufficiently distanced from each other were boxed out to create an image stack

through the acquired image series using the PARTICLE software package. The images in each

stack were uniformly masked to include information with an 80-nm diameter then colored for

visualization purposes and movie production (Movies S1 3 – S3). Images within each stack were

also subjected to a density threshold using a significance level cutoff of 3σ as described in other

work (2, 3). Contour maps of the remaining density were compiled for movie production and

quantitative analysis. Images sequences and contour maps were imported into the iMovie 10.0.7

software package (Apple, Inc.) and the movies were compiled to show images cycling at 0.5-

second intervals. The movies were exported .mov format.

28

3D reconstructions

Individual DLPs were selected from EM images using the PARTICLE software package

utilizing a box size of 120 nm. The selected particles were output as MRC image stacks and

imported into the RELION software package (4) for 3D reconstruction calculations. Within the

RELION package we used refinement parameters included a pixel size of 5 Å, a reference model

low-pass filtered to 50 Å, and a regularization parameter of T= 4. We enforced icosahedral

symmetry over an angular search space of 7.5° while implementing refinement procedures for 25

cycles outputting the reconstructions in Figure 3.5 A and B with a resolution of ~2.8 and ~2.5

nm, respectively. Slices through each of the reconstructions revealed internal densities of the

particles. The slices were taken at ~20 nm intervals ending at the midsection of each structure.


Recent advances in the design and manufacture of nanometer-thick materials has spurred a new

era in the imaging field (5, 6). When used in combination with microfluidic devices(7), it is now

possible to observe electrochemical processes(8), the growth of nanoparticles (9), drug delivery

systems (10), and live cellular targeting events (11) at unprecedented resolution. We refer to this

new innovative mode of imaging as “in situ” microscopy as experiments can now be directly

observed “inside” a transmission electron microscope (TEM). Although in situ imaging has

gained a great deal of momentum in the material science field, its application to biological

systems is not as widespread. To address this issue, we developed a unique platform to directly

observe movements in biological complexes using in situ TEM. Here, we present the first time-

resolved movies and mathematical of a model virus system moving in solution at the nanoscale.

Specifically, we used purified rotavirus double-layered particles (DLPs) in our experiments since

29

DLPs can be enzymatically activated in vitro by the addition of nucleotide triphosphates (NTPs)

to initiate viral mRNA synthesis (12, 13).

To observe rotavirus DLPs in a self-contained liquid environment, we used the Poseidon 200 in

situ specimen holder that can accommodate nanoliter volumes of solution between two tightly

sealed silicon nitride (SiN) microchips (Figure 3.1A). These chips contained central imaging

windows (~30-nm thick) that are transparent to the electron beam and form an environmental

chamber at the tip of the holder. The bottom chip that defines the base of the assembly contained

a 400 × 400 micron array of microwells within the central imaging window (10). Each

microwell (10 × 10 microns) was etched down to accommodate a solution thickness of 150 nm

(Figure 3.1B). To use this system optimally, we first tethered purified, enzymatically activated

DLPs (Figure 3.2A) to the microchips that were coated with a tunable substrate. This tethering

step was used to limit diffusion of the viral particles in solution during imaging (Figure 3.1 C, D,

E). The coating applied to the microchips contained Nickel-nitrilotriacetic acid (Ni-NTA), a

substance that binds with high affinity to histidine (His)-tagged recombinant proteins (14, 15).

After the Ni-NTA coating was applied to the microchips, we added 3-μl aliquots of protein A

solution (0.01 mg ml-1) containing 50 mM HEPES (pH 7.5), 150 mM NaCl, 20 mM CaCl2 and

20 mM MgCl2. The excess solution was removed from the microchips followed by the addition

of guinea pig antisera (3-μl aliquots) containing IgG antibodies against the VP6 viral protein,

diluted to 0.01 mg ml-1 in the same HEPES buffer solution. In parallel as negative controls, we

also prepared 1) DLPs samples lacking ATP, a nucleotide needed for enzymatic activation, and

2) protein-A coated microchips lacking IgG antibodies against VP6. All samples were imaged

using cryo-EM to assess the specificity of the method. We found that within a 2-minute

incubation period, the antibody-decorated microchips could recruit DLPs diluted to 0.1 mg ml-1

30

Figure 3.1. Microfluidic system for in situ TEM imaging of virus particles. (A) The

microfluidic chamber was assembled by placing O-ring fittings in the empty tip of the holder

(steps 1 - 3). A flat chip was placed over top of the wet specimen chip (step 4) and the assembly

was covered with a metal face-plate held in place by 3 brass screws (step 5 - 6). (B) Schematic

of a cross-section through the fluidic chamber assembly positioned within the EM beam.

Integrated microwells (10 × 10 m) etched into the chip can each accommodate a solution layer

of 150 nm thick. (C) Representative image of DLPs in liquid. (D) The base chip contained a

transparent imaging window having array of microwells (400 × 400 µm). DLPs were tethered to

the base chips and image series were recorded in the EM. (E) Representative DLPs (1 - 3) with

viral mRNAs transcripts (black arrows). Width of individual panels is ~120 nm. Individual DLPs

were contrast-inverted, masked by a diameter of ~80 nm and colorized for visualization

purposes. Please see Movies 1 – 3.

31

Figure 3.2 Tethering system captures DLPs. Tethering DLPs to functionalized surfaces. (a)

Schematic to illustrate the immunocapture procedure used to tether asynchronously transcribing RV

DLPs to antibody (IgG)-decorated surfaces via protein A adaptors. (b) Transcribing DLPs tethered to

EM grids or SiN microchips in the presence of VP6-specific IgGs showed varying lengths of

associated mRNA transcripts (white arrows) in cryo-EM images. (c) DLPs prepared in the absence of

nucleotides needed for transcription do not show associated mRNA in cryo-EM images. (e) EM

specimens prepared in the absence of IgGs generally failed to recruit DLPs. Scale bar is 100 nm.

32

in 50 mM HEPES buffer solution (pH 7.5) containing 150 mM NaCl, 20 mM CaCl2 and 20

mM MgCl2, and NTPs (Figure 3.2 B, C). DLPs lacking nucleotides could still bind to antibody-

decorated surfaces but viral mRNA transcripts were not present in the images (Figure 3.2 C).

Microchips prepared without antibodies did not recruit appreciable quantities of DLPs (Figure

3.2D).

To biochemically verify transcriptional competence of the viral particles, the purified DLPs

were subjected to in vitro mRNA synthesis in Eppendorf tubes in the presence of [32P]-UTP for

30 minutes at 37°C. Each reaction contained DLPs, NTPs, and [32P]-UTP. Negative control

reactions contained all transcription cocktail components except ATP. Radiolabeled mRNA

products were only detected in the reactions containing complete transcription cocktail; no

radiolabeled products were seen in the reaction lacking ATP. Therefore, our functional analysis

confirmed that purified DLPs could be enzymatically activated. These results were consistent

with previous experiments (15) and suggested that the system could be used to recruit and image

active DLPs in solution. Therefore, we proceeded to record images of the virus particles using in

situ TEM. For in situ imaging experiments, we loaded microchips containing tethered DLPs into

the Poseidon 200 specimen holder (Figure 3.1A, B). A metal faceplate, held in place by 3 brass

screws, was placed on top of the assembled unit. The enclosed DLPs were examined using a FEI

Spirit Bio-Twin TEM equipped with a LaB6 filament and operating at 120kV. We recorded at

least five image series at time intervals up to 10 seconds (4 frames second-1) while optimizing

output conditions. Images were recorded on an Eagle 2k HS CCD camera employing low-dose

conditions (~0.1 electrons per Å2) at a 60,000× magnification, yielding a final sampling of 5 Å

per pixel. Image series proved reproducible between replicates.

33

Figure 3.3 DLPs imaged over time. Representative particle images of DLPs contained in liquid

were selected from the image series recorded over 10 seconds. The selected particles were then

contrast-inverted, and colorized for visualization purposes. Scale bar is 30 nm. Please see

associated Movies 1 – 3.

34

To produce time-resolved movies of the viral particles in liquid, we selected from the image

series representative DLPs having different lengths of RNA strands associated with their exterior

shells (Figure 3.1C, E). The selected particles were contrast-inverted and arbitrarily colorized

(blue, gray, and green) for tracking purposes (Figure 3.1E). Images sequences of the DLPs

within a 10-second time interval were compiled using the iMovie (Apple, Inc.) software package

(Figure 3.3, Movies 1 – 3). The compiled movies revealed real-time movements of the DLPs in

the surrounding liquid. We attributed the overall particle movements at the molecular level to

Brownian motion, biological activity, beam-induced movement, and beam damage – or some

combination of these factors.

To better visualize mobile units in the DLPs, we segmented the strongest features in the

particles by applying a density threshold filter of 3σ where σ is the standard deviation in the

contrast values of each particle image (3). By focusing on these strongest features, we could

attenuate noise in the images due to variations in the liquid flow and to some extent, radiolysis.

After applying the density-threshold, we were able to visualize the highest electron scattering

components within each particle image in the series (Figure 3.4A). This density included both

protein subunits and viral RNA (>12 MDa) that constitutes a significant portion of the DLP (~50

MDa). Representative contour maps within a 10 seconds timeframe were compiled using the

iMovie software package (Movies 4 – 6).

35

Figure 3.4. Visualizing and quantifying DLP mobility in liquid. (A) Representative DLPs (1 –

3) with associated viral mRNA strands (white arrows) were selected from an image series and a

subjected to a density threshold filter. Contour maps were generated based on a 3 σ -cutoff,

highlighting pixel displacements in the particles. Cumulative differences in the mobile units of the

viral particles were represented in heat maps using a scale of 0 – 4 to indicate the number of pixel

displacements within a 10-second interval. (B) Histograms that represent the total number of pixels

displacements within 10 seconds revealed non-uniform movements between the DLPs. Means

pixel displacement values are also summarized. The plot shows the correlation (R2=0.99) between

the mean pixel displacements and the lengths of external RNA associated with each particle. Please

see Movies 4 – 6.

36

The compiled movies of the contoured regions showed the strongest visual differences in the

individual virus particles and their associated mRNAs. This information provided a new time-

resolved view of biological assemblies in liquid at the nanoscale. We also noted the presence of

the viral mRNA transcripts (~10 Å in width) associated with the viral particles at the beginning

and the end of the image series (Movies 4 – 6). The conserved presence of these fragile mRNA

strands throughout the image series supports the idea that beam damage, or beam induced

movements do not impose a stronger influence on the observed particle mobility than Brownian

motion or biological activity. However, we continue to delineate the individual contribution of

these factors in our analysis.

To quantify the pixel movements in the particle images over time, the binary thresholded

images were combined into time series. The temporal derivative of each pixel returned values of

1 when a pixel changed from empty to full, and -1 when the pixel changed from full to empty.

Summing the absolute value of each pixel’s temporal derivative returned a value that represents

how frequently each pixel changed value. When applied to the entire time series, we could

quantify how much the entire particle moved or changed over the imaging duration (Figure

3.4B). This algorithm unlike a nearest neighbor, cluster detection, or cross-correlation

calculations has no assumptions about whether a particle moved into or out of a feature.

Summary data for the three representative particles demonstrated that Particle 1 (blue) exhibited

statistically more pixel displacements than the other particles examined over a 10 seconds

timeframe. Particle 2 (gray) showed the least overall mean pixel displacements, while Particle 3

(green) displayed an intermediate value of mean pixel displacements (Figure 3.4B). A heat map

summarizing the positions of the cumulative movements for each particle is given in Figure

3.4A.

37

Figure 3.5. 3D structures of DLPs reveal internal densities in liquid and ice specimens. (A)

A 3D reconstruction was calculated for actively transcribing particles in a liquid environment.

Cross-sections of the reconstruction are show at ~20-nm intervals to the midsection. The

diameter of the reconstruction are ~80-nm. (B) A 3D reconstruction was calculated for frozen-

hydrated particles using the same image processing steps and while imposing icosahedral

symmetry during the refinement procedure. The diameters of the resulting reconstructions are

~80-nm. Scale bar is 30 nm.

