The hierarchical organisation of the 3D genome
The human genome contains approximately 3.2 billion base pairs of DNA per haploid set, packaged into 46 chromosomes, which must be compacted approximately 10,000-fold to fit within the 5-10 micron nucleus. This compaction is not random. Chromosome conformation capture techniques — particularly Hi-C, which maps genome-wide chromatin contacts by proximity ligation — have revealed a hierarchical organisation that operates at three distinct spatial scales:
At the largest scale (megabase range), chromosomes are partitioned into A and B compartments. The A compartment corresponds to gene-rich, actively transcribed, euchromatic regions with high acetylation (H3K27ac, H3K9ac) and enrichment in RNA Pol II. The B compartment corresponds to gene-poor, heterochromatic, late-replicating regions with high H3K9me2/3 and HP1. A and B compartments interdigitate in a checkerboard pattern: A-compartment regions preferentially contact other A-compartment regions, and B-compartment regions preferentially contact other B-compartment regions, even across megabase distances and even on different chromosome arms. This compartmentalisation reflects the nuclear radial organisation — B compartment chromatin is enriched at the nuclear lamina (peripheral heterochromatin), while A compartment chromatin is enriched at nuclear pore complexes and in the nuclear interior.
At the intermediate scale (100 kilobases to 2 megabases), chromosomes are organised into topologically associating domains (TADs). TADs were first described by Dixon et al. (2012,Nature) and Nora et al. (2012, Nature) using Hi-C and micro-C. They are defined as genomic regions within which chromatin contacts are significantly enriched relative to contacts across the domain boundary. TADs are largely conserved across cell types and are partially conserved across species (mouse and human TADs share approximately 75% of boundaries). Their stability suggests they reflect a fundamental structural feature of genome organisation rather than cell-type-specific regulatory choices.
At the smallest scale (kilobases), individual chromatin loops mediate specific enhancer-promoter contacts. These loops are highly cell-type-specific and are regulated by the transcription factor and co-activator landscape — changing with differentiation, activation, and stress.
Cohesin and the loop extrusion model
The physical basis of chromatin loop formation was clarified by the loop extrusion model, developed by Sanborn et al. (2015, PNAS), Fudenberg et al. (2016, Cell Reports), and directly visualised by Ganji et al. (2018, Science) using single-molecule imaging on reconstituted DNA substrates. The model proposes that cohesin — a ring-shaped SMC (structural maintenance of chromosomes) complex — translocates along DNA, extruding a progressively larger loop until it encounters a barrier.
Cohesin's architecture is a tripartite ring: SMC1 and SMC3 (the ATPase subunits) form a V-shaped dimer through their hinge domains, with their head ATPase domains brought together by the kleisin subunit RAD21 (SCC1 in yeast) and two HEAT repeat subunits SA1/SA2 (STAG1/2). The ring is loaded onto DNA by NIPBL (SCC2/Mis4 in yeast) working with MAU2 (SCC4). The ATPase cycle of SMC1/SMC3, powered by ATP binding and hydrolysis, drives DNA translocation through a mechanism analogous to other SMC family motors (condensin, SMC5/6). WAPL (and its co-factor PDS5) acts as the cohesin unloader, opening the kleisin-SMC1 interface to release cohesin from chromatin. The balance between NIPBL-mediated loading and WAPL-mediated unloading determines cohesin residence time on chromatin and, consequently, the size of the loops it extrudes.
During loop extrusion, cohesin progressively enlarges the chromatin loop until one or both of the loop's anchors encounter CTCF (CCCTC-binding factor), which acts as a roadblock. Cohesin stalls at CTCF binding sites in a convergent orientation — meaning that two convergently oriented CTCF motifs (one facing right, one facing left) define the preferred loop anchor pairs. This orientation dependence, revealed computationally by Rao et al. (2014,Cell) and mechanistically by studies showing that inverting CTCF motifs disrupts specific loop anchors, is one of the strongest arguments for the loop extrusion model.
CTCF: the zinc-finger insulator and loop anchor
CTCF (CCCTC-binding factor) is an 11-zinc-finger transcription factor that recognises a 19 base-pair core motif (the CTCF motif, recognisable in ChIP-seq as a sharp, highly positional signal). CTCF binds approximately 55,000-65,000 sites in the human genome, with the majority at TAD boundaries and loop anchors. Its zinc fingers 4-7 contact the core motif; zinc fingers 1-3 and 8-11 contribute flanking sequence specificity and protein interactions.
