Cyclic AMP Dependent Protein Kinase a (PKA) Mutant Associated with Fibrolamellar Hepatocellular Carcinoma (FLHCC): Structure, Dynamics and in Cell Studies

Cyclic AMP dependent Protein Kinase A (PKA) mutant associated with Fibrolamellar Hepatocellular carcinoma (FLHCC): Structure, dynamics and in cell studies.

A Dissertation SUBMITTED TO THE FACULTY OF THE UNIVERSITY OF MINNESOTA

Adak Karamafrooz

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

■ David D Thomas

Laurie Parker

This work is dedicated to:

All genuine souls who have remained loyal to their integrity, despite all odds.

…and to my dear ones who bear with me all the way through.

“To strive, to seek, to find, and not to yield…”

ALFRED, LORD TENNYSON

List of Tables

3.1 Michaelis-Menten Kinetics constants for PKA-CDNAJB1and PKAWT ...... 110 3.2 Midpoint of thermal denaturation (Tm) for PKA-CDNAJB1and PKAWT...... 110 µ 3.3 Dissociation constants ( M) of nucleotides and inhibitor peptide PKI5-24 ...... 110 5.1 Subset of PRKACA-Dnajb1 direct Peptide Substrates identified by KALIP ...... 127 5.2 subset of PRKACA-Dnajb1 direct Protein Substrates identified by KALIP ...... 128 5.3 Canonical pathways that are enriched in both protein and peptide KALIP ...... 128 5.4 Subset of PRKACA-Dnajb1 substrates affected by the two PKA inhibitors ...... 129 6.2 Catalytic parameters derived from coupled enzyme assay for Human PKA and Cushing’s disease related PKA mutants W196R and L205R ...... 142

6.3 Comparison of the Thermodynamic parameters for the two Cushing's disease mutant compared to the wild type PK ...... 143

List of Figures

CHAPTER 1 1.1 Position of the catalytic and glycine rich loop in the conformation of the PRKACA...... 102 1.2 Hydrophobic C-spine and R-spine in the conformation of PRKACA and their relative position compared to important loops in the PKA catalytic subunit ...... 102 1.3 Glycine-rich loop of the C subunit in its closed and open conformation ...... 103 1.4 Post translational modifications in the N-terminal tail of PKA-Cα...... 103 1.5 hydrogen bonding of the tip of the loop in the PKA-C:PKI ...... 103 1.6 Reaction pathway for catalysis ...... 103 1.7 Regulatory and catalytic subunit complex in PKA ...... 104 1.8 Sequence alignment of the hinge of different R subunit isoforms ...... 104 1.9 A kinase anchoring protein (AKAP) bound to the regulatory subunit ...... 104 1.10 A schematic representation of cAMP signaling pathway inside the cell ...... 105 1.11 A simplified scheme of the possible role of cAMP/PKA pathway in cancer ...... 105

CHAPTER 3 3.1 PKA DANJB1 (Cyan, PDB:4WB7) crystal structure superimposed on PKAWT ...... 106 3.2 Statistical analysis of the chemical shift changes in PKA DANJB1 ...... 106 DANJB1 3.3 A. Chemical shift analysis of PKA in complex with PKI5-24 vs PKIFL ...... 107 WT 3.3 B. Chemical shift analysis of PKA in complex with PKI5-24 vs PKIFL ...... 107 3.4 Mutual information and allosteric networks in PKAWT and PKA DANJB1 ...... 108 3.5 Emergence of allosteric hubs upon ATP binding in PKAWT and PKA DANJB1 ...... 109

CHAPTER 4 4.1 Structures of J-PKAcα chimera and wild-type PKAcα...... 111 4.2 RMSF per residue. RMSF for both J-PKACα and wild-type PKACα ...... 112 4.3 Movement of the J-domain in J-PKAcα chimera...... 113 4.4 Top four clusters by population from cluster analysis ...... 114 4.5 Top clusters from cluster analysis modeled into the RIIβ holoenzyme ...... 115

iii

4.6 Residue-specific T1/T2 ratio of NMR relaxation of J-PKACα ...... 116 S.1 Backbone assignment of PKA-DNAJB1 ...... 117 S.2 1H-15N TROSY-HSQC spectra for PKADNAJB1 ...... 118 S.3 Mutual information and allosteric networks in the binary and ternary forms of PKA WT ...... 119

CHAPTER 5 5.1 A. Schematic presentation of Kinase assay linked with phosphoproteomic (KALIP)...... 120 5.1 B. Overlap of the in-cell phospho-proteins ...... 120 5.2 Comparison of some of enriched canonical pathways derived from phosphorylation analysis of in-cell PRKACA and PRKACA-Dnajb1 ...... 121

5.3 Comparison of Phosphorylation sites in each of PRKACA and its mutant PRKACA-Dnajb1...... 122

5.4 Comparison of canonical pathways derived from IPA phosphorylation analysis of PRKACA and PRKACA-Dnajb1 Peptide direct substrates ...... 123

5.5 Comparison of canonical pathways derived from IPA phosphorylation analysis of PRKACA and PRKACA-Dnajb1 Protein direct substrates ...... 124

5.6 A. Pattern of downregulated phosphorylation sites in each species ...... 125 5.6 B. Comparison of phosphorylated substrates percentage in PRKACA-Dnajb1 in the presence of inhibitors PKI and rp-cAMP ...... 125

5.7 Inhibition of canonical pathways derived from IPA phosphorylation analysis of PRKACA and PRKACA-Dnajb1 ...... 126

S.1 Western blot of kinases expressed in HEK293 cells ...... 130

CHAPTER 6

6.1 Location of the Cushing’s disease mutations at the interface of R-Cα binding site ...... 142 6.2 Affinity pull-down assay using His-tagged RIIB and untagged W196R ...... 143 6.3 Inhibitory mechanism using Non-Radioactive Protein Kinase Assays ...... 143 6.4 TROSY-HSQC for PKA-Human Isoform II (Apo form) ...... 144 6.5 TROSY-HSQC of Isoform 2 in APO and AMPPNP bound form overlaid ...... 144 6.6 PKA-Human isoform II in complex with nucleotide and PKI ...... 145

6.7 TROSY-HSQC spectra of 15N, 13C perdeuterated Human PKA bound to ATPγN and PKI Ternary complex) ...... 145 5-24 (

6.8 PKA mutant “L205R” in the apo and in ATP-γN bound (binary) and PKI5-25 (ternary) form ...... 146 6.9 CONCISE analysis applied to PKA mutant “L205R” ...... 147 6.10 TROSY-HSQC spectra of 15N-labeled W196R ...... 147 6.11 Comparison of correlation matrix of the sidechain movements in W196R with that of PKA-WT ...... 148

List of Abbreviations:

ADP: Adenosine Diphosphate AKAP: A kinase anchoring protein ATP: Adenosine Triphosphate

ATPγN: Adenylyl-imidodiphophate cAMP: cyclic adenosine monophosphate CHESCA: chemical shift covariance analysis CNC: Carney complex CONCISE: COordinated ChemIcal Shifts bEhavior CREB: cAMP Response Element-binding protein SCP: chemical shift perturbation EPK: Eukaryotic Protein kinase EPR: Electron Paramagnetic Relaxation ERK: Extracellular signal-Regulated Kinase FL-HCC: Fibrolamellar Hepatocellular Carcinoma FLC: Fibrolamellar Hepatocellular Carcinoma FPLC: Fast protein liquid chromatography FRET: Förster resonance energy transfer HSQC: Heteronuclear Single Quantum Correlation Spectroscopy ITC: isothermal titration calorimetry KALIP: Kinase Assay Linked with Phosphoproteomic LC-MS: Liquid Chromatography-Mass spectrometry MAPK: Mitogen-Activated Protein Kinase mTOR: mammalian target of rapamycin MD: Molecular Dynamics NMR: Nuclear Magnetic Resonance NMT: N-myristoyltransferase NOE: Nuclear Overhauser Effect

NOESY: Nuclear Overhauser Enhancement Spectroscopy PCR: Polymerase Chain Reaction PKA: cAMP-dependent protein kinase A PKA-C: catalytic subunit of cAMP-dependent protein kinase A

PKA-Cα: catalytic subunit of cAMP-dependent protein kinase A PKI: Protein Kinase Inhibitor

PKI5-24: 20 amino acid fragment of Protein Kinase Inhibitor PTMs: post-translational modifications RMSD: Root Mean Square Deviation RMSF: Root Mean Square Fluctuation SAXS: Small angle x-ray scattering SMOAC: Sequential enrichment by Metal Oxide Affinity Chromatography TMS: Tetramethylsilane TOCSY: TOtal Correlated SpectroscopY TROSY: Transverse Optimized Spectroscopy

vii

Contents

Dedication ...... i List of Tables ...... ii List of Figures ...... iii List of Abbreviations ...... v i 0 Preface ...... 3 1 cAMP-dependent Protein Kinase A (PKA): history, structure and physiology ...... 7 1.1 Architecture of PKA-Cα ...... 7

1.2 Post-translational modifications in PKA-Cα ...... 10

1.3 Kinetics and Catalysis mechanism in cAMP dependent PKA ...... 11

1.4 Substrate recognition motif ...... 12

1.5 Regulators and inhibitors of PKA in cell ...... 13

1.6 Compartmentalization of PKA and cAMP signaling ...... 14

1.7 PRKACA and Human disease ...... 16

2 Application of Solution NMR Spectroscopy in Biological systems ...... 19 2.1 NMR parameters...... 20

2.2 Chemical shift ...... 21

2.3 Two and Multi-dimensional NMR ...... 22

2.4 NMR resonance assignment ...... 24

2.5 Detecting Molecular Interaction by NMR ...... 27

2.6 Protein Dynamics by NMR ...... 30

3 Conformational dynamics and allostery in the DNAJB1-PRKACA chimeric transcript linked to fibrolamellar hepatocellular carcinoma...... 34 3.1 Preface ...... 35

3.2 Introduction ...... 35

3.3 Material and Methods ...... 37

3.4 Results ...... 41

3.5 Conclusion ...... 44

4 Conformational Landscape of the PRKACA-DNAJB1 Chimeric Kinase, the Driver for Fibrolamellar Hepatocellular Carcinoma ...... 46 4.1 Introduction ...... 48

4.2 Results ...... 49

4.3 Discussion ...... 57

4.4 Material and Methods ...... 59

5 Using an integrated Phosphoproteomics method for identifying Human Protein Kinase A (PRKACA) and its oncogenic mutant DNAJB1-PRKACA substrates...... 63 5.1 Preface ...... 64

5.2 Introduction ...... 64

5.3 Material and Methods ...... 67

5.4 Results ...... 73

5.5 Discussion ...... 77

5.6 Conclusion ...... 81

6 PKA in Cushing’s disease ...... 82

References ...... 89 Figures and tables ...... 102

Preface

Since phosphorylation has been discovered to be a universal means of

communication in the cell, Protein kinases have been known to play crucial role in all

critical cell processes including metabolism, gene expression, and cell proliferation.

Constituting some 2% of the proteome, kinases regulate almost all biochemical pathways

and may phosphorylate up to 30% of the proteome. Chemically, protein kinases transfer the γ-phosphoryl group of a nucleotide triphosphate (e.g., ATP) to the hydroxyl group of a serine, threonine, tyrosine or histidine residue of the substrate protein. Over 500 protein kinases have been identified in the human genome (~ 1.7 % of genes) [1], pointing to the

biological importance of phosphorylation of cellular targets. Protein phosphorylation is one

of the most common post-translational modifications and is controlled by the opposing

actions of various specific protein kinases and phosphatases (PPs). The phosphorylation

status of a protein may affect its enzymatic activity, cellular localization, dynamics and/or

association with other proteins.

Protein kinase A (also known as the cyclic AMP-dependent protein kinase or A

kinase) is an eminent member of the AGC protein kinase family that transfers the

phosphoryl group to the side chains of serine (Ser) or Threonine (Thr) residues in the

substrates. Since protein phosphorylation affects many cellular processes, its regulation

must be specific and act on a defined subset of cellular targets to ensure signal fidelity.

Specificity is achieved through assembly of distinct complexes of enzymes at subcellular

localizations by various anchoring and scaffolding proteins. Being a regulatory Enzyme

itself, protein kinase A (PKA) is tightly regulated in space and time through different

mechanisms which will be discussed in chapter one. given the complicated task of cell

signaling, sophisticated regulatory mechanisms ensure that signaling enzymes encounter

3 their intracellular substrates in the right place and at the right time. Protein Kinase A catalytic subunit (known as PKA-C or PRKACA) has been extensively studied. Since its discovery as one of the earliest kinases [2], several structural, functional and dynamic studies on PKA-C have been published and it has served as a prototype for all kinases as general.

PKA as well many other signaling enzymes have a modular organization, being composed of domains with binding or catalytic functions that is often spatially separated from regions that serve as docking or substrate sites for other molecules. Every part of

PKA-C surface seems committed to some type of protein-protein interaction, and these interactions are as essential to its function as is phosphoryl transfer. Moreover, spatial control is achieved through association with anchoring proteins that ensure specificity in signal transduction by placing enzymes close to their cognate effectors and substrates inside the cell, therefore PKA role as a scaffold is also important and should not be underestimated.

Despite the amount of work that has been done on protein kinase A since its crystal structure was resolved [3,4] there are still new variants of this enzyme that introduce unknown aspects about the catalytic or regulatory mechanism of PKA. Several disease- related PKA-C mutants have been reported in recent years, which makes PKA the center of researcher’s attention again. Recent genomic studies have identified recurrent mutations in the catalytic subunit of PKA in tumors associated with Cushing’s syndrome, a kidney disorder leading to excessive cortisol production, and in tumors associated with fibrolamellar hepatocellular carcinoma (FL-HCC), a rare liver cancer [5]. Detection of a single, consistent genetic deletion in one copy of chromosome 19 [6,7] results in the formation of a chimeric gene, DNAJB1-PRKACA, which combines the first exon of

DNAJB1, the heat-shock protein 40 with exons 2 through 10 of PRKACA, a catalytic

subunit of protein kinase A. This chimera has been found to be the main driver of a rare

and highly aggressive form of liver cancer (Fibrolamellar hepatocellular carcinoma or

FLHCC) that affects predominantly young patients without underlying cirrhosis or disease

[8]. The kinase resulting from the chimeric transcript is fully functional (PKA-CDNAJB1) and

comprises 405 residues with 69 amino acids of DNAJB1 replacing the first 14 residues of

the kinase N-terminal A-helix [6] . A recent X-ray crystallography study revealed that the

structure of the kinase catalytic core remains essentially unchanged upon fusion of

DNAJB1 with PRCAKA and the structure of the ternary complex is virtually

superimposable to that of PKA-C. Additionally, In-vivo enzymatic assays [6] as well as in

vitro studies using classical coupled enzyme assays (using kemptide as substrate)

showed that PKA-CDNAJB1 activity is almost indistinguishable from wild-type PKA-C (PKA-

CWT). knowing these facts about this chimeric oncogenic kinase, and in order to further

investigate the structural features of the chimera and elucidate the effects of the DNAJB1

appendix on its kinetics and substrate binding, I have used different spectroscopic and

biophysical methods such as Nuclear Magnetic Resonance (NMR) spectroscopy, isothermal titration calorimetry (ITC), Förster resonance energy transfer (FRET), Electron

Paramagnetic Relaxation (EPR) and Small angle x-ray scattering (SAXS). NMR spectroscopy was extensively used as the main technique for structural assignment and allostery, as well as relaxation studies, and for binding assays and cooperativity studies I used ITC. Other in-solution studies such as EPR and SAXS were also used to further investigate the conformational dynamics of PKA-CDNAJB1.

The overall goal of this thesis is to understand the structure, dynamics and function

of PKA-CDNAJB1 in vitro and in vivo, using human PKA-CWT as reference for the studies. I have utilized NMR for most of structural and dynamics studies, and ITC for binding

experiments; Phosphoproteomics were used for in vivo studies and FRET for in cell

localization of the PKA-CDNAJB1 and PKA-CWT.

Chapters 1-3 outlines the background and theory of PKA, allostery, binding and solution state nuclear magnetic resonance (NMR). In chapters 4-6 I will include with permission the reprints and publications that I was involved.

Chapter I

cAMP-dependent Protein Kinase A (PKA): history, structure and physiology

1.1. Architecture of PKA-Cα (PRKACA)

Cyclic AMP dependent protein kinase A (PKA) was originally discovered by pioneer work of biochemists Edmond H. Fischer and Edwin G. Krebs, who were working on the role of protein phosphorylation and control of glycogen metabolism. Fischer and

Krebs named their new enzyme cAMP-dependent protein kinase and purified the kinase in 1968 [9]. Fischer and Krebs won the Nobel Prize in Physiology or Medicine in 1992 for this discovery and subsequent work on phosphorylation and dephosphorylation, particularly how it relates to cAMP dependent protein kinase activity [10]. Since then, there has been concerted efforts by several investigators to discern the molecular basis of activation, substrate binding/recognition and regulation of PKA-C, with perhaps the most surprising finding of catalytic (C) subunit that was constrained in an inactive state by a range of regulatory (R) subunits [11-13]. The inactive PKA holoenzyme exists as a tetramer of two regulatory subunits which form a complex with two catalytic subunits [14].

This tetrameric protein complex constitutes the type I and type II PKA holoenzymes [15].

This unique type of regulatory tetramer complex was found in only one other AGC kinase

(CK2) among over 540 human protein kinases. In contrast, the vast majority of protein kinases contain intramolecular regulatory sequences that reside within the same polypeptide chain as the catalytic core of the enzyme [16].

Despite their remarkable diversity in the sequence, substrate specificity and regulation of kinases, all kinases share a highly conserved catalytic core [17,18]. The kinase domain can be conceptualized as two functional modules: a highly conserved ATP- binding and catalytic module, located between the small and large lobes, found in all typical and atypical protein kinase structures; and substrate/peptide binding groove

7 localized primarily to the large lobe. Very little is known about this region of kinases as detailed knowledge would require many kinase structures co-crystallized with full-length substrates, and as yet no single such structure exists.

Conserved protein kinase A core and active site cleft

The basic features of the C subunit were described when the structure of a binary complex containing the inhibitor peptide from the heat-stable protein kinase inhibitor (PKI

5–24) was solved (Knighton et al., 1991a,b). The subsequent elaboration of ternary complex

(containing both nucleotide and peptide inhibitor) revealed the detailed features of the

ATP-binding site; additional structures of the enzyme in complex with peptide substrates and adenosine, revealed a set of different conformational states that correlate with opening and closing of the active site cleft.

The overall architecture of protein kinase A consists of a conserved small (N- terminal) lobe, featuring a β-sheet with five strands; and the large (C-terminal) lobe, consisting mostly of α-helices and loops. The nucleotide binding site is located deeply between these two lobes, in a hydrophobic core. The nucleotide (ATP) acts as a dynamic and allosteric activator by coupling the two lobes of apo PKA, enhancing the enzyme dynamics synchronously, and priming it for catalysis [19,20]. NMR and crystallography studies have shown that PKA catalytic core toggles between an open and close conformation during catalysis, these dynamic features are associated with the active enzyme and reflect movements that are thought to be an essential part of catalysis and product release. Two hydrophobic “spines” connecting the N-lobe with the C-lobe have also been discovered, named regulatory or “R” spine and Catalytic or “C” spine [21]. R and C spines are highly conserved spatial motifs indicating an active kinase but missing in inactive kinases. R-spine comprises four non-consecutive hydrophobic residues, two from the N-lobe: Leu106 from the β4 strand and Leu91 from the C-helix; and two from the

C-lobe: Phe185 from the Activation Loop and Tyr164 from the Catalytic Loop (figure.1.1).

It can be thought of as a hydrophobic “spine” that links the two lobes. It is widely accepted that, once the R-spine of an Eukaryotic Protein kinase (EPK) is assembled, its N- and C- lobes perform a “breathing motion,” opening and closing the active site cleft [19, 23, 24], a motion that has been known to be necessary for catalysis (As seen in figure 1.2, assembly of the R-spine brings together most of the essential motifs of the kinase core and leaves the enzyme poised for catalysis.

Other than this conserved catalytic core, two additional motifs were discovered in

PKA-C: an N-terminal helix, αA, comprising of approximately 40 amino acids that sits adjacent to αC and αE and a myristoylation group on the N-terminus. The C-terminal tail contains the last sixty amino acids and wraps around the kinase from the bottom of the C- lobe all the way up behind the C-helix. Although this C-terminal tail is not shared between all kinases, these have been found to be the hallmark of the AGC kinase family [22]. most of the conserved residues converge at the active site cleft between the two lobes, and contribute either directly or indirectly to ATP binding and catalysis [25]. Many of these residues are contained within three loops: the glycine-rich loop located in the small lobe

(residues 47-58), the catalytic loop (Asp166–Asn171) and the Mg2+ positioning loop

(Asp184-Phe187) both located in the large lobe (figure1.1). Lys168, conserved in the protein kinases specific for serine and threonine, interacts directly with the γ-phosphate of

ATP, while Asp184 chelates the Mg2+ that bridges the β and γ phosphates of ATP and is essential for both ATP binding and catalysis [27].

The catalytic segment is by no means sufficient for an active enzyme. In addition to the activation core (residues 40–300 of the PKA catalytic subunit), the PKA catalytic subunit is flanked by two dynamic tails: The N-terminal and the C-terminal tails. The C-

terminal tail wraps around both the C-lobe and the N-lobe and is a conserved feature of

9 all protein kinases that belong to the AGC subfamily. The C-tail functions as a cis- regulatory element and is an essential part of the active kinase [28].

1.2. Post-translational modifications in PKA-C:

Myristoylation: The N-terminus of PKA-C is modified by several post-translational modifications (PTMs), including myristoylation of Gly1, deamidation of Asn2, as well as phosphorylation of Ser10 (figure 1.4). These PTMs have been demonstrated experimentally to have functional consequences for both catalytic subunit function and conformation, as well as subcellular localization [29]. N-myristoylation is an acylation process absolutely specific to the N-terminal amino acid glycine in proteins. The cytosolic enzyme responsible for this activity, N-myristoyltransferase (NMT) (NMT, E.C 2.3.1.97) catalyzes the covalent linkage of a tetradecanoyl lipid group to the N-terminal Gly1 of the cAMP-dependent protein kinase A catalytic subunit (PKA-Cα). Studies have shown that upon elevation of cAMP in the cytosol, a significant percentage of PKA-C is released from regulatory subunits[30]. Liberated PKA-C becomes associated with the membrane via N- terminal myristoylation. This membrane association does not require the interaction between PKA-R and scaffold proteins called A-kinase anchoring proteins or AKAPs. It slows the mobility of PKA-C and enriches kinase activity on the membrane. Membrane- residing PKA substrates are preferentially phosphorylated compared to cytosolic substrates. NMR studies show that myristoylated PKA-C conformation is in equilibration between two major populations: a myr-in state in the absence of membrane, indication a tucked-in position of the myristoylated moiety in the hydrophobic pocket of PKA-C, and a myr-out state which represents a more mobile, unwounded N-terminus. In the presence of a membrane or lipid bilayer, myristoyl group is extruded and inserts into the hydrocarbon region of the lipid bilayer [31].

Phosphorylation: The protein kinases themselves are typically regulated by

phosphorylation and concurrent structural rearrangements. Accurate physiological

function requires posttranslational modification and deformable structures, as seen in a

prototypical example of cyclic AMP-dependent protein kinase (PKA). PKA is activated by

phosphorylation, is inhomogeneously phosphorylated when expressed in bacteria, and

exhibits a wide range of dynamic properties. PKA-C is also regulated by phosphorylation:

there are two phosphorylation sites on PKA catalytic subunit, one near the C-terminus

(Ser338) and the other one is in the activation loop (Thr197). Each phosphate is essential

for the active enzyme in mammalian cells [32]: Thr197, located in the activation loop near

the cleft interface, as in many other protein kinases, is thought to activate PKA by

anchoring the activation loop away from the active site and freeing access of substrate

and by contributing to the structuring of the catalytic residues. Ser338 is phosphorylated

co-translationally and is known to be essential for correct folding of the active enzyme as

it is being synthesized. Two additional phosphates are present that are associated with

expression in E. coli: Ser139 and Ser10, none of them have any apparent effect on activity

or solubility of the enzyme [33].

1.3. Kinetics and Catalysis mechanism in cAMP dependent PKA:

The reactions catalyzed by protein kinases require both ATP and a substrate

protein and, thus, can be viewed as bisubstrate kinetic mechanisms. PKA is a

Serine/Threonine kinase, and has shown to incorporate a random kinetic mechanism

using Kemptide as a substrate. However, PKA has a preference for binding ATP prior to

substrate based on thermodynamic grounds [34,35]. The kinetic properties of the enzyme

are highly sensitive to Mg2+ [34-36]. Protein kinases are capable of binding two metals, the dissociation constant of the second metal binding event (~2 mM) is about 2 orders of

magnitude higher than that for the first, so that this secondary site is only partially occupied

under physiological metal concentrations. Owing to its high concentration in the cell

compared to other divalent metal ions (0.5 mM), Mg2+ is considered the physiological

activator of protein kinases. Nonetheless, other divalent metal ions can replace Mg2+ in the active site and, in some cases, support catalysis in vitro [37]. A schematic representation of the reaction pathway for catalysis is shown in figure 1.6. The phosphoryl

transfer step is rapid (.500 sec-1) [35,36] and, this step requires the fully closed

conformation with the tip of the glycine-rich loop hydrogen bonded (Figure 1.5). In protein

kinases, the adenine ring is positioned at the base of the cleft in a deep hydrophobic

pocket between the two lobes. The γ-phosphate is then positioned at the edge of the active

site cleft between two critical elements : the glycine rich loop and the catalytic loop while

the Activation Loop positions the C Helix and the substrate docking surfaces and helps to

control opening and closing of the active site cleft. In this conformation, however, the

nucleotide product cannot be released. The slower, rate-determining step in vivo could be

the off-rate of the phosphorylated protein as well as the off-rate of ADP. This step requires

an opening of the glycine-rich loop and, most likely, at least a partial displacement of the

carboxyl terminal tail. Kinases in general are not efficient catalyst and their substrates

should be tethered close by, and not limited by diffusion as in most other metabolic

enzymes.

1.4. Substrate recognition motif:

Protein kinase A is a dual-specific protein kinase, phosphorylating both serine and

threonine side chains. A shared consensus sequence of Arg-Arg-X-Ser/Thr-Y or Arg-X-

X-Arg-X-X-Ser/Thr-Y is common in most PKA substrates, where Y is a large hydrophobic residue and X is any residue. Protein kinases rely partly on local residues for high affinity

and in general, they phosphorylate peptide regions based on the residues immediately

flanking the site of phosphorylation. This site is called the P-site, and residues N-terminal

to this site are sequentially numbered P-1, P-2, P-3, etc. Residues C-terminal to the P-site

are termed P+1, P+2, P+3, etc. Amino acids that are approximately four residues or less

from the phosphorylation site in either the N or C-terminal directions are the most significant for directing phosphoryl transfer. A standard substrate peptide for PKA, based on this consensus sequence, is Kemptide (LRRASLG). In addition to local sequence

elements, there are also distal recognition elements that are very important in binding

affinity of substrates. The natural substrates are likely to utilize additional binding

determinants not present in the immediate environment of the active site.

1.5. Regulators and inhibitors of PKA in cell:

PKA is distinguished from all other cAMP receptors for its ubiquitous distribution in eukaryotic organisms and its pleiotropic effects. PKA primarily exists as an inactive

tetrameric holoenzyme, constituting of two regulatory (R) and two catalytic (C) subunits

(figure 1.7). it is the tetrameric holoenzyme complex that reflects the basal physiological

state of the enzyme. In all mammals, there are four functionally non-redundant R-subunits

in form of four isoforms: RIα, RIβ, RIIα and RIIβ, all having a stable dimerization domain at the N-terminus. This domain is followed by a flexible linker that is classified as an

intrinsically disordered region. Embedded within the linker is an inhibitor site that docks to

the active site cleft in the holoenzyme. The very stable helical dimerization domain is multi- functional. In addition to creating a stable dimer, it serves as a docking site for A-kinase anchoring proteins or AKAPs (Figure 1.9). AKAPs are polyvalent scaffold proteins that are characterized by an amphipathic helix that binds with high affinity to the dimerization/docking (D/D) domain of the R-subunits [38]. RII subunits are considered both substrates and inhibitors, where RI and PKI (heat stable protein kinase A inhibitor) are considered pseudosubstrate because the autophosphorylation site is replaced by an

Alanine in the place of a phosphorylatable Serine (Figure 1.8). Unlike R subunits, PKI inhibition of the C subunit is not relieved by cAMP.

1.6. Compartmentalization of PKA and cAMP signaling:

The subcellular localization and trafficking of PKA is critical in signaling specificity

and efficiency, and several competing mechanisms have been described that may

modulate this process. The localization of holoenzyme is most often suggested to be a

mechanism to bring PKA closer to selective substrates, increasing substrate specificity as

well as the rate of phosphorylation. localizing mechanisms could also serve to place PKA

close to sites of cAMP generation and allow regional changes in cAMP to promote PKA

activation. AKAPs are anchoring proteins that bind with nanomolar affinities to the type II

regulatory subunits (RΙΙ), and these AKAPs are themselves associated with cytoplasmic

membranes, cytoskeletal components, or cytoplasmic organelles [39,40]. several novel

AKAPs have been isolated that can also bind to type I regulatory subunits although with

much lower(micromolar) [41]. Myristoylation of the amino terminal of Cα and Cβ of PKA

also play a role in membrane association and thus PKA-C localization (explained in post

translational modifications section of this chapter). In addition to these mechanisms, the

heat-stable inhibitor of PKA, PKI, has also been shown to facilitate nuclear export of C

subunits back to the cytoplasm, and this may help to terminate the transcriptional

response. C subunit localizes to both cytoplasmic and nuclear compartments and

promotes gene activation, however, while the C subunit is in the holoenzyme form no

nuclear activity has been detected. The R/C holoenzyme exceeds the size that is capable

of diffusing at significant rates through the nuclear pore complex (NPC). In addition,

cytoplasmic AKAPs are potentially capable of binding holoenzyme and preventing its

redistribution to the nucleus [42]. Upon dissociation of C subunit from the holoenzyme and

release of the active C subunit, this 40-kDa monomer is now presumed to be small enough

14 to diffuse into the nucleus where it can phosphorylate nuclear targets such as CREB and other transcription factors and modulate their activity.

1.7. PRKACA and Human disease

Cyclic adenosine monophosphate, identified in 1957 as the first intracellular second messenger of extracellular ligand action, is now established as a universal regulator of metabolism and gene expression in all life forms [43]. a number of human diseases that were linked to abnormal cAMP responses, such as irregular cell growth and proliferation, have already begun to be elucidated at the molecular level. PKA activation readily oversees the capacity of cells to either proliferate or differentiate, doing so by altering their genetic and epigenetic landscapes, and by remodeling their actin cytoskeletons.

Importantly, these cellular events enable PKA to govern a diverse array of physiological processes, including embryogenesis and development, cardiac and neuronal function, steroidogenesis, and immune homeostasis, as well as the responses of tissues to a host of hormones, neurotransmitters, and peptides [44]. A family of enzymes called adenylate cyclases (AC) catalyzes cAMP formation from ATP. In vertebrates, AC comprise both membrane-bound isoforms and soluble isoform [45]. In addition to AC expression and activity, cAMP homeostasis is regulated by a superfamily of phosphodiesterases (PDE) that degrade intracellular cyclic nucleotides, therefore an efficient cAMP signaling pathway is a result of an interplay between multiple enzymes as well as compartmentalized cAMP pools in the cell (Figure 1.10).

