Andrey Ptitsyn Sidra medical and Research Center Biological - PowerPoint PPT Presentation

The Structure of Biological Pathways in Time Life works on AC power Andrey Ptitsyn Sidra medical and Research Center

Biological pathways: untangling the hairballs

Mammalian molecular clock • BMAL1 and CLOCK (NPAS2) form heterodimers that act as positive transcriptional regulators • PERIOD (Per1, Per2, Per3) and CRYPTOCHROME (Cry1, Cry2) family members serve as negative transcriptional regulators • downstream targets, such as the albumin D site binding protein (DBP), can further activate transcription while others, such as E4BP4, repress transcription • The serine/threonine kinases, casein kinase Iε (CK1 ε) and glycogen synthase kinase 3β (GSK3 β), phosphorylate BMAL1, PER, and other proteins exposing them for degradathion through ubiquitin/proteasomal pathway

Circadian rhythms • Physiological rhythms respond to environmental cues • Suprachiasmic nucleus (SCN) of the brain is the body’s central oscillator light • SCN responds to light/dark cycle • Daily activity rhythm continues even in total darkness feeding fasting heme treatment • The central circadian oscillator may dexamethasone act through sympathetic outputs and streptozotocin controlled secretion of circulating gene ablation glucocorticoids, melatonin, and other ??? mediators, thereby “synchronizing” the circadian rhythms of the body’s tissues fat and organs

First experiment • Age-matched male Black 6 mice on chow • entrained to 12/12 alternation of lighting • samples collected for every 4h starting from 8am • total of 12 samples collected to represent a 48h expression profile • 3-5 mice are sacrificed at each time point, samples pooled • liver, white fat and brown fat are collected (bone is added later in a similar experiment • entraining conditions are kept throughout data collection period

Algorithms to identify periodicity Raw data (time series expression profile) Pre-processing Phase assignment Fourier transformation Fisher’s g -test Autocorrelation analysis Pt-test Other permutation tests Time domain Frequency domain

Fisher’s g -test DFT            I ( ) , , ,... Y x , x , x ,... x   2  0 1 2 N / 2 1 0 1 2 N 1 N 1   1           i t I ( ) x t e , 0 , N  t 0 500 Based on periodogram 400 Signal to noise ratio 300    200 max I  k k 100 g ,     N 2  0 I k 1 2 3 4 5 k 1 Fisher’s formula produces p -value for significance of oscillation: 1     x   n !          p n 1 P g x  1 1 px  ,     p ! n p !  p 1

Autocorrelation and Phase Assignment       N 1   x x x x  i f 0 R ( f )     N 1  2 x x i 0 15000 10000 5000 0 1423439_at 1 3 5 7 9 11 13 15 17 1423439_at -5000 -10000 -15000 -20000          N 1   2 x x y y     i f 0 y i cos * i   R ( f ) , where      N 1     p x x y y i i 0

Pt-test Original profile 250 200 Permutated periodograms Permutated profiles 150 100 50 0 1 2 3 4 5 6 7 8 9 10 11 12 -50 -100 -150 -200 Original periodogram 180 160 140 120 100 80 60 40 20 0 1 2 3 4 5 Significance is estimated by comparing specific frequency peak in original and multiple randomized periodograms

Analysis of circadian signal in phase continuum (real data) 1.5 1.2 1.5 0.8 1 0.6 1 0.8 1 0.4 0.6 0.2 0.4 0.5 0 0.5 1 2 3 4 5 6 7 8 9 10 11 12 0.2 -0.2 0 0 -0.4 1415803_at 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 0 1415996_at -0.2 -0.6 1 2 3 4 5 6 7 8 9 10 11 12 1416087_at -0.4 -0.5 -0.8 -0.6 -1 -0.5 -0.8 -1 -1.2 -1 -1.5

Heat map 1.5 Visualization 1 0.5 1415803_at 0 1415996_at 1 2 3 4 5 6 7 8 9 10 11 12 1416087_at -0.5 -1 -1.5

Murine transcriptome

Is there a non-oscillating fraction at all? Simulation experiment

Does aging affect the rhythm?

Can we knock out the rhythm? Bray, M., et al., Disruption of the circadian clock within the cardiomyocyte influences myocardial contractile function, metabolism, and gene expression Am J Physiol Heart Circ Physiol 294:1036-1047, 2008.

Reanalysis vs. original report

Mosquito Dbt Pdp P+ Per dCLK tr - Dbt tr + Per Tim Dbt Per Cyc Tim Tim Vri P+ Sgg Cytoplasm Nucleus 1.5 1 0.5 Cytochrome oxidase 0 NADH Dehydrogenase 1 2 3 4 5 6 7 8 9 10 11 12 -0.5 -1 -1.5

Are those rhythms only driven by light? Plant Human Yeast Mosquito Mouse NADH reductase/cytochrome oxidase

Oscillation in Biopathways: Life works on AC power Glucocorticoid receptor pathway, murine bone GILZ 1.5 1 0.5 0 1 2 3 4 5 6 7 8 9 10 11 12 -0.5 -1 -1.5

