The ability of immune cells to detect and respond to danger signals is fundamental in immune function. While the receptors involved differ across immune cells types (14) the distribution of these receptors, and other co-stimulatory molecules, is critically important to appropriate activation.

For many years it was thought that receptors, such as B cell receptor (BCR) on B cells, T cell receptor (TCR) on T cells and Toll like receptors (TLRs) on macrophages, were evenly distributed across their respective cell surfaces only to aggregate upon activation (15). However, despite electron microscopy results suggesting the non-random distribution of immunoglobulin molecules on resting B cells (16) as early as 1975, it was not until the advent of super-resolution imaging technologies that it was explicitly shown that many, if not all, receptors cluster within the plasma membrane in the steady state (17–19).

For example, in a seminal work using PALM imaging ( Box 3 ) the TCR and a key T cell signalling adaptor molecule, Linker for Activation of T cells (Lat), were shown to reside in clusters upon the plasma membrane, termed protein islands (21). In more recent technically and visually stunning expansions of the characterisation of TCR distribution it was shown that these TCR islands are found across the whole live T cell membrane in culture (22) and in the lymph node (23).

Using other variants of SMLM ( Box 3 ) other immune signalling molecules have also been shown to form clusters in the steady state including; CD4 and Lck on T cells (24–26), IgM, IgG and IgD on B cells (27–29), IgE on mast cells (30), TLR4 (31–33), signal regulatory protein α, Fc gamma receptor I and II on human macrophages (34), β2 integrins on human neutrophils (35) and NKG2D on NK cells (36).

The majority of these studies revealed not only the pre-activation clustering of these signalling molecules, but a consistent activation induced redistribution of these clusters. Interestingly, evidence from STORM, PALM and FLIM/FRET ( Box 4 ) imaging of T and B cell membranes reveals this redistribution to be a concatenation, but not coalescence, of these clusters (21, 38, 39) ( Figure 2 ). This concatenation of protein islands, as opposed to a complete merging, is thought to play an important regulatory function. As such, it is thought that signalling occurs only at the boundaries of clusters that contain distinct compositions of important signalling molecules. For example, in B cells IgM and CD45 are found together on an island separated from islands containing Lyn and CD19 (40, 41). This is important as Lyn is required for some forms of signalling via IgM (42, 43). Thus, the two islands must come together, and exchange components, during activation, however, if complete coalescence of the islands was allowed dysregulated activation could result.

It has been known for over a decade that the network of cortical actin just under the cell membrane plays a critical role in the separation of protein islands (44, 45). However, it was the application of super resolution imaging technologies that allowed elucidation of the underlying molecular mechanism of control. As such, STORM imaging has shown that treatment of B cells with an actin-depolymerizing compound (latrunculin A) increases the proximity of the aforementioned IgM and IgD containing islands (28) and the lateral mobility of BCR and CD19 containing islands (27, 46, 47).

Once a danger signal is detected, immune cells undergo dramatic cellular and molecular changes in order to play their part in the immune response. In addition to revealing previously unseen mechanisms of immune cell danger detection, super resolution imaging has also added to our understanding of the molecular changes during immune cell activation.

Some immune cells, such as cytotoxic T cells and natural killer (NK) cells, respond to activation with the release of lytic granules at a synapse between them and their target cell. These granules are designed to induce apopotosis in the target cell. Unsurprisingly, given their lethality, the formation, trafficking and release of these modified lysosomes is tightly controlled. The ability of sub-diffraction limit imaging to visually untangle the dense, intricate and highly dynamic network of cortical actin and lytic granules underlying the synapse and the plasma membrane in general has revolutionised our understanding of immune cell killing (48).

For example, in recent years a number of super-resolution imaging technologies, including 3D-SIM (49), STED (50, 51), SMLM and TIRF ( Box 5 ) (53, 54), and Lattice light sheet microscopy ( Box 6 ) (12, 56) have all been used to observe the rapid and intricate movement of actin and lytic granules towards, and within, the immune synapse of both T and NK cells. As such, it was revealed that upon activation the network of actin that normally forms a mesh too dense for lytic granules to traverse dilates or dissolves at the immune synapse allowing microtubule-guided granule release ( Figure 3 ) This process takes approximately one or thirty minutes in T and NK cells, respectively (50, 54, 57).

Furthermore, not only is the intricate actin network important for facilitating degranulation, it also appears to play an important role in regulating the number of granules released. This control is critically important as it minimises healthy by-stander cell killing (58) and potentially also influences the number of target cells an individual cytotoxic lymphocyte can kill (59). Recent imaging studies have shown that T and NK cells do not release their entire granule payload during initial degranulation. In fact, they may release as little as one tenth of their total granules (60, 61). Actin likely regulates degranulation via two mechanisms; one, it limits the transport of granules to the immune synapse (60) and two, TIRF imaging recently observed the reformation of the dense actin network underlying the immune synapse soon after degranulation, restricting further granule release (12).

