The Nucleolus and Ribosome Assembly

The nucleolus is the site within the nucleus where ribosomal RNA is synthesized, processed, and assembled with ribosomal proteins into the large and small ribosomal subunits. It is not bounded by a membrane, which distinguishes it fundamentally from organelles like the mitochondrion or the ER. Instead, it exists because a huge local concentration of RNA polymerase I, ribosomal DNA (rDNA), nascent rRNA transcripts, and hundreds of processing and assembly factors creates a distinct biochemical and biophysical environment — a droplet-like phase separated from the surrounding nucleoplasm, held together by multivalent interactions between RNA and RNA-binding proteins with intrinsically disordered regions. Each human cell contains roughly 200 to 400 copies of the ribosomal RNA gene, organized in tandem repeats on the short arms of five acrocentric chromosome pairs (chromosomes 13, 14, 15, 21, and 22 in humans). These chromosomal loci are the nucleolar organizer regions, or NORs. When a cell exits mitosis, RNA polymerase I begins transcribing rDNA at active NORs, and the nucleolus reforms de novo around this transcriptional activity. Because nucleolar existence is contingent on transcription, the nucleolus is best understood not as a "thing" but as a process made visible — a structure that exists only because rRNA synthesis is happening continuously. The product of this factory is not one molecule but an entire macromolecular machine: the ribosome. A eukaryotic ribosome comprises a small subunit (40S in mammals) and a large subunit (60S), which together form the 80S ribosome. The small subunit contains the 18S rRNA plus about 33 ribosomal proteins; the large subunit contains 28S, 5.8S, and 5S rRNAs plus roughly 47 ribosomal proteins. Three of these four rRNAs (18S, 5.8S, and 28S) are transcribed together as a single long precursor inside the nucleolus and then cleaved apart; the fourth, 5S rRNA, is transcribed separately by RNA polymerase III, largely outside the nucleolus proper, and imported into the assembly line. This article builds toward Ribosomes: Structure and Function, where the finished machine and its catalytic mechanism are treated in full.

Step 1 — Transcription of the 45S pre-rRNA. RNA polymerase I binds rDNA promoters within active NORs, assisted by the transcription factors UBF (upstream binding factor) and SL1 (selectivity factor 1), and transcribes a single long precursor transcript. In humans this is called the 47S (or 45S, depending on convention) pre-rRNA, containing the 18S, 5.8S, and 28S rRNA sequences separated by internal and external transcribed spacers that will be discarded. A single nucleolus can contain hundreds of rDNA genes being transcribed simultaneously, each studded with dozens of RNA polymerase I molecules, producing a visually striking "Christmas tree" pattern of increasingly long nascent transcripts along each gene, first captured directly by electron microscopy (see Evidence below).

Step 2 — Co-transcriptional processing begins in the fibrillar center and dense fibrillar component. Even before transcription of a single rRNA gene finishes, small nucleolar ribonucleoprotein particles (snoRNPs) bind the nascent transcript. Box C/D snoRNPs guide 2′-O-methylation of specific ribose sugars; box H/ACA snoRNPs guide pseudouridylation of specific uridines. These chemical modifications, numbering well over 100 sites across the mature rRNAs, fine-tune rRNA folding and ribosome function. The very first snoRNP to assemble, U3 snoRNP, is essential for the first cleavages that separate the transcript into pre-40S and pre-60S processing streams.

Step 3 — Cleavage of pre-rRNA into subunit-specific precursors. A cascade of endonucleolytic and exonucleolytic cuts, carried out by dedicated nucleases, removes the external and internal transcribed spacers and separates the 18S rRNA (destined for the small subunit) from the 5.8S and 28S rRNAs (destined for the large subunit). This occurs in an ordered, hierarchical sequence — cut one spacer, and only then can the next enzyme access its site — which acts as a built-in checkpoint: if an earlier step fails, later cuts cannot proceed, and defective intermediates are degraded rather than passed forward.

