Reimagining biogenic amine signaling in the brain and beyond

Abstract

Monoaminergic neurotransmission has long been recognized as essential for the development, maintenance, and plasticity of the nervous system, with classical models defining serotonin, dopamine, and histamine as extracellular messengers acting through cell-surface receptors. Broadening this view, emerging evidence reveals that biogenic amines also covalently modify proteins—a process termed monoaminylation—to directly influence intracellular signaling. The discovery of non-canonical monoamine signaling across subcellular compartments offers new insight into brain–body communication. Here, we review the evolving signaling landscape of protein monoaminylations and highlight new chemical-biological tools for probing their impact on neural development, plasticity, and disease.

Neurotransmission-dependent and independent roles for monoamines in the brain and beyond

Evolutionarily conserved biogenic amines, such as serotonin, histamine, dopamine, norepinephrine, and epinephrine are important neuromodulators in both central and peripheral systems. They orchestrate many functions that allow organisms to adapt to their environment [1] including arousal and feeding, as well as complex cognitive processes including attention, learning, and memory [2–4]. Imbalances in monoamine levels and their signaling are linked to many neurological and neurodevelopmental disorders, such as autism spectrum disorder, attention-deficit/hyperactivity disorder, intellectual disability, and stress-related conditions including substance use disorders and depression [5–7]. Late-onset neurodegenerative disorders, including Parkinson’s and Alzheimer’s disease, are also associated with dysregulated monoaminergic signaling [1,7–10]. Indeed, several classes of drugs for monoamine-related brain disorders alter monoamine balances [11,12].

Outside of the brain, all monoamines serve as trophic factors or hormones that control basic aspects of physiology and are central to embryonic development, as well as to proper function of the immune, metabolic, cardiovascular, and gastrointestinal systems [1,11–15]. For example, serotonin has been highlighted as a potent regulator of pancreatic islet hormone secretion, glucose homeostasis, and nutrient metabolism [16,17]. Immunoregulatory functions have been ascribed to monoamines as well; histamine, for example, is an important mediator of inflammation, and serotonin modulates cytokine secretion in monocytes and macrophages [11,18–20]. Aberrant levels of monoamines or disruption of their signaling pathways are associated with many somatic disorders, including metabolic, immunological, cardiac, digestive, and respiratory diseases [1,14,21]. Most monoamines, and their multitude of actions, are ancient and have been conserved throughout evolution from unicellular organisms to invertebrates and mammals [22,23]. Intriguingly, numerous unicellular organisms (including yeast, protozoa, and bacteria), which lack a canonical nervous system, can import, synthesize, and metabolize monoamines [24–30]. Even in the absence of canonical GPCRs, in these organisms, monoamines influence pathways regulating growth, metabolism, motility, chemotaxis, and adherence [22,27,28,31–34].

In mammals, the canonical signaling pathways activated by monoamines are reasonably well understood [35–37]. Monoamines engage cognate cell surface receptors (Figure 1A); their binding leads to activation of intracellular signaling cascades that elicit diverse cellular outputs, including calcium-dependent signaling and transcriptional regulation [38]. While these pathways rely on vesicular packaging and synaptic release of monoamines, it has long been recognized that non-vesicular pools of monoamines exist in various cell types, including neurons, that can traverse subcellular compartments to reach the nucleus [39,40]. The existence of these pools suggested a non-canonical and receptor-independent role for biogenic amines. Indeed, early evidence indicated that monoamines can be covalently attached to proteins by transglutaminase enzymes [41–43], including modification of GTPases [44–46]. The physiological impacts of these posttranslational modifications (PTMs) include altering immune/inflammatory responses and glucose homeostasis [44,45,47–49].

Figure 1. Canonical and non-canonical monoaminergic signaling.

(A) Canonical monoamine neurotransmission involves the release of monoamine-filled vesicles into the synaptic space where they elicit diverse signaling responses by binding cognate receptors on the cell surface, typically GPCRs, that then drive intracellular signaling cascades to influence neuronal activity/plasticity. (B) Beyond their canonical role as extracellular neurotransmitters activating cell surface receptors, monoamines also operate as TG2-mediated post-translational modifications on proteins, enabling a non-canonical, epigenetic mode of regulation. Protein monoaminylations influence protein activity and protein-protein interactions by serving as a docking site for reader proteins. Histone proteins have recently emerged as a robust target of TG2-mediated monoaminylations, at histone H3 Gln 5, which have been shown to influence transcriptional programs related to neurodevelopment [50,90], drug-relapse-related behavior [51,54,55], circadian rhythmicity [52], and tumorigenesis [86,88]. This figure was created with BioRender.