CTCF serves two related but distinguishable functions at its binding sites:
As a loop anchor, CTCF provides the roadblock for cohesin loop extrusion, defining the precise genomic co-ordinates of looped enhancer-promoter contacts. The strongest loops in the human genome correspond to CTCF-CTCF contacts where both anchors have high CTCF ChIP- seq signal and convergent orientation. Deletion of a single CTCF binding site can abolish a specific enhancer-promoter loop and reduce the target gene's expression 5-10 fold.
As a TAD insulator, CTCF at boundaries prevents the formation of enhancer-promoter contacts that cross domain boundaries. This insulation is mechanistically connected to loop extrusion: CTCF at boundaries blocks cohesin from extruding loops that would bridge two TADs, thereby maintaining the physical separation of regulatory domains. The clearest demonstration of this insulation function comes from cancer genomics: in T-cell acute lymphoblastic leukaemia (T-ALL), a deletion at the NOTCH1 locus removes a CTCF boundary, allowing an upstream enhancer to contact the NOTCH1 promoter across the former boundary and activate the oncogene (Mansour et al., 2014, Science). The TAD boundary deletion, not a coding mutation, is the oncogenic event.
A/B compartment switching in inflammation
TADs and loops operate within the compartment framework, and compartments themselves are dynamic. LPS stimulation of macrophages causes large-scale A/B compartment reorganisation over hundreds of megabases, as reported by Phanstiel et al. (2017, Cell). Some B-compartment regions (silent, peripheral heterochromatin) switch to the A compartment (active, nuclear interior) upon inflammatory activation, corresponding precisely to the loci of LPS-induced genes. This compartment switch involves lamina detachment, histone acetylation, and chromatin opening at these loci — a coordinated change in nuclear architecture that accompanies and facilitates the transcriptional response.
Conversely, some A-compartment regions can shift toward the B compartment in response to anti-inflammatory signals or during resolution of inflammation. The compartment switch model means that anti-inflammatory treatments that affect multiple co-regulated gene clusters simultaneously — rather than gene-by-gene — can be explained by bulk compartment reorganisation rather than individual gene regulation. This is relevant to interpreting spirulina's broad anti-inflammatory effects on gene expression profiles.
CTCF, oxidative stress, and cysteine sensitivity
CTCF contains multiple cysteine residues (11 zinc fingers each require two cysteines per zinc coordination, yielding 22 zinc-ligating cysteines, plus additional non-zinc cysteines). The zinc fingers' structural integrity depends on the reduced state of the co-ordinating cysteines — oxidation of zinc-chelating cysteines releases zinc and unfolds the zinc finger. Several studies have shown that oxidative stress conditions reduce CTCF's DNA-binding affinity. Yusufzai and Felsenfeld (2004) described oxidative sensitivity of CTCF's insulation function, and more recent work on CTCF in the context of oxidative stress in ageing and disease has confirmed that CTCF binding site occupancy decreases under chronic oxidative conditions.
If CTCF DNA binding is compromised by oxidative modification of its zinc-co-ordinating cysteines, loop anchors are weakened, cohesin extrusion barriers are lost, and loops can extend across former TAD boundaries — reproducing aspects of the cancer-associated boundary deletions described above. This is a potentially significant mechanism by which chronic oxidative stress (as in obesity, metabolic syndrome, or chronic inflammation) could alter 3D genome architecture in ways that activate genes normally insulated by CTCF barriers.
Phycocyanin's antioxidant action — reducing intracellular ROS via Nrf2 activation and direct radical quenching — would be predicted to preserve CTCF cysteine redox state and maintain CTCF DNA-binding activity. This would help preserve TAD boundary integrity and prevent pathological enhancer-promoter rewiring caused by oxidative CTCF loss-of-function. The mechanistic logic is sound. Direct evidence in the form of CTCF ChIP-seq in spirulina- supplemented cells under oxidative stress has not been published.