Step 4 — Assembly with ribosomal proteins and import of 5S rRNA. As processing proceeds, ribosomal proteins — synthesized on cytoplasmic ribosomes and imported through the nuclear pore complex with the help of dedicated karyopherins — bind their rRNA partners in a defined temporal order, not randomly. The 5S rRNA, transcribed separately by RNA polymerase III in the nucleoplasm, is loaded together with ribosomal proteins RPL5 and RPL11 as a pre-formed 5S RNP that docks onto the maturing pre-60S particle. Roughly 200 distinct assembly factors — ATPases, GTPases, kinases, and RNA helicases — transiently bind and release in sequence, acting like a scaffold and quality-control crew rather than becoming part of the final ribosome.

Step 5 — Maturation in the granular component and nucleoplasmic quality control. Pre-40S and pre-60S particles migrate outward from the dense fibrillar component into the granular component, the outermost and largest nucleolar subregion, where the bulk of protein assembly and particle compaction occurs. Immature particles carry placeholder proteins and assembly factors that must be stripped off and exchanged for final ribosomal proteins before export is permitted — a proofreading mechanism ensuring that incompletely folded or incorrectly assembled subunits are retained and either fixed or degraded rather than released into the cytoplasm.

Step 6 — Nuclear export. Mature-enough pre-40S and pre-60S particles are handed to export receptors (notably exportin-1/CRM1 acting through adaptor proteins) and pass through nuclear pore complexes into the cytoplasm. Final maturation steps — including the very last proofreading cleavages and the loading of a handful of remaining proteins — occur in the cytoplasm, so that a subunit only becomes translation-competent after passing several checkpoints on both sides of the nuclear envelope. The two subunits do not join into a functional 80S ribosome until they encounter an mRNA and initiation factors during translation initiation, which is covered in Ribosomes: Structure and Function.

Why no membrane is needed. The nucleolus's three subregions are not membrane-bound compartments but concentration gradients of biochemical activity, maintained by liquid-liquid phase separation. Nucleolar proteins such as nucleophosmin (NPM1) and fibrillarin contain intrinsically disordered, low-complexity domains that make many weak, multivalent contacts with rRNA and with each other. These contacts drive spontaneous demixing from the surrounding nucleoplasm, similar to oil droplets separating from water, but tuned so that different proteins partition into different sub-droplets (fibrillar center vs. dense fibrillar component vs. granular component) based on their particular stickiness and the maturation stage of RNA they bind. This explains why nucleoli behave like liquid droplets that can fuse, drip, and reform, and why they vanish within minutes if RNA polymerase I transcription is inhibited (for example by the drug actinomycin D at low doses) — remove the RNA scaffold, and the phase separation collapses.

Barbara McClintock and the discovery of the nucleolar organizer (1934). Working with maize chromosomes, McClintock identified a specific constricted region on certain chromosomes that was always physically associated with the nucleolus and was required for its formation. She named this locus the "nucleolar organizer" and showed cytologically that chromosome breaks separating this region from the rest of the chromosome produced fragments that failed to organize a nucleolus, while the fragment retaining the organizer region did. This was the first direct evidence linking a specific chromosomal locus to nucleolar formation, decades before anyone knew the region contained rRNA genes.

Ritossa and Spiegelman: molecular identification of rDNA (1965). Ferruccio Ritossa and Sol Spiegelman used DNA-RNA hybridization in Drosophila to show that ribosomal RNA specifically hybridizes to DNA located at the nucleolar organizer region on the X and Y chromosomes. Crucially, they compared flies with one, two, or three doses of the nucleolar organizer (using a partial deletion, "bobbed," mutant background) and found that rRNA gene dosage — the number of rDNA repeats — scaled with the number of NOR copies, directly proving that the NOR is the physical location of the ribosomal RNA genes themselves, not merely a region that happens to sit near the nucleolus.

