Figures
Abstract
Background
The current practice in newborn medicine is to subjectively assess when a premature infant is ready to feed by mouth. When the assessment is inaccurate, the resulting feeding morbidities may be significant, resulting in long-term health consequences and millions of health care dollars annually. We hypothesized that the developmental maturation of hypothalamic regulation of feeding behavior is a predictor of successful oral feeding in the premature infant. To test this hypothesis, we analyzed the gene expression of neuropeptide Y2 receptor (NPY2R), a known hypothalamic regulator of feeding behavior, in neonatal saliva to determine its role as a biomarker in predicting oral feeding success in the neonate.
Methodology/Principal Findings
Salivary samples (n = 116), were prospectively collected from 63 preterm and 13 term neonates (post-conceptual age (PCA) 26 4/7 to 41 4/7 weeks) from five predefined feeding stages. Expression of NPY2R in neonatal saliva was determined by multiplex RT-qPCR amplification. Expression results were retrospectively correlated with feeding status at time of sample collection. Statistical analysis revealed that expression of NPY2R had a 95% positive predictive value for feeding immaturity. NPY2R expression statistically significantly decreased with advancing PCA (Wilcoxon test p value<0.01), and was associated with feeding status (chi square p value = 0.013).
Conclusions/Significance
Developmental maturation of hypothalamic regulation of feeding behavior is an essential component of oral feeding success in the newborn. NPY2R expression in neonatal saliva is predictive of an immature feeding pattern. It is a clinically relevant biomarker that may be monitored in saliva to improve clinical care and reduce significant feeding-associated morbidities that affect the premature neonate.
Citation: Maron JL, Johnson KL, Dietz JA, Chen ML, Bianchi DW (2012) Neuropeptide Y2 Receptor (NPY2R) Expression in Saliva Predicts Feeding Immaturity in the Premature Neonate. PLoS ONE 7(5): e37870. https://doi.org/10.1371/journal.pone.0037870
Editor: Ann M. Moormann, University of Massachusetts Medical School, United States of America
Received: September 27, 2011; Accepted: April 30, 2012; Published: May 21, 2012
Copyright: © 2012 Maron et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: NICHD K08 HD 059819-03, the Charles H. Hood Foundation (www.hria.org/tmfservices/tmfgrants/hood.html), and the Richard Saltonstall Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The current practice in newborn intensive care units is to have caregivers subjectively assess when a premature infant is ready to feed by mouth. When the assessment is inaccurate, morbidities may include choking, hypoxia, aspiration, oral aversion, and prolonged hospitalization [1]–[3]. These associated morbidities are estimated to cost millions of U.S. health care dollars annually. A barrier to the development of an objective diagnostic assay is the ability to directly assess the complex neuromuscular and neurological sequences required for the development of safe and effective oral feeding skills in the premature neonate [4]–[6]. Furthermore, feeding behavior is the result of a complex interplay of feedback loops, satiety, and cellular signaling that is under the direct control of the hypothalamus [7], [8]. While hypothalamic regulation of feeding behavior is an area of active investigation, such research has frequently been conducted on genetically altered animal models with post-mortem examinations to allow for direct visualization and staining of the hypothalamus and its associated neurological pathways [9]–[12].
Neuropeptide Y (NPY) and its family of five receptors are known to be associated with hypothalamic regulation of feeding behavior, metabolism, and energy homeostasis in both rodents and humans [9]–[12]. In particular, knock-out mice for the gene neuropeptide Y 2 receptor, NPY2R, exhibit hyperphagia and excessive weight gain [9]. As a result, research on the gene has focused on genotyping analysis for the identification of single nucleotide polymorphisms (SNPs) in obese human subjects [13], [14] and drug development for manipulation of its expression as a novel therapeutic treatment for obesity [15], [16]. Although primarily expressed in the arcuate nucleus of the hypothalamus [17], NPY2R may also be found in other tissues including trabecular bone [18], vascular tissue [19], and colonic mucosa [20].
