Human milk banks (HMBs) are an essential clinical service, providing safe, screened donor human milk (DHM) to clinically vulnerable infants, which can enable mothers to establish their own milk supply. While HMB processes share similarities with other medical products of human origin (e.g., blood, platelets), HMBs remain marginalised services with limited resilience to external or internal pressures. As demand for DHM increases, understanding how to mitigate these vulnerabilities is essential. In November 2022, a workshop organised by the Human Milk Foundation brought together UK milk bank leaders, academics, and planners. The workshop aimed to understand existing pressures faced by UK HMBs and create a network to support strategic, evidence-based responses. Emergent themes highlighted that HMBs are stretched, facilitated largely by goodwill. Most services are reliant on limited workforces with limitations in service continuity, training, and succession planning. COVID-19 exacerbated pressures, but led to HMBs cooperating more. Proposed responses included investment in staffing, training and IT resources, greater public awareness and inter-HMB cooperation, and a national HMB Risk Register. Service continuity needs to be urgently addressed through external multiagency engagement, operational support, and research prioritization (including qualitative, technological and implementation science), defining optimal service design and financial resourcing to meet nationally agreed needs. Without intervention, the UK milk bank network may be unable to respond to future significant pressures, including specialist infant feed or infant formula shortages. With appropriate investment, a comprehensive National Service can be created to provide equitable services underpinned by robust research and innovation.

When maternal milk is unavailable, donor human milk (DHM) from human milk banks (HMBs) is the optimal alternative, as recommended by the World Health Organisation. The microbiota of DHM could contain opportunistic pathogens, which means rigorous microbiological screening for DHM, prior to pasteurisation, is recommended to safeguard recipients. Here, an analysis of 6863 DHM samples from 1419 donors at the Hearts Milk Bank between 2017 and 2023 showed approximately 70.1% of samples exhibited a total viable count (TVC) between 10³-10⁵ CFU/mL, while 18.3% yielded no growth; 11.5% of samples exceeded the 10⁵ CFU/mL threshold. Staphylococcus was the most prevalent genus, with S. epidermidis found in 61.5% of samples. A significant (p < 0.05) negative co-occurrence was observed between S. epidermidis and Gram-negative opportunistic pathogens. Overall, 16.8% of DHM samples failed to meet UK microbiological screening criteria, with 68.3% of these failures due to exceeding TVC thresholds. S. epidermidis accounted for approximately 10.2% of the total failed samples. The majority of DHM samples met the current microbiological criteria specified in the National Institute for Health and Care Excellence (NICE) clinical guidance (CG93), “Donor milk banks: service operation”. The core species in DHM reflects microorganisms typically found on the skin. These findings highlight that the current UK thresholds and criteria could potentially be modified to increase the available supply of DHM without increasing microbiological risk.

Background: The human gut microbiota develops in concordance with its host over a lifetime, resulting in age-related shifts in community structure and metabolic function. Little is known about whether these changes impact the community’s response to microbiome-targeted therapeutics. Providing critical information on this subject, faecal microbiomes of subjects from six age groups, spanning from infancy to 70-year-old adults (n = six per age group) were harvested. The responses of these divergent communities to treatment with the human milk oligosaccharide 2’-fucosyllactose (2’FL), fructo-oligosaccharides (FOS), and lactose was investigated using the Ex vivo SIFR® technology that employs bioreactor fermentation and is validated to be predictive of clinical findings. Additionally, it was evaluated whether combining faecal microbiomes of a given age group into a single pooled microbiome produced similar results as the individual microbiomes. Results: First, marked age-dependent changes in community structure were identified. Bifidobacterium levels strongly declined as age increased, and Bifidobacterium species composition was age-dependent: B. longum, B. catenulatum/pseudocatenulatum, and B. adolescentis were most prevalent for breastfed infants, toddlers/children, and adults, respectively. Metabolomic analyses (LA-REIMS) demonstrated that these age-dependent differences particularly impacted treatment effects of 2’FL (more than FOS/lactose). Further analysis revealed that while 2’FL enhanced production of short-chain fatty acids (SCFAs) and exerted potent bifidogenic effects, regardless of age, the specific Bifidobacterium species enhanced by 2’FL, as well as subsequent cross-feeding interactions, were highly age-dependent. Furthermore, single-pooled microbiomes produced results that were indicative of the average treatment response for each age group. Nevertheless, pooled microbiomes had an artificially high diversity, thus overestimating treatment responses (especially for infants), did not recapitulate interindividual variation, and disallowed for the correlative analysis required to unravel mechanistic actions. Conclusions: Age is an important factor in shaping the gut microbiome, with the dominant taxa and their metabolites changing over a lifetime. This divergence affects the response of the microbiota to therapeutics, demonstrated in this study using 2’FL. These results evidence the importance of screening across multiple age groups separately to provide granularity of how therapeutics impact the microbiome and, consequently, human health.

