Brain Research, 578 (1992) 297-304 (~ 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-899302/$05.00

297

BRES 17653

Properties of favored patterns in spontaneous spike trains and responses of favored patterns to electroacupuncture in evoked trains Yi-Qing Chen and Yun-Hui Ku Department of Physiology, Beijing Medical University, Beijing (People's Republic of China) (Accepted 10 December 1991) Key words: Favored pattern; Spontaneous spike train; Evoked spike train; Locus coeruleus; Substantia nigra zona compacta

By using 'the modified detection method', our previous study has shown that all spontaneous spike trains recorded from several areas of brain and spinal cord have favored patterns (FPs). The present study further shows that: (1) all newly detected spike trains from substantia nigra zona compacta, nucleus periventricularis hypothalami and nucleus hypothalamicus posterior also have FPs, and some spike trains from neurons in the same nucleus have a common favored pattern (CF, i.e. they share the same FP), indicating that FP and CF in spike trains are common phenomena; (2) all serial correlation coefficients of FP repetitions (in serial order) in different spike trains detected are less than 0.3 (close to 0), revealing that the repetition of FPs is a renewal process; (3) in different periods of the spike trains evoked by electroacupuncture (EA), the number of different FPs and the number of repetitions of the same representative FP either increase or decrease along with the change of firing rate. The tendencies of these changes are very similar, but after EA the repetitions of different FPs in the same spike trains change differently, showing that different (hidden) responses exist at the same time. The above results suggest that the FPs in spike trains may represent various neural codes, and 'the modified detection method of FP' can pick up more information from spike trains than the firing rate analysis, hence it is a very useful tool for the study of neural coding. INTRODUCTION

dex of neuronal activity and 'the modified FP-detection m e t h o d ' can be used for studies of neuroscience and

In the contemporary world the most traditional methods of spike train analysis (mainly analyzing the changes in firing rate) are not deepgoing. Clearly, such kinds of methods can not pick up more information from the spike trains, which are generally accepted as the carriers of information8. We presented a 'modified m e t h o d ' for detecting favored pattern (FP) in spike trains 7 after Dayhoff and Gerstein 1"2 and carried out studies with simulated trains (containing k n o w n interpolated patterns) and real spike trains. O u r previous study 7 has shown that: (1) 'the modified detection m e t h o d ' is reliable and appropriate for detecting favored patterns (FPs) in different simulated and real spike trains; (2) all 44 spontaneous spike trains detected have FPs, some spike trains recorded from n e u r o n s in the same nucleus have a c o m m o n fragment of FP (i.e. they share the same FP fragment), while the FPs in spike trains from different nuclei are different from each other; (3) the FPs can remain unchanged From beginning to end in 35-min records, and their repetitions are relatively stable; (4) microinjection of noraaal saline or normal serum into the same nucleus durn g recording has no significant influence on the Jccurrence of FPs in 35-min records of spike trains; the ~bove results suggest that the FP may be used as an in-

neural coding. O n the basis of the above, the purposes of the present study are: (1) to apply 'the modified method' to more spike trains from different nuclei to see if the FP and the c o m m o n fragment of FP are c o m m o n p h e n o m e n a ; (2) to analyze the properties of the FPs in spontaneous spike trains; (3) to test if 'the modified m e t h o d ' is also applicable to the evoked spike trains, and to investigate the regularities of FP-changes in evoked spike trains.

MATERIALS AND METHODS Modified FP detection method 7 'The modified FP detection method' includes several procedures: 1. Sampling. The spike trains recorded were fed to a window discriminator (model DSC-3). After the discharge of a single neuron was isolated and transformed into a rectangular impulse train, the output of DSC-3 was fed to a microcomputer (Super PC/AT), where the interval histograms were made and FPs were detected. The sampling point in the computer is 18/~s. Each sampling spike train includes 3-5 min of activity. 2. Analyzing the interval histogram to select the favored interval length. 3. The quantized Monte Carlo statistical method. Firstly, the interval values were quantized at a chosen quantization resolution (quantizaton bin width). Then the computer program implemented the quantized Monte Carlo method which can identify FPs not having extra or missing spikes, and a statistical test (the shuffling

5orrespondence: Y.H. Ku, Department of Physiology, Beijing Medical University, Beijing 100083, People's Republic of China.

