Author response:
eLife Assessment
This is a potentially important study comparing LTP mechanisms between primates and rodents. The experimental methods have some possible confounds, and the power (replicates) and design of the statistical methods could be strengthened, hence the support for the central claims of species differences is currently incomplete.
We thank the Editor and the Reviewers for taking the time to carefully review our manuscript and for providing constructive comments and suggestions, as well as the opportunity to revise our work.
Public Reviews:
Reviewer #1 (Public review):
Summary:
This is an important paper examining LTP induced by theta-burst stimulation in hippocampal slices from macaques and rats. While both species show theta-burst-late-LTP, only the non-human primate theta-burst-late-LTP showed synaptic tagging and capture that converts early-LTP into late-LTP in an independent synaptic pathway.
Strengths:
Synaptic tagging is a fundamental feature of repeated 100 Hz-tetanus-induced LTP, whereas theta-burst induction is arguably more physiologically relevant. Thus, synaptic tagging during theta-burst may differ in the two species, a distinction that may prove important in the mechanisms underlying the cognitive differences between the species.
Weaknesses:
Bursts repeated at the frequency (~5 Hz) of the endogenous theta rhythm induce strong LTP, primarily because this frequency disables feed-forward inhibition and allows sufficient postsynaptic depolarization to activate voltage-sensitive NMDA receptors. Therefore, the species differences may be due to differences in inhibition, rather than in molecular mechanisms of maintenance. One way to assess the relative strengths of this early induction mechanism in rats and macaques is to examine the "depolarization envelope" during the sequential bursts, which may be determined from the recordings already obtained. (Larson and Munkácsy, Theta-burst LTP, Brain Res 2015 Sep 24:1621:38-50. doi: 10.1016/j.brainres.2014.10.034)
Another issue is that the PKMzeta-antisense oligodeoxynucleotides block the synthesis of the kinase. However, Mei F, Nagappan G, Ke Y, Sacktor TC, Lu B (2011), BDNF Facilitates L-LTP Maintenance in the Absence of Protein Synthesis through PKMzeta. PLoS ONE 6(6):e21568, provided evidence that BDNF and theta-burst stimulation can act to increase PKMzeta by a protein synthesis-independent mechanism, presumably through decreased degradation. Therefore, the absence of an effect of the PKMzeta-antisense does not exclude the possibility that persistently increased PKMzeta is the mechanism of theta-burst-late-LTP maintenance in mice or macaques. This issue is worth discussing.
We sincerely thank the reviewer for the positive evaluation of our study and for highlighting the significance of examining synaptic tagging and capture following theta-burst stimulation (TBS) in rodents and non-human primates.
We agree that TBS is a physiologically relevant induction paradigm and that differences in inhibitory circuit dynamics may also contribute to the species-specific effects observed in our study. As highlighted by Larson and Munkácsy (2015), repeated bursts delivered at theta frequency (~5 Hz) can transiently suppress feed-forward inhibition through GABAB receptor-mediated mechanisms, thereby enhancing postsynaptic depolarization and facilitating NMDA receptor activation. We therefore agree that species differences in inhibitory regulation and burst-evoked depolarization may contribute to the distinct expression of synaptic tagging and capture observed between rats and non-human primates.
We further agree that analysis of the “depolarization envelope” during sequential bursts may provide additional insight into the relative strengths of early induction mechanisms. We will therefore perform these analyses using the existing recordings and compare the depolarization envelope between rodents and NHPs in the revised manuscript. Following the reviewer’s suggestion, we will expand the Discussion section to acknowledge the potential contribution of inhibitory circuit dynamics and depolarization envelope differences during sequential bursts.
Importantly, however, we believe that differences in downstream molecular maintenance mechanisms also contribute to these species-specific effects. In support of this, our molecular analyses revealed enhanced recruitment of plasticity-related proteins and transcriptional pathways in NHP hippocampus following TBS, including increased expression of BDNF and PKCζ. These findings suggest that both induction-related network properties and downstream molecular stabilization mechanisms may collectively contribute to the enhanced associative plasticity observed in NHPs.
