DNA, mRNA, protein: the central dogma

All known cellular life stores hereditary information in DNA and uses two related processes to convert that information into the working machinery of the cell.

• Transcription. A segment of DNA is read by an enzyme (RNA polymerase) and copied into a molecule of messenger RNA (mRNA).
• Translation. The mRNA is read by a ribosome in triplet codons; each codon corresponds to one amino acid; the ribosome assembles the amino acids into a protein.

This flow — DNA → mRNA → protein — was first described by Francis Crick in 1957 as the central dogma of molecular biology. The dogma is structural: the same chain of information conversion happens in nearly every cell of nearly every organism, with bacteria, archaea, and eukaryotes using only slightly different protein machinery to do it.

Proteins are the workhorses of the cell — enzymes, receptors, structural fibers, transporters, antibodies. The information for which proteins to make, where, and when is stored in DNA. Editing DNA changes what proteins the cell can make. Editing or modifying mRNA changes what proteins get made in a single instance without altering the DNA itself. Each editing strategy targets a different level of this flow.

DNA is a polymer of four nucleotides: adenine (A), thymine (T), cytosine (C), and guanine (G). The molecule is a double helix: two complementary strands held together by base pairing — A pairs with T, C pairs with G.

A few quantitative features:

• The human genome has roughly $3 \times 10^9$ base pairs across 23 chromosomes (46 in a diploid cell).
• Only about 1–2% of the human genome codes for proteins; the rest is regulatory sequence, repetitive elements, introns, and sequence of less-understood function.
• The genome is organized into genes, each of which (in protein-coding genes) carries the information for one protein (sometimes multiple via alternative splicing).
• Genes contain exons (coding regions) and introns (non-coding regions removed during mRNA processing).
• Surrounding the gene are regulatory sequences — promoters, enhancers, silencers — that control when and where the gene is transcribed.

The complementary structure of the double helix is the structural reason DNA can be copied faithfully: separate the two strands, and each can serve as a template for synthesizing its partner. This is how cells replicate their DNA before division, and how laboratory techniques (PCR, sequencing) work.

Transcription happens inside the nucleus (in eukaryotes). RNA polymerase binds to a promoter sequence upstream of a gene, separates the two DNA strands, and synthesizes an RNA copy of one strand. The result is a pre-mRNA that undergoes processing — adding a 5' cap, splicing out introns, adding a 3' poly-A tail — to produce the mature mRNA.

The mature mRNA is exported to the cytoplasm, where it encounters a ribosome. The ribosome reads the mRNA in codons (three-base groups), each of which specifies one amino acid. The genetic code maps the 64 possible codons onto the 20 standard amino acids (with three codons serving as 'stop' signals).

Key numbers:

• 4 bases × 4 bases × 4 bases = 64 codons.
• 20 standard amino acids encoded.
• The code is redundant — most amino acids are coded by multiple codons. A single-base change might or might not change the amino acid (silent vs missense mutation).
• The code is nearly universal across all known cellular life, with minor variations in mitochondria and some organisms.

The mRNA is read 5' to 3', codon by codon, until a stop codon. The completed amino-acid chain folds into a three-dimensional protein structure determined largely by its sequence.

Information flows from stored DNA through transient mRNA to functional protein.

A mutation is any change in the DNA sequence. Mutations are categorized by their effect on the encoded protein.

• Silent mutation. A base change that produces a different codon for the same amino acid (because the code is redundant). The protein is unchanged.
• Missense mutation. A base change that produces a codon for a different amino acid. The protein has one amino acid replaced; effect depends on which amino acid and where.
• Nonsense mutation. A base change that creates a stop codon mid-gene. The protein is truncated, usually non-functional.
• Insertion or deletion (indel). Adding or removing bases. If the number is not a multiple of 3, all downstream codons shift — a frameshift — and the protein from that point is garbage.
• Structural variants. Larger rearrangements: duplications, inversions, translocations of chromosome segments.

