Cannon discussed their findings mainly in relation to Hughlings

J.

Phya8ol.

(1962),

164,

pp.

1-16

1

With

8

text-figure8

Printed

in

Great

Britain

SENSORY

CONNEXIONS

TO

THE

HYPOTHALAMUS

AND

MID-

BRAIN,

AND

THEIR

ROLE

IN

THE

REFLEX

ACTIVATION

OF

THE

DEFENCE

REACTION

BY

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM*

From

the

National

Institute

for

Medical

Research,

Mill

Hill,

London,

N.

W.

7

(Received

11

October

1961)

It

has

long

been

known

that

the

behavioural

and

autonomic

responses

characteristic

of

flight

or

attack

are

integrated

by

structures

in

the

hypo-

thalamus

and

nearby

mid-brain

(Cannon

&

Britton,

1925;

Bard,

1928).

The

responses

appear

in

acutely

decerebrated

or

decorticated

cats

without

any

deliberate

stimulus

being

applied,

and

have

been

termed

'sham

rage'.

Bard

(1928)

mentioned

the

possibility

that

the

responses

might

arise

in

such

preparations

from

afferent

impulses

set

up

at

wound

edges

and

impinging

on

the

brain-stem

structures

concerned;

but

both

he

and

Cannon

discussed

their

findings

mainly

in

relation

to

Hughlings

Jackson's

(1898)

concept

that

elimination

of

higher

parts

of

the

brain

releases

the

activity

of

lower

parts.

They

thus

gave

the

impression

that

some

sponta-

neous

or

tonic

activity

of

the

brain

stem

was

being

revealed

in

these

experiments.

Hess,

whose

detailed

work

led

to

the

location

of

precise

brain-stem

areas

for

integration

of

many

basic

behavioural

patterns,

also

seemed

to

regard

responses

such

as

flight

or

attack-which

he

called

the

defence

reaction-

as

automatic

activities

of

the

hypothalamus

which

could

be

inhibited

or

facilitated

from

higher

parts

of

the

brain

(Hess

&

Briigger,

1943).

Yet

Woodworth

&

Sherrington

(1904)

had

observed

long

before

that

several

components

of

the

whole

reaction

to

nociceptive

stimulation

in

the

cat

could

be

obtained

as

a

stereotyped

response

to

stimulation

of

the

sciatic

nerve

shortly

after

pre-collicular

decerebration.

They

characterized

this

as

a

reflex

response

and

called

it

the

pseudaffective

reflex.

Recently

it

has

been

shown

that

if

section

of

the

brain

stem

is

made

at

a

slightly

higher

level,

so

as

to

spare

the

hypothalamus,

the

reflex

response

then

obtained,

even

in

an

acute

preparation,

comprises

most

of

the

features

of

the

defence

reaction

(Abrahams,

Hilton

&

Zbrozyna,

1960b).

In

the

chronic

decorti-

cate

preparation,

as

first

exemplified

by

Goltz's

(1892)

dogs,

such

reactions

persist

as

stereotyped

responses

to

given

stimuli,

quite

unchanged

so

long

*

Present

address:

Department

of

Physiology,

Aberdeen

University.

1

Physiol.

164

2

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

as

the

animal

survives.

Thus,

there

is

every

reason

to

consider

these

complex,

autonomic

and

behavioural

patterns

of

response

as

reflexes.

It

is

surprising,

therefore,

that,

following

the

discovery

of

sensory

pathways

which

impinge

upon

the

brain

stem

(Starzl,

Taylor

&

Magoun,

1951

b),

the

role

of

such

pathways

has

been

considered

mainly

in

relation

to

the

concept

of

the

reticular

activating

system;

and

their

possible

significance

as

the

afferent

pathways

of

specific

brain-stem

reflexes

has

never

yet

been

seriously

considered.

This

was

the

hypothesis,

however,

which

led

us

to

carry

out

experiments

to

see

whether

the

brain-stem

regions

which

constitute

the

integrative

centre

for

the

defence

reaction

have

the

sensory

input

necessary

for

them

to

function

as

a

reflex

centre,

in

the

usually

accepted

sense

of

this

term.

It

was

clear

from

previous

work

that

potentials

could

be

evoked

in

part

of

the

mid-brain

structures

involved,

in

response

to

nociceptive

stimula-

tion

(French,

Verzeano

&

Magoun,

1953

a)

and

flashes

of

light

(Ingvar

&

Hunter,

1955),

and

in

part

of

the

hypothalamic

structures

in

response

to

nociceptive

stimulation

(Dell,

1952;

Feldman,

van

der

Heide

&

Porter,

1959).

But

a

large

part

of

the

centre

for

the

defence

reaction

remained

to

be

explored

systematically

for

responses

evoked

by

different

sensory

modalities.

In

the

present

experiments,

we

have

found

that

responses

can

be

evoked

in

all

parts

of

the

integrative

centre

for

the

defence

reaction

to

brief

electrical

pulses

applied

to

the

superficial

radial

nerve

or

to

the

skin

itself,

to

single

flashes

of

light

to

the

eye

and

to

single

clicks.

Thus,

the

afferent

connexions

do

exist

which

would

enable

these

brain-stem

regions

to

function

as

a

reflex

centre

for

the

defence

reaction.

METHODS

Most

of

the

experiments

were

performed

on

cats

anaesthetized

with

chloralose

in

a

single

intravenous

dose

of

60

mg/kg,

after

preliminary

induction

with

ethyl

chloride

and

ether.

In

two

cats

pentobarbitone

(30

mg/kg)

was

used

instead.