As a second messenger, cyclic AMP serves in multiple downstream pathways. Most prominent, it activates the cAMP-dependent protein kinase A (PKA). PKA, ubiquitous in mammalian cells, is a critical signaling node and regulates a plethora of biological processes. Cushing’s Disease and Carney Complex (CNC) Disease are caused by gain- of-function mutations in the catalytic and regulatory subunits, respectively [46,47]. In some instances, single mutations in either subunit lead to a phenotype where PKA activity is not tightly regulated. In these patients the basal PKA activity levels are high even in the

absence of cAMP. Perhaps the most striking PKA-linked disease mutation so far is a

fusion protein between DNAJB1 and the PKA Cα subunit, which was discovered to be a driver of a childhood liver cancer, fibrolamellar hepatocellular carcinoma (FLHCC) [48,49].

In other cases, inhibition of PKA can lead to an invasive phenotype suggesting that PKA can function as both a tumor driver and a tumor suppressor. Recently, a point mutation in the RIβ dimerization/docking domain was found in patients diagnosed with a neurodegenerative disease. This mutation disrupts dimerization, prevents AKAP binding and thus abolishes targeting even though the residual activity of the complex is robustly regulated by cAMP. The variety of mutations that lead to dysfunctional PKA signaling is consistent with its central role in regulating so many biological events. The consequences of dysfunctional PKA signaling will be both diverse and cell specific.

The first time that cAMP-dependent PKA itself was found to be directly involved in oncogenesis in an inherited disease was in the case of multiple endocrine neoplasia syndrome known as Carney complex (CNC) [50-52]. Studies have shown that non-

inherited mutations of the regulatory subunit RIα are associated with lupus [53].

Competitive PCR revealed a marked reduction of RIα and RIβ mRNA and protein levels

in these cells [53]. Role of cAMP signaling in cancer genesis and treatment has also been

established in several cases, wherein PKA influences malignant phenotypes in a

paradoxical manner. At present, the best defined contribution of PKA signaling to

neoplastic transformation is found in endocrine-associated tumors, including those arising

in the kidney, pituitary, thyroid, and testis, where elevated activation of PKA is highly

associated with tumor aggression [54]. Likewise, PKA activation has also been linked to

the induction of EMT programs due to its ability to promote cytoskeletal remodeling and

migratory behaviors in malignant cells (Figure 1.11) [55, 56]; it also serves as a critical

mediator of EMT programs activated by hypoxia [57] and as a potential driver of

chemoresistance in breast cancer cells [58].

Recent clinical studies, either measuring autoantibodies for PKA [59] or its enzymatic

activity [60] in serum patients, strongly suggest that PKA may function as a cancer marker

for various human cancers and can be used in cancer detection and for monitoring

response to therapy. Overall, based on the above considerations, PKA selective targeting in antitumor strategies has become very attractive and perused in various attempts.

Chapter II

Application of Solution NMR Spectroscopy in Biological systems Nuclear magnetic resonance (NMR) is, at the present time, an established method in a

variety of scientific fields such as physics, chemistry, biology, and medicine. However, it

took more than 60 years to reach this interdisciplinary status. In its early years, NMR was

a rather insensitive method: for instance, pure liquids were required to detect 13C NMR spectra. Stronger electromagnets were designed to reach 100 MHz for the 1H frequency

until the emergence of superconducting magnets in the early 1960s. The introduction of

an additional frequency axis led to correlation maps between spins (either via J-coupling

or NOEs) and to powerful tools for resonance assignment (2D NMR). Today, NMR is

unique in the versatility of the multidimensional experiments that can be implemented. 2D

NMR was quickly transferred to the field of biomolecules, and the credibility of NMR as

structural tool for proteins was strengthened over the years as its performance increased: 3D

NMR was introduced first on unlabeled proteins followed quickly by a new set of triple

resonance experiments using 15N and 13C labeled samples [61]. NMR spectrometers devoted

to structural biology benefit from several recent technological achievements: (1) higher

magnetic field (≥ 950 MHz) can be reached using new superconducting material, (2)

cryoprobes, in which the transmit/receive coils are maintained at low temperature to reduce

the noise, have become standard equipment, (3) the design of the spectrometer electronics

leads to superb experimental long-term stability, and (4) alternate processing methods are

possible with the increased power of computers. In recent years, biological NMR has evolved

toward more diverse applications.

the number of published structures solved by NMR has stagnated over the years in

comparison with the structures solved by X-ray diffraction. Solving a protein structure by

X-ray can be quite fast once suitable crystals have been obtained, however, NMR can

provide other types of information that is hardly amenable by crystallography: dynamics

can be investigated by NMR over a wide range of time scales [62], from slow exchange

where the two interconverting species are visible to fast motion using relaxation

measurements. In the field of drug discovery [63], chemical shift mapping provides

information on which part of the protein is interacting with the ligand and NMR is very

powerful at screening or optimizing hits. In conclusion, the ecological niche of NMR is

currently not restricted to protein structure determination but covers a wider range of

relevant information [64].

2.1. NMR Parameters

An NMR spectrum can only be observed for nuclei that possess a net spin. In this respect,

the most abundant nucleus in a protein, hydrogen, is well suited as its most abundant

isotope (1H) has spin ½. In contrast, carbon, nitrogen, and oxygen are not easily visible by

NMR, at least for their most abundant isotopes (12C, 14N, and 16O). However, “isotope- labeling” technique enriches the protein with isotopes such as 13C and 15N, making them

detectable by NMR. Although these strategies were very expensive two decades ago, uniform or selective labeling is now cost-effective.

NMR experiments are carried out in a static magnetic field B0 (several Tesla) aligned

conventionally along the +z axis. As a result of this field, the space is no longer isotropic

and all interactions experienced by the spins will depend on the orientation of the molecule

with respect to the magnetic field B0. In mathematical terms, the anisotropic NMR

interactions are described by second-rank tensors or 3 × 3 matrices. However, in liquid

state NMR, the molecule under investigation is rotating freely with a correlation time τc (1–

50 ns) much smaller than the acquisition time: if this rotation is isotropic, all interactions

will average out and only the isotropic component will be observed. This explains the

20 sharpness of resonance typically seen in solution NMR spectra as compared with solid- state spectra.

2.2. Chemical Shift

The atomic-resolution power of NMR is intrinsically linked to the occurrence of chemical shift. In an NMR spectrum, the magnitude or intensity of the resonance is displayed along a single frequency axis (in the case of 1D NMR) or several axes (for multidimensional

NMR). Chemical shift is usually expressed not in Hz but in ppm relative to a standard:

0 ( ) = 10 . 0 6 𝑣𝑣 − 𝑣𝑣 𝛿𝛿 𝑝𝑝𝑝𝑝𝑝𝑝 𝑣𝑣 where ν is the signal frequency in Hz and ν0 that of a reference compound so that chemical shifts in ppm can be compared between data sets recorded at different field strength.

Several calibration standards are available (e.g. tetramethylsilane or TMS), however, to avoid any additional compound that might interfere with the protein, most spectroscopists use, as a calibration intermediate, the water line although its position is temperature- and pH-dependent.

Measuring chemical shift value is the most amenable task of NMR spectroscopy. The wealth of information provided by chemical shift data depends on the availability of the individual resonance assignments. If the chemical shifts of compound A change when compound B is added to the sample, we already know that A and B are interacting. If the resonances of A have been assigned (see below), then these changes can be interpreted at the atomic level. Through such an experiment applied to a protein-ligand interaction

[65], we can learn what parts of the small molecule are interacting and to which part of the macromolecular target the small molecule is bound.

The external magnetic field B0 induces currents in the electronic clouds in the protein; in turn, these circulating currents generate a local induced field Bind. As a result, the different spins sense the vector sum of the two fields. The magnetic field “sensed” by the nucleus

is generally different from the applied field B0: this additional contribution (or screening)

arises from the interaction of the surrounding electrons with the applied field. The electron

density around each nucleus varies according to its chemical properties (nature, bonds,

etc.).

= +

𝐵𝐵�⃗𝑙𝑙𝑙𝑙𝑙𝑙 𝐵𝐵�⃗0 𝐵𝐵�⃗𝑖𝑖𝑖𝑖𝑖𝑖

A proton located in the plane of the aromatic ring (Ha) experiences a stronger field whereas another facing the ring (Hb) perceives a weaker field. Ha is said to be downfield shifted whereas Hb is upfield shifted. Bloc: local magnetic field, B0: applied magnetic field, Bind: induced field by other nuclei.

2.3. Two and Multi-dimensional NMR

Any structural investigation starts with the recording of a standard one-dimensional NMR

spectrum. This spectrum bears resemblance with spectra obtained with any optical

spectroscopy (infrared, visible, ultraviolet): absorption is plotted as function of a frequency

or wavelength. For each nucleus, the NMR spectrum displays a signal at a given

resonance frequency; but the wealth of information that can be obtained by NMR relies on

multidimensional NMR. Since 1970 has been widely used to correlate the resonance

frequencies of several nuclei. In contrast to optical spectroscopy, the information content of a

NMR frequency is rather low whereas a correlation experiment mediated by an interaction (J-

22 coupling or NOE) provides the nature of the partners as well as the interaction strength. During the preparation, the spins are allowed to recover from the previous experiment, they evolve then during a variable evolution delay (t1). The key step in a 2D NMR experiment is the mixing, which allows the magnetizations (or the coherences in NMR jargon) to exchange (A B) through any interaction (J-coupling or NOE). Finally, the signal is sampled during the detection⇄ period (t2). Although two frequency labeling periods are present, the signal is indirectly detected during t1, because of the “memory” of the spins.

As a matter of fact, as long as the delays are not longer than the corresponding relaxation times, the spins remember their previous evolution: the signal detected at the very end of the pulse sequence (during t2) is modulated either in amplitude or in phase as a function of t1. The resulting data set will be a (n × m) matrix of points, corresponding to n time increments along t1 and m increments along t2. After applying a 2D Fourier transform to the time domain data, a two-dimensional NMR spectrum is obtained.

1H-15N heteronuclear single quantum correlation spectrum (HSQC) of B. subtilis l,d-transpeptidase LdtBs (169 residues, concentration 0.7 mm], pH 6.5) recorded at 600 MHz (image adapted from Bodenhausen G. et al, Chem Phys Lett 69, 185–189). The horizontal axis corresponds to the 1H frequency and the vertical one to the 15N frequency. A correlation peak is observed for each backbone 1H-15N pair. 23

his generic 2D NMR scheme can be used to generate a so-called homonuclear spectrum

(it correlates 1H with 1H frequencies) or more generally heteronuclear ones (1H - 15N or 1H

- 13C). pulse sequences are identified by their acronym: COSY [66] or TOCSY [67] for J- coupling 1H — 1H correlation, NOESY [68] for NOE 1H — 1H correlation, HSQC [69] for

1H - X correlation via the JHX coupling. An HSQC spectrum is nowadays often the very

first NMR spectrum recorded on a new protein under investigation: the chemical shift

dispersion is a reliable proof of the compactness of the protein whereas the line-width of

each signal provides information on the aggregation state of the protein [70]. An HSQC is

a very robust experiment that requires only a on a small amount of 15N-labeled material

(less than 2 mg for a 20 kDa protein) and only half an hour of spectrometer time.

The modular design of 2D NMR can be easily extended to 3D and even 4D NMR. A 3D

NMR experiment is described as:

Preparation – Evolution (t1) – Mixing #1–Evolution(t2) – Mixing#2-Detection (t3)

The resulting spectrum is a three-dimensional spectrum, with three frequency axes (F1,

F2, and F3), which correlates three different nuclei. The overall sensitivity of a 3D NMR

relies on the efficiency of the individual transfers and the most sensitive experiments uses

exclusively large 1J couplings. This observation has led to the design of the triple

resonance experiments that will be discussed in the next section.

2.4. NMR Resonance Assignment

NMR resonance assignment is a prerequisite for studies where one aims at deriving

information at the atomic level. Although changes in the spectrum can be monitored even

without assignment, the wealth of information is greatly enhanced for assigned signals.

As mentioned previously, the unique value of NMR lies in the fact that distinct signals can be resolved even for chemically identical groups that are located in different environments

in a protein. Consequently, for well-resolved spectra (narrow lines and optimal digital

resolution), one expects to discern one signal for each active spin (1H, 13C, or 15N). Before any data can be obtained from spectral parameters, the resonances should be assigned, i.e. a one-to-one correspondence between a nucleus in the molecule and a resonance in

the spectrum should be established.

Strategies for sequential resonance assignment in proteins.

As proteins are linear copolymers, these strategies aim at correlating the resonance

frequencies of one residue with those of the following one. 2D-1H-1H correlation experiments can be employed on unlabeled proteins and 3D triple resonance experiments are better suited for 15N-13C uniformly labeled molecules. The first approach uses two

correlations experiments: one using 1H-1H J-couplings (COSY or TOCSY) and one

25 using 1H-1H NOE (NOESY). One has to resort to NOE to link two adjacent residues as

3 no JHH coupling is available for this purpose: note that NOE is a through-space effect and that nuclei close in space but not belonging to an adjacent residue may lead to fallacious correlations. This issue is resolved in the 3D triple resonance strategy, which relies exclusively on J-couplings. Triple resonance experiments always works in pairs as illustrated here: HN(CO)CA that correlates the HN and the N of residue (i+1) with the Cα of residue (i) and HNCA that correlates the HN and the N of residue (i+1) with both the Cα of residue (i) and that of residue (i+1). In the HN(CO)CA, the carbonyl 13C act as a relay but its frequency is not detected. The experimental combination links each (HN, N) pairs with the preceding and following Cα's and reciprocally each Cα's with the adjacent (HN, N) pairs. Triple resonance experiments establish connectivities between adjacent residues using 1J and 2J couplings: experiments are always used in pairs (HNCO and HN(CA)CO,

HNCA and HN(CO)CA …) to connect residue (i) with residue (i+1). The acronyms used refer to the correlated nuclei and a nucleus denoted with parentheses is used as a relay but not identified. With this set of triple resonance experiments, the backbone resonance

(usually up to the Cβ) can be assigned. Extending the assignment to the entire side-chain is a more tedious and time-consuming task. In a number of cases, where the X-ray structure is already available, a NMR study is initiated not to confirm the conformation but to answer questions that could not be addressed by other means. With only the backbone assignment, valuable information can already be obtained: the location of the secondary structure elements (α-helices and β-sheets), the flexibility of the backbone over several time scales (ns, ms …), the affinity and binding site of a ligand. To identify the side-chain resonance, a combination of several 3D experiments are employed, some based on J- coupling transfer (HCCH-TOCSY experiments) [71], some based on NOE effects (13C edited NOESY). Because spectral overlap is more severe for backbone nuclei, side-chains

are generally less completely assigned, an issue that can impact on the precision of the

derived structures.

2.5. Detecting Molecular Interactions by NMR

Protein–protein interactions play a key role in numerous cellular processes. Even when

two partners have been structurally characterized, cocrystallization of the complex may

be difficult because of the low affinity or some local disorder. NMR can complement these

studies primarily for weak interactions (Kd > 100 μm). There are various NMR tools for

complex studies: chemical shift perturbation, paramagnetic relaxation enhancement,

intermolecular NOE, H/D exchange rates and residual dipolar couplings (see figure

below). A prerequisite is the resonance assignment of one of the partners, at least for the

backbone. The proteins are expressed and purified separately and, by selective isotopic

labeling methods (13C versus 12C or 15N versus 14N), the spectrum of either partner can

be hidden.

NMR detection of molecular interactions. Protein interaction/titration using a ligand. Image adapted from ref [72].

A protein (P) is interacting with a ligand (L) (another protein, a small molecule, a RNA

fragment …) with an association constant Ka (inset A in the figure above). This thermodynamic constant is related to two kinetic constants, the on-rate constant kon and

the off-rate constant koff. Ka = kon/koff = [PL]/[P].[L]. The affinity between the two molecules and structural information on the complex can be obtained using several NMR parameters.

An unlabeled binding partner is added to a 15N labeled protein: this causes changes in the

environment of the protein and therefore the chemical shifts of nuclei at the binding

interface (inset B). A paramagnetic tag is added to the ligand (inset C): resonances in the

protein that are close to the tag are broadened (or even disappear) while others remains

unaffected. Paramagnetic relaxation enhancement (PRE) provides structural restraints up

to 30 Å. Short internuclear distances can be detected between nuclei belonging to both

partners using intermolecular NOEs (inset D): discrimination between intra- and

intermolecular NOE can be simplified by specific isotope labeling of a single partner.

Residual dipolar couplings provide information of the orientation of internuclear vector with

respect to the molecular frame of alignment. If the protein and its ligand align in a different

manner when free or bound (inset E), RDCs provide powerful long-range restraints for

their relative orientation in the complex when bound.

Widely used, the chemical shift perturbation (CSP) method has been introduced in the

early 90's and is based on the observation of the spectrum of one molecule (a 15N-1H HSQC

spectrum for instance) with increasing concentration of the partner (titration). When the

complex is formed, the two molecules are in equilibrium between their free and bound states.

This equilibrium is described by the dissociation constant Kd. During the titration experiment,

three exchange regimes can be observed depending on the exchange rate of the complex

formation and the chemical shift difference between the free and bound states. In the slow

exchange regime, two sets of signals are detected for the two states and their integral can be

used to monitor their population. From a practical point of view, this regime is less convenient

than the fast one because the complexed signals have to be reassigned de novo. When the

exchange between the bound and free forms is fast as compared with the chemical shift

differences, the HSQC correlation peaks move in a continuous manner and no new resonance

assignment is needed. Analysis of the perturbation reveals which amino acids are located at

the interaction interface. It should be noted that some residues at the far end of the protein

may also be slightly perturbed if its global fold is affected. The intermediate regime gives rise

to detrimental peak broadening that may prevent the signal observation: one possible work-

around is a change in the experimental temperature.

Paramagnetic probes are unique tools for studying macromolecular complexes because

of their capability to provide long-range structural restraints (as far as 30 Å, a value to be compared with the 5–7 Å range of NOE). Paramagnetic tags (nitroxide radicals, Mn2+

chelates, or lanthanides) are introduced site-specifically in one of the partners. Two effects can be monitored on the NMR spectrum of the other protein: paramagnetic relaxation enhancements (PRE) are detected as resonance line-broadening whereas pseudocontact shifts (PCS) alter the resonance frequency. The PRE effects are predominantly caused by two additional relaxation mechanisms (electron-nucleus dipolar or Curie-spin relaxation), which share the same dependence on the distance from the paramagnetic center (1/r6). The paramagnetic species should be chosen with care to induce moderate broadening without washing out too many signals. Some nuclei such as Gd3+ give large

PRE but also no shifts, whereas others combine the two influences [73]. Most lanthanides are introduced using tags covalently bound to a thiol group in the protein, which should engineered by site-directed mutagenesis to contain a single Cys residue. The sketch in

Figure above highlights the critical choice of the tag position: it should not interfere with the interaction interface but should be close enough to the partner protein. Short distances are detected by means of NOE inside proteins and this is also applicable to protein complexes. The separate expression and purification of the interaction partner offer the

29 unique opportunity of using isotope labeling to discriminate between intra- and intermolecular contacts. This approach is best suited for high-affinity complexes (Kd <

50 nm) where the two proteins remain in contact for a longer period of time. For lower∼ affinity, less internuclear nOes can be detected because of line broadening. The CSP,

PRE, and intermolecular nOe data can be used to map the interaction interface and model the complex starting from the known structure of the individual partners obtained independently by X-ray diffraction or NMR. Once the various evidences of the intermolecular interaction have been collected, they can be entered into docking software that will predict the preferred orientation of one molecule to the other.

2.6. Protein Dynamics by NMR

The biological function of a protein is intricately linked to both its structure and dynamics.

Time-dependent fluctuations in the conformation occur during enzymatic activities, protein folding, regulation etc. NMR is sensitive to fluctuations in many distinct time windows ranging from picoseconds to seconds and can generally be complemented (only for fast motions) by in silico protein dynamics simulations. The first repercussion of dynamics on an NMR spectrum is the resonance line width: a small molecule exhibits narrow lines whereas large proteins have broad signals. This line-broadening primarily arises from the slow tumbling of large proteins but is mitigated by protein flexibility: fast internal fluctuations (> ns) narrow the signals whereas slower motions (ms range) act in the opposite direction. Besides the line-width effect, an array of NMR experiments provides more detailed information on both the global and internal dynamic in proteins [66]. Another

NMR experiment becomes more adapted for slower processes (10 ms to 5 s): the exchange spectroscopy method or EXSY [75]. Although the pulse sequence is identical to NOESY, the spins but not only the magnetizations are transferred in the present case.

The process (A B) should be in the slow exchange regime, i.e. kex << νA–νB, with two

⇄ 30

distinct signals visible for A and B. The 2D cross-peak intensities (A→B and B→A) are

quantified for different values of the exchange time T, with an upper limit set to the T1

relaxation time of the resonances. For example, the interaction of the SH3 domain ( 60 aa) from the Fyn tyrosine kinase with proline-rich peptides was studied using EXSY ∼[76] dissociation rate constants (koff) as well as thermodynamic parameters were derived from the NMR data combined with isothermal titration calorimetry (ITC). If the exchange process is no longer in the slow exchange regime, the two lines start broadening and then merge. Note that the spectral appearance depends on the kinetic parameters and not directly on the binding affinity, though tighter binding yields generally longer-lived bound states. The broadening caused by microsecond to millisecond molecular motion can be

exploited in the CPMG relaxation dispersion experiment. By applying a variable number

of 180° refocusing pulses, the dephasing caused by the A to B magnetization jumps can

be suppressed: the transverse relaxation rate R2obs (which is the inverse of the line-

width) decreases with more 180° pulses. For a two-site exchange model (A B), the

dispersion profiles are governed by the rate of the exchange process, the chemical⇄ shift

difference and the relative population pA and pB. This experiment remains operative in

cases with strongly skewed populations (pA << pB), where the minor species is nearly

invisible in the NMR spectrum [77]. This methodology can detect low-populated excited

states associated with local unfolding events in proteins: the chemical shifts of the

intermediate state extracted indirectly from CPMG experiments allow its conformation to

be characterized. CPMG relaxation has been used to investigate mutated proteins, where

the replacement of a single residue slightly destabilizes the global fold [78] or to protein-

peptide complexes with a small ( 10%) mole fraction of bound peptide [79]. Excited

invisible states are present along ∼the folding/refolding pathways and in conformational

changes during catalysis or ligand binding. The rate at which the magnetizations relax to

equilibrium after excitation is governed by the global and internal dynamics of proteins:

Proteins and peptides undergo internal motions over a wide range of time scales from picoseconds to milliseconds and longer. Fast motions falling in the picosecond to nanosecond range generally arise from motions of individual bonds or small groups of atoms that are faster than the overall tumbling of the molecule. These fast motions contribute to NMR relaxation parameters such as T1, T2 and NOE measurements. Internal motions of larger groups of atoms, correlated motions of several residues, segmental motions, and the like usually occurs on the nanosecond to microsecond time scales, whereas conformational reorganization of the entire molecule, like folding-unfolding, occurs on the slower time scale of microseconds to milliseconds. These motions contribute primarily to transverse magnetization as chemical exchange.

the overall rotational diffusion occurs in the ns time scale and the internal fluctuations on

the picosecond time scale. Studies of motions of N-H bonds from amide groups provide

site-specific probes for all protein residues (except Pro). Two mechanisms contribute to

the relaxation in the 15N-1H pair, the 15N chemical shift anisotropy and the 15N-1H dipolar interaction discussed earlier. Three relaxation parameters for the 15N-1H pair (longitudinal

1 15 relaxation R1, transverse relaxation R2, and { H}- N NOE) are combined to obtain the

2 timescale (correlation time, τm) and the amplitude (order parameter, S ) of the internal motion. Such a separation of the internal and global motions is possible only if they are not correlated [80], an assumption supported by their frequency difference. in contrast to

chemical exchange, NMR relaxation cannot see internal fluctuations that are slower than

the global rotation of the protein. The backbone dynamics can be complemented by

2 relaxation studies using side-chain probes ( H relaxation in CD3 groups of aliphatic

32 sidechains) [81]. When proteins are comprised of multiple domains separated by flexible linkers, their location can be identified by 15N relaxation. Modern NMR spectroscopy offers a rich assortment of techniques to study protein dynamics. Some biologists have expressed criticism of NMR-derived protein dynamics because some time scales are either extremely fast or slow as compared with the biological process. One should concede that picosecond backbone fluctuations are loosely correlated with enzymatic reactions in the millisecond range and that real-time NMR can mainly visualize artificially slow biological events. With these limits in mind, NMR remains the only experimental technique to report protein dynamics at atomic resolution over such a wide range of time scale.

Chapter III

Conformational dynamics and allostery in the DNAJB1-PRKACA

chimeric transcript linked to fibrolamellar hepatocellular

carcinoma

Adak Karamafrooz,1† Yingjie Wang†, Simon M. Sandford,2 Susan S. Taylor,3 and Gianluigi

Veglia1,4*

1Department of Biochemistry, Molecular Biology, and Biophysics- University of Minnesota,

Minneapolis, MN 55455;

2Laboratory of Cellular Biophysics, Rockefeller University, 1230 York Avenue, New York,

NY 10065, USA.

3Department of Chemistry and Biochemistry and Pharmacology, University of California at San Diego, CA 92093

4Department of Chemistry– University of Minnesota, Minneapolis, MN 55455.

†These authors contributed equally to the work.

3.1. PREFACE

An aberrant mis-splicing of the DNAJB1 and PRCAKA genes give rise to a chimeric

construct where the N-terminal helix is fused to the J domain (J-PKA-C). This chimera has been found to be the main driver of a rare form of liver cancer (Fibrolamellar hepatocellular carcinoma or FLHCC) that targets young adults. In cellular cultures and standard kinetic assays, the isolated enzyme displays a catalytic efficiency similar to that of the wild-type

PKA-C. In an effort to understand the molecular etiology of disease, we used solution

NMR spectroscopy to analyze the internal network of interactions of this aberrant kinase as well as its interactions with protein kinase inhibitors, a mechanism of regulation for the kinase. While in the crystal structure the J-domain is well ordered and tucked onto the C-

lobe of the enzyme, the solution conformation is better represented by an ensemble of

structures with a quite dynamic J-domain. These motions affect the intra-molecular

allosteric network and determine significant changes in the allosteric cooperativity

between the nucleotide and the pseudo-substrate inhibitor PKI. The change in the

allosteric cooperativity may have repercussion on the chimeric kinase regulation and its

export from the nucleus to the cytoplasm. These results suggest that the chimera

dysfunction may be also linked to dysregulation of its phosphorylation activity in the

nucleus.

3.2. INTRODUCTION

Fibrolamellar hepatocellular carcinoma (FLHCC) is a rare and highly aggressive

form of liver cancer that affects predominantly young patients without underlying cirrhosis

or disease [82] . This tumor does not respond to chemotherapy [83, 84], therefore, surgical

resection remains the mainstay of therapy [85]. The molecular pathogenesis of FLHCC is

still unknown; however, a recurrent chimeric gene was recently identified in FLHCC

patients and proposed as the main driver for the progression of the disease [48, 49, 86,87].

The chimeric gene originates from a ~400-kilobase deletion on chromosome 19 that

causes an in-frame fusion of exon 1 from the DNAJB1 gene, encoding a member of the

heat shock 40 protein family, with exons 2-to-10 from the PRKACA gene, encoding for the

adenosine 3′,5′- monophosphate cAMP–dependent protein kinase A catalytic subunit

alpha (PKA-C) [49]. The kinase resulting from the chimeric transcript is fully functional

(PKA-CDNAJB1) and comprises 405 residues with 69 amino acids of DNAJB1 replacing the

first 14 residues of the kinase N-terminal A-helix [49].

In-vivo enzymatic assays [49] as well as in vitro studies using classical coupled

enzyme assays (using kemptide as substrate) showed that PKA-CDNAJB1 activity is almost

indistinguishable from wild-type PKA-C (PKA-CWT). Additionally, a recent X-ray

crystallography study revealed that the structure of the kinase catalytic core remains essentially unchanged upon fusion of DNAJB1 with PRCAKA and the structure of the ternary complex is virtually superimposable to that of PKA-CWT. In the X-ray structure, the

N-terminal helix (A-helix) of PKA-CDNAJB1 is longer than in the wild-type enzyme and

DNAJB1 appendix is tacked into the large lobe of the enzyme (Figure 3.1). Although the

X-ray structure revealed the atomic details of the chimera, it fails to explain its oncogenic function.

To further investigate the structural features of the chimera and elucidate the effects of the DNAJB1 appendix on its kinetics and substrate binding, we used multidimensional solution NMR spectroscopy in concert with phosphoryl transfer kinetic assays and thermodynamic binding experiments. NMR measurements revealed that in solution the kinase exists as an ensemble of conformations with a very dynamic DNAJB1 appendix [88] Chemical shift mapping reveal that the allosteric network of communication in the apo, intermediate, and closed forms of PKA-CDNAJB1 is affected by the presence of

the DNAJB1 appendix. Finally, binding studies using full-length and truncated versions of

protein kinase inhibitor peptide (PKI) suggest that substrate or regulatory binding may be

perturbed the DNAJB1 appendix in PKADNAJB1.

3.3. MATERIAL AND METHODS

Protein expression, purification, and NMR sample preparations.

PKADNAJB1: High-level expression of PKADNAJB1 in E. coli was achieved by the construction of pET28a (+) vector that contained the protein gene subcloned prior to a phage T7 RNA polymerase promoter. A TEV protease cleavage sequence was engineered between the histidine (His) tag sequence and the target gene. Uniformly 2H/15N-labeled PKADNAJB1 was expressed in the E. coli cell line BL21(DE3) by growing the bacteria in M9 medium

15 containing NH4Cl as the sole nitrogen source. Protein expression was induced by

addition of 0.4 mM IPTG and carried out overnight at 24°C before harvesting the cells.

PKADNAJB1 purification was carried out by affinity chromatography using TALON Metal

Affinity Resin (Clontech) followed by His-Tag removal using a modified TEV protease

protocol [93]. An additional purification step using FPLC with cation exchange column

(HiTrap® Q-SP, GE Healthcare Life Sciences) was used to separate the three isoforms of

PKADNAJB1 , corresponding to the three different phosphorylation states using a linear

gradient from buffer A [20 mM KH2PO4 (pH 6.5)] to 30% buffer B [20 mM KH2PO4 and

1.0M KCl (pH 6.5)] with a flow rate of 2.0 ml/min. The three isoforms displayed identical

kinetic parameters as assessed by coupled enzyme assays.

Dnajb1: The 69 amino acid sequence corresponding to the DNAJB1 heat shock protein

fragment DNAJB11-69 was cloned into the pET-28a(+) vector. An affinity His-tag was engineered after the sequence separated by a thrombin cleavage site. E. coli strain BL-

21 was used as a bacterial host for the overexpression. The purified fusion protein was

cleaved using thrombin protease, diluting 500 units of enzyme in 0.5 ml into cold PBS

buffer (140 mM NaCl; 2.7 mM KCl; 10 mM Na2HPO4; 1.8 mM KH2PO4; pH=7.3). The cleavage reaction was monitored by SDS-PAGE, and upon completion, the thrombin was inactivated by adding of 1mM PMSF.

