A Scalable Approach for Protein False Discovery Rate Estimation in Large Proteomic Data Sets

Calculating the number of confidently identified proteins and estimating false discovery rate (FDR) is a challenge when analyzing very large proteomic data sets such as entire human proteomes. Biological and technical heterogeneity in proteomic experiments further add to the challenge and there are strong differences in opinion regarding the conceptual validity of a protein FDR and no consensus regarding the methodology for protein FDR determination. There are also limitations inherent to the widely used classic target–decoy strategy that particularly show when analyzing very large data sets and that lead to a strong over-representation of decoy identifications. In this study, we investigated the merits of the classic, as well as a novel target–decoy-based protein FDR estimation approach, taking advantage of a heterogeneous data collection comprised of ∼19,000 LC-MS/MS runs deposited in ProteomicsDB (https://www.proteomicsdb.org). The “picked” protein FDR approach treats target and decoy sequences of the same protein as a pair rather than as individual entities and chooses either the target or the decoy sequence depending on which receives the highest score. We investigated the performance of this approach in combination with q-value based peptide scoring to normalize sample-, instrument-, and search engine-specific differences. The “picked” target–decoy strategy performed best when protein scoring was based on the best peptide q-value for each protein yielding a stable number of true positive protein identifications over a wide range of q-value thresholds. We show that this simple and unbiased strategy eliminates a conceptual issue in the commonly used “classic” protein FDR approach that causes overprediction of false-positive protein identification in large data sets. The approach scales from small to very large data sets without losing performance, consistently increases the number of true-positive protein identifications and is readily implemented in proteomics analysis software.Shotgun proteomics is the most popular approach for large-scale identification and quantification of proteins. The rapid evolution of high-end mass spectrometers in recent years (1–5) has made proteomic studies feasible that identify and quantify as many as 10,000 proteins in a sample (6–8) and enables many lines of new scientific research including, for example, the analysis of many human proteomes, and proteome-wide protein–drug interaction studies (9–11). One fundamental step in most proteomic experiments is the identification of proteins in the biological system under investigation. To achieve this, proteins are digested into peptides, analyzed by LC-MS/MS, and tandem mass spectra are used to interrogate protein sequence databases using search engines that match experimental data to data generated in silico (12, 13). Peptide spectrum matches (PSMs)¹ are commonly assigned by a search engine using either a heuristic or a probabilistic scoring scheme (14–18). Proteins are then inferred from identified peptides and a protein score or a probability derived as a measure for the confidence in the identification (13, 19).Estimating the proportion of false matches (false discovery rate; FDR) in an experiment is important to assess and maintain the quality of protein identifications. Owing to its conceptual and practical simplicity, the most widely used strategy to estimate FDR in proteomics is the target–decoy database search strategy (target–decoy strategy; TDS) (20). The main assumption underlying this idea is that random matches (false positives) should occur with similar likelihood in the target database and the decoy (reversed, shuffled, or otherwise randomized) version of the same database (21, 22). The number of matches to the decoy database, therefore, provides an estimate of the number of random matches one should expect to obtain in the target database. The number of target and decoy hits can then be used to calculate either a local or a global FDR for a given data set (21–26). This general idea can be applied to control the FDR at the level of PSMs, peptides, and proteins, typically by counting the number of target and decoy observations above a specified score.Despite the significant practical impact of the TDS, it has been observed that a peptide FDR that results in an acceptable protein FDR (of say 1%) for a small or medium sized data set, turns into an unacceptably high protein FDR when the data set grows larger (22, 27). This is because the basic assumption of the classical TDS is compromised when a large proportion of the true positive proteins have already been identified. In small data sets, containing say only a few hundred to a few thousand proteins, random peptide matches will be distributed roughly equally over all decoy and “leftover” target proteins, allowing for a reasonably accurate estimation of false positive target identifications by using the number of decoy identifications. However, in large experiments comprising hundreds to thousands of LC-MS/MS runs, 10,000 or more target proteins may be genuinely and repeatedly identified, leaving an ever smaller number of (target) proteins to be hit by new false positive peptide matches. In contrast, decoy proteins are only hit by the occasional random peptide match but fully count toward the number of false positive protein identifications estimated from the decoy hits. The higher the number of genuinely identified target proteins gets, the larger this imbalance becomes. If this is not corrected for in the decoy space, an overestimation of false positives will occur.This problem has been recognized and e.g. Reiter and colleagues suggested a way for correcting for the overestimation of false positive protein hits termed MAYU (27). Following the main assumption that protein identifications containing false positive PSMs are uniformly distributed over the target database, MAYU models the number of false positive protein identifications using a hypergeometric distribution. Its parameters are estimated from the number of protein database entries and the total number of target and decoy protein identifications. The protein FDR is then estimated by dividing the number of expected false positive identifications (expectation value of the hypergeometric distribution) by the total number of target identifications. Although this approach was specifically designed for large data sets (tested on ∼1300 LC-MS/MS runs from digests of C. elegans proteins), it is not clear how far the approach actually scales. Another correction strategy for overestimation of false positive rates, the R factor, was suggested initially for peptides (28) and more recently for proteins (29). A ratio, R, of forward and decoy hits in the low probability range is calculated, where the number of true peptide or protein identifications is expected to be close to zero, and hence, R should approximate one. The number of decoy hits is then multiplied (corrected) by the R factor when performing FDR calculations. The approach is conceptually simpler than the MAYU strategy and easy to implement, but is also based on the assumption that the inflation of the decoy hits intrinsic in the classic target–decoy strategy occurs to the same extent in all probability ranges.In the context of the above, it is interesting to note that there is currently no consensus in the community regarding if and how protein FDRs should be calculated for data of any size. One perhaps extreme view is that, owing to issues and assumptions related to the peptide to protein inference step and ways of constructing decoy protein sequences, protein level FDRs cannot be meaningfully estimated at all (30). This is somewhat unsatisfactory as an estimate of protein level error in proteomic experiments is highly desirable. Others have argued that target–decoy searches are not even needed when accurate p values of individual PSMs are available (31) whereas others choose to tighten the PSM or peptide FDRs obtained from TDS analysis to whatever threshold necessary to obtain a desired protein FDR (32). This is likely too conservative.We have recently proposed an alternative protein FDR approach termed “picked” target–decoy strategy (picked TDS) that indicated improved performance over the classical TDS in a very large proteomic data set (9) but a systematic investigation of the idea had not been performed at the time. In this study, we further characterized the picked TDS for protein FDR estimation and investigated its scalability compared with that of the classic TDS FDR method in data sets of increasing size up to ∼19,000 LC-MS/MS runs. The results show that the picked TDS is effective in preventing decoy protein over-representation, identifies more true positive hits, and works equally well for small and large proteomic data sets.     

