Some chloroplast-stealing dinoflagellates remodel their kleptoplastids, form an extensive network with stollen mitochondria, and likely benefit from photosynthetic genes that they inherited or acquired from prey. These adaptations allow these protists to exploit kleptoplastids for months and shed light on past plastid acquisitions.

Everything in life requires energy. Whether harnessed directly from sunlight or acquired by consuming other organisms, energy is essential for survival. Some organisms have developed a clever way to gain energy -instead of investing in building their own photosynthetic machinery, they 'steal' chloroplasts (that is, plastids) from their prey and use them to harvest energy from the sun, a process called kleptoplasty 1 . In a new study published in this issue of Current Biology, Rao et al. 2 show the kleptoplastic Ross Sea dinoflagellate (RSD) can uptake, remodel, and even increase the size of its prey's nucleus, mitochondria, and plastids for energy. However, after approximately one week, the prey nucleus is lost, leading to a decrease in plastid volume and photosynthetic rate. Despite this, the plastid, and the closely associated prey mitochondria, continue to contribute functionally to the RSD for more than two months. Kleptoplasty has evolved independently across nearly all major groups of single-celled protists, as well as in multicellular animals such as sea slugs (Elysia) and acoel flatworms 1,[3][4][5] . This strategy serves as an evolutionary shortcut, temporarily acquiring plastids and, in many cases, other organelles from prey rather than (or prior to) permanently acquiring them. Such temporary acquisition allows organisms flexibility and a metabolic advantage, particularly in environments with unpredictable food availability. It also avoids the complexity, metabolic burden, and regulatory challenges associated with maintaining endosymbionts or acquiring a permanent organelle. However, these interactions are often short-lived, on the order of days to months, as the host is not able to replicate or indefinitely maintain the stolen organelles and must continuously encounter the right prey and replace the organelles for survival. While kleptoplasty is often portrayed as a free meal, it also has its costs, which include evolving metabolic dependencies upon photosynthetic metabolism 6 . Over repeated interactions spanning evolutionary time, horizontal gene transfer (HGT) from prey to host genomes are thought to be a key mechanism for gradually stabilizing the kleptoplast-host relationship 7 . Such gene transfers can establish essential plastid maintenance genes into the host genome, transforming kleptoplasty from a transient solution into a stepping stone towards stable, photosynthetic organelles 7 .

Indeed, kleptoplasty may reflect an early, transient stage of a much deeper evolutionary process: the origin of plastids themselves. Primary plastids originated from a single primary endosymbiotic event in which a heterotrophic eukaryote engulfed a cyanobacterium 8 . Over time, extensive HGT moved most plastid genes into the host nucleus, reducing the cyanobacterium into the simplified, genome-reduced plastids seen in modern red, green, and glaucophyte algae of the Archaeplastida 8 . Aside from this major event, only one other primary plastid acquisition is known in the rhizarian Paulinella chromatophore, and it remains in an earlier stage that is still far from the reduced form seen in the Archaeplastida 9 . Subsequent evolutionary events involved eukaryotes acquiring plastids from primary algae, resulting in a secondary plastid, and further by secondary plastids becoming sources for tertiary plastid acquisitions 8 .

Kleptoplasty has traditionally been viewed as an evolutionary dead-end for plastids, while endosymbiosis in the strict sense of host-symbiont dynamics has been used as a model for complex (i.e., secondary and above) plastid acquisitions. However, if one observes all extant plastid-bearing groups, almost none possess members that host endosymbiotic algae, but many of these lineages have members that are kleptoplastidic [10][11][12] . In addition, many groups of 'algae' possess members that have lost their plastids, and some have basal lineages that never had one. Further, some lineages have passed through the cycle of plastid acquisition and loss multiple times. As a result, the evolutionary history of plastids in eukaryotes is highly complex, resulting in a mosaic of nuclear-encoded plastidtargeted genes from different sources in certain lineages. Therefore, in some eukaryotic lineages, kleptoplasty may benefit from past evolutionary acquisitions, while in others (e.g., ciliates and perhaps Rapaza viridis) it begins with a 'clean slate'. Since the products of evolutionary trajectories are unknowable, extant relationships involving kleptoplasty may be interpreted as either stable adapted states, as evolutionary stepping stones, or as both.

Found globally, dinoflagellates are an ancient phylum of protists, originating >300 million years ago, and contributing significantly to their ecosystems 13 . They represent perhaps one of the most unique eukaryotic lineages due to their complex history, involving repeated acquisitions and losses of plastids 14 . Approximately half of known dinoflagellates are photosynthetic, possessing a stable plastid (Figure 1), and most of these are thought to be mixotrophic, also able to simultaneously ingest prey 14 . The remaining either never had a plastid or have lost their plastid (Figure 1), and are either heterotrophic, kleptoplastidic (also mixotrophs) (Figure 1), or parasitic 14 . Most photosynthetic dinoflagellates contain highly reduced, peridinin-type secondary plastids derived from red alga. These plastids typically contain fewer than 20 genes arranged in small minicircles, each encoding one or a few genes 14 . Remaining genes essential to plastid function have been transferred to the uniquely complex dinoflagellate nucleus 13,14 .