We also thank the reviewer for the important point regarding PKMζ antisense experiments and the study by Mei et al. (2011). We agree that the absence of an effect of PKMζ antisense oligodeoxynucleotides does not necessarily exclude a role for persistently elevated PKMζ in the maintenance of theta-burst late-LTP. As demonstrated by Mei et al., BDNF together with theta-burst stimulation can maintain late-LTP in the absence of protein synthesis, potentially through stabilization of PKMζ protein levels by reducing degradation rather than through de novo synthesis. However, these findings are not directly comparable to our study, since our experiments involved theta-burst stimulation alone without exogenous BDNF application. Interestingly, our results suggest species-specific differences in the interaction between BDNF and PKMζ signaling pathways. In rats, TrkB/Fc-mediated blockade of BDNF impaired TBS-LTP maintenance, whereas PKMζ inhibition alone had no significant effect. In contrast, in NHP hippocampal slices, inhibition of either BDNF signaling or PKMζ alone failed to abolish late-LTP, whereas simultaneous inhibition of both pathways disrupted LTP maintenance.
These findings suggest that endogenous BDNF signaling and PKMζ may operate through partially redundant or compensatory mechanisms, particularly in the primate hippocampus. Therefore, although our findings indicate that de novo PKMζ synthesis may not be strictly required under the present experimental conditions, we cannot fully exclude the possibility that protein synthesis-independent stabilization or maintenance of PKMζ contributes to theta-burst late-LTP maintenance in rodents or NHPs. We will now clarify this point in the revised Discussion section.
Reviewer #2 (Public review):
Summary:
This study compares theta-burst stimulation (TBS)-induced synaptic plasticity in hippocampal CA1 slices from rats and non-human primates (Macaca fascicularis). The authors report that while TBS induces persistent LTP in both species, only primate hippocampal slices exhibit synaptic tagging and capture (STC) under these conditions. They further show increased BDNF and PKMζ expression following TBS in primates and propose that a redundant BDNF/PKMζ signaling architecture supports persistent plasticity in primates, whereas rodent TBS-LTP depends primarily on BDNF. The work aims to identify species-specific specializations in associative plasticity with implications for translational neuroscience.
Strengths:
The topic is potentially important because direct comparisons of hippocampal plasticity mechanisms between rodents and primates are rare.
Weaknesses:
(1) Limited biological replication in the primate experiments
The manuscript's strongest claims rely on data obtained from 36 slices from 7 monkeys, qPCR analyses with n=3 biological replicates, and Western blot analyses with n=3 biological replicates. The effective sample size for species-level conclusions is therefore not large. The manuscript frequently treats slices as independent observations while drawing conclusions about species differences. This is particularly problematic for electrophysiological experiments because multiple slices appear to originate from the same animals. The statistical unit should be the animal, not the slice, unless nested analyses are performed.
The authors should (1) report the number of animals contributing to each experiment, (2) provide animal-level analyses, (3) use mixed-effects or hierarchical models where appropriate, and (4) clarify whether multiple slices from the same monkey contributed to the same experimental condition. Without these analyses, the evidence for species-specific mechanisms remains weaker than presented.
We thank the reviewer for this important and thoughtful comment regarding statistical interpretation and biological replication. We agree that, particularly for electrophysiological experiments where multiple slices may originate from the same animal, the effective sample size for species-level conclusions should be considered at the animal level rather than solely at the slice level.
In the revised manuscript, we will clearly indicate the number of biological replicates (animals) together with the number of slices contributing to each electrophysiological experiment, as well as the biological replicates used for qPCR and Western blot analyses. We will also clarify whether multiple slices from the same NHP/rat contributed to the same experimental condition. These details will be incorporated into the figures and figure legends wherever appropriate.
In addition, we will perform animal-level analyses by averaging slice responses within each animal prior to statistical comparison and, where appropriate, apply hierarchical or mixed-effects statistical models to account for the nested structure of slices within animals.