Motivated by these findings, recent investigations have asked whether this class of PTMs might also occur on histones, the core packaging proteins of chromatin. Indeed, transglutaminase 2 (TG2)-mediated serotonylation [50], dopaminylation [51], and histaminylation [52] of histone H3 have all now been described, both within monoaminergic and non-monoaminergic brain cells. This work also demonstrates the potential relevance of these PTMs for normal neural development and plasticity, as well as for brain disorders and their treatment [51,53–55]. In this review, we thus trace the historical and functional significance of transglutaminases and protein monoaminylations, examine the emergence of histone monoaminylations in both the central nervous system and peripheral tissues, and highlight novel tools for studying these unconventional PTMs.

Early evidence for protein monoaminylations mediated by transglutaminase enzymes

In the late 1950s, it was first proposed that calcium-dependent mammalian enzymes, later identified as transglutaminases, can catalyze covalent linkages between amine groups of mono- and polyamines, and the γ-carboxamides of glutamine residues within certain protein substrates [47,56,57]. Although these initial findings strongly hinted at potential regulatory roles for aminyl PTMs within cells, the neuroscience field largely shifted its focus towards that of monoaminergic neurotransmission over the latter half of the 20^th century, phenomena that are indeed critical for neural communication and plasticity. Nonetheless, subsequent discoveries in both the brain and periphery revealed the existence of large extravesicular pools of monoamines in both nuclear and cytoplasmic compartments, suggestive of potential neurotransmission-independent functions for monoamines in cellular development, gene regulation, and oxidative stress control [22,58,59]. Most critically, perhaps, serotonin was demonstrated to be transamidated onto small GTPases by TG2 [45,49] in platelets, which was found to directly alter their intracellular signaling properties. These data provided the first clear evidence of a receptor-independent mechanism for protein monoaminylations in mammalian cells.

Transglutaminases (TGs) are a family of related enzymes, which have historically been characterized to form N^e(g-glutamyl)lysine bonds [42,56]. TGs were first thought to function primarily in generating cross-linked protein products, which would then be resistant to proteolytic degradation, thus stabilizing target proteins involved in, for example, the extracellular matrix [60]. Human TGs (hTGs) total nine members and include TGs 1–7, factor XIII, and protein band 4.2[61]. Excluding protein band 4.2 (which is believed to only be structurally related to TGs), all members of the hTG family carry out this reaction in a calcium-dependent manner. Of the nine members of the hTG family, TG2 is the most ubiquitously expressed. It is found in most tissue types and throughout most subcellular locales [62].

Mechanistically, TG2 functions similarly to cysteine proteases. A catalytic triad, consisting of a Cys-His-Asp, activates the thiol for nucleophilic attack on a glutamine donor [56]. Access to the TG2 active site is heavily dependent on the local environment [57,63–65]. Many factors, including calcium, GTP concentration [57,65], and local redox potential [64] have been discovered to induce a substantial conformational change [63,66] that affects active site accessibility and regulates TG2’s function (Figure 1B). Among these factors, calcium is thought to be the most crucial activator of the enzyme. There are six known calcium binding sites within TG2, and in vitro biochemical studies indicate that activation of the enzyme requires [Ca²⁺] in the tens of mM range [67]. TG2 activity is also thought to be regulated by binding of GTP to the closed (i.e., inactive) state (Figure 1B). Interestingly, the enzyme has been shown to act as a GTPase [57], although the significance of this function remains elusive. The activity of the enzyme is also sensitive to redox conditions. The human enzyme contains 20 cysteine residues and oxidation of key pairs of these to disulfides has been shown to inactivate the enzyme [68]. Given the reducing nature of the intracellular milieu, it is thought that such redox regulation is restricted to the extracellular pool of the enzyme [64]. Taken together, TG2 regulation is clearly multifactorial with the potential for a complex interplay and even hierarchy between all these regulatory inputs; for example, GTP inhibition can be overridden by calcium binding [57]. With respect to the intracellular transamidation activity of TG2, it seems likely that it must be tightly linked to overall cellular physiology, including fluctuations in intracellular calcium levels, which, it should be noted, increase during canonical biogenic amine signaling [37,69] and can reach concentrations in the hundreds of mM range in intracellular microdomains [56,57,64]. It should also be stressed that it is currently unknown whether there are interaction partners for TG2 that could modulate the impact of calcium or GTP on its enzymatic activation. Similarly, it is unclear whether other PTMs of the enzyme (i.e., beyond oxidation) may regulate its basal activity or its requirement for calcium.

