Profiling of serum metabolome of breast cancer: multi-cancer features discriminate between healthy women and patients with breast cancer.

Introduction: The progression of solid cancers is manifested at the systemic level as molecular changes in the metabolome of body fluids, an emerging source of cancer biomarkers. Methods: We analyzed quantitatively the serum metabolite profile using high-resolution mass spectrometry. Metabolic profiles were compared between breast cancer patients (n=112) and two groups of healthy women (from Poland and Norway; n=95 and n=112, respectively) with similar age distributions. Results: Despite differences between both cohorts of controls, a set of 43 metabolites and lipids uniformly discriminated against breast cancer patients and healthy women. Moreover, smaller groups of female patients with other types of solid cancers (colorectal, head and neck, and lung cancers) were analyzed, which revealed a set of 42 metabolites and lipids that uniformly differentiated all three cancer types from both cohorts of healthy women. A common part of both sets, which could be called a multi-cancer signature, contained 23 compounds, which included reduced levels of a few amino acids (alanine, aspartate, glutamine, histidine, phenylalanine, and leucine/isoleucine), lysophosphatidylcholines (exemplified by LPC(18:0)), and diglycerides. Interestingly, a reduced concentration of the most abundant cholesteryl ester (CE(18:2)) typical for other cancers was the least significant in the serum of breast cancer patients. Components present in a multi-cancer signature enabled the establishment of a well-performing breast cancer classifier, which predicted cancer with a very high precision in independent groups of women (AUC>0.95). Discussion: In conclusion, metabolites critical for discriminating breast cancer patients from controls included components of hypothetical multi-cancer signature, which indicated wider potential applicability of a general serum metabolome cancer biomarker.

Association of serum metabolome profile with the risk of breast cancer in participants of the HUNT2 study.

Background: The serum metabolome is a potential source of molecular biomarkers associated with the risk of breast cancer. Here we aimed to analyze metabolites present in pre-diagnostic serum samples collected from healthy women participating in the Norwegian Trøndelag Health Study (HUNT2 study) for whom long-term information about developing breast cancer was available. Methods: Women participating in the HUNT2 study who developed breast cancer within a 15-year follow-up period (BC cases) and age-matched women who stayed breast cancer-free were selected (n=453 case-control pairs). Using a high-resolution mass spectrometry approach 284 compounds were quantitatively analyzed, including 30 amino acids and biogenic amines, hexoses, and 253 lipids (acylcarnitines, glycerides, phosphatidylcholines, sphingolipids, and cholesteryl esters). Results: Age was a major confounding factor responsible for a large heterogeneity in the dataset, hence age-defined subgroups were analyzed separately. The largest number of metabolites whose serum levels differentiated BC cases and controls (82 compounds) were observed in the subgroup of younger women (<45 years old). Noteworthy, increased levels of glycerides, phosphatidylcholines, and sphingolipids were associated with reduced risk of cancer in younger and middle-aged women (≤64 years old). On the other hand, increased levels of serum lipids were associated with an enhanced risk of breast cancer in older women (>64 years old). Moreover, several metabolites could be detected whose serum levels were different between BC cases diagnosed earlier (<5 years) and later (>10 years) after sample collecting, yet these compounds were also correlated with the age of participants. Current results were coherent with the results of the NMR-based metabolomics study performed in the cohort of HUNT2 participants, where increased serum levels of VLDL subfractions were associated with reduced risk of breast cancer in premenopausal women. Conclusions: Changes in metabolite levels detected in pre-diagnostic serum samples, which reflected an impaired lipid and amino acid metabolism, were associated with long-term risk of breast cancer in an age-dependent manner.

Combining Low-Dose Computer-Tomography-Based Radiomics and Serum Metabolomics for Diagnosis of Malignant Nodules in Participants of Lung Cancer Screening Studies.

Radiomics is an emerging approach to support the diagnosis of pulmonary nodules detected via low-dose computed tomography lung cancer screening. Serum metabolome is a promising source of auxiliary biomarkers that could help enhance the precision of lung cancer diagnosis in CT-based screening. Thus, we aimed to verify whether the combination of these two techniques, which provides local/morphological and systemic/molecular features of disease at the same time, increases the performance of lung cancer classification models. The collected cohort consists of 1086 patients with radiomic and 246 patients with serum metabolomic evaluations. Different machine learning techniques, i.e., random forest and logistic regression were applied for each omics. Next, model predictions were combined with various integration methods to create a final model. The best single omics models were characterized by an AUC of 83% in radiomics and 60% in serum metabolomics. The model integration only slightly increased the performance of the combined model (AUC equal to 85%), which was not statistically significant. We concluded that radiomics itself has a good ability to discriminate lung cancer from benign lesions. However, additional research is needed to test whether its combination with other molecular assessments would further improve the diagnosis of screening-detected lung nodules.

