Make-up of spinal cord circuits which process inputs from the femoral-saphenous vein

MISSING IMAGE

Material Information

Title:
Make-up of spinal cord circuits which process inputs from the femoral-saphenous vein
Physical Description:
ix, 116 leaves : ill. ; 29 cm.
Language:
English
Creator:
Yates, Billy J., 1960-
Publication Date:

Subjects

Subjects / Keywords:
Afferent Pathways   ( mesh )
Neural Pathways   ( mesh )
Saphenous Vein -- innervation   ( mesh )
Neuroscience thesis Ph.D   ( mesh )
Dissertations, Academic -- Neuroscience -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D)--University of Florida, 1986.
Bibliography:
Bibliography: leaves 106-115.
Statement of Responsibility:
by Billy J. Yates.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000578651
oclc - 16989812
notis - ADA6431
System ID:
AA00011760:00001


This item is only available as the following downloads:


Full Text















THE MAKE-UP OF SPINAL CORD CIRCUITS
WHICH PROCESS INPUTS FROM
THE FEMORAL-SAPHENOUS VEIN






By

BILLY J. YATES


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY




UNIVERSITY OF FLORIDA


1986
















ACKNOWLEDGEMENTS


The studies presented in this dissertation could not have been

performed without the guidance, assistance and encouragement of a

small army of individuals. Much of the guidance and encouragement

came from the members of my committee: Floyd Thompson, John Munson,

Chuck Vierck and Paul Davenport. Special thanks go to Floyd, from

whom I learned many electrophysiological techniques and a great deal

of philosophy. I also could not have completed this work without the

advice concerning neuroanatomical techniques that I obtained from

Parker Mickle, Paul Reier, John Houle, Celeste Wirsig and Roger Reep.

I thank Alan Light for his suggestions regarding the staining of

neurons. I wish to express my utmost gratitude to Deb Dalziel, Bill

Hedden, Albis Acosta, Dan Orlando, Marty Yungmann, Linda Cheshire,

Josepha Cheong, Shoobha Daftary and Ellen Grygotis for superb

technical assistance.

Data were analyzed using the Statistical Analysis System (SAS,

SAS Institute Inc., Cary, North Carolina). Computing was done using

the facilities of the Northeast Regional Data Center of the State

University System of Florida.

Financial support was provided by NIH grant R01 HL 25619, U.S.

Air Force Aerospace Medicine Contract F33615-82-D-0627, the

Department of Neuroscience, the Department of Neurological Surgery,

the Division of Sponsored Research and the Center for Neurobiological

Sciences.



















TABLE OF CONTENTS


Page

ACKNOWLEDGEMENTS ..................................... ii

LIST OF TABLES ........................................ v

LIST OF FIGURES ....................................... vi

ABSTRACT ............................................. viii

CHAPTERS

I GENERAL INTRODUCTION ........................... 1

II TRACING OF AFFERENT PATHWAYS FROM THE FEMORAL-
SAPHENOUS VEIN TO THE DORSAL ROOT GANGLIA USING
TRANSPORT OF HORSERADISH PEROXIDASE ........... 5

Introduction ..................... ................ 5
Methods and Materials ........................... 6
Results ..*........ ............................... 9
Discussion ....................................... 19

III PROPERTIES OF SPINAL CORD PROCESSING OF FEMORAL-
SAPHENOUS VENOUS AFFERENT INPUT REVEALED BY
ANALYSIS OF EVOKED POTENTIALS ................. 25

Introduction ................................... 25
Methods and Materials ............................ 26
Results .................. ........... ........ ... 33
Discussion ...................... ..... ............. 41

IV PHYSIOLOGICAL PROPERTIES OF SINGLE SPINAL CORD
NEURONS ACTIVATED BY STIMULATION OF THE
FEMORAL-SAPHENOUS VEIN ........................ 44

Introduction ................................... 44
Methods and Materials ............................ 44
Results ............................... ...... ... 49
Discussion ....................................... 65











V THE VARIABILITY IN THE RESPONSES OF NEURONS
LOCATED IN DIFFERENT LAMINAE FOLLOWING
STIMULATION OF THE FEMORAL-SAPHENOUS VEIN ..... 70

Introduction ................. .................. 70
Methods and Materials ........................... 70
Results ... .... ................... ........... .. 72
Discussion ....................................... 92

VI GENERAL DISCUSSION ............................ 101

REFERENCES .............................................. 106

BIOGRAPHICAL SKETCH .................................. 116


















LIST OF TABLES


Page

2-1 Locations of cell bodies in the dorsal root
ganglia giving rise to femoral-saphenous
venous afferents ................................ 13

2-2 Conduction velocities of primary femoral-
saphenous venous afferents ....................... 23

