Optogenetic approaches for understanding homeostatic and degenerative processes in Drosophila.

Many organs and tissues have an intrinsic ability to regenerate from a dedicated, tissue-specific stem cell pool. As organisms age, the process of self-regulation or homeostasis begins to slow down with fewer stem cells available for tissue repair. Tissues become more fragile and organs less efficient. This slowdown of homeostatic processes leads to the development of cellular and neurodegenerative diseases. In this review, we highlight the recent use and future potential of optogenetic approaches to study homeostasis. Optogenetics uses photosensitive molecules and genetic engineering to modulate cellular activity in vivo, allowing precise experiments with spatiotemporal control. We look at applications of this technology for understanding the mechanisms governing homeostasis and degeneration as applied to widely used model organisms, such as Drosophila melanogaster, where other common tools are less effective or unavailable.

Introduction

Adult organisms maintain their tissues and organs in a dynamic equilibrium, where cells die and are constantly replaced in a process of tissue homeostasis [1]. Homeostasis maintains the functionality of tissues and organ systems through stem cell mediated repair. This repair recruits stem cell pools which are maintained in specialized niches and regulated by extracellular signals or injury [2]. Stem cell niches have a complex, signal-rich environment which preserves tissue-specific stem cells. In aging organisms, these stem cell pools are depleted, initiating the degenerative process, where tissue homeostasis becomes less active [3, 4]. Mutations affecting maintenance signals can lead to disease, but recent studies into drug combinations show therapeutic potential, as they stimulate symmetric stem cell divisions, replenish the stem cell pool, and boost tissue regeneration [5, 6]. Given the importance of homeostasis in disease progression and aging, it is important to develop methods to study the complex interplay of these processes. Optogenetics is one such recent method being applied extensively to this field of research.

Optogenetics has been a particular success in Drosophila due to the many tools for gene manipulation. The most common approach has been to use the yeast-derived GAL4/UAS expression system to drive expression of optogenetic transgenes in a variety of tissues [7], especially when combined with specific integration sites to assure equal expression from different transgenes [8]. These can be combined with other binary expression systems such as lexA and QUAS to express different transgenes in different tissues [9, 10]. Temporal control of transgene expression can also be achieved using the temperature sensitive allele of the transcriptional repressor Gal80ts [11], or drug inducible promoters such as RU486 inducible GeneSwitch Gal4 system [12]. The ease of fly genetics allows expression in traditional genetic mutant backgrounds as well as in combination with RNAi knockdowns available as a library of fly lines; resources which have recently been expanded to include guide RNAs for CRISPR/Cas9 and dCas9-VPR for gene knockout and overexpression [13]. To manipulate genes directly, without overexpression, many genes in the Drosophila genome have been modified with a landing site for recombination-mediated cassette exchange (RMCE) [14, 15]. This landing site can be used to introduce any tag as an additional exon allowing expression of a tagged version of the protein at endogenous levels, and has been used to introduce an optogenetic cassette directly into the genome [16].

Optogenetics encompass a toolkit of techniques to control neuronal firing, gene expression, and protein function through light-responsive proteins or protein domains [17-29]. The term “optogenetics” was used to describe approaches that combine optical and genetic manipulation of specific cell types or specific cellular processes [30]. Optogenetic tools have been applied to successfully address a number of questions in neuroscience and cell biology that require spatiotemporal and neurochemical precision in vitro or in vivo. This toolkit has expanded rapidly, generating many potential applications for research questions across disciplines which were previously difficult to address [31]. In this review, we focus on recent advances in optogenetic tools for the study of mechanisms behind homeostatic and degenerative processes. We discuss how innovative applications of optogenetics in stem cell biology, neurodegenerative diseases, and homeostasis are being used to fill existing knowledge gaps and suggest future directions for the utilization of this suite of tools.