TG2-mediated monoaminylations of histones: contributions to neural development, plasticity, and disease

Histone serotonylation

Genetic information in eukaryotes is stored in the form of chromatin, a complex of genomic DNA with histone and non-histone proteins. The structure and organization of chromatin are regulated by various mechanisms, such as chemical modifications of DNA and histones, that establish different gene expression patterns from the same genome [70–72]. Many proteins orchestrate chromatin-templated epigenetic regulation, including ‘writers’ and ‘erasers’ that control the installation and removal of DNA and histone modifications, respectively, and ‘readers’ that mediate their biological effects [71]. The interplay of these macromolecules to establish the epigenome underlies cell identity. Epigenetic dysregulation, by contrast, can lead to pathologies, such as developmental disorders and cancer [73–75].

Histones, as well as the myriad of effector proteins that engage with chromatin, are modified with numerous different PTMs (or ‘marks’) [72]. Both the sequence location and chemical nature of these marks influence the functional state of chromatin, either by directly altering structure and stability, or by recruiting reader proteins to the vicinity of a given nucleosome (the basic repeating unit of chromatin) [76]. Histones are among the most highly conserved proteins known, reflecting the importance of essentially every residue in tuning the functionality of chromatin. The unstructured N- and C-terminal tail regions of histones have an extraordinarily high density of PTMs. The N-terminal tail of histone H3 is especially rich in modifications; until recently, with the notable exception of glutamines 5 and 19 (H3Q5 and H3Q19), every residue that could conceivably be modified had been shown to be [77]. Since these highly conserved glutamines are surrounded by residues whose modifications play critical roles in gene regulation [78], it was initially speculated that they might also be modified in some way. Attention then turned to glutamine transamidation, driven in part by earlier in vitro studies showing that recombinant histones are cross-linking substrates of TG2 [79–81]. Since intracellular (i.e., non-vesicular) pools of monoamines had also previously been documented, these biogenic amines were initially investigated, beginning with serotonin. Treatment of HeLa cells with a clickable analog of serotonin, propargyl-5-HT, led to monoaminylation of histone H3 [50]. A series of biochemical studies using recombinant TG2 established that this occurs on both H3Q5 and H3Q19 [50,82]. Mass spectrometry and antibodies developed against serotonylated H3Q5 (H3Q5ser), as well as the combinatorial mark with both H3K4me3 and H3Q5ser (H3K4me3Q5ser), revealed that both marks are evolutionarily conserved and are widely distributed in animals (both vertebrates and invertebrates) and across tissue/cell types. These tools were used in conjunction with various genomics methodologies to show that the dual mark (i.e., H3K4me3Q5ser) correlates with permissive gene expression during neural differentiation [50]. Biochemical studies have additionally indicated that H3Q5ser can augment the function of H3K4me3 by enhancing its binding affinity for the TAF3 [50,83] subunit of the general transcription factor IID (TFIID), and by preventing its removal by dedicated demethylase enzymes [83]. Together, these results argue that H3Q5ser plays a critical role in the maintenance and fine-tuning of gene expression programs during neural development [50,82,83].

More recently, it was observed that H3Q5ser also plays an important role in several brain-related physiological functions, including stress regulation (Figure 2A) [53]. More specifically, the dorsal raphe nucleus (DRN), the primary hub of serotonergic projection neurons in the central nervous system, displays robust transcriptional changes following chronic social defeat stress – an etiologically relevant rodent model for the study of human depression – in both male and female mice [53]. The biological pathways influenced by chronicstress-associated gene expression displayed substantial overlap between male and female mice, and these shared pathways are significantly enriched for psychiatric and mood-related disorders, including major depressive disorder. These alterations in gene expression coincided with disruptions in H3 serotonylation dynamics in the DRN of both male and female mice. Similar results were observed in postmortem DRN tissues from individuals diagnosed with major depressive disorder, where major depressive disorder subjects displayed altered levels of H3 serotonylation compared to demographically matched controls. Interestingly, male mice deemed to be stress-resilient following chronic social defeat stress displayed significant attenuation of these H3Q5ser dynamics [53], indicating that patterns of differential H3Q5ser enrichment observed in stress-susceptible mice may contribute to maladaptive behaviors elicited by chronic stress. Animals classified as stress-susceptible, in contrast to stress-resilient counterparts, exhibited abnormal accumulation within the DRN that persisted long after stress exposure [53]. Of note, this phenotype was completely reversed by chronic fluoxetine treatments—a classical SSRI—mirroring observations in humans treated with antidepressants prior to death. Fluoxetine exposure also significantly restored both transcriptional and behavioral impairments in susceptible mice. Finally, it was shown that directly reducing levels of H3 serotonylation in DRN prior to chronic social defeat stress (using viral vector-mediated dominant negative strategies) promoted behavioral resilience and rescued stress-mediated gene expression programs, like those observed in response to chronic fluoxetine treatments. While some studies suggest that H3 monoaminylation levels are largely dictated by intracellular monoamine concentrations [52], once established in neuronal chromatin, it remains unclear how quickly these marks are turned over, especially given the relatively slow kinetics of histone turnover observed in both neurons and glia [84]. As such, SSRI treatments may function to increase serotonin release from 5-HTergic neurons, thereby reducing intracellular 5-HT concentrations in the DRN, eventually leading to loss, or restoration, of the mark within these cells. However, if histone serotonylation remains relatively stable during initial antidepressant treatment treatments (i.e., acute fluoxetine, exposures which do not rescue behavioral deficits in the chronic social defeat stress model [53]), then its accumulation may not be fully resolved by acute administrations of these drugs. If true, then chronic treatments with antidepressants may be required to facilitate the full restoration of normal H3 serotonylation levels in DRN serotonergic cells.