Maximizing the extraction of true, high-quality, nonredundant features from biofluids analyzed via LC-MS systems is challenging. Here, the R packages IPO and AutoTuner were used to optimize XCMS parameter settings for the retrieval of metabolite or lipid features in both ionization modes from either faecal or urine samples from two cohorts (n = 621). The feature lists obtained were compared with those where the parameter values were selected manually. Three categories were used to compare feature lists: 1) feature quality through removing false positives, 2) tentative metabolite identification using the Human Metabolome Database (HMDB) and 3) feature utility such as analyzing the proportion of features within intensity threshold bins. Furthermore, a PCA-based approach to feature filtering using QC samples and variable loadings was also explored under this category. Overall, more features were observed after automated selection of parameter values for all data sets (1.3- to 3.7-fold), which propagated through comparative exercises. For example, a greater number of features (on average 51 vs 45%) had a coefficient of variation (CV) < 30%. Additionally, there was a significant increase (7.6-10.4%) in the number of faecal metabolites that could be tentatively annotated, and more features were present in higher intensity threshold bins. Considering the overlap across all three categories, a greater number of features were also retained. Automated approaches that guide selection of optimal parameter values for preprocessing are important to decrease the time invested for this step, while taking advantage of the wealth of data that LC-MS systems provide.

Breastfeeding, which is recognised as the optimum nutrition for infants, offers numerous benefits. However, circumstances can arise when infants are unable to be breastfed from birth. In such cases, the World Health Organisation (WHO) recommends donor human milk (DHM) as the safest alternative. Current practices freeze DHM and transport it under a cold supply chain, which can create logistical challenges. Here, we investigated the efficacy of freeze-drying as a method for determining the compositional stability of DHM. The samples were freeze-dried and stored at-20 degrees C, 4 degrees C and ambient temperature, with sampling at 1, 3, 6, 9, and 12 months. The macronutrient composition was assessed before and after freeze-drying, protein and lipid profiles were studied using MALDI-TOF MS, and the metabolite profile was analysed through LA-REIMS. The findings revealed that freeze-drying did not significantly alter the macronutrient composition and that microbiological safety was preserved. Lipid, protein, and metabolite fingerprints remained consistent across storage conditions over 12 months. This work provides a broad insight into the compositional stability of DHM after freeze-drying. It suggests the applicability of freeze-drying for long-term preservation without a cold supply chain. The use of freeze-dried DHM may broaden its use in emergency situations and resource-limited settings.

The identification of microorganisms in environmental science is a key component in the process of understanding community structure, function, and interactions. For the past two decades, this process has relied on the use of molecular profiling methods to sequence DNA and RNA. Although informative, this approach is limited in terms of functional understanding of microbial communities in environmental processes and systems. Mass spectrometry (MS) offers novel analytical approaches to both culture-dependent and culture-independent microbial identification and functional profiling, and holds potential to provide ever greater insight into microbial community structure and function in the environmental sciences. This chapter explores the requirement for microbial identification in the environmental sciences and MS methods for both culture-dependent and culture-independent identification. It goes on to discuss the role of MS in functional profiling of microbial communities in environmental science and combines MS with identification methods in several case studies. Finally, the chapter ends with a discussion of the future role and opportunities of MS in microbial research within the environmental sciences.

