Goodman has previously outlined the evidence implicating retinoids in the etiology of schizophrenia ( Goodman, 1994; Goodman, 1998), and in the current paper expands the focus to include two other members of the nuclear receptor (NR) superfamily—thyroid hormone and (to a lesser extent) estrogen (Palha and Goodman, 2005). The NR superfamily is a phylogenetically ancient system of ligand-activated transcription factors that contributes to fundamental biological processes in metazoans (i.e., from sponges to humans; Escriva et al., 2004). In humans, the best known NRs include those for steroid hormones (glucocorticoid, sex hormones), the seco-steroid vitamin D, thyroid hormone, and the retinoid (vitamin A) family. Evolution has recycled nuclear receptors many times via gene duplication (Escriva et al., 1997), and as a consequence the NR-related pathways underpin a broad sweep of biological functions.
For those interested in generating hypotheses linking NR-related pathways to schizophrenia, the dense cross-talk between members of the NR family and many biological pathways makes for a very wide creative canvas. However, clues from epidemiology can help prioritize NR-related candidates. Sex differences in schizophrenia suggest a role for estrogen (Hafner, 2003; Riecher-Rossler and Seeman, 2002). In particular, the lower incidence of schizophrenia in women (McGrath et al., 2004), the later age of onset, and the better course of illness (Leung and Chue, 2000) implicate sex hormones in this group of disorders. Vitamin D also draws strength from epidemiology, as low prenatal vitamin D may explain the seasonal birth effect (McGrath, 1999), and could contribute to the increased incidence rates in dark-skinned migrants to cold climates (Cantor-Graae and Selten, 2005).
When it comes to exploring NR-related pathways in neuropsychiatric disorders, the real "heavy lifting" involves the translation of hypotheses into more precise descriptive and experimental studies. Hypotheses need to be sharpened by the data, and refined. For example, are there critical windows during the lifespan when disruption of NR-related pathways influences disease risk (e.g., prenatal, peri-onset, all of life)? If ligand concentration influences disease susceptibility, what is the direction of this change (e.g., high or low thyroid hormone, hypo- or hypervitaminosis A)? With respect to genetic susceptibility, should the researcher look for polymorphisms in (a) genes that influence ligand levels; (b) gene coding for the NR; (c) the NR response elements scattered ubiquitously throughout the genome; (d) the thousands of genes that are functionally activated by the NR ligands; or (e) all of the above?
Despite the broad canvas for hypothesis generation, good progress is being made testing these research questions. Gene linkage studies between candidate NR and disease outcomes are a good place to start; however, such studies do not allow for the rejection of NR hypotheses involving disruption of the ligand during development (e.g., low prenatal vitamin D could influence the offspring’s neurocognitive development regardless of maternal or fetal haplotypes).
Transgenic animal models based on NR-related theories of human brain disorders have been very informative—they can help reverse engineer previously hidden layers of complexity (LaMantia, 1999). Disrupting prenatal NR ligand concentration can also help unravel pathways of interest. For example, following up clues from epidemiology, our group has discovered that low prenatal vitamin D alters brain development in the rat (Feron et al., 2005). Adult rats exposed to developmental vitamin D deficiency have enlarged ventricles (Feron et al., 2005) and a behavioral phenotype of interest to schizophrenia research (Becker and Grecksch, 2006; Burne et al., 2004).
Finally, NR-related theories of schizophrenia need clinically based analytic epidemiology. In individuals with schizophrenia, is there any direct evidence of disruption of these pathways? However, if the exposure of interest is during early development (e.g., low maternal thyroid hormone, altered vitamin D or A), then the research community has to rely on banked biological specimens (maternal sera, dried blood spots, etc.). These are rare resources, and limited sample size can result in inconclusive results (McGrath et al., 2003).
Schizophrenia research needs new ideas. In recent years, NRs have provided fertile ground for hypothesis generation. Now is the time to roll up our collective sleeves and examine these candidates in an efficient, but urgent, fashion.
References:Becker A, Grecksch G. Pharmacological treatment to augment hole board habituation in prenatal Vitamin D-deficient rats.
Behav Brain Res. 2006 Jan 7;166(1):177-83. Epub 2005 Sep 26.
Burne TH, Becker A, Brown J, Eyles DW, Mackay-Sim A, McGrath JJ. Transient prenatal Vitamin D deficiency is associated with hyperlocomotion in adult rats.
