Spreading depression appeared during orthodromic-HFS with monophasic but not biphasic pulses and with a higher rather than a lower stimulation frequency
Electrical pulses of biphase and monophase were delivered through the concentric bipolar electrodes placed at the Schaffer collaterals for the orthodromic-HFS (O-HFS) and at the alveus fibers for the antidromic-HFS (A-HFS), respectively (Fig. 1a and 1b). For the stimulations of a single pulse with a same current intensity (Fig. 1c), the mean amplitude of orthodromically-evoked population spike (OPS) induced by a monophasic pulse (9.44 ± 1.25 mV, n = 5) was ~ 23% greater than that induced by a biphasic pulse (7.69 ± 1.74 mV, n = 6; t-test, P < 0.05). The result was consistent with the theory that a monophasic pulse is more efficient than a biphasic pulse in eliciting neuronal firing, because the second positive phase of a biphasic pulse may have a reverse effect to the first negative phase thereby reducing the ability of the negative pulse to activate neurons in a certain extent [8, 9].
During the O-HFS of 1-min 200 Hz biphasic pulses, OPS events only appeared in the initial several seconds of O-HFS (Fig. 2a). After the disappearance of OPS, multiple unit activity (MUA) continued to the end of the O-HFS with a firing rate of unit spikes higher than baseline. A silent period (10–30 s) without MUA appeared immediately following the end of O-HFS, indicating that the unit spikes during the O-HFS period were induced by the stimulation. During the period of O-HFS, an antidromic-test (A-test) pulse was applied every 5 seconds (i.e., 0.2 Hz) at the alveus fibers to evaluate the excitability of the CA1 neurons. The single A-test pulses and orthodromic-test (O-test) pulses were also applied before and after O-HFS to evaluate the baseline and the recovery of neuronal activity. Large antidromically-evoked population spike (APS) evoked by A-test pulses persisted throughout the 1-min O-HFS, and the mean amplitude of these APS (7.26 ± 5.59 mV, n = 6) was ~ 17% greater than the corresponding baseline level (6.19 ± 3.19 mV, n = 6; paired t-test, P < 0.05). These results indicated that the sustained O-HFS increased the excitability of the CA1 neurons. About 4 min after the end of O-HFS, both test APS and test OPS evoked by single pulses recovered to baseline level. In addition, no SD event appeared in all of the 6 rats with the 200 Hz biphasic O-HFS.
However, SD events appeared in 4 of the 5 rats with the 200 Hz monophasic O-HFS (Fig. 2b). The initial neuronal responses induced by the monophasic O-HFS was similar to that induced by the biphasic O-HFS: large OPS appeared at first, then OPS disappeared and dense MUA appeared. However, an SD appeared later with a slow waveform lasting 4.31 ± 3.11 s (n = 4). At the same time, the MUA disappeared completely and the A-test pulses were no longer able to induce APS, indicating a silence of neuronal activity. The MUA did not appear until 3.51 ± 2.47 min (n = 4) after the end of O-HFS (Fig. 2b, bottom). By this time, the test APS recovered to ~ 80% of baseline level. The test OPS did not recover even ~ 25 min after the end of O-HFS, while the test APS had almost recovered to baseline level (89.5 ± 9.7%, n = 4. Figure 2b, middle).
The 16 channels arranged in the four shanks of recording electrode (RE) showed the spread of SD waveforms (Fig. 3a). The waveforms of baseline OPS and APS along the shanks indicated the locations of each recording channel in the different stratums of CA1 region [17]. Because the signal recording in this study was AC-coupled (0.3–5000 Hz), the SD waveform appeared as a trough similar to previous reports [18]. The SD trough appeared first in the stratum radiatum (S. rad.) of hippocampal CA1 region (Fig. 3b), accompanied by a burst of population spikes (60–80 spikes/s) in the stratum pyramidale (S. pyr.) that was prominent in the filtered signals greater than 10 Hz (Fig. 3b right). Then the SD trough propagated slowly to the CA1 layers of S. pyr. and stratum oriens (S. ori.) at a speed of 89.9 ± 50.9 µm/s (n = 4) in the perpendicular direction, characterized by the movement of the negative peak of SD trough along the recording shanks (Fig. 3b, hollow triangles). Also, the SD trough moved at a speed of 826 ± 627 µm/s (n = 4) in the S. pyr. layer transversely among the recording shanks (Fig. 3a right, blue dotted line). The characteristics of the SD events, including the waveform, the accompanied burst of population spikes, the slow travelling speed and the silence of neuronal electrical activity, were consistent with previous reports [18, 19].