298 test) at the 1% significance level was used. The number of FPs detected was equal to the number of FPs in the unshuffled spike trains minus the maximum number in 99 shuffles (for details see ref. 1). In view of the fact that neural spike trains contain fast and slow firing mode, we treated the fast firing (including 'bursts') and slow firing mode separately: the range for the fast firing was from several tens of microseconds to several milliseconds, that for the low firing was from several tens of milliseconds to several seconds. 4. Choosing templates. Since the interval lengths of the FP identified by the quantized Monte Carlo method varied from occurrence to occurrence, each interval length of the FP (template) used for template matching (the next step) was averaged respectively, i.e. the template chosen for template matching is an averaged template. 5. The template statistical method. This method was based on a 'sliding' comparison of the chosen template with the entire spike train, which can determine the locations of patterns matched with the template pattern in the spike train and determine the number of matches, and also a statistical test (the shuffling test) at the 1% significance level was used. The number of pattern repetitions was equal to the number of matches in the unshuffled spike train minus the maximum number in 99 shuffles. For the detection of imperfect matches, our modified template method allows intervals of the template in sequence matching with the corresponding intervals of the spike train, respectively, using a certain tolerance of variation (0.2 of the interval length). If the successive interval lengths of the matched fragment of spike train are all within the tolerance of the corresponding interval lengths of the template, the fragment of spike train is said to 'match' the template. In the present study, a section of the spike train that approximately matches the template pattern (i.e. the former may contain one extra or one missing spike) can still be considered a match (see Fig. 1).

ill

splke t r o i n

---,.-slide

temprate r

1

intervol I

r 1

~

'

it]

~ntervol II

i ntervol

Ill tolerence range

Fig. 1. Showing the mechanism of template matching which is a sliding comparison of the chosen template with the entire spike train: if the successive interval lengths of the matched fragment of spike train were all within the tolerance (the hatched area) of the corresponding interval lengths of the template, the fragment of spike train was said to 'match' the template (the former may contain one extra or one missing spike).

Source of spike trains All spike trains used were already recorded extracellularly from nucleus hypothalamicus posterior 6, substantia nigra zona compacta l°, nucleus periventricularis hypothalami6, and locus coeruleus 3 in our previous electrophysiological study in chloral hydrateanesthetized, curare-immobilized and artificially ventilated rats (for details of surgical operation, recording techniques and location of recording sites, please refer to the corresponding papers3"6'l°).

TABLE I

Favored patterns in spike trains from the neurons of SNc, HPE and HP Only part of the FPs are displayed in the Table, the rest are not displayed because of the limitation of the Table size. The spike trains from neurons in the same animal were recorded 30 min apart. The number (n) of CF repetitions is equal to the number of matches in the unshuffled spike train (c) minus the maximum number in 99 shuffles (m), i.e. n = c-re. HP, nucleus hypothalamicus posterior; HPE, nucleus periventricularis hypothalami; SNc, substantia nigra zona compacta.

Exp. No. neuron No.

Firing rate (spikes~s)

Common favored patterns (CFs) Sequence of interval-length (ms)

Sequence of quantized interval-length

Number (n) of CF repetitions (times/lO00 spikes)

SNcl7-1 -2 -3 SNc24-3 -4 -5 SNc25-2 -3 -4 SNc40-1

2.50 3.29 0.93 1.88 5.07 2.82 2.91 1.12 2.66 1.72

179, 153, 140, 130, 149, 132, 181, 132, 145, 123,

185, 107, 122, 140, 139, 146, 170, 159, 130, 135,

182 105 123 116 146 137 169 166 152 130

1,1,1 1,1,1 1,1,1 1,1,I 1,1,1 1,1,1 1,1,I 1,1,1 1,1,1 1,1,1

61 63 9 34 150 28 116 20 43 31

HPE11-1 HPE76-2 HPE91-1

1.82 5.78 3.36

98, 72, 56,

19, 31, 25,

20 35 28

1,0,0 1,0,0 1,0,0

12 24 53

HP03-1 HP05-1 HP07-2 HP09-2

2.20 3.51 5.98 13.24

66, 59, 69, 57,

59, 69, 69, 68,

19 38 46 18

1,1,0 1,1,1 1,1,1 1,1,0

9 14 90 9

Quantization bin width (ms)

144

72

54

Number of different FPs in one spike train 48 25 9 20 42 35 21 2 1 35 32 26 9 30 29 75 45

299 TABLE II

The variations of FP-lengths in different spike trains Pattern repetitions (the number of pattern repetitions in unshuffled spike trains): pattern repetitions with one extra or one missing spike are included. The largest CV (coefficient of variation) of FP-length in the Table is 9.2%. HP, nucleus hypothalamicus posterior; HPE, nucleus periventricularis hypothalami; SNc, substantia nigra zona compacta.