38

To build upon these measurements, we also quantified the lengths of the mRNA strands

surrounding each particle, starting from the edge of the DLPs and extending to a 150-nm region

in all directions. The total length of mRNA that surrounded Particle 1 (blue) was ~21 nm.

Particle 2 (gray) was surrounded by ~8 nm of mRNA, while ~13 nm of mRNA and surrounded

Particle 3 (green). We plotted the mean pixel displacements for each particle versus the relative

lengths of associated mRNAs. These values represented a combination of more than 10,000 pixel

movements in total. Upon examining a linear regression analysis of these measurements, we

found a strong correlation (R2=0.99) existed (Figure 3.4B). This information suggested that

particles having more mobile units were associated with greater quantities of viral mRNAs.

Although, the asynchronous nature of the DLPs does not permit us to fully determine the history

of each particle, the use of these newly developed tools and analysis can be used to probe more

deeply into viral transcription events in future experiments.

To validate the integrity of the DLPs used in our analysis, we calculated a 3D image

reconstruction on the initial image in the series. We used the PARTICLE software package to

select 30 individual DLPs from the image as previously described (10). For reference, when

imposing icosahedral (60-fold symmetry) during the reconstruction phase, 30 particles equates to

1800 asymmetric particles. The selected particles were then exported into the RELION software

package (16). Using RELION, we calculated a 3D structure of the DLPs while using icosahedral

symmetry operators over an angular search space of 7.5° throughout 25 cycles of refinement

(Figure 3.5A). Refinement parameters in RELION included a DLP reference model (17). low-

pass filtered to 50 Å, a pixel size of 5 Å, and a regularization parameter of T = 4. Consistent with

previous 3D structures of enzymatically activated particles (10), the resulting reconstruction of

the DLPs in liquid showed dynamic features in their core interiors at ~2.8 nm-resolution (Figure

39

3.3A).

As an additional control, we analyzed images of frozen-hydrated active DLPs. We

implemented the same computing procedures and calculated a reconstruction using 38 particles

from a single EM image while enforcing icosahedral symmetry. For reference, when enforcing

icosahedral (60-fold symmetry) during the reconstruction phase, 38 particles equates to 2280

asymmetric particles. We chose to use slightly more particles in the cryo-EM calculations as we

anticipated more views of the DLPs would be needed in the static DLPs to equate with the fluid

system. The resulting 3D structure was strikingly similar to the DLP reconstruction in liquid and

resolved to ~2.5 nm. (Figure 3.5B). A notable difference between reconstructions calculated for

DLPs in ice and liquid is that the liquid structure had a continuum of density throughout the

particles while the ice structures had more rigidly defined segmented features visible in cross-

sections through the density maps (Figure 3.5A, B). These differences may be due to a lack of

fluidity in the captured particles resulting from the specimen freezing process. Overall, the 3D

structural information in both the liquid and ice preparations complements the time-resolved

measurements by ensuring the proper integrity of the activated viral assemblies.

As we continue to make progress toward real-time molecular imaging of biological systems,

one item that can elevate our efforts is the use of direct electron detectors. These new detectors

can operate in movie-mode at lower magnifications to further minimize issues of electron dose

and frame rate. In addition, the use of thinner silicon nitride windows or graphene liquid cells

may enable us to gather more information of how single-stranded mRNA transcripts emerge

from viral particles. Similarly, reducing the thickness of the liquid cell in general could enable us

to view protein-protein interactions or other viral processes. On course with data presented here,

molecular imaging of biological assemblies at the nanoscale inspires a new era in microscopy,

40

with the potential to further decode life process and improve our knowledge of human health

worldwide.

Acknowledgements

The authors acknowledge the following funding sources that supported this work, 1R21AI113402-

01 NIAID/NIH (D.F.K., S.M.M) and 1R01AI116815-01 NIAID/NIH (D.F.K., S.M.M).

Author Contribution

D.F.K. conceived this project and supervised the completion of all experiments. A.C.V. and

D.F.K. prepared the manuscript. A.C.V., A.R., and M.J.D. performed the experiments. S.P.

measured virus movies for pixel movements. S.M. provided reagents and technical support.

41

References

1. B. L. Gilmore et al., Visualizing viral assemblies in a nanoscale biosphere. Lab Chip 13,

216-219 (2013).



3. J. Frank, Three-dimensional electron microscopy of macromolecular assemblies :

visualization of biological molecules in their native state. (Oxford University Press,

Oxford ; New York, ed. 2nd, 2006), pp. xiv, 410 p.



5. E. A. Ring, D. B. Peckys, M. J. Dukes, J. P. Baudoin, N. de Jonge, Silicon nitride

windows for electron microscopy of whole cells. J Microsc-Oxford 243, 273-283 (2011).

6. J. M. Yuk et al., High-resolution EM of colloidal nanocrystal growth using graphene

liquid cells. Science 336, 61-64 (2012).

7. E. A. Ring, N. de Jonge, Microfluidic system for transmission electron microscopy.

Microsc Microanal 16, 622-629 (2010).

8. A. Radisic, P. M. Vereecken, J. B. Hannon, P. C. Searson, F. M. Ross, Quantifying

electrochemical nucleation and growth of nanoscale clusters using real-time kinetic data.

Nano Lett 6, 238-242 (2006).

9. J. Park et al., Direct observation of nanoparticle superlattice formation by using liquid

cell transmission electron microscopy. ACS nano 6, 2078-2085 (2012).

10. M. J. Dukes et al., Improved microchip design and application for in situ transmission

electron microscopy of macromolecules. Microsc Microanal 20, 338-345 (2014).

11. E. S. Pohlmann et al., Real-time visualization of nanoparticles interacting with

glioblastoma stem cells. Nano Lett 15, 2329-2335 (2015).

12. J. A. Lawton, M. K. Estes, B. V. Prasad, Three-dimensional visualization of mRNA

release from actively transcribing rotavirus particles. Nat Struct Biol 4, 118-121 (1997).

13. J. A. Lawton, M. K. Estes, B. V. Prasad, Comparative structural analysis of

transcriptionally competent and incompetent rotavirus-antibody complexes. Proc Natl

Acad Sci U S A 96, 5428-5433 (1999).

14. K. Degen, M. Dukes, J. R. Tanner, D. F. Kelly, The development of affinity capture

devices-a nanoscale purification platform for biological in situ transmission electron

microscopy. Rsc Adv 2, 2408-2412 (2012).

15. B. L. Gilmore et al., Visualizing viral assemblies in a nanoscale biosphere. Lab on a chip

13, 216-219 (2013).

16. S. H. Scheres, A Bayesian view on cryo-EM structure determination. J Mol Biol 415,

406-418 (2012).

17. X. Zhang et al., Near-atomic resolution using electron cryomicroscopy and single-

particle reconstruction. Proc Natl Acad Sci U S A 105, 1867-1872 (2008).

42

Chapter 4

Preparation of Disease-related Protein Assemblies for Single Particle Electron

Microscopy

A.Cameron Varano1,2, Naoe Harafuji3, William Dearnaley1,

Lisa Guay-Woodford3, and Deborah F. Kelly 1*

1. Virginia Tech Carilion Research Institute, Virginia Tech



3. Center for Translational Science, Children’s National Health System

*Correspondance to: Deborah F. Kelly 2 Riverside Circle, Roanoke, VA 24016; 540-526-2031;

[email protected];

43

Abstract

Electron microscopy (EM) is a rapidly growing area of structural biology that permits us to decode

biological assemblies at the nanoscale. To examine biological materials for single particle EM

analysis, purified assemblies must be obtained using biochemical separation techniques. Here we

describe effective methodologies for isolating histidine (his)-tagged protein assemblies from the

nucleus of disease-relevant cell lines. We further demonstrate how Isolated assemblies are

visualized using single particle EM techniques and provide representative results for each step in

the process.

Introduction

Electron microscopy (EM) allows us to examine and characterize biological entities ranging

from relatively large, uniform virus structures to smaller, non-symmetrical proteins (1,2).

Determining the intricate details of disease-related proteins using high-resolution EM imaging

can reveal new targets for rational drug design. The presented protocols here are adapted from

our recent EM structural studies on native BRCA1 assemblies formed in human cancer cells

(3,4). A new aspect of the present work is the application of these methods for disease-related

protein assemblies beyond human cancer. Here we demonstrate the utility of our established

protein separation techniques for instances in which the target protein is expressed in low

abundance. Specifically, we provide detailed methods to produce and characterize protein

assemblies incorporating the C-terminal domain (CTD) of the Fibrocystin/polyductin protein

(FPC) (5,6).

There are many known mutations in the FPC protein that are implicated in autosomal

recessive polycystic kidney disease. At the cellular level, the longest mRNA product of the FPC

44

gene encodes a membrane-bound protein that undergoes Notch-like proteolytic cleavage to

generate a functional carboxy-terminal domain (CTD-FPC). The CTD-FPC translocates to the

cell’s nucleus to carry out a variety of gene-related functions. As the FPC protein is naturally

expressed in low abundance in the kidney epithelia, this presents a technical barrier to

understand the molecular basis of disease-related mechanisms (7).

To address this issue and facilitate structural analysis on FPC assemblies, we employed a protein

enrichment strategy by over-expressing a his-tagged (6x-His) version of the CTD-FPC in mouse

kidney cells (mIMCD-3 line; ATCC; (8)). Nuclear assemblies that incorporated the His-tagged

recombinant protein were isolated and imaged using previously established single particle EM

techniques (3). The overall steps of the process that we describe include: 1) a nuclear extraction

step, 2) a nickel chromatography isolation procedure, 3) EM specimen preparation and data

collection, 4) processing the image data. Overall, this work demonstrates a means to assess disease-

related protein assemblies for structural analysis while providing new avenues to explore protein

interfaces for drug discovery purposes.

Materials

All reagent solutions described in our procedures work optimally when prepared with ultrapure

water. In general, working reagents should be prepared and stored on ice or at 4°C. When working

to obtain structural information of protein complexes it is important to avoid excessive procedures

that can create “bubbles” in solution or at the air-water interface, as these disturbances can affect

protein integrity.

45

Cytoplasmic and Nuclear Extraction

Cellular nuclear material is obtained by using a commercially available kit, NE-PER Nuclear and

Cytoplasmic Extraction Reagents (Thermo Scientific). For this procedure, we use a recommended

packed cell volume of 100 μL. For different cell volumes, please consult manufactures

recommendations. Reagents are prepared according to the following steps, and we recommend

following the procedures described by the manufacturer to obtain the nuclear fraction.

1. 1.5 mL tubes

2. 600 μL tubes

3. Using forceps, place 1 tablet of cOmplete mini protease inhibitor 100x (Roche Diagnostics) on

a piece of weigh paper, then crush the tablet. Carefully transfer crushed tablet to 600 μL tube.

Then add 100 μL of water to the crushed tablet. Pipette up and down gently to dissolve into

solution.

4. Prepare the cytoplasmic extraction reagent (CER) 1 solution in a pre-chilled 1.5 mL tube label.

First, add 1000 μL of CER 1 reagent, followed by 10 μL of the protease inhibitor solution. Finally,

add 10 μL of Halt phosphatase inhibitor 100x (Thermo Scientific).

5. Prepare the nuclear extraction reagent (NER) solution in the pre-chilled 1.5 mL tube. Pipette

500 μL of NER solution into the tube followed by 5 μL of protease inhibitor solution. (Note 1)

Nickel Chromatography Purification

Nickel chromatography utilizes the affinity of a recombinant protein with a polyhistidine (His) tag

to associate with nickel cations. The protein of interest is then eluted off using buffers containing

imidazole (Note 2). To perform these steps, prepare the following buffers in 50 mL conical tubes

using ultrapure water and chilled to 4°C:

46

20 mM HEPES Buffer (pH, 7.4): 0.2383 g HEPES (20mM), fill to 50 mL with ultrapure water.

20 mM HEPES Buffer (pH, 7.4) with salts: 0.2383 g HEPES (20mM), 0.4091g NaCl2 (140 mM),

100 μL of 1M MgCl2 (2mM), 11.098mg CaCl2 (2mM), fill to 50 mL with ultrapure water.

