Related Articles

The translation efficiency of most Saccharomyces cerevisiae genes remains fairly constant across poor and rich growth media. This observation has led us to revisit the available data and to examine the potential utility of a protein abundance predictor in reinterpreting existing mRNA expression data. Our predictor is based on large-scale data of mRNA levels, the tRNA adaptation index, and the evolutionary rate. It attains a correlation of 0.76 with experimentally determined protein abundance levels on unseen data and successfully cross-predicts protein abundance levels in another yeast species (Schizosaccharomyces pombe). The predicted abundance levels of proteins in known S. cerevisiae complexes, and of interacting proteins, are significantly more coherent than their corresponding mRNA expression levels. Analysis of gene expression measurement experiments using the predicted protein abundance levels yields new insights that are not readily discernable when clustering the corresponding mRNA expression levels. Comparing protein abundance levels across poor and rich media, we find a general trend for homeostatic regulation where transcription and translation change in a reciprocal manner. This phenomenon is more prominent near origins of replications. Our analysis shows that in parallel to the adaptation occurring at the tRNA level via the codon bias, proteins do undergo a complementary adaptation at the amino acid level to further increase their abundance.

Author Summary

DNA microarrays measuring gene expression levels have been a mainstay of systems biology research, but since proteins are more direct mediators of cellular processes, protein abundance levels are likely to be a better indicator of the cellular state. However, as proteomic measurements are still lagging behind gene expression measurements, there has been considerable effort in recent years to study the correlations between gene expression (and a plethora of protein characteristics) and protein abundance. Addressing this challenge, the current study is one of the first to introduce a predictor for protein abundance levels that is tested and validated on unseen data using all currently available large-scale proteomic data. The utility of this predictor is shown via a comprehensive set of tests and applications, including improved functional coherency of complexes and interacting proteins, better fit with gene phenotypic data, cross-species prediction of protein abundance, and most importantly, the reinterpretation of existing gene expression microarray data. Finally, our revisit and analysis of the existing large-scale proteomic data reveals new key insights concerning the regulation of translation efficiency and its evolution. Overall, a solid protein abundance prediction tool is invaluable for advancing our understanding of cellular processes; this study presents a further step in this direction.

The nematode Caenorhabditis elegans is a popular model system in genetics, not least because a majority of human disease genes are conserved in C. elegans. To generate a comprehensive inventory of its expressed proteome, we performed extensive shotgun proteomics and identified more than half of all predicted C. elegans proteins. This allowed us to confirm and extend genome annotations, characterize the role of operons in C. elegans, and semiquantitatively infer abundance levels for thousands of proteins. Furthermore, for the first time to our knowledge, we were able to compare two animal proteomes (C. elegans and Drosophila melanogaster). We found that the abundances of orthologous proteins in metazoans correlate remarkably well, better than protein abundance versus transcript abundance within each organism or transcript abundances across organisms; this suggests that changes in transcript abundance may have been partially offset during evolution by opposing changes in protein abundance.

Author Summary

Proteins are the active players that execute the genetic program of a cell, and their levels and interactions are precisely controlled. Routinely monitoring thousands of proteins is difficult, as they can be present at vastly different abundances, come with various sizes, shapes, and charge, and have a more complex alphabet of twenty “letters,” in contrast to the four letters of the genome itself. Here, we used mass spectrometry to extensively characterize the proteins of a popular model organism, the nematode Caenorhabditis elegans. Together with previous data from the fruit fly Drosophila melanogaster, this allows us to compare the protein levels of two animals on a global scale. Surprisingly, we find that individual protein abundance is highly conserved between the two species. So, although worms and flies look very different, they need similar amounts of each conserved, orthologous protein. Because many C. elegans and D. melanogaster proteins also have counterparts in humans, our results suggest that similar rules may apply to our own proteins.

A quantitative comparison of two animal proteomes shows a striking correlation of protein abundance levels, a better correlation than transcript levels. Are the latter more variable during evolution?

A proteome of the bio-entity, including cell, tissue, organ, and organism, consists of proteins of diverse abundance. The principle that determines the abundance of different proteins in a proteome is of fundamental significance for an understanding of the building blocks of the bio-entity. Here, we report three regular patterns in the proteome-wide distribution of protein abundance across species such as human, mouse, fly, worm, yeast, and bacteria: in most cases, protein abundance is positively correlated with the protein's origination time or sequence conservation during evolution; it is negatively correlated with the protein's domain number and positively correlated with domain coverage in protein structure, and the correlations became stronger during the course of evolution; protein abundance can be further stratified by the function of the protein, whereby proteins that act on material conversion and transportation (mass category) are more abundant than those that act on information modulation (information category). Thus, protein abundance is intrinsically related to the protein's inherent characters of evolution, structure, and function.

