Affective Disorders as Complex Dynamic Diseases – a Perspective from Systems Biology

Abstract

Understanding mental disorders and their neurobiological basis encompasses the conceptual management of “complexity” and “dynamics”. For example, affective disorders exhibit several fluctuating state variables on psychological and biological levels and data collected of these systems levels suggest quasi-chaotic periodicity leading to use concepts and tools of the mathematics of nonlinear dynamic systems. Regarding this, we demonstrate that the concept of “Dynamic Diseases” could be a fruitful way for theory and empirical research in neuropsychiatry. In a first step, as an example, we focus on the analysis of dynamic cortisol regulation that is important for understanding depressive disorders. In this case, our message is that extremely complex phenomena of a disease may be explained as resulting from perplexingly simple nonlinear interactions of a very small number of variables. Additionally, we propose that and how widely used complex circuit diagrams representing the macroanatomic structures and connectivities of the brain involved in major depression or other mental disorders may be “animated” by quantification, even by using expert-based estimations (dummy variables). This method of modeling allows to develop exploratory computer-based numerical models that encompass the option to explore the system by computer simulations (in-silico experiments). Also inter- and intracellular molecular networks involved in affective disorders could be modeled by this procedure. We want to stimulate future research in this theoretical context.

References

• 1 WHO .(ed) ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines: Clinical Description and Diagnostic Guidelines.. WHO-Geneva; 1992
• 2 Huber MT, Braun HA, Krieg JC. On episode sensitization in recurrent affective disorders: the role of noise.  Neuropsychopharmacology. 2003;  28 (S 01) S13-S20
• 3 Leverich GS, Post RM. Life charting the course of bipolar disorder.  Current Review of Mood and Anxiety Disorders. 1996;  1 48-61
• 4 Belair J, Glass L, an der Heiden U. et al. .(Eds). Dynamical disease – mathematical analysis of human illness. Woodbury, USA: American Institute of Physics; 1995
• 5 Katerndahl D, Ferrer R, Best R. et al . Dynamic patterns in mood among newly diagnosed patients with major depressive episode or panic disorder and normal controls. Prim Care Companion.  J Clin Psychiatry. 2007;  9 (3) 183-187
• 6 Scafetta N. Fractal and diffusion entropy analysis of time series: Theory, concepts, applications and computer codes for studying fractal noises and Lévy walk signals.. Saarbrücken: VDM Verlag Dr. Müller; 2010
• 7 Noble D. The Music of Life: Biology Beyond Genes.. Oxford: Oxford University Press; 2006
• 8 Posener JA, DeBattista C, Williams GH. et al . 24-Hour monitoring of cortisol and corticotropin secretion in psychotic and nonpsychotic major depression.  Arch Gen Psychiatry. 2000;  57 (8) 755-760
• 9 Holsboer F, Ising M. Stress hormone regulation: biological role and translation into therapy.  Annu Rev Psychol. 2010;  61 81-109
• 10 Young EA, Abelson J, Lightman SL. Cortisol pulsatility and its role in stress regulation and health.  FrontNeuroendocrinol. 2004;  25 69-76
• 11 Posener JA, DeBattista C, Veldhuis JD. et al . Process irregularity of cortisol and adrenocorticotropin secretion in men with major depressive disorder.  Psychoneuroendocrinol. 2004 Oct;  29 (9) 1129-1137
• 12 Brown EN, Meehan PM, Dempster AP. A stochastic differential equation model of diurnal cortisol patterns.  Am J Physiol Endocrinol Metab. 2001;  280 (3) E450-E461
• 13 Kitano H. Systems Biology: a brief overview.  Science. 2002;  295 1662-1664
• 14 NIH . Systems Biology. 2007;  http://www.nigms.nih.gov/Initiatives/SysBio/
• 15 Breen G, McGuffin P, Simmons A. Towards “systems psychiatry”.  Rev Bras Psiquiatr. 2008 Jun;  30 (2) 97-98
• 16 Gallinat J, Obermayer K, Heinz A. Systems Neurobiology of the dysfunctional brain.  Pharmacopsychiatry. 2007;  40 (S 01) S40-S44
• 17 Gebicke-Haerter P. Systems biology in molecular psychiatry.  Pharmacopsychiatry. 2008;  41 (S 01) S19-S27
• 18 Tretter F, Albus M. Systems biology and psychiatry – modeling molecular and cellular networks of mental disorders.  Pharmacopsychiatry. 2008;  41 (S 01) S2-S18
