Although an adverse early-life environment has been linked to an increased risk of developing the metabolic syndrome, the molecular mechanisms underlying altered disease susceptibility as well as their relevance to humans are largely unknown. Importantly, emerging evidence suggests that these effects operate within the normal range of birth weights and involve mechanisms of developmental palsticity rather than pathology.
Method
To explore this further, we utilised a non-human primate model Macaca fascicularis (Cynomolgus macaque) which shares with humans the same progressive history of the metabolic syndrome. Using microarray we compared tissues from neonates in the average birth weight (50-75th centile) to those of lower birth weight (5-25th centile) and studied the effect of different growth trajectories within the normal range on gene expression levels in the umbilical cord, neonatal liver and skeletal muscle.
Results
We identified 1973 genes which were differentially expressed in the three tissue types between average and low birth weight animals (P < 0.05). Gene ontology analysis identified that these genes were involved in metabolic processes including cellular lipid metabolism, cellular biosynthesis, cellular macromolecule synthesis, cellular nitrogen metabolism, cellular carbohydrate metabolism, cellular catabolism, nucleotide and nucleic acid metabolism, regulation of molecular functions, biological adhesion and development.
Conclusion
These differences in gene expression levels between animals in the upper and lower percentiles of the normal birth weight range may point towards early life metabolic adaptations that in later life result in differences in disease risk.
Background
Clinical, experimental and epidemiological studies have highlighted a link between the early-life environment and the health and well-being of offspring in later life. An adverse maternal environment has been linked to an increased risk of developing metabolic and cardiovascular disorders including type 2 diabetes, obesity, hyperlipidemia, insulin resistance and hypertension [1–7]. An important feature of these studies is that these relationships exist within the normative birth range and do not depend on extremes of birth weight. This has led to the proposal that later life disease risk is the result of maladaptive consequences of plastic mechanisms which would normally be adaptive [8, 9].
It is proposed that developmental plasticity determines the trajectory of development through epigenetic processes such that the fetus attempts to match its later phenotype to the environment [10]. It has been proposed that low birth weight is a marker of a poor early life nutritional environment [11] and thus a smaller fetus is more likely to develop a metabolic capacity appropriate for a low nutrient postnatal environment. But, if faced with a high nutrient environment it is more likely to become obese and insulin resistant [12]. Although, epigenetic processes have been increasingly implicated largely from rodent studies involving nutritional manipulation of the dam [13, 14] the molecular mechanisms underlying altered disease susceptibility are largely unknown. There is also some evidence that these developmental trajectories, and associated long-term gene expression and epigenetic changes can be reversed by the administration of the adipokine leptin to the neonatal rat although the concentrations used were higher than physiological levels [10, 12, 15, 16]. These data suggest that a better understanding of the molecular events associated with impaired early life development may help in designing future intervention strategies.
To identify the possible molecular pathways associated with variations in the fetal environment, we have utilised a non-human primate (NHP) model, the Macaca fascicularis (Cynomolgus macaque) to elucidate whether variations within the normal birth weight range are associated with differential gene expressions patterns. Cynomolgus macaques share with humans the same progressive history of the metabolic syndrome [17] which makes this model directly relevant to humans and importantly, Cynomolgus macaque is a monotocous species in which spontaneous variation in fetal growth rather than experimental manipulation can be investigated. This study therefore we have investigated the effect of spontaneous lower birth weight on gene expression in key tissues (umbilical cord, hepatic tissue and skeletal muscle) from female Cynomolgus macaque neonates.
Methods
Collection of Umblical cords
Sixty-five pregnant Cynomolgus macaque dams, sired naturally by one male, were monitored prior to delivery at the Vietnam Primate Breeding and Development Corporation. After birth, dams were sedated (ketamine-HCl; 7 mg/kg) to facilitate collection of the umbilical cord. The cords were collected and immediately snap-frozen in liquid nitrogen and stored at -80°C for later analyses. Neonates were weighed at birth and promptly returned to the dams. All animal procedures were approved by Nafovanny, subsidiary of the Ministry of Forestry, Vietnam, and performed in accordance with the guidelines set by the national advisory committee for laboratory animal research (NACLAR) of Singapore.
Collection of hepatic and skeletal muscle samples
The normative birth range was assessed from these 65 pregnancies and 8 neonates were selected based on their birth weights to comprise 2 groups: 1) lower birth weight group (LBW); n = 4 classified as those that were within the 5th to 25th birth weight percentile, birth weight range 299-317 g and 2) average birth weight group (ABW); n = 4 classified as those that were within the 50th to 75th birth weight percentile, birth weight range 358-398 g. The normal gestation of Cynomolgus macaque is approximately 155-170 days [18]. We have estimated the gestational age based on early ultrasound measurements (greatest length of the embryo at the time of pregnancy detection) and used those pregnancies where fetuses were within normal distribution for full term Cynomolgus macaques [18]. On postnatal day 5, neonates were sedated with an intramuscular injection of ketamine-HCl (15 mg/kg), and exsanguinated under anesthesia. Liver and skeletal muscle (biceps femoris) were collected and immediately snap frozen in liquid nitrogen and stored at -80°C for later analyses
Preparation of Total RNA
Total RNA was isolated from umbilical cords and neonatal tissues using TRIzol reagent according to the manufacturer's instructions (Invitrogen). RNA integrity was confirmed by bio analyzer 2100 (Agilent Technologies, Santa Clara, USA). An RNA Integrity Number (RIN) value of 7.5 above was considered acceptable and used in further experiments.
