Can You Feel the Love Tonight: A Dipeptidergic Circuit for Pleasant Touch

Soft and gentle touches such as the maternal caressing of infants foster a sense of well-being, promote social bonding, and have long-lasting impacts on the brain and behaviors. Dr. Xiang Yu’s group recently identified a dipeptidergic excitatory circuit, from tachykinin 1 (Tac1)-expressing neurons in the ventrolateral periaqueductal gray (l/vlPAG) to oxytocin neurons in the paraventricular hypothalamus (PVH), that mediates the enduring effects of gentle touch on social behaviors. Their work entitled “Social touch-like tactile stimulation activates a tachykinin 1-oxytocin pathway to promote social interactions” is published in Neuron [1].

In this study, the authors first designed a novel tactile stimulation paradigm, which they termed social-touch-like (ST), by gently applying caress-like strokes at 3 cm/s with cotton wool to the back of juvenile mice several times (5 min each) a day for several days, starting from postnatal day 15. Such low-speed strokes, mimicking maternal licking and grooming, are thought to activate low-threshold unmyelinated peripheral afferent fibers knowns as C-fibers to transmit positive affect [2]. Previously, Dr. Yu’s group has found that early sensory experience bi-directionally regulates oxytocin neuronal activity [3]. Consistent with these earlier findings, they found that ST stimulation led to upregulated oxytocin mRNA and protein levels and increased oxytocin neuron activity, as measured by the expression of the immediate-early gene Egr1 and by patch-clamp recordings from oxytocin neurons in acute brain slices. Impressively, this increased oxytocin neuronal excitability was still detectable weeks after the initial treatment, indicating long-term potentiation of oxytocin neurons (Fig. 1A). Behaviorally, ST-treated juvenile mice, when they grew up and were tested as adults, showed a clear preference for the cotton wool used in ST stimulation, compared to home-cage corn bedding, and spent more time interacting with littermates in agreement with the pro-social effects of oxytocin neurons [4] (Fig. 1A).

Fig. 1.

Activation of a dipeptidergic pathway from l/vlPAG tac1⁺ neurons to PVH oxytocin neurons. A Both ST and chemogenetic activation of PVH oxytocin neurons by hm3Dq lead to better social interaction and conditioned place preference. ST induces higher firing of l/vlPAG tac1⁺ neurons and PVH oxytocin neurons, which depends on substance P-TACR1 signaling. B l/vlPAG tac1⁺ neurons release glutamate and substance P in the PVH to activate oxytocin neurons. However, only the substance P-TACR1 signaling contributes to the reduction of transient K⁺ current. I_A channels: transient, A-type, K⁺ current channels.

Crucially, stroking at a faster speed of 10 cm/s, which likely evoked fewer C-fibers and were less pleasant, did not upregulate the spontaneous activity of oxytocin neurons. Furthermore, the behavioral and electrophysiological effects of ST stimulation were abolished in oxytocin-knockout animals as ST treatment neither promoted oxytocin neuronal activity nor conditioned a preference for cotton wool or promoted social interactions in oxytocin-knockout mice (Oxt^−/−) while it remained effective in control oxytocin-heterozygous mice (Oxt^+/−). These results indicate that oxytocin expression is required for ST stimulation to impact neuronal activity and behavior. On the other hand, chemogenetic activation of oxytocin neurons with intraperitoneal injection of clozapine-N-oxide once per day during postnatal days 15 to 21 in a cotton wool context, was sufficient to condition a preference for cotton wool and promoted social interactions, similar to ST stimulation (Fig. 1A). Together, these results firmly establish potentiated oxytocin neuronal activity as the likely mediator underlying the behavioral effects of gentle social touch.

Next, leveraging their previous findings that developmental sensory deprivation reduces oxytocin neuronal activity, the authors sensitized the effects of ST stimulation on oxytocin neurons by treating juvenile mice whose whiskers had been trimmed during postnatal days 12 to 18. Compared to the ~54% increase in the whisker-intact group, ST stimulation induced a ~138% increase in oxytocin neuronal activity in whisker-deprived mice. As an essential control, ST treatment in whisker-deprived juvenile mice had no effect on the firing rate of PVH vasopressin neurons, demonstrating that the ST potentiation of oxytocin neuronal activity is specific. Using this sensitized model, the authors performed more in-depth electrophysiological characterizations. They found that ST stimulation upregulated excitatory but not inhibitory synaptic inputs onto oxytocin neurons and increased the intrinsic excitability of oxytocin neurons by reducing the K⁺ channel current density of these neurons.

Where then do these potentiated excitatory inputs originate? To answer this question, the authors analyzed brain regions activated by ST stimulation by performing whole-brain immunostaining of the immediate-early gene Egr1 one day after the tactile stimulation. This experiment was complemented by pseudorabies virus-mediated retrograde tracing of upstream inputs to PVH oxytocin neurons. This approach revealed the l/vlPAG as densely labeled by both the retrograde pseudorabies virus and Egr1 signals. Moreover, the l/vlPAG is ideally situated anatomically within a neural network that transmits peripheral sensory information [5], suggesting that it might be the sought-after node that relays the tactile information of ST stimulation to promote PVH oxytocin neuronal activity.