Oscar Miller's electron micrographs: watching transcription directly (late 1960s). Oscar L. Miller Jr. developed a spreading technique that gently lysed nuclei and spread their chromatin content onto grids for electron microscopy without shearing away nascent transcripts. Applied to amphibian oocyte nucleoli — chosen because oocytes contain hugely amplified rDNA, up to a million extra copies, to stockpile ribosomes for early embryogenesis — this technique produced now-iconic images. Each rDNA gene appeared as an axis studded with a graded series of increasingly long fibers, each fiber being a nascent RNA transcript attached to an RNA polymerase molecule, giving the classic "Christmas tree" or "feather" appearance. This was direct, visual, unambiguous proof that RNA polymerases move along rDNA synthesizing progressively longer transcripts, exactly as the transcription model predicted, and it let researchers count the density of polymerases per gene and estimate elongation rates directly from images rather than inferring them indirectly.

Fluorescence recovery after photobleaching (FRAP) reveals liquid, not solid, structure (2000s). A pivotal shift in understanding the nucleolus came from live-cell imaging studies (notably from groups including Tom Misteli's and later Timothy Mitchison's, among others building on this foundation) that fused nucleolar proteins such as fibrillarin or nucleophosmin to GFP and photobleached a spot within the nucleolus. Fluorescence recovered within seconds to a couple of minutes, showing that nucleolar proteins are highly mobile and constantly exchanging with the surrounding nucleoplasm rather than being locked into a rigid scaffold. Complementary experiments showed that two nucleoli can fuse into one spherical droplet when they touch, and that nucleoli round up after being deformed — hallmark behaviors of liquids governed by surface tension, not solids. Together with the observation that low-dose actinomycin D (which selectively blocks RNA polymerase I) causes nucleoli to dissolve within tens of minutes, this evidence base established the now-standard model: the nucleolus is a phase-separated liquid condensate assembled around an actively transcribed RNA scaffold, not a pre-built compartment that gets filled.

Example 1 — Estimating rRNA gene dosage from a nucleolar organizer deletion. A diploid organism has two nucleolar organizer regions, one per homologous chromosome, each carrying 250 tandem copies of the rRNA gene. A geneticist recovers a partial deletion allele, bobbed-1 (bb⁻), that removes 70% of the rDNA repeats on one homolog only. Wild-type diploid rDNA dosage: 250 + 250 = 500 copies. Heterozygote (one wild-type NOR, one bb⁻ NOR): 250 + (250 × 0.30) = 250 + 75 = 325 copies, or 65% of wild-type dosage. If a second independent deletion allele removing 40% of repeats is combined in trans (bb⁻/bb⁻ compound heterozygote, each homolog independently deleted): (250 × 0.30) + (250 × 0.60) = 75 + 150 = 225 copies, or 45% of wild-type dosage. Ritossa and Spiegelman's hybridization method predicts, and experimentally confirmed, that the amount of rRNA-hybridizing DNA detected scales linearly with these copy numbers. Phenotypically in Drosophila, reduced rDNA dosage below roughly 50% of wild-type produces the visible bobbed phenotype (short, thin bristles) because ribosome output becomes limiting in rapidly growing tissues such as bristle-forming cells — a direct, quantitative link between gene copy number, ribosome supply, and an organismal phenotype.