Microarray analysis
For microarray analysis, RNA from 6 groups of samples (Cord: ABW and LBW; Liver: ABW and LBW; skeletal muscle: ABW and LBW) were labeled using QuickAmp 1-color labeling kit (Agilent Technologies) according to manufacturer's protocol. The Cy3 labeled cRNA were subsequently hybridized to Agilent Rhesus Macaque (G2519F-015421) Gene Expression microarray. The Rhesus macaque gene expression microarray used in this study represented 43,803 Rhesus monkey probes synthesized as 60-mers spotted using the Agilent SurePrint technology (Agilent Technologies). The microarrays were scanned with Agilent High resolution Scanner and the images were feature extracted using FE software version 10.5 (Agilent Technologies).
Data analysis was performed using Genespring GX ver10 (Agilent Technologies). The raw signal intensity from each samples is global normalized to 75th percentile and base-line transformed. Probes flagged with present call in at least 75% of the samples in any of the 6 groups were used for subsequent data analysis.
Two-way ANOVA was performed with p value cut-off at 0.05 to identify genes that are differentially expressed in the tissue type and birth weight. Due to limited annotation of Rhesus Monkey genome, the microarray probes are annotated against human genes and ontology. For mapping against the human genome, the probes from the monkey microarray were re-annotated using Agilent eARRAY. The probe sequences were aligned to human genome (hg18) using BLAST based algorithm and the associated human annotation was extracted from the database.
To study the gene expression profile of the birth weight in each of the tissue type, Welch T-Test with p-value cut off of 0.05 and fold change of 1.5X was performed between the ABW and LBW samples in each of the tissue groups.
Complete microarray data is available at the Gene Expression Omnibus (GEO) database under the accession number GSE32069.
Network Analysis
The microarray data was imported into Pathway Studio version 7 (Ariadne Genomics, Rockville, USA) for network analysis. Gene Set Enrichment Analysis (GSEA) was performed on ABW vs LBW in the respective tissue using Kolmogorov-Smirnov algorithm with p-value cut-off at 0.05. In addition, Network Enrichment Analysis (NEA) was performed to identify expression hub of the treatment. Sub network was generated by connecting entities to their expression target network using the Resnet 7 Mammalian database.
Quantitative RT-PCR
Five up-regulated and two down regulated genes were selected for verification using qRT-PCR. The primers were designed using the Cynomolgus if available or Rhesus macaque sequences using the primer 3 software if not available [19]. The gene symbols and the primers are listed in Table 1.
Table 1
Sequences of primers used for qRT-PCR:
PASK F 5' CTACTCCGGGAGCTGCTATC 3'
PASK R5' AGCAGCAGAACAGAGGTGTG 3'
116936 F5' GCACATCTGCCTGAAGTGAA 3'
116936 R5' GAGCAGCTTGTCCAGGAAGT 3'
ADK F5' TGGTGGCTCTACCCAGAACT 3'
ADK R5' CATCTACATGGGCTTCAGCA 3'
ELMO F5' AGCTCTGTGTGGCTTGGTTT 3'
ELMO R5' CGGTGTGAATAACGGAGTCCT 3'
SIX1 F5' GTTTAAGAACCGGAGGCAAA 3'
SIX1 R5' GGAGAGAGTTGGTTCTGCTTG 3'
SLC12A9 F5' GGCTTCAACAGCAGTACCCT 3'
SLC12A9 R5' AAGAGGACAGCAAAGACGCT 3'
RBL1 F5' TAGCCTGACCAACATGGAGA 3'
RBL1 R5' GTTCAAGCAATTCTGCCTCA 3'
Uni18SrRNA F5' AGTCCCTGCCCTTTGTACACA 3'
Uni18SrRNA R5' GATCCGAGGGCCTCACTAAAC 3'
We used skeletal muscle RNA to verify the array. Briefly, total RNA was extracted as mentioned above from four ABW and four LBW neonates and were reverse transcribed using Applied Biosystem's high-capacity cDNA reverse transcription kit using 1 μg of total RNA in a reaction volume of 20 μl. The PCR reactions were carried out in 25 μl of SYBR Green Master Mix with 200 ng of cDNA using 7500 real time PCR system (Applied Biosystems, CA, USA). The comparative Ct method was used to calculate the relative gene expression [20]. 18S RNA was used as the internal control which was validated using the method described in Schmittgen and Livak [20] and found to be stable and consistent.
Results
Microarray analysis
Two-way factorial ANOVA identified 1973 genes which were differentially expressed in the three tissue types between ABW and LBW neonates (P < 0.05). Of these, 1141 genes were up regulated in the LBW samples while 832 genes were down-regulated in the LBW samples compared to ABW (Figure 1).
Figure 1
Hierarchical clustergram of 1973 genes (1141 up regulated in LBW and 832 down regulated in LBW) identified by ANOVA (P < 0.05) in all the three tissues (umbilical cord (C), liver (L) and skeletal muscle (B) analyzed. The relative expression is reflected by the intensity of the color (Green = down regulated, red = up regulated)
Gene expression profiles of umbilical cord, liver and skeletal muscle were also compared using Welch-t-test. There were 250 genes significantly and differently expressed in the liver, 850 genes significantly and differently expressed in the skeletal muscle and 891 genes significantly and differently expressed in cord samples (P < 0.05, >1.5 fold difference) (Figure 2, 3). The top 50 genes whose gene expression levels changed in each tissue based on their fold difference based on Welch T-test are given in Tables 2, 3, 4.
