Proteomic Search Research Articles

A comprehensive and scalable database search system for metaproteomics.

BackgroundMass spectrometry-based shotgun proteomics experiments rely on accurate matching of experimental spectra against a database of protein sequences. Existing computational analysis methods are limited in the size of their sequence databases, which severely restricts the proteomic sequencing depth and functional analysis of highly complex samples. The growing amount of public high-throughput sequencing data will only exacerbate this problem. We designed a broadly applicable metaproteomic analysis method (ComPIL) that addresses protein database size limitations.ResultsOur approach to overcome this significant limitation in metaproteomics was to design a scalable set of sequence databases assembled for optimal library querying speeds. ComPIL was integrated with a modified version of the search engine ProLuCID (termed “Blazmass”) to permit rapid matching of experimental spectra. Proof-of-principle analysis of human HEK293 lysate with a ComPIL database derived from high-quality genomic libraries was able to detect nearly all of the same peptides as a search with a human database (~500x fewer peptides in the database), with a small reduction in sensitivity. We were also able to detect proteins from the adenovirus used to immortalize these cells. We applied our method to a set of healthy human gut microbiome proteomic samples and showed a substantial increase in the number of identified peptides and proteins compared to previous metaproteomic analyses, while retaining a high degree of protein identification accuracy and allowing for a more in-depth characterization of the functional landscape of the samples.ConclusionsThe combination of ComPIL with Blazmass allows proteomic searches to be performed with database sizes much larger than previously possible. These large database searches can be applied to complex meta-samples with unknown composition or proteomic samples where unexpected proteins may be identified. The protein database, proteomic search engine, and the proteomic data files for the 5 microbiome samples characterized and discussed herein are open source and available for use and additional analysis.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-016-2855-3) contains supplementary material, which is available to authorized users.

The intricacy of biomarker complexity-the identification of a genuine proteomic biomarker is more complicated than believed.

Several reasons have been put forward to explain the irreproducibility of proteomic biomarker search. However, these reasons pertain to almost every part of biomarker search across the entire analytical workflow but are entirely experimental or methodological. However, in this article we point out that there is a further cause of such irreproducibility. This is not an additional methodological or experimental cause but arises directly from the biology of protein expression. It arises from the fact that disease changes the diversity within protein families. This cause of irreproducibility has been very little studied in relation to proteomic biomarker search. Gene expression is highly variable even in healthy people. Therefore, multiple proteoforms are also to be expected when gene expression is disrupted by disease, proteoforms that may be differently altered by pathology. In consequence, it is illogical to expect that the whole protein family produces a reliably usable biomarker. It is more reasonable to expect that a specific proteoform fulfills this role. Appropriate sample pre-fractionation methods and data analyses could help to identify this version, carrying the modification or the epitope required.

Adaptation of Decoy Fusion Strategy for Existing Multi-Stage Search Workflows.

A number of proteomic database search engines implement multi-stage strategies aiming at increasing the sensitivity of proteome analysis. These approaches often employ a subset of the original database for the secondary stage of analysis. However, if target-decoy approach (TDA) is used for false discovery rate (FDR) estimation, the multi-stage strategies may violate the underlying assumption of TDA that false matches are distributed uniformly across the target and decoy databases. This violation occurs if the numbers of target and decoy proteins selected for the second search are not equal. Here, we propose a method of decoy database generation based on the previously reported decoy fusion strategy. This method allows unbiased TDA-based FDR estimation in multi-stage searches and can be easily integrated into existing workflows utilizing popular search engines and post-search algorithms. Graphical Abstract ᅟ.

Human Proteomic Variation Revealed by Combining RNA-Seq Proteogenomics and Global Post-Translational Modification (G-PTM) Search Strategy.

Mass-spectrometry-based proteomic analysis underestimates proteomic variation due to the absence of variant peptides and posttranslational modifications (PTMs) from standard protein databases. Each individual carries thousands of missense mutations that lead to single amino acid variants, but these are missed because they are absent from generic proteomic search databases. Myriad types of protein PTMs play essential roles in biological processes but remain undetected because of increased false discovery rates in variable modification searches. We address these two fundamental shortcomings of bottom-up proteomics with two recently developed software tools. The first consists of workflows in Galaxy that mine RNA sequencing data to generate sample-specific databases containing variant peptides and products of alternative splicing events. The second tool applies a new strategy that alters the variable modification approach to consider only curated PTMs at specific positions, thereby avoiding the combinatorial explosion that traditionally leads to high false discovery rates. Using RNA-sequencing-derived databases with this Global Post-Translational Modification (G-PTM) search strategy revealed hundreds of single amino acid variant peptides, tens of novel splice junction peptides, and several hundred posttranslationally modified peptides in each of ten human cell lines.