URH49 is a mammalian protein that is 90% identical to the DExH/D box protein UAP56, an RNA helicase that is important for splicing and nuclear export of mRNA. Although Saccharomyces cerevisiae and Drosophila express only a single protein corresponding to UAP56, mRNAs encoding URH49 and UAP56 are both expressed in human and mouse cells. Both proteins interact with the mRNA export factor Aly and both are able to rescue the loss of Sub2p (the yeast homolog of UAP56), indicating that both proteins have similar functions. UAP56 mRNA is more abundant than URH49 mRNA in many tissues, although in testes URH49 mRNA is much more abundant. UAP56 and URH49 mRNAs are present at similar levels in proliferating cultured cells. However, when the cells enter quiescence, the URH49 mRNA level decreases 3–6-fold while the UAP56 mRNA level remains relatively constant. The amount of URH49 mRNA increases to the level found in proliferating cells within 5 h when quiescent cells are growth-stimulated or when protein synthesis is inhibited. URH49 mRNA is relatively unstable (T½ = 4 h) in quiescent cells, but is stabilized immediately following growth stimulation or inhibition of protein synthesis. In contrast, there is much less change in the content or stability of UAP56 mRNA following growth stimulation. Our observations suggest that in mammalian cells, two UAP56-like RNA helicases are involved in splicing and nuclear export of mRNA. Differential expression of these helicases may lead to quantitative or qualitative changes in mRNA expression.

We review the results of attempts to correlate protein abundance with mRNA expression levels, focusing on yeast.

Attempts to correlate protein abundance with mRNA expression levels have had variable success. We review the results of these comparisons, focusing on yeast. In the process, we survey experimental techniques for determining protein abundance, principally two-dimensional gel electrophoresis and mass-spectrometry. We also merge many of the available yeast protein-abundance datasets, using the resulting larger 'meta-dataset' to find correlations between protein and mRNA expression, both globally and within smaller categories.

Several epidemiological models predict a positive relationship between host population density and abundance of directly transmitted macroparasites. Here, we generalize these, and test the prediction by a comparative study. We used data on communities of gastrointestinal strongylid nematodes from 19 mammalian species, representing examination of 6670 individual hosts. We studied both the average abundance of all strongylid nematodes within a host species, and the two components of abundance, prevalence and intensity. The effects of host body weight, diet, fecundity and age at maturity and parasite body size were controlled for directly, and the phylogenetically independent contrast method was used to control for confounding factors more generally. Host population density and average parasite abundance were strongly positively correlated within mammalian taxa, and across all species when the effects of host body weight were controlled for. Controlling for other variables did not change this. Even when looking at single parasite species occurring in several host species, abundance was highest in the host species with the highest population density. Prevalence and intensity showed similar patterns. These patterns provide the first macroecological evidence consistent with the prediction that transmission rates depend on host population density in natural parasite communities.

A long-standing assumption in evolutionary biology is that the evolution rate of protein-coding genes depends, largely, on specific constraints that affect the function of the given protein. However, recent research in evolutionary systems biology revealed unexpected, significant correlations between evolution rate and characteristics of genes or proteins that are not directly related to specific protein functions, such as expression level and protein–protein interactions. The strongest connections were consistently detected between protein sequence evolution rate and the expression level of the respective gene. A recent genome-wide proteomic study revealed an extremely strong correlation between the abundances of orthologous proteins in distantly related animals, the nematode Caenorhabditis elegans and the fruit fly Drosophila melanogaster. We used the extensive protein abundance data from this study along with short-term evolutionary rates (ERs) of orthologous genes in nematodes and flies to estimate the relative contributions of structural–functional constraints and the translation rate to the evolution rate of protein-coding genes. Together the intrinsic constraints and translation rate account for approximately 50% of the variance of the ERs. The contribution of constraints is estimated to be 3- to 5-fold greater than the contribution of translation rate.

Background

Phospholipid hydroperoxide glutathione peroxidases (PHGPx), the most abundant isoforms of GPx families, interfere directly with hydroperoxidation of lipids. Biochemical properties of these proteins vary along with their donor organisms, which has complicated the phylogenetic classification of diverse PHGPx-like proteins. Despite efforts for comprehensive analyses, the evolutionary aspects of GPx genes in invertebrates remain largely unknown.

Results

We isolated GPx homologs via in silico screening of genomic and/or expressed sequence tag databases of eukaryotic organisms including protostomian species. Genes showing strong similarity to the mammalian PHGPx genes were commonly found in all genomes examined. GPx3- and GPx7-like genes were additionally detected from nematodes and platyhelminths, respectively. The overall distribution of the PHGPx-like proteins with different biochemical properties was biased across taxa; selenium- and glutathione (GSH)-dependent proteins were exclusively detected in platyhelminth and deuterostomian species, whereas selenium-independent and thioredoxin (Trx)-dependent enzymes were isolated in the other taxa. In comparison of genomic organization, the GSH-dependent PHGPx genes showed a conserved architectural pattern, while their Trx-dependent counterparts displayed complex exon-intron structures. A codon for the resolving Cys engaged in reductant binding was found to be substituted in a series of genes. Selection pressure to maintain the selenocysteine codon in GSH-dependent genes also appeared to be relaxed during their evolution. With the dichotomized fashion in genomic organizations, a highly polytomic topology of their phylogenetic trees implied that the GPx genes have multiple evolutionary intermediate forms.

