Unifying immunology with informatics and multiscale biology.

The immune system is a highly complex and dynamic system. Historically, the most common scientific and clinical practice has been to evaluate its individual components. This kind of approach cannot always expose the interconnecting pathways that control immune-system responses and does not reveal how the immune system works across multiple biological systems and scales. High-throughput technologies can be used to measure thousands of parameters of the immune system at a genome-wide scale. These system-wide surveys yield massive amounts of quantitative data that provide a means to monitor and probe immune-system function. New integrative analyses can help synthesize and transform these data into valuable biological insight. Here we review some of the computational analysis tools for high-dimensional data and how they can be applied to immunology.

Structural basis for mechanical force regulation of the adhesin FimH via finger trap-like beta sheet twisting.

The Escherichia coli fimbrial adhesive protein, FimH, mediates shear-dependent binding to mannosylated surfaces via force-enhanced allosteric catch bonds, but the underlying structural mechanism was previously unknown. Here we present the crystal structure of FimH incorporated into the multiprotein fimbrial tip, where the anchoring (pilin) domain of FimH interacts with the mannose-binding (lectin) domain and causes a twist in the beta sandwich fold of the latter. This loosens the mannose-binding pocket on the opposite end of the lectin domain, resulting in an inactive low-affinity state of the adhesin. The autoinhibition effect of the pilin domain is removed by application of tensile force across the bond, which separates the domains and causes the lectin domain to untwist and clamp tightly around the ligand like a finger-trap toy. Thus, beta sandwich domains, which are common in multidomain proteins exposed to tensile force in vivo, can undergo drastic allosteric changes and be subjected to mechanical regulation.

Clonal Deletion Prunes but Does Not Eliminate Self-Specific αß CD8(+) T Lymphocytes.

It has long been thought that clonal deletion efficiently removes almost all self-specific T cells from the peripheral repertoire. We found that self-peptide MHC-specific CD8(+) T cells in the blood of healthy humans were present in frequencies similar to those specific for non-self antigens. For the Y chromosome-encoded SMCY antigen, self-specific T cells exhibited only a 3-fold lower average frequency in males versus females and were anergic with respect to peptide activation, although this inhibition could be overcome by a stronger stimulus. We conclude that clonal deletion prunes but does not eliminate self-specific T cells and suggest that to do so would create holes in the repertoire that pathogens could readily exploit. In support of this hypothesis, we detected T cells specific for all 20 amino acid variants at the p5 position of a hepatitis C virus epitope in a random group of blood donors.

Virus-specific CD4(+) memory-phenotype T cells are abundant in unexposed adults.

Although T cell memory is generally thought to require direct antigen exposure, we found an abundance of memory-phenotype cells (20%-90%, averaging over 50%) of CD4(+) T cells specific to viral antigens in adults who had never been infected. These cells express the appropriate memory markers and genes, rapidly produce cytokines, and have clonally expanded. In contrast, the same T cell receptor (TCR) specificities in newborns are almost entirely naïve, which might explain the vulnerability of young children to infections. One mechanism for this phenomenon is TCR cross-reactivity to environmental antigens, and in support of this, we found extensive cross-recognition by HIV-1 and influenza-reactive T lymphocytes to other microbial peptides and expansion of one of these after influenza vaccination. Thus, the presence of these memory-phenotype T cells has significant implications for immunity to novel pathogens, child and adult health, and the influence of pathogen-rich versus hygienic environments.

γδ T cells recognize a microbial encoded B cell antigen to initiate a rapid antigen-specific interleukin-17 response.

γδ T cells contribute uniquely to immune competence. Nevertheless, how they function remains an enigma. It is unclear what most γδ T cells recognize, what is required for them to mount an immune response, and how the γδ T cell response is integrated into host immune defense. Here, we report that a noted B cell antigen, the algae protein phycoerythrin (PE), is a murine and human γδ T cell antigen. Employing this specificity, we demonstrated that antigen recognition activated naive γδ T cells to make interleukin-17 and respond to cytokine signals that perpetuate the response. High frequencies of antigen-specific γδ T cells in naive animals and their ability to mount effector response without extensive clonal expansion allow γδ T cells to initiate a swift, substantial response. These results underscore the adaptability of lymphocyte antigen receptors and suggest an antigen-driven rapid response in protective immunity prior to the maturation of classical adaptive immunity.