Integrated Transcriptomic-Proteomic Analysis Using a Proteogenomic Workflow Refines Rat Genome Annotation

Proteogenomic re-annotation and mRNA splicing information can lead to the discovery of various protein forms for eukaryotic model organisms like rat. However, detection of novel proteoforms using mass spectrometry proteomics data remains a formidable challenge. We developed EuGenoSuite, an open source multiple algorithmic proteomic search tool and utilized it in our in-house integrated transcriptomic-proteomic pipeline to facilitate automated proteogenomic analysis. Using four proteogenomic pipelines (integrated transcriptomic-proteomic, Peppy, Enosi, and ProteoAnnotator) on publicly available RNA-sequence and MS proteomics data, we discovered 363 novel peptides in rat brain microglia representing novel proteoforms for 249 gene loci in the rat genome. These novel peptides aided in the discovery of novel exons, translation of annotated untranslated regions, pseudogenes, and splice variants for various loci; many of which have known disease associations, including neurological disorders like schizophrenia, amyotrophic lateral sclerosis, etc. Novel isoforms were also discovered for genes implicated in cardiovascular diseases and breast cancer for which rats are considered model organisms. Our integrative multi-omics data analysis not only enables the discovery of new proteoforms but also generates an improved reference for human disease studies in the rat model.

Peptide identification in “shotgun” proteomics using tandem mass spectrometry: Comparison of search engine algorithms

High-throughput proteomics technologies are gaining popularity in different areas of life sciences. One of the main objectives of proteomics is characterization of the proteins in biological samples using liquid chromatography/mass spectrometry analysis of the corresponding proteolytic peptide mixtures. Both the complexity and the scale of experimental data obtained even from a single experimental run require specialized bioinformatic tools for automated data mining. One of the most important tools is a so-called proteomics search engine used for identification of proteins present in a sample by comparing experimental and theoretical tandem mass spectra. The latter are generated for the proteolytic peptides derived from a protein database. Peptide identifications obtained with the search engine are then scored according to the probability of a correct peptide-spectrum match. The purpose of this work was to perform a comparison of different search algorithms using data acquired for complex protein mixtures, including both annotated protein standards and clinical samples. The comparison was performed for three popular search engines: commercially available Mascot, as well as open-source X!Tandem and OMSSA. It was shown that the search engine OMSSA identifies in general a smaller number of proteins, while X!Tandem and Mascot deliver similar performance. We found no compelling reasons for using the commercial search engine instead of its open source competitor.

Novel methodologies for biomarker discovery in atherosclerosis.

Identification of subjects at increased risk for cardiovascular events plays a central role in the worldwide efforts to improve prevention, prediction, diagnosis, and prognosis of cardiovascular disease and to decrease the related costs. Despite their high predictive value on population level, traditional risk factors fail to fully predict individual risk. This position paper provides a summary of current vascular biomarkers other than the traditional risk factors with a special focus on the emerging -omics technologies. The definition of biomarkers and the identification and use of classical biomarkers are introduced, and we discuss the limitations of current biomarkers such as high sensitivity C-reactive protein (hsCRP) or N-terminal pro-brain natriuretic peptide (NT-proBNP). This is complemented by circulating plasma biomarkers, including high-density lipoprotein (HDL), and the conceptual shift from HDL cholesterol levels to HDL composition/function for cardiovascular risk assessment. Novel sources for plasma-derived markers include microparticles, microvesicles, and exosomes and their use for current omics-based analytics. Measurement of circulating micro-RNAs, short RNA sequences regulating gene expression, has attracted major interest in the search for novel biomarkers. Also, mass spectrometry and nuclear magnetic resonance spectroscopy have become key complementary technologies in the search for new biomarkers, such as proteomic searches or identification and quantification of small metabolites including lipids (metabolomics and lipidomics). In particular, pro-inflammatory lipid metabolites have gained much interest in the cardiovascular field. Our consensus statement concludes on leads and needs in biomarker research for the near future to improve individual cardiovascular risk prediction.

Quantitative proteomics of rat and human pancreatic beta cells.