Conclusion

Comparative analysis of invertebrate GPx genes provides informative evidence to support the modular pathways of GPx evolution, which have been accompanied with sporadic expansion/deletion and exon-intron remodeling. The differentiated enzymatic properties might be acquired by the evolutionary relaxation of selection pressure and/or biochemical adaptation to the acting environments. Our present study would be beneficial to get detailed insights into the complex GPx evolution, and to understand the molecular basis of the specialized physiological implications of this antioxidant system in their respective donor organisms.

By characterizing dynamic changes in yeast protein abundance following osmotic shock, this study shows that the correlation between protein and mRNA differs for transcripts that increase versus decrease in abundance, and reveals physiological reasons for these differences.

The correlation between protein and mRNA change is very high at transcripts that increase in abundance, but negligible at reduced transcripts following NaCl shock.Modeling and experimental data suggest that reducing levels of high-abundance transcripts helps to direct translational machinery to newly made transcripts.The transient burst of transcript increase serves to accelerate changes in protein abundance.Post-transcriptional regulation of protein abundance is pervasive, although most of the variance in protein change is explained by changes in mRNA abundance.

Natural microenvironments change rapidly, and living creatures must respond quickly and efficiently to thrive within this flux. At all cellular levels—signaling, transcription, translation, metabolism, cell growth, and division—the response is dynamic and coordinated. Some aspects of this response, such as dynamic changes of the transcriptome, are well understood. But other aspects, like the response of the proteome, have remained obscured primarily because of previous limitations in technology. Without coordinated time-course data, it has remained impossible to correctly characterize the correlations and dependencies between these two essential levels of cell biology.

This work presents an extended picture of the coordinated response of the transcriptome and proteome as cells respond to an abrupt environmental change. To assay proteomic dynamics, we developed a strategy for large-scale, multiplexed quantitation using isobaric tags and high mass accuracy mass spectrometry. This sensitive yet efficient platform allows for the expedient collection of quantitative time-course proteomic data at six time points, sufficiently reproducible to permit meaningful interpretation of variation across biological replicates. Time-course transcriptome data were generated from paired biological samples, allowing us to examine the relationships between changes in mRNA and protein for each gene in terms of direction and intensity, as well as the characteristics of the temporal profiles for each gene.

It was immediately obvious that a single measure of correlation across the entire data set was a meaningless metric. We therefore analyzed relationships between mRNA and protein for different subsets of data. In response to osmotic shock, hundreds of transcripts are highly induced, and their temporal pattern reveals a transient peak of maximal induction, which resolves into a new elevated level as cells acclimate (Figure 2). For this group of genes, there is extremely high correlation between peak mRNA change and protein change (R2∼0.8). But the dynamics of the molecules differ: while mRNA levels transiently overshoot their final levels, proteins gradually rise in abundance toward their new, elevated state. We observed, however, that a measure of efficiency connects the two profiles. The time it takes for a protein to acclimate to its new state correlates with the magnitude of the excess mRNA induction. Thus, the cell imparts an urgency to protein induction by transiently producing excess transcript.

The most surprising result, however, involves transcripts that decrease in abundance. In response to osmotic shock, the cell transiently reduces over 600 transcripts, many of which are among the most highly expressed in unstressed cells. But protein levels for these genes remain, for the most part, almost completely unchanged. The stark absence of protein repression is independent of basal protein abundance, independent of reported protein half-lives, reproducible across biological replicates, and validated by quantitative western blots. Furthermore, since we do detect a handful of proteins whose abundance is significantly reduced, our technology is capable of identifying protein loss. Thus, we conclude that transcript reduction serves another purpose besides reducing protein levels.

To explore alternate interpretations of the consequence of transcriptional repression, we devised a mass-action kinetic model, which describes protein changes based on mRNA dynamics in the context of transient changes in the rates of cell division. The model successfully recapitulated the observed data, allowing us to alter modeling parameters to test various hypotheses.

In response to osmotic shock, overall rates of translation temporarily decrease and cell growth transiently arrests before resuming at a slower rate. We reasoned that mRNA reduction might lower the rate of new protein synthesis, but that retarded production is balanced by reduced cell division. We explored both aspects of this logic with our model.

As expected, removing cell division from our model led to a calculated decrease of protein levels, indicating that reduced growth is necessary for maintaining protein levels. However, when we computationally held mRNA levels stable and calculated protein levels in the absence of mRNA repression, we did not find the expected increase in protein abundance.