While cytotoxic lymphocytes respond to activation with the direct killing of target cells, B lymphocytes direct killing via the mass production of specific antibodies. This requires transformation into ‘antibody factories’, including dramatic increases in cell size, proliferation and RNA synthesis (62). The process also involves the spreading of chromatin from its naïve location, predominantly at the nuclear periphery, to a more dispersed configuration. This chromatin spreading is thought to promote transcription factor binding and gene expression important in the transformation to antibody secreting cells (63, 64).

While super resolution investigations confirmed chromatin spreading (13, 65), the ability to visualise the chromatin fibre to <20 nm resolution revealed that not only do the fibres spread within the nucleus, they also decompact, meaning there is more accessible DNA between normally tightly compacted nucleosomes. These processes were shown to be regulated independently, and furthermore it was chromatin decompaction, not spreading, that was important in regulating transcription factor binding (13). By inserting a fluorescent Halo-Tag downstream of two transcription factors, CTCF and JunD, almost unimaginably detailed three-dimensional single molecule tracking revealed the binding and diffusion behaviours of these factors during their DNA interrogations. As such, in a naïve B cell JunD collides with DNA roughly 130 times before finding a suitable and accessible binding site. This search time is roughly halved upon B cell activation (13). This was elegantly shown to be independent of chromatin spreading and reliant upon nucleosome decompaction using drug treatment or energy depletion, respectively (13). The dwell or residence times of CTCF determined by single molecule tracking was confirmed using FRAP imaging ( Box 7 ).

These studies are examples of how super resolution imaging has already fundamentally changed our understanding of central immunological processes, in this case by revealing the molecular underpinnings of immune cell activation. While earlier technologies did elude to many of these mechanisms, the fact that actin fibres, nucleosomes and indeed protein islands (22) are frequently separated by less that 200 nm means they can only be meaningfully visualised, and thus more completely understood, in super resolution.

Recombination of the antigen receptor genes (Igh, Igk and Igl in B cells and Tcrg, Tcrd, Tcrb and Tcra in T cells) is key to generating a wide antigen receptor repertoire. The tightly controlled process involves removal of the intervening DNA between genes from three segment pools, known as variable (V), diversity (D) and joining (J). The result is in an exon that encodes the antigen-binding domain of an antigen receptor. Recombination relies on a series of remarkable genomic manoeuvres, including relocalisation of the antigen receptor gene from the periphery to the centre of the nucleus (67, 68), removal of genomic domain boundaries within the gene (69) and a contraction of the gene to bring linearly distant V regions into close physical proximity with the D-J region for recombination (68, 70–72). These processes have been extensively examined using molecular and genetic manipulations, but also imaging technologies. For example, DNA FISH has been used to quantify the nuclear position and contraction of the Igh locus in B cell progenitors (67).

While these studies have added to our understanding of antigen receptor recombination there are still significant gaps in our knowledge of the process. For example, while locus contraction brings the V region into proximity with the D-J region, how the V region that will ultimately form part of the functional exon is ‘selected’ from numerous candidates within the distal region is unclear. Diffusion fitting a fractional Langevin motion model (73) within the viscoelastic nuclear environment is currently the best explanation of how this process may work (74, 75).

Recent super resolution imaging breakthrough technologies provide an opportunity to reveal the mechanics of this long-standing immunological puzzle. These technologies, including ORCA and Hi-M ( Box 8 ) (80, 81), leverage the development of complex pools of synthetic fluorescently labelled oligonucleotides (OligoPaint) alongside sequential super resolution STORM imaging to reveal the nanoscale configuration of genomic regions. In visually stellar works building on super resolution examinations of chromosome scale genome organisation (83–85), both ORCA and Hi-M have been used to visualise the nanoscale (2-15 kB resolution) organisation of specific genomic regions (up to 700 kB in size) within individual cells of whole Drosophila embryo sections. These works revealed previously undetectable relationships between genome organisation, epigenetic states and transcription (80, 81).

Given the unprecedented resolution, throughput and applicability of these new super resolution technologies it is conceivable that they could be used to examine the nanoscale genome organisation of antigen receptor loci in thousands of adaptive immune cell progenitors of any species. Taking mouse Igh as an example 700 probes would be sufficient to cover the entirety of the expansive 2 million base pair locus at a 3 kb resolution. Given the 113 V_H region gene segments are mostly separated from each other by at least 5 kb (86), 3 kb resolution would be sufficient to reveal the location of all V_H gene segments, along with the rest of the locus, within thousands of individual cells. While fixation required by sequential imaging prohibits a live view of the recombination process, a compilation analysis from the thousands of single cells could reveal an unparalleled view of the local genomic environment in which recombination occurs (80, 84, 87). This could reveal patterns of order, indicative of stable, consistent position or interactions, or disorder, potentially revealing regions undergoing random diffusion. For example, it could be that locus contraction consistently brings particular regions into physical proximity, relative to all others. Alternatively, it could be that the entire locus diffuses with minimal physical constraints and no discernible patterns of interaction. Either way these breakthrough super resolution technologies could enable a greater understanding of the role of diffusion and physical proximity during recombination.

