The Collection,Storage,and Processing of Human Milk Samples as a Source of mRNA: A State of the Science Review

Abstract

Background:

Human milk contains messenger ribonucleic acid (mRNA), a key player in protein production and a source of gene expression information for understanding lactation physiology. The mRNA is renowned for its fragility and exists in several milk fractions, creating practical challenges for mRNA isolation and analysis. While research teams have developed protocols for the use of human milk samples in transcriptomic applications, it is unclear whether there are best practices in sample collection, storage, and processing procedures for the collection of the highest quality mRNA.

Research Aim:

We aim to review current practices in the collection, storage, and processing of human milk relevant for its use in transcriptomic applications (using mRNA) in the last 10 years.

Method:

The PubMed database was systematically searched for publications addressing methodological considerations or the use of mRNA derived from human milk samples.

Results:

Most sources described the collection of human milk from small cohorts of term mothers in a mature lactational stage. While fresh collection of human milk was common, some was frozen or preserved. Several centrifugation protocols exist, and considerations for collection, storage, and processing may vary by transcriptomic application. There is inconsistency in the description and reporting of mRNA quality used in analyses that obfuscates the determination of best practices for best mRNA quality.

Conclusions:

Research into the effects of human milk collection, processing, and storage on mRNA quality metrics is warranted.

Keywords

breastfeeding lactation human milk methodology mRNA qPCR RNA sequencing RNA-seq ScRNA-seq transcriptomics

Background

Human milk feeding is recommended by the American Academy of Pediatrics and the World Health Organization as the ideal primary source of nutrition for infants exclusively until at least 6 months of age (Meek & Noble, 2022). Continued human milk feeding is recommended for 2 years or beyond and is strongly associated with numerous beneficial maternal and infant health outcomes (Meek & Noble, 2022). Human milk is well recognized as a complex, nutritive substance that contains a variety of components including carbohydrates, protein, fat, vitamins, minerals, enzymes, and hormones. Additionally, it may be considered a living tissue that is rich in mammary and immune cells and contains a multitude of bioactive molecules critical to developmental programming (Hassiotou, Geddes, et al., 2013; Tingö et al., 2021). Decades of intensive study have revealed that human milk contains biological markers that provide information about the health of mothers and infants, as well as keys to understanding processes by which human milk feeding is able to support health (Dastmalchi et al., 2022; Hassiotou, Geddes, et al., 2013). The transcriptome, or entire collection of gene reads present in a cell or source, has been recognized as a valuable resource in the field of lactation physiology (The National Institutes of Health, 2020). During the last 2 decades, study and sequencing of expressed transcripts within human milk, human milk transcriptome sequencing, has grown in popularity for its ability to provide insights into gene expression and inferred function (Lemay, Ballard, et al., 2013; Maningat et al., 2009; Twigger et al., 2015).

Research seeking to understand milk cellular function may use a methodology designed to sequence ribonucleic acid (RNA), specifically messenger RNA (mRNA), that act as a blueprint for cell protein production (Lemay, Ballard, et al., 2013; Schroeder et al., 2006). These mRNA are often referred to as coding RNAs as many mRNA code for proteins. Other types of non-coding RNA exist in milk that act as cellular regulators, like microRNAs (miRNA), long non-coding RNAs (lncRNA), and circular RNAs (circRNA; Tingö et al., 2021). Human milk may contain RNA in several places, including the cellular components of the milk, skim (or whey), lipid (or fat/milk fat globules) and extracellular vesicle components (including exosomal), and therefore research employing human milk transcriptome sequencing methods may vary (Alsaweed et al., 2015; Maningat et al., 2009; Raymond et al., 2022).

Practical considerations may limit the feasibility of human milk transcriptomic research in population-based cohorts. Unlike deoxyribonucleic acid (DNA), RNA is renowned as fragile, and its components are susceptible to degrading enzymes, like RNases (Schroeder et al., 2006). The gold standard of RNA quality assessment is calculation of the RNA integrity number (RIN), which is performed using an electrophoresis tool like the Agilent Bioanalyzer and results in a numerical value from 1 (degraded) to 10 (intact; Schroeder et al., 2006). Previous research suggests that RNA in fresh (unfrozen) human milk degrades markedly as the distance from the collection site to the laboratory increases, with sites up to 25 miles away often providing RINs considered unsuitable for whole transcriptome sequencing (Lemay, Hovey, et al., 2013). This suggests that research teams conducting studies in animals or in laboratory-adjacent settings may be able to process milk samples individually or quickly, while collection from early postpartum (outpatient, potentially location-bound) participants could present challenges in sample batching and transport. For many tissue types, snap-freezing to -80 °C may be employed to protect against cellular degradation during transport to the laboratory or long-term storage. While the snap-freezing of human milk at the time of collection has been performed in previously conducted studies using human milk transcriptome sequencing, it is currently unclear whether snap-freezing effectively stabilizes the sample or compromises its integrity (Barboza et al., 2012).

Heterogeneity in approaches to human milk transcriptome sequencing pose significant challenges to the field of lactation science. The presence and use of a multitude of methodologies related to the collection, storage, and processing of samples may obfuscate the appraisal and consumption of published research. Furthermore, it presents a barrier to scientific teams attempting to plan and reliably execute their own human milk studies. Additional challenges likely include variable description or disclosure of study protocols and procedures. Here, we sought to review current practices in the collection, storage, and processing of human milk samples relevant to transcriptome sequencing of coding RNA (mRNA) in the last 10 years.

Methods

Design

This research is a state of the science review with a focus on procedures used in the collection, storage, and processing of human milk samples for the study of gene expression or transcriptomic applications. We present a synthesis of the most recent evidence guiding current procedures. A formal review protocol is available in the online Supplementary Materials (S1). No IRB review was necessary for this review of the literature.

Sample: Defining the Articles Reviewed

Inclusion criteria specified that sources must either evaluate sample collection, storage, or processing methodology relevant to (mRNA) human milk transcriptome sequencing or use the methodology in pursuit of a primary study aim. Due to an overwhelming surplus of sources to describe human milk RNA sequencing (including rapidly growing interest in miRNA), this review focuses specifically on the sequencing of coding RNA (mRNA). For non-coding RNA (miRNA, lncRNA, etc.), a high-quality systematic review exists to meet this need (Tingö et al., 2021). As an additional inclusionary note, although the primary focus of this review was transcriptomic (global gene expression), we included sources using RT-qPCR techniques (targeted expression) for their potential to reveal critical insights into how collection, processing, or storage of samples may affect mRNA quantity or quality in human milk. Non-primary sources (i.e., reviews), sources without peer-review, those examining primarily viral RNA (non-host) or non-coding RNA, those examining RNA from exosomes (often serving as packages for miRNA), and animal studies were excluded from this synthesis. Sources that included both animal and human data were included if all other inclusion criteria were met. Lastly, those published in a language other than English were not reviewed by our English-speaking study team. Sources were first screened against inclusion and exclusion criteria independently by two reviewers. The final sample size for this state of the science review includes 24 sources (Figure 1; Page et al., 2021).

Figure 1.

Search and Screen for Included Sources.

Data Collection: The Search Strategy and Process

The authors initially met with a senior academic librarian at the University of Wisconsin – Madison to collaboratively develop a rigorous search strategy. Articles published through September 12, 2024 were considered for review, with date and English filters applied to return sources published during the last 10 years (2013–2023, with extension to 2024). Primary sources either addressing sample collection, processing, or storage for (mRNA) human milk transcriptome sequencing, or those using the methodology in pursuit of a primary research aim, were identified in the PubMed, Web of Science, Embase, and Cochrane databases, and considered potentially relevant. The full search strategies are available in the online Supplementary Materials (S2). Following the search, all results were uploaded to the Covidence platform for two-author review (Veritas Health Innovation, n.d.).

