Synaptic memory and CaMKII

In the paper

“Although it is difficult to entirely exclude the possibility of a minor presynaptic component of NMDAR-dependent LTP, the evidence is overwhelming that its expression is primarily postsynaptic.”

What the sources say

• Supported

  In “Long-Term Potentiation: The Last 25 Years” (page unavailable), the review describes glutamate-uncaging experiments as strong evidence for postsynaptic LTP expression. It says those findings do not exclude an additional presynaptic mechanism, while noting that there is no need to invoke one during the first hour—the period most studied. (LLM, conf=0.91)

  cites [17] source
• ?Not checked — no full text of the sourcecites [18] source
• ?Not checked — no full text of the sourcecites [19] source
• Partly supported

  In the Introduction’s “The AMPAR code and LTP” section (page unavailable), the paper describes NMDAR-dependent LTP as resulting from increased postsynaptic AMPAR density and/or conductance, supporting primarily postsynaptic expression. Our copy does not address whether a minor presynaptic component is difficult to exclude; a search for “presynaptic” found no discussion of that qualification. (LLM, conf=0.84)

  cites [22] sourceExperimental — accuracy not yet measured
• Disputed

  The Introduction says presynaptic mechanisms can play an important, even dominant, role in some experimental situations. The Discussion concludes that early LTP can be entirely presynaptic, entirely postsynaptic, or a combination of both. The paper therefore does not treat presynaptic expression as merely a minor possibility or postsynaptic expression as generally predominant. (LLM, conf=0.93)

  Checked by a reasoning model — may be wrong.

  cites [31] sourceExperimental — accuracy not yet measured

In the paper

“The prevailing model posits that LTP-mediated covalent modification of the intracellular CTD of GluA1 results in the capture of these modified GluA1-containing receptors by preexisting “slots” in the postsynaptic density (PSD) (17, 18, 212).”

What the sources say

• Disputed

  In “Proposed Working Models for LTP” (page unavailable), the paper describes capture by preexisting PSD slots as one of three possible models, not the prevailing model. It explicitly says that C-terminal domains would be modulatory but not essential to that capture; modification of other receptor domains or auxiliary subunits would play the essential role. (LLM, conf=0.85)

  Checked by a reasoning model — may be wrong.

  cites [17] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [18] source
• ?Not checked — no full text of the sourcecites [212] source

In the paper

“CaMKII phosphorylates Kalirin at T75 (Ref. 315, but see Ref. 317), and this phosphorylation is sufficient to enhance synaptic transmission.”

What the sources say

• Disputed

  The Results identify T95, not T75, as the CaMKII phosphorylation site on kalirin-7; the Discussion likewise names Thr 95. The Results show that normal kalirin expression is required for activity-dependent enhancement of synaptic transmission, but do not show that phosphorylation alone is sufficient to enhance it. Page numbers are not available in our copy. (LLM, conf=0.90)

  Checked by a reasoning model — may be wrong.

  cites [315] sourceExperimental — accuracy not yet measured
• Disputed

  In Results and Discussion (‘Kalirin phosphorylation is brain region specific’; page unavailable), the paper says its in vitro studies found CaMKII phosphorylation at T79, not the T75 site named in the claim. The cited passage also does not show that phosphorylation at T75 is sufficient to enhance synaptic transmission. (LLM, conf=0.83)

  Checked by a reasoning model — may be wrong.

  cites [317] sourceExperimental — accuracy not yet measured

In the paper

“In support of these results, both sLTP (379) and LTP (121, 379) are greatly diminished in a CaMKII-dead mutant (K42K or K42M), despite its translocation.”

What the sources say

• Partly supported

  The sections on tetanus-induced spine structural plasticity and LTP report severely impaired spine enlargement and LTP, while the postsynaptic-dynamics section reports preserved translocation. The mutant studied is K42R, however, not the K42K or K42M mutant named in the claim. Page numbers are unavailable in our copy. (LLM, conf=0.88)

  cites [379] sourceExperimental — accuracy not yet measured
• Disputed

  The Discussion (page unavailable) says that substantial LTP remained with the CaMKIIα K42R mutation, contrary to the claim that LTP was greatly diminished in the kinase-dead mutant. The Results report loss of long-term sLTP with a different mutant, T286A; they do not establish the claimed sLTP result for K42K or K42M, or its translocation. (LLM, conf=0.85)

  Checked by a reasoning model — may be wrong.

  cites [121] sourceExperimental — accuracy not yet measured

In the paper

“There is even evidence that at CA1 excitatory synapses both NMDAR-dependent and NMDAR-independent forms of LTP can coexist under certain conditions (49–51).”

What the sources say

• ?Not checked — no full text of the sourcecites [49] source
• ?Not checked — no full text of the sourcecites [50] source
• Supported

  The paper reports that TEA application in the CA1 region produces two distinct potentiations: one induced through NMDA receptor channels and another distinct from NMDA receptor-dependent LTP. The Discussion likewise describes separate potentiations with different expression mechanisms under these conditions. The supplied passages do not give page numbers. (LLM, conf=0.91)

  cites [51] source

In the paper

“long-term potentiation (LTP), in which brief repetitive synaptic stimulation results in long-lasting increases in synaptic strength, was gaining popularity as a cellular model for learning and memory (16–22).”

What the sources say

• ?Not checked — no full text of the sourcecites [16] source
• Supported

  The review’s “Long-Term Potentiation: The Last 25 Years” section (page unavailable) describes repetitive activation of receptors on a spine producing a long-lasting increase in its AMPAR response. It also compares that enhancement with the effect of synaptic stimulation paired with postsynaptic depolarization, supporting the citing sentence’s general description of LTP. (LLM, conf=0.87)

  cites [17] source
• ?Not checked — no full text of the sourcecites [18] source
• ?Not checked — no full text of the sourcecites [19] source
• Supported

  In “The search for the cellular and molecular basis of synaptic plasticity” (page not provided), the paper describes repetitive activation producing a long-lasting increase in synaptic responses during LTP. Its Introduction (page not provided) also links interest in LTP to the study of memory. This supports the citing sentence’s description of LTP as a model for learning and memory. (LLM, conf=0.90)

  cites [20] source
• ?Not checked — no full text of the sourcecites [21] source
• Partly supported

  The Introduction describes LTP as a change in synaptic strength associated with learning and memory, and the ‘AMPAR code and LTP’ section describes enhanced synaptic AMPARs during LTP. The passages do not establish the claim’s specific induction condition: brief repetitive synaptic stimulation. (LLM, conf=0.84)

  cites [22] sourceExperimental — accuracy not yet measured

In the paper

“For instance, weak repetitive synaptic stimulation can elicit an NMDA-dependent decaying potentiation referred to as short-term potentiation (STP) (99–101).”

What the sources say

• Supported

  The Results report that a weak repetitive tetanus elicited STP that decayed back to baseline. The Characterization section describes STP in hippocampal CA1 as short-lasting, decremental potentiation dependent on postsynaptic NMDA receptor activation. (LLM, conf=0.93)

  cites [99] source
• Supported

  The Results report that a brief two-stimulus burst in a test pathway, paired with conditioning-pathway stimulation, elicited STP that decayed over approximately 20 minutes. A separate Results passage reports that an NMDA-receptor antagonist fully blocked the potentiation. The paired stimulation is an important condition of the finding. (LLM, conf=0.88)

  cites [100] source
• ?Not checked — no full text of the sourcecites [101] source

In the paper

“However, once Ca2+/CaM dissociates from a kinase phosphorylated at T286, its autonomous activity quickly phosphorylates T305 and T306, blocking rebinding of Ca2+/CaM and exerting a brake on further Ca2+/CaM stimulation, in what is at times referred to as inhibitory autophosphorylation (80, 84–86).”

What the sources say

• Partly supported

  The Results report that phosphorylation at T286 produces autonomous activity and that removal of Ca2+ triggers a rapid inhibitory burst of autonomous autophosphorylation. Figure 8 depicts phosphorylation of a serine and a threonine after Ca2+/calmodulin dissociates, reducing further Ca2+/calmodulin-dependent activity. Our copy does not identify those inhibitory residues as T305 and T306 or establish that inhibition occurs by blocking Ca2+/calmodulin rebinding. (LLM, conf=0.88)

  cites [80] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [84] source
• ?Not checked — no full text of the sourcecites [85] source
• ?Not checked — no full text of the sourcecites [86] source

In the paper

“This autophosphorylation is primarily intersubunit autophosphorylation within a holoenzyme, although at very high kinase concentration it can also occur by an interholoenzyme mechanism (FIGURE 2B4) (71, 79–81).”