Figure 2. Histone monoaminylations in health and disease.

(A) In mice, serotonylation of histone H3 in the dorsal raphe nucleus is dynamically altered by stress and antidepressant exposure and causally regulates depression-linked gene expression and behavior [53]. (B) Across studies in liver, hepatocellular carcinoma, and ependymoma, H3Q5 serotonylation (H3Q5ser) was found to be elevated in malignant tissues, where it drives oncogenic transcriptional programs, enhances proliferative and stemness pathways, and promotes tumorigenesis through aberrant chromatin activation mediated by TGM2-dependent serotonylation [86–88]. (C) In mice, placental H3Q5 serotonylation has been showed to establish a monoamine-linked epigenetic axis that programs fetal neurodevelopmental gene expression, shaping early brain maturation and influencing offspring susceptibility to later-life neuropsychiatric outcomes [90]. (D) In mice and rats, dopaminylation of histone H3 in the ventral tegmental area and nucleus accumbens accumulates following chronic cocaine or opioid exposure, reprograms transcriptional networks governing reward circuitry, and drives relapse-associated behaviors [51,54,55]. (E) H3Q5 dopaminylation in the dorsal hippocampal formation (dHF) is dynamically regulated across pregnancy and the postpartum period, coupling dopaminergic signaling to transcriptional networks that support maternal neuroplasticity, motivation, and affective adaptation in mice [91]. (F) Histaminylation of histone H3 oscillates in phase with histamine signaling in the mouse tuberomammillary nucleus (TMN) to regulate circadian gene expression, demonstrating that H3Q5his functions as a chromatin-based integrator of monoaminergic rhythmicity [52]. Abbreviations: 5-HT – serotonin; DA – dopamine; DRN – dorsal raphe nucleus; HA – histamine; NAc – nucleus accumbens; VTA – ventral tegmental area; TG2 – transglutaminase 2. This figure was created with BioRender.

Research has also begun to characterize histone serotonylation in the context of other forms of brain homeostasis and tumorigenesis (Figure 2B). For example, a recent study in mice exploring histone serotonylation in the context of astrocyte-neuron communication identified an activity-dependent transporter, SLC22A3, which imports 5-HT into astrocytes and contributes to histone serotonylation in these cells [85]. Such phenomena appear to broadly regulate transcriptional programs that contribute to sensory processing in the olfactory bulb [85]. In another recent publication, researchers explored histone serotonylation dynamics using a mouse model of ependymoma, an aggressive neural-derived tumor that spawns from ependymal cells [86]. 5-HTergic neurons were shown to suppress the growth of ependymoma tumors, suggesting that 5-HTergic signaling may be involved in ependymoma tumorigenesis. To explore if histone serotonylation plays a role in this process, the authors employed a series of epigenomic methodologies to demonstrate that this PTM contributes importantly to the regulation of ependymoma gene transcriptional programs. Using viral-mediated gene therapy approaches, they found that attenuation of the mark significantly halted tumor growth and increased animal survival, suggesting that H3Q5ser dynamics play a key role in ependymoma tumorigenesis and that directly reducing levels of the mark in these tumors may serve as a novel treatment strategy. Indeed, additional groups have also reported that H3Q5 serotonylation plays dynamic roles in regulating other cancers including hepatocellular carcinoma [87] and neuroendocrine prostate cancer [88] (Figure 2B).