Optogenetic approaches

A series of discoveries enabled the development of optogenetics, from basic characterisation of light-activated protein pumps to applications in neuronal firing. In 1971, well before optogenetics was established as a concept, the bacteriorhodopsin was among the first protein pumps that were characterised as being activated by visible light photons [32]. One of the earliest applications of a light-induced technique was to study gene expression through the ability of phytochrome B to bind reversibly to the transcription factor, PIF3 [33-35]. Researchers also developed other light-inducible methods which use optically gated ion channels, are non-invasive, and are able to control Drosophila behaviour [36] and neuronal activity with Rhodopsin [37]. These studies led to the discovery of several optically sensitive genetic proteins that can manipulate cellular physiology. In 2005, Deisseroth, Boyden, and colleagues demonstrated the ability to precisely control neural activity at a millisecond timescale by successfully introducing and activating light-sensitive, microbial opsins in mammalian neurons in vitro [38]. Deisseroth et al. coined the term “Optogenetics”[30], the short history of which was reviewed by Deisseroth [24] and by Boyden (from a first-person perspective) [39]. Optogenetics in its early stages was applied to behavioural neuroscience, but subsequently expanded to cell biology [40].

Optogenetic approaches can be separated into two broad categories. The first focuses on the use of photoactivatable protein domains to regulate protein localization or oligomerization [41]. Examples of these systems include cryptochrome 2 (CRY2) [42, 43], phytochrome B (phyB) [34, 44, 45], light, oxygen, and voltage (LOV) [46], and Dronpa [47, 48]. Another example is the functional expression of photoactivated adenylyl cyclases (PACs) for light-induced manipulation of the cAMP messenger molecule in animal cells [49]. For a review of these single and dual-protein approaches for multiscale control of protein function in signalling biology, please see [50]. The second category involves the use of antibody-like recombinant binders for a broader range of protein targets using intrabodies or nanobodies to bind to endogenous proteins, targets which were previously inaccessible by the approaches in the first category [51-53]. The light-programmable control of these recombinant binders greatly expands the list of potential protein targets that can be targeted. These protein domains can be fused to genes of interest, providing a simple modular system, as they require low levels of activation light, are reversible, and are often independent of exogenous cofactors [54]. Given the growing number of new optogenetic approaches, researchers can keep track of different systems using the OptoBase database [55]. The suite of optogenetic tools available has progressed significantly in the past 15 years and can be used to tackle complex questions in cell biology (Fig. 1).

Fig. 1

Progression of optogenetic systems in cell biology. a Characterisation of the bacteriorhodopsin protein pump. Light photons activate the pump to move protons across the membrane of the cell. The retinal molecule (highlighted in red) changes its conformation after absorbing light photons. b First demonstrated use of microbial opsins in mammalian neurons. This use of channelrhodopsins was the first to probe neural coding with millisecond precision [37]. c Manipulation of the cellular cyclic AMP (cAMP) messenger molecule using photoactivated adenylyl cyclases (PACs). PACs act as modular photoreceptors with blue light-sensing domains and are bound to catalytic domains which produce cAMP activity, allowing cAMP activity to be modulated in Drosophila neuronal populations [48]. d PhytochromeB-PIF system for control of protein–protein interactions for cell signalling. The phyB-PIF system can be used to translocate target proteins to the membrane and back with micrometre spatial resolution [44]. The excitation speed (timescale of when the system activates after illumination with stimulatory light) is within seconds. The reversion speed (timescale of when the system resets in the dark or after illumination with inhibitory light) is within seconds when illuminated at 750 nm or hours in the dark [40]. e CRY2/CIB system for control of protein dimerization. The CRY/CIB system improves upon the phyB-PIF system by not requiring exogenous cofactors (like the bilin cofactor) while still maintaining fast temporal and subcellular spatial resolution [156]. f Dronpa fluorescent protein which is adapted for use as a light switch for controlling protein interactions. Dronpa domains are fused to the ends of the protein of interest. Cyan light illumination dissociates the tetrameric form of Dronpa, releasing the domains and uncaging the protein of interest [47]. This switch has been applied to studies of adhesion processes and control of protein kinases [157, 158]. g Optogenetic control of endogenous proteins using highly specific intrabodies (iB) [52]. The schematic represents light-induced recruitment to the membrane and movement from the cytoplasm to the nucleus. This system allows multiplexed protein regulation and simultaneous monitoring with visible-light biosensors. Previous systems could only target over-expressed exogenous proteins which could result in unintended aggregation or competition with endogenous proteins

Compared to classic genetic tools, optogenetic tools have the following advantages: (1) unmatched precision for spatial, temporal, and neurochemical regulation of proteins in in vitro and in vivo systems [54, 56, 57]; (2) great flexibility and reversibility, where the same approach can be used to turn pathways on or off in a variety of tissues or brain regions [56, 58]; and (3) the possibility of multiplexing [41, 59].

