Activation of Gαq Signaling Enhances Memory Consolidation and Slows Cognitive Decline

Abstract

Perhaps the most devastating decline with age is the loss of memory. Therefore, identifying mechanisms to restore memory function with age is critical. Using C. elegans associative learning and memory assays, we identified a gain-of-function G_αq signaling pathway mutant that forms a long-term (cAMP response element binding protein [CREB]-dependent) memory following one conditioned stimulus-unconditioned stimulus (CS-US) pairing, which usually requires seven CS-US pairings. Increased CREB activity in AIM interneurons reduces the threshold for memory consolidation through transcription of a set of previously identified "long-term memory" genes. Enhanced G_αq signaling in the AWC sensory neuron is both necessary and sufficient for improved memory and increased AIM CREB activity, and activation of G_αq specifically in aged animals rescues the ability to form memory. Activation of G_αq in AWC sensory neurons non-cell autonomously induces consolidation after one CS-US pairing, enabling both cognitive function maintenance with age and restoration of memory function in animals with impaired memory performance without decreased longevity.

Figure 1. Activation of G_αq enhances consolidation and prevents age-related decline in memory performance via increased CREB activity

A) Diagram of G_αq pathway. Animals bearing hypomorphic mutations in components of this signaling pathway are defective for normal associative learning and memory (indicated by *). * = Defect in learning after 1 CS-US pairing, * = Defect in short-term memory, * = Defect in learning after 7 CS-US pairings. For performance indices, see Figure S1. B) Day 1 adult animals with a gain-of-function (js126) mutation in egl-30 exhibit extended memory (~24–30 hrs) following 1 CS-US pairing when compared to the memory of wild-type animals (~2hrs). Mean ± SEM. n ≥ 9–12 per timepoint. #p<0.0001 when compared to wild-type animals at same timepoint (data not shown). C) egl-30(js126) animals lacking functional CREB [egl-30(js126);crh-1(tz2)] do not display extended memory, indicating the extended memory is LTAM (CREB-dependent). Mean ± SEM. n ≥ 24–26 per genotype. ****p<0.0001. D) CREB activity is elevated in AIM/SIA of naive egl-30(js126);pCRE::GFP animals, while GFP is undetectable in naïve wild-type pCRE::GFP animals. CREB activity is increased immediately following LTAM (7x) training in wild-type animals, while there is no observable increase in CREB activity following training (1x) in egl-30(js126);pCRE::GFP animals. E) Quantification of naïve wild-type pCRE::GFP vs. naïve egl-30(js126);pCRE::GFP. Mean ± SEM. n > 30 animals per genotype. ****p<0.0001. F) egl-30(js126) animals maintain the ability to form a long-term memory after 1 CS-US pairing at Day 5 of adulthood. Mean ± SEM. n ≥ 10 per genotype. ****p<0.0001. G) Naïve (untrained) egl-30(js126) animals maintain elevated CREB activity in the AIM at Day 5 of adulthood, while wild-type worms lack AIM GFP. H) Quantification of naïve wild-type pCRE::GFP vs naïve egl-30(js126);pCRE::GFP on Day 5. Mean ± SEM. n > 26 animals per genotype. ****p<0.0001. See also Figure S1 and S2.

Figure 2. Increased CREB activity in the AIM neurons decreases the threshold for consolidation in animals with enhanced G_αq signaling by increasing transcription of “memory genes”

A) Rescue of CREB in the SIA fails to restore memory formation to egl-30(js126);crh-1(tz2) animals, while (B) CREB rescue in the AIM restores memory performance to a level comparable to egl-30(js126) animals. Mean ± SEM. n ≥ 11–12 per genotype. ****p<0.0001, n.s.= p>0.05. C) Expression of previously identified “CREB/LTAM dependent genes” (Lakhina et al., 2015), which are upregulated in wild-type animals following LTAM training in a CREB-dependent manner (middle) but remain unchanged following LTAM training in crh-1(tz2) mutants (right), are elevated in naïve egl-30(js126) animals (left). Expression of these genes in naïve egl-30(js126) animals correlates with their expression LTAM-trained wild-type animals (Pearson correlation = 0.56), and is anti-correlated with crh-1(tz2) mutants after LTAM training (−0.21). Individual columns represent the expression of “CREB/LTAM genes” for a single microarray. D) Quantification of egl-30(js126);pCRE::GFP animals up to 6 hours after training shows that CREB activity does not significantly increase after conditioning relative to naïve egl-30(js126);pCRE::GFP animals. Mean ± SEM. n ≥ 25–40 per timepoint. n.s = p>0.05. E) Inhibiting transcription by administration of actinomycin D during the pre-conditioning starve and conditioning has no detectable effect on long-term memory in egl-30(js126) animals. Mean ± SEM. n ≥ 10 per treatment. n.s.= p>0.05. F) Inhibiting translation by treatment with cycloheximide during conditioning and 1 hr post-conditioning abolishes long-term memory without affecting learning (See Figure S2F) in egl-30(js126) animals. Mean ± SEM. n ≥ 15 per treatment. ****p<0.0001. See also Figure S2.