Behav Brain Res. 2004 Oct 5;154(2):549-55.
Cantor-Graae E, Selten JP. Schizophrenia and migration: a meta-analysis and review.
Am J Psychiatry. 2005 Jan;162(1):12-24. Review.
Escriva H, Bertrand S, Laudet V. The evolution of the nuclear receptor superfamily.
Essays Biochem. 2004;40:11-26. Review.
Escriva H, Safi R, Hanni C, Langlois MC, Saumitou-Laprade P, Stehelin D, Capron A, Pierce R, Laudet V. Ligand binding was acquired during evolution of nuclear receptors.
Proc Natl Acad Sci U S A. 1997 Jun 24;94(13):6803-8.
Feron F, Burne TH, Brown J, Smith E, McGrath JJ, Mackay-Sim A, Eyles DW. Developmental Vitamin D3 deficiency alters the adult rat brain.
Brain Res Bull. 2005 Mar 15;65(2):141-8.
Goodman AB. Retinoid dysregulation as a cause of schizophrenia.
Am J Psychiatry. 1994 Mar;151(3):452-3. No abstract available.
Goodman AB. Three independent lines of evidence suggest retinoids as causal to schizophrenia.
Proc Natl Acad Sci U S A. 1998 Jun 23;95(13):7240-4. Review.
Hafner H. Gender differences in schizophrenia.
Psychoneuroendocrinology. 2003 Apr;28 Suppl 2:17-54.
LaMantia AS. Forebrain induction, retinoic acid, and vulnerability to schizophrenia: insights from molecular and genetic analysis in developing mice.
Biol Psychiatry. 1999 Jul 1;46(1):19-30. Review.
Leung A, Chue P. Sex differences in schizophrenia, a review of the literature.
Acta Psychiatr Scand Suppl. 2000;401:3-38. Review.
McGrath J. Hypothesis: is low prenatal vitamin D a risk-modifying factor for schizophrenia?
Schizophr Res. 1999 Dec 21;40(3):173-7. Review.
McGrath J, Eyles D, Mowry B, Yolken R, Buka S. Low maternal vitamin D as a risk factor for schizophrenia: a pilot study using banked sera.
Schizophr Res. 2003 Sep 1;63(1-2):73-8.
McGrath J, Saha S, Welham J, El Saadi O, MacCauley C, Chant D. A systematic review of the incidence of schizophrenia: the distribution of rates and the influence of sex, urbanicity, migrant status and methodology.
BMC Med. 2004 Apr 28;2:13. Review.
Palha JA, Goodman AB. Thyroid hormones and retinoids: A possible link between genes and environment in schizophrenia.
Brain Res Brain Res Rev. 2005 Nov 30; [Epub ahead of print]
Riecher-Rossler A, Seeman MV. Oestrogens and schizophrenia--introduction.
Arch Women Ment Health. 2002 Nov;5(3):91-2. No abstract available.
View all comments by Alan Mackay-SimView all comments by John McGrath
This review article builds on previous papers hypothesizing a role for thyroid hormone and retinoids in the development of schizophrenia (Goodman, 1998). Retinoids are vitamin A related compounds that act on cells to alter gene expression. In early development, neuronal systems are particularly susceptible to regulation by retinoids, and a key role in defining cells as nerves, nerve cell growth, and connectivity is well established (Maden, 2002; McCaffery et al., 2003). Similarly, thyroid hormone exerts its effects by regulating gene expression and is essential for the regulation of metabolism, growth, and differentiation of the brain. Given this central role in neuronal development, and the numerous neurodevelopmental phenomena associated with schizophrenia, it is perhaps not surprising that thyroid hormone and retinoid-signaling pathways have emerged as important vulnerability targets in schizophrenia. Evidence is also emerging that retinoid-signaling pathways may be important in the function of adult neurons (Crandall et al., 2004; Lane and Bailey, 2005; O' Reilly et al., 2006).
