A Probabilistic Model of the Genotype/Phenotype Relationship

Does Life Play the Dice?

Jean-Pierre Hugot

• Publisher: CRC Press
• Year: 2018
• Print ISBN: 9781138320727
• eBook ISBN: 9780429841668
• Edition: 1
• Topic: Biological Sciences
• Subtopic: Biology
• Index: Biological Sciences

A Probabilistic Model of the Genotype/Phenotype Relationship provides a new hypothesis on the relationship between genotype and phenotype. The main idea of the book is that this relationship is probabilistic, in other words, the genotype does not fully explain the phenotype. This idea is developed and discussed using the current knowledge on complex genetic diseases, phenotypic plasticity, canalization and others.

Contents

Dedication

Preface

Warning

Introduction: Complex Genetic Diseases

PART I:The Genotype/Phenotype Relationship

1. The Beginnings of Life

2. Genetic Program and Self-Organization

3. The Phenotype

4. The Role of Genes

5. The Share of the Environment

6. Analogy Between Organisms and Particles

7. Phenotypic Plasticity

8. Behavior

PART II: Heredity

9. Reproduction and Phenotypic Canalization

10. Fertilization and Overlapping of Genotypes

11. Nature of Canalization

12. Canalization: Experimental Data

13. Implications of Canalization

14. Canalization and Instinct

PART III: Evolution

15. Limits to Natural Selection

16. Instinct and Evolution

17. Biological Function

18. Differentiation and Phenotypic Coherence

19. Speciation

20. Phenotypic Innovation

Conclusion

In Summary

Bibliography

Thanks

Index

Warning

As a paediatrician and researcher, I specialised in Crohn’s disease, a disabling disease characterised by chronic diarrhea and an alteration of well being. My reflection on this illness and complex genetic disorders in general have fuelled my approach and serve as an introduction to this work. However, this book is not directly linked to my work as a doctor and researcher. It has only a light bearing on diseases but on the contrary questions other biological problems of a more general nature. As a result the reader may deduce that I am not an expert in the multiple fields broached herein. This work is the attempt of an amateur to be enlightened. Consequently, certain mistakes or misinterpretations may be noted by specialists. For this I apologize in advance and hope they will nevertheless find some interest in these pages.

The work is not scientific in the usual sense of the word by research biologists, professional category to which I belong. Thus, it does not present a hypothesis, an experimental plan and a discussion of original results from experiments. On the contrary, it is based on ideas and concepts. The book should therefore be taken as a reflection. On the other hand, it talks about biology and is based on observed data. Hypotheses and models are generated which can be tested by the experimental method. It therefore has a link with science. I would willingly give it the slightly old-fashioned term of ‘Essay on natural philosophy’ to illustrate the link with a deductive rather than experimental approach.

As the reader will discover, the ideas developed in this book feed on analogies between the concepts of quantum physics and those of biology. An analogic approach has been used since it is an effective method for reflection. It is nevertheless only a tool. It is not an ideological choice, and it is only for its operational qualities simply that the analogical approach has been used. Like the microscope, it provides an original view point to look at the living. It is in no way to research or promote any mystical or esoteric thought as may have been done by others.

Introduction: Complex Genetic Diseases

No thought is more important than that which focusses on the transmissibility of aptitude and character.

Nietzsche

In the simple model corresponding to Mendel's Laws and classic genetics, the correlation between the genotype and the phenotype is perfect. ¹ In other words, a given genotype corresponds to a given phenotype and vice versa. A pea is smooth or wrinkled according to the existence of a molecular shape (called allele) ² or the other of the gene. The relationship is so unequivocal that the allele, since Morgan, carries the name of the phenotypical trait “smooth” or “wrinkled” whereas it expresses in reality only information on the sequence of the DNA. This direct and symmetric relationship [genotype phenotype], allows reliable and simple predictions. They are expressed by Mendel's laws for the transmission of phenotype from one generation to another during sexual reproduction. It is classic genetics that we learn at school.

In medicine, there are several thousand illnesses, often of a very low incidence (known as orphan diseases) which have a Mendelian transmission. In the most typical case there is a perfect correlation between the phenotype and the genotype: people with the mutation (or morbid allele) always have the disease and those with the disease are always carriers of the mutation. To mention just two classic and caricatural examples, mutations in the CFTR gene (coding for a chloride ion-transmembrane channel) are associated with cystic fibrosis and a specific mutation of the b chain of hemoglobin is associated with sickle cell anemia.