Exp. No-neuron No Interval-lengths in template (ms)

B (times~number

A (times/min)

HP03-1 SNc17-2 HPE91-1

69,133,215 55,152,54 56,26,28

Length of pattern repetitions (measured from pattern repetitions B)

Pattern repetitions

9.3 10.1 14.1

of spikes)

3(+ S.E.M. (ms)

CV (%)

40/562 38/800 51/751

403.9 + 5.5 265.3 + 3.6 107.7 + 1.4

8.6% 8.4% 9.2%

Analyzing the properties of the FPs in spontaneous spike trains

RESULTS

1. ~( _+ S.E, and CV (coefficient of variation) of FP-length were measured. 2. Estimating the serial correlation coefficients of FP repetitions in spike trains (based on the first spike of each FP repetition) and plotting the serial correlograms.

FPs and CFs in spike trains recorded from substantia nigra zona compacta (SNc), nucleus periventricularis hypothalami (HPE) and nucleus hypothalamicus posterior (HP).

Examining the regularities of FP-changes in spike trains evoked by electroacupuncture 1. Electroacupuncture (EA). Two acupuncture needles made of stainless steel were inserted into each hind leg, one in the 'Zusanli' acupoint (S 36, 3 mmm lateral to the anterior tubercle of the tibia) and the other in the 'Sanyingao' (Sp 6, 5 mm proximal to the medial malleolus, at the posterior border of the tibia). Electrical stimulation (8 or 10 Hz, 3V) was delivered via two needles by a medical pulse stimulator (Model DM). Before recording, the stimulus artifacts of EA were inspected and then entirely eliminated by several procedures. Usually after recording a 5 min-spontaneous discharge, a 3 min-spike train was recorded every 5 min for 25-45 rain, immediately after the first 5 min-spontaneous discharge EA was delivered for 5 or 10 min. The FPs were detected and counted from these spike trains, and the number of pattern repetitions per min was used for plotting Figs.

2. The response to EA and restoration in number of different FPs and number of the representative-FPrepetitions were compared with that in firing rate. The representative FP was chosen according to two criteria: (1) having the highest or higher rate of repetition, (2) being present at the beginning and the end of the spike train (in spite of disappearance sometimes). Because of the great variation in the firing rate of evoked spike trains, it is necessary to try several bin widths and determine the optimum and normalized bin width which is applicable to detect FPs in different periods of evoked spike train.

T a b l e I shows that all the spike trains d e t e c t e d h a v e FPs and s o m e spike trains f r o m n e u r o n s in t h e s a m e nucleus s h a r e the s a m e F P (the c o m m o n FP, C F ) . A l l spike trains of SNc (n = 10) and H P E (n = 3) h a v e q u a n t i z e d CF (1,i,1

and 1,0,0 r e s p e c t i v e l y ) , while pairs o f H P

spike trains h a v e q u a n t i z e d C F (1,1,1 o r 1,1,0).

Properties o f FPs in spontaneous spike trains Table II shows t h e variations of F P - l e n g t h in d i f f e r e n t spike trains. T h e largest C V of F P - l e n g t h in the Table is 9,2%. T h e serial c o r r e l a t i o n coefficients o f F P r e p e t i t i o n s (in serial o r d e r ) in the H P - , SNc- and H P E - s p i k e trains are less t h a n 0.3 (close to z e r o ) , indicating that the n a t u r e of P R ( p a t t e r n r e p e t i t i o n ) i n t e r d e p e n d e n c e in t h e s e spike trains is a r e n e w a l process (Table III, Fig. 2).

Responses o f FPs in different spike trains to acupuncture (EA) T h e e v o k e d spike trains d e t e c t e d w e r e d i v i d e d into 3

TABLE III

The serial correlation coefficient of different spike trains Pattern repetitions (the number of pattern repetitions in unshuffled trains), pattern repetitions with one extra or one missing spike are included. All the tmax are less than t0.05 (i.e. P > 0.05), indicating that all spike trains studied are renewal processes. HP, nucleus hypothalamicus posterior; HPE, nucleus periventricularis hypothalami; SNc substantia nigra zona compacta; rrm., the minimum value of r; rmax, the maximum value of r; tr, the t value of r; tr = r ~/(n-2)l~/(1-r2); tmax the t value of rmax.