The NMR samples of PKADNAJB1 complexes were prepared using 15N labeled PKADNAJB1

and in a 2:1 ratio with PKI5-24 and PKIFL, in the presence of 12 mM of ATPγN for formation

of the ternary complex. For the binary complex formation, 12 mM ATPγN was added to

DNAJB1 180 µM solution of PKA . The proteins were solubilized in 250 µl of 95% H2O and 5%

D2O buffer solution containing 20 mM potassium phosphate, 180 mM KCl, 10 mM MgCl2,

10mM DTT and protease inhibitor (pH = 6.5). Uniformly 2H,15N,13C labeled DNAJB11-69

was prepared in 250 µl of 95% H2O and 5% D2O in 20 mM potassium phosphate buffer with 180 mM KCl and 10 mM MgCl2,10mM DTT and protease inhibitor (pH 7.0) to a final

concentration of 200 µM.

NMR Spectroscopy

Backbone resonance assignments of PKA-CDNAJB1. The NMR experiments were

carried out on a 900 MHz Bruker spectrometer equipped with triple resonance cryogenic

probe. The temperature was held constant at 300 K. DNAJB11-69 amide fingerprint and side chains were assigned using a combination of 3D CBCA(CO)NH and HNCACB experiments. All NMR data were processed using the NMRPipe [94].

Analysis of Chemical Shift Perturbations (CSP). The chemical shift trajectories were

monitored using the CONCISE (COordinated ChemIcal Shifts bEhavior) method [89] to

measure the change in structural equilibrium associated with each PKADNAJB1 complex

(apo, ATPγN, ADP, ATPγN/PKI5–24 and ATPγN/PKIFL). Using principal components

analysis (PCA), the method identifies a set of residues whose chemical shifts respond

linearly to the open-to-closed conformational transition of the kinase. Each one of these

residues provides a measure of the equilibrium position for all the PKADNAJB1 complexes, scoring them along the first principal component. The equilibrium position for a given

PKADNAJB1 variant is given by the average of the PC scores over all linear residues. To

identify the largest group of residues that responded to ligand binding in a correlated

fashion, an adapted version of the chemical shift covariance analysis (CHESCA) [90] was

applied to the PCA projection of the chemical shifts. First, the correlation matrix between

all linear residues PC1 projections was constructed and used to build a dendrogram

through hierarchical clustering. A threshold of 0.98 level of correlation coefficient was

applied to form the subset to trace the equilibrium position for each state.

Statistical analysis of Chemical shifts (CHESCA):

To identify purely allosteric sites, it becomes essential to identify which residues within a

protein respond in a correlated manner to allosteric modulators. Chemical shifts serve as

excellent sensors of local chemical and structural changes, and are especially useful for

measuring populations of states. CHESCA analysis is based on two simple but essential

notions: subtle but functionally relevant structural changes are effectively probed by

accurately measured NMR chemical shift variations, and when a system is subject to a

set of perturbations, residues that belong to the same allosteric network exhibit a

concerted response to the perturbation set [90,91] Chemical shift correlation analysis

was performed on two inhibited forms of PKADNAJB1 as well as PKAWT (Figure 3.2).

Molecular Dynamics Simulations

Systems Setup. Parallel MD simulations were set up to compare the wild type PKA-CWT

and the PKADNAJB1. These simulations were repeated for the apo, the binary (ATP bound),

WT and the ternary complexes (ATP and PKI5-24). The systems with the PKA-C complexes were built directly from the X-ray structures starting from the close configurations (PDB

ID: 1ATP) [95]. The PKADNAJB1 chimera were built from their close conformation PDB ID:

4WB7 [96]. All simulations were performed using GROMACS 4.6 [97] in CHARMM36 [98]

force field. The all-atom structure were solvated in a rhombic dodecahedron solvent box

with a TIP3P [98] water molecule layer extended approximately 10 Å away from the surface of the proteins. Counter ions (K+ and Cl-) were added to ensure electrostatic neutrality corresponding to an ionic concentration of ~150 mM. The LINCS [99] algorithm

was applied to constrain all covalent H-bonds to the equilibrium length, and particle-mesh

Ewald7 was used to treat long-range electrostatic interactions with a real-space cutoff of

10 Å. All of the systems were minimized using steepest descent algorithm, and then were

gradually heated to 300 K at a constant volume over 1 ns, using harmonic restraints with

a force constant 1000 kJ/(mol*Å2) on heavy atoms of both proteins and nucleotides. Over the following 12 ns of simulations at constant pressure (1 atm) and temperature (300 K), the restraints were gradually released. The systems were equilibrated for an additional 20 ns without positional restraints. A Parrinello-Rahman barostat [101] was used to keep the

pressure constant, while a V-rescale thermostat with a time step of 2 fs was used to keep

the temperature constant. Each system was simulated for 1.05 µs, with snapshots

recorded every 20 ps. A total of 6.3 µs and 315000 conformations were utilized for the

analyses.

Mutual Information and Allosteric Network Analysis. To monitor the allosteric effects originating from the DNAJB1 appendix and the catalytic core of the PKADNAJB1 chimera,

GSATools [102,103] was used to compute mutual information of PKA-CWT and the

PKADNAJB1. GSATools is a set of GROMACS utilities that translate the ensemble of protein

conformations into alignment of structural strings using a String Alphabet algorithm. The

RMSF of local structures in the MD trajectories was computed and encoded into a set of aligned structural strings. Then correlations of local motions are computed as the mutual

information between the two columns of these structural strings and pruned at a cutoff of

0.20. These matrices of mutual information and their differences were further mapped onto

the crystal structure with Xpyder [104] plugin for PyMol. Graph analysis was applied to

detect hubs in the networks and key allosteric communication pathways in PKA-CWT and

PKADNAJB1.

3.4. Results

Enzymatic kinetics and binding thermodynamics.

Table 1 shows the results of coupled enzyme assay using kemptide as substrate. PKA-

CDNAJB1 is slightly less thermally stable compared to its wild type counterpart (table 2), but

the trend of stepwise increasing thermal stability upon adding nucleotide and inhibitor PKI

is still observed.

NMR spectroscopy

Assignment: The assignments of the fingerprint resonances of the catalytic core of the

PKA-CDNAJB1 were transferred from human PKA-CWT, which resonate to virtually identical

chemical shifts. To assign the resonances of the DNAJB1 appendix, we recombinantly

expressed the 69-residue construct (DNAJB11-69) in E. coli bacteria. As previously noted

[92], this domain folds independently from the entire chaperone and its resonances overlay with those in the full-length PKA-CDNAJB1 chimera (Figs. S1 and S2). Approximately

62 amide resonances were unequivocally assigned, with the exception of resonances in the A-helix. The latter residues are broadened beyond detection and probably undergoing conformational exchange in the µs-ms time scale.

To study the dynamic response of nucleotide and substrate binding of the chimera, we monitored the chemical shift changes associated with the major conformational states:

Apo, nucleotide-bound (ATPγN, known as binary form) and pseudo-substrate bound PKI5-

24. In addition, to understand the effects of the regions peripheral to the PKI recognition sequence, we also studied the binding with the full-length PKI. As for PKA-CWT, the

chemical shift perturbations (CSPs) associated with amide groups follow linear trajectories

along the coordinates of the open to closed states (Figure 3.2). Using CONCISE

(COordiNated ChemIcal Shifts bEhavior), we calculated the distribution of the amide resonances along the coordinates of the open-to-closed states. Specifically, CONCISE

enables one to quantitatively measure the population shifts associated with ligand

titrations and estimate the degree of collectiveness of the protein residues’ response to

ligand binding [89]. To improve the statistical significance of the degree of linearity, we

included data from both ATPγN and ADP binding (Figure 3.2). Plots of the probability

density distributions versus the principal component 1 (PC1) of the chemical shift changes

analyzed using CHESCA shows the positions of the four states correspond to equilibrium

positions, defining the collective response of the residues to ligand binding. The

distributions of the populations of the resonances in the different ligated states is

essentially similar to that of the PKA-CWT, indicating the that PRKACA portion undergoes similar conformational transitions. In the CONCISE analysis, we found that for most of the resonances the linear trend between open and closed states is disrupted upon binding to the full-length peptide inhibitor (PKI-FL) (Figure 3.3, A). When we included the PKA-

Dnajb1 C /ATPγN/PKIFL to the CONCISE analysis, we found that the equilibrium shifts towards a slightly more open state and the distribution of the resonances is broader, indicating that the resonances do not reach collectively a defined state (Figure 3.2).

To trace the correlated chemical shifts changes in more detail, we utilized CHESCA

(CHEmical Shift Covariance Analysis). According to the CHESCA results, correlation is substantially reduced in the presence of the bound form of chimera to the PKI-FL. What is more obvious in comparing the correlation matrix in three major forms in PKADNAJB1, is

that correlation specifically in the appendix part and majority of the enzyme core is

disrupted upon complex formation with PKI-FL (Figure 3.3, A). This effect is observed in

PKAWT, however, with considerably less pronounced changes (Figure 3.3, B).

MD simulations suggest that the DNAJB1 appendix may induce the allosteric

changes at the catalytic core. To understand the structural and allosteric effects of

DNAJB1 appendix in Chimera, we first analyzed the allosteric networks in WT and

Chimera, both in the apo state. Specifically, we calculated the mutual information matrices

of Ca atoms, and then performed graph analysis on the MI matrix to map the correlation

network and key hubs. In WT, the correlation is mainly between the N-terminus and the

A-helix, and little correlation exists inside the catalytic lobes (Figurer 3.4 A,B). Whereas in

Chimera, allostery is dramatically enhanced: not only the DNAJB1 appendix shows significant dynamical correlation with the catalytic lobe and the C-terminal, but also the correlation inside the catalytic lobes is much stronger compared to WT (Figure 3.4 C); specifically, the C terminal tail, Gly-rich loop and the activation loop becomes central hubs in the allosteric network (Figure 3.4 D, Figure 3.5).

Next, we investigated the change of allostery upon binding nucleotides and substrates. In

WT, the correlation increased in the binary form bound with ATP (Fig 3.3 A, Fig S3A), and get even stronger throughout the protein in the ternary form (Fig S3B). In contrast, the allostery of Chimera showed minimal changes in the catalytic core upon binding nucleotides, whereas the allosteric hubs in the DnaJB1 appendix shift from the tip to the hinge of the appendix (Fig 3.3B, Fig S4B). To elucidate the detailed changes at the kinase domain that contributes to the catalytic function, we further mapped the increase of mutual information upon binding nucleotides onto the structure and compared the emergence of allosteric hubs. In WT, the emerging hubs exist mainly at the catalytic core (Fig 3.4A), and the enhanced correlation throughout the catalytic core is consistent with the positive

cooperativity found in WT, and it also agrees with the transition from the dynamically

uncommitted state to the dynamically committed state. In contrast, in Chimera minimal

changes occur in the catalytic core (Fig 3.4B) upon binding nucleotides. The similar

allostery of Chimera in the catalytic core in all three forms agrees with the decrease of

positive cooperativity in Chimera, indicating that the apo form already assumed the

dynamically committed state. Taken together, the DNAJB1 appendix diminishes the

population of the dynamically uncommitted state and enhances the inherent allostery in

the apo form, and this might explain the increased activity of the oncogenic mutant.

3.5. Discussion

Our structural and dynamic studies show that chimera adopts a broad ensemble of

conformations in solutions, with the N-terminal DNAJB1 appendix undergoing faster

reorientation dynamics than the core of the enzyme. The latter explains the ability to form

R/C complex, as the dynamics of the DNAJB1 appendix can affect the binding of Chimera

to the regulatory subunit. In fact, the higher B-factors for the DNAJB1[105] implied the presence of conformational dynamics in this domain. The latter is supported by the SAXS and NMR studies that reveal the interconversion between a ground state, where the

DNAJB1 appendix is tucked onto the large lobe, and an extended conformation, where the DNAJB1 appendix is dislodged from the core of the enzyme and freely diffusing in the bulk solvent.

3.6. Conclusions

From our studies it emerges that allosteric cooperativity of PKA-C is a central mechanism

to its function and dysfunction. NMR, kinetic, and thermodynamic studies revealed that

the motions of the J-domain appendix fused to the A-helix of PKA-C affects the allosteric

network of interactions in the chimeric kinase disrupting the allosteric cooperativity

44 between nucleotide binding and pseudo-substrate. The latter may have repercussion on the phosphorylation of nuclear substrates as well as the exporting mechanism from the nucleus to the cytoplasm. Based on these results, we anticipate that both regulation mechanism and substrate recognition will be affected as well and the evolution of the disease over time can be a multifactorial event.

Chapter IV

Conformational Landscape of the PRKACA-DNAJB1 Chimeric Kinase, the Driver for Fibrolamellar Hepatocellular Carcinoma

Michael D. Tomasini1, Yingjie Wang2,3, Adak Karamafrooz3, Geoffrey Li2, Thijs Beuming5,

Jiali Gao2,6, Susan S. Taylor 4,7, Gianluigi Veglia2,3 & Sanford M. Simon

1Laboratory of Cellular Biophysics, The Rockefeller University, 1230 York Avenue, New York, NY, 10065, USA. 2Department of Chemistry, University of Minnesota, Minneapolis, MN, 55455, USA. 3Department of Biochemistry, Molecular Biology, and Biophysics. University of Minnesota, Minneapolis, MN, 55455, USA. 4Department of Pharmacology, University of California, San Diego, CA, 92093, USA. 5Schrödinger Inc., 120 West 45th Street, New York, NY, 10036, USA. 6Theoretical Chemistry Institute, Jilin University, Changchun, Jilin Province, 130028, People’s Republic of China. 7Department of Chemistry and Biochemistry, University of California, San Diego, CA, 92093, USA.

Reprinted with permission from Nature Scientific Reports

4.1. Introduction: Fibrolamellar Hepatocellular Carcinoma (FLC) is a rare liver

cancer usually detected in adolescents and young adults [106]. It does not respond well to chemotherapy, and surgical resection is the primary means of treatment [107,108]. In

FLC tumors there is a single, consistent genetic deletion in one copy of chromosome 19

[109,110]. This results in the formation of a chimeric gene, DNAJB1-PRKACA, which

combines the first exon of DNAJB1, the heat-shock protein 40 with exons 2 through 10 of

PRKACA, a catalytic subunit of protein kinase A [109]. There are 3500 statistically

significant changes in the transcriptome and proteome when comparing the FLC to the

adjacent normal tissue [111]. However, this chimera is the only detected structural

alteration of the genome [110]. Expression of the chimera in mouse liver either by

recreating the deletion by CRISPR/Cas [114], or expression, in trans, by a transposon is

sufficient to produce FLC[112]. Thus, it is considered to be the driver for the disease. The chimeric protein encoded by DNAJB1-PRKACA comprises of the J-domain of DnaJB1

(the amino-terminal 69 residues), fused to the carboxyl-terminal residues of the PKAcα, the protein encoded by PRKACA [109].The chimeric protein, J-PKAcα, is enzymatically active. Protein Kinase A (PKA) exists as a heterotetramer comprised of two catalytic (C) subunits, either PRKACA (encoding PKAcα), PRKACB (PKAcβ), or PRKACG (PKAcγ) and a regulatory (R) subunit dimer which comes as either RI (PRKAR1A (RIα), PRKAR1B

(RIβ)) or RII (PRKAR2A (RIIα), PRKAR2B (RIIβ)) variants. In the holoenzyme, the R- subunits inhibit catalytic activity. A dimerization domain at the N-terminus of each R- subunit is joined by a flexible linker to two tandem cyclic nucleotide binding domains. An inhibitory peptide embedded within the linker of each R-subunit binds in the active site of the C-subunits. When the second messenger 3′,5′-cyclic adenosine monophosphate

(cAMP) binds the R-subunits, the C-subunits are able to phosphorylate many downstream targets8 involved in a variety of cellular processes9. PKAcα is made up of a conserved catalytic core (residues 40–300) consisting of a β-strand rich small lobe and a primarily α-

helical large lobe [115,116]. The substrate binding site is mostly in the large lobe while

ATP binds in the cleft between the two lobes [117]. The catalytic core is flanked by a 50 amino acid C-terminal segment that wraps around both lobes of the core, while the small

lobe is preceded by an N-terminal α-helix termed the A-helix. In the chimeric protein, the

lower portion at the amino end of the A-helix is fused to the carboxyl end of the J-domain of DnaJB1 (Fig. 1). In wild-type PKAcα the first exon (residues 1–14) constitutes a myristoylation motif where a myristoyl group on the amino-terminal glycine extends into a hydrophobic cleft in the large lobe. This keeps the A-helix packed closer against the catalytic core of the enzyme compared to non-myristoylated wild-type PKAcα where the first 14 residues are quite flexible [118]. Myristoylation of wild-type PKAcα is thought to be associated with membrane binding of both RII PKA holoenzymes [119] and allosteric regulation of the ATP active site [120,121].

A recent crystal structure of the chimeric J-PKAcα protein with ATP and inhibitor peptide

(a portion of the cAMP-dependent kinase inhibitor PKI5-24) demonstrated that the architecture of the core is conserved compared to wild-type PKAcα [122]. The J-domain extends the A-helix and is tucked beneath the large lobe of PKAcα (Fig. 1).

The catalytic core of both PKAcα and J-PKAcα are virtually superimposable in the crystal structures. However, it is possible that the addition of the J-domain could alter conformational dynamics and thus influence substrate binding, enzymatic activity, or regulation [123]. Indeed, the observed B-factors, an indication of the fluctuation of the atomic coordinates relative to their average positions, in the J-domain of the chimera

crystal structure were larger than the rest of the protein indicating increased dynamics

[122]. It is unresolved how the physiological properties of the chimera cause FLC.

Transformation is not only the consequence of the increased transcription of the DNAJB1-

PRKACA as a result of expression from the DNAJB1 promoter since expression of the

DNAJB1-PRKACA is sufficient for transformation in mouse liver, but expression of the

PRKACA is not [112]. It could be the result of altered localization of the J-PKAcα fusion

protein or interaction with other proteins through the J-domain, effects of the J-domain on

the dynamics or specificity of the catalytic core, the absence of myristoylation which could

affect either localization or altered kinase dynamics, or the lack of proper regulation of

PKA activity by the R-subunits? Any or all of these could contribute to the transformed

phenotype. To start exploring these possibilities we performed atomistic molecular

dynamics (MD) simulations of both wild-type PKAcα and the J-PKAcα chimera in the ATP-

bound, ADP-bound, and Apo states as well as bound to both nucleotide and substrate

(ATP and the pseudosubstrate PKI5-24) to explore how the addition of the J-domain in the

FLC chimera affects the kinase conformational dynamics. We find the J-domain of J-

PKAcα samples a wide range of conformations in all substrate/nucleotide binding modes tested. Some conformations show similarity to the crystal structure with the J-domain tucked underneath the large lobe of the kinase, while others show extended conformations that deviated significantly from the crystal structure with the J-domain positioned away from the large lobe. These predictions were confirmed by NMR analysis of the chimeric protein. Structural modeling of different conformational states of the J-PKAcα chimera into a holoenzyme with two RIIβ subunits did not indicate any constraints on the movement of the J-domain.

4.2. Results

To explore the impact of replacing exon 1 of PRKACA with exon 1 of DNAJB1 on the conformational landscape of Protein Kinase A, we performed 1 μs MD simulations on wild-type PKAcα and the J-PKAcα chimera in the ATP-bound, ADP-bound, and Apo

states as well as the tertiary ATP and PKI5-24 bound states starting from their respective

crystal structure. Each simulation rapidly relaxed to a relatively stable state distinct from

the starting structure. We quantified the time required for relaxation to this state by

calculating the root-mean-squared-deviation (RMSD) of the backbone atoms and

comparing it to the average structure over the last 50 ns of simulation time (Supplementary

Fig. S1). The structures were aligned using the relatively immobile helices E and F

(residues 140 to 160 for helix E and 217 to 233 for helix F with residue numbering from

the native PKAcα). The large initial RMSD changes in the J-PKAcα chimera (i.e. 9 Å to

2.5 Å for ATP and PKI-bound) and the non-myristoylated wild-type PKAcα (i.e. 5 Å to 1.5

Å for ATP-bound) indicate a change in the protein conformation, and are primarily the

result of the motions occurring at their N-termini: the J-domain and the loop preceding the

A-helix for the chimera and wild-type respectively. This can be shown by computing the

RMSD of the backbone atoms beginning at residue 15 (70 in J-PKAcα chimera numbering)

which omits the N-terminal regions (Supplementary Fig. S1). Omitting these regions, the

RMSD values for all proteins change by less than 2 Å during the simulations. All of the simulations shifted away from the crystal structure to a new steady-state following

approximately 200 ns. Thus, all subsequent analysis focuses only on snapshots obtained

after 200 ns.

The degree of local mobility along the kinase was determined by calculating the root-

mean-square fluctuation (RMSF) of each residue averaged over the final 800 ns of the

simulation and is shown in Fig. 2. For native PKAcα (Fig. 2b), in agreement with previous

studies [118,124], we observed large fluctuations in catalytically important loops such as

the Gly-rich loop (residues 50 to 55) and portions of the activation loop (residues 192 to

199). Large fluctuations were also seen in the loop connecting the A-helix to the small lobe

(residues 32–37), Lys81 of the B-helix, the loop connecting helices H and I (residues 284

to 287), and residues in the C-terminal tail. Residues of the G-helix (residues 241–245)

and Arg133, which are involved in substrate positioning, showed high RMSF values in all

50 simulations except when bound to PKI. The decreased conformational flexibility of those two regions is likely the result of interactions that formed with PKI and stabilized those areas. In simulations of non-myristoylated PKAcα, the N-terminal residues (1–8) were highly dynamic (Supplementary Fig. S2). Overall, the local conformations follow the trend that the Apo state is more dynamic, the intermediate state (nucleotide bound) shows restricted dynamics especially around the nucleotide binding pocket, and the ternary complex is increasingly rigid. This is in agreement with previous simulations and NMR studies of PKAcα which showed the t degree of fluctuations in the Apo state which decreased as the protein subsequently bound nucleotide and substrate [118,125].

The chimera showed similar fluctuations to PKAcα in the loop following the A-helix, the

Gly-rich loop, the activation loop, the G-helix, and residues in the C-terminal tail (Fig. 2a).

Additionally, J-PKAcα demonstrated large fluctuations in the J-domain as a whole. A plot of the differential RMSF between the chimera and native protein highlights the difference in mobility (Fig. 2c). In the ATP- and ADP-bound states, native PKAcα showed increased fluctuations of Asn36 relative to the chimera, which was attenuated in the Apo and ATP-

PKI forms. The Apo state of PKAcα had a high amount of fluctuation in Arg133, which was the result of multiple breaking and reforming events of a hydrogen bond between Arg133 and Glu230 which did not occur in the other simulations. The motions of the loop connecting helices H and I are reduced in the chimera compared to PKAcα for the Apo and ADP-bound form, but were not observed in the ATP-bound chimera. This is likely due to the interaction of this loop with the N-terminus of the J-domain. Lys340 and Asn341 form intermittent hydrogen bonds with residues in the J-domain which were not observed in the ATP or ATP-PKI bound forms of the J-PKAcα chimera. The hydrogen bonds result in decreased conformational dynamics of the loop between helices H and I. Finally, there were greater fluctuations in the C-terminal tail of the chimera, especially in the ADP-bound

form in which Phe327, which forms the back of the nucleotide binding pocket, was

observed to move away from the core of the protein (Supplementary Fig. S3).

To characterize the motions of the J-domain we defined three vectors (Fig. 3a). We chose

the first vector

(ν1) along the A-helix and the second vector (ν2) at the N-terminal end of the A-helix, which extends into the J-domain. The bending motion of the A-helix is described by the angle between these two vectors (θ1). The third vector (ν3) points from the end of the extended

A-helix towards the end of the J-domain. The up/down motion on the J-domain relative to the core of the enzyme is described by the angle between ν2 and ν3 (θ2). Finally, we define a dihedral angle (θ3) given by the positions of the Cα atoms of residues Lys29 – Leu160

– Glu140 – Lys−19 that describes the shearing motion between the J-domain and the large lobe.

In the presence of the J-domain there is a kink in the A-helix centered near Arg10 (Arg65 in chimera numbering) that is most pronounced in the ATP-PKI bound form. The bending of this helix, given by θ1, is 161° in the crystal structure [122], and samples values as low as 77° in the simulation (Fig. 3b). The kink is less prominent in the ADP and Apo forms of the chimera, yet both forms still sample values of θ1 distinct from the crystal structure

(minimum values of θ1 = 133° and θ1 = 124° respectively). Likewise, for movement of the amino-terminus (θ2) the ATP-PKI bound chimera showed the greatest range of motion

sampling angles from 62° to 121° (Fig. 3c). The larger values of θ2 sampled by ATP-PKI bound chimera correspond to a downward movement of the J-domain away from the large lobe (movement away from the initial starting structure as indicated by the black x). Both movements may be related to the shearing motion (θ3) of the J-domain, which also shows

the greatest range of motion in the ATP-PKI bound simulation. During the initial 200 ns of

the simulation (lighter colors in Fig. 3b,c) the J-domain of ATP-PKI bound chimera

transitions from tucked under the large lobe, similar to what is observed in the crystal

structure, to a more extended conformation with the N-terminal portion of the J-domain

rotated away from the crystal structure conformation. The ATP, ADP, and Apo forms of

the chimera also show movement e J-domain away from its position in the crystal

structure, though to a lesser extent than ATP-PKI bound chimera.

The simulations are consistent with the substrate/nucleotide binding state having an

influence on the range of motions of the J-domain (Fig. 3). However, it is conceivable that the 1 μs simulation time was not enough to sample all of the possible conformations of a given binding state. This was tested by performing three more 1 μs simulations, two of the

ATP-bound and one ADP-bound chimera. The aggregate data for the J-domain motions

from all seven 1 μs simulations is shown in Fig. 4d,e. While the J-domain in the second 1

μs simulation of the ADP-bound chimera (CH ADP T2) sampled conformations similar to

the first simulation (CH ADP T1), the additional ATP-bound chimera simulations sampled

other conformational states. The second ATP-bound chimera simulation (CH ATP T2)

sampled conformations that much resembled the ATP-PKI bound chimera with J-domain

extended away from the large domain, as indicated by decreased values of θ1 and θ3

(minimum value of θ1 in CH ATP T1: 119° vs minimum value of θ1 in CH ATP T2: 93° and minimum value of θ3 in CH ATP T1: 123° vs minimum value of θ3 in CH ATP T2: 61°). The third ATP-bound chimera simulation (CH ATP T3) did not show as great a range of conformational space, but did sample larger values of θ2 indicating a downward shift of

the J-domain relative to the crystal structure.

As the conformational space of the J-domain in the ATP-bound chimera showed variations

between the three 1 μs simulations, we attempted to obtain converged results by carrying

out a series of ten shorter simulations starting from different initial configurations taken

from the three 1 μs simulations. Four conformations were taken from CH ATP T1 (at 200

ns, 466 ns, 733 ns, and 1000 ns), three conformations from CH ATP T2 (at 200 ns, 600

ns, and 1000 ns) and three conformations from CH ATP T3 (at 200 ns, 600 ns, and 1000

ns) and each were extended for 300 ns. The results showed a wide range of states that

the J-domain could sample in the ATP-bound chimera, especially in the rotation of the J-

domain indicated by θ3 values ranging from 24° to 180° (Supplementary Fig. S4).

To further probe the range of conformations that are accessible to the J-domain, and gain

insight into the relative populations of the conformations, we performed a clustering

analysis based on the RMSD distance of the backbone atoms in each chimera structure.

All chimera simulations were combined for a total aggregate time of 10 μs. Using an RMSD

cutoff of 5 Å, the chimera simulations separated into 21 clusters with the first four clusters

(Fig.4) accounting for 91% of the total conformations (Supplementary Fig. S5). Pooling

the wild type simulations and performing a cluster analysis with the same parameters

resulted in a total of 2 clusters with the vast majority (greater than 99%) populating the

first cluster indicating a larger conformational ensemble of structures in J-PKAcα

compared to the wild type protein. The first J-PKAcα cluster comprises the great majority

of conformations accounting for 62% of all structures. A variety of conformations for the J-

domain was observed in the top four clusters (Fig. 4). The first cluster shows similarities

to the crystal structure (compare to Fig. 1) with similar angles for the J-domain (Table 1).

Cluster 2 shows the kink near Arg10 in the A-helix resulting in a decreased value of θ1

compared to the crystal structure (119° vs 164°). This is also accompanied by a downward

motion (increase in θ2 compared to the crystal structure) and a rotation (decrease in θ3

compared to the crystal structure) of the J-domain. Cluster 3 shows the J-domain tucked

underneath the large lobe, similar to the crystal structure, but there is also a kink near

Arg10 similar to cluster 2. Finally, cluster 4 shows a rotation of the J-domain in the opposite

direction from the rotation in cluster 2.

We examined the frequency of transitions between the different clusters. At each time point we determined if the initial conformation transitioned to a different state

(Supplementary Fig. S6) or remained in the same state (Supplementary Fig. S6). The frequency of transitions is given by a color code. The time steps where the state did not change were much more frequent than transitions to different states. In general, the transitions between clusters are symmetrical, i.e. cluster 1 is most likely to transition to cluster 4 and cluster 4 is most likely to transition to cluster 1. A number of transitions were not preferred. For example, in the 10 μs of aggregate simulation time, there was only one transition between the two most populated clusters, cluster 1 and cluster 2. Both clusters, much more frequently transition to cluster 3, which then shows a moderate number of transitions back to both clusters 1 and 2.

All of these studies are consistent with the J-domain being freely mobile and able to explore a variety of conformations in the isolated chimera kinase. We wanted to examine if the same mobility of movement was possible for the chimeric kinase in the context of the holoenzyme. Using a crystal structure of PKAcα bound to the R-subunit(RIIβ)[126], the J-domain from the top two clusters of the cluster analysis were modeled into the holoenzyme by aligning the backbone atoms of helices E and F and with the wild type

PKAcα contained in the holoenzyme. In the holoenzyme the RIIβ subunits do not demonstrate any steric hindrance to the motions of the J-domain either when it is in a ‘J- in’ state (as in cluster 1) tucked underneath the core of the kinase or when in a ‘J-out’ state (as in cluster 2) when the domain is positioned far away from the catalytic core (Fig. 5). This would suggest that while in the holoenzyme the J-domain can be similarly mobile as in the isolated C-subunit. As a caveat, the first 103 N-terminal residues of RIIβ structure, containing the AKAP binding domain, were not resolved in the crystal

structure. Thus, it is possible that this N-terminal domain is located such that it would

interact with the J-domain and alter its conformational dynamics.

The MD simulation results indicate that the J-domain of the chimeric kinase is freely

mobile. To test experimentally the conformational dynamics of the J-domain, we used

NMR nuclear spin relaxation measurements. The resonance assignments for the catalytic

core of the J-PKAcα fingerprint were transferred from the previous chemical shift

assignments of the wild-type PKAcα amide resonances [127]. To assign the J-domain, we recombinantly synthesized and purified a DnaJB1 peptide from residue 1 to 76. The chemical shift assignments of the DnaJB1 peptide were then compared to the corresponding resonances of the J-domain in the full-length J-PKAcα. Ambiguous or overlapped peaks were assigned using the TROSY-version [128] of triple-resonance experiments [129] with U-2H,15N,13C labeled J-PKAcα. The direct correspondence between the resonances of the isolated constructs and the full-length J-PKAcα indicates that this domain folds independently [130]. To assess the global motions of J-PKAcα, we

15 measured both the N longitudinal (T1) and transverse (T2) relaxation times26

(Supplementary Fig. S7). From the analysis of the relaxation times of the core of the

enzyme (i.e., without the J-domain), the ternary form of the J-PKAcα chimera reorients

faster (22.9 ns) than the ternary form of wild-type PKAcα (24.0 ns). Given that the mass

of the J-PKAcα chimera is larger than that of the wild-type PKAcα, one might expect the

chimera to have a slower rotational correlation time in NMR. In fact, a theoretical

calculation using HYDRONMR27 estimates a correlation time of 29.4 ns for the wild-type

enzyme (PDB: 1ATP); while it predicts 37.4 ns for the J-PKAcα chimera (PDB: 4WB7).