The gut microbiota is implicated in the pathogenesis of colorectal cancer (CRC). We aimed to map the CRC mucosal microbiota and metabolome and define the influence of the tumoral microbiota on oncological outcomes. A multicentre, prospective observational study was conducted of CRC patients undergoing primary surgical resection in the UK (n = 74) and Czech Republic (n = 61). Analysis was performed using metataxonomics, ultra-performance liquid chromatography-mass spectrometry (UPLC-MS), targeted bacterial qPCR and tumour exome sequencing. Hierarchical clustering accounting for clinical and oncological covariates was performed to identify clusters of bacteria and metabolites linked to CRC. Cox proportional hazards regression was used to ascertain clusters associated with disease-free survival over median follow-up of 50 months. Thirteen mucosal microbiota clusters were identified, of which five were significantly different between tumour and paired normal mucosa. Cluster 7, containing the pathobionts Fusobacterium nucleatum and Granulicatella adiacens, was strongly associated with CRC (PFDR = 0.0002). Additionally, tumoral dominance of cluster 7 independently predicted favourable disease-free survival (adjusted p = 0.031). Cluster 1, containing Faecalibacterium prausnitzii and Ruminococcus gnavus, was negatively associated with cancer (PFDR = 0.0009), and abundance was independently predictive of worse disease-free survival (adjusted p = 0.0009). UPLC-MS analysis revealed two major metabolic (Met) clusters. Met 1, composed of medium chain (MCFA), long-chain (LCFA) and very long-chain (VLCFA) fatty acid species, ceramides and lysophospholipids, was negatively associated with CRC (PFDR = 2.61 × 10−11); Met 2, composed of phosphatidylcholine species, nucleosides and amino acids, was strongly associated with CRC (PFDR = 1.30 × 10−12), but metabolite clusters were not associated with disease-free survival (p = 0.358). An association was identified between Met 1 and DNA mismatch-repair deficiency (p = 0.005). FBXW7 mutations were only found in cancers predominant in microbiota cluster 7. Networks of pathobionts in the tumour mucosal niche are associated with tumour mutation and metabolic subtypes and predict favourable outcome following CRC resection.

Understanding the impact of long-term physiological and environmental stress on the human microbiota and metabolome may be important for the success of space flight. This work is logistically difficult and has a limited number of available participants. Terrestrial analogies present important opportunities to understand changes in the microbiota and metabolome and how this may impact participant health and fitness. Here, we present work from one such analogy: the Transarctic Winter Traverse expedition, which we believe is the first assessment of the microbiota and metabolome from different bodily locations during prolonged environmental and physiological stress. Bacterial load and diversity were significantly higher during the expedition when compared with baseline levels (p < 0.001) in saliva but not stool, and only a single operational taxonomic unit assigned to the Ruminococcaceae family shows significantly altered levels in stool (p < 0.001). Metabolite fingerprints show the maintenance of individual differences across saliva, stool, and plasma samples when analysed using flow infusion electrospray mass spectrometry and Fourier transform infrared spectroscopy. Significant activity-associated changes in terms of both bacterial diversity and load are seen in saliva but not in stool, and participant differences in metabolite fingerprints persist across all three sample types.

For almost a century, it has been accepted that human milk contains viable microbial cells. However, for a considerable amount of this period, it was believed that they were the result of exogenous contamination, primarily from the skin or non-sterile handling. Early work using culture-dependent methods, supported by molecular profiling, however, identified the presence of lactic acid bacteria from an endogenous origin. This provided evidence that the human milk microbiota consisted of microorganisms that were not found solely on the skin surface and therefore could not result from contamination. Through the advent of next-generation sequencing, the field of microbiota research has caused a paradigm shift away from a typical focus on the presence of pathogenic microorganisms in human milk. This had led to a broad appreciation that the human milk microbiota consists of several hundred species of non-pathogenic commensal microbes – with many anaerobic microbial taxons being found only in the gastrointestinal tract outside of human milk. Nevertheless, as our appreciation of the complexity and diversity of the human milk microbiota has improved, many questions relating to the functional basis of host–microbiota interactions in the newborn infant’s gastrointestinal tract remain outstanding. To address these, mechanistic studies will be required in which the utilisation of isolated microorganisms will be essential. As such, a return to culture-dependent methods in the new paradigm of culturomics will be required. In this review, we bring together the current understanding of the human milk microbiota and how culturomics could play a fundamental role in furthering our understanding.