Figure 2
Heat map depicting the differentially expressed genes in skeletal muscle (850 genes), liver (210 genes) and cord (891 genes) (P < 0.05, >= 1.5 fold difference). A. 733 genes (584 genes up regulated in LBW skeletal muscle, 149 genes down regulated in LBW skeletal muscle), B. 19 genes (16 genes up regulated in LBW skeletal muscle and 15 genes up regulated in LBW liver, 3 genes down regulated in LBW skeletal muscle and 4 genes down regulated in liver). C 182 genes (115 genes up regulated in LBW liver and 67 genes down regulated in LBW liver). D 5 genes (2 genes up regulated in LBW cords and 3 genes up regulated LBW liver, 3 genes down regulated in LBW cords and 2 genes down regulated in LBW liver). E. 788 genes (338 genes up regulated in LBW cord and 450 genes down regulated LBW cord). F. 94 genes (60 genes up regulated in LBW skeletal muscle and 22 up regulated in LBW cord, 34 genes down regulated in LBW skeletal muscle and 72 genes down regulated in LBW cord. G. 4 genes (4 genes up regulated in LBW skeletal muscle and liver, 4 genes down regulated in LBW cord).
Figure 3
Venn diagram depicting the differentially expressed genes in skeletal muscle (850 genes), liver (210 genes) and cord (891 genes) t-test unpaired unequal variance, LBWvsABW, p < 0.05, >= 1.5 fold difference.
Table 2
List of top 50 genes significantly differentially regulated in the skeletal muscle based on fold difference (t-test).
Gene Symbol
2Way ANOVA P value
Fold Change in Skeletal muscle, Welch t-test
Regulation in LBW
Fold change Liver, Welch t-test
Regulation in LBW
Fold change in Cords, Welch t-test
Regulation in LBW
Gene name
XM_116936
0.017913306
10.054285
up
1.080386
up
2.185324
up
PREDICTED: Homo sapiens similar to RIKEN cDNA 4832428D23 gene (LOC196541) mRNA [XM_116936]
Homo sapiens chromosome 5 open reading frame 23 (C5orf23), mRNA [NM_024563]
PASK
0.03426426
5.977056
up
2.995584
up
19.17474
up
Homo sapiens PAS domain containing serine/threonine kinase (PASK), mRNA [NM_015148]
WDR8
0.004681234
5.6804295
down
2.978104
down
8.061375
down
Homo sapiens WD repeat domain 8 (WDR8), mRNA [NM_017818]
ADK
1.63E-04
5.639505
down
1.368648
down
9.386349
down
Homo sapiens adenosine kinase (ADK), transcript variant ADK-short, mRNA [NM_001123]
APOBEC3G
0.040698703
4.8567457
down
1.669566
down
1.105857
down
Homo sapiens apolipoprotein B mRNA editing enzyme, catalytic polypeptide-like 3G (APOBEC3G), mRNA [NM_021822]
CCL2
0.002121243
4.757898
down
1.287543
down
2.895307
down
Homo sapiens chemokine (C-C motif) ligand 2 (CCL2), mRNA [NM_002982]
A_01_P013390
0.029869065
4.6248507
down
1.099004
down
1.527467
down
BC044226
0.04677891
4.3998837
down
1.966762
down
1.502299
down
Homo sapiens myosin binding protein H, mRNA [BC044226]
CCL11
0.008992519
4.050988
down
3.178148
down
1.002582
up
Homo sapiens chemokine (C-C motif) ligand 11 (CCL11), mRNA [NM_002986]
POU4F3
0.02459994
3.8642302
up
1.615214
up
1.991698
down
Homo sapiens POU domain, class 4, transcription factor 3 (POU4F3), mRNA [NM_002700]
XR_011794
0.010519872
3.7855186
up
3.637105
up
2.76764
up
PREDICTED: Macaca mulatta similar to poly(A)-specific ribonuclease (PARN)-like domain containing 1 (LOC707835), mRNA [XR_011794]
H1FOO
0.011109326
3.7804163
up
1.404401
up
1.449255
down
Homo sapiens H1 histone family, member O, oocyte-specific (H1FOO), mRNA [NM_153833]
C17orf75
0.001362531
3.7583363
down
3.447702
down
1.443745
down
Homo sapiens chromosome 17 open reading frame 75 (C17orf75), mRNA [NM_022344]
RNF216
0.001650785
3.7307158
down
1.504036
down
6.793567
down
Homo sapiens TRIAD3 protein (TRIAD3), transcript variant 1, mRNA [NM_207111]
GTSF1
0.017040279
3.6632807
up
4.653995
up
2.487166
up
Homo sapiens family with sequence similarity 112, member B (FAM112B), mRNA [NM_144594]
C13orf33
0.006214988
3.5870998
down
2.664488
down
1.061963
up
Homo sapiens chromosome 13 open reading frame 33 (C13orf33), mRNA [NM_032849]
CA2
0.033340864
3.5736616
down
3.483002
down
1.194329
up
Homo sapiens carbonic anhydrase II (CA2), mRNA [NM_000067]
CO645773
0.03704907
3.5384166
down
2.022003
down
1.474969
down
ILLUMIGEN_MCQ_30118 Katze_MMPB Macaca mulatta cDNA clone IBIUW:22572 5' similar to Bases 1 to 42 highly similar to human RARRES3 (Hs.17466), mRNA sequence [CO645773]
Homo sapiens keratin associated protein 10-11 (KRTAP10-11), mRNA. [NM_198692]
SHANK2
0.023280138
3.3547363
up
1.255762
up
1.189415
down
Homo sapiens SH3 and multiple ankyrin repeat domains 2 (SHANK2), transcript variant 1, mRNA [NM_012309]
CCDC81
0.004982437
3.2717588
down
1.533572
down
4.912844
down
Homo sapiens coiled-coil domain containing 81 (CCDC81), mRNA [NM_021827]
CCL11
0.008992519
3.1781478
down
4.050988
down
1.002582
up
