Cellular organization: organelles

Intuition [Beginner]

A eukaryotic cell is a compartmentalised machine. Instead of letting all reactions happen in one shared space, the cell divides its interior into membrane-bound organelles, each specialised for a particular set of tasks. This division increases efficiency: enzymes and substrates are concentrated together, incompatible reactions are separated, and conditions (pH, ion concentrations, redox state) can be tuned for each compartment.

The nucleus stores the cell's DNA and is the site of transcription (making RNA from DNA templates). The nuclear envelope is a double membrane penetrated by nuclear pore complexes, which regulate what enters and exits. The nucleolus, a substructure within the nucleus, assembles ribosomal RNA.

The endoplasmic reticulum (ER) is a network of membrane-enclosed tubes and sacs extending from the nuclear envelope. The rough ER (studded with ribosomes) synthesises proteins destined for secretion, for the plasma membrane, or for organelles of the endomembrane system. The smooth ER synthesises lipids and detoxifies drugs.

The Golgi apparatus receives proteins from the ER, modifies them (adding carbohydrate groups, trimming sugar chains, proteolytic cleavage), sorts them, and packages them into vesicles for delivery to their final destinations.

Mitochondria are the power plants of the cell. They oxidise pyruvate (from glucose) through the citric acid cycle and use the released electrons to generate ATP through oxidative phosphorylation. Mitochondria have their own DNA and ribosomes, reflecting their evolutionary origin as free-living bacteria that were engulfed by an ancestral eukaryotic cell (endosymbiosis).

Lysosomes are acidic (pH ~5) compartments containing digestive enzymes (acid hydrolases) that break down macromolecules, worn-out organelles (autophagy), and ingested material. Peroxisomes contain oxidases that produce hydrogen peroxide and catalase that decomposes it, protecting the cell from oxidative damage.

Visual [Beginner]

The endomembrane system connects organelles through vesicular traffic:

The secretory pathway flows: nucleus (mRNA) to cytoplasm (ribosomes on rough ER) to ER lumen to Golgi to final destination. This is a directional pathway, but retrograde transport also occurs to recycle membrane components and escaped ER proteins.

Worked example [Beginner]

Tracing a secretory protein from synthesis to secretion. Consider insulin, produced by beta cells in the pancreas.

Step 1. Synthesis on rough ER. The insulin mRNA is translated by ribosomes that dock on the rough ER membrane. The nascent polypeptide enters the ER lumen through a translocon (a protein channel). A signal sequence at the N-terminus of preproinsulin targets the ribosome to the ER; this signal is cleaved off inside the lumen, producing proinsulin.

Step 2. Folding and disulfide bond formation in the ER. Proinsulin folds into its native conformation. Three disulfide bonds form between cysteine residues. ER chaperone proteins (BiP, calnexin) assist folding, and a quality-control system ensures only correctly folded proteins proceed.

Step 3. Vesicular transport to Golgi. Correctly folded proinsulin is packaged into COPII-coated vesicles that bud from the ER and fuse with the cis-Golgi (the receiving face).

Step 4. Modification in Golgi. In the trans-Golgi network, a convertase enzyme cleaves proinsulin at two sites, removing the C-peptide and producing mature insulin (two polypeptide chains, A and B, linked by disulfide bonds). This proteolytic activation is the key modification step.

Step 5. Packaging and secretion. Mature insulin is concentrated into secretory granules that bud from the trans-Golgi. When blood glucose rises, the granules fuse with the plasma membrane (exocytosis), releasing insulin into the bloodstream.

Check your understanding [Beginner]

Formal definition [Intermediate+]

The endomembrane system comprises the nuclear envelope, ER, Golgi apparatus, lysosomes, vacuoles, transport vesicles, and plasma membrane. These compartments are connected by vesicular traffic and share a common evolutionary origin. Notably, mitochondria, peroxisomes, and chloroplasts are not part of the endomembrane system.

Protein sorting signals are short amino acid sequences or structural motifs that direct proteins to their correct compartments:

Signal type	Location	Target	Example
Signal sequence	N-terminus (cleaved)	ER lumen / secretory pathway	Preproinsulin signal peptide
Signal-anchor	Internal	ER membrane (type II)	Asialoglycoprotein receptor
Nuclear localization signal (NLS)	Internal	Nucleus	PKKKRKV (SV40 large T antigen)
Nuclear export signal (NES)	Internal	Cytoplasm	LQLPPLERLTL (HIV Rev)
Mitochondrial targeting signal	N-terminus (cleaved)	Mitochondrial matrix	MLSLRQSIRFFKPATRT
Peroxisomal targeting signal 1 (PTS1)	C-terminus (not cleaved)	Peroxisome	SKL (Ser-Lys-Leu)
KDEL / HDEL	C-terminus	ER retention	KDEL (Lys-Asp-Glu-Leu)

Vesicular transport moves cargo between compartments via coated vesicles:

• COPII-coated vesicles: ER to Golgi (anterograde). Coat proteins: Sec23/24 (cargo selection) and Sec13/31 (outer scaffold).
• COPI-coated vesicles: Golgi to ER (retrograde) and intra-Golgi. Mediated by the coatomer complex and ARF1 GTPase.
• Clathrin-coated vesicles: Trans-Golgi to lysosomes, plasma membrane to endosomes (endocytosis). Adaptor proteins (AP complexes) link cargo to clathrin.