As for live imaging, the fixation required for ORCA and Hi-M would also obstruct downstream examination of the physiological impacts of the visualised genomic organisation. For example, even if as hypothesised the distal V_H gene segments are revealed to contract to, then diffuse near, the D-J region, the ultimately selected segment could not be confirmed within fixed cells. However, here it is worth remembering that the near universal applicability of these technologies mean they can not only expand our understanding of steady state conditions, but also be applied to the genetic and molecular manipulation systems used previously to understand recombination. Thus, perturbations to antigen receptor loci genomic organisation could be re-examined using these novel technologies validating and expanding previous conclusions.

Immune cells make up arguably the most diverse cellular system in complex organisms. This diversity requires numerous lineage decisions as an immune cell differentiates from a haematopoietic stem cell. Be they step-wise and absolute or fluid and continuous (88), these decisions are directed by transcription factors (89). Some of these factors as so influential that the expression of a single transcription factor gene can alter a cells lineage fate (90, 91). The cellular consequences of these lineage decisions have been well explored, in part using imaging (7). However, the molecular events underlying immune cell lineage decisions remain largely unexplored.

Recent application of cutting-edge super resolution imaging technologies in other systems have demonstrated the power of these technologies to reveal molecular insights into transcriptional regulation, and thus potentially lineage decisions. For example, in a recent technical masterpiece single molecule tracking combined with target loci locking microscopy was used to reveal the single molecule resolution, real time kinetics of transcription and its regulators in mouse embryonic stem cells (92). As such, phage genome sequences that can be recognised by fluorescently tagged phage coat proteins were engineered into the 3’ UTR of two pluripotency transcription factor genes ( Box 8 ). This allowed single molecule visualisation of nascent mRNA. In the same cells, RNA polymerase II or other transcriptional regulatory factors (Sox2, Cdk9, Brd4 or Mediator) were fluorescently labelled. This allowed a phenomenally detailed examination of the relationship between the numbers, dynamics and positioning of these factors relative to transcription, revealing hierarchical, highly dynamic (2-10 second turnover) but relatively small clusters (<20 molecules) of all factors at sites of transcription.

As mentioned above single molecule tracking has been previously performed in immune cells (13). However, these experiments were not in the context of lineage decisions or concurrent with transcriptional visualisation. Here we outline experiments applying the visualisation systems used in Drosophila transcriptional regulation to immune cell lineage decisions. While ultimately these experiments could be conducted in genetically engineered primary cells, there are numerous in vitro systems in which immune cells can be induced to make lineage decisions. For example, the monocytic cell line THP-1 can be induced to differentiate into M1 or M2 macrophages by treatment with propidium monoazide (93). Within this system expression of lineage defining transcription factors, such as STAT1, 3 or 6, could be visualised (94). When the expression of these transcription factors is first detected the locus could be target locked and the relationship between transcription and single molecules of select regulatory factors could be examined. This could reveal how single molecules can regulate expression of these lineage defining transcription factors, and thus influence the fate of the immune lineage.

One obvious weakness of this methodology is the inability to visualise regulatory events prior to transcription initiation. Many of these events are likely just as lineage defining as those after transcription begins. As previously outlined, there are a number of technologies that allow visualisation of specific loci in cells ( Box 8 ), however, none have yet been combined with live single molecule tracking of regulatory factors and transcription. Like so many recent molecular technologies one recent breakthrough in visualising loci in live cells utilises catalytically dead Cas9 (dCas9). As such, in a system known as Chimeric Array of gRNA Oligonucleotides (CARGO), numerous guide RNAs are introduced into the cell to guide fluorescently tagged dCas9 to a locus of interest (10) ( Box 8 ). While the presence of dCas9 was shown not to dramatically impact local genome organisation (10), it is likely that dCas9 will obstruct other regulatory factors at sites of interest. Thus, other methods of visualising loci of interest prior to transcription will be required if the regulatory events prior to transcription are to be studied at the nanoscale.

While there is still work to be done, recent developments in super resolution imaging have revealed the behaviours of lineage defining molecules, be it transcriptional regulators or genomic loci, in almost unimaginable detail. If, or perhaps when, they are ultimately applied to immune cells, these single molecule scale technologies will provide an unprecedented view of entire antigen receptor gene loci and potentially allow us to watch as a single transcriptional regulator changes the fate of an entire lineage.