Homo sapiens chemokine (C-C motif) ligand 11 (CCL11), mRNA [NM_002986]
CO647386
0.020776557
3.1651435
down
1.030158
down
1.631509
down
ILLUMIGEN_MCQ_40418 Katze_MMPB2 Macaca mulatta cDNA clone IBIUW:21432 5' similar to Bases 185 to 778 highly similar to human CXCL2 (Hs.75765), mRNA sequence [CO647386]
GPR98
0.024752488
3.1570897
down
1.10821
down
3.919889
down
Homo sapiens G protein-coupled receptor 98 (GPR98), transcript variant 1, mRNA [NM_032119]
TMEM59L
5.34E-04
3.137047
up
1.57364
up
1.407988
up
Homo sapiens transmembrane protein 59-like (TMEM59L), mRNA [NM_012109]
Homo sapiens heat-responsive protein 12 (HRSP12), mRNA [NM_005836]
CN643639
2.57E-05
19.177528
down
1.710693
down
1.132373
up
ILLUMIGEN_MCQ_8235 Katze_MMBR Macaca mulatta cDNA clone IBIUW:3333 5' similar to Bases 1 to 682 highly similar to human Unigene Hs.513885, mRNA sequence [CN643639]
PASK
0.03426426
19.17474
up
5.977056
up
2.995584
up
Homo sapiens PAS domain containing serine/threonine kinase (PASK), mRNA [NM_015148]
PREDICTED: Macaca mulatta similar to transcription factor B2, mitochondrial (LOC710669), mRNA [XR_011737]
CP
1.42E-04
13.285389
down
1.122256
down
1.274657
down
Homo sapiens ceruloplasmin (ferroxidase) (CP), mRNA [NM_000096]
TOLLIP
1.59E-04
12.683504
down
1.156986
down
1.00844
down
Homo sapiens toll interacting protein (TOLLIP), mRNA [NM_019009]
DARS
3.99E-05
12.632657
down
3.152794
down
1.10567
down
Homo sapiens aspartyl-tRNA synthetase (DARS), mRNA [NM_001349]
CKLF
4.24E-04
12.476327
down
1.130294
down
1.064736
down
Homo sapiens chemokine-like factor (CKLF), transcript variant 1, mRNA [NM_016951]
STT3B
1.13E-04
12.457355
down
2.027118
down
1.030927
down
Homo sapiens STT3, subunit of the oligosaccharyltransferase complex, homolog B (S. cerevisiae) (STT3B), mRNA [NM_178862]
TXNDC11
4.40E-05
12.377014
down
1.525444
down
1.011564
down
Homo sapiens thioredoxin domain containing 11 (TXNDC11), mRNA [NM_015914]
CR603105
9.80E-04
12.209639
down
1.284419
down
1.155491
down
full-length cDNA clone CS0DF006YN22 of Fetal brain of Homo sapiens (human) [CR603105]
XR_010672
5.09E-05
11.634537
down
1.602301
down
1.059423
up
PREDICTED: Macaca mulatta similar to Molybdenum cofactor synthesis protein 2 large subunit (Molybdopterin synthase large subunit) (MPT synthase large subunit) (MOCS2B) (MOCO1-B) (LOC703049), mRNA [XR_010672]
GDE1
0.03917646
11.536806
down
1.190901
up
1.20255
up
Homo sapiens membrane interacting protein of RGS16 (MIR16), mRNA [NM_016641]
CFHR2
2.07E-04
11.37657
down
1.208862
down
1.032327
down
Homo sapiens complement factor H-related 2 (CFHR2), mRNA [NM_005666]
RPL30
0.03067462
11.374878
down
1.173423
up
1.096937
up
Homo sapiens ribosomal protein L30 (RPL30), mRNA [NM_000989]
XM_495885
0.002429009
11.322625
down
1.165255
down
1.156577
down
PREDICTED: Homo sapiens similar to ribosomal protein S12 (LOC440055), mRNA [XM_495885]
Homo sapiens angio-associated, migratory cell protein (AAMP), mRNA [NM_001087]
ESF1
1.02E-07
9.745275
down
1.620818
down
1.008674
down
Homo sapiens ESF1, nucleolar pre-rRNA processing protein, homolog (S. cerevisiae) (ESF1), mRNA [NM_016649]
DOCK7
0.01799708
9.74066
down
1.174452
up
1.171709
up
Homo sapiens dedicator of cytokinesis 7 (DOCK7), mRNA [NM_033407]
DDX3Y
4.78E-06
9.687978
down
1.356089
down
1.134329
down
Homo sapiens DEAD (Asp-Glu-Ala-Asp) box polypeptide 3, Y-linked (DDX3Y), mRNA [NM_004660]
XIAP
3.65E-05
9.587557
down
1.882389
down
1.280735
up
Homo sapiens baculoviral IAP repeat-containing 4 (BIRC4), mRNA [NM_001167]
TRPC4AP
9.51E-05
9.4549885
down
1.340843
down
1.07043
up
Homo sapiens transient receptor potential cation channel, subfamily C, member 4 associated protein (TRPC4AP), transcript variant 1, mRNA [NM_015638]
ADK
1.63E-04
9.386349
down
5.639505
down
1.368648
down
Homo sapiens adenosine kinase (ADK), transcript variant ADK-short, mRNA [NM_001123]
NM_0010022 92
3.92E-04
9.258961
down
1.016391
up
1.220083
down
Homo sapiens chromosome 1 open reading frame 139 (C1orf139), transcript variant 2, mRNA. [NM_001002292]
Of the 250 genes which were differently expressed in the liver, 182 genes were unique to the liver (115 genes up regulated in the LBW group and 67 genes down-regulated in the LBW group). There were 19 genes which were significantly and differently expressed between liver and skeletal muscle (16 up regulated in LBW skeletal muscle and 15 genes up regulated in liver and 3 down-regulated in the LBW skeletal muscle and 4 down regulated in LBW liver). There were 5 genes which are significantly and differently expressed between liver and cord (3 up regulated in LBW liver and 2 genes up regulated in LBW cord and 2 down regulated in LBW liver and 3 down regulated in the cord).