SNARE proteins mediate vesicle-target membrane fusion. v-SNAREs on the vesicle (e.g., VAMP) pair with t-SNAREs on the target membrane (e.g., syntaxin + SNAP-25). The pairing pulls the two membranes together, overcoming the energy barrier for lipid bilayer fusion.

Key theorem with proof [Intermediate+]

Theorem (Signal-sequence sufficiency for ER targeting). The N-terminal signal sequence is both necessary and sufficient to target a protein to the ER lumen. If the signal sequence from a secretory protein is genetically fused to a normally cytosolic protein, the fusion protein is directed to the ER. Conversely, deletion of the signal sequence from a secretory protein causes it to remain in the cytosol.

Proof (experimental, by chimeric protein analysis).

Necessity: Deleting the signal sequence from preprolactin (a secreted protein) causes the truncated protein to be translated on free ribosomes and released into the cytosol, as demonstrated by cell-free translation assays (Blobel and Dobberstein, 1975).

Sufficiency: Fusing the preprolactin signal sequence to the N-terminus of a cytosolic protein (e.g., globin) redirects the fusion protein to the ER lumen, where the signal sequence is cleaved and the globin moiety is translocated across the membrane. This was demonstrated using in vitro translation systems with added microsomal membranes (isolated ER vesicles).

The mechanism involves: (1) the signal recognition particle (SRP) binds the signal sequence as it emerges from the ribosome; (2) SRP pauses translation and docks the ribosome at the SRP receptor on the ER membrane; (3) the ribosome engages the translocon (Sec61 complex); (4) SRP is released and translation resumes, feeding the growing polypeptide through the translocon into the ER lumen; (5) signal peptidase cleaves the signal sequence.

This is the signal hypothesis of Blobel and Sabatini (1971), confirmed experimentally and recognised by the Nobel Prize in Physiology or Medicine to Blobel in 1999.

Worked example: the KDEL retrieval pathway

Soluble ER resident proteins (such as BiP and protein disulfide isomerase) carry a C-terminal KDEL sequence (Lys-Asp-Glu-Leu in mammals). If these proteins escape the ER via COPII vesicles to the Golgi, the acidic pH of the Golgi causes KDEL to bind to the KDEL receptor in the Golgi membrane. The receptor-ligand complex is packaged into COPI-coated retrograde vesicles that return to the ER, where the neutral pH releases the protein. This retrieval system maintains ER resident protein concentration despite constant vesicular leakage.

Bridge. The signal hypothesis is the foundational reason that eukaryotic cells maintain chemically distinct compartments: each organelle's unique protein composition depends on address codes that direct newly synthesised polypeptides to the correct membrane. This is exactly the logic that extends beyond the ER to mitochondrial targeting sequences recognised by TOM/TIM complexes, peroxisomal PTS1 signals decoded by Pex5, and nuclear localisation sequences ferried by importins through nuclear pore complexes. The signal-sequence principle builds toward 17.05.01 pending transcription and translation, whose protein products — histones, DNA polymerases, transcription factors — must be actively imported into the nucleus by karyopherin-mediated transport. The central insight is that a short amino acid motif functions as a molecular postal code, and the cell dedicates an infrastructure of receptors, GTPases, and translocons to read and execute that code. The pattern recurs throughout cell biology: specificity in a system of many compartments emerges not from physical barriers alone but from active, signal-dependent sorting.

Exercises [Intermediate+]

Protein targeting and the secretory pathway [Master]

The signal recognition particle (SRP) is a ribonucleoprotein complex — six polypeptides (SRP9, SRP14, SRP19, SRP54, SRP68, SRP72) assembled on a 300-nucleotide 7S RNA scaffold — that recognises signal sequences as they emerge from the ribosomal exit tunnel. SRP54 is the key subunit: its methionine-rich M-domain forms a hydrophobic pocket that accommodates the signal peptide's core of 7-15 hydrophobic residues, while its N-domain contacts the ribosome near the exit tunnel. When SRP54 binds a signal sequence, SRP9/14 simultaneously pauses translation by inhibiting elongation factor binding. The ribosome-nascent chain-SRP complex then diffuses to the ER membrane, where the SRP receptor (SR, a heterodimer of SRα and SRβ, both GTPases) captures it. GTP binding by both SRP54 and SRα drives complex formation; GTP hydrolysis after translocon engagement releases SRP for another round.

The Sec61 translocon is a heterotrimer (Sec61α, β, γ) that forms a protein-conducting channel in the ER membrane. Sec61α has ten transmembrane helices arranged in two pseudo-symmetric halves that hinge open laterally toward the lipid bilayer when a signal peptide inserts. The ribosome docks on the cytoplasmic face with its exit tunnel aligned over the channel pore. As translation resumes, the nascent chain feeds through the channel into the ER lumen, with the signal peptide serving as both the initial targeting signal and a start-transfer signal that orients the polypeptide. For membrane proteins, subsequent transmembrane segments act as stop-transfer or signal-anchor sequences, partitioning into the lipid bilayer through the Sec61 lateral gate. Post-translational translocation in yeast (for proteins too short for co-translational targeting) uses the Sec62/63 complex to recruit BiP, which acts as a molecular ratchet in the ER lumen.