Finally, given that 5-HT, among other monoamines, circulates peripherally, it should come as no surprise that histone serotonylation was also recently detected in peripheral organs, including the placenta (Figure 2C), thereby hinting that histone serotonylation plays broader roles outside of the CNS. For example, in a recent pair of studies, it was observed that placental histone serotonylation is developmentally regulated and depends on the SERT/SLC6A4 transporter [89,90]. Disruption of the mark in placenta, via SERT knockout, resulted in downstream consequences on progeny by disrupting gene expression programs involved in synaptic signaling, neuronal proliferation, and other key neurodevelopmental pathways [90]. These data further highlight how neurotransmission-independent mechanisms of 5-HT action, particularly at the maternal-fetal interface, may influence neurodevelopment.

Histone dopaminylation and histaminylation

The work on histone serotonylation prompted researchers to ask whether other biogenic monoamines can be enzymatically attached to histone glutamines. This led to the discovery of H3Q5 dopaminylation (H3Q5dop) in the ventral tegmental area (VTA) of the brain. It should be noted that H3Q5dop, like H3Q5ser, is also widely distributed in the brain within regions receiving relevant inputs – in this case, dopaminergic inputs including the nucleus accumbens (NAc) and medial prefrontal cortex (mPFC). Interestingly, it was found that H3Q5dop levels accumulate in the VTA during prolonged abstinence from addictive substances (Figure 2D), such as cocaine and heroin, and that this promotes persistent transcriptional programs that contribute to relapse vulnerability [51,54,55]. In rats exposed to prolonged abstinence from cocaine self-administration, it was additionally found that H3Q5dop similarly displays accumulation in the NAc, but not the mPFC, and that such accumulation is critical for establishing drug-induced gene expression programs that contribute to drug craving [55].

Besides its maladaptive roles in drug craving, H3Q5dop was also recently identified as a core epigenetic mechanism mediating reproductive experience-induced neural plasticity in the maternal brain [91]. In this preprint study from our group, it was found that H3Q5dop in the dorsal hippocampal formation (dHF) of both mice and humans persistently reshapes transcriptional profiles in a dopamine-dependent manner in response to pregnancy and postpartum experiences to direct pro-adaptive cellular and behavioral plasticity (Figure 2E). In contrast, following chronic periods of postpartum stress, these pro-adaptive alterations are attenuated, leading in mice to disruptions in cognition and offspring care. Surprisingly, chemogenetic manipulations of dopamine release from the VTA to the dHF of virgin female mice was shown to be sufficient to mimic parity-related effects, demonstrating that dopamine release and the downstream deposition of the H3Q5dop mark are sufficient to recapitulate epigenomic and behavioral features of reproductive experience.

More recently, our group also documented the existence of H3Q5 histaminylation (H3Q5his), again installed by TG2 [52]. Antibodies raised against H3Q5his revealed enrichment of the mark in the region of the brain that synthesizes histamine (the tuberomammillary nucleus or TMN). Studies in mice revealed that levels of H3Q5his in the TMN are dynamic as a function of diurnal sleep-wake cycles (Figure 2F). While the origins of this in vivo rhythmicity remain to be fully elucidated, in vitro analyses indicated that TG2 can convert H3Q5his into H3Q5ser or H3Q5dop, and it is conceivable that this “exchange” capability plays a role in H3Q5his dynamics [52]. Interestingly, H3Q5his rhythmicity opposes H3K4 methylation levels (i.e., they are out of phase), which might relate to the in vitro findings that H3Q5his––unlike H3Q5ser––inhibits the activity of MLL and SETD1 methyltransferase complexes, which install H3K4 methyl marks, by antagonizing WDR5 (a component of all MLL1–4 and SETD1A/B methyltransferase complexes) binding interactions with histone H3 [52]. Importantly, disruption of H3 monoaminylation dynamics in the TMN disrupted both circadian gene expression and rhythmic behavior, indicating essential roles for these marks in guiding normal patterns of physiological plasticity in the brain.