Of the 850 genes significantly and differently expressed in skeletal muscle, 733 genes were specific to the skeletal muscle; i.e. showed altered regulation only in the skeletal muscle (584 genes up regulated in the LBW samples and 149 genes down regulated in the LBW group). There were 94 genes which were significantly and differently expressed between skeletal muscle and cord (60 up regulated in LBW skeletal muscle and 22 genes up regulated in LBW cord and 34 genes down regulated in the LBW skeletal muscle and 72 genes down regulated in the LBW cord).
Of the 891 genes significantly and differently expressed in umbilical cord; 788 genes showed altered regulation only in umbilical cord (338 genes up regulated and 450 genes down regulated in LBW group). There were 4 genes which are significantly and differently expressed between liver, skeletal muscle and umbilical cord (4 genes up regulated in the LBW skeletal muscle and liver while the same four genes were down regulated in LBW cord) Figure 2, 3.
Functional classification of genes
Gene ontology was used to classify genes based on functional significance. The main component of the Gene Ontology (GO) annotation taken into consideration was metabolic function. Genes were classified into fifteen functional categories: Cellular lipid metabolic process (43 genes), Cellular biosynthetic process (95 genes), Cellular macromolecule synthesis (222 genes), Cellular nitrogen compound metabolic process (10 genes), Cellular carbohydrate metabolic process (6 genes), Cellular catabolic process (9 genes), Nucelobase, Nucleoside, nucleotide and nucleic acid metabolic process (216 genes), Other cellular metabolic process (29 genes), Other metabolic process (36 genes), Transport (141 genes), Regulation of molecular functions (28 genes), Biological adhesion (27 genes), Developmental process (74 genes) Other biological processes (252 genes) and Non classified (795) with p-value of (p > 0.05). The genes enriched for each GO term were further classified into the number of genes up regulated or down regulated in each tissue with a fold difference of 1.5 or greater (Table 5) (Additional file 1).
Table 5
Gene ontology classification to group genes using Genespring ver10 (Agilent Tech, Santa Clara) of similar functional families.
Skeletal muscle
Liver
Cord
Up regulated
Down regulated
Up regulated
Down regulated
Up regulated
Down regulated
Cellular lipid metabolic process
15
6
4
4
13
11
Cellular biosynthesis process
8
14
6
4
16
40
Cellular macromolecule synthesis
21
32
19
14
45
63
Cellular nitrogen metabolic process
4
1
4
0
7
1
Cellular carbohydrate metabolic process
2
0
0
0
2
1
Cellular catabolic process
1
2
0
0
2
1
Nucleobase, Nucleoside, nucleotide and nucleic acid metabolic process
19
17
15
4
66
43
Other cellular metabolic process
3
2
1
3
3
15
Other metabolic process
5
7
5
2
9
11
Transport
30
17
21
14
39
26
Regulation of molecular functions
4
2
1
1
12
8
Biological adhesion
6
8
4
5
8
10
Developmental process
19
11
7
4
20
16
Other biological process
38
38
25
19
65
59
Quantitative RT-PCR
To validate the micoarray results we carried out quantitative RT-PCR (qRT-PCR) using the same RNA samples in the microarray analysis. We selected seven genes (a novel gene XM_116936, PAS domain containing serine/threonine kinase (PASK), Adenisine kinase (ADK) transcript variant ADK-short, ELMO/CED-12 domain containing 1 (ELMOD1), Sine oculins homeobox homolog 1 (SIX1) Retinoblastoma like-1 (RBL1) and Solute carrier family 12 (potassium/chloride transporters) member 9 (SLC12A9) for this validation of the array using skeletal muscle cDNA. Of these 5 genes were significantly up regulated in the array and 2 genes were significantly down regulated in the array. Our results from the qPCR complement our results from the microarray (Table 6). The fold differences along with the values which derived from the microarray are presented in Table 6.
Table 6
Verification of seven genes from the microarray using Real-time RT-PCR analysis in skeletal muscle.
Gene Symbol and description
2 way ANOVA p-value (Birth weight)
Microarray Fold change (t-Test)
Regulation in LBW
qPCR-Fold Change
XM_116936
0.017913306
10.054285
Up regulated
PREDICTED: Homo sapiens similar to RIKEN cDNA 4832428D23 gene
4.780893
PASK:
0.03426426
5.977056
Up regulated
Homo sapiens PAS domain containing serine/threonine kinase
14.55481
ADK:
1.63E-04
-5.639505
Down regulated
Homo sapiens adenosine kinase, (transcript variant ADK-short)
-3.57235
ELMOD1:
Homo sapiens ELMO/CED-12 domain containing 1.
0.011425177
-1.7373136
Down regulated
-1.71015
SIX1:
Homo sapiens sine oculis homeobox homolog 1 (Drosophila).
0.045728132
1.1253903
Up in low birth weight
1.246574
RBL1:
Homo sapiens retinoblastoma-like 1 (p107), (transcript variant 1)
0.011705314
1.5320477
Up regulated
3.023726
SLC12A9:
Homo sapiens solute carrier family 12 (potassium/chloride transporters), member 9.
