Transcription looks like a massive molecular machine clamping onto DNA, prying the two strands apart, and threading one strand through a deep groove where it’s read and copied into a fresh RNA molecule. At the molecular scale, it’s a precise, mechanical process: a bulky enzyme grips the DNA, bends it sharply, and races along at speeds up to 4,300 nucleotides per minute in human cells. Here’s what that actually looks like at each stage.
The Setup: Building a Launch Pad on DNA
Before any RNA gets made, a cluster of proteins has to find the right starting point on the DNA and assemble into what’s called a preinitiation complex. Picture a stretch of DNA lying relatively still. First, a recognition protein lands on a specific sequence in the promoter region, a short stretch of DNA that signals “start here.” This protein bends the DNA sharply, creating a physical landmark that other proteins can find.
From there, additional factors arrive one by one in a specific order. A stabilizing pair locks the first protein in place. Then a bridging factor arrives and lays down a platform for the main enzyme, RNA polymerase II, to dock. RNA polymerase II doesn’t show up alone. It arrives already bound to another factor, and together they settle onto the still-closed double helix. Finally, two more factors join and trigger the architectural shift that pries the DNA strands apart. The full assembly forms a ring-like structure: three large complexes encircle a core of six proteins and the polymerase, all wrapped around and gripping the DNA from multiple angles.
What the Enzyme Looks Like Up Close
RNA polymerase II is enormous by molecular standards, made up of twelve protein subunits. Its most distinctive feature is a deep, positively charged groove called the cleft, which runs along one face of the enzyme. Think of it like a channel carved into the surface. Above this cleft sits a hinged structure called the clamp, which can swing open or closed.
Before transcription starts, double-stranded DNA sits above the cleft. Once the strands separate, the single template strand drops down into the groove and reaches the active site, the spot where new RNA nucleotides are added. The clamp then closes around the DNA, locking the enzyme in place so it doesn’t fall off mid-job. Deep inside the cleft, there’s a structure called the wall where the newly made RNA peels away from the DNA template. At this wall, the DNA makes a sharp 90-degree turn and exits the enzyme. So if you could see inside the working polymerase, you’d find a short hybrid stretch where RNA is temporarily paired with DNA, and then the two splitting apart as the DNA bends and leaves through a separate exit channel.
The Copying Process in Motion
Once the enzyme locks on and the DNA strands separate, transcription enters its elongation phase. This is the part most people picture when they think of transcription: the polymerase sliding along DNA, reading one strand and assembling a complementary RNA molecule one nucleotide at a time.
In human cells, this happens fast. RNA polymerase II moves at an average rate of 1,300 to 4,300 base pairs per minute. For context, a typical human gene might span tens of thousands of base pairs, so copying a gene can take anywhere from a few minutes to over an hour depending on its length. Bacterial RNA polymerase is actually slower in raw speed, clocking around 600 to 1,200 base pairs per minute in single-molecule experiments, though bacteria compensate by having much shorter genes and no interruptions in their coding sequences.
As the polymerase moves, a small “bubble” of separated DNA travels with it. Ahead of the enzyme, the double helix unwinds. Behind it, the two DNA strands snap back together. The growing RNA strand trails out of the enzyme like a tail, getting longer with each passing second. The process isn’t perfectly smooth. The polymerase pauses frequently, sometimes backing up a few nucleotides before resuming. These pauses aren’t errors; they serve as regulatory checkpoints and help the enzyme navigate tricky sequences or wait for signals from the cell.
How It Differs in Bacteria and Human Cells
The most striking visual difference between transcription in bacteria and in human cells is what happens to the RNA as it emerges from the polymerase. In bacteria, there’s no membrane separating the DNA from the rest of the cell. Ribosomes, the protein-building machines, can grab onto the growing RNA strand while it’s still being made. Under an electron microscope, this looks like a “Christmas tree” pattern: the polymerase sits on the DNA, the RNA extends outward, and ribosomes are already strung along the RNA like beads, translating it into protein in real time. Transcription and translation happen simultaneously, physically coupled in the same space.
In human cells, the picture is completely different. DNA sits inside the nucleus, sealed off by a double membrane. The RNA has to be finished, processed (its non-coding sections spliced out, a cap added to one end, a tail to the other), and then exported through nuclear pores before ribosomes ever touch it. So if you could peer into a human cell, transcription and translation would be happening in entirely separate compartments, potentially minutes or hours apart.
What Transcription Looks Like Under a Microscope
For decades, transcription was something scientists could only infer from indirect experiments. That changed with techniques that make active transcription visible in living cells. The most widely used approach involves tagging RNA with small molecular loops that glow when a fluorescent protein binds to them. Researchers insert a series of these loops (sometimes 128 copies in a row) into a gene of interest. As the gene gets transcribed, each loop in the emerging RNA grabs a fluorescent protein, creating a bright dot at the transcription site that’s visible under a confocal microscope.
What you actually see through the microscope is a speckled nucleus with individual bright spots blinking on and off. Each spot represents a single gene being actively transcribed. The spots appear when transcription starts and fade when it stops. By filming these spots over time, researchers can measure exactly how long a gene stays active, how frequently it fires, and how transcription pulses on and off in individual cells. The classic version of this is the “Christmas tree” image from electron microscopy of ribosomal genes: a central DNA fiber with progressively longer RNA strands branching off it, each representing a polymerase that started at a different time, the earliest ones having produced the longest strands.
The Overall Shape of the Process
If you zoomed out and watched the whole thing unfold, transcription looks like a three-act sequence. First, a crowd of proteins converges on a specific spot on the DNA over the course of seconds to minutes, building a launch complex through a precise series of handoffs. Second, the polymerase fires and begins racing along the gene, trailing a growing RNA strand behind it while maintaining a tight grip on the DNA template through its clamp mechanism. Third, when the polymerase reaches a termination signal, it releases both the completed RNA and the DNA, the bubble closes, and the enzyme falls off.
Across the genome, thousands of these events are happening simultaneously. Some genes have dozens of polymerases lined up one behind the other, all copying the same stretch of DNA at once. Others fire rarely, producing just a handful of RNA copies per hour. The result, at the scale of the whole nucleus, is a dynamic landscape of blinking transcription sites, each one a tiny molecular factory converting genetic information into the RNA instructions that drive the cell’s behavior.