Previously, our laboratory published on the diverse range of gene expression changes occurring in the premature neonate that can be detected in as little as 10 µL of saliva. Specifically, we showed that expression of NPY2R in neonatal saliva statistically significantly changed over time as newborns learned to orally feed [21]. This work suggested that the hypothalamus plays a central role in feeding behavior in the premature neonate. Further, through noninvasive salivary gene expression analyses, we can monitor, in real-time, hypothalamic regulation as it directly relates to the learning process of oral feeding. Here, we hypothesized that the physiological hyperphagia and exponential weight gain observed in healthy term newborns is associated with decreased expression of NPY2R. Persistence of NYP2R in salivary samples would suggest immature hypothalamic regulation, indicative of failed oral feeding trials. In this study, we independently explored NPY2R's expression profile in relation to feeding status and post-conceptual age (PCA), and determined its role as a biomarker in neonatal saliva to objectively predict successful oral feeding in the newborn.
Results
Demographics and Sample Characteristics
One hundred and sixteen salivary samples (10 to 50 µL) were collected from 76 newborns with post-conceptual ages (PCAs) ranging from 26 4/7 to 41 4/7 weeks. There were 63 preterm and 13 term infants in this data set. Pertinent clinical information for all subjects is shown in Table 1, while a detailed summary of clinical sequelae for each individual subject is available in Table S1. In general, subjects in the study were either healthy term newborns, or premature infants who had resolution of their acute clinical sequelae and were “feeders and growers” in the NICU prior to discharge home. Of the 76 subjects enrolled in this study, 31 had between two and five salivary samples analyzed, at either different PCAs and/or feeding statuses. The number of samples collected from subjects at each predefined feeding stage, as well as the percentage of infants expressing NPY2R, was as follows: Stage 1: no feeds (NPO) (n = 17; PCA 25 3/7 to 36 1/7 weeks; 59% NPY2R expression); Stage 2: partial per gastric feeds (PPG) (n = 21, PCA 28 2/7 to 41 3/7 weeks; 57% NPY2R expression); Stage 3: full per gastric feeds (FPG) (n = 36, PCA 28 5/7 to 37 5/7 weeks; 67% NPY2R expression); Stage 4: partial oral feeds (PPO) (n = 24, PCA 33 5/7 to 41 2/7 weeks; 50% NPY2R expression); Stage 5: full oral feeds (FPO) (n = 18, PCA 33 3/7 to 40 2/7 weeks; 17% NPY2R expression).
Multiplex Reverse Transcription-quantitative Polymerase Chain Reaction (RT-qPCR) Characteristics
Every effort was made to adhere to the Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) guidelines [22]. Multiplex RT-qPCR amplification was performed on all extracted total RNA samples for the gene NPY2R, along with three reference genes: glyceraldehyde-3-phosphate dehydrogenase (GAPDH), tyrosine 3-monoxygenase/tryptophan 5-monooxygenase activation protein, zeta polypeptide (YWHAZ), and hypoxanthine phosphoribosyltransferase 1 (HPRT1). Reference genes were selected based upon microarray data from previous studies conducted in our laboratory that revealed that each of these genes maintains a relative constant and consistent range of expression across different PCAs. We selected one gene known to be expressed at a relatively high level (YWHAZ), one that had an average expression level (GAPDH), and one that demonstrated a low level of expression (HPRT1) at the limit of our detection level on the multiplex platform.
Prior to multiplex RT-qPCR, three samples were tested simultaneously on both uniplex and multiplex platforms for all genes to assess reaction efficiencies. There was no difference in reaction efficiency between uniplex and multiplex assays. Of the three test samples run in both formats, quantification cycle (Cq) values were within 1 cycle for all genes. Thus, we were confident that running the samples on a multiplex platform would not impact results. The reference gene YWHAZ had the highest level of expression in our samples (mean Cq: 27.7), followed by GAPDH (mean Cq: 30.3), and HPRT1 (mean Cq: 36.8). Mean delta Cq and standard deviation values between reference genes for all reactions were as follows: GAPDH-YWHAZ: 2.3+/−1.85; HPRT1-GAPDH 7.2+/−1.83; HPRT1-YWHAZ 9.5+/−1.93. These results demonstrate that reaction efficiencies across experiments were similar and reproducible. Of the 116 samples analyzed, 95 had amplification of all three reference genes, while 21 revealed amplification of GAPDH, and YWHAZ only. Of these 21 samples, six were positive for NPY2R expression and 15 had no detectable NPY2R in the sample.