1.31E-05
1.3670695
Up regulated
1.694303
Discussion
In the present study, we have identified genes involved in key metabolic signaling pathways in three tissue types in a non-human primate model, that were differentially expressed according to the birth weight of the animal. Importantly, this differential expression was across the normal birth weight spectrum and therefore likely to represent adaptive pathways that the fetuses uses to predict its postnatal environment. The identification of novel signaling pathways that appear to be regulated by the early life environment is a key step in designing future experimental paradigms to understand the association between birth weight and disease risk. Metabolic disease particularly, has been strongly with early life adversity [21, 22]. Our data begin to shed light on the key signaling pathways that are vulnerable to subtle changes in the early life environment.
The strength of our study, despite its small size, is that we have focused on infants whose growth was not experimentally manipulated but lay within normal birth weight range. Many experimental models have manipulated pregnancy in an effort to produce fetal growth restriction. Such studies have shown that offspring which are born growth restricted catch-up in growth with their normally nourished counterparts and in adulthood are obese, hypertensive, hyperinsulinemic, leptin resistant and display sedentary behavior [23–25]. Investigations into underlying mechanisms and the determination of gene expression levels that may explain these altered phenotypes have produced conflicting results [26–28] which may reflect variations in the model systems used and the gender of the animals used [28]. Taken together, although these studies have established the link between early life nutritional adversity to later pathophysiology, there are limitations in the interpretation of rodent studies in development as applied to humans.
In the present study we aimed to study the molecular associates of growth variation within the normative range and exclude pathology. This is because the growing literature on developmental outcomes highlights that the importance of variation in risk is associated with non-pathological developmental environments. Accordingly we studied relatively small infants born between the 5th and 25th centile but excluded the smallest neonates, which may reflect obstetrical abnormalities. These infants were compared to infants in the middle of the normative range (50th-75th centile) and accordingly we excluded infants who may have had macrosomia as a result of the mother's being over-nourished by being maintained in captivity. Thus we are confident that we have excluded pathological influences and demonstrated that within the normative range patterns of gene expression may vary considerably with variable birth weights. Indeed we found a number of genes with more than a 10 fold shift in expression levels. There are important implications to this observation. Historically, experimental and epidemiological focus has been on the extremes of birth weight (either small for gestational age or large for gestational age) and there has been little focus within the normal birth weight range continuum. What is evident from the present study is that relatively small changes in the birth phenotype may be associated with profound changes in molecular physiology. In turn this also suggests the presence of highly evolved processes by which the fetus can adjust its development in response to subtle cues from mother [29].
The Cynomolgus, as in the human, has monotocous pregnancies with haemochorial placentae; they have omnivorous diets and monogastric digestion. They also share with humans the same progressive history of the metabolic syndrome [17]. We have identified alterations in the levels and expression patterns of a number of genes involved in different metabolic processes including cellular lipid metabolism, cellular biosynthesis, cellular macromolecule synthesis, cellular nitrogen metabolism, cellular carbohydrate metabolism, cellular catabolism, nucleotide and nucleic acid metabolism, biological adhesion and development.Recently, transcriptional profiling in rats subjected to gestational under nutrition was performed in young adult male rats where 249 genes were shown to be differentially expressed in the liver [28]. We compared the genes which are significantly altered in our array with those identified in the rat array study and have identified twelve similar or closely related genes from those identified in the rat: Tribbles homolog 2 (Trib2); 3-hydroxyanthanillate dioxygenase (Haao), transmembrane serine protease 6 (Tmmprss6); Thioredoxin domain containing10 (Txndc10); tralation initiation factor 4A3 (Eifa3); Ribosomal protein L31 (Rpl31); Danse 1-like 2 (Dnase1l2); Quinolate phosphoribosyl transferase (Qprt); general transcription factor IIa 2 (Gtf2a3); General transcription factor II H 3 [28]. Only four of these genes (Trib2, Trip10, EIF4A3 and Dnase1l2) were altered in the same direction in the livers of LBW macaques as in the rat array. We have also compared our observations with those identified from LBW term placentas by McCarthy and colleagues [30] and found similarities in expression changes in the genes Procollagen-lysine, 2-oxoglutarate 5-dioygense (PLOD2); Soluble interleukin-1 receptor accessory protein (IL1RAP); Solute carrier family 2 (facilitated glucose transporter) member3 (SLC2A3); Myosin V1(MYH6); Ribosomal protein S6(RPS6) and Latexin (LXN). The Tribbles homolog 2 (Trib2) is also increased in the LBW infants and suggests another possible way in way insulin/IGF-1 signalling might be impaired during development. Tribbles belongs to a family of kinase-like proteins and are reported to be negative regulators of Akt, the principle target of insulin signaling [31, 32].
Studies using animal models such as rodents to understand the developmental origins of metabolic diseases have shown that epigenetic changes in genes correlates with expression changes including metabolic enzymes such as PEPCK, transcriptional factors such as PPARα which regulate fat metabolism and factors associated with insulin action (e.g. PI3 kinase, PKC-ζ), the key regulatory genes which are responsible for bringing these changes are not known [15, 33]. One aim in our study was to identify whether there were shifts within the expression of key early regulators and from the array we have identified one such key regulatory gene the PAS Kinase (PASK), an evolutionarily conserved PAS domain containing serine/threonine kinase whose expression is altered as a result of adverse early developmental conditions. PAS domains are evolutionarily conserved and appear from archaea, bacteria to eukaryotes and are present in many signaling proteins where they act as signal sensor domain [34]. The PASK gene, whereby expression was up regulated in the LBW animals, has been shown to be a metabolic sensor based on mice knock-out studies; mice lacking PASK are resistant to high-fat induced obesity, hepatic steatosis and are resistant to insulin [35].