Measurement

We identified and collected relevant variables from the sources that include (1) source demographic information, (2) first author and year of publication, (3) the number of included participants and whether they were term or preterm populations, (4) the date of antenatal or postpartum sample collection, (5) the experimental method (i.e., qPCR, RNA-seq), (6) sample storage conditions provided, (7) which milk fraction was used (i.e. cells, milk fat globule), (8) time of day the sample was collected, (9) the time, temperature and speed (x g or rpm) of initial centrifugation, (10) the quantity of human milk collected, (11) which isolation kit (if any) was used, and (12) descriptions of RNA concentration, purity, or integrity assessment. Two authors participated in data extraction from sources that met full-text inclusion criteria using a piloted form and table of variables.

Data Analysis

Data was synthesized quantitatively through categorization and comparison for a total of 24 sources (N = 24) to provide a summary and analysis of source demographics and variables. Summaries of extracted data are available, with demographics listed in Table 1 and measured variables in Table 2. Further synthesis was conducted within sub-groups according to experimental method (i.e., qPCR, RNA-seq, scRNA-seq) when relevant.

Table 1.

Demographic Information for Included Sources.

First Author, Publication Date	Country of Origin, Institution	Journal	Study Aim, Question or Purpose	Sample (N)	Research Design
Adams et al. (2022)	The University of Cincinnati, USA	The Journal of Nutrition	Characterization of mammary gland transcriptomes in DHA-treated and non-treated mothers.	N = 14	Quantitative Intervention Study
Alsaweed et al. (2015)	The University of Western Australia, Australia	Journal of Cellular Biochemistry	Determination of total RNA and microRNA content between skim, lipid, and cellular fractions of human milk.	N = 29	Methodological Investigation
Blans et al. (2017)	Aarhus University, Denmark	Journal of Extracellular Vesicles	Description of an approach to isolate the extracellular vesicles in milk.	N = 7	Methodological Investigation
Cai et al. (2017)	University of Manitoba, Canada	Journal of Pediatric Gastroenterology and Nutrition	Determination of iron expression transporters responsible for iron excretion in human milk.	N = 9	Quantitative Observational Study
Canale et al. (2021)	University of California, USA	Heliyon	Investigation of maternal mental health status (anxiety/depression) and human milk characteristics (gene expression/bacterial composition).	N = 8	Quantitative Cohort Study
De Andrés et al. (2018)	ProbiSearch	Frontiers in Microbiology	Exploration of the feasibility of microarray and RT-qPCR techniques to search for specific molecules as new targets responsive to the oral intake of a probiotic in the human mammary gland affected by mastitis.	N = 31	Quantitative Intervention Study
Gil-Kulik et al. (2023)	Medical University of Lublin, Poland	International Journal of Molecular Sciences	Evaluation of perinatal factors on gene expression in human milk stem cells.	N = 42	Quantitative Observational Study
Gleeson et al. (2022)	Carnegie Mellon University, USA	Science Advances	Characterization of cells in mature milk from healthy human donors.	N = 10	Quantitative Observational Study
Kothapalli et al. (2018)	Cornell University, USA	Prostaglandins, Leukotrienes and Essential Fatty Acids	Characterization of the second alternative transcript for fatty acid desaturase 2.	N = 1	Quantitative Observational Study
Kubota et al. (2014)	Chiba University, Japan	British Journal of Nutrition	Assessment of mRNA expression in human milk cells to identify changes in the concentration of cytokines in milk after the consumption of fructooligosaccharides.	N = 64	Quantitative Intervention Study
LeMaster et al. (2023)	Children’s Mercy Research Institute, USA	Communications Biology	Characterization of 62 analytes of the soluble and cellular components of human milk in early lactation.	N = 10	Quantitative Observational Study
Lemay, Hovey, et al. (2013)	University of California Davis, USA	BMC Genomics	Assessment of lactating rhesus macaques as a model for lactating humans and determination of whether RNA from milk fractions was representative of RNA extracted from mammary tissues.	N = 5	Methodological Investigation
Lemay, Ballard, et al. (2013)	University of California Davis, USA	Plos One	Characterization of the three stages of lactation: colostral, transitional and mature milk production using RNA-sequencing of human milk fat globules.	N = 12	Quantitative Observational Study
Carli et al. (2020)	University of Colorado Anschutz, USA	Journal of Mammary Gland Biology and Neoplasia	Description of methods for the collection and cryopreservation of human milk-derived mammary epithelial cells and bioinformatic code for processing scRNA-seq data.	N = 2	Methodological Investigation
Mohammad and Haymond (2013)	Children’s Nutrition Research Center, USA	American Journal of Physiology Endocrinology and Metabolism	Study of mRNA expression of lipid metabolic enzymes in human mammary epithelial cells in conjunction with the measurement of milk fatty acid composition during secretory activation.	N = 7	Quantitative Observational Study
Mohammad et al. (2014)	Children’s Nutrition Research Center, USA	American Journal of Physiology Endocrinology and Metabolism	To test whether incorporation of 13C carbons from glucose into fatty acids and triglycerides was greater 1) in milk compared to plasma, 2) during a high-carbohydrate diet compared to high-fat diet, and 3) during feeding compared to fasting.	N = 7	Quantitative Intervention Study
Nyquist et al. (2022)	Broad Institute of MIT and Harvard, USA	Proceedings of the National Academy of Sciences of the United States of America	Characterization of human milk-derived cells using scRNA-seq to better understand cellular dynamics and longitudinal lactational heterogeneity.	N = 15	Quantitative Observational Study
Sharp et al. (2016)	Institute for Frontier Materials, Australia	Functional and Integrative Genomics	Demonstration of the utility of RNA obtained directly from human milk cells to detect mammary epithelial cell-specific gene expression.	N = 10	Quantitative Observational Study
Slupecka-Ziemilska (2017)	The Kielanowski Institute of Animal Physiology and Nutrition, Poland	Journal of Physiology and Pharmacology	Investigation of the influence of preterm delivery on ghrelin and obestatin concentrations (in serum and milk) and gene expression in mammary epithelial cells.	N = 40	Quantitative Cohort Study
Tian et al. (2022)	Jilin University, China	Nutrients	Exploration of the differences in polyunsaturated fatty acids among milk groups with different levels of FADS gene expression.	N = 50	Quantitative Observational Study
Twigger et al. (2015)	The University of Western Australia, Australia	Scientific Reports	Description of the variation in milk cell populations between participants. Exploration of dyad characteristics with gene expression.	N = 66	Quantitative Observational Study
Twigger et al. (2018)	Institute for Stem Cell Research, Germany	Nutrients	Examination of the presence and localization of immune proteins such as perforin, granulysin, and granzymes in human milk cells at the protein and mRNA level.	N = 24	Quantitative Observational Study
Twigger et al. (2022)	University of Cambridge, England	Nature Communications	Report of a single-cell transcriptomic analysis of 110,744 viable breast cells isolated from human milk or non-lactating tissue.	N = 9	Quantitative Observational Study
Zhao et al. (2020)	Kunming Institute of Zoology, China	Nature Communications	Investigation of RNA-binding proteins with the potential to control post-transcriptional expression of essential genes in lactation.	N = 60	Quantitative Observational Study

Note. Country of Origin refers to the first author’s country of origin. FAD = fatty acid desaturase.

Table 2.

Description of Included Sources.

