When navigating the world, learned information allows us to operate efficiently and adaptively within an ever-changing environment. Memory describes the capacity of an organism to acquire, store and retrieve information following experience (Squire, 2009). This information guides behaviour and informs future decisions. How memory information is stored, when it is retrieved, and its effect on behaviour is heavily influenced not only by the relevant sensory cues but also by internal signals governed by other physiological systems that guide appropriate behaviour (Daniel, Roberts and Dohanich, 1999; Schwabe and Wolf, 2013, Verma, et al., 2016) This integration between both central and peripheral systems is crucial for physiological and behavioural adaptions that facilitate homeostatic control (Buchanan and Tranel, 2009).Fig. 1.Fig. 2.Fig. 3.Fig. 4..

The immune system is composed of specialised cell types and signalling molecules and is an essential system for the overall maintenance of bodily homeostasis, primarily by protecting against pathogens (Rankin and Artis, 2018). In recent years it has become apparent that the nervous system and immune system are highly connected (Dantzer and Kelley, 1989, Osterhout, et al., 2022). The early idea of “immune privilege” is no longer a common perspective and critical communication between the two systems has been well documented (Louveau, Harris and Kipnis, 2015). Though the blood–brain barrier (BBB) still confers important protection over the brain microenvironment by restricting the flux of substances and cell entry, communication between the brain and the broader immune system is well established (Galea, 2021).

More generally, brain-body bidirectional communication is a key element of physiological regulation. Fluctuations in homeostasis in response to a changing environment requires synchrony between behavioural and physiological modifications (Sammons et al., 2024). To give an example, increased energy demands require both metabolic shifts as well as altered feeding behaviour (Myers et al., 2021). More recently, it has been shown that the brain can form memories of cold experiences in order to pre-emptively increase thermogenetic metabolism when a cold temperature is predicted by the brain (Muñoz Zamora et al., 2025). Changes in immune status during inflammation and infection also requires essential brain-body interactions. The brain can both sense and respond to immune challenges (Jagot, et al., 2023, Jin, et al., 2024).

One of the most common examples of immune modulation of brain function and behaviour is during sickness (Hart, 1988). Sickness behaviour pertains to changes in behaviour and cognition due to immune interactions with the brain (Elmquist et al., 1997, Dantzer, et al., 2008). Release of immune signalling molecules, known as cytokines, in response to pathogens in the periphery can act on peripheral nerves such as the vagal nerves or even cytokine receptors at the BBB engaging important communication with the central nervous system (CNS) (Bluthé et al., 1994, Ek, et al., 1998, Dantzer, et al., 2000, Dantzer, 2004). Sickness behaviours include weakness, lethargy, social withdrawal, loss of appetite and even impairments in memory and attention (Konsman, Parnet and Dantzer, 2002; Ilanges, et al., 2022, Osterhout, et al., 2022). This is believed to be an adaptive strategy to redirect energy away from other physiological sources and direct it towards fighting the infection (Kluger et al., 1975, Dantzer, 2001). Sickness results in overt changes in behaviour and environmental interaction, which necessitates changes in information processing. Therefore, immune signalling must impart significant influence over memory function. Both memory and immune systems are crucial for environmental adaption, homeostasis and ultimately survival and key collaboration between these two systems is essential (Yirmiya and Goshen, 2011).

Here we review the bi-directional interactions between the brain and the immune system, in sickness and disease but also in normal physiology, and we discuss the influence of these integral neuroimmune interactions on both the “writing-in” and “reading-out” of memories.

Neuro-immune interactions in neuronal function and synaptic plasticity.

We often examine the immune system’s influence on higher-order cognitive function through the lens of sickness or disease states. However, to fully understand the role of the immune system in memory function we need to step away from this pathology-centric view, and shift our focus to neuro-immune interactions in normal physiology, plasticity and behaviour (Marin and Kipnis, 2013). The nervous and immune systems are distinctly characterised by specialised cell types and signalling molecules but there is an important cross-over in function and expression.

Neuronal plasticity describes a set of mechanisms by which neuronal stimuli can alter synaptic connections through the strengthening or weakening of transmission (Bliss and Collingridge, 1993). This adaptive mechanism partly underpins the brain’s ability to change in response to environmental experience, to acquire and store information and is therefore indispensable for memory (Martin, Grimwood and Morris, 2000). Following neuronal activation, numerous synaptic modifications occur both dependent and independent of protein synthesis that permit these changes in wiring that support the encoding and consolidation of memory information (Matus, 2000). These changes include structural plasticity in the form of immediate cytoskeletal modifications and alterations in receptor trafficking leading to facilitation or depression of synaptic transmission known as LTP or LTD, respectively (Bliss and Lømo, 1973). It is this plasticity that permits experiences to be translated into lasting changes that are thought to represent the memory information (Chklovskii et al., 2004, Kim and Linden, 2007, Poo, et al., 2016). In this section, we detail the influence of immune mediators on neuronal and synaptic physiology and plasticity.