Prognostic Value of Molecular Intratumor Heterogeneity in Primary Oral Cancer and Its Lymph Node Metastases Assessed by Mass Spectrometry Imaging.

Different aspects of intra-tumor heterogeneity (ITH), which are associated with the development of cancer and its response to treatment, have postulated prognostic value. Here we searched for potential association between phenotypic ITH analyzed by mass spectrometry imaging (MSI) and prognosis of head and neck cancer. The study involved tissue specimens resected from 77 patients with locally advanced oral squamous cell carcinoma, including 37 patients where matched samples of primary tumor and synchronous lymph node metastases were analyzed. A 3-year follow-up was available for all patients which enabled their separation into two groups: with no evidence of disease (NED, n = 41) and with progressive disease (PD, n = 36). After on-tissue trypsin digestion, peptide maps of all cancer regions were segmented using an unsupervised approach to reveal their intrinsic heterogeneity. We found that intra-tumor similarity of spectra was higher in the PD group and diversity of clusters identified during image segmentation was higher in the NED group, which indicated a higher level of ITH in patients with more favorable outcomes. Signature of molecular components that correlated with long-term outcomes could be associated with proteins involved in the immune functions. Furthermore, a positive correlation between ITH and histopathological lymphocytic host response was observed. Hence, we proposed that a higher level of ITH revealed by MSI in cancers with a better prognosis could reflect the presence of heterotypic components of tumor microenvironment such as infiltrating immune cells enhancing the response to the treatment.

Intra-Tumor Heterogeneity Revealed by Mass Spectrometry Imaging Is Associated with the Prognosis of Breast Cancer.

Intra-tumor heterogeneity (ITH) results from the coexistence of genetically distinct cancer cell (sub)populations, their phenotypic plasticity, and the presence of heterotypic components of the tumor microenvironment (TME). Here we addressed the potential association between phenotypic ITH revealed by mass spectrometry imaging (MSI) and the prognosis of breast cancer. Tissue specimens resected from 59 patients treated radically due to the locally advanced HER2-positive invasive ductal carcinoma were included in the study. After the on-tissue trypsin digestion of cellular proteins, peptide maps of all cancer regions (about 380,000 spectra in total) were segmented by an unsupervised approach to reveal their intrinsic heterogeneity. A high degree of similarity between spectra was observed, which indicated the relative homogeneity of cancer regions. However, when the number and diversity of the detected clusters of spectra were analyzed, differences between patient groups were observed. It is noteworthy that a higher degree of heterogeneity was found in tumors from patients who remained disease-free during a 5-year follow-up (n = 38) compared to tumors from patients with progressive disease (distant metastases detected during the follow-up, n = 21). Interestingly, such differences were not observed between patients with a different status of regional lymph nodes, cancer grade, or expression of estrogen receptor at the time of the primary treatment. Subsequently, spectral components with different abundance in cancer regions were detected in patients with different outcomes, and their hypothetical identity was established by assignment to measured masses of tryptic peptides identified in corresponding tissue lysates. Such differentiating components were associated with proteins involved in immune regulation and hemostasis. Further, a positive correlation between the level of tumor-infiltrating lymphocytes and heterogeneity revealed by MSI was observed. We postulate that a higher heterogeneity of tumors with a better prognosis could reflect the presence of heterotypic components including infiltrating immune cells, that facilitated the response to treatment.

The Lipid Composition of Serum-Derived Small Extracellular Vesicles in Participants of a Lung Cancer Screening Study.