The

electrodes

used

for

recording

were

placed

by

conventional

stereotactic

methods.

The

exact

electrode

positions

were

determined

after

fixation

of

the

brain

in

8itu

by

examination

of

frozen

sections

of

the

brain,

stained

by

the

Luxol

Fast

Blue

technique

of

Kluver

&

Barrera

(1953).

Metal

micro-electrodes

were

used,

with

tip

diameters

ranging

from

1

to

10

g.

They

were

made

from

tungsten

wire

or

stainless-steel

surgical

needles

by

electrolytic

erosion

(Hubel,

1957;

Green,

1958).

The

conventional

techniques

of

amplification

and

recording

were

used.

Auditory

stimuli

consisted

of

clicks,

generated

by

passing

a

square

wave

through

a

rochelle-

salt

crystal,

or

in

later

experiments,

through

a

crystal

microphone

insert.

Two

sources

of

visual

stimuli

were

used,

either

a

miniature

neon

lamp,

or

a

miniature

incandescent

lamp

(L.E.S.

12

V,

0-75

W,

Radiospares)

driven

directly

from

a

stimulator.

When

the

incandes-

cent

bulb

was

used,

the

flash

produced

was

monitored

by

a

phototransistor

placed

in

close

apposition

to

the

bulb.

The

output

from

the

phototransistor

was

displayed

on

one

beam

of

the

double-beam

oscilloscope.

In

most

experiments

the

light

source

was

mounted

close

to

one

eye,

the

pupil

of

which

had

previously

been

dilated

by

the

local

instillation

of

a

1

%

AFFERENT

PATHWAYS

FOR

DEFENCE

REFLEX

solution

of

atropine

sulphate.

When

small,

restricted

areas

of

the

retina

were

to

be

stimulated,

the

light

source

was

mounted

on

a

perimeter,

and

by

the

use

of

apertures

of

varying

dia-

meter

the

angle

subtended

by

the

light

source

could

be

controlled.

The

eye

being

illuminated

was

secured

to

a

ring

by

a

few

stitches

in

the

sclera,

and

the

other

eye

was

occluded

by

a

tin-foil

cup.

For

cutaneous

stimulation

single

shocks

were

delivered

either

to

the

superficial

radial

nerve,

or

through

a

pair

of

needles

inserted

subcutaneously.

In

some

experiments

the

action

potential

generated

in

the

superficial

radial

nerve

was

monitored

by

a

pair

of

electrodes

placed

proximal

to

the

site

of

stimulation.

When

cutaneous

stimulation

led

to

small

reflex

muscular

movements,

either

decamethonium

iodide

(100

,ug/kg)

or

gallamine

triethiodide

(Flaxedil,

3

mg/kg)

was

injected

and

the

animal

maintained

on

artificial

respiration.

In

some

experiments

the

cerebral

cortex

was

removed

under

chloralose

anaesthesia,

either

by

undercutting

with

a

scalpel

or

by

means

of

a

suction

apparatus.

In

all

these

experiments,

arterial

blood

pressure

was

continuously

monitored

by

a

Statham

strain-gauge

manometer

connected

to

one

femoral

artery.

RESULTS

Distribution

of

evoked

potentials

In

a

previous

investigation

it

had

been

concluded

that

certain

regions

of

the

hypothalamus,

central

grey

matter

and

mid-brain

tegmentum

function

as

a

reflex

centre

for

the

defence

reactioh

(Abrahams

et

al.

1960b).

These

regions

have

been

explored

in

cats

anaesthetized

with

chloralose.

All

regions

were

explored

in

eight

cats,

the

hypothalamus

alone

in

nine,

and

the

mid-brain

alone

in

three.

Electrical

potentials

were

evoked

in

all

these

regions

by

cutaneous,

auditory

and

visual

stimuli.

This

has

been

demonstrated

under

other

experimental

conditions

for

the

posterior

hypothalamus,

central

grey

matter

and

mid-brain

tegmentum

(Starzl

et

al.

1951

b;

Dell,

1952;

Ingvar

&

Hunter,

1953;

Feldman

et

al.

1959).

We

recorded,

in

addition,

responses

from

more

anterior

regions

of

the

hypothalamus,

extending

into

the

pre-

optic

region

of

the

brain

stem.

Their

distribution

is

illustrated

in

Fig.

1,

in

which

they

are

plotted

on

diagrammatic

sections

of

the

pre-optic

region

(A)

and

at

the

chiasmatic

(B)

and

tuberal

(C)

levels

of

the

hypothalamus.

The

sections

on

the

right

show,

for

comparison,

the

region

at

each

level

from

which

the

defence

reaction

is

elicited

on

electrical

stimulation.

It

can

be

seen

that

the

centre

for

the

defence

reaction

occupies

only

a

small

part

of

the

region

from

which

evoked

responses

can

be

obtained.

This

was

found

to

be

true

also

for

the

posterior

hypothalamus,

the

central

grey

matter

and

mid-brain

tegmentum.

Potentials

were

consistently

elicited

in

all

regions

explored

in

response

to

cutaneous

stimulation.

In

the

experiments

in

which

the

superficial

radial

nerve

was

stimulated

directly,

the

action

potential

was

monitored

and

it

was

found

that

the

responses

resulted

from

conduction

centrally

by

nerve

fibres

of

the

AS

group.

Responses

to

auditory

stimuli

were

only

seen

in

8

of

the

20

experiments,

1-2

3

4

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

Cut.

Aud.

Vis.

C

Cut.

Aud.

Vis.

Fig.

1.

