Brain Research, 595 (1992) 87-97 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00

87

BRES 18220

Long duration ventral root potentials in the neonatal rat spinal cord in vitro; the effects of ionotropic and metabotropic excitatory amino acid receptor antagonists S.W.N. Thompson

c G. Gerber

b, L . G . S i v i l o t t i a a n d C . J . W o o l f a

a Department of Anatomy and Developmental Biology, UniversityCollege London, London (UK), t, 2nd Department of Anatomy, Semmelweis University of Medicine, Budapest (Hungary) and c Neuropharmacology Laboratory, Sandoz Institute for Medical Research, London (UK)

(Accepted 9 June 1992)

Key words: N-Methyl-n-aspartate receptor; Metabotropic glutamate receptor; Slow ventral root potential; Cumulative depolarization

Long duration, primary afferent evoked ventral root potentials (VRP's) have been recorded in vitro from hemisected spinal cords prepared from 8-12-day-old rat pups. Single shock stimulation of a dorsal r¢ot at stimulus strengths sufficient to recruit C/group IV afferent fibres evoked a long duration (11.9:1:1.2 s) ipsilateral VRP in all preparations. This long duration VRP consisted of two components, (i) a slow wave, time to peak 137.0+5.1 ms, the amplitude of which was reduced to 8.7% of mean control value in the presence of the N-methyl-D-aspartate (NMDA) antagonist D-AP5(40 #,M), (ii) a prolonged wave with a time to peak of 2.0+0.2 s which was partially resistant to D-AP5(40 p,M). Both the slow and the prolonged waves were unaffected following superfusion with the metabotropic excitatory amino acid (EAA) receptor antagonist L-AP3 (100-200/~M). Low frequency (1-10 Hz) repetitive stimulation (20 s duration) of high threshold dorsal root afferents evoked a temporal summation of synaptic activitywhich generated a progressivelydepolarizing VRP. This cumulative VRP was graded with frequencyof stimulation (0.89+0.13 to 1.25+0.19 mV). The cumulative VRP was followed by a post-stimulus depolarization which outlasted the period of repetitive stimulation by tens of seconds (47.6ff8.4 to 91.2+ 19.9 s). In the presence of AP5 the amplitude of the cumulative VRP was depressed to 54.5+ 11.5% of control values when low frequency (1.0 Hz) stimulation was used. The proportion of the cumulative VRP resistant to D-AP5 increased as the frequency of stimulation was increased to 10 Hz. The decay time of the post-stimu!t~.~ depolarization was unaffected by AP5. Neither the amplitude nor the post-stimulus depolarization of the cumulative VRP was affected by 200/zM L-AP3. It t~ suggested that both an AP5 sensitive and AP5 insensitive potential contribute to the long duration VRP evoked in the neonatal rat spinal cord following single shock high threshold afferent stimulation. Moreover, the AP5 insensitive prolonged depolarization is manifest followingsustained low frequency st~.mu:,i and higher frequency inputs.

INTRODUCTION

and E A A dependent slow post-synoptic depolarizations have been demonstrated ~6'17'25'43'45'47. Moreover,

Ventral root potentials of tens of seconds duration can be recorded from immature rat spinal cords in response to high threshold afferent fibre stimulation Ln. Variously implicated in the generation of these slow ventral root potentials (VRP's) are (i) tachykinins such as substance P (SP) 28'3°'47 and (ii) agonists at the N-

recent evidence shows that following low frequency high threshold stimulation, an N M D A receptor mediated prolonged depolarization is evoked in ventral horn neurones which via a cumulative depolarization, leads to the phenomenon of windup 43. In the hippocampus it has been demonstrated that glutamate acting at the metabotropic excitatory amino acid (EAA) receptor site can evoke a slow depolarization lasting for several hundred milliseconds 4°. This effect has shown to be associated with the inhibition of two separate K + conductances, IAnp and IK(M)7. Activation of the metabotropic receptor by trans-l-amino-

methyl-D-aspartate ( N M D A ) excitatory amino acid (EAA) receptor site t2'47. Intracellular recordings from the neonatal rat spinal cord in vitro and from the adult cat in vivo have demonstrated slow post-synaptic depolarizations in superficial 49 and deep dorsal horn cells 21,32,45 and in ventral horn neurones 43. Both SP

Correspondence: S.W.N. Thompson, Neurophmmacology Laboratory, Sandoz l,.stitute for Medical Research, 5 Gower Place, London WC1E 6BN, UK. Fax: (44) (71) 387 4116.