We then considered the possibility that one function of the regulated repression of these highly abundant transcripts was to liberate proteins essential for translation, such as ribosomes or translation initiation factors. To explore this, we examined a mutant lacking the Dot6p/Tod6p transcriptional repressors, which fails to properly repress ∼250 genes in response to osmotic shock. In the wild type, the mRNA for a Dot6p/Tod6p target (ARX1) decreased seven-fold, and the remaining transcript was generally unassociated with poly-ribosomes. In the mutant, however, the mRNA levels were reduced only two-fold, while the remaining transcript continued to bind ribosomes. Therefore, failure to reduce transcript levels led to a persistent association with poly-ribosomes, thereby consuming translational machinery.

Our hypothesis is, therefore, that widespread changes in the transcriptome promote efficient translation of new proteins. Transcript increase serves to increase abundance of the encoded proteins, while reduction of some of the most abundant and highly translated mRNAs supports this project by liberating translational capacity. While it is not clear what factors are the limiting elements, it is clear that a full picture of cellular biology requires exploring the dynamics of the cellular response.

The transcriptome and proteome change dynamically as cells respond to environmental stress; however, prior proteomic studies reported poor correlation between mRNA and protein, rendering their relationships unclear. To address this, we combined high mass accuracy mass spectrometry with isobaric tagging to quantify dynamic changes in ∼2500 Saccharomyces cerevisiae proteins, in biological triplicate and with paired mRNA samples, as cells acclimated to high osmolarity. Surprisingly, while transcript induction correlated extremely well with protein increase, transcript reduction produced little to no change in the corresponding proteins. We constructed a mathematical model of dynamic protein changes and propose that the lack of protein reduction is explained by cell-division arrest, while transcript reduction supports redistribution of translational machinery. Furthermore, the transient ‘burst' of mRNA induction after stress serves to accelerate change in the corresponding protein levels. We identified several classes of post-transcriptional regulation, but show that most of the variance in protein changes is explained by mRNA. Our results present a picture of the coordinated physiological responses at the levels of mRNA, protein, protein-synthetic capacity, and cellular growth.

Despite the fact that tRNA abundances are thought to play a major role in determining translation error rates, their distribution across the genetic code and the resulting implications have received little attention. In general, studies of codon usage bias (CUB) assume that codons with higher tRNA abundance have lower missense error rates. Using a model of protein translation based on tRNA competition and intra-ribosomal kinetics, we show that this assumption can be violated when tRNA abundances are positively correlated across the genetic code. Examining the distribution of tRNA abundances across 73 bacterial genomes from 20 different genera, we find a consistent positive correlation between tRNA abundances across the genetic code. This work challenges one of the fundamental assumptions made in over 30 years of research on CUB that codons with higher tRNA abundances have lower missense error rates and that missense errors are the primary selective force responsible for CUB.

Author Summary

Codon usage bias (CUB) is a ubiquitous and important phenomenon. CUB is thought to be driven primarily due to selection against missense errors. For over 30 years, the standard model of translation errors has implicitly assumed that the relationship between translation errors and tRNA abundances are inversely related. This is based on an implicit and unstated assumption that the distribution of tRNA abundances across the genetic code are uncorrelated. Examining these abundance distributions across 73 bacterial genomes from 20 different genera, we find a consistent positive correlation between tRNA abundances across the genetic code. We further show that codons with higher tRNA abundances are not always “optimal” with respect to reducing the missense error rate and hence cannot explain the observed patterns of CUB.

The analysis of a human thyroid serial analysis of gene expression (SAGE) library shows the presence of an abundant SAGE tag corresponding to the mRNA of thyroglobulin (TG). Additional, less abundant tags are present that can not be linked to any other known gene, but show considerable homology to the wild-type TG tag. To determine whether these tags represent TG mRNA molecules with alternative cleavage, 3′-RACE clones were sequenced. The results show that the three putative TG SAGE tags can be attributed to TG transcripts and reflect the use of alternative polyadenylation cleavage sites downstream of a single polyadenylation signal in vivo. By screening more than 300 000 sequences corresponding to human, mouse and rat transcripts for this phenomenon we show that a considerable percentage of mRNA transcripts (44% human, 22% mouse and 22% rat) show cleavage site heterogeneity. When analyzing SAGE-generated expression data, this phenomenon should be considered, since, according to our calculations, 2.8% of human transcripts show two or more different SAGE tags corresponding to a single gene because of alternative cleavage site selection. Both experimental and in silico data show that the selection of the specific cleavage site for poly(A) addition using a given polyadenylation signal is more variable than was previously thought.

The two-dimensional gel electrophoresis of polypeptides synthesized in vitro from poly(A)+ RNA showed that mRNA populations change during sporulation of Physarum polycephalum. The differential hybridization of a cDNA library prepared from poly(A)+ RNA isolated from sporulating cells revealed that of 846 clones, 64 corresponded to sporulation-specific mRNAs. Further analysis demonstrated that these clones contained seven different sequences: three abundant sequences composing 3.2, 1.8, and 1.2% of the library and four other less abundant sequences. It is probable that all the major mRNAs specifically expressed in early stages of sporulation were identified. The most abundant mRNA from this group coded for a hydrophobic protein that contained a signal peptide. This protein is 47% similar to another Physarum protein, which was encoded by the most abundant plasmodium-specific mRNA. The plasmodial mRNA was degraded during sporulation and was replaced by the sporulation mRNA. These two proteins are thus encoded by members of a gene family whose expression is developmentally regulated.