Diagrammatic

coronal

sections

at

three

levels

of

cat's

hypothalamus

to

show

points

at

which

evoked

potentials

have

been

recorded.

A,

pre-optic

region;

B,

level

of

optic

chiasma;

C,

tuberal

region

of

hypothalamus.

Extreme

right-hand

section

shows

regions

concerned

with

defence

reaction

(from

Abrahams

et

al.

1960b),

hatched

region

being

that

from

which

responses

are

regularly

obtained

and

dots

indicating

sites

of

responses

obtained

in

individual

experiments.

The

three

panels

to

the

left

of

each

semisection

show

distribution

of

evoked

potentials

in

response

to

cutaneous

(Cut.),

auditory

(Aud.)

and

visual

stimulation

(Vis.).

CA,

anterior

commissure;

Ch,

optic

chiasma;

Fx,

fornix;

RPO,

pre-optic

area.

AFFERENT

PATH

WAYS

FOR

DEFENCE

REFLEX

but

in

these

8

experiments

the

responses

were

found

in

all

the

regions

explored.

This

suggests

that

the

absence

of

responses

in

the

remaining

experiments

was

due

to

damage

of

the

middle

ear

by

the

ear

bars

used

to

fix

the

head.

Responses

were

evoked

by

visual

stimuli

in

all

but

one

experiment.

In

any

single

experiment

responses

were

not

always

seen

in

all

the

regions

explored,

but

as

shown

in

Fig.

1,

when

the

results

from

all

20

experiments

were

put

together,

the

responses

were

distributed

throughout

the

centre

.~~~~

Fig.

2.

Potentials

evoked

by

electrical

stimulation

of

different

areas

of

skin.

Above:

diagram

indicating

site

of

cat's

skin

stimulated,

with

the

potential

evoked.

Below:

diagrammatic

coronal

section

of

cat's

brain

through

tuberal

region

of

hypothalamus

showing

position

of

recording

electrode

tip,

medial

to

fornix.

for

the

defence

reaction.

The

amplitude

of

these

potentials

depended

on

the

area

of

retina

illuminated.

An

increase

of

the

illuminated

area

from

1

to

100

was

sufficient

to

increase

the

amplitude

of

the

evoked

potential.

The

electrical

responses

obtained

by

recording

from

any

one

point

in

the

brain

stem

were

little

altered

by

changing

the

position

of

the

source

of

stimulation.

Figure

2

shows

the

responses

recorded

at

a

single

point

in

the

hypothalamus

on

electrical

stimulation

of

six

widely

separated

regions

of

5

6

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

the

skin,

on

each

of

the

extremities,

at

the

root

of

the

tail

and

between

the

scapulae.

Similarly,

when

small

restricted

regions

(10)

of

the

retina

of

one

eye

were

illuminated,

as

shown

on

the

perimeter

diagram

in

Fig.

3,

similar

evoked

responses

were

seen

at

a

single

recording

site

whichever

retinal

region

was

illuminated.

This

evidence

of

convergence

within

a

single

sensory

system

was

com-

plemented

by

evidence

of

convergence

between

systems.

The

experiments

Fig.

3.

Potentials

evoked

by

flashes

of

light,

illuminating

10

of

retina.

Above:

perimeter

diagram

with

evoked

potentials

superimposed

on

the

position

of

the

light

source.

Below:

diagrammatic

coronal

section

of

cats'

brain

through

tuberal

region

of

hypothalamus

showing

position

of

recording

electrode

tip,

lateral

to

fornix.

AFFERENT

PATHWAYS

FOR

DEFENCE

REFLEX

of

Starzl

et

al.

(1951

b),

Scheibel,

Scheibel,

Mollica

&

Moruzzi

(1955)

and

Amassian

&

Wailer

(1959)

had

shown

considerable

convergence

of

the

sensory

pathways

which

relay

into

the

mesencephalon

and

diencephalon,

and

this

was

readily

demonstrated

in

the

present

experiments.

In

four

cats

the

effect

was

examined

of

a

response

evoked

by

one

sensory

system

upon

the

response

evoked

shortly

after

by

a

different

system.