Systematic analyses of drugs and disease indications in RepurposeDB reveal pharmacological, biological and epidemiological factors influencing drug repositioning.

Increase in global population and growing disease burden due to the emergence of infectious diseases (Zika virus), multidrug-resistant pathogens, drug-resistant cancers (cisplatin-resistant ovarian cancer) and chronic diseases (arterial hypertension) necessitate effective therapies to improve health outcomes. However, the rapid increase in drug development cost demands innovative and sustainable drug discovery approaches. Drug repositioning, the discovery of new or improved therapies by reevaluation of approved or investigational compounds, solves a significant gap in the public health setting and improves the productivity of drug development. As the number of drug repurposing investigations increases, a new opportunity has emerged to understand factors driving drug repositioning through systematic analyses of drugs, drug targets and associated disease indications. However, such analyses have so far been hampered by the lack of a centralized knowledgebase, benchmarking data sets and reporting standards. To address these knowledge and clinical needs, here, we present RepurposeDB, a collection of repurposed drugs, drug targets and diseases, which was assembled, indexed and annotated from public data. RepurposeDB combines information on 253 drugs [small molecules (74.30%) and protein drugs (25.29%)] and 1125 diseases. Using RepurposeDB data, we identified pharmacological (chemical descriptors, physicochemical features and absorption, distribution, metabolism, excretion and toxicity properties), biological (protein domains, functional process, molecular mechanisms and pathway cross talks) and epidemiological (shared genetic architectures, disease comorbidities and clinical phenotype similarities) factors mediating drug repositioning. Collectively, RepurposeDB is developed as the reference database for drug repositioning investigations. The pharmacological, biological and epidemiological principles of drug repositioning identified from the meta-analyses could augment therapeutic development.

Environments Tune and Select Cellular Diversity.

Technical advances in single-cell sequencing data and their application to greater samples is revealing substantial cell-to-cell variation in expression levels and propagation of this variation between molecules across cells. New quantitative approaches that apply mechanistic and statistical models in a systems-wide approach are illuminating the drivers of phenotypic diversity.

Automated cell type discovery and classification through knowledge transfer.

MOTIVATION: Recent advances in mass cytometry allow simultaneous measurements of up to 50 markers at single-cell resolution. However, the high dimensionality of mass cytometry data introduces computational challenges for automated data analysis and hinders translation of new biological understanding into clinical applications. Previous studies have applied machine learning to facilitate processing of mass cytometry data. However, manual inspection is still inevitable and becoming the barrier to reliable large-scale analysis. RESULTS: We present a new algorithm called utomated ell-type iscovery and lassification (ACDC) that fully automates the classification of canonical cell populations and highlights novel cell types in mass cytometry data. Evaluations on real-world data show ACDC provides accurate and reliable estimations compared to manual gating results. Additionally, ACDC automatically classifies previously ambiguous cell types to facilitate discovery. Our findings suggest that ACDC substantially improves both reliability and interpretability of results obtained from high-dimensional mass cytometry profiling data. AVAILABILITY AND IMPLEMENTATION: A Python package (Python 3) and analysis scripts for reproducing the results are availability on https://bitbucket.org/dudleylab/acdc . CONTACT: brian.kidd@mssm.edu or joel.dudley@mssm.edu. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Variants in TRIM22 That Affect NOD2 Signaling Are Associated With Very-Early-Onset Inflammatory Bowel Disease.