Protein folding in the ER is monitored by the calnexin/calreticulin cycle, a glycan-dependent quality-control system. Oligosaccharyltransferase (OST) transfers a preassembled Glc₃Man₉GlcNAc₂ oligosaccharide from dolichol pyrophosphate to asparagine residues in N-X-S/T sequons on nascent glycoproteins. Glucosidases I and II remove the outer two glucose residues, producing a monoglucosylated glycan recognised by the lectin chaperones calnexin (membrane-bound) and calreticulin (soluble). These chaperones bind the glycoprotein, preventing aggregation and allowing disulfide bond formation by protein disulfide isomerase (PDI). Glucosidase II then removes the final glucose, releasing the glycoprotein. If folding is incomplete, UDP-glucose glucosyltransferase (UGGT) acts as a folding sensor: it reglucosylates exposed hydrophobic patches on incompletely folded glycoproteins, returning them to the calnexin/calreticulin cycle. This cycle repeats until proper folding is achieved, at which point UGGT no longer recognises the protein and it exits to the Golgi.

Proteins that fail to fold after prolonged cycling are diverted to ER-associated degradation (ERAD). ER mannosidases (ERManI, EDEM proteins) trim mannose residues from the N-glycan, creating a timer signal that the protein has spent too long in the ER. Lectins OS-9 and XTP3-B recognise the trimmed glycan and deliver the substrate to the Hrd1 ubiquitin ligase complex. Hrd1, a multi-spanning membrane RING-finger E3 ligase, forms a retrotranslocation channel through which the substrate is extracted to the cytosol. The AAA-ATPase p97 (VCP/Cdc48), together with its cofactors Ufd1-Npl4, pulls the ubiquitinated substrate into the cytosol for degradation by the 26S proteasome.

When ERAD capacity is overwhelmed, the unfolded protein response (UPR) activates three ER transmembrane sensors. IRE1, a kinase-endoribonuclease, splices XBP1 mRNA to produce the active XBP1s transcription factor, upregulating chaperones, ERAD components, and lipid biosynthesis genes. PERK phosphorylates eIF2α, attenuating global translation while allowing preferential translation of ATF4, which induces CHOP and apoptotic genes under prolonged stress. ATF6 transits to the Golgi where S1P and S2P proteases cleave it, releasing its cytosolic domain as a transcription factor. If the UPR cannot restore proteostasis, CHOP-induced apoptosis eliminates the stressed cell — a mechanism relevant to pancreatic β-cell death in type 2 diabetes, where chronic high insulin demand overwhelms the ER folding capacity.

COPII-coated vesicles mediate anterograde transport from the ER to the Golgi. Vesicle formation begins when Sec12, a guanine nucleotide exchange factor (GEF), activates Sar1 by exchanging GDP for GTP. Sar1-GTP inserts its amphipathic N-terminal helix into the ER membrane, recruiting the Sec23/24 heterodimer. Sec24 is the primary cargo adaptor: it recognises di-acidic (DxE), di-hydrophobic (FF, FY, YY), and LxxLE sorting signals on transmembrane cargo, and captures soluble cargo via receptors such as ERGIC-53. The Sec13/31 heterotetramer polymerises into an outer lattice that deforms the membrane into a bud. Sec23 also acts as a GTPase-activating protein (GAP) for Sar1; hydrolysis after vesicle release disassembles the coat. Large cargo such as procollagen rods (300-400 nm) require the adaptor TANGO1/cTAGE5 and Sec16 to generate oversized COPII carriers, challenging the earlier assumption that all COPII vesicles are 60-70 nm spheres.

Retrograde transport from the Golgi to the ER uses COPI-coated vesicles, assembled by the coatomer complex (seven subunits: α, β, β', γ, δ, ε, ζ) and the small GTPase ARF1. COPI recognises two classes of retrieval signals: the C-terminal KDEL/HDEL sequence on soluble ER proteins (captured by the KDEL receptor, which binds ligand at Golgi pH 6.0-6.5 and releases it at ER pH 7.2), and the C-terminal KKXX or KXKXXX dibasic motifs on transmembrane ER proteins (recognised directly by α-COP and β'-COP). Clathrin-coated vesicles at the trans-Golgi network use adaptor protein complexes (AP-1, AP-3) to sort cargo toward endosomes and lysosomes, with each AP complex recognising specific tyrosine-based (YXXØ) or dileucine-based ([DE]XXXL[LI]) sorting motifs.

SNARE proteins mediate the final step of vesicle fusion. v-SNAREs on the transport vesicle pair with t-SNAREs on the acceptor membrane to form a four-helix coiled coil — the trans-SNARE complex. Progressive zippering from the N-terminal layers toward the C-terminal membrane anchors pulls the two bilayers within approximately 1 nm of each other, overcoming the hydration repulsion barrier. The energy released by SNARE complex formation (roughly 35-40 k_BT per complex) is sufficient to drive lipid mixing. After fusion, the NSF ATPase (with SNAP co-factors) disassembles the cis-SNARE complex, recycling the SNAREs for another round. Tethering complexes (COG, GARP, exocyst, HOPS) provide an additional layer of specificity by capturing vesicles before SNARE engagement, ensuring that vesicles fuse only with the correct target membrane.

