Molecular biology, biochemistry and genomics: pre-mRNA processing and P450 monooxygenases

Background

From structural perspectives, plant and insect cytochrome P450 monooxygenases (P450s) are membrane-bound proteins sharing high conservation in their secondary and tertiary folding patterns with the membrane-bound forms found in vertebrates and soluble forms found in bacteria. Embedded within the external skeleton formed by higher order α-helix and β-pleated sheet structures, the internal catalytic sites in various P450s are defined by a series of varying length loop sequences (designated as substrate recognition sequences (SRS)) that specify the size of the catalytic site and orientation of compounds over the catalytic heme. In the correct orientation and in proximity to the catalytically important I-helix, substrates incorporate the heme-bound oxygen through a reaction cycle that utilizes electrons transferred from P450 reductase and cytochrome b5. In the incorrect orientation, inhibitors position themselves within the catalytic site, block the reaction cycle and interfere with substrate oxygenation. 

From biochemical perspectives, the ability to incorporate oxygen at very specific points in a substrate’s structure is essential to numerous synthetic and catabolic pathways that utilize simple alkyl and aromatic hydroxylations or more complex epoxidations, aryl migrations, decarboxylations and carbon-carbon bond cleavages. Because of the involvement of P450s in many critical metabolic conversions, the restricted substrate specificity of many plant and insect monooxygenases and the chemical versatility of the entire group of monoxygenases have generated significant interest in defining functions for those not yet characterized, in understanding catalytic site constraints of those in biologically important pathways and in modifying them for crop improvement and biopharmaceutical production. 

From genomics and evolutionary perspectives, these types of goals are especially challenging since plant P450 gene families have duplicated and diverged to unprecedented degrees as new pathways have evolved for the synthesis of plant defense toxins and other secondary metabolites. From functional and transcriptional perspectives, these goals are also challenged by the fact that closely related P450s sometimes vary in only a few catalytic site residues affecting their substrate specificities and their genes may vary in only a few promoter elements affecting their expression patterns.

Research Projects

P450 Biochemistry and Functional Genomics

Our research projects in this area are focused on defining the functions and structures of a variety of plant P450s with biological roles important in the production of defense toxins, phenylpropanoids, fatty acids and signaling molecules as well as insect P450s with biological roles in the destruction of plant defense toxins and synthetic insecticides. With many P450s being integral to the synthetic and detoxicative functions in these organisms, we have sought to better define the biochemical parameters dictating the substrate specificities of individual P450s, the molecular programs defining their transcriptional responses to internal and external cues and the genomic events enabling their acquisition of novel biochemical functions. 

Bringing together structural, biochemical and genomics perspectives into a comprehensive view of each P450 being studied, we have focused on developing and interfacing computational and biochemical technologies important for understanding, defining and manipulating P450 functions, irrespective of their plant, insect or vertebrate origin. Stepping beyond the phylogenetic primary sequence alignments most frequently used to assign P450 nomenclature, we have developed and improved molecular modeling protocols allowing us to more accurately predict three-dimensional structures for plant and insect P450s based on available mammalian P450 crystal structures (Baudry et al., 2006; Rupasinghe and Schuler, 2006). We have also developed low-throughput and high-throughput docking procedures for predictions of a compound’s binding mode within various catalytic sites. Interfacing computational predictions of protein structure and substrate binding mode with biochemical reality, we have optimized several heterologous expression systems for production of P450s, P450 reductase and cytochrome b5 (Duan and Schuler, 2006) and utilized these for defining biochemical functions of individual P450s. Among our available systems, baculovirus-mediated insect cell expression has been used for substrate binding and activity analyses, inhibitor analyses and scaled-up product characterizations, yeast expression has been used for substrate activity analyses and metabolic engineering, and bacterial expression has been used for metabolic engineering and scaled-up production for crystallization trials and isotopic labelings. Our biochemical analyses, which are tied back to computational predictions and forward to evolutionary analyses, support our structure predictions by generating and testing numerous catalytic site, proximal surface and membrane anchor mutants. Together, these data have provided for very thorough appreciations of the catalytic site and proximal surface residues modulating P450 activities and substrate ranges. 

Incorporating these computational and biochemical perspectives with the genomics perspectives needed to understand the physiological roles of specific P450s, we have used a variety of techniques to detail and manipulate a range of plant P450 transcripts expressed in response to abiotic and biotic stresses and insect P450 transcripts expressed in response to natural and synthetic toxins. Providing perspective on the types of biological systems that we are currently studying, these include:

Arabidopsis thaliana P450s

Projects in the large collection of 246 Arabidopsis P450s have focused on defining biochemical and physiological functions for the subfamilies mediating metabolisms of fatty acids, phenylpropanoids and defense toxins. Molecular modeling, substrate docking and biochemical analysis have allowed us to detail the catalytic site conservations and differences between members of the CYP86 and CYP94 families that modulate hydroxylations on nonoxygenated vs. oxygenated fatty acids (Rupasinghe et al., 2007) and members of the CYP98A subfamily that modulate hydroxylations on lignin precursors. Biochemical analyses tied to functional analyses of T-DNA knockouts have allowed us to describe the developmental and stress-related effects of inactivating particular Arabidopsis P450 genes.
 