BACKGROUND & AIMS: Severe forms of inflammatory bowel disease (IBD) that develop in very young children can be caused by variants in a single gene. We performed whole-exome sequence (WES) analysis to identify genetic factors that might cause granulomatous colitis and severe perianal disease, with recurrent bacterial and viral infections, in an infant of consanguineous parents. METHODS: We performed targeted WES analysis of DNA collected from the patient and her parents. We validated our findings by a similar analysis of DNA from 150 patients with very-early-onset IBD not associated with known genetic factors analyzed in Toronto, Oxford, and Munich. We compared gene expression signatures in inflamed vs noninflamed intestinal and rectal tissues collected from patients with treatment-resistant Crohn's disease who participated in a trial of ustekinumab. We performed functional studies of identified variants in primary cells from patients and cell culture. RESULTS: We identified a homozygous variant in the tripartite motif containing 22 gene (TRIM22) of the patient, as well as in 2 patients with a disease similar phenotype. Functional studies showed that the variant disrupted the ability of TRIM22 to regulate nucleotide binding oligomerization domain containing 2 (NOD2)-dependent activation of interferon-beta signaling and nuclear factor-κB. Computational studies demonstrated a correlation between the TRIM22-NOD2 network and signaling pathways and genetic factors associated very early onset and adult-onset IBD. TRIM22 is also associated with antiviral and mycobacterial effectors and markers of inflammation, such as fecal calprotectin, C-reactive protein, and Crohn's disease activity index scores. CONCLUSIONS: In WES and targeted exome sequence analyses of an infant with severe IBD characterized by granulomatous colitis and severe perianal disease, we identified a homozygous variant of TRIM22 that affects the ability of its product to regulate NOD2. Combined computational and functional studies showed that the TRIM22-NOD2 network regulates antiviral and antibacterial signaling pathways that contribute to inflammation. Further study of this network could lead to new disease markers and therapeutic targets for patients with very early and adult-onset IBD.

Allosteric coupling in the bacterial adhesive protein FimH.

The protein FimH is expressed by the majority of commensal and uropathogenic strains of Escherichia coli on the tips of type 1 fimbriae and mediates adhesion via a catch bond to its ligand mannose. Crystal structures of FimH show an allosteric conformational change, but it remains unclear whether all of the observed structural differences are part of the allosteric mechanism. Here we use the protein structural analysis tool RosettaDesign combined with human insight to identify and synthesize 10 mutations in four regions that we predicted would stabilize one of the conformations of that region. The function of each variant was characterized by measuring binding to the ligand mannose, whereas the allosteric state was determined using a conformation-specific monoclonal antibody. These studies demonstrated that each region investigated was indeed part of the FimH allosteric mechanism. However, the studies strongly suggested that some regions were more tightly coupled to mannose binding and others to antibody binding. In addition, we identified many FimH variants that appear locked in the low affinity state. Knowledge of regulatory sites outside the active and effector sites as well as the ability to make FimH variants locked in the low affinity state may be crucial to the future development of novel antiadhesive and antimicrobial therapies using allosteric regulation to inhibit FimH.

The bacterial fimbrial tip acts as a mechanical force sensor.

There is increasing evidence that the catch bond mechanism, where binding becomes stronger under tensile force, is a common property among non-covalent interactions between biological molecules that are exposed to mechanical force in vivo. Here, by using the multi-protein tip complex of the mannose-binding type 1 fimbriae of Escherichia coli, we show how the entire quaternary structure of the adhesive organella is adapted to facilitate binding under mechanically dynamic conditions induced by flow. The fimbrial tip mediates shear-dependent adhesion of bacteria to uroepithelial cells and demonstrates force-enhanced interaction with mannose in single molecule force spectroscopy experiments. The mannose-binding, lectin domain of the apex-positioned adhesive protein FimH is docked to the anchoring pilin domain in a distinct hooked manner. The hooked conformation is highly stable in molecular dynamics simulations under no force conditions but permits an easy separation of the domains upon application of an external tensile force, allowing the lectin domain to switch from a low- to a high-affinity state. The conformation between the FimH pilin domain and the following FimG subunit of the tip is open and stable even when tensile force is applied, providing an extended lever arm for the hook unhinging under shear. Finally, the conformation between FimG and FimF subunits is highly flexible even in the absence of tensile force, conferring to the FimH adhesin an exploratory function and high binding rates. The fimbrial tip of type 1 Escherichia coli is optimized to have a dual functionality: flexible exploration and force sensing. Comparison to other structures suggests that this property is common in unrelated bacterial and eukaryotic adhesive complexes that must function in dynamic conditions.