1M Imidazole: 0.34g imidazole, fill to 5mL with ultrapure water.

Wash Buffer (5mM imidazole): 10 mL HEPES 20mM (pH, 7.4) with salt, 50 μL 1M imidazole.

Elution Buffer (200mM imidazole): 1 mL of 1M imidazole, 4 mL HEPES 20mM (pH, 7.4) with

salt.

Carbon-coated TEM Grids

Carbon-coated TEM grids can be prepared by conventional methods in-house or purchased from

commercial suppliers, such as Ted Pella, Inc. The following supplies are needed to prepare

negatively stained EM specimens.

Preparation of uranyl formate heavy metal stain (1% w/v)

1. Boil 3 mL of ultrapure water in a 10 mL beaker on hot plate

2. Using tongs, transfer to stir plate and add 22.5 mg of Uranyl Formate and a stir bar.

3. Stir for 5 minutes

4. Add 4.2 μL of 5N NaOH

5. Stir another 5 minutes

6. Draw-up in a 5 mL syringe

7. Filter through 0.2 μm filter (Fisher) into a 15 mL conical tube covered in aluminum foil.

47

Methods

Cytoplasmic and Nuclear Extraction Procedures

The following information is based on the guidelines provided by Thermo Scientific for optimal

use of their NE-PER Nuclear and Cytoplasmic Extraction Reagents.

1. Pellet cells using centrifugation at 100 x g for 5 minutes at 4°C.

2. Add ice-cold CER 1 to the resulting cell pellet.

3. Vortex the sample for 15 seconds followed by a 10-minute incubation on ice.

4. Add 55 μL ice-cold CER 2 buffer to the sample solution.

5. Vortex the tube for 5 seconds then incubate on ice for 1 minute.

6. Vortex the tube for 5 seconds on the highest setting. Centrifuge the tube for 5 minutes at

16,000 x g in a microcentrifuge at 4°C. (Note 3)

7. Carefully decant the supernatant containing the cytoplasmic components to a waste beaker.

(Note 4)

8. Resuspend the insoluble pellet containing nuclear components in ice-cold NER solution.

9. Vortex for 15 seconds then place on ice. Repeat the vortexing-incubation cycle in 10

minute intervals for a total of 40 minutes.

10. Centrifuge at 16,000 x g for 10 minutes at 4°C.

11. Carefully transfer the supernatant containing the nuclear extract (NE) to a 1.5 mL tube and

place on ice. (Note 5)

Nickel Chromatography Purification

Immobilized-metal affinity chromatography (IMAC) exploits the principles of specific protein

interactions with chelated metal groups, held in place by immobilized surfaces or beads. One

48

popular association is that formed by the interaction of polyhistidine residues (His tag) with

functionalized Nickel- Nitrilotriacetic acid (Ni-NTA) resin. Below we describe procedures for

isolating His-tagged CTD-FPC protein assemblies from the nuclear extract prepared in materials

section.

1. Dilute the NE in 20 mM HEPES buffer without salts and store on ice. (Note 6)

2. Pipette 400 μL of Ni-NTA agarose (QIAGEN) slurry (Note 7) into a 1.5 mL centrifuge

tube. Centrifuge the mixture for 2 minutes at 700 g.

3. Carefully remove the supernatant and discard.

4. Add 1 mL of the wash buffer containing 20mM HEPES buffer, and 5mM imidazole to the

resin and invert to mix.

5. Centrifuge for 2 minutes at 700 x g.

6. Remove the supernatant and discard.

7. Resuspend the Ni-NTA resin into 1 mL of wash buffer and centrifuge at 700 x g for 2

minutes. Remove supernatant.

8. Add the diluted NE to Ni-NTA resin.

9. Gently mix the material on a clinical rotator at 4ºC for 60 minutes. (Note 8)

10. Wash the Ni-NTA resin using the wash buffer to remove background proteins. (Note 9).

Collect the wash buffer that flows though the resin.

11. Elute the His-tagged proteins of interest using the elution buffer. The high concentration

of imidazole in this buffer will compete off the His-tagged FPC complexes. Typically, the

eluted material is collected in multiple fractions. Each fraction is roughly equal to 1 bed

volume, which is ~200 μL in our experiments.

a. Add 1 mL of the elution buffer, carefully pipetting it along the tube wall.

49

b. Collect increments of 200 μL in 1.5 mL tube. A total 5 fractions are usually

collected.

12. All sample collected should be stored on ice and analyzed for total protein concentration

using a standard Bradford assay.

The His-tagged CTD-FPC protein often elutes in fraction 2 and 3 (See Figure 4.1).

50

Figure 4.1. Representative immunoblot. Western blot analysis probed with anti-His antibodies (Abcam) shows samples obtained from the Ni-NTA purification procedure. Lanes include the nuclear extract, column flow-through, wash and eluted fractions (1-5). His-tagged CTD-FPC elutes in fractions 2 and 3.

51

EM Specimen Preparation

Glow Discharge Procedure for Continuous Carbon Grids

1. Place continuous carbon coated copper grids on glass slide using either parafilm wrap or

double-sided tape to keep the grids in place.

2. Open dome of Pelco easiGlow (Ted Pella, Inc.)

3. Insert slide into machine

4. Wipe rubber rim of machine to remove any dust.

5. Replace dome

6. Use stylist to tap auto run (Note 10)

7. After the cycle is completed, open the dome and remove glow-discharged grids

52

Figure 4.2. Applying sample to EM grid. Biological material was carefully added to glow-

discharged continuous carbon grids. (A) Aliquots (3 μL each) of sample was applied to glow

discharged EM grids with the grid secured in forceps. (B) When washing each sample, turn the

grid to the side to blot off excess solution onto Whatman #1 filter paper.

53

Preparation of Negatively Stained EM Specimens

1. Using parafilm, place muliple 200 μL drops of ultrapure water and 2- 200 μL drops of 1%

uranyl formate in a row. These drops will be used for specimen washing and staining steps.

2. Carefully, pick-up a fresh glow-discharged grid with forceps. The tips of the forceps, should

only touch the edge of the grid.

3. Place 3 μL of sample on the grid and incubate at room temperature for 1 minute (Figure

4.2a).

4. Blot off the excess sample onto Whatman #1 filter paper (Figure 4.2b).

5. Wash the face of the grid 3 times with ultrapure water. Gently touch the face of the grid on

each droplet then again blot off the ultrapure water using Whatman #1 filter paper. Use a

new drop for each wash step. Do not fully immerse the grid into the water droplet.

6. Wash the face of the grid 1 time with 1% uranyl formate, utilizing the same method as the

water wash.

7. Stain with 1% uranyl formate for 30 seconds (Figure 4.3a).

8. Blot away excess solution using a vacuum hose. Be cautious not to touch the vacuum hose

directly to the grid as it may damage the sample (Figure 4.3b). Store the samples in a labeled

petri dish until ready to image.

54

Figure 4.3. Staining EM grid. (A) To negatively stain the samples with 1% uranyl formate,

touch the face of the grid onto uranyl formate droplet without submersing the grid into the stain.

(B) Dry the stained specimens with a vacuum hose from the backside of the grid. The vacuum

hose should not come in contact with the grid as it may damage the sample.

55

Figure 4.4. Representative structural data and output. (A) A representative micrograph of

purified CTD-FPC assemblies taken at 68,000x magnification. Scale bar is 20 nm. (B) Class

averages obtained from multi-reference alignment procedures implement in the SPIDER

software package. (C) 3D reconstruction of the CTD-FPC complex calculated using RELION

software package with each view rotated 90°.

56

TEM Image Collection

1. Negatively stained CTD-FPC complexes can be examined using a FEI Spirit Bio-Twin

TEM equipped with a LaB6 filament and operating at 120 kV.

2. Grids are loaded into single tilt specimen holder at room temperature (FEI Company.).

3. Images are recorded using an Eagle 2k HS CCD camera employing low-dose conditions

(~1 electrons per Å2) at 68,000× magnification. The final sampling at the specimen level

is 4.4 Å per pixel.

4. All images are acquired under the same conditions using a defocus value of -1.5 μm (Figure

4.4a).

Data Analysis and Representative CTD-FPC Results

1. Prior to particle selection, the original images are normalized and CTF-corrected using the

standard routines in SPIDER software package (2).

2. Individual complexes from the images are manually selected using the WEB interface of

the SPIDER software package, and employing a box size of 300 Å.

3. Multi-reference alignment routines are implemented outputting 2D classes (Figure 4.4b)

as previously described (3).

4. Representative image stack containing selected particles are imported into the RELION

software package.

5. A spherical structure can be used as a reference map to reconstruct CTD-FPC complexes

through 25 refinement iterations using an angular sampling interval of 7.5°. Other

parameters input into RELION include a magnification of 68,000x, a pixel size of 4.4 Å,

57

and a regularization parameter of T=4. A representative 3D reconstruction of the His-

tagged CTD-FPC complex is shown in Figure 4.4C.

Notes

1. Halt Phosphatase inhibitor is not used when doing a nickel column purification, as it may

interfere with the nickel resin's binding capacity.

2. Imidazole is a salt, which competes with the polyhistidine (His) tag for the Nickel-NTA resin.

All buffers and samples must be kept at 4°C or on ice to maintain protein integrity.

3. Place the tube in a specific orientation, for example, so that the lid hinge is located outside of

rotator. This helps to locate the pellet in the tube since the pellet created is not always easily

distinguishable.

4. Our protein of interest is a nuclear protein therefore we do not need keep the cytoplasmic

fraction.

5. Extracts can be stored at -80°C for 30 days or used immediately.

6. This dilution step lowers the salt concentration of the sample, bringing it closer to a

physiological level.

7. The slurry is a 50% resin and buffer mixture. Pipetting a volume of 400 μL results in a 200 μL

bed volume of Ni-NTA resin. Before we begin the purification the Ni-NTA slurry is separated

from the storage buffer and equilibrated with the wash buffer.

8. The amount of volume remaining is a small quantity. The meniscus of the buffer solution

should be just above the resin.

9. This volume is dependent upon the bed volume. The wash volume is ~15 times the bed

volume.

58

10. Auto run should be set at 15 mA, glow 1 minutes and hold 10 seconds.

Acknowledgements

This work was supported by NIH/NCI grant R01CA193578 to D.F.K.

Author Contribution

D.F.K. and L.G-W. conceived this project and supervised the completion of all experiments.

A.C.V. prepared the manuscript. A.C.V., N.H. and W.D. performed the experiments

59

References

1. Taylor KA, Glaeser RM (2008) Retrospective on the early development of cryoelectron

microscopy of macromolecules and a prospective on opportunities for the future. J Struct

Biol 3:214-23

2. Frank J, Radermacher M, Penczek P, Zhu J, Li Y, Ladjadj M, et al. (1996) SPIDER and

WEB: processing and visualization of images in 3D electron microscopy and related fields. J

Struct Biol 1:190-9

3. Gilmore BL, Winton CE, Demmert AC, Tanner JR, Bowman S, Karageorge V, et al. (2015)

A Molecular Toolkit to Visualize Native Protein Assemblies in the Context of Human

Disease. Sci Rep:14440

4. Winton CE (2016) A microchip platform for structural oncology applications. NPJ Breast

Cancer:16016

5. Kim I, Li C, Liang D, Chen XZ, Coffy RJ, Ma J, et al. (2008) Polycystin-2 expression is

regulated by a PC2-binding domain in the intracellular portion of fibrocystin. J Biol Chem

46:31559-66

6. Boddu R, Yang C, O'Connor AK, Hendrickson RC, Boone B, Cui X, et al. (2014) Intragenic

motifs regulate the transcriptional complexity of Pkhd1/PKHD1. J Mol Med (Berl) 10:1045

7. Guay-Woodford LM (2014) Autosomal recessive polycystic kidney disease: the prototype of

the hepato-renal fibrocystic diseases. J Pediatr Genet 2:89-101

8. Rauchman MI, Nigam SK, Delpire E, Gullans SR (1993) An osmotically tolerant inner

medullary collecting duct cell line from an SV40 transgenic mouse. Am J Physiol 3 Pt

2:F416-24

60

Chapter 5

Cryo-EM-on-a-chip: custom-designed substrates for the 3D analysis of

macromolecules

Nick A. Alden1‡, A. Cameron Varano1,2 ‡, William J. Dearnaley1,4,5,6, Maria J.