Many immune molecules have pleiotropic functions, and some are found on neuronal cells and can directly affect the function and plasticity of neurons and synapses (Boulanger, 2009, Garay and McAllister, 2010). Classically, Shatz and Boulanger identified major histocompatibility (MHC) class I immune protein expression that is regulated by neuronal activity and acts as an important player in the molecular mechanisms governing plasticity (Boulanger and Shatz, 2004). Although MHCI proteins are primarily known for their role in antigen presentation and adaptive immune responses, they are widely expressed throughout the central nervous system including in cortical and hippocampal areas (Elmer and McAllister, 2012). Notably, MHCI mRNA expression is regulated by neuronal activity, peaking during windows of high plasticity such as sensory critical periods (Corriveau, Huh and Shatz, 1998). This implicated their importance in brain development and activity-dependent synaptic plasticity. Moreover, MHC proteins have been shown to regulate NMDA receptor function and AMPA receptor trafficking (Nelson et al., 2013). Transgenic mice lacking normal expression of MHCI proteins exhibit abolished NMDAR-dependent LTD in the hippocampal CA1 which was accompanied by deficits in hippocampal-dependent memory such as contextual and object memories (Nelson et al., 2013).

Complement proteins are immune proteins of the complex complement immunological surveillance system that is involved in orchestrating immune responses against pathogens and eliminating debris (Ricklin et al., 2010). Yet these classical immune proteins are also expressed on neurons under physiological conditions and govern neuronal-glial communication (Thomas, et al., 2000, Mastellos, 2014, Parker et al., 2022). Importantly, complement proteins such as C1q or C3 have been implicated in developmental synaptic elimination (Stevens et al., 2007). CX3CL1 (fractalkine) is a chemokine involved in important neuronal-glial interactions. Its receptor, C-X3-C Motif Chemokine Receptor 1 (CX3CR1) is a transmembrane G-protein-coupled receptor exclusively expressed by microglia in the brain but also widely expressed by peripheral immune cells (Maciejewski-Lenoir, et al., 1999, Gerlach, et al., 2016, Burgess, et al., 2019). CX3CL1 is primarily expressed by neurons in both membrane-bound and soluble forms. CX3CR1 is important in the regulation of microglial state plasticity, migration and proliferation throughout development and synaptic pruning (Harrison, et al., 1998, Maciejewski-Lenoir, et al., 1999). Mice lacking CX3CR1 have increased dendritic spine number and reduced excitatory postsynaptic potentials (EPSPs) (Paolicelli et al., 2011). Behaviourally, studies have demonstrated both altered fear memory (Reshef, et al., 2014, Schubert, et al., 2018) and spatial memory deficits (Maggi et al., 2011) in CX3CR1-deficient mice.

In addition to the direct expression of immune proteins by neurons, immune molecules expressed by resident or infiltrating immune cells can directly impact the function of neurons. Neurons express receptors for cytokines interleukin-1beta (IL-1β), IL-6, TNF and IL-17 (Yang, Huh and Choi, 2023). TNF is constitutively released by glial cells and can modulate synaptic strength through alterations in AMPA receptor trafficking and exocytosis of GABAa receptors (Beattie et al., 2002).

The cytokine IL-17 is best known for its role in inflammatory signalling and host defence, recent evidence has highlighted novel neuromodulatory roles for IL-17 in metabolism and feeding behaviour (Douglas et al., 2023, Douglas, et al., 2024, Gallo, et al., 2024). Gamma-delta T-cells are an important non-inflammatory source of IL-17 which has key physiological roles and has been suggested to play a role in memory (Ribeiro, et al., 2019, Alves de Lima, et al., 2020). Mice lacking either meningeal gamma-delta T-cells or IL-17A exhibited short-term memory deficits as measured through a short-term Y-maze task and reduced synaptic plasticity and hippocampal LTP (Ribeiro et al., 2019). It remains unclear how exactly meningeal IL-17 influences brain function and behaviour but indicative evidence suggests it may mediate its effects on memory through interaction with microglia. IL-17 promotes BDNF release from glial cells and exogenous application of BDNF to mice lacking IL-17 was sufficient to rescue deficits in LTP and short-term memory (Ribeiro et al., 2019). In human studies, administration of IL-6 to healthy participants results in an enhancement of sleep-associated memory consolidation (Benedict et al., 2009).

Memory function requires plasticity and strengthening of synaptic transmission and these works demonstrate a level of interaction between immune molecules and neuronal function that ultimately affects synaptic function and memory expression.