Given the long-term advantages of exclusive breastfeeding to infants and their mothers, there is both an individual and public health benefit to its promotion and support. Data on the composition of human milk over the course of a full period of lactation for a single nursling is sparse, but data on human milk composition during tandem feeding (feeding children of different ages from different pregnancies) is almost entirely absent. This leaves an important knowledge gap that potentially endangers the ability of parents to make a fully informed choice on infant feeding. We compared the metataxonomic and metabolite fingerprints of human milk samples from 15 tandem feeding dyads to that collected from ten exclusively breastfeeding single nursling dyads where the nursling is under six months of age. Uniquely, our cohort also included three tandem feeding nursling dyads where each child showed a preferential side for feeding—allowing a direct comparison between human milk compositions for different aged nurslings. Across our analysis of volume, total fat, estimation of total microbial load, metabolite fingerprinting, and metataxonomics, we showed no statistically significant differences between tandem feeding and single nursling dyads. This included comparisons of preferential side nurslings of different ages. Together, our findings support the practice of tandem feeding of nurslings, even when feeding an infant under six months.

The pregnancy vaginal microbiome contributes to risk of preterm birth, the primary cause of death in children under 5 years of age. Here we describe direct on-swab metabolic profiling by Desorption Electrospray Ionization Mass Spectrometry (DESI-MS) for sample preparation-free characterisation of the cervicovaginal metabolome in two independent pregnancy cohorts (VMET, n = 160; 455 swabs; VMET II, n = 205; 573 swabs). By integrating metataxonomics and immune profiling data from matched samples, we show that specific metabolome signatures can be used to robustly predict simultaneously both the composition of the vaginal microbiome and host inflammatory status. In these patients, vaginal microbiota instability and innate immune activation, as predicted using DESI-MS, associated with preterm birth, including in women receiving cervical cerclage for preterm birth prevention. These findings highlight direct on-swab metabolic profiling by DESI-MS as an innovative approach for preterm birth risk stratification through rapid assessment of vaginal microbiota-host dynamics.

Of the many metabolites involved in any clinical condition, only a narrow range of biomarkers is currently being used in the clinical setting. A key to personalized medicine would be to extend this range. Metabolic fingerprinting provides a more comprehensive insight, but many methods used for metabolomics analysis are too complex and time-consuming to be diagnostically useful. Here, a rapid evaporative ionization mass spectrometry (REIMS) system for direct ex vivo real-time analysis of biofluids with minor sample pretreatment is detailed. The REIMS can be linked to various laser wavelength systems (such as optical parametric oscillator or CO2 laser) and with automation for high-throughput analysis. Laser-induced sample evaporation occurs within seconds through radiative heating with the plume guided to the MS instrument. The presented procedure includes (i) laser setup with automation, (ii) analysis of biofluids (blood/urine/stool/saliva/sputum/breast milk) and (iii) data analysis. We provide the optimal settings for biofluid analysis and quality control, enabling sensitive, precise and robust analysis. Using the automated setup, 96 samples can be analyzed in ~35–40 min per ionization mode, with no intervention required. Metabolic fingerprints are made up of 2,000–4,000 features, for which relative quantification can be achieved at high repeatability when total ion current normalization is applied. With saliva and feces as example matrices, >70% of features had a coefficient of variance ≤30%. However, to achieve acceptable long-term reproducibility, additional normalizations by, e.g., LOESS are recommended, especially for positive ionization. This high-throughput protocol for direct ex vivo real-time metabolic fingerprinting of biofluids uses a laser system and an automated sampling platform. It includes REIMS procedures for analyzing blood, urine, stool, saliva, sputum and breast milk.

