Phase separation in colloidal and macromolecular systems has, for a century or more, been an area of active research driven not only by curiosity but also by problems in formulation science and by a quest to understand its role in biological compartmentalisation. With the development of formulations with improved performance in areas like cleaning, personal care, and pharmaceuticals, more complex compositions of the formulations were required. This created challenges in terms of miscibility and, thus to the creation of homogeneous solutions. In fact, combining several beneficial components in a formulation can often be the main obstacle in creating a product, and there are extensive efforts in industry to avoid phase separation phenomena in formulations. Whereas achieving homogeneity in a formulation is typically the main problem, there are also cases where instability is the desired function. Thus, phase-separated systems can be beneficial for the separation and purification of systems ranging from proteins to viruses and bacteria, and they can also include functions of deposition. The study of mixed colloid systems, including lipids, surfactants, polysaccharides, proteins, nucleic acids, etc., performed over >100 years, got a strong revival when, a couple of decades ago, it was realized that colloidal phase separation has a vital biological function, for example in the creation of membrane-less organelles in cells by formation of dynamic liquid droplets, termed biomolecular condensates [[1], [2], [3]]. It is interesting to note that membrane-less compartments such as the nucleolus were described already in the 1830s [4,5].

Liquid-liquid phase separation (LLPS) – the formation of biomolecular condensates or dynamic droplets – provides a mechanism for the formation of membrane-less compartments in cells, which play a significant biological role [1,6]. This has led to efforts to characterise and establish the role of such biomolecular condensates in normal cells as well as in various disease-related pathological states of different cells. Clearly, there is a considerable motivation to describe the specific properties of biological condensates and to understand the physical-chemical principles based on a colloid science description. In the present review, we focus on the basic fundamental mechanisms of DNA and DNA-protein phase separation motivated by its particular importance for chromatin systems and the role of LLPS for compartmentalisation and regulation of the genetic material in the cell nucleus. We review DNA phase separation from a basic colloid physical chemistry approach and emphasise the general principles that govern the phase separation of all DNA systems.

There are two main types of phase separation, segregation and association, depending on the interactions; they will be defined below [7,8]. For the case in focus here, DNA systems and, in particular, chromatin, association is the main driving force, but cases of segregative aspects can also be noted. In several cases of liquid-liquid phase separation (LLPS) in biological systems, this distinction has not been made, and therefore, the driving forces are unclear. The different phases (there can be more than two) are water-based, often with very high water contents. In biological systems, the phase separation of interest often leads to discrete particles dispersed in an aqueous continuum (denoted droplets or biomolecular condensates). There are several issues to be discussed in this context: Why - under certain conditions - is there a formation of discrete droplets rather than a separation into two separate macroscopic phases? What are the droplet sizes, and what is the size distribution? How are the droplets stabilized? Do the droplets have a liquid-like or solid-like character? These issues are still largely open but are intensely discussed in the current literature.

It appears that the common scientific basis of macromolecular phase separation in different fields, like formulation science and organization in biological systems, is frequently underutilized. The present review is an attempt to overcome this barrier to some extent. Whereas the focus will be on DNA and chromatin phase separation systems that under suitable conditions, display dynamic droplet formation through LLPS, we will anchor our discussion on broader aspects of macromolecular phase separation. In the present review, the term LLPS is considered from a general phase separation context but mainly refers to such phase separation leading to formation of dynamic droplets . The field is vast, and there will be no attempt to treat the subject exhaustively. Over the years, several insightful reviews have been presented, and regarding several aspects, the readers are referred to those for more in-depth information. Here can be mentioned, for example, the monograph of Albertsson [9] and recent reviews by Esquena [[10], [11], [12]], both focusing on what we will term segregative systems, as well as the article by Piculell and Lindman discussing also associative systems [7]. Other reviews addressing associative systems that have a particular focus on DNA systems undergoing LLPS are due to King [13] and Shakya [14]. The work by Korolev et al. [15] pays attention to the polyelectrolyte aspects of chromatin compaction and phase separation. For general reviews on LLPS formation of biomolecular condensates and their role in biology, refer to [[1], [2], [3]], as well as the work by Hildebrand and Dekker that reviews phase separation as a mechanism for chromatin compartmentalisation [16]. For historical developments and general treatises of interactions in colloidal systems, a number of accounts are available [[17], [18], [19], [20], [21]].

Phase separation can be in terms of different aggregation states, but here, we will mainly consider liquid-liquid phase separation. The balance between liquid-liquid and liquid-solid phase separation is, in many cases, delicate and non-trivial and needs to be clarified, and this also concerns the formation of liquid-crystalline phases. The most comprehensive early account of the field is the mentioned monograph by Albertsson, published in 1960 (with an enlarged edition in 1971). Beijerinck in 1896 observed that if, for example, aqueous solutions of gelatin and agar were mixed, a turbid solution, which separated into two phases, was observed, with two layers enriched in one of the components [22]. Another early report is also due to Beijerinck [23]. Later pioneering work included that of Albertsson, who published an extensive account of such segregated systems and tabulated phase diagrams for a large number of systems and described their use for separation and purification purposes, ranging from proteins to viruses and organelles. Additional important work was due to Tolstoguzov and co-workers [24,25]. Tolstoguzov also introduced the concept of water-in-water emulsions for the case that metastable systems of dispersion type formed. This field was significantly advanced by Esquena [[9], [10], [11]], Nicolai [[24], [25], [26], [27]] and others, mainly with respect to stabilisation.

As discussed by Albertsson, another type of macromolecular phase separation, which was described at about the same time as segregation, was found by Beijerinck. It was found that in mixtures of polymers, they were collected in one of the phases. This associative phase separation induced by attractive interactions has also been investigated for a long time. Detailed studies were performed by Bungenberg de Jong and co-workers [[26], [27], [28]]. They found, for example, such phase separation in mixtures of gelatin and gum arabic at a pH where gelatin is positively charged and gum Arabic negative. For the phase separation between two oppositely charged species, they coined the term complex coacervation. We will use the concept of associative phase separation.

With its high negative charge, DNA strongly associates with cationic co-solutes and is prone to phase separation in the presence of added multivalent cations and cationic colloids (surfactants/proteins). Interestingly, as has been described recently, DNA can also play an essential role in the stabilisation of segregating systems [29]. This review focuses on simple and complex systems containing DNA, of course, a matter of broad biological significance; this will be treated in some detail below.