Evidence

of

interaction

between

responses

was

obtained

in

tests

carried

out

in

the

200

msec

~~~~~~~"A

Fig.

4.

Interaction

of

potentials

evoked

by

different

stimulus

modalities.

Poten-

tials

recorded

dorsal

to

mammillary

body

(as

shown

in

diagrammatic

coronal

section

on

right),

in

response

to

click

(at

arrow)

followed

by

electrical

stimulation

of

forepaw

skin

(stimulus

shown

by

artifact).

hypothalamus,

superior

colliculus,

central

grey

matter

and

adjacent

tegmentum

in

that

the

second

response

was

reduced

or

abolished

when

the

interval

between

the

stimuli

was

sufficiently

short.

In

Fig.

4

this

is

shown

for

a

response

to

cutaneous

stimulation

following

one

to

a

click.

When

skin

stimulation

follows

the

click

with

an

interval

of

340

msec

between

click

and

cutaneous

stimulation,

the

response

to

the

latter

is

just

dis-

tinguishable

above

the

background.

In

several

interaction

experiments,

when

the

second

response

was

reduced

to

this

extent,

a

brief

spike

of

short

latency

still

remained,

possibly

due

to

impulses

in

fibres

running

near

the

tip

of

the

electrode.

7

8

V.

C.

ABBAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

Characteristics

of

the

evoked

potentials

The

evoked

potentials

were

complex,

of

long

duration

and

were

con-

siderably

affected

by

the

fluctuations

of

potential

which

occurred

in

the

absence

of

deliberate

stimulation.

The

potentials

evoked

by

cutaneous

and

auditory

stimulation

were

not

consistent

in

form

from

one

animal

to

another,

neither

were

they

related

to

the

location

of

the

electrode

tip.

Potentials

evoked

by

visual

stimulation

were

more

consistent,

frequently

appearing

as

a

double-peaked

negative

wave.

A

_

B

C

D_|

A

C)1°

30

nisec

Fig.

5.

Appearance

of

evoked

potentials

after

administration

of

small

doses of

chloralose

to

a

cat

anaesthetized

with

pentobarbitone

(30

mg/kg).

Potentials

recorded

at

single

site

in

hypothalamus

in

response

to

shocks

delivered

to

forepaw

skin

(at

arrow).

Records

taken

before

(A)

and

after

i.v.

administration

of

chloral-

ose,

3-5

mg/kg

(B),

7

mg/kg

(C),

and

10-5

mg/kg

(D).

Whereas

potentials

in

lemniscal

pathways

in

the

brain

stem

are

rela-

tively

resistant

to

anaesthetics,

those

in

extralemniscal

pathways

are

readily

reduced

or

abolished

by

agents

such

as

nitrous

oxide,

ether,

pentobarbitone

and

cyclopropane

(French,

Verzeano

&

Magoun,

1953

b;

Randt,

Collins,

Davis

&

Dillon,

1958).

In

two

cats

anaesthetized

with

pentobarbitone

(30

mg/kg)

extensive

explorations

revealed

only

occasional,

small

evoked

potentials.

Further,

in

cats

under

chloralose

anaesthesia

the

potentials

were

considerably

reduced

by

intravenous

injection

of

3-4

mg/kg

of

pentobarbitone.

In

cats

anaesthetized

with

pentobarbitone,

chloralose

was

found

to

antagonize

the

depression

of

the

activity

of

the

extralemniscal

system

produced

by

the

barbiturate.

The

successive

sweeps

of

Fig.

5

show

the

gradual

appearance

of

a

response

to

cutaneous

stimulation

as

small

doses

of

chloralose

(3-5

mg/kg)

were

given

intra-

venously

to

a

cat

anaesthetized

with

pentobarbitone.

AFFERENT

PATH

WAYS

FOR

DEFENCE

REFLEX

This

is

evidence

of

a

facilitatory

effect

of

chloralose

on

the

neurones

of

the

afferent

pathway,

but,

in

addition,

these

experiments

show

one

of

the

sites

where

chloralose

exerts

a

blocking

action.

Electrical

stimulation

via

electrodes

stereotactically

placed

in

the

same

brain-stem

regions

readily

elicits

all

the

autonomic

components

of

the

defence

reaction

in

cats

under

chloralose.

Nevertheless,

even

following

intense

cutaneous

stimulation

in

such

cats

it

is

not

possible

to

obtain

the

autonomic

components

of

the

defence

reaction

reflexly

(Abrahams

et

al.

1960

b).

It

therefore

appears

that

one

of

the

sites

of

action

of

chloralose

is

at

the

synapses

between

afferent

and

efferent

pathways

in

the

brain-stem

centre

for

the

defence

reaction.

Latencies

of

evoked

potentials

In

individual

animals

differences

were

recorded

in

the

latencies

of

the

responses,

according

to

the

site

of

the

recording

electrode

and

the

sensory

system

being

activated.

The

latencies

also

depended

on

the

polarity

of

the

initial

component,

being

shorter

when

this

was

negative-going.

In

19

experiments,

in

which

latencies

were

measured

of

responses

to

electrical

stimulation

of

the

skin

of

the

fore

or

hind

limbs,

the

values

lay

between

6

and

16*5

msec.

In

most

experiments

the

latencies

tended

to

be

shorter

in

the

central

grey

matter

than

in

the

hypothalamus,

and

so

in

two

cats

a

large

number

of

observations

were

made

of

latencies

in

both

brain-stem

regions.

Only

responses

with

an

initial

negative-going

component

were

compared.

The

mean

value

for

the

latency

in

the

central

grey