Nevertheless, most hereditary characteristics do not follow a simple genotype/phenotype correlation. Distortions in the relationship between genotype and phenotype are common and have been recognized for a long time (including by Mendel ³ ) thus some people with mutations may not have the expected disease. Phenotype modulator factors are then used to explain these discrepancies. These additional variables make it possible to reestablish an explicit genotype/phenotype relationship. Some phenotype modulating factors are genetic. This is the case of the foetal haemoglobin gene in sickle cell anemia. The genetically programmed persistence of foetal haemoglobin limits the severity of the disease.

Environmental factors may also contribute to the phenotype. This is the case of bronchial infection with Pseudomonas aeruginosa in cystic fibrosis, which precipitates the progression of pulmonary disease. Some pathologies may even depend entirely on exposure to an environmental risk factor that must be added to the genetic risk for the disease to be expressed. This is the case of phenylketonuria where exposure to a highprotein diet is necessary for clinical signs to appear in mutated subjects. For celiac disease, wheat consumption by genetically predetermined people is required to cause disease. In the latter two examples, exclusion diets (phenylalanine for phenylketonuria or gluten in wheat for celiac disease) are in fact very effective in controlling the expression of the disease in genetically predisposed individuals. In all these situations, the link between the genotype and the phenotype is no longer direct and implies taking into account additional risk factors in relation to the morbid genotype.

By going further into complexity, the phenotype can be the combination of a very large number of genes and many environmental factors. This is the case for diseases known as complex genetic disorders. These diseases are common in human populations. These include diabetes mellitus, cardiovascular diseases, psychiatric disorders, autoimmune diseases, degenerative diseases, cancer, etc. To avoid multiplying examples, we will often take Crohn's disease, an inflammatory bowel disease, as an illustration. Recent genome exploration work for these pathologies shows that predisposing genes are numerous. At least several tens or hundreds for each disease, at least 140 for Crohn's disease. ⁴

Thus, in total, several thousand polymorphisms have been associated with complex diseases. ⁵ However, each of the susceptibility alleles has a weak individual effect. Thus, most genetic polymorphisms associated with complex human diseases carry an estimated relative risk between 1 and 1.5. In other words, the likelihood of developing the disease for a person with a risky genetic variant is no more than 1.5 times higher than that of a person not carrying the variant. In comparison, the same relative risk for a typical Mendelian disease is theoretically infinite. The disease appears to result from a multitude of low-risk alleles scattered throughout the genome.

The genetics of complex diseases then leads from an exact deterministic genetic model to a fuzzy probabilistic model. Instead of predicting a disease every time from simple genetic information, the geneticist must now take into account a very large amount of information, not only genetic but also environmental, to arrive at an approximation of a risk. Complex diseases require us to change the way we think about the relationship between genotype and phenotype.

Geneticists, however, believe that the link between the genotype and the phenotype is not probabilistic in nature. It is our incomplete knowledge of the genetic determinism of diseases that explains why we have to use probabilities to approach reality. Accordingly, there are still unknown risk factors which, once discovered, will lead to a complete knowledge of the determinism of complex diseases, making the probabilities of the model disappear.

In fact, we have not explored all the variations present on the genome. The polymorphisms studied to date are mainly polymorphisms with two simple alleles and common in the general population. Studies on more complex common gene variants (copy-number variations) are less numerous but produce comparable results: a large number of genetic polymorphisms with a low effect. ⁶ There remains the question of the rare genetic variants not yet systematically explored to date in most complex diseases.

Rare genetic variants could, if they proved to have a strong phenotypic effect, allow us to return to determinism closer to the Mendelian model and thus more explicit. ⁷ The current question is therefore whether rare mutations with high penetrance and which can therefore have a strong predictive effect on the disease remain to be discovered. Many laboratories are currently searching for rare genetic variations left out of early pan-genomic studies. ⁸

There is no doubt that such mutations will be discovered, at least for some extreme morbid phenotypes. Thus, for Crohn's disease, in infants, it is possible to identify mutations with a strong effect, such as those of the interleukin-10 receptor. ⁹ Such discoveries may nevertheless remain confined to a few rare cases and not be generalizable to the majority of patients. Although the results of large-scale sequencing of the genomes of patients are not yet known, I am, in fact, rather pessimistic about our ability to discover rare and highly penetrant genetic variations accounting for a significant part of disease risk for two reasons.

My first argument is that the analysis of the genealogical trees of the majority of the patients shows that the transmission of the disease is far from the proportions defined by Mendel. Indeed, according to a model of a rare mutation with a strong phenotypic effect, even in the presence of a strong genetic heterogeneity (in the extreme, a specific mutation per family of patients) and a large number of neo-mutations, we expect to s...

Table of contents

• Cover
• Halftitle Page
• Title Page
• Copyright
• Dedication
• Preface
• Table of Contents