Resident immune cells are found within the brain parenchyma as well as brain-border tissues such as the meninges (Korin et al., 2017). Notably, the brain harbours microglia, specialised macrophage-like cells characterised by their highly ramified morphology and dynamic processes (Murabe and Sano, 1983, Nimmerjahn et al., 2005). As the CNS representatives of the immune system, they participate in important immune signalling to provide neuroprotection in response to immune challenges, invading pathogens or damage (Eyo and Wu, 2019). Furthermore, and much like their peripheral counterparts, microglia are an important source of cytokines and chemokines and possess phagocytic capacities used to engulf debris, pathogens or necrotic and apoptotic cells to maintain tissue homeostasis throughout the lifespan (Stence et al., 2001, Hanisch, 2002, Ransohoff and Perry, 2009). From this perspective, microglia are ideally positioned to communicate changes in immune status within the brain parenchyma, engaging in critical neuronal communication and modulation that may underpin behavioural changes, including alterations in learning and memory that often accompany inflammation and sickness. Their active surveying of the brain microenvironment and close communication with neurons through fractalkine (CX3CR1) signalling, cytokine and neuromodulation release and even direct neuronal modulation though ATP and neurotransmitter release permits translation of immune signals into neuronal modulation (Murabe and Sano, 1983, Nimmerjahn et al., 2005). Yet beyond these classically described immunological functions, microglia have other integral roles in brain development and the regulation of neuronal function under homeostatic conditions (Paolicelli and Ferretti, 2017; Thion, Ginhoux and Garel, 2018; Badimon et al., 2020). Microglia epitomize neuro-immune interactions, balancing neuronal interactions and immune functions to maintain homeostasis (Li and Barres, 2018).

Perhaps one of the most well-described roles of microglia in brain development is the postnatal sculpting of neuronal circuits (Stevens, et al., 2007, Paolicelli, et al., 2011, Schafer, et al., 2012). The early postnatal brain contains supernumerary synapses. Throughout development and as the individual is exposed to sensory input as well as the contribution of spontaneous activity, circuits are refined accordingly, and synapses are formed and eliminated as appropriate (Hua and Smith, 2004). This developmental synaptic pruning appears to be vital for circuit modelling and establishing accurate connectivity and cognitive function (Stevens et al., 2007).

Memory can be considered as a terminal yet ongoing stage of brain development, and the synapse is a crucial structural locus of memory-related plasticity (Chklovskii et al., 2004, Tonegawa, et al., 2015). Given the association of microglia with synaptic remodelling and engulfment of synaptic material, it is not unexpected that microglia may play a role in memory (Wang, Wang and Gu, 2021). Beyond the pronounced synaptic remodelling and elimination described during postnatal development, several studies have detailed microglia synaptic interactions across the lifespan that, at a molecular level, modulate synaptic plasticity and at a behavioural level, impact learning and memory (Parkhurst, et al., 2013, Wang, et al., 2020). Microglia ablation using Cre-dependent diphtheria toxin expression results in impaired synapse elimination and formation following motor learning, while CX3CR1 KO mice exhibit enhanced hippocampal-dependent memory (Reshef et al., 2014). It must be noted, however, that CX3CR1 also regulates peripheral immune cell function and infiltration that may contribute to observed effects (Garré et al., 2017).

Microglial release of the neurotrophin BDNF influences synaptic plasticity through modulation of synaptic proteins (Parkhurst et al., 2013). Experience-driven microglial remodelling of the extracellular matrix via IL-33 signalling has been shown to promote synaptic plasticity and remote memory precision (Nguyen et al., 2020). Similarly, Microglial Rac1, a RhoGTPase involved in cytoskeletal remodelling, has been shown to also regulate microglia-synapse interactions and influence synaptic plasticity (Socodato et al., 2023).

Recent work investigating the role of microglia in natural forgetting found that inhibition of microglial activity or depletion following contextual fear conditioning prevented a reduction in freezing behaviour indicative of forgetting and increased memory engram reactivation during recall test (Wang et al., 2020). Furthermore, microglial TLR2/4 activation has been suggested to underpin Rac1 driven destabilization of fear memories (Chen et al., 2024).

The impact of microglial modulation of synapse and neural function on wide-ranging behaviours is becoming increasing clear and newer approaches employing optogenetic and chemogenic tools for direct microglial modulation provide even strong evidence for microglia modulation of behaviour, relevant for both homeostatic behavioural control and sickness or inflammation driven behavioural alterations (Klawonn, et al., 2021, Yi, et al., 2021, Nagarajan and Capecchi, 2023, Zhao et al., 2024).