What the sources say

• ?Not checked — no full text of the sourcecites [71] source
• ?Not checked — no full text of the sourcecites [79] source
• Partly supported

  The discussion (page unavailable in our copy) identifies threonine 286 as the autophosphorylation site responsible for autonomous kinase activity. It does not establish that phosphorylation is primarily between subunits of one holoenzyme or that phosphorylation between holoenzymes occurs at very high kinase concentrations; searches for those mechanisms found no supporting passage. (LLM, conf=0.82)

  cites [80] sourceExperimental — accuracy not yet measured
• ?Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  cites [81]

In the paper

“The binding of Ca2+/CaM peels off the regulatory segment from the surface of the kinase, thus unlocking constraints on catalytic competency and providing access to substrates and CaMKII interacting and anchoring proteins (e.g., GluN2B, Tiam1, and CaMKIIN) (61).”

What the sources say

• Supported

  The Results section, “A proposed model for persistent CaMKII activity” (page unavailable), states that Ca2+/CaM binding releases autoinhibition and puts the kinase in an “on” state. The Results section on αD-helix rotation explains that removal of the regulatory segment accommodates binding partners including GluN2B, Tiam1, and CaMKIIN, supporting the cited sentence’s description of access to substrates and interacting proteins. (LLM, conf=0.91)

  cites [61] source

In the paper

“It is NMDAR dependent, but unlike classical LTP the coincidence time window is seconds, not milliseconds.”

What the sources say

• Partly supported

  The Introduction describes BTSP as acting within a seconds-long window, and the Results section on initial membrane voltage says BTSP induction requires NMDA-type glutamate receptors, citing prior work and reporting an experiment consistent with that requirement. The Discussion contrasts BTSP with the short-timescale synchrony of traditional Hebbian plasticity, but our copy does not specify a millisecond coincidence window for classical LTP. (LLM, conf=0.83)

  cites [103] sourceExperimental — accuracy not yet measured
• Supported

  The indexed copy (section labelled Abstract; page unavailable) reports that NMDAR blockade reduced BTSP potentiation by 84%, supporting NMDAR dependence without implying complete blockade. It also reports a seconds-long pairing window and describes the plasticity kernel as two orders of magnitude longer than that expected for standard plasticity. (LLM, conf=0.91)

  cites [104] source
• Partly supported

  The Introduction (page unavailable) says BTSP affects inputs active several seconds before a burst and calls it a departure from conventional plasticity schemes, supporting the seconds-scale distinction. The available passages do not state that BTSP is NMDAR dependent or explicitly quantify classical LTP’s window in milliseconds. (LLM, conf=0.84)

  cites [105] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [106] source

In the paper

“Self-association occurring in transfected cells was used to show that the I205K mutant that blocks docking of GluN2B to the kinase also blocks clustering, suggesting that the regulatory segment and GluN2B have an overlapping docking site (134).”

What the sources say

• Supported

  The ‘Determinants within the catalytic domain’ section reports that Ile205Lys, previously shown to prevent CaMKII binding to NR2B (GluN2B), also prevented self-association in transfected HEK293 cells. The Discussion identifies Ile205 as part of a targeting site that interacts with NR2B and is proposed to interact with the Thr286 region of the autoregulatory segment, supporting the claim’s suggested overlap. (LLM, conf=0.92)

  cites [134] source

In the paper

“Two aspects of kinase structure reduce clustering, one is the CaMKIIβ subunit and another is T286 phosphorylation, but these are countered by the high CaMKII concentration at the PSD (134, 135).”

What the sources say

• Partly supported

  The section “Role of Thr286 autophosphorylation in CaMKII self-association” (page unavailable) reports that a phosphorylation-mimicking Thr286 mutation strongly reduced self-association, supporting the T286 part of the claim. The passages examined do not show that the CaMKIIβ subunit reduces clustering or that a high CaMKII concentration at the PSD counteracts either effect; the Discussion describes a concentrating effect of self-association, not that counteraction. (LLM, conf=0.84)

  cites [134] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [135] source

In the paper

“In the forebrain the ratio is roughly 3 α to 1 β, and holoenyzmes are composed of both subunits (155–157).”

What the sources say

• ?Not checked — no full text of the sourcecites [155] source
• Partly supported

  The subunit-composition results report a 3:1 α:β ratio in material precipitated with the α-specific antibody and show that forebrain holoenzymes containing both subunits exist. The Discussion also says the preparation contains α-only homomers, so the claim is too broad if it means all holoenzymes contain both subunits; the paper reports a 2:1 ratio for the heteromers. (LLM, conf=0.86)

  cites [156] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [157] source

In the paper

“Inactive CaMKIIα, but not β, binds to Shank3, and when it is activated it dissociates (161).”

What the sources say

• Supported

  The Results report that Shank3 binds CaMKIIα but not CaMKIIβ. They also show that its NTD-ANK region binds autoinhibited CaMKIIα and that T286 autophosphorylation disrupts the interaction. (LLM, conf=0.91)

  cites [161] source

In the paper

“They inhibit kinase activity with high affinity by binding across the entire substrate binding site (FIGURE 2B7) (61, 147, 153), and therefore also block GluN2B binding (152, 154).”

What the sources say

• Supported

  In Results (‘The CaMKII kinase domain has a single binding site’), CaMKIIN binds with high affinity (Kd = 39 ± 24 nM), and the structures place it on the same binding surface as GluN2B. Results (‘Testing the S and T site binding model’) reports that high levels of CaMKIIN completely inhibit substrate phosphorylation. The shared single site supports the stated inference that CaMKIIN blocks GluN2B binding; no page numbers are available in our copy. (LLM, conf=0.88)

  cites [61] source
• Partly supported

  The Results and discussion report that AIP potently inhibits CaMKII and binds a substrate-binding site. The paper studies AIP, not CaMKIINtides, and does not establish binding across the entire substrate-binding site. The citing sentence attaches different references to the GluN2B claim, so that part is not attributed to this source. (LLM, conf=0.84)

  cites [147] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [153] source

In the paper

“The linker length affects CaM affinity and Ca2+ spike frequency sensitivity (69, 158) as well as rates of autophosphorylation (159).”

What the sources say

• ?Not checked — no full text of the sourcecites [69] source
• ?Not checked — no full text of the sourcecites [158] source
• Supported

  Citation 159 is attached to the claim about autophosphorylation rates, not the preceding claims about CaM affinity or Ca2+ spike frequency sensitivity. The Concluding remarks state that longer linkers are expected to reduce the rate of activating trans-autophosphorylation, shifting the balance toward inhibitory phosphorylation; the Results and discussion also describe this rate relationship in a kinetic model. (LLM, conf=0.91)

  cites [159] source

In the paper

“Upon activation CaMKIIβ dissociates from F-actin because of competition with Ca2+/CaM for the F-actin binding region, and dissociation is made more persistent by specific autophosphorylation of CaMKIIβ in the F-actin binding region (171–174).”

What the sources say

• ?Not checked — no full text of the sourcecites [171] source
• Partly supported

  The Results sections on phosphorylation of the F-actin-binding region report that Ca2+/CaM reduces CaMKIIβ binding to F-actin and that autophosphorylation in that region prevents binding from recovering after Ca2+ chelation (page not provided). But the paper describes competition as a possible steric-hindrance mechanism involving adjacent binding domains, not an established competition for the same F-actin-binding region. (LLM, conf=0.88)

  cites [172] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [173] source
• Partly supported

  The Results section, “Activation of CaMKII Leads to a Loss of Bundling Activity” (page not provided), reports that Ca2+/calmodulin activation reverses CaMKIIβ-mediated F-actin bundling and that autophosphorylation contributes to a longer-lasting reversal. It does not establish that Ca2+/calmodulin competes specifically for the F-actin-binding region, or identify specific phosphorylation sites there as the cause of persistent dissociation. (LLM, conf=0.84)

  cites [174] sourceExperimental — accuracy not yet measured

In the paper

“It is proposed that this CaMKIIα-Shank3 complex may serve as a nearby tethered reservoir pool of CaMKII (161).”