88 Ap

1,3-cyclopentane-dicarboxylic acid (ACPD) has been shown to increase neuronal firing in the spinal cord 23. Whether any contribution to slow synaptic signalling in the spinal cord is made by glutamate acting at the metabotropic receptor site needs to be determined. In the present paper we have used the isolated spinal cord preparation from 8-12-day-old rat pups (b.w. 25-35 g) to describe the contribution of NMDA and non-NMDA EAA receptors in the generation of an ipsilateral afferent-evoked slow VRP and the subsequent prolonged depolarizations evoked by low frequency repetitive afferent stimulation. A preliminary account of this study has been presented in abstract form.

I~,I ~

,~ ,.,....' .................... "~..;7,.'r.;~,-.'-.. ~.';'..-~. ,-:;~.::--.. r;~ .~..~';'.T~.f~;~ ~

J

I,

Ii A,3

IA

, ',,

I

slow I

"\

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

MATERIALS AND METHODS Hemisected spinal cords were prepared as previously described 29 from 8-12-day-old rat pups, (b.w. 20-35 g) following urethane anaesthesia (1.8 g/kg) and coolin~ on ice/water slurry. The spinal cords were placed in a perspex incubating chamber (vol. approx. 1.0 ml) and perfused at 5 ml/min with oxygenated modified Krehs selution (in mM: NaC! 124, KCI 1.9, MgSO4 1.3, NaHCO 3 26, KH2PO 4 1.5, CaCI 2 2.0, Glucose 10.0) at room temperature (18-22°C). All drugs (Tocris Neuramin) were dissolved in this modified Krebs and held in secondary reservoirs which could be connected to the recording chamber at the same flow rate. Glass suction electrodes were used to stimulate the dorsal roo~s with constant current stimuli at distances of 10-15 mm from the dorsal root entry zone. This distance was considered adequate to prevent any undue synchronisation of the afferent volley which may occur as a result of proximal spread of stimulus current and consequent decrease in conduction distance s . Stimulation intensities used (50 #A, 50 p.s; 500 ~A, 50 ~s; 500/zA, 500 ~s) were the same as those used in a previous study with the same prcp,~ratiom 4,~. The 3 standard intensities used were associated with the successive recruitment of A /3/'group l / l l ; A ~/group Ill and C/group IV afferent fibre groups. Snug fitting glass suction electrodes placed over the ventral roots in close apposition to the ventral horn were used to record the ventral root potential (VRP). Stable DC recordings of the evoked synoptic and action potentials could be recorded for several hours by this method. Sportaneous activity was routinely observed as low amplitude, rapid deflections of the VRP. Ventral root responses were conventionally amplified and either played out on a Gould brush chart recorder or digitized using a CED 1401 interface and recorded onto PC using SCAN software (J. Dempster, Strathclyde University). All values in the text are given as mean :t: S.E.M. Statistical significance of data was assessed using a paired Studeut's t-test where appropriate.

RESULTS Fig. 1 shows an example of the VRP recorded from an L5 ventral root following electrical stimulation of the ipsilateral L5 dorsal root. The afferent evoked VRP showed progressive recruitment of several distinct waves which corresponded with the successive recruitment of afferent fibre ~,roups. In all preparations, single shock stimulation ~,~fdorsal roots at low intensity (50 ~A, 50 ~s) whicl: corresponded to the recruitment of afferent fibres conducting in the A

-

i)

ii) ,/

1.mV

!

A.

slpw

]

I

50.ms

Fig. 1. The ipsilateral ventral root potentials (VRP) recorded following single shock stimulation of the L5 dorsal root at increasing stimulus intensities which recruited higher threshold afferent fibre groups. Stimulation intensities were as follows; 50 I~A, 50 p,s (A/3),

500 izA, 50 izs (An), 500 ~A, 500 ~s (C). Peaks i) and ii) represent tt~e early phase of the response. The arrowhead gives the approxi. mate position of the time to peak of the slow peak of the late phase of the response. The time to peak of this slow response was calculated from the point of maximal depression of the wave in the presence of the specific N-methyl-o-aspartate (NMDA) receptor antagonist D-AP5.

