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.

Cells in human milk are an untapped source, as potential "liquid breast biopsies", of material for investigating lactation physiology in a non-invasive manner. We used single cell RNA sequencing (scRNA-seq) to identify milk-derived mammary epithelial cells (MECs) and their transcriptional signatures in women with diet-controlled gestational diabetes (GDM) with normal lactation. Methodology is described for coordinating milk collections with single cell capture and library preparation via cryopreservation, in addition to scRNA-seq data processing and analyses of MEC transcriptional signatures. We comprehensively characterized 3740 cells from milk samples from two mothers at two weeks postpartum. Most cells (>90%) were luminal MECs (luMECs) expressing lactalbumin alpha and casein beta and positive for keratin 8 and keratin 18. Few cells were keratin 14+ basal MECs and a small immune cell population was present (<10%). Analysis of differential gene expression among clusters identified six potentially distinct luMEC subpopulation signatures, suggesting the potential for subtle functional differences among luMECs, and included one cluster that was positive for both progenitor markers and mature milk transcripts. No expression of pluripotency markers POU class 5 homeobox 1 (POU5F1, encoding OCT4) SRY-box transcription factor 2 (SOX2) or nanog homeobox (NANOG), was observed. These observations were supported by flow cytometric analysis of MECs from mature milk samples from three women with diet-controlled GDM (2-8 mo postpartum), indicating a negligible basal/stem cell population (epithelial cell adhesion molecule (EPCAM)-/integrin subunit alpha 6 (CD49f)+, 0.07%) and a small progenitor population (EPCAM+/CD49f+, 1.1%). We provide a computational framework for others and future studies, as well as report the first milk-derived cells to be analyzed by scRNA-seq. We discuss the clinical potential and current limitations of using milk-derived cells as material for characterizing human mammary physiology.

Randomization to a Provided Higher-Complex-Carbohydrate Versus Conventional Diet in Gestational Diabetes Mellitus Results in Similar Newborn Adiposity.

OBJECTIVE: Nutrition therapy for gestational diabetes mellitus (GDM) has conventionally focused on carbohydrate restriction. In a randomized controlled trial (RCT), we tested the hypothesis that a diet (all meals provided) with liberalized complex carbohydrate (60%) and lower fat (25%) (CHOICE diet) could improve maternal insulin resistance and 24-h glycemia, resulting in reduced newborn adiposity (NB%fat; powered outcome) versus a conventional lower-carbohydrate (40%) and higher-fat (45%) (LC/CONV) diet. RESEARCH DESIGN AND METHODS: After diagnosis (at ∼28-30 weeks' gestation), 59 women with diet-controlled GDM (mean ± SEM; BMI 32 ± 1 kg/m2) were randomized to a provided LC/CONV or CHOICE diet (BMI-matched calories) through delivery. At 30-31 and 36-37 weeks of gestation, a 2-h, 75-g oral glucose tolerance test (OGTT) was performed and a continuous glucose monitor (CGM) was worn for 72 h. Cord blood samples were collected at delivery. NB%fat was measured by air displacement plethysmography (13.4 ± 0.4 days). RESULTS: There were 23 women per group (LC/CONV [214 g/day carbohydrate] and CHOICE [316 g/day carbohydrate]). For LC/CONV and CHOICE, respectively (mean ± SEM), NB%fat (10.1 ± 1 vs. 10.5 ± 1), birth weight (3,303 ± 98 vs. 3,293 ± 81 g), and cord C-peptide levels were not different. Weight gain, physical activity, and gestational age at delivery were similar. At 36-37 weeks of gestation, CGM fasting (86 ± 3 vs. 90 ± 3 mg/dL), 1-h postprandial (119 ± 3 vs. 117 ± 3 mg/dL), 2-h postprandial (106 ± 3 vs. 108 ± 3 mg/dL), percent time in range (%TIR; 92 ± 1 vs. 91 ± 1), and 24-h glucose area under the curve values were similar between diets. The %time >120 mg/dL was statistically higher (8%) in CHOICE, as was the nocturnal glucose AUC; however, nocturnal %TIR (63-100 mg/dL) was not different. There were no between-group differences in OGTT glucose and insulin levels at 36-37 weeks of gestation. CONCLUSIONS: A ∼100 g/day difference in carbohydrate intake did not result in between-group differences in NB%fat, cord C-peptide level, maternal 24-h glycemia, %TIR, or insulin resistance indices in diet-controlled GDM.

Human milk imparts higher insulin concentration in infants born to women with type 2 diabetes mellitus.