What the sources say

• Supported

  The Introduction says inactive CaMKIIα is enriched in the Shank3 subcompartment of the postsynaptic density. The section “Recruitment of CaMKIIα to Shank3 condensates through direct interaction” proposes that CaMKIIα ‘stored’ by Shank3 can rapidly shift to NMDAR upon synapse activation, supporting the proposed nearby reservoir role (page unavailable). (LLM, conf=0.91)

  cites [161] source

In the paper

“Brief large-amplitude signals favor potentiation, whereas more modest long-duration signals favor depression (181).”

What the sources say

• Supported

  The Discussion states that substantially increasing the amplitude of a brief postsynaptic calcium signal selectively activates LTP, while substantially increasing its duration selectively activates LTD. This supports the claim's contrast between brief large-amplitude and more modest long-duration signals. (LLM, conf=0.94)

  cites [181] source

In the paper

“Interestingly, deletion of CaMKIIβ impairs LTP (121, 172, 175), as does the deletion of CaMKIIα.”

What the sources say

• Supported

  The Results section (‘The role of CaMKII in basal synaptic transmission and LTP’) reports that deleting CaMKIIα made LTP absent and that deleting CaMKIIβ blocked LTP similarly. This directly supports the attributed claim. (LLM, conf=0.95)

  cites [121] source
• Supported

  The Results section (‘Dissociation of CaMKIIβ from F-actin is necessary for functional and structural LTP’) reports that CaMKIIβ shRNA decreased LTP and says this was consistent with a study in CaMKIIβ knockout animals. The citation is attached to the CaMKIIβ claim, not the subsequent CaMKIIα clause. (LLM, conf=0.90)

  cites [172] source
• Supported

  In “Camk2b−/− mice show impaired LTP” (Fig. 3b), the paper reports a significant LTP deficit after CaMKIIβ deletion and says it is as severe as the deficit after loss of CaMKIIα. The same section reports impaired LTP in CaMKIIα-knockout mice under a stronger protocol (Fig. 3d). It also notes that the stronger protocol overcomes the CaMKIIβ-knockout deficit, so the claim does not hold for every induction protocol. (LLM, conf=0.93)

  cites [175] source

In the paper

“Activating VDCCs, which are present in spines, evokes cytosolic spine Ca2+ transients (185–187) and the activation of cytosolic spine CaMKII (188).”

What the sources say

• Supported

  The cited paper supports the Ca2+ portion of the sentence: its main-text passage labeled Abstract reports spine-head Ca2+ signals mediated exclusively by voltage-gated Ca2+ channels. The sentence attaches a separate citation (188) to CaMKII activation, so that portion is not attributed to this source. (LLM, conf=0.93)

  cites [185] source
• ?Not checked — no full text of the sourcecites [186] source
• ?Not checked — no full text of the sourcecites [187] source
• Supported

  Citation 188 is attached to the CaMKII-activation part of the sentence, not the spine Ca2+ transient claim. The paper’s “Channel specific activation of CaMKII” section reports that blocking L-type voltage-sensitive calcium channels abolished CaMKII activation in dendritic spines during depolarization. It specifies that this effect depends on L-type channels rather than voltage-dependent calcium channels generally; page numbers are unavailable in our copy. (LLM, conf=0.91)

  cites [188] source

In the paper

“Expression of active CaMKII (121, 130–132, 209, 210, 212) or direct delivery into the cell via the recording electrode (199) has repeatedly revealed a roughly two- to threefold enhancement in the AMPAR EPSC but little change in the NMDAR EPSC (130–132).”

What the sources say

• Partly supported

  In Results, ‘CaMKII constitutive activity and the maintenance of LTP’ (Fig. 10; page unavailable), expression of constitutively active CaMKIIα increased AMPAR EPSCs to about 250% of control. The cited passages do not report NMDAR EPSCs for that active-kinase experiment, so they do not establish the claimed little change in NMDAR EPSCs. Direct delivery through a recording electrode is attributed to a different citation in the citing sentence. (LLM, conf=0.83)

  cites [121] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [130] source
• Partly supported

  The Results (Fig. 2; page not specified in our copy) report that expression of active αCaMKII increased AMPAR EPSCs to 3.73 ± 0.9 times control, with no significant change in NMDAR EPSCs. This supports the selective enhancement, but the reported mean is above the claimed roughly two- to threefold range. The recording-electrode delivery method is cited separately in the citing sentence, not attributed to this paper. (LLM, conf=0.87)

  cites [131] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [132] source
• ?Not checked — no full text of the sourcecites [209] source

In the paper

“Just using a T286D mutation would be problematic because it generates a constitutively active kinase that then results in the phosphorylation of T305/T306, which can counteract T286D and alter resident time at the synapse (88, 210, 211).”

What the sources say

• ?Not checked — no full text of the sourcecites [88] source
• Partly supported

  The Results describe T286D as an active form of CaMKII and hypothesize that it might autophosphorylate T305/T306; they do not establish that this phosphorylation necessarily occurs. The Results and Discussion show that T305/T306 phosphorylation can counteract T286D-associated potentiation, and the Discussion says it reduces binding to the postsynaptic density. The paper does not report a change in residence time at the synapse (page numbers unavailable in our copy). (LLM, conf=0.86)

  cites [210] sourceExperimental — accuracy not yet measured
• Partly supported

  The Results and Discussion (pages unavailable) describe T286D as constitutively autonomous and report that T305/306 phosphorylation is expected to follow its activity and can inhibit further Ca²⁺/calmodulin stimulation and reduce GluN2B binding. They also report a possible further reduction in synaptic localization, significant by t test but not ANOVA. The paper does not measure synaptic residence time, so that specific effect is not established. (LLM, conf=0.84)

  cites [211] sourceExperimental — accuracy not yet measured

In the paper

“Most AMPARs in CA1 hippocampal pyramidal cells are tetraheteromeric assemblies of either GluA1/GluA2 subunits or GluA2/GluA3 subunits with a preponderance of GluA1/GluA2 subunits (264–266).”

What the sources say

• Supported

  The Results section (“Synaptic transmission is mediated primarily by GluA1A2 heteromeric receptors”; page unavailable) estimates that about 80% of receptors at excitatory synapses of CA1 pyramidal neurons are GluA1A2 heteromers and about 16% are GluA2A3 heteromers. The Introduction describes AMPARs as heterotetrameric receptors, supporting the assembly description. (LLM, conf=0.91)

  cites [264] source
• ?Not checked — no full text of the sourcecites [265] source
• Partly supported

  The immunoprecipitation results report that most complexes in the CA1/CA2 pyramidal-neuron preparation contain GluR1/GluR2 or GluR2/GluR3. Our copy does not establish a four-subunit assembly for those complexes or a clear preponderance of GluR1/GluR2 over GluR2/GluR3, and the preparation includes CA2 as well as CA1 (pages unavailable in the indexed copy). (LLM, conf=0.84)

  cites [266] sourceExperimental — accuracy not yet measured

In the paper

“It is suggested that activity-dependent synaptic recruitment of AMPARs may involve two steps (259, 261). First, AMPARs are delivered to the PSD, but this is not enough to increase synaptic strength. Second, the newly recruited AMPARs are added to AMPAR clusters in the nanocolumn.”

What the sources say

• ?Not checked — no full text of the sourcecites [259] source
• Supported

  The Discussion sections “Paradoxical functional effects of AMPA receptor recruitment” and “AMPA receptor recruitment and plasticity” propose that receptors recruited to the PSD may need to be concentrated in functional nanodomains to potentiate established synapses. This supports the citing sentence’s hedged two-step account. The paper calls these sites nanodomains rather than nanocolumns and notes that PSD recruitment alone can activate weak or silent synapses. (LLM, conf=0.88)

  cites [261] source

In the paper

“Intriguingly, AMPAR responses were reduced ∼50%, similar to the reduction seen by deleting CaMKII”

What the sources say

• Partly supported

  The Results (Fig. 9 caption; page not provided) report that replacement with GluN2Bmut left AMPAR responses at 48.5% of control. The Results and Discussion also report an approximately 50% reduction after CaMKIIα deletion. Our copy does not identify GluN2Bmut as L1298A-R1300Q, so that specific mutation designation remains unverified. (LLM, conf=0.82)

  cites [121] sourceExperimental — accuracy not yet measured

In the paper

“Replacement with the mutant GluN2B had no effect on NMDAR synaptic currents, indicating that it is functional.”