With the stress inducibilities of many Arabidopsis P450 transcripts defined previously in microarray analyses, we have begun analyzing the circadian fluctuations of individual genes to determine the range of P450 transcripts affected by the circadian clock and the extent to which stress signaling cascades swamp normal cyclical variations. Our analyses have demonstrated that many P450 transcripts in the flavonoid, carotenoid, brassinosteroid, jasmonate and glucosinolate biosynthetic pathways are circadian-regulated (Pan et al., 2009), including several capable of responding to jasmonic acid and other biotic stress signaling molecules. Transcript analyses at varying times in the circadian cycle have shown that, for several P450 genes, the magnitudes of their inducibilities change throughout the day in response to stress signaling molecules indicating that the responses of some biosynthetic pathways vary with and, sometimes, supercede the circadian cycle. Bioinformatics analysis has allowed us to identify over-represented promoter motifs in those responding to both circadian and stress signaling cues and these are being tested for functionality in vitro and in vivo.

Helicoverpa, Apis and Anopheles P450s

Projects on a collection of insect P450s, including Anopheles gambiae (malarial mosquito) and Apis mellifera (honey bee) have focused on identifying those capable of detoxifying natural plant toxins and understanding their relationships to those detoxifying synthetic insecticides. Our work with many of these insect P450s combines expression in heterologous systems with molecular modeling to predict and define the range of compounds metabolized by individual P450 sequences and to determine the effects of allelic variations on catalytic activities toward various classes of plant compounds and insecticides. These analyses have provided concrete evidence for the convergent evolution of P450s involved in the detoxification of plant compounds and insecticides (Li et al., 2004, 2007; Rupasinghe et al., 2007; Chiu et al., 2008) and highlighted the importance of avoiding repeated insecticide selections. For others, these analyses have provided evidence for selective evolution in some P450 catalytic sites that have restricted access to the heme oxygen and directed metabolism toward highly select groups of plant toxins or that have varied proximal surface residues to enhance interactions with critical electron transfer partners needed for activity (Mao et al., 2006, 2007, 2008). 

In current projects, we have begun characterizing and mapping P450 allelic variants in A. gambiae and other mosquitoes vectoring parasites for human diseases. Predictive structures built for several members of the A. gambiae CYP6Z subfamily, which are over-expressed in insecticide-resistant stock strains, have provided information on a range of catalytic site and proximal surface changes that have potential to affect enzymatic functions. In conjunction with these studies, we are using high-throughput computational screening to identify inhibitors for several P450s capable of metabolizing insecticides and likely to mediate the acquisition of insecticide resistance in various mosquito populations. Other projects have begun characterizing the small collection of 46 honey bee P450s and identified several in the expanded CYP6AS subfamily that metabolize quercitin, the main flavonoid in honey and pollen (Mao et al., 2009). Several involved in the metabolism of the insecticide tau-fluvalinate have also been identified. 

Biopharmaceutically-relevant P450s 

Projects in biopharmaceutically-important P450s are focused on identifying plant P450s and their associated proteins involved in the synthesis of important anti-cancer and anti-viral alkaloids in Catharanthus roseus and Camptotheca acuminata, two medicinal plants. With many parts of the complex alkaloid synthetic pathways in these plants containing P450s whose protein expression characteristics have not yet been tested in bacterial and yeast systems, we have begun cloning and optimizing expression of components in the early and late alkaloid pathways. Protein candidates for these components are being identified by bioinformatic and computational assessments of transcriptome profiles that have recently become available for these medicinal plants. With structural and biochemical analyses of individual candidates completed, future projects with our collaborators will focus on engineering of complete pathways in yeast and bacteria. 

P450 Structural Biology

Projects in this area are focused on coupling molecular modeling, protein expression and solid-state NMR capabilities to define the backbone structures for plant and human P450s. For these, full-length P450s (including their hydrophobic membrane anchor) are expressed in E. coli under conditions suitable for uniformly-labeling proteins with 13C,15N amino acids (Rupasinghe et al., 2007) or selectively-labeling proteins with subsets of 13C,15N amino acids. Interfacing of scored molecular models with experimental solid-state NMR data has provided us and our collaborators with the ability to discriminate between good and bad P450 models and to significantly enhance structure determinations from solid-state NMR datasets. 

Plant Intron Splicing

Projects in this area are focused on defining the range of splicing factors mediating constitutive and alternative splicing during plant growth and development. Using Arabidopsis T-DNA knockout/knockout lines and RNA silencing lines depleted for different classes of splicing factors, we have been characterizing the roles of a number of second-step splicing factors as well as general splicing factors. We are particularly interested in the second-step splicing factors since they are encoded by multiple divergent genes that have potential to differentially regulate splicing, in contrast to those in yeast and humans that are encoded by single genes. Additional projects have begun to examine the stress responses of the second-step splicing factors and to determine whether some of these alter splice site choices in the stressed plant cells. 

Coupled with information on the cis acting sequences required for recognition of plant AU-rich introns, these studies will identify the trans acting factors involved in the recognition of plant AU rich introns and, in the long run, help define the optimal sequences needed to express transgenes and transgenic proteins in heterologous plants.