Sip1, a Novel RS Domain-Containing Protein Essential for Pre-mRNA Splicing

Previous studies have shown that protein-protein interactions among splicing factors may play an important role in pre-mRNA splicing. We report here identification and functional characterization of a new splicing factor, Sip1 (SC35-interacting protein 1). Sip1 was initially identified by virtue of its interaction with SC35, a splicing factor of the SR family. Sip1 interacts with not only several SR proteins but also with U1-70K and U2AF65, proteins associated with 5′ and 3′ splice sites, respectively. The predicted Sip1 sequence contains an arginine-serine-rich (RS) domain but does not have any known RNA-binding motifs, indicating that it is not a member of the SR family. Sip1 also contains a region with weak sequence similarity to the Drosophila splicing regulator suppressor of white apricot (SWAP). An essential role for Sip1 in pre-mRNA splicing was suggested by the observation that anti-Sip1 antibodies depleted splicing activity from HeLa nuclear extract. Purified recombinant Sip1 protein, but not other RS domain-containing proteins such as SC35, ASF/SF2, and U2AF65, restored the splicing activity of the Sip1-immunodepleted extract. Addition of U2AF65 protein further enhanced the splicing reconstitution by the Sip1 protein. Deficiency in the formation of both A and B splicing complexes in the Sip1-depleted nuclear extract indicates an important role of Sip1 in spliceosome assembly. Together, these results demonstrate that Sip1 is a novel RS domain-containing protein required for pre-mRNA splicing and that the functional role of Sip1 in splicing is distinct from those of known RS domain-containing splicing factors.Pre-mRNA splicing takes place in spliceosomes, the large RNA-protein complexes containing pre-mRNA, U1, U2, U4/6, and U5 small nuclear ribonucleoprotein particles (snRNPs), and a large number of accessory protein factors (for reviews, see references 21, 22, 37, 44, and 48). It is increasingly clear that the protein factors are important for pre-mRNA splicing and that studies of these factors are essential for further understanding of molecular mechanisms of pre-mRNA splicing.Most mammalian splicing factors have been identified by biochemical fractionation and purification (3, 15, 19, 31–36, 45, 69–71, 73), by using antibodies recognizing splicing factors (8, 9, 16, 17, 61, 66, 67, 74), and by sequence homology (25, 52, 74).Splicing factors containing arginine-serine-rich (RS) domains have emerged as important players in pre-mRNA splicing. These include members of the SR family, both subunits of U2 auxiliary factor (U2AF), and the U1 snRNP protein U1-70K (for reviews, see references 18, 41, and 59). Drosophila alternative splicing regulators transformer (Tra), transformer 2 (Tra2), and suppressor of white apricot (SWAP) also contain RS domains (20, 40, 42). RS domains in these proteins play important roles in pre-mRNA splicing (7, 71, 75), in nuclear localization of these splicing proteins (23, 40), and in protein-RNA interactions (56, 60, 64). Previous studies by us and others have demonstrated that one mechanism whereby SR proteins function in splicing is to mediate specific protein-protein interactions among spliceosomal components and between general splicing factors and alternative splicing regulators (1, 1a, 6, 10, 27, 63, 74, 77). Such protein-protein interactions may play critical roles in splice site recognition and association (for reviews, see references 4, 18, 37, 41, 47 and 59). Specific interactions among the splicing factors also suggest that it is possible to identify new splicing factors by their interactions with known splicing factors.Here we report identification of a new splicing factor, Sip1, by its interaction with the essential splicing factor SC35. The predicted Sip1 protein sequence contains an RS domain and a region with sequence similarity to the Drosophila splicing regulator, SWAP. We have expressed and purified recombinant Sip1 protein and raised polyclonal antibodies against the recombinant Sip1 protein. The anti-Sip1 antibodies specifically recognize a protein migrating at a molecular mass of approximately 210 kDa in HeLa nuclear extract. The anti-Sip1 antibodies sufficiently deplete Sip1 protein from the nuclear extract, and the Sip1-depleted extract is inactive in pre-mRNA splicing. Addition of recombinant Sip1 protein can partially restore splicing activity to the Sip1-depleted nuclear extract, indicating an essential role of Sip1 in pre-mRNA splicing. Other RS domain-containing proteins, including SC35, ASF/SF2, and U2AF65, cannot substitute for Sip1 in reconstituting splicing activity of the Sip1-depleted nuclear extract. However, addition of U2AF65 further increases splicing activity of Sip1-reconstituted nuclear extract, suggesting that there may be a functional interaction between Sip1 and U2AF65 in nuclear extract.     

