A tight grip on differentiation: Nuclear constriction by microtubules regulates hematopoietic stem cells

Maintenance of the mature blood cells requires controlled cell fate decisions by hematopoietic stem and progenitor cells (HSPCs). While our knowledge of the gene expression changes that facilitate differentiation has made a leap forward, less is known about the cellular triggers that induce them. Biedzinski et al (2020) now uncover a new intracellular mechanism that drives myeloid differentiation: Microtubule bundles squeeze the nucleus of HSPCs and form large invaginations, thus causing changes in chromatin organization. These microtubule‐induced nuclear shape changes result in gene expression profiles that favor myeloid differentiation.

A new study shows that microtubule‐induced nuclear shape changes promote myeloid progenitor differentiation.

Since hematopoietic stem cells (HSCs) were identified several decades ago (Jacobson et al, 1951), their differentiation has been extensively studied. They maintain blood homeostasis by a strictly regulated mechanism of self‐renewal and differentiation of their intermediate progenitor stages. The process of differentiation is accompanied by a reduced self‐renewal potential and increased lineage identity. A major branching point occurs when HSPCs form myeloid progenitors of myeloid, erythroid, and megakaryocytic potential and lymphoid progenitors with lymphoid potential. However, there is evidence that also an early myeloid‐only commitment step exists, besides the classical erythro‐myeloid versus lymphoid split (Perie et al, 2015). Recent technical advances in single‐cell analysis could directly link gene expression profiles to the function of single cells. Subpopulations of HSPCs turned out to be very heterogeneous in terms of gene expression (Notta et al, 2016). Hence, the classical model of hematopoiesis is increasingly challenged (Zhang et al, 2018).

Lineage decisions are made according to an interplay of cytokines (Rieger et al, 2009) and physical cues transmitted by the extracellular matrix and stromal cells in the bone marrow niche (Pinho & Frenette, 2019). Transcription factors with mainly cell type‐restricted expression patterns have been identified as key components in the orchestration of myeloid cell maturation of HSPCs. Nevertheless, not much is known about the biophysical mechanisms that determine whether and how HSPCs differentiate.

Remarkably, the nuclei of HSPCs undergo dramatic morphological changes during their differentiation steps to myeloid or lymphoid lineages (Biedzinski et al, 2020). The laboratory of M. Théry and S. Brunet closely analyzed the morphology of differentiating HSPCs. Nuclei of myeloid progenitors are significantly bigger than lymphoid progenitors or stem cells. In addition, myeloid progenitors have large invaginations cutting the nuclear surface like canyons. Surprisingly, these nuclear deformations are not a consequence of a softening of the nuclear lamina, as the proportion of lamin A/C to lamin B is increased in myeloid progenitors, while nuclei with reduced lamin A/C are frequently more fragile and easily deformed by mechanical force (Pajerowski et al, 2007). Instead, the observed nuclear invaginations appear to originate from the cytoplasmic microtubule cytoskeleton‐mediated constriction of the nuclear membrane. Accordingly, the microtubule organizing center is also pulled into and buried deep in the nuclear envelope folds. For nuclear constriction, microtubule bundles have to generate mechanical force or at least have to resist the progressive force of nuclear expansion during myeloid differentiation.

The authors observed an enrichment of the dynactin subunit p150Glued and the dynein‐activating adaptor BicD2 on the nuclear surface of differentiating HSPCs. An involvement of microtubules and dynein motors was further corroborated by microtubule‐ and dynein‐affecting drugs that efficiently prevent nuclear constriction (Fig 1).

Figure 1. Nuclear constriction changes gene expression and induces myeloid differentiation.

Microtubule bundles constrain the nucleus and form deep invaginations in the nuclear envelope (1). Invaginations cause a release of lamin B from the nucleoskeleton at the construction site and induce peripheral loss of heterochromatin (2). Chromatin reorganization alters gene expression and promotes myeloid differentiation (3).

Interestingly, constriction of the nucleus is further translated to the nucleoskeleton. Lamin A/C at the nuclear lamina remained unaffected by lineage decisions and maintained a homogeneous distribution along the nuclear membranes of HSPCs, lymphoid progenitors, and myeloid progenitors. In contrast, Lamin B was homogeneously distributed along the nuclear lamina of HSPCs, but was absent in the constricted areas with concave nuclear curvature in myeloid progenitors.

Lamins can interact with chromatin either directly or through histones and other lamin‐associated proteins. By these interactions, heterochromatin is anchored to the nuclear membrane. Indeed, coinciding with inward nuclear curvature and reduction of lamin B, the authors observed a loss of facultative and constitutive heterochromatin in the nuclear periphery during differentiation toward the myeloid lineage. Moreover, the redistribution of lamin B and the peripheral loss of heterochromatin was prevented by the inhibition of nuclear constriction. From the microtubule cytoskeleton, a cue for myeloid differentiation is transferred to the nucleoskeleton and is further transmitted to the genetic information by lamin B. A previous study in the early Drosophila embryo showed that microtubule polymerization in bundles organized by dynein induces nuclear deformations, which in turn increase chromatin dynamics (Hampoelz et al, 2011). Therefore, a related mechanism might be employed during Drosophila development, even though a different set of proteins is involved.

Interestingly, Biedzinski et al now found that microtubules are causative in the regulation of chromatin rearrangements across the nuclear membrane during differentiation. Inhibition of nuclear constriction, for example, by the dynein inhibitor ciliobrevin D, significantly altered gene expression. Characteristic genes of myeloid cells were downregulated, whereas representative genes of other lineages, for example, megakaryocytes, were upregulated. These results therefore implicate that mechanical force applied to the nucleus by the cytoplasmic microtubule cytoskeleton can influence gene expression and differentiation.

Previous studies have established a connection between the cytoplasmic cytoskeleton, nuclear morphology, chromatin structure, and even gene expression. In mammalian oocytes, Formin 2‐nucleated actin filaments induce nuclear envelope fluctuations, which correlate with chromatin motion and result in changes in gene expression (Almonacid et al, 2019).

Even viruses can successfully hijack the cytoskeleton for the modification of gene expression. The human cytomegalovirus (HCMV) uses microtubules to polarize nuclear proteins and to trigger nuclear F‐actin formation. This series of events compartmentalizes the nucleus, thus leading to a separation of heterochromatin and viral DNA for efficient virus replication (Procter et al, 2020).

Taken together, the study by the laboratory of M. Théry and S. Brunet adds an important chapter to the fast‐emerging field of nuclear mechano‐transduction. A common ground with previous studies is the finding that changes in the mechanical properties of the perinuclear cytoskeleton and the nuclear lamina are critical for a cellular response. However, in the current study the cells are not adapting to external mechanical cues or a physical microenvironment, for example, to protect the genetic material. Remarkably, mechano‐transduction also seems to be intrinsically used to recapitulate a cellular program in differentiation, although further studies are needed to elucidate the underlying triggers for the observed cytoskeletal rearrangements.

The EMBO Journal (2020) 39: e107086.