matter

was

shorter

than

that

in

the

hypothalamus

by

3-9

msec

(t

=

8-696,

P

<

0-001)

in

one

experiment

and

6*1

msec

(t

=

11*05,

P

<

0.001)

in

the

other.

The

latencies

of

the

responses

evoked

in

the

hypothalamus

were,

of

course,

longer

than

those

recorded

in

the

somatic

sensory

cortex.

In

2

experiments

potentials

were

recorded

simultaneously

at

a

point

in

the

hypothalamus

and

at

a

site

on

the

precruciate

region

of

the

cerebral

cortex

where

the

shortest

latencies

were

found.

The

latency

of

the

cortical

responses

was

always

less

than

that

of

the

corresponding

hypothalamic

response,

the

mean

difference

being

4

msec

in

one

experiment

and

5

1

msec

in

the

other.

The

latencies

of

the

responses

to

auditory

stimulation

ranged

from

10

to

22

msec,

there

being

no

consistent

difference

in

the

values

recorded

in

the

various

brain-stem

regions.

The

responses

to

visual

stimulation

appeared

with

longer

latencies-from

40

to

50

msec

in

all

regions-except

for

the

hypothalamus

where

they

were

as

short

as

30

msec

in

one

experiment

and

as

long

as

68

msec

in

another.

Evoked

potentials

after

removal

of

the

cerebral

cortex

and

basal

ganglia

In

four

cats

under

chloralose

anaesthesia

decerebration

was

performed

at

a

high

level

with

the

intention

of

removing

most

of

the

brain

lying

9

10

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

dorsal

and

anterior

to

the

hypothalamus.

The

recording

electrode

was

then

inserted

into

the

various

regions

of

the

brain

stem.

In

all

experiments

cutaneous

stimulation

evoked

responses

in

the

relevant

mid-brain

regions:

that

is,

in

the

central

grey

matter

and

the

adjacent

tegmentum.

This

is

illustrated

by

the

results

shown

in

Fig.

6.

Auditory

stimulation

evoked

responses

in

these

regions

in

only

one

cat,

but

in

no

experiment

were

res-

ponses

obtained

to

visual

stimulation.

In

3

of

the

experiments

the

line

of

section

encroached

upon

the

hypothalamus,

and

evoked

potentials

were

20

msec

8

T

Fig.

6.

Responses

evoked

by

cutaneous

stimulation

(at

arrow),

recorded

at

four

points

in

the

mid-brain

of

a

cat

decerebrated

under

chloralose

anaesthesia.

GC,

central

grey

matter;

CS,

superior

colliculus.

not

recorded

in

this

region.

In

the

remaining

cat

the

hypothalamus

was

intact

and

potentials

were

evoked

in

this

part

of

the

brain

stem

by

cuta-

neous

stimulation,

but

not

by

auditory

or

visual

stimuli.

The

responses

to

cutaneous

stimulation

were

recorded

from

sites

both

in

the

tuberal

and

pre-optic

regions,

an

example

from

the

tuberal

region

being

shown

in

Fig.

7.

In

five

cats

under

chloralose

anaesthesia

the

cerebral

cortex

was

removed

in

stages

by

suction.

A

recording

electrode

had

previously

been

intro-

duced

into

the

hypothalamic

region

and

evoked

potentials

obtained.

This

AFFERENT

PATHWAYS

FOR

DEFENCE

REFLEX

electrode

was

left

in

situ

during

removal

of

the

cortex.

The

advantage

of

this

method

is

that

it

allows

the

potential

to

be

followed

through

successive

stages

of

the

decortication,

although

it

is

difficult

to

avoid

damage

to

the

brain

stem

by

movements

against

the

electrode,

particularly

when

the

cortex

in

the

vicinity

of

the

electrode

is

being

removed.

Nevertheless,

in

every

experiment

the

cutaneous

evoked

responses

survived

removal

of

the

pericruciate

area

and

the

orbitofrontal

cortex

anterior

to

it,

regions

which

include

the

primary

and

secondary

cortical

receiving

areas.

Similarly,

50

msec

100#V

Fig.

7.

Above:

potential

evoked

by

cutaneous

stimulation

in

hypothalamus

of

cat

decerebrated

under

chloralose

anaesthesia.

Below:

diagrammatic

coronal

section

through

tuberal

region

of

hypothalamus

showing

position

of

recording

electrode,

medial

to

fornix.

The

brain

dorsal

and

lateral

to

the

dotted

line

had

been

removed

before

insertion

of

the

recording

electrode.

visual

responses

survived

removal

of

the

striate

region

and

a

large

area

of

surrounding

cortex,

including

the

whole

of

the

occipital

lobe.

In

2

out

of

the

5

experiments

the

response

to

cutaneous

stimulation

survived

the

removal

of

nearly

the

whole

of

the

cerebral

cortex.

In

one

of

these

the

visual

response

also

survived

(Fig.

8):

in

this

experiment

only

the

super-

ficial

layers

of

the

cerebral

cortex

had

been

removed.

11

12

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

50

msec

Fig.

8.

Potentials

evoked

by

flash

of

light

before

and

after

almost

complete

removal

of

superficial

layers

of

cerebral

cortex

of

cat

under

chloralose

anaesthesia.

Above:

diagrammatic

coronal

section

through

tuberal

region

of

hypothalamus

showing

position

of

recording

electrode,

medial

to

fornix,

and

extent

of

removal

of

cerebral

cortex

(shown

by

dotted

line)

at

this

level.

Below:

potentials

