Hebbian activity only temporarily stabilizes synaptic transmission at CA3-CA1 synapses in the developing hippocampus

Prolonged low frequency (0.05-1 Hz) stimulation of previously non-stimulated (naive) CA3-CA1 synapses in the developing hippocampus results in a profound synaptic depression explained by a postsynaptic AMPA silencing. It has been proposed that Hebbian activity can stabilize the synapses by preventing such depression. Using field recordings, we have examined to which extent strong repeated high frequency tetanization simulating Hebbian activity results in such prevention . The tetanization resulted within minutes in a field EPSP potentiation to 1 50- 170% of the naive field EPSP level which remained unaltered if stimulation was suspended. If test pulse stimulation (0.2 Hz) was allowed to continue after the te tanization the field EPSP continuously decreased a nd was after 2700 stimuli depressed by 75% from the po tentiated level. This depression did not differ in relative terms from that induced in naive synapses (by 82% from the naive level). The long-lasting component of this depression revealed by a subsequen t 30 min stimulus interruption (by 59% from the potentiated level) did not differ from that of naive synapses ( by 66% from the naive level). This equal relative degree of depression of tetanized and naive synapses was also observed following 2700 stimuli a t 1 Hz. On the other han d, when examined at earlier time points during the test pulse stimulati on (e.g. after 400 -900 stimuli) tetanized synapses were less depressed than naive synapses, and the long -lasting depression after 900 stimuli at 1 Hz was only ha lf that observed in naive synapses. This effect of tetanization was observed independently of whether the 1 Hz stimulation was commenced 15 min or 2 hours after the tetanization . In conclusion, while a strong preceding tetanization results in a partial sta bilization of transmission at CA3-CA1 sy napses in the deve loping hippocampus, this effect appears only temporary. This temporary effect is not linked to t ime after tetanization but to the number of low frequency stimuli given.

During brain development synapses are continuously generated and thereafter selected in an activity dependent manner to establish an appropriate mature pattern of synaptic connectivity. While little is known about the mechanisms that initiate the break -up of es tablished synapses or that protect synapses from eliminati on, activity dependent synaptic plasticity which may differ in important respects from th at in the mature nervous system are likely of critical importance. In fact, when CA3-CA1 synapses are su bjected to low frequency stimulation in the developing, but not in the mature , hippocampus, they become considerably depressed even when stimulated only a few times per minute . This depression was found to be initiated to much the same extent per stimulus over a large frequency range (0.05 -1 Hz) and to be explained by AMPA silencing (Abrahamsson et al., 2007; Strandberg et al., 2009; Xiao et al., 2004) . In the developing hippocampus even rather sparse activity may thus result in many CA3 -CA1 synapses losing their AMPA receptors (AMPARs), an absence that if occurring too frequently may be an initial s tep towards synapse elimination (Bastrikova et al., 2008; Becker et al., 2008; Kamikubo et al., 2006). The depression induced by low frequency stimulation can reverse in 20-30 min by a stimulus interruption or reverse by high frequency tetanization (Hebbian activity) (Abrahamsson et al., 2007, 2008; Xiao et al., 2004). When the depression is reversed by stimulus interruption a resumed low frequency stimulation easily depresses the synapses preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint again (Abrahamsson et al., 2008) . That is, while presynaptic silence can allow a synapse to maintain its AMPARs, th ese receptors can easily be lost again if the synapse is activated, even sporadically. In contrast, when the depression is reversed after a high-frequency tetanizat ion the synapses appear to stabilize at the naive level present prior to the initiation of stimulation of these synapses , a level when essentially no synapses are AMPA silent (Abrahamsson et al., 2008; Xiao et al., 2004) . Hebbian activity in the 2 nd postnatal week thus seems to pro duce plasticity that differs distinctly from that in the more mature brain in that no actual lasting potentiation takes place . Inste ad, Hebbian activity de -depresses the synapses and appears to transform the synapses into a state in w hich they are stable towards the depressive effect of low frequency stimulation (Groc et al., 2006) . Such a transformation would protect the synapse s from easily losing their AMPARs and possibly be protective against later elimination. Synaptic plasticity in the developing brain may th us have the character t hat synapses are easily silenced and possibly later eliminated if they are active without participating in Hebbian activity but are on the other hand protected against silencing after having participated in Hebbian activity. However, whether such a protection is permanent o r onl y temporary is unkno wn since the stabilization of the naive AMPA signaling level after tetanization has only been followed up to 60 min post -tetanus. In the present study we have used more prolonged low frequency stimulation following high frequency tetanization to examine in what manner participation in Hebbian activity alters the sensitivity of CA3-CA1 synapses in the developing hippocampus to low frequency stimulation.