Example 2 — Estimating nucleolar transcriptional output in a growing cell. Consider a cultured human cell with 300 active rRNA gene copies (a mix of active NORs across its five acrocentric chromosome pairs; not all copies are transcribed at once — typically about half are silenced by heterochromatin). Suppose each active gene carries an average of 80 RNA polymerase I molecules simultaneously transcribing it (consistent with the dense polymerase loading seen in Miller spreads), and each polymerase elongates at roughly 40 nucleotides per second. Transcripts completed per gene per hour: the 47S pre-rRNA precursor is about 13,000 nucleotides long, so one full transcript takes 13,000 ÷ 40 = 325 seconds ≈ 5.4 minutes. With 80 polymerases per gene queued along its length, once the pipeline is full a new completed transcript emerges roughly every (325 s ÷ 80) ≈ 4 seconds per gene, i.e., about 15 transcripts per gene per minute. Across 150 active genes (half of 300): 150 × 15 = 2,250 pre-rRNA molecules completed per minute, or 135,000 per hour, from this one cell's active NORs. Since each pre-rRNA precursor eventually yields one 18S, one 5.8S, and one 28S rRNA (destined for one 40S and one 60S subunit, pending assembly with proteins and 5S rRNA), this single cell is capable of initiating well over 100,000 new ribosomal subunit precursors per hour — consistent with independent estimates that actively growing mammalian cells assemble on the order of several thousand ribosomes per minute. This calculation illustrates why even modest reductions in active NOR number or Pol I loading (as in ribosomopathies, discussed below) can create a substantial ribosome supply bottleneck.

Barbara McClintock's identification of the nucleolar organizer emerged from her painstaking cytogenetic work on maize chromosomes at Cornell and later Cold Spring Harbor Laboratory. Working before DNA was known to be the genetic material, she and her contemporaries relied entirely on light microscopy of stained, squashed chromosomes to correlate visible chromosomal landmarks with heritable traits and cellular structures. Her 1934 paper describing the nucleolar organizer showed extraordinary rigor: by studying chromosome fragments produced by X-ray-induced breaks, she demonstrated that nucleolus formation tracked with a specific physical chromosomal segment, establishing a causal (not merely correlative) link years before the molecular nature of that segment — tandemly repeated rRNA genes — was understood. This same meticulous, hypothesis-driven cytogenetic approach later led her to discover transposable elements, for which she received the Nobel Prize in Physiology or Medicine in 1983, decades after the initial work.

Cancer diagnosis and biology. Pathologists have used enlarged, irregular nucleoli as a microscopic marker of malignancy since the 19th century, long before the mechanism was known; today this is understood mechanistically, since oncogenic signaling (e.g., MYC overexpression) drives ribosome biogenesis to fuel unchecked growth, making nucleolar hypertrophy a direct, visible readout of oncogenic drive. Drugs such as CX-5461 selectively inhibit RNA polymerase I transcription and are in clinical trials as a strategy to starve cancer cells of ribosome supply while sparing normal cells relatively better, because tumor cells are often more dependent on high Pol I output.

Ribosomopathies. A class of human genetic diseases, including Diamond-Blackfan anemia, Treacher Collins syndrome, and Shwachman-Diamond syndrome, arise from mutations in ribosomal proteins or assembly factors that impair nucleolar ribosome production. Curiously, these diseases produce tissue-specific defects (especially in blood and craniofacial development) despite ribosomes being needed in every cell, illustrating that some tissues have unusually low tolerance for reduced ribosome supply during rapid developmental proliferation.

Antibiotic and antifungal development. Because rRNA processing and nucleolar assembly pathways differ enough between species, understanding nucleolar assembly factors in pathogenic fungi and parasites has opened avenues for selective drugs that disrupt ribosome biogenesis in the pathogen while sparing the host's nucleolar machinery.

Basic cell biology and synthetic biology. The nucleolus remains the paradigm system for studying membraneless organelles and liquid-liquid phase separation, informing broader research into stress granules, P-bodies, and other biomolecular condensates now implicated in neurodegenerative diseases such as ALS and frontotemporal dementia, where aberrant phase transitions of RNA-binding proteins are a central disease mechanism.

Agricultural and reproductive biology. Amplification of rDNA and massive nucleolar activity in oocytes (used originally by Miller to visualize transcription so clearly) is exploited in reproductive biology and assisted reproduction research, since oocyte ribosome stockpiles determine early embryo developmental competence before the embryo's own genome activates.

Ask the tutor: Why does inhibiting nucleolar rRNA transcription trigger a p53-dependent stress response, and how do cancer cells exploit or evade this checkpoint?