Data set description: This data set is composed by label-free alternate-scanning LC-MS/MS proteomics analysis human and Wistar rat pancreatic islet endocrine cells.The mass spectrometry data of the human and rat pancreatic beta cells and the resulting proteome search output from ProteinLynx GlobalSERVER (PLGS) have been deposited to the ProteomeXchange Consortium [1] via the PRIDE partner repository with the dataset identifiers PXD001539 (human) and PXD001816 (rat). From these mass spectrometry data, ‘relative molar amount units’ between cell types and across species were calculated.Biological relevance: These data provide a quantitative view on the unfractionated proteomes of human and rat beta and alpha cells. It is likely biased towards the proteins with higher molar abundance, relating to core functional pathways, but also includes several proteins with an islet-enriched expression. The quality of the cell preps is state-of-the-art, and the label-free quantitation is both precise and accurate, allowing detailed quantitative analysis.

Processing Shotgun Proteomics Data on the Amazon Cloud with the Trans-Proteomic Pipeline

Cloud computing, where scalable, on-demand compute cycles and storage are available as a service, has the potential to accelerate mass spectrometry-based proteomics research by providing simple, expandable, and affordable large-scale computing to all laboratories regardless of location or information technology expertise. We present new cloud computing functionality for the Trans-Proteomic Pipeline, a free and open-source suite of tools for the processing and analysis of tandem mass spectrometry datasets. Enabled with Amazon Web Services cloud computing, the Trans-Proteomic Pipeline now accesses large scale computing resources, limited only by the available Amazon Web Services infrastructure, for all users. The Trans-Proteomic Pipeline runs in an environment fully hosted on Amazon Web Services, where all software and data reside on cloud resources to tackle large search studies. In addition, it can also be run on a local computer with computationally intensive tasks launched onto the Amazon Elastic Compute Cloud service to greatly decrease analysis times. We describe the new Trans-Proteomic Pipeline cloud service components, compare the relative performance and costs of various Elastic Compute Cloud service instance types, and present on-line tutorials that enable users to learn how to deploy cloud computing technology rapidly with the Trans-Proteomic Pipeline. We provide tools for estimating the necessary computing resources and costs given the scale of a job and demonstrate the use of cloud enabled Trans-Proteomic Pipeline by performing over 1100 tandem mass spectrometry files through four proteomic search engines in 9 h and at a very low cost.

Construction and assessment of individualized proteogenomic databases for large-scale analysis of nonsynonymous single nucleotide variants.

Next-generation sequencing projects focusing on genomes and transcriptomes identify millions of single nucleotide variants (SNVs), many of which result in single amino acid substitutions. These nonsynonymous (ns) SNVs are typically not incorporated into protein sequence databases used to identify MS/MS data. Here, we perform a comparative analysis of the assembly of nsSNV-containing proteogenomic databases. We use a comprehensive transcriptome and proteome dataset of HeLa cells from the literature to derive and to incorporate SNVs into databases applicable to proteomics search engines, and to assess their performance in the identification of nsSNVs. We assemble the databases by (1) translation of SNV-containing transcripts into all possible reading frames, (2) translation of predicted reading frame, (3) prediction of nsSNVs and subsequent incorporation into canonical protein sequences. We show substantial differences between generated databases in terms of represented nsSNVs and theoretical search space, affecting sensitivity and specificity of database search. We query the databases with >2.2M high-resolution MS/MS spectra using MaxQuant software and identify 451 variant peptides, containing 401 nsSNVs. We conclude that prediction of reading frame and, if applicable, SNV effect result in comprehensive yet compact databases necessary to retain sensitivity in large-scale analysis of nsSNVs called from transcriptomics data.

Identification of potential drug targets by subtractive genome analysis of Bacillus anthracis A0248: An in silico approach