Abstract Early stage diagnosis of colorectal cancer (CRC) is a key predictor of patient survival, highlighting the need for accurate diagnostic biomarkers. Laser Assisted - Rapid Evaporative Ionization Mass Spectrometry (LA-REIMS) demonstrates potential for direct from sample metabolic profiling of feces. Here, we present an optimized LA-REIMS approach for fecal metabolomic analysis in CRC for the first time. A prospective, observational cohort biomarker discovery study was performed at an NHS hospital trust in the UK. Patients referred through the national two week wait cancer pathway were prospectively recruited. The primary outcome was to differentiate between adenomas or CRC and non-disease controls, the secondary outcome was to compare LA-REIMS to fecal immunochemical testing (FIT) against the gold-standard of colonoscopy or CT colonography. Those under the age of 18, pregnant, or unable to undergo colonic investigation were excluded. Fecal samples were analyzed using FIT (Kyowa Medex Co Ltd, Japan), with a detection limit of 7µg haemoglobin/g feces considered a positive result, as well as a LA-REIMS setup (Xevo G2-S QTof (Waters Corporation)), with optimized laser and instrumentation parameters (improved signal-to-noise and reduced carry-over between samples). After pre-processing, univariate and multivariate statistics were carried out in MetaboAnalyst 4.0. Random forest classification (RFC) with leave one out cross-validation (LOOCV) was performed.244 subjects (120 females, mean age 67.5 (range 21-93)) were included, of which n=135 were non-disease controls, n=82 adenoma patients, and n=27 CRC patients. Partial least square-discriminant analysis (PLS-DA) demonstrated a distinction between control vs CRC (Q2= 0.60087, R2= 0.89376). RFC with LOOCV for cancer vs control showed a sensitivity of 34.6% and specificity of 99.2% compared to 81.8% sensitivity and 87.2% specificity with FIT. PLS-DA of control vs adenoma (Q2= 0.64284, R2= 0.8898) demonstrated good separation. RFC with LOOCV for adenoma vs control had a sensitivity of 61.0 % and specificity of 94.7% compared to 33.8 % sensitivity and 87.2% specificity with FIT. Univariate analysis identified a feature tentatively assigned to the glycerolipid class (area under the receiver operating characteristic curve (AUROC) 0.986 (CI=0.957-1) to be more abundant in CRC vs control (false discovery rate (FDR) corrected p<0.001). A tentatively assigned organic acid (AUC 0.916, CI=0.868-0.953) was associated with control (FDR corrected p<0.001). Multiple features in the mass-to-charge ratio (m/z) range 200-400 were found to be significantly different between control vs adenoma. Annotations of the features are underway using liquid chromatography - MS/MS and spiked standards. LA-REIMS allows direct from sample metabolite profiling of feces and has applications as a novel screening tool for the early detection of CRC. Citation Format: Petra Paizs, Monika Widlak, Alvaro Perdones-Montero, Maria Sani, Lauren Ford, James L. Alexander, Simon Cameron, Ramesh Arasaradnam, James M. Kinross, Zoltan Takats. High-throughput fecal metabolic profiling for the early detection of colorectal cancer using a direct mass spectrometry assay [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 271.

Mass spectrometry has established itself as a powerful tool in the chemical, biological, medical, environmental, and agricultural fields. However, experimental approaches and potential application areas have been limited by a traditional reliance on sample preparation, extraction, and chromatographic separation. Ambient ionization mass spectrometry methods have addressed this challenge but are still somewhat restricted in requirements for sample manipulation to make it suitable for analysis. These limitations are particularly restrictive in view of the move toward high-throughput and automated analytical workflows. To address this, we present what we consider to be the first automated sample-preparation-free mass spectrometry platform utilizing a carbon dioxide (CO2) laser for sample thermal desorption linked to the rapid evaporative ionization mass spectrometry (LA-REIMS) methodology. We show that the pulsatile operation of the CO2 laser is the primary factor in achieving high signal-to-noise ratios. We further show that the LA-REIMS automated platform is suited to the analysis of three diverse biological materials within different application areas. First, clinical microbiology isolates were classified to species level with an accuracy of 97.2%, the highest accuracy reported in current literature. Second, fecal samples from a type 2 diabetes mellitus cohort were analyzed with LA-REIMS, which allowed tentative identification of biomarkers which are potentially associated with disease pathogenesis and a disease classification accuracy of 94%. Finally, we showed the ability of the LA-REIMS system to detect instances of adulteration of cooking oil and determine the geographical area of production of three protected olive oil products with 100% classification accuracy.