evoked

(a)

before

and

(b)

after

decortication.

DISCUSSION

This

investigation

has

been

primarily

concerned

with

the

afferent

path-

ways

by

which

the

defence

reaction

can

be

elicited

reflexly.

The

term

'defence

reaction'

is

here

used

to

include

not

only

the

fully

developed

res-

ponses

of

flight

or

attack,

but

also

the

alerting

response;

for

the

evidence

suggests

that

the

regions

in

the

hypothalamus,

central

grey

matter

and

mid-brain

tegmentum

concerned

in

the

integration

of

all

these

responses

are

the

same;

the

responses

simply

represent

different

stages

in

a

graded

reaction.

Hess

&

Briigger

(1943)

showed

that

the

response

to

threshold

electrical

stimulation

of

these

regions

of

the

brain

is

alerting,

and

that

if

the

stimulus

is

prolonged

or

its

intensity

increased

the

reaction

progresses

to

flight

or

attack.

When

the

visceral

signs

of

the

defence

reaction

are

taken

into

account,

most

of

them

are

seen

to

be

fully

developed

during

the

AFFERENT

PATHWAYS

FOR

DEFENCE

REFLEX

early

stages

of

the

response,

when

the

only

outward

sign

is

alerting,

no

matter

whether

the

alerting

is

produced

by

direct

electrical

stimulation

of

the

brain

stem

or

by

a

stimulus

from

the

external

environment

such

as

a

loud

noise.

This

applies

not

only

to

such

features

of

the

response

as

pupil-

lary

dilatation

and

the

rise

in

arterial

blood

pressure,

which

are

not

specific

to

the

defence

reaction;

but

also

to

the

atropine-sensitive

muscle

vaso-

dilatation

that

Abrahams,

Hilton

&

Zbrozyna

(1960

a,

b)

have

shown

to

be

a

characteristic

and

invariable

component,

which

is

well

developed

during

alerting.

The

appearance

of

evoked

potentials

in

all

these

regions

in

response

to

cutaneous,

auditory

and

visual

stimuli

provides

direct

evidence

of

relays

from

all

three

systems

converging

on

the

appropriate

brain-stem

regions.

It

might

be

questioned

whether

the

widespread

distribution

of

the

evoked

potentials

arose

in

part

from

an

action

of

chloralose.

However,

when

explorations

have

been

made

in

similar

regions

of

the

cat's

brain

without

the

use

of

any

anaesthetic,

the

same

distribution

of

evoked

potentials

has

been

observed

(e.g.

Starzl

et

al.

1951

b;

Feldman

et

at.

1959).

Our

results,

therefore,

may

be

taken

to

indicate

that

the

afferent

connex-

ions

exist

which

would

enable

the

brain-stem

regions

concerned

to

act

not

simply

as

an

integrative

centre

but,

indeed,

as

a

reflex

centre

for

the

defence

reaction.

The

characteristic

features

of

the

evoked

potentials

were

their

long

latencies,

sensitivity

to

barbiturate

anaesthesia

and

persistence

after

acute

removal

of

the

cerebral

cortex.

Potentials

with

these

features

have

been

reported

previously

in

the

mid-brain

of

the

cat

and

monkey

(Starzl

et

al.

1951b;

French

et

at.

1953a,

b).

Starzl

et

al.

(1951b),

who

coined

the

term

'afferent

collateral

system'

for

the

multineurone

pathway

giving

rise

to

evoked

potentials

of

this

kind,

believed

this

to

be

an

afferent

system

which

activates

large

regions

of

the

brain

stem,

so

that

these

in

turn

can

maintain

the

cerebral

cortex

in

a

state

of

'alertness'.

Thus

the

function

of

these

parts

of

the

brain

stem

has

come

to

be

considered

mainly

in

relation

to

the

concept

of the

ascending

reticular

activating

system

(Lindsley,

Bowden

&

Magoun,

1949;

Starzl,

Taylor

&

Magoun,

1951

a).

It

was

recognized

that

these

parts

of

the

brain

stem

can

influence

motor,

auto-

nomic

and

endocrine

activity

(Magoun,

1958),

but

attention

has

been

concentrated

on

one

single

manifestation

of

their

activity.

Yet

there

is

evidence

from

several

sources,

established

over

many

years,

which

points

to

the

role

of

these

regions

as

centres

of

complex,

co-ordinated

reflex

responses,

integrating

the

autonomic

and

behavioural

patterns

of

alimen-

tary

and

sexual

reflexes,

as

well

as

of

the

defence

reaction

(Goltz,

1892;

Woodworth

&

Sherrington,

1904;

Bard

&

Rioch,

1937;

Bard,

1940).

The

reflex

centre

for

the

defence

reaction

itself

occupies

a

major

part

of

the

13

14

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

brain-stem

regions

hitherto

considered

mainly

in

relation

to

the

reticular

activating

system

(Abrahams

et

al.

1960b)

and

the

behavioural

alerting

which

results

from

activation

of

these

regions

is

readily

explained

as

a

manifestation

of

the

defence

reaction.

The

afferent

limb

of

the

uncondi-

tioned

reflex

is

thus

seen

to

lie

in

the

pathway

which

has

been

termed

the

afferent

collateral

system.

In

cats

with

extensive

mid-brain

lesions

responses

to

noxious

stimuli

are

grossly

impaired

(Hunsperger,

1956;

Sprague,

Chambers

&

Stellar,

1961).

Sprague

et

al.

(1961),

who

kept

their

cats

for

long

periods of

time

after