Since HYDRONMR assumes a rigid body reorientation of the molecule, the experimental

observation of a faster rotational correlation time for the J-PKAcα chimera is consistent

with the J-domain having a high degree of flexibility, in agreement with the MD results.

Importantly, the plots of the T1/T2 versus residues reveal two distinct regions for J-PKAcα

(Fig.6): the core of the enzyme, with T1/T2 values averaging around 200, and the N- terminal region encompassing the J-domain with an average T1/ T2 of 55. These data suggest that the protein does not tumble as a rigid body; rather the catalytic core is disjointed from the J-domain, (residues −69 to −1) which undergoes faster global reorientation.

4.3. Discussion

There is compelling evidence that the J-PKAcα chimera is the main driver for the development of FLC4–6. There are a number of possible explanations for how expression of this chimera could lead to transformation. One possibility is that the DNAJB1 promoter, which drives the chimera, may be expressing more of the C-subunit, and this is sufficient for transformation. A second possibility is that the loss of the amino-terminal myristoylation motif, affects the catalytic rate or localization of PKAcα. A third possibility is that the presence of the J-domain at the amino terminus alters the subcellular localization or binding partners of the C-subunit and/or the PKA holoenzymes. A fourth possibility is that the addition of the J-domain at the amino terminus affects the dynamics and thereby potentially the specificity of the kinase for nucleotide or substrate. A final possibility is the

J-domain alters the finely tuned regulation of the PKA holoenzyme. We favor the hypothesis that the addition of the J-domain in some way alters the kinase dynamics, activity, regulation, or localization since the cancer has never been reported with just loss of the amino terminus, just loss of myristoylation on the second glycine, nor with the addition of any other amino terminus. Here, we have focused attention on the dynamic effects of adding the J-domain to the amino-terminal end of PKAcα, the C-subunit of protein kinase A.

In the crystal structure of the RIIβ wild-type holoenzyme the N-terminus of the C-subunit,

where the J-domain addition occurs, is positioned away from the major symmetrical

interface between the C- and R-subunits in the holoenzyme[126], and the presence of the

J-domain does not prevent formation of the holoenzymes. Thus, rather than affecting the

interactions with the regulatory subunits, it is possible that addition of the J-domain alters

the conformational landscape of the chimera and the resulting holoenzymes. The higher

B-factors in the J-domain [122] implied a large degree of conformational flexibility. Both

simulations and experiment have shown that point mutations in PKAcα far from the active

site can affect kinase activity allosterically though decoupling of dynamics[133,134],

implying that motions of the J-domain could influence the allosteric network in J-PKAcα

and thereby alter its activity. In this work, we performed several microseconds of MD

simulations exploring the conformational states adopted by the kinase and J-domains. We

found that a primary conformation, accounting for 62% of all structures sampled by

chimera MD trajectories, resembled the crystal structure where the J-domain was tucked

underneath the large lobe of the kinase. Nonetheless, we also found an ensemble of

chimera conformations in which the J-domain was dislodged from the core of the enzyme

allowing it to swing freely in solution. These states were not captured by the X-ray crystal

structure, but were confirmed using NMR. The flexible conformations appear to be

independent of the nucleotide/substrate binding mode as they were observed in ATP,

ADP, Apo, and ATP-PKI bound states. Focusing on the ATP-bound state, we showed that

1 μs of simulation time was not sufficient to obtained a converged ensemble of J-domain

configurations as different 1 μs simulations showed a wide variability in the J-domain

motions (Fig. 3). Because the conserved core of the wild-type enzyme and the chimeric

fusion show little structural differences and the canonical function is not affected by the J-

domain appendix, the presence of alternate conformations may constitute a way to target

the chimera selectively. This opens up the possibility to develop novel small molecule

inhibitors directed at the region of the fusion. However, as our results show, care must be

taken when designing inhibitors so a pass the wide range of motion of the J-domain, and

in particular, the A-helix where the fusion occurs.

4.4. Materials and Method

Molecular Dynamics simulations were performed to observe the structural dynamics of J-

PKAcα chimera and determine how they differ from wild-type PKAcα. Each protein (wild-

type and chimera) was simulated in the ATP-bound, ADP-bound, and Apo forms as well

as the tertiary state bound to both ATP and pseudosubstrate PKI5-24. Each simulation was built from the crystal structure, PDB ID: 4DFX15 in which the protein is in complex with the ATP mimic AMP-PNP, 2 Mg2+ ions and the 20 residue peptide SP20. SP20 corresponds to the endogenous PKA inhibitor PKI5-24 with two mutations (N20A and A21S) which convert PKI5-24 from inhibitor to substrate. The crystal structure of PKAcα also includes the myristoylation at the N-terminus as well as the mutation K7C. In all wild-type

PKAcα simulations the cysteine at position 7 was mutated back to a lysine using the

Schrodinger Maestro software suite (Maestro 10 (2016) Schrödinger, Inc., Portland, OR).

AMP-PNP was either mutated to ATP or ADP or removed for the Apo state. 2 Mg2+ ions

were retained for the ATP simulations, 1 for the ADP simulations, and removed for the

Apo simulations. SP20 was mutated back to PKI5-24 in simulations of the tertiary state.

The chimeric J-PKAcα systems were built using chain A of the crystal structure, PBD ID:

4WB7[122] in which the protein is bound to ATP, several Zn2+ ions, and PKI5-24. As with

the wild-type PKAcα simulations, J-PKAcα chimera was simulated in the ATP, ADP, and

Apo states as well as the tertiary state with ATP and PKI5-24 bound. Zn2+ ions associated with nucleotide were converted to Mg2+ ions, with other Zn2+ ions removed. All structures were phosphorylated at S139, T197 and S338 (amino acid numbers for the native PKAcα or S194, T252 and S393 in J-PKAcα chimera numbering). Structures were processed

59 using the Protein Preparation Wizard in Maestro (Maestro 10 (2016) Schrödinger, Inc.,

Portland, OR), solvated in a rectangular box with SPC waters, and sodium or chloride ions were added for electrostatic neutrality. Simulations were performed using the Desmond

MD Package[135] using the OPLS3 force field [136]. Each system was subject to energy minimization using the steepest decent method. An initial 100 ps Brownian Dynamics simulation at constant volume and a temperature of 10 K with heavy atoms constrained was performed. Subsequent equilibration included a 12 ps simulation at constant volume and at 10 K with heavy atoms restrained, followed by a 12 ps simulation at constant pressure with heavy atoms restrained, and finally a heating simulation in which the restraints were gradually tem heated to 300 K over 24 ps. Production simulations were performed for 1 μs or 300 ns with snapshots saved every 50 ps.

Clustering of Conformations. All simulations were aggregated to generate distinct clusters of conformational states. Helices E (Residues 140–160) and F (Residues 217–

233) where used to align all trajectories to the initial crystal structure. Analysis was performed on the protein backbone atoms first using the GROMACS analysis tool ‘gmx rms’ to compute the root-mean-square-deviation between all structures in the trajectory.

The GROMACS utility ‘gmx cluster’ was then used to cluster the structures using a cutoff of 5 Å and the gromos clustering methodology [137]. Representative structures were taken as the structure with the smallest average RMSD distance to every other member of the cluster.

Protein expression, purification, and NMR sample preparations.

High-level expression of J-PKAcα in E. coli was achieved by the construction of pET28a(+) vector that contained the protein gene subcloned prior to a phage T7 RNA polymerase promoter. A TEV protease cleavage sequence was engineered between the Histidine

(His) tag sequence and target gene. Uniformly 2H,15N-labeled J-PKAcα was expressed

using the E. coli BL-21(DE3) bacterial strain in M9 medium containing 15NH4Cl as the

sole nitrogen source and transferred to an 80% 2H containing M9 media. His-tagged J-

PKAcα was purified using TALON Metal Affinity Resin (Clontech). The His-tag was

removed using a modified TEV protease protocol [138] (Supplementary Fig. S8) followed

by an additional purification step using FPLC with an ion exchange column (HiTrap Q-SP).

The wild-type PKAcα was expressed in the same media and purified as described

previously34. The 69 amino acid sequence corresponding to the DNAJB1 fragment was

cloned into pET-28a(+) vector. An affinity His-tag was engineered after the sequence

separated by a thrombin cleavage site. E. coli strain BL-21 was used as a bacterial host

for the overexpression. The purified fusion protein was cleaved using thrombin protease,

diluting 500 units o ml into cold PBS buffer (140 mM NaCl; 2.7 mM KCl; 10 mM Na2HPO4;

1.8 mM KH2PO4; pH = 7.3). reaction was monitored by SDS-PAGE. Upon completion the

thrombin was inactivated by adding of 1 mM PMSF. The NMR samples of ternary

complexes were prepared using 2H,15N labeled J-PKAcα and wild-type PKAcα in a 2:1

ratio with PKI5-24 and in the presence of 12 mM of ATPγN. The proteins were solubilized in 250 μl of 95% H2O and 5% D2O buffer solution, containing 20 mM potassium phosphate,

180 mM KCl, 10 mM MgCl2 (pH = 6.5). The final sample concentrations were 180 μM and

250 μM for the chimera and the wild-type kinase complexes, respectively. Uniformly

2H,15N,13C labeled DnaJB1 was prepared in 250 μl of 95% H2O and 5% D2O in 20 mM

potassium phosphate buffer with 180 mM KCl and 10 mM MgCl2 (pH 7.0) to a final

concentration of 200 μM.

NMR Spectroscopy. NMR experiments were carried out on 850 and 900 MHz Bruker

spectrometers equipped with a cryogenic probe. The temperature was held constant at

300 K. Resonance assignments of the amide fingerprint of J-PKAcα was carried out by

transferring the previous assignments from wild-type PKAcα. The DnaJB1 amide

fingerprint and side chains were assigned using a combination of 3D CBCACONH35 and

HNCACB36 experiments. Longitudinal (T1) and transverse (T2) relaxation times were

evaluated by monitoring the signal intensity decay of the amide resonances in the series

of 2D spectra using the pulse sequences described by Zhu et al.[142]. The spectra were

recorded in an interleaved manner for both T1 and T2. The spectral dimensions were 3466

Hz (F1) and 14423 Hz (F2). For both T1 and T2, the relaxation decay was sampled for 8

different delays. For each T1 value, the order of the relaxation delay was randomized (T1

delays: 0, 10, 100, 300, 500, 700, 900, 1200 and 1700 ms) with 32 scans per FID. For T2, the relaxation delays were 8.1, 16.2, 24.3, 32.4, 40.5, 48.6, 56.7, and 64.8 ms with 96 scans per FID. Least-squares fitting of the decay curves was carried out with the downhill simplex algorithm implemented into SPARKY (Goddard and Kneller). Backbone assignments of J-PKAcα was achieved by overlay the resonances of the previously assigned wild-type PKAcα and transferring the resonance assignments of DnaJB1. All

NMR data were processed using the NMRPipe suite of programs [143].

Chapter V

Using an integrated Phosphoproteomics method for identifying Human Protein Kinase A (PRKACA) and its oncogenic mutant DNAJB1-PRKACA substrates

Karamafrooz A1, Blankenhorn J1, DD Thomas,1 Parker L1

1Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, Minneapolis, Minnesota, 55455, USA

5.1. Preface:

A segmental deletion resulting in DNAJB1-PRKACA gene fusion is now recognized as the signature genetic event of fibrolamellar hepatocellular carcinoma (FL-HCC), a rare but lethal liver cancer that primarily affects adolescents and young adults. This genetic deletion results in the fusion of the first exon of the heat shock protein 40, DNAJB1, which encodes the J domain, with exons 2–10 of the catalytic subunit of protein kinase

A, PRKACA, consequently producing an enzymatically active chimeric protein Dnajb1-

PRKACA. Here we implemented an integrated in-vivo and in vitro phosphoproteomic methods to identify direct substrates targeted by Human PRKACA and its oncogenic mutant Dnajb1-PRKACA in the cell. The unique in vitro kinase reaction is carried out in a highly efficient manner using a pool of peptides derived directly from HEK93 cellular kinase substrates and then dephosphorylated as substrate candidates; PRKACA and

Danjb1-PRKACA were used as kinases to re-phosphorylate these substrates. The resulting newly phosphorylated peptides are then isolated and identified by mass spectrometry. A further comparison of these in vitro phosphorylated peptides with phosphopeptides derived from endogenous proteins isolated from HEK293-T cells, in which the kinase or empty carrier vector reveals new candidate protein substrates. The kinase assay linked with phosphoproteomics strategy was then applied to identify unique substrates of PRKACA, a serine-Threonine kinase and its oncogenic mutant Dnajb1-

PRKACA in human cells.

5.2. Introduction:

Fibrolamellar hepatocellular carcinoma (FL-HCC) is a primary liver cancer that occurs in young people without a gender bias or underlying liver disease. It accounts for less than

1% of all primary liver cancers, but represents the majority of HCCs in patients younger than 30 years of age [144,145].

Surgery is the mainstay of treatment in FL-HCC. There is no clear benefit of any

chemotherapeutic agents that have been shown to have some efficacy in HCC [146]. The

gross anatomic features of FLC are characteristic of an expanding, heterogeneous tumor

mass with areas of increased vascularity and necrosis, including areas of fibrosis similar

to focal nodular hyperplasia, a benign vascular–fibrotic lesion in the liver [147].

Due to the rarity of FL-HCC and a lack of representative experimental systems, the

molecular basis for carcinogenesis in these tumors has been elusive and difficult to study.

Recently, however there has been a breakthrough in understanding the pathogenesis and

expression profile of the FL-HCC. Honeyman and colleagues discovered a novel, chimeric transcript that is present in all studied samples of FL-HCC [148]. Detection of a single,

consistent genetic deletion in one copy of chromosome 19 [148,149] results in the

formation of a chimeric gene, DNAJB1-PRKACA, which combines the first exon of

DNAJB1, the heat-shock protein 40 (HSP40), with exons 2 through 10 of PRKACA, the

catalytic subunit of protein kinase A. This gives rise to a chimeric construct where the N-

terminal helix is fused to the J domain of DNAJB1(J-PKA-C). This chimera has been found

to be the main driver of FL-HCC, and while some gene expression changes have been

characterized, the detailed molecular pathogenesis producing those changes remains

poorly understood.

Protein kinases and their substrates represent the largest signaling network that regulates

protein–protein interactions, subcellular localization, and ultimately cellular functions.

Dysregulation of signaling networks leads to a variety of disease states e.g. human

malignancies, diabetes, and immune disorders [150, 151]. Overexpression of PKA is

considered a hallmark of most human tumors, and PKA may function as a cancer marker

for various human cancers [154]. PKA selective targeting in antitumor strategies has

become very attractive and protein kinases represent a significant class of potential drug

targets. Today, more than 40 FDA approved small molecule kinase inhibitors have been

65 used in US, with many more in development or in clinical trials [152,153]. However, potency and selectivity have remained the main problem with these compounds, in addition, resistance to these drugs can develop. Consequently, understanding the downstream pathways of cAMP signaling would help to find new viable targets in many diseases associated with PKA malfunction or overactivation.

PKA itself is one of the best understood kinases in the human kinome, and there is evidence to support the hypothesis that its kinase activity plays a role in FL-HCC

[148,149]. However, it is upstream of a very broad set of cellular pathways, and its downstream effects are governed by particular regulatory interactions that determine its substrates under a given set of biological conditions. As the substrates of protein kinase

A (PKA) or its regulators may represent a pool of potential “druggable” targets, there is good reason to carefully examine its signaling pathway in the context of the DNAJB1-

PRKACA fusion protein. Although much is known about PKA, the precise connection between the fusion enzyme, which may alter substrate recognition and/or protein-protein interactions, is not well understood. Previous work has suggested that the catalytic activity of the chimeric fusion is not significantly different than that of the native PKA enzyme [148,

195], yet the phenotypic differences observed in FL-HCC suggest that increased PKA activity is not the only contributor to oncogenic transformation [195]. DNAJB1 (otherwise known as HSP40) is a chaperone protein with numerous protein-protein interaction partners. Its fusion with a kinase could potentially bring novel proteins in contact with PKA and promote their aberrant phosphorylation. Indeed, in a recent study it was shown that

PRKACA-Dnajb1 appears to function as a scaffold via its HSP40 component, to assemble signaling elements that contribute to the pathogenicity of FL-HCC [154]. However, its direct, bona fide substrates have not yet been fully elucidated.

Several high throughput approaches for the identification of phosphopeptides with mass spectrometry-based proteomics have been described [155,146], however, these do not

66 typically enable clear assignment of direct substrates relative to those that may have been phosphorylated by other kinases downstream of the target kinase’s activation. Therefore, we sought to understand whether this chimeric fusion between HSP40 and PKA resulted in direct phosphorylation of a different set of substrates that would lead to its activation of unique pathways. In this study, we employed an integrated strategy termed kinase assay linked with phosphoproteomics (KALIP) [157,158] for comparing the specificity and identifying direct substrates of cAMP-dependent protein kinase A (PKA) and the chimeric mutant DNAJB1-PRKACA (Figure 5.1). We identified peptides from proteins phosphorylated in cells expressing PKA or the chimeric mutant, as well as from in vitro kinase reactions with the purified native and chimeric mutant kinases using either digested peptides or whole proteins as the substrate pool. These experiments identified phosphorylated proteins associated uniquely with the chimeric fusion enzyme, and enabled pathway analysis that revealed a number of pathways activated in the presence of the chimeric fusion but not the native PKA. We also identified bona fide direct substrates of the chimeric fusion enzyme that may enable selection of specific alternative targets downstream of PKA activity itself for treatment of FL-HCC.

5.3. Materials and Methods:

Recombinant Enzyme expression and purification

High-level expression of PRKACA-Dnajb1 and Human PKA in E. coli was achieved by the construction of pET28a (+) vector that contained the protein gene subcloned prior to a phage T7 RNA polymerase promoter. Each species was expressed in the E. coli cell line

BL21(DE3) by growing the bacteria in LB medium at 37 . Protein expression was induced by addition of 0.4 mM IPTG and carried out overnight at℃ 24°C before harvesting the cells.

Affinity purification was carried out using Ni-NTA affinity Resin (Thermo Scientific HisPur™

Ni-NTA Resin, Catalog number: 88221). Activity of purified enzymes were confirmed using non-radioactive protein kinase assay (Promega, catalog number V5340).

Human Cell Culture and Transfection

Human embryonic kidney 293 cells (HEK293-T) were maintained in Modified Dulbecco’s

Medium (DMEM) supplemented with 10% fetal bovine serum, 100 μg/ mL of streptomycin, and 100 IU/mL of penicillin in 5% CO2 at 37 °C. Cells were grown to 80-90% confluency before transfection with plasmids. Cells were treated with 0.1 mM sodium pervanadate for

15 minutes in 4 prior to harvesting the cells. The cells were then washed twice with PBS and frozen in -80℃ for further use.

HEK 293-T cells were transiently transfected with 10 µg plasmids containing either human

PRKACA gene (pcrDNA3.1-PRKACA, Addgene) (pcrDNA3.1-PRKACA was a gift from

Sanford Simon [148] (Addgene plasmid # 100890 ), PRKACA-Dnajb1 gene (pcrDNA3.1-

Chimera was a gift from Sanford Simon (Addgene plasmid # 100891) [5] or empty pcrDNA3.1 vector using Lipofectamine™ 3000 transfection kit (ThermoFisher Scientific).

Cells were harvested after incubating for 36 hours at 37 and 5% carbon dioxide in transfection media (DMEM/ 10% FBS). ℃

Inhibitor treatment

Inhibitors were purchased from Tocris. Cell permeable PKI 14-22 amide, myristoylated

(Cat number: 2546; The non-myristoylated version inhibits protein kinase A with a Ki = 36 nM) with 6.4 µM final concentration of inhibitor in each well. Small molecule non hydrolysable inhibitor cAMPS-Rp, triethylammonium salt (Cat. No. 1337) with 370 µM final concentration in each well. The inhibitor rp-cAMP was added at the time of transfection and 12 hours after transfection. PKI14-22 was added 12 hours and 24 hours after transfection (The PKI peptide is prone to in-cell protease interaction.). The cells were maintained in Modified Dulbecco’s Medium (DMEM) supplemented with 10% fetal bovine serum, 100 μg/ mL of streptomycin, and 100 IU/mL of penicillin in 5% CO2 at 37 °C. Cells were grown to ~80% confluency before transfection with plasmids and harvested after

68 incubation for 36 hours at 37 and 5% carbon dioxide in transfection media (DMEM/ 10%

FBS). Cells were treated with℃ 0.1 mM sodium pervanadate for 15 minutes in 4 prior to harvesting the cells. ℃

Western Blotting

After measuring protein concentration of lysates, 50 μg of protein from each sample was reduced with 10 mM DTT and 1× loading buffer (Bio-Rad Cat#161-0710 and Cat.#161-

0737) and heated to 95 °C for 5 min. Proteins were separated on 4–15% Mini-PROTEAN®

TGX™ Precast Protein Gels and transferred to PVDF membranes. After blocking with

TBS-T milk containing 0.1% Tween 20 and fat-free powdered milk for 1 hour the membranes were probed against the proteins of interest using the appropriate primary

(anti human PKA-C mouse antibody) and secondary (donkey anti mouse IR Dye 680) antibodies (Figure S1).

Peptide Preparation

Peptides were prepared as previously described. Cell pellets that had been frozen at −80

°C were lysed in lysis buffer (20% acetonitrile, 7M Urea, 2M thiourea, 0.4 M TEAB, 4mM

DTT, phosphatase inhibitor cocktail) and probe sonicated to shear DNA. The cell debris was cleared by centrifugation at 16,000g for 30 min at 4 °C, and the supernatant containing the soluble proteins was collected. Protein concentrations were measured using the

Bradford assay, and up to 1 mg of protein was reduced using TCEP and alkylated with iodoacetamide (40 mM) for 1 hour at RT in the dark. Proteins were then digested overnight with proteomics grade trypsin (Pierce™ Trypsin Protease, MS Grade-Catalog number

90058) at a ratio of 1:50 by mass (enzyme/substrate) at 37 °C. On the next day, samples were acidified with 10% trifluoroacetic acid (TFA), desalted with Waters Oasis HLB cartridges and concentrated to dryness using a Speed-Vac.

SMOAC Enrichment of Phosphopeptides

Sequential enrichment by Metal Oxide Affinity Chromatography (SMOAC) (Thermofisher

Scientific) was used to enrich phosphopeptides according to manufacturer’s instructions

[159]. This method takes advantage of a two-step consecutive enrichment of

phosphopeptides using TiO2 and Fe-NTA spin columns, with a simple and robust protocol

to produce clean, reproducible peptide mixtures for MS [159]. The recovered

phosphopeptides from this step were concentrated to dryness using a Speed-Vac.

Kinase reaction

For peptide KALIP, peptides were first alkylated with iodoacetamide (40 mM) for 1 hour at

RT in dark. After quenching the reaction using 10mM DTT, peptides were digested

overnight at a ratio of 1:50 by mass (enzyme/total lysate protein) at 37 . On the next day,

samples were acidified with 10% trifluoroacetic acid (TFA), desalted℃ with Waters Oasis

HLB cartridges and concentrated to dryness using a Speed-Vac. Samples were then

resuspended in 100 µL lambda phosphatase reaction buffer (50mM Hepes, 100mM NaCl,

2mM DTT, 1mM MnCl2, pH 7.5). Lambda phosphatase (New England Biolabs P0753S) was added to each sample (2800-3000 units) and incubated overnight at 30 . The phosphatase was deactivated by heating at 75 for 10 minutes. Samples were incubated℃ in kinase assay reaction buffer (100Mm Tris HCl℃ pH 7.8, 10Mm MgCl2, 5Mm DTT, 2Mm

ATP) containing the purified recombinant kinase (~5000 units) at 30 for 90 minutes.

Reactions were quenched by addition of 0.5% TFA to a pH below 3. The℃ samples were then desalted using HLB cartridges and concentrated to dryness using a Speed-Vac.

For the native protein KALIP experiment, samples were treated with lambda phosphatase

(2800-3000 units) and incubated at 37 for 90 minutes. Since it was desirable to keep

the proteins in their native, folded state ℃as best as possible, in order to stop the reaction a

cocktail of two mini tablets of phosphatase inhibitor (Pierce™ Phosphatase Inhibitor Mini

Tablets, Catalog number A32957) and two tablets of protease inhibitor (Roche

cOmplete™, Mini, EDTA-free Protease Inhibitor Cocktail 04693159001) dissolved in 5 ml

kinase reaction buffer was prepared and 150 µl of this cocktail was added to each sample,

stirring gently for 10 minutes. The purified recombinant kinase was then added to the

sample (5000 units) while gently stirring for 60 minutes at 30 . The kinase reaction was

quenched by heat inactivation (75 for 10 minutes) and dialyzed℃ overnight against trypsin digestion buffer (100 mM Tris-HCl,℃ 10 mM CaCl2, 5mM DTT, pH 7.8) in pur-a-lyzer midi dialysis tubes (Sigma PURD10005) for buffer exchange. Samples were then resolubilized using up to 2M urea, alkylated with iodoacetamide (40 mM) for 1 hour at RT in dark and digested using trypsin as explained for peptide KALIP previously. Peptide solutions were desalted using Waters Oasis HLB cartridges and concentrated to dryness using a Speed-

Vac.

Mass Spectrometric Data Acquisition and Data Analysis

The phosphopeptide samples were then run using MS/MS on an Orbitrap Fusion mass spectrometer as previously described [233]. To explain briefly, the samples were run on a

ThermoScientific Easy NanoLC LC 1000 system coupled to a high resolution Orbitrap

Fusion Tribrid Mass Spectrometer. On the LC side the samples were run on a reverse phase C18 column with a [Water+.01% Formic acid] to [Acetonitrile+.01% Formic Acid] gradient of 2%-30% that increased linearly over a period of 60 minutes. A flow rate of

200nL/min flowed into the mass spectrometer which operated in data-dependent mode with a resolution of 60,000 and a scan range of 300-1500 m/z. The top 12 most abundant ions were selected with a dynamic exclusion time of 15 seconds for High Collision

Dissociation (HCD) and fragments were analyzed.

Post LC-MS Analysis

All tools discussed herein, except ProteinPilot. area available through the GalaxyP platform of UMN (galaxy.msi.umn.edu).

The mass spectrometer produced raw files which were converted into .mgf files by the

Thermo Raw tool on GalaxyP. Those .mgf files were then searched by ProteinPilot 5.0 to find what peptides had been seen by the mass spectrometer, and what proteins those peptides came from. The search algorithm was performed against the Uniprot reviewed human proteome database merged with a cRAP common lab contaminants file

(thegpm.org/crap/). The following parameters were used: trypsin digestion, iodoacetamide alkylation, urea denaturation, and an emphasis on discovering phosphorylated peptides.

The search algorithm returned a file containing all the peptides that were discovered by the mass spectrometer, and the proteins they came from. That file was then run through the Kinamine tool (https://toolshed.g2.bx.psu.edu/view/jfb/st_kinamine/27b9583c0290) to extract only the phospho peptides and the proteins that those phospho peptides came from.

The mass spectrometry experiment was run with two or three different replicates for each condition. To remove some of the random variability, only those peptides that were found in all three replicates were used for further data analysis. The Commonality Finder

(https://testtoolshed.g2.bx.psu.edu/view/jfb/stcommon_and_dif/1875d4b62014) tool was used to extract only those peptides found in all replicates.

Additionally, we wanted to discover which proteins were ONLY phosphorylated by wild- type PKA or ONLY phosphorylated by the chimera mutant. For this, difference finder

(https://testtoolshed.g2.bx.psu.edu/view/jfb/difference_finder/5daecc1cd7db) was used to subtract all PKA substrates from PRKACA-Dnajb1 set, and subtract all PRKACA-Dnajb1 substrates from the PKA set. Furthermore, any substrates found in the negative (kinase not added) sets were subtracted out. This resulted in having a set of substrates that were only phosphorylated by PKA and a set of substrates that were only phosphorylated by the chimera mutant.

Finally, in order to determine if the overall preferred substrate motif changed between the chimera mutant and the wilt-type PKA, the substrates were put through the KINATEST-ID algorithm (https://testtoolshed.g2.bx.psu.edu/view/jfb/st_kinatest/95c0e1ee15b7) to discover their preferred substrate motif.

5.4. RESULTS

Cell-based phosphoproteomic profiles and signaling pathway comparison

In order to compare the signaling profiles of human PRKACA and its oncogenic mutant

PRKACA-Dnajb1(chimera), constructs containing each protein were transiently expressed in HEK293 cells (along with vector-only control). Expression was verified by Western blot for human cAMP-dependent protein kinase A catalytic subunit. Cells were treated briefly with sodium orthovanadate to enhance phosphopeptides stability and lysed in denaturing lysis buffer. Phosphopeptides were enriched using as described in the Materials and

Methods and analyzed on an Orbitrap Fusion LC-MS/MS. Mass spectrometry data were processed and analyzed using Protein Pilot 5.0 software (SCIEX). The high confidence phosphorylation sites observed per identified protein were extracted from the Feature

Summary tab in the report from Protein Pilot 5.0, and input for Ingenuity Pathways

Analysis. The pathways observed as significant were classified into three major groups: a) pathways associated with cell proliferation and translation/transcription control, b) inflammatory and oxidative stress response, and c) DNA damage response. Eukaryotic translation initiation factors (EIFs), the tRNA charging pathway, ERK/MAPK Signaling, mTOR and Wnt/β-catenin signaling were the pathways most clearly enriched in PRKACA-

Dnajb1expressing cells (Figure 5.2). These observations are consistent with previous reports on human FL-HCC tumor tissue gene expression profiles, [195,196] transcriptome analysis of FL-HCC from tumor samples, [217] and recent proteomics studies on the FL-

HCC intratumor environment (which included a mouse model of the DNAJ fusion as well

as patient samples) [234]. These results support the usefulness of HEK293 cells as a

model for examining direct substrates of these kinases in this context.

Identification of direct in vitro Peptide Substrates of PKACA and PRKACA-Dnajb1

Cyclic-amp dependent protein kinase A (PKA) displays a number of phosphorylation

motifs that have variations on the basic R/K-R/K-X-S/T motif, including R-X-R-R-X-S-Φ

(where Φ is a hydrophobic residue), R-R-X-S-Φ and R-R-X-S/T. Peptide KALIP is performed on a predigested substrate pool, in order to limit the degree of false positive hits and any promiscuous kinase activity, a negative control (vector only) was running in parallel and the phosphopeptides were enriched prior to LC-MS/MS.

Comparison of the number of phosphosites identified as the result of kinase reactions versus the control indicates the efficiency of phosphatase treatment and following procedure (Figure 5.3). In case of PRKACA-Dnajb1, analysis of the phosphorylation motif

and the sequences flanking the unique phosphorylated Serine/Threonine residues by

WebLogo 3.1 (Figure 5.3,B and D) produce logos or motifs that, in peptide KALIP, contain

hydrophobic residues upstream of the phosphorylation site and

Serine/Glycine/Alanine/Leucine at the -1 to -3 position. However, in case of protein KALIP

the generated motif (Figure 5.3 D) shows enrichment of charged aminoacids both

upstream and downstream of the phosphorylation site which resembles the expected PKA

motifs reported. In case of PRKACA, the peptide motif contains more basic residues

downstream the phosphorylation site(S/T), which is closer to what is expected from a

known PKA motif. This difference in the flanking aminoacids between the peptide and

protein motifs suggests the role of 3D protein-protein interactions in enzyme-substrate

interactions in the cell.

