Pilot multi-level review

Nuclear Speckles

Primary experiments distinguish speckle assembly, gene association and RNA-processing effects in specific cultured-cell systems.

Reading level

Graduate / Professional Undergraduate Lay Public

What a speckle is, and how assembly was tested

Nuclear speckles are bodies without an enclosing membrane, enriched in RNA-processing proteins. Ilik and colleagues combined SRRM2 truncation with SON knockdown in human HAP1 cells, and combined SON/SRRM2 knockdown in HEK293 cells. Speckle-associated proteins became broadly dispersed, supporting a structural role for these two core proteins in those models. [1]

The same study showed that the commonly used SC35 monoclonal antibody primarily marks SRRM2 in speckles, rather than its historically assigned target SRSF2. Marker identity matters: an antibody signal identifies its target distribution, not every function of the body it labels. [1]

Gene association is a specific measurement

Kim and colleagues studied heat-shock genes in human HAP1 and hamster CHOK1 cells, including engineered HSPA1B reporters. RNA FISH and live imaging associated speckle proximity with more RNA. The timing supported an effect following association, but increased RNA can reflect altered degradation as well as transcription. These Hsp70 genes are intronless, so this result alone does not demonstrate improved splicing. [2]

A 2026 perturbation tests RNA processing

Małszycki and colleagues reported that acute removal of SON and SRRM2 in engineered human cells impaired splicing and reduced expression of GC-rich genes with short introns in GC-rich genomic regions. Their abstract reports that genes outside those regions were spared. This supports a restricted processing role in the tested system; it is not a claim that every gene requires speckles. [3]

What these experiments leave to resolve

Core-protein depletion changes both protein availability and speckle organization. Assigning each RNA phenotype specifically to spatial organization therefore needs controls that separate those effects. Neither a stained focus nor one locus-specific RNA response establishes a universal liquid material state or disease mechanism. [1] [2] [3]

A 2024 subcompartment result

Evidence update · 12 October 2026

Zhang and colleagues report distinct SRRM2/SON subcompartments and alternative-splicing effects on different target subsets. The authors connect SRRM2 organization to self-interactions and protein–RNA coacervation. These observations extend the structural model; they do not establish that every RNA-processing effect depends selectively on a liquid phase. SRRM2 phase separation drives assembly of nuclear speckle subcompartments (2024).

Primary sources for this update

Source access: primary abstract; full Methods controls not appraised. The study-specific findings above do not certify associated Word drafts, raw data or every proposed nuclear-body mechanism.

Primary studies behind this guide

  1. Ilik et al. (2020), eLife 9:e60579. SON and SRRM2 are essential for nuclear speckle formation. DOI: 10.7554/eLife.60579. Primary research.
  2. Kim et al. (published online 2019; 2020 issue), Journal of Cell Biology 219:e201904046. Gene expression amplification by nuclear speckle association. DOI: 10.1083/jcb.201904046. Primary research.
  3. Małszycki et al. (2026), Cell 189:2024–2039.e25. Nuclear speckles enable processing of RNA from GC-rich isochores. DOI: 10.1016/j.cell.2026.01.011. Primary research.

Scientific text and the identity and fit of the cited studies were checked on 3 October 2026. This guide covers the experiments described here; it is not an exhaustive literature review. Source-access limitations and outstanding work remain.