BackgroundBacillus anthracis is a gram positive, spore forming, rod shaped bacteria which is the etiologic agent of anthrax – cutaneous, pulmonary and gastrointestinal. A recent outbreak of anthrax in a tropical region uncovered natural and in vitro resistance against penicillin, ciprofloxacin, quinolone due to over exposure of the pathogen to these antibiotics. This fact combined with the ongoing threat of using B. anthracis as a biological weapon proves that the identification of new therapeutic targets is urgently needed. MethodsIn this computational approach various databases and online based servers were used to detect essential proteins of B. anthracis A0248. Protein sequences of B. anthracis A0248 strain were retrieved from the NCBI database which was then run in CD-hit suite for clustering. NCBI BlastP against the human proteome and similarity search against DEG were done to find out essential human non-homologous proteins. Proteins involved in unique pathways were analyzed using KEGG genome database and PSORTb, CELLO v.2.5, ngLOC – these three tools were used to deduce putative cell surface proteins. ResultsSuccessive analysis revealed 116 proteins to be essential human non-homologs among which 17 were involved in unique metabolic pathways and 28 were predicted as membrane associated proteins. Both types of proteins can be exploited as they are unlikely to have homologous counterparts in the human host. ConclusionBeing human non-homologous, these proteins can be targeted for potential therapeutic drug development in future. Targets on unique metabolic and membrane-bound proteins can block cell wall synthesis, bacterial replication and signal transduction respectively.

Evaluation of proteomic search engines for the analysis of histone modifications.

Identification of histone post-translational modifications (PTMs) is challenging for proteomics search engines. Including many histone PTMs in one search increases the number of candidate peptides dramatically, leading to low search speed and fewer identified spectra. To evaluate database search engines on identifying histone PTMs, we present a method in which one kind of modification is searched each time, for example, unmodified, individually modified, and multimodified, each search result is filtered with false discovery rate less than 1%, and the identifications of multiple search engines are combined to obtain confident results. We apply this method for eight search engines on histone data sets. We find that two search engines, pFind and Mascot, identify most of the confident results at a reasonable speed, so we recommend using them to identify histone modifications. During the evaluation, we also find some important aspects for the analysis of histone modifications. Our evaluation of different search engines on identifying histone modifications will hopefully help those who are hoping to enter the histone proteomics field. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium with the data set identifier PXD001118.

Programmed Cell Death Protein 5 Interacts with the Cytosolic Chaperonin Containing Tailless Complex Polypeptide 1 (CCT) to Regulate β-Tubulin Folding

Programmed cell death protein 5 (PDCD5) has been proposed to act as a pro-apoptotic factor and tumor suppressor. However, the mechanisms underlying its apoptotic function are largely unknown. A proteomics search for binding partners of phosducin-like protein, a co-chaperone for the cytosolic chaperonin containing tailless complex polypeptide 1 (CCT), revealed a robust interaction between PDCD5 and CCT. PDCD5 formed a complex with CCT and β-tubulin, a key CCT-folding substrate, and specifically inhibited β-tubulin folding. Cryo-electron microscopy studies of the PDCD5·CCT complex suggested a possible mechanism of inhibition of β-tubulin folding. PDCD5 bound the apical domain of the CCTβ subunit, projecting above the folding cavity without entering it. Like PDCD5, β-tubulin also interacts with the CCTβ apical domain, but a second site is found at the sensor loop deep within the folding cavity. These orientations of PDCD5 and β-tubulin suggest that PDCD5 sterically interferes with β-tubulin binding to the CCTβ apical domain and inhibits β-tubulin folding. Given the importance of tubulins in cell division and proliferation, PDCD5 might exert its apoptotic function at least in part through inhibition of β-tubulin folding.

Proteome profile and proteogenomics of the organohalide-respiring bacterium Dehalococcoides mccartyi strain CBDB1 grown on hexachlorobenzene as electron acceptor

Dehalococcoides mccartyi strain CBDB1 is an obligate organohalide-respiring anaerobic bacterium that is able to transform a wide range of highly relevant halogenated organic contaminants and has been detected in natural and engineered environments. In order to understand the molecular principles, we generated a protein profile of CBDB1 cultivated in a synthetic anaerobic medium with hexachlorobenzene as the sole electron acceptor. The proteome characterization through two different proteomic techniques resulted in 8491 distinct peptides corresponding to 1023 proteins that covered 70% of the annotated 1458 protein-coding sequences. From the 32 annotated reductive dehalogenases homologous genes (rdhA) we were able to detect 16 RdhA proteins. High-quality MS spectra were further used to build a spectra library to provide a platform for precise and fast proteomic searches. In addition, the derived proteome data set was used to apply a proteogenomic approach that led to a refinement of genome annotation of CBDB1. This was mostly due to protein prolongation (13 cases) and detection of translated sequences without an obvious link to existing annotations (5 cases). These data go beyond the current knowledge of the bacterial proteome and provide an expanded platform for a better understanding of the functional cellular. Anaerobic reductive dehalogenation of chlorinated organic molecules is a key process for the recycling of halogenated organic substances in natural habitats and highly relevant for groundwater bioremediation. The ability to couple reductive dehalogenation of halogenated compounds to ATP-generation via a respiratory chain is one remarkable capability prevalent in the model organism D. mccartyi strain CBDB1. Therefore, a reference proteome map was generated by comprehensive LC-MS analysis and converted into a SpectraST library to provide a platform for precise and fast shotgun proteomic searches. A proteogenomics approach led to a refinement of gene annotation of CBDB1 and will extend the current knowledge of the bacterial proteome. The approach provides an expanded platform for a better understanding of the functional cellular metabolism and thereby constitutes a tool for comparative studies of protein expression as well as further research on organohalide respiration.

