The Genetics of Schizophrenia, The Small RNA’s and The Recourse to The Method
“Ofelia was convinced: There is
Only one Earth and The Earth’s Earth is wherever the same “
A. Carpenter – “The Recourse to
This article reviews the progress in the Psychiatric Genetics in general and the Genetics of Schizophrenia in particular, together with the new discoveries in the field of Small RNA’s. It addresses their impact on the understanding of Schizophrenia’s Etiological Models and it brings to light the failure of long standing endeavors to build generally valid models. This direction did not achieve the expected breakthrough in modeling complex illnesses like Schizophrenia. The article raises the question of elaborating different tools that would switch the attention from general models of diseases to modeling particular and complex treatment plans for specific ill individuals.
Progress in Psychiatric Genetics
As a whole the field of Psychiatric Genetics did not benefit from an impressive reputation: Many linked loci were initially claimed with enthusiasm and subsequently withdrawn. An overwhelming number of association studies were not replicated.
Too many chromosomal regions were claimed to be responsible about schizophrenia.
The transmission pattern of psychiatric illness is generally very complex and occurring at the encounter of genetic and environmental interactions. The phenotypes of psychiatric illnesses are at times unclear and factors like co morbidity with another illness may complicate the picture. The accepted idea is that there are several combinations of genetic and environmental interactions which could lead to same end results (the concept of equifinality).One of the pertinent questions is if the genetic heterogeneity is a decisive factor? The general opinion is that it is complicated but not insurmountable.
An example to sustain this conclusion is that in the Hearing Loss there were more than 50
Loci mapped while about a dozen of them could be identified.
The most complicating factor in going to the essence of complex disorders etiology seems to be the epistatic work of multiple genes that could be implicated.
If this is the case then the risk of the disease will considerably diminish while you move
Toward more and more distant relatives (schizophrenia and bipolar disorders fit this
Respectable authors like Risch concluded that when considering the relative risk of Schizophrenia in the relatives of affected patients the most valid model seems to be Gottesman and Shields of multigenic inheritance with 3-5 interacting loci(none of which increases the susceptibility for more than 2-3 times).
A very interesting work is the one done by Risch who did a genome scan in siblings with autism and identified alleles sharing in every chromosomal region. This lead to another hypothesis of relative big number of genetic areas that could interact together with the environment and life stressors and cause the complex illness. In this model one can have more than 20 loci involved each of them responsible of a minor effect.
Another confusing element is that at times the same affected genetic area could be expressed by different clusters of diseases. The best known example is the Velocardiofacial syndrome which is caused by a deletion of approximately 3 Mb on chromosome 22q11 and could be associated with several complex disorders like Major depression , Schizophrenia , Bipolar Disorders , ADHD , Autism Spectrum Disorders ,etc.
Linkage Studies were used in the psychiatric research because their ability to localize loci for the Mendelian transmitted diseases .Different study designs were utilized: Like for example traditional and sib-pair in large and isolated populations. Most of them were not replicated .However some authors like Gershon sustain the idea that even when the linkage is not replicated the results should not be discounted because the statistical power to find a significant locus is approximately 5-10%. In this case most of the similar samples will not replicate the linkage even if it is accurate.
This explanation (that leaves the jury out in regard to the linkage studies utility in understanding complex diseases) is not taking into account the participation of environment and possibly multiple genetic areas into the appearance of the disease.
The use of Endophenotypes or trait markers was a basic work idea meant to identify a trait that is more common in the ill population versus the general population and is also displayed by the unaffected family members (carriers of the predisposing allele).The chosen trait should be heritable with a high frequency by the close relatives and stable over time (including not being changed with the giving of medication for treatment).
Examples of traits like that are: low response to alcohol in alcoholic offspring’s or in case of Schizophrenia relatives: the deficit in smooth pursuit of eye tracking movements.
The candidate genes approach was embraced by many researchers due to possibility to verify certain genetic hypotheses which related to the specific insights about the diseases.
The studies were simple to design and using epidemiological tools would enhance their accuracy .However the population stratification could lead to false positives even in big samples of patients. The researched allele variant was desirable to be functional and biologically relevant. Newer approaches have the ability to test the effects of two markers and their statistical interactions.