Most experiment al details h ave previously been described in (Strandberg et al., 2009) and (Strandberg & Gustafsson, 2011) . In brief, experiments were performed on h ippocampal slices from 8–12-day-old Wistar rats kept an d killed in accordance wi th the guideli nes of the G othenburg ethical committee for animal research. The brain was removed, placed in an ice –cold solution and transverse hippocampal slices (400 µm thick) were cut with a vibratome. After typically 2 -5 hours of storage at 25ºC a single slice was transferred to a recording chamber where it was kept submerged in a constant flow (~ 2 ml per minute) at ~30ºC. The perfusion ACSF contained (in mM): 124 NaCl, 3 KCl, 4 CaCl 2, 4 Mg Cl2, 26 NaHCO 3, 1.25 NaH2PO4, and 10 D–glucose. Picrotoxin (100 µM) was always present in t he perfusion ACSF to block GABAA receptor-mediated activity. Electrical stimulation of Schaffer collateral afferents was carried out in the stratum radiatum delivered through a tungsten microelectrode (resistance ~0.1 MΩ) insulated except at its tip . Usually, two stimulating electrodes were positioned on either side of the recording electrode to provide for two independent synaptic inputs to the same dendritic region. Field EPSP recordings were made by me ans of a glass micropipette (~2 MΩ, filled with 1 M NaCl) in th e stratum radiatum. The field EPSP magnitude was measured, using linear regression over the first 0.8 ms, as the initial slope of the field EPSP rising phase. With the stimulation intensities u sed (20 -50 µA) the naive field EPSPs were generally subthreshol d for spike generation. The magnitude of the presynaptic volley was estimated by linear regression of the negative slope of the initial positive -negative deflection. T he field EPSP slope measurements in each experiment were linearly adjusted for by changes in the magnitude of the presynaptic volley. For validation of this “volley correction” of the field EPSP, see (Strandberg et a l., 2009). For the estimation of the naive level of synaptic transmission only the very 1st evoked field EPSP was used while for the estimation of the amount of depression attained at a certain time during the stimulation the average of 20 field EPS PS was used. Because the field EPSP obtained after stimulus interruption could be quite small, the field EPSP level after stimulus interruption was calculated as the average of the first three field EPSPs obtained when stimulation was resumed. The high -frequency tetaniz ation protocol consisted of three tetanization events 7.5 min apart, each event consisting of three 20 -impulse 50 Hz tetani at 0.1 Hz. To ascertain that the induction conditions were strong enough to fulfill the cooperativity (Hebbian) requir ement for LTP induction which may otherwise constitute a problem in the developing hippocampus (Harris & Teyler, 1984; Liao & Malinow, 1996) the tetanization of the synaptic test input was always made in conjun ction with tetanization of a second conditioning synaptic input (as in (Abrahamsson et al., 2008) ). Data are preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint expressed as means ± SEM. Statistical significance for independent samples was evaluated using Student’s t–test. Drugs Chemicals were from Sigma –Aldrich (Stockholm, Sweden).

Does high frequency te tanization stabilize synaptic transmission at the naive level? High-frequency tetanization of previously non - stimulated ( naive) CA3 -CA1 synapses in slices from the 2 nd postnatal week rat hippocampus does not result in any stable potentiation of the field EPSP. Instead, when foll owed by test p ulse stimuli (0.05-0.2 Hz) for 30 -60 min post-tetanus there is only a transient (5 -15 min) enhancement above the naive level bef ore the field EPSP seemingly stabilizes at th e naive level (Abrahamsson et al., 2008). In the present experiments we have used this procedure of applying tetanization at the naive level and have thereafter exposed t he synapses to very prolonged test pulse ( 0.2 Hz ) stimulation (2700 stimuli). Figure 1A sh ows that after applying our strong tetanization protocol (see Methods and text - Fig. 1) the field EPSP was after the 3 rd tetanization event potentiated to 151% ( ± 7.3% , n = 6) o f th e naive level and thereafter decayed back towards the naive level. However, as shown in Figure 1B , the field EPSP d id not stabilize at th e naive level but decreased continuously during the ~ 4-hour recording period to 37% of the naive level, i.e., to a depression of 63 ± 2.6% from the naive level (n = 6). Thus, not even strong Heb bian activity can stabilize the naive AMPA signalling level at CA3- CA1 synapses in the developing hippocampus. Test pulse stimulation of naive CA3-CA1 synapses