We acknowledge that the number of non-human primates (NHPs) available for this study was inherently limited because of the substantial ethical, logistical, financial, and technical challenges associated with primate electrophysiology and tissue collection. Consequently, achieving sample sizes comparable to rodent studies is often not feasible in NHP research. Nevertheless, to further strengthen the biological robustness of the findings, we are currently in the process of obtaining additional NHP brain samples and plan to repeat key experiments in an additional 3-4 animals. We believe these revisions and additional experiments will substantially strengthen the statistical rigor and overall interpretation of the study.
(2) The central STC conclusion requires stronger controls
The most important result is that TBS supports STC in primates but not rats (Figures 1F-G). However, several alternative explanations are not excluded. For example, only a single interval (30 min) between TBS and WTET is examined. Classical STC studies characterize tag duration, PRP availability window, and temporal asymmetry. The current work does not determine whether primates exhibit longer tag persistence, increased PRP synthesis, altered capture efficiency, or merely a shifted temporal window. A temporal series (e.g., {plus minus}15, {plus minus}30, {plus minus}60, {plus minus}90 min) would substantially strengthen the mechanistic interpretation.
We thank the reviewer for this insightful comment regarding the mechanistic interpretation of the STC findings. In the present study, we selected the 30 min interval based on well-established classical STC paradigms in rodents, where this interval reliably falls within the effective tagging and capture window. Using this experimentally validated interval allowed us to directly compare whether TBS is sufficient to support STC in primates versus rats under equivalent experimental conditions. Accordingly, the primary objective of this study was to determine whether TBS-induced STC varies across species, rather than to comprehensively define the temporal dynamics of the tagging window.
We agree, however, that the current experiments do not distinguish whether the primate-specific effect reflects prolonged tag persistence, enhanced plasticity-related protein (PRP) synthesis, altered capture efficiency, or a shifted temporal window. Addressing these possibilities would indeed require systematic temporal interval analyses (e.g., ±15, ±30, ±60, and ±90 min), which represent important future directions. Such experiments are particularly challenging in non-human primates because the availability of primate tissue and experimental resources for large-scale electrophysiological studies remains limited and is currently beyond our experimental capacity due to substantial ethical, logistical, financial, and technical constraints.
Nevertheless, we fully agree with the reviewer that these experiments are important for advancing the mechanistic interpretation of the findings. Similar temporal analyses have recently proven informative in our rodent studies (Chong YS, Ang SR, Sajikumar S. Commun Biol. 2025;8:553). Importantly, we are currently in the process of obtaining additional non-human primate samples and plan to extend the present work by examining an additional 60 min temporal interval to further characterize the temporal properties of synaptic tagging and capture in non-human primates.
(3) Species differences may reflect tissue quality or preparation differences
The manuscript compares 5-7 week-old rats with 5-7 year-old monkeys. These are very different developmental stages. Moreover, euthanasia methods, extraction procedures, and post-mortem handling are different. These factors can affect BDNF expression, protein synthesis, LTP magnitude, and transcriptional responses. The authors should discuss these caveats more explicitly.
We thank the reviewer for raising this important and insightful point. We agree that differences in developmental stage between the experimental groups represent an important consideration when interpreting potential species-dependent effects. In the present study, rat experiments were performed in 5-7 week-old animals, whereas non-human primate (NHP) tissues were obtained from 5-7-year-old monkeys. This difference largely reflects the practical, ethical, and logistical constraints associated with NHP research and tissue availability. We acknowledge that these ages are not developmentally equivalent and that maturation state may influence BDNF signaling, protein synthesis capacity, synaptic plasticity thresholds, and transcriptional responses relevant to late-LTP and STC mechanisms.
We also recognize that differences in euthanasia procedures, tissue extraction, slice preparation, and postmortem handling between rodent and primate tissues may influence tissue physiology and electrophysiological properties. Although extensive care was taken to optimize tissue viability and maintain stable recordings within each species, these variables cannot be completely excluded as contributing factors to the observed differences.
Accordingly, we will revise the Discussion section to more explicitly acknowledge these limitations and clarify that our findings support potential species-dependent differences under the present experimental conditions, rather than definitive intrinsic species-specific mechanisms. Nevertheless, despite the inherent challenges associated with NHP electrophysiological studies, we believe that the present findings provide an important initial framework for understanding the translational relevance of synaptic tagging and capture mechanisms across species.
