LETN and NPM1 tango in human nucleoli

In a recent paper published in Cell Research, Wang et al. identify a nucleolar-localized long noncoding RNA (lncRNA), LETN (lncRNA essential for tumor cell proliferation via NPM1) and depict a mutually dependent NPM1–LETN interaction that is essential for maintaining nucleolar function and sustaining cancer cell proliferation.

RNAs are engaged with proteins during their life cycles, so are long noncoding RNAs (lncRNAs). Due to their large sizes and dynamic conformations, functional lncRNAs interact with different sets of proteins to ensure their proper processing and subcellular localizations. However, functions of proteins are generally thought to be independent of RNAs. In a recent study in Cell Research, Wang et al.¹ uncovered that lncRNA LETN is required for the protein NPM1 to execute its multiple roles.

LETN is a polyadenylated, primate-specific lncRNA with ~4500 nt in length (Fig. 1a). Its expression is low in most normal human tissues, but is upregulated in embryonic tissues and multiple types of tumor. LETN is largely localized in the nucleolus and is critical for cancer cell proliferation and tumor development. Knockdown or overexpression of LETN reduced or accelerated tumor cell proliferation, respectively. Mainly responsible for rRNA biogenesis, the nucleolus is the biggest nuclear condensate assembled via liquid–liquid phase separation (LLPS) and composed of three morphologically distinct sub-compartments: fibrillar center (FC), dense fibrillar component (DFC) and granular component (GC) regions.² Prior work has suggested that lncRNAs are involved in ribosome biogenesis, e.g., SLERT is localized to FC/DFC regions and promotes polymerase I (Pol I) transcription.³ Wang et al. showed that LETN was mainly enriched in GC regions and that LETN depletion led to GC distortion without affecting FC/DFCs.¹

Fig. 1: LETN and NPM1 are mutually dependent for nucleolar functions in primates.

a LETN is a primate-specific, nucleolar-localized lncRNA that binds the C-terminal region of NPM1 to facilitate NPM1 pemtamerization at the N-terminal domain with an unknown mechanism. b LETN promotes human cancer cell proliferation by driving main functions of NPM1, such as maintaining rRNA biogenesis and supporting GC integrity of the nucleolus. c Two newly evolved Ser residues of human NPM1 are essential for the NPM1–LETN function in human cells, whereas mouse NPM1 that lacks these two residues functions independently of LETN, which is absent in rodents.

GC distribution of LETN relies on direct interaction with the GC protein NPM1, which functions as pentamers in nucleolar structures,^4,5 ribosome biogenesis and chromatin remodeling.⁶ The authors found that LETN interacted with NPM1 via the C-terminal domain containing the nucleolar localization signal (NoLS) (Fig. 1a). Repression of NPM1 expression resulted in similar phenotypes as seen by loss of LETN: decreased cell proliferation, disturbed nucleoli, suppressed chromatin compaction and reduced Pol I transcription, indicating that NPM1 requires LETN for functional activation. It is worth noting that an even GC distribution of NPM1 in live cells vs a ring-like GC distribution in fixed cells were observed.¹ It will be of interest to clarify these distinct NPM1 subcellular distribution patterns.

The authors next perfomed multiple assays to examine a mutual dependent role of LETN and NPM1. First, they dissected the RNA–protein interaction in detail. On one hand, NPM1 bound to three regions in LETN, 180 nt in length for each. Deleting all three regions completely blocked the LETN–NPM1 interaction. Re-introducing the first 2000 nt of LETN containing these regions, but not a simple combination of the three fragments, partially rescued cell proliferation. On the other hand, LETN promoted NPM1 pentamerization via its N-terminal domain both in vitro and in cells (Fig. 1a). Consequently, phenotypes related to pentamerized NPM1, such as the interaction between NPM1 and histone proteins⁷ as well as nucleolar morphology⁶ were found to be impaired upon LETN depletion.

Next, the authors showed that LETN nucleolar localization was dependent on the full-length NPM1 (NPM1-F); meanwhile, LETN carried out its function mainly via NPM1 (Fig. 1b). LETN overexpression/knockdown enhanced/suppressed cell proliferation and chromatin condensation only in NPM1-F-expressing cells but not in cells expressing N- or C-terminally truncated NPM1 (NPM1-ΔN/ΔC).¹ Mechanistically, NPM1-ΔN failed to form oligomers whereas NPM1-ΔC failed to recruit LETN to nucleoli. Thus, the interaction between LETN and NPM1, rather than each factor alone, is necessary for their functions. However, how the interaction between the C-terminal fragment of NPM1 and LETN affects NPM1 pentamerization at the N-terminus is warranted for future study.

The authors also examined whether the regulation of ribosome biogenesis by NPM1–LETN interaction is physiologically relevant in cancer context, as hyperactive rRNA production is related to fast cell proliferation.⁸ As expected, liver cancer patients with better prognosis displayed lower levels of NPM1 and LETN, whereas patients with worse prognosis exhibited higher levels of both factors. Such a reverse correlation thus supports the view that the NPM1–LETN interaction in ribosome biogenesis is associated with cancer progression.

Finally, the mutual dependency between NPM1 and LETN was supported from an evolutionary point of view. While NPM1 is highly conserved across species, LETN is only present in primates (Fig. 1c). In NPM1-depleted human HUH7 cells, re-introducing either mouse or human NPM1 (mNPM1 or hNPM1) largely restored cell proliferation and rRNA biogenesis. Intriguingly, the rescue effects of hNPM1 are dependent on LETN, but those of mNPM1 are not. Further, LETN was required for pentamerization of hNPM1 but not mNPM1. It turned out that hNPM1 C-terminal region harbors four different residues from mNPM1, among which two Ser residues play a key role in the NPM1–LETN interaction (Fig. 1a–c). The hNPM1-214A/217A mutant lost its capabilities of pentamerization or binding with LETN. In an elegant reciprocal set of experiments, conversion of mNPM1 to its human homolog at these two residues (mNPM1-214S/217S) largely reactivated its binding with LETN, as well as its pentamerization. These results together suggest that the two Ser residues of hNPM1 are key to NPM1–LETN function in human cells.

Collectively, the human-specific protein–lncRNA interaction presents an intriguing regulation for acquiring a hyperactive nucleolus in fast cell proliferation relevant to cancers. The mutual requirement of hNPM1 and LETN raises the notion that this newly emerged lncRNA in evolution is indispensable for hNPM1’s major functions. Similarly, a recent study suggests that lncRNA orthologs undergo distinct species-specific processing for primate-specific functions.⁹ Thus, the prevalent expression and processing of lncRNAs in primates appears to be more meaningful in gaining primate-specific roles than previously thought. Additionally, it also raises some questions that need to be further addressed. For instance, how could ~200 copies of LETN per cell adequately modulate ~8,800,000 copies of NPM1?¹⁰ Do the other two of the four different residues between hNPM1 and mNPM1 play any role in NPMs’ function? How does LETN arise during evolution, and whether it is common to other primate-specific lncRNAs? Do changes in NPM1 residues co-evolve with LETN emergence? Given the species-specific expression of lncRNAs, it is possible that mutually independent protein–lncRNA pairs may be present in lower organisms but absent in primates. Nonetheless, the current study provides clear evidence that an evolutionarily non-conserved lncRNA is indispensable for NPM1 functions in human nucleoli. Further studies of additional protein–lncRNA pairs will certainly broaden our understanding of lncRNA functions and action modes of previously characterized proteins.