Recent novel techniques like SNP’s scored on DNA chips allow the simultaneous testing of thousands of candidate genes and necessitate large samples of patients.
A recent hopeful approach is the ability of Psychopharmacogenetics to individualize the treatment plan based on the patient’s genotype at significant markers (Serotonin and Dopamine transporters, Cyp450, etc).
Molecular Genetics of Schizophrenia – A brief review
Most of the relatives of Schizophrenia patients do not develop the disease .However the available data from the adoption studies point toward the genetic factors versus the environmental ones. Twin studies endorse the same finding by showing a clear increased rate of disease in the monozygotic twins compared with dizygotic ones.
Is is also clear that what is inherited it is a predisposition to the illness where the environmental factors are necessary to contribute to the illness appearance in the majority of the cases. There are known risk factors like intrauterine events, perinatal complications, early developmental delays, life stressors, drug abuse, etc.
The linkage studies ruled out a single major locus and suggest several regions that could be susceptibility loci and demand possible replication and further investigation.
These regions include the chromosomal regions: 5p, 5q, 6p, 6q, 8p, 10p, 13q, 15q, etc.
As previously said the fundamental question is how reliable are these linkages?
One way to answer this question would be to come up with a statistical threshold that would help reject the false positives.
A way to go into this direction was developing the concept of “Genome Wide Significance” in order to elaborate a more reliable opinion if the linkage is real or not.
The concept is based on the probability threshold that declares
Linkage after testing many DNA markers that are used in a genome scan.
Lander and Krugylak came up with 3 levels of Genome Wide Significance:
- suggestive linkage
- significant linkage
- confirmed linkage(replication in an independent sample)
The expectation of a highly significant linkage in Schizophrenia is unlikely due to the lack of a homogenous phenotype and a known mode of transmission.
The strongest findings for Schizophrenia are for 13q and 6p which meet the criteria for replicated linkage. Multi center analyses support an association between Schizophrenia and polymorphisms at the loci for D3 dopamine receptor and 5HT2a receptor gene.
Several fundamental limitations lead to failure to advance the mission of finding the Schizophrenia genetic base:
- Wide variability in the clinical phenotypes of the cases included in the studies
- Insufficient use of the endophenotypes and sub clinical types of Schizophrenia in the research
The future should be oriented into the direction of using the disturbances of different neurobiological functions associated with Schizophrenia like for example : Impaired gating of the auditory evoked responses and ocular motor dysfunctions as endophenotytpes for linkage studies.
Any good model for transmission of Schizophrenia needs take in consideration the MZ and DZ twins rates and the familial rates of disease transmission.
Admitting the hypothesis that Schizophrenia follows a classic Mendelian transmission would in the case of a dominant gene imply that 50% the offspring of one Schizophrenic patient would have the disease when the real rate of inheritance is 13%.
If the illness would be transmitted by a fully penetrant recessive gene that would imply that every child born to Schizophrenic parents would have the disease when the rate of inheritance in this case is only 36-50%.
If the illness would be transmitted by a Mendelian mechanism there would be a certain number of first degree relatives ill. But most of the patients have no first degree relatives ill.
Another possible model would be the Single Major locus model (SML) in which the pair of genes at a single locus are responsible of the illness transmission. This model however underestimates the risk for MZ twins and for the offspring’s of two schizophrenic patients.
Another model assumes the existence of a latent trait which could cause the illness or some related phenotypic traits of it. The latent trait is transmitted by a Mendelian mechanism with high penetrance but may not always occur in the same exact pattern. Based on this idea the appearance of the illness is a rare outcome of an often met context. A good way to illustrate this idea is to use the example of SPEM (smooth pursuit eye movements) which is a task where one need to track a moving object in space only by moving the eyes and not the head.65-80% of Schizophrenic patients show eye saccadic movements intrusions to differ from 40-45% of their 1st degree relatives and only 8% of the general population. This would suggest that schizophrenia and the SPEM dysfunction are independent expressions of a single gene. This model however can not
Give a good estimate for the illness risk among MZ twins and offspring’s of 2 schizophrenic parents.