Images

The folding of many cellular proteins occurs co-translationally immediately outside the ribosome exit tunnel, where ribosomal proteins and other associated factors coordinate the synthesis and folding of newly translated polypeptides. Here, we show that the large subunit protein L29, which forms part of the exit tunnel in Escherichia coli, is required for the productive synthesis of an array of structurally diverse recombinant proteins including the green fluorescent protein (GFP) and an intracellular single-chain Fv antibody. Surprisingly, the corresponding mRNA transcript level of these proteins was markedly less abundant in cells lacking L29, suggesting an unexpected regulatory mechanism that links defects in the exit tunnel to the expression of genetic information. To further highlight the importance of L29 in maintaining protein expression, we used mutagenesis and selection to obtain L29 variants that enhanced GFP expression. Overall, our results suggest that the ribosomal exit tunnel proteins may be key targets for optimizing the overproduction of active, structurally complex recombinant proteins in bacterial cells.

Background

Techniques for measuring protein abundance are rapidly advancing and we are now in a situation where we anticipate many protein abundance data sets will be available in the near future. Since proteins are translated from mRNAs, their expression is expected to be related to their abundance, to some degree.

Results

We have developed a web tool, called PARE (Protein Abundance and mRNA Expression; ), to correlate these two quantities. In addition to globally comparing the quantities of protein and mRNA, PARE allows users to select subsets of proteins for focused study (based on functional categories and complexes). Furthermore, it highlights correlation outliers, which are potentially worth further examination.

Conclusion

We anticipate PARE will facilitate comparative studies on mRNA and protein abundance by the proteomics community.

Background

Growth of the ocean's most abundant primary producer, the cyanobacterium Prochlorococcus, is tightly synchronized to the natural 24-hour light-dark cycle. We sought to quantify the relationship between transcriptome and proteome dynamics that underlie this obligate photoautotroph's highly choreographed response to the daily oscillation in energy supply.

Methodology/Principal Findings

Using RNA-sequencing transcriptomics and mass spectrometry-based quantitative proteomics, we measured timecourses of paired mRNA-protein abundances for 312 genes every 2 hours over a light-dark cycle. These temporal expression patterns reveal strong oscillations in transcript abundance that are broadly damped at the protein level, with mRNA levels varying on average 2.3 times more than the corresponding protein. The single strongest observed protein-level oscillation is in a ribonucleotide reductase, which may reflect a defense strategy against phage infection. The peak in abundance of most proteins also lags that of their transcript by 2–8 hours, and the two are completely antiphase for some genes. While abundant antisense RNA was detected, it apparently does not account for the observed divergences between expression levels. The redirection of flux through central carbon metabolism from daytime carbon fixation to nighttime respiration is associated with quite small changes in relative enzyme abundances.

Conclusions/Significance

Our results indicate that expression responses to periodic stimuli that are common in natural ecosystems (such as the diel cycle) can diverge significantly between the mRNA and protein levels. Protein expression patterns that are distinct from those of cognate mRNA have implications for the interpretation of transcriptome and metatranscriptome data in terms of cellular metabolism and its biogeochemical impact.

The abundance and cell cycle dependent expression of the mRNA for human nonhistone protein HMG-17 were studied in synchronized HeLa cells. Slot blot analysis indicates that the HMG-17 mRNA is a very abundant message, significantly more so than histone or actin mRNA. RNA prepared from tissue culture cells contains higher amounts of HMG-17 transcripts than RNA prepared from liver suggesting a correlation between the rate of cell division and HMG-17 mRNA levels. HMG-17 mRNA is present in the cells throughout the cell cycle however there is a significant increase in the mRNA levels late in S phase suggesting that the protein is deposited on chromatin after nucleosome assembly. Synthesis of the HMG-17 transcript is not coupled to DNA replication suggesting that the cell cycle related expression during late S phase is regulated in a different manner from that of the nucleosomal histones.