The result of statistical test showed that the SD incidence during 200 Hz monophasic O-HFS (4/5) was significantly greater than the incidence during 200 Hz biphasic O-HFS (0/6; Fisher’s exact test, P < 0.05). In addition, with a decrease of the O-HFS frequency from 200 to 100 Hz, no SD events were observed with monophasic O-HFS (five rats) and with biphasic O-HFS (six rats). Therefore, given the data of monophasic O-HFS only, the SD incidence during 100 Hz O-HFS (0/5) was significantly lower than that during 200 Hz O-HFS (4/5; Fisher’s exact test, P < 0.05).
These results indicated that O-HFS of monophasic pulses with a higher stimulation frequency may generate SD events in the hippocampal CA1 region and affect the orthodromic pathway persistently. The generation of OPS by the stimulation at afferent fibers involves both the axonal conductions and the synaptic transmissions. Therefore, the non-recovery of OPS after the monophasic O-HFS could have been caused by potential damages in the axons and/or synapses. To confirm whether the monophasic HFS could cause damages in axons, we next inspected the responses of CA1 neurons to the A-HFS at their own axons (i.e., the alveus) without involving synaptic transmissions.
More attenuation of population spike amplitudes during A-HFS of monophasic pulses
During the 1-min 200 Hz A-HFS of biphasic pulses, APS was able to follow each stimulation pulse with a large amplitude only at the initial period and then the APS amplitude decreased rapidly (Fig. 4a). Single test pulses showed that ~ 1 min after the end of A-HFS, the test APS recovered to ~ 70% of baseline level. And, ~ 2 min after the end of A-HFS, the test APS recovered to baseline level (92.2 ± 21.0%, n = 4).
When the A-HFS was applied with monophasic pulses, the neuronal responses at the initial period were similar to that of biphasic pulses -- each pulse induced a large APS. However, at the late period of A-HFS, the monophasic pulses hardly induced APS (Fig. 4b). In addition, the test APS only recover to 34.5 ± 12.1% (n = 5) of baseline level even ~ 20 min after the end of A-HFS.
The APS amplitude decreased rapidly at the onset of A-HFS and became steady within several seconds and persisted until the end of A-HFS, which may be caused at least partially by the depolarization block of axons [5, 20]. We calculated the APS amplitudes normalized to baseline value during the A-HFS period to evaluate the changes of APS (Fig. 4c). The decrease of APS amplitudes at the initial 1 s (ΔA1s) of A-HFS was significantly greater with monophasic A-HFS (86.9 ± 8.7%, n = 5) than with biphasic A-HFS (73.5 ± 8.0%, n = 4; t–test, P < 0.05. Figure 4d). Also, the decrease of APS amplitudes at the end of A-HFS (ΔA60s) was significantly greater with monophasic A-HFS (98.6 ± 1.2%, n = 5) than with biphasic A-HFS (94.1 ± 3.4%, n = 4; t–test, P < 0.05. Figure 4e).
The partial recovery of test APS after monophasic A-HFS together with the faster and larger degree of APS attenuation during A-HFS indicated that the monophasic A-HFS may cause conduction failures in a portion of axons even after the A-HFS.
In addition, no SD event was observed in all the 9 rats applied 1-min 200 Hz A-HFS (4 with biphasic pulses and 5 with monophasic pulses). That is, the SD incidence during 200 Hz monophasic A-HFS (0/5) was significantly smaller than that during 200 Hz monophasic O-HFS (4/5; Fisher’s exact test, P < 0.05).