Conclusions
In summary, this paper has identified significant variation in gene expression in multiple tissues in primate newborns of different growth trajectories but within the normative range of birth size. Further detailed analyses may improve our understanding of how alterations in such genes due to adverse early life environment predisposes towards metabolic syndrome. These data give strength to the hypothesis that developmental plasticity operating within the normative range of birth weights can influence metabolic and other physiological systems in a way that might have later health consequences. It emphasizes that the concept of developmental programming need not involve pathological changes in growth trajectories to have molecular and presumably functional consequences.
Declarations
Acknowledgements
BSE, RK, PDG, KC, SM are supported by Agency for Science, Technology and research (Singapore). PDG, DMS, MHV are funded by the National Research Centre for Growth and Development, University of Auckland (New Zealand).
Authors’ Affiliations
(1)
Growth, Development and Metabolism Programme, Singapore Institute for Clinical Sciences, Brenner Centre for Molecular Medicine
(2)
Liggins Institute and the National Research Centre for Growth and Development, The University of Auckland
(3)
Division of Molecular Genetics & Cell Biology, School of Biological Sciences, Nanyang Technological University
(4)
Department of Anatomy, Faculty of medicine and Health Sciences, UAE University, Tawam Medical Campus
References
Barker DJ, Winter PD, Osmond C, Margetts B, Simmonds SJ: Weight in infancy and death from ischaemic heart disease. Lancet. 1989, 2 (8663): 577-80.View ArticlePubMedGoogle Scholar
Barker DJP: The fetal and infant origins of adult disease, London. British Medical Journal Publishing. 1993, 1-343.Google Scholar
Barker DJ, Gluckman PD, Godfrey KM, Harding JE, Owens JA, Robinson JS: Fetal nutrition and cardiovascular disease in adult life. Lancet. 1993, 341 (8850): 938-41. 10.1016/0140-6736(93)91224-A.View ArticlePubMedGoogle Scholar
Phillips P, Wilson D, Taylor A, Esterman A, Wakefield M: Cardiovascular risk factors in South Australians with diabetes. Aust J Public Health. 1994, 18 (4): 445-9.View ArticlePubMedGoogle Scholar
Curhan GC, Chertow GM, Willett WC, Spiegelman D, Colditz GA, Manson JE, Speizer FE, Stampfer MJ: Birth weight and adult hypertension and obesity in women. Circulation. 1996, 94 (6): 1310-5.View ArticlePubMedGoogle Scholar
Curhan GC, Willett WC, Rimm EB, Spiegelman D, Ascherio AL, Stampfer MJ: Birth weight and adult hypertension, diabetes mellitus, and obesity in US men. Circulation. 1996, 94 (12): 3246-50.View ArticlePubMedGoogle Scholar
Burdge GC, Lillycrop KA: Nutrition, epigenetics, and developmental plasticity: implications for understanding human disease. Annu Rev Nutr. 2010, 30: 315-39. 10.1146/annurev.nutr.012809.104751.View ArticlePubMedGoogle Scholar
Gluckman PD, Hanson MA, Bateson P, Beedle AS, Law CM, Bhutta ZA, Anokhin KV, Bougnères P, Chandak GR, Dasgupta P, Smith GD, Ellison PT, Forrester TE, Gilbert SF, Jablonka E, Kaplan H, Prentice AM, Simpson SJ, Uauy R, West-Eberhard MJ: Towards a new developmental synthesis: adaptive developmental plasticity and human disease. Lancet. 2009, 373 (9675): 1654-7. 10.1016/S0140-6736(09)60234-8.View ArticlePubMedGoogle Scholar
Gluckman PD, Hanson MA: Living with the past: evolution, development, and patterns of disease. Science. 2004, 305 (5691): 1733-6. 10.1126/science.1095292.View ArticlePubMedGoogle Scholar
Gluckman PD, Hanson MA, Beedle AS: Early life events and their consequences for later disease: a life history and evolutionary perspective. Am J Hum Biol. 2007, 19 (1): 1-19. 10.1002/ajhb.20590.View ArticlePubMedGoogle Scholar
Hales CN, Barker DJ: The thrifty phenotype hypothesis. Br Med Bull. 2001, 60: 5-20. 10.1093/bmb/60.1.5.View ArticlePubMedGoogle Scholar
Gluckman PD, Hanson MA, Cooper C, Thornburg KL: Effect of in utero and early-life conditions on adult health and disease. N Engl J Med. 2008, 359 (1): 61-73. 10.1056/NEJMra0708473.View ArticlePubMedPubMed CentralGoogle Scholar
Burdge GC, Hanson MA, Slater-Jefferies JL, Lillycrop KA: Epigenetic regulation of transcription: a mechanism for inducing variations in phenotype (fetal programming) by differences in nutrition during early life?. Br J Nutr. 2007, 97 (6): 1036-46. 10.1017/S0007114507682920.View ArticlePubMedPubMed CentralGoogle Scholar
Lillycrop KA, Slater-Jefferies JL, Hanson MA, Godfrey KM, Jackson AA, Burdge GC: Induction of altered epigenetic regulation of the hepatic glucocorticoid receptor in the offspring of rats fed a protein-restricted diet during pregnancy suggests that reduced DNA methyltransferase-1 expression is involved in impaired DNA methylation and changes in histone modifications. Br J Nutr. 2007, 97: 1064-73. 10.1017/S000711450769196X.View ArticlePubMedPubMed CentralGoogle Scholar