Authors,PublicationYear	Participant Population Type(n, total N)	Time of Sample Collection	RNASequencing	Sample Storage Conditions	Milk Fraction	Time of Day	Centrifugation (Initial Spin Speed)	Quantity(ml)	IsolationKit	RNAIntegrityAssessment
Adams et al. (2022)	Preterm N = 14	Four weeks post-enrollment	RNA-seq	Fresh milk samples were centrifuged and dispersed in TRIzol. Processed immediately or stored at -80 °C	Milk fat globule	No control for time of day	1500 g x 10 min	NA	Promega Maxwell 16 Integrated System	8.4–9.4 RNA integrity, rRNA ratio between 1.3 and 2.6, RNA quantity > 1μ g.
Alsaweed et al. (2015)	NA, N = 29	3–158 weeks, longitudinal (collected 1–4 x per participant) N = 49 samples	None, examined total RNA count	Fresh milk samples immediately processed	Cells, milk fat globule, skim	No control for time of day	720 g x 20 min	14–135	Eight commercially available kits were used	Concentration and purity (260:280) were measured using a spectrophotometer.
Blans et al. (2017)	Term N = 7	Mid-lactation	Electrophoresis	Samples were stored < 24 hr at 4 °C until isolation	Milk extracellular vesicles	NA	1250g x 35 minutes	NA	miRCURY RNA Isolation Kit	Total RNA Profiles provided using Agilent Bioanalyzer.
Cai et al. (2017)	Term N = 9	5–6 months PP	RT-qPCR	Fresh milk was transported to the laboratory in a cooler surrounded by ice pads	Milk fat globule	NA	1600 g x 10 min	15	QIAgen RNeasy Mini Kit/TRIzol protocol hybrid	Quantity measured by Nanodrop Spectrophotometer. Quality measured by gel electrophoresis.
Canale et al. (2021)	Term, N = 8	Peak Lactation (30–55 days)	RNA-seq	Frozen, samples stored in liquid nitrogen and shipped to the lab. Kept at -80 °C until processing	Cells	NA	NA	30	Qiagen RNeasy Micro Kit	Integrity examined using Agilent 4200 TapeStation System.
De Andrés et al. (2018)	NA, N = 31	Mature stage milk	Microarray, RT-qPCR	Fresh milk	Cells	NA	620 g x 20 min	10	AllPrep DNA/RNA/Protein Mini Kit	RNA concentration and purity checked with Nanodrop and quality with Agilent 2100 bioanalyzer. Samples used if (260:280) between 1.8–2.1 and RIN > 6.
Gil-Kulik et al. (2023)	Term and preterm, N = 42	2–136^th day PP, mean = 9.5 days	qPCR	NA	Cells	Morning	805 g x 20 min at 20 ºC	5	Protocol using TRI Reagent, isopropanol, chloroform, and ethyl alcohol	RNA extract was spectrophotometrically evaluated. 1 µg of RNA was used for analyses.
Gleeson et al. (2022)	NA, N = 6–7 for qRT-PCR NA, N = 3 for ScRNA-seq	Mature stage milk (4 weeks or more, mean 32 weeks PP)	RT-qPCR, scRNA-seq	Fresh milk was pumped, placed on ice and immediately transferred.	Cells	NA	805 g x 20 min at 20 °C	NA	RNeasy mini kit	1500 ng of RNA used in RT-qPCR. Flow cytometry for cell viability. (average 70%).
Kothapalli et al. (2018)	NA, N = 1	NA	RT-PCR assay	Sample was anonymous and collected from a human milk bank	Milk fat globule	NA	835 g x 10 min at 4 °C	NA	QIAshredder and RNeasy Kit	Concentration measured by Qubit 2.0 fluorometer and 63 µg was used for sequencing library preparation after testing for RIN using an Agilent Bioanalyzer.
Kubota et al. (2014)	NA, N = 64	Colostrum during hospitalization, 1 month PP	Microarray	Collected in sterile tubes and processed within 24 hr. Spun first, then frozen	Cells	NA	Whey/Cellular Elements Separation: 2500 rpm x 5 min at room temperature Fat/Whey Separation: 10,000 rpm x10 min at 4 °C	NA	RNeasy mini kit Samples with low RNA yields were concentrated with RNeasy micro kit	Quantity and quality of RNA were determined by BioSpec-nano and the Agilent 2100 BioAnalyzer. High quality RNA were used in analyses. OD (260:280) > 1.8.
LeMaster et al. (2023)	NA, N = 36, Term and preterm, N = 10 for scRNA-seq	2–7 days PP, 8–16 days PP	scRNA-seq	Freshly collected, processed within 1 hr if at room temperature and 4 hr if kept on ice	Cells	Daytime	800 g x 15 min at 4 °C	3–10	NA	Cells counted using an Automated Cell Counter. Cells with > 25% mitochondrial expression were filtered out.
Lemay, Hovey, et al. (2013)	Term, N = 5	Samples collected two times per participant at 90 days PP and 180 days PP	Electrophoresis	Fresh samples were used, immediately transported to the laboratory on ice	Cells, milk fat globule	Mid-day (11 am–1 pm)	2311 g x 10 min at 4 °C	NA	Qiagen Universal Kit	RNA yield and quality were measured by Nanodrop spectrophotometry and integrity by Agilent Bioanalyzer. Mean RIN values are provided.
Lemay, Ballard, et al. (2013)	NA, N = 12	2–130 days of lactation Colostrum (41–52 hr), Transitional (39–56 hr), Mature (33–130 days)	RNA-seq, qPCR	Fresh, immediate processing	Milk fat globule	NA	Tested hard and soft spin protocols. Hard spin yielded preferable results 15,000g at 4 °C	Five early, 10 mature	Promega Maxwell 16 Integrated System	Concentration, 28S:18S ratio and RIN Tested using Agilent 2100 bioanalyzer. RIN > 7.0 RNA > 9– 10 ng/µl.
Carli et al. (2020)	Term, N = 5 N = 2 samples used for scRNA-seq	Two weeks PP	scRNA-seq	Fresh milk was collected, placed in a cooler on ice and immediately transported to the clinic	Cells	Morning	500 g x 10 min	< 30	NA	Flow cytometric analysis for cell viability.
Mohammad and Haymond (2013)	Term, N = 7	Six hr PP, then q12 hr for 4 days and weekly for 6 weeks, 14 times per participant	Microarray	Samples combined with 20 µl RNAse inhibitor, spun and frozen at -80 °C. TRIzol added to the fat layer	Milk fat globule	Serial, following Day 7, samples collected at 10 am	3000 rpm x 15 min at 4 °C	3–5 during the first 4 days, 10 afterward.	NA	Total RNA concentration was measured using NanoDrop spectrophotometer. RNA quality assessed using Experion TMRNA StdSens Analysis Kit.
Mohammad et al. (2014)	NA, N = 7	10 +/- 2 weeks	Microarray	Samples combined with 20 µl RNAse inhibitor, spun and frozen at -80 °C. TRIzol added to the fat layer	Milk fat globule	9 am, 4 pm	NA	15	NA	Total RNA concentration was measured using NanoDrop spectrophotometer. RNA quality assessed using Experion TMRNA StdSens Analysis Kit.
Nyquist et al. (2022)	Term, N = 15	Variable 3–632 days PP, Colostrum (3–6 days), Transitional (10–14 days), Mature (15–18 days) and Late PP (5–90 weeks)	scRNA-seq	Fresh milk samples were iced and processed within 6 hr. Methods for collecting cells, including holding 4 °C overnight, and a single freeze/thaw of whole milk were tested	Cells	Mornings 6 am–9 am	350 g x 10 min	0.5 ml minimum, sample volumes provided in supplementary materials	NA	Fresh processed cells, sorted cells, or live-enriched cells clustered together on principal component space, suggesting little gain by additional processing prior to ScRNA-seq.
Sharp et al. (2016)	NA, N = 10 total, Colostrum N = 2, Midlactation N = 5, Mastitis N = 1, Involution N = 2	Variable: Colostrum (24 hr prepartum), midlactation (24, 48, 101 days), mastitis (Day 23) and involution (Day 7, 14 days)	RT-qPCR	Fresh (defined as < 3 hr) or Frozen (-80 °C)	Cells	NA	2000 g x 5 minutes at 4 °C	NA	Qiagen RNeasy Micro Kit	Cell staining to determine live/dead status.
Slupecka-Ziemilska (2017)	Term and preterm, N = 40	PP day 3	RT-PCR	Fresh	Cells	Between 10:00–11:00, 2 hr after eating a meal	1500 g x 15 min at 4 °C	2	NucleoSpin RNA XS Kit	Light microscope used for viability/cell count. RNA integrity verified by electrophoresis. 10 ng of RNA used. OD 230, 260 and 280 by Nanodrop.
Tian et al. (2022)	Term, N = 50	42–45 days PP, mature milk samples	RT-qPCR	Samples preserved in a 4 °C icebox and transferred to –80 before processing (for maximum 2 weeks). Samples thawed at room temperature prior to centrifugation	Cells	9–11 am	15 min at 4 °C	10	The TRIzol method	1µl of RNA suspension on nucleic acid quantifier to determine concentration and purity. The 1% agarose gel electrophoresis technique was used to detect the integrity of total RNA extracted.
Twigger et al. (2015)	Term and preterm, N = 66	PP	RT-PCR	Samples expressed, transported to the laboratory immediately and shielded from light	Cells	NA	800 g x 20 min at 20 °C	5–240	RNEasy extraction kit	Nanodrop 1000 used for quantity and quality (260:280) 1.8–2.2 Total cell content and viability Neubauer haemocytometer by Trypan Blue exclusion.
Twigger et al. (2018)	NA, N = 36 N = 24 for QRT-PCR, Term	Five days before delivery (N = 1), 4–142 days PP	qRT-PCR	Samples brought directly to the laboratory and processed. Cells were stored at -80 °C until extraction	Cells	NA	800 g x 20 min at 20 °C. Cell pellet washed 1 x and resuspended	55	Mini RNeasy extraction kit	All RNA was high quality, concentration and purity checked with a 260/280 ratio between 1.8 and 2.2 Total cell count and viability information in table.
Twigger et al. (2022)	N = 9	2–12 months of lactation	ScRNA-seq	Fresh milk samples were immediately transported on ice to the laboratory and processed in < 2 hr. Diluted 1:1 with sterile PBS	Cells	NA	870 g x 20 min at 20 °C. Pellet washed in cold PBS twice and 100–550 µl mammary epithelial cell growth medium was added to the pellet, depending on size	38–102.5	NA	Cells loaded onto a Neubauer improved counting chamber and examined on an immunofluorescence microscope.
Zhao et al. (2020)	Term, N = 60	Days 3–5 PP	qRT-PCR	Fresh, processed within 3 hr	Milk fat globule	Early morning	1000 g x 10 min at 4 ºC	3–5	TRIzol	2% agarose gel electrophoresis.