Assay Results
For the purpose of this study, we performed three separate statistical analyses on the data. The first analysis considered all 116 data points; the second analysis considered only one sample per subject in order to ensure that multiple sampling from an individual was not skewing the data (n = 76); the third analysis removed samples that did not have amplification of all three housekeeping genes in order to eliminate those samples that had a theoretical risk of a false negative result (n = 95). NPY2R expression in neonatal saliva for all 116 samples had a 95% positive predictive value (CI: 85%–99%) of an immature feeding pattern with an inability to sustain full oral feeds (Stage 5). The negative predictive value of the assay was 27% (CI: 17%–41%). The sensitivity was 59% (CI: 49%–69%) with 83% specificity (CI: 58%–96%). There was a statistically significant difference between NPY2R expression and PCA. Neonates who expressed NPY2R were younger than those infants who did not express the gene (p value<0.01) (Figure 1). However, among term infants, there was a statistically significant difference between infants who could and could not orally feed (p value = 0.037) (Figure 2). Expression of NPY2R was associated with feeding status (p value = 0.013) (Figure 3). As infants matured through the feeding stages, they were less likely to express the gene.
Neither of the two additional statistical analyses altered our primary results. Limiting the data set to include only one salivary sample per subject revealed a positive predictive value of 97% (CI: 85%–100%), with a negative predictive value of 38% (CI: 23%–55%). Sensitivity of the assay was 62% (CI: 50%–75%) with specificity of 93% (CI: 66%–100%). NPY2R expression remained statistically significantly associated with advancing PCA (Wilcoxon rank sum test p value<0.01) and feeding status (chi square p value = 0.004). Further, eliminating the samples that did not have amplification of all three housekeeping genes in order to reduce the potential impact of false negative results had a positive predictive value of 95% (CI: 84%–99%), negative predictive value of 33% (CI:20%–50%), with a sensitivity of 67% (CI: 55%–77%) and specificity of 81% (51%–95%). NPY2R expression was marginally statistically significantly associated with advancing PCA (Fisher's exact test p value = 0.054). There remained a nonrandom association between NPY2R expression and feeding status (chi square p value = 0.076).
Discussion
Our results demonstrate that detection of NPY2R in neonatal saliva has a 95% positive predictive value for determining that an infant cannot sustain full oral feeds, implying that this gene plays an important role in neonatal feeding behavior. Our data suggest that NPY2R is a novel biomarker that could be used as part of an objective assay prior to the introduction of oral feeds, particularly for those infants at risk for aspiration, hypoxia, and long-term feeding aversion.
Although the precise molecular mechanisms by which NPY2R influences neonatal feeding are currently unknown, our data demonstrate that expression of NPY2R is independent of the presence of enteral nutrition if given by catheter. We were only able to detect down-regulation of the gene once infants were able to take at least some feeds by mouth. Thus, stimulation of the gastrointestinal tract alone is not enough to cause decreased gene expression. Similarly, although NPY2R expression is statistically significantly negatively correlated with advancing PCA, here too, it appears that the PCA of the infant is not the sole driving force behind NPY2R expression. Some subjects in this study could successfully orally feed prior to 34 weeks' PCA, while others at 41 weeks' PCA could not. In fact, there was a statistically significant difference in NPY2R expression between infants who could and could not orally feed at term gestation. Thus, while advancing gestational age correlates with decreased NPY2R expression, the relationship is complex and not mutually inclusive. Of note, neonates in this study had a wide range of ages and clinical sequelae. The deliberate incorporation of a diverse patient population was essential to determine the applicability and accuracy of NPY2R as a diagnostic salivary biomarker.
We believe that NPY2R's relatively weak negative predictive power confirms the complexities of oral feeding in the newborn. The physical limitations of some infants, including lung disease, reliance on oxygen and poor gastrointestinal motility, may prohibit neonates from sustaining oral feeds who otherwise appear to have appropriate hypothalamic signaling. In addition, subjective assessments by the caregivers that infants could not safely feed and therefore did not allow the neonates to attempt oral feeds may have skewed the data. Thus, future work must be performed to identify additional biomarkers in neonatal saliva that will incorporate all aspects of feeding in the neonate, including maturation of oral musculature and innervation.