The quantitative analysis of vesicular traffic uses kinetic rate equations. The steady-state concentration of a cargo protein in each compartment depends on the rates of forward and retrograde transport. For a linear pathway ER to Golgi to plasma membrane:

$$\frac{d[\text{Golgi}]}{dt}=k_{\text{in}}[\text{ER}]-k_{\text{out}}[\text{Golgi}]$$

At steady state: $[\text{Golgi}{]}_{\text{ss}}=(k_{\text{in}}/k_{\text{out}})[\text{ER}]$, and the transit time through the Golgi is $\tau =1/k_{\text{out}}$. Experimental measurements give $\tau \approx 20$-$30$ minutes for VSV-G protein in mammalian cells.

Mitochondrial structure-function and the endosymbiotic theory [Master]

Mitochondria are bounded by two membranes that differ radically in composition and function. The outer membrane (OMM) is permeable to molecules up to roughly 5 kDa due to porins (VDAC — voltage-dependent anion channels) that form aqueous pores. The inner membrane (IMM) is tightly impermeable: it has a high cardiolipin content (approximately 20% of phospholipid, unique among eukaryotic membranes) that stabilises the protein complexes of the electron transport chain 17.04.02 pending. The IMM folds into cristae — invaginations that increase surface area approximately fivefold. Cristae are not simple folds: they are connected to the peripheral IMM by narrow tubular openings called cristae junctions (12-40 nm diameter), which restrict diffusion of proteins and metabolites between the intercristae space and the intermembrane space proper. The MICOS complex (mitochondrial contact site and cristae organising system), anchored by Mic60/Mitofilin and Mic10, maintains cristae junction architecture.

The mitochondrial matrix contains the citric acid cycle enzymes, the mitochondrial genome, and the protein-folding chaperones Hsp60/Hsp10 and mtHsp70. The human mitochondrial genome is a 16,569 base-pair circular DNA molecule encoding 37 genes: 13 protein-coding genes (all subunits of the oxidative phosphorylation complexes I, III, IV, and V), 22 transfer RNAs, and 2 ribosomal RNAs. The remaining approximately 1,500 mitochondrial proteins are nuclear-encoded, synthesised on cytosolic ribosomes, and imported. This genetic asymmetry is the primary evidence against viewing mitochondria as autonomous symbionts; over evolutionary time, most endosymbiont genes transferred to the host nucleus.

The endosymbiotic theory, proposed by Lynn Sagan (Margulis) in 1967, posits that mitochondria descend from an α-proteobacterium engulfed by an archaeal host cell approximately 1.5-2 billion years ago ^{[Sagan 1967]}. The evidence is cumulative: mitochondria have their own circular DNA, distinct from the nuclear genome; their ribosomes (55S) resemble bacterial 70S ribosomes more than eukaryotic 80S ribosomes; mitochondrial translation is sensitive to chloramphenicol and tetracycline (bacterial ribosome inhibitors) but not cycloheximide; phylogenetic analysis of mitochondrial rRNA places mitochondria within the α-proteobacteria, with Rickettsia prowazekii as the closest living relative; the double membrane reflects the engulfment event; and cardiolipin, a hallmark of bacterial membranes, is abundant in the IMM but absent from other eukaryotic membranes.

Mitochondrial protein import uses two major translocase systems. The TOM complex (Translocase of the Outer Membrane) is the entry point for essentially all nuclear-encoded mitochondrial proteins. Its core subunit Tom40 forms a β-barrel channel, while peripheral receptors Tom20 and Tom22 recognise the N-terminal amphipathic helix of presequence-containing proteins. On the inner membrane, the TIM23 complex handles presequence-bearing precursors destined for the matrix or the IMM. Translocation through TIM23 requires the membrane potential (Δψ ≈ 180 mV, negative inside), which electrophoretically drives the positively charged presequence through the channel. In the matrix, the presequence-associated motor (PAM) — comprising mtHsp70, its nucleotide exchange factor Mge1, and the scaffold Tim44 — uses ATP hydrolysis to ratchet the polypeptide in. Mitochondrial processing peptidase (MPP) cleaves the presequence, and the mature protein folds, often with Hsp60/Hsp10 chaperonin assistance. A second inner-membrane translocase, TIM22, handles polytopic inner-membrane carriers (adenine nucleotide translocase, phosphate carrier, citrate carrier) using a distinct pathway that depends on Δψ but not on the PAM motor.

Mitochondrial shape is dynamic, controlled by opposing fission and fusion reactions. Fission is mediated by Drp1, a cytosolic dynamin-related GTPase recruited to the OMM by receptor proteins Mff, Fis1, and MiD49/51. Drp1 oligomerises into helical filaments that constrict and sever the membrane. ER-mitochondria contact sites mark the positions where fission occurs: the ER wraps around the mitochondrion and constricts it to approximately 150 nm before Drp1 is recruited, suggesting a two-step mechanism. Fusion of the outer membrane uses the mitofusins MFN1 and MFN2, which tether adjacent mitochondria through anti-parallel coiled-coil interactions, then undergo a conformational change powered by GTP hydrolysis that pulls the two OMMs together. Inner-membrane fusion is mediated by OPA1, a dynamin-related GTPase that exists in long (L-OPA1, membrane-anchored) and short (S-OPA1, proteolytically cleaved) isoforms; balanced processing by the AAA proteases YME1L and OMA1 maintains the ratio required for fusion. Mutations in MFN2 cause Charcot-Marie-Tooth type 2A (peripheral neuropathy), and mutations in OPA1 cause autosomal dominant optic atrophy — both reflecting the sensitivity of neurons to impaired mitochondrial dynamics.