• 19 Tretter F, Gallinat J, Müller WE. (Eds) Systems biology and psychiatry: The functional architecture of molecular networks in mental disorders – data and models.  Pharmacopschiatry. 2008;  41 (S 01) S1-S106
• 20 Tretter F, Gebicke-Haerter P, Heinz A. (Eds) Systems biology of addiction.  Pharmacopsychiatry. 2009;  42 (S 01) S1-S152
• 21 Tretter F, Rujescu D, Pogarell O. et al. . (Eds) Systems biology of the synapse.  Pharmacopsychiatry. 2010;  43 (S 01) S1-S98
• 22 Tretter F, Gebicke-Haerter PJ, Mendoza ER. et al. .(Eds) Systems Biology in Psychiatric Research: From High-Throughput Data to Mathematical Modeling.. Weinheim: Wiley-VCH Publisher; 2010
• 23 Noble D. Multilevel modelling in Systems Biology: From Cells to Whole Organs.. In:, Szallasi Z,, Periwal V,, Stelling J, (Eds). System, modelling in Cellular Biology. Cambridge MA: MIT Press; 2006: 297-312
• 24 Mackey MC, Glass L. Oscillations and chaos in physiological control systems.  Science. 1977;  197 287-289
• 25 Mackey MC, an der Heiden U. The dynamics of recurrent inhibition.  J Math Biol. 1984;  19 211-225
• 26 Mackey MC, Milton JG. Dynamical diseases.  Ann NY Acad Sci. 1987;  504 16-32
• 27 Prank K, Kloppstech M, Nowlan SJ. et al . Random Secretion of Growth Hormone in Humans.  Phys Rev Lett. 1996;  77 1909-1911
• 28 Young EA, Abelson J, Lightman SL. Cortisol pulsatility and its role in stress regulation and health.  FrontNeuroendocrinol. 2004;  25 69-76
• 29 an der Heiden U. Chaos in health and disease.. In:, Tschacher W,, Schiepek G,, Brunner EJ, (Eds). Selforganization and clincial psychology. Berlin: Springer; 1992: 55-87
• 30 Bewernick BH, Hurlemann R, Matusch A. et al . Nucleus accumbens deep brain stimulation decreases ratings of depression and anxiety in treatment-resistant depression.  Biol Psychiatry. 2010 Jan 15;  67 (2) 110-116
• 31 Drevets WC, Price JL, Furey ML. Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression.  Brain Struct Funct. 2008 September;  213 (1–2) 93-118
• 32 Mayberg HS. Limbic-cortical dysregulation: a proposed model of depression.  J Neuropsychiatry Clin Neurosci. 1997;  9 471-481
• 33 Mayberg HS. Defining neurocircuits in depression.  Psychiatr Ann. 2006;  36 (4) 259-268
• 34 Mayberg HS. Targeted electrode-based modulation of neural circuits for depression.  J Clin Invest. 2009;  119 717-725 · doi:10.1172/JCI38454
• 35 Monk CS, Klein RG, Telzer EH. et al . Amygdala and nucleus accumbens activation to emotional facial expressions in children and adolescents at risk for major depression.  Am J Psychiatry. 2008;  165 (1) 90-98
• 36 Nestler EJ, Carlezon WA. The Mesolimbic Dopamine Reward Circuit in Depression.  J Biol Psychiatry. 2006;  59 1151-1159
• 37 Price JL, Drevets WC. Neurocircuitry of mood disorders.  Neuropsychopharmacol Rev. 2010;  35 192-216
• 38 Schildkraut JJ. The catecholamine hypothesis of affective disorders: a review of supporting evidence.  Am J Psychiatry. 1965;  122 609-622
• 39 Coppen A. The biochemistry of affective disorders.  Br J Psychiatry. 1967;  113 1237-1264
• 40 Schatzberg AF, Rothschild AJ, Langlais PJ. et al . A corticosteroid/dopamine hypothesis for psychotic depression and related states.  J Psychiatr Res. 1985;  19 (1) 57-64
• 41 Luscher B, Shen Q, Sahir N. The GABAergic deficit hypothesis of major depressive disorder.  Mol Psychiatry. 2010 Nov 16;  (Epub ahead of print)
• 42 Janowsky DS, Overstreet DH. The role of acetylcholine mechanisms in mood disorders.. New York: Raven Press; 1995
• 43 Krystal JH. N-methyl-D-aspartate glutamate receptor antagonists and the promise of rapid-acting antidepressants.  Arch Gen Psychiatry. 2010;  67 1110-1111
• 44 Tretter F. Systemtheorie im klinischen Kontext. Lengerich: Pabst; 2005
• 45 Coyle JT, Duman RS. Finding the intracellular signaling pathways affected by mood disorder treatments.  Neuron. 2003;  38 157-160
• 46 Best J, Reed M, Nijhout HF. Models of dopaminergic and serotonergic signalling.  Pharmacopsychiatry. 2010;  Vol 43 (Supplement) S61-S66
• 47 Qui Z, Miller GW, Voit EO. Computational modelling of synaptic neurotransmission as a tool for assessing dopamine hypotheses of schizophrenia.  Pharmacopsychiatry. 2010;  43 (S 01) S50-S60