What the sources say

• Supported

  In Results, under “The synaptic actions of CaMKII require binding to NMDARs” (page not provided), the authors report that mutant GluN2B fully rescued NMDAR synaptic currents and conclude that the mutation did not affect NMDAR function. Figure 9 likewise shows no significant difference in NMDAR EPSCs for the mutant replacement. (LLM, conf=0.93)

  cites [121] source
• Supported

  The Introduction identifies the knock-in mutation as GluN2B L1298A/R1300Q. In the Results section on basal synaptic transmission, the authors report that NMDAR-mediated synaptic response measurements were indistinguishable between wild-type and knock-in mice and conclude that postsynaptic NMDAR function was unchanged under basal conditions. (LLM, conf=0.89)

  cites [120] source

In the paper

“Furthermore, AMPARs are highly mobile in the extrasynaptic membrane and can diffuse into and out of the synapse (20).”

What the sources say

• Supported

  The section on AMPAR surface diffusion describes the high mobility of extrasynaptic AMPARs and their exchange with synaptic membranes by lateral diffusion. It also reports independent confirmation that synaptic AMPARs rapidly exchange with the extrasynaptic pool. (LLM, conf=0.94)

  cites [20] source

In the paper

“CaMKII phosphorylates SynGAP and decreases its binding to PSD-95 (302, 304).”

What the sources say

• ?Not checked — no full text of the sourcecites [302] source
• Supported

  In the Results section on phosphorylation at Ser1108 and Ser1138, active CaMKII disrupted the SynGAP–PSD-95 interaction, whereas it did not disrupt binding by the phospho-deficient SynGAP mutant. The authors conclude that CaMKII phosphorylation at those sites triggers SynGAP dissociation from PSD-95. (LLM, conf=0.95)

  cites [304] source

In the paper

“Phosphomimic mutations greatly diminished both the binding of the TARP CTD to PSD-95 and the rescue of AMPAR transmission in AMPAR-deleted cells.”

What the sources say

• Partly supported

  The Results and Significance passages report that deleting TARP γ-8’s PDZ-binding motif prevented rescue of synaptic currents in AMPAR-deleted cells; a separate passage says phosphor-null mutations did not fully rescue basal transmission. Neither passage reports the claimed effect of phosphomimic serine mutations, and the PSD-95 binding experiment described in our copy concerns GluK1/Neto1 rather than the TARP CTD. No page numbers are provided in our copy. (LLM, conf=0.82)

  cites [288] sourceExperimental — accuracy not yet measured
• Supported

  The Discussion (page unavailable) states that the phosphomimetic γ−8 S10D mutant has diminished PSD-95 binding and impaired capacity to support AMPAR synaptic transmission. The Results (page unavailable) report significantly reduced AMPAR currents with S10D in cells where endogenous AMPARs had been removed. The paper also notes that a different, two-serine phosphomimetic mutant did not have these effects, so the finding does not apply to every such mutation. (LLM, conf=0.92)

  cites [289] source

In the paper

“CaMKII phosphorylation of the serines neutralizes the charge, releasing the CTD and providing access to the PDZ domains of PSD-95 (284, 285, 287).”

What the sources say

• Partly supported

  The Results section identifies TARPγ-8 serines phosphorylated by CaMKIIα. The Discussion describes phosphorylation-related electrostatic changes releasing the cytoplasmic domain from membrane lipids and enabling PSD-95 interaction, but specifically for TARPγ-2. It does not establish that CaMKII phosphorylation of TARPγ-8 neutralizes charge and exposes its PSD-95-binding site. (LLM, conf=0.86)

  cites [284] sourceExperimental — accuracy not yet measured
• Partly supported

  The Discussion (‘Roles of multiple phosphorylation sites in TARPs’; page unavailable) says acidic phosphorylated serines inhibit the electrostatic interaction of stargazin’s basic residues with negatively charged lipid bilayers. The Results (‘Lipid bilayers inhibit binding of stargazin to PSD-95’; page unavailable) say phosphorylation permits dissociation from lipids and binding to PSD-95. Our copy does not identify CaMKII as the kinase responsible for this mechanism, so that attribution is unsupported. (LLM, conf=0.86)

  cites [285] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [287] source

In the paper

“Since the binding of AMPAR/TARPs to the PSD protein PSD-95 is known to be important for synaptic targeting and LTP (278, 279)”

What the sources say

• ?Not checked — no full text of the sourcecites [278] source
• Partly supported

  In “The role of TARPs in AMPAR surface trafficking and synaptic targeting” (page not provided), the review says stargazin binds PSD-95 and that this interaction is important for AMPAR synaptic targeting. The same section reports that a phospho-null stargazin construct prevents LTP, but it does not establish that binding to PSD-95 itself is important for LTP. (LLM, conf=0.84)

  cites [279] sourceExperimental — accuracy not yet measured

In the paper

“since overexpression of PSD-95 on its own potentiates AMPAR EPSCs and occludes LTP (131, 297, 298) and the S73D mutant causes a similar potentiation.”

What the sources say

• ?Could not verify

  The Results (Figure 2) attribute increased AMPAR EPSCs to expression of activated αCaMKII, not PSD-95. Searches for PSD-95 overexpression, LTP occlusion, and the S73D mutant found no passage addressing those claims in our copy. (LLM, conf=0.76)

  cites [131] sourceExperimental — accuracy not yet measured
• Supported

  The citation is attached to the claim about PSD-95 overexpression, not the subsequent S73D clause. The Results report larger AMPA EPSCs in PSD-95-expressing hippocampal cells, and the Discussion states that PSD-95 expression occluded LTP. No page numbers are available in our copy. (LLM, conf=0.93)

  cites [297] source
• Partly supported

  The section on PSD-95-induced synaptic potentiation reports that PSD-95 expression potentiates AMPAR responses and prevents further LTP; the Discussion describes this as occlusion. The mutant-results section discusses C3,5S, AAAA, Double, and PDZ1-2 constructs, but the passages returned for searches on S73D and Ser73 do not support the claim that S73D causes similar potentiation. Page numbers were not available in our copy. (LLM, conf=0.89)

  cites [298] sourceExperimental — accuracy not yet measured

In the paper

“Interfering with actin polymerization blocks both sLTP (247) as well as LTP (310).”

What the sources say

• Not checked yet

  Checked before — its earlier verdict is held back until a new check runs.

  cites [247] source
• Partly supported

  Citation 310 supports the LTP portion of the sentence, not the separately cited sLTP portion. In “LTP induction is inhibited by actin polymerization inhibitors” (page unavailable), the paper reports that latrunculin B reduced potentiation rather than completely blocking LTP, so “blocks” overstates its finding. (LLM, conf=0.88)

  cites [310] sourceExperimental — accuracy not yet measured

In the paper

“Furthermore, knockdown of SynGAP results in the accumulation of synaptic AMPARs, which occludes chemLTP (304).”

What the sources say

• Supported

  In the Results section on AMPAR trafficking and spine enlargement (Figure 6), SynGAP knockdown produced elevated basal AMPAR levels in spines and prevented further receptor increases after LTP induction. The passage describes this as occlusion, matching the claim about chemLTP. (LLM, conf=0.94)

  cites [304] source

In the paper

“On average there are ∼250 PSD-95 molecules in the PSD (65) and, on average, ∼50 AMPARs, although this number is highly variable (143, 290, 291).”

What the sources say

• ?Not checked — no full text of the sourcecites [65] source
• Partly supported

  The Analysis of AMPAR- and NMDAR-Type Structures by EM Tomography section reports variability in AMPAR counts, including a relationship with PSD size, but estimates 105 ± 52 AMPARs in an average-sized PSD—not approximately 50. The ∼250 PSD-95 figure has a separate citation in the citing sentence and is not attributed to this paper. (LLM, conf=0.86)

  cites [143] sourceExperimental — accuracy not yet measured
• Partly supported

  The paper reports substantial variability in AMPA-receptor content across rat hippocampal synapses (Figure 8), supporting the variability qualifier. It does not establish approximately 50 AMPARs as an overall average: its estimates differ by synapse type, and it calculates an average of 81 functional receptors for juvenile mossy-fiber synapses. The PSD-95 figure is attached to a different citation and is not assessed here. (LLM, conf=0.85)

  cites [290] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [291] source

In the paper

“Furthermore, alanine substitutions of the critical serines in the TARP CTD, although reducing AMPAR EPSCs, do not affect the magnitude of LTP (288).”