Figure 3. Activation of G_αq solely in the AWC is necessary and sufficient for long-term memory formation after 1 CS-US pairing

A) Expression of a gain-of-function EGL-30 (Podr-1::egl-30(Q205L)) in the AWC results in extended (24hr) memory following 1 CS-US pairing. Mean ± SEM. n ≥ 9 per genotype. ****p<0.0001. B) Extended memory of Podr-1::egl-30(Q205L) requires downstream G_αq signaling components, as egl-8(n488);Podr-1::egl-30(Q205L) animals do not display extended memory. Mean ± SEM. n ≥ 9–10 per genotype. ****p<0.0001. C–E) Extended memory in Podr-1::egl-30(Q205L) is CREB-dependent long-term memory. Podr-1::egl-30(Q205L) animals do not display extended memory without functional CREB (C, [crh-1(tz2);Podr-1::egl-30(Q205L)]), when AIM function is disrupted by an mbr-1 mutation (D, [mbr-1(qa5901);Podr-1::egl-30(Q205L)]), or when AIA/AIY function is disrupted by a ttx-3 mutation (E, [ttx-3(ot22);Podr-1::egl-30(Q205L)]). Mean ± SEM. n ≥ 9–15 per genotype. ****p<0.0001. F) Repressing EGL-30 activity in AWC of egl-30(js126) animals by AWC-specific expression of a gain-of-function inhibitor of G_αq signaling (egl-30(js126);Podr-1::goa-1(Q205L)) blocks the ability to form long-term memory. Mean ± SEM. n ≥ 9 per genotype. ****p<0.0001. See also Figure S3 and S4.

Figure 4. Enhanced G_αq signaling in the AWC cell non-autonomously regulates AIM CREB activity via neuropeptidergic signaling

A) CREB activity is elevated in the AIM of Day 1 adult naïve (untrained) Podr-1::egl-30(V180M);pCRE::GFP animals, which express an AWC-specific V180M egl-30 gain-of-function allele of egl-30 (allele present egl-30(js126)). B) Quantification of Day 1 naïve Podr-1::egl-30(V180M);pCRE::GFP vs naïve wild-type pCRE::GFP. Mean ± SEM. n > 30 animals per genotype. ****p<0.0001. C) Knockdown of neuropeptidergic signaling (egl-30(js126);Podr-3::unc-31 RNAi)) from the AWC in egl-30(js126) animals abolishes the ability to form long-term memory after one CS-US pairing. Mean ± SEM. n ≥ 12 per genotype. ****p<0.0001. D) Neuropeptidergic signaling is necessary for normal long-term memory formation: knockdown of neuropeptidergic signaling (Podr-3::unc-31 RNAi) in the AWC of wild-type animals renders them unable to form long-term memory after 7 CS-US pairings. Mean ± SEM. n ≥ 9 per genotype. ****p<0.0001. E) Increasing neuropeptide secretion from the AWC via expression of a constitutively active PKC-1 (Podr-3::pkc-1(A160E)) results in extended memory after 1 CS-US pairing. Mean ± SEM. n ≥ 10 per genotype. ****p<0.0001. See also Figure S4.

Figure 5. Activation of G_αq signaling solely in the AWC slows cognitive decline and restores cognitive function to aged animals

A) Podr-1::egl-30(Q205L) animals maintain the ability to form long-term memory after one CS-US pairing on Day 5 of adulthood. Mean ± SEM. n ≥ 13–14 per genotype. ****p<0.0001. B) Naïve (untrained) Podr-1::egl-30(V180M);pCRE::GFP animals maintain elevated CREB activity in the AIM on Day 5 of adulthood. C) Quantification of Day 5 naïve Podr-1::egl-30(V180M);pCRE::GFP vs naïve wild-type pCRE::GFP. Mean ± SEM. n > 40 animals per genotype. ****p<0.0001. D) 1 hour of heat shock (HS) at 34°C induces expression of a GFP-tagged gain-of-function EGL-30 in HS Inducible Podr-1::egl-30(Q205L) animals. E) Heat shock of HS Inducible Podr-1::egl-30(Q205L) animals at the L1 larval stage enables them to form long-term memories after one CS-US pairing at Day 1 of adulthood, while non-transgenic siblings are unaffected. Mean ± SEM. n ≥ 3 per genotype. ****p<0.0001. F) Schematic of induction of AWC-specific gain-of-function EGL-30 in aged animals. One hour of heat shock at 34°C on Day 4 of adulthood is sufficient to induce transgene expression that is detectable in Day 5 animals (1 hr HS). G) Animals without transgene induction (No HS) fail to form a long-term memory at Day 5 of adulthood. Mean ± SEM. n ≥ 4 per genotype. n.s = p>0.05. H) Induction of AWC-specific gain-of-function EGL-30 in aged (Day 4) animals enables long-term memory formation after one CS-US pairing on Day 5 of adulthood. Mean ± SEM. n ≥ 13 per genotype. ****p<0.0001. I) egl-30(js126) mutants have a significantly shortened lifespan (****p<0.0001), as previous reported (Ch’ng et al., 2008), while the lifespan of transgenic Podr-1::egl-30(Q205L) animals does not significantly differ from wild-type siblings (p = 0.47). n ≥ 100 per genotype. See also Figure S5.

Figure 6. Model summarizing regulation of memory formation by increased G_αq signaling

Activation of G_αq signaling in the AWC increases the release of neuropeptides that signal to the downstream LTAM network, resulting in CREB activation in the AIM and enhanced memory performance.