Note. NA = information not listed or located, RT-qPCR = real time polymerase chain reaction (PCR), RNA-seq = RNA sequencing, ScRNA-seq = single cell RNA sequencing, PP = postpartum.

Results

Study Characteristics

Through our literature search, we identified a total of 24 journal articles meeting our inclusion criteria. Publications examining human milk mRNA or pursuing transcriptome sequencing remained relatively consistent throughout this period, with 11 studies published before 2018 and 13 published in 2018 or later (see Table 2). We found that the earliest published studies (prior to 2018) used techniques including basic RNA quantification (N = 1; Alsaweed et al., 2015) and electrophoresis (N = 2; Blans et al., 2017; Lemay, Hovey, et al., 2013), microarrays (N = 3; Kubota et al., 2014; Mohammad & Haymond, 2013; Mohammad et al., 2014), qPCR (N = 5; Cai et al., 2017; Lemay, Ballard, et al., 2013; Sharp et al., 2016; Slupecka-Ziemilska, 2017; Twigger et al., 2015), and RNA-seq (N = 1; Lemay, Ballard, et al., 2013). While qPCR (N = 7; De Andrés et al., 2018; Gil-Kulik et al., 2023; Gleeson et al., 2022; Kothapalli et al., 2018; Tian et al., 2022; Twigger et al., 2018; Zhao et al., 2020) and RNA-seq (N = 2; Adams et al., 2022; Canale et al., 2021) remain commonly used in later published work (2018 and later), recent years have brought a rise to the use of single-cell RNA sequencing technologies (N = 4; Carli et al., 2020; Gleeson et al., 2022; LeMaster et al., 2023; Nyquist et al., 2022; Twigger et al., 2022).

Human milk samples in the included studies appeared to be mainly sourced from small groups of postpartum mothers delivering term infants. Most studies (N = 10; Blans et al., 2017; Cai et al., 2017; Canale et al., 2021; Carli et al., 2020; Lemay, Hovey, et al., 2013; Mohammad & Haymond, 2013; Nyquist et al., 2022; Tian et al., 2022; Twigger et al., 2018; Zhao et al., 2020) sampled only mothers of term infants, although some included preterm infants (N = 1; Adams et al., 2022), combined populations of term/preterm (N = 4; Gil-Kulik et al., 2023; LeMaster et al., 2023; Slupecka-Ziemilska, 2017; Twigger et al., 2015), or did not directly specify. Most milk samples were provided by participants from the mature lactational stage (> 2 weeks postpartum, N = 10; Adams et al., 2022; Blans et al., 2017; Cai et al., 2017; Canale et al., 2021; Carli et al., 2020; De Andrés et al., 2018; Gleeson et al., 2022; Lemay, Ballard, et al., 2013; Tian et al., 2022; Twigger et al., 2022), with fewer from those producing colostrum (N = 2; Mohammad et al., 2014; Slupecka-Ziemilska, 2017), although many sample designs included collection across multiple lactational stages (Alsaweed et al., 2015; Gil-Kulik et al., 2023; Kubota et al., 2014; LeMaster et al., 2023; Lemay, Ballard, et al., 2013; Mohammad & Haymond, 2013; Nyquist et al., 2022; Sharp et al., 2016; Twigger et al., 2018). Several studies sampled participants throughout a wide range of dates (e.g., 4–142 days) and a variety were non-specific in their reporting or did not disclose the timing of sample collection using a method that allowed for perception of lactation timeline. The majority of sources featured cross-sectional study designs, yet several (N = 8) suggest that samples may have been collected longitudinally (Alsaweed et al., 2015; Kubota et al., 2014; LeMaster et al., 2023; Lemay, Ballard, et al., 2013; Lemay, Hovey, et al., 2013; Mohammad & Haymond, 2013; Sharp et al., 2016; Twigger et al., 2018). Small sample size design characterized most studies, with approximately half of studies (N = 12) reporting results based on 10 or fewer participants (Blans et al., 2017; Cai et al., 2017; Canale et al., 2021; Carli et al., 2020; Gleeson et al., 2022; Kothapalli et al., 2018; LeMaster et al., 2023; Lemay, Hovey, et al., 2013; Mohammad & Haymond, 2013; Mohammad et al., 2014; Sharp et al., 2016; Twigger et al., 2018). Designs with longitudinal data collection and using single-cell sequencing technologies (N = 5; Carli et al., 2020; Gleeson et al., 2022; LeMaster et al., 2023; Nyquist et al., 2022; Twigger et al., 2018) frequently belonged to the < 10 participant category.