Images

PARN is one of several deadenylase enzymes present in mammalian cells, and as such the contribution it makes to the regulation of gene expression is unclear. To address this, we performed global mRNA expression and half-life analysis on mouse myoblasts depleted of PARN. PARN knockdown resulted in the stabilization of 40 mRNAs, including that encoding the mRNA decay factor ZFP36L2. Additional experiments demonstrated that PARN knockdown induced an increase in Zfp36l2 poly(A) tail length as well as increased translation. The elements responsible for PARN-dependent regulation lie within the 3′ UTR of the mRNA. Surprisingly, changes in mRNA stability showed an inverse correlation with mRNA abundance; stabilized transcripts showed either no change or a decrease in mRNA abundance. Moreover, we found that stabilized mRNAs had reduced accumulation of pre–mRNA, consistent with lower transcription rates. This presents compelling evidence for the coupling of mRNA decay and transcription to buffer mRNA abundances. Although PARN knockdown altered decay of relatively few mRNAs, there was a much larger effect on global gene expression. Many of the mRNAs whose abundance was reduced by PARN knockdown encode factors required for cell migration and adhesion. The biological relevance of this observation was demonstrated by the fact that PARN KD cells migrate faster in wound-healing assays. Collectively, these data indicate that PARN modulates decay of a defined set of mRNAs in mammalian cells and implicate this deadenylase in coordinating control of genes required for cell movement.

Author Summary

Almost all cellular mRNAs terminate in a 3′ poly(A) tail, the removal of which can induce both translational silencing and mRNA decay. Mammalian cells encode many poly(A)-specific exoribonucleases, but their individual roles are poorly understood. Here, we undertook an analysis of the role of PARN deadenylase in mouse myoblasts using global measurements of mRNA decay rates. Our results reveal that a discrete set of mRNAs exhibit altered mRNA decay as a result of PARN depletion and that stabilization is associated with increased poly(A) tail length and translation efficiency. We determined that stabilization of mRNAs does not generally result in their increased abundance, supporting the idea that mRNA decay is coupled to transcription. Importantly, knockdown of PARN has wide ranging effects on gene expression that specifically impact the extracellular matrix and cell migration.

We provide a large-scale dataset on absolute protein and matching mRNA concentrations from the human medulloblastoma cell line Daoy. The correlation between mRNA and protein concentrations is significant and positive (Rs=0.46, R2=0.29, P-value<2e16), although non-linear.Out of ∼200 tested sequence features, sequence length, frequency and properties of amino acids, as well as translation initiation-related features are the strongest individual correlates of protein abundance when accounting for variation in mRNA concentration.When integrating mRNA expression data and all sequence features into a non-parametric regression model (Multivariate Adaptive Regression Splines), we were able to explain up to 67% of the variation in protein concentrations. Half of the contributions were attributed to mRNA concentrations, the other half to sequence features relating to regulation of translation and protein degradation. The sequence features are primarily linked to the coding and 3′ untranslated region. To our knowledge, this is the most comprehensive predictive model of human protein concentrations achieved so far.

mRNA decay, translation regulation and protein degradation are essential parts of eukaryotic gene expression regulation (Hieronymus and Silver, 2004; Mata et al, 2005), which enable the dynamics of cellular systems and their responses to external and internal stimuli without having to rely exclusively on transcription regulation. The importance of these processes is emphasized by the generally low correlation between mRNA and protein concentrations. For many prokaryotic and eukaryotic organisms, <50% of variation in protein abundance variation is explained by variation in mRNA concentrations (de Sousa Abreu et al, 2009).

Given the plethora of regulatory mechanisms involved, most studies have focused so far on individual regulators and specific targets. Particularly in human, we currently lack system-wide, quantitative analyses that evaluate the relative contribution of regulatory elements encoded in the mRNA and protein sequence. Existing studies have been carried out only in bacteria and yeast (Nie et al, 2006; Brockmann et al, 2007; Tuller et al, 2007; Wu et al, 2008). Here, we present the first comprehensive analysis on the impact of translation and protein degradation on protein abundance variation in a human cell line. For this purpose, we experimentally measured absolute protein and mRNA concentrations in the Daoy medulloblastoma cell line, using shotgun proteomics and microarrays, respectively (Figure 1). These data comprise one of the largest such sets available today for human. We focused on sequence features that likely impact protein translation and protein degradation, including length, nucleotide composition, structure of the untranslated regions (UTRs), coding sequence, composition of the translation initiation site, presence of upstream open reading frames putative target sites of miRNAs, codon usage, amino-acid composition and protein degradation signals.

Three types of tests have been conducted: (a) we examined partial Spearman's rank correlation of numerical features (e.g. length) with protein concentration, accounting for variation in mRNA concentrations; (b) for numerical and categorical features (e.g. function), we compared two extreme populations with Welch's t-test and (c) using a Multivariate Adaptive Regression Splines model, we analyzed the combined contributions of mRNA expression and sequence features to protein abundance variation (Figure 1). To account for the non-linearity of many relationships, we use non-parametric approaches throughout the analysis.

We observed a significant positive correlation between mRNA and protein concentrations, larger than many previous measurements (de Sousa Abreu et al, 2009). We also show that the contribution of translation and protein degradation is at least as important as the contribution of mRNA transcription and stability to the abundance variation of the final protein products. Although variation in mRNA expression explains ∼25–30% of the variation in protein abundance, another 30–40% can be accounted for by characteristics of the sequences, which we identified in a comparative assessment of global correlates. Among these characteristics, sequence length, amino-acid frequencies and also nucleotide frequencies in the coding region are of strong influence (Figure 3A). Characteristics of the 3′UTR and of the 5′UTR, that is length, nucleotide composition and secondary structures, describe another part of the variation, leaving 33% expression variation unexplained. The unexplained fraction may be accounted for by mechanisms not considered in this analysis (e.g. regulation by RNA-binding proteins or gene-specific structural motifs), as well as expression and measurement noise.