Analysis of Proteins That Rapidly Change Upon Mechanistic/Mammalian Target of Rapamycin Complex 1 (mTORC1) Repression Identifies Parkinson Protein 7 (PARK7) as a Novel Protein Aberrantly Expressed in Tuberous Sclerosis Complex (TSC)

Many biological processes involve the mechanistic/mammalian target of rapamycin complex 1 (mTORC1). Thus, the challenge of deciphering mTORC1-mediated functions during normal and pathological states in the central nervous system is challenging. Because mTORC1 is at the core of translation, we have investigated mTORC1 function in global and regional protein expression. Activation of mTORC1 has been generally regarded to promote translation. Few but recent works have shown that suppression of mTORC1 can also promote local protein synthesis. Moreover, excessive mTORC1 activation during diseased states represses basal and activity-induced protein synthesis. To determine the role of mTORC1 activation in protein expression, we have used an unbiased, large-scale proteomic approach. We provide evidence that a brief repression of mTORC1 activity in vivo by rapamycin has little effect globally, yet leads to a significant remodeling of synaptic proteins, in particular those proteins that reside in the postsynaptic density. We have also found that curtailing the activity of mTORC1 bidirectionally alters the expression of proteins associated with epilepsy, Alzheimer''s disease, and autism spectrum disorder—neurological disorders that exhibit elevated mTORC1 activity. Through a protein–protein interaction network analysis, we have identified common proteins shared among these mTORC1-related diseases. One such protein is Parkinson protein 7, which has been implicated in Parkinson''s disease, yet not associated with epilepsy, Alzheimers disease, or autism spectrum disorder. To verify our finding, we provide evidence that the protein expression of Parkinson protein 7, including new protein synthesis, is sensitive to mTORC1 inhibition. Using a mouse model of tuberous sclerosis complex, a disease that displays both epilepsy and autism spectrum disorder phenotypes and has overactive mTORC1 signaling, we show that Parkinson protein 7 protein is elevated in the dendrites and colocalizes with the postsynaptic marker postsynaptic density-95. Our work offers a comprehensive view of mTORC1 and its role in regulating regional protein expression in normal and diseased states.The mechanistic/mammalian target of rapamycin complex 1 (mTORC1)¹ is a serine/threonine protein kinase that is highly expressed in many cell types (1). In the brain, mTORC1 tightly coordinates different synaptic plasticities — long-term potentiation (LTP) and long-term depression (LTD) — the molecular correlates of learning and memory (2–5). Because mTORC1 is at the core of many synaptic signaling pathways downstream of glutamate and neurotrophin receptors, many hypothesize that dysregulated mTORC1 signaling underlies cognitive deficits observed in several neurodegenerative diseases (3, 6–17). For example, mTORC1 and its downstream targets are hyperactive in human brains diagnosed with Alzheimer''s disease (AD) (18–20). Additionally in animal models of autism spectrum disorder (ASD), altered mTORC1 signaling contributes to the observed synaptic dysfunction and aberrant network connectivity (13, 15, 21–27). Furthermore, epilepsy, which is common in AD and ASD, has enhanced mTORC1 activity (28–32).Phosphorylation of mTORC1, considered the active form, is generally regarded to promote protein synthesis (33). Thus, many theorize that diseases with overactive mTORC1 arise from excessive protein synthesis (14). Emerging data, however, show that suppressing mTORC1 activation can trigger local translation in neurons (34, 35). Pharmacological antagonism of N-methyl-d-aspartate (NMDA) receptors, a subtype of glutamate receptors that lies upstream of mTOR activation, promotes the synthesis of the voltage-gated potassium channel, Kv1.1, in dendrites (34, 35). Consistent with these results, in models of temporal lobe epilepsy there is a reduction in the expression of voltage-gated ion channels including Kv1.1 (30, 31, 36). Interestingly in a model of focal neocortical epilepsy, overexpression of Kv1.1 blocked seizure activity (37). Because both active and inactive mTORC1 permit protein synthesis, we sought to determine the proteins whose expression is altered when mTORC1 phosphorylation is reduced in vivo.Rapamycin is an FDA-approved, immunosuppressive drug that inhibits mTORC1 activity (38). We capitalized on the ability of rapamycin to reduce mTORC1 activity in vivo and the unbiased approach of mass spectrometry to identify changes in protein expression. Herein, we provide evidence that mTORC1 activation bidirectionally regulates protein expression, especially in the PSD where roughly an equal distribution of proteins dynamically appear and disappear. Remarkably, using protein–protein interaction networks facilitated the novel discovery that PARK7, a protein thus far only implicated in Parkinson''s disease, (1) is up-regulated by increased mTORC1 activity, (2) resides in the PSD only when mTORC1 is active, and (3) is aberrantly expressed in a rodent model of TSC, an mTORC1-related disease that has symptoms of epilepsy and autism. Collectively, these data provide the first comprehensive list of proteins whose abundance or subcellular distributions are altered with acute changes in mTORC1 activity in vivo.     