Damaged mitochondria are selectively removed by mitophagy, a specialised form of autophagy. The PINK1-Parkin pathway is the best-characterised mechanism: PINK1, a serine/threonine kinase normally imported into healthy mitochondria and degraded by the IMM protease PARL, accumulates on the OMM of damaged mitochondria that have lost Δψ. Surface-accumulated PINK1 autophosphorylates and recruits Parkin, a cytosolic E3 ubiquitin ligase. Parkin ubiquitinates OMM proteins including VDAC1, MFN1, and MFN2, marking the mitochondrion for autophagic engulfment. Ubiquitin chains are recognised by autophagy receptors p62/SQSTM1, OPTN, and NDP52, which simultaneously bind LC3 on the phagophore membrane, linking the ubiquitinated mitochondrion to the autophagic machinery.

Lysosomes, peroxisomes, and autophagy [Master]

Lysosomes are membrane-bound organelles (0.1-1 μm diameter) containing approximately 60 acid hydrolases — proteases (cathepsins B, D, L), glycosidases, lipases, nucleases, and phosphatases — that collectively degrade all major classes of biological macromolecules. Their optimal pH of approximately 5.0 is maintained by the vacuolar H+-ATPase (V-ATPase), a rotary proton pump whose V₁ cytoplasmic sector hydrolyses ATP and whose V₀ transmembrane sector conducts protons. The electrochemical gradient generated by proton pumping is dissipated by a counter-ion conductance (primarily the CLC-7 Cl⁻/H⁺ antiporter), preventing the buildup of a positive interior potential that would stall the pump. This V-ATPase mechanism builds on the active transport principles introduced in 17.02.02.

Lysosomal enzymes are targeted via the mannose-6-phosphate (M6P) pathway. In the cis-Golgi, the enzyme GlcNAc-1-phosphotransferase (a heterohexameric complex of α₂β₂γ₂ subunits; the α and β subunits are encoded by GNPTAB) recognises a conformational signal patch on lysosomal hydrolases and transfers N-acetylglucosamine-1-phosphate to a high-mannose oligosaccharide. The uncovering enzyme removes the GlcNAc, exposing the M6P tag. In the trans-Golgi network, the M6P receptor captures tagged enzymes and packages them into clathrin-coated vesicles. After delivery to late endosomes (pH approximately 6.0), the lower pH causes receptor-ligand dissociation; the receptor recycles while the enzymes proceed to lysosomes. Deficiency in GlcNAc-1-phosphotransferase causes I-cell disease, in which lysosomal enzymes are secreted rather than targeted.

Lysosomal storage diseases (LSDs) result from deficiency of individual acid hydrolases, causing substrate accumulation. Tay-Sachs disease results from mutations in HEXA encoding the α subunit of β-hexosaminidase A; GM2 ganglioside accumulates in neuronal lysosomes, producing progressive neurodegeneration with onset at 4-6 months and death by age 3-4. Gaucher disease, the most common LSD, results from mutations in GBA encoding β-glucocerebrosidase; glucocerebroside accumulates in macrophage lysosomes, causing hepatosplenomegaly, bone pain, and cytopenias. Enzyme replacement therapy with recombinant imiglucerase (modified with terminal mannose residues to target macrophage lysosomes via mannose receptors) is effective for non-neurological manifestations. GBA mutations are also the strongest genetic risk factor for Parkinson's disease, linking lysosomal dysfunction to neurodegeneration.

The lysosomal surface is a signalling platform for mTORC1 (mechanistic target of rapamycin complex 1), the master regulator of cell growth. When amino acids are abundant, the Rag GTPase heterodimers (RagA/B with RagC/D) adopt a configuration that recruits mTORC1 from the cytosol to the lysosomal surface via the Ragulator scaffold. At the lysosome, the small GTPase Rheb (activated by growth-factor signalling through PI3K/Akt and inhibited by the TSC1/2 complex) directly activates the mTOR kinase domain. Amino acid sensing occurs through multiple channels: Sestrin2 binds leucine and dissociates from GATOR2; CASTOR1 binds arginine; SLC38A9 transports arginine out of the lysosome and signals luminal amino acid levels. Active mTORC1 phosphorylates S6K1 and 4E-BP1, promoting protein synthesis, lipid synthesis, and nucleotide synthesis while inhibiting autophagy via ULK1 phosphorylation. Under starvation, mTORC1 dissociates from the lysosome and autophagy activates.

The transcription factors TFEB and TFE3 control lysosomal biogenesis and autophagy gene expression through the CLEAR (Coordinated Lysosomal Expression and Regulation) network. When mTORC1 is active, it phosphorylates TFEB on Ser211, creating a binding site for 14-3-3 proteins that retain TFEB in the cytosol. Under starvation or lysosomal stress, calcineurin dephosphorylates TFEB, enabling nuclear translocation and activation of CLEAR target genes — a homeostatic feedback loop that expands lysosomal capacity on demand.