Molecular components of exosomes and other classes of small extracellular vesicles (sEV) present in human biofluids are potential biomarkers with possible applicability in the early detection of lung cancer. Here, we compared the lipid profiles of serum-derived sEV from three groups of lung cancer screening participants: individuals without pulmonary alterations, individuals with benign lung nodules, and patients with screening-detected lung cancer (81 individuals in each group). Extracellular vesicles and particles were purified from serum by size-exclusion chromatography, and a fraction enriched in sEV and depleted of low-density lipoproteins (LDLs) was selected (similar sized vesicles was observed in all groups: 70-100 nm). The targeted mass-spectrometry-based approach enabled the detection of 352 lipids, including 201 compounds used in quantitative analyses. A few compounds, exemplified by Cer(42:1), i.e., a ceramide whose increased plasma/serum level was reported in different pathological conditions, were upregulated in vesicles from cancer patients. On the other hand, the contribution of phosphatidylcholines with poly-unsaturated acyl chains was reduced in vesicles from lung cancer patients. Cancer-related features detected in serum-derived sEV were different than those of the corresponding whole serum. A high heterogeneity of lipid profiles of sEV was observed, which markedly impaired the performance of classification models based on specific compounds (the three-state classifiers showed an average AUC = 0.65 and 0.58 in the training and test subsets, respectively).

Serum Metabolite Profiles in Participants of Lung Cancer Screening Study; Comparison of Two Independent Cohorts.

Serum metabolome is a promising source of molecular biomarkers that could support early detection of lung cancer in screening programs based on low-dose computed tomography. Several panels of metabolites that differentiate lung cancer patients and healthy individuals were reported, yet none of them were validated in the population at high-risk of developing cancer. Here we analyzed serum metabolome profiles in participants of two lung cancer screening studies: MOLTEST-BIS (Poland, n = 369) and SMAC-1 (Italy, n = 93). Three groups of screening participants were included: lung cancer patients, individuals with benign pulmonary nodules, and those without any lung alterations. Concentrations of about 400 metabolites (lipids, amino acids, and biogenic amines) were measured by a mass spectrometry-based approach. We observed a reduced level of lipids, in particular cholesteryl esters, in sera of cancer patients from both studies. Despite several specific compounds showing significant differences between cancer patients and healthy controls within each study, only a few cancer-related features were common when both cohorts were compared, which included a reduced concentration of lysophosphatidylcholine LPC (18:0). Moreover, serum metabolome profiles in both noncancer groups were similar, and differences between cancer patients and both groups of healthy participants were comparable. Large heterogeneity in levels of specific metabolites was observed, both within and between cohorts, which markedly impaired the accuracy of classification models: The overall AUC values of three-state classifiers were 0.60 and 0.51 for the test (MOLTEST) and validation (SMAC) cohorts, respectively. Therefore, a hypothetical metabolite-based biomarker for early detection of lung cancer would require adjustment to lifestyle-related confounding factors that putatively affect the composition of serum metabolome.

Classification of Thyroid Tumors Based on Mass Spectrometry Imaging of Tissue Microarrays; a Single-Pixel Approach.

The primary diagnosis of thyroid tumors based on histopathological patterns can be ambiguous in some cases, so proper classification of thyroid diseases might be improved if molecular biomarkers support cytological and histological assessment. In this work, tissue microarrays representative for major types of thyroid malignancies-papillary thyroid cancer (classical and follicular variant), follicular thyroid cancer, anaplastic thyroid cancer, and medullary thyroid cancer-and benign thyroid follicular adenoma and normal thyroid were analyzed by mass spectrometry imaging (MSI), and then different computation approaches were implemented to test the suitability of the registered profiles of tryptic peptides for tumor classification. Molecular similarity among all seven types of thyroid specimens was estimated, and multicomponent classifiers were built for sample classification using individual MSI spectra that corresponded to small clusters of cells. Moreover, MSI components showing the most significant differences in abundance between the compared types of tissues detected and their putative identity were established by annotation with fragments of proteins identified by liquid chromatography-tandem mass spectrometry in corresponding tissue lysates. In general, high accuracy of sample classification was associated with low inter-tissue similarity index and a high number of components with significant differences in abundance between the tissues. Particularly, high molecular similarity was noted between three types of tumors with follicular morphology (adenoma, follicular cancer, and follicular variant of papillary cancer), whose differentiation represented the major classification problem in our dataset. However, low level of the intra-tissue heterogeneity increased the accuracy of classification despite high inter-tissue similarity (which was exemplified by normal thyroid and benign adenoma). We compared classifiers based on all detected MSI components (n = 1536) and the subset of the most abundant components (n = 147). Despite relatively higher contribution of components with significantly different abundance and lower overall inter-tissue similarity in the latter case, the precision of classification was generally higher using all MSI components. Moreover, the classification model based on individual spectra (a single-pixel approach) outperformed the model based on mean spectra of tissue cores. Our result confirmed the high feasibility of MSI-based approaches to multi-class detection of cancer types and proved the good performance of sample classification based on individual spectra (molecular image pixels) that overcame problems related to small amounts of heterogeneous material, which limit the applicability of classical proteomics.