Targeting Synaptic Pathology with a Novel Affinity Mass Spectrometry Approach

We report a novel strategy for studying synaptic pathology by concurrently measuring levels of four SNARE complex proteins from individual brain tissue samples. This method combines affinity purification and mass spectrometry and can be applied directly for studies of SNARE complex proteins in multiple species or modified to target other key elements in neuronal function. We use the technique to demonstrate altered levels of presynaptic proteins in Alzheimer disease patients and prion-infected mice.One prominent pathological feature of neuropsychiatric disorders such as Alzheimer disease (AD)¹ is severe synaptic loss (1–3). Previous reports of AD patients have shown that presynaptic dysfunction might occur early in the disease process (1, 4). Cortical synapse pathology has also been shown to correlate to the severity of dementia more closely than other pathological hallmarks of AD such as plaques and neurofibrillary tangles (5, 6). The SNARE proteins are essential components for the regulation of neurotransmitter exocytosis at the presynaptic site (7). Animal models suggest that changed expression or modification of SNARE complex proteins (synaptosomal-associated protein 25 (SNAP-25), syntaxin-1, and vesicle-associated membrane protein (VAMP)) alters synaptic function and is an interesting target for the development of therapeutics for neuropsychiatric illness (8, 9). The constituents of the SNARE complex are either localized in synaptic vesicles (VAMPs) or anchored at the presynaptic plasma membrane (SNAP-25 and syntaxin). The SNARE proteins are tightly assembled, and subsequent neurotransmitter release of the complex is quickly dissociated by N-ethylmaleimide-sensitive factor (7, 10–12). Because they are both strongly associated into complexes and membrane associated, the SNARE proteins are difficult to analyze via mass spectrometry, which is incompatible with most detergents necessary for the solubilization of proteins. Each SNARE complex protein exists in several isoforms that are differently distributed within the central nervous system (13–18). Post-translational modifications and truncated variants of the SNARE proteins make investigation of the protein expression even more complicated.In this study we developed an approach for the characterization and concurrent quantification of SNARE complex proteins that combines affinity purification by immunoprecipitation and mass spectrometry (IP-MS). We used precipitation with monoclonal antibodies against SNAP-25 to target the SNARE complex proteins and nanoflow LC–tandem mass spectrometry (LC-MS/MS) to characterize the co-immunoprecipitated interaction partners. Selected reaction monitoring (SRM) on a triple quadrupole mass spectrometer coupled to a microflow LC system was used for quantification of the SNARE proteins. To demonstrate the usability of the IP-MS method, we performed a comparison of SNARE complex protein levels in brain tissue from AD patients and age-matched controls, as well as a study of SNARE complex protein levels in brain tissue from prion-infected mice.     