Macroautophagy proceeds through a defined sequence controlled by ATG (autophagy-related) gene products. Initiation begins with the ULK1 complex (ULK1-ATG13-FIP200-ATG101), which activates when mTORC1-mediated inhibitory phosphorylation is relieved and AMPK-mediated activating phosphorylation occurs under low energy. The activated ULK1 complex phosphorylates components of the class III PI3K complex (VPS34-Beclin-1-ATG14L), which produces PI(3)P at the phagophore assembly site. Two ubiquitin-like conjugation systems expand the phagophore: ATG7 (E1-like) and ATG10 (E2-like) conjugate ATG12 to ATG5, forming the ATG12-ATG5-ATG16L1 complex (E3-like ligase); and ATG7 with ATG3 conjugate LC3 to phosphatidylethanolamine (PE), forming LC3-II, which inserts into the phagophore membrane. LC3-II serves as a docking site for selective autophagy receptors (p62/SQSTM1, NBR1, OPTN, NDP52) that contain LC3-interacting region motifs and ubiquitin-binding domains, linking ubiquitinated cargo to the growing autophagosome. Upon closure, the autophagosome fuses with a lysosome via SNAREs (STX17, SNAP29, VAMP8) and the HOPS tethering complex.

Peroxisomes are single-membrane organelles that perform β-oxidation of very-long-chain fatty acids (VLCFAs, greater than C22), synthesise plasmalogens (ether phospholipids critical for myelin), and detoxify glyoxylate. Peroxisomal β-oxidation differs from mitochondrial β-oxidation: peroxisomal acyl-CoA oxidase transfers electrons directly to O₂, producing H₂O₂ (rather than feeding electrons into an electron transport chain), and the chain-shortening process stops at medium-chain length, requiring carnitine shuttling of products to mitochondria for completion. Catalase decomposes the H₂O₂. Peroxisomal matrix proteins are imported post-translationally in a folded state — unique among organelles. Most carry a C-terminal PTS1 signal (the tripeptide SKL or variant), recognised by the cytosolic receptor Pex5. Pex5 delivers cargo to the docking complex (Pex13/Pex14), inserts into the membrane as a transient translocation pore, and is then monoubiquitinated by the RING finger complex (Pex2/Pex10/Pex12) and extracted by the AAA ATPases Pex1/Pex6. Mutations in PEX genes cause Zellweger spectrum disorders: peroxisome biogenesis fails, VLCFAs accumulate, plasmalogens are deficient, and patients present with severe neurological impairment.

Organelle contact sites and lipid transfer [Master]

Membrane contact sites (MCS) are regions where two organelle membranes approach within 10-30 nm without fusing, tethered by protein complexes that bridge the gap. MCS are functional platforms for non-vesicular lipid exchange, calcium signalling, and organelle positioning. The ER makes contact sites with virtually every other organelle — mitochondria, plasma membrane, Golgi, endosomes, lipid droplets, and peroxisomes — consistent with its role as the cell's primary site of lipid synthesis and the largest membrane-bound compartment by surface area.

ER-mitochondria contact sites, historically called MAMs (mitochondria-associated membranes), occupy roughly 5-20% of the mitochondrial surface. The IP3R-GRP75-VDAC complex constitutes the calcium transfer channel: inositol 1,4,5-trisphosphate receptor (IP3R) on the ER releases Ca²⁺ from stores, GRP75 (an Hsp70 chaperone) bridges IP3R to VDAC on the mitochondrial outer membrane, and VDAC passes Ca²⁺ into the intermembrane space. The mitochondrial calcium uniporter (MCU) complex on the inner membrane then transports Ca²⁺ into the matrix. Ca²⁺ microdomains at these contact sites reach concentrations of 10-100 μM, far above the cytosolic average of approximately 100 nM, and are sufficient to activate mitochondrial dehydrogenases (pyruvate dehydrogenase, isocitrate dehydrogenase, α-ketoglutarate dehydrogenase), stimulating ATP production in response to ER calcium release. MFN2 also tethers ER to mitochondria: MFN2 on the ER engages MFN1 or MFN2 on the OMM in trans, and MFN2 knockout reduces contacts by approximately 50% and impairs calcium transfer.

ER-plasma membrane contact sites mediate store-operated calcium entry (SOCE), the primary mechanism by which cells replenish ER calcium stores. STIM1, the ER calcium sensor, contains a luminal EF-hand domain that binds Ca²⁺ at resting levels. When ER calcium is depleted, Ca²⁺ dissociates from STIM1, causing STIM1 to oligomerise and translocate to ER-PM junctions. At these junctions, STIM1's cytoplasmic domain binds and gates ORAI1, a tetrameric plasma membrane calcium channel, allowing sustained Ca²⁺ influx from the extracellular space. The incoming Ca²⁺ is taken up by SERCA pumps into the ER, refilling the store. ORAI1 and STIM1 mutations cause severe combined immunodeficiency, reflecting the essential role of SOCE in T-cell activation.