OBJECTIVE: Human milk (HM) insulin plays many roles for the infant, especially for the newborn. We hypothesized HM insulin in women with type 2 diabetes (T2DM) would be higher than BMI-matched women with either gestational diabetes (GDM) or normal glucose tolerance (NGT). In T2DM, we also assessed macronutrient composition and relationships between maternal glycemic control and HM insulin. STUDY DESIGN: HM was characterized at 2-weeks postpartum among three BMI-matched groups: T2DM (n= 12), diet-controlled GDM (n= 12), and NGT (n= 12). In T2DM, additional fasting and postprandial HM samples were collected while wearing a continuous glucose monitor (CGM), as well as fasting and 90-minute postprandial samples after a standardized meal at 1-2 weeks postpartum. RESULTS: Fasting HM insulin was two times higher in T2DM compared to GDM and NGT (p < .001), which were not different from each other. Among T2DM, fasting (p < .001) and postprandial (p = .01) HM insulin levels were between 2 and 5× higher than plasma. Postprandial HM insulin (p = .03) and glucose (p < .001) were increased compared to fasting. Mean nocturnal glucose (p < .01) and maternal hemoglobin A1c (p < .01) positively associated with fasting HM insulin. CONCLUSIONS: These data are the first to show HM insulin concentrations are doubled in T2DM compared to BMI-matched GDM and NGT. In HM of T2DM, insulin increases postprandially, may be concentrated relative to plasma, and is influenced by maternal glycemic control, with potential clinical implications that merit further study.

Social Determinants of Health Challenges Are Prevalent Among Commercially Insured Populations.

OBJECTIVES: To evaluate the prevalence of social determinants of health (SDoH) factors in a large commercially-insured population and to characterize the prevalence of common conditions (eg, diabetes, behavioral health issues) and addressable health services utilization concerns (eg, lack of preventive care) for which employers offer no- and low-cost benefit programs. METHODS: We identified groups with SDoH challenges within a commercially-insured population of 5.1 M through administrative data and self-report. Using medical claims and health assessment data, we identified populations with SDoH needs who had common conditions for which employers often provide no- or low-cost benefit programs (ie, diabetes, behavioral health conditions, high-risk pregnancy, overweight/obesity). Additionally, we sought populations with common addressable health services utilization concerns such as avoidable emergency room visits, lack of preventive care services, or non-adherence to medications. We used univariate analyses to describe the prevalence of SDoH risks in the population of interest. RESULTS: Twenty-seven percent of this commercially-insured population live in a zip code where the median income is at or below 200% of the Federal Poverty Line. Respondents identified cost (55%) and family, school, or work responsibilities (26%) as key barriers to care. ER overutilization rates are higher in lower income zip codes than wealthier zip codes (34% vs 9%) as is the prevalence of diabetes, overweight/obesity, and behavioral issues, and decreased use of preventive services. Fifteen percent of the study population live in a low-access food area. There is considerable variability in access to employer-sponsored resources to address these needs (70% of employers provide behavioral health programs; 63% provide telehealth programs, but only 1% offer healthy food programs and less than 0.5% offer either child care or transportation support programs). CONCLUSIONS: Commercially insured populations could benefit from employer-sponsored programs or benefits that address key SDoH barriers such as financial support, healthy food programs, child-care, and transportation.

Effect of type 2 diabetes mellitus on placental expression and activity of nutrient transporters and their association with birth weight and neonatal adiposity.

AIMS: Infants born to women with Type 2 Diabetes Mellitus (T2DM) are at risk of being born large for gestational age due to excess fetal fat accretion. Placental nutrient transport determines fetal nutrient availability, impacting fetal growth. The aims of the study were to evaluate the effect of T2DM on placental insulin signaling, placental nutrient transporters and neonatal adiposity. METHODS: Placentas were collected from BMI-matched normoglycemic controls (NGT, n = 9) and T2DM (n = 9) women. Syncytiotrophoblast microvillous (MVM) and basal (BM) plasma membranes were isolated. Expression of glucose (GLUT1, -4), fatty acid (FATP2, -4, -6, FAT/CD36), amino acid (SNAT1, -2, -4, LAT1, -2) transporters, insulin signaling, and System A transporter activity was determined. Neonatal fat mass (%) was measured in a subset of neonates born to T2DM women. RESULTS: GLUT1 protein expression was increased (p = 0.001) and GLUT4 decreased (p = 0.006) in BM from T2DM. MVM FATP6 expression was increased (p = 0.02) and correlated with birth weight in both T2DM and NGT groups (r = 0.65, p = 0.02). BM FATP6 expression was increased (p = 0.01) in T2DM. In MVM of T2DM placentas, SNAT1 expression was increased (p = 0.05) and correlated with birth weight (r = 0.84, p = 0.004); SNAT2 was increased (p = 0.01), however System A transporter activity was not different between groups. MVM LAT1 expression was increased (p = 0.01) in T2DM and correlated with birth weight (r = 0.59, p = 0.04) and neonatal fat mass (r = 0.76, p = 0.06). CONCLUSION: In pregnancies complicated by T2DM placental protein expression of transporters for glucose, amino acids and fatty acids is increased, which may contribute to increased fetal growth and neonatal adiposity.

Obstructive Sleep Apnea Is Associated With Altered Glycemic Patterns in Pregnant Women With Obesity.