in both long -lasting depression and in depression that reverses within a subsequent 30 min period of stimulus interrupt ion (Strandberg & Gustafsson, 2011) . The 0.2 Hz -induced depre ssion of the tetanized synapses described above also partially reversed leaving a long-lasting depression of 38 ± 5.8% from the naive level (n = 6) (Fig. 1B). For comparison, when we exposed naive synapses to a sim ilar prolonged 0.2 Hz stimul ation (2700 stimuli) this stimulation resulted in a depression of 82 ± 1.8% from the naive level (n = 5) and in a long-lasting depression of 66 ± 4.6% from the naive level (Fig. 1C) . Tetanized synapses were thus less depressed than the naive sy napses both with respect to the final level of depression reached during the Figure 1. Effect of high frequency tetanization on test pulse-induced depression. A, after 20 stimuli at 0.2 Hz to previously non -stimulated ( naive) synapses to establish the naive field EPSP level, three 20-impulse 50 Hz tetani were given at 0.05 Hz (thick arrow) after which the 0.2 Hz stimulation was resumed . This tetanization event was repeated twice with 7.5 min interval. After the third tetanization event the test pulse stimulation ( 0.2 Hz) was first given three times 5 min after the tetaniz ation and thereafter resumed 15 min after the tetanization. Note t hat the field EPSP did not decay until the stim ulation was resumed 15 min after tetanization (n = 6 experiments). B, same as in A, but on a lo nger time scale. Note that the field EPSP decreases substantially below the naive level (100%). After 2700 stimuli, test pulse stimulation was interrupted for 30 min to allow the depression to reverse after which the test pulse stimulation was resumed. C, test pulse stimulation (0.2 Hz) applied to previously naive synapses (n = 5 experiments). 160 120 80 40 0 Normalised fEPSP (%) 6050403020100 Time (min) 160 120 80 40 0 Normalised fEPSP (%) 300250200150100500 Time (min) 120 100 80 60 40 20 0 Normalised fEPSP (%) 250200150100500 Time (min) A B C preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint stimulation (63% vs 82% , p < 0.001) and to the long-lasting depression (38% vs 66% , p < 0.02), indicating a partial stabilizing action of Hebbian activity (Fig. 2A). Which is the “naive” level after the tetanization? The above comparison between the 0.2 Hz -induced depressions obtained before and after tetanization assumes that the initial potentiation is a transie nt phenomenon whose impact on the fi eld EPSP is essentially over within 15 min of stimulation and which does not re verse during the stimulus interruption. Figure 1A shows that the prolonged 0.2 Hz stimulation was n ot initiated immediately after the 3rd tetanization e vent but that 1 5 min of stimulus interruption were given allowing the initial potentiation to largely decline (note the transien t character of the initial potentiation following the 1 st tetanization even t). As indicated in figure 1A, however, the field EPSP was as potentiated when the stimulation was resum ed after the 15 min of stimulus inte rruption as when measured following three 0.2 Hz stimuli delivered ~ 3 min after the 3 rd tetanization event. In other experiments (see below) we used a 2-hour stimulus interruption after the tetanization without observing any decreas e in the field EPSP . The tetanization-induced potentiation above the naive level is thus not transient by itself but declines in a stimulat ion dependent manner with a time course th at becomes more prolonged with the s uccessive tetanization eve nts (Fig. 1A) . To assess the stabili zing action of Hebbian activity on stimulus -induced depression we have th erefore also normalized the depression obse rved after tetanization to the very first field EP SP magnitude obtained when stimulation w as resumed 15 min after the 3 rd tetanization event . We will hereafter refer to this field EPSP magnit ude as the post- tetanus-naive level. Using this normalization procedure, the 0.2 Hz stimulation of tetanize d synapses resulted in a depression and in a long - lasting depression of 75 ± 2.7% and 58 ± 5.8% from the post-tetanus-naive level, respectively (Fig. 2B). As evaluated from this post-tetanus-naive level the tetani zed synapses thus are as depressed percent wise as the naive synapses both with respect to the depression reached during the stimulation (75% vs 82%, p = 0.06) and to the level of long-lasting depression (58% vs 66% , p > 0.40). Moreover, the reversible com ponent of the depression now becomes proportionally similar at tetanized synap ses (17 ± 3.3 %) as at the naive synapses (18 ± 1.7%). However, when measured at earlier time points during the stimulation, such as ~ 30 min after tetanization the tetanized syn apses 120 100 80 60 40 20 0 Normalised fEPSP (%) 250200150100500 Time (min) 0.2 Hz naïve 0.2 Hz normalised to peak of initial potentiation A B 160 120 80 40 0 Normalised fEPSP (%) 250200150100500 Time (min) 0.2 Hz naïve 0.2 Hz after tetanisation were significantly less depressed (34 ± 3.7% vs 54 ± 2.9 %, p < 0.005). Thus, even wh en evaluated using this normalization it appears that tetanization

in a partial stabilization of the AMPA signalling, albeit more temporarily. In fact, while it takes ~ 30 min to obtain a depression of 54% from the naive level at naive synapses it takes about three times as long to obtain the same amount of depression from the post -tetanus-naive level at the tetanized synapses (Fig. 2B). Does high frequency tetanization affect 1 Hz- induced depression? We next determined to what extent Hebbian activity also might exer t some stabilizing action on the depression ind uced by the common LTD-inducing 1 Hz stimulation. When applied to naive synapses 900 stimuli at 1 Hz fail to produce significantly more depression than the same number of stimuli at 0.05-0.2 Hz (Strandberg et al., 2009) . To determine Figure 2. Comparison be tween test pulse -induced depression of naive and tetanized synapses. A, the depressions shown in figure 1B and C, respectively, are plotted su perimposed using the naive level as

