The list below is an experiment that I am trying. After engaging in a heated discussion on Twitter about preprints and the value of peer review and the quality of the published peer-reviewed literature, I am taking a closer look at papers that pique my interest during my weekly scan of the newly added articles in PubMed. I do this scan as a participant in BioMed News. My topic is proteomics and disease. Through BioMed News, I receive a list of articles added to PubMed each week and identified through machine learning as likely relevant to my topic. Not all of the articles identified by machine learning are relevant to the topic. I select those relevant to the topic without any assessment of their value or validity. My selections represent a list vetted by topic. You can see those papers here: http://biomed.news/reports. My proteomics and disease selections are under the heading “bims-prodis Proteomics in disease.”
Below are papers that I found particularly intriguing from the total list. Many of these are not included in the proteomics and disease list, because they do not fit the selection criteria that I use, which includes using human samples. However, I found the titles intriguing, so I read the abstract and when possible looked at the full text of the paper. After giving them a closer look, I divided the papers into those that I give a thumbs up and those that I give a thumbs down.
Part of my motivation is curiosity. I simply want to learn more about the fields that are covered by these papers or am interested in these studies. Part of my motivation for doing this is to personally assess the quality of the published literature without biasing my selected reading by choosing papers based on specific journals. I want to get a better sense of peer-review quality (either by reading comments posted as part of transparent review or by inferring the quality from the published study). I decided to share my assessment by either classifying the article with a thumbs up or with a thumbs down. In giving papers a thumbs up or thumbs down, physiological relevance is an important criterium for biological studies. I also considered scientific contribution and data quality in my thumbs up or down choice.
Thumbs Up
J. A. Blakely-Ruiz, et al., Metaproteomics reveals persistent and phylum-redundant metabolic functional stability in adult human gut microbiomes of Crohn’s remission patients despite temporal variations in microbial taxa, genomes, and proteomes. Microbiome. 2019 Feb 11;7(1):18. doi: 10.1186/s40168-019-0631-8. PubMed
Summary and comments: The authors found that the microbial composition of the gut was highly individualized in the patients. Within each patient, the microbial composition and the proteomic composition changed over time. By matching the metaproteome to biochemical reactions involved in metabolism, the authors found that the overall metabolic activities of the microbiota were stable despite the time-dependent changes in the microbiota composition in each patient. Many products of microbial metabolism are bioactive to humans, thus analysis of just the genomic data to understand the populations of microbes is not sufficient to predict the production of microbial products of metabolism. Analysis of the metaproteome (the proteins produced by the microbial populations) is also critical.
W. Wasenang, et al., Serum cell-free DNA methylation of OPCML and HOXD9 as a biomarker that may aid in differential diagnosis between cholangiocarcinoma and other biliary diseases. Clin. Epigenetics. 2019 Mar 4;11(1):39. doi: 10.1186/s13148-019-0634-0. PubMed
Summary and comments: The authors tested DNA methylation at 3 genes to determine if, individually or in combination, methylation of those genes had predictive power in identifying patients with different types of disease of the biliary system, which connects the liver to the gall bladder and the gall bladder to the intestine and pancreas. The data were encouraging that combined analysis of DNA methylation of two of the genes (OPCML and HOXD9) in DNA isolated from the patient’s serum from a blood sample could be helpful in making a differential diagnosis of cholangiocarcinoma (a potentially fatal cancer) or other types of biliary disease.