OBJECTIVE: Often unrecognized, obstructive sleep apnea (OSA) worsens over pregnancy and is associated with poorer perinatal outcomes. The association between OSA in late pregnancy and metabolic biomarkers remains poorly understood. We tested the hypothesis that OSA in pregnant women with obesity is positively correlated with 24-hour patterns of glycemia and IR despite controlling for diet. DESIGN: Pregnant women (32 to 34 weeks' gestation; body mass index, 30 to 40 kg/m2) wore a continuous glucose monitor for 3 days. OSA was measured in-home by WatchPAT 200™ [apnea hypopnea index (AHI), oxygen desaturation index (ODI; number per hour)]. Fasting blood was collected followed by a 2-hour, 75-g, oral glucose tolerance test to measure IR. Association between AHI and 24-hour glucose area under the curve (AUC) was the powered outcome. RESULTS: Of 18 women (29.4 ± 1.4 years of age [mean ± SEM]), 12 (67%) had an AHI ≥5 (mild OSA). AHI and ODI were correlated with 24-hour glucose AUC (r = 0.50 to 0.54; P ≤ 0.03) and mean 24-hour glucose (r = 0.55 to 0.59; P ≤ 0.02). AHI and ODI were correlated with estimated hepatic IR (r = 0.59 to 0.74; P < 0.01), fasting free fatty acids (fFFAs; r = 0.53 to 0.56; P < 0.05), and waking cortisol (r = 0.49 to 0.64; P < 0.05). CONCLUSIONS: Mild OSA is common in pregnant women with obesity and correlated with increased glycemic profiles, fFFAs, and estimates of hepatic IR. OSA is a potentially treatable target to optimize maternal glycemia and metabolism, fetal fuel supply, and pregnancy outcomes.

Postprandial Triglycerides Predict Newborn Fat More Strongly than Glucose in Women with Obesity in Early Pregnancy.

OBJECTIVE: Maternal obesity (OB) accounts for the majority of large-for-gestational-age infants, and newborn percent fat (NB%fat) correlates strongest with childhood OB. In addition to maternal glucose, fasting triglycerides (TGs) may contribute, but postprandial triglycerides (PPTGs) are unstudied. It was hypothesized that fasting TGs and PPTGs are higher in women with OB compared with women with normal weight (NW) throughout pregnancy, correlate more strongly with NB%fat than glucose, and may relate to dietary chylomicron TGs. METHODS: Fasting TGs and PPTGs, free fatty acids, glucose, and insulin were prospectively measured 10 times over 4 hours after a controlled liquid breakfast early (14-16 weeks) and later (26-28 weeks) in pregnancy in 27 mothers with NW and 27 with OB. NB%fat was measured by dual x-ray absorptometry. RESULTS: Fasting TGs and PPTGs were already ≥ 30% higher in mothers with OB at 14 to 16 weeks (P < 0.001) versus mothers with NW. In mothers with OB, a simple 1-hour (r = 0.71; P < 0.01) or 2-hour (r = 0.69; P < 0.01) PPTG at 14 to 16 weeks correlated strongest with NB%fat. In mothers with NW, the increase in TGs from early to later pregnancy correlated strongest with NB%fat (r = 0.57; P < 0.01). Maternal glucose did not statistically add to prediction models. CONCLUSIONS: These novel data suggest that 1- or 2-hour PPTGs might be a new target for early intervention in pregnancies with OB to prevent excess newborn adiposity and attenuate child OB risk.

Validity of uniaxial accelerometry during activities of daily living in children.

The purpose of this study was to examine the validity of treadmill-based equations of two commonly used uniaxial accelerometers to estimate energy expenditure (EE) during activities of daily living in children. Twelve subjects with mean (SD) age11.4 (0.4) years engaged in a choreographed routine consisting of three activities (sweeping, bowling, and basketball) of 4min duration while wearing a Manufacturing Technology, Inc. (MTI) accelerometer, Caltrac accelerometer, and a portable gas analyzer (Cosmed K4b(2)). The equations of Trost et al. and Sallis et al. were used to convert activity counts to estimations of EE for the MTI and Caltrac, respectively. Correlation coefficients between Caltrac predictions of EE and measured EE from indirect calorimetry ranged from r=0.22 to 0.72 for individual activities. Correlations between MTI EE predictions and indirect calorimetry ranged from r=0.50 to 0.68 for individual activities. When the activities were pooled the correlations between EE from uniaxial accelerometers and EE from indirect calorimetry were moderately strong (MTI, r=0.78 and Caltrac, r=0.82). Inter-accelerometer (counts min(-1)) correlations were 0.08, -0.54, 0.63, and 0.79 for sweeping, bowling, basketball, and pooled data, respectively. The overall mean difference, or bias, and 95% confidence intervals (CI) for each uniaxial accelerometer compared to indirect calorimetry were as follows: Caltrac, bias = 2.80 (2.36, 3.24) kcal min(-1); MTI, bias = 0.88 (0.23, 1.53) kcal min(-1). Both accelerometers significantly underestimated measured EE ( P<0.05). Uniaxial accelerometers provide potential for the measurement of physical activity (PA) and EE in children. Future studies refining accelerometry predictions of PA and EE are warranted.