4-1 Responses of spinal neurons receiving inputs
from the femoral-saphenous vein .................. 53

5-1 Spontaneous firing rates of neurons located in
different laminae ................................ 78

5-2 Locations of neurons receiving convergent inputs
from the femoral-saphenous vein, muscle and skin .. 91


















LIST OF FIGURES


Page

2-1 Cell bodies in the dorsal root ganglia labeled
by the application of HRP to the femoral-
saphenous vein ................. ............. 11

2-2 Diameters of cell bodies in the dorsal root
ganglia of the kitten giving rise to venous
afferents .................................... .... 16

2-3 Diameters of cell bodies in the dorsal root
ganglia of the adult cat giving rise to venous
afferents ......................................... 18

3-1 Minimal spinal cord processing time for inputs
from muscle, skin and the femoral-saphenous vein .. 35

3-2 Potentials recorded along a typical electrode
track through L6 following stimulation of the
femoral-saphenous vein ............................ 38

3-3 The location of the focus of the short latency
negative waves elicited by stimulation of the
femoral-saphenous vein ........................... 40

4-1 The effects of stimulation of the femoral-saphenous
vein on the excitabilities of interneurons ........ 51

4-2 Effects of stimulation of muscle, cutaneous and
venous afferents on the excitability of a single
neuron ........................ ............ 59

4-3 The response amplitude versus stimulus intensity
relationship for cutaneous afferents .............. 63

5-1 Examples of recording sites marked through the
extracellular iontophoresis of HRP ................ 74

5-2 Locations of sites at which responses were recorded
from venous afferent-activated interneurons ....... 77










5-3 Locations of units excited or both excited and
inhibited by stimulation of femoral-saphenous
venous afferents ..... ........................ ....... 80

5-4 Procedures used to determine whether a unit is
activated at monosynaptic latencies by primary
femoral-saphenous venous afferents ............... 83

5-5 Locations of units which were activated at
different minimal latencies following stimulation
of the femoral-saphenous vein ..................... 86

5-6 Locations of units activated for different
durations following the stimulation of the femoral-
saphenous vein ................................... 89

5-7 Reconstructions of venous afferent-activated
interneurons that were intracellularly stained
using HRP ......................................... 94



















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



THE MAKE-UP OF SPINAL CORD CIRCUITS
WHICH PROCESS INPUTS FROM
THE FEMORAL-SAPHENOUS VEIN


By


BILLY J. YATES


August 1986


Chairman: Floyd J. Thompson
Major Department: Neuroscience


Several characteristics of the afferents which conduct inputs

from the femoral-saphenous vein to the spinal cord and the neurons

within the cord which process these inputs were described.

Experiments using the retrograde transport of horseradish peroxidase

(HRP) from the vein to the dorsal root ganglia were performed to

determine the number, distribution and sizes of the cell bodies from

which the femoral-saphenous venous afferents originate. These

experiments showed that afferents arising along the entire length of

the vein project to very localized spinal levels and that these

afferents are few in number. Most of the cell bodies labeled by the

application of HRP to the femoral-saphenous vein were small in size


viii










(diameter less than 35 micrometers). However, some large cell bodies

(diameter greater than 50 micrometers) were also noted. It was

estimated that over 70% of the femoral-saphenous venous afferents are

C fibers; at least 13% were estimated to be A fibers. The largest

venous afferents were predicted to conduct action potentials at

approximately 60 m/sec.

The minimal spinal cord processing time for inputs elicited by

stimulating large fibers arising in the femoral-saphenous vein was

approximately the same as for inputs from skin; thus, di- or tri-

synaptic pathways appeared to be the shortest circuits separating the

primary afferents and motoneurons. Mappings of field potentials as

well as the recording of single unit activity from reconstructed

sites suggested that most of the first neurons activated by this

afferent input are found in lamina V. Thus, it is likely that many

of the large primary femoral-saphenous venous afferents terminate in

this region.

Recordings were done from single spinal neurons driven by large

venous afferents. The interneurons characterized were excited or

both excited and inhibited for long durations following stimulation

of the femoral-saphenous vein. These neurons were located

predominantly in laminae V and VII. Most neurons could also be

activated by stimulation of large muscle and cutaneous afferents.

Some of the neurons were intracellularly stained; the reconstructed

cells had large cell bodies and extensive dendritic fields.


















CHAPTER I
GENERAL INTRODUCTION

The study of the neural substrate of relatively simple reflex

pathways allows an investigator to gain insight into how the central

nervous system organizes and coordinates motor behavior. A spinal

reflex which appears to be produced by relatively simple central

nervous system circuitry is elicited when receptors in the walls of

the femoral-saphenous vein are stimulated (Thompson and Barnes, 1979;

Thompson and Yates, 1983, 1984, 1986; Thompson et al., 1982, 1983).