What the sources say

• Partly supported

  The passage supplied under Significance (page not given) reports that phosphor-null mutations in tethered TARP γ-8 did not fully rescue basal synaptic transmission, while LTP was normal. This supports the direction of both effects, but the paper's finding is limited to tethered γ-8 with homomeric GluA1 receptors; the passage does not identify the substitutions as alanine, and it cautions that phosphorylation might matter for heteromeric AMPARs. (LLM, conf=0.82)

  cites [288] sourceExperimental — accuracy not yet measured

In the paper

“The authors find that Tiam1 has a CaMKII binding domain. This domain is homologous to the GluN2B CTD, and thus the Tiam1-CaMKII complex formation results in constitutive CaMKII activity.”

What the sources say

• Partly supported

  The Results sections on Tiam1 binding and reciprocal activation (page unavailable) identify a CaMKII-binding region in Tiam1 and report that the formed complex maintains CaMKII activity without Ca2+/calmodulin. They describe that region as homologous to CaMKII’s autoinhibitory domain, not to the GluN2B C-terminal domain. The claimed GluN2B homology is therefore unsupported by our copy. (LLM, conf=0.88)

  cites [314] sourceExperimental — accuracy not yet measured

In the paper

“A caveat to this proposal is that the expression levels of Tiam1 in the CA1 region of the hippocampus are very low and deleting Tiam1 has no effect on synaptic transmission (320).”

What the sources say

• Partly supported

  The CA1 neurotransmission section reports low Tiam1 expression and no detectable change in AMPAR- or NMDAR-mediated currents after Tiam1 shRNA knockdown. The Discussion likewise limits the finding to baseline glutamatergic transmission at Schaffer collateral–CA1 synapses. The claim says Tiam1 was deleted and extends the result to synaptic transmission generally; the cited experiment tested knockdown, not deletion. (LLM, conf=0.88)

  cites [320] sourceExperimental — accuracy not yet measured

In the paper

“The RhoGEFs that have received the most attention are kalirin-7, Trio, and Tiam, which are all expressed in spines and phosphorylated by CaMKII (314–317).”

What the sources say

• Partly supported

  The Introduction describes Tiam1 and Kalirin-7/Trio as CaMKII substrates involved in spine formation and development. The Results place Tiam1 in a stimulated spine, but the cited passages do not establish that all three are expressed in spines; the paper also does not say they have received the most attention. (LLM, conf=0.78)

  cites [314] sourceExperimental — accuracy not yet measured
• Partly supported

  The Results and Discussion establish that kalirin-7 is present in dendritic spines and phosphorylated by CaMKII at T95. The cited passages do not establish that Trio or Tiam are phosphorylated by CaMKII or that Trio is expressed in spines, so this paper supports only the kalirin-7 portion of the sentence. (LLM, conf=0.87)

  cites [315] sourceExperimental — accuracy not yet measured
• Partly supported

  In “Discovery and expression” (page unavailable), the review places Kal7 in dendritic spines; “Protein structure and signaling” says CaMKII is among the kinases that target Kalirin. The Tiam1 “Electrophysiology and synaptic function” section places Tiam1 in spines and reports calcium-dependent phosphorylation, but does not identify CaMKII as the kinase. The cited passages do not establish that Trio is expressed in spines or phosphorylated by CaMKII, so the claim that all three share both properties is not supported. (LLM, conf=0.87)

  cites [316] sourceExperimental — accuracy not yet measured
• Partly supported

  The Results and Discussion section (page unavailable) states that Kal7 is concentrated at the postsynaptic density and that in vitro studies show CaMKII phosphorylates Kal7 at T79. The paper does not establish that Trio or Tiam are expressed in spines or phosphorylated by CaMKII, so it supports the claim only for kalirin-7. (LLM, conf=0.90)

  cites [317] sourceExperimental — accuracy not yet measured

In the paper

“However, such a mechanism may well be involved in the dentate gyrus, where Tiam1 expression is high and does contribute to synaptic transmission (320).”

What the sources say

• Supported

  The Significance Statement says Tiam1 expression is enriched in dentate gyrus granule neurons. The Discussion reports that inhibiting Tiam1 in those neurons significantly reduces glutamatergic synapse function, supporting the claim that Tiam1 contributes to synaptic transmission there. Page numbers are not provided in our copy. (LLM, conf=0.90)

  cites [320] source

In the paper

“Importantly, although the maintenance of this segregation in the LLPS study required T286 autophosphorylation, it did not require the phosphorylation of any other component of the reconstituted system.”

What the sources say

• Supported

  The section “Segregation of AMPAR and NMDAR by LLPS-mediated mechanism of CaMKII” states that segregation does not require phosphorylation of the reconstituted components, while persistent segregation after calcium recedes requires CaMKII T286 phosphorylation. This supports the claim's distinction between T286 autophosphorylation and phosphorylation of other components. No page number is provided in our copy. (LLM, conf=0.93)

  cites [323] source

In the paper

“Remarkably, when the CTD of neuroligin 1 is included, it partitions with the AMPAR/TARPs but not with the NMDARs, reminiscent of nanocolumns in the PSD (251–253)”

What the sources say

• Supported

  In the “NLGN co-segregates together with AMPAR” section (Fig. 5B–C), the neuroligin-1 carboxyl tail moves into the STGc/PSD-95 phase while the NMDAR component GluN2Bc segregates into a distinct phase. The same section connects this arrangement to positioning AMPAR beneath the presynaptic active zone, supporting the nanocolumn comparison. (LLM, conf=0.94)

  cites [323] source
• Supported

  The citation supports the sentence’s comparison to nanocolumns in the postsynaptic density. The paper describes postsynaptic receptor and scaffold concentrations aligned with presynaptic nanoclusters, forming a transsynaptic nanocolumn (Abstract; page unavailable). (LLM, conf=0.88)

  cites [251] source
• Partly supported

  In “Lateral patterning of cytosolic PSD components” (page unavailable), the paper reports that AMPARs and NMDARs occupy separate synaptic zones in EM tomography and that TARPs associate with AMPAR scaffolds. But it does not report that the CTD of neuroligin-1 partitions with AMPAR/TARPs rather than NMDARs; its “Candidate mechanisms of trans-synaptic alignment” section instead describes neuroligin-1 as broadly distributed over the PSD. (LLM, conf=0.82)

  cites [252] sourceExperimental — accuracy not yet measured
• Partly supported

  The Discussion (page not specified) reports that AMPAR and NMDAR clusters barely colocalize at the nanoscale, supporting the comparison to distinct PSD domains. It does not report that the CTD of neuroligin 1 partitions with AMPAR/TARPs rather than NMDARs. (LLM, conf=0.83)

  cites [253] sourceExperimental — accuracy not yet measured

In the paper

“Sixth, whereas the recovery from bleaching spine SEP-GluA1 following glutamate uncaging shows a modest dependence on CaMKII, single SEP-GluA1 exocytotic events during glutamate uncaging appear to be independent of CaMKII (250).”

What the sources say

• Supported

  In the Results (Fig. 2D; page not provided), the CaMKII inhibitor KN62 partially blocked SEP-GluA1 fluorescence recovery after bleaching and stimulation. The Results also report that KN62 did not alter the stimulation-induced increase in individual exocytosis events in spines or dendrites (Fig. 4A–B). (LLM, conf=0.94)

  cites [250] source

In the paper

“Additional support for the vesicle-centric model is the finding that AMPAR-containing vesicles (obtained by immunoisolation) are associated with such proteins as synaptotagmin-1 (341), which is involved in LTP (195).”

What the sources say

• Supported

  The Abstract states that the authors immunoisolated AMPAR-containing vesicles and identified synaptotagmin-1 on them. This supports the association attributed to this paper; the citing sentence assigns synaptotagmin-1’s involvement in LTP to a separate citation. (LLM, conf=0.94)

  cites [341] source
• Supported

  Citation 195 is attached to the statement that synaptotagmin-1 is involved in LTP, not to the immunoisolation finding cited to 341. The Abstract reports that blocking postsynaptic synaptotagmin-1 together with synaptotagmin-7 abolished LTP and describes them as redundant sensors for AMPA-receptor exocytosis during LTP. (LLM, conf=0.94)

  cites [195] source

In the paper

“while on the other hand Stx-4 is reported to be required for LTP (336, 338).”