Exp. Noneuron No

Interval-lengths in template (ms)

Pattern repetitions A (times/min)

B (times~number

The absolute value of serial correlation coefficient (r) and tr (measured from pattern repetitions B)

of spikes) HP03 - 1 SNc 17-2 HPE91-1

69,133,215 55,152,54 56,26,28

9.3 10.1 14.1

40/562 38/800 51/751

rmm

rmax

tmax

to.o5

0.0025 0.0004 0.0009

0.218 0.269 0.253

1.865 1.626 1.793

2.030 2.021 2.008

300

groups, according to different changes in firing frequency induced by EA, i.e. excitatory response (increase in firing frequency), inhibitory response (decrease in firing frequency), excitatory-inhibitory response (increase followed by decrease in firing frequency). In order to guarantee the reliability of the results, only those dischargerecords which had an extremely good signal/noise ratio were chosen for detection, hence there were a few spike trains in one group. However, the evoked changes of all spike trains in each group were very similar.

1. Excitatory response group (LC neurons) (1) In different periods of the evoked spike trains, the number of various FPs and the number of repetitions of the same representative FP increased and restored along with the change of firing rate (Fig. 3). (2) The various FPs in the same spike train responded

differently. Taking the LC101-2 as an example, more than 6 types of responses were observed (Fig. 4). It is noteworthy that in addition to the excitatory response there were different types of inhibitory responses as shown in Fig. 4 A - D . A n d the 3 evoked spike trains

in this group shared the common types of responses (Fig. 4A,Ea).

2. Inhibitory response group (SNc neurons) (1) In different periods of the evoked spike trains, the number of various FPs and the number of repetitions of the same representative FP decreased and restored also along with the change of firing rate (Fig. 5). (2) The responses of various FPs in the same spike train (e.g. SNc 17-2) to EA also showed several types (Fig. 6). Notably, in Fig. 6D the quantized FP (3,1,1) was entirely inhibited during EA, and restored in a unique pattern (appearance-disappearance). And the 3 evoked spike trains in this group also shared the common type of response (Fig. 6A). 3. Excitatory-inhibitory response group (HPE and SNc neurons) (1) The firing rate increased and then decreased identically, accompanied by the corresponding changes in number of different FPs and number of repetitions of the same representative FP (Fig. 7). (2) The changes (induced by EA) of various FPs in the same spike train (e.g. HPE91-1) reveal several patterns too. New FPs emerge during EA just as in the excitatory response group, and the unique pattern of res-

15

0,6

A LC

S

(I)HP03-I

n=3

0,4

u_

o

v

5

-- 0 Ul

EIA -

10

15

20 25

30

35

40

m c

0.4I (2)SNC17-2

U q'-

-0,

o

:°0,2

IIIIIll

0

5

4vv,

~

~

10

15

n 3

20

-5

0

5

tO

15

20

25

30

35

410 (r'qln)

I Ijll[lill[

25

30

35

40

45

100

occ

(3)HPEgl~I 0.2 .^

LC

50

z Jlllllllillllllllllllllll

0,4

^M^

_

el9 ~

75

oa, e o..g

50

.... I

-0,2

~

~_ o -0,4 m

B

iIo 15 20 2L5 310 315 410 (r~,n)

(.~ -.-, +, 150

0,6

0 qU¢.~

200_

5

45

0,6

S

Q.' o.,

0

.

5

m

-5

0

[

-5

C

0

LC

5

n 3

I0 t5 20 25 30 35 Time

410

(rn,n)

-0.6 5

10

15

20 25

30

35

40

45

order

Fig. 2. The serial correlogram of favored pattern (FP) repetitions in spike trains from different nuclei. Abbreviations, see the legend of Table I.

Fig. 3. Effect of electroacupuncture (EA) on FPs in spike trains of locus coeruleus (LC) neurons (excitatory response). A: firing rate; B: number of different FPs; C: number of pattern repetitions per rain. ** P < 0.01 as compared with control, * P < 0.05 as compared with control.