N- and O-Glycosylation in the Murine Synaptosome

We present the first large scale study characterizing both N- and O-linked glycosylation in a site-specific manner on hundreds of proteins. We demonstrate that a lectin-affinity fractionation step using wheat germ agglutinin enriches not only peptides carrying intracellular O-GlcNAc, but also those bearing ER/Golgi-derived N- and O-linked carbohydrate structures. Liquid chromatography-MS (LC/MS) analysis with high accuracy precursor mass measurements and high sensitivity ion trap electron-transfer dissociation (ETD) were utilized for structural characterization of glycopeptides. Our results reveal both the identity of the precise sites of glycosylation and information on the oligosaccharide structures possible on these proteins. We report a novel iterative approach that allowed us to interpret the ETD data set directly without making prior assumptions about the nature and distribution of oligosaccharides present in our glycopeptide mixture. Over 2500 unique N- and O-linked glycopeptides were identified on 453 proteins. The extent of microheterogeneity varied extensively, and up to 19 different oligosaccharides were attached at a given site. We describe the presence of the well-known mucin-type structures for O-glycosylation, an EGF-domain-specific fucosylation and a rare O-mannosylation on the transmembrane phosphatase Ptprz1. Finally, we identified three examples of O-glycosylation on tyrosine residues.

Proteomic analysis of Girdin-interacting proteins in migrating new neurons in the postnatal mouse brain

Neural stem cells continuously generate new neurons in the ventricular–subventricular zone (V–SVZ) of the postnatal and adult mammalian brain. New neurons born in the rodent V–SVZ migrate toward the olfactory bulb (OB), where they differentiate into interneurons. To reveal novel intracellular molecular mechanisms that control postnatal neuronal migration, we performed a global proteomic search for proteins interacting with Girdin, an essential protein for postnatal neuronal migration. Using GST pull-down and LC–MS/MS shotgun analysis, we identified cytoskeletal proteins, cytoskeleton-binding proteins, and signal-transduction proteins as possible participants in neuronal migration. Our results suggest that Girdin and Girdin-interacting proteins control neuronal migration by regulating actin and/or microtubule dynamics.

Host-centric Proteomics of Stool: A Novel Strategy Focused on intestinal Responses to the Gut Microbiota

The diverse community of microbes that inhabits the human bowel is vitally important to human health. Host-expressed proteins are essential for maintaining this mutualistic relationship and serve as reporters on the status of host-microbiota interaction. Therefore, unbiased and sensitive methods focused on host proteome characterization are needed. Herein we describe a novel method for applying shotgun proteomics to the analysis of feces, focusing on the secreted host proteome. We have conducted the most complete analysis of the extracellular mouse gut proteome to date by employing a gnotobiotic mouse model. Using mice colonized with defined microbial communities of increasing complexity or a complete human microbiota ('humanized'), we show that the complexity of the host stool proteome mirrors the complexity of microbiota composition. We further show that host responses exhibit signatures specific to the different colonization states. We demonstrate feasibility of this approach in human stool samples and provide evidence for a "core" stool proteome as well as personalized host response features. Our method provides a new avenue for noninvasive monitoring of host-microbiota interaction dynamics via host-produced proteins in stool.

CustomProDB: an R package to generate customized protein databases from RNA-Seq data for proteomics search

Summary: Database search is the most widely used approach for peptide and protein identification in mass spectrometry-based proteomics studies. Our previous study showed that sample-specific protein databases derived from RNA-Seq data can better approximate the real protein pools in the samples and thus improve protein identification. More importantly, single nucleotide variations, short insertion and deletions and novel junctions identified from RNA-Seq data make protein database more complete and sample-specific. Here, we report an R package customProDB that enables the easy generation of customized databases from RNA-Seq data for proteomics search. This work bridges genomics and proteomics studies and facilitates cross-omics data integration.Availability and implementation: customProDB and related documents are freely available at http://bioconductor.org/packages/2.13/bioc/html/customProDB.html.Contact: bing.zhang@vanderbilt.eduSupplementary information: Supplementary data are available at Bioinformatics online.

