Functional Genomics and Experimental Modeling of Genetic Variation

Instructor Background and Core Philosophy

• Instructor Profile: Archer Fox is an academic jointly appointed between the School of Human Sciences and the School of Molecular Sciences.

• Academic Background: Fox's primary field of study was molecular genetics. However, his focus is not on classical genetics but on understanding the molecular changes resulting from genetic variation and the mechanisms controlling gene expression (the "on and off" switches of genes).

• Current Research Interest: Fox defines himself as a molecular cell biologist. He investigates the spatial localization of molecules within cells and how this relates to gene regulation.

• Course Objective: The lecture aims to address functional genomics and the challenges of a society where entire genomes are sequenced. The focus is on interpreting variation—understanding how genetic templates affect the production and quantity of molecules, even when protein sequences remain unchanged.

Introduction to Functional Genomics

• Definition and Scope: Functional genomics involves investigating the functional significance of genetic variation discovered through sequencing. It aims to bridge the gap between knowing a variant exists and understanding its biological impact.

• The Genomic Landscape: Most human genetic variation does not alter protein sequences. Consequently, a major scientific challenge is interpreting the impact of variants that occur in non-coding regions.

• Research Problem: In a society with accessible genome sequencing, researchers must distinguish between harmless variation and variants that contribute to disease or specific phenotypes.

Categories of Non-Protein-Coding Variation

• Synonymous Variants: These are mutations within the coding sequence (exons) of a gene that change the DNA sequence but do not change the amino acid residue encoded. This occurs due to the redundancy of the genetic code (e.g., a change from $A$ to $C$ in a leucine codon like $CUU$, which might still result in leucine).

• Non-Coding Genes: These are genes that are transcribed into RNA but never translated into protein. These "non-coding RNAs" remain functional within the cell.

• Regulatory Elements: Variants occurring in promoters or enhancers. These elements control the level of protein production (making more or less) rather than the structure of the protein itself.

• Intronic Variants: Changes located within introns (the sequences between exons) that can interfere with the splicing process, potentially leading to the inclusion or exclusion of specific exons in the mature mRNA.

Mechanisms of Phenotypic Change via Synonymous Variants

Even if the protein sequence remains identical, a single base pair change in the DNA/RNA can affect the phenotype through several mechanisms:

• RNA Stability: A variant can change the lifespan of an mRNA transcript. If an mRNA is rapidly degraded in the cytoplasm, the ribosome has less time to translate it, resulting in lower protein levels.

• Transcription Factor (TF) Binding Sites: TF binding sites are not always upstream of the transcription start site; they can actually be located within exons. A synonymous change might disrupt a binding site, altering the transcription rate.

• Splicing Regulation: Changes within an exon can still influence how the spliceosome recognizes exon-intron boundaries, leading to alternative splicing.

• Speed of Translation: Variants can influence how efficiently or quickly the ribosome binds to and moves along the RNA transcript.

• Case Study: $DRD2$ Gene: Researchers found a synonymous variant ($C957T$) in the $DRD2$ gene implicated in a psychiatric disorder. By blocking all cellular transcription with drugs and measuring the rate of degradation, they proved that this specific variant caused the mRNA to degrade significantly faster than the wild-type transcript.

Regulatory Elements and the ENCODE Project

• Enhancers and Promoters: Promoters sit at the transcription start site. Enhancers are "upstream regulatory sequences" that can be located far from the gene. DNA loops back so that these enhancers come into physical proximity with the promoter and RNA polymerase complex.

• The Significance of Non-Coding Variants: Approximately $90\%$ of variants associated with complex, multigenic, or chronic diseases are found outside the protein-coding regions (the open reading frame).

• The ENCODE Project (Encyclopedia of DNA Elements): A massive American government-funded consortium designed to map every functional element in the human genome.

  • Goal: To provide a resource where researchers can look up a variant and see if it falls within a known functional element (protector, enhancer, etc.).

  • Methodology: The project utilized approximately $50$ different human cell lines.

  • ChIP-seq (Chromatin Immunoprecipitation Sequencing): A technique used in ENCODE to identify transcription factor binding sites. It involve cross-linking proteins to DNA, isolating the complexes, and sequencing the bound DNA.

  • Chromatin Accessibility: The project mapped areas of "open chromatin" versus "closed (heterochromatin)" to determine visibility to regulatory molecules.