What the sources say

• ?Not checked — no full text of the sourcecites [336] source
• Supported

  The Results section, “Stx4 is Required for Long-Term Potentiation in CA1 Hippocampus,” reports that blocking postsynaptic Stx4 function almost completely abolished LTP in CA1 pyramidal neurons. The paper explicitly concludes that acute postsynaptic block of Stx4 function is sufficient to prevent LTP. (LLM, conf=0.94)

  cites [338] source

In the paper

“On the one hand Stx-3 is reported to be required for LTP (Ref. 334, but see Ref. 339)”

What the sources say

• Supported

  The Results section, “Postsynaptic Syntaxin-3 is Required for AMPAR Delivery During LTP,” concludes that Stx-3 is critical for AMPAR delivery during LTP. The Discussion also reports that Stx-3 knockdown impaired LTP and that wild-type Stx-3 rescued the effect. (LLM, conf=0.94)

  cites [334] source
• Supported

  This paper is cited in the “but see” qualification, not as the report that Stx-3 is required for LTP. Its Results section (“CA1 long-term potentiation is preserved in syntaxin-3 cKO mice”) reports normal theta-burst-induced LTP after syntaxin-3 deletion, while its Discussion acknowledges a previous knockdown study that found impaired LTP using a different induction protocol. (LLM, conf=0.91)

  cites [339] source

In the paper

“If this were the case, it has been argued that the system would be unstable and quickly saturate (366–371).”

What the sources say

• ?Not checked — no full text of the sourcecites [366] source
• Partly supported

  In §II, the paper says that if synapses are allowed only to increase, all will eventually saturate. It also says that its activity-dependent modification threshold is necessary for a bounded state. This supports the general concern about unchecked synaptic gain, but the paper does not address CaMKII or establish that saturation would occur quickly. (LLM, conf=0.84)

  cites [367] sourceExperimental — accuracy not yet measured
• Partly supported

  In “The Stability Problem” (page not provided), the paper argues that unconstrained strengthening of excitatory synapses creates positive feedback that can drive neuronal activity to saturation. It does not make that argument specifically about a net additive contribution from CaMKII; its discussion of CaMKII in “Signaling Pathways Underlying Synaptic Scaling” concerns different homeostatic effects. (LLM, conf=0.84)

  cites [368] sourceExperimental — accuracy not yet measured
• Partly supported

  In the section on power-law forgetting (page unavailable), the paper says that imbalances between synaptic strengthening and weakening spoil its memory-performance result, and that a repeatedly potentiated synapse can reach a point where it becomes no stronger. It does not connect this to CaMKII adding to a cell’s preexisting excitatory drive, or say that such a gain would make the system unstable and quickly saturate. (LLM, conf=0.76)

  cites [369] sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction says that Hebbian plasticity is a positive-feedback process whose synaptic weights are typically unstable until they reach saturation. The Results discuss CaMKII as a possible mechanism of Hebbian LTP, but do not establish that CaMKII adds a net gain to preexisting excitatory drive or that this specific arrangement would quickly saturate. (LLM, conf=0.83)

  cites [370] sourceExperimental — accuracy not yet measured
• Partly supported

  Section 4.1, Pair-based STDP rules (page unavailable), says that an additive plasticity model has an unstable fixed point and that individual synaptic weights move to their upper or lower bounds. It does not make this argument about CaMKII or say saturation is quick. It also says the neuron's mean firing rate is stabilized, so instability of individual weights should not be described as instability of the whole system. (LLM, conf=0.85)

  cites [371] sourceExperimental — accuracy not yet measured

In the paper

“The third mechanism, homeostasis, occurs on a slower timescale of many hours or days (368). In this case the cell senses the excess level of activity and through a cellwide mechanism scales down all synapses in a multiplicative manner to restore excitatory drive to a set point.”

What the sources say

• Supported

  The citation is attached to the timescale statement. In “Synaptic Scaling: the Phenomenon” (page unavailable), the paper says compensatory changes occur over many hours and firing rates return to control values over that period; “Signaling Pathways Underlying Synaptic Scaling” reports scaling after 4–24 hours of activity blockade, with effects continuing as deprivation is prolonged. (LLM, conf=0.91)

  cites [368] source

In the paper

“The first step involves Ca2+/CaM peeling away the autoinhibitory segment from the kinase domain. This results in the phosphorylation of T286 and also the binding of CaMKII to GluN2B even without prior autophosphorylation, both of which maintain CaMKII in an active state (64, 78, 119–121, 221).”

What the sources say

• ?Not checked — no full text of the sourcecites [64] source
• ?Not checked — no full text of the sourcecites [78] source
• Partly supported

  The Introduction (page not provided) says Ca2+/CaM exposes CaMKII’s catalytic site and that binding to NR2B then keeps CaMKII active independently of its phosphorylation state. It mentions Thr286 autophosphorylation as a requirement for a proposed second NR2B binding site, but does not say that Ca2+/CaM exposure results in Thr286 phosphorylation or that this phosphorylation maintains activity. (LLM, conf=0.83)

  cites [119] sourceExperimental — accuracy not yet measured
• Supported

  The Discussion section on postsynaptic CaMKII targeting describes Ca2+/calmodulin binding, T286 phosphorylation, and GluN2B binding as ways to remove the inhibitory pseudosubstrate from CaMKII’s catalytic site. It also explains that GluN2B binding keeps CaMKII subunits active and can sustain autophosphorylation, including when T286 was previously unphosphorylated. Page numbers are not available in our copy. (LLM, conf=0.87)

  cites [120] source
• Partly supported

  The Introduction (page not available) says Ca2+/CaM relieves CaMKII autoinhibition and that T286 autophosphorylation produces kinase activity that persists after Ca2+/CaM is removed. The Discussion identifies GluN2B as a CaMKII binding partner, but the paper does not establish that binding occurs without prior autophosphorylation or that the binding itself maintains CaMKII in an active state. (LLM, conf=0.86)

  cites [121] sourceExperimental — accuracy not yet measured

In the paper

“Second, NMDAR-dependent long-term depression (LTD) depresses synaptic transmission and can reverse established LTP (374, 375) (see sect. 9.2.3).”

What the sources say

• ?Not checked — no full text of the sourcecites [374] source
• Partly supported

  In the section headed “NMDA Receptor-Dependent LTD” (page unavailable), the review describes LTD of basal synaptic responses and says that blocking NMDARs blocks LTD. It also says synaptic activity can reverse LTP, but calls that phenomenon “depotentiation”; it does not establish that NMDAR-dependent LTD is what reverses established LTP. (LLM, conf=0.86)

  cites [375] sourceExperimental — accuracy not yet measured

In the paper

“As long as Ca2+/CaM is present, binding and phase separation are maintained even with a phosphonull mutant of T286 (T286A) (78, 169, 323, 377) or with a kinase-dead mutant of CaMKII (169, 323, 378, 379), although the binding of CaMKII to GluN2B is modestly enhanced by T286 phosphorylation (378).”

What the sources say

• ?Not checked — no full text of the sourcecites [78] source
• ?Not checked — no full text of the sourcecites [169] source
• Partly supported

  Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

  cites [323] sourceExperimental — accuracy not yet measured

In the paper

“Second, spine CaMKII activity as monitored with fluorescence resonance energy transfer (FRET) imaging (188) during LTP only remained elevated for ∼1 min after the induction of LTP, suggesting that conformational changes associated with CaMKII activation due to either Ca2+/CaM or autophosphorylation are transient.”

What the sources say

• Partly supported

  Split into 2 atoms — 1 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 0 unverifiable, 1 not checked. Expand the claim to see per-atom verdicts.

  cites [188] sourceExperimental — accuracy not yet measured

In the paper

“Further support for the role of T286 phosphorylation in maintaining the condensates and the binding of CaMKII to GluN2B in the absence of Ca2+/CaM is the finding that phosphatase PP2A rapidly dephosphorylates CaMKII, which results in the dissolution of the CaMKII-GluN2B complex (161).”