Non-vesicular lipid transfer at MCS is mediated by lipid transfer proteins (LTPs) that solubilise a single lipid molecule in a hydrophobic pocket and shuttle it between membranes. At ER-Golgi contacts, the ceramide transfer protein (CERT) extracts ceramide (synthesised in the ER) via its START domain, transports it across the cytosolic gap, and delivers it to the Golgi for sphingomyelin synthesis. OSBP (oxysterol-binding protein) exchanges cholesterol (from ER to Golgi) for PI4P (from Golgi to ER) at ER-Golgi contacts: the PI4P gradient (high in Golgi, low in ER, where SAC1 phosphatase hydrolyses it) drives the net cholesterol transfer. This counter-exchange model — the energy of the PI4P gradient powers cholesterol transport against its concentration gradient — is a general principle for several ORP-family LTPs.

Lipid droplets (LDs) are neutral lipid storage organelles that originate from the ER. Triacylglycerol synthesis by DGAT1 and DGAT2 at the ER membrane produces a lipid lens between the two leaflets of the bilayer; when this lens reaches a critical size, it buds outward as a lipid droplet, surrounded by a phospholipid monolayer derived from the outer ER leaflet. Seipin (BSCL2) forms a ring-shaped oligomer at ER-LD junctions that controls droplet geometry, ensuring uniform size. Mutations in seipin cause congenital generalized lipodystrophy. ATGL (adipose triglyceride lipase) accesses stored triacylglycerol at ER-LD contact sites, releasing fatty acids for mitochondrial β-oxidation 17.04.02 pending.

The phosphoinositide code provides each organelle with a distinctive lipid identity that recruits specific effector proteins. Each compartment has a characteristic phosphoinositide species: PI(4,5)P₂ at the plasma membrane, PI4P at the Golgi, PI(3)P on early endosomes, and PI(3,5)P₂ on late endosomes. The conversion between these species — by kinases (PI4K, PIP5K, VPS34, PIKfyve) and phosphatases (SAC1, INPP5B, FIG4) — marks organelle maturation. PI(3)P on early endosomes recruits EEA1 (a tethering factor with a FYVE domain that specifically binds PI(3)P), which captures incoming vesicles and promotes fusion. As the early endosome matures, PI(3)P is converted to PI(3,5)P₂, recruiting different effectors that redirect traffic toward lysosomal fusion.

Synthesis. Putting these together, the organelle landscape of the eukaryotic cell is a self-organising system maintained by three interlocking mechanisms: protein targeting codes that direct nascent polypeptides to the correct compartment, vesicular trafficking machinery that moves cargo between endomembrane compartments, and membrane contact sites that enable non-vesicular lipid and ion exchange. The central insight is that organelle identity is not a static property of the membrane but a dynamic steady state, actively maintained by GTPase switches, phosphoinositide kinases, and tethering factors whose continuous activity is required to prevent identity collapse. This is exactly the reason why mutations in single trafficking components — a Sar1 GTPase, an OPA1 dynamin, a PEX gene — produce devastating human diseases: the system has no redundancy for its load-bearing elements. The foundational reason the eukaryotic cell can perform oxidative phosphorylation, protein quality control, and calcium signalling simultaneously is that these incompatible chemistries are physically separated by membranes whose identity is maintained by the targeting and trafficking machinery described here. The bridge is between the static picture of organelles visible by electron microscopy and the dynamic molecular reality of constant flux, and the pattern recurs at every scale from the single protein to the whole cell.

Connections [Master]

• Biomolecules in cells 17.01.01. Introduced the four classes of biomolecules that compose every organelle: membrane phospholipids form the compartments, proteins carry out enzymatic and transport functions, and nucleic acids in the nucleus and mitochondria encode the genetic information for the proteins that build and operate each organelle.

• Cell membranes: structure 17.02.01. Established the phospholipid bilayer as the fundamental barrier. Each organelle is bounded by one (lysosomes, peroxisomes, ER) or two (nucleus, mitochondria) bilayer membranes with distinct lipid and protein compositions. Membrane asymmetry, fluidity, and curvature determine organelle function, vesicle budding, and SNARE-mediated fusion.

• Membrane transport: passive and active 17.02.02. The V-ATPase that acidifies lysosomes and the SERCA pump that refills ER calcium stores are active transporters whose mechanisms build on the ion transport principles described in 17.02.02. The proton and calcium gradients maintained by these pumps are energy stores that power lysosomal degradation and SOCE signalling, respectively.

• Cellular respiration: glycolysis and CAC 17.04.01. The citric acid cycle takes place in the mitochondrial matrix, whose double-membrane structure and protein import machinery are described here. Pyruvate enters the matrix through the mitochondrial pyruvate carrier, linking cytosolic glycolysis to mitochondrial oxidative metabolism.

• Oxidative phosphorylation 17.04.02 pending. The electron transport chain complexes are embedded in the inner mitochondrial membrane, whose cristae folds maximise surface area for ATP production. Calcium transfer at ER-mitochondria contact sites stimulates matrix dehydrogenases, directly coupling the contact site machinery to the energetic output of oxidative phosphorylation.

• Amino acids and protein chemistry 15.12.01 pending. Disulfide bond formation in the ER (mediated by Ero1 and PDI) and N-linked glycosylation (the calnexin cycle) are chemical modifications of polypeptides whose chemistry is described in 15.12.01 pending. The ER is the primary site of both modifications in the cell.