Our combined model including mRNA concentration and sequence features can explain 67% of the variation of protein abundance in this system—and thus has the highest predictive power for human protein abundance achieved so far (Figure 3B).

Transcription, mRNA decay, translation and protein degradation are essential processes during eukaryotic gene expression, but their relative global contributions to steady-state protein concentrations in multi-cellular eukaryotes are largely unknown. Using measurements of absolute protein and mRNA abundances in cellular lysate from the human Daoy medulloblastoma cell line, we quantitatively evaluate the impact of mRNA concentration and sequence features implicated in translation and protein degradation on protein expression. Sequence features related to translation and protein degradation have an impact similar to that of mRNA abundance, and their combined contribution explains two-thirds of protein abundance variation. mRNA sequence lengths, amino-acid properties, upstream open reading frames and secondary structures in the 5′ untranslated region (UTR) were the strongest individual correlates of protein concentrations. In a combined model, characteristics of the coding region and the 3′UTR explained a larger proportion of protein abundance variation than characteristics of the 5′UTR. The absolute protein and mRNA concentration measurements for >1000 human genes described here represent one of the largest datasets currently available, and reveal both general trends and specific examples of post-transcriptional regulation.

Transcription, mRNA decay, translation and protein degradation are essential processes during eukaryotic gene expression, but their relative global contributions to steady-state protein concentrations in multi-cellular eukaryotes are largely unknown. Using measurements of absolute protein and mRNA abundances in cellular lysate from the human Daoy medulloblastoma cell line, we quantitatively evaluate the impact of mRNA concentration and sequence features implicated in translation and protein degradation on protein expression. Sequence features related to translation and protein degradation have an impact similar to that of mRNA abundance, and their combined contribution explains two-thirds of protein abundance variation. mRNA sequence lengths, amino-acid properties, upstream open reading frames and secondary structures in the 5' untranslated region (UTR) were the strongest individual correlates of protein concentrations. In a combined model, characteristics of the coding region and the 3'UTR explained a larger proportion of protein abundance variation than characteristics of the 5'UTR. Further, we used data from human and six other organisms (bacteria, yeast, worm, fly, and plant) and established that steady-state abundances of proteins show significantly higher correlation across these diverse phylogenetic taxa than the abundances of their corresponding mRNAs (p=0.0008, paired Wilcoxon). These data suggest strong selective pressure to maintain protein abundances during evolution, even when mRNA abundances diverge. The absolute protein and mRNA concentration measurements for >1000 human genes and for other organisms represent one of the largest datasets currently available, and reveal both general trends and specific examples of post-transcriptional regulation.

Summary

Genome-wide studies in Saccharomyces cerevisiae concluded that the dominant determinant of protein evolutionary rates is expression level, where highly-expressed proteins evolve most slowly. To determine how this constraint affects the evolution of protein interactions, we directly measure evolutionary rates of protein interface, surface and core residues by structurally mapping domain interactions to yeast genomes. We find that mRNA level and protein abundance, though correlated, report on pressures affecting regions of proteins differently. Pressures proportional to mRNA level slow evolutionary rates of all structural regions and reduce the variability in rate differences between interfaces and other surfaces. In contrast, the evolutionary rate variation within a domain is less dependent on protein abundance. Distinct pressures may be associated primarily with the cost (mRNA level) and functional benefit (protein abundance) of protein production. Interfaces of proteins with low mRNA levels may have higher evolutionary flexibility, and could constitute the raw material for new functions.

Biased usage of amino acids near exon-intron boundaries is phylogenetically widespread and characteristic of species for which there are expected to be problems defining exons.

Background

In mammals, splice-regulatory domains impose marked trends on the relative abundance of certain amino acids near exon-intron boundaries. Is this a mammalian particularity or symptomatic of exonic splicing regulation across taxa? Are such trends more common in species that a priori have a harder time identifying exon ends, that is, those with pre-mRNA rich in intronic sequence? We address these questions surveying exon composition in a sample of phylogenetically diverse genomes.

Results

Biased amino acid usage near exon-intron boundaries is common throughout the metazoa but not restricted to the metazoa. There is extensive cross-species concordance as to which amino acids are affected, and reduced/elevated abundances are well predicted by knowledge of splice enhancers. Species expected to rely on exon definition for splicing, that is, those with a higher ratio of intronic to coding sequence, more introns per gene and longer introns, exhibit more amino acid skews. Notably, this includes the intron-rich basidiomycete Cryptococcus neoformans, which, unlike intron-poor ascomycetes (Schizosaccharomyces pombe, Saccharomyces cerevisiae), exhibits compositional biases reminiscent of the metazoa. Strikingly, 5 prime ends of nematode exons deviate radically from normality: amino acids strongly preferred near boundaries are strongly avoided in other species, and vice versa. This we suggest is a measure to avoid attracting trans-splicing machinery.