301 n

ca

iO0

75

r.,

'

-5

(A) qFP(I,2,1> '

0

100,

5

~o

tO

15 20 25 TIP's' (nln)

30 35 40

(B) qFP(I, L3)

r a

o_ b_

100 F

~ ~.

t~

7.5

-5

75

0

5

tO

15 20 25 Time (mln)

40

(C) qFP(2,P~2)

v A '

30 35

n.,

0. b. o._

'

o

'

-5

0

5

75

10 15 20 25 Tit,~e (nln)

,

,

0

5 .tO

,

,

,

,

15 20 25 Tl~e (mln)

E a ~

30 35

40

(Ea~ qrP(3,1,3)

c a

100

~

(I]~ qrP(O,l,2) '

I00_

~ "

ca

~

-5

"2

"~

~_

0

n,

'~

I0

o-

Orb)

,

'

30 3'5 40

qFP(],4,4)

I

25 -5

0

5

to

t5

20

25

30 35

40

Time (mln)

"

0 -5

0.

II

. 5 . tO . . 15

' 3'5 40 ' 2 0 2'5 30 Time (mln)

Fig. 4. Different responses of various quantized favored patterns (qFP) in the same spike train (LCI01-2) to EA. A: the '+,-' ('excitationinhibition') response of the qFP(1,2,1) which exists at the beginning and the end of the spike trains (in spite of disappearance sometimes). B: the '+' response of the qFP(1,1,3) which disappears immediately after EA. C: the '-,+" response of the qFP (2,2,2). D: the '-' response to EA and its unique restoration (alternation of 'appearance-disappearance'). Ea: the qFP(3,1,3) only appears during EA. Eb: the qFP(1, 4,4) appears during EA and remains for a short period after EA.

toration (appearance-disappearance-appearance) (Fig. 8D) is similar to Fig. 6D in the inhibitory response group. The HPE91-1 and SNc40-1 spike trains also share common types of responses (Fig. 8A,D).

tions in different spike trains also alters from 9 to 150 times/1000 spikes, showing the high complexity of information communication in the central nervous system.

DISCUSSION

Properties of FPs in spontaneous spike trains

FPs and CFs in spontaneous spike trains recorded from SNc, HPE and HP Consistent with our previous study 7, all spike trains newly detected have FPs and spike trains from neurons of the same nucleus share the same FP (the common FP, CF), indicating that FP and CF in spike trains are common phenomena. The presence of CFs or CFFs (common fragment of FP, shown in our previous study 7) further proves that FPs are non-stochastic phenomena and suggests that the spike trains sharing the same FP may belong to the same kind of neurons or represent the same functional status. Notably, the number of different FPs in various spike trains varies from 1 to 75 and the number of FP repeti-

Since our previous study has measured the variations of the interval lengths in FP repetitions, the present study measured the variations of FP-lengths in spike trains recorded from different nuclei. The largest CV of FP-length (length of FP repetitions) was 9.2%. In addition, the serial correlation coefficients of FP-repetitions in these spike trains were estimated, they all were less than 0.3 (close to zero), revealing that the nature of PR (pattern repetition) interdependence in these spike trains is a renewal process.

Responses of FPs in different spike trains to electroacupuncture Considering that we have no prior knowledge of changes of FPs evoked by EA in different spike trains,

302 and in order to examine the reliability of the results obtained and to test if using FP as an index of neuronal activity may pick up more information than using the firing rate, we compared the EA-evoked changes (and restorations) in the number of different FPs and the number of the representative-FP repetitions with that in the firing rate, because the analysis of firing rate is generally used for the study of evoked spike trains and the regularity of change in firing rate induced by stimulus is known, i.e. increase in firing rate denotes excitation of the neuron, while decrease represents inhibition; moreover, the physiological significance of the changes in firing rate of the EA-evoked spike trains detected have already been investigated in our previous studies3X6*10. These known results provide the basis for further analysis of FP-changes in these evoked spike trains. The regularities of FP changes in 3 groups of EAevoked spike trains are as follows: 1. In different periods of the evoked spike trains the number of various FPs and the number of repetitions of same representative FP increase or decrease (and restore) along with the change of firing rate. It seems likely that the number of various FPs signify the number of different inputs (or outputs), while the number of FP repetitions signify the degree of each input (or output)

45,

activity. The time profiles in 3 groups of EA-evoked spike trains are different: in the excitatory response group of LC spike trains, the excitatory responses of the firing rate, the number of various FPs and the number of repetitions of the same representative FP to EA begin to decrease during EA; in the inhibitory response group of SNc spike trains, the inhibitory response to EA persists for a period after EA (i.e. there is an after effect); while in the excitatory-inhibitory response group of HPE and SNc spike trains, there is excitation during EA and inhibition after EA.