Sample Properties

We found that during the last 10 years, investigators pursued research in the field of human milk mRNA transcriptomics using cells (N = 16; Alsaweed et al., 2015; Canale et al., 2021; Carli et al., 2020; De Andrés et al., 2018; Gil-Kulik et al., 2023; Gleeson et al., 2022; Kubota et al., 2014; LeMaster et al., 2023; Lemay, Hovey, et al., 2013; Nyquist et al., 2022; Sharp et al., 2016; Slupecka-Ziemilska, 2017; Tian et al., 2022; Twigger et al., 2015, 2018, 2022), the milk fat globule (N = 9) (Adams et al., 2022; Alsaweed et al., 2015; Cai et al., 2017; Kothapalli et al., 2018; Lemay, Ballard, et al., 2013; Lemay, Hovey, et al., 2013; Mohammad et al., 2014; Mohammad & Haymond, 2013; Zhao et al., 2020), skim (N = 1; Alsaweed et al., 2015) and milk extracellular vesicles (MEV; N = 1) (Blans et al., 2017). Research teams used a variety of fresh (N = 16; Adams et al., 2022; Alsaweed et al., 2015; Cai et al., 2017; Carli et al., 2020; De Andrés et al., 2018; Gleeson et al., 2022; Kubota et al., 2014; LeMaster et al., 2023; Lemay, Ballard, et al., 2013; Lemay, Hovey, et al., 2013; Sharp et al., 2016; Slupecka-Ziemilska, 2017; Twigger et al., 2015, 2018, 2022; Zhao et al., 2020), preserved (N = 5; Blans et al., 2017; Canale et al., 2021; Mohammad & Haymond, 2013; Mohammad et al., 2014; Nyquist et al., 2022; Tian et al., 2022), or combination (N = 1; Nyquist et al., 2022) sample preparations. Notably, one study sourced human milk from a human milk bank (N = 1; Kothapalli et al., 2018). Many did not specify expression method or the time of day that the sample was collected, but in those that did, morning collection was common (morning N = 5; Carli et al., 2020; Gil-Kulik et al., 2023; Nyquist et al., 2022; Slupecka-Ziemilska, 2017; Tian et al., 2022). Others practiced either mid-day (N = 1; Lemay, Hovey, et al., 2013) or serial collections (N = 2; Mohammad & Haymond, 2013; Mohammad et al., 2014). The sample volume collected ranged drastically from 0.5 ml to 240 ml (Nyquist et al., 2022; Twigger et al., 2015). While many studies collected sample volumes of 1 fluid ounce (30 ml) or less (N = 13; Cai et al., 2017; Canale et al., 2021; Carli et al., 2020; De Andrés et al., 2018; Gil-Kulik et al., 2023; LeMaster et al., 2023; Lemay, Ballard, et al., 2013; Mohammad & Haymond, 2013; Mohammad et al., 2014; Nyquist et al., 2022; Slupecka-Ziemilska, 2017; Tian et al., 2022; Zhao et al., 2020), several studies did not disclose the sample volumes of human milk used in analyses or sampled across a considerable range (e.g., 6–240 ml). Notably, 10–15ml of mature milk was common among those studies that isolated milk fat globules (Cai et al., 2017; Lemay, Ballard, et al., 2013; Mohammad & Haymond, 2013; Mohammad et al., 2014) while volumes for cellular isolation varied dramatically (0.5–240 ml). In those collecting colostrum, small volumes were ubiquitous (0.5 ml–5 ml; Gil-Kulik et al., 2023; Nyquist et al., 2022; Slupecka-Ziemilska, 2017).

Sample Processing and Storage

Human milk samples were routinely centrifuged for separation of milk components (cells, skim, fat) prior to RNA isolation and sequencing. A variety of sample centrifugation protocols were used, and the results suggest that the milk component targeted for isolation (i.e., cells, milk fat globule, extracellular vesicle) and planned experimental method (i.e., scRNA-seq) were important considerations in protocol selection. For cellular components, 620 g–15,000 g for 5–20 min (Alsaweed et al., 2015; De Andrés et al., 2018; Gil-Kulik et al., 2023; Gleeson et al., 2022; Lemay, Hovey, et al., 2013; Sharp et al., 2016; Slupecka-Ziemilska, 2017; Tian et al., 2022; Twigger et al., 2015, 2018) or 2500 rpm for 5-min (Kubota et al., 2014) protocols were used. Temperature settings varied (20 °C N = 3, 4 °C N = 4). Notably, those pursuing single-cell RNA sequencing used centrifugation protocols with lower force, 350–870 g for 10–20 min at 4 °C or 20 °C (Carli et al., 2020; Gleeson et al., 2022; LeMaster et al., 2023; Nyquist et al., 2022; Twigger et al., 2022). In studies isolating the milk fat globule, centrifugation speeds ranged from 720–2311 g for 10–20 min (Adams et al., 2022; Alsaweed et al., 2015; Cai et al., 2017; Kothapalli et al., 2018; Lemay, Ballard, et al., 2013; Lemay, Hovey, et al., 2013; Zhao et al., 2020) or 3,000 rpm for 10 min (Mohammad & Haymond, 2013; Mohammad et al., 2014) at 4 °C. The singular study examining the milk extracellular vesicles listed a centrifugation protocol of 1250 g x 35 min (Blans et al., 2017).

Later processing steps included RNA isolation, RNA quantification, and integrity testing. Results indicated that a variety of commercially available RNA isolation kits were used, but that Qiagen kits were most common (N = 9; Alsaweed et al., 2015; Cai et al., 2017; Canale et al., 2021; De Andrés et al., 2018; Kubota et al., 2014; Lemay, Hovey, et al., 2013; Sharp et al., 2016; Twigger et al., 2015, 2018). Of those examining the milk fat globule, approximately half (N = 3) opted for an extraction protocol including TRIzol or a similar reagent (Cai et al., 2017; De Andrés et al., 2018; Zhao et al., 2020). Few studies explicitly reported RNA integrity numbers (N = 4; Adams et al., 2022; De Andrés et al., 2018; Lemay, Ballard, et al., 2013; Lemay, Hovey, et al., 2013), quantities (N = 7; Adams et al., 2022; Gil-Kulik et al., 2023; Gleeson et al., 2022; Kothapalli et al., 2018; Lemay, Ballard, et al., 2013; Slupecka-Ziemilska, 2017; Tian et al., 2022), or purity (N = 5; Alsaweed et al., 2015; De Andrés et al., 2018; Kubota et al., 2014; Twigger et al., 2015, 2018). Ultimately, the total number of studies that did not provide descriptive or quantitative measures of RNA quantity or integrity beyond mention of method (Nanodrop, Agilent Bioanalyzer) used was high. For sources using scRNA-seq, several provided quality control metrics to describe cellular count and viability (Carli et al., 2020; Gleeson et al., 2022; LeMaster et al., 2023; Nyquist et al., 2022; Twigger et al., 2022).

Discussion

Following review and synthesis of 24 published sources describing recent methodological inquiries and applications of human milk mRNA-based experimental technique, we report significant heterogeneity in sample collection, processing, storage, and reporting. It is unclear whether the application of diverse study protocols reliably produces human milk mRNA suitable for sequencing. Based on our synthesis, we will discuss relevant considerations and present several opportunities for further development of rigorous methods in human milk mRNA handling, processing, and storage.

Sample Collection

Human milk is well known to be a dynamic tissue, and therefore samples collected from small cohorts, across lactational stages, and by differing expression methods, may result in heterogeneity. Collection of mature milk samples from small cohorts were the most common among the studies included in this review. While the high cost of RNA-sequencing services has depreciated considerably in recent years, (Hu et al., 2021) it is possible that project budget or feasibility constraints limited cohort sizes.

We can also speculate that parent availability and mobility, expressed volume, and RNA quality may favor inclusion of mature lactating populations of term-born infants in studies to reduce the challenges inherent in the collection of samples in populations producing colostrum and transitional milk. As with RNA from other tissue types, RNA in human milk samples decreases over time (Lemay, Hovey, et al., 2013). Colostrum collected antenatally and immediately postpartum may be feasible if lactating populations are inpatient or accessible in ways that facilitate transport of samples to a laboratory for processing. Notably, in healthcare systems designed to discharge maternal populations shortly after birth, samples collected after 2–4 days of life may be difficult to acquire as maternal populations move into residences that may not facilitate sample collection and transportation (i.e., longer transport times) necessary to preserve human milk RNA. Notably, the lactational stage itself may affect the quality of sample RNA. Data on RNA integrity suggest that RNA quality increases with progressive lactational stage, with 27% of colostrum samples, 60% of transitional samples and 78% of mature milk samples providing RNA suitable for RNA-seq (RIN > 7.0; Lemay, Ballard, et al., 2013). For this reason, studies seeking collection of colostrum or transitional samples may consider the minimization of other factors known to affect RNA degradation, such as transport time to the laboratory, for best results.