Theoretical Validation: Identifying the "Needle in the Haystack"

• The Data Deluge: Genome sequencing is now inexpensive ($\approx \$1,000$ per genome), leading to a surplus of data.

• Rare Disease Diagnosis: Sequencing the genomes of a mother, father, and affected child can isolate the specific mutation causing a rare disease.

  • Anecdote: A young boy in Sydney with a rare disease was diagnosed via whole-genome sequencing of his family. This allowed doctors to repurpose an existing drug for a different condition, which successfully treated him.

• The ExAC Database (Exome Aggregation Consortium): A collection of over $60,000$ genomes from individuals who are supposedly disease-free.

  • Utility: Researchers use this to find which variants are not tolerated in the population. If a specific genomic position shows zero variation across $60,000$ people, it is likely a critical site where mutations are lethal or deleterious (Probability of Loss of Function).

Model Systems: Human Cell Lines and iPS Cells

To prove a variant causes a phenotype, researchers use different model systems since they cannot manipulate human embryos.

• Primary Cells: Cells donated from biopsies or bone marrow.

  • Limitation: They have a limited number of divisions before undergoing apoptosis because their telomeres shorten with every division.

• Immortalized Cell Lines: Mostly derived from cancer biopsies. They grow indefinitely in a lab environment ($37^\circ C$, $5\% \, CO_2$).

  • Pros: Cheap, quick, and possess a human genome.

  • Cons: Grown in 2D monolayers (unlike 3D tissue); carry many background mutations from the original cancer.

• Transformed Normal Cells: Primary cells that have been "tricked" into lengthening their telomeres via the introduction of specific enzymes.

• Induced Pluripotent Stem Cells (iPS cells):

  • Process: Adult skin fibroblasts are reprogrammed using specific transcription factors to become stem cells.

  • Applications:

    1. Personalized Medicine: Test multiple drugs on a patient's own differentiated cells in a dish to find the most effective/least toxic one.

    2. Gene Repair: Fix a disease-causing variant in the iPS cells, differentiate them into the required cell type (e.g., retinal cells), and potentially transplant them back into the patient.

Non-Human Model Organisms

Researchers choose organisms that mirror the pathology of the human disease being studied.

• Yeast ($Saccharomyces \, cerevisiae$): Used for studying fundamental biochemical processes and "complementation" experiments.

  • Complementation Example: In a study on chromosome instability, researchers identified $621$ essential yeast genes. They replaced these with human orthologs and found $58$ cases where the yeast survived normally with the human gene. Researchers then introduced human variants into these "humanized" yeast to see which ones caused cell death.

• $C. \, elegans$ (Microscopic Worm):

  • Attributes: Incredibly simple ($1,000$ cells total, exactly $302$ of which are neurons). Same number of protein-coding genes as humans ($\approx 25,000$).

  • Utility: Excellent for tracking the exact lineage of every cell; used extensively in neurobiology and aging research.

• $Drosophila$ (Fruit Fly):

  • Attributes: $65\%$ of human genes have a fly ortholog.

  • Utility: Quick life cycle and easy to cultivate in large numbers in small jars for genetic screening.

• Zebrafish:

  • Attributes: Vertebrates; fully transparent during the larval stage.

  • Case Study: Used to study Adult-Onset Limb Girdle Muscular Dystrophy. By injecting RNA with human variants into zebrafish embryos, researchers could visualize perturbed muscle structures similar to those in human biopsies.

• Mice:

  • Pros: Mammals with the most similar physiology and genome to humans.

  • Cons: Expensive ($\approx \$50,000$ per transgenic line), ethically sensitive, and long-term.

  • Experimental Types: Germline knockout (from fertilization), adult-onset knockout, tissue-specific manipulation, and global overexpression.

Case Study: The $FTO$ Gene and Obesity

• The Variant: A highly correlated variant associated with obesity was found in the intron of the $FTO$ gene via GWAS (Genome-Wide Association Studies).

• The Challenge: Despite various mouse experiments (global overexpression, tissue-specific changes, etc.) and human data analysis, it remains difficult to prove that this specific intronic variant causes obesity.

• Complexity: The study illustrates that correlations are high, but functional proof is difficult because the variant likely affects the RNA regulation of the gene rather than the protein sequence.