This new article updates the evidence supporting the hypothesis that key components of thyroid hormone and retinoid-signaling pathways are candidate genes that may be involved in the pathology of schizophrenia. Given the prominent role of retinoids in neuronal development, it is perhaps not surprising that mapping of several human chromosomal regions associated with susceptibility to schizophrenia has shown a high coincidence of genes involved in the retinoid signaling cascade at these loci. Deficits in these signaling pathways could cause severe neurodevelopmental deficits, but could also cause more subtle changes in the expression of genes involved in neurotransmission. Of particular interest are the recent studies on altered gene expression in postmortem brain tissues. Several candidate genes that are thought to contribute to increased vulnerability in schizophrenia show altered expression in such studies. That myelination-related genes are differentially regulated in chronic schizophrenia further supports this process as important in the pathology of schizophrenia. Myelination is the process whereby neurons are enveloped in an insulating sheath that permits rapid and effective neuronal signaling. Failures of myelination can severely impair the function of neurons. It is not clear from such postmortem studies which genes are causally related to schizophrenia or which genes might be reactive changes once the disease is manifest. In addition, the finding that some antipsychotic drugs are themselves capable of regulating retinoid-signaling pathways further complicates the interpretation of gene expression studies (Langlois et al., 2001; Ethier et al., 2004).
An important consideration for this hypothesis is whether thyroid hormone or specific retinoids can actually activate the nuclear receptors proposed by Pahla and Goodman, thereby altering gene expression. Although it is tempting to speculate that retinoids and thyroid hormone modulate the expression of a variety of genes through RXR/THR heterodimers, it is important to note that RXR and THR are nonpermissive partners. Nonpermissive nuclear receptor partners can only be activated by the partner receptor’s ligand. In the case of RXR/THR heterodimers, this means that only thyroid hormone can activate gene transcription. RXR ligands, such as 9-cis-retinoic acid or docosahexaenoic acid, cannot (Shulman and Mangelsdorf, 2005). This is thought to allow the THR component of the RXR/THR heterodimer to prevent RXR signaling enabling precise regulation of thyroid hormone target genes (Shulman and Mangelsdorf, 2005).
In addition to inducing gene transcription via RXR/RXR homodimers, 9-cis-retinoic acid can also serve as a ligand for RAR, which can stimulate gene transcription via nonpermissive RAR/RXR heterodimers. However, in vivo, RAR/RXR-mediated gene transcription is believed to be regulated by all-trans-retinoic acid (Soprano et al., 2004) because 9-cis-retinoic acid has been difficult to detect under normal physiological conditions (Horton and Maden, 1995; Kurlandsky et al., 1995; Ulven et al., 2000; Ulven et al., 2001; Werner and DeLuca, 2001). Rather, docosahexaenoic acid, a fatty acid found in the brain, may be the functional RXR ligand in this tissue (de Urquiza et al., 2000; Lengqvist et al., 2004).
While the interaction between thyroid hormone and retinoid signaling pathways is an attractive hypothesis, further work is still required to identify genes that are causally related to schizophrenia, rather than changes in expression that are an adaptation to the disease or pharmacological intervention. There also needs to be more clarity on whether the author’s proposition that retinoids provide a link between genes and environment comes into play early during development or during adulthood.
References:Crandall J, Sakai Y, Zhang J, Koul O, Mineur Y, Crusio WE, McCaffery PJ (2004). 13-cis-retinoic acid suppresses hippocampal cell division and hippocampal-dependent learning in mice. Proc Natl Acad Sci U S A 101: 5111-5116. Abstract
de Urquiza AM, Liu S, Sjöberg M, Zetterström RH, Griffiths W, Sjövall J, Perlmann T (2000). Docosahexaenoic acid, a ligand for the retinoid X receptor in mouse brain. Science. 290: 2140-4. Abstract
Ethier I, Beaudry G, St-Hilaire M, Milbrandt J, Rouillard C, Levesque D (2004). The transcription factor NGFI-B (Nur77) and retinoids play a critical role in acute neuroleptic-induced extrapyramidal effect and striatal neuropeptide gene expression. Neuropsychopharmacology 29: 335-46. Abstract
Goodman AB (1998). Three independent lines of evidence suggest retinoids as causal to schizophrenia. Proc Natl Acad Sci U S A 95: 7240-4. Abstract
Horton C, Maden M (1995). Endogenous distribution of retinoids during normal development and teratogenesis in the mouse embryo. Developmental dynamics : an official publication of the American Association of Anatomists. 202: 312-23. Abstract
Kurlandsky SB, Gamble MV, Ramakrishnan R, Blaner WS (1995). Plasma delivery of retinoic acid to tissues in the rat. The Journal of biological chemistry. 270: 17850-7. Abstract
Lane MA, Bailey SJ (2005). Role of retinoid signaling in the adult brain. Progress in Neurobiology 72: 275-293. Abstract
Langlois MC, Beaudry G, Zekki H, Rouillard C, Levesque D (2001). Impact of antipsychotic drug administration on the expression of nuclear receptors in the neocortex and striatum of the rat brain. Neuroscience 106: 117-28. Abstract
Lengqvist J, Mata De Urquiza A, Bergman AC, Willson TM, Sjövall J, Perlmann T, Griffiths WJ (2004). Polyunsaturated fatty acids including docosahexaenoic and arachidonic acid bind to the retinoid X receptor alpha ligand-binding domain. Molecular & cellular proteomics : MCP. 3: 692-703. Abstract
Maden M (2002). Retinoid signaling in the development of the central nervous system. Nat Rev Neurosci 3: 843-53. Abstract
McCaffery PJ, Adams J, Maden M, Rosa-Molinar E (2003). Too much of a good thing: retinoic acid as an endogenous regulator of neural differentiation and exogenous teratogen. Eur J Neurosci 18: 457-72. Abstract
O' Reilly K, Shumake J, Gonzalez-Lima F, Lane M, Bailey S (2006). Chronic Administration of 13-Cis-Retinoic Acid Increases Depression-Related Behavior in Mice.