///group 1 / I I range, resulted in tile appearance of a short duration, low amplitude VRP. This response consisted of two components; (a) a first peak, mean amplitude 1.22 :t 0.3 mV and time to peak 6.29:1:0.59 ms, and (b) a second peak, mean amplitude 0.72 :t: 0.1 mV and time to peak 20.0-t- 1.66 ms. These two peaks constituted the early response and are presumed to represent the synaptic and spiking potentials evoked through the mono- and di-synaptic pathways, respectively (see Table I). As the ~timulus was increased to an intensity sufficient to recruit afferent fibres conducting in the A a / g r o u p III range (500 IzA, 50 #s) a long duration slow VRP was evoked in addition to the early VRP. Time to peak of this slow VRP (measured from the peak depression of this wave in the presence of the EAA receptor antagonist AP5, see below) was 137.0 _+ 5.19 ms, its mean amplitude at this latency was 0.4 _+ 0.07 mV. When the stimulus intensity to the dorsal

89 TABLE I Properties of the four components of the C-fibre evoked VRP and their selective antagonism by ionotropic EAA receptor antagonists All values are mean 5: S.E.M. values in parentheses are numbers of preparations. C-evoked VRP

Time to peak ms

Control amplitude mV

AP5 % of control amplitude

CNQX amplitude mV

CNQX % of control amplitude

First rJeak Seeo,~dpeak Slowpeak Prolonged peak

6.295:0.59 (10) 20.05:1.6 (10) 137.0:!:5.1 (10) 2.035:0.2s (10)

1.225:0.3 (6) 0.725:0.1 (I0) 0.405:0.07(10) 0.205:0.04 (10)

96.15:1.1 (10) 77.05:11.5 * (10) 8.7+ 2.7 **(10) 78.75:7.2 * (10)

0.05 (3) 0.14(3) 0.08(3) -

4.3:t:0.7 (3) 18.4+6.2(3) 16.6-t-2.8(3)

* P < 0.05; * * P < 0.001.

B

A

CONTROL

501JAMdl-AP5 I!

Prolonged

f

t

............

..., ..... , ...........

. ........

,...,..,.,

.......

..* ........

,,.,.,

WASH

I .mV

1.mV

2.S I

I '

SO.ms

Fig. 2. The effect of the NMDA receptor antagonist t),L-AP5 (50 I~M) upon the early and late phases of the C-fibre evoked VRP. A: both peaks of the early response remain relatively unaffected by the antagonist (i, ii). The slow peak of the late response is substantially depressed (slow). B: control, the full duration of the C-fibre evoked response is shown. In the presence of D,x.-AP5 a prolonged wave is unmasked (prolonged) which is partially resistant to the NMDA receptor antagonist. In the presence of D-AP5 the time to peak and amplitude may be measured (arrowhead). Wash response is also shown.

90 root was sufficient to recruit afferent fibres conducting in the C/group IV afferent fibre range (500/~A, 500 /~s), a longer duration VRP was evoked (duration 11.9 + 1.2 s, n = 11; this is probably an underestimate since several responses outlasted the period of response capture and were not included in this estimate, Figs. 2 and 3). Under control conditions in the absence of blockers this long duration VRP consisted of two partially superimposed components. The time to peak and mean amplitudes of these two late components could only be resolved following application of the EAA receptor antagonist AP5. A slow VRP (time to peak, 137 ms) and a prolonged VRP, time to peak 2.03 _+0.2 s, n = 10, partially resistant to the EAA receptor antagonist AP5, are present (see below, also Table I). These two longer duration peaks constituted the late response. The components of the C-fibre evoked VRP are shown in graphical form in Fig. 3.