Solares1,2,5,6, Yanping Liang1, John Damiano3, Jennifer McConnell3, Madeline

Dukes3, and Deborah F. Kelly1, 4, 5, 6

1. Virginia Tech Carilion Research Institute, Virginia Tech, Roanoke, VA 24016, United States

2. Translational Biology, Medicine, and Health Graduate Program, Virginia Tech, Blacksburg,

VA 24061, United States

3. Application Science, Protochips Inc., Morrisville, NC 27560, United States

4. Department of Biomedical Engineering, Pennsylvania State University, University Park, PA,

16802, United States

5. Huck Institutes of the Life Sciences, Pennsylvania State University, University Park, PA,

16802, United States

6. Center for Structural Oncology, Pennsylvania State University, University Park, PA, 16802,

United States

61

Abstract

A multi-disciplinary scientific approach is needed in the fight against human cancer. Blending

tools from seemingly unrelated areas, such as materials science and molecular biology,

researchers may better address the challenges that confound our current understanding of disease

processes. Here we developed custom-designed substrates composed of Silicon nitride (SiN) to

study the three-dimensional (3D) attributes of tumor suppressor proteins. New on-chip

preparation strategies were used to evaluate the molecular properties of native protein complexes

isolated from human cancer cells. Combined techniques of cryo-Electron Microscopy (EM) and

molecular modeling revealed a new modified form of the p53 tumor suppressor present in

aggressive brain tumors, glioblastoma multiforme (GBM). Taken together, our findings provide

a new design template to customize EM substrates for biomedical applications in the evaluation

of disease-related macromolecules.

Introduction

Cryo-Electron Microscopy (EM) has revolutionized the manner in which we analyze biological

processes. Early pioneers of the technique were awarded the 2017 Nobel Prize in Chemistry,

bringing to light its important contribution to life sciences. Although the spatial resolution

achievable with cryo-EM has skyrocketed recently, the basic principles that govern this

sophisticated imaging technique remain intact. Simply, a focused stream of electrons is produced

by a field emission gun source. The collimated electrons then penetrate and scatter off of frozen-

hydrated molecules that are spread out upon a support film, only a few nanometers in thickness

(1). As the scattered electrons are focused by electromagnetic lenses, the amplitudes and phases

of the exit waves are collected upon a detector and output as projection images. Each snapshot

62

contains exquisite details of the sample’s molecular structures. Particle images with different

angular information are used to mathematically reconstruct a three-dimensional (3D)

representation of the original object (2-4).

Over the last decade, enormous effort in the field has been devoted to automating data

collection routines and downstream computing procedures (3). The culmination of these efforts

has spurred an exciting new era for the structural determination of biological machines. One

critical opportunity for improving the current work-flow is the design of new substrates and

materials to prepare fragile specimens. Typical preparation procedures involve weakly attaching

macromolecules to carbon-based support films. These films are often comprised of micron-sized

“holes” in which biological particles become frozen in time and space. More recently, alternative

materials are gaining popularity in the cryo-EM field, including gold foils containing micron-

sized holes or graphene-based substrates (5), both of which have proven useful for sample

preparation procedures. One recurring bottle-neck that still confounds the field, regardless of the

support material, is the grid screening process. This time-consuming step limits automation

procedures due to inconsistencies in freezing conditions and ice thicknesses between specimens.

To reduce specimen-related barriers in the field, we propose the use of radically different

cryo-EM substrates – precisely engineered to help control for ice thickness and particle

distribution. These new custom window designs are integrated into Silicon nitride (SiN)

microchips and hold great potential to minimize variability among frozen-hydrated specimens.

The engineered micro-well or micro-post topographies spanning the SiN viewing window

promote enhanced particle capture and retention during specimen preparation steps for cryo-EM

studies. In addition to being more physically reinforced than traditional grids, custom-designed

63

microchips offer the option to work with native proteins from patient samples for structural

studies. In the context of cryo-EM, we refer to these microchips as Cryo-SiN.


Simian rotavirus DLPs

Simian rotavirus (strain SA11-4F) DLPs were prepared as described in previous work (6).

Briefly, transcription activation reactions were performed at 37°C using the following

components: 1 μg DLPs prepared in 100 mM Tris-HCl pH 7.5, 6 mM MgAc, 4 mM DTT, 2 mM

each of ATP, GTP, CTP, UTP, and 1μl RNasin (Promega Corp., Madison, WI). After an

incubation period of 30 minutes, aliquots of the transcription reactions (2 - 3 μl each) were used

for cryo-EM experiments.

Isolation of protein assemblies

The HCC1937 cells used in this study was purchased from ATCC and independently

characterized by ATCC as mutated BRCA15382insC-BARD1. The HCC1937 cells were cultured in

RPMI 1640 (Mediatech) supplemented with 10% fetal bovine serum (ATCC) and 0.5×

penicillin-streptomycin (Thermo Fisher Scientific). The U87MG cells were used this study

cultured in DMEM (Life Technologies Corporation) supplemented with 10% fetal bovine serum

(FBS; Atlas Biologicals, Inc.), 100 μg/ml of streptomycin, and 100 IU/ml of penicillin. Cell

lines were incubated at 37°C in 5% CO2. Immobilized metal affinity chromatography was used

to enrich for either BRCA15382insC-BARD1 or p53 complexes. Approximately 1 million cells

were collected using trypsin-EDTA (Thermo Fisher Scientific) and pelleted by centrifugation

(500xg; 5 minutes). The nuclear fraction was extracted using the NE-PER kit (Thermo Fisher

64

Scientific). The soluble nuclear material was incubated with 200 μl of Ni-NTA agarose beads

(Qiagen) for 1 hour at 4°C with gentle rotations. The beads with bound material were transferred

to a 3ml column and washed with 3ml of 20 mM HEPES buffer (pH 7.2; 140 mM NaCl, 2 mM

CaCl2, 2 mM MgCl2, and 10 mM imidazole). The p53 complexes were eluted in the same

HEPES buffer supplemented with 60 mM imidazole, while the BRCA15382insC-BARD1

complexes were eluted in the HEPES buffer supplemented with 150 mM imidazole. Protein

concentrations were determined using Bradford assays (Thermo Fisher Scientific).

SDS-PAGE and immunoblotting

Following enrichment, protein fractions were analyzed by SDS-PAGE stained and immunoblot

analysis. Proteins were separated on 3 to 8% Tris-Acetate NuPAGE mini gels (Thermo Fisher

Scientific). The gels were washed 3 times with ultra-pure water and then stained by SimplyBlue

SafeStain solutions (Invitrogen). Gels were washed with deionized water overnight on an orbital

shaker and then in 20% NaCl aqueous solution overnight to achieve maximum sensitivity.

Antibodies against p53 (DO-1; Santa Cruz, sc-126) and ubiquitin linkages (D7A11; Cell

Signaling, #5621) were used for western blot detection.

EM specimen preparation and data collection

Aliquots of purified proteins or virus assemblies (~300 µl each of 0.02 mg/ml) were added to 96-

well microtiter plates. Clean SiN microchips were either glow-discharged or functionalized with

20% Ni-NTA lipid monolayers as previously described (7). Following an appropriate incubation

period (10 seconds to 1 minute), the excess aqueous solution was blotted away and the microchip

was loaded into a FEI Mark III vitrobot (ThermoFisher Scientific) for plunge freezing. Blotting

65

times were typically 4 - 6 seconds or ~2 seconds/sample volume. Alternatively, ~2-µl aliquots of

biological material may be added direct to the microchip surface prior to plunge-freezing. In

parallel, holey carbon EM grids (Protochips, Inc.) were prepared by adding 2-µl aliquots of each

sample directly to the grid and incubating for one minute. Grids were loaded into the same

freezing device prepared under the same blotting conditions. Frozen-hydrated specimens were

examined using a FEI Spirit BioTwin TEM (Thermo Fisher Scientific) equipped with a LaB6

filament and operating at 120 kV under low-dose conditions (<5 electrons/Å2). Images were

recorded using an Eagle 2k HS CCD camera having a pixel size of 30-µm (Thermo Fisher

Scientific) at various magnifications. For the p53 protein assemblies, a nominal magnification of

~68,000x was used for a final sampling of 4.4 Å / pixel.

Image processing and movie production

For the p53 reconstruction, particles were selected from EM images using automated routines

implemented in the RELION software package. A filtered (30 Å) structure of the p53 core bound

to DNA (pdb code, 2AC0) was used as an initial model. Class averages were calculated for the

data and the density map contained 12,144 particles. The initial model assisted in the first round

of 3D refinement to assign orientation values to each particle. Up to 25 subsequent refinement

iterations were executed in RELION implementing a regularization parameter of T=4, a pixel

size of 4.4 Å, and a mask value of 80 Å. Following the refinement procedures, particle data was

divided into two halves and the resolution for each half converged to a common numerical value.

The 0.5-FSC criteria provided an estimate of the resolution in RELION, which was verified

using the RMEASURE program to be approximately 10 Å. During the refinement procedures,

we imposed the C2-symmetry operator, bringing the total particle equivalency to 24,288. The

66

final map was masked at ~80 Å in diameter and the Chimera software package was used to

visualize all density maps. Threshold values are included as part of the accompanying

information for the EM map deposition.

To highlight components of the p53 structural assemblies and the molecular models, we

used the Chimera software package to produce movies of the data. Each movie contained the p53

EM reconstruction and models that were imported and manually aligned. The density map with

the atomic models in place were rotated and cross-sectioned during the movie production

procedures. To generate cross-sectional views, slices through the output were generated at 110

frames, then reversed as frame slices were replaced. The structure was rotated about the x– or y–

axis by ~1 degree per frame, while up to 90 frames were used in total. Movies were output in

.mov format.


The development of custom-designed EM substrates

An exciting new direction for the cryo-EM field involves the structure determination of

medically-relevant macromolecules for therapeutic design purposes. One example of this

approach includes our recent studies on BRCA1-protein assemblies isolated from breast cancer

cells (Figure 5.1) (8, 9). Other successful examples include the high-resolution work of Scheres

and colleagues, who determined 3D structures of tau filaments isolated from brains of dementia

patients (10). Hence, developing new tools to evaluate native protein assemblies is important to

advancing the biomedical community. Cryo-SiN is highly appealing for this application due to

features of durability, consistent flatness, and electron-transparent imaging windows (8).

67

Figure 5.1. Workflow from cancer cells to protein structure. Our workflow to capture and

image native proteins isolated from human cancer cells for cryo-EM structural analysis. Data

collected from cryo-EM images is used to reconstruct EM density maps, such as the BRCA1-

BARD1 model shown here (9).

68

One strategy to prepare cryo-EM samples on custom microchips utilizes a 96-well

microtiter plate (Figure 5.2). In this procedure, the wells of the plate may contain

macromolecules of interest in a variety of buffer conditions or activation states. Glow-discharged

or functionalized microchips are dipped into the sample and incubated in the aqueous solution

(~200 - 300 l) for up to one minute (Figure 5.2A). Upon removing the microchip from the

sample, the excessive aqueous solution that remains on the surface of the microchip is blotted

away using Whatman number 1 filter paper. The microchip is then loaded onto a freezing device

and plunged into a slurry of liquid ethane using standard protocols (8, 9). Optimal specimen

preservation conditions were achieved using an FEI Mark III vitrobot (Figure 5.2A), and

included a blotting step of ~2 sec/l of sample volume and a dwell time of ~1 second prior to

plunge-freezing.