Applications of optogenetics

Neural signalling

Before optogenetic tools were developed in the 2000s, Francis Crick suggested in 1979 that a light-based approach would be useful to study neural signalling, so it was only natural that it was the first area of application for early optogenetics [31]. One of the first optogenetic systems used a combination of proteins including photo-reversible Drosophila rhodopsin NinaE, arrestin-2, and the α subunit of photoreceptor G-protein to sensitize cultured hippocampal neurons to light stimulation [37]. Expression of channelrhodopsins in 2 distinct classes of presynaptic motoneurons, which innervate one postsynaptic cell, was used to study presynaptic homeostatic plasticity at the neuromuscular junction, demonstrating the usefulness of light-activated channels for targeted experiments [60]. Newer and improved tools, such as the red-light activated channelrhodopsin, CsChrimson, have been applied to study dopamine release in the ventral nerve cord and medial protocerebrum of the larval Drosophila central nervous system [61, 62], nociceptive-like escape behaviour (in response to noxious stimuli and painful experiences), and the curling and rolling escape response in the larvae [63]. Further improvements such as the addition of the Split-Gal4 system were used to map the behavioural phenotypes attributed to at least one third of all descending neurons in the fly [64]. A wide range of behaviours are being dissected using optogenetics, such as sensorimotor pathways underlying behaviours, back-walking/back-crawling [65], turning/head casting [66], escaping [67], and oviposition [68]. In vertebrates, optogenetics studies address motor control [69-72], reward system [73], learning, and memory [74, 75].

Neurodegeneration

Optogenetics can be used to study protein misfolding common in neurodegenerative diseases. For Alzheimer’s disease (AD) and Parkinson’s disease (PD), existing Drosophila models can be divided into three broad categories: (1) Drosophila orthologs of human disease genes; (2) transgenic constructs carrying alleles of human disease-causing genes; and (3) models used to study the effects of environmental stressors on Aβ toxicity [76]. Studies have also combined approaches, developing models that use both Drosophila orthologs of human disease genes and overexpression of human transgenes. Expression using the GAL4-UAS system and accumulation of Aβ-42 peptide alone is sufficient to cause neurodegeneration and behavioural defects in Drosophila models [77]. A novel optogenetic approach used the Cry2, blue-light sensitive oligomerizing protein to investigate the consequences of formation of Aβ oligomers in Drosophila, C. elegans, and D. rerio [78]. Both expression and induced oligomerization of Aβ reduced lifespan and healthspan, but the effects could be separated into metabolic problems induced by expression alone and physical damage caused by light-induced Aβ oligomerization [79, 80].

Similarly, PD is characterized by the aggregation of α-synuclein, or Lewy bodies, within cells of the central nervous system. α-synuclein is hypothesised to contribute to toxicity and subsequent death of dopaminergic neurons in the substantia nigra pars compacta which is responsible for motor control [81]. To study the fixed stages of PD, an optogenetic PD model was constructed by introduction of the halorhodopsin (NpHR) gene, a light-gated chloride channel which inhibits neurons, into the substantia nigra compacta [82]. Unlike the classical animal model of PD, this novel optogenetic approach avoids causing damage to the nigrostriatal system and can be used as a PD model for various stages of the disease by adjustment of illumination parameters [82].

Other studies looked at TDP-43, a protein excluded from the nucleus into insoluble cytoplasmic inclusions, observed in post-mortem patient tissue in neurodegenerative disorders including Amyotrophic Lateral Sclerosis (ALS) and Frontotemporal dementia (FTD) [83, 84]. An optogenetic model of TDP-43 proteinopathy was developed by expressing Cry2 fused to human TDP-43 (optoTDP43). This in vivo model induced TDP-43 cytoplasmic aggregation and recapitulated important features of patient pathology such as progressive motor dysfunction, addressing the link between TDP-43 cytoplasmic aggregation and toxicity.

Stem cells, development, and tissue regeneration

The process of gastric epithelial homeostasis and regeneration has become an important model for understanding the regulation of stem cells [85, 86]. As these are adult tissues, genetic approaches must circumvent developmental stages using temperature-sensitive or tissue-specific overexpression, knockdown, or CRISPR/Cas9 [87-89]. One approach used signalling pathway components fused to CRY2 to regulate signalling using light in the adult fly midgut stem cell model [16]. A limitation of traditional overexpression or knockout techniques is that they affect stem cell regulation across the entire lifespan of the fly. This approach avoided the developmental effects of modulated stem cell signalling by raising flies in the dark and using light only at adult stages to activate the targeted pathway when required. Targeting intestinal stem cell signalling pathways with optogenetics allows perturbation at specific ages of Drosophila (e.g., young, adult, and old) to establish an age-based link to the effects of signals on gastric homeostasis.