A single-base change can have effects ranging from undetectable to lethal. The most striking examples are diseases caused by single point mutations:

• Sickle-cell disease — a single A-to-T mutation in the beta-globin gene replaces glutamic acid with valine; the resulting hemoglobin polymerizes under low oxygen, deforming red blood cells.
• Cystic fibrosis — most cases caused by a three-base deletion that removes one phenylalanine residue from the CFTR chloride channel.
• Huntington's disease — caused by an expansion of a CAG repeat in the huntingtin gene.

These examples motivate genome editing therapies: if one base or one short sequence is the cause, correcting it could restore function. The structural challenge is doing so reliably in the relevant cells without unintended changes elsewhere.

Three levels of intervention are conceptually possible.

• DNA editing (most permanent). Change the genome sequence in the cell. Every subsequent mRNA from that gene reflects the change; every daughter cell inherits the change.
• mRNA modulation. Add a synthetic mRNA (mRNA vaccines, mRNA replacement therapies); use antisense oligonucleotides to silence or splice-modify endogenous mRNA. Changes the protein output without altering the genome. Effects are temporary.
• Protein modulation. Conventional drugs that bind proteins and modify their activity. Effects last only while the drug is present.

Which level is appropriate depends on the disease and the cell type involved.

• Heritable monogenic diseases with cells that turn over (blood cells from bone marrow) are good candidates for DNA editing of the stem cells. The edited stem cell's descendants carry the correction; the correction is permanent.
• Acute conditions (some infections, transient suppressions) suit mRNA or protein-level intervention.
• Diseases of post-mitotic cells (neurons, cardiomyocytes) face a constraint: the cells do not divide, so editing must reach them in place and the edit must be efficient enough on a per-cell basis to produce phenotype change.

The choice of level matters for the size of the technical challenge, the durability of the effect, and the ethics of the intervention (DNA editing of germline cells is qualitatively different from somatic editing because the change is heritable across generations). The next lesson examines how CRISPR/Cas9 became the workhorse for DNA editing and what its biology actually is.

The same DNA sequence is present in nearly every cell of an organism. But a liver cell looks and acts nothing like a neuron. The difference is gene regulation — which genes are transcribed in which cells, when, and at what level.

Regulation operates at many levels:

• Transcription factors — proteins that bind to regulatory DNA sequences and turn nearby genes on or off.
• Epigenetic modifications — chemical marks on the DNA (methylation) and on the histone proteins around which DNA wraps. These marks change which genes are accessible without changing the underlying sequence.
• Non-coding RNAs — RNA molecules that do not code for protein but regulate other RNAs (microRNAs, long non-coding RNAs).
• mRNA stability and translation efficiency — how long an mRNA persists, how efficiently it is translated.

A practical consequence: many diseases involve regulatory dysfunction, not sequence changes. The DNA is fine; the regulation is wrong. Therapies for these diseases may need to act on transcription factors, epigenetic marks, or mRNA dynamics — not on the DNA itself. Editing DNA in these contexts may not help, and may have unintended downstream effects on regulation.

The genome editing therapies in clinical use today (Casgevy for sickle-cell) target sequence-driven diseases where correcting or compensating for the mutation produces clinical benefit. Many other diseases are not yet candidates for the same approach. Distinguishing 'this is a CRISPR target' from 'this requires regulatory intervention' is an active area of biological research.

Three structural points for the rest of the cursus.

• Information flows from DNA through mRNA to protein. Editing DNA changes all subsequent transcription; editing mRNA changes one round of translation; modulating protein changes activity in the moment. The choice of level determines durability, reversibility, and technical difficulty.
• Single-base changes can have outsize phenotypic effects. This is what motivates editing therapies: many diseases have a known causal mutation; correcting it could restore function.
• Not every disease is a sequence disease. Many involve regulation, expression timing, or epigenetic state. Sequence editing alone does not address these. The space of editing therapies is a subset of the space of diseases.

The next lesson examines the CRISPR/Cas9 system — its biological origin as a bacterial immune system, the components needed to repurpose it for human cells, and the structural constraints (off-target effects, delivery, double-strand break repair pathways) that the editing therapy field has been working through since 2012.