The custom-designed microchips illustrated here are 2 mm x 2 mm in x- and y-

dimension with regions of the SiN imaging windows (500 m x 500 m) having a pitch of 10

m between windows (Figure 5.2B). Custom window designs and arrays can vary in spacing in

x- and y- as well as pitch. The imaging array of the microchips were further incised down to

produce 50-nm thick micro-wells or areas spanning the 50-nm thick film with 150 nm micro-

posts. The presence of the micro-wells or micro-posts across the membrane may serve to

improve sample detention and control for uniform ice distribution across the imaging array

(Figure 5.2C). Cryo-SiN microchips can also be tailored to accommodate specimen-specific

properties, such as particle size and sample thicknesses up to ~5-m. Eukaryotic cells, such as

neurons or brain tumor stem cells, have been directly cultured upon SiN microchips which can

also be used to examine features in whole cells (11, 12).

69

Figure 5.2. The preparation of frozen-hydrated macromolecules using the “Cryo-EM-on-a-

chip” technique. (A) Microchips are dipped into the sample containing the macromolecule of

interest before the excess aqueous buffer solution is wicked away using Whatman paper. Chips

are plunge-frozen into liquid ethane and used for downstream imaging and analysis. (B)

Schematic diagram of a Cryo-SiN microchip with integrated micro-wells. The 2 mm x 2 mm

microchip has a central imaging region of 500 µm x 500 µm and can be engineered with a

variety of micro-wells designs. (C) Side view schematic of a microchip. The height of the

microchip is ~300 µm, while each individual well has a depth of 150 µm and a pitch of 10 µm.

The micro-wells may aid in the encapsulation of macromolecules in vitreous ice.

70

Case study #1 – Active virus assemblies

As the frozen-hydrated specimens are prepared, they can be easily transferred under liquid

nitrogen from storage vessels to an EM specimen holder (Figure 5.3A). The microchip transfer

procedure entails its careful removal from a cryogenic storage buttons (Ted Pella, Inc.) and

centering the specimen into the tip of the holder, similar to the procedures used for conventional

EM grids. The chip is secured in the holder using a bronze grid-clip with minor modifications.

The specimen holder with the secured sample is transferred into the EM, while maintaining a

temperature of approximately -180ºC.

To better understand performance differences between Cryo-SiN and commercially

available holey carbon grids (Protochips, Inc.), we used transcriptionally-active rotavirus double-

layered particles (DLPs, ~2.5 MDa) as a model system. Simian DLPs (SA11 strain) were

purified as previously described and contained in aqueous buffer solution (100 mM Tris-HCl pH

7.5, 6 mM MgAc, 4 mM DTT, 2 mM each of ATP, GTP, CTP, UTP, and 1μl RNasin (Promega

Corp., Madison, WI)). Samples (2 - 3 l aliquots) from the same biochemical preparation were

frozen on the two respective substrates using the same conditions and protocols. Images were

recorded using an Eagle 2k HS CCD camera integrated into a FEI Spirit Bio-Twin TEM

equipped with a LaB6 filament and operating at 120 kV (Figure 5.3B). Images were recorded at

a magnification of 30,000x at a defocus value of approximately –1-m using an electron dose of

~5 electrons / Å2. We observed enhanced visual contrast in the DLPs prepared on Cryo-SiN in

comparison to the holey carbon substrate, as previous noted (13).

71

Figure 5.3. Cryo-SiN specimen transfer and image comparisons. (A) Using the aid of a clip

tool, Cryo-SiN specimen are transferred from cryogenic storage buttons to the tip of a EM

specimen holder, then secured in place using a modified clip ring. (B) Comparisons between

Cryo-SiN chips and holey carbon films demonstrate the high-contrast of Cryo-SiN chips for

macromolecules like transcriptionally-active rotavirus particles. (C) Enhanced image contrast is

maintained with Cryo-SiN chips for smaller molecules like BRCA1-associated protein

assemblies in comparison with holey carbon films used to image the same assemblies.

72

Case study #2 – Native protein assemblies isolated from breast cancer cells

Building upon the finding of enhanced visual contrast, we investigated whether the same effect

held true for samples isolated from human cancer cells. We prepared cryo-EM samples of the

breast cancer susceptibility protein (BRCA1)-associated protein assemblies isolated from breast

cancer cells (HCC1937 line). A full description of the biochemical purification scheme used to

obtain the BRCA1 complexes was recently published (9). This protein specimen is an appealing

model as its molecular weight and particle diameter (~300 kDa, 100 Å) is much smaller than that

of rotavirus assemblies (~2.5 MDa, 800 Å). Purified BRCA1 assemblies were prepared in

aqueous buffer solution containing 50 mM HEPES (pH 7.5), 150 mM NaCl, 10 mM MgCl2, 10

mM CaCl2. Aliquots (~2 l each) of the purified complexes were added to either Cryo-SiN

microchips or conventional holey carbon grids, each coated with 20% Nickel-Nitrilotriacetic acid

(Ni-NTA) lipid monolayers (please see MATERIALS AND METHODS). Microchips contained

custom designed micro-wells having a 500-m x 500-m window array and a 10-m pitch of

(Figure 5.2B, C). The same cryogenic preservation procedures were used for each specimen and

samples were examined under low-dose conditions (~5 electrons / Å2).

Specimens were examined using the same FEI Spirit Bio-Twin EM equipped with a

LaB6 filament and operating at 120 kV. Images were recorded using a magnification of 50,000x

at a defocus value of approximately –1 m. Images of Cryo-SiN specimens consistently revealed

particles with higher visible contrast compared with images of holey carbon samples (Figure

5.3C). Contrast values were inverted in the images of BRCA1 assemblies displayed in Figure

5.3C for ease in particle identification. Protein assemblies identified on Cryo-SiN substrates

contained stronger edge boundaries (dark halos) that facilitate particle detection in automated

image processing routines, such as those implemented in the RELION software package (3, 14).

73

This halo was not present in particles images prepared on holey carbon films. Similarly, an

enhanced white halo effect surrounded the viral assemblies displayed in Figure 5.2B, the Cryo-

SiN samples.

This halo effect was not present in the samples prepared on holey carbon films. Also, specimens

prepared on holey carbon substrates drifted more during imaging routines than Cryo-SiN

specimens examined at the same magnification and within the same session.

Structural analysis of a novel p53 assembly produced in glioblastoma multiforme

To further test the robustness of the Cryo-SiN microchips, we sought to acquire structural

information of a novel protein assembly isolated from a different cancer source. The tumor

suppressor protein, p53, participates in DNA damage response and is often called the

“guardian of the genome” (15). During its life cycle, modifications to p53 can stimulate its repair

response in the nucleus through the process of ubiquitination. As p53 is mutated in

approximately half of all human tumors (15, 16), it is important to investigate its structure-

function relationship in the context of human disease. To facilitate this opportunity, we used

Cryo-SiN microchips having integrated micro-wells to prepare p53 assemblies isolated from

glioblastoma multiforme (GBM) cells (U87MG line). A full description of protein isolation and

imaging procedures is provided in the MATERIALS AND METHODS section.

Based on the characteristic morphology and physical dimensions of the particles noted in

our EM images, we identified a dimeric form of ubiquitinated-p53 (~116 kDa, 60-Å diameter) in

our biochemical preparation (Figure 5.4A). The p53 protein forms dimers and tetramers upon

DNA in the nucleus of the cells during stress response activities (17). According to SDS-PAGE

and western blot analysis, wild-type p53 migrated at ~50 kDa and multiple bands on the gel

74

suggested post-translational modifications were present on p53 (Figure 5.4B). A positive signal

for ubiquitin modifications was confirmed by western blot analysis, accounting for much of the

higher molecular weight material. As p53 is subject to phosphorylation at many different

residues, other modifications are likely to occur in glioblastoma cells during disease progression.

We posit that p53 dimers and monomers were both present in our biochemical

preparation, albeit difficult to visualize individual monomers in the EM images. Class averages

of the dimeric particles, however, were well-defined with consistent features and particles

adopted multiple views in the images (Figure 5.4C). The final EM reconstruction refined to ~10-

Å according to the 0.5-Fourier Shell Correlation (FSC) criteria and the data was not limited in

the angular distribution of particles (Figure 5.4D). Molecular models from a crystal structure of a

p53 multimer engaging DNA was used to interpret the structure (pdb code, 2AC0; (18)). The

ubiquitin model (pdb code, 1UBQ, (19)) was placed in the density map in a biologically relevant

manner that enables the proper residues of ubiquitin to be in proper proximity to the p53 subunits

(Figure 5.4A). Overall, these results demonstrate the useful nature of Cryo-SiN microchips to

prepare native p53-DNA repair assemblies derived from human cancer cells. This technical

advancement provides the foundation for future work to analyze functional differences among

wild-type and mutated forms tumor suppressors as disease-related molecular targets.

75

Figure 5.4: Structural and biochemical analysis of p53 assembly isolated from human

cancer cells. (A) Cross sections through the 3D density map of the p53 dimer (grey structure;

pdb code, 2A0C) isolated from GBM cells (U87MG line). The density located proximal to the

dimer structure fits two ubiquitin monomers (yellow; pdb code, 1UBQ) and suggests a

biologically relevant configuration to attach to each p53 subunit. Scale bar is 15 Å. (B) p53

migrated to ~50 kDa according to a denaturing gel and western blot analysis. Extra bands that

migrated slower than ~50 kDa appear to be ubiquitinated-p53, according to western blot

detection. (C) EM image of the p53 preparation. Corresponding class averages show consistent

features and alternative views of the protein assembly. Scale bar is 100 Å; box size is 200 Å. (D)

According to the 0.5-FSC criteria, the ubiquitinated-p53 density map resolves to 10- Å and

particles were not limited in their angular distribution (inset distribution plot).

76

Conclusions

As the production of carbon support films can be highly variable, a valuable asset for the EM

community is the design of new flat substrates that are transparent to the electron beam. Using

custom-designed Cryo-SiN microchips, we observed enhanced image contrast in biological

specimens compared with similarly prepared conventional holey carbon films. Improved contrast

in particle images correlates with more accurate computing procedures and the potential to

produce EM structures with less time-consuming technical barriers. Due to the difficulty in

manufacturing uniformly flat and durable holey carbon film, charging effects may ensue during

imaging, which cause beam-induced movements. Adding amorphous carbon layers to blanket the

perforated film may mitigate charging effects, although these measures may become obsolete

through the use of Cryo-SiN. Other new support materials were also proven effective at reducing

beam-induced movements, which may cause resolution-limiting effects (13, 20, 21).

We demonstrate that custom-designed SiN microchips are a powerful tool to harvest

biological samples and prepare them for cryo-EM structural studies. Enhanced contrast was seen

in two case studies involving highly symmetric virus assemblies and non-symmetric cancer-

related proteins. As these model samples were effortless to prepare and assess using the custom-

designed substrates, we turned to a more challenging complex that is often mutated in aggressive

and incurable brain tumors.

Frozen-hydrated specimens of p53 assemblies were prepared from GBM cells.

Corresponding biochemical analysis and downstream computing procedures indicated a dimeric

p53 complex engaging DNA. The 10-Å structure of the native protein assembly was

ubiquitinated in a manner that is consistent with DNA damage response. As post-translational

modifications on p53 are important for its activation and nuclear repair processes, the captured

77

assemblies suggest a snapshot of the tumor suppressor performing one of its most essential

functions. This conformational view of p53 is important as certain ubiquitinated forms of the

protein enable erroneous DNA repair events, supporting the growth and proliferation of cancer

cells. Taking together, the use of custom-designed microchips may transform our view of

medically-relevant protein targets in the context of human health and disease.

Acknowledgements

This work was supported by the National Institutes of Health and the National Cancer Institute

[R01CA193578, R01CA227261, R01CA219700 to D.F.K.]. Additional support was provided by

the University of Virginia-Virginia Tech Carilion Seed Fund Award and the Cartledge Charitable

Foundation.

Author Contributions

D.F.K. conceived this project and supervised the completion of all experiments. N.A. and A.C.V.

prepared the manuscript. A.C.V., N.A., W.D., M.J.D. and Y.L. performed the experiments. J.D.,

J.M. and M.D. designed the microchips. ‡These authors contributed equally.

78

References

1. K. A. Taylor, R. M. Glaeser, Electron diffraction of frozen, hydrated protein crystals.

Science 186, 1036-1037 (1974).