Developmental processes such as tissue morphogenesis can be studied using optogenetic approaches. For example, morphogenesis during embryonic Drosophila development has been studied using optogenetics in a variety of ways [90]. A non-exhaustive list of optogenetic approaches to study Drosophila development includes: the structure of centrosomes [91], Erk signalling [27, 92–94], Toll signalling [95], adhesion [56, 96–98], cell shapes [99], GTPase activity [45, 100–110], anterior-to-posterior patterning through Bicoid [111], Wnt [112], and Notch signalling [113]. One approach that looked at phosphatidylinosital-4,5 biphosphate [PI(4,5)P2] during Drosophila ventral furrow formation showed that it was required for apical constriction, an essential tissue invagination process [114, 115]. The study observed that optogenetics can be utilized to control cell contractility in Drosophila adding tissue mechanics to light-regulatable processes [116].

Optogenetic control of cell differentiation is also applicable to regenerative medicine and tissue engineering. These systems typically provide reversible changes through protein–protein interactions or temporary changes which only last for the lifetime of the modified proteins. Tissue engineering requires long-term changes in gene expression through a spatially dependent “on” switch. Polstein et al. engineered such an optogenetic switch that would permanently express a transgene upon illumination by blue light using the CRY2 system and incorporating a myogenic differentiation factor in vivo in a mouse model [117]. Their result demonstrated that sustained transgene activation of in vivo cell differentiation and tissue morphogenesis was possible after an initial dose of illumination.

Nutrient and metabolic homeostasis

To optimize their fitness levels, animals vary their food choices with their body needs. For lab-raised Drosophila, yeast is the primary source of protein and nutrients. When investigating the chemosensory channels mediating yeast feeding, researchers found that gustatory receptor neurons (GRNs) of the taste peg, a sensory structure, and sensillar GRNs sustain and initiate yeast feeding, respectively [118]. The response of the yeast-sensitive GRNs is intensified following deprivation from amino acids, in turn modulating nutrient homeostasis. This result was determined using the FlyPAD assay, which allows automated, high-throughput analysis of feeding behaviour with temporal resolution [119]. Incorporation of optogenetics into the FlyPAD assay, known as the optoPAD, led to a high-throughput system that allowed direct manipulation of neurons to induce appetitive or aversive effects on feeding by activating or inhibiting GRNs. [120]. The adjustable light pulse of the optoPAD system, which can be dynamically linked to observed behaviour, makes it ideal for silencing neurons or activating reinforcement circuits. OptoPAD allowed “virtual gustatory realities” and induced appetitive behaviours by activating or inhibiting sweet and bitter neurons, respectively [120]. Another system, the Sip-Triggered Optogenetic Behavior Enclosure (‘STROBE’), addressed the challenge of manipulating feeding circuits in freely moving animals when the timing of sensory inputs is affected by the animals’ behavior. STROBE was able to temporally couple neuron activation with feeding events to find that coincident activation of sweet neurons upon food contact drives appetitive behavior and activation of bitter neurons drives aversive behavior [121].

Emerging optogenetic tools focus on detection of metabolic pathway components and modulation of whole-body energy homeostasis. To control phosphoinositide (PI) metabolism, Idevall-Hagren et al. exploited the CRY2 system to recruit key catalytic modules (PI3-kinases) to the plasma membrane, resulting in lipid gradients and cell polarization [122]. To activate fat thermogenesis, Tajima et al. designed an implantable wireless optogenetic device to selectively trigger Ca2+ [123]. This device enables targeted activation, as opposed to systemic activation, which can increase overall blood pressure. The device takes advantage of the spatiotemporal control allowed by optogenetic tools to activate fat thermogenesis without external stimuli, resulting in protection against obesity. These studies demonstrate how optogenetic systems can target various metabolic modules to have broad applications for studying other homeostatic processes.