Towards novel tool development to ascertain the full scope of protein monoaminylations in the brain and beyond

As discussed in previous sections, the discovery of protein monoaminylations has led to an expanding understanding of monoamine signaling across multiple cell types and subcellular compartments. However, for the most part, these studies employed candidate-based approaches to identify and characterize monoaminyl PTMs. More recently, in aiming to capture the full scope of protein monoaminylomes in the brain and beyond, efforts have been directed towards the development of novel chemical biological tools to better address the diversity of these modifications using in cellulo and in vivo systems. While the initial discoveries of protein monoaminylations relied heavily on radioisotope labeling of monoamines, monoamine analogs (e.g., propargyl-5-HT, monodansylcadaverine, biotin cadaverine), and/or site-specific antibodies to detect monoaminylated substrates in cells and/or tissues, such approaches often suffer from experimental limitations. For example, in cases where monoaminylations are induced by exogenous application of monoamines or their analogs in cellular/tissue systems, it remains difficult to ascertain the in vivo relevance of identified PTMs, as they may not exist endogenously without exogenous stimulations. Also, while antibodies remain a powerful tool, in situations where antibodies are used selectively to assess proteins that contain the respective PTMs, researchers are often prohibited from gaining a more comprehensive understanding of the full repertoire of endogenously monoaminylated substrates in their systems of interest. To address these limitations, novel tools have begun to emerge that more faithfully capture the dynamic range of protein monoaminylations in vivo and allow exploring how intracellular monoamine dynamics may contribute to these PTMs as a consequence of environmentally-induced cellular plasticity.

Chemical probes to detect monoaminylated proteins in cells and tissues

Traditional methods used to detect modified proteins include immuno- and mass spectrometry-based methodologies. Antibodies, however, are generally restricted to specific modification sites that have already been identified, and targeted mass spectrometry––while remaining the gold standard for detection of endogenously modified proteins in cells/tissues––can be challenging for a variety of technical reasons, including detection bias of high vs. low stoichiometry PTMs, peptide fragmentation issues, the unknown co-occurrence of PTMs on a given peptide. As such, in recent years, several studies have established novel chemical probes that take advantage of unique chemistries of monoamines to capture proteins that have been covalently modified by serotonin or dopamine.

Two such studies recently developed dopamine probes that allowed researchers to observe that protein dopaminylation can also occur on cysteine residues [92,93], albeit through a non-enzymatic mechanism that remains of unknown biological significance (Figure 3A). These first require oxidation of dopamine, or an alkynylated dopamine analog, into one of many reactive intermediates, followed by a nucleophilic attack of a cysteine thiol. Once incorporated, the alkyne handle can be used in click chemistry reactions to conjugate either fluorophores to dopaminylated cysteines for visualization in cells, or the incorporation of biotin handles for substrate protein pulldowns and identification via mass spectrometry. These approaches, however, rely on reactive intermediates interacting with reduced thiols, which can lead to many false positives, and they exclusively capture cysteine dopaminylations, which may or may not exist endogenously (i.e., some non-enzymatic PTMs are likely to occur non-specifically in vitro during sample preparation). Nevertheless, one of these probes was recently used to identify two candidate cysteines (Cys280 and Cys311) within the human tau protein that may be dopaminylated and might be predicted––owing to their location within the protein––to aid in aggregation prevention and progressive neurodegeneration, thus highlighting a potential neuroprotective role for cysteine dopaminylation in tauopathies [92]. In addition, dopaminylation probes have been developed that specifically detect glutamine-linked protein dopaminylation events in cells/tissues [94,95]. These probes are more useful for exploring monoaminylation events that are enzymatically catalyzed by TG2 (Figure 3B). For example, these probes exploit the chemical properties of 1,2 catechols (e.g., dopamine, norepinephrine), whereby they can undergo a strain-promoted cycloaddition reaction with cyclooctynes following oxidation to orthoquinones. As a result, upon treatments of cellular/tissue protein lysates with the oxidizing agent periodate (and treatments with iodoacetamide to cap reactive cysteines), dopaminylated proteins can be selectively labeled with a biotinylated cyclooctyne probe. It should be noted that while theoretically noradrenylated substrates should also be pulled down using this approach, in a recent preprint study, the methodology has been optimized to solely capture dopaminylated substrates [94]. Following substrate labeling, biotin-tagged dopaminylated proteins can then be efficiently pulled down and identified using mass spectrometry. Using this approach, it was found that >1,000 proteins across multiple brain regions in mouse––many of them synaptic (both pre- and post-), as well as glial-enriched––are putative substrates of dopaminylation, suggesting that such phenomena may impose downstream effects on intracellular protein signaling within the CNS [94]. The precise transport mechanisms that allow for post-synaptic uptake of dopamine to establish these PTMs in brain (both in neurons and glia) remain elusive, especially given that many of these cells do not express the canonical transporter for dopamine. With that, it is likely that volume transmission through other high capacity, low affinity transporters present in these cells, such as SLC22A3 (aka organic cation transporter 3/OCT-3) or SLC29A4 (aka plasma monoamine transporter/PMAT), may facilitate such uptake to allow for mark deposition [85].