level (100%) for both curves. B, same plot as in A, but the depression of the tetanized synapses is plotted using the post -tetanus-naive level as reference level (100%). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint whether this is true also for tetanized synapses we 100 80 60 40 20 0 Normalised fEPSP (%) 806040200 Time (min) 200 150 100 50 0 Normalised fEPSP (%) 120100806040200 Time (min) 100 80 60 40 20 0 Normalised fEPSP (%) 806040200 Time (min) 1 Hz naïve 1 Hz normalised to peak of initial potentiation A B C first examined to what extent this equal potency of 1 Hz and test pulse stimulation to depress naive synapses also extends to the depression induced by 2700 stimuli. We th erefore applied 2700 stimuli at 1 Hz to naive synapses which resulted in a depression and a long -lasting depression of 89 ± 3.0% (n = 5) and of 51% ± 4.6% (n = 5) from the naive level, respectively (Fig. 3A). While the 1 Hz stimulation compared to the 0.2 Hz stimulation thus on average produced a somewhat greater depression (89% vs 82%) and a somewhat smaller long-lasting depression (51% vs 66%) none of these differences were statistically significant. However, the reversal from the 1 Hz-induced depression (38 ± 3.1% of the naive level) was significantly l arger (p < 0.004 ) than the reversal from the 0.2 Hz -induced depression (1 8 ± 1.7% of the naive level). In fact, while prolongation of 1 Hz st imulation of naive synapses from 900 to 2700 stimu li resulted in an increase of the depression from 60% (Strandberg & Gustafsson, 2011) to 89% from the naive level it did not lead to any significant increase in long - lasting depression (45% vs 51%, p = 0.19). In contrast, a similar prolongation of the 0.2 Hz stimulation of naive synapses resulted in a greater depression (63% vs 82%) but also to an even greater increase of the long-lasting depression (38% vs 66%). Thus, while the equal potency of 1 Hz and 0.2 Hz stimulation to depress naive synapses largely holds true also with respect to this more prolonged sti mulation, the 1 Hz stimulatio n is considerably less effective than the 0.2 Hz stimulation to convert the additional depression into a long-lasting depression. We thereafter applied the 2700 stimuli at 1 Hz to tetanized synapses which resulted in a depres sion and a long -lasting depression of 65 ± 9.8% and of 17 ± 14% (n = 5) from the naive level, respectively, and in a reversible depr ession of 47 ± 5.5% of the naive level (Fig. 3B). Thus, 1 H z stimulation fail ed to produce more depression than the 0.2 Hz stimulation also when applied to tetaniz ed synapses. Moreover, like the naive synapses the 1Hz stimulation of tetanized synapses resulted in a significantly greater revers ible depression than the 0.2 Hz stimulation (25 ± 3.9% of the naive level, p < 0.01) . As evaluated from the post -tetanus-naive level, the 1 Hz stimulation result ed in a depre ssion and a long -lasting depression of 81 ± 4.0% and of 54 ± 4.4% from that level , respectively, and a reversible depression of 27 ± 2.1 % of that level. Thus, also when evaluated in this manner the 1 Hz stimulation fail ed to produce s ignificantly more depression than the 0.2 Hz stimulation and the reversal of depression was significantly greater (p < 0.04). To evaluate the possible stabilizing action of Hebbian activity against 1 Hz -induced depression Figure 3. Effect of high frequency teta nization on 1 Hz-induced depression. A, 2700 stimuli at 1Hz was applied to previously naive synapses. After 30 min of stimulus interruption to allow the depression to reverse, 0.2 Hz test pulse stimulation was given (n = 5 experiments). B, same procedure as in Fig. 1 A, B but the stimulation started 15 min after the third tetanization event was at 1 Hz (n = 5 experiments). C, the depressions shown in A and B are plotted superimposed using the post -tetanus-naive level a s

level (100%) for the depression of the tetanized synapses. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint we comp ared the depression induced at naive 120 100 80 60 40 20 0 Normalised fEPSP (%) 6050403020100 Time (min) 1 Hz 900 15 min after tetanisation 1 Hz 900 120 min after tetanisation synapses from the naive level with that induced at tetanized synapses from the post-tetanus-naive level (Fig. 3C) . Like the 0.2 Hz -induced depression neither the depression (81%) nor the lo ng-lasting depression (54%) differed significantly from the corresponding depressions of the naive synapses (89% and 51%, respectively). On the other hand, the reversal from depression was significantly smaller for the tetanized synapses than for the naive synapses (27% vs 38%, p < 0.03) indicating that the relative failure of th e prolonged 1Hz st imulation to