What the sources say

• Supported

  The Results section says PP2A rapidly dephosphorylates CaMKIIα when Ca²⁺ becomes low and that purified PP2A blocks its interaction with GluN2B. The Ca²⁺ oscillation section reports that this binding is quickly reversed by PP2A when Ca²⁺ is removed, supporting the described dissolution of the complex. Page numbers are not available in the supplied copy. (LLM, conf=0.92)

  cites [161] source

In the paper

“The higher concentrations of inhibitor that do reverse LTP (154, 361) could result from the displacement of GluN2B bound to CaMKII.”

What the sources say

• Supported

  The Discussion reports that 20 µM tatCN21 disrupts the CaMKII/NMDAR complex and reverses saturated LTP, whereas 5 µM inhibits CaMKII activity without reducing the complex or synaptic strength. The Results further report reduced binding involving the receptor’s NR2B (GluN2B) subunit at 20 µM. This supports the proposed explanation, though the paper does not establish displacement as the sole mechanism. (LLM, conf=0.91)

  cites [154] source
• Supported

  The Discussion (page unavailable) says the inhibitors interfere with CaMKII binding to GluN2B and that the study cannot distinguish disruption of that binding from blockade of kinase activity. It also identifies concentration and time as factors in reversing LTP. Together, these passages support displacement as a possible explanation, though the paper does not establish it as the mechanism. (LLM, conf=0.84)

  cites [361] source

In the paper

“Finally, another proposal also focuses on CaMKII and the phosphorylation of T305/T306, which selectively promotes LTD (402).”

What the sources say

• Supported

  The Results section ('CaMKII T305/306AV mutation impairs LFS-induced LTD but not HFS-induced LTP'; page not provided) reports that preventing T305/306 phosphorylation significantly impaired LTD while leaving LTP normal. It concludes that pT305/306 is selectively required for normal LTD expression at the excitatory synapse tested. (LLM, conf=0.94)

  cites [402] source

In the paper

“LTP is associated with a long-lasting increase in spine size (247, 248), including a late increase in the size of the PSD (56, 213, 414, 415) and additionally a slow increase in the size of the presynaptic bouton (213, 414).”

What the sources say

• Supported

  The cited paper supports the long-lasting increase in spine size: its Abstract reports persistent enlargement of stimulated small spines, and the passage describing Figure 4 links long-lasting enlargement to LTP. In the citing sentence, separate citations attach to the PSD and presynaptic-bouton claims; those claims are not attributed to this source. (LLM, conf=0.92)

  cites [247] source
• Supported

  The citation is offered for the first clause, the long-lasting increase in spine size, rather than the separately cited PSD and presynaptic-bouton changes. In “LTP leads to plasticity in spine head and neck geometry” (p. 11), the paper reports that spine head volume increased after LTP induction and then reached a stable plateau above baseline. (LLM, conf=0.89)

  cites [248] source
• Partly supported

  The Results section reports delayed PSD enlargement after structural LTP, and the Discussion describes PSD remodeling beginning about 60 minutes after induction. Our copy also reports persistent spine enlargement, but the passages located for presynaptic bouton growth do not report a slow increase in bouton size. That part of the claim is unsupported by this source. (LLM, conf=0.87)

  cites [56] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [213] source
• Partly supported

  The Introduction says spine enlargement after LTP can persist for hours to days, and the Results report PSD enlargement at 2 hours. The Results also show enlargement of presynaptic active zones, but do not establish a slow increase in the size of the presynaptic bouton itself. Page numbers are not available in our copy. (LLM, conf=0.87)

  cites [414] sourceExperimental — accuracy not yet measured
• Partly supported

  The Introduction and PSD-reorganization section describe spine enlargement after LTP and PSD growth that lags spine-volume growth by hours. The Discussion mentions presynaptic bouton volume change, but does not establish the claimed slow increase in bouton size. The paper also reports that late-phase spine volume was not significantly different from controls. (LLM, conf=0.86)

  cites [415] sourceExperimental — accuracy not yet measured

In the paper

“BTSP has been reproduced in slices (104) (FIGURE 12, A–C). In this case the plateau potential is induced by strong postsynaptic depolarization.”

What the sources say

• Supported

  The passage describing hippocampal slice preparations from adult mice reports that the authors tested whether the plasticity could be reproduced there. It says the postsynaptic plateau potential was generated by somatic current injection alone (300 ms, 300–600 pA), supporting the description of strong postsynaptic depolarization. (LLM, conf=0.94)

  cites [104] source

In the paper

“However, more recent studies have found evidence that high kinase concentration enables subunit exchange and transphosphorylation between subunits within a holoenzyme can occur in the absence of Ca2+/CaM but on a much slower time course (406).”

What the sources say

• Supported

  In Results and discussion (no page given), reducing CaMKII concentration from 8 to 1 µM reduced the rate of subunit exchange under activating conditions. The paper also reports that constitutively active CaMKII exchanged subunits without Ca2+/CaM and phosphorylated subunits in unactivated holoenzymes; its pThr286 signal rose much more slowly than the exchange signal. (LLM, conf=0.90)

  cites [406] source

In the paper

“It is interesting to note that CaMKIIα, but not CaMKIIβ, can self-associate, at least at reduced pH or high concentration and this requires displacement of the autoinhibitory domain (133–135).”

What the sources say

• ?Not checked — no full text of the sourcecites [133] source
• Partly supported

  The section on pH-dependent clustering reports that αCaMKII forms clusters at reduced pH in the presence of elevated Ca²⁺, and the Discussion proposes release of the autoregulatory domain as a step in self-association. Our copy does not establish that CaMKIIβ cannot self-associate or that high concentration alone permits CaMKIIα self-association; it presents the domain-displacement mechanism as a model rather than a demonstrated requirement. Page numbers are unavailable in our copy. (LLM, conf=0.85)

  cites [134] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [135] source

In the paper

“Although autophosphorylation does occur by transphosphorylation between subunits within a holoenzyme, the substrate subunit needs to be bound to Ca2+/CaM because that is what exposes T286 (82, 83), thus challenging the model of perpetuating CaMKII activity.”

What the sources say

• ?Not checked — no full text of the sourcecites [82] source
• Supported

  The Discussion states that autophosphorylation occurs between subunits within a holoenzyme and that Thr 286 is not phosphorylated when the substrate subunit cannot bind calmodulin. The Results support the proposed explanation that Thr 286 is shielded without bound Ca2+/calmodulin; they also describe a C289P mutant that bypasses this normal requirement. (LLM, conf=0.93)

  cites [83] source

In the paper

“Protein synthesis is also proposed to be required for “late” LTP (55–57) (see sect. 3.3).”

What the sources say

• Supported

  The paper states that consolidation of LTP requires protein synthesis (Abstract; page unavailable). It also reports that protein-synthesis inhibitors abolished the gradual long-term phase of LTP-associated spine enlargement while largely sparing the immediate phase. This supports the citing sentence’s qualified proposal about late LTP. (LLM, conf=0.91)

  cites [55] source
• Supported

  The Introduction states that consolidating early LTP into late LTP requires the synthesis and transport of new proteins. The Results further report that protein-synthesis inhibitors reduced late-phase spine enlargement, consistent with the citing sentence. (LLM, conf=0.94)

  cites [56] source
• Supported

  The Introduction explicitly distinguishes longer-lasting, protein synthesis-dependent L-LTP from shorter-lasting, protein synthesis-independent E-LTP. The Results section further reports that protein-synthesis inhibitors abolished the response induced by the authors’ L-LTP protocol. (LLM, conf=0.94)

  cites [57] source

In the paper

“Although it can be argued as to whether this occurs through interholoenzyme phosphorylation or through subunit exchange, this finding shows that phosphorylation of inactive holoenzymes occurs (not bound Ca2+/CaM), thus establishing a phosphorylation event for maintaining CaMKII activity in the face of protein turnover.”

What the sources say

• Supported

  The Results and discussion section reports Thr 286 phosphorylation in subunits of previously unactivated holoenzymes mixed with constitutively active CaMKII. It also describes experiments without Ca2+/CaM; the Introduction says the resulting spread of activation provides a mechanism for maintaining active CaMKII despite protein turnover. The paper presents that maintenance as a proposed mechanism, not a direct test of protein turnover. (LLM, conf=0.90)

  cites [406] source

In the paper

“This subunit exchange process requires CaMKII activation and is initiated after phosphorylation of T305/T306 in the CaM footprint, which blocks rebinding of CaM and frees the CaM-binding element to interact with the hub of the holoenzyme (407).”