Neuropsychopharmacology. 2006 Jan 4; [Epub ahead of print]
Shulman AI, Mangelsdorf DJ (2005). Retinoid x receptor heterodimers in the metabolic syndrome. The New England journal of medicine. 353: 604-15. Abstract
Soprano DR, Qin P, Soprano KJ (2004). Retinoic acid receptors and cancers. Annual Review of Nutrition 24: 201-221. Abstract
Ulven SM, Gundersen TE, Sakhi AK, Glover JC, Blomhoff R (2001). Quantitative axial profiles of retinoic acid in the embryonic mouse spinal cord: 9-cis retinoic acid only detected after all-trans-retinoic acid levels are super-elevated experimentally. Developmental dynamics : an official publication of the American Association of Anatomists. 222: 341-53. Abstract
Ulven SM, Gundersen TE, Weedon MS, Landaas VO, Sakhi AK, Fromm SH, Geronimo BA, Moskaug JO, Blomhoff R (2000). Identification of endogenous retinoids, enzymes, binding proteins, and receptors during early postimplantation development in mouse: important role of retinal dehydrogenase type 2 in synthesis of all-trans-retinoic acid. Developmental biology. 220: 379-91. Abstract
Werner EA, DeLuca HF (2001). Metabolism of a physiological amount of all-trans-retinol in the vitamin A-deficient rat. Arch Biochem Biophys 393: 262-70. Abstract
View all comments by Sarah J. BaileyView all comments by Michelle A. Lane
This paper by Huang, Bahn, and colleagues makes for very interesting reading and provides an early glimpse into the future of proteomic studies of schizophrenia and other mental disorders. Although some interesting new leads have been provided regarding particular proteins and peptides, these will need replication, as the authors themselves acknowledge. Thus, we should not get caught up in those details at this time, but rather appreciate this work for its greater contribution, which is in the modern theoretical framework that drives the study.
First and foremost, it is refreshing to see a focus on a syndrome, such as psychosis, rather than traditional focus on DSM-based diagnostic boundaries. This approach is one that our group has also endorsed in recent years in light of overlapping linkage, association, and gene expression data in schizophrenia and bipolar disorder. It just makes sense that biomarkers will work best for symptoms or other lower-level traits or states rather than hierarchical diagnoses with questionable validity. In turn, biomarker work performed in this manner may subsequently inform the derivation of novel, biologically based, valid diagnostic categories.
The use of CSF samples is laudatory, since this provides a good balance between access to the central nervous system and ease of sample collection, whereas other modern biomarker studies of blood or postmortem brain tissue do not allow for such balance. The protein-chip technology itself is also impressive, although as with any cutting-edge technology in its infancy, major improvements in resolution and throughput will be needed to move this technique into the mainstream.
In summary, this manuscript should be read carefully by all those among us who are interested in the frontiers of biomarker research on mental disorders. However, in contrast to the authors' statement that "An
alternative approach to genetic studies is to screen for disease markers (biomarkers)," we might be wise to treat biomarker research as a useful adjunct to—rather than replacement for—analyses of genetic polymorphisms.
View all comments by Stephen J. Glatt