D. Holub, et al., Mass spectrometry amyloid typing is reproducible across multiple organ sites. Biomed. Res. Int. 2019 Jan 31;2019:3689091. doi: 10.1155/2019/3689091. PubMed
Summary and comments: This article is about using mass spectrometry-based methods on formalin-fixed paraffin-embedded tissue samples from autopsies of patients with various types of amyloidosis. Amyloidosis is a disorder arising from the inappropriate folding of proteins and the formation of tissue-damaging fibrils called amyloids. Treatment relies on knowing which protein is the cause of the amyloids. Although tissue-staining (histochemistry) techniques can identify the presence of amyloids, they cannot always identify the protein responsible. The authors compared the result of identification of the responsible protein by mass spectrometry to identification by immunohistochemistry (tissue staining with antibodies to specific proteins). Their analysis indicated that mass spectrometry was vastly superior, identifying the amyloid-responsible protein in all 22 samples and immunohistochemistry identifying the protein in only 8 of the samples.
F. Chen, et al. Neoantigen identification strategies enable personalized immunotherapy in refractory solid tumors. J. Clin. Invest. 2019 Mar 5;130. pii: 99538. doi: 10.1172/JCI99538. PubMed
Summary and comments: This study reports the results of a clinical trial that tested two different methodological pipelines for identifying tumor neoantigens and identifying neoantigen-reactive T cells in the patients. One of the pipelines involved targeted sequencing the patient’s tumor DNA, epitope prediction, and then peptide synthesis. The other pipeline involved generation of a neoepitope library from hotspot mutations in selected tumor-associated proteins from cancer databases. The patient’s own blood cells were used to test for reactivity to the neopeptides or to screen the library. Each pipeline was successful in identifying neoepitopes for a subset of the patients tested in each approach. Six of the 9 patients with T cells that recognized the neoantigens were treated with personalized neoantigen-loaded dendritic cells and neoantigen-reactive T cells. One patient with treatment-refractory cancer that had metastasized to the lung achieved complete remission. A patient with metastatic pancreatic cancer exhibited regression of many lesions. The four other patients showed extended stabilization of their cancers. The version I read is a “preview” version of the accepted manuscript before copyediting and compilation.
L. Brueggeman, et al., Drug repositioning in epilepsy reveals novel antiseizure candidates. Ann. Clin. Transl. Neurol. 2018 Dec 11;6(2):295-309. doi: 10.1002/acn3.703 PubMed
Summary and comments: These authors started with biopsy tissue from the brains of epilepsy patients who underwent and compared the transcripts expressed in tissue that exhibited neuronal spiking and tissue that did not. Although only 3 patients had matched spiking and nonspiking tissue samples, there were 6 patients in the study. The researchers identified 689 transcripts from 628 genes that were expressing at different amounts in the spiking (epileptic) and nonspiking brain tissue. They called this set of transcripts the “epilepsy transcriptional signature.” A panel of drugs had been previously tested on cell lines to determine how they affected the gene expression. They evaluated how the drug-induced transcriptional signatures in the cell lines correlated with the epilepsy transcriptional signature. Fro this comparison, they predicted that 184 compounds would have a therapeutic effect on epilepsy and 208 compounds would increase seizures. Using a fish model of drug-induced epilepsy, the authors tested 4 compounds predicted to have anti-epileptic activity. Three of the 4 compounds reduced seizures in the fish, although one stimulated movement in control fish that had not been exposed to the seizure-inducing drug. This study shows how patient data can be used to predict potential new applications of existing medications, which can be tested in preclinical studies using model organisms.
L. Loyez, et al.,In situ cancer diagnosis through online plasmonics. Biosens. Bioelectron. 2019 Feb 18;131:104-112. doi: 10.1016/j.bios.2019.01.062 PubMed
Summary and comments: Part of my reason for selecting this paper was to learn what plasmonics is and how it could be applied to medicine. Plasmonics is an electromagnetic phenomenon involving metals. The plasmonic signal can be detected. The authors designed and tested a biosensor consisting of a grated structure imprinted on an optical fiber (like those used for fiber optic delivery in telecommunications). This fiber was coated with a thin layer of metal (gold) and onto the metal was applied a layer of antibodies that would react with specific proteins on lung cancer cells. The concept is that this biosensor can be inserted into the lungs using a catheter and a region suspected of cancer can be assessed without having to perform a biopsy. The antibodies used on the biosensor detect cytokeratin-17, which is abundant on lung cancer cells. Testing the biosensor on human lung tissue, from patients who had surgical removal of a lobe of the lung because of lung cancer, showed that the biosensor properly differentiated healthy tissue from cancerous tissue. Although the biosensor was only tested in human lungs removed from patients, testing in pigs demonstrated that the biosensor could be successfully maneuvered into different regions of the lung and produce a signal. This type of a device could be modified for other cancers by incorporating other proteins specific to other cancers and could reduce the need for invasive surgical procedures and enable earlier detection of solid cancers in tissues.