The polygenic models state that the etiological genes are located at two or more loci .The oligogenic type of polygenic models implicates a specific number of loci while the multifactorial polygenic model (MFP) implicates a large unspecified number of loci (one
Suspicion the participation of a large number of genes when each one of them has a very limited effect).
This MFP model assumes that everyone has a certain vulnerability to Schizophrenia and when the sum of the genetic effects together with the environmental influence pass over a specific threshold the illness occurs.
None of these models taken alone could explain the high concordance rate among the MZ twins versus the low rate of the illness appearance in the 1st degree relatives of the patients. However when the concept of Epistasis (some gene interfering with the phenotypic expression of another gene) is included together with the MFP model then these together seem to be a fairly good answer for the above objections.
This combined model explains well the higher risk for the relatives and the abrupt decrease in the risk as the number of shared genes decreases in the more distant relatives.
Small RNAs make it Big
In 1969 Davidson and Britten hypothesized that RNAs could decide which genes are to be active and which genes are not in the eukaryotic cells.
The subsequent discovery of transcription factors(~1850 in humans) made Davidson hypothesis to be forgotten.
Signs that RNAs could be more versatile than it was thought were again shown in 1990 when it was determined that small RNAs could inactivate the expression of diverse genes in plants and animal cells.
There are several distinct classes of these small RNAs:
- MicroRNAs (miRNAs)
- Small interfering RNAs (siRNAs)
- Repeat associated small interfering RNAs (rasiRNAs).
They all are distinguished by their origins. The miRNAs are supposed to regulate 1/3 of all human genes .Small RNAs (21-30 nt) are involved in a big number of biological pathways.
The production and function of small RNAs needs a set of proteins like: Double stranded RNA (dsRNA)-specific endonucleases (such as Dicer), dsRNA binding proteins and small RNA-Binding proteins called Argonaute proteins.
Together the small RNAs and their associated proteins achieve the RNA pathways silencing in different occasions and circumstances like for example: pathways regulating transcription, chromatin structure, genome integrity, mRNA stability.
In 1990, 2 groups over expressed a pigment synthesis enzyme to get deep purple petunia flowers but the experiment ended up generating predominantly white flowers.
The phenomenon which occurred was called “co suppression” because the transgenic and endogenous genes were repressed. This was a breakthrough and the beginning of RNA silencing knowledge.
By the year 2000 RNA silencing was described to many eukaryotes when RNA interference (RNAi) is the best known RNA silencing mechanism (because of its involvement in the human clinical trials testing RNAi based drugs).
The most important moment in RNA silencing was the discovery that dsRNA (double stranded RNA – these are specific endonucleases like dsRNA binding proteins) is the actual trigger of mRNA destruction .It is the sequence of dsRNA that determines which mRNA is to be destroyed.
The ds RNA is converted into siRNA fragments of the original dsRNA(21-25 nt length) that guide protein complexes to complementary mRNA targets where expression is
MicroRNAs (miRNAs) derive from long unstructured transcripts (pri-miRNA) containing hairpin structures of ~70 nt length.
The hairpins are cut out of the pri-miRNA by the dsRNA endonucleases Drosha acting with its dsRNA – binding protein DGCR8 in humans.
The mature mRNA is excised from the pri-miRNA by another dsRNA-specific endonucleases Dicer acting with a dsRNA binding protein partner the tar binding protein (TRBP) in humans.
The human genome may contain ~1000 miRNAs while few of each are unique to humans and also make us the humans to be unique.
There are 53 miRNAs unique to primates.
Only 6-7 nt of the 21 nt within miRNA or siRNA are involved in binding specificity for the small RNA protein complex they guide. This small region has been called “the seed sequence”.
The pattern of nucleic acid interaction is related to the way the small RNAs alone and paired with their RNA target is bound by a member of the argonaute family of proteins. The Argonaute family of proteins are specified for binding the small RNAs that mediate the RNA silencing .The small RNAs that act on RNA silencing act somehow like restriction enzymes. When a small RNA pairs with its RNA target it directs cleavage of a single phosphodiester bond in the target RNA .This cleavage is very specific and occurs only when the RNA is bound to the right Argonaut protein.