Conclusion

Constraints on amino acid composition near exon-intron boundaries are phylogenetically widespread and characteristic of species where exon localization should be problematic. That compositional biases accord with sequence preferences of splice-regulatory proteins and are absent in ascomycetes is consistent with selection on exonic splicing regulation.

The assessment of microbial diversity and distribution is a major concern in environmental microbiology. There are two general approaches for measuring community diversity: quantitative measures, which use the abundance of each taxon, and qualitative measures, which use only the presence/absence of data. Quantitative measures are ideally suited to revealing community differences that are due to changes in relative taxon abundance (e.g., when a particular set of taxa flourish because a limiting nutrient source becomes abundant). Qualitative measures are most informative when communities differ primarily by what can live in them (e.g., at high temperatures), in part because abundance information can obscure significant patterns of variation in which taxa are present. We illustrate these principles using two 16S rRNA-based surveys of microbial populations and two phylogenetic measures of community β diversity: unweighted UniFrac, a qualitative measure, and weighted UniFrac, a new quantitative measure, which we have added to the UniFrac website (http://bmf.colorado.edu/unifrac). These studies considered the relative influences of mineral chemistry, temperature, and geography on microbial community composition in acidic thermal springs in Yellowstone National Park and the influences of obesity and kinship on microbial community composition in the mouse gut. We show that applying qualitative and quantitative measures to the same data set can lead to dramatically different conclusions about the main factors that structure microbial diversity and can provide insight into the nature of community differences. We also demonstrate that both weighted and unweighted UniFrac measurements are robust to the methods used to build the underlying phylogeny.

Although several studies have provided important insights into the general principles of biological networks, the link between network organization and the genome-scale dynamics of the underlying entities (genes, mRNAs, and proteins) and its role in systems behavior remain unclear. Here we show that transcription factor (TF) dynamics and regulatory network organization are tightly linked. By classifying TFs in the yeast regulatory network into three hierarchical layers (top, core, and bottom) and integrating diverse genome-scale datasets, we find that the TFs have static and dynamic properties that are similar within a layer and different across layers. At the protein level, the top-layer TFs are relatively abundant, long-lived, and noisy compared with the core- and bottom-layer TFs. Although variability in expression of top-layer TFs might confer a selective advantage, as this permits at least some members in a clonal cell population to initiate a response to changing conditions, tight regulation of the core- and bottom-layer TFs may minimize noise propagation and ensure fidelity in regulation. We propose that the interplay between network organization and TF dynamics could permit differential utilization of the same underlying network by distinct members of a clonal cell population.

Ostertagia ostertagi is a gastrointestinal parasitic nematode that affects cattle and leads to a loss of production. In this study, we present the first large-scale genomic survey of Ostertagia ostertagi by the analysis of expressed transcripts from three stages of the parasite: third-stage larvae, fourth-stage larvae and adult worms. Using an in silico approach, 2,284 genes were identified from over 7,000 expressed sequence tags and abundant transcripts were analyzed and characterized by their functional profile. Of the 2,284 genes, 66% had similarity to other known or predicted genes while the rest were novel and potentially represent genes specific to the species and/or stages. Furthermore, a subset of the novel proteins were structurally annotated and assigned putative function based on orthologs in C. elegans and corresponding RNA interference phenotypes. Hence, over 70% of the genes were annotated using protein sequences, domains and pathway databases. Differentially expressed transcripts from the two larval stages and their functional profiles were also studied leading to a more detailed understanding of the parasite’s life-cycle. The identified transcripts are a valuable resource for genomic studies of O. ostertagi and can facilitate the design of control strategies and vaccine programs.

The full-length 6.14-kb polycistronic glf-zwf-edd-glk mRNA from Zymomonas mobilis appears to be processed by endonucleolytic cleavage, resulting in the formation of several discrete transcripts. Northern analysis and transcript mapping revealed that the processed transcripts correspond to functional mono-, di-, or tricistronic messages. The relative abundance of the gene-specific, functional messages was measured. Expression of zwf and edd correlated well with functional message levels. Disproportionally high levels of the glk-specific mRNAs might compensate for the instability of glucokinase by allowing increased translation. The relative abundance of the discrete transcripts was shown to be a function of their respective decay rates. Northern analysis of the fate of the 6.14-kb transcript after inhibition of transcription by rifampin showed that the abundance of shorter, more stable transcripts increased at the expense of longer, less stable transcripts. This is suggestive of endonucleolytic mRNA processing. The most abundant 5' and 3' transcript ends were found to lie within secondary structures that probably impart stability to the most abundant mRNAs.

Images