Acute Effects of Growth Hormone on the Cellular Immunologic Landscape in Pediatric Patients.

INTRODUCTION: Growth hormone (GH) and the immune system have multiple bidirectional interactions. Data about the acute effects of GH on the immune system are lacking. The objective of our study was to evaluate the acute effects of GH on the immune system using time-of-flight mass cytometry. METHODS: This was a prospective study of pediatric patients who were being evaluated for short stature and underwent a GH stimulation test at a tertiary care center. Blood samples for immunologic markers, i.e., complete blood count (CBC) and time of flight mass cytometry (CyTOF), were collected at baseline (T0) and over the course of three hours (T3) of the test. Differences in immune profiling in patients by timepoint (T0, T3) and GH response (growth hormone sufficient (GHS) versus growth hormone deficient (GHD)) were calculated using a two-way ANOVA test.  Results: A total of 54 patients (39 boys and 15 girls) aged five to 18 years were recruited. Twenty-two participants tested GHD (peak GH <10 ng/ml). The CyTOF analysis showed a significant increase from T0 to T3 in granulocyte percentage, monocyte count, and dendritic cell (DC) count; in contrast, a significant decrease was seen in T lymphocytes (helper and cytotoxic) and IgD+ B lymphocytes. The CBC analysis supported these findings: an increase in total white blood cell count, absolute neutrophil count, and neutrophil percentage; a decrease in absolute lymphocyte count, lymphocyte percentage, absolute eosinophil count, and absolute monocyte count. No significant differences were found between CBC/CyTOF measurements and GH status at either time. CONCLUSIONS: This study provides the first high-resolution map of acute changes in the immune system with GH stimulation. This implies a key role for GH in immunomodulatory function.

Serum autoantibodies to myelin peptides distinguish acute disseminated encephalomyelitis from relapsing-remitting multiple sclerosis.

BACKGROUND AND OBJECTIVE: Acute disseminated encephalomyelitis (ADEM) and relapsing-remitting multiple sclerosis (RRMS) share overlapping clinical, radiologic and laboratory features at onset. Because autoantibodies may contribute to the pathogenesis of both diseases, we sought to identify autoantibody biomarkers that are capable of distinguishing them. METHODS: We used custom antigen arrays to profile anti-myelin-peptide autoantibodies in sera derived from individuals with pediatric ADEM (n = 15), pediatric multiple sclerosis (Ped MS; n = 11) and adult MS (n = 15). Using isotype-specific secondary antibodies, we profiled both IgG and IgM reactivities. We used Statistical Analysis of Microarrays software to confirm the differences in autoantibody reactivity profiles between ADEM and MS samples. We used Prediction Analysis of Microarrays software to generate and validate prediction algorithms, based on the autoantibody reactivity profiles. RESULTS: ADEM was characterized by IgG autoantibodies targeting epitopes derived from myelin basic protein, proteolipid protein, myelin-associated oligodendrocyte basic glycoprotein, and alpha-B-crystallin. In contrast, MS was characterized by IgM autoantibodies targeting myelin basic protein, proteolipid protein, myelin-associated oligodendrocyte basic glycoprotein and oligodendrocyte-specific protein. We generated and validated prediction algorithms that distinguish ADEM serum (sensitivity 62-86%; specificity 56-79%) from MS serum (sensitivity 40-87%; specificity 62-86%) on the basis of combined IgG and IgM anti-myelin autoantibody reactivity to a small number of myelin peptides. CONCLUSIONS: Combined profiles of serum IgG and IgM autoantibodies identified myelin antigens that may be useful for distinguishing MS from ADEM. Further studies are required to establish clinical utility. Further biological assays are required to delineate the pathogenic potential of these antibodies.