Vickers MH, Gluckman PD, Coveny AH, Hofman PL, Cutfield WS, Gertler A, Breier BH, Harris M: The effect of neonatal leptin treatment on postnatal weight gain in male rats is dependent on maternal nutritional status during pregnancy. Endocrinology. 2008, 149 (4): 1906-13. 10.1210/en.2007-0981.View ArticlePubMedGoogle Scholar
Wagner JE, Kavanagh K, Ward GM, Auerbach BJ, Harwood HJ, Kaplan JR: Old world nonhuman primate models of type 2 diabetes mellitus. ILAR J. 2006, 47 (3): 259-71.View ArticlePubMedGoogle Scholar
Tarantal AF, HendrickX AG: Charecterization of prenatal growth and development in the Crab-eating macaque (Macaca fasicularis) by ultrasound. The Anatomical record. 1988, 222: 177-184. 10.1002/ar.1092220210.View ArticlePubMedGoogle Scholar
Rozen S, Skaletsky H: Primer3 on the WWW for general users and for biologist programmers. Methods mol Biol. 2000, 132: 365-386.PubMedGoogle Scholar
Schmittgen TD, Livak KJ: Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc. 2008, 3 (6): 1101-8. 10.1038/nprot.2008.73.View ArticlePubMedGoogle Scholar
Gluckman PD, Hanson MA: Maternal constraint of fetal growth and its consequences. Semin Fetal Neonatal Med. 2004, 9 (5): 419-25. 10.1016/j.siny.2004.03.001.View ArticlePubMedGoogle Scholar
Godfrey KM, Gluckman PD, Hanson MA: Developmental origins of metabolic disease: life course and intergenerational perspectives. Trends Endocrinol Metab. 2010, 21 (4): 199-205. 10.1016/j.tem.2009.12.008.View ArticlePubMedGoogle Scholar
Vickers MH, Breier BH, Cutfield WS, Hofman PL, Gluckman PD: Fetal origins of hyperphagia, obesity, and hypertension and postnatal amplification by hypercaloric nutrition. Am J Physiol Endocrinol Metab. 2000, 279 (1): E83-7.PubMedGoogle Scholar
Vickers MH, Reddy S, Ikenasio BA, Breier BH: Dysregulation of the adipoinsular axis-a mechanism for the pathogenesis of hyperleptinemia and adipogenic diabetes induced by fetal programming. J Endocrinol. 2001, 170 (2): 323-32. 10.1677/joe.0.1700323.View ArticlePubMedGoogle Scholar
Vickers MH, Breier BH, McCarthy D, Gluckman PD: Sedentary behavior during postnatal life is determined by the prenatal environment and exacerbated by postnatal hypercaloric nutrition. Am J Physiol Regul Integr Comp Physiol. 2003, 285 (1): R271-3.View ArticlePubMedGoogle Scholar
Lillycrop KA, Phillips ES, Jackson AA, Hanson MA, Burdge GC: Dietary protein restriction of pregnant rats induces and folic acid supplementation prevents epigenetic modification of hepatic gene expression in the offspring. J Nutr. 2005, 135 (6): 1382-6.PubMedGoogle Scholar
Lillycrop KA, Rodford J, Garratt ES, Slater-Jefferies JL, Godfrey KM, Gluckman PD, Hanson MA, Burdge GC: Maternal protein restriction with or without folic acid supplementation during pregnancy alters the hepatic transcriptome in adult male rats. Br J Nutr. 2010, 103 (12): 1711-9. 10.1017/S0007114509993795.View ArticlePubMedGoogle Scholar
Morris TJ, Vickers M, Gluckman P, Gilmour S, Affara N: Transcriptional profiling of rats subjected to gestational undernourishment: implications for the developmental variations in metabolic traits. PLoS One. 2009, 4 (9): e7271-10.1371/journal.pone.0007271.View ArticlePubMedPubMed CentralGoogle Scholar
Gluckman PD, Hanson MA, Morton SM, Pinal CS: Life-long echoes--a critical analysis of the developmental origins of adult disease model. Biol Neonate. 2005, 87 (2): 127-39. 10.1159/000082311.View ArticlePubMedGoogle Scholar
McCarthy C, Cotter FE, McElwaine S, Twomey A, Mooney EE, Ryan F, Vaughan J: Altered gene expression patterns in intrauterine growth restriction: potential role of hypoxia. Am J Obstet Gynecol. 2007, 196 (1): 70.e1-6. 10.1016/j.ajog.2006.08.027.View ArticleGoogle Scholar
Du K, Herzig S, Kulkarni RN, Montminy M: TRB3: a tribbles homolog that inhibits Akt/PKB activation by insulin in liver. Science. 2003, 300 (5625): 1574-7. 10.1126/science.1079817.View ArticlePubMedGoogle Scholar
Hegedus Z, Czibula A, Kiss-Toth E: Tribbles: a family of kinase-like proteins with potent signalling regulatory function. Cell Signal. 2007, 19 (2): 238-50. 10.1016/j.cellsig.2006.06.010.View ArticlePubMedGoogle Scholar
Godfrey KM, Lillycrop KA, Burdge GC, Gluckman PD, Hanson MA: Epigenetic mechanisms and the mismatch concept of the developmental origins of health and disease. Pediatr Res. 2007, 61: 5R-10R. 10.1203/pdr.0b013e318045bedb.View ArticlePubMedGoogle Scholar
Rutter J, Michnoff CH, Harper SM, Gardner KH, McKnight SL: PAS kinase: an evolutionarily conserved PAS domain-regulated serine/threonine kinase. Proc Natl Acad Sci USA. 2001, 98 (16): 8991-6. 10.1073/pnas.161284798.View ArticlePubMedPubMed CentralGoogle Scholar
Hao HX, Cardon CM, Swiatek W, Cooksey RC, Smith TL, Wilde J, Boudina S, Abel ED, McClain DA, Rutter J: PAS kinase is required for normal cellular energy balance. Proc Natl Acad Sci USA. 2007, 104 (39): 15466-71. 10.1073/pnas.0705407104.View ArticlePubMedPubMed CentralGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