Common substrates in KALIP-peptide and in-cell analysis: The peptide substrates are

found in both in-vitro (kinase peptide substrates) and in-cell (transient cell expression of

kinases) kinase reactions (direct peptide substrates). A subset of unique PRKACA-Dnajb1

74 direct peptide substrates are presented in table 5.1 (for the whole set of substrates refer to tables S1-S4).

Upregulated canonical pathways in PRKACA-Dnajb1 profile:

Some of the most dramatically enriched canonical pathways previously observed in the in-cell as well as in vitro analysis include those associated with a) proliferation and replication: regulation of eIF4 and p70S6K signaling, mTOR, AMPK, VEGF, ERK/MAPK,

EIF2 and Wnt/β-catenin signaling; b) those modulating cell adhesion, migration, and junction signaling: Gap Junction, Integrin, Germ Cell-Sertoli Cell Junction Signaling and remodeling of epithelial adherent junction; c) DNA damage/repair and checkpoint control:

Cell cycle control of chromosomal replication, role of BRCA1 in DNA Damage Response,

CDK5 Signaling and role of CHK Proteins in Cell Cycle Checkpoint Control (Figure 5.4). another pathway that is upregulated in both in-cell and in vitro phosphoenrichment analysis is Breast Cancer Regulation by Stathmin1. Stathmin1 (STMN1, also known as

Op18, p18, p19, or metablastin) is a candidate oncoprotein and prognosis marker in several kinds of cancers [222] and might serve as a prognostic marker and a potential therapeutic target for FL-HCC.

Identification of direct in vitro Protein Substrates of PKACA and PRKACA-Dnajb1

Protein level kinase reactions can be advantageous when specific protein conformations are needed to facilitate kinase-substrate docking and subsequent phosphorylation. These sets of proteins substrates are phosphorylated by PRKACA-Dnajb1 or PKACA in vitro and are also observed in the in-cell phosphopeptides profile (Table 5.2). A full set of these native direct substrates are presented in table S3, and those exclusive to PRKACA-Dnajb1 are listed in table S1.

Upregulated canonical pathways in PRKACA-Dnajb1 profile: Some of the most enriched pathways in native state KALIP phosphoproteomic study includes those associated with cell proliferation and tumorigenesis such as mTOR, AMPK, ERK/MAPK, Chemokine

Signaling, Telomere Extension by Telomerase, Integrin-linked kinase (ILK), Endometrial

Cancer Signaling and PI3K/AKT Signaling. Those canonical pathways associated with

DNA damage/repair and cell cycle arrest include DNA Double-Strand Break Repair by

Non-Homologous End Joining, cell cycle control of chromosomal replication, p53 signaling

(also apoptotic) and CDK5 Signaling. Finally, those pathways modulating cell adhesion,

migration, and junction signaling include ILK (EMT mediator), regulation of actin-based

motility by Rho, remodeling of epithelial adherens junctions and germ cell-sertoli cell

junction signaling (Figure 5.5). The canonical pathways that are upregulated in both

protein and peptide KALIP are presented in Table 5.4.

Inhibition of cAMP-dependent protein kinase A (PRKACA) signaling

In order to inspect which phosphorylation sites and/or pathways are affected by inhibition

of PRKACA and PRKACA-Dnajb1, two PKA inhibitors were selected and applied to the

cell media: rp-cAMP which is a potent inhibitor of cAMP-PKA, and protein kinase A

Inhibitor peptide (PKI14-22). The mechanism of inhibition for each of these inhibitors is

different: rp-cAMP competitively inhibits the cAMP-induced activation of cAMP-dependent

protein kinase (PKA) by stabilizing the inactive conformation or holoenzyme [227,228].

The nucleotide binding site in PKACA is located deeply between the two lobes, in a

hydrophobic core; while Protein kinase inhibitor peptide (PKI) is an endogenous

thermostable peptide that potently and specifically inhibit the activity of the free catalytic

subunit of PKA and export the free catalytic subunit of cAMP-dependent protein kinase

from the nucleus (The peptide inhibitor PKI14-22 used in this study lacks the nuclear export sequence). The two inhibitors were selected for their different inhibitory mechanism, both are known for their selective and potent inhibition of cAMP defendant protein kinase A

(PKA). Treatment of HEK293 cells with each of PKA inhibitors resulted in a noticeable reduction of phosphopeptides (Fig.5.6), while the western blot confirms the expression of both PKA and PRKACA-Dnajb1 in treated HEK293 cells (supporting figure S.1). Cells

were all treated with sodium pervanadate prior to collection to prevent phosphatase

activity. A subset of inhibited substrates is presented as table 5.3. More in dept analysis

of the substrates affected by kinase inhibition classifies them into main pathways of DNA

double strand break repair, epithelial adherens junction remodeling and cell to cell

signaling, EIF2 and VEGF signaling (table. S4). These observations is in agreement with

previous KALIP peptide and protein phosphorylation analysis, confirming that these

substrates and their associated pathways are directly affected by PRKACA-Dnajb1.

Some of the canonical pathways that are either diminished or deteriorated as a result of

cAMP signaling inhibition are shown in figure 5.7 (These canonical pathways are also

confirmed by the previous in-cell study). Overall pattern of Inhibitory effect of rp-cAMP on

PRKACA-Dnajb1 is different from PKA, and rp-cAMP appears to be a much more potent inhibitor of PKA than its mutant PRKACA-Dnajb1.This could be due to the fact that rp- cAMP binds most effectively to the holoenzyme (Cα2R2 form) were presence of the

Dnajb1 appendix and its dynamic can affect its conformational dynamic landscape and its binding to the regulatory subunit [235]. Nevertheless, the trend of signal reduction is mostly the same for both kinases. Specifically, pathways associated with cell motility, migration and cytoskeleton organization or cell junction are visibly affected by the inhibitors (fig.5.7). The EIF2 signaling pathway which is also one of the most predominant ones in PRKACA-Dnajb1 in both KALIP and in-cell phosphopeptides profiles, is inhibited in the presence of rp-cAMP, but to a lesser extent in the presence of PKI14-22.

5.5. DISCUSSION

Kinases have become one of the most intensively pursued classes of drug target with

approximately 30 distinct kinase targets being developed to the level of a Phase I clinical

trial. Most of these targets are being investigated for the treatment of cancer. However,

deregulation of kinase function has been implicated in other disorders as well, including

immunological, neurological, metabolic and infectious disease. The genetic or epigenetic

77 alteration of kinase may result in increased specific activity of the kinase itself, its overexpression, or the loss of negative regulation. Most frequently, tumor cells harbor somatic point mutations at structurally conserved residues, or mutation hotspots, which constitutively upregulate kinase activity [160]. PKA, also known as cAMP-dependent protein kinase A, is a multi-unit protein kinase that mediates signal transduction of G- protein coupled receptors through its activation upon cAMP binding. The ubiquitous expression of PKA subunit genes, and the myriad of mechanisms by which cAMP is regulated within a cell suggest the critical role of PKA signaling in cellular function. PKA signaling is involved in the control of a wide variety of cellular processes from metabolism to ion channel activation, cell growth and differentiation, gene expression and apoptosis.

Importantly, since it has been implicated in the initiation and progression of many tumors,

PKA has been one of the most sough-after targets in cancer therapy and drug design. the cAMP pathway interacts with other intracellular signaling pathways, including cytokine and

Ras-Raf-Erk pathways [161-163]. These signaling connections play an important role in cancer biology and a combined blockade of such signaling pathways is considered a relevant strategy for therapeutic intervention [164,165]. The major nuclear targets of PKA are the transcription factors of the cAMP response element binding (CREB) family [166]. several lines of evidence obtained from studies on leukemia, fusion oncoproteins, viral oncoproteins and endocrine tumors support the notion that CREB is involved in oncogenesis [167-169]. As a result, cAMP signaling effectors serve as an interesting target in tumor development and prevention studies.

A remarkable increase in the enrichment of EIF2 and EIF4/p70S6K signaling pathways was observed upon expression of PRKACA-Dnajb1 in the cells (Fig.5.2), both involved in cellular growth, proliferation and development. Studies in the past two decades have shown that a number of initiation factors (higher expressions levels of eIF4A, eIF4E,

78 eIF4G, as well as low expressions levels of 4E-BPs or phosphorylation of EIF2) are involved in various types of cancers [236-238].

Another major pathway enriched in PRKACA-Dnajb1 expressing cells is ERK

(extracellular-regulated kinase)/MAPK (mitogen activated protein kinase) signaling, a key pathway that transduces cellular information on meiosis/mitosis, growth, differentiation and carcinogenesis within a cell. Consistent with our finding, heterogeneous activation of the ERK cascade within the FL-HCC intratumor environment is reported in a recent study

[234]. Aberrant activation of signaling through the RAS-RAF-MEK-ERK (MAPK) pathway is implicated in numerous cancers, making it an attractive therapeutic target [183,184].

MAPK/ERK pathway is a critical pathway for human cancer cell survival, dissemination, and resistance to drug therapy and can demonstrates both oncogene and tumor suppressor effects depending on the tissue-specific tumor microenvironment [184].

A major signaling pathway elevated in our in-cell study was Wnt/β-catenin (fig.5.2).

Interestingly, Kastenhuber and colleagues observed that expression of the DNAJB1–

PRKACA fusion alone led to increased membranous ß-catenin and phosphorylation of ß- catenin at PKA phosphorylation site S675 in mice [195]. Our in vitro direct substrate assay for PRKACA-Dnajb1 confirms the phosphorylation of catenin-β (CTNNB), albeit on T551 and S552, the latter has been reported previously to be a substrate of PKA and AKT

[205,206]. Different studies show that phosphorylation on S552 and S675 contributes to

β-catenin stabilization and transcriptional activation in cancers [205,206]. The Wnt/ β- catenin signaling pathway deregulation has also been attributed to HCC (Hepatocellular carcinoma) development [200-202] and it has been shown that concurrent abnormal activation of the phosphoinositide 3-kinase/Akt (PI3/Akt) and Wnt/β-catenin signaling pathway, promotes the epithelial-to-mesenchymal (EMT) transition and cancer progression [206,207].

The NRF2-mediated oxidative stress response and protein ubiquitination signaling were

another observed elevated pathway in PRKACA-Dnajb1 expressing cells (fig.5.2). In a recent study, it was showen that Nrf2 target genes can play a potential role in hepatic injury, inflammation and hepatocarcinogenesis [203]. In agreement with the upregulated oxidative stress response pathway observed in our in-cell study, Kastenhuber and colleagues have similarly reported the upregulation of enzymes involved in reactive oxygen response (ROS) in human and murine FL-HCC tumor [195]. The Nrf2‐mediated antioxidant response is one of the major cellular defense mechanisms that protect against oxidative stress. Oxidative stress has been known to be the major causes underlying HCC development too, by driving DNA damage, reactive oxygen species (ROS), reactive nitrogen species (RNS) production and altered protein expression [203]. located at the crossroad of multiple defensive responses influencing cell fate during xenobiotic, oxidative, and metabolic stress, the NRF2/KEAP1 pathway acts as a double-edged sword in many tumors, promoting cancer growth, survival, metastasis formation and therapy resistance [198,199]. Likewise, Vascular Endothelial growth factor (VEGF) signaling is

among the pathways activated in hypoxia and Inflammation, which is noticeably elevated

in cells expressing PRKACA-Dnajb1 (fig.5.2). The expression of VEGF is particularly found to be correlated to hypoxia in the tumor cells. VEGF also induces tumor angiogenesis; hence its inhibitors are effective in inhibiting tumor growth and metastasis and are attractive drug targets in tumor suppression [208].

Eukaryotic mismatch repair and cell cycle control of chromosomal replication pathways are among the enriched pathways upon overexpression of PRKACA-Dnajb1 gene in cells.

Several studies have linked upregulation of DNA repair genes with chemo- and radio- resistance in multiple tumor types [45] and with the ability of tumors to metastasize [189,

190]. Therefore, while loss of DNA repair function is significant in cancer initiation, gain of function of similar genes and re-activation of lost repair pathways is involved in disease

80 progression [191]. Simon and colleagues have also confirmed the upregulation of DNA damage checkpoint in the FL-HCC transcriptome analysis [217]. DNA repair pathways, in a large sense, are overexpressed in primary tumors associated with high risk of distant metastasis. Analysis of the biological pathways associated with metastatic potential points to cell adhesion, angiogenesis, cell cycle regulation, initiation of DNA synthesis, and DNA repair [191,192]. Pathways related to breast cancer is also enriched in the in-cell phosphoproteomic analysis (fig.5.2), an observation also reported by transcriptome analysis of FL-HCC tumor tissues [217].

5.6. Conclusion

Using robust phosphoproteomic techniques, we were able to determine the most likely bona fide direct substrates of native PKA and those arising from the fusion mutation

PRKACA-Dnajb1. We also examined downstream signaling uniquely associated with phosphorylation by the chimeric fusion and identified several pathways that were only altered in the presence of the chimeric fusion and not native PKA overexpression.

Moreover, from a pharmacological standpoint one of the greatest challenges in developing kinase specific inhibitors is the occurrence of off target effects, causing adverse and unforeseen reactions in the clinical trials. We have applied KALIP using PKA potent inhibitors to delineate substrates of kinase of interest. This strategy can be the first step in the designing specific substrates/inhibitors as direct targets of kinases, or even for the purpose of targeting upstream regulators.

Chapter VI

Protein Kinase A and Cushing’s disease

This section has been focused on mutants of cAMP dependent protein kinase A (PKA) that have been associated with Cushing’s disease. The contribution to the first mutant

Lys205Arg has been published in 2019 in the Science Advances (Cushing’s syndrome driver mutation disrupts protein kinase A allosteric network, altering both regulation and substrate specificity. Sci. Adv. 2019;5: August 2019.). The next mutant Trp196Arg structure, dynamics and thermodynamic stability has been extensively studied and presented in this chapter.

Introduction: Endocrine cells are under the strict control of hormones and local messengers that act primarily through G-protein-coupled receptors. Several of these receptors activate the cAMP/PKA pathway, leading in most cases to both increased hormone production and cell replication. Thus, somatic mutations causing constitutive activation of the cAMP/PKA pathway have been shown in several instances to be responsible for tumor formation and excessive hormone production, ultimately leading to endocrine Disorders [239-242]. Moreover, germline-inactivating mutations in the

PRKAR1A gene, coding for the PKA RIa subunit, have been linked to Carney complex, a rare syndrome characterized by excess of several hormones—most commonly cortisol excess (Cushing’s syndrome) [243]. Endogenous Cushing syndrome (CS) is a rare disorder. The clinical phenotype of patients with Cushing's syndrome is characterized by severe comorbidities, due to the excessive cortisol production. The presence of resistant hypertension, type 2 diabetes, and vertebral osteoporotic fractures are common features of the syndrome that are frequently found in association with typical catabolic signs such as easy bruising, purple striae, and proximal muscle weakness [244]. If not adequately

82 treated, severe and prolonged hypercortisolism could lead to an increased morbidity and mortality, mainly due to cardiovascular diseases [245, 246] and infectious complications

[247]. Recently, somatic mutations of the PRKACA gene, encoding for the catalytic subunit α (Cα) of the PKA, have been identified in more than one third of patients with

Cushing's syndrome due to sporadic adrenocortical adenomas [248]. In different studies, the most common genetic alteration found in patients was a missense mutation that leads to constitutive activation of the Cα subunit, resulting in cAMP-independent activity of the

PKA and enhanced cortisol production [249]. Moreover, patients with the mutation in the PRKACA gene had a more pronounced cortisol hypersecretion than nonmutated subjects. In a study on patients with non-secreting adenomas (n = 32), subclinical hypercortisolism (n = 36), Cushing's syndrome (n = 64), androgen-producing tumors (n =

4), adrenocortical carcinomas (n = 5), and primary bilateral macronodular adrenal hyperplasias (n = 8) sequencing analysis of the tissues showed mutations of exon 7 of the

PRKACA gene in 22 of 64 adenomas with Cushing's syndrome (34%). No mutations were found in nonsecreting adenomas and in samples from patients with subclinical hypercortisolism, nor in the remaining 24 tumoral samples of androgen-producing tumors, adrenocortical carcinomas, and primary bilateral macronodular adrenal hyperplasias. In addition, no mutations were found in normal adrenal tissue taken from peritumoral specimens [249]. All identified mutations involve regions that are located at the interface between the catalytic and regulatory subunits, with Leu205Arg being the most prevalent mutation. All identified mutations involve regions that are located at the interface between the catalytic and regulatory subunits (Fig.6.1).In silico analysis of the most frequent mutation (Leu206Arg), suggested that the exchange of Leucine by a bulky and positively charged amino acid such as Arginine would lead to a steric hindrance with Val115 and

Tyr228 in the regulatory subunit, likely impairing holoenzyme formation and, hence, leading to constitutive PKA activation [248]. This hypothesis was supported by functional

83 studies of the Leu206Arg variant showing high basal PKA activity, lack of suppression by the regulatory subunit, and absence of regulation by cAMP [248].

We have studied two of these single mutations in PRKACA that have been reported to be present in tissues from patients with Cushing’s syndrome: the prevalent L205R and less frequent mutation W196R. The enzyme activity, thermal stability, structural and dynamics of these mutants have been extensively studied and presented in the following section.

Kinetic assay: Standard kinetic analysis uses the coupled enzyme assay in which the hydrolysis of ATP by PKA is coupled to the reaction involving the conversion of phosphoenolpyruvate to lactate. The reaction is set up in such a manner that the hydrolysis of ATP is the rate limiting step as to monitor the activity of PKA directly.

A coupled reaction uses one of the products as a reactant for an additional enzyme. That enzyme is typically easy to measure, however, there are lots of considerations with this type of assay. There must be enough of the second enzyme present, so that it isn't limiting the rate. The reactants for the second reaction also must be in excess, so the rate is limited only by the production of the reactant for the second enzyme.

ATP ADP

Kemptide Kemptide + P W196R

+ ADP ATP NADH NAD

Phosphoenolpyruvate Pyruvate Lactate Pyruvate Lactate Dehydrogenase Kinase

Depletion of NADH will then be monitored spectroscopically at 340 nm as the reaction proceeds.

Using standard PKA substrate “kemptide” and a standard NADH-coupled assay, the kinetic parameters of both mutants were measured. The data were fitted to the Michaelis-

Menten equation to obtain Km values:

As the data in table 6.1 shows, both PKA mutants are less efficient enzymes compared to

the wild type PKA.

Thermodynamic (thermal) stability: Circular dichroism (CD) is an excellent spectroscopic

technique for following the unfolding and folding of proteins as a function of temperature.

circular dichroism is defined as the difference between the absorption of left-handed, εR,

and right-handed, εL, circularly polarized light:

𝐿𝐿 𝑅𝑅 ΔA is a function of wavelength, so for a∆ 𝐴𝐴measurement𝐴𝐴 − 𝐴𝐴 to be meaningful the wavelength at

which it was performed must be known.

One of its principal applications is to determine the effects of mutations and ligands on

protein and polypeptide stability If the change in CD as a function of temperature is

reversible, analysis of the data may be used to determine the van't Hoff enthalpy (ΔH) and

entropy (ΔS) of unfolding, the midpoint of the unfolding transition (Tm) and the free energy

(ΔG) of unfolding. Binding constants of protein-protein and protein-ligand interactions may also be estimated from the unfolding curves. Analysis of CD spectra obtained as a function of temperature is also useful to determine whether a protein has unfolding intermediates

Secondary structure can also be determined by CD spectroscopy in the "far-UV" spectral region (190-250 nm). At these wavelengths the chromophore is the peptide bond, and the signal arises when it is in a regular, folded environment.

Circular Dichroism spectroscopy was used to measure the thermal stability of PKA-

Human. Protein was dialyzed against the CD buffer (10 mM Pipes (1,4- piperazinebis(ethanesulfonic acid)), pH 7.0, and 150 mM NaCl ) overnight and three forms of PKA were measured: Apo, binary (in complex with nucleotide ATP and Mg ion) and

2+ ternary form (in complex with nucleotide ATP, Mg and PKI 5-24). Samples were incubated over a temperature range of 25 to 75 ˚C at a rate of 1 ˚C/min in a rectangular quartz cuvette in a Jasco J-815 spectropolarimeter. Spectra were acquired at 222 nm in a stepwise

fashion at 1 ˚C increments following an equilibration time of 10 s. A blank spectrum

acquired with all reaction components except PKA was subtracted in selected

experiments. The data were fitted to a two-state sigmoidal unfolding model using Origin

8.0 (Microcal). The midpoint of the curve was taken as the melting temperature (Tm).

As the data in table 6.2 shows, the thermodynamic stability of the mutant W196R is less

than the mutant L205R and wild type PKA (Human).

Since the PKA mutants associated with the Cushing’s disease are located at the interface

between the regulatory and catalytic subunits binding site, it was assumed that these

mutants-particularly those with bulky positive sidechains e.g. L205R or W196R-would

likely hinder the interaction with the regulatory subunits. To our surprise, this assumption

was not true for the W196R mutant. Using a simple pull-down assay with affinity

chromatography, the regulatory subunit RΙΙβ was co-eluted with the W196R, indicating

that the mutant interacts with the regulatory subunit (Figure 6.2). However, the binding

constant of the binding is not clear, it is possible that the binding is less efficient compared

to the wild type PKA. Likewise, binding of the mutant with the known PKA inhibitors was

monitored using non-radioactive gel shift assay (figure 6.3). The PepTag® Non-

Radioactive Protein Kinase Assay (Promega) provide use brightly colored fluorescent

peptide substrate that is highly specific for PKA (PepTag® A1 Peptide-LRRASLG).

Phosphorylation of the peptide alters the net charge from +1 to –1, allowing phosphorylated and nonphosphorylated versions of the substrate to be rapidly separated on an agarose gel at neutral pH. Using fluorescent detection, less than 2ng of purified kinase can be detected in less than 2 hours. It was apparent that W196R is inhibited by both PKA inhibitors (Figure 6.3).

Conformational studies using NMR

Using the same method previously explained in chapter III, the Human PKA-Cα and

W196R mutant were labelled and purified. The gene is inserted in pET28 a(+) and has

Thrombin cleavage site after His tag. Primers were designed for inserting a TEV protease

cleavage site right before the PKA-Human gene.

TROSY-HSQC of Human PKA:

2D NMR spectra of 15N labeled PKA-Human in three forms of Apo, Binary and Ternary were recorded. Isoform II was selected for complex studies, but isoform III was also acquired and overlaid to isoform II to detect possible changes in peak position. The experiment condition for each acquisition are as follows: Isoform 2 in apo (non-nounded) form was recorded on 7001 magnet, NS: 64, S/N: 33, PH: 6.5 (Figure 6.4), Isoform 2 in

APO and AMPPNP bound form (nucleotide bound or binary form), Recorded on 7001, NS:

64, S/N: 33, PH: 6.5 (Figure 6.5) and in complex with AMPPNP and PKI (ternary

complex)(Figure 6.6).

double labeled, Perdeuterated PKA-Human Triple resonance assignment: 3D

experiments: HNCA, HNCACB, HN(CO)CA were performed on 850 MHZ magnet. Each

spectrum was processed, and assignments were transferred from Mouse PKA (formerly

assigned by Jonggul Kim) to the new spectra. Remaining peaks were assigned by me and

Cristina Olivieri (Figure 6.7).

TROSY-HSQC of PKA mutants associated with Cushing’s disease:

Mutant L205R: In order to analyze the intermediate conformations of the “L205R” using

CONSICE (Chapter III), a series of HSQC spectra were recorded using nucleotides and

inhibitor peptide PKI. The chemical shift changes in the peaks in each spectrum will be

used to monitor the intermediate conformations and fast dynamic motions (figures 6.8).

We applied CONCISE to the conformational transitions of L205R as mapped by [1H,15N]-

HSQC. The HSQC data used in our analysis were in different states of Apo, binary form

(complex with AMPPNP) and ternary form (AMP-PNP and PKI5-25 bound). Results are shown in figure 6.9. The contribution of current study is published in Sci. Adv. 2019 [250]

Mutant W196R: TROSY-HSQC spectra of 15N-labeled W196R was recorded for three main states of Apo, Binary (nucleotide bound) and ternary (Closed) forms (Figure 6.10).

Ternary complex (nucleotide and inhibitor bound form) is less correlated in W196R compared to the WT-PKA, another indication that a single mutation can completely re- wire the allostery of a protein.

References: 1- Manning, G., et al., (2002) The Protein Kinase Complement of the Human Genome. Science, 298(5600): p. 1912-1934. 2- Walsh, D.A., J.P. Perkins, and E.G. Krebs, (1968) An adenosine 3',5'-monophosphate- dependant protein kinase from rabbit skeletal muscle. J Biol Chem,. 243(13): p. 3763-5. 3- Knighton, D. R.; Zheng, J.; Ten Eyck, L. F.; Ashford, V. A.; Xuong, N.-h.; Taylor, S. S.; Sowadski, J. M. (1991) Science, 253, 407. 4- Knighton, D. R.; Zheng, J.; Ten Eyck, L. F.; Xuong, N.-h.; Taylor, S. S.; Sowadski, J. M. (1991) Science, 253, 414. 5- Cheung et al, (2015) Structural insights into mis-regulation of protein kinase A in human tumors. PNAS, 112 (5) 1374-1379. 6- Honeyman, J. N. et al. (2014) Detection of a recurrent DNAJB1-PRKACA chimeric transcript in fibrolamellar hepatocellular carcinoma. Science 343, 1010–1014 7- Darcy, D. G. et al. (2015) The genomic landscape of fibrolamellar hepatocellular carcinoma: whole genome sequencing of ten patients. Oncotarget 6, 755–770 8- Craig et al, (1980) Fibrolamellar carcinoma of the liver: a tumor of adolescents and young adults with distinctive clinico-pathologic featuresCancer. 15;46(2):372-9. 9- Walsh, D.A., Perkins, J.P., Krebs, E.G., (1968). An adenosine 3′,5′-monophosphate- dependent protein kinase from rabbit skeletal muscle. J. Biol. Chem. 243, 3763–3765. 10- Fischer, E.H., Krebs, E.G., 1955. Conversion of phosphorylase b to phosphorylase a in muscle extracts. J. Biol. Chem. 216, 121–132. 11- Corbin, J.D., Sugden, P.H., Lincoln, T.M., Keely, S.L., (1977). Compartmentalization of adenosine 3′,5′-monophosphate and adenosine 3′,5′-monophosphate-dependent protein kinase in heart tissue. J. Biol. Chem. 252, 3854–3861. 12- Corbin, J.D., Sugden, P.H.,West, L., Flockhart, D.A., Lincoln, T.M., McCarthy, D., (1978). Studies on the properties and mode of action of the purified regulatory subunit of bovine heart adenosine 3′:5′-monophosphate-dependent protein kinase. J. Biol. Chem. 253, 3997–4003. 13- Ringheim, G.E., Taylor, S.S., (1990). Dissecting the domain structure of the regulatory subunit of cAMP-dependent protein kinase I and elucidating the role of MgATP. J. Biol. Chem. 265, 4800–4808. 14- Rannels, S.R., Corbin, J.D., (1980). Studies of functional domains of the regulatory subunit from cAMP-dependent protein kinase isozyme I. J. Cyclic Nucleotide Res. 6, 201–215. 15- Taylor, S.S., Ilouz, R., Zhang, P., Kornev, A.P., (2012). Assembly of allosteric macromolecular switches: lessons from PKA. Nat. Rev. Mol. Cell Biol. 13, 646–658. 16- Manning, G., Whyte, D.B., Martinez, R., Hunter, T., Sudarsanam, S., (2002)a. The protein kinase complement of the human genome. Science 298, 1912–1934. 17- Hanks, S.K., A.M. Quinn, and T. Hunter, (1988) The protein kinase family: conserved features and deduced phylogeny of the catalytic domains. Science,. 241(4861): p. 42-52. 18- Knight JDR, Qian B, Baker D, Kothary R (2007) Conservation, Variability and the Modeling of Active Protein Kinases. PLoS ONE 2(10): e982. 19- Masterson LR, et al. (2011) Dynamics connect substrate recognition to catalysis in protein kinase A. Nat Chem Biol. May;7(5):319

20- Srivastava AK et al., (2014). Synchronous Opening and Closing Motions Are Essential for cAMP-Dependent Protein Kinase A Signalling. Structure, 22(12);1735-43. 21- Taylor SS, Kornev AP (2011) Protein kinases: Evolution of dynamic regulatory proteins. Trends Biochem Sci 36(2):65–77. 22- Kannan, N., et al., (2008)The hallmark of AGC kinase functional divergence is its C- terminal tail, a cis-acting regulatory module .Proceedings of the National Academy of Sciences of the United States of America, 2008. 105(26): p. 9130-9130. 23- Johnson DA, Akamine P, Radzio-Andzelm E, Madhusudan M, Taylor SS (2001) Dynamics of cAMP-dependent protein kinase. Chem Rev 101(8):2243–2270. 24- Huse M, Kuriyan J (2002) The conformational plasticity of protein kinases. Cell 109(3): 275–282. 25- Taylor SS. Et al, (1993) cAMP-dependent protein kinase defines a family of enzymes Philos Trans R Soc Lond B Biol Sci. Jun 29;340(1293):315-2. 26- Taylor et al. (2004) PKA: a portrait of protein kinase dynamics. Biochimica et Biophysica Acta 1697, 259–269 27- Taylor et al, (1999) Catalytic Subunit of Cyclic AMP-Dependent Protein Kinase: Structure and Dynamics of the Active Site Cleft Pharmacol. Ther. Vol. 82, Nos. 2–3, pp. 133–141. 28- Kannan N, Haste N, Taylor SS, Neuwald AF. (2007) The hallmark of AGC kinase functional divergence is its C-terminal tail, a cis-acting regulatory module. Proc Natl Acad Sci U S A.; 104:1272–1277. 29- Johnson DA. et al, (2001) Dynamics of cAMP-dependent protein kinase. Chem Rev. 101(8):2243-70 30- Tillo et al, (2017). Liberated PKA catalytic subunits associate with the membrane via myristoylation to preferentially phosphorylate membrane substrates. Cell Rep. Apr 18; 19(3): 617–629. 31- Gaffarogullari et al, (2011). A MYRISTOYL/PHOSPHOSERINE SWITCH CONTROLS cAMP-DEPENDENT PROTEIN KINASE ASSOCIATION TO MEMBRANES. J Mol Biol.August 26; 411(4): 823–836. 32- Taylor SS. Et al, (2013). PKA: Lessons Learned after Twenty Years. Biochim Biophys Acta. July ; 1834(7): 1271–1278. 33- Seifert et al, (2002). Phosphorylation and Flexibility of Cyclic-AMP-Dependent Protein Kinase (PKA) Using 31P NMR SpectroscopyBiochemistry.41, 5968-5977 34- Cook, P. F.; Neville, M. E.; Vrana, K. E.; Hartl, F. T.; Roskoski, J., R. Biochemistry 1982, 21, 5794. 35- Grant, B.; Adams, J. A. Biochemistry (1996), 35, 2022. 36- JOSEPH A. ADAMS AND SUSAN S. TAYLOR , 1993.Divalent metal ions influence catalysis and active-site. Protein Science (1993), 2, 2177-2186 37- Joseph A.Adams,(2001). Kinetic and Catalytic Mechanisms of Protein Kinases. Chem. Rev., 101, 2271-2290. 38- Sarma, G. N., Kinderman, F. S., Kim, C., von Daake, S., Chen, L., Wang, B. C. & Taylor, S. S. (2010) Structure of D-AKAP2:PKA RI complex: insights into AKAP specificity and selectivity. Structure 18, 155–166. (doi:10.1016/j.str.2009.12.012) 39- Chen, Q., Lin, R. Y., and Rubin, C. S. (1997) J. Biol. Chem. 272, 15247–15257 40- Huang, L. J., Durick, K., Weiner, J. A., Chun, J., and Taylor, S. S. (1997) D-AKAP2, a novel protein kinase A anchoring protein with a putative RGS domain. Proc. Natl. Acad. Sci. U. S. A. 94, 11184–11189.