in long-lasting rather than in rev ersible depression is less pronounced after tetanization. When measur ed after ~ 6 min of stimulatio n at 1 Hz, i.e., after a similar number of stimuli as the early m easurement using 0.2 Hz stimulation, the tetanized synapses were less depressed than the naive synapses (40 ± 3.3% vs 48 ± 2. 6%) but this difference did not reach statistical significance ( p = 0.08). Does the 1 Hz -induced long-lasting depression of tetanized synapses depend on time after tetanization? Since tetanization failed to induce any clear stabilizing e ffect on the amount of depression produced during on-going 1 Hz stimulation we next examined whether it may have affected the amount of long -lasting depression induced by this stimulation. We ha ve previously shown that 900 stimuli at 1 Hz of naive synapses results in a depression of 62% from the naive level, subdivided in reversible and long -lasting depression of 17% and 45% from the naive level, respectively (Strandberg & Gustafsson, 20 11). In the presen t experiments we applied this stimulation 15 min or 2 hours post -tetanus, respectively, in order both to compare the resulting depressions against each other and to compare them to the depression induced at naive synapses. The two hours of stimulus inter ruption did not affect the field EPSP since when stimulation was resumed after these two hours its magnitude (167 ± 6.4%, n = 5) was no different from that observed when examined by a few 0.2 Hz stimuli 4 min after the 3rd tetanization (169 ± 4.7%, n = 5, p = 0.77). Figure 4 shows that when normalized to the post -tetanus-naive level the depressions induced by 900 stimuli at 1 Hz applied 15 min and 2 hours post -tetanus, respectively, were quite super imposable and did not differ significantly with respect to either depression, long- lasting depression, or reversible depression . Compared to the naive synapses the depression after 900 stimuli was however somewhat smaller (50 ± 3.3%, n = 10, p < 0.0 02) (combined data of both 15 min and 2 hours po st-tetanus). In contrast, the long -lasting depression was s ubstantially smaller (24 ± 5 .3%, n = 8 , p < 0.00 5) and the reversible depression substantially larger (28 ± 3.2%, n = 8, p < 0.02) compared to the naive synapses. Thus, tetanization provides for a temporary stabilization also against 1 Hz stimulation, this effect however being most evident with respect to the a bility of the stimulation to convert depression into a long-lasting depression.

The present study shows that Hebbian activity doe s not stabilize AMPA signalling in developing CA3 - CA1 synap ses. Thus, even after strong repeated high frequency tet anization test pulse stimulation (1/5s) depresses the field EPSP far below its naive pre-tetanus level without a stable level of AMPA signalling being reached even after 3-4 hours of test pulse stimulation . This result contrasts with the stable potentiatio n observed in the mature hippocampus after 30 -60 min of stimulation at similar rates of test pulse stimulation (e.g. 1/7.5 s , (Volianskis & Jensen, 2003) ). Thus, the stabilizing action of Hebbian activity against te st pulse induced depression noted in earlier reports (Abrahamsson et al., 2008; Xiao et al., 2004) is only temporary . We also found that th e tetanization-induced transient potentiation above the naive AMPA signalling level did not decline if test pulse stimulation was suspended. High Figure 4. The effect of tetanization on 1 Hz -induced depression is not dependent on time after tetanization. 900 stimuli at 1Hz were given 15 min (n = 5 experiments, 1 experiment is lacking data after the 30 min st imulus inter ruption) or 120 min (n = 5 experiments, 1 experiment is lacking data after the 30 min st imulus inter ruption) af ter our standard tetanization protocol was applied (see text-Fig. 1 A). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint frequency teta nization thus leads to two seemingly disparate effects on the developing synapses; i) to a de-depression of any preceding test pulse induced depression back to the naive field EPSP level, and ii) to a potentiation exceeding the naive level. Both these effects are long-lasting if the synapse s are presynaptically silent and are transient if the synapses are exposed to test pulse stimulation. Are there any AMPA stable CA3-CA1 synapses in the 2nd postnatal week? Since test pulse induced depression in the 2 nd postnatal week appeared to saturate at about 40% from the naive level it was suggested that the CA3- CA1 synapse population could be subdivided into AMPA labile and stable AMPA-mature synapses, the latter synapses stabilized by prior participation in Hebbian activity (Abrahamsson et al., 2008; Groc et al., 2006) . We showed that more prolonged test pulse stimulation (900 stimuli at 0 .05-0.2 Hz) resulted in depression to ~ 60% from the naive level, and we now demonstrate that further 0.2 Hz stimulation (2700 stimuli in tot al) results in depression to ~ 80% from the naive level. Thus, while this further depression might involve synapses less AMPA labile than those silenced by briefer test pulse stimulation , this large depression nevertheless suggests that a subdivision into AMPA stable and AMPA labile synapse in the developing hippocampus is no longer tenable . Furthermore, as shown prese ntly, Hebbian activity does not create AMPA stable synapses at this developmental stage. Application of the protein kinase A (PKA) activator forskolin was previously found to de-depress an on- going test pulse induced depression and to stabilize the field EPSP at the naive level (Abrahamsson et al, 2008). Such an action by forskolin agrees with the demonstration that t etanization-induced “LTP” in the 2 nd postnatal week relies o n PKA a ctivation (Yasuda et al., 2003) and that GluR2 long-containing AMPARs at this developmental stage can be synaptically incorpor ated via NMDAR/PKA activation (Esteban et al., 2003; Kolleker et al., 2003; Qin et al., 2005; Zhu et al., 2000) . Our results would