Applying channelrhodopsin-2-assisted circuit mapping [6], the authors functionally characterized the synaptic connection between l/vlPAG neurons and PVH oxytocin neurons. They found that a specific subset of l/vlPAG neurons that express Tac1 form monosynaptic excitatory connections onto ~47% of PVH oxytocin neurons. In addition, Tac1⁺ neurons synthesize and secrete both glutamate and neuropeptides, including substance P, which acts on tachykinin receptor 1 (TACR1). Intriguingly, optogenetic stimulation of l/vlPAG Tac1⁺ axon terminals in the PVH increased the firing rates of oxytocin neurons, recapitulating the effect of ST treatment. Moreover, this potentiating effect was only abolished by simultaneous blockade of both glutamatergic receptor and TACR1 signaling, indicating that both substance P and glutamate released by l/vlPAG Tac1⁺ neurons play a role in the induction of oxytocin neuronal plasticity (Fig. 1B). Mimicking another electrophysiological effect of ST, optogenetic activation of l/vlPAG Tac1⁺ terminals also significantly reduced the transient K⁺ current density of PVH oxytocin neurons, which relied on the release of substance P and TACR1 signaling (Fig. 1B). On the other hand, bath application of a TACR1 agonist, a substance P analog, was sufficient to increase the spontaneous firing of PVH oxytocin neurons in brain slices. Together, these results strongly suggest that ST stimulation likely promotes long-lasting changes in oxytocin neuronal activity through activating l/vlPAG Tac1⁺ neurons and the release substance P.

To test this directly in vivo, the authors carried out patch-clamp recordings and found that ST stimulation in juvenile mice produced a persistent increase in the spontaneous firing rates of l/vlPAG Tac1⁺ neurons in adults, similar to the effect seen in PVH oxytocin neurons (Fig. 1A). Moreover, ablation of l/vlPAG Tac1⁺ neurons in juvenile mice reduced the basal PVH oxytocin neuronal activity, while chemogenetic inhibition of l/vlPAG Tac1⁺ neurons during ST treatment in whisker-deprived juvenile mice attenuated the PVH oxytocin neuronal activity recorded afterward. These results demonstrate that l/vlPAG Tac1⁺ neuronal activity is required for ST-induced oxytocin neuron plasticity. Furthermore, intraperitoneal injection of a TACR1 antagonist 30 min prior to ST stimulation also prevented the increase in oxytocin neuronal activity in whisker-deprived juvenile mice, consistent with the proposed role of substance P release on oxytocin neuronal plasticity (Fig. 1A). Finally, acute chemogenetic activation of PVH-projecting l/vlPAG Tac1⁺ neurons in adult mice was sufficient to promote social interaction and condition a preference for the context of ST, fully recapitulating the behavioral effects of ST stimulation. Taken together, these results compellingly support the conclusion that ST treatment activates l/vlPAG Tac1⁺ neurons to promote oxytocin neuronal activity and behavioral changes. Thus, a dipeptidergic circuit from l/vlPAG Tac1⁺ neurons to PVH oxytocin neurons mediates the effects of ST stimulation.

Along with identifying a novel circuit for social touch, Yu et al. brought forward an exciting theme of neuropeptide-dependent neuronal plasticity that awaits further exploration. For example, ST treatment in juvenile mice led to increased l/vlPAG Tac1⁺ and PVH oxytocin neuronal activity that persisted into adulthood. What might be the detailed cellular mechanisms that translate a brief sensory experience into long-lasting changes in neuronal excitability? The authors showed that chemogenetic activation of oxytocin neurons in juvenile mice alone was sufficient to recapitulate the behavioral effects of ST treatment. It is still unclear whether such manipulation leads to stable increases in l/vlPAG Tac1⁺ and PVH oxytocin neuronal activity, as in ST stimulation. Furthermore, as bath application of a TACR1 agonist, a substance P analog, was sufficient to elicit oxytocin neuron plasticity in brain slices, it will be interesting to test whether chemogenetic activation of l/vlPAG Tac1⁺ neurons, which presumably release substance P, works just as well as activation of oxytocin neurons to generate the pro-social effects of ST. Given that ST reduced transient K⁺ conductance and bath application of a TACR1 antagonist blocked the inhibition of K⁺ current induced by light activation of l/vlPAG Tac1⁺ terminals, it is also interesting to test whether the inhibition of K⁺ current by substance P is long-lasting, and thus may contribute to the memory of sensation including chronic pain [7]. More broadly, it is essential to investigate whether there is a specific developmental window during which sensory experiences, including touch and other modalities, shape the connectivity and excitability of this dipeptidergic circuit. Alternatively, sensory-driven plasticity may still be present in adults, even those sensory-deprived as juveniles.

Notably, abnormal sensory processing and aversion to gentle touch have been reported for patients with autism spectrum disorder (ASD), a disease in which the core symptoms are social interaction deficits [8, 9]. Given that oxytocin has been proposed as a potential treatment for ASD patients [10], figuring out the ins and outs of this l/vlPAG Tac1⁺ neurons–to–PVH oxytocin neurons circuit newly identified by Yu et al. may thus generate novel treatment options for treating ASD patients. In summary, Dr. Xiang Yu’s group identified a dipeptidergic excitatory circuit that mediates the enduring effects of gentle touch on social behaviors. Their captivating work paves the way for future investigations that could lead to new therapeutic developments for alleviating social behavior deficits in neuropsychiatric patients.