41- Banky et al, (1998). Dimerization/Docking Domain of the Type Ia Regulatory Subunit of cAMP-dependent Protein Kinase. THE JOURNAL OF BIOLOGICAL CHEMISTRY,Vol. 273, No. 52, Issue of December 25, pp. 35048–35055. 42- Wiley et al, (1999) Role of Regulatory Subunits and Protein Kinase Inhibitor (PKI) in Determining Nuclear Localization and Activity of the Catalytic Subunit of Protein Kinase A. The Journal of Biological Chemistry, Vol. 274, No. 10, Issue of March 5, pp. 6381–6387. 43- Beavo JA, Brunton LL.(2002) Cyclic nucleotide research – still expanding after half a century. Nat Rev Mol Cell Biol. 3:710–8. 44- Taskén K, Aandahl EM (2004) Localized effects of cAMP mediated by distinct routes of protein kinase A. Physiol Rev. 84(1):137-67. 45- Dessauer CW. (2009) Adenylyl cyclase – A-kinase anchoring protein complexes: the next dimension in cAMP signaling. Mol Pharmacol 76:935–41. 46- Beuschlein et al, (2014). On stitutive Activation of PKA Catalytic Subunit in Adrenal Cushing's Syndrome. N Engl J Med 2014; 370:1019-1028. 47- Stratakis, C. A. et al (2002). Protein-kinase a and human disease: The core of cAMP- dependent signaling in health and disease. Hormone and Metabolic Research, 34(4), 169-175. 48- Simon et al., (2015) Transcriptomic characterization of fibrolamellar hepatocellular carcinoma. PNAS November 3,112 (44) E5916-E5925. 49- Honeyman, J. N. et al. (2014) Detection of a recurrent DNAJB1-PRKACA chimeric transcript in fibrolamellar hepatocellular carcinoma. Science 343, 1010–1014. 50- Stratakis, C.A. et al. (2001) Clinical and molecular features of the Carney complex: diagnostic criteria and recommendations for patient evaluation. J. Clin. Endocrinol. Metab. 86, 4041–4046 3 51- Kirschner, L.S. et al. (2000) Genetic heterogeneity and spectrum of mutations of the PRKAR1A gene in patients with the Carney complex. Hum. Mol. Genet. 9, 3037–3046 4 52- Kirschner, L.S. et al. (2000) Mutations of the gene encoding the protein kinase A type I-α regulatory subunit in patients with the Carney complex. Nat. Genet. 26, 89–92. 53- Laxminarayana, D. and Kammer, G.M. (2000) mRNA mutations of type I protein kinase A regulatory subunit α in T lymphocytes of a subject with systemic lupus erythematosus. Int. Immunol. 12, 1521–1529. 54- Yu B, et al. (2012) Protein kinase A alterations in endocrine tumors. J Horm Metab Res.; 44(10):741-8. 55- Howe AK., (2004) Regulation of actin-based cell migration by cAMP/PKA. Biochim Biophys Acta. Jul 5; 1692(2-3):159-74. 56- Jiang P, Enomoto A, Takahashi M. (2009) Cell biology of the movement of breast cancer cells: intracellular signaling and the actin cytoskeleton. Cancer Lett. Nov 1; 284(2):122-30. 57- Shaikh D. et al, (2012) cAMP-dependent protein kinase is essential for hypoxia-mediated epithelial-mesenchymal transition, migration, and invasion in lung cancer cells. Cell Signal. 24(12):2396-406. 58- de Leeuw R et al., (2013) PKA phosphorylation redirects ERα to promoters of a unique gene set to induce tamoxifen resistance., Oncogene. Jul 25; 32(30):3543-51. 59- Mani S, Stratakis CA et al. (2006) Autoantibody cancer biomarker: extracellular protein kinase A. Cancer Res. ;66:8971-4. 60- Wang H, Li M, Lin W, Wang W, Zhang Z, Rayburn ER et al. (2007) Extracellular activity of cyclic AMP-dependent protein kinase as a biomarker for human cancer detection: distribution characteristics in a normal population and cancer patients. Cancer Epidemiol Biomarkers Prev; 16:789-95. 61- Bax A. (1994) Multidimensional nuclear magnetic resonance methods for protein studies. Curr Opin Struc Biol 4, 738–744.

62- Mittermaier A. K., Kay L. E. (2009) Observing biological dynamics at atomic resolution using NMR. Trends Biochem. Sci. 34, 601–611. 63- Pellecchia M., et al. (2008) Perspectives on NMR in drug discovery: a technique comes of age. Nature Reviews Drug Discovery 7, 738–745. 64- Dominique Marion (2013) An Introduction to Biological NMR Spectroscopy. Molecular & Cellular Proteomics November 1, 2013, First published on July 6, 12 (11) 3006-3025. 65- Pellecchia M., et al. (2008) Perspectives on NMR in drug discovery: a technique comes of age. Nature Reviews Drug Discovery 7, 738–745. 66- Aue W. P., Bartholdi E., Ernst R. R. (1976) Two-dimensional spectroscopy. Application to nuclear magnetic resonance, J. Chem. Phys. 64, 2229–2246. 67- Davis D. G., Bax A.(1985) Assignment of complex 1H NMR spectra via two-dimensional homonuclear Hartmann-Hahn spectroscopy. J. Am. Chem. Soc. 107, 2820–2821. 68- Kumar A., Ernst R. R., Wüthrich K. (1980) A two-dimensional nuclear Overhauser enhancement (2D NOE) experiment for the elucidation of complete proton-proton cross- relaxation networks in biological macromolecules. Biochem. Biophys. Res. Commun. 95, 1– 6. 69- Bodenhausen G., Ruben D. J. (1980) Natural abundance nitrogen-15 NMR by enhanced heteronuclear spectroscopy. Chem Phys Lett 69, 185–189. 70- Kwan A. H., Mobli M., Gooley P. R., King G. F., Mackay J. P. (2011) Macro-molecular NMR spectroscopy for the non-spectroscopist. FEBS Journal 278, 687–703. 71- Bax A., Clore G. M., Driscoll P. C., Gronenborn A. M., Ikura M., Kay L. E. (1990) Practical aspects of proton-carbon-carbon-proton 3-dimensional correlation spectroscopy of 13C- labeled proteins. J. Magn. Reson. 87, 620–627. 72- Dominique Marion (2013) An Introduction to Biological NMR Spectroscopy. Mol Cell Proteomics;12:3006-3025. 73- Otting G. (2008) Prospects for lanthanides in structural biology by NMR. J. Biomol. NMR 42, 1–9. 74- Kleckner I. R., Foster M. P. (2011) An introduction to NMR-based approaches for measuring protein dynamics. BBA - Proteins and Proteomics 1814, 942–968. 75- Williamson M. P. (2009) Applications of the NOE in Molecular Biology. Annu. Rep. NMR Spectrosc. 65, 77–109. 76- Demers J. P., Mittermaier A. K. (2009) Binding Mechanism of an SH3 Domain Studied by NMR and ITC. J. Am. Chem. Soc. 131, 4355–4367. 77- Skrynnikov N. R., Dahlquist F. W., Kay L. E. (2002) Reconstructing NMR spectra of ”invisible“ excited protein states using HSQC and HMQC experiments. J. Am. Chem. Soc. 124, 12352–12360. 78- Demers J. P., Mittermaier A. K. (2009) Binding Mechanism of an SH3 Domain Studied by NMR and ITC. J. Am. Chem. Soc. 131, 4355–4367. 79- Hansen D. F., Vallurupalli P., Lundström P., Neudecker P., Kay L. E. (2008) probing chemical shifts of invisible states of proteins with relaxation dispersion NMR spectroscopy: How well can we do? J. Am. Chem. Soc. 130, 2667–2675. 80- Lipari G., Szabo A.(1982) Model-free approach to the interpretation of nuclear magnetic relaxation in macromolecules: 1. Theory and range of validity. J. Am. Chem. Soc. 104, 4546– 4559. 81- Choy W. Y., Shortle D., Kay L. E. (2003) Side chain dynamics in unfolded protein states: an NMR based 2H spin relaxation study of D131D. J. Am. Chem. Soc. 125, 1748–1758. 82- J. R. Craig, R. L. Peters, H. A. Edmondson, and M. Omata, (1980) “Fibrolamellar carcinoma of the liver: a tumor of adolescents and young adults with distinctive clinico-pathologic features,” Cancer, vol. 46, no. 2, pp. 372–379.

83- Weeda VB.et al., (2013) Fibrolamellar variant of hepatocellular carcinoma does not have a better survival than conventional hepatocellular carcinoma--results and treatment recommendations from the Childhood Liver Tumour Strategy Group (SIOPEL) experience. Eur J Cancer. Aug;49(12):2698-704. 84- Katzenstein HM et al.,(2003) Fibrolamellar hepatocellular carcinoma in children and adolescents. Cancer. Apr 15;97(8):2006-12. 85- Mavros MN, Mayo SC, Hyder O, Pawlik TM. (2012) A systematic review: treatment and prognosis of patients with fibrolamellar hepatocellular carcinoma. J Am Coll Surg. Dec;215(6):820-30.. 86- Graham RP et al, (2015) DNAJB1-PRKACA is specific for fibrolamellar carcinoma. Mod Pathol. 2015 Jun;28(6):822-9 87- Kastenhuber et al.,(2017) DNAJB1-PRKACA fusion kinase interacts with β-catenin and the liver regenerative response to drive fibrolamellar hepatocellular carcinoma. Proc Natl Acad Sci U S A. Dec 12;114(50):13076-13084. 88- Tomasini, M.D., Wang, Y., Karamafrooz, A. et al. (2018). Conformational Landscape of the PRKACA-DNAJB1 Chimeric Kinase, the Driver for Fibrolamellar Hepatocellular Carcinoma. Sci Rep 8, 720. 89- Cembran et al, (2014) NMR mapping of protein conformational landscapes using coordinated behavior of chemical shifts upon ligand binding. Phys. Chem. Chem. Phys. 16, 6508–6518. 90- R. Selvaratnam et al, (2011) Mapping allostery through the covariance analysis of NMR chemical shifts Proc. Natl. Acad. Sci. U.S.A. 108,6133–6138 91- Boulton et al, (2014) A Tool Set to Map Allosteric Networks through the NMR Chemical Shift Covariance Analysis Scientific Reports volume 4, Article number: 7306 92- Stark JL, et al. (2014) Structure and function of human DnaJ homologue subfamily a member 1 (DNAJA1) and its relationship to pancreatic cancer. Biochemistry. 53:1360–1372 93- Nallamsetty S et al., (2004) Efficient site-specific processing of fusion proteins by tobacco vein mottling virus protease in vivo and in vitro. Protein Expr Purif. Nov;38(1):108-15. 94- Delaglio et al (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J Biomol NMR. Nov;6(3):277-93. 95- Zheng, J. et al. (1993). 2.2 A refined crystal structure of the catalytic subunit of cAMP- dependent protein kinase complexed with MnATP and a peptide inhibitor. Acta Crystallogr. D. Biol. Crystallogr. 49, 362–5 . 96- Hess, B., Kutzner, C., van der Spoel, D. & Lindahl, E. (2008). GROMACS 4: Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation. J. Chem. Theory Comput. 4, 435–447. 97- Best, R. B. et al. (2012) Optimization of the additive CHARMM all-atom protein force field targeting improved sampling of the backbone φ, ψ and side-chain χ(1) and χ(2) dihedral angles. J. Chem. Theory Comput. 8, 3257–3273. 98- Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. (1983) Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. 79, 926 99- Hess, B., Bekker, H., Berendsen, H. J. C. & Fraaije, J. G. E. M. (1997).LINCS: A linear constraint solver for molecular simulations. J. Comput. Chem. 18, 1463–1472. 100- Darden, T., York, D. & Pedersen, L. (1993). Particle mesh Ewald: An N log(N) method for Ewald sums in large systems. J. Chem. Phys. 98, 10089. 101- Parrinello, M. & Rahman, A. (1980). Crystal Structure and Pair Potentials:⋅ A Molecular- Dynamics Study. Phys. Rev. Lett. 45, 1196–1199. 102- Pandini, A., Fornili, A., Fraternali, F. & Kleinjung, J. (2012) Detection of allosteric signal transmission by information-theoretic analysis of protein dynamics. FASEB J. 26, 868–81

103- Pandini, A., Fornili, A., Fraternali, F. & Kleinjung, J. (2013).GSATools: analysis of allosteric communication and functional local motions using a structural alphabet. Bioinformatics 29, 2053–5. 104- Pasi, M., Tiberti, M., Arrigoni, A. & Papaleo, E. (2012) xPyder: a PyMOL plugin to analyze coupled residues and their networks in protein structures. J. Chem. Inf. Model. 52, 1865–74 105- Cheung, J. et al. (2015) Structural insights into mis-regulation of protein kinase A in human tumors. Proc. Natl. Acad. Sci. U. S. A. 112, 1374–9. 106- Torbenson M. (2012) Fibrolamellar Carcinoma: Update. Scientifica (Cairo). 2012;2012:1– 15. 107- Katzenstein HM, et al. Fibrolamellar hepatocellular carcinoma in children and adolescents. Cancer. 2003;97:2006–2012. 108- Weeda VB, et al.(2013) Fibrolamellar variant of hepatocellular carcinoma does not have a better survival than conventional hepatocellular carcinoma–results and treatment recommendations from the Childhood Liver Tumour Strategy Group (SIOPEL) experience. Eur. J. Cancer.;49:2698–2704. 109- Honeyman JN, et al.(2014) Detection of a recurrent DNAJB1-PRKACA chimeric transcript in fibrolamellar hepatocellular carcinoma. Science.;343:1010–1014. 110- Darcy DG, et al.(2015) The genomic landscape of fibrolamellar hepatocellular carcinoma: whole genome sequencing of ten patients. Oncotarget. 6:755–770. 111- Simon EP, et al. (2015) Transcriptomic characterization of fibrolamellar hepatocellular carcinoma. Proc. Natl. Acad. Sci. USA.112:E5916–25. 112- Kastenhuber, E. R.et al.(2017) DNAJB1–PRKACA fusion kinase interacts with β-catenin and the liver regenerative response to drive fibrolamellar hepatocellular carcinoma. Proc. Natl. Acad. Sci. 114 13076–13084 113- Shabb JB. (2001) Physiological substrates of cAMP-dependent protein kinase. Chem. Rev. 101:2381–2411. 114- Walsh DA, Van Patten SM.(1994) Multiple pathway signal transduction by the cAMP- dependent protein kinase. FASEB J. Off. Publ. Fed. Am. Soc. Exp. Biol.;8:1227–1236. 115- Knighton DR, et al. (1991) Crystal structure of the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science. 253:407–414. 116- Zheng J, et al. (1993) Crystal structures of the myristylated catalytic subunit of cAMP- dependent protein kinase reveal open and closed conformations. Protein Sci. 2:1559–1573. 117- Knighton DR, et al. (1991) Structure of a peptide inhibitor bound to the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science (80-.).253:414–420. 118- Cembran A, et al. (2012) Conformational equilibrium of N-myristoylated cAMP-dependent protein kinase A by molecular dynamics simulations. Biochemistry.51:10186–10196. 119- Gaffarogullari EC, et al. (2011) A myristoyl/phosphoserine switch controls cAMP- dependent protein kinase association to membranes. J. Mol. Biol.411:823–836. 120- Bastidas AC, et al. (2012) Role of N-terminal myristylation in the structure and regulation of cAMP-dependent protein kinase. J. Mol. Biol.422:215–229. 121- Zhang P, et al. (2015) An Isoform specific Myristylation Switch Targets Type II PKA Holoenzymes to Membranes. Structure (London, England: 1993) 23:1563–1572. 122- Cheung J, et al. (2015) Structural insights into mis-regulation of protein kinase A in human tumors. Proc. Natl. Acad. Sci. USA.112:1374–1379. 123- Kornev AP, Taylor SS. (2015) Dynamics-Driven Allostery in Protein Kinases. Trends Biochem. Sci.40:628–647. 124- Masterson LR, et al.(2011) Dynamically committed, uncommitted, and quenched states encoded in protein kinase A revealed by NMR spectroscopy. Proc. Natl. Acad. Sci. USA.108:6969–6974.

125- Masterson LR, et al. (2010) Dynamics connect substrate recognition to catalysis in protein kinase A. Nat. Chem. Biol.6:821–828. 126- Zhang P, et al. (2012) Structure and allostery of the PKA RIIbeta tetrameric holoenzyme. Science. 335:712–716. 127- Kim J, et al. (2015) Dysfunctional conformational dynamics of protein kinase A induced by a lethal mutant of phospholamban hinder phosphorylation. Proc. Natl. Acad. Sci. 112:3716– 3721. 128- Pervushin K, Riek R, Wider G, Wuthrich K. (1997) Attenuated T2 relaxation by mutual cancellation of dipole-dipole coupling and chemical shift anisotropy indicates an avenue to NMR structures of very large biological macromolecules in solution. Proc. Natl. Acad. Sci. USA. 94:12366–12371. 129- Ikura M, Kay LE, Bax A.(1990) A novel approach for sequential assignment of 1H, 13C, and 15N spectra of proteins: heteronuclear triple-resonance three-dimensional NMR spectroscopy. Application to calmodulin. Biochemistry.29:4659–4667. 130- Stark JL, et al. (2014) Structure and function of human DnaJ homologue subfamily a member 1 (DNAJA1) and its relationship to pancreatic cancer. Biochemistry. 53:1360–1372. 131- Farrow NA, Zhang O, Forman-Kay JD, Kay LE. (1994) A heteronuclear correlation experiment for simultaneous determination of 15N longitudinal decay and chemical exchange rates of systems in slow equilibrium. J. Biomol. NMR. 4:727–734. 132- García de la Torre J, Huertas ML, Carrasco B. (2000) HYDRONMR: Prediction of NMR Relaxation of Globular Proteins from Atomic-Level Structures and Hydrodynamic Calculations. J. Magn. Reson.147:138–146. 133- Ahuja LG, Kornev AP, McClendon CL, Veglia G, Taylor SS. (2017) Mutation of a kinase allosteric node uncouples dynamics linked to phosphotransfer. Proc. Natl. Acad. Sci. USA. 114:E931–E940. 134- Srivastava AK, et al. (2014) Synchronous opening and closing motions are essential for cAMP-dependent protein kinase A signaling. Structure.22:1735–1743. 135- Bowers, K. et al. Scalable Algorithms for Molecular Dynamics Simulations on Commodity Clusters. in ACM/IEEE SC 2006 Conference (SC’06) 43–43 (IEEE, 2006). 136- Harder E, et al. (2016) OPLS3: A Force Field Providing Broad Coverage of Drug-like Small Molecules and Proteins. J. Chem. Theory Comput.12:281–296. 137- Daura X, et al. (1999) Peptide Folding: When Simulation Meets Experiment. Angew. Chemie Int. Ed. 38:236–240. 138- Nallamsetty S, et al. (2004) Efficient site-specific processing of fusion proteins by tobacco vein mottling virus protease in vivo and in vitro. Protein Expr. Purif. 38:108–115. 139- Yonemoto WM, McGlone ML, Slice LW, Taylor SS. (1991) Protein phosphorylation. Part A. Protein kinases: assays, purification, antibodies, functional analysis, cloning, and expression. Methods Enzymol. 200:1–763 140- Grzesiek S, Bax A. (1992) Correlating backbone amide and side chain resonances in larger proteins by multiple relayed triple resonance NMR. J. Am. Chem. Soc.114:6291–6293. 141- Wittekind M, Mueller L. HNCACB (1993) a High-Sensitivity 3D NMR Experiment to Correlate Amide-Proton and Nitrogen Resonances with the Alpha- and Beta-Carbon Resonances inProteins. J. Magn. Reson. Ser. B. 101:201–205. 142- Sze KH. (2000) Protein dynamics measurements by TROSY-based NMR experiments. J. Magn. Reson.143:423–426. 143- Delaglio F, et al. (1995) NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR.6:277–293. 144- Eggert T, McGlynn KA, Duffy A,Manns MP, Greten TF, Altekruse SF. (2013) Fibrolamellar hepatocellular carcinoma in the USA, 2000–2010: A detailed report on

frequency, treatment and outcome based on the surveillance, epidemiology, and end results database. United European Gastroenterol J; 1:351–357. 145- Torbenson M. (2012) Fibrolamellar carcinoma: date. Scientifica (Cairo) 2012;2012:743790. 146- Ang CS, et al. (2013) Clinicopathologic characteristics and survival outcomes of patients with fibrolamellar carcinoma: Data from the fibrolamellar carcinoma consortium. Gastrointest Cancer Res; 6:3–9. 147- Jesper B. Andersen, (2015) Fibrolamellar Hepatocellular Carcinoma: A Rare but Distinct Type of Liver Cancer. Gastroenterology. Volume 148, Issue 4, Pages 707-710 148- Honeyman JN et al., (2014) Detection of a recurrent DNAJB1-PRKACA chimeric transcript in fibrolamellar hepatocellular carcinoma. Science. 343:1010–1014. 149- Darcy, D. G. et al. (2015). The genomic landscape of fibrolamellar hepatocellular carcinoma: whole genome sequencing of ten patients. Oncotarget 6, 755–770 150- Blume-Jensen, P.; Hunter, T. (2001) Oncogenic kinase signalling. Nature, 411 (6835), 355−65. 151- Hunter, T. (2009) Tyrosine phosphorylation: thirty years and counting. Curr. Opin. Cell Biol., 21 (2), 140−6. 152- Ferguson, F. M.; Gray, N. S. (2018) Kinase inhibitors: the road ahead. Nat. Rev. Drug Discovery 17 (5), 353−377. 153- New drugs at FDA: CDER’s new molecular entities and new therapeutic biological products. https://www.fda.gov/Drugs/ DevelopmentApprovalProcess/DrugInnovation/default.htm (January 2019). 154- Sapio et al, (2014) Targeting protein kinase A in cancer therapy: an update. EXCLI Journal 13:843-855 155- Bodenmiller B, Aebersold R (2010) Quantitative analysis of protein phosphorylation on a system-wide scale by mass spectrometry-based proteomics. Methods Enzymol. 470:317–334. 156- Morandell S, et al. (2010) QIKS—quantitative identification of kinase substrates. Proteomics 10:2015–2025. 157- Xue, L.; Geahlen, R. L.; Tao, W. A.(2013) Identification of Direct Tyrosine Kinase Substrates Based on Protein Kinase Assay-Linked Phosphoproteomics. Mol. Cell. Proteomics, 12 (10), 2969−2980. 158- Xue, L.et al, (2012) Sensitive kinase assay linked with phosphoproteomics for identifying direct kinase substrates. Proc. Natl. Acad. Sci. U. S. A., 109 (15), 5615−20. 159- Antharavally et al. (2013) A versatile mass spectrometry sample preparation procedure for complex protein samples. 160- Torkamani A, Verkhivker G, Schork NJ.(2009) Cancer driver mutations in protein kinase genes. Cancer Lett.;281(2):117–127. 161- Yu S, Huang H, Iliuk A, Wang WH, Jayasundera KB, Tao WA et al.(2013) Syk inhibits the activity of protein kinase A by phosphorylating tyrosine 330 of the catalytic subunit. J Biol Chem. 288:10870-81. 162- Kostenko S, Shiryaev A, Dumitriu G, Gerits N, Moens U.(2011) Cross-talk between protein kinase A and the MAPK-activated protein kinases RSK1 and MK5. J Recept Signal Transduct Res. 31:1-9. 163- Follin-Arbelet V, Torgersen ML, Naderi EH, Misund K, Sundan A, Blomhoff HK.(2013) Death of multiple myeloma cells induced by cAMP-signaling involves downregulation of Mcl-1 via the JAK/STAT pathway. Cancer Lett; 335:323-31. 164- Awada A, Aftimos PG. (2013) Targeted therapies of solid cancers: new options, new challenges. Curr Opin Oncol. 25:296-304.

165- Colzani M, Noberini R, Romanenghi M, Colella G, Pasi M, Fancelli D et al. (2014) Quantitative chemical proteomics identifies novel targets of the anti-cancer multikinase inhibitor E-3810. Mol Cell Proteomics. 13:1495-509. 166- Mayr B, Montminy M.(2001) Transcriptional regulation by the phosphorylation-dependent factor CREB. Nat Rev Mol Cell Biol;2:599-609. 167- Siu YT1, Jin DY. (2007) CREB--a real culprit in oncogenesis. FEBS J. Jul;274(13):3224-32. 168- André Steven and Barbara Seliger, (2016) Control of CREB expression in tumors: from molecular mechanisms and signal transduction pathways to therapeutic target. Oncotarget. 7:35454-35465. 169- Sakamoto KM, Frank DA. (2009) CREB in the pathophysiology of cancer: implications for targeting transcription factors for cancer therapy. Clin Cancer Res. 15:2583-7. 170- Sonenberg N, Hinnebusch AG. (2009). Regulation of translation initiation in eukaryotes: mechanisms and biological targets. Cell 136:731–745. 171- . Ma XM, Blenis J. (2009). Molecular mechanisms of mTOR-mediated translational control. Nat Rev Mol Cell Biol 10:307–318. 172- Wek RC, Jiang HY, Anthony TG. (2006). Coping with stress: eIF2 kinases and translational control. Biochem Soc Trans 34:7–11. 173- M. Bhat, N. Robichaud, L. Hulea, N. Sonenberg, J. Pelletier, and I. Topisirovic,(2015) “Targeting the translation machinery in cancer,” Nature Reviews Drug Discovery, vol. 14, no. 4, pp. 261–278. 174- J. Kong and P. Lasko, (2012) “Translational control in cellular and developmental processes,” Nature Reviews Genetics, vol. 13, no. 6, pp. 383–394. 175- D. Ruggero, (2013) “Translational control in cancer etiology,” Cold Spring Harbor Perspectives in Biology, vol. 5, no. 2. 176- Wang X, Li W, Williams M, Terada N, Alessi DR, Proud CG. (2001). Regulation of elongation factor 2 kinase by p90(RSK1) and p70 S6 kinase. EMBO J 20:4370 – 4379. 177- Kenney JW, Moore CE, Wang X, Proud CG. (2014). Eukaryotic elongation factor 2 kinase, an unusual enzyme with multiple roles. Adv Biol Regul 55:15–27. 178- R. Spilka, C. Ernst, A. K. Mehta, and J. Haybaeck,(2013) “Eukaryotic translation initiation factors in cancer development and progression,” Cancer Letters, vol. 340, no. 1, pp. 9–21. 179- Q. Zheng, J. Ye, and J. Cao, (2014). “Translational regulator eIF2α in tumor,” Tumor Biology, vol. 35, no. 7, pp. 6255–6264. 180- S. Wang, I. B. Rosenwald, M. J. Hutzler et al.,(1999) “Expression of the eukaryotic translation initiation factors 4E and 2α in non-Hodgkin's lymphomas,” The American Journal of Pathology, vol. 155, no. 1, pp. 247–255. 181- M. V. T. Lobo, M. E. Martín, M. I. Pérez et al.,(2000) “Levels, phosphorylation status and cellular localization of translational factor eIF2 in gastrointestinal carcinomas,” Histochemical Journal, vol. 32, no. 3, pp. 139–150. 182- S. Tejada, M. V. T. Lobo, M. García-Villanueva et al.,(2009) “Eukaryotic initiation factors (eIF) 2α and 4E expression, localization, and phosphorylation in brain tumors,” Journal of Histochemistry and Cytochemistry, vol. 57, no. 5, pp. 503–512. 183- De Luca A, Maiello MR, D’Alessio A, Pergameno M, Normanno N.(2012) The RAS/RAF/MEK/ERK and the PI3K/AKT signalling pathways: role in cancer pathogenesis and implications for therapeutic approaches. Expert Opin Ther Targets. 16:17–27. 184- Germann UA et al, (2017) Targeting the MAPK Signaling Pathway in Cancer: Promising Preclinical Activity with the Novel Selective ERK1/2 Inhibitor BVD-523. Mol Cancer Ther. Nov;16(11):2351-2363. 185- Velasco et al., (2016) The use of cannabinoids as anticancer agents. Progress in Neuro- Psychopharmacology and Biological Psychiatry Volume 64, 4 January, pp 259-266.

186- Malfitano et al., (2011). update on the endocannabinoid system as an anticancer target Expert Opin Ther Targets, 15 (3) (2011), pp. 297-308. 187- Bandana Chakravarti, Janani Ravi and Ramesh K. Ganju, Oncotarget. (2014) Cannabinoids as therapeutic agents in cancer: current status and future implications Aug; 5(15): 5852–5872. 188- Bouwman P, Jonkers J., (2012) The effects of deregulated DNA damage signalling on cancer chemotherapy response and resistance. Nat Rev Cancer. Sep; 12(9):587-98. 189- LaTulippe E, Satagopan J, Smith A, Scher H, Scardino P, Reuter V, Gerald WL., (2002) Comprehensive gene expression analysis of prostate cancer reveals distinct transcriptional programs associated with metastatic disease. Cancer Res. Aug 1; 62(15):4499-506. 190- Wei Q, Cheng L, Xie K, Bucana CD, Dong Z. J., (1997) Direct correlation between DNA repair capacity and metastatic potential of K-1735 murine melanoma cells. Invest Dermatol. Jan; 108(1):3-6. 191- Sarasin A, Kauffmann A., (2008) Overexpression of DNA repair genes is associated with metastasis: a new hypothesis. Mutat Res. Jul-Aug; 659(1-2):49-55. 192- A Kauffmann et al., (2008) High expression of DNA repair pathways is associated with metastasis in melanoma patients., Oncogene volume27, pages565–573. 193- Zbyslaw Sondka, Sally Bamford, Charlotte G. Cole, Sari A. Ward, Ian Dunham & Simon A. Forbes., (2018). The COSMIC Cancer Gene Census: describing genetic dysfunction across all human cancers. Nature Reviews Cancervolume 18, pages696–705 194- Peng Xia, Yuening Liu, and Zhaokang Cheng. (2016) Signaling Pathways in Cardiac Myocyte Apoptosis. Biomed Res Int. 195- Kastenhuber et al, (2017) DNAJB1–PRKACA fusion kinase interacts with β-catenin and the liver regenerative response to drive fibrolamellar hepatocellular carcinoma PNAS December 12, 114 (50) 13076-13084 196- Stewart P.A. et al. (2019) The Galaxy Platform for Reproducible Affinity Proteomic Mass Spectrometry Data Analysis. In: Evans C., Wright P., Noirel J. (eds) Mass Spectrometry of Proteins. Methods in Molecular Biology, vol 1977. Humana Press, New York, NY. 197- Perez et al,(2019) High-throughput Identification of FLT3 Wild-type and Mutant Kinase Substrate Preferences and Application to Design of Sensitive In Vitro Kinase Assay Substrates. Mol Cell Proteomics. Mar;18(3):477-489 198- Wu et al,(2019) Nrf2 in cancers: A double‐edged sword. Cancer Medicine; 8:2252–2267 199- Panieri E. and Saso L., (2019) Potential Applications of NRF2 Inhibitors in Cancer Therapy. Oxidative Medicine and Cellular Longevity, Article ID 8592348. 200- EI-Serag, H.B.; Rudolph, K.L. (2007) Hepatocellular carcinoma: Epidemiology and molecular carcinogenesis. Gastroenterology, 132, 2557–2576. 201- Moon, R.T.; Bowerman, B.; Boutros, M.; Perrimon, N. (2002) The promise and perils of Wnt signaling through beta-catenin. Science, 296, 1644–1646. 202- Villanueva, A.; Newell, P.; Chiang, D.Y.; Friedman, S.L.; Llovet, J.M. (2007) Genomics and signaling pathways in hepatocellular carcinoma. Semin. Liver Dis., 27, 55–76. 203- Raghunath et al., (2018) Dysregulation of Nrf2 in Hepatocellular Carcinoma: Role in Cancer Progression and Chemoresistance. Cancers, 10, 481. 204- Consolato M. Sergi. (2016) Hepatocellular Carcinoma, Fibrolamellar Variant: Diagnostic Pathologic Criteria and Molecular Pathology date. A Primer. Diagnostics (Basel). Mar; 6(1): 3. 205- Taurin S. et al., Phosphorylation of b-Catenin by Cyclic AMP-dependent Protein Kinase.The Journal of Biological Science,Vol. 281, NO. 15, pp. 9971–9976. 206- Sastre-Perona, A. et al, (2016) β-catenin signaling is required for RAS-driven thyroid cancer through PI3K activation. Oncotarget.; 7:49435-49449.