the

lesions

had

been

made,

found

that

their

animals

did

not

respond

to

electric

shocks,

to

inhalation

of

high

concentrations

of

ammonia

or

even

to

attack

by

dogs.

They

attributed

this

result

to

interruption

of

the

classical

sensory

(lemniscal)

pathways.

It

is

clear

from

the

diagrams

they

present,

however,

that

their

lesions

occupied

most

of

the

mid-brain

structures

which

we

have

identified

as

the

reflex

centre

for

the

defence

reaction,

which

would

have

resulted

in

the

destruction

of

a

large

part

of

the

reflex

centre

and

of

the

sensory

inflow

leading

to

it.

Conversely,

when

the

part

of

the

brain

above

the

brain-stem

centre

for

the

defence

reaction

regions

has

been

removed,

all,

or

at

least

many,

of

the

features

of

the

defence

reaction

are

regularly

obtained

as

stereotyped

reflex

responses.

These

are

seen

on

mild

cutaneous

stimulation

of

the

chronic

decorticate

or

decerebrate

cat

or

dog

(Goltz,

1892;

Dusser

de

Barenne,

1920;

Keller,

1932;

Bard

&

Rioch,

1937),

and

in

response

to

a

loud

noise,

even

in

the

chronic

mid-brain

cat

(Bard

&

Macht,

1958).

Consonant

with

these

earlier

observations,

the

potentials

evoked

by

activation

of

these

sensory

systems

were

found

to

converge

on

the

appro-

priate

parts

of

the

brain

stem

and

to

survive

decortication

or

high

decere-

bration.

They

were

obtained

in

the

hypothalamus

inresponse

to

cutaneous

stimulation

when

all

the

brain

above

had

been

removed,

showing

that

the

cutaneous

afferent

pathways

were

intact

and

functioning.

When

the

line

of

section

encroached

on

the

hypothalamus,

evoked

potentials

were

no

longer

obtained.

This

failure

of

conduction

in

the

afferent

pathway

could

explain

why

Abrahams

et

al.

(1960b)

were

able

to

obtain

active

muscle

vasodilata-

tion

in

only

a

proportion

of

their

acute

decerebrate

cats,

when

eliciting

the

pseudaffective

reflex

response

to

peripheral

nerve

stimulation.

The

defence

reactions

of

chronic

decorticate

cats

and

dogs

are

a

striking

feature

of

their

response

to

gentle

handling

(Goltz

1892;

Dusser

de

Barenne,

1920).

From

Goltz's

description

of

his

decorticate

dogs,

for

instance,

their

response

to

such

a

trivial

stimulus

as

stroking

the

skin

is

clearly

identical

with

the

fully

developed

defence

reaction.

This

response

was

elicited

with

undiminished

vigour,

day

after

day,

as

long

as

the

animals

lived.

Thus,

in

the

decorticate

animals

in

which

the

reflex

mechanism

underlying

the

AFFERENT

PATHWAYS

FOR

DEFENCE

REFLEX

defence

reaction,

including

the

collateral

afferent

pathway,

is

intact,

but

the

projections

of

the

classical

sensory

pathways

are

removed,

defence

reflexes

are

continually

produced,

as

stereotyped

responses,

even

to

hardly

noxious

stimuli.

A

characteristic

feature

of

the

decorticate

preparation,

therefore,

is

the

absence

of

inhibition

as

normally

manifested

in

the

phenomenon

of

habituation.

In

the

normal

cat

the

alerting

produced

by

any

sudden

stimulus,

such

as

the

sound

of

a

buzzer,

soon

disappears

with

repetition

of

the

stimulus.

This

is

true

also

for

the

defence

reactions

in

response

to

weak

noxious

stimuli.

Here

we

can

see

a

possible

significance

of

the

slowness

of

conduction

in

the

collateral

afferent

system,

reflected

in

the

long

latencies

of

the

evoked

responses,

as

compared

with

the

much

shorter

latencies

of

the

cortical

evoked

responses.

The

inhibitory

pathway

which

underlies

the

phenomenon

of

habituation

involves

the

cerebral

cortex.

Nevertheless,

impulses

set

up

in

this

inhibitory

pathway

must

arrive

at

the

reflex

centre,

and

so

depress

its

excitability,

before

impulses

travelling

in

the

multineurone

pathway

impinge

on

the

centre.

It

is

possible

that

transmission

is

inhibited

in

the

multi-neurone

pathway

itself.

SUMMARY

1.

In

cats

anaesthetized

with

chloralose,

cutaneous,

auditory

and

visual

stimuli

evoke

potentials

in

widespread

regions

of

the

hypothalamus,

central

grey

matter

and

mid-brain

tegmentum.

2.

These

potentials

appear

after

relatively

long

latencies.

They

are

reduced

or

abolished

by

small

amounts

of

pentobarbitone,

but

survive

removal

of

the

cerebral

cortex.

3.

Convergence

of

the

afferent

pathways

is

indicated

by

considerable

interactions

obtained

between

potentials

evoked

by

different

sensory

systems.

There

is

also

convergence

within

the

afferent

pathways

of

a

single

system.

4.

The

multineurone

pathway

giving

rise

to

these

potentials

is

apparently

identical

with

that

originally

termed

the

afferent

collateral

system,

which

is

usually

considered

in

relation

to

the

concept

of

the

ascending

reticular

activating

system.

It

is

suggested,

however,

that

the

significance

of

this

pathway

lies

chiefly

in

connexion

with

complex

reflexes

organized

at