The potential to monitor hypothalamic regulation of feeding behavior through salivary gene expression analysis is also a major conclusion of this study. Real-time gene expression of NPY2R may be obtained noninvasively, in as little as 10 µl of saliva, from human subjects. We speculate that our ability to readily detect NPY2R in neonatal saliva is a direct result of the permeability of the arcuate nucleus to the blood brain barrier [17]. As a filtrate of blood, saliva has been called a ‘window into the body’ with a biomarker profile similar to plasma [23].
The analysis of NPY2R gene expression as it relates to oral feeding in the premature neonate provides a window into developmental biology that has the potential to transcend into other areas of medicine. The early newborn period represents a unique time in the human lifespan when hyperphagia and exponential weight gain is normal. The average newborn will gain 200% of his or her birth weight by one year of age. Our results suggest that developmentally, the human neonate down-regulates NPY2R expression in order to demonstrate the necessary hyperphagia required for this exponential growth. We speculate that understanding the normal developmental expression of the gene, combined with its ease of monitoring in human saliva, has the potential to help elucidate pathological conditions, including obesity.
Previous studies on neonatal oral feeding almost exclusively focused on oral motor skills, the suck-swallow reflex, neurodevelopment and neuronal integration [4], [5]. Here, we speculate that in addition to these essential developmental milestones, satiety, as regulated, in part, by NPY2R expression, also plays a key role in oral feeding success. This is based both upon the numerous published reports linking NPY2R expression with hunger, satiety and feeding behavior in both humans and animals [9]–[14], and our previous work on the neonatal salivary transcriptome demonstrating that NPY2R had statistically significantly gene expression changes over time as infants matured and learned to orally feed [21]. By demonstrating that detection of NPY2R in neonatal saliva is predictive of failed oral feeding trials, we have substantiated our hypothesis that NPY2R does play a role in the attainment of successful newborn oral feeding.
In conclusion, NPY2R is a predictive novel biomarker in neonatal saliva that may be monitored noninvasively in order to objectively determine when an infant can safely feed by mouth. While future research is needed to further elucidate the precise molecular role of NPY2R in relation to human newborn feeding, its presence in neonatal saliva nevertheless remains 95% predictive of an immature feeding pattern and an inability to sustain full oral feeds. This work lays the foundation for the development of an objective diagnostic test to accurately and safely predict oral readiness to feed in this vulnerable population, significantly improving clinical care, and reducing millions of dollars of health care costs associated with feeding morbidities in the premature neonatal population.
Materials and Methods
Ethics Statement
This study was approved by the Tufts Medical Center Institutional Review Board. Written parental consent was obtained for all neonatal subjects enrolled.
Demographics and Sample Characteristics
Salivary samples were collected from premature and term neonates with a diverse range of clinical sequelae at various feeding stages during hospitalization: Stage 1: no feeds (NPO); Stage 2: partial gastric feedings (PPG); Stage 3: full gastric feedings (FPG); Stage 4: partial oral feeds (PPO); Stage 5: full oral feeds (FPO). Infant feeding stage at time of collection was determined solely by the caregivers and was not influenced by study participation. The number of samples collected from a subject was dependent upon their clinical course. In general, premature infants had more than one salivary sample obtained as they matured through the feeding process, while healthy term infants had only one sample obtained at feeding stage 5, FPO. Although salivary samples were collected prospectively from all enrolled subjects, the correlation between salivary NPY2R gene expression and feeding status was made retrospectively once all samples were obtained and analyzed.
Salivary Collection and mRNA Extraction
All salivary samples were collected and processed with previously described standardized techniques that aim to simulate routine bedside care of the neonates [24]. Samples were stored at 4°C for a minimum of 48 hours prior to total RNA extraction with the use of the RNA Protect Saliva Mini Kit (Qiagen, Valencia, CA USA) per manufacturer's instructions. On column DNase digestion was performed on all samples to limit DNA contamination. Final elution volume was approximately 14 µL. Samples were stored at −80°C pending further analysis.