Molecular Heterogeneity of Papillary Thyroid Cancer: Comparison of Primary Tumors and Synchronous Metastases in Regional Lymph Nodes by Mass Spectrometry Imaging.

Intra-tumor heterogeneity results from both genetic heterogeneity of cancer (sub)clones and phenotypic plasticity of cancer cells that could be induced by different local microenvironments. Here, we used mass spectrometry imaging (MSI) to compare molecular profiles of primary tumors located in the thyroid gland and their synchronous metastases in regional lymph nodes to analyze phenotypic heterogeneity in papillary thyroid cancer. Two types of cancerous (primary tumor and metastasis) and two types of not cancerous (thyroid gland and lymph node) regions of interest (ROIs) were delineated in postoperative material from 11 patients, then the distribution of tryptic peptides (spectral components) was analyzed by MSI in all tissue regions. Moreover, tryptic peptides identified by shotgun proteomics in corresponding tissue lysates were matched to components detected by MSI to enable their hypothetical protein annotation. Unsupervised segmentation of all cancer ROIs revealed that different clusters dominated in tumor ROIs and metastasis ROIs. The intra-patient similarity between thyroid and tumor ROIs was higher than the intra-patient similarity between tumor and metastasis ROIs. Moreover, the similarity between tumor and its metastasis from the same patients was lower than similarities among tumors and among metastases from different patients (inter-patient similarity was higher for metastasis ROIs than for tumor ROIs). Components differentiating between tumor and its metastases were annotated as proteins involved in the organization of the cytoskeleton and chromatin, as well as proteins involved in immunity-related functions. We concluded that phenotypical heterogeneity between primary tumor and lymph node metastases from the same patient was higher than inter-tumor heterogeneity between primary tumors from different patients.

Discrimination of normal oral mucosa from oral cancer by mass spectrometry imaging of proteins and lipids.

Identification of biomarkers for molecular classification of cancer and for differentiation between cancerous and normal epithelium remains a vital issue in the field of head and neck cancer. Here we aimed to compare the ability of proteome and lipidome components to discriminate oral cancer from normal mucosa. Tissue specimens including squamous cell cancer and normal epithelium were analyzed by MALDI mass spectrometry imaging. Two molecular domains of tissue components were imaged in serial sections-peptides (resulting from trypsin-processed proteins) and lipids (primarily zwitterionic phospholipids), then regions of interest corresponding to cancer and normal epithelium were compared. Heterogeneity of cancer regions was higher than the heterogeneity of normal epithelium, and the distribution of peptide components was more heterogeneous than the distribution of lipid components. Moreover, there were more peptide components than lipid components that showed significantly different abundance between cancer and normal epithelium (median of the Cohen's effect was 0.49 and 0.31 in case of peptide and lipid components, respectively). Multicomponent cancer classifier was tested (vs. normal epithelium) using tissue specimens from three patients and then validated with a tissue specimen from the fourth patient. Peptide-based signature and lipid-based signature allowed cancer classification with a weighted accuracy of 0.85 and 0.69, respectively. Nevertheless, both classifiers had very high precision (0.98 and 0.94, respectively). We concluded that though molecular differences between cancerous and normal mucosa were higher in the proteome domain than in the analyzed lipidome subdomain, imaging of lipidome components also enabled discrimination of oral cancer and normal epithelium. Therefore, both cancer proteome and lipidome are promising sources of biomarkers of oral malignancies.

Molecular descriptor data explain market prices of a large commercial chemical compound library.

The relationship between the structure and a property of a chemical compound is an essential concept in chemistry guiding, for example, drug design. Actually, however, we need economic considerations to fully understand the fate of drugs on the market. We are performing here for the first time the exploration of quantitative structure-economy relationships (QSER) for a large dataset of a commercial building block library of over 2.2 million chemicals. This investigation provided molecular statistics that shows that on average what we are paying for is the quantity of matter. On the other side, the influence of synthetic availability scores is also revealed. Finally, we are buying substances by looking at the molecular graphs or molecular formulas. Thus, those molecules that have a higher number of atoms look more attractive and are, on average, also more expensive. Our study shows how data binning could be used as an informative method when analyzing big data in chemistry.

Ligand-Based Virtual Screening in a Search for Novel Anti-HIV-1 Chemotypes.

