Transcription produces RNA, a single-stranded copy of a gene’s DNA sequence. The primary product is a freshly built RNA strand assembled one nucleotide at a time by an enzyme called RNA polymerase. Depending on the gene being read, that RNA molecule can be messenger RNA (mRNA), ribosomal RNA (rRNA), transfer RNA (tRNA), or one of several smaller regulatory RNA types, each with a distinct job in the cell.
The Main RNA Products
Not all genes code for proteins. When people think of transcription, they usually picture mRNA, the molecule that carries protein-building instructions from DNA to the cell’s ribosomes. But mRNA is only one of three major RNA types produced. Ribosomal RNA makes up the structural and catalytic core of ribosomes themselves, and transfer RNA acts as the adapter that matches amino acids to the correct spot on the mRNA during protein assembly. All three are essential for turning genetic information into functional proteins.
In human and other eukaryotic cells, each type gets its own dedicated copying enzyme. RNA polymerase I produces rRNA, RNA polymerase II synthesizes mRNA, and RNA polymerase III handles tRNA. Bacteria simplify the process with a single RNA polymerase that produces all three.
Beyond these major classes, transcription also generates smaller, less well-known RNA molecules. Small nuclear RNAs (snRNAs), typically 100 to 300 nucleotides long, combine with proteins to form the machinery that splices mRNA. Small nucleolar RNAs (snoRNAs) chemically modify other RNA molecules. MicroRNAs (miRNAs), just 21 to 23 nucleotides long, silence genes by blocking mRNA from being translated into protein. Many of these are transcribed from their own genes, though some are encoded within the non-coding stretches (introns) of other genes.
How the RNA Strand Is Built
RNA polymerase reads one strand of the DNA double helix and assembles a complementary RNA strand from four building blocks: the ribonucleoside triphosphates ATP, GTP, CTP, and UTP. These molecules do double duty in the cell. ATP and GTP are also familiar as energy carriers, but during transcription they serve as raw materials. Each time the polymerase adds a nucleotide, it clips off two phosphate groups, releasing energy that drives the bond formation. The new RNA strand grows in one direction only, from its 5′ end toward its 3′ end.
To begin, the polymerase must find the right starting point on the DNA. In bacteria, a subunit called the sigma factor recognizes short signal sequences in the promoter region, specifically a six-nucleotide stretch about 35 base pairs upstream and another about 10 base pairs upstream of the start site. In human cells, the process is more elaborate: a protein called TBP (TATA-binding protein) latches onto a promoter element known as the TATA box, then a cascade of additional transcription factors assembles in a specific order before RNA polymerase II can dock and begin copying.
Once positioned, the polymerase pries apart roughly 14 base pairs of the DNA double helix, creating a small opening called the transcription bubble. Inside this bubble, the template DNA strand is exposed, and the polymerase reads it one base at a time, matching each with the complementary RNA nucleotide. As it moves forward, the bubble travels with it, zipping the DNA back together behind it.
Speed and Accuracy
In human cells, RNA polymerase II moves fast. Averaged estimates place its speed between roughly 1,000 and 4,000 nucleotides per second, though some measurements suggest bursts above 50,000 nucleotides per minute for a single gene copy. Speed varies depending on the gene, the local DNA structure, and how many polymerases are working along the same stretch.
Transcription is less accurate than DNA replication, and that’s partly by design. Because RNA molecules are temporary and cells produce many copies, the occasional error matters less than a permanent mutation in DNA. In the bacterium E. coli, the error rate sits around 1 mistake per roughly 12,000 nucleotides copied. Yeast cells are about 20 times more accurate, making only about 1 error per 285,000 nucleotides. This difference likely reflects evolutionary pressure: organisms with smaller populations tend to evolve tighter quality control.
Where Transcription Happens
In bacteria, which lack a nucleus, transcription and protein synthesis happen in the same open compartment. Ribosomes can latch onto the growing mRNA and start translating it into protein before transcription is even finished. This simultaneous process makes bacterial gene expression extremely fast.
In human and other eukaryotic cells, transcription is confined to the nucleus. The RNA must be processed and then exported through nuclear pores before ribosomes in the cytoplasm can translate it. That physical separation creates an opportunity for quality control and regulation that bacteria don’t have.
Processing the Raw Transcript
In eukaryotic cells, the RNA that rolls off the polymerase isn’t the final product. The initial transcript, called pre-mRNA, undergoes three major modifications before it’s ready for use.
First, a chemical cap is added to the front (5′) end of the molecule while transcription is still underway. This cap protects the RNA from being chewed up by enzymes and later helps ribosomes recognize it. Second, when the polymerase reaches a specific signal sequence near the end of the gene, the RNA is cut and a tail of roughly 200 adenine nucleotides, the poly(A) tail, is added. This tail stabilizes the molecule and helps shuttle it out of the nucleus.
Third, the non-coding segments (introns) scattered throughout the pre-mRNA are precisely snipped out, and the remaining coding segments (exons) are stitched together in a process called splicing. The snRNA molecules mentioned earlier are central players here, forming the spliceosome complexes that carry out this cut-and-paste work. All three modifications, capping, polyadenylation, and splicing, happen while the transcript is still being made, coordinated through a section of the RNA polymerase itself that acts as a docking platform for processing machinery.
The result is a mature mRNA molecule, shorter than the original transcript, chemically protected at both ends, and carrying an uninterrupted set of protein-coding instructions ready for translation.
Why Multiple RNA Types Matter
The fact that transcription produces so many kinds of RNA reflects how cells use genetic information in layers. Only about 2% of the human genome codes for protein, yet a much larger fraction is transcribed into RNA. The non-coding RNAs produced during transcription regulate which genes get turned on or off, modify other RNA molecules, and build the molecular machines (like ribosomes and spliceosomes) that keep the cell running. Transcription, in other words, doesn’t just create the message. It creates much of the machinery that reads it.