Whether the milk expression technique affects the integrity of human milk samples for transcriptome sequencing remains unclear. We found that descriptions of collection methods (i.e., hand expression, hand/mechanical breast pump), interval since last expression, and time of sample collection were often vaguely described or omitted. If research teams seek to isolate RNA from cellular or milk fat globule sources, time of day and expression practice may be relevant. Human milk cellular content has been observed to fluctuate with circadian rhythm and may be higher in hindmilk than foremilk (Alhussien & Dang, 2018; Hassiotou, Hepworth, Williams, et al., 2013). Nursing frequency may also affect a cellular sample with higher numbers of apoptotic cells potentially accumulating with longer between-expression intervals or individuals with higher breast capacities (Martin Carli et al., 2021). Because mRNA is quickly degraded, apoptotic cells may lack high-quality mRNA. Similarly, it is reasonable to speculate that RNA in the milk fat globule may be most accessible in milk with higher fat fractions, such as hindmilk (Hassiotou, Hepworth, Williams, et al., 2013). Whether inter-breast variation is a factor is also unclear, and therefore research teams may consider adherence to best practice protocols for general human milk collection whenever possible (McGuire et al., 2021).

Sample Quantity

Sample quantity is a critical consideration in the planning of human milk gene expression or transcriptomic studies. Research teams must collect the smallest amount of human milk necessary to ensure minimal burden to research participants while still collecting a sample that will reliably yield RNA appropriate for analyses. Although many studies did not describe the quantity of human milk collected, we found that those studying colostrum collected smaller amounts (maximum 5ml) than those working with transitional or mature milk (commonly < 1 oz, 10–15 ml). This may reflect sample availability and concern for vulnerable populations, as lactating people develop the ability to produce greater volumes of milk in mature lactational stages. Additional work studying use of smaller quantities may be warranted, considering that best practices suggest < 1 ml colostrum be collected when possible. Minimization of colostrum collection quantity is a paramount consideration, particularly for at-risk infants requiring critical care (McGuire et al., 2021). Higher sample quantities may provide higher cellular counts for those seeking mRNA from cellular sources. Notably, cell count fluctuates in type and quantity with lactational stage, and therefore higher volumes of mature milk, rather than colostrum, might be necessary to isolate appropriate components. For example, leukocyte content is highest in colostrum, representing up to 70.4% of total milk cells and falling rapidly to a maximum of 1.7% in transitional and 1.5% in mature milk samples (Hassiotou, Hepworth, Metzger, et al., 2013). Those requiring high quantities of RNA may consider isolation of milk cells or milk fat globules, as they have been shown to contain more RNA than skim portions (Alsaweed et al., 2015). Future studies may seek to determine the minimum sample volumes of colostrum, transitional, and mature milk samples suitable for supporting human milk transcriptomics while accounting for known high inter-individual variation in human milk components (Hassiotou, Hepworth, Williams, et al., 2013).

Fresh Collection or Preservation

The results of our synthesis suggest that freshly collected or frozen human milk may be appropriately used if correctly handled for specific experimental applications. Stabilization techniques may offer a lucrative advantage, as they may preserve or prevent RNA from RNase degradation and broach significant feasibility concerns in the sampling of population-based or disease phenotype cohorts of childbearing people. Although most studies used fresh human milk samples (Adams et al., 2022; Alsaweed et al., 2015; Cai et al., 2017; Carli et al., 2020; De Andrés et al., 2018; Gleeson et al., 2022; Kubota et al., 2014; LeMaster et al., 2023; Lemay, Ballard, et al., 2013; Lemay, Hovey, et al., 2013; Sharp et al., 2016; Slupecka-Ziemilska, 2017; Twigger et al., 2015, 2018, 2022; Zhao et al., 2020), several that were subjected to preservation methods such as freezing (Blans et al., 2017; Canale et al., 2021; Nyquist et al., 2022; Tian et al., 2022), or the addition of a stabilizing additive were used in analyses. Our ability to determine the effect of these preservation methods on RNA quantity or quality is severely limited in cases where RIN or other quality assurance metrics were not reported. Notably, the appropriateness of sample preservation and processing technique may vary based on experimental application and whether viability or lysis is a particular concern (as with ScRNA-seq). While freshly collected samples for ScRNA-seq have been shown to yield reliable high-quality data, holding samples at 4 °C overnight or freezing following cell isolation using a cell cryopreservation protocol have demonstrated success (Carli et al., 2020; Nyquist et al., 2022; Twigger et al., 2022). Samples frozen prior to cell isolation may result in poorer quality cells (Nyquist et al., 2022). Similarly, concern surrounding RNA integrity may vary based on whether a qPCR or RNA-seq application is pursued. While RIN values > 7 are strongly preferred for RNA-seq (Puchta et al., 2020), samples with quality scores > 5 may be acceptable for qPCR (Fleige & Pfaffl, 2006). These less stringent integrity requirements may explain the infrequent reporting of RNA integrity in studies using qPCR. Further research is needed to understand the effects of preservation techniques on human milk RNA of the various milk fractions (cellular, skim, lipid) and whether cellular lysis and loss of mRNA occur. It is possible that rapid freeze protocols or use of a stabilizing agent is more feasible when RNA integrity is not a paramount concern, or for use in select experimental applications. Scientists within the human milk and/or transcriptomics field may consider development and validation of remote collection procedures to support inclusion of difficult-to-access (e.g., rural, low-incidence/special) populations.

A variety of extraction and centrifugation protocols exist to guide the fractionation of human milk samples prior to RNA extraction. In general, the softest spins of 350–870 g were used for ScRNA-seq applications, likely to protect the viability and prevent lysis of cells. In the isolation of milk fat globules, research teams used slightly higher spins of 720–2311 g in contrast to the harder spins of cellular isolation at 620–15,000 g. Notably, one team found that a harder spin protocol (15,000 g) produced the highest quality cellular RNA (Lemay, Ballard, et al., 2013). Temperature settings during centrifugation varied between room temperature (20 °C) and refrigerator temperature (4 °C). Considering that many fresh samples were stored on ice for transport to the laboratory, understanding the effects of centrifuge temperature on RNA quality may be warranted, particularly for protocols involving longer spin cycles (20 min). Investigators may consider approaching RNA extraction using a column-based phenol-free method, as it demonstrates high performance for all three fractions of human milk (Alsaweed et al., 2015).

Limitations

We respectfully acknowledge that this synthesis of recent work has limitations. Results suggest that reporting of RIN, RNA quantity, and purity is not a standard practice for all experimental applications. Without this data, our ability to provide insight into the success of human milk sample collection, storage, and processing procedures for gene expression quantification or transcriptomic applications was limited. Furthermore, despite our best efforts to extract this information, it is possible that RNA quality metrics were reported elsewhere, in previous publications or locations that evaded our extraction procedures. Additional considerations outside the scope of this review, including downstream bioinformatic analysis, should be synthesized and critically evaluated in consideration of the fact that data cleaning and handling are widely recognized for the capacity to affect results. Finally, it is possible that applicable publications were published after our search or not retrieved by our search procedures.

Conclusions

The field of human milk science is in critical need of rigorously conducted studies that clearly describe the effects of sample collection, processing, and storage on human milk mRNA isolated from the cellular, skim, and milk fat globule fractions. Without this knowledge, it is unclear whether prospective studies designed to examine mammary gene expression can expect RNA with quality suitable to support various transcriptomic applications. Future studies may seek to investigate novel methods—including remote human milk collection, freezing, and transport of the sample—that facilitate large, multi-center trials. Research teams previously engaged in human milk transcriptomic work may also consider publishing study protocols with inclusion of data that describes RNA sample quantity, quality, and purity. Considering the heterogeneity in reporting practices, we also suggest that a human milk reporting checklist may be beneficial to ensure that future studies can provide the most complete and accurate descriptions of their work.

Supplemental Material

sj-docx-1-jhl-10.1177_08903344251387654 – Supplemental material for The Collection, Storage, and Processing of Human Milk Samples as a Source of mRNA: A State of the Science Review

Supplemental material, sj-docx-1-jhl-10.1177_08903344251387654 for The Collection, Storage, and Processing of Human Milk Samples as a Source of mRNA: A State of the Science Review by Katelyn Desorcy-Scherer and Kerry McNamara in Journal of Human Lactation

Footnotes

Acknowledgements

The authors would like to acknowledge the expertise and guidance of Mary Hitchcock, Senior Academic Librarian at the University of Wisconsin – Madison.