peak of the slow VRP could be measured from the latency of peak depression (137.0 :i: 5.1 ms). At tiffs latency the slow peak of the later response was significantly reduced to 8.7% of control values (P < 0.001, see Table i). No further reduction in peak amplitudes was recorded in preparations superfused with the antagonist for periods of up to 1.0 h. In the presence of D-AP5, stimulation of DR's at a high stimulus intensity also revealed the longer latency prolonged response which was partially resistant to the EAA antagonist. In the presence of 40/~M D-AP5 this prolonged potential had a time to peak of 2.03 _+0.2 s. The amplitude of the response measured at this latency in the presence of the antagonist represented 78.7% of that measured at a latency of 2.0 s in control medium (0.2 + 0.04 mV, Table I). In 6 preparations following 15 min application of 50 M D,L-AP5 the areas of the early response (measured between ~ and 25 ms) and the slow response (measured between 25 and 2000 ms) were compared to control values. The area of the slow C-fibre evoked VRP was significantly reduced to 27.6 + 6.8% of control (P < 0.001). The area of the early response was not affected significantly. The data from one experiment are given in Fig. 4 and the mean reduction in area of the C-fibre evoked slow peak is summarized in Table II. Bath application of the non-NMDA receptor antagonist 6-nitro-7-cyanoquinoxaline-2,3-dione (CNQX) (8 ~M) substantially and reversibly depressed the 1st and 2nd peaks of the early VRP evoked following low threshold afferent stimulation to 4.3 + 0.7% and 18.4

Effect of ionotropic EAA receptor antagonists on the VRP In all preparations the C-fibre evoked VRP was markedly depressed by the NMDA receptor antagonist AP5. The depressant effect varied for each different component of the VRP (Fig. 2) and occurred within 6 min of superfusion with 40 ttM D-AP5 or 50 ~M D,L-AP5. The amplitude of the 1st peak of the early response remained unaffected (96.1% :1: 1.1%, n = 10 of control). The 2nd peak of the early response was significantly reduced to 77.0% + 11.5% of control, n = 10 (P < 0.05). In the presence of AP5 the mean time to

~ peak i)

C-Evoked VRP amplitude

onse

peak ii)

Late ~sponse

Slow'peak

Prolongedpeak

iiiiiiii!iiiiiii!iiii!!iii!i!i

-~.'-'-~-~--"

....

~-:~'"'"'"'."-','.',"

500

" '~

"~.~'~'."

1000

" "~L2..Ld

....

: :'."

1500

".*...

2000 ms

Time

Fig. 3. Diagramatical representation of the 4 components of the VRP following single shock stimulation to the ipsilateral dorsal root. Peaks i) and ii) constitute the early phase of the reflex and are present following recruitment of low threshold A/3 afferent fibres. The slow wave occurs following recruitment of A 8 afferent fibres. The prolonged wave occurs following C-afferent fibre recruitment. The slow wave and early portion of the prolonged wave represent the late portion of the VRP.

91

Ap

Late VRP Area ( 2 5 - 2 0 0 0 m s ) Early VRP Area ( 0 - 2 5 m s )

• o

Control

/ 8 pM CNQX

AP5 1400

20

18

1200 o

16

•

o

'~ 14

oo

•°o °°o°° o

o "-"

12

'~

10

DO

1000..

°

o

•

Ooo0ooO

o o0OO

•00

•

• ° •

o Oo

o

°°

°0° o

•

•

•

• ° L

6

C o(

'800

600

• •

i

Oo

2 0

t ,5

= 10

Control

400

200

•

0

"" k

•

0

•

~

o0 o

•

io

I. 20

15

0 25

TIME (min) Fig. 4 The area of the early phase (measured 0-25 ms) and of the slow wave of the late phase (measured 25-2000 ms) of the C-fibre evoked V R P following superfusion with 50 ~M D,L-AP5 during the period shown by the bar.

:[: 6.2% of control values, respectively, n = 3 (Fig. 5). CNQX also substantially depressed the slow peak of the C-evoked VRP to 16.6 :t: 2.8% of control value ( n - 3). In the presence of CNQX the prolonged response was absent.

Effect of metabotropic EAA receptor antagonists on the VRI' The chemical compound L-AP3 has been proposed as a putative antagonist at the glutamate metabotropic site 38. In 6 preparations the effect of L-AP3 (100-200 /~M) was tested upon the high threshold-evoked VRP. Fig. 6 shows typical results from two different experi-

I

I g0ms

Fig. 5. The effect of the non-NMDA receptor antagonist CNQX (8 /zM) upon the early (A/3) and late phases (C) of the single shockevoked VRP. The slow peak of the late phase was reduced to 16.6% of control value. The prolonged peak of the late phase was absent in the presence of CNQX.

ments. No effect of L-AP3, at either concentration upon any portion of the early or late C-fibre evoked VRP was observed, either when the antagonist was applied on its own or following application of the ionotropic NMDA receptor antagonist APS.