4. D. J. De Rosier, A. Klug, Reconstruction of three dimensional structures from electron

micrographs. Nature 217, 130-134 (1968).

5. R. M. Glaeser, How good can cryo-EM become? Nat Methods 13, 28-32 (2016).

6. D. A. Kam J, Tanner JR, McDonald SM and Kelly DF, Structural dynamics of viral

nanomachines. Technology 2, 44-48 (2014).

7. B. L. Gilmore et al., Preparation of Tunable Microchips to Visualize Native Protein

Complexes for Single-Particle Electron Microscopy. Methods Mol Biol 1764, 45-58

(2018).

8. C. E. Winton et al., Tunable substrates improve imaging of viruses and cancer proteins.

Micros Today 25, 22-27 (2017).

9. Y. Liang et al., Structural analysis of BRCA1 reveals modification hotspot. Sci Adv 3,

e1701386 (2017).

10. A. W. P. Fitzpatrick et al., Cryo-EM structures of tau filaments from Alzheimer's disease.

Nature 547, 185-190 (2017).

11. E. A. Ring, D. B. Peckys, M. J. Dukes, J. P. Baudoin, N. de Jonge, Silicon nitride

windows for electron microscopy of whole cells. Journal of microscopy 243, 273-283

(2011).

12. E. S. Pohlmann et al., Real-time visualization of nanoparticles interacting with

glioblastoma stem cells. Nano Lett 15, 2329-2335 (2015).

13. D. A. Tanner JR, Dukes MJ, Melanson LA, McDonald SM, Kelly DF, Cryo-SiN - An

Alternative Substrate to Visualize Active Viral Assemblies. J Analyt Molecul Tech 1,

(2013).

14. S. H. Scheres, Semi-automated selection of cryo-EM particles in RELION-1.3. J Struct

Biol 189, 114-122 (2015).

15. P. A. Muller, K. H. Vousden, p53 mutations in cancer. Nat Cell Biol 15, 2-8 (2013).

16. D. Malkin et al., Germ line p53 mutations in a familial syndrome of breast cancer,

sarcomas, and other neoplasms. Science 250, 1233-1238 (1990).

17. P. N. Friedman, X. Chen, J. Bargonetti, C. Prives, The p53 protein is an unusually shaped

tetramer that binds directly to DNA. Proceedings of the National Academy of Sciences of

the United States of America 90, 3319-3323 (1993).

18. M. Kitayner et al., Structural basis of DNA recognition by p53 tetramers. Mol Cell 22,

741-753 (2006).

19. S. Vijay-Kumar, C. E. Bugg, W. J. Cook, Structure of ubiquitin refined at 1.8 A

resolution. J Mol Biol 194, 531-544 (1987).

20. D. Rhinow, W. Kuhlbrandt, Electron cryo-microscopy of biological specimens on

conductive titanium-silicon metal glass films. Ultramicroscopy 108, 698-705 (2008).

21. C. Yoshioka, B. Carragher, C. S. Potter, Cryomesh: a new substrate for cryo-electron

microscopy. Microscopy and microanalysis : the official journal of Microscopy Society of

79

America, Microbeam Analysis Society, Microscopical Society of Canada 16, 43-53

(2010).

80

Chapter 6

Cryo SiN Platform Reveal BRCA1 Complexes in Metastatic Cancer

Cameron Varano1,2, Nick Alden1, William Dearnaley1, Yanping Liang1, and

Deborah F. Kelly1,3

1. Virginia Tech Carilion Research Institute, Virginia Tech, Roanoke, VA 24016, USA.



3. Department of Biological Sciences, Virginia Tech, Blacksburg, VA 24061, USA.

81

Abstract

The breast cancer susceptibility protein (BRCA1) is known to be important in the prevention of

cancer development. Here we adapt silicon nitride chips with microwells for cryo-electron

microscopy to study the three-dimensional architecture of native BRCA1 complexes derived in

metastatic breast cancer cells. We further our previous BRCA1 structural studies to compare a

spectrum of BRCA1 complexes. Here we show that wild-type BRCA1 complexes have

additional density in the C-terminal domain compared to wild-type complexes from primary

breast cancer. The additional density is due to p53 monomers bound to the BRCT domain of

BRCA1.

Introduction

When breast cancer is treated early the 5-year survival rate is ~80%. The survival rate drops to

only 24% when breast cancer is found in later stages and spreads throughout the body.

Importantly, this low survival rate has not improved since 2003 (1). During metastasis, cancer

cells migrate from the primary site to other tissues. Common sites of metastasis for breast cancer

include the bone, brain, liver, and lungs (2). Individuals harboring germline mutations in the

breast cancer susceptibility gene, BRCA1, are ~60% more likely to develop these invasive

cancers (3). Following initial treatment, BRCA1-related cancers often return in less than 5 years

and rapidly metastasize (1). Although disease propensity is higher for individuals with BRCA1

mutations many individuals living with these mutations do not develop cancer. Meaning the

genomic status of BRCA1 cannot act as the sole predictor of breast cancer. We hypothesize that

visualizing the structure of BRCA1 will further our understanding into its function and role in

82

Figure 6.1. Schematic of the primary structures of p53, BRCA1 and BARD1. p53 is 393

amino acids and comprised of 3 domains- the transactivation domain (TAD), DNA-binding

domain (DBD), and C-terminus domain (CT). BRCA1 and BARD1 bind at their respective

RING domains. BRCA1 is 1863 amino acids. The RING domain is located at the N-terminus;

the nuclear localization sequence is located in the central region of the protein; the tandem

BRCT domains are located in the C-terminus domain. BARD1 is 777 amino acids with a RING

domain at the N-terminus and tandem BRCT domains at the C-terminus.

83

the development of cancer. To that end we examine the 3D structures of native BRCA1 protein

complexes from three different cancer cell lines.

BRCA1 is a highly flexible protein comprised of 1863 amino acids with a predicted

molecular weight of 208 kDa. The N-terminus of BRCA1 contains a RING domain (residues 1 –

300) that forms a functional heterodimer in the nucleus with its binding partner (Figure 6.1), the

BRCA1-Associated RING domain protein (BARD1) (4). Adjacent to the RING domain is the

flexible central portion of BRCA1 that contains nuclear localization sequences (NLS). This

central portion of BRCA1 represents nearly 70% of the protein’s mass. Following the central

region of BRCA1 is the C-terminal domain containing two motifs collectively known as the

BRCT domains (BRCA1 C-terminal domain, 214 residues) (Figure 6.1). Structures for the

BRCT domain have been determined by x-ray crystallography (5). Additionally, a portion of the

BRCA1-BARD1 RING domain structure has been determined through NMR (6). However,

examining a flexible, native protein complex such as BRCA1-BARD1 is best suited for single

particle electron microscopy. In 2017, we published the first full length structure of wild-type

BRCA1-BARD1 using single particle electron microscopy (7).

BRCA1 is categorized as a tumor suppressor protein due to its role in the cell cycle and

repairing genomic material. An essential component to its function as a tumor suppressor is its

role as an E3 ubiquitin ligase, which is coordinated by the RING domains of BRCA1 and

BARD1 (8). BRCA1, like other E3 ligases, serves as the platform for the E2 ligase to transfer a

ubiquitin to the substrate. The RING domain is where E2 ligase docks. The tandem BRCT

domains serves as the dock for the substrate. Within the BRCT domain is a known hydrophobic

binding pocket (9).

84

There are several known proteins that interact with BRCT domain of BRCA1 (10-14), of

importance to this study is p53. While the function of BRCA1 and p53 in response to DNA

damage have independently been well characterized, very little known about the interaction of

these two tumor suppressors(15). However, the homology between the BRCT domains of

BRCA1 and p53-binding protein 1 (53BP1) has been characterized (14). As its name suggests,

53BP1 is binds to p53. This homology supports the possible interaction of p53 at the BRCT

domain of BRCA1. This gap in knowledge has been limited by our ability to capture and image

native protein complexes.

In order to capture native complexes, we developed tunable microchips composed of

silicon nitride (SiN) (16, 17). This technology allows us to study protein complexes, such as

BRCA1-BARD1 produced in normal and diseased states. In this system, the microchips are

decorated in a layered manner comprised of a specialized lipid mixture containing functionalized

Ni-NTA polar head groups (Figure 6.2). These lipid coatings provide the added benefit of

segregating fragile proteins from the disruptive air-water interface during the vitrification

process. Additionally, the microwells provide added control during the blotting processes. This

results in more consistent ice thickness across the microchip and increased imaging capabilities.

Here we use the microchip capture approach to prepare cryo-EM specimens of native

BRCA1-BARD1 protein assemblies from metastatic breast cancer cells. Frozen-hydrated

specimens were examined using cryo-EM imaging routines. The molecular blueprints of the

metastatic complexes are compared to BRCA1 complexes derived from the primary site. These

findings provide insights into the structural properties of BRCA1 beyond the genome.

85

Figure 6.2. Schematic of silicon nitride microchip. (A) The silicon nitride microchip window

(550 nm by 550 nm) has integrated microwells. (B) The silicon nitride chip is 200 nm thick with

wells 150 nm deep. The schematic shows a monolayer (gold and orange), the orange represents

the incorporated nickel. (C) A side view of the microchip with integrated microwells

86


Authentication of cells, cell culture, and BRCA1 enrichment procedures

The MDA-MB-361 line used in this study was purchased from ATCC and independently

characterized by ATCC as BRCA1 wild-type cells. For all experiments, cells were used within 6

months of resuscitation. Cells were cultured in RPMI 1640 (Mediatech) supplemented with 20%

fetal bovine serum (ATCC) and 0.5× penicillin-streptomycin (Thermo Fisher Scientific) and

incubated at 37°C and 5% CO2. Phosphorylated BRCA1-BARD1 complexes were enriched

using immobilized metal affinity chromatography to which phosphorylated BRCA1 naturally

binds (18) (Figure 6.3A). Approximately 1 million cells were collected using trypsin-EDTA

(Thermo Fisher Scientific) and pelleted by centrifugation (500g; 5 min). The soluble nuclear

fraction was isolated using the NE-PER kit (Thermo Fisher Scientific). To enrich BRCA1

complexes within the soluble nuclear, material was incubated at 4°C with 200 μL of Ni-NTA

agarose beads (Qiagen) with gentle rotation for 1 hour. The beads with bound material were

transferred to a 3mL column and washed with 3mL of 20 mM HEPEs buffer (pH 7.2; 140 mM

NaCl, 2 mM CaCl2, 2 mM MgCl2, and 5 mM imidazole). Phosphorylated BRCA1-BARD1

complexes were eluted in the same HEPEs buffer supplemented with 200 mM imidazole. Protein

concentrations of elution fractions were determined using a Bradford assay (Thermo Fisher

Scientific).

87

Figure 6.3. An overview of the project work flow- from cells to biochemical evaluations and

imaging. (A) Metastatic breast cancer cells are pelleted then nuclear material is chemically

extracted. From the nuclear material, BRCA1 complexes are enriched by immobilized metal

affinity chromatography (IMAC) and then analyzed and imaged. (B) Coomassie stained SDS-

PAGE gel of fraction 2 which shows BRCA1 at 250 kDa and BARD1 at 87 kDa. (C) Co-IP

results display interactions between BRCA1, BARD1 and p53 in fraction 2. pSer2 serves as the

negative control in the deplete material (DEP) and normal rabbit antibody (IgG) serves as the

positive control. (D) A representative micrograph show a fairly homogenous sample. Scale bar is

60 nm. The associated 2D classifications show the clamp-like motif of BRCA1 complexes.

88

Coomassie blue staining

The enriched protein fractions were analyzed by SDS-PAGE with Coomassie blue staining

(Figure 6.3D). Proteins were separated on 3 to 8% tris-acetate NuPAGE mini gels (Thermo

Fisher Scientific). The gels were washed 3 times with ultra-pure water and then stained by

SimplyBlue SafeStain solutions (Invitrogen) for 6 min. Gels were washed with deionized water

overnight on an orbital shaker and then stored in 20% NaCl overnight to achieve maximum

sensitivity.