When siRNAs and miRNAs pair only partially with their targets they cannot achieve mRNA cleavage but they block translation of mRNA into protein.
There is need for several miRNAs to bind to the same target to block translation.
The coordinated expression of mRNA becomes a way to control gene expression.
Line et al found that miRNAs can alter the stability of hundreds of mRNAs.
It is thought that small RNAs make mRNAs less stable moving the mRNAs to P-Bodies (the sites of mRNAs destruction).This also explains the small RNA direct translational repression. By sequestering mRNAs into the P-Bodies small RNAs would block translation.
However Filipowicz argue that repression of mRNA translation by miRNAs is just a consequence of relocalization of mRNAs from cytosol to the P-Bodies and sustains that miRNA block translational initiation while the relocalization of mRNA to the P-Body is just a consequence of that.
Once in the P-Body the mRNA is degraded releasing the miRNA programmed protein complex to return to cytosol for a new cycle of mRNA repression.
RNA interference has also been implicated in silencing parasitic DNA sequences such as transposons and repetitive sequences .A specialized RNA silencing pathway senses the “aberrant RNA” transcribed from such sequences and then initiates post-transcriptional or transcriptional silencing.
Transcriptional silencing directed by small RNAs is also associated with the formation of heterochromatin .In mammals heterochromatin DNA is also hypermethylated .Chicken and mouse cells lacking Dicer fail to assemble silent heterochromatin at their centromeres.
RNA and the genetics of Schizophrenia
In higher organisms only 2-3% of the genetic transcripts code for proteins .Some of the noncoding RNA may as shown play a key role in gene expression regulation.
Malter identified highly conserved regions across several vertebrate species in a 12 mb orthologous genetic region where 80% of the highly conserved regions were located in introns or intergenetic regions which shows that the nonprotein coding regions could
have an important regulatory role that has to be yet elucidated.
Taft and Matick observed that the increased biological complexity is positively correlated with the relative genome wide expansion of non protein coding DNA sequences and poorly correlated with its number of protein coding sequences.
As was previously shown the rate of protein synthesis is related to the amount of mRNA available which is dependent on the rate of DNA transcription ,mRNA degradation,etc.
Noncoding RNA may regulate protein synthesis by slowing or speeding the mRNA silencing. While dysregulated these control mechanisms could be critical in the appearance of certain diseases.
According to Perkins et al another possible explanation for the lack of success of the genetic schizophrenia related studies relates to the fact that the regulation of the genes involved in the appearance of schizophrenia could be the one responsible of the illness and not the diverse genetic polymorphisms as it used to be believed.
In this hypothesis the dysregulation of translation and transcription (due to the noncoding RNA) is responsible of a default in the gene expression and the development of the illness. This dysregulation according to Krichevsky may happen during the development of the brain when the microRNAs appear to have an important regulatory role.
The pathological changes occurring this way may last for the whole life .
Regulating pseudogenes (noncoding RNA) may exist at very different chromosomal locations which is different from the one of the regulatory genes themselves.
So it could be that some areas identified through genetic linkage studies as associated with schizophrenia could be just regulatory noncoding RNA(Perkins).
BDNF is a very important factor whose polymorphisms were associated with schizophrenia .
Lewis et al showed that 2 new miRNAs (has-mir1 and has-mir206) may target the 3-UTR of BDNF interfering with the BDNF mRNA translation during brain development or later.
However the 2 miRNAs are not mapping exactly in areas shown to be significant in linkage study meta-analysis of Lewis.
The D2 receptor is a major site for schizophrenia neuroleptic medication and DRD2 dopamine receptor gene a candidate for schizophrenia risk.
However association studies are not significant. Known DRD2 synonymous polymorphisms have never been examined in the association studies. But Duan et al showed that synonymous mutations in DRD2 affect the mRNA stability and the synthesis of the receptor.