Figure 3. Novel tools to study monoamine and protein monoaminylation dynamics in subcellular compartments.

(A) To capture the full scope of protein monoaminylations in cells/tissues, several chemical probes have emerged. These approaches rely on dopamine oxidation to reactive intermediates that react with cysteine thiols, allowing visualization or pulldown of labeled proteins, but they carry risks of false positives and exclusively detect cysteine dopaminylation. Such probes have revealed that dopamine can modify cysteine residues through a non-enzymatic mechanism, although the physiological significance remains unclear [92,93]. (B) More recent probes selectively detect glutamine-linked dopaminylation, which is enzymatically catalyzed by TG2 and is likely to mediate specific signaling functions [94]. (C) Two complementary chemical probes were developed to selectively label serotonin-modified proteins via their 5-hydroxyindole moiety. AzTAD is activated via UV-light to generate a reactive triazolinedione for covalent tagging of serotonylated glutamines (e.g. H3Q5ser), whereas AlkTAD is activated via diazotization chemistry to form the reactive species under mild conditions [96]. Each probe carries a conjugation handle (azide or alkyne) for downstream attachment of fluorophores or biotin, enabling imaging, enrichment, and proteomic analysis. Abbreviations: 5-HT – serotonin; DA – dopamine. This figure was created with BioRender.

Additional methodological advances that have been reported recently include probes to interrogate monoaminylated histones and non-histone proteins for use in tracking [95], enrichment and/or imaging of histone monoaminylations in cells [96] (Figure 3C). Using such approaches, it was recently found that H3Q5 monoaminylations appear to be enriched in human breast and colon cancers [96], likely contributing to crosstalk with H3K4me3––versus other cancer cell lines that appear to lack such enrichment–– although future studies are needed to fully elucidate the precise contributions of histone monoaminylations to neoplasia. Collectively, such labeling approaches have proven useful in illuminating the unique diversity and dynamics of protein monoaminylations in cells/tissues, as well as the pathways in which they operate.

Concluding remarks and future perspectives

Monoamine neurotransmitters are critical regulators of synaptic transmission, neurodevelopment, and brain plasticity. Canonical receptor-mediated mechanisms of monoamine neurotransmission are well established; however, growing evidence suggests that monoamines also act non-canonically as protein modifiers, influencing protein activity, subcellular localizations, protein-protein and protein-DNA interactions. This expanded view also highlights how certain small molecule therapeutics may contribute to signaling dynamics and gene regulation within the brain. For example, SSRIs—long considered to be the standard treatment for mood disorders, such as major depressive disorder —display only partial efficacy, implying that roles for serotonin beyond that of canonical neurotransmission may contribute to disease etiology and modulate therapeutic efficacy. Supporting this, our lab demonstrated that in mice, the dual H3K4me3Q5ser modification in the DRN [53], disrupted by chronic stress, maladaptively regulates transcriptional programs underlying depression-related phenotypes. These effects can be reversed by chronic, but not acute, SSRI treatments, paralleling human observations. Thus, maladaptive TG2-mediated histone serotonylation may promote chromatin states that increase depression vulnerability. Future therapies aimed at directly reversing H3 monoaminylation dynamics, such as targeting TG2 or H3Q5ser readers, may then prove more effective than current SSRIs, which primarily alter serotonin levels without directly impacting the marks.

Although TG2 is the primary H3Q5 monoaminylase in human cells, how it regulates monoaminylation remains largely unclear. TG2 lacks DNA-binding or reader domains, suggesting that it requires accessory proteins to localize to chromatin (Figure 4A). While TG2’s many cellular roles are well documented, its binding partners remain poorly defined. Cofactors such as calcium and GDP can shift TG2 conformations, restricting activity, but the extent to which proteins or intrinsic PTMs modulate TG2’s localization, catalysis, or substrate preference is unknown. Moreover, the biological impact of TG2’s ability to exchange distinct post-translational modifications (PTMs) on the same substrate remains to be fully resolved. Such PTM ‘rewriting’ by a single enzyme is highly unusual, as the cycling of histone modifications typically require multi-subunit complexes that coordinate the ‘writing’ and ‘erasing’ of histone marks. From an epigenetic perspective, such enzymatic versatility may offer new insights into the histone code hypothesis, which posits that specific PTM combinations govern chromatin dynamics and gene-expression patterns in cells. One could imagine that in addition to serving distinct epigenetic roles on chromatin, H3Q5 monoaminylations may function as an epigenetic regulatory hub, with the exchange of these PTMs allowing for rapid responses to local monoamine flux (e.g., through alterations in synaptic release dynamics) induced by external stimuli. Furthermore, although TG2 is the best-studied transglutaminase in mammals, whether the other human transglutaminases also act as monoaminylases on histones or other substrate proteins remains to be explored.