thus sugg est that any such NMDAR/PKA activation even by strong repeated high frequency tetanization of these synapses has no long -term stabilizing action on the AMPA signaling. Tetanization-induced p otentiation a bove the naive level While the p resent study show s that tetanization does not stabilize AMPA signaling it does indicate that the synapses become less labile. The evaluation of such a p artial stabilizing effect of Hebbian activity is howev er hampered by our poor understanding of what unde rlies the potentiation exceeding the naive level and its lability compared to that of the pre-existing AMPA signaling . Such a potentiation has previously been observed in the mature hippocampus and thought to un derlie the so called short-term potentiation (STP) (Volianskis & Jensen, 2003) , and has been attributed to a presynaptic modification (Volianskis & Jensen, 2003). While we at present cannot exclude such a mechanism also in the developing hippocampus it has been observed, albeit in somewhat ol der animals (P14) than used here, that whereas LTP in GluR1 knock -out mice was not much less than in wild-type mice the STP was absent (Jensen et al., 2003). GluR1 subunits are abundantly present in CA1 pyramidal cel ls also in the 2 nd postnatal week (Li et al., 2003; Zhu et al., 2000) but require CaMKII, which is present only at a low level at this early time period (Kelly et al., 1987) , for their stable synaptic insertion (Esteban et al., 2003; Hayashi et al., 2000) . However, the lack of S TP in GluR1 knock -out mice may sugg est that NMDAR/PKA activity may insert GluR1- containing AMPARs into the synaptic membrane but only in a labile manner such that these receptors are remo ved by test pulse stimulation. Application of the PKA activator fo rskolin also results in a transient po tentiation above the naive level (Abrahamsson et al., 2008) . Both the de -depression to the naive level and the potentiation abov e that level may thus be explained by a n NMDAR/PKA mediated insertion of AMPARs into th e synaptic membrane, GluR2long–containing and GluR1 - containing AMPARs, respectively. Comparison between depressions of naive vs tetanized synapses If we assume that the potentiation above the naive level is as labile as indicated by its decay after a single tetanization event (<10-15 min ) (Fig. 1A) (see also (Abrahamsson et al., 2008) ), all subsequent depression would be only that of the pre-existing AMPA signaling and be evaluated from the naive level. Hebbian ac tivity would th us appear to result in a substantial partial stabilization of both test pulse and 1 Hz -induced depression (Figs. 2 and 3 ). For example, after 2700 stimuli at 0.2 Hz the depression at tetanized synapses is 63% preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint from the naive level compare d to 82% at naive synapses, indicating that about twice as much of the pre-existing AMPA signaling remains after t his number of stimuli at tetani zed than at naive synapses. However, at the other extreme , we could assume that the potentiation above the naive level is as labile as the pre-existing AMPA signaling . This assumption has also the advantage that it compa res the overall lability of AMPA sig naling after a strong Hebbian induction vs that of naive synapses. The depression of the tetanized synapses should then be evaluated from the post-tetanus-naive level. Evaluated in this manner there was after 2700 stimuli no difference in lability between tetanized and naive synapses (Figs 2B and 3C). Furthermore, despite using this assumption that may overestimate the lability of the pre -existing AMPA signaling, the depression of tetanized synapses measured at earlier time points was less than the depressi on of naive synapses. It thus seems safe to conclude that tetanization does cause some st abilization of the pre-existing AMPA signalling, although to what extent this occurs is uncertain. Comparison between long -lasting depressions of naive vs tetanized synapses We previously showed that 900 stimuli delivered at both 0.2 and 1 Hz resulted in a long -lasting depression of ~ 40% from the naive level (Strandberg & Gustafsson, 2011) . We presently found tha t a dditional 1800 stimuli at 0.2 Hz substantially increased this depression to 66% from the naive level suggesting that prolonged test pulse stimulation can res ult in long -lasting silence in about two-thirds of the CA3 -CA1 synapses. After tetanization this long -lasting depression was found to be subs tantially reduced, at least as evaluated from the naive level (Fig. 2A). We believe ho wever that this procedure lea ds to an underestimation of the long -lasting depression at tetanized sy napses since the potentiation as observed after a single tetanization event is fully recuperated after 20 min of stimulus interruption after having been de - potentiated by 10 min of 0.2 Hz stimulation (Abrahamsson, Gustafsson and Hanse, unpublished observations). If we make the conservative assumption that after 2700 stimuli the de- potentiation of the potentiation (above the naive level) is reversed to much th e same extent as the depression of the pre-existing AMPA signalling the long-lasting depression of t etanized synapses should instead be evaluated from the post -tetanus- naive level. Adopting this procedure, the long- lasting depression at tetanized synapses did not in relative terms differ from that at naive synapses neither using 0.2 Hz (Fig. 2B) nor using 1Hz stimulation (Fig. 3C), indicating that Hebbian activity does not affect the ability of 2700 stimuli to