Co-immunoprecipitation analysis

Protein-protein interactions were detected by co-IP experiments were performed on isolated

BRCA1-BARD1 protein fractions (Figure 6.3B). First, antibodies against BRCA1 (5 μg Cell

Signaling Technology; 14823) or normal rabbit IgG (5 μg Cell Signaling Technology; 2729)

were diluted in PBS-T (0.05% Tween-20) before incubating with 0.75 mg Dynabeads Protein G

(Thermo Fisher). The mixtures were incubated with rotation for 30 minutes at 4 °C. Antibody-

coated beads were washed in 20 mM HEPES buffer (pH 7.2) containing 140 mM NaCl, 10 mM

CaCl2, 10 mM MgCl2 prior to adding enriched nuclear fraction 2 (E2) supplemented with

protease and phosphatase inhibitors (Thermo Fisher). This material was incubated with the

antibody-coated beads overnight at 4 °C with gentle rotation. The beads were then washed

followed by elution with 40μL of elution buffer (26 μL ultra-pure water, 10 μL NuPAGE 4x

LDS sample buffer (Thermo Fisher) and 4μL of DTT (0.5 M)). A 3–8% NuPAGE Tris-Acetate

mini gel with NuPAGE Tris-AcetateSDS running buffer was used to separate the proteins.

Following separation by SDS PAGE, the proteins were transferred onto an Immobilon-P

membrane (Millipore) in a Mini-PROTEAN Tetra system (Bio-Rad). The membranes were

89

incubated on a rocker at room temperature for 1 hour in blocking solution (5% NFDM or 5%

bovine serum albumin (SCBT)). The blots were incubated overnight with primary antibody

(diluted in 5% NFDM or bovine serum albumin solution) at 4 °C. Following three washes with

TBS-T (0.05% Tween-20), the membrane was incubated for 1 hour at room temperature with

secondary antibodies (goat anti-rabbit or goat anti-mouse secondary antibodies conjugated to

horseradish peroxidase (Jackson ImmunoResearch)). The membrane was washed three times

with TBS-T, prior to adding ECL Prime western blotting reagent (GE Healthcare) for detection.

The membrane was imaged using a ChemiDoc MP (Bio-Rad). The following primary antibodies

were used in our analysis: BRCA1-A8X9F (Cell Signaling Technology, #14823; a-RING),

BARD1 (SCBT, sc-11438), phospho-p53 (Sigma-Aldrich, P8982) and RNA Polymerase II H5

(pSer2-specific) (Covance, MMS-129).

EM specimen preparation and imaging

Samples of isolated BRCA1-BARD1 assemblies from metastatic breast cancer cells [0.2 mg/ml

in 20 mM Hepes buffer (pH 7.2), 140 mM NaCl, 10 mM CaCl2, 10 mM MgCl2] were applied to

affinity SiN microchips (Protochips). Affinity microchips were decorated with a lipid monolayer

with integrated nickel. BRCA1 complexes were tethered to the decorated grids by incubating Ni-

NTA eluate for 2 min, followed rapid freezing using the Vitrobot with a blotting time of 7 to 10

seconds. Specimens were examined using a FEI TEM (FEI Company) equipped with a LaB6

filament and operating at 120 kV under low-dose conditions. A representative image is shown in

Figure 6.3E. Images were recorded using an Eagle 2k HS CCD camera (FEI Company) with a

pixel size of 30 µm at a magnification of ×68,000 for a final sampling of 4.4 Å/pixel.

90

Figure 6.4. The structural spectrum of BRCA1 complexes in breast cancer. (A) The three

structures show the density maps of wild-type BRCA1 from primary cancer cells (cyan), mutated

BRCA1 from primary cancer cells (gray) and wild-type BRCA1 from metastatic cancer cells

(yellow). All three have the atomic structure for the RING domain (pink) and BRCT domain

(blue for wild-type and red for mutated). The metastatic structure also includes the atomic

structure for p53 DBD. Scale bar is 60 nm (B) Rotational views of the BRCA1 complexes from

metastatic cells. Scale bar is 60 nm (C) Cross sectional views of the C-terminal domain of the

metastatic structure show the fit of the BRCT domain and p53 DBD. (D) The angular distribute

shows that there is minimal preferential orientation (E) The 0.5 FSC determines the resolution of the

structure to be 10.8 Å.

91

Image processing

Individual complexes, particles, were selected from the EM images using the SPIDER software

package (19). 2D class averages were generated from selected particles using reference-free

alignment routines (Figure 6.3E). The particles were exported as an image stack into the

RELION software package(20). The RELION software package was used to refine and

reconstruct the individual complexes using an initial model from experimental data (wild-type

BRCA1-BARD1) with a low pass filter of 20 Å and a box size of 100 Å. The initial round of

refinement was based on the model, while later iterations were heavily based on the experimental

data with a regularization parameter of T=4. We followed standard reconstruction routines, using

a pixel size of 4.4 Å to produce 3D structures masked at ~350 Å. RELION parameters were set

to output 5 classes and run 25 iterations of refinement. The statistical output from RELION

indicated that the data could be combined into a single map. Statistically quivalent contour levels

were used to compare the EM maps to the previous structures (7) in the Chimera software

package (21). Particle heterogeneity was evaluated at both the 2D and 3D classification steps.

The final density maps are shown in Figure 6.4A. The resolution was determined by dividing the

particle data for each reconstruction into two halves and calculating separate density maps. We

used the 0.5 FSC criteria in RELION to determine the final resolution. The final structure

BRCA1-BARD1 from the metastatic cells (10.8 Å) contained 1615 particles.

Movie production

The Chimera software package (21) was used to produce the supplemental movie. A python file

of the scene was exported, then a Chimera command file was produced for the structure. These

files contain a list of command line instructions that Chimera parses and applies to the scene. At

92

the beginning of each movie, a trio of labels are generated to indicate the location of the RING

and BRCT domains and the DNA binding domain of p53. For cross sectional views, camera

slices were produced over a period of 150 frames. This procedure was reversed, as frame slices

were replaced. Next, the density model was removed to highlight the relationship between the

p53 DBD and the BRCA1 BRCT domains. The atomic models were rotated about the y axis by

1° per frame over 360 frames. The slicing procedure was repeated on the atomic models and then

reversed again over 125 frames each direction. To conclude the movie the density map

reappears.

RESULTS

BRCA1-BARD1 complexes bind to phosphorylated p53

BRCA1-BARD1 heterodimers are naturally produced in the nucleus, which can be enriched by

nickel chromatography (Figure 6.3A). Active BRCA1-BARD1 heterodimers are phosphorylated

which naturally bind to the Ni-NTA beads (22). The enrichment was analyzed by SDS–

polyacrylamide gel electrophoresis (PAGE) analysis; the phosphorylated form of BRCA1

migrated at ~250 kDa and BARD1 migrated at ~87 kDa (Figure 6.3B). Co-immunoprecipitation

(co-IP) experiments verified the interaction between BRCA1 and BARD1. Protein G magnetic

beads were decorated with antibodies against BRCA1 (A8X9F; Cell Signaling Technologies)

and then the beads were incubated with the enriched protein fraction from the Ni-NTA

chromatography. The precipitated material was analyzed using Western blot detection. We

identified BRCA1-BARD1 interactions by probing the blots with antibodies against BRCA1 and

BARD1 (Figure 6.3C). As the workflow indicates, we simultaneously analyzed the BRCA1

complexes from fraction 2 biochemically and by single particle analysis. The initial rounds of

EM imaging indicated additional density in the complexes from the metastatic cells. Based on

93

these results we probed the immunoblots for known binding partners of BRCA1. While BRCA1-

BARD1 complexes associate with multiple proteins, the co-IP results from the metastatic cells

showed interaction with phosphorylated p53.

Imaging wild-type BRCA1 in metastatic disease

In order to determine the 3D architecture of BRCA1-BARD1 natively formed in metastatic

breast cancer cells we used single particle EM imaging technology. The sample preparation and

imaging routines used in these studies were previously implemented to examine wild-type and

mutated BRCA1 complexes isolated from cells derived from the primary disease site (7). Here

we imaged the sample under low-dose cryo-EM conditions (<5e-/Å2) using an FEI Spirit

BioTWIN transmission electron microscope (TEM) operating at 120kV (Figure 6.3D).

Individual complexes were analyzed using the SPIDER software package (19) using standard

reference free alignment techniques. The 2D class averages for the wild-type BRCA1-BARD1

complexes showed clamp-like structure with a diameter of ~120 Å (Figure 6.3D) similar to those

previously found in wild-type BRCA1-BARD1(Figure 6.4A) (7). The particles were then

imported into the RELION software package (20) to reconstruct and refine an EM density map.

The 3D structure of BRCA1-BARD1 exhibited the clamp-like motif indicated in the 2D

averages. Atomic models of the RING [Protein Data Bank (PDB) code, 1JM7 (6)] and the BRCT

[PDB code, 1JNX (5)] domains were placed within respective regions of the density map. The

regions of the complex were distinguished using antibody-labeling experiments(23). The wild-

type structures from both the primary and metastatic cancer cells have similar overall molecular

architecture. However, the complex from metastatic cells have additional density in the BRCT

domains. Based on the results of the co-IP experiments the atomic structure of the DNA Binding

94

Domain (DBD) of p53 was placed in the C-terminal region of the metastatic density map. The

DBD has a similar sequence to that of a peptide known to interact with the BRCT binding pocket

(9). The DBD was oriented in a manner that mimicked the known binding of the peptide to the

BRCT domain. The orientation and quality of fit can be seen in the cross-sectional views of the

map with ribbon structures. (Figure 6.4C). The angular distribution graph demonstrates that the

particles did not have preferential orientation on the microchips (Figure 6.4D). The metatstatic

structure was refined to 10.8 Å according to the 0.5 Fourier shell correlation (FSC) criteria in

RELION (Figure 6.4E).

Structural spectrum of BRCA1-BARD1 complexes

We hypothesized that the architecture of wild-type BRCA1-BARD1 complexes at the metastatic

site differ from wild-type and mutant BRCA1-BARD1 complexes at the primary site. The

foundation of this study is built on our work revealing 3D structure of wild-type BRCA1-

BARD1 (7), which displays a clamp-like motif similar to other E3 ligase structures (24, 25).

Both the wild-type BRCA1-BARD1 and mutated BRCA1-BARD1 complexes were extracted

from the cells derived from the primary site (HCC70 and HCC1937, ATCC). These BRCA1-

BARD1 complexes were extracted and enriched using the same procedures conditions used on

complexes derived from the metastatic derived cells. While the metastatic cells were imaged

under frozen-hydrated conditions, the primary and mutant complexes were examined using

negative stain. However, consistent image processing routines were implemented in SPIDER and

RELION to generate the density maps. The resulting maps were contoured to statistically

equivalent levels in Chimera. A comparison of the maps reveals various levels of density in the

C-terminal domain. Examining wild-type BRCA1-BARD1 from primary breast cancer (cyan)

95

shows a tight fit of the density around the BRCT domain atomic models. Unsurprisingly, the

density map representing the mutated structure (gray) has a truncated C-terminal domain. The

mutation (BRCA15382insC) in this particular cell line results in a ~10-kDa truncation due to a

frameshift mutation in the C-terminus occurring at residue S1755 (7). The 3D density map of the

wild-type BRCA1-BARD1 complexes derived from the metastatic site contained additional

density in the C-terminal domain. BRCA1 binds with multiple proteins at in this region (10-14).

Co-IP experiments revealed the density represented a p53 monomer, which can indeed occupy

the density. Currently, there is no atomic structure of full-length p53. However, the DNA

Binding Domain (DBD) has been solved by x-ray crystallography (26) and comprises more than

50% of the fμLl protein (Figure 6.1). Additionally, the DBD contains a surface exposed peptide

sequence similar to a peptide known to bind in the BRCT binding pocket. Therefore, we

oriented the DBD into the density map to reflect this interaction. The orientation allows for the

remaining components of p53 to fit within the resolved density map. This association between

BRCA1 and p53 is unique from other maps.