An example is the DRD2 DNA polymorphism SNP C957T that altered the receptor tertiary structure and decreased the stability of mRNA and also of the proteic synthesis.
Another synonymous polymorphism SNP C1101A disrupted the mRNA translation rate.
Singh and Petronis proposed that the environmental influence in Schizophrenia is due to epigenetic modification of protein coding gene expression which could be mediated by small RNAs.
Frankle in his synaptic hypothesis of Schizophrenia related the illness to altered connectivity inside the brain.
One protein which is involved in the normal dentritic density and function at the level of cortical pyramidal cells is Reelin .The Reelin gene was not linked to schizophrenia by studies. However in the postmortem studies of patients brains there is evidence of Reelin mRNA levels to be clearly low.
It is clear that all the RNA evidence related to schizophrenia that is available in the literature raise more signs of questions than give any answers.
However this inquiry opens new directions of research and brings in some hope which could related to the possible use of miRNA medications in the future.
The linkage studies in their classic design are probably not completely fit to address the complexity of the schizophrenic illness.
The pool of schizophrenia genes has been around from ancient times and has not suffered extinction due to natural selection. This observation supports perhaps the MFP + Epistasis Model of schizophrenia.This model is more adequate to explain also the heterogenous phenotype and the difference in genetic constellations found in diverse ethnic groups.
As much as it is speculative in its nature the superimposing of the RNA hypothesis on the MFA + Epistasis Model makes the things to look several times more complicated .Also this new revealed quest offers new directions of inquiry.
Are the classic tools good enough to face this newly enriched complexity? Do ones need to look for other complex tools from the field of Bioinformatics or Theory of Complex Systems and Emergence? The application of Chaos Theory and Nolinear Dynamics to the field of Psychiatry did not bring the expected breakthrough in our understanding of these very complex diseases .Would these lately mentioned tools cause a “recourse to the method” with a switch of the paradigm from building general models to the elaboration of individualized highly complex patients treatment plans?
Would having the knowledge of what genes are involved influence the treatment of a certain disease? Our knowledge of what genes are involved in the diseases like Lesch-Nyhan or Huntington did not advance our ability to treat the behavioral disturbances associated with these illnesses a lot.
However the existence of miRNAs trials brings in a new approach and together with it some hope.
Stoltenberg, S. F., & Burmeister, M. (2000). Recent progress in psychiatric genetics – some hope but no hype. Human Molecular Genetics, 9(6) Review; 927-935.
NIMH Workgroup (1999). Genetics and mental disorders. National Institute of Mental Health’s Genetics Workgroup. Biol. Psychiatry, 45; 559-553.
Van Camp, G., Willems, P. J., & Smith, R. J. (1997). Nonsyndromic hearing impairment: unparalleled heterogeneity. Am. J. Hum. Genet., 60; 758-764.
Risch, N. (1990). Linkage strategies for genetically complex traits. I. Multilocus models. Am. J. Hum. Genet., 46; 222-228.
Risch, N., Spiker, D., Lotspeich, L., Nouri, N. Hinds, D., Hallmayer, J., Kalaydjieva, L., McCague, P., Dimiceli, S., Pitts, T. et al. (1999). A genomic screen of autism: evidence for a multilocus etiology. Am. J. Hum. Genet., 65; 493-507.
Kendler, K. S., Karkowski, L. M., & Walsh, D. (1998). The structure of psychosis: latent class analysis of probands from the Roscommon Family Study. Arch. Gen. Psychiatry, 55; 492-499.
Gershon, E. S., Badner, J. A., Goldin, L. R., Sanders, A. R., Cravichik, A. & Detera-Wadleigh, S. D. (1998). Closing in on genes for manic-depressive illness and schizophrenia. Neuropsychopharamacology, 18;233-242.
Schuckit, M. A. (1998). Biological, psychological and environmental predictors of the alcoholism risk: a longitudinal study. J. Stud. Alcohol, 5_; 485-494.
Schuckit, M. A. (1987). Studies of populations at high risk for the future development of alcoholism. Prog. Clin. Biol. Res., 241; 83-96.
