The survival of the pathogenic bacterium, Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis, depends on its ability to sense and respond to a hostile and constantly changing environment within the host. Cell-wall components--such as peptidoglycan, arabinogalactan, mycolic acids and lipoarabinomannan--perform critical structural and biological functions that allow Mtb to survive and persist inside of human cells. Despite the critical roles of the cell wall in the pathogenicity of Mtb, surprisingly little is known about how this complex structure is built and remodeled in response to environmental cues. Recent evidence, however, implicates membrane-bound serine/threonine protein kinases (STPKs) in the coordinating cell-wall synthesis with growth and division. These eukaryotic-like STPKs play critical roles in Mtb signal transduction by sensing extracellular stimuli and catalyzing the reversible phosphorylation of cytoplasmic target proteins.

To further understand the mechanisms that regulate STPKs and cell wall architecture in Mtb, I biochemically and structurally characterized the extracellular sensor domain of one STPK receptor kinase, PknH. Initial analysis of recombinantly produced PknH sensor domain by size-exclusion chromatography and small-angle x-ray scattering indicated the protein was soluble but completely unfolded. On-column oxidative refolding produced a properly folded protein that crystallized as a monomer with a novel protein fold made up of six alpha helices, seven beta strands, and two disulfide bonds. The PknH sensor domain has a large conserved cleft with a mixed polar and hydrophobic surface. These results indicated PknH binds a small molecule ligand which may affect its quarternary structure and/or its localization.

In order to determine what ligand binds to the PknH sensor domain, I conducted native gel binding assays using purified Mtb arabinogalactan, lipoarabinomannan, and lipomannan. These assays indicated that in vitro, the PknH sensor domain does not bind to any of these purified cell wall components. Affinity chromatography with His₆-tagged PknH incubated with Mtb H37Rv cell lysate followed by intact mass, LC-MS, and ESI-MS analysis indicated that PknH does not bind glycolipids such as phosphoinositol or Ac₁PIM₂. Although the ligand for PknH was not identified, these studies set the stage for future researchers to identify the ligand and discover the function of PknH.

Intracellular pathogens depend on their hosts for survival, and pathogen effectors that function at the host interface are critical to this process. Pathogens regulate their hosts through a combination of offensive and defensive strategies - capturing host machinery to carry out virulence functions while evading detection and actively inhibiting the host immune response. Here I report important advances resulting from studies of two relationships, including the oomycete Hyaloperonospora arabidopsidis (Hpa), which infects the model plant Arabidopsis thaliana, and the retrovirus HIV-1, which infects humans.

Hpa secretes a family of virulence effectors into Arabidopsis cells where they interact with host targets. Activation of the plant immune response against Hpa requires physical recognition of effectors by host Resistance proteins (R-proteins), and Hpa escapes host detection through genetic variation of effector recognition surfaces. To understand the basis for RPP1 recognition, I determined the 2.3-Å resolution crystal structure of one Hpa effector, Arabidopsis thaliana Recognized 1 (ATR1), which is recognized by the Arabidopsis R-protein Recognition of Peronospora Parasitica 1 (RPP1). We found that RPP1 recognizes distributed surfaces of ATR1, and different alleles of RPP1 recognize distinct surfaces of the effector. ATR1 belongs to an ancient family of conserved oomycete effectors that evolves rapidly through surface polymorphisms to escape host recognition while maintaining a conserved structural core.

To explore the mechanisms Human immunodeficiency virus (HIV) pathogenesis, I focused on the HIV accessory protein, Tat, which hijacks host machinery to stimulate viral transcription. HIV Tat recruits human positive transcription elongation factor b (P-TEFb) to the HIV LTR, and this recruitment is required for elongation of viral transcripts. Using a proteomics-based approach, we discovered that P-TEFb associates with several new components, including AFF4, ELL2, and homologs ENL and AF9, that form a larger Super Elongation Complex (SEC). Through in vitro reconstitution using purified recombinant proteins, I showed that AFF4 is the major scaffold that directly binds other subunits via short interaction domains. Using limited proteolysis and binding assays to structurally characterize the SEC, I found that AFF4 is an extensively disordered scaffold that remains flexible even upon binding to SEC components. Short, hydrophobic peptides on an otherwise hydrophilic, serine-rich landscape form the binding domains on AFF4. Moreover, ELL2 and ENL function as bridging components that link the SEC to another transcriptional regulator, the PAF complex (PAFc). This work establishes the overall architecture of the SEC and provides insight into how Tat, by binding P-TEFb, can recruit a large ensemble of scaffolded transcriptional regulators to stimulate viral transcription.