(4) Statistical reporting is incomplete
Many comparisons report exactly Wilcoxon p = 0.0313 and U-test p = 0.0022, across numerous experiments. This suggests very small sample sizes and discrete nonparametric distributions. The manuscript should report exact n values for each comparison, effect sizes, and confidence intervals.
Second, many genes and proteins are tested. No correction for multiple testing is described. The authors should state whether corrections were applied, and if not, justify this choice.
We thank the reviewer for this important comment regarding statistical reporting and interpretation. We agree that the repeated occurrence of identical exact p-values in several nonparametric analyses reflects the relatively small sample sizes and the discrete nature of the statistical distributions. This issue is particularly relevant for the NHP experiments, where biological replication is inherently limited because of the substantial ethical, logistical, financial, and technical challenges associated with obtaining and processing primate tissue.
In the revised manuscript, we will provide exact n values for all comparisons, including the number of biological replicates (animals) and slices where applicable. We will also include additional statistical details, including effect sizes and confidence intervals where appropriate, to improve transparency and facilitate interpretation of the reported findings. Furthermore, we are currently in the process of obtaining additional NHP samples and will attempt to include more biological replicates in the revised version to further strengthen the robustness of the analyses.
We also agree that the issue of multiple testing should be addressed more explicitly, particularly because multiple genes and proteins were examined. In the revised manuscript, we will clearly state the statistical correction methods applied for multiple comparisons where appropriate. For analyses in which corrections were not applied, we will provide justification, noting that several experiments were based on hypothesis-driven candidate targets rather than exploratory large-scale screening analyses. These statistical considerations will be clarified in the Methods and Results sections.
(5) Interpretation and significance
The study addresses an important and understudied question: whether associative synaptic plasticity mechanisms differ between rodents and primates. The finding that TBS can support STC in the primate hippocampus is potentially novel and impactful. However, the mechanistic evidence remains incomplete, the molecular analyses are underpowered, and several key controls are missing. At present, the data support the conclusion that under the specific experimental conditions tested, TBS-induced plasticity in primate hippocampal slices exhibits greater associative persistence than in rat slices.
The stronger claims regarding evolutionary specialization, fundamentally distinct plasticity rules, altered STC thresholds, and redundant BDNF/PKMζ architecture require additional experimental support.
We thank the reviewer for this thoughtful and balanced assessment of our work. We agree that the present data primarily support the conclusion that, under the specific experimental conditions examined, TBS-induced plasticity in primate hippocampal slices exhibits greater associative persistence than that observed in rat slices. We also agree that broader interpretations regarding evolutionary specialization, fundamentally distinct plasticity rules, altered STC thresholds, and potentially redundant BDNF/PKMζ-related mechanisms require additional mechanistic investigation and experimental validation.
Accordingly, we will moderate these interpretations throughout the revised manuscript and clearly state that these conclusions remain preliminary. We will further emphasize that additional experiments, including increased biological replication, expanded temporal analyses, and further mechanistic investigations, will be necessary to more conclusively define the basis of the observed species-dependent differences. Within our current experimental capacity, we are actively working to obtain additional non-human primate samples and plan to incorporate additional biological replicates and key follow-up experiments in the revised version to further strengthen the robustness of the findings.
At the same time, we believe the present study provides an important initial contribution to an understudied area by directly examining synaptic tagging and capture mechanisms in the primate hippocampus. Given the limited availability of non-human primate electrophysiological data in the field, these findings may offer a valuable framework for future studies investigating the translational and evolutionary relevance of associative synaptic plasticity mechanisms across species.