R. Wakasaki, et al., Glomerular filtrate proteins in acute cardiorenal syndrome. JCI Insight. 2019 Feb 21;4(4). pii: 122130. doi: 10.1172/jci.insight.122130. PubMed
Summary and comments: Patients who have a heart attack and receive cardiopulmonary resuscitation (CPR) are at risk of developing a kidney condition called acute cardiorenal syndrome. Identifying signals from the injured heart that may cause this kidney complication have been difficult. The authors used 2-photon imaging in living control mice or mice that had been subjected to experimentally induced heart attack and then CPR. With this image-guided approach, they collected fluid from a kidney structure called Bowman’s urinary space. The proteins in this fluid were compared between the control mice and the mice subjected to heart attack and CPR using mass spectrometry analysis. This allowed the researchers to determine that proteins from the heart were more common in the fluid from the mice subjected to heart attack and CPR than in the fluid from the control mice. Using genetically altered mice, the heart protein LIM (also known as CSRP3) was identified as specifically increased in the urine of mice subjected to heart attack and CPR. LIM was also increased in the circulation of mice after cardiac arrest and CPR, but not in mice with a kidney injury without a heart injury. Injecting LIM into mice resulted in impaired kidney function 28 days later. Importantly, LIM was increased in the circulation of 5 patients who were critically ill after having heart attacks and CPR.
A. Banerjee, et al., Proteomic analysis reveals that wildtype and alanine-expanded nuclear poly(A)-binding protein exhibit differential interactions in skeletal muscle. J. Biol. Chem. 2019 Mar 5. pii: jbc.RA118.007287. doi: 10.1074/jbc.RA118.007287 PubMed
Summary and comments: Proteins produced from genes abnormally repeated codons have stretches of the protein with a single amino acid repeated multiple times. These are referred to as expanded proteins and cause diseases, such as a type of muscular dystrophy called oculopharyngeal musclar dystrophy (OPMD). The authors introduced a tagged form of the protein PABPN1, which is expanded in OPMD, into skeletal muscles of mice. They introduced some mice with the the normal (wild-type) form and some with the expanded form. After isolating the proteins using the tags, the authors compared the proteins that interacted with wild-type PABPN1 and expanded PABPN1 and selected one (TDP-43) that interacted specifically with the expanded form to investigate in detail. Their analysis indicated that the interaction between the expanded form of PABPN1 and TDP-43 impaired the function of TDP-43, which changed the processing of the mRNA encoding some proteins. Testing the function of one of these altered proteins indicated that the altered mRNA processing changed the function of the protein. Thus, the toxicity associated with PABPN1 may relate to the interaction of PABPN1 with TDP-43 and impairment of mRNA processing.
S. Mondal and P. R. Thompson, Protein arginine deiminases (PADs): Biochemistry and chemical biology of protein citrullination. Acc. Chem. Res. 2019 Mar 7. doi: 10.1021/acs.accounts.9b00024. PubMed
Summary and comments: Without full text access, I can only see the graphical abstract and written abstract. The authors claim to have generated reversible and irreversible inhibitors of these enzymes and then developed selective inhibitors of specific members of this family of enzymes. They tested inhibitors that blocked the activity of all PADs (pan-PAD inhibitors) in animal models of rheumatoid arthritis, lupus, and ulcerative colitis. Citrullination is a post-translational chemical modification of proteins. Compared with phosphorylation or ubiquitylation, this post-translational modification is less well characterized.