ORCID iD

Katelyn Desorcy-Scherer

Author Contributions

Katelyn Desorcy-Scherer: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing

Kerry McNamara: Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Writing - review & editing

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Disclosures and Conflicts of Interest

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The authors have no conflicts of interest to disclose. The authors worked collaboratively with K.M.D.S., Assistant Professor, as an academic mentor to K.M., DNP student

Supplemental Material

Supplementary Material may be found in the "Supplemental material" tab in the online version of this article.

References

Adams

J. M.

Valentine

C. J.

Karns

R. A.

Rogers

L. K.

Murase

Fowler

G. N.

Nommsen-Rivers

L. A.

(2022). DHA supplementation attenuates inflammation-associated gene expression in the mammary gland of lactating mothers who deliver preterm. Journal of Nutrition, 152(6), 1404–1414. https://doi.org/10.1093/jn/nxac043

Alhussien

M. N.

Dang

A. K.

(2018). Diurnal rhythm in the counts and types of milk somatic cells, neutrophil phagocytosis and plasma cortisol levels in Karan Fries cows during different seasons and parity. Biological Rhythm Research, 49(2), 187–199.

Alsaweed

Hepworth

A. R.

Lefèvre

Hartmann

P. E.

Geddes

D. T.

Hassiotou

(2015). Human milk MicroRNA and total RNA differ depending on milk fractionation. Journal of Cellular Biochemistry, 116(10), 2397–2407. https://doi.org/10.1002/jcb.25207

Barboza

Pinzon

Wickramasinghe

Froehlich

J. W.

Moeller

Smilowitz

J. T.

Ruhaak

L. R.

Huang

Lönnerdal

German

J. B.

Medrano

J. F.

Weimer

B. C.

Lebrilla

C. B.

(2012). Glycosylation of human milk lactoferrin exhibits dynamic changes during early lactation enhancing its role in pathogenic bacteria-host interactions. Molecular and Cellular Proteomics, 11(6), Article 15248. https://doi.org/10.1074/mcp.M111.015248

Blans

Hansen

M. S.

Sørensen

L. V.

Hvam

M. L.

Howard

K. A.

Möller

Wiking

Larsen

L. B.

Rasmussen

J. T.

(2017). Pellet-free isolation of human and bovine milk extracellular vesicles by size-exclusion chromatography. Journal of Extracellular Vesicles, 6(1), Article 1294340. https://doi.org/10.1080/20013078.2017.1294340

Cai

Eck

Friel

J. K.

(2017). Gene expression profiles suggest iron transport pathway in the lactating human epithelial cell. Journal of Pediatric Gastroenterology and Nutrition, 64(3), 460–464. https://doi.org/10.1097/MPG.0000000000001303

Canale

Ramanathan

Pelligrini

Rochette

N. C.

Nadel

B. B.

Gee

(2021). Analysis of gene expression from human breastmilk cells: A comparison between low and high producers, and the influence of anxiety and depression on milk production, gene expression and bacterial production. Heliyon, 7(11), Article e08335. https://doi.org/10.1016/j.heliyon.2021.e08335

Carli

J. F. M.

Trahan

G. D.

Jones

K. L.

Hirsch

Rolloff

K. P.

Dunn

E. Z.

Friedman

J. E.

Barbour

L. A.

Hernandez

T. L.

Maclean

P. S.

Monks

Mcmanaman

J. L.

Rudolph

M. C.

(2020). Single cell RNA sequencing of human milk-derived cells reveals sub-populations of mammary epithelial cells with molecular signatures of progenitor and mature states: a novel, non-invasive framework for investigating human lactation physiology. Journal of Mammary Gland Biology and Neoplasia, 25, 367–387. https://doi.org/10.1007/s10911-020-09466-z/Published

Dastmalchi

Jones

H. N.

Lemas

D. J.

(2022). Assessment of human milk in the era of precision health. In Matthews

D. E.

Norman

(Eds.), Current opinion in clinical nutrition and metabolic care (Vol. 25, pp. 292–297). Lippincott Williams and Wilkins. https://journals.lww.com/co-clinicalnutrition/fulltext/2022/09000/Assessment_of_human_milk_in_the_era_of_precision.4.aspx?context=LatestArticles

10.

De Andrés

Jiménez

Espinosa-Martos

Rodríguez

J. M.

García-Conesa

M. T.

(2018). An exploratory search for potential molecular targets responsive to the probiotic lactobacillus salivarius PS2 in women with mastitis: Gene expression profiling vs. Interindividual variability. Frontiers in Microbiology, 9, Article 2166. https://doi.org/10.3389/fmicb.2018.02166

11.

Fleige

Pfaffl

M. W.

(2006). RNA integrity and the effect on the real-time qRT-PCR performance. Molecular Aspects of Medicine, 27(2–3), 126–139. https://doi.org/10.1016/j.mam.2005.12.003

12.

Gil-Kulik

Leśniewski

Bieńko

Wójcik

Więckowska

Przywara

Petniak

Kondracka

Świstowska

Szymanowski

Wilińska

Wiliński

Płachno

B. J.

Kostuch

Rahnama-Hezavach

Szuta

Kwaśniewska

Bogucka-Kocka

Kocki

(2023). Influence of perinatal factors on gene expression of IAPs family and main factors of pluripotency: OCT4 and SOX2 in human breast milk stem cells—a preliminary report. International Journal of Molecular Sciences, 24(3), Article 2476. https://doi.org/10.3390/ijms24032476

13.

Gleeson

J. P.

Chaudhary

Fein

K. C.

Doerfler

Hredzak-Showalter

Whitehead

K. A.

(2022). Profiling of mature-stage human breast milk cells identifies six unique lactocyte subpopulations. Science Advances, 8(26), Article 6865. https://doi.org/10.1126/sciadv.abm6865

14.

Hassiotou

Geddes

D. T.

Hartmann

P. E.

(2013). Cells in human milk: State of the science. Journal of Human Lactation, 29(2), 171–182. https://doi.org/10.1177/0890334413477242

15.

Hassiotou

Hepworth

A. R.

Metzger

Tat Lai

Trengove

Hartmann

P. E.

Filgueira

(2013). Maternal and infant infections stimulate a rapid leukocyte response in breastmilk. Clinical and Translational Immunology, 2(4), Article e3. https://doi.org/10.1038/cti.2013.1

16.

Hassiotou

Hepworth

A. R.

Williams

T. M.

Twigger

A. J.

Perrella

Lai

C. T.

Filgueira

Geddes

D. T.

Hartmann

P. E.

(2013). Breastmilk cell and fat contents respond similarly to removal of breastmilk by the infant. PLoS One, 8(11), Article e78232. https://doi.org/10.1371/journal.pone.0078232

17.

Chitnis

Monos

Dinh

(2021). Next-generation sequencing technologies: An overview. Human Immunology, 82(11), 801–811. https://doi.org/10.1016/j.humimm.2021.02.012

18.

Kothapalli

K. S. D.

Park

H. G.

Guo

Sun

Zou

Hyon

S. S.

Qin

Lawrence

Ran-Ressler

R. R.

Zhang

J. Y.

Brenna

J. T.

(2018). A novel FADS2 isoform identified in human milk fat globule suppresses FADS2 mediated Δ6-desaturation of omega-3 fatty acids. Prostaglandins Leukotrienes and Essential Fatty Acids, 138, 52–59. https://doi.org/10.1016/j.plefa.2018.06.004

19.

Kubota

Shimojo

Nonaka

Yamashita

Ohara

Igoshi

Ozawa

Nakano

Morita

Inoue

Arima

Chiba

Nakamura

Ikegami

Masuda

Suzuki

Kohno

(2014). Prebiotic consumption in pregnant and lactating women increases IL-27 expression in human milk. British Journal of Nutrition, 111(4), 625–632. https://doi.org/10.1017/S0007114513003036

20.

LeMaster

Pierce

S. H.

Geanes

E. S.

Khanal

Elliott

S. S.

Scott

A. B.

Louiselle

D. A.

McLennan

Maulik

Lewis

Pastinen

Bradley

(2023). The cellular and immunological dynamics of early and transitional human milk. Communications Biology, 6(1), Article 539. https://doi.org/10.1038/s42003-023-04910-2

21.