Rheinberger, H. J., & Müller-Wille, S. “Gene”, The Stanford Encyclopedia of Philosophy (Winter 2004 Edition), Edward N. Zalta (ed.) URL = <http://plato.stanford.edu/archives/win2004/entries/gene/>.
Beurton, P. (2000). A unified view of the gene, or how to overcome reductionism. In Peter Beurton, Raphael Falk, and Hans-Jörg Rheinberber (eds.), The Concept of the Gene in Development and Evolution. Historical and Epistemological Perspectives . Cambridge University Press, Cambrige, 286-314.
Crick, F. (1958). On protein synthesis. Symposium of the Society of Experimental Biology, 12; 138-163.
Gros, F. (1991). Les secrets du geÿne. Nouvelle eÿdition revue et augmeenteÿe, Editions Odile Jacob, Paris.
Jacob, F. (1976). The Logic of Life. Vanguard, New York.
Kühn, A. (1941). Über eine Gen-Wirkkette der Pigmentbildung bei Insekten. Nachrichten der Adedemie der Wissenschaften in Göttingen, Mathematisch-Physikalische Klasse, 231-261.
Morgan, T. H. (1935). The relation of genetics to physiology and medicine. Les prix Nobel en 1933. Imprimerie Royale, Stockholm, 1-6.
Morange, M. (2000). La function des geÿnes. Comptes rendus de lAcademie des sciences. SeriesIII Sciences de la vie, 323(12) Dec; 1147-1153.
Zamore, P. D., & Haley, B. (2005). Ribo-gnome: The big world of Small RNA’s. Science, 309(5740) Sept.; 1519-1524.
Berry, N., Jobanputra, V., & Pal, H. (2003). Molecular genetics of schizophrenia: a critical review. Psychiatry Neurosci, 28(6); 415-429.
Perkins, D. O., Jeffries, C., & Sullivan, P. (2004). Expanding the ‘central dogma’: the regulatory role of nonprotein coding genes and implications for the genetic liability to schizophrenia, Molecular Psychiatry, 10; 69-78.
Mattick, J. S. (2001). Non coding RNA’s: the architects of eukaryotic complexity. EMBO Rep, 2; 986-991.
Taft, R. J., & Mattick, J. S. Increasing biological complexity is positively correlated with the relative genome-wide expansion of non-protein-coding DNA sequences. arXiv org 2004.
Krichevsky A. M., King, K. S., Donahue, C. P., Kharpko, K., & Kosik, K. S. (2003). A microRNA array reveals extensive regulation of microRNA’s during brain development. RNA, 9; 1274-1281.
Lewis, C. M., Levinson, D. F., Wise, L. H., DeLisi, L. E., Straub, R. E., Hovatta I., et al. (2003) Genome scan meta-analysis of schizophrenia and bipolar disorder, Part II: schizophrenia. Am J Hum Gene, 73; 34-48.
Duan, J., Wainwright, M. S., Comeron, J. M., Saitou, N., Sanders, A. R, Gelernter, J. Et al. (2003). Synonymous mutations in the human dopamine receptor D2 (DRD2) affect mRNA stability and synthesis of the receptor. Hum Mol Genet, 12; 205-216.
Singh, S. M., Murphy, B., & O’Reilly, R. (2002). Epigenetic contributors to the discordance of monozygotic twins. Clin Genet, 62; 97-103.
Petronis, A., Gottesman II, Kan P., Kennedy, J. L., Basile, B. S., Paterson, A. D. et al. (2003). Monozygotic twins exhibit numerous epigenetic differences: clues to twin discordance? Schizophr Bull, 29; 169-178.
Alina R.Rais MD , Assistant Professor of Psychiatry, University of Toledo Medical Center , Toledo,OH.
Tanvir Singh MD , Assistant Professor of Psychiatry , UTMC , Toledo ,Ohio
Theodor B.Rais MD Associate Professor of Psychiatry , UTMC , Toledo , Ohio.
Address any correspondence to : Alina R.Rais MD
Ruppert Health Center
3120 Glendale Avenue
Toledo , Ohio 43615
Copyright Priory Lodge Education Limited 2007
First Published October 2007