Multiplex qRT-PCR
All RT-qPCR experiments were performed on the Life Technologies 7900 instrument with the use of the Path-ID™ Multiplex One-Step RT-PCR Kit (Life Technologies, Carlsbad, CA USA). Standard stock sequences of reference genes were provided by Life Technologies and were VIC labeled, primer limited. They were as follows: GAPDH-VIC (Hs03929097), HPRT1-VIC (Hs01003267), and YWHAZ-VIC (Hs03044281). Gene sequences for NPY2R were custom made with the use of Primer Express Software v 1 to ensure optimal G-C content and melting temperatures. The custom sequence for NPY2R (Sequence accession number: NM_000910.2) was as follows: Forward Primer: GGC TTT CCT CTC GGC CTT C; Reverse Primer TGT CAC GGA CAC CTC AGA GTG; Probe 6FAM-CTG TGA GCA GCG GTT GGA TGC CAT-TAMRA. The amplicon is located at base pairs 1496 to 1563 of the gene and to the best of our knowledge, contains no SNPs. There are only two known exons for the gene NPY2R. The entire amplicon used in this study was contained within one exon.
For each salivary sample, NPY2R was run in triplicate, multiplexed one time each with the three reference genes. Negative controls with nuclease-free water were performed on each plate. The total volume for each reaction was 25 µL including 2.5 µL of template in each well. The thermal cycle profile for all reactions was as follows: 48°C for 10 minutes, 95°C for 10 minutes, followed by 40 cycles of PCR with a 15 second denaturing cycle at 95°C, followed by 45 seconds of annealing and extension at 60°C.
For the purposes of this study, only the expression of NPY2R in neonatal saliva was considered. The gene was considered expressed if it amplified along with ≥2 of the reference genes. Similarly, NPY2R was considered not expressed if it did not amplify in the presence of ≥2 of the reference genes. Lacking any reference values for salivary NPY2R, it is difficult to provide a clinical interpretation based upon normalized relative quantitative values. Therefore, we analyzed the data in the context of expression of the gene in order to provide our most accurate and biologically relevant assessment. The binary nature of the assay makes it well suited for the development of a rapid diagnostic assay. Finally, average Cq for each reference gene was calculated, along with mean delta Cq and standard deviation values between housekeeping genes to assess efficiencies and variability between reactions.
Statistical Analyses
Statistical analyses included the Fisher's exact test to determine the relationship of NPY2R expression and gestational age, as well as a Wilcoxon rank sum test to compare gestational age in the groups of infants that did or did not express NPY2R (Figure 1). Chi-square test was used to assess the association between NPY2R gene expression and feeding status (Stages 1–5) (Figure 3). We further examined NPY2R expression at term gestation (≥37 weeks' gestation) with the use of a Fisher's exact test to compare expression of the gene in saliva between infants who could and could not successfully feed (Figure 2). Sensitivity, specificity, negative and positive predictive values, along with each respective confidence interval, of the assay were then calculated. Additional statistical analyses were performed to (1) limit the analysis to those samples that had amplification of all three reference genes and to (2) ensure that repeat measures from the same individuals in this study did not skew the data. In the latter analysis, we used only one sample per infant, determined by a random computer number generator to reduce the risk of bias.
Supporting Information
Table S1.
Complete subject demographic and clinical information.
https://doi.org/10.1371/journal.pone.0037870.s001
(XLSX)
Acknowledgments
We would like to thank the families who graciously participated in this study, as well as the staff of the Mother Infant and Neonatal Intensive Care Units at Tufts Medical Center, Boston, MA, USA. We would also like to thank Jessica Oh for help in preparing this manuscript.
Author Contributions
Conceived and designed the experiments: JLM KLJ JAD. Performed the experiments: JAD. Analyzed the data: JLM KLJ MLC. Contributed reagents/materials/analysis tools: JLM DWB. Wrote the paper: JLM KLJ MLC DWB.
References
- 1. Samara M, Johnson S, Lamberts K, Marlow N, Wolke D (2010) Eating problems at age 6 years in a whole population sample of extremely preterm children. Dev Med Child Neurol 52: e16–e22.
- 2. Lau C (2006) Oral feeding in the preterm infant. Neo Reviews 7: e19–e27.
- 3. Mizuno K, Ueda A (2005) Neonatal feeding performance as a predictor of neurodevelopmental outcome at 18 months. Dev Med Child Neurol 47: 299–304.
- 4. Delaney AL, Arvedson JC (2008) Development of swallowing and feeding: prenatal through first year of life. Dev Disabil Res Rev 14: 105–117.