Reviewer #3 (Public review):
Summary:
In this manuscript, the authors have undertaken an investigation of differences between two mammalian species, the brown rat and the crab-eating macaque, in the mechanisms supporting a well-established model of long-term Hebbian synaptic plasticity, Schaffer collateral to CA1 Long-term potentiation (LTP) in the hippocampus. LTP has been long-studied and deeply characterised due to its potential importance in modeling a strong candidate process for the central mechanism of learning and memory. LTP was first discovered in lagomorphs (rabbits), but has since been much more widely studied in rodents (mostly rats and mice), and there has been some complementary work revealing LTP in non-human primates and even in humans, revealing largely overlapping canonical mechanisms of induction, expression, and maintenance. More specifically, this study puts a particular focus on the fascinating associative features of this form of lasting synapse-specific modification, in which a synaptic input can be stimulated with a relatively weak induction protocol that will not produce lasting plasticity on its own, but can undergo lasting LTP if paired with stronger stimulation on a separate synaptic input to the same neuron. This associativity mechanism is particularly attractive within the Hebbian synaptic plasticity framework as it provides a candidate mechanism for associative forms of learning in which stimulus-stimulus, stimulus-reward, stimulus-punishment, or action-outcome associations are formed. A particularly attractive feature of this associative LTP is that there can also be a substantial time-lag between the strong stimulation of one pathway and the weaker stimulation of the other synaptic input, which only undergoes lasting LTP by hijacking the proteins synthesized as a result of strong stimulation elsewhere. This observation has led to the famous tagging and capture hypothesis as an explanation of how such synapse-specific change can be achieved on both stimulated inputs but not on other synaptic inputs, given the potential requirement for cell-wide protein synthesis. This theory, for which there is very strong experimental evidence, posits that a protein tag is left at synapses that have been stimulated with sufficient vigor in recent history, serving as a key mechanism to ensure that those weakly stimulated synapses will undergo change when a larger-scale LTP event occurs due to stronger stimulation elsewhere within a relevant time window. Again, this idea is attractive as it can explain how we might form associations between events that occur slightly separated in time. The manuscript goes on to show that an induction protocol that is particularly physiologically relevant, theta burst stimulation, produces this tag and capture associative effect in ex vivo slices of Macaque hippocampus, much more readily than in side-by-side ex vivo slices of rat hippocampus. Moreover, the manuscript delves into the importance of well-characterised LTP maintenance mechanisms, including PKMzeta and BDNF, which are key factors that ensure that altered synaptic change is maintained for long periods of time despite substantial molecular turnover in the neuron. The observation in this manuscript is that a degree of redundancy for these mechanisms exists in the primate species but not the rodent species, as both mechanisms need to be inhibited to return LTP to baseline in the Macaque, but only one needs to be inhibited to have that effect in the rat. A major emphasis of this study is that there may be a step-wise difference in associative learning mechanisms between rodents and primates that may contribute to their differing cognitive capacities, although I believe a lot more evidence would be required to reach that conclusion.
Strengths:
The strengths of this study are that it is technically very proficient and is from a laboratory that has a long history of seminal work on synaptic tagging and capture. The cross-species comparison, particularly involving non-human primates, is also very hard to achieve, and a major strength here is the side-by-side comparison of slices from rat and monkeys. Further strengths of the study are the use of a number of experimental strategies, including both observation and intervention, to demonstrate differential involvement of LTP maintenance mechanisms. A final major strength is conceptual, as it is undoubtedly useful not only to identify shared mechanisms of plasticity between commonly used model organisms and either humans or much more closely related species such as old world monkeys, but also to reveal differences that have the potential to contribute to differences in memory/cognition.