All other photosynthetic dinoflagellates with stable plastids have undergone serial secondary or even tertiary replacements of plastids. In these cases, the original red alga-derived plastid has been replaced by those derived from green alga (e.g., Lepidodinium spp.), by engulfed diatom cells (known as 'dinotoms' in Kryptoperidiniaceae), or by haptophyte-derived plastids (Kareniaceae genera; Karenia, Karlodinium, and Takayama) 14 . Dinotoms are generally not considered to possess fully domesticated plastids, since they permanently host a plastid, nucleus, and mitochondrion of their reduced diatom 'endosymbiont'. Remarkably, the origin of the dinotom organelles can be traced to multiple independent acquisition events 15 . The precise mechanism behind these serial plastid replacements in dinoflagellates remains unclear. However, the existence of closely related kleptoplastic species within both dinotom-11 and haptophyte-derived 10 lineages makes them ideal models for studying organelle acquisition and evolutionary maintenance.

The kleptoplastic RSD dinoflagellate belongs to the Kareniaceae family, and is closely related to species harbouring haptophyte-derived plastids. RSD may contain a non-photosynthetic relict peridinin-type plastid since it still expresses genes with peridinin-plastid targeting peptides, along with several plastid-targeting proteins acquired horizontally from various algal sources 10 (Figure 1). Conceivably, the presence of a relict peridinin plastid could have maintained the fitness of nuclearencoded, plastid-targeted genes following the loss of its photosynthetic functionality. Impressively, RSD can steal and retain plastids from the polar bloom-forming haptophyte Phaeocystis antarctica for periods as long as 29.5 months 16 . The longevity of these stolen plastids is likely due to genes RSD already has in its possession from previous HGT events 10 . Interestingly, RSD lacks certain essential genes, such as those encoding photosystem II -thus, when prey nuclei degrade, RSD presumably sustains energy through photosystem I cyclic electron transport until a new plastid can be acquired 10 .

Using advanced imaging techniques and molecular approaches, Rao et al. tracked how RSD and its kleptoplasts respond under starvation conditions 2 . Initially, RSD engulfs its prey and encloses plastids, nuclei, mitochondria, and potentially other organelles within a symbiosome membrane 2 (Figure 1). Each RSD cell can harbor numerous stolen plastids, and like in other kleptoplastic species, these prey organelles increase in volume and in photosynthetic capabilities 2 . RSD can remodel the prey mitochondria and place it in close proximately to the newly elongated plastids 2 . The mitochondria remain in tight association with the plastids throughout starvation (Figure 1), suggesting their critical role in RSD plastid maintenance 2 . Once host cells lose prey nuclei, plastids are no longer able to divide; however, photosystem II activity persists even after nine weeks of starvation 2 .

Kleptoplasty is not only a biological curiosity, but a valuable model for understanding cellular integration, metabolic plasticity, and complex plastid acquisitions. As in other kleptoplastidic dinoflagellates, for example, Dinophysis spp. 17 , RSD demonstrates elaborate structural adaptation for stolen organelles, maintained by an array of plastid-targeted proteins that originate from both a lost ancestral plastid(s) and past kleptoplastidic associations. Innovations that facilitate adaptation to exploiting photosynthetic metabolism prior to the stable acquisition of the organelle can also be found in the organelle-stealing ciliate Mesodinium rubrum, which steals a transcriptionally active nucleus that allows exploitation of its prey's full metabolic potential 6 . In the euglenozoan Rapaza viridis, adaptations that facilitate control of its kleptoplastids include HGT from diverse prey and developing a novel plastid-targeting system 12 . Unraveling the mechanisms by which host cells regulate, remodel, and maintain stolen organelles reveals startling adaptations towards a photosynthetic lifestyle prior to acquiring the genetic machinery to fully support it, thereby offering insights into ancient The ancestral dinoflagellate with peridinin chloroplasts, derived from a secondary endosymbiotic event (not shown here), that subsequently lost its plastids and became a phagotrophic heterotrophic dinoflagellate, then evolved to become kleptoplastidic, feeding on a range of prey likely spanning evolutionary time and perhaps multiple speciation events, finally acquiring new nuclear-encoded, plastid-targeted genes via horizontal gene-transfer (HGT) from several of these prey species, and becoming the extant Ross Sea dinoflagellate (RSD) that steals organelles from the haptophyte Phaeocystis antarctica. RSD appears to benefit from a mosaic of plastid-targeted genes acquired through past HGT events and has evolved cellular innovations to better exploit stolen organelle metabolism (allowing organelles to enlarge, forming a closer association between the plastid and mitochondrion, and an extensive mitochondrial network linking plastids). Note the ability to phagocytize prey has been maintained across evolutionary time as an ancestral trait. Each cartoon represents one or more speciation events. The existence of a vestigial ancestral peridinin plastid in the RSD and its ancestors is unknown but plausible (represented with a question mark). endosymbiotic events. Future research into these extraordinary interactions promises exciting discoveries at the intersection of evolution, ecology, and cell biology.

Vocal communication relies on accurate vocal output and real-time self-monitoring through auditory feedback. A recent study in marmoset monkeys explores the underlying neural mechanisms, revealing bidirectional communication between frontal and auditory cortices during vocalization. These findings suggest a key role for these brain areas in vocal monitoring and error detection.

Vocal communication is fundamental for social interactions across species, including humans, allowing individuals to exchange information. This process involves two components: a sender and a receiver, with vocal signals transmitted in dyadic interactions from one to the other over time. A crucial aspect of effective communication is the accuracy of vocal output. While the underlying system of vocal production is complex, it can be broadly divided into two key components: the motor production of sound and the continuous self-monitoring of that output. This self-monitoring process involves detecting and correcting potential errors by comparing the intended vocal output with the received auditory feedback. In other words, vocalizations, including speech, must be constantly monitored for accuracy. When a mismatch occurs, it is rapidly detected through auditory feedback and, when possible, adjusted in real time 1 . However, how the brain monitors vocal output and detects errors remains largely unknown. A new study in a recent issue of Current Biology by