207- Song Y.et al, (2017) The Wnt/β-catenin and PI3K/Akt signaling pathways promote EMT in gastric cancer by epigenetic regulation via H3 lysine 27 acetylation. Tumour Biol. Jul;39(7). 208- Ramakrishnan S. et al. (2014) Vascular Endothelial growth factor signaling in hypoxia and Inflammation. J Neuroimmune Pharmacol. Mar; 9(2): 142–160. 209- Attwooll C, Lazzerini Denchi E, Helin K. (2004) The E2F family: specific functions and overlapping interests. EMBO J.;23(24):4709–4716. 210- Ren B, Cam H, Takahashi Y, et al. E2F integrates cell cycle progression with DNA repair, replication, and G(2)/M checkpoints. Genes Dev.2002;16(2):245–256. 211- Dimova DK, Dyson NJ. (2005) The E2F transcriptional network: old acquaintances with new faces. Oncogene.;24(17):2810–2826 212- Yang M, Wu S, Jia J, May WS. (2011) JAZ mediates G1 cell cycle arrest by interacting with and inhibiting E2F1. Cell Cycle.;10 (14):2390–2399. 213- Jiang X, Nevins JR, Shats I, Chi JT. (2015) E2F1-mediated induction of NFYB attenuates apoptosis via joint regulation of a pro-survival transcriptional program. PLoS One.;10(6):e0127951. 214- Baiz D, Dapas B, Farra R, et al. (2014) Bortezomib effect on E2F and cyclin family members in human hepatocellular carcinoma cell lines. World J Gastroenterol.;20(3):795– 803. 215- Li W, Ni GX, Zhang P, Zhang ZX, Li W, Wu Q. (2010) Characterization of E2F3a function in HepG2 liver cancer cells. J Cell Biochem.;111(5):1244–1251. 216- Huang YL et al., (2019) Promising diagnostic and prognostic value of E2Fs in human hepatocellular carcinoma. Cancer Manag Res. Feb 19;11:1725-1740. 217- Simon E. et al, (2015) Transcriptomic characterization of fibrolamellar hepatocellular carcinoma. PNAS November 3, 112 (44) E5916-E5925. 218- Chua et al, (2017) Cancer-type dependent expression of CK2 transcripts. Plos One, December 4,. 219- Zhang HX, Jiang SS, Zhang XF, Zhou ZQ, Pan QZ, Chen CL, et al. (2015) Protein kinase CK2alpha catalytic subunit is overexpressed and serves as an unfavorable prognostic marker in primary hepatocellular carcinoma. Oncotarget.; 6(33):34800–17. 220- Roessler S, Jia HL, Budhu A, Forgues M, Ye QH, Lee JS, et al. (2010) A unique metastasis gene signature enables prediction of tumor relapse in early-stage hepatocellular carcinoma patients. Cancer Res.;70(24):10202–12. 221- Park et al, (2004) Modulation of β-Catenin Phosphorylation/Degradation by Cyclin- dependent Kinase 2, The Journal of Biological Chemistry. 279, 19592-19599. 222- Biaoxue, R., Xiguang, C., Hua, L. et al. (2016) Stathmin-dependent molecular targeting therapy for malignant tumor: the latest 5 years’ discoveries and developments. J Transl Med 14, 279 doi:10.1186/s12967-016-1000-z 223- Scmidt et al, Critical role for NF-κB-induced JunB in VEGF regulation and tumor angiogenesis. EMBO J. 2007 Feb 7; 26(3): 710–719. 224- Groot et al, (1991) Activation of junB by PKC and PKA signal transduction through a novel cis-acting element.Nucleic Acids Res. Feb 25; 19(4): 775–781. 225- Joseph A. Caruso, (2018) Low-Molecular-Weight Cyclin E in Human Cancer: Cellular Consequences and Opportunities for Targeted Therapies. Cancer Research. 226- Lis, J. T., Mason, P., Peng, J., Price, D. H. and Werner, J. (2000). P-TEFb kinase recruitment and function at heat shock loci. Genes Dev. 14, 792-803. 227- Dostmann WR. (RP)-(1995) cAMPS inhibits the cAMP-dependent protein kinase by blocking the cAMP-induced conformational transition. FEBS Lett.Nov 20;375(3):231-4. 228- Badireddy S. et al, (2010) Cyclic AMP analog blocks kinase activation by stabilizing inactive conformation: conformational selection highlights a new concept in allosteric

inhibitor design. Molecular & Cellular Proteomics March 1, 2011, First published on November 16,10 (3). 229- Tréfier A. et al, (2018) G Protein-Coupled Receptors As Regulators of Localized Translation: The Forgotten Pathway? Front Endocrinol (Lausanne). Feb 2;9:17. 230- Jun Zhang et al., (2007) Suppression of breast cancer metastasis through the inhibition of VEGF-mediated tumor angiogenesis. Cancer Ther. 5: 273–286. 231- Michael A. Harding a,d, Dan Theodorescu, RhoGDI (2010) signaling provides targets for cancer therapy. EUROPEAN JOURNAL OF CANCER 46 1252 – 1259. 232- Signaling Pathways Induced by G-protein-coupled Receptors .Enrique Rozengurt, in Physiology of the Gastrointestinal Tract (Fifth Edition), 2012. 233- Perez M. et al, (2018) Molecular & Cellular Proteomics March 1, 2019, First published on December 12,18 (3) 477-489. 234- Turnham et al.(2019) eLife;8:e44187. 235- Cao et al., (2019) Structures of the PKA RIα holoenzyme with the FLHCC driver J-PKAcα or wild type PKAcα. Structure 27, 816–828. 236- Liang, S, Zhou Chen, Y, Ke, G. (2014) Decreased expression of EIF4A1 after preoperative brachytherapy predicts better tumor-specific survival in cervical cancer. Int J Gynecol Cancer. 24: 908–915. 237- Heikkinen T. et al, (2013) Eukaryotic translation initiation factor 4E (eIF4E) expression is associated with breast cancer tumor phenotype and predicts survival after anthracycline chemotherapy treatment. Breast Cancer Res Treat. 2013 Aug;141(1):79-88. 238- Trigka, E.A., et al. (2013). A detailed immunohistochemical analysis of the PI3K/AKT/mTOR pathway in lung cancer: Correlation with PIK3CA, AKT1, K-RAS or PTEN mutational status and clinicopathological features. Oncology Reports, 30, 623-636. 239- Lyons, J. et al. (1990) Two G protein oncogenes in human endocrine tumors. Science 249, 655–659. 240- Parma, J. et al.(1993) Somatic mutations in the thyrotropin receptor gene cause hyperfunctioning thyroid adenomas. Nature 365, 649–651. 241- Vallar, L., Spada, A. & Giannattasio, G. (1987) Altered Gs and adenylate cyclase activity in human GH-secreting pituitary adenomas. Nature 330, 566–56. 242- Landis, C. A. et al. (1989) GTPase inhibiting mutations activate the alpha chain of Gs and stimulate adenylyl cyclase in human pituitary tumours. Nature 340,692–696. 243- Kirschner, L. S. et al. Mutations of the gene encoding the protein kinase A type I-alpha regulatory subunit in patients with the Carney complex. (2000) Nat. Genet. 26, 89–92. 244- Newell-Price J, Bertagna X, Grossman AB, Nieman LK. (2006) Cushing's syndrome. Lancet.367:1605–1617. 245- Lindholm J, Juul S, Jørgensen JO, et al. (2001) Incidence and late prognosis of cushing's syndrome: A population-based study. J Clin Endocrinol Metab. ; 86:117–123. 246- Arnaldi G, Mancini T, Tirabassi G, Trementino L, Boscaro M. (2012) Advances in the epidemiology, pathogenesis, and management of Cushing's syndrome complications. J Endocrinol Invest. 35:434–448. 247- Plotz CM, Knowlton AI, Ragan C. (1952) The natural history of Cushing's syndrome. Am J Med.13:597–614. 248- Beuschlein F, Fassnacht M, Assié G, et al. (2014) Constitutive Activation of PKA Catalytic Subunit in Adrenal Cushing's Syndrome. N Engl J Med.370:1019–1028.

100

249- Dalamzi et al., (2014) Novel Somatic Mutations in the Catalytic Subunit of the Protein Kinase A as a Cause of Adrenal Cushing's Syndrome: A European Multicentric Study, The Journal of Clinical Endocrinology & Metabolism, Volume 99, Issue 10, E2093–E2100. 250- Walker et al, (2019) Cushing’s syndrome driver mutation disrupts protein kinase A allosteric network, altering both regulation and substrate specificity. Sci. Adv. 2019;5: August 2019.

101

Figures:

Figure 1.1: Position of the catalytic and glycine rich loop in the conformation of the PRKACA.

Figure 1.2: Hydrophobic C-spine and R-spine in the conformation of PRKACA and their relative position compared to important loops in the PKA catalytic subunit.

102

Figure 1.3: Glycine-rich loop of the C subunit in its closed and open conformation

Figure 1.4: Post translational modifications in the N-terminal tail of PKA-Cα.

Figure 1.5: hydrogen bonding of the tip of the loop in the PKA-C:PKI (5– 24):ATP ternary complex. Hydrogen bonds are indicated by the dashed lines.

Figure 1.6: Reaction pathway for catalysis. The kinetic parameters for this pathway were obtained using a heptapeptide substrate.

103

Figure 1.7: Regulatory and catalytic subunit complex in PKA.

Figure 1.8: Sequence alignment of the hinge of different R subunit isoforms. Amino acid sequences immediately NH2-terminal to the autoinhibitory sites (underlined) in R subunits are compared with that of PKI.P (the phosphorylation site), P-3, P-4, P-5, P-6, P-11, and and P+1 sites are marked by arrows.(Image adapted from Xiaodong Cheng et al. J. Biol. Chem. 2001;276:4102-4108)

Figure 1.9: A kinase anchoring protein (AKAP) bound to the regulatory subunit and catalytic subunits pf PKA.

104

Figure 1.10: A schematic representation of cAMP signaling pathway inside the cell.

Adenylate cyclases (AC) produce cAMP from adenosin-tri-phosphate (ATP). High levels of cytosolic cAMP lead to activation of protein kinase A (PKA). PKA stimulation induces the phosphorylation of transcription factors, such as CREB, ICER/CREM, ATF-1, and CBP to drive camp-driven genes. Phosphodiesterase(PDE) decreases intracellular cAMP levels and counterbalances the intracellular cAMP effect. ATF, cAMP-dependent transcription factor; CBP, cAMP-binding protein; CNG, cyclic nucleotide- gated ion channel; CREB, cAMP response element-binding protein; ICER, inducible cAMP early repressor; P, phosphorylation.

Figure 1.11: A simplified scheme of the possible involvement of cAMP/PKA pathway in cancer. Adenylyl cyclase (AC) may be activated in several ways, including G-protein-coupled receptors (GPCR), leading to the formation of cAMP, that binds to the regulatory (R) subunits of PKA, thus releasing the catalytic subunits. PKA catalytic subunit in turn phosphorylates target proteins in cell and nucleus, leading to cellular responses.

105

Figure 3.1. PKA DANJB1 (Cyan, PDB:4WB7) crystal structure superimposed on PKAWT(Red, PDB:1ATP).

Open  Closed Open  Closed

Figure 3.2: Statistical analysis of the chemical shift changes in PKA DANJB1 (left) and PKAWT (right). CONCISE analysis: Plots of the probability density distributions versus the principal component 1 (PC1) of the chemical shift changes. The plot indicates the progression of all the chemical shifts of the PKADNAJB1 from the open (Apo) to the closed (Ternary) states.

106

PKADANJB1 in complex with ATPγN and inhibitor PKI PKADANJB1 in complex with ATPN and inhibitor PKI 5-24 FL

DANJB1 Figure 3.3- A: Chemical shift analysis of PKA in complex with PKI5- 24(left) vs PKIFL (right).

WT γ WT γ PKA in complex with ATP N and inhibitor PKI5-24 PKA in complex with ATP N and inhibitor PKIFL

WT Figure 3.3- B: Chemical shift analysis of PKA in complex with PKI5-24(left) vs PKIFL (right).

Figure 3.3: CHESCA correlation matrix showing the degree of correlated chemical shift changes for 15N-labeled residues. The correlation coefficient, rij, indicates the degree of correlated chemical shift changes along the linear trajectories.

107

Figure 3.4. Mutual information and allosteric networks in PKAWT and PKA DANJB1. (A) Mutual information matrices of WT. (B) Allosteric networks of WT. The correlation between residues are shown in green lines, and the key hubs are shown in blue spheres. (C) Mutual information matrices of Chimera (D) Allosteric networks of Chimera. C-Terminal, G-Loop (denoting Gly-rich Loop) and A-Loop (denoting Activation Loop) become central hubs and are labeled.

108

Figure 3.5. Emergence of allosteric hubs upon ATP binding in PKAWT and PKA DANJB1. (A). In WT, residues in the G-Loop and A-Loop emerge as allosteric hubs, suggesting the enhanced correlation throughout the catalytic core upon ATP binding. (B). In Chimera, residues primarily at the C-Terminal emerge as allosteric hubs, whereas the little changes occur at the catalytic core.

109

Table 3.1: Michaelis-Menten Kinetics constants for PKA-CDNAJB1and comparison with PKAWT

K /k -1 µ -1 cat m V (µM S ) K ( M) K (s ) -1 -1 max m cat µ ( S M ) PKA WT 0.20± 0.02 40±9 20 ± 2 0.50 PKA-CDNAJB1 0.49± 0.02 58±10 22.±2 0.38

Table 3.2: Midpoint of thermal denaturation (Tm) for PKA-CDNAJB1and comparison with PKAWT

Tm (°C) APO +ATP +ATP/PKI

PKAWT (Human) 53.2±0.5 55.03±0.73 60.21±0.43 PKA-CDNAJB1 45.48 ± 0.06 48.55 ± 0.08 52.20 ± 0.14

µ Table 3.3: Dissociation constants ( M) of nucleotides and inhibitor peptide PKI5-24

ATPγN/ ADP ATPγN PKI5-24 PKI5-24 PKA WT (Human) 24 ± 1 83.5 ± 8 6.1 ± 0.1 0.125 ± 0.005 PKA-CDNAJB1 24.7 ± 1 81 ± 15 3.74 ± 0.47 0.27 ± 0.03

110

Figure 4.1. Structures of J-PKAcα chimera and wild-type PKAcα. (a) J-PKAcα chimera (PDB ID: 4WB7) with the major domains labeled. (b) Wild-type PKAcα (PDB ID: 4DFX). The coloring scheme is as follows: blue = J-domain, Cyan = N-terminal A-helix, Green = Small Lobe, Yellow = Hinge between large and small lobes, Orange = Large Lobe, Red = C-terminal domain, Purple = PKI. ATP, Mg2+ ions, and the myristoylation are shown in licorice representation and colored according to atom.

111

Figure 4.2. RMSF per residue. RMSF for both J-PKACα chimera (a) and wild-type PKACα (b). Appended residues in the J-domain of J-PKACα are numbered negative with residue 15 the first residue that is common to both J-PKAcα chimera and wild-type PKACα. (c) Difference in RMSF (chimera – wild-type) beginning at residue 15 where the residues of the chimera and wild-type are equivalent. Positive numbers indicate greater fluctuations in residues of the chimera while negative numbers indicate greater fluctuation in the wild-type PKACα.

112

Figure 4.3. Movement of the J-domain in J-PKAcα chimera. (a) Vectors defined to compute the motions of the J-domain as expressed on the crystal structure. The straightness of the A-helix is given by θ1, the angle between vectors ν1 (Cα of Lys29 and Cα of Val15) and ν2 (Cα of Lys15 and Cα of Pro2). The up and down motion is given by θ2, the angle between vectors ν2 (Cα of Lys15 and Cα of Pro2) and ν3 (Cα of Pro2 and Cα of Lys−19). The azimuthal rotation of the J-domain is given by θ3 defined as the dihedral angle between the Cα atoms of Lys29 – Leu160 – Glu140 – Lys−19. (b) Scatterplots of θ1 vs θ3 for J-PKAcα chimera for the different binding modes. Lighter colors indicate angles sampled during the first 200 ns of simulation time while darker colors are from the final 800 ns. (c) Scatterplots of θ2 vs θ3 for J-PKAcα chimera for the different binding modes. Lighter colors indicate angles sampled during the first 200 ns of simulation time while darker colors are from the final 800 ns. (d) and (e) Motions of the J-domain for nine different 1 μs

113 simulations. θ1, θ2, and θ3, are as defined in Fig. 3. (d) θ1 vs θ3 (e) θ2 vs θ3. All data are from the final 800 ns of simulation time. The black x in the scatterplots corresponds to the crystal structure, PDB ID: 4WB7.

Figure 4.4. Top four clusters by population from cluster analysis. The coloring scheme is the same as in Fig. 1. Representative structures are the structure with the smallest average RMSD distance to every other member of the cluster. Cluster 1 comprises 62% of all conformations, Cluster 2 comprises 13% of conformations, Cluster 3 comprises 9% of conformations, and Cluster 4 comprises 7% of conformations. The cluster analysis found a total of 21 clusters.

114

Figure 4.5. Top clusters from cluster analysis modeled into the RIIβ holoenzyme. (a) Cluster 1 in a ‘J-in’ state and (b) Cluster 2 in a ‘J-out’ state from Fig. 4. The two RIIβ subunits are colored in orange and red, the two C-subunits without the J-domains are colored blue and gray, and the J- domain is colored in cyan.

115

Figure 4.6. Residue-specific T1/T2 ratio of NMR relaxation of J-PKACα. Experimental values of residues in J-domain are much smaller than the values of other domains, suggesting a significantly higher flexibility of the appendix.

116

Chapter VI Supporting Information:

Figure S1. Backbone assignment of PKA-DNAJB1: DNAJB11-69 amide fingerprint and side chains were assigned using a combination of 3D CBCACONH and HNCACB experiments. Resonance assignment of the amide fingerprint of PKA-DNAJB1 was carried out by transferring the previous assignments from the PKA-WT. Given the almost perfect overlay of the HSQC spectra with the DNAJB1, we transferred the assignments of the latter to the PKA-DNAJB1. A) backbone assignment of DNAJB1. B) backbone walk for residues Arg-13 to Asp-17 for Dnajb1 is shown. The correlations are HN(CA)CB and HN(COCA)CB.

117

Figure S2. (A)1H-15N TROSY-HSQC spectra for PKADNAJB1, with peaks from the DNAJB1 appendix colored in red, and peaks from the PKA-Cα in blue. (B). Crystal structure of the DNAJB1 appendix. (C). Crystal structure of PKADNAJB1.

118

Figure S3. Mutual information and allosteric networks in the binary (A) and ternary (B) forms of WT. The mutual information matrices is shown in the left, and the allosteric network and key hubs are shown in the right. The correlation between residues are shown in green lines, and the key hubs are shown in spheres. Residues throughout the catalytic core emerge as allosteric hubs in the binary and ternary form.

119

In vitro

Figure 5.1-A: schematic presentation of Kinase assay linked with phosphoproteomic (KALIP)

Figure 5.1-B: Overlap of the in-cell phospho-proteins for PRKACA-Dnajb1, PKACA and empty Vector (control). Each data set is cleared of background data(vector) and the other kinase dataset prior to further analysis.

PKA Empty vector Dnajb1-PKA PKA-WT

PKA-Dnajb1 “Omics” analysis tools (IPA*)

Vector

HEK293-T cells were Phosphorylation data sets Each data set is cleared of transiently transfected with for each species background data(vector) and plasmids containing Kinase the other kinase dataset. or empty plasmid.

120

-log(p-value) 25 0

Figure 5.2: Comparison of some of enriched canonical pathways derived from phosphorylation analysis of in-cell PRKACA and PRKACA-Dnajb1 phosphoenrichment profile (left) and pathways exclusive to PRKACA-Dnajb1 with no matching value in PRKACA (right).

121

Peptide Phosphorylation A C PRKACA 5000 4500 /T

S 4000 3500 3000 2500 2000 robability

1500 P 1000 500

Phosphosite Number Phosphosite 0

B Protein Phosphorylation 1400 D PRKACADnajb1 1200

1000 S/T 800 600 400 200 Probability Phosphosite Number 0

Figure 5.3: Comparison of Phosphorylation sites in each of PRKACA and its mutant PRKACA- Dnajb1 with the control reaction in peptide KALIP (A) and Protein KALIP (C). Phosphorylation motif derived from KALIP in vitro protein assay for wild type PKA (C) and PRKACA-Dnajb1 (D).

122

Figure 5.4: Comparison of canonical pathways derived from IPA phosphorylation analysis of PRKACA and PRKACA-Dnajb1 Peptide direct substrates, also present in in-cell phosphopeptides profile (peptide KALIP and in-cell) The arrows on the right column is indicative of pathways enriched in PRKACA-Dnajb1 relative to PKACA: green arrow presents a higher than 1.2 ratio, yellow dashes are between 1.17-1.0, and red arrows present a lower that 1:1 ratio.

123

Figure 5.5: Comparison of canonical pathways derived from IPA phosphorylation analysis of PRKACA and PRKACA-Dnajb1 protein direct substrates, also present in in-cell phosphopeptides profile (KALIP-protein and in-cell). The arrows on the right column is indicative of pathways enriched in PRKACA-Dnajb1 relative to PKACA: green arrow presents a higher than 1.2 ratio, yellow dashes are between 1.17-1.0, and red arrows present a lower that 1:1 ratio.

124

A B

100 80 60 40 20 0

Figure 5.6. A: pattern of downregulated phosphorylation sites in each species shown as the heatmap where red and black represent the phosphorylated and unphosphorylated substrates respectively. B: Comparison of phosphorylated substrates percentage in PRKACA- Dnajb1 in the presence of inhibitors PKI and rpcAMP.

125

Figure 5.7: Inhibition of canonical pathways derived from IPA phosphorylation analysis of PRKACA and PRKACA-Dnajb1 in HEK293 cell phosphopeptides profile.

126

Table 5.1: Subset of PRKACA-Dnajb1 direct peptide Substrates identified by KALIP

127

Table 5.2: subset of PRKACA-Dnajb1 direct Protein Substrates identified by KALIP

Table 5.3: Canonical pathways that are enriched in both protein and peptide KALIP

Canonical pathway Synaptogenesis Signaling Pathway  Telomere Extension by Telomerase  Regulation of Actin-based Motility by Rho  AMPK Signaling  Insulin Receptor Signaling  CDK5 Signaling  Prostate Cancer Signaling  Dopamine Receptor Signaling  Cell Cycle Control of Chromosomal Replication  ILK Signaling  mTOR Signaling  ERK/MAPK Signaling  Remodeling of Epithelial Adherens Junctions  Germ Cell-Sertoli Cell Junction Signaling  EIF2 Signaling 

128

Table 5.4: Subset of PRKACA-Dnajb1 substrates affected by the two PKA inhibitors

129

Chapter V supporting information

vector

Figure S.1: western blot of kinases expressed in HEK293 cells Primary antibody: anti human PKACA antibody, Secondary antibody: mouse/goat anti human antibody. Samples: PKA (WT), PRKACA-Dnajb1 (Chimera) and pcDNA3.1 (empty vector).

130

Table S1: Direct protein substrates exclusive to PRKACA-Dnajb1

ID Phospho Site Symbol Entrez Gene Name Location Type(s) API5_HUMAN 462 API5 apoptosis inhibitor 5 Cytoplasm other API5_HUMAN 464 API5 apoptosis inhibitor 5 Cytoplasm other BCKD_HUMAN 31 BCKDK branched chain ketoacid dehydrogenase kinase Cytoplasm kinase BCKD_HUMAN 32 BCKDK branched chain ketoacid dehydrogenase kinase Cytoplasm kinase BCKD_HUMAN 33 BCKDK branched chain ketoacid dehydrogenase kinase Cytoplasm kinase ICAL_HUMAN 133 CAST calpastatin Cytoplasm peptidase ICAL_HUMAN 216 CAST calpastatin Cytoplasm peptidase COF1_HUMAN 3 CFL1 cofilin 1 Nucleus other COF1_HUMAN 24 CFL1 cofilin 1 Nucleus other CTNA1_HUMAN 654 CTNNA1 catenin alpha 1 Plasma Membrane other CTNA1_HUMAN 655 CTNNA1 catenin alpha 1 Plasma Membrane other CTNA2_HUMAN 653 CTNNA2 catenin alpha 2 Plasma Membrane other CTNA2_HUMAN 654 CTNNA2 catenin alpha 2 Plasma Membrane other CTNB1_HUMAN 551 CTNNB1 catenin beta 1 Nucleus transcription regulator CTNB1_HUMAN 552 CTNNB1 catenin beta 1 Nucleus transcription regulator DCTP1_HUMAN 24 DCTPP1 dCTP pyrophosphatase 1 Cytoplasm enzyme DDX10_HUMAN 831 DDX10 DEAD-box helicase 10 Nucleus enzyme DTD1_HUMAN 196 DTD1 D-tyrosyl-tRNA deacylase 1 Cytoplasm enzyme DYHC1_HUMAN 1230 DYNC1H1 dynein cytoplasmic 1 heavy chain 1 Cytoplasm peptidase EF1B_HUMAN 106 EEF1B2 eukaryotic translation elongation factor 1 beta 2 Cytoplasm translation regulator EF1D_HUMAN 86 EEF1D eukaryotic translation elongation factor 1 delta Cytoplasm translation regulator EIF3C_HUMAN 39 EIF3C eukaryotic translation initiation factor 3 subunit C Cytoplasm translation regulator 4ET_HUMAN 351 EIF4ENIF1 eukaryotic translation initiation factor 4E nuclear import factor 1 Cytoplasm translation regulator 4ET_HUMAN 352 EIF4ENIF1 eukaryotic translation initiation factor 4E nuclear import factor 1 Cytoplasm translation regulator 4ET_HUMAN 353 EIF4ENIF1 eukaryotic translation initiation factor 4E nuclear import factor 1 Cytoplasm translation regulator H31T_HUMAN 29 HIST3H3 histone cluster 3 H3 Nucleus other ROA1_HUMAN 6 HNRNPA1 heterogeneous nuclear ribonucleoprotein A1 Nucleus enzyme HNRL2_HUMAN 161 HNRNPUL2 heterogeneous nuclear ribonucleoprotein U like 2 Nucleus other HS90B_HUMAN 226 HSP90AB1 heat shock protein 90 alpha family class B member 1 Cytoplasm enzyme HECT, UBA and WWE domain containing 1, E3 ubiquitin protein HUWE1_HUMAN 1084 HUWE1 Nucleus transcription regulator ligase IKKB_HUMAN 69 IKBKB inhibitor of nuclear factor kappa B kinase subunit beta Cytoplasm kinase ILF3_HUMAN 382 ILF3 interleukin enhancer binding factor 3 Nucleus transcription regulator ILKAP_HUMAN 379 ILKAP ILK associated serine/threonine phosphatase Cytoplasm phosphatase I2BP2_HUMAN 360 IRF2BP2 interferon regulatory factor 2 binding protein 2 Nucleus transcription regulator IRS4_HUMAN 933 IRS4 insulin receptor substrate 4 Plasma Membrane other IRS4_HUMAN 931 IRS4 insulin receptor substrate 4 Plasma Membrane other IRS4_HUMAN 1231 IRS4 insulin receptor substrate 4 Plasma Membrane other KH RNA binding domain containing, signal transduction associated KHDR1_HUMAN 18 KHDRBS1 Nucleus transcription regulator 1 LARP1_HUMAN 1040 LARP1 La ribonucleoprotein domain family member 1 Cytoplasm translation regulator LMNB1_HUMAN 158 LMNB1 lamin B1 Nucleus other MAP1B_HUMAN 1501 MAP1B microtubule associated protein 1B Cytoplasm other TAU_HUMAN 720 MAPT microtubule associated protein tau Plasma Membrane other TAU_HUMAN 721 MAPT microtubule associated protein tau Plasma Membrane other TAU_HUMAN 739 MAPT microtubule associated protein tau Plasma Membrane other MARCS_HUMAN 26 MARCKS myristoylated alanine rich protein kinase C substrate Plasma Membrane other MARCS_HUMAN 101 MARCKS myristoylated alanine rich protein kinase C substrate Plasma Membrane other MRP_HUMAN 22 MARCKSL1 MARCKS like 1 Cytoplasm other MATR3_HUMAN 150 MATR3 matrin 3 Nucleus other MATR3_HUMAN 598 MATR3 matrin 3 Nucleus other MATR3_HUMAN 759 MATR3 matrin 3 Nucleus other MCM2_HUMAN 41 MCM2 minichromosome maintenance complex component 2 Nucleus enzyme MDC1_HUMAN 448 MDC1 mediator of DNA damage checkpoint 1 Nucleus other MDC1_HUMAN 449 MDC1 mediator of DNA damage checkpoint 1 Nucleus other MDN1_HUMAN 4538 MDN1 midasin AAA ATPase 1 Nucleus other 131