Cardiac homeostasis

Optogenetic pacing of the heart allows non-invasive as well as highly precise spatiotemporal electrical stimulation. Transgenic flies expressing a light-gated cation channel, channelrhodopsin-2 (ChR2), paced contractions optogenetically at different developmental stages, including the larval, pupal, and adult stages. Cardiac function and its response to pacing stimulation were tracked using a high-speed and ultra-high-resolution optical coherence imaging system, serving as a powerful tool for basic heart research [124]. A hypersensitive variant of ChR2 (ChR2.XXL) was able to elevate the Drosophila larval heart rate over a range of temperatures and calcium levels, showing the potential of using light as a pacemaker in mammalian heart transplant conditions [125]. Further optogenetic studies using this variant showed that activation of cholinergic, dopaminergic, and serotonergic neurons also causes the release of cardioactive substances that increase heart rate over a range of temperatures [124-128].

Future directions in application of optogenetics to study homeostasis

Targeting regulatory signalling pathways of intestinal stem cells

Moving forward, optogenetic fusion constructs can be used with major components in signalling pathways to apply spatiotemporal control of homeostasis. For example, the Keap1–Nrf2 regulator pathway has potential for perturbation with optogenetics. Nrf2 is a master regulator of the cellular redox state. It controls the proliferative activity of ISCs and promotes intestinal homeostasis. Loss of Nrf2 in ISCs causes reactive oxygen species (ROS) accumulation and accelerates age-related degeneration of the intestinal epithelium [129]. In ISCs, loss of Keap1, a negative regulator for Nrf2, decreases intracellular ROS levels and regulates redox balance to control ISC proliferation as Keap1 homodimers promote proteasomal degradation of Nrf2 in the cytoplasm [130]. Post-translational modification of Keap1 by ROS (or electrophiles) leads to the release of Nrf2, and its translocation into the nucleus [131]. Nrf2 enters the nucleus, where it dimerizes with Maf family transcription factors, inducing expression of target genes [132].

The Keap1–Nrf2 pathway could be perturbed with optogenetics at each of these three points: (1) during the homodimerization of Keap1 (as illustrated in Fig. 2a); (2) the nuclear translocation of Nrf2 (Fig. 2b); or (3) the dimerization of Nrf2 with a transcriptional activator. In this pathway, an optogenetic approach which can take advantage of protein dimerization would be appropriate (to prevent Keap1 dimerization, induce Nrf2 clustering to modify its size and ability to move into the nucleus, or prevent Nrf2 binding to a transcriptional activator). For example, CRY2-based tools would be useful here, because the light-sensitive CRY2 protein, fused to the protein of interest, can bind to a partner protein (such as CIB) and modulate dimerization. Alternatively, another approach to knockdown Nrf2 is to use light-controlled degradation tools which regulate protein stability by fusing the protein of interest to a degradation module and light-responsive domain like LOV [133-135]. Given the interaction of ROS with Keap1, future studies could combine this application with existing optogenetic tools of ROS production [136].

Fig. 2

Potential optogenetic manipulation of the Keap1–Nrf2 pathway. The Keap1–Nrf2 pathway can progress to degrade Nrf2 in the cytoplasm after ubiquitination. Alternatively, if Nrf2 is activated by ROS or electrophiles, it will dissociate from the Keap1 dimer and translocate into the nucleus. Subsequently, Nrf2 will dimerize with the transcriptional activator, Maf. It will begin transcription of target genes that contain the antioxidant response element (ARE) motif. a Potential optogenetic perturbation of Keap1 (with a fused optogenetic construct such as CRY2/CIB) could induce clustering [16, 137] and prevent Keap1 from forming a homodimer and functioning normally. This perturbation could act as a potential gain-of-function of Nrf2 target gene transcription. b Potential optogenetic perturbation of Nrf2 (with a fused construct such as CRY2/CIB) could cause Nrf2 to cluster and prevent movement through the nuclear membrane. Such regulation would act as a potential loss-of-function of Nrf2 target gene transcription

As summarized in Table 1, stem cell signalling pathways that have already been studied using optogenetics include Wnt [137], RTK (receptor tyrosine kinase) [138], Ras/ERK (extracellular signal regulated kinase) [22], Toll, EGFR (epidermal growth factor receptor) [16, 139], BMP (bone morphogenetic protein) [140], and FGF (fibroblast growth factor) [141]. Conversely, a non-exhaustive list of other potential signalling pathways which have not yet been studied with optogenetics (at the time of this review), but have significant potential include Notch, Hedgehog, Hippo/YAP (Yes-associated protein), and JAK/STAT (Janus kinase/signal transducers and activators of transcription) (Table 2). The application of optogenetics to regulate Nrf2 and other pathways involved in stem cell activity will provide researchers with more precise tools to study age-related degeneration and aging (for example, comparing median-age flies to elderly flies).