Appendix

The model contains two time dependent variables:

x (t)=concentration of ACTH in the blood at time t,

y (t)=concentration of cortisol in the blood at time t.

The rates of change of these two state variables are denoted by

respectively (mathematically speaking these are the derivatives of x and y with respect to t, or the velocities by which these concentrations change).

The fact that the increase of cortical concentration is the greater the greater the ACTH-concentration, is expressed by the equation

where g (x) is a function (in the mathematical sense) such that the larger x, the larger is g (x). An example of such a function is g (x)=dx ^m /(k+x ^m ) with positive constant (with respect to time) parameters d, m, k.

Another constant parameter is τ which denotes the length of the time interval between the release of ACTH (=x) from the pituitary until the release of cortisol (y) from the adrenal gland into the blood. The cortisol molecules decay with a rate constant a, and hence the above equation for the change of y is to be extended to

Not very differently we can propose an equation for the rate of change of ACTH:

However, since cortisol (y) inhibits (both across the hypothalamus and the pituitary) the production of ACTH (x), the function f (y) must be decreasing, i. e., the higher the value of y the lower the value of f (y). An example of such a function is given by

with positive constant parameters c, K, n.

The mathematical model consisting of the two equations (1) and (2) is called in mathematical terms a “nonlinear system of two differential delay equations”. It is nonlinear because the gain functions f and g are nonlinear. It contains the two delays δ and τ, however, without loss of generality δ=0 can be assumed.

The interactive structure described by the equations (1) and (2) allows essentially for two kinds of dynamics: (i) a stable equilibrium, i. e., temporally constant solutions x (t)=constant, y (t)=constant such that after any perturbation of x or y the system returns to this equilibrium; (ii) a stable periodic oscillator (sometimes called limit cycle), i. e., both x (t) and y (t) increase and decrease in a periodic fashion and, moreover, after external perturbations away from this periodicity the system returns to the same periodic behavior (to a stable clock). Whether the situation (i) or alternatively (ii) occurs depends only on the values of the temporally constant parameters of the model.

More complex behavior arises when this oscillatory system is coupled to a second oscillator (“coupled oscillators”). This is definitely the case since it is well known that the activity of the hypothalamus is influenced by the suprachiasmatic nucleus, a tiny region of the brain considered to be the origin of the circadian rhythm. This rhythmic influence is taken into account in the mathematical model by including a periodic function into the differential equation (2):

Now the model consists of equation (3) together with equation (1). From the point of view of systems theory the operational structure of this system is that of 2 coupled oscillators. In [Fig. 2d] a solution of this system is presented. At least qualitatively there is coincidence with the rather irregular (quasi-chaotic) empirical data shown in [Fig. 2c]. Another interpretation of the term A sin (ω t) could be that the pituitary is in itself an autonomous oscillator mimicked by this term. The pattern observable in [Fig. 2d] is just one of a large variety of qualitatively different solution types of System (2)−(3) including different ratios of phase locking and different types of chaos (for more examples see an der Heiden 1992 [29]). Including the effect of the immune system sketched in the diagram of [Fig. 2a] and/or other influences from the brain into the mathematical model presented here would still substantially increase the complexity of its behavior.

Correspondence