Together, these studies provide new insights into pathogenic mechanisms. By characterizing the structural underpinnings of interactions at two different host-pathogen interfaces, we better understand how one effector, Hpa ATR1, escapes host detection and how another effector, HIV-1 Tat, finds and recruits multiple host factors to serve a virulence function.

Proteins populate structural ensembles. Defining these ensembles and understanding the role of the interconversions between structures is a grand challenge of structural biology. My work addresses that challenge through the development and application of new methods to reveal sparsely populated structures. Quantitative electron-density map interpretation, implemented in Ringer, provides an objective, systematic method to identify previously undiscovered alternate side chain substates that mediate conformational transitions in proteins.

Next, I applied these methods to study the role of the interconversions of an enzyme, the human proline isomerase CypA, between two conformations during its catalytic cycle. Using the dual strategies of ambient-temperature X-ray crystallographic data collection and automated electron-density sampling, I defined the previously undiscovered minor state as a network of alternate side chain conformations. A conservative mutation outside the active site inverts the equilibrium between the substates and causes large, parallel reductions in the conformational interconversion rates and the catalytic rate.

The temperature dependent differences in electron density observed with CypA led me to critically examine the assumption that crystal freezing does not significantly bias protein structure. I found extensive remodeling of the crystal lattice upon freezing. Crystal freezing also leads to improved packing through reduction of small voids and a reduction in protein volume. I used real-space electron density sampling to show that these voids can be transiently populated by alternate conformations in the room temperature ensemble. This work shows how crystal freezing biases our understanding of protein packing and can lead to differences in the spatial distribution of the dynamic features of protein side chains.

These studies highlight the importance of conformational diversity in protein function. By looking beyond the model and into the map, we can find that polysteric regions often populate conformations that resemble the structures populated along reaction or evolutionary trajectories. Thus, understanding polysterism yields insights into where a protein might visit during its reaction cycle and where it has been during its evolution.

Mycobacterium tuberculosis (Mtb) coordinates a wide variety of metabolic and cellular responses to changing external environments throughout the multiple stages of infection. Signaling kinases are critical for these responses. The Mtb genome encodes 11 Serine/Threonine Protein Kinases (STPKs) that function as important nodes of this sensing and response network, but the chemical and structural changes that mediate kinase activation have not been elucidated.

Autophosphorylation activates several of the Mtb STPKs, and kinase dimerization can activate receptor kinases for autophosphorylation through an allosteric dimer interface. Inter-kinase phosphorylation has been reported, but the function and specificity of these interactions remain unknown.

In this study, a biochemical approach was used to comprehensively map the cross-kinase trans-phosphorylation activity of the Mtb STPKs. The results reveal a pattern of kinase interactions that suggests each protein plays a distinct regulatory role in controlling cellular processes by phosphorylating other kinases.

The PknB and PknH STPKs act in vitro as master regulators that are activated only through autophosphorylation and also phosphorylate other STPKs. In contrast, the signal-transduction kinases PknE, PknJ and PknL are phosphorylated by the master regulatory STPKs and phosphorylate other kinases. The substrate STPKs PknA, PknD, PknF and PknK are phosphorylated by upstream STPKs, but do not phosphorylate other kinases. The delineation of the Mtb STPK signaling networks reveals for the first time the specific network of STPK phosphorylation that may mediate the intracellular signaling circuitry.

STPKS are activated and inhibited by phosphorylation at different residues. The regulatory role of the extensive Mtb STPK trans-phosphorylation network is unknown. Through mass spectrophotometry and mutagenesis, the amino acids targeted by each phosphorylation were identified. I find that key activation loop residues are the targets of both autophosphorylation and trans-phosphorylation. Mutation of the two conserved threonines in the activation loops of nine Mtb STPKs renders the kinases inactive. These results demonstrate that activation loop phosphorylation is a common mechanism of Mtb STPK activation.

To explore the structural implications of activation loop phosphorylation, I determined the crystal structures of the phosphorylated and unphosphorylated Mtb PknH kinase domain. The PknH kinase domain forms a back-to-back dimer observed previously in the structures of Mtb PknB and PknE. Amino-acid substitutions in the dimer interface fail to block kinase dimerization or autophosphorylation. Unexpectedly, the PknH activation loop is folded in the unphosphorylated form and disordered in the active, phosphorylated enzyme. These structures revealed that the back-to-back kinase dimer is a surprisingly stable structure that does not undergo global conformational changes with phosphorylation or nucleotide binding. Unlike the well-established, conformational regulatory mechanisms of eukaryotic STPKs, PknH activation may be a biochemical process mediated by changes in nucleotide affinity or activation loop disorder rather than remodeling of the overall kinase-domain architecture.

