We propose a mechanistic model to explain the role of Lsr2 in cells. In this model, Lsr2 can bind different genomic regions in a sequence-dependent manner, pulling these regions into the co-condensate, thereby exerting control over gene expression while appearing as a punctae in the cell.

21/

Lsr2 co-condensation responds to the overall binding energy landscape rather than targeting specific motifs. While individual Lsr2 molecules preferentially bind 10 bp AT-rich DNA motifs, the observed condensation arises from the collective behavior of many Lsr2 molecules.

19/

Our model predictions for DNA polymers align excellently with experimental observations. Our model allows us to connect the base pair level binding of Lsr2 to DNA with the emerging co-condensation patterns that involve many protein molecules and long stretches of DNA.

17/

In line with our biochemical assays, Lsr2 alone phase separates at ~10 µM, while adding DNA induces condensation well below the saturation concentration for bulk Lsr2 phase separation. The results suggest that protein-protein and protein-DNA interactions suffice in explaining the exp. observations.

We observed that the majority of Lsr2 condensates formed at the center of the DNA molecules where the AT-rich sequence was present. In contrast, on the control construct lacking an AT-rich sequence, Lsr2 did not show any preference for condensate formation.

13/

i) 18 kb DNA with relatively homogeneous AT-content and ii) 21 kb DNA with a 3 kb long AT-rich segment in the middle iii) 21 kb DNA with a moderate AT content region. All the constructs were labelled with Cy5 at one end for directionality.

12/

Upon flowing in Lsr2, we observed the molecules initially coating the DNA, forming some kind of nucleoprotein filaments. This was followed by DNA bridging and cluster formation at specific locations along the DNA.

10/

We also employed a single-molecule assay to visualize the formation of Lsr2 condensates on DNA in real-time. Individual dsDNA molecules were labelled with biotin on both ends and immobilized on a PEG-passivated glass surface.

9/

From our bulk LLPS and single molecule studies, it became clear that dimerization, the ability to oligomerize and the weak molecular interactions mediated via the IDR region are essential for phase separation.

8/

Furthermore, we investigated how each of these features individually contributes to the formation of Lsr 2 condensates. To this end, we created a bunch of Lsr 2 mutant variants that either lacked the IDR region or the ability to form dimers and/or oligomers.

7/

The Lsr2-DNA co-condensates exhibited classic characteristics of LLPS, including fusion and reversibility, as demonstrated by FRAP experiments.

5/

Initially, we observed that purified Lsr2 readily undergoes phase separation at very high concentrations (>=10 μM). Interestingly, the addition of DNA reduced the Lsr2 concentration required for condensate formation by over 10-fold!

4/

In this study, we aimed to investigate the mechanism of action of Lsr 2, a critical nucleoid-associated protein (NAP) from Mycobacterium tuberculosis(Mtb). Lsr2 plays a key role in gene regulation and contributes to the pathogenicity of Mtb.

2/