Latent Transforming growth factor Binding Protein 2 (LTBP2) is related to phenotypic changes suggestive for HFPEF in stable outpatients with coronary artery disease and preserved ejection fraction

Purpose: The Latent Transforming growth factor Binding Protein 2 (LTBP2) was recently discovered in an unbiased proteomics search for novel markers in patients with acute dyspnea. The highest levels of LTBP2 are found in lung tissue and plasma levels of LTBP2 are a powerful predictor of all-cause mortality in patients presenting with acute dyspnea. We evaluated the relationship between plasma levels of LTBP2 and phenotypic changes suggestive for heart failure with preserved ejection fraction (HFPEF) including parameters of neurohormonal activation, Left Ventricular (LV) remodeling and filling pressures as well as exercise capacity. Methods: We studied 136 stable patients with CAD and LVEF≥50% (mean age 68±8 years, 76% men, LVEF 63±8%, 37% in NYHA class II/III). All patients underwent resting echocardiography for evaluation of LVEF and volumes, TDI derived longitudinal systolic and diastolic velocities function and E/E', as well as a six minute walking distance test (6minWD), maximal bicycle spiroergometry (peakVO2 and VE/VCO2 slope, n=104) and blood sampling. A proteomic approach for antibody-free targeted protein quantification based on high-end mass spectrometry was used to measure LTBP2 levels. Results: Levels of LTBP2 were higher in NYHA II/III as compared to NYHA I patients (p<0.05) and were correlated with NT-proBNP levels (r=0.49, p<0.01). ROC analysis showed a comparable modest AUC of 0.61 (p<0.05) for NT-proBNP and 0.60 (p<0.05) for LTBP2 to discern NYHA I versus NYHA II/III patients. LTBP2 levels were correlated with LV volumes (r=-0.38, p<0.01 for LVEDV, r=-0.33, p<0.01 for LVESV) as well as with LV longitudinal systolic velocities (r=-0.35, p<0.01), LV early diastolic velocities (r=-0.31,p<0.01) and E/E' (r=0.28,p<0.01). In linear regression analysis, including age and creatinine, LVESV and E/E' were independent predictors for LTBP2 levels (all p<0.05). Higher levels of LTBP2 were associated with lower 6minWD (r=-0.43, p<0.001), lower peakVO2 (r=-0.38, p<0.001) and higher VE/VCO2 slope (r=0.36, p<0.001). In multiple linear regression analysis including age, creatinine, LVESV and E/E', LTBP2 levels appeared to be an independent predictor for both 6minWD and peakVO2 (p<0.01). Conclusion: In stable patients with CAD and preserved LV function, higher levels of LTBP2 are related to phenotypic changes suggestive for HFPEF including higher neurohormonal activation, adverse LV remodeling with loss of longitudinal function, higher LV filling pressures and a lower exercise capacity. Therefore, the role of LTBP2 in the development, diagnosis and outcome of HFPEF deserves further study.

Evolution of Chloroplast J Proteins

Hsp70 chaperones are involved in multiple biological processes and are recruited to specific processes by designated J domain-containing cochaperones, or J proteins. To understand the evolution and functions of chloroplast Hsp70s and J proteins, we identified the Arabidopsis chloroplast J protein constituency using a combination of genomic and proteomic database searches and individual protein import assays. We show that Arabidopsis chloroplasts have at least 19 J proteins, the highest number of confirmed J proteins for any organelle. These 19 J proteins are classified into 11 clades, for which cyanobacteria and glaucophytes only have homologs for one clade, green algae have an additional three clades, and all the other 7 clades are specific to land plants. Each clade also possesses a clade-specific novel motif that is likely used to interact with different client proteins. Gene expression analyses indicate that most land plant-specific J proteins show highly variable expression in different tissues and are down regulated by low temperatures. These results show that duplication of chloroplast Hsp70 in land plants is accompanied by more than doubling of the number of its J protein cochaperones through adding new J proteins with novel motifs, not through duplications within existing families. These new J proteins likely recruit chloroplast Hsp70 to perform tissue specific functions related to biosynthesis rather than to stress resistance.