Weaknesses:
The findings of this study are a very useful building block for understanding how generalisable mechanisms of LTP are. However, arriving at really substantial conclusions from these findings is challenging, as there are a number of variables that are unaccounted for in this study that may explain the differences that have been observed between rats and monkeys. One example of a potential confound to these interpretations is that rats are nocturnal/crepuscular animals, and macaques are diurnal animals. Thus, to undertake a like-for-like comparison, it would be necessary for the rats to be on a reversed light-dark cycle to ensure that the wake cycle of the rat (dark) is being compared with the wake cycle of the monkey (light). It is possible that the authors have done this, but it is not mentioned in the methods section. The reason this is important is that there is a substantial body of work indicating that different mechanisms are at play in hippocampal LTP during wake and sleep. Transcripts and proteins related to synaptic function are dramatically differentially regulated during sleep-wake cycles, and phosphorylation states of key proteins involved in plasticity are also altered. Moreover, synaptic tagging and capture are specifically disrupted by sleep deprivation. Perhaps the authors have already considered this factor and appropriately reversed the light-dark cycle of their rat subjects, in which case a clarification in the manuscript would be useful. Nevertheless, I have used this as an example because there is a variety of potential confounds that may explain the difference between SC-CA1 TBS LTP in rats and monkeys, e.g., circadian rhythms, degree of enrichment, natural light vs indoor lighting, diet, degree of inbreeding, strain, etc. Thus, to make strong conclusions about the potential for differences in plasticity rules/mechanisms and how those may contribute to differences in cognition, I think it would be necessary to compare a wider variety of species, including a good representation of each order (e.g., nocturnal rats and diurnal squirrels, new and old world primates) and not just a single exemplar. I understand, of course, that this is really pushing the boundaries of practicality, but I see no other way to make a strong conclusion or to generalise to mechanisms or properties of plasticity in rodent’s vs primates. Thus, while I believe the manuscript presents really admirable work, I am not sure the findings are at all easy to interpret.
We thank the reviewer for this thoughtful and insightful comment, as well as for the encouraging appreciation of our long-duration plasticity recordings and associative plasticity experiments, which are both technically demanding and time-intensive. We fully agree that interpretation of cross-species differences in synaptic plasticity requires careful consideration of multiple biological and environmental variables, including circadian state, enrichment conditions, strain differences, diet, lighting conditions, and species-specific behavioral ecology.
Regarding the specific concern related to circadian phase and sleep-wake state, the reviewer raises an important point. Rats are nocturnal animals, whereas macaques are diurnal, and hippocampal plasticity mechanisms are known to be influenced by circadian rhythms and sleep-dependent regulation of synaptic proteins and signaling pathways. Previous studies have demonstrated modulation of LTP, synaptic tagging and capture and protein synthesis in rats across normal sleep-wake cycles. We therefore agree that these factors may influence plasticity outcomes and should be carefully considered in comparative studies.
Studies have further shown that theta frequency is highly sensitive to sleep-related manipulations. Specifically, theta frequency decreases immediately after sleep, remains elevated during sleep deprivation, and rapidly declines following recovery sleep. In aged animals, these effects appear comparatively attenuated, suggesting reduced sleep-dependent modulation of theta dynamics with aging. Therefore, disruption of normal circadian or sleep-wake patterns may significantly alter theta activity and associated plasticity mechanisms within a species and may not accurately reflect physiological baseline states (Utku Kaya et al., 2026).
In our experiments, recordings from rats and macaques were performed during their respective active phases under standardized laboratory housing conditions, and we will further clarify these details in the revised Methods section. Nevertheless, we acknowledge that circadian state and related physiological variables cannot be completely excluded as contributing factors to the observed differences between species.
More broadly, we agree with the reviewer that the present study does not permit definitive conclusions regarding universal “rodent versus primate” rules of synaptic plasticity. Our intention was not to propose a generalized dichotomy between rodents and primates, but rather to report that, under the experimental conditions used here, SC-CA1 TBS-LTP and associated synaptic tagging mechanisms differed between rats and macaques. We agree that broader evolutionary or cognitive interpretations would require systematic comparative analyses across multiple species, including both nocturnal and diurnal rodents as well as diverse primate species. Such studies would provide a stronger framework for distinguishing conserved versus species-specific mechanisms of plasticity.
At the same time, we believe the present findings remain important because they provide one of the first direct experimental comparisons of SC-CA1 TBS-LTP-associated plasticity mechanisms between rodents and non-human primates under controlled ex vivo conditions. Although the interpretation should be done cautiously, the observed differences raise the possibility that certain metaplastic or protein synthesis-dependent mechanisms may not be fully conserved across species. Accordingly, we will revise the Discussion section to better emphasize the exploratory and comparative nature of the study, while explicitly acknowledging the limitations and potential confounding factors highlighted by the reviewer.