DISCUSSION

In summary, we present the first cryo-EM structure of BRCA1-BARD1 isolated from human

metastatic breast cancer cells. Additionally, we provide a spectrum of structures from wild-type

and mutated complexes derived from both primary and metastatic sites. By contouring at

statistically equivalent levels we were able to directly compare these structures. While each of

the 3D density maps maintained the clamp-like motif exhibited in other E3 ubiquitin ligases (24,

25), unique differences were noted. Using the wild-type structure from primary breast cancer

96

cells as the baseline structure, we noted various levels of density in the C-terminal domain in the

mutated and metastatic structures. While the absence of density in a truncated protein is

expected, it provides a high level of contrast for the additional density in the complex extracted

from the metastatic complexes.

To further explore the BRCA1 complexes from metastatic cells, we preformed biochemical

assays. Western Blot analysis of co-IP experiments shows that BRCA1 (~250 kDa) associates

with BARD1 (~87 kDa) and phosphorylated p53 monomer (~68 kDa). The manner in which

BRCA1 associates with p53 has remained unclear. In an early study by Zhang et al. (1998) data

from mutated or truncated recombinant expressions of both BRCA1 and p53 suggested that C-

terminal domain of p53 bound to BRCA1 close to the N-terminal domain adjacent to the RING

domain (15). A later study performed by Joo et al. found that it was the DBD of p53 which

associated with BRCT domains of BRCA1 (14). Both studies relied on the manipulation and

often extreme truncation of the two proteins. So, while these studies came to different

conclusions they both highlight the need to study protein complexes in their full and native

states. Our data shows additional density in the C-terminal domain, which sufficiently

accommodates for a protein at least 50 kDa in size. This combined with the co-IP data, suggest

that BRCA1 and p53 do associate at the BRCT domains of BRCA1. To more fully understand

the interaction orientation of the complex the cryo-EM data will need to be resolved to a higher

level. The functional implication of this association is ongoing. One possibility could be that

BRCA1 is acting as an E3 ubiquitin ligase and p53 is the substrate. Future directions of this work

also include examining a possible “similarity trap.” A similarity trap requires two proteins with

highly homologous binding domains. The binding protein of one of the two proteins has

disruption in its affinity with its functional binding partner. Consequently, it has a non-functional

97

or over association with the protein having a similar domain. This idea has led to the

consideration of carcinogenic protein-protein interactions(27). For example, the BRCT domain

is a functional binding motif found in many proteins involved in DNA repair pathways (28-30);

one such example is 53BP1. Interestingly, molecular dynamic modeling studies have shown that

while the DBD of p53 has an affinity for both BRCA1 and 53BP1, the affinity for 53BP1 is

higher. The researchers then showed that affinity equilibrium can be shifted through various

mutations or alterations to either protein.

The structural spectrum presented here supports the need to examine proteins implicated in

the progression of disease beyond the genomic level. Additionally, we present a technology to

protect and allow for cryo-EM analysis of these complexes in their native form. The microchip

with integrated wells provide a platform which is able to tether the complex away from the

destructive air water interface. The wells allow for protection during the blotting process and

result in consistent ice formation.

Overall, our results provide a perspective on the structural variations of native BRCA1

complexes in breast cancer which is currently missing in the field. We found that p53 monomers

are bound to BRCA1 in a manner unique to the complexes derived from the metastatic site.

These findings indicate the varied nature of BRCA1 complexes in varies stages of disease.

Further investigation into the structural spectrum of BRCA1 complexes paired with functional

assays can provide new insights into the relationship of BRCA1 and breast cancer.

Acknowledgements This work was supported by funds from the Commonwealth Health Research Board (2080914),

the Concern Foundation (303872), NIH/National Cancer Institute (1R01CA193578-01A1), and

the University of Virginia–Virginia Tech Carilion Neuroscience Seed Fund Award.

98

Author Contribution

D.F.K. conceived this project and supervised the completion of all experiments. A.C.V. prepared

the manuscript. A.C.V., N.A., W.D. and Y.L. performed the experiments.

99

References

1. SEER Cancer Statistics Review (1975-2013 http://seer.cancer.gove/csr/1975_2013/). 2. . (2016). 3. M. C. King, J. H. Marks, J. B. Mandell, Breast and ovarian cancer risks due to inherited mutations

in BRCA1 and BRCA2. Science 302, 643-646 (2003). 4. J. E. Meza, P. S. Brzovic, M. C. King, R. E. Klevit, Mapping the functional domains of BRCA1.

Interaction of the ring finger domains of BRCA1 and BARD1. J Biol Chem 274, 5659-5665 (1999). 5. X. Zhang et al., Structure of an XRCC1 BRCT domain: a new protein-protein interaction module.

Embo J 17, 6404-6411 (1998). 6. P. S. Brzovic, P. Rajagopal, D. W. Hoyt, M. C. King, R. E. Klevit, Structure of a BRCA1-BARD1

heterodimeric RING-RING complex. Nat Struct Biol 8, 833-837 (2001). 7. Y. Liang et al., Structural analysis of BRCA1 reveals modification hotspot. Sci Adv 3, e1701386

(2017). 8. K. L. Lorick et al., RING fingers mediate ubiquitin-conjugating enzyme (E2)-dependent

ubiquitination. Proceedings of the National Academy of Sciences of the United States of America 96, 11364-11369 (1999).

9. S. J. Campbell, R. A. Edwards, J. N. Glover, Comparison of the structures and peptide binding specificities of the BRCT domains of MDC1 and BRCA1. Structure 18, 167-176 (2010).

10. X. Yu, C. C. Chini, M. He, G. Mer, J. Chen, The BRCT domain is a phospho-protein binding domain. Science 302, 639-642 (2003).

11. B. Wang et al., Abraxas and RAP80 form a BRCA1 protein complex required for the DNA damage response. Science 316, 1194-1198 (2007).

12. X. Yu, J. Chen, DNA damage-induced cell cycle checkpoint control requires CtIP, a phosphorylation-dependent binding partner of BRCA1 C-terminal domains. Mol Cell Biol 24, 9478-9486 (2004).

13. M. Ouchi et al., BRCA1 phosphorylation by Aurora-A in the regulation of G2 to M transition. J Biol Chem 279, 19643-19648 (2004).

14. W. S. Joo et al., Structure of the 53BP1 BRCT region bound to p53 and its comparison to the Brca1 BRCT structure. Genes Dev 16, 583-593 (2002).

15. H. Zhang et al., BRCA1 physically associates with p53 and stimulates its transcriptional activity. Oncogene 16, 1713-1721 (1998).

16. B. L. Gilmore et al., Preparation of Tunable Microchips to Visualize Native Protein Complexes for Single-Particle Electron Microscopy. Methods Mol Biol 1764, 45-58 (2018).

17. C. E. Winton et al., A microchip platform for structural oncology applications. NPJ Breast Cancer 2, (2016).

18. R. Scully et al., BRCA1 is a component of the RNA polymerase II holoenzyme. Proc Natl Acad Sci U S A 94, 5605-5610 (1997).

19. J. Frank et al., SPIDER and WEB: processing and visualization of images in 3D electron microscopy and related fields. J Struct Biol 116, 190-199 (1996).

20. S. H. Scheres, RELION: implementation of a Bayesian approach to cryo-EM structure determination. J Struct Biol 180, 519-530 (2012).

21. E. F. Pettersen et al., UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25, 1605-1612 (2004).

22. R. Scully et al., BRCA1 is a component of the RNA polymerase II holoenzyme. Proceedings of the National Academy of Sciences of the United States of America 94, 5605-5610 (1997).

http://seer.cancer.gove/csr/1975_2013/

100

23. B. L. Gilmore et al., A Molecular Toolkit to Visualize Native Protein Assemblies in the Context of Human Disease. Sci Rep 5, 14440 (2015).

24. G. Nalepa, M. Rolfe, J. W. Harper, Drug discovery in the ubiquitin-proteasome system. Nat Rev Drug Discov 5, 596-613 (2006).

25. C. E. Berndsen, C. Wolberger, New insights into ubiquitin E3 ligase mechanism. Nat Struct Mol Biol 21, 301-307 (2014).

26. M. Kitayner et al., Structural basis of DNA recognition by p53 tetramers. Mol Cell 22, 741-753 (2006).

27. J. Liu, Y. Pan, B. Ma, R. Nussinov, "Similarity trap" in protein-protein interactions could be carcinogenic: simulations of p53 core domain complexed with 53BP1 and BRCA1 BRCT domains. Structure 14, 1811-1821 (2006).

28. P. Bork et al., A superfamily of conserved domains in DNA damage-responsive cell cycle checkpoint proteins. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 11, 68-76 (1997).

29. I. Callebaut, J. P. Mornon, From BRCA1 to RAP1: a widespread BRCT module closely associated with DNA repair. FEBS Lett 400, 25-30 (1997).

30. E. V. Koonin, S. F. Altschul, P. Bork, BRCA1 protein products ... Functional motifs. Nat Genet 13, 266-268 (1996).

102

Chapter 7

Conclusions

In this body of work, we utilize single particle EM to examine macromolecules implicated in

human disease, specifically rotavirus double layered particles, fibrocystin/polyductin complex,

and BRCA1 complexes. Our findings presented here not only further our understanding of these

important proteins, but present new methodologies to more comprehensively study native protein

structures.

Rotavirus double layered particles are large icosahedral complexes. Previous structural

studies of rotavirus were of transcriptionally inactive particles, as it limited, the confirmation of

the particles. While the structural insights were highly resolved, the biological insights were

limited. First, we executed traditional cryo-EM routines to compare structures of rotavirus in

transcriptionally active states compared to inactive states. Our data showed internal structural

changes in the virus particles based on transcriptional status. Additionally, the data showed that

when symmetry was not enforced the inactive particles retained icosahedral symmetry, while the

active particles did not display any symmetry. These findings led to the development of a liquid

cell methodology that allows viral particles to be evaluated under temporally dynamic

conditions. This new imaging process revealed that not only were the active particles mobile, but

that the data suggests their mobility was related to the level of transcriptional activity.

We applied previously established methodologies to enrich and examine

fibrocystin/polyductin complex (FPC) using EM. FPC is a multidomain protein containing a

large extracellular domain, transmembrane domain and a small intracellular C-terminal domain.

This membrane bound protein undergoes Notch-like proteolytic cleavage, resulting in a

functional carboxy-terminal domain (CTD-FPC). The CTD-FPC is translocated to the nucleus

103

where it is involved in a variety of gene-related functions. Previous molecular studies have

established a link between FPC and autosomal recessive polycystic kidney disease (ARPKD),

which is a leading cause of end stage renal disease in children. Despite its known connection

with ARPKD, little structural information has been resolved due to its low abundance. In order

to enable structural analysis on FPC assemblies, we over-expressed a his-tagged (6x-His) version

of the functional carboxy-terminal domain of FPC. This allowed us to employ classical EM

techniques to resolve an initial density EM map of this important protein. Our findings not only

provide important structural information, additionally they further establish the role of EM in

understanding protein complexes vital to human health.

BRCA1 is an asymmetric and flexible protein which associates with its binding partner

BARD1 to form a functional heterodimer. We previously resolved the structure of native wild-

type BRCA1-BARD1 isolated from cells derived at the primary disease site. Here, we resolved

the structure of native wild-type BRCA1 complexes isolated from metastatic cells to further

expand our understanding of BRCA1 and cancer progression. To do this, we adapted silicon

nitride microchip technology for cryo-EM imaging. The microchip protected and stabilized the

protein complex without disrupting its confirmation. Interestingly, we found that the BRCA1

complexes isolated from the metastatic cells have additional density in the C-terminal domain,

which can be attributed to an interaction with p53 monomers.

In summary, our structural examination of rotavirus, FPC and BRCA1 complexes has

revealed the need to implement single particle EM methodologies that evaluate native

macromolecule complexes in biologically relevant conditions. These findings will help create

frameworks to gain insights into the molecular world, where small differences can have large

impacts on human health.

imaging technology a. cameron varano virginia polytechnic ... › bitstream › handle › 10919 ›...

Documents