the

level

of

the

hypothalamus

and

mid-brain,

such

as

the

defence

reaction,

and

that

it

constitutes

the

afferent

limb

of

the

un-

conditioned

reflex.

REFERENCES

ABRAHAMS,

V.

C.,

HiLTON,

S.

M.

&

ZBROZYNA,

A.

(1960a).

Reflex

activation

of

vasodilator

nerve

fibres

to

skeletal

muscle

in

decerebrate

and

intact

cats.

J.

Phy8iol.

152,

54-5p.

ABRAHAMS,

V.

C.,

HITON,

S.

M.

&

ZBROZYNA,

A.

(1960b).

Active

muscle

vasodilatation

produced

by

stimulation

of

the

brain

stem:

its

significance

in

the

defence

reaction.

J.

Phy8iol.

154,

491-513.

15

16

V.

C.

ABRAHAMS,

S.

M.

HILTON

AND

J.

L.

MALCOLM

AMIASSAN,

V.

E.

&

WALLER,

H.

J.

(1959).

Spatiotemporal

patterns

of

activity

in

individual

reticular

neurons.

In

Reticular

Formation

of

the

Brain,

pp.

69-108.

London:

Churchill.

BARD,

P.

(1928).

A

diencephalic

mechanism

for

the

expression

of

rage

with

special

reference

to

the

sympathetic

nervous

sytem.

Amer.

J.

Phyaiol.

84,

490-515.

BARD,

P.

(1940).

The

hypothalamus

and

sexual

behaviour.

Res.

Publ.

nerv.

ment.

Di8.

20,

551-579.

BARD,

P.

&

MACHT,

M.

B.

(1958).

The

behaviour

of

chronically

decerebrate

cats.

In

Neurological

Basis

of

Behaviour,

pp.

57-71.

London:

Churchill.

BARD,

P.

&

RIocH,

D.

McK.

(1937).

A

study

of

four

cats

deprived

of

neocortex

and

addi-

tional

portions

of

the

forebrain.

Johns

Hopk.

Hosp.

Bull.

60,

65-125.

CANNON,

W.

B.

&

BRITTON,

S.

W.

(1925).

Studies

on

the

conditions

of

activity

in

endocrine

glands;

pseudoaffective

medulliadrenal

secretion.

Amer.

J.

Physiol.

72,

283-294.

DELL,

P.

(1952).

Correlations

entre

le

systeme

vegetatif

et

le

systeme

de

la

vie

de

relation.

Mesencephale,

dienc6phale

et

cortex

cerebrale

(1).

J.

Physiol.

Path.

gen.

44,

471-557.

DusSER

DE

BARENNE,

J.

G.

(1920).

Recherches

experimentales

sur

les

fonctions

du

syst6me

nerveux

central,

faites

en

particulier

sur

deux

chats

dont

neopallium

avait

et6

enlev6.

Arch.

neerl.

Physiol.

4,

31-123.

FELDMAN,

S.,

VAN

DER

HEIDE,

C.

S.

&

PORTER,

R.

W.

(1959).

Evoked

potentials

in

the

hypothalamus.

Amer.

J.

Physiol.

196,

1163-1167.

FRENCH,

J.

D.,

VERZEANO,

M.

&

MAGOUN,

H.

W.

(1953a).

An

extralemniscal

sensory

system

in

the

brain.

Arch.

Neurol.

Chicago,

69,

505-518.

FRENCH,

J.

D.,

VERZEANO,

M.

&

MAGOUN,

H.

W.

(1953b).

A

neural

basis

for

the

anaesthetic

state.

Arch.

Neurol.

Chicago,

69,

519-529.

GOLTZ,

F.

M.

(1892).

Der

Hund

ohne

Grosshirn.

Pflug.

Arch.

ges.

Physiol.

51,

570-614.

GREEN,

J.

D.

(1958).

A

simple

micro-electrode

for

recording

from

the

central

nervous

system.

Nature,

Lond.,

182,

962.

HESS,

W.

R.

&

BRUGGER,

M.

(1943).

Das

subcorticale

Zentrum

der

affectiven

Abwehr-

reaktion.

Helv.

physiol.

acta,

1,

33-52.

HUBEL,

D.

H.

(1957).

Tungsten

microelectrode

for

recording

from

single

units.

Science,

125,

549-550.

HuNSPERGER,

R.

W.

(1956).

Affektreaktionen

auf

elektrische

Reizung

im

Hirnstamm

der

Katze.

Helv.

physiol.

acta,

14,

70-92.

INGVAR,

D.

H.

&

HUNTER,

J.

(1953).

Influence

of

visual

cortex

on

light

impulses

in

the

brain

stem

of

the

anaesthetised

cat.

Acta

physiol.

scand.

33,

194-218.

JACKSON,

J.

H.

(1898).

Remarks

on

the

relations

of

different

divisions

of

the

central

nervous

system

to

one

another

and

to

parts

of

the

body.

Brit.

med.

J.

i,

65-69.

KELLER,

A.

D.

(1932).

Autonomic

discharges

elicited

by

physiological

stimuli

in

mid-brain

preparations.

Amer.

J.

Physiol.

100,

576-586.

KLUVER,

H.

&

BARRERA,

E.

(1953).

A

method

for

the

combined

staining

of

cells

and

fibres

in

the

nervous

system.

J.

Neuropath,

12,

400-403.

LINDSLEY,

D.

B.,

BOWDEN,

J.

&

MAGOUN,

H.

W.

(1949).

Effect

upon

the

EEG

of

acute

injury

to

the

brain

stem

activating

system.

Electroenceph.

cdin.

Neurophysiol.

1,

475-486.

MAGOUN,

H.

W.

(1958).

The

Waking

Brain,

Springfield:

Thomas.

RANDT,

C.

T.,

COLLINS,

W.

F.,

DAVIS,

H.

S.

&

DILLON,

W.

N.

(1958).

Differential

suscepti-

bility

of

afferent

pathways

to

anaesthetic

agents

in

the

cat.

Amer.

J.

Physiol.

192,

305-310.

SCHEIBEL,

M.,

SCHEIBEL,

A.,

MOLLICA,

A.

&

MoRuzzI,

G.

(1955).

Convergence

and

inter-

action

of

afferent

impulses

on

single

units

of

reticular

formation.

J.

Neurophysiol.

18,

309-331.

SPRAGUE,

J.

M.,

CHAMBERS,

W.

W.

&

STELLAR,

E.

(1961).

Attentive,

affective

and

adaptive

behaviour

in

the

cat.

Science,

133,

165-173.

STARZL,

T.

E.,

TAYLOR,

C.

W.

&

MAGOUN,

H.

W.

(1951a).

Ascending

conduction

in

reticular

activating

system,

with

special

reference

to

the

diencephalon.

J.

Neurophysiol.

14,

461-477.

STARZL,

T.

E.,

TAYLOR,

C.

W.

&

MAGOUN,

H.

W.

(1951b).

Collateral

afferent

excitation

of

reticular

formation

of

brain

stem.

J.

Neurophysiol.

14,

479-496.

WOODWORTH,

R.

S.

&

SHERRINGTON,

C.

S.

(1904).

A

pseudaffective

reflex

and

its

spinal

path.

J.

Physiol.

31,

234-243.