VRP evoked following low frequency dorsal root st emulation. The effect of repeated low frequency (1-10 Hz, 20 s duration) stimulation of high threshold dorsal root

TABLE II Mean area of C-fibre evoked slow portion of the late response (measured 25-2000 ms) with percentage of control value following application of ionotropic and metabotropic EAA receptor antagonists either separately (A) or together (B)

Number of observations given in parentheses. A Control area m V" s

AP3 m V. s

% of control

AP5 m V" s

% of control

Wash

% of control

1.22+0.21 (3)

1.18+0.34 (3)

93.8+3.7 (3)

0.36+0.1 (3)

27.7+_6.5(3)

1.41 +_0.6 (3)

91.0+_8.4(3)

B Control area inV. s

AP5 mV. s

% of control . . . . . .

AP5 + AP3 mV. s . . . .

% of control

Wash

% of control

1.79+_0.3 (3)

0.74+0.1 (3)

41.2+- 1.8 (3)

0.79+_0.25(3)

41.2+_6.8(3)

1.71 +_0.3 (3)

95.7+_4.0(3)

92 afferents was examined in all preparations. Repetitive stimulation at 1-10 Hz evoked a temporal summation of synaptic activity generating a progressively depolarizing VRP whose amplitude was graded with the frequency of stimulation (Fig. 7). No comparable depolarization was evoked when dorsal roots were stimulated at a low intensity (50/.tA, 50 ~s) at frequencies up to 10 Hz. The amplitude of the cumulative VRP depolarization measured following 20 s stimulation, ranged between 0.89 _+0.13 mV at 1.0 Hz to 1.25 _+0.19 mV at 10.0 Hz. Following cessation of the repetitive stimulation, the depolarization was maintained for tens of Late VRP Area (25-2000ms)

•

:~ Early VRP Area ( 0 - 2 5 m s )

A

AP5

AP3

3000 D

>

?.~o

2000

~o • OO

•

OOO

•

_

"

lOO0"o •

•

o•O 0

•

eO

~eeOoo

- - e o o O o_ -,,, •

o" •

~ ' - ~ . o • Oo • b ~ " ~

o

o

o~ •O

(I . . . . . . . . . . . . .

10

(I

.O.. ; ........... O_.L..................L............

20

:}0

,I (I

seconds (Fig. 8). The duration of this post-stimulus depolarization ranged from 47.6 ± 8.4 s following 1.0 Hz stimulation to 91.2 ± 19.9 s following 10.0 Hz stimulation. Although the absolute duration appeared to be dependent ul.~on frequency of stimulation, half decay time was not (Fig. 8).

Effect of EA,4 receptor antagonists on the cumulative Fig. 8 shows the effect of 40 p,M D-AP5 upon the amplitude of the C-fibre evoked cumulative VRP and the post-stimulus depolarization duration. The amplitude of the cumulative VRP was significantly depressed to 54.5 _+ 11.5% of control following 1.0 Hz stimulation (P < 0.001, n = 11) and to 58.8 ± 6.9% of control following 10 Hz stimulation (P < 0.001, n - 8). Both the half decay time and the absolute duration of the post-stimulus depolarization were however unaffected following superfusion with the ionotropic antagonist. The mean absolute and percentage changes for both parameters are given in Table Ill. No change in the amplitude of the C-fibre evoked cumulative VRP measured at 20 s was observed at any frequency of stimulation following 15 min superfusion with the metabotropic EAA receptor antagonist L-AP3 (100 ~M) (Fig. 9). Nor was any change in the duration of the post-stimulus depolarization observed at any stimulation frequency following superfusion with L-AP3 (Table IV).

50

DISCUSSION

TIME (rain}

13 APII

AP5

2 0 ~-

•

--. 16~

o • tO00

%•.~

"° o

~J

600 •

O•

4 ' •

0 0

'%."

.

• 00

.................... lO 20

c gO

~ gee

30

I i

0

~

40

50

TIME (min) Fig. 6. The effect of metabotropic and ionotropic excitatory amino acid (EAA) receptor antagonists upon the VRP. A: the area of the slow wave of the late phase of the VRP (measured between 25 and 2000 ms) evoked following C-fibre stimulation in the presence of 40 ttM D-AP5 alone and together with 100 # M L-AP3. B: the area of the early phase (measured 0-25 ms) and the late phase of the C-fibre-evoked VRP following bath application of 100 /zM L-AP3 and 40 g M D-AP5 at the times shown by the bars.