Life on AC power Leptin Signaling Pathway-Liver

Life on AC power Leptin Signaling Pathway-Brown Fat

Life on AC power Leptin Signaling Pathway-White Fat

Mathematical model 𝑒𝑜 1 𝑒𝑜 1 𝑒𝑢 = 𝑞𝑠 𝑞 − 𝑠 𝑒1 ; 𝑒𝑢 = 𝑞𝑏 sin(𝜕𝑢 + 𝛽 1 ) − 𝑐 sin(𝜕𝑢 + 𝛽 2 ); 𝑒𝑜 2 𝑒𝑜 2 𝑒𝑢 = 1 − 𝑞 𝑠 𝑞 − 𝑠 𝑒2 ; 𝑒𝑢 = 1 − 𝑞 𝑏 sin(𝜕𝑢 + 𝛽 1 ) − 𝑑 sin(𝜕𝑢 + 𝛽 3 ); 𝑜 1 (𝑢) = 𝐵 cos(𝜕𝑢 + 𝛾 1 ); 𝑜 2 (𝑢) = 𝐶 cos(𝜕𝑢 + 𝛾 2 ); 2 2 𝑞𝑏 𝑐 − 2𝑞𝑏𝑐 𝐵 = − 𝜕 2 cos 𝛽 1 − 𝛽 2 𝜕 𝜕 2 2 1 − 𝑞 2 𝑏 𝑑 − 2(1 − 𝑞)𝑏𝑑 𝐶 = − cos 𝛽 3 − 𝛽 1 𝜕 2 𝜕 𝜕 𝑑 sin 𝛽 3 − (1 − 𝑞)𝑏 sin 𝛽 1 𝑐 sin 𝛽 2 − 𝑞𝑏 sin 𝛽 1 𝛾 2 − 𝛾 1 = 𝑏𝑠𝑑𝑢𝑏𝑜 − 𝑏𝑠𝑑𝑢𝑏𝑜 𝑑 cos 𝛽 3 − (1 − 𝑞)𝑏 cos 𝛽 1 𝑐 cos 𝛽 2 − 𝑞𝑏 cos 𝛽 1 The phase lag between isoforms may have values varying between 0 and 2p. In the middle of this range, when b 2 -b 1 =p the amplitude of n is reduced to 0.

Function of miRNA

Affy target sequences 11101 cacacaagga gccaaacaca gccaataggc agagagttga gggattcacc caggtggcta 11161 caggccaggg gaagtggctg caggggagag acccagtcac tcaggagact cctgagttaa 11221 cactgggaag acattggcca gtcctagtca tctctcggtc agtaggtccg agagcctcca 11281 ggccctgcac agccctccct tctcacctgg ggggaggcag gaggtgatgg agaagccttc 11341 ccatgccgct cacaggggcc tcacgggaat gcagcagcca tgcaattacc tggaactggt 11401 cctgtgttgg ggagaaacaa gttttctgaa gtcaggtatg gggctgggtg gggcagctgt 11461 gtgttggggt ggcttttttc tctctgtttt gaataatgtt tacaatttgc ctcaatcact 11521 tttataaaaa tccacctcca gcccgcccct ctccccactc aggccttcga ggctgtctga 11581 agatgcttga aaaactcaac caaatcccag ttcaactcag actttgcaca tatatttata 11641 tttatactca gaaaagaaac atttcagtaa tttataataa aagagcacta ttttttaatg 11701 aaaaaaaaaa gtgacttgag 1456212_x_at 1455899_x_at 1416576_at 2.5 2 1.5 1.5 1 2 1 1.5 0.5 0.5 0 1 1 2 3 4 5 6 7 8 9 10 11 12 0 0.5 -0.5 1 2 3 4 5 6 7 8 9 10 11 12 -0.5 0 -1 -1 1 2 3 4 5 6 7 8 9 10 11 12 -1.5 -0.5 -1.5 -1 -2 -2 -2.5 -1.5 -2.5

miRNA expression patterns miR-187 Acyl-CoA synthetase miRNA I II Inhibitor of growth Fibroblast growth factor 23 Tumor necrosis factor, alpha-induced protein 3 III IV

Math again Let 𝑦 𝑞 be rate of transcription, 𝑦 𝑒 - rate of mRNA degradation and 𝑦 - abundance of transcript.

Function of miRNA miRNA miRNA Transcription factors Transcription factors

miRNA too early miRNA miRNA Transcription factors Transcription factors

miRNA too late miRNA miRNA Transcription factors Transcription factors

Summary of observations • Oscillation of baseline expression is an immanent property of all genes, not a function of some; • Phase and amplitude of expression are important characteristics of expressed genes and specific to tissues and conditions; • Alternative expression variants can have different oscillatory properties; • miRNAs play important role is oscillatory regulation of gene expression; • Accounting for the oscillatory properties of gene expression is essential for understanding and modeling of biological processes

Acknowledgements • Jeffrey M. Gimble (PBRC) • Sanjin Zvonic (Tulane) • Randall L. Mynatt (PBRC) • Steven A. Conrad (LSU HSC Shreveport) • L. Keith Scott (LSU HSC Shreveport) • Kai Florian Storch (Harvard->McGill) • John Hogenesch (UPenn) • Benjamin Tu (UT Southwestern Medical Branch) • Morey Haymond (BCM) • Eugene Selkov Sr. (Argonne National Laboratory) • Roberto Refinetti (University of South Carolina) • Franz Halberg (Halberg Chronobiology Center, Univ. of Minnesota) • Molly Bray (UAB)