Beyond these specialised macrophages, there are also bona fide immune cells found within the brain and brain-border tissues (Korin et al., 2017). T-cells migrate from the blood and populate areas such as the meninges and choroid plexus and can infiltrate the brain under certain conditions, a classic feature of neuroinflammation (Ransohoff, Pia Kivisäkk, and Grahame Kidd, 2003; Smolders, et al., 2018, Alves de Lima, et al., 2020). For example, in disease states such as multiple sclerosis (MS), both T cells and B cells are found in the CNS (Machado-Santos et al., 2018). Brain infiltration of neutrophils, cells of the innate immune system, is seen during infections of the CNS and in neurodegenerative conditions such as Alzheimer’s disease (Kanashiro et al., 2020). However, under homeostatic conditions, meningeal T-cells have been shown to support memory. Depletion of meningeal T cells using Anti-VLA4 antibodies that inhibit T cell migration, resulted in an impairment of learning and memory, where T cell derived IL-4 was shown to play a crucial role (Derecki et al., 2010).

While the infiltration of immune cells during neuroinflammation is well characterised, an important population of brain-resident CD4 + T cells have been identified in the healthy mouse and human brain in the absence of any inflammation (Pasciuto, et al., 2020, Liston et al., 2022). Given their presence in the non-inflamed brain, they likely have important physiological roles. Mice lacking CD4 T cells had aberrant synaptic pruning and learning and memory deficits and these tissue-resident CD4 T cells may be crucial for proper microglial maturation (Liston, Dooley and Yshii, 2022).

Mast cells, which are bone marrow-derived innate immune cells, migrate into the brain throughout development and are present in brain tissues in the absence of inflammation and have been found to play a role in the masculinization of the brain and adult sexual behaviours (Silver, et al., 1996, Lenz, et al., 2013, Lenz, et al., 2018). Relatedly, mast cells are found in the hippocampus and mice deficient in mast cells show deficits in learning and memory (Nautiyal et al., 2012) The presence of these cells in the undisturbed brain indicates an important physiological role.

Astrocytes, though not considered classic immune cells, are key participants in the regulation immune responses in the brain (Colombo and Farina, 2016). In a similar manner to the resident and infiltrating immune cells, astrocytes possess the capacity to both sense and respond to immune signalling through immune receptors and the release of cytokines, chemokines and neuromodulators, thereby engaging in crucial cell cross-talk under physiological conditions and in disease (Matejuk and Ransohoff, 2020, Sutter and Crocker, 2022, Fisher and Liddelow, 2024). The neuromodulatory functions of astrocytes have been well defined (Fellin, 2009, Ota et al., 2013, Squadrani, et al., 2024). Calcium signalling in astrocytes in response to neuronal activity as well as the release of neurotransmitters and ATP can feedback onto and modulate neuronal activity and synaptic plasticity (Wang, et al., 2006, Fellin, 2009). Importantly, there is increasing evidence for an astrocytic role in learning and memory formation (Kol, et al., 2020, Squadrani, et al., 2024). Astrocytes are particularly responsive to increases in TNF which can result in functional modifications of hippocampal excitatory synapses and changes in memory (Habbas et al., 2015). However, even in the absence of overt neuroinflammation astrocytes have also been shown to contribute to memory. Manipulation of astrocytic activity through excitatory DREADD expression in CA1 astrocytes disturbs remote memory formation by disrupting hippocampal to cortical communication (Kol et al., 2020). Recently, learning-associated astrocyte ensembles have been defined that show elevated c-Fos levels following learning and reactivation of these astrocytic ensembles can provoked memory recall (Williamson et al., 2025).

In addition to these central and infiltrating immune cells, peripheral immune function also appears to support cognition. Severe combined immune deficient (scid) mice, which lack adaptive immunity, show impairments in learning and memory which can be alleviated through passive transfer of T-cells (Kipnis, et al., 2004, Brynskikh, et al., 2008).

Neurogenesis describes the generation of new neurons from neural stems cells and is thought to be key in supporting hippocampal memory formation (Shors, et al., 2001, Shors, et al., 2002, Josselyn and Frankland, 2012, Akers, et al., 2014). Neurogenesis occurs at a high level during development but continues at a reduced rate throughout life into adulthood in the olfactory bulb and the subventricular zone of the DG (Gage, 2019). Evidence suggests important immune regulation of neurogenesis that facilitates memory (Ziv, et al., 2006, Ziv and Schwartz, 2008). Immune deficient mice that lack of mature T-cells exhibit impaired hippocampal neurogenesis that was partly rescued upon reconstitution of immune system (Ziv et al., 2006). An impairment of neurogenesis has also been reported following under inflammatory conditions (Ekdahl et al., 2003). Microglia may be keys player in the regulation of neurogenesis as mice lacking CX3CR1 displayed altered hippocampal memory (Reshef, et al., 2014, Diaz-Aparicio, et al., 2020).