R. Zaman, et al., Current strategies in extending half-lives of therapeutic proteins.J. Control Release. 2019 Mar 5. pii: S0168-3659(19)30091-4. doi: 10.1016/j.jconrel.2019.02.016. PubMed
Summary and comments: I could only read the abstract, see the graphical abstract, and access the detailed outline of this paper. From that information, this article is a review about the different strategies to improve the pharmacokinetic properties of proteins and peptides that are used as medication. The review provides detailed coverage of strategies based on the attachment of other biomolecules (primarily fusion proteins), the attachment of polymers, and combination approaches that rely on both polymers and biomolecules. Strategies based on the introduction of point mutations to render proteins protease resistant and strategies based on nanoparticle-mediated delivery are described in shorter sections.
Thumbs Down
H. Komura, et al., Alzheimer Aβ assemblies accumulate in excitatory neurons upon proteasome inhibition and kill nearby NAKα3 neurons by secretion. iScience. 2019 Jan 29. pii: S2589-0042(19)30018-5. doi: 10.1016/j.isci.2019.01.018. PubMed
Summary and comments: The authors used engineered viruses to infect mixed cultures of neurons from rat hippocampus with mutant forms of a protein that is associated with Alzheimer’s disease. Most of the neurons were infected with the engineered virus and produced the mutant protein. In the presence of an inhibitor of the proteasome, the protein formed large aggregates (amylospheroids) that accumulated selectively in excitatory neurons rather than inhibitory neurons or in glial cells in the culture. Inside the proteasome-inhibited neurons, the amylospheroids accumulated in the trans-Golgi network. Amylospheroids were also secreted into the culture medium. Application of amylospheroid-containing culture medium to rat hippocampal neuronal cultures reduced the number of cells that stained positive for the sodium-potassium ATPase subunit alpha3. This study is entirely performed with cultured cells and relies on the proteasome inhibitor to induce the formation of the amylospheriods. The physiological relevance is not clear.
J. L. Hentze, et al., Methylation and ovarian cancer: Can DNA methylation be of diagnostic use? Mol. Clin. Oncol. 2019 Mar;10(3):323-330. doi: 10.3892/mco.2019.1800. PubMed
Summary and comments: DNA methylation is a naturally occurring chemical modifications to DNA that affects the expression of genes. Differences in the methylation of specific genes may be useful in diagnosing disease. The article by Hentze et al. is a review of the current status of DNA methylation in ovarian cancer. The short answer to the title question is “No, not yet at least.” Currently, there are insufficient data indicating that DNA methylation is a useful diagnostic approach for identifying ovarian cancer patients. No DNA methylation profiles have been found that are predictive. I am not convinced this review was needed at this time.
P. Zhao, et al., The imbalance in the complement system and its possible physiological mechanisms in patients with lung cancer.
BMC Cancer. 2019 Mar 6;19(1):201. doi: 10.1186/s12885-019-5422-x. PubMed
Summary and comments: Some complement proteins were higher in the circulation of lung cancer patients than control subjects. Some complement proteins were lower in the lung cancer tissue than in control tissue. Co-culturing a hepatocyte cell line with lung cancer cell lines or normal lung epithelial cells increased production of some complement proteins by the hepatocytes. Importance of these findings is unclear. Hepatocytes and lung epithelial cells are not found in the same organ. Whether lung epithelial cells release a factor that travels in the circulation and affects the liver is not addressed. Additional problems with this paper include a description of a c panel in Figure 3, but this figure only has an a and b panel. The co-culture experiments may have been performed only once. The open review process shows reviewers had concerns about this study. I am not convinced this study adds much to the scientific literature. I reported the missing figure panel through PubPeer and the authors are addressing the problem.