Cancer-associated fibroblasts are the main contributors to epithelial-to-mesenchymal signatures in the tumor microenvironment.

Epithelial-to-mesenchymal transition (EMT) is associated with tumor initiation, metastasis, and drug resistance. However, the mechanisms underlying these associations are largely unknown. We studied several tumor types to identify the source of EMT gene expression signals and a potential mechanism of resistance to immuno-oncology treatment. Across tumor types, EMT-related gene expression was strongly associated with expression of stroma-related genes. Based on RNA sequencing of multiple patient-derived xenograft models, EMT-related gene expression was enriched in the stroma versus parenchyma. EMT-related markers were predominantly expressed by cancer-associated fibroblasts (CAFs), cells of mesenchymal origin which produce a variety of matrix proteins and growth factors. Scores derived from a 3-gene CAF transcriptional signature (COL1A1, COL1A2, COL3A1) were sufficient to reproduce association between EMT-related markers and disease prognosis. Our results suggest that CAFs are the primary source of EMT signaling and have potential roles as biomarkers and targets for immuno-oncology therapies.

Donor strand exchange and conformational changes during E. coli fimbrial formation.

Fimbriae and pili are macromolecular structures on the surface of Gram negative bacteria that are important for cellular adhesion. A 2.7Å resolution crystal structure of a complex of Escherichia coli fimbrial proteins containing FimH, FimG, FimF, and FimC provides the most complete model to date for the arrangement of subunits assembled in the native structure. The first three proteins form the tip of the fimbriae while FimC is the chaperone protein involved in the usher/chaperone assembly process. The subunits interact through donor strand complementation where a ß-strand from a subunit completes the ß-sandwich structure of the neighboring subunit or domain closer to the tip of the fimbria. The function of FimC is to provide a surrogate donor strand before delivery of each subunit to the FimD usher and the growing fimbria. Comparison of the subunits in this structure and their chaperone-bound complexes show that the two FimH domains change their relative orientation and position in forming the tip structure. Also, the non-chaperone subunits undergo a conformational change in their first ß-strand when the chaperone is replaced by the native donor strand. Some residues move as much as 14Å in the process. This structural shift has not been noted in structural studies of other bacterial adhesion sub-structures assembled via donor strand complementation. The domains undergo a significant structural change in the donor strand binding groove during fimbrial assembly, and this likely plays a role in determining the specificity of subunit-subunit interactions among the fimbrial proteins.

Computation of conformational coupling in allosteric proteins.

In allosteric regulation, an effector molecule binding a protein at one site induces conformational changes, which alter structure and function at a distant active site. Two key challenges in the computational modeling of allostery are the prediction of the structure of one allosteric state starting from the structure of the other, and elucidating the mechanisms underlying the conformational coupling of the effector and active sites. Here we approach these two challenges using the Rosetta high-resolution structure prediction methodology. We find that the method can recapitulate the relaxation of effector-bound forms of single domain allosteric proteins into the corresponding ligand-free states, particularly when sampling is focused on regions known to change conformation most significantly. Analysis of the coupling between contacting pairs of residues in large ensembles of conformations spread throughout the landscape between and around the two allosteric states suggests that the transitions are built up from blocks of tightly coupled interacting sets of residues that are more loosely coupled to one another.

Phase 2 trial of a DNA vaccine encoding myelin basic protein for multiple sclerosis.