To establish the effects of stepwise phosphorylation of an Mtb STPK, I determined the structures of the PknK kinase domain modified with 0, 1, or multiple phosphoryl groups. PknK is one of the two solution STPKs in Mtb and is phosphorylated by multiple upstream kinases. Mass spectrophotometry revealed that the trans-phosphorylation and autophosphorylation reactions resulted in varying numbers of phosphate groups on this substrate protein. Crystal structures revealed that, like PknH, PknK does not undergo conformational remodeling following activation loop phosphorylation. Unlike many eukaryotic homologs, the unphosphorylated forms of PknH and PknK bind ATP analogs. These two examples suggest that phosphorylation activates the Mtb STPKs by directly changing the properties of the activation loop. Based on these results, I propose the testable new idea that activation loop phosphorylation may regulate the Mtb STPKs by directly changing the affinities for protein substrates or nucleotides. The absence of conformational remodeling of PknH and PknK upon activation loop phosphorylation implies that the prokaryotic and eukaryotic STPKs are regulated by different mechanisms.

Mycobacterium tuberculosis (Mtb) cell wall, built on a cross-linked sugar-peptide polymer called peptidoglycan, protects the bacterial cell from adverse environments. Peptidoglycan homeostasis is maintained by extracellular peptidoglycan synthases and hydrolases. Intricate coordination of their activities is required to maintain structural integrity of the cell wall during growth, division, and response to stress. The sensor protein kinase B (PknB) is likely to play a critical role in monitoring the state of the peptidoglycan outside the cell and inducing subsequent metabolic changes inside. In this study, computational, biochemical, and structural approaches were used to characterize the peptidoglycan hydrolases of Mtb and to investigate the molecular mechanism of peptidoglycan signaling through PknB.

Peptidoglycan hydrolases are critical players in bacterial growth, division, cell shape determination, and peptidoglycan fragment-mediated communication. Computational analysis identified 22 mycobacterial peptidoglycan hydrolases based on homology to known enzymes from model organisms. The peptidoglycan degradation machinery of Mtb includes 4 N-acetylmuramoyl-L-alanine amidases, 8 lytic transglycosidases, and 10 peptidases of various specificities. Ten of these enzymes form a core set of mycobacterial peptidoglycan hydrolases, while four of them are essential for growth in Mtb. Comprehensive biochemical and structural investigation of the Mtb peptidoglycan hydrolases was initiated by cloning and heterologously expressing constructs representing all 22 Mtb peptidoglycan hydrolases in Escherichia coli. Robust expression was observed for all but one target protein. Twelve were successfully purified on large scale.

The peptidoglycan amidases Rv3717 and Rv3915 share similar catalytic cores yet have non-redundant functions in peptidoglycan turnover. Hydrolase activity assays using polymerized peptidoglycan sacculi and soluble peptidoglycan fragments elucidated contributions of individual amino acid residues, metal binding, and disulfide bond formation to catalysis. The structure of product-bound Rv3717 suggested a mechanism that limits this enzyme's activity on polymerized sacculi.

Peptidoglycan D,D-peptidases Rv2911, Rv3330 and Rv3627 are low molecular weight penicillin-binding proteins that participate in peptidoglycan maturation and degradation. Surprisingly, these three enzymes were inactive on peptidoglycan sacculi or peptidoglycan fragments, yet were active on beta-lactams meropenem and Bocillin. The structure of Rv3330 solved in complex with meropenem revealed a potential peptide-binging groove distant from the active site. The observed lack of activity of low molecular weight penicillin-binding proteins suggests a requirement for an activator. Discovery of such factors will significantly advance our understanding of Mtb peptidoglycan homeostasis.

PknB is an essential sensor kinase that controls cell wall biosynthesis. Its homologs in Gram-positive bacteria have been implicated in binding peptidoglycan fragments and in mediating bacterial responses to cell wall stress. To investigate the mechanism of peptidoglycan recognition by PknB, the structure of its extracellular sensor domain was solved. It consists of four 70-amino acid PASTA repeat domains that adopt an extended conformation. The last repeat domain contains a hydrophobic pocket with a conserved tryptophan, a signature of a ligand-binding site. The structure of PknB extracellular domain suggests that ligand-dependent localization and oligomerization control kinase activity.