A Pipeline for Determining Protein–Protein Interactions and Proximities in the Cellular Milieu

It remains extraordinarily challenging to elucidate endogenous protein-protein interactions and proximities within the cellular milieu. The dynamic nature and the large range of affinities of these interactions augment the difficulty of this undertaking. Among the most useful tools for extracting such information are those based on affinity capture of target bait proteins in combination with mass spectrometric readout of the co-isolated species. Although highly enabling, the utility of affinity-based methods is generally limited by difficulties in distinguishing specific from nonspecific interactors, preserving and isolating all unique interactions including those that are weak, transient, or rapidly exchanging, and differentiating proximal interactions from those that are more distal. Here, we have devised and optimized a set of methods to address these challenges. The resulting pipeline involves flash-freezing cells in liquid nitrogen to preserve the cellular environment at the moment of freezing; cryomilling to fracture the frozen cells into intact micron chunks to allow for rapid access of a chemical reagent and to stabilize the intact endogenous subcellular assemblies and interactors upon thawing; and utilizing the high reactivity of glutaraldehyde to achieve sufficiently rapid stabilization at low temperatures to preserve native cellular interactions. In the course of this work, we determined that relatively low molar ratios of glutaraldehyde to reactive amines within the cellular milieu were sufficient to preserve even labile and transient interactions. This mild treatment enables efficient and rapid affinity capture of the protein assemblies of interest under nondenaturing conditions, followed by bottom-up MS to identify and quantify the protein constituents. For convenience, we have termed this approach Stabilized Affinity Capture Mass Spectrometry. Here, we demonstrate that Stabilized Affinity Capture Mass Spectrometry allows us to stabilize and elucidate local, distant, and transient protein interactions within complex cellular milieux, many of which are not observed in the absence of chemical stabilization.Insights into many cellular processes require detailed information about interactions between the participating proteins. However, the analysis of such interactions can be challenging because of the often-diverse physicochemical properties and the abundances of the constituent proteins, as well as the sometimes wide range of affinities and complex dynamics of the interactions. One of the key challenges has been acquiring information concerning transient, low affinity interactions in highly complex cellular milieux (3, 4).Methods that allow elucidation of such information include co-localization microscopy (5), fluorescence protein Förster resonance energy transfer (4), immunoelectron microscopy (5), yeast two-hybrid (6), and affinity capture (7, 8). Among these, affinity capture (AC)¹ has the unique potential to detect all specific in vivo interactions simultaneously, including those that interact both directly and indirectly. In recent times, the efficacy of such affinity isolation experiments has been greatly enhanced through the use of sensitive modern mass spectrometric protein identification techniques (9). Nevertheless, AC suffers from several shortcomings. These include the problem of 1) distinguishing specific from nonspecific interactors (10, 11); 2) preserving and isolating all unique interactions including those that are weak and/or transient, as well as those that exchange rapidly (10, 12, 13); and 3) differentiating proximal from more distant interactions (14).We describe here an approach to address these issues, which makes use of chemical stabilization of protein assemblies in the complex cellular milieu prior to AC. Chemical stabilization is an emerging technique for stabilizing and elucidating protein associations both in vitro (15–20) and in vivo (3, 12, 14, 21–29), with mass spectrometric (MS) readout of the AC proteins and their connectivities. Such chemical stabilization methods are indeed well-established and are often used in electron microscopy for preserving complexes and subcellular structures both in the cellular milieu (3) and in purified complexes (30, 31), wherein the most reliable, stable, and established stabilization reagents is glutaraldehyde. Recently, glutaraldehyde has been applied in the “GraFix” protocol in which purified protein complexes are subjected to centrifugation through a density gradient that also contains a gradient of glutaraldehyde (30, 31), allowing for optimal stabilization of authentic complexes and minimization of nonspecific associations and aggregation. GraFix has also been combined with mass spectrometry on purified complexes bound to EM grids to obtain a compositional analysis of the complexes (32), thereby raising the possibility that glutaraldehyde can be successfully utilized in conjunction with AC in complex cellular milieux directly.In this work, we present a robust pipeline for determining specific protein-protein interactions and proximities from cellular milieux. The first steps of the pipeline involve the well-established techniques of flash freezing the cells of interest in liquid nitrogen and cryomilling, which have been known for over a decade (33, 34) to preserve the cellular environment, as well as having shown outstanding performance when used in analysis of macromolecular interactions in yeast (35–39), bacterial (40, 41), trypanosome (42), mouse (43), and human (44–47) systems. The resulting frozen powder, composed of intact micron chunks of cells that have great surface area and outstanding solvent accessibility, is well suited for rapid low temperature chemical stabilization using glutaraldehyde. We selected glutaraldehyde for our procedure based on the fact that it is a very reactive stabilizing reagent, even at lower temperatures, and because it has already been shown to stabilize enzymes in their functional state (48–50). We employed highly efficient, rapid, single stage affinity capture (36, 51) for isolation and bottom-up MS for analysis of the macromolecular assemblies of interest (52–54). For convenience, we have termed this approach Stabilized Affinity-Capture Mass Spectrometry (SAC-MS).     