Lemay

D. G.

Ballard

O. A.

Hughes

M. A.

Morrow

A. L.

Horseman

N. D.

Nommsen-Rivers

L. A.

(2013). RNA sequencing of the human milk fat layer transcriptome reveals distinct gene expression profiles at three stages of lactation. PLoS One, 8(7), Article e67531. https://doi.org/10.1371/journal.pone.0067531

22.

Lemay

D. G.

Hovey

R. C.

Hartono

S. R.

Hinde

Smilowitz

J. T.

Ventimiglia

Schmidt

K. A.

Lee

J. W.

Islas-Trejo

Ivo Silva

Korf

Medrano

J. F.

Barry

P. A.

Bruce German

(2013). Sequencing the transcriptome of milk production: Milk trumps mammary tissue. BMC Genomics. http://www.biomedcentral.com/1471-2164/14/872

23.

Maningat

P. D.

Sen

Rijnkels

Sunehag

A. L.

Hadsell

D. L.

Bray

Haymond

M. W.

Haymond

M. W.

(2009). Gene expression in the human mammary epithelium during lactation: the milk fat globule transcriptome. Physiological Genomics, 37, 12–22. https://doi.org/10.1152/physiolgenomics.90341.2008.-The

24.

Martin Carli

J. F.

Trahan

G. D.

Rudolph

M. C

. (2021). Resolving human lactation heterogeneity using single milk-derived cells, a resource at the ready. Journal of Mammary Gland Biology and Neoplasia, 26(1), 3–8. https://doi.org/10.1007/s10911-021-09489-0

25.

McGuire

M. K.

Seppo

Goga

Buonsenso

Collado

M. C.

Donovan

S. M.

Müller

J. A.

Ofman

Monroy-Valle

O’Connor

D. L.

Pace

R. M.

Van De Perre

(2021). Best practices for human milk collection for COVID-19 research. Breastfeeding Medicine, 16(1), 29–38. https://doi.org/10.1089/bfm.2020.0296

26.

Meek

J. Y.

Noble

(2022). Policy statement: Breastfeeding and the use of human milk. Pediatrics, 150(1). https://doi.org/10.1542/peds.2022-057988; https://publications.aap.org/pediatrics/article/150/1/e2022057988/188347/Policy-Statement-Breastfeeding-and-the-Use-of?autologincheck=redirected

27.

Mohammad

M. A.

Haymond

M. W.

(2013). Regulation of lipid synthesis genes and milk fat production in human mammary epithelial cells during secretory activation. American Journal of Physiology, Endocrinology and Metabolism, 305, 700–716. https://doi.org/10.1152/ajpendo.00052.2013

28.

Mohammad

M. A.

Sunehag

A. L.

Haymond

M. W.

(2014). De novo synthesis of milk triglycerides in humans. American Journal of Physiology. Endocrinology and Metabolism, 306, 838–847. https://doi.org/10.1152/ajpendo.00605.2013

29.

Nyquist

S. K.

Gao

Haining

T. K. J.

Retchin

M. R.

Golan

Drake

R. S.

Kolb

Mead

B. E.

Ahituv

Martinez

M. E.

Shalek

A. K.

Berger

Goods

B. A.

Przytycka

Twigger

A.-J.

(2022). Cellular and transcriptional diversity over the course of human lactation. Proceedings of the National Academy of Sciences of the United States of America, 119(15), e2121720119. https://doi.org/10.1073/pnas.2121720119

30.

Page

M. J.

McKenzie

J. E.

Bossuyt

P. M.

Boutron

Hoffmann

T. C.

Mulrow

C. D.

Shamseer

Tetzlaff

J. M.

Akl

E. A.

Brennan

S. E.

Chou

Glanville

Grimshaw

J. M.

Hróbjartsson

Lalu

M. M.

Loder

E. W.

Mayo-Wilson

McDonald

. . . Moher

(2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. The BMJ, 372, Article n71. https://doi.org/10.1136/bmj.n71

31.

Puchta

Boczkowska

Groszyk

(2020). Low RIN value for RNA-seq library construction from long-term stored seeds: A case study of barley seeds. Genes, 11(10), Article 1190. https://doi.org/10.3390/genes11101190

32.

Raymond

Lefebvre

Texari

Pruvost

Metairon

Cottenet

Zollinger

Mateescu

Billeaud

Picaud

J. C.

Silva-Zolezzi

Descombes

Bosco

(2022). Longitudinal human milk miRNA composition over the first 3 mo of lactation in a cohort of healthy mothers delivering term infants. Journal of Nutrition, 152(1), 94–106. https://doi.org/10.1093/jn/nxab282

33.

Schroeder

Mueller

Stocker

Salowsky

Leiber

Gassmann

Lightfoot

Menzel

Granzow

Ragg

(2006). The RIN: An RNA integrity number for assigning integrity values to RNA measurements. BMC Molecular Biology, 7, 3. https://doi.org/10.1186/1471-2199-7-3

34.

Sharp

J. A.

Lefèvre

Watt

Nicholas

K. R.

(2016). Analysis of human breast milk cells: gene expression profiles during pregnancy, lactation, involution, and mastitic infection. Functional and Integrative Genomics, 16(3), 297–321. https://doi.org/10.1007/s10142-016-0485-0

35.

Slupecka-Ziemilska

Wolinski

Herman

A. P.

Romanowicz

Dziegielewska

Borszewska-Kornacka

M. K.

(2017). Influence of preterm delivery on ghrelin and obestatin concentrations in maternal plasma, milk and their expression in mammary epithelial cells. Journal of Physiology and Pharmacology, 68(5), 693–698.

36.

The National Institutes of Health. (2020). Transcriptome fact sheet. National Human Genome Research Institute. https://www.genome.gov/about-genomics/fact-sheets/Transcriptome-Fact-Sheet#:~:text=A%20transcriptome%20is%20a%20collection,readouts%20present%20in%20a%20cell

37.

Tian

Lin

Chen

Liu

Xie

(2022). Association between FADS gene expression and polyunsaturated fatty acids in breast milk. Nutrients, 14(3), Article 457. https://doi.org/10.3390/nu14030457

38.

Tingö

Ahlberg

Johansson

Pedersen

S. A.

Chawla

Sætrom

Cione

Simpson

M. R.

(2021). Non-coding RNAs in human breast milk: A systematic review. Frontiers in Immunology, 12, Article 725323. https://doi.org/10.3389/fimmu.2021.725323

39.

Twigger

A. J.

Engelbrecht

L. K.

Bach

Schultz-Pernice

Pensa

Stenning

Petricca

Scheel

C. H.

Khaled

W. T.

(2022). Transcriptional changes in the mammary gland during lactation revealed by single cell sequencing of cells from human milk. Nature Communications, 13(1), Article 562. https://doi.org/10.1038/s41467-021-27895-0

40.

Twigger

A. J.

Hepworth

A. R.

Tat Lai

Chetwynd

Stuebe

A. M.

Blancafort

Hartmann

P. E.

Geddes

D. T.

Kakulas

(2015). Gene expression in breastmilk cells is associated with maternal and infant characteristics. Scientific Reports, 5, Article 12933. https://doi.org/10.1038/srep12933

41.

Twigger

A. J.

Küffer

G. K.

Geddes

D. T.

Filgueria

(2018). Expression of granulisyn, perforin and granzymes in human milk over lactation and in the case of maternal infection. Nutrients, 10(9), Article 1230. https://doi.org/10.3390/nu10091230

42.

Veritas Health Innovation. (n.d.). Covidence systematic review software. www.covidence.org

43.

Zhao

Wang

G. D.

Zhang

Zou

Xiang

Peng

Zhou

Yang

Shen

C. K. J.

Wang

Jiao

(2020). TDP-43 facilitates milk lipid secretion by post-transcriptional regulation of Btn1a1 and Xdh. Nature Communications, 11(1), Article 341. https://doi.org/10.1038/s41467-019-14183-1