Table 1

Stem cell signalling pathways already targeted with optogenetics

Signalling pathways	Function	Protein interaction	Optogenetic system used
Wnt/Wg	ISC maintenance [159], cell fate specification, cell polarity [160]	Beta-catenin (Arm) oligomerization	CRY2 [137]
RTK	Cell proliferation, differentiation, cell cycle control, cell metabolism [161]	Membrane receptor oligomerization	CLICR [138]
Ras/ERK	Cell proliferation, cell cycle arrest, differentiation [162]	Plasma membrane recruitment	OptoSOS [22]
Toll	Immune response, cell proliferation [163]	Membrane receptor oligomerization	CRY2 [16]
EGFR	Cell division, survival, migration [164, 165]	Membrane receptor oligomerization	CRY2 [16]
BMP	Embryogenesis, adult tissue maintenance [166]	Dimerization, recruitment of SMAD proteins	optoBMP [140]
FGF	Cell proliferation, migration, patterning [167]	Membrane receptor homodimerization	OptoFGFR1/CRY2 [141]

Each pathway plays a unique but interconnected function in regulating stem cell and homeostatic activity. With an understanding of the protein interactions for each pathway, specific optogenetic systems have been applied, optimized, or developed to regulate activity

Table 2

Stem cell signalling pathways with potential for optogenetic manipulation

Signalling pathways	Function	Protein interaction
Nrf/Keap1	Redox regulation [129, 168]	Dimerization, nucleartranslocation
Notch	Neural differentiation, ISC maintenance, enterocyte differentiation [169, 170]	Nuclear translocation
Hedgehog	Embryo pattern formation, progenitor and stem cell proliferation [171]	Protein–protein interactions
Hippo/YAP	Cell proliferation, apoptosis [172]	Phosphorylation
JAK/STAT	Cell proliferation, stem cell maintenance [173]	Dimerization, nuclear translocation

This non-exhaustive list highlights signalling pathways which have not yet been targeted with optogenetics, despite their important roles in homeostasis. The specific protein interactions of these signalling pathways inform the type of optogenetic system that researchers can use for spatiotemporal control and investigation

Models for protein-folding diseases

Given the recent success of optogenetics-based models, these tools have significant potential to create models of protein-folding diseases or proteopathies. One potential new model would involve optogenetically induced aggregation of amylin to study Type 2 diabetes. One of the pathological hallmarks of Type 2 diabetes is the presence of islet amyloid deposits composed of amylin (human islet amyloid polypeptide precursor or hIAPP) [142]. Amylin is a 37-amino-acid polypeptide that is toxic to pancreatic islet cells [143]. Levels of human amylin cause different extents of oligomerization, which correspond to diabetes disease stages. Diabetes phenotypes differ according to amylin overproduction levels [144], and there are similarities between Type 2 diabetes and Alzheimer’s disease [145]. The toxicity of amylin fibrils is similar to the toxicity of Amyloid-β proteins in Alzheimer’s disease [143]. Amylin oligomerization in the pancreas and amyloid formation in the brain contributes to the degeneration in AD and Type 2 diabetes [146].

Since the fat body organ in Drosophila plays a similar function in energy storage and metabolism to the human liver, it is a useful in vivo model to study amylin aggregation. Expression of human and mouse amylin to induce amylin aggregates in central nervous system (CNS) and fat body regions have been completed [147]. In addition, researchers have completed a 3D crystal structure analysis of hIAPP in Drosophila fat body tissue [148], providing useful insights for manipulation of the aggregation process. Future studies could apply optogenetic systems which regulate protein oligomerization, such as CRY2, PhyB, and LOV, to induce amylin aggregation. The optogenetic advantage of spatiotemporal control enables three variables for researchers to control when studying amylin and its effects: (1) levels of amylin aggregation; (2) age or time points at which aggregation is induced;, and (3) chosen areas of fat body tissue or central nervous system (CNS) to induce aggregation. Optogenetics would enable control of these three important dimensions for a useful Type-2 diabetes in vivo model of amylin-based pancreas degeneration in Drosophila.