MDN1_HUMAN 5017 MDN1 midasin AAA ATPase 1 Nucleus other MAP2_HUMAN 29 METAP2 methionyl aminopeptidase 2 Cytoplasm peptidase NUCL_HUMAN 69 NCL nucleolin Nucleus other NUCL_HUMAN 496 NCL nucleolin Nucleus other NUCL_HUMAN 498 NCL nucleolin Nucleus other NELFA_HUMAN 277 NELFA negative elongation factor complex member A Nucleus other MK67I_HUMAN 218 NIFK nucleolar protein interacting with the FHA domain of MKI67 Nucleus other NOL10_HUMAN 475 NOL10 nucleolar protein 10 Nucleus other NOP56_HUMAN 537 NOP56 NOP56 ribonucleoprotein Nucleus other NOP56_HUMAN 538 NOP56 NOP56 ribonucleoprotein Nucleus other NOP56_HUMAN 563 NOP56 NOP56 ribonucleoprotein Nucleus other NOP56_HUMAN 569 NOP56 NOP56 ribonucleoprotein Nucleus other NPM_HUMAN 243 NPM1 nucleophosmin 1 Nucleus transcription regulator NUMA1_HUMAN 1225 NUMA1 nuclear mitotic apparatus protein 1 Nucleus other PDCD4_HUMAN 93 PDCD4 programmed cell death 4 Nucleus other PININ_HUMAN 443 PNN pinin, desmosome associated protein Plasma Membrane other MYPT1_HUMAN 908 PPP1R12A protein phosphatase 1 regulatory subunit 12A Cytoplasm phosphatase MYPT1_HUMAN 852 PPP1R12A protein phosphatase 1 regulatory subunit 12A Cytoplasm phosphatase AAKB1_HUMAN 108 PRKAB1 protein kinase AMP-activated non-catalytic subunit beta 1 Nucleus kinase KAPCA_HUMAN 11 PRKACA protein kinase cAMP-activated catalytic subunit alpha Cytoplasm kinase KAPCA_HUMAN 49 PRKACA protein kinase cAMP-activated catalytic subunit alpha Cytoplasm kinase PRC2A_HUMAN 456 PRRC2A proline rich coiled-coil 2A Cytoplasm other PUM1_HUMAN 709 PUM1 pumilio RNA binding family member 1 Cytoplasm other PUM1_HUMAN 711 PUM1 pumilio RNA binding family member 1 Cytoplasm other PYGL_HUMAN 430 PYGL glycogen phosphorylase L Cytoplasm enzyme SPF45_HUMAN 169 RBM17 RNA binding motif protein 17 Nucleus other RL13_HUMAN 106 RPL13 ribosomal protein L13 Nucleus other RL13_HUMAN 107 RPL13 ribosomal protein L13 Nucleus other RL17_HUMAN 5 RPL17 ribosomal protein L17 Cytoplasm other RS3A_HUMAN 153 RPS3A ribosomal protein S3A Nucleus other RS3A_HUMAN 154 RPS3A ribosomal protein S3A Nucleus other RTN4_HUMAN 181 RTN4 reticulon 4 Cytoplasm other RTN4_HUMAN 182 RTN4 reticulon 4 Cytoplasm other SLTM_HUMAN 289 SLTM SAFB like transcription modulator Nucleus other U520_HUMAN 225 SNRNP200 small nuclear ribonucleoprotein U5 subunit 200 Nucleus enzyme RU2A_HUMAN 180 SNRPA1 small nuclear ribonucleoprotein polypeptide A' Nucleus other SNX6_HUMAN 316 SNX6 sorting nexin 6 Cytoplasm transporter SRP14_HUMAN 68 SRP14 signal recognition particle 14 Cytoplasm other SRRM2_HUMAN 1601 SRRM2 serine/arginine repetitive matrix 2 Nucleus other SRRM2_HUMAN 1600 SRRM2 serine/arginine repetitive matrix 2 Nucleus other F10A1_HUMAN 75 ST13 ST13 Hsp70 interacting protein Cytoplasm other F10A1_HUMAN 76 ST13 ST13 Hsp70 interacting protein Cytoplasm other F10A1_HUMAN 79 ST13 ST13 Hsp70 interacting protein Cytoplasm other TM230_HUMAN 24 TMEM230 transmembrane protein 230 Cytoplasm other TM230_HUMAN 25 TMEM230 transmembrane protein 230 Cytoplasm other TOP2B_HUMAN 1344 TOP2B DNA topoisomerase II beta Nucleus enzyme TRAD1_HUMAN 414 TRAFD1 TRAF-type zinc finger domain containing 1 Other other RMP_HUMAN 372 URI1 URI1 prefoldin like chaperone Nucleus transcription regulator RMP_HUMAN 373 URI1 URI1 prefoldin like chaperone Nucleus transcription regulator RMP_HUMAN 375 URI1 URI1 prefoldin like chaperone Nucleus transcription regulator WDR26_HUMAN 121 WDR26 WD repeat domain 26 Cytoplasm other WDR43_HUMAN 431 WDR43 WD repeat domain 43 Nucleus other XRCC6_HUMAN 477 XRCC6 X-ray repair cross complementing 6 Nucleus enzyme ZC3HE_HUMAN 325 ZC3H14 zinc finger CCCH-type containing 14 Nucleus other

132

Table S2: Direct peptide substrates of PRKACA-Dnajb1

Phospho ID Entrez Gene Name Symbol Site ACTB_HUMAN actin beta ACTB KILTERGYpSFTTTAERE 199 ACTB_HUMAN actin beta ACTB AGDDAPRAVFPpSIVGRPRH 33 ACTG_HUMAN actin gamma 1 ACTG1 AGFAGDDAPRAVFPpSIVGRPR 33 afadin, adherens junction formation 1082 AFAD_HUMAN factor AFDN AHNK_HUMAN AHNAK nucleoprotein AHNAK 1452 Bet1 golgi vesicular membrane 50 BET1_HUMAN trafficking protein BET1 CYBP_HUMAN calcyclin binding protein CACYBP 126 chaperonin containing TCP1 5 TCPB_HUMAN subunit 2 CCT2 chaperonin containing TCP1 193 TCPB_HUMAN subunit 2 CCT2 KCRB_HUMAN creatine kinase B CKB YALKSMpTEAEQQQL 180 cytoplasmic linker associated 313 CLAP2_HUMAN protein 2 CLASP2 PYRG1_HUMAN CTP synthase 1 CTPS1 573 DBNL_HUMAN drebrin like DBNL 269 DBNL_HUMAN drebrin like DBNL 270 DBNL_HUMAN drebrin like DBNL 271 DBNL_HUMAN drebrin like DBNL 272 defective in cullin neddylation 1 9 DCNL5_HUMAN domain containing 5 DCUN1D5 DDX27_HUMAN DEAD-box helicase 27 DDX27 565 DIS3 homolog, exosome endoribonuclease and 3'-5' 587 RRP44_HUMAN exoribonuclease DIS3 DnaJ heat shock protein family 16 DNJB1_HUMAN (Hsp40) member B1 DNAJB1 LARGApSDEEIKRAY DnaJ heat shock protein family 11 DNJC5_HUMAN (Hsp40) member C5 DNAJC5 RQRSLSpTSGESLYH eukaryotic translation elongation 140 EF1B_HUMAN factor 1 beta 2 EEF1B2 AKKPALVAKpSSILLD eukaryotic translation elongation 141 EF1B_HUMAN factor 1 beta 2 EEF1B2 AKKPALVAKSpSILLD eukaryotic translation elongation 484 EF2_HUMAN factor 2 EEF2 LVKTGpTITTF eukaryotic translation initiation 492 EIF3A_HUMAN factor 3 subunit A EIF3A eukaryotic translation initiation 316 EIF3B_HUMAN factor 3 subunit B EIF3B eukaryotic translation initiation 1077 IF4G1_HUMAN factor 4 gamma 1 EIF4G1 eukaryotic translation initiation 1080 IF4G1_HUMAN factor 4 gamma 1 EIF4G1 ELAV1_HUMAN ELAV like RNA binding protein 1 ELAVL1 41 ELAV1_HUMAN ELAV like RNA binding protein 1 ELAVL1 42 FKBP4_HUMAN FKBP prolyl isomerase 4 FKBP4 78 FKBP4_HUMAN FKBP prolyl isomerase 4 FKBP4 224 FLNA_HUMAN filamin A FLNA 343 high density lipoprotein binding 317 VIGLN_HUMAN protein HDLBP

133

heterogeneous nuclear 70 ROA2_HUMAN ribonucleoprotein A2/B1 HNRNPA2B1 heterogeneous nuclear 71 ROA2_HUMAN ribonucleoprotein A2/B1 HNRNPA2B1 heterogeneous nuclear 209 HNRPC_HUMAN ribonucleoprotein C HNRNPC heterogeneous nuclear 379 HNRPK_HUMAN ribonucleoprotein K HNRNPK heterogeneous nuclear 59 HNRPU_HUMAN ribonucleoprotein U HNRNPU heat shock protein 90 alpha family 169 HS90A_HUMAN class A member 1 HSP90AA1 AGGpSFTVRT heat shock protein 90 alpha family 63 HS90A_HUMAN class A member 1 HSP90AA1 YEpSLTDPSK heat shock protein 90 alpha family 211 HS90A_HUMAN class A member 1 HSP90AA1 VKKHpSQFIGY

heat shock protein 90 alpha family 641 HS90A_HUMAN class A member 1 HSP90AA1 INPDHpSIIETL heat shock protein 90 alpha family 164 HS90B_HUMAN class B member 1 HSP90AB1 AGGpSFTVRA heat shock protein 90 alpha family 206 HS90B_HUMAN class B member 1 HSP90AB1 VVKKHpSQFIG heat shock protein 90 alpha family 459 HS90B_HUMAN class B member 1 HSP90AB1 ELLRYHpTSQSG heat shock protein 90 alpha family 467 HS90B_HUMAN class B member 1 HSP90AB1 GDEMpTSLS heat shock protein 90 alpha family 468 HS90B_HUMAN class B member 1 HSP90AB1 GDEMTpSLS

heat shock protein family A (Hsp70) 393 HSP74_HUMAN member 4 HSPA4 KVREFpSITD heat shock protein family A (Hsp70) 756 HSP74_HUMAN member 4 HSPA4 LQNKQpSLTMD heat shock protein family A (Hsp70) 758 HSP74_HUMAN member 4 HSPA4 LQNKQSLpTMD heat shock protein family D 252 CH60_HUMAN (Hsp60) member 1 HSPD1 KH RNA binding domain containing, signal transduction 20 KHDR1_HUMAN associated 1 KHDRBS1 KIF1A_HUMAN kinesin family member 1A KIF1A 416 KIF1A_HUMAN kinesin family member 1A KIF1A 1370 526 IMB1_HUMAN karyopherin subunit beta 1 KPNB1 527 IMB1_HUMAN karyopherin subunit beta 1 KPNB1 microtubule actin crosslinking factor 3927 MACF1_HUMAN 1 MACF1 MAP1B_HUMAN microtubule associated protein 1B MAP1B 1256 MAP1B_HUMAN microtubule associated protein 1B MAP1B 1242 MATR3_HUMAN matrin 3 MATR3 533 minichromosome maintenance 229 MCM2_HUMAN complex component 2 MCM2 minichromosome maintenance 497 MCM6_HUMAN complex component 6 MCM6 minichromosome maintenance 498 MCM6_HUMAN complex component 6 MCM6 minichromosome maintenance 410 MCM7_HUMAN complex component 7 MCM7 non-SMC condensin I complex 277 CND3_HUMAN subunit G NCAPG

134

580 NUCL_HUMAN nucleolin NCL TLFVKGLpSEDTT NDRG1_HUMAN N-myc downstream regulated 1 NDRG1 332 NDRG1_HUMAN N-myc downstream regulated 1 NDRG1 333

NUP93_HUMAN nucleoporin 93 NUP93 GPPGRpSSLDN 176

NUP93_HUMAN nucleoporin 93 NUP93 GPPGRSpSLDN 177

OSB11_HUMAN oxysterol binding protein like 11 OSBPL11 174 platelet activating factor LIS1_HUMAN acetylhydrolase 1b regulatory PAFAH1B1 152 subunit 1 protein disulfide isomerase family A PDIA6_HUMAN PDIA6 375 member 6 phosphatidylethanolamine binding PEBP1_HUMAN PEBP1 54 protein 1

PROF1_HUMAN profilin 1 PFN1 RTKpSTGGAP 92

PROF1_HUMAN profilin 1 PFN1 RTKSpTGGAP 93 PKN2_HUMAN protein kinase N2 PKN2 PPRASpSLGE 583 protein phosphatase 1 regulatory MYPT1_HUMAN PPP1R12A 445 subunit 12A protein phosphatase 1 regulatory MYPT1_HUMAN PPP1R12A 910 subunit 12A protein kinase cAMP-activated KAPCA_HUMAN PRKACA IKTLGpTGSFG 52 catalytic subunit alpha protein phosphatase 2 PTPA_HUMAN PTPA 348 phosphatase activator DHPR_HUMAN quinoid dihydropteridine reductase QDPR 223 DHPR_HUMAN quinoid dihydropteridine reductase QDPR 224 DHPR_HUMAN quinoid dihydropteridine reductase QDPR 226 RAI14_HUMAN retinoic acid induced 14 RAI14 296 RAGP1_HUMAN Ran GTPase activating protein 1 RANGAP1 453 RAGP1_HUMAN Ran GTPase activating protein 1 RANGAP1 454 SYRC_HUMAN arginyl-tRNA synthetase 1 RARS1 336 RFC4_HUMAN replication factor C subunit 4 RFC4 9 RL18A_HUMAN ribosomal protein L18a RPL18A 123 RL27_HUMAN ribosomal protein L27 RPL27 86 RL6_HUMAN ribosomal protein L6 RPL6 143 spliceosome associated factor 3, SART3_HUMAN SART3 215 U4/U6 recycling protein SERPH_HUMAN serpin family H member 1 SERPINH1 138 135

SERPH_HUMAN serpin family H member 1 SERPINH1 139 SERPH_HUMAN serpin family H member 1 SERPINH1 141 SF3B1_HUMAN splicing factor 3b subunit 1 SF3B1 322 SIN3 transcription regulator family SIN3A_HUMAN SIN3A 689 member A SWI/SNF related, matrix associated, actin dependent SMCA1_HUMAN SMARCA1 644 regulator of chromatin, subfamily a, member 1 SWI/SNF related, matrix associated, actin dependent SMCA5_HUMAN SMARCA5 629 regulator of chromatin, subfamily a, member 5 SWI/SNF related, matrix associated, actin dependent SNF5_HUMAN SMARCB1 111 regulator of chromatin, subfamily b, member 1 SNX2_HUMAN sorting nexin 2 SNX2 119 SOS Ras/Rac guanine nucleotide SOS1_HUMAN SOS1 1134 exchange factor 1 JIP4_HUMAN sperm associated antigen 9 SPAG9 705 SRRM2_HUMAN serine/arginine repetitive matrix 2 SRRM2 857 SRRM2_HUMAN serine/arginine repetitive matrix 2 SRRM2 1458 SRRM2_HUMAN serine/arginine repetitive matrix 2 SRRM2 2449 SRRM2_HUMAN serine/arginine repetitive matrix 2 SRRM2 2692 STXB1_HUMAN syntaxin binding protein 1 STXBP1 509 TBCD4_HUMAN TBC1 domain family member 4 TBC1D4 569 TBCD4_HUMAN TBC1 domain family member 4 TBC1D4 570 treacle ribosome biogenesis factor TCOF_HUMAN TCOF1 734 1 THOC3_HUMAN THO complex 3 THOC3 273 thyroid hormone receptor TR150_HUMAN THRAP3 253 associated protein 3

TLE family member 4, TLE4_HUMAN TLE4 206 transcriptional corepressor TPIS_HUMAN triosephosphate isomerase 1 TPI1 249 trafficking protein particle complex TPC12_HUMAN TRAPPC12 184 12 TIF1B_HUMAN tripartite motif containing 28 TRIM28 473

TIF1B_HUMAN tripartite motif containing 28 TRIM28 784 TBA1B_HUMAN tubulin alpha 1b TUBA1B VFHpSFGGGTGSG 140 TBA1B_HUMAN tubulin alpha 1b TUBA1B VFHSFGGGpTGSG 145 TBA1B_HUMAN tubulin alpha 1b TUBA1B VFHSFGGGTGpSG 147 TBB5_HUMAN tubulin beta class I TUBB QNKNSpSYFV 339 TBB4B_HUMAN tubulin beta 4B class IVb TUBB4B 138 TBB4B_HUMAN tubulin beta 4B class IVb TUBB4B 143 TBB4B_HUMAN tubulin beta 4B class IVb TUBB4B 145 TBB4B_HUMAN tubulin beta 4B class IVb TUBB4B 339

136

TBB4B_HUMAN tubulin beta 4B class IVb TUBB4B 371 TBB4B_HUMAN tubulin beta 4B class IVb TUBB4B 372 ubiquitin like modifier activating UBA1_HUMAN UBA1 MAKNGpSEADI 46 enzyme 1 UBP4_HUMAN ubiquitin specific peptidase 4 USP4 445 XRCC6_HUMAN X-ray repair cross complementing 6 XRCC6 37 XRN2_HUMAN 5'-3' exoribonuclease 2 XRN2 PSISPNpTSFTS 478 XRN2_HUMAN 5'-3' exoribonuclease 2 XRN2 479 XRN2_HUMAN 5'-3' exoribonuclease 2 XRN2 678

137

Table S3: Common PRKACA and PRKACA-Dnajb1 direct protein substrates

Phospho ID Symbol Entrez Gene Name Location Type(s) Site ACINU_HUMAN 400 ACIN1 apoptotic chromatin condensation inducer 1 Nucleus enzyme AN32A_HUMAN 17 ANP32A acidic nuclear phosphoprotein 32 family member A Other other transcription BCLF1_HUMAN 177 BCLAF1 BCL2 associated transcription factor 1 Nucleus regulator transcription BCLF1_HUMAN 397 BCLAF1 BCL2 associated transcription factor 1 Nucleus regulator transcription BCLF1_HUMAN 658 BCLAF1 BCL2 associated transcription factor 1 Nucleus regulator CCNL1_HUMAN 352 CCNL1 cyclin L1 Nucleus other

PYRG1_HUMAN 574 CTPS1 CTP synthase 1 Nucleus enzyme

PYRG1_HUMAN 575 CTPS1 CTP synthase 1 Nucleus enzyme

DTD1_HUMAN 197 DTD1 D-tyrosyl-tRNA deacylase 1 Cytoplasm enzyme translation EF1D_HUMAN 133 EEF1D eukaryotic translation elongation factor 1 delta Cytoplasm regulator translation EF1D_HUMAN 162 EEF1D eukaryotic translation elongation factor 1 delta Cytoplasm regulator EMAL4_HUMAN 94 EML4 EMAP like 4 Cytoplasm other FA50A_HUMAN 276 FAM50A family with sequence similarity 50 member A Nucleus other FBRL_HUMAN 124 FBL fibrillarin Nucleus enzyme transcription HNRPK_HUMAN 216 HNRNPK heterogeneous nuclear ribonucleoprotein K Nucleus regulator HNRNPUL HNRL2_HUMAN 228 heterogeneous nuclear ribonucleoprotein U like 2 Nucleus other 2

ENPL_HUMAN 306 HSP90B1 heat shock protein 90 beta family member 1 Cytoplasm other

LMNB1_HUMAN 302 LMNB1 lamin B1 Nucleus other MATR3_HUMAN 188 MATR3 matrin 3 Nucleus other MATR3_HUMAN 766 MATR3 matrin 3 Nucleus other MDC1_HUMAN 455 MDC1 mediator of DNA damage checkpoint 1 Nucleus other MDC1_HUMAN 453 MDC1 mediator of DNA damage checkpoint 1 Nucleus other MDHC_HUMAN 241 MDH1 malate dehydrogenase 1 Cytoplasm enzyme MDN1_HUMAN 5015 MDN1 midasin AAA ATPase 1 Nucleus other NDRG1_HUMAN 330 NDRG1 N-myc downstream regulated 1 Nucleus kinase NOP56_HUMAN 543 NOP56 NOP56 ribonucleoprotein Nucleus other transcription NPM_HUMAN 70 NPM1 nucleophosmin 1 Nucleus regulator transcription NPM_HUMAN 242 NPM1 nucleophosmin 1 Nucleus regulator NU188_HUMAN 1708 NUP188 nucleoporin 188 Nucleus other NU188_HUMAN 1709 NUP188 nucleoporin 188 Nucleus other PDCD4_HUMAN 94 PDCD4 programmed cell death 4 Nucleus other PELP1_HUMAN 481 PELP1 proline, glutamate and leucine rich protein 1 Nucleus other Plasma transmembra PGRC1_HUMAN 181 PGRMC1 progesterone receptor membrane component 1 Membrane ne receptor PNPO_HUMAN 164 PNPO pyridoxamine 5'-phosphate oxidase Cytoplasm enzyme PNPO_HUMAN 165 PNPO pyridoxamine 5'-phosphate oxidase Cytoplasm enzyme KAPCA_HUMAN 202 PRKACA protein kinase cAMP-activated catalytic subunit alpha Cytoplasm kinase KAPCA_HUMAN 260 PRKACA protein kinase cAMP-activated catalytic subunit alpha Cytoplasm kinase KAPCA_HUMAN 325 PRKACA protein kinase cAMP-activated catalytic subunit alpha Cytoplasm kinase KAPCA_HUMAN 326 PRKACA protein kinase cAMP-activated catalytic subunit alpha Cytoplasm kinase

138

PRC2A_HUMAN 761 PRRC2A proline rich coiled-coil 2A Cytoplasm other TEBP_HUMAN 113 PTGES3 prostaglandin E synthase 3 Cytoplasm enzyme PUF60_HUMAN 206 PUF60 poly(U) binding splicing factor 60 Nucleus other RL8_HUMAN 130 RPL8 ribosomal protein L8 Cytoplasm other S10AD_HUMAN 32 S100A13 S100 calcium binding protein A13 Cytoplasm other SRRM1_HUMAN 693 SRRM1 serine and arginine repetitive matrix 1 Nucleus other SRRM1_HUMAN 694 SRRM1 serine and arginine repetitive matrix 1 Nucleus other SRRM1_HUMAN 695 SRRM1 serine and arginine repetitive matrix 1 Nucleus other SRRM1_HUMAN 696 SRRM1 serine and arginine repetitive matrix 1 Nucleus other SRRM1_HUMAN 872 SRRM1 serine and arginine repetitive matrix 1 Nucleus other SRRM1_HUMAN 874 SRRM1 serine and arginine repetitive matrix 1 Nucleus other SRRM2_HUMAN 315 SRRM2 serine/arginine repetitive matrix 2 Nucleus other SRRM2_HUMAN 316 SRRM2 serine/arginine repetitive matrix 2 Nucleus other SRRM2_HUMAN 317 SRRM2 serine/arginine repetitive matrix 2 Nucleus other SRRM2_HUMAN 318 SRRM2 serine/arginine repetitive matrix 2 Nucleus other SRRM2_HUMAN 322 SRRM2 serine/arginine repetitive matrix 2 Nucleus other SRRM2_HUMAN 323 SRRM2 serine/arginine repetitive matrix 2 Nucleus other LA_HUMAN 94 SSB small RNA binding exonuclease protection factor La Nucleus enzyme STMN1_HUMAN 16 STMN1 stathmin 1 Cytoplasm other STMN1_HUMAN 63 STMN1 stathmin 1 Cytoplasm other transcription TCEA1_HUMAN 57 TCEA1 transcription elongation factor A1 Nucleus regulator LAP2B_HUMAN 66 TMPO thymopoietin Nucleus other LAP2B_HUMAN 67 TMPO thymopoietin Nucleus other Extracellula ZC3HD_HUMAN 265 ZC3H13 zinc finger CCCH-type containing 13 other r Space ZFR_HUMAN 1054 ZFR zinc finger RNA binding protein Nucleus other transcription ZRAB2_HUMAN 120 ZRANB2 zinc finger RANBP2-type containing 2 Nucleus regulator

139

Table S4: Canonical Pathways in PRKACA-Dnajb1 inhibited by kinase inhibitors.

Ingenuity Canonical -log(p-value) Molecules Pathways DNA Double-Strand 8.31 NBN,PARP1,PRKDC,XRCC5,XRCC6 Break Repair by Non- Homologous End Joining Remodeling of Epithelial 6.07 ACTA1,ACTA2,ACTC1,TUBB4B,VCL,ZYX Adherens Junctions Cellular Effects of 5.26 ACTA1,ACTA2,ACTC1,ITPR3,MYH11,PRKACA,PRKAR2A Sildenafil (Viagra) Epithelial Adherens 5 ACTA1,ACTA2,ACTC1,MYH11,TUBB4B,VCL,ZYX Junction Signaling EIF2 Signaling 4.95 ACTA1,ACTA2,ACTC1,EIF3C,RPL22,RPL3,RPL36A,RPL3 6AL Tight Junction Signaling 4.74 ACTA1,ACTA2,ACTC1,MYH11,PRKACA,PRKAR2A,VCL Sertoli Cell-Sertoli Cell 4.39 ACTA1,ACTA2,ACTC1,PRKACA,PRKAR2A,TUBB4B,VCL Junction Signaling Mechanisms of Viral Exit 4.32 ACTA1,ACTA2,ACTC1,VPS4A from Host Cells Telomere Extension by 4.3 NBN,XRCC5,XRCC6 Telomerase Calcium Signaling 4.15 ACTA1,ACTA2,ACTC1,ITPR3,MYH11,PRKACA,PRKAR2A Gap Junction Signaling 4.14 ACTA1,ACTA2,ACTC1,ITPR3,PRKACA,PRKAR2A,TUBB4 B VEGF Signaling 3.82 ACTA1,ACTA2,ACTC1,ELAVL1,VCL Paxillin Signaling 3.76 ACTA1,ACTA2,ACTC1,NCK1,VCL Germ Cell-Sertoli Cell 3.73 ACTA1,ACTA2,ACTC1,TUBB4B,VCL,ZYX Junction Signaling Clathrin-mediated 3.4 AAK1,ACTA1,ACTA2,ACTC1,CD2AP,HSPA8 Endocytosis Signaling Integrin Signaling 3.27 ACTA1,ACTA2,ACTC1,NCK1,VCL,ZYX Sirtuin Signaling 3.03 NBN,PARP1,PRKDC,SMARCA5,VDAC2,XRCC5,XRCC6 Pathway Death Receptor 3 ACTA1,ACTA2,ACTC1,PARP1 Signaling FAK Signaling 2.88 ACTA1,ACTA2,ACTC1,VCL Fcγ Receptor-mediated 2.84 ACTA1,ACTA2,ACTC1,NCK1 Phagocytosis in Macrophages and Monocytes BER pathway 2.75 FEN1,PARP1 Amyloid Processing 2.7 MAPT,PRKACA,PRKAR2A ILK Signaling 2.64 ACTA1,ACTA2,ACTC1,MYH11,VCL Cell Cycle Control of 2.58 CDK11B,CDK17,TOP2B Chromosomal Replication Granzyme B Signaling 2.57 PARP1,PRKDC Role of CHK Proteins in 2.56 MDC1,NBN,RFC1 Cell Cycle Checkpoint Control RAR Activation 2.53 PARP1,PML,PRKACA,PRKAR2A,SNW1 MSP-RON Signaling 2.52 ACTA1,ACTA2,ACTC1 Pathway Molecular Mechanisms 2.5 CDK11B,CDK17,NBN,PRKACA,PRKAR2A,PRKDC,RALB of Cancer P1

140

Mismatch Repair in 2.46 FEN1,RFC1 Eukaryotes Actin Cytoskeleton 2.37 ACTA1,ACTA2,ACTC1,MYH11,VCL Signaling PXR/RXR Activation 2.34 ALDH3A2,PRKACA,PRKAR2A Netrin Signaling 2.31 NCK1,PRKACA,PRKAR2A Caveolar-mediated 2.22 ACTA1,ACTA2,ACTC1 Endocytosis Signaling Agrin Interactions at 2.18 ACTA1,ACTA2,ACTC1 Neuromuscular Junction Role of BRCA1 in DNA 2.15 MDC1,NBN,RFC1 Damage Response GPCR-Mediated 2.15 ITPR3,PRKACA,PRKAR2A Integration of Enteroendocrine Signaling Exemplified by an L Cell Tec Kinase Signaling 2.07 ACTA1,ACTA2,ACTC1,GTF2I eNOS Signaling 2.05 HSPA8,ITPR3,PRKACA,PRKAR2A Sonic Hedgehog 2.03 PRKACA,PRKAR2A Signaling Regulation of Actin- 2.01 ACTA1,ACTA2,ACTC1 based Motility by Rho Crosstalk between 2.01 ACTA1,ACTA2,ACTC1 Dendritic Cells and Natural Killer Cells

141

PKA-CA

Regulatory Subunit

Figure 6.1. Location of the Cushing’s disease mutations at the interface of R-Ca binding site. These mutations are assumed to impact the activity and regulation of the PKA catalytic subunit and consequently, affect cyclic AMP (cAMP) signaling.

Table 6.2: Catalytic parameters derived from coupled enzyme assay for Human PKA and Cushing’s disease related PKA mutants W196R and L205R.

142

Table 6.3: A) Comparison of the Thermodynamic parameters for the two Cushing's disease mutant compared to the wild type PKA. Measured at wavelength 222 nm, Temperature scan was from 20°C to 70°C at 1 degree/min to unfold the protein. B) Human PKA is thermally more stable compared to its Mouse counterpart in Apo and Binary form. This thermodynamic stability can be a result of eight amino acid difference, which are located at N terminal lobe.

B APO MgATP MgATP-PKI5-24 Species TM TM (°C) TM (°C)

(°C) Human PKA 53.22±0.50 55.03±0.73 60.21±0.43

Mouse PKA 48.9 °C 53.17 °C 61.21 °C

Control

RII W196R

A WT

PK + Control PKA PKI + PKA W196R PKA+ H89 w196R+PKI W196R + H89 - Figure 6.3. Inhibitory mechanism using Non- Radioactive Protein Kinase Assays (Gel shift assay). PKI and H89 are PKA potent inhibitors used in the

assay.

Figure 6.2. Affinity pull-down assay using His-tagged RIIB and untagged W196R. after mixing the catalytic and the regulatory subunits, the mix was passed through the Ni-NTA resin. The column was washed twice using high salt concentration and cAMP (to detach the R-C subunits). Upon elution with Imidazole, the RIIb subunit was detached from the resin as shown in figure. The 14% SDS gel was used to display the protein content of fractions.

143

Figure 6.4. TROSY-HSQC for PKA-Human Isoform II (Apo form)

Figure 6.5. TROSY-HSQC of Isoform II in APO and ATPγN -bound form, overlaid.

Recorded on 7001, NS: 64, S/N: 33, PH: 6.5

144

Figure 6.6: PKA-Human isoform II in complex with nucleotide and PKI PKI/PKA. Conc ratio: 2:1, ATPγN Con: 12 mM, Recorded on 7001, NS: 64, S/N: 33, PH: 6.5

Figure 6.7: TROSY-HSQC spectra of 15N, 13C perdeuterated Human PKA bound to ATPγN and PKI Ternary complex). 5-24 ( TROSY-HSQC, NS: 16, Temp: 300 K, S/N : 87, PKA Conc.: 560-630 µM

145

Figure 6.8: Up: PKA mutant “L205R” in the apo and in ATP-γN bound (binary) form. TROSY- HSQC spectra of PKA mutant “L205R” in Apo (red) and in complex with 12 mM ATPγN (blue/Green). Some affected residues upon binding to the ATP-γN have been zoomed out. Down: PKA mutant “L205R” in the apo and in PKI5-25 and ATP-γN bound form (ternary form). Blue: Apo form, yello-red: ternary complex.

146

Open (Apo) Closed

Figure 6.9: CONCISE analysis applied to PKA mutant “L205R”. Probability distribution for each of the three states is depicted.

Figure 6.10: TROSY-HSQC spectra of 15N-labeled W196R. Some peaks show a linear fashion in chemical shift difference (indication of fast exchange), where in others a slow exchange is observed. Details of these residue movements are analyzed.

147

Titration with Nucleotide ATPγN Apo, ATPγN bond and PKI5-24 bound Apo, ATPγN bond and PKI5-24 bound

W196R mutant W196R mutant PKA-WT

Figure 6.11: Comparison of correlation matrix of the sidechain movements in W196R with that of PKA-WT. Overall Correlation is considerably decreased in mutant W196R.

148