EA ce 0 VI z GB 3 x 2lti

10

2 z

5 0 Iii-u 15 -5

ki _o fr +i%

5

12 z \ 23

10

15

20

hid

(B) qFP(1.2,3)

EA

r 10

5 5

n=3

&

0 :I

-5

0

5

15

2

SNc

0

(A) qFP(l,l,l)

15 t ce 0

SNc

2 \e

r

10

15

20

EA

(Cl qFP(L1.2)

EA

CD) qFP(3,l.l)

(min)

n=3

(c 0 : -0 ?rl + x e2

Fig. 5. Effect of EA on FPs in spike trains of substantia nigra zona compacta (SNc)-neurons (inhibitory response). A: firing rate. B: number of different FPs. C: number of pattern repetitions per min. * P < 0.05 as compared with control.

^c \z VI 2 P

10 5 0 kIL..l 15 -5 0

5

10

15

20

bin)

Fig. 6. Different responses of various quantized favored patterns (qFP) in the same spike train (SNc17-2). A: the disappearance of the qFP(1.1.1) immediately after EA. B: the disappearance of the qFP(1,2,3) during EA and immediately after EA. C: the inhibition-excitation response of the qFP(1,1,2). D: the qFP(3,l.l) disappears during EA and appears sometimes afterwards.

303

2. Some FPs are present in the spontaneous spike train and still exist after g r e a t changes i n d u c e d by E A

(in spite of disappearance sometimes). These FPs may be the primary FPs of the spike train. 3. After E A , different changes in number (per min) of pattern repetitions of various FPs in the same spike train take place. The following phenomena are noteworthy: (1) In each group the evoked spike trains usually contain some FPs which reveal the c o m m o n types of changes in number (per min) of pattern repetitions. These types of FP-changes may be the characteristic FP-responses to E A in each group. (2) During E A some new FPs (absent in spontaneous discharge) emerge in the spike trains of both the excitatory response group and excitatory-inhibitory group, may be suggesting that besides the inputs (or outputs)

present in the ongoing activity increase, the new inputs (or outputs) activated by E A at the same time add to the response. (3) In addition to the primary response of excitation, there are different types of inhibition in the excitatory response group of LC spike trains, as shown in Fig. 4 A - D , indicating that EA-stimulation has dual effects (the primary excitation and the weaker inhibition) on LC neurons. It has been proved 3 that the inhibitory effect is due to the activation of the fl-endorphinergic neurons in the nucleus arcuatus hypothalami by E A , and the projection fibers of these neurons to LC release fl-endorphin

o_ LL

15

[

o

45 C -iS -iS

E A

[--1

(A)

qFP(1,1,1)

5

X 0 -5

0

5

l0

15

20

25 (mln)

7,5

O) (~ C~ L

15

-.. 21: E m v

U.L

F

~ 2,5

lO I

-5

0

5

I

I

I

i0

15

20

a,

CL -o

q.i _Q

E_ 0

5

E [ , "8

60

~

i

TM

q-

_c

0

5

I0

15

20

"~

10

£

5

o

3O

0-2I Z

0

10

15

20

85 (mln)

15

90

c

~_ 0

'

0

tn -P

'

I

25 (mln)

-5

~-

(B) qFP(I,2,1)

o

0

q.i

rat A

cz

0

(C) qFP(3,L1)

LI

-5

25 (mln)

EA

0

5

I

10

15

i

20

25 (min)

15 [

Q) -p ._ c 88-£

8

o ~ ~ o_E

4

(b ~_v E s Z

o 4s

HPE91-1

~ E A

~

O

-

2

1

0

5

I0 Time

15

20

25 (rain)

Fig. 7. Effect of EA on FPs in spike trains of nucleus periventricularis hypothalami (HPE)-neuron and SNc-neuron (excitatory-inhibitory response). A: firing rate. B: number of different FPs. C: number of pattern repetitions per min.