Optogenetics can also augment models of neurodegenerative diseases, where protein misfolding plays a major role. The mechanisms by which protein aggregation relates to neurodegeneration are still subject to debate. Optogenetic control of protein aggregation can help to provide some insight into this gap. For example, in Huntington’s disease patients, the Huntingtin protein mis-folds and aggregates because of an expanded polyQ sequence in the gene [149]. Studies could take advantage of the control and visible readouts of protein aggregation with models that use optogenetic fusion constructs with the Huntingtin protein or any other proteins that aggregate. These models would adopt a similar approach to previously developed Amyloid models.

Adopting a closed-loop approach

In addition to new models of disease, a closed-loop approach to designing optogenetic tools for cell signalling networks will also be an important step for the future of optogenetics in homeostasis (Fig. 3). Closed-loop control is an engineering concept that is defined as using the difference between the measured output and a desired target to guide changes in the control inputs to a system [150]. This stands in contrast to open-loop control, where the measured output of a system has no effect on the control input. In neuroscience, this approach has been adopted and involves perturbing neural systems to have real-time control over neural dynamics and animal behaviour [120, 121, 150, 151].

Fig. 3

Open- and closed-loop approaches for optogenetics. a Open-loop approach. The control input is the light stimulus and the output is a selected measure of interest for pathway activity. The measured output of interest has no impact on the control input. b Closed-loop approach. The target is a predetermined level of a measure of interest (e.g., pathway activation). When the measured output level differs from the expected target level, the controller considers the “error margin” between both levels, and adjusts the input accordingly. Examples of adjustments include changing the light wavelength or pulse frequency of illumination. The measured output influences the control input via a feedback loop

Given the complexity of cell signalling networks, a similar closed-loop approach would yield important insights for the interplay between signalling pathways. A closed-loop approach requires three parts: (1) availability of a target; (2) system control inputs; and (3) measured system outputs [150]. The system control input would be light stimulus that is modulated based on measured output [150]. In cell signalling networks, the measured system output could be any of the downstream proteins or effects of the signalling pathway (such as protein activity, gene expression or membrane modification). The target would be a selected level of downstream activity of the signalling pathway. “Closing the loop” with optogenetics would help to elucidate the delicate feedback mechanisms involved in homeostatic signalling.

In cell biology, the future challenge will be to develop accurate readouts of the output of interest. For example, in the Wnt signalling pathway, a potential measured output would be the recruitment of Dishevelled (Dvl) proteins to the plasma membrane or stabilization of β-catenin when canonical Wnt signalling is activated [152]. This activity would then be used to create automatic feedback for modulating the optogenetic light stimulus. The diverse range of output types to measure and target in cell signalling will remain a challenge to be solved when applying a closed-loop control approach.

Even though optogenetics provides unparalleled spatiotemporal control, researchers should keep in mind the limitations of this tool, controls required and alternate approaches to ensure experiments are designed appropriately and data are interpreted carefully. Some examples include: (1) excitation with high frequency blue light is toxic to tissues, causes excessive heating [153], and has limited penetration to deeper tissues [154]. (2) Light can affect normal physiological processes such as brain function and temperature changes associated with light illumination during in vivo optogenetic manipulations could have both electrophysiological and behavioural consequences [155]. (3) This can be overcome by targeting optogenetic tools that are excited by near-infrared light (780–1100 nm) which causes less thermal heating and damage, in addition to displaying deep penetration [154]. (4) Careful experimental design to control the time course of light-driven responses, such as using pulses of light instead of continuous illumination which generates more heat, can help to remove confounding factors [155]. (5) The use of controls is required to substantiate that the optogenetic manipulations are physiological. (6) Different optogenetic constructs have inherent limitations such as specific on/off kinetics, excitation and reversal time (Fig. 1) that might render them unsuitable for certain applications.

Conclusion

In this review, we discussed recent advancements and applications of optogenetic approaches in homeostasis. The rapidly growing optogenetic toolkit is enabling the perturbation of a diverse range of targets and in vivo processes. Where other approaches and techniques may fail or lack in spatiotemporal control and specificity, optogenetics yields significant potential to uncover new insights for stem cells, developmental biology, protein-folding diseases, and cell signalling circuitry.