In a search for new anti-HIV-1 chemotypes, we developed a multistep ligand-based virtual screening (VS) protocol combining machine learning (ML) methods with the privileged structures (PS) concept. In its learning step, the VS protocol was based on HIV integrase (IN) inhibitors fetched from the ChEMBL database. The performances of various ML methods and PS weighting scheme were evaluated and applied as VS filtering criteria. Finally, a database of 1.5 million commercially available compounds was virtually screened using a multistep ligand-based cascade, and 13 selected unique structures were tested by measuring the inhibition of HIV replication in infected cells. This approach resulted in the discovery of two novel chemotypes with moderate antiretroviral activity, that, together with their topological diversity, make them good candidates as lead structures for future optimization.

Probing a chemical space for fragmental topology-activity landscapes (FRAGTAL): application for diketo acid and catechol HIV integrase inhibitor offspring fragments.

Fragmental topology-activity landscapes (FRAGTAL), a new concept for encoding molecular descriptors for fragonomics into the framework of the molecular database records is presented in this paper. Thus, a structural repository containing biological activity data was searched in a substructure mode by a series of molecular fragments constructed in an incremental or decremental manner. The resulted series of database hits annotated with their activities construct FRAGTAL descriptors encoding a frequency of the certain fragments among active compounds and/or their activities. Actually, this method might be interpreted as a simplified adaptation of the frequent subgraph mining (FSM) method. The FRAGTAL method reconstructs the way in which medicinal chemists are used to designing a prospective drug structure intuitively. A representative example of the practical application of FRAGTAL within the ChemDB Anti-HIV/OI/TB database for disclosing new fragments for HIV-1 integrase inhibition is discussed. In particular, FRAGTAL method identifies ethyl malonate amide (EMA) as the diketo acid (DKA) related arrangement. Since new molecular constructs based on the EMA fragment are still a matter of future investigations we referred to this as anthe DKA offspring.

Mapping fragmental drug-likeness in the MoStBioDat environment: intramolecular hydrogen bonding motifs in ß-ketoenols.

A detailed knowledge of hydrogen bond geometry and its directional preferences is vital for in silico investigations of the ligand-receptor short-range non-covalent interactions. The spatial arrangement of the carbonyl and hydroxyl groups seems to determine the capability of ß-ketoenol derivatives to recognize the surrounding environment by forming inter- and intra-molecular hydrogen bonds (IHB). In the current study we examined the application of the MoStBioDat platform for a massive database screening of the IHB motifs in ß-ketoenol subunits (O=C-C=C-OH). Then, the virtual 3D structural data derived from ZINC and PubChem repository were compared to the experimentally determined CSD data. Differences specific for each database were discovered, which indicated inaccuracies in the simulated data.

MoStBioDat--molecular and structural bioinformatics database.

Computer simulations play a crucial role in contemporary chemical investigations, generating enormous amounts of data. The constraint of sharing data and results is regarded as a major impediment in drug discovery. Among the steepest barriers to overcome in the high throughput screening studies is the limited number of suitable, freely accessible repositories for storing drug and drug target data. By offering a uniform data storage and retrieval mechanism, various data might be compared and exchanged easily. This paper presents the stages of the MoStBioDat software platform development, originally designed for the efficient storage, management and access of SDF and PDB data. The detailed architecture and software implementation of this project are described, indicating also the disadvantages of the solutions chosen. The current implementation of the first prototype is written in Python, an open-source, high-level, object-oriented scripting language. The modular architecture of the package enables future extension with the necessary functionalities. The main objective of the MoStBioDat is to serve as an alternative, extensible open-source database derived partly from SDF and PDB files.

The use of MoStBioDat for rapid screening of molecular diversity.

MoStBioDat is a uniform data storage and extraction system with an extensive array of tools for structural similarity measures and pattern matching which is essential to facilitate the drug discovery process. Structure-based database screening has recently become a common and efficient technique in early stages of the drug development, shifting the emphasis from rational drug design into the probability domain of more or less random discovery. The virtual ligand screening (VLS), an approach based on high-throughput flexible docking, samples a virtually infinite molecular diversity of chemical libraries increasing the concentration of molecules with high binding affinity. The rapid process of subsequent examination of a large number of molecules in order to optimize the molecular diversity is an attractive alternative to the traditional methods of lead discovery. This paper presents the application of the MoStBioDat package not only as a data management platform but mainly in substructure searching. In particular, examples of the applications of MoStBioDat are discussed and analyzed.