AMPK and cohesin loading factors
AMPK has been shown to phosphorylate multiple components of the cohesin loading and regulation machinery. NIPBL (the cohesin loader) contains multiple AMPK consensus phosphorylation sequences (LXRXXS motif). Phosphorylation of NIPBL under energy stress conditions has been reported to reduce cohesin loading efficiency, shortening loop extrusion tract lengths and reducing the frequency of long-range contacts. In the context of chromatin architecture, this would predict that AMPK activation leads to shorter loops and more compact TAD structures, reducing the reach of enhancers and potentially limiting the transcriptional activation of genes that depend on long-range enhancer contacts for full expression.
This AMPK-cohesin axis connects spirulina's metabolic effects to chromatin architecture in an unexpected way. The prediction is that spirulina supplementation, by activating AMPK, would favour more compact chromatin loops and reduced long-range enhancer action. Genes that require long-range enhancer contacts for full expression (including many inflammatory genes whose enhancers are positioned megabases from their promoters) would be preferentially affected, while genes regulated by proximal enhancers (promoter-proximal elements within a few kilobases) would be less affected. This spatial selectivity provides another mechanistic dimension to spirulina's preferential anti-inflammatory effects.
The evidence base here is somewhat indirect: AMPK phosphorylation of NIPBL has been reported in proteomics datasets but the functional consequences for loop length and chromatin architecture have not been studied in the context of spirulina or phycocyanin specifically.
TAD boundary disruption in cancer: the architectural oncology framework
Beyond the NOTCH1 example in T-ALL, TAD boundary disruptions have been found in gliomas (PDGFRA activation by a gained enhancer after boundary deletion), medulloblastoma (GFI1/GFI1B oncogene activation), Burkitt's lymphoma (MYC-IgH translocation creates a new SE-MYC contact crossing a former TAD boundary), and multiple other cancer types. The identification of TAD boundary integrity as a tumour suppressor mechanism — parallel to classical tumour suppressors like RB1 and TP53 — opens a new dimension in understanding cancer epigenomics.
Preserving CTCF binding and TAD boundary integrity through antioxidant mechanisms is thus an interesting cancer-prevention hypothesis. Chronic oxidative stress progressively erodes CTCF binding at boundaries; dietary antioxidants (including spirulina's phycocyanin) could theoretically slow this erosion. This is a speculative prevention hypothesis with no direct clinical evidence, but it is mechanistically coherent with the CTCF-oxidation chemistry and the cancer-associated boundary disruption literature.
The compartment-switching explanation for broad transcriptional effects
A recurring observation in spirulina research is that its anti-inflammatory effects are broad — affecting not individual cytokines but clusters of co-regulated inflammatory genes simultaneously. Conventional interpretations invoke NF-kappaB suppression as the common thread. The compartment-switching framework offers a complementary explanation: if spirulina's antioxidant and anti-inflammatory effects prevent the A-compartment shift of inflammatory gene clusters during chronic low-grade inflammation (as occurs in metabolic syndrome, obesity, and ageing), the result would be globally attenuated expression of multiple co-localised inflammatory genes without requiring individualised suppression of each gene's promoter or enhancer.
This hypothesis predicts that spirulina's anti-inflammatory effects should preferentially affect gene clusters that co-localise in the same A/B compartment domains, rather than randomly distributed individual genes. Testing this prediction would require Hi-C or compartment analysis in peripheral blood mononuclear cells from spirulina-supplemented subjects — an experiment that is technically feasible but has not been performed.
Summary
The 3D genome is organised hierarchically: A/B compartments at megabase scale, TADs at 100 kilobase-to-megabase scale, and chromatin loops at kilobase scale. Cohesin drives loop extrusion (loaded by NIPBL, unloaded by WAPL) until convergently oriented CTCF sites define loop anchors and TAD boundaries. CTCF insulation prevents aberrant enhancer-promoter contacts across boundaries; loss of CTCF binding by deletion or oxidative damage can activate oncogenes (NOTCH1 in T-ALL) by exposing promoters to distant enhancers. A/B compartment switching accompanies inflammatory activation of macrophages, co-ordinating the expression of multiple gene clusters simultaneously. Spirulina's antioxidant activity preserves CTCF cysteine redox state and DNA-binding capacity, maintaining TAD boundary integrity. AMPK activation by spirulina reduces cohesin loop extension by phosphorylating NIPBL, potentially limiting long-range enhancer action at inflammatory gene loci. These mechanisms are biochemically grounded and mechanistically coherent; direct demonstration using spirulina in Hi-C or chromatin architecture assays remains to be performed.