Identification of Tumor-associated Autoantigens for the Diagnosis of Colorectal Cancer in Serum Using High Density Protein Microarrays

There is a mounting evidence of the existence of autoantibodies associated to cancer progression. Antibodies are the target of choice for serum screening because of their stability and suitability for sensitive immunoassays. By using commercial protein microarrays containing 8000 human proteins, we examined 20 sera from colorectal cancer (CRC) patients and healthy subjects to identify autoantibody patterns and associated antigens. Forty-three proteins were differentially recognized by tumoral and reference sera (p value <0.04) in the protein microarrays. Five immunoreactive antigens, PIM1, MAPKAPK3, STK4, SRC, and FGFR4, showed the highest prevalence in cancer samples, whereas ACVR2B was more abundant in normal sera. Three of them, PIM1, MAPKAPK3, and ACVR2B, were used for further validation. A significant increase in the expression level of these antigens on CRC cell lines and colonic mucosa was confirmed by immunoblotting and immunohistochemistry on tissue microarrays. A diagnostic ELISA based on the combination of MAPKAPK3 and ACVR2B proteins yielded specificity and sensitivity values of 73.9 and 83.3% (area under the curve, 0.85), respectively, for CRC discrimination after using an independent sample set containing 94 sera representative of different stages of progression and control subjects. In summary, these studies confirmed the presence of specific autoantibodies for CRC and revealed new individual markers of disease (PIM1, MAPKAPK3, and ACVR2B) with the potential to diagnose CRC with higher specificity and sensitivity than previously reported serum biomarkers.Colorectal cancer (CRC)¹ is the second most prevalent cancer in the western world. The development of this disease takes decades and involves multiple genetic events. CRC remains a major cause of mortality in developed countries because most of the patients are diagnosed at advanced stages because of the reluctance to use highly invasive diagnostic tools like colonoscopy. Actually only a few proteins have been described as biomarkers in CRC (carcinoembryonic antigen (CEA), CA19.9, and CA125 (1–3)), although none of them is recommended for clinical screening (4). Proteomics analysis is actively used for the identification of new biomarkers. In previous studies, the use of two-dimensional DIGE and antibody microarrays allowed the identification of differentially expressed proteins in CRC tissue, including isoforms and post-translational modifications responsible for modifications in signaling pathways (5–8). Both approaches resulted in the identification of a collection of potential tumoral tissue biomarkers that is currently being investigated.However, the implementation of simpler, non-invasive methods for the early detection of CRC should be based on the identification of proteins or antibodies in serum or plasma (9–13). There is ample evidence of the existence of an immune response to cancer in humans as demonstrated by the presence of autoantibodies in cancer sera. Self-proteins (autoantigens) altered before or during tumor formation can elicit an immune response (13–19). These tumor-specific autoantibodies can be detected at early cancer stages and prior to cancer diagnosis revealing a great potential as biomarkers (14, 15, 20). Tumor proteins can be affected by specific point mutations, misfolding, overexpression, aberrant glycosylation, truncation, or aberrant degradation (e.g. p53, HER2, NY-ESO1, or MUC1 (16, 21–25)). In fact, a number of tumor-associated autoantigens (TAAs) were identified previously in different studies involving autoantibody screening in CRC (26–28).Several approaches have been used to identify TAAs in cancer, including natural protein arrays prepared with fractions obtained from two-dimensional LC separations of tumoral samples (29, 30) or protein extracts from cancer cells or tissue (9, 31) probed by Western blot with patient sera, cancer tissue peptide libraries expressed as cDNA expression libraries for serological screening (serological analysis of recombinant cDNA expression libraries (SEREX)) (22, 32), or peptides expressed on the surface of phages in combination with microarrays (17, 18, 33, 34). However, these approaches suffer from several drawbacks. In some cases TAAs have to be isolated and identified from the reactive protein lysate by LC-MS techniques, or in the phage display approach, the reactive TAA could be a mimotope without a corresponding linear amino acid sequence. Moreover, cDNA libraries might not be representative of the protein expression levels in tumors as there is a poor correspondence between mRNA and protein levels.Protein arrays provide a novel platform for the identification of both autoantibodies and their respective TAAs for diagnostic purposes in cancer serum patients. They present some advantages. Proteins printed on the microarray are known “a priori,” avoiding the need for later identifications and the discovery of mimotopes. There is no bias in protein selection as the proteins are printed at a similar concentration. This should result in a higher sensitivity for biomarker identification (13, 35, 36).In this study, we used commercially available high density protein microarrays for the identification of autoantibody signatures and tumor-associated antigens in colorectal cancer. We screened 20 CRC patient and control sera with protein microarrays containing 8000 human proteins to identify the CRC-associated autoantibody repertoire and the corresponding TAAs. Autoantibody profiles that discriminated the different types of CRC metastasis were identified. Moreover a set of TAAs showing increased or decreased expression in cancer cell lines and paired tumoral tissues was found. Finally an ELISA was set up to test the ability of the most immunoreactive proteins to detect colorectal adenocarcinoma. On the basis of the antibody response, combinations of three antigens, PIM1, MAPKAPK3, and ACVR2B, showed a great potential for diagnosis.     