In the present paper we have investigated some of the properties of the long duration C-fibre-evoked ipsilateral ventral root potential in 8-12-day-old rat spinal cords. Previous studies have investigated reflex activity in either very young ( < postnatal day 5) rat spinal cord preparations t°'tt'3t'4t or in adult hamster spinal cord preparations 2. In the neonatal rat spinal cord immediately postnatal, the central connections from primary afferent C-fibres to spinal cord cells and those between intrinsic interneurones are poorly developed 4'13. Inappropriate intersegmental ipsi- and contralateral reflexes may also be observed at early developmental stages 35 and an exuberant distribution and concentration of both peptide and amino-containing neurones and receptor sites is present 3,6'22. The distribution of SP within the rat spinal cord does not, for example, achieve the adult pattern until between postnatal days 14 and 212°'22. In the present study we have used hemisected spinal cord preparations from rats in the second week of life. At this age the neurochemical and anatomical appearance of C-fibre terminals in the

93 Cor:trol ii

i

2Hz

IHz

IOHz

40pM D-AP5

l]llIHlflllflllflfl .,, Wosh

v; Fig. 7. The cumulative VRP evoked following law frequency C-afferent fibre stimulation of the dorsal root. Each trace represents 20 s stimulation to the dorsal root at C-fibre strength at the given frequency. The amplitude of the cumulative VRP (CU-VRP) was measured 500 ms following the final stimulus. A i-iii: VRP amplitude is graded with frequency of stimulation. B i-iii: 40 v.M D-AP5 depressed but did not completely eliminate the cumulatively depolarizing VRP, especially at higher stimulation frequencies. C: recovery of CU-VRP following 30 rain wash.

dorsal horn begins to approach that of the adult t4, the physiological function of C-fibres is virtually fully established ~3't4 and dorsal horn neurones develop Cfibre-evoked action potential discharges 13.

lonotropic EAA receptor antagonists and the single shock evoked VRP We have demonstrated the presence of several distinct components of the VRP following a single shock stimulus to the corresponding ipsiiateral dorsal root. These responses are illustrated in Fig. 10 and may be distinguished by (i) the stimulus threshold required for activation, (ii) time to peak and (iii) pharmacological

profile. The early component (region 1) of the reflex consists of the 1st and 2nd peaks of the response. They are observed following both high and low threshold afferent stimulation, with a time to peak of 6.5 ms for tile first, presumably monosynaptic, peak. This early region of the VRP is resistant to NMDA receptor antagonists but is eliminated by the non-NMDA class of ionotropic EAA receptor antagonists. The insignificant reduction in the area of the early response observed following superfusion with APSe(Fig. 4) is due to an early portion of the AP5 sensitive slow component being included within this area. Alternatively an AP5 sensitive component to the monosynaptic transmission

TABLE II!

Amplitude of the C-fibre-evoked cumulative depolarization and duration of the post-stimulation depolarization measured following 20 s stimulation at 1, 2, 5 and I0 Hz Percentage of control values are given following superfusion with 40 # M D-AP5. All values are mean _+S.E.M., numbers in parentheses.

Frequency of DR stimulation (20 s duration) C-evoked CU-FRP Amplitude (mV) Duration(s)

Control 1.0 Hz 0.89+ 0.1 (11) 47.6 + 7.4 (7)

2.0 Hz 0.88+ 0.2 (8) 71.6 + 4.6(5)

5.0 Hz 0.76+ 0.1 (5) 73.3 +19.0(3)

10.0 Hz 1.25+ 0.1 (10) 91.2 +19.9 (4)

AP5 Amplitude(mV) % o f Control Duration (s) % of Control * P < 0.001; ** P < 0.05.

0.42+ 0.1 (11) 54.5 +11.5% * 56.0 + 11.4 (6) 119.1%

0.47+ 0.1 (8) 61.4 +14.9% ** 58.5 + 16.9 (4) 81.7%

-

0.70+ 0.1 (8) 58.8 + 6.95% * 95.3 + 19.9 (3) 104.5%

94 •

1.5

Control L-AP3

• Control 0 AP5

1.5

1.0 v

o.5

~"

0.0

I

I

I

I

I

2

4

6

8

10

1.0

,-.---J

12