What the sources say

• Partly supported

  The Introduction says phosphorylation of Thr 305/306 prevents Ca2+/CaM rebinding, and the Results and discussion describe activation-triggered exchange and a model in which the freed CaM-binding element interacts with the hub (pages not provided). But the paper says Thr 305/306 phosphorylation potentiates exchange; it does not establish that exchange must be initiated after that phosphorylation. (LLM, conf=0.84)

  cites [407] sourceExperimental — accuracy not yet measured

In the paper

“Remaining to be demonstrated, if this mechanism is to propagate CaMKII activity, is that an inactive holoenzyme can be phosphorylated on T286 in the absence of Ca2+/CaM. This is what was demonstrated in the studies claiming to exhibit subunit exchange (406).”

What the sources say

• Partly supported

  The Results and discussion section (Figure 5B; page unavailable) reports T286 phosphorylation in previously unactivated wild-type CaMKII mixed with active CaMKIIT286D. But the Figure 5 caption says the mixing experiments used wild-type CaMKII in the presence of Ca2+/CaM and ATP, so they do not demonstrate phosphorylation in its absence. (LLM, conf=0.89)

  cites [406] sourceExperimental — accuracy not yet measured

In the paper

“These in turn increase autophosphorylation of T286 (or T287 in some isoforms) that further increases constitutive activity (429, 430).”

What the sources say

• Partly supported

  The “Mutations in CAMK2” section reports increased Thr286 autophosphorylation for two CAMK2A variants, and the Discussion says Thr286 autophosphorylation makes CAMK2A autonomously active (page numbers unavailable). But the paper describes heterogeneous effects: other variants reduce phosphorylation, and the constitutively active Thr286Pro variant cannot be phosphorylated at Thr286. The claim therefore overgeneralizes the mechanism to these mutations as a group. (LLM, conf=0.88)

  cites [429] sourceExperimental — accuracy not yet measured
• Partly supported

  The Results section reports increased Thr286/Thr287 phosphorylation in three mutants, although two increases were not statistically significant and no phosphorylation signal was observed for a fourth. The Discussion proposes that the variants may increase autophosphorylation and thereby increase Ca²⁺-independent activity; the citing sentence presents that proposed chain as established for the mutations. (LLM, conf=0.86)

  cites [430] sourceExperimental — accuracy not yet measured

In the paper

“Several dozen human mutations in CaMKII have been identified, and their molecular characteristics support many of the biochemical behaviors of CaMKII described above.”

What the sources say

• Partly supported

  The Introduction and Discussion describe biochemical effects of the human CAMK2A E183V mutation, including reduced kinase activity and altered protein interactions. The paper does not establish that several dozen human CaMKII mutations have been identified or characterize their molecular effects as a group; the Discussion instead focuses on one mutation. No page numbers are available in our copy. (LLM, conf=0.88)

  cites [428] sourceExperimental — accuracy not yet measured
• Partly supported

  The Distribution of the Variants section reports 19 distinct CAMK2A or CAMK2B variants in 24 people, not several dozen distinct mutations. The Mutations in CAMK2 section reports that variants alter autophosphorylation, supporting the point that their molecular characteristics bear on CaMKII biochemical behavior. The cited paper therefore supports the substance of the biochemical point but not the stated count. (LLM, conf=0.88)

  cites [429] sourceExperimental — accuracy not yet measured
• Partly supported

  The Discussion reports five de novo CAMK2A/CAMK2B variants identified in this study, not several dozen. The Results link some of those variants to reduced autoinhibition and increased Thr286/Thr287 autophosphorylation, supporting the claim that human variants illuminate CaMKII biochemistry, but not its stated number or breadth. Page numbers are unavailable in our copy. (LLM, conf=0.88)

  cites [430] sourceExperimental — accuracy not yet measured

In the paper

“Consistent with a surface groove accommodating both substrates and anchoring proteins, it also exhibits reduced interactions with GluN2B and other interacting proteins.”

What the sources say

• Supported

  The Results section (page unavailable) reports that CaMKIIα-E183V significantly reduced binding to GluN2B and several other CaMKII-associated proteins compared with wild type. The surface-groove language frames the citing authors’ interpretation; the attributed finding about reduced interactions is supported. (LLM, conf=0.91)

  cites [428] source

In the paper

“Most of these molecules are in the pallium, just beneath the core of the PSD and not close enough to be bound to GluN2B (139).”

What the sources say

• Partly supported

  The Results describe CaMKII labeling as concentrated particularly in the PSD pallium under stimulation. The Discussion says CaMKII molecules in the distal pallium are too far from NR2B (GluN2B) in the PSD core to bind it directly. It does not establish that most CaMKII molecules are too far away; it also reports labeling in the core. (LLM, conf=0.86)

  cites [139] sourceExperimental — accuracy not yet measured

In the paper

“The number of NMDARs at spine synapses and perhaps filopodia synapses is much less variable, averaging roughly 20, and this number is largely independent of PSD size (143, 438, 439).”

What the sources say

• Partly supported

  The Analysis of AMPAR- and NMDAR-Type Structures by EM Tomography reports 23 ± 5 NMDAR-type structures in controls, nearly constant across PSD sizes; the Discussion estimates 20–30 per PSD. The study examined spine synapses, but our copy does not establish that the same pattern holds at filopodia synapses, even tentatively. (LLM, conf=0.86)

  cites [143] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [438] source
• ?Not checked — no full text of the sourcecites [439] source

In the paper

“Expression of the mutant in neurons decreases spine density and synaptic transmission.”

What the sources say

• Supported

  The spine-density results report significantly fewer dendritic spines in cultured hippocampal neurons expressing CaMKIIα-E183V. The synaptic-transmission results report reduced AMPAR-mediated mEPSC frequency and conclude that the mutation disrupts synaptic transmission. (LLM, conf=0.94)

  cites [428] source

In the paper

“The mutation is in the autoregulatory segment, and it is likely that the disruption of its conformation by the presence of a proline reduces autoinhibition and produces increased autonomy that has a gain-of-function effect in the cytosol (432).”

What the sources say

• Supported

  The Results locate p.Arg292Pro in the α-helical autoregulatory domain and report increased kinase activity. The Discussion proposes that the proline substitution disrupts autoinhibition, producing constitutive activity; the Results also describe gain-of-function effects in an exclusively cytosolic CAMK2GArg292Pro isoform. The claim preserves the paper’s uncertainty about the proposed mechanism. (LLM, conf=0.90)

  cites [432] source

In the paper

“The number of AMPARs at dendritic spine synapses is highly variable but averages roughly 50 (143, 438, 439), and this number strongly correlates with the size of the PSD.”

What the sources say

• Partly supported

  The Analysis of AMPAR- and NMDAR-Type Structures by EM Tomography reports a strong positive correlation between AMPAR-type structure count and PSD diameter (r = 0.87; Fig. 6A). It estimates 105 ± 52 AMPARs in an average-size PSD, not an average of roughly 50. Page numbers were not supplied for these passages. (LLM, conf=0.90)

  cites [143] sourceExperimental — accuracy not yet measured
• ?Not checked — no full text of the sourcecites [438] source
• ?Not checked — no full text of the sourcecites [439] source

In the paper

“The finding that strong hyperpolarization before the plateau potential blunts the potentiation (103) favors the former but does not rule out the latter.”

What the sources say

• Supported

  The citation is attached to the finding about hyperpolarization, not to the citing authors’ subsequent comparison of mechanisms. In Results, under “Dependence on initial Vm vs. initial synaptic weights,” the paper reports that strong hyperpolarization just before plateau initiation decreased synaptic potentiation (Figure 4—figure supplement 2). (LLM, conf=0.92)

  cites [103] source

In the paper

“Although there is no change in the number of PSD-95 molecules (56), saturating LTP results in a two- to threefold enhancement of the AMPAR EPSC, as does the expression of constitutively active CaMKII.”

What the sources say

• Supported

  The Results section reports that the amount of PSD-95 did not persistently change for up to 30 minutes after LTP induction. Citation (56) is attached to the PSD-95 statement, not to the subsequent claim about the magnitude of AMPAR EPSC enhancement or constitutively active CaMKII. (LLM, conf=0.92)

  cites [56] source