PDZ domains have long been thought to cluster into discrete functional classes defined by their peptide-binding preferences. We used protein microarrays and quantitative fluorescence polarization to characterize the binding selectivity of 157 mouse PDZ domains with respect to 217 genome-encoded peptides. We then trained a multidomain selectivity model to predict PDZ domain-peptide interactions across the mouse proteome with an accuracy that exceeds many large-scale, experimental investigations of protein-protein interactions. Contrary to the current paradigm, PDZ domains do not fall into discrete classes; instead, they are evenly distributed throughout selectivity space, which suggests that they have been optimized across the proteome to minimize cross-reactivity. We predict that focusing on families of interaction domains, which facilitates the integration of experimentation and modeling, will play an increasingly important role in future investigations of protein function.

A general purpose transformation vector, designated pPha-T1, was constructed for use with the diatom Phaeodactylum tricornutum Bohlin. This vector harbors the sh ble cassette for primary selection on medium containing the antibiotic zeocin, and a multiple cloning site flanked by the P. tricornutum fcp A promoter. pPha-T1 was used to establish the utility of three selectable marker genes and two reporter genes for P. tricornutum transformation. The nat and sat-1 genes confer resistance to the antibiotic nourseothricin, and npt II confers resistance to G418. Each of these genes was effective as a selectable marker for identifying primary transformants. These markers could also be used for dual selections in combination with the sh ble gene. The reporter genes uid A and gfp were also introduced into P. tricornutum using pPha-T1. Gus expression in some transformants reached 15 g· g Ϫ 1 of total soluble protein and permitted excellent cell staining, while GFP fluorescence was readily visible with standard fluorescence microscopy. The egfp gene, which has optimal codon usage for expression in human cells, was the only version of gfp that produced a strong fluorescent signal in P. tricornutum. The codon bias of the egfp gene is similar to that of P. tricornutum genes. This study suggests that codon usage has a significant effect on the efficient expression of reporter genes in P. tricornutum. The results presented here demonstrate that a variety of selectable markers and reporter genes can be expressed in P. tricornutum , enhancing the potential of this organism for exploring basic biological questions and industrial applications.

Most microalgae are obligate photoautotrophs and their growth is strictly dependent on the generation of photosynthetically derived energy. We show that the microalga Phaeodactylum tricornutum can be genetically engineered to thrive on exogenous glucose in the absence of light through the introduction of a gene encoding a glucose transporter (glut1 or hup1). This demonstrates that a fundamental change in the metabolism of an organism can be accomplished through the introduction of a single gene. This also represents progress toward the use of fermentation technology for large-scale commercial exploitation of algae by reducing limitations associated with light-dependent growth.

Diatoms and related algae have plastids that are surrounded by four membranes. The outer two membranes are continuous with the endoplasmic reticulum and the inner two membranes are analogous to the plastid envelope membranes of higher plants and green algae. Thus the plastids are completely compartmentalized within the ER membranes. The targeting presequences for nuclear‐encoded plastid proteins have two recognizable domains. The first domain is a classic signal sequence, which presumably targets the proteins to the endoplasmic reticulum. The second domain has characteristics of a transit peptide, which targets proteins to the plastids of higher plants. To characterize these targeting domains, the presequence from the nuclear‐encoded plastid protein AtpC was utilized. A series of deletions of this presequence were fused to Green Fluorescent Protein (GFP) and transformed into cells of the diatom, Phaeodactylum tricornutum. The intracelluar localization of GFP was visualized by fluorescence microscopy. This work demonstrates that the first domain of the presequence is responsible for targeting proteins to the ER lumen and is the essential first step in the plastid protein import process. The second domain is responsible to directing proteins from the ER and through the plastid envelope and only a short portion of the transit peptide‐like domain is necessary to complete this second processing step. In vivo data generated from this study in a fully homologous transformation system has confirmed Gibbs' hypothesis regarding a multistep import process for plastid proteins in chromophytic algae.