in long-lasting depression. Nonetheless, when the briefer stimulat ion of only 900 stimuli at 1 Hz w as used the long -lasting depression wa s substantially smaller than that observed at naive synapses. Moreover, when measured at earlier t ime points the depression was smaller at tetanized than at naive synapses (F ig. 2B). Thus, Hebbian activity at least results in a temporary partial stabilization against depression. An inhibitory influence of a preceding LTP on LTD induction has previously been described in older animals as a time-delimited effect, preventing LTD induction w ithin 60 min after LTP induction (Montgomery & Madison, 2002; Peineau et al., 2007). For example, us ing organot ypic slice cultures Montgomery and Madison described protection from depression o f recently unsilenced synapses lasting for about an hour. In th e present study we fou nd that the depression/ long-lasting depression produced by 900 stimuli at 1 Hz was similar whether the stimulation was applied 15 min or 2 hours after tetanization. The presently observed stabilizing action of Hebbian activity thus remain ed if the synapses were not stimulated and it dissipated with prolonged stimulation. Prolonged 0.2 Hz stim ulation is more effective than 1 Hz stimulation to produce long -lasting depression at naive synapses While 2700 stimuli resulted on average in a larger depression when 1 Hz than 0.2 Hz was used, the 1 Hz stimulation resulted in a significantly l arger reversal and in a smaller long-lasting depression. Moreover, after the stimulus interruption the reversed depression did not readily return to the pre-stimulus interruption level when stimu lation was resumed. We previously showed th at while NMDARs ar e not neces sary for inducing depression during on -going stimulation they are both necessary and su fficient for inducing the long- lasting depression (Strandberg & Gustafsson, 2011). Since 1 Hz stimulation result s in LTD of NMDAR-mediated synaptic transmission (Selig et al., 1995; Xiao et al., 1994) a possible explanatio n might be that the prol onged 1 Hz, but not 0. 2 Hz, stimulation of developing synapses results in such an NMDA LTD of sufficient magnitude to impair preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint induction of the long-lasting depression. The reduced depression observed when the stimulation was resumed after stimulus interruption is also similar to that observed whe n NMDARs a re blocked specifically during this time period (Strandberg & Gustafsson, 2011) , consistent with a 1 Hz -induced prolonged depression of NMDAR - mediated transmission. Functional considerations The glutamate synapse appears to acquire its AMPARs early in an NMDAR independent manner, but the AMPARs may easily be l ost upon synaptic activation. In the first po stnatal weeks the main role for Hebbian activity may therefore be to stabilize AMPA signalling in th e synapses that partake in such combined pre - and postsynap tic activity. T he present result s hows that in the 2 nd postnatal week CA3 -CA1 synapses that participate in Hebbian activity are initially less easily depressed when ex posed to low frequency activity . Nevertheless, even after extensive participation in Hebbian activity they do not in the long run appear significantly m ore stable than before . To maintai n its AMPA signalling the synapse must then either be pre synaptically silent, or continuously participate in Hebbian activity, synaptic activity outside this context leading to AMPA silencing and possible elimin ation. This would all ow for a dynamic build -up and refinement of the synaptic circuitry during development, only al lowing synapses that throughout this early developmental period are activated in proper relat ion to other synapses to survive.

Abrahamsson, T., Gustafsson, B., & Hanse, E. (2007). Reversible synaptic depression in developing rat CA3 CA1 synapses explained by a novel cycle of AMPA silencing- unsilencing. J Neurophysiol, 98(5), 2604-2611. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=17804578 Abrahamsson, T., Gustafsson, B., & Hanse, E. (2008). AMPA silencing is a prerequisite for developmental long-term potentiation in the hippocampal CA1 region. J Neurophysiol, 100(5), 2605-2614. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=18799599 Bastrikova, N., Gardner, G. A., Reece, J. M., Jeromin, A., & Dudek, S. M. (2008). Synapse elimination accompanies functional plasticity in hippocampal neurons. Proc Natl Acad Sci U S A, 105(8), 3123-3127. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=18287055 Becker, N., Wierenga, C. J., Fonseca, R., Bonhoeffer, T., & Nägerl, U. V. (2008). LTD induction causes morphological changes of presynaptic boutons and reduces their contacts with spines. Neuron, 60(4), 590-597. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=19038217 Esteban, J. A., Shi, S. H., Wilson, C., Nuriya, M., Huganir, R. L., & Malinow, R. (2003). PKA phosphorylation of AMPA receptor subunits controls synaptic trafficking underlying plasticity. Nat Neurosci, 6(2), 136-143. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=12536214 Groc, L., Gustafsson, B., & Hanse, E. (2006). AMPA signalling in nascent glutamatergic synapses: there and not there! Trends Neurosci, 29(3), 132-139. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=16443288 Harris, K. M., & Teyler, T. J. (1984). Developmental onset of long-term potentiation in area CA1 of the rat hippocampus. J Physiol, 346, 27-48. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=6699775 Hayashi, Y., Shi, S. H., Esteban, J. A., Piccini, A., Poncer, J. C., & Malinow, R. (2000). Driving AMPA receptors into synapses by LTP and CaMKII: requirement for GluR1 and PDZ domain interaction. Science, 287(5461), 2262-2267. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=10731148 Jensen, V., Kaiser, K. M., Borchardt, T., Adelmann, G., Rozov, A., Burnashev, N., Brix, C., Frotscher, M., Andersen, P., Hvalby, O., Sakmann, B., Seeburg, P. H., & Sprengel, R. (2003). A juvenile form of postsynaptic hippocampal long-term potentiation in mice deficient for the AMPA receptor subunit GluR-A. J Physiol, 553(Pt 3), 843-856. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=14555717 Kamikubo, Y., Egashira, Y., Tanaka, T., Shinoda, Y., Tominaga-Yoshino, K., & Ogura, A. (2006). Long-lasting synaptic loss after repeated induction of LTD: independence to the means of LTD induction. Eur J Neurosci, 24(6), 1606-1616. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=17004924 Kelly, P. T., Shields, S., Conway, K., Yip, R., & Burgin, K. (1987). Developmental changes in calmodulin-kinase II activity at brain synaptic junctions: alterations in holoenzyme composition. J Neurochem, 49(6), 1927-1940. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=2824699 Kolleker, A., Zhu, J. J., Schupp, B. J., Qin, Y., Mack, V., Borchardt, T., Kohr, G., Malinow, R., Seeburg, P. H., & Osten, P. (2003). Glutamatergic plasticity by synaptic delivery of GluR-B(long)-containing AMPA receptors. Neuron, 40(6), 1199-1212. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=14687553 Li, A. J., Suzuki, M., Suzuki, S., Ikemoto, M., & Imamura, T. (2003). Differential phosphorylation at serine sites in glutamate receptor-1 within neonatal rat hippocampus. Neurosci Lett, 341(1), 41-44. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=12676339 Liao, D., & Malinow, R. (1996). Deficiency in induction but not expression of LTP in hippocampal slices from young rats. Learn Mem, 3(2-3), 138-149. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=10456084 Montgomery, J. M., & Madison, D. V. (2002). State-dependent heterogeneity in synaptic depression between pyramidal cell pairs. Neuron, 33(5), 765-777. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=11879653 Peineau, S., Taghibiglou, C., Bradley, C., Wong, T. P., Liu, L., Lu, J., Lo, E., Wu, D., Saule, E., Bouschet, T., Matthews, P., Isaac, J. T., Bortolotto, Z. A., Wang, Y. T., & Collingridge, G. L. (2007). LTP inhibits LTD in the hippocampus via regulation of GSK3beta. Neuron, 53(5), 703-717. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=17329210 Qin, Y., Zhu, Y., Baumgart, J. P., Stornetta, R. L., Seidenman, K., Mack, V., van Aelst, L., & Zhu, J. J. (2005). State-dependent Ras signaling and AMPA receptor trafficking. Genes Dev, 19(17), 2000-2015. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=16107614 Selig, D. K., Hjelmstad, G. O., Herron, C., Nicoll, R. A., & Malenka, R. C. (1995). Independent mechanisms for long-term depression of AMPA and NMDA responses. Neuron, 15(2), 417-426. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=7544143 Strandberg, J., & Gustafsson, B. (2011). Critical and complex role of N-methyl-D-aspartate receptors in long-term depression at CA3-CA1 synapses in the developing hippocampus. Neuroscience, 192, 54-66. https://doi.org/10.1016/j.neuroscience.2011.06.081 Strandberg, J., Wasling, P., & Gustafsson, B. (2009). Modulation of low-frequency-induced synaptic depression in the developing CA3-CA1 hippocampal synapses by NMDA and metabotropic glutamate receptor activation. J Neurophysiol, 101(5), 2252-2262. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=19225168 Volianskis, A., & Jensen, M. S. (2003). Transient and sustained types of long-term potentiation in the CA1 area of the rat hippocampus. J Physiol, 550(Pt 2), 459-492. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=12794181 Xiao, M. Y., Wasling, P., Hanse, E., & Gustafsson, B. (2004). Creation of AMPA-silent synapses in the neonatal hippocampus. Nat Neurosci, 7(3), 236-243. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=14966524 Xiao, M. Y., Wigström, H., & Gustafsson, B. (1994). Long-term depression in the hippocampal CA1 region is associated with equal changes in AMPA and NMDA receptor-mediated synaptic potentials. Eur J Neurosci, 6(6), 1055-1057. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=7952275 Yasuda, H., Barth, A. L., Stellwagen, D., & Malenka, R. C. (2003). A developmental switch in the signaling cascades for LTP induction. Nat Neurosci, 6(1), 15-16. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=12469130 Zhu, J. J., Esteban, J. A., Hayashi, Y., & Malinow, R. (2000). Postnatal synaptic potentiation: delivery of GluR4-containing AMPA receptors by spontaneous activity. Nat Neurosci, 3(11), 1098-1106. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Ci tation&list_uids=11036266 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 11, 2024. ; https://doi.org/10.1101/2024.01.10.574846doi: bioRxiv preprint