(D) qFP(1,3,3)

5

P -5

0 -5

~-

, _~AE A

0

5

I0 Time

15

20

25 (rain)

Fig. 8. Different responses of various quantized favored patterns (qFP) in the same spike train (HPE91-1). A: the 'excitation-inhibition' response of qFP(1,1,1). B: the inhibitory response of qFP(1, 2,1). C: the qFP(3,1,1) evoked by EA disappears immediately after EA and is restored gradually. D: the qFP(1,3,3) evoked by EA disappears immediately after EA and appears sometimes afterwards.

304 acting upon L C neurons. But in our previous study 3 (using the firing rate analysis) this inhibitory effect could only be uncovered by microinjecting fl-endorphin antiserum into L C before E A . This shows that 'the modified detection m e t h o d of F P ' can pick up m o r e (hidden) information from spike trains than the firing rate analysis. (4) A unique restoration of F P response to E A can be seen in spike trains of all 3 groups, i.e. after a p e r i o d of inhibition the F P restored in an a p p e a r a n c e - d i s a p p e a r a n c e alternative pattern, such as D in Figs. 4, 6 and 8, respectively. It m a y be speculated that the recurrent inhibition is involved in the neuronal circuit of response. In the case of SNc dopaminergic neurons, it has been r e p o r t e d that their dendrites can release d o p a m i n e ( D A ) , in turn inhibiting their cell bodies; thus possibly when SNc dopaminergic neurons re-excite after inhibition (induced by E A ) , D A (released from their dendrites) inhibits their cell bodies and such a process repeats sev-

eral times; as a result, exhibiting the unique restoration. (5) For some different FPs in the same spike train, the time profiles as in Figs. 4, 6 and 8 are similar, e.g. the spike train of LC101-2 contains quantized F P ' I , 2 , 1 ' '1, 1,1' '1,1,2' '2,1,1' whose time profiles are similar to Fig. 4A, while the time profiles of quantized FP '1,1,3' '1, 3,1' '2,1,2' are similar to Fig. 4 B . . . etc., suggesting that some informational inputs may be from the same origin or in the same active status. In summary, the different types of FP response to E A in various spike trains m a y reflect the complex situation 9 which results from interactions among numerous inputs, local circuits containing excitatory or inhibitory interneurons and self-regulation, etc. This is additional information being picked up, and provides the topics for further study.

REFERENCES

6 Ku, Y.-H. and Gao, G.-Y., The response of neurons in nucleus periventricularis hypothalami to electroacupuncture, Acupunct. Res., 6 (1981) 265-269. 7 Ku, Y.-H. and Wang, X.-Q., Favored patterns in spontaneous spike trains, Brain Res., 559 (1991) 241-248. 8 Perkel, D.H., Spike trains as carriers of information. In F.O. Schmitt (Ed.), Neurosciences Second Study Program, Rockefeller Univ. Press, New York, 1970, 587 pp. 9 Pinsker, H.M. and Willis Jr., W.D., (Eds.), Information Processing in the Nervous System, Raven, New York, 1980. 10 Yu, Y.-X., Sun, D.-Y., Gu, T. and Ku Y.-H., The effect of electroacupuncture on spontaneous discharge of neurons in pars compacta of substantia nigra, J. Beijing Med. College, 3 (1979) 149-151.

1 Dayhoff, J.E. and Gerstein, G.L., Favored patterns in spike trains. I. Detection, J. Neurophysiol., 49 (1983) 1334-1348. 2 Dayhoff, J.E. and Gerstein, G.L., Favored patterns in spike trains. II. Application, J. Neurophysiol., 49 (1983) 1349-1363. 3 Gao, Y.-S. and Ku, Y.-H., Mechanism underlying the inhibitory effect of rat nucleus arcuatus hypothalami on unit discharge of locus coeruleus with reference to electroacupuncture, Acta Physiol. Sin., 35 (1983) 163-171. 4 Geffen, L.B., Jessell, T.M., Cuello, A.C. and Iversen, L.L., Release of dopamine from dendrites in rat substantia nigra, Nature, 260 (1976) 258-260. 5 Glaser, E.M. and Ruchkin, D.S., Principles of Neurobiological Signal Analysis, Academic Press, New York, 1976.

Acknowledgement. This work was supported by a grant from the State Education Commission.

Properties of favored patterns in spontaneous spike trains and responses of favored patterns to electroacupuncture in evoked trains.

By using 'the modified detection method', our previous study has shown that all spontaneous spike trains recorded from several areas of brain and spin...
611KB Sizes 0 Downloads 0 Views