Figure 4. Lingering questions with respect to TG2-mediated protein monoaminylations in health and disease.

(A) Although TG2 is widely recognized as the primary H3Q5 monoaminylase in human cells, the mechanisms governing its regulation remain unclear. Despite extensive studies on TG2’s cellular roles, its binding partners and pathways linking cytoplasmic activity to nuclear functions remain poorly defined. Co-factor binding (e.g., Ca²⁺, GDP) can induce open versus closed conformations and partially regulate catalytic activity, but the influence of additional proteins or intrinsic PTMs on TG2’s localization, substrate selectivity, and monoamine preference remains unclear. These gaps highlight the need for further studies to define how TG2 activity is tuned to regulate chromatin-based processes. (B) In vivo studies have established serotonin, dopamine, and histamine as bona fide donors for protein monoaminylations. It is plausible, however, that other amine-containing molecules – including trace amines (e.g., tyramine, tryptamine, octopamine), polyamines (e.g., spermidine, spermine, putrescine), and even non-endogenous compounds (e.g., amphetamine, mescaline) – may also serve as TG2-dependent donors. If confirmed, such modifications could represent an expanded regulatory mechanism influencing intracellular signaling, chromatin organization, and neural or peripheral plasticity. (C) Therapies such as selective serotonin reuptake inhibitors (SSRIs) that are used to address apparent monoamine dysregulation in neuropsychiatric disorders (including depression and schizophrenia) show, at best, mixed efficacy. This begs the question as to whether protein monoaminylations, rather than canonical monoaminergic signaling (or in addition to canonical signaling) plays a role in the etiology of these conditions. Abbreviations: 5-HT – serotonin; DA – dopamine; HA – histamine. This figure was created with BioRender.

Another question to be addressed in future work is how monoaminylation levels and stoichiometries are regulated to drive phenotypic outcomes. Differential monoamine biosynthesis, release, or transporter expression across brain regions—or in response to stimuli—may shape protein monoaminylations via altered donor availability. In addition, while serotonin, dopamine, and histamine have been validated as protein monoaminylation donors in vivo, other amines—including trace amines (tyramine, tryptamine, octopamine), polyamines (spermidine, spermine, putrescine), or even nonendogenous amines (amphetamine, mescaline)—may also act as TG2-dependent donors (Figure 4B). If confirmed, such modifications could substantially expand the known scope of mechanisms supporting intracellular signaling, chromatin regulation, and neural or peripheral plasticity.

In summary, much remains to be discovered regarding how monoamine-derived PTMs are established, regulated, and integrated into physiology and disease (Figure 4C). It is increasingly clear that canonical views of monoaminergic signaling—long centered on neurotransmission—must be revised to incorporate protein monoaminylations as a potentially equally important consequence of biogenic amine signaling. This perspective does not in any way diminish the importance of the many critical receptor-mediated pathways that have been uncovered over the past century. Rather, we would argue, it highlights the need to frame protein monoaminylation phenomena as “another side of the same coin,” perhaps even reflecting the evolutionary origins of these critical signaling molecules.

OUTSTANDING QUESTIONS.

How are TG2-mediated protein monoaminylations regulated in cells?

How does TG2 discriminate between monoamines in cells?

How is nuclear TG2 activity regulated?

What is the stoichiometry of histone monoaminylations relative to other histone PTMs?

How do transglutaminases import into the nucleus, and how do they localize to chromatin?

Are there other transglutaminases beyond TG2 that can transamidate histones?

Can additional monoamines (e.g., norepinephrine, tryptamine, octopamine, tyramine), polyamines, and/or non-endogenous free-amine containing molecules serve as donors for protein monoaminylation?

Highlights.

• Emerging evidence reframes biogenic amines, traditionally known for their roles in neurotransmission, as chemical donors for key protein posttranslational modifications that shape cellular signaling and transcription.

• Histone monoaminylations, such as H3 Gln5 serotonylation, dopaminylation, and histaminylation, represent a novel epigenetic class influencing molecular and behavioral phenotypes.

• Transglutaminase-2 installs, removes, and exchanges these modifications, thereby regulating neural transcription, development, and plasticity.

• Advances in chemical biology are expanding the repertoire of monoaminylated proteins in the brain, revealing their critical roles in modulating neural function and plasticity.