Sgt1 Dimerization Is Negatively Regulated by Protein Kinase CK2-mediated Phosphorylation at Ser361

The kinetochore, which consists of centromere DNA and structural proteins, is essential for proper chromosome segregation in eukaryotes. In budding yeast, Sgt1 and Hsp90 are required for the binding of Skp1 to Ctf13 (a component of the core kinetochore complex CBF3) and therefore for the assembly of CBF3. We have previously shown that Sgt1 dimerization is important for this kinetochore assembly mechanism. In this study, we report that protein kinase CK2 phosphorylates Ser³⁶¹ on Sgt1, and this phosphorylation inhibits Sgt1 dimerization.The kinetochore is a structural protein complex located in the centromeric region of the chromosome coupled to spindle microtubules (1, 2). The kinetochore generates a signal to arrest cells during mitosis when it is not properly attached to microtubules, thereby preventing chromosome missegregation, which can lead to aneuploidy (3, 4). The molecular structure of the kinetochore complex of the budding yeast Saccharomyces cerevisiae has been well characterized; it is composed of more than 70 proteins, many of which are conserved in mammals (2).The centromere DNA in the budding yeast is a 125-bp region that contains three conserved regions, CDEI, CDEII, and CDEIII (5, 6). CDEIII (25 bp) is essential for centromere function (7) and is bound to a key component of the centromere, the CBF3 complex. The CBF3 complex contains four proteins, Ndc10, Cep3, Ctf13 (8–15), and Skp1 (14, 15), all essential for viability. Mutations in any of the CBF3 proteins abolish the ability of CDEIII to bind to CBF3 (16, 17). All of the kinetochore proteins, except the CDEI-binding Cbf1 (18–20), localize to the kinetochores in a CBF3-dependent manner (2). Thus, CBF3 is a fundamental kinetochore complex, and its mechanism of assembly is of great interest.We have previously found that Sgt1 and Skp1 activate Ctf13; thus, they are required for assembly of the CBF3 complex (21). The molecular chaperone Hsp90 is also required to form the active Ctf13-Skp1 complex (22). Sgt1 has two highly conserved motifs that are required for protein-protein interaction: the tetratricopeptide repeat (21) and the CHORD protein and Sgt1-specific motif. We and others have found that both domains are important for the interaction of Sgt1 with Hsp90 (23–26), which is required for assembly of the core kinetochore complex. This interaction is an initial step in kinetochore activation (24, 26, 27), which is conserved between yeast and humans (28, 29).We have recently shown that Sgt1 dimerization is important for Sgt1-Skp1 binding and therefore for kinetochore assembly (30). In this study, we have found that protein kinase CK2 phosphorylates Sgt1 at Ser³⁶¹, and this phosphorylation inhibits Sgt1 dimerization. Therefore, CK2 appears to regulate kinetochore assembly negatively in budding yeast.     

A New in Vivo Cross-linking Mass Spectrometry Platform to Define Protein–Protein Interactions in Living Cells