Two major outstanding questions in microbiome research ask what microbes are present in a community and how they interact with each other and their hosts. Recent, rapid improvements in nucleic acid (DNA and RNA) sequencing allow us to study the composition and function of microbiomes in unprecedented detail, leading to a step change in our understanding of host–microbe interactions. This chapter gives a broad overview of the basic toolkit available to modern microbiologists and microbial ecologists, exploring their application to key questions about microbiome structure and function. We cover tools based on nucleic acid sequencing (e.g. amplicon sequencing, metagenomics, metatranscriptomics) as well as approaches targeting larger molecules such as metabolomics and proteomics. We discuss the use of microbial culture as a means of measuring functional capacity of individual microbes, or building artificial communities to understand emergent properties of consortia. We emphasise the advantages of combining multiple techniques alongside robust experimental design to garner powerful quantitative estimates of microbiome structure, and how this relates to host–microbe interactions.

Background: The introduction of high-risk human papillomavirus (hrHPV) testing as part of primary cervical screening is anticipated to improve sensitivity, but also the number of women who will screen positive. Reflex cytology is the preferred triage test in most settings but has limitations including moderate diagnostic accuracy, lack of automation, inter-observer variability and the need for clinician-collected sample. Novel, objective and cost-effective approaches are needed. Methods: In this study, we assessed the potential use of an automated metabolomic robotic platform, employing the principle of laser-assisted Rapid Evaporative Ionisation Mass Spectrometry (LA-REIMS) in cervical cancer screening. Findings: In a population of 130 women, LA-REIMS achieved 94% sensitivity and 83% specificity (AUC: 91.6%) in distinguishing women testing positive (n = 65) or negative (n = 65) for hrHPV. We performed further analysis according to disease severity with LA-REIMS achieving sensitivity and specificity of 91% and 73% respectively (AUC: 86.7%) in discriminating normal from high-grade pre-invasive disease. Interpretation: This automated high-throughput technology holds promise as a low-cost and rapid test for cervical cancer screening and triage. The use of platforms like LA-REIMS has the potential to further improve the accuracy and efficiency of the current national screening programme. Funding: Work was funded by the MRC Imperial Confidence in Concept Scheme, Imperial College Healthcare Charity, British Society for Colposcopy and Cervical Pathology, National Research Development and Innovation Office of Hungary, Waters corporation and NIHR BRC.

Sparse data exist regarding the normal range of composition of maternal milk beyond the first postnatal weeks. This single timepoint, observational study in collaboration with the ‘Parenting Science Gang’ citizen science group evaluated the metabolite and bacterial composition of human milk from 62 participants (infants aged 3–48 months), nearly 3 years longer than previous studies. We utilised rapid evaporative ionisation mass spectrometry (REIMS) for metabolic fingerprinting and 16S rRNA gene metataxonomics for microbiome composition analysis. Milk expression volumes were significantly lower beyond 24 months of lactation, but there were no corresponding changes in bacterial load, composition, or whole-scale metabolomic fingerprint. Some individual metabolite features (~14%) showed altered abundances in nursling age groups above 24 months. Neither milk expression method nor nursling sex affected metabolite and metataxonomic fingerprints. Self-reported lifestyle factors, including diet and physical traits, had minimal impact on metabolite and metataxonomic fingerprints. Our findings suggest remarkable consistency in human milk composition over natural-term lactation. The results add to previous studies suggesting that milk donation can continue up to 24 months postnatally. Future longitudinal studies will confirm the inter-individual and temporal nature of compositional variations and the use of donor milk as a personalised therapeutic.