- 5. Amaizu N, Shulman RJ, Schanler RJ, Lau C (2008) Maturation of oral feeding skills in preterm infants. Acta Paediatr 97: 61–7.
- 6. Barlow SM (2009) Central pattern generation involved in oral and respiratory control for feeding in the term infant. Curr Opin Otolaryngol Head Neck Surg 17: 187–93.
- 7. Panksepp J (1974) Hypothalamic regulation of energy balance and feeding behavior. Fed Proc 33: 1150–1165.
- 8. King BM (2006) The rise, fall, and resurrection of the ventromedial hypothalamus in the regulation of feeding behavior and body weight. Physiol Behav 87: 221–224.
- 9. Naveilhan P, Hassain H, Canals JM, Ekstrand AJ, Larefalk A, et al. (1999) Normal feeding behavior, body weight and leptin response require the neuropeptide Y Y2 receptor. Nat Med 5: 1188–93.
- 10. Lin S, Boey D, Herzog H (2004) NPY and Y receptors: lessons from transgenic and knockout models. Neuropeptides 38: 189–200.
- 11. Huang X-F, Yu U, Li Y, Tim S, Deng C, et al. (2008) Ventromedial hypothalamic NPY Y2 receptor in the maintenance of body weight in diet-induced obesity in mice. Neurochem Res 33: 1881–1888.
- 12. Butler AA, Cone RD (2001) Knockout models resulting in the development of obesity. Trends Genet 17: S50–54.
- 13. Takiguchi E, Fukano C, Kimura Y, Tanaka M, Tanida K, et al. (2010) Variation in the 5′-flanking region of the neuropeptide Y2 receptor gene and metabolic parameters. Metabolism 59: 1591–1596.
- 14. Campbell CD, Lyon HMN, Nemesh J, Drake JA, Tuomi T, et al. (2007) Association studies of BMI and Type 2 diabetes in the Neuropeptide Y pathway: A possible role for NPY2R as a candidate gene for type 2 diabetes in men. Diabetes 56: 1460–1467.
- 15. Kamiji MM, Inui A (2007) NPY Y2 and Y4 receptors selective ligands: Promising anti-obesity drugs? Curr Top Med Chem 7: 1734–1742.
- 16. Sato N, Ogino Y, Mashiko S, Ando M (2009) Modulation of neuropeptide Y receptors for the treatment of obesity. Expert Opin Ther Pat 19: 1401–1415.
- 17. Yoo SB, Ryu V, Park EY, Kim B-T, Kang DW, et al. (2011) The arcuate NPY, POMC, and CART expressions responding to food deprivation are exaggerated in young female rats that experienced neonatal maternal separation. Neuropeptides 45: 343–349.
- 18. Shi Y-C, Lin S, Wong IPL, Baldock PA, Aljanova A, et al. (2010) NPY neuron-specific Y2 receptors regulate adipose tissue and trabecular bone but not cortical bone in homeostasis in mice. PLoS One 5: e11361.
- 19. Uddman R, Möller S, Nilsson T, Nyström S, Ekstrand J, et al. (2002) Neuropeptide Y Y1 and neuropeptide Y Y2 receptors in human cardiovascular tissues. Peptides 23: 927–934.
- 20. Wang L, Gourcerol G, Yuan P-Q, Wu S, Million M, et al. (2010) Peripheral peptide YY inhibits propulsive colonic motor function through Y2 receptor in conscious mice. Am J Physiol Gastrointest Liver Physiolog 298: G45–G56.
- 21. Maron JL, Johnson KL, Rocke DM, Cohen MG, Liley AJ, et al. (2010) Neonatal salivary analysis reveals global developmental gene expression changes in the premature infant. Clin Chem 56: 409–16.
- 22. Bustin ST, Benes B, Garson JA, Hellemans J, Huggett J, et al. (2009) The MIQE guidelines: Minimum information of publication of quantitative real-time PCR experiments. Clin Chem 55: 611–622.
- 23. Zimmermann BG, Park NJ, Wong DT (2007) Genomic targets in saliva. Ann NY Acad Sci 1098: 184–91.
- 24. Dietz JA, Johnson KJ, Wick HC, Bianchi DW, Maron JL (2011) Optimal techniques for mRNA extraction from neonatal salivary supernatant. Neonatology 101: 55–60.