Protein–protein interactions (PPIs) are fundamental to the structure and function of protein complexes. Resolving the physical contacts between proteins as they occur in cells is critical to uncovering the molecular details underlying various cellular activities. To advance the study of PPIs in living cells, we have developed a new in vivo cross-linking mass spectrometry platform that couples a novel membrane-permeable, enrichable, and MS-cleavable cross-linker with multistage tandem mass spectrometry. This strategy permits the effective capture, enrichment, and identification of in vivo cross-linked products from mammalian cells and thus enables the determination of protein interaction interfaces. The utility of the developed method has been demonstrated by profiling PPIs in mammalian cells at the proteome scale and the targeted protein complex level. Our work represents a general approach for studying in vivo PPIs and provides a solid foundation for future studies toward the complete mapping of PPI networks in living systems.Protein–protein interactions (PPIs)¹ play a key role in defining protein functions in biological systems. Aberrant PPIs can have drastic effects on biochemical activities essential to cell homeostasis, growth, and proliferation, and thereby lead to various human diseases (1). Consequently, PPI interfaces have been recognized as a new paradigm for drug development. Therefore, mapping PPIs and their interaction interfaces in living cells is critical not only for a comprehensive understanding of protein function and regulation, but also for describing the molecular mechanisms underlying human pathologies and identifying potential targets for better therapeutics.Several strategies exist for identifying and mapping PPIs, including yeast two-hybrid, protein microarray, and affinity purification mass spectrometry (AP-MS) (2–5). Thanks to new developments in sample preparation strategies, mass spectrometry technologies, and bioinformatics tools, AP-MS has become a powerful and preferred method for studying PPIs at the systems level (6–9). Unlike other approaches, AP-MS experiments allow the capture of protein interactions directly from their natural cellular environment, thus better retaining native protein structures and biologically relevant interactions. In addition, a broader scope of PPI networks can be obtained with greater sensitivity, accuracy, versatility, and speed. Despite the success of this very promising technique, AP-MS experiments can lead to the loss of weak/transient interactions and/or the reorganization of protein interactions during biochemical manipulation under native purification conditions. To circumvent these problems, in vivo chemical cross-linking has been successfully employed to stabilize protein interactions in native cells or tissues prior to cell lysis (10–16). The resulting covalent bonds formed between interacting partners allow affinity purification under stringent and fully denaturing conditions, consequently reducing nonspecific background while preserving stable and weak/transient interactions (12–16). Subsequent mass spectrometric analysis can reveal not only the identities of interacting proteins, but also cross-linked amino acid residues. The latter provides direct molecular evidence describing the physical contacts between and within proteins (17). This information can be used for computational modeling to establish structural topologies of proteins and protein complexes (17–22), as well as for generating experimentally derived protein interaction network topology maps (23, 24). Thus, cross-linking mass spectrometry (XL-MS) strategies represent a powerful and emergent technology that possesses unparalleled capabilities for studying PPIs.Despite their great potential, current XL-MS studies that have aimed to identify cross-linked peptides have been mostly limited to in vitro cross-linking experiments, with few successfully identifying protein interaction interfaces in living cells (24, 25). This is largely because XL-MS studies remain challenging due to the inherent difficulty in the effective MS detection and accurate identification of cross-linked peptides, as well as in unambiguous assignment of cross-linked residues. In general, cross-linked products are heterogeneous and low in abundance relative to non-cross-linked products. In addition, their MS fragmentation is too complex to be interpreted using conventional database searching tools (17, 26). It is noted that almost all of the current in vivo PPI studies utilize formaldehyde cross-linking because of its membrane permeability and fast kinetics (10–16). However, in comparison to the most commonly used amine reactive NHS ester cross-linkers, identification of formaldehyde cross-linked peptides is even more challenging because of its promiscuous nonspecific reactivity and extremely short spacer length (27). Therefore, further developments in reagents and methods are urgently needed to enable simple MS detection and effective identification of in vivo cross-linked products, and thus allow the mapping of authentic protein contact sites as established in cells, especially for protein complexes.Various efforts have been made to address the limitations of XL-MS studies, resulting in new developments in bioinformatics tools for improved data interpretation (28–32) and new designs of cross-linking reagents for enhanced MS analysis of cross-linked peptides (24, 33–39). Among these approaches, the development of new cross-linking reagents holds great promise for mapping PPIs on the systems level. One class of cross-linking reagents containing an enrichment handle have been shown to allow selective isolation of cross-linked products from complex mixtures, boosting their detectability by MS (33–35, 40–42). A second class of cross-linkers containing MS-cleavable bonds have proven to be effective in facilitating the unambiguous identification of cross-linked peptides (36–39, 43, 44), as the resulting cross-linked products can be identified based on their characteristic and simplified fragmentation behavior during MS analysis. Therefore, an ideal cross-linking reagent would possess the combined features of both classes of cross-linkers. To advance the study of in vivo PPIs, we have developed a new XL-MS platform based on a novel membrane-permeable, enrichable, and MS-cleavable cross-linker, Azide-A-DSBSO (azide-tagged, acid-cleavable disuccinimidyl bis-sulfoxide), and multistage tandem mass spectrometry (MSⁿ). This new XL-MS strategy has been successfully employed to map in vivo PPIs from mammalian cells at both the proteome scale and the targeted protein complex level.