• Cytoskeleton and contractile proteins 17.03.02 pending. Microtubule-based transport by kinesin and dynein is the primary mechanism for positioning organelles and for vesicular trafficking between compartments. The Golgi apparatus localises at the microtubule-organising centre, endosomes and lysosomes move along microtubule tracks, and mitochondrial distribution depends on microtubule motor proteins. The cytoskeletal machinery is the physical substrate for the organelle localisation and membrane trafficking described here.

• Photosynthesis: light and dark reactions 17.04.03 pending. Chloroplast structure is inseparable from photosynthetic function: the thylakoid membrane system organised into stacked grana and unstacked stroma lamellae physically separates PSII from PSI, which is essential for the independent operation of the two photosystems. The endosymbiotic origin of chloroplasts from cyanobacteria, described here, explains why the photosynthetic electron transport chain resembles bacterial respiratory chains more closely than eukaryotic ones.

Historical & philosophical context [Master]

Robert Brown first described the nucleus in 1831, observing an opaque spot in orchid cells ^{[Brown 1831]}. Mitochondria were visualised by light microscopy in the 1890s; Richard Altmann called them "bioblasts" in 1890, and Carl Benda coined "mitochondrion" in 1898, but their function remained unknown for decades. The Golgi apparatus was reported by Camillo Golgi in 1898 using his silver-staining method, though its existence was disputed until electron microscopy confirmed it in the 1950s. Christian de Duve discovered lysosomes in 1955 using differential centrifugation of rat-liver homogenates, finding that acid phosphatase activity sedimented in a fraction distinct from mitochondria ^{[de Duve 1955]}.

George Palade's electron microscopy studies at the Rockefeller Institute in the 1950s mapped the secretory pathway ^{[Palade 1975]}. Using thin-section EM, Palade visualised ribosomes on the rough ER surface, transport vesicles between ER and Golgi, and secretory granules at the plasma membrane. His pulse-chase experiments with Jamieson demonstrated the directional flow of secretory proteins: ER to Golgi to secretory vesicles to extracellular space. Palade shared the 1974 Nobel Prize with Albert Claude and de Duve.

The signal hypothesis was proposed by Blobel and Sabatini in 1971 and confirmed experimentally by Blobel and Dobberstein in 1975 ^{[Blobel & Dobberstein 1975]}. Using cell-free translation systems supplemented with microsomal membranes, they demonstrated that the N-terminal signal sequence is both necessary and sufficient for ER targeting. Blobel received the 1999 Nobel Prize.

Lynn Sagan (later Margulis) proposed the endosymbiotic theory for the origin of mitochondria and chloroplasts in 1967 ^{[Sagan 1967]}. The theory gained decisive support from molecular phylogenetics in the 1970s-80s, when sequencing of mitochondrial rRNA genes placed mitochondria firmly within the α-proteobacteria.

Randy Schekman used yeast genetics to identify the SEC genes required for vesicle budding and fusion, while James Rothman used biochemical reconstitution to identify the coat proteins and SNAREs ^{[Schekman 2013]}. They shared the 2013 Nobel Prize with Thomas Sudhof, who identified synaptotagmin as the calcium sensor for synaptic vesicle release.

Bibliography [Master]

@article{BlobelDobberstein1975,
  author = {Blobel, G. and Dobberstein, B.},
  title = {Transfer of proteins across membranes. {I.} {P}resence of proteolytically processed and unprocessed immunoglobulin light chains},
  journal = {J. Cell Biol.},
  volume = {67},
  year = {1975},
  pages = {835--851}
}

@article{Palade1975,
  author = {Palade, George},
  title = {Intracellular aspects of the process of protein synthesis},
  journal = {Science},
  volume = {189},
  year = {1975},
  pages = {347--358},
  note = {Nobel Lecture}
}

@article{Sagan1967,
  author = {Sagan, Lynn},
  title = {On the origin of mitosing cells},
  journal = {J. Theor. Biol.},
  volume = {14},
  year = {1967},
  pages = {225--274}
}

@article{deDuve1955,
  author = {de Duve, C. and Pressman, B. C. and Gianetto, R. and Wattiaux, R. and Appelmans, F.},
  title = {Tissue fractionation studies. 6. {I}ntracellular distribution patterns of enzymes in rat-liver tissue},
  journal = {Biochem. J.},
  volume = {60},
  year = {1955},
  pages = {604--617}
}

@article{Schekman2013,
  author = {Schekman, R.},
  title = {Genetic and biochemical analysis of the secretory pathway in yeast},
  journal = {Nobel Lecture},
  year = {2013}
}

@article{Rothman2013,
  author = {Rothman, J. E.},
  title = {The principles of membrane trafficking},
  journal = {Nobel Lecture},
  year = {2013}
}

@article{Sudhof2013,
  author = {Sudhof, T. C.},
  title = {A molecular machine for neurotransmitter release},
  journal = {Nobel Lecture},
  year = {2013}
}

@book{Alberts2022,
  author = {Alberts, B. and others},
  title = {Molecular Biology of the Cell},
  edition = {7th},
  publisher = {Garland Science},
  year = {2022}
}

@book{Lodish2023,
  author = {Lodish, H. and others},
  title = {Molecular Cell Biology},
  edition = {9th},
  publisher = {W. H. Freeman},
  year = {2023}
}