OBJECTIVE: To evaluate the efficacy and safety of BHT-3009 in relapsing-remitting multiple sclerosis (MS) and to confirm that BHT-3009 causes immune tolerance. METHODS: BHT-3009 is a tolerizing DNA vaccine for MS, encoding full-length human myelin basic protein. Relapsing-remitting MS patients were randomized 1:1:1 into three groups: placebo, 0.5 mg BHT-3009, or 1.5 mg BHT-3009, given intramuscularly at weeks 0, 2, 4, and every 4 weeks thereafter until week 44. The primary end point was the 4-week rate of occurrence of new gadolinium-enhancing lesions on brain magnetic resonance images from weeks 28 to 48. Protein microarrays were used to measure levels of anti-myelin autoantibodies. RESULTS: Compared with placebo, in the 267 patient analysis population the median 4-week rate of new enhancing lesions during weeks 28 to 48 was 50% lower with 0.5 mg BHT-3009 (p = 0.07) and during weeks 8 to 48 was 61% lower with 0.5 mg BHT-3009 (p = 0.05). The mean volume of enhancing lesions at week 48 was 51% lower on 0.5 mg BHT-3009 compared with placebo (p = 0.02). No significant improvement in magnetic resonance imaging lesion parameters was observed with 1.5 mg BHT-3009. Dramatic reductions in 23 myelin-specific autoantibodies in the 0.5 mg BHT-3009 arm were observed, but not with placebo or 1.5 mg BHT-3009. CONCLUSIONS: In relapsing-remitting MS patients, treatment with the lower dose (0.5 mg) of BHT-3009 for 44 weeks nearly attained the primary end point for reduction of the rate of new enhancing magnetic resonance imaging lesions (p = 0.07) and achieved several secondary end points including a reduction of the rate of enhancing magnetic resonance imaging lesions from weeks 8 to 48 (p = 0.05). Immunological data in a preselected subgroup of patients also indicated that treatment with 0.5 mg induced antigen-specific immune tolerance. The greater dose was ineffective.

Lyme Disease Patient Trajectories Learned from Electronic Medical Data for Stratification of Disease Risk and Therapeutic Response.

Lyme disease (LD) is the most common tick-borne illness in the United States. Although appropriate antibiotic treatment is effective for most cases, up to 20% of patients develop post-treatment Lyme disease syndrome (PTLDS). There is an urgent need to improve clinical management of LD using precise understanding of disease and patient stratification. We applied machine-learning to electronic medical records to better characterize the heterogeneity of LD and developed predictive models for identifying medications that are associated with risks of subsequent comorbidities. For broad disease categories, we identified 3, 16, and 17 comorbidities within 2, 5, and 10 years of diagnosis, respectively. At a higher resolution of ICD-9 codes, we identified known associations with LD including chronic pain and cognitive disorders, as well as particular comorbidities on a timescale that matched PTLDS symptomology. We identified 7, 30, and 35 medications associated with risks of these comorbidities within 2, 5, and 10 years, respectively. For instance, the first-line antibiotic doxycycline exhibited a consistently protective association for typical symptoms of LD, including backache. Our approach and findings may suggest new hypotheses for more personalized treatments regimens for LD patients.

Reflecting health: smart mirrors for personalized medicine.

Inexpensive embedded computing and the related Internet of Things technologies enable the recent development of smart products that can respond to human needs and improve everyday tasks in an attempt to make traditional environments more "intelligent". Several projects have augmented mirrors for a range of smarter applications in automobiles and homes. The opportunity to apply smart mirror technology to healthcare to predict and to monitor aspects of health and disease is a natural but mostly underdeveloped idea. We envision that smart mirrors comprising a combination of intelligent hardware and software could identify subtle, yet clinically relevant changes in physique and appearance. Similarly, a smart mirror could record and evaluate body position and motion to identify posture and movement issues, as well as offer feedback for corrective actions. Successful development and implementation of smart mirrors for healthcare applications will require overcoming new challenges in engineering, machine learning, computer vision, and biomedical research. This paper examines the potential uses of smart mirrors in healthcare and explores how this technology might benefit users in various medical environments. We also provide a brief description of the state-of-the-art, including a functional prototype concept developed by our group, and highlight the directions to make this device more mainstream in health-related applications.