The femoral-saphenous venous afferents include fibers excited by

intravenous perfusion pressures as low as 2-3 mm Hg (Thompson et al.,

1983). Electromyographic recordings following low-threshold

electrical stimulation of these afferents showed that they can elicit

the simultaneous contraction of flexor and extensor hindlimb muscles

(Thompson et al., 1982). It is interesting that both flexors and

extensors are simultaneously excited by stimulation of afferents

arising from the femoral-saphenous vein, as other known hindlimb

reflexes involve the reciprocal facilitation and inhibition of

motoneurons. Stimulation of cutaneous afferents or other

exteroceptors generally results in flexor facilitation and extensor

inhibition (Sherrington, 1910; Lloyd, 1943a; Hagbarth, 1952).

Activation of muscle afferents typically produces the reciprocal

facilitation and inhibition of agonists and antagonists (Lloyd,











1943a; Laporte and Lloyd, 1952; Matthews, 1969). The unique

segmental reflex elicited by venous afferent stimulation has been

suggested to provide a rapid means of controlling the counterpressure

applied to the high-capacitance intramuscular veins; changes in this

counterpressure are potentially important in compensating for

orthostatic blood shifts (Thompson and Yates, 1983, 1984, 1986).

It is well established that other inputs from the internal

environment, in addition to those from the femoral-saphenous vein,

converge with somatic inputs in the spinal cord on interneurons that

are part of somatic reflex pathways (Downman, 1955; Evans and

McPherson, 1958, 1959, 1960; Evans, 1963). Spinal neurons which send

their axons into ascending tracts, including the postsynaptic dorsal

column pathway (Rigamonti and Hancock, 1974; Rigamonti et al., 1978),

the spinocervical pathway (Rigamonti and Michelle, 1977) and the

spinoreticular and spinothalamic pathways (Hancock et al., 1975;

Foreman and Weber, 1980; Blair et al., 1981, 1984; Milne et al.,

1981; Rucker and Holloway, 1982; Ammons et al., 1984; Rucker et al.,

1984), also receive visceral inputs. These inputs arise in part from

Pacinian corpuscles that sometimes discharge in phase with

cardiovascular pulsations (Gammon and Bronk, 1935), thermoreceptors

that are located at least in part along blood vessels (Thompson and

Barnes, 1970; Rawson and Quick, 1972; Riedel, 1976), vascular

receptors that respond to visceral movement or distention (Bessou and

Perl, 1960; Floyd and Morrison, 1974; Floyd et al., 1976),

proprioceptors found in the intercostal muscles and in the diaphragm











(Yamamoto et al., 1960; Remmers and Tsiaras, 1973; Davenport et al.,

1985; Marlot et al., 1985), mechanoreceptors located in the vagina

(Komisaruk and Wallman, 1977; Henry, 1983) and chemoreceptors

innervating blood vessels (Moore and Moore, 1933; Lim et al., 1962;

Potter et al., 1962; Besson et al., 1972). It would be of great

interest to compare the spinal cord circuits which process inputs

from the femoral-saphenous vein with the circuits that process other

inputs from the internal environment.

The present study examined several of the characteristics of the

afferents which conduct inputs from the femoral-saphenous vein to the

spinal cord and the neurons within the cord which process these

inputs. Experiments using the retrograde transport of horseradish

peroxidase from the vein to the dorsal root ganglia were performed to

determine the number, distribution and sizes of the cell bodies from

which the femoral-saphenous venous afferents originate. The results

of these experiments are described in Chapter II. In addition, field

potential mappings were performed to determine the location of the

first interneurons excited by the venous afferents. These mappings

also demonstrated the shortest pathway between the primary afferents

and motoneurons; this information is presented in Chapter III. The

electrophysiological properties of the venous afferent-activated

interneurons are discussed in Chapter IV. Interneurons with

different locations in the cord were shown to be affected by venous

afferent stimulation in different ways; the variability in how the

excitabilities of neurons in the different laminae of Rexed were

altered by vein stimulation is presented in Chapter V. Finally,








4


Chapter VI considers the hypothesis that the characteristics of the

spinal cord circuitry which processes inputs from the

femoral-saphenous vein are similar to the characteristics of the

circuitry, described in previous studies, which processes other

inputs from the internal environment.


















CHAPTER II
TRACING OF AFFERENT PATHWAYS FROM THE FEMORAL-SAPHENOUS VEIN TO
THE DORSAL ROOT GANGLIA USING TRANSPORT OF HORSERADISH PEROXIDASE

Introduction

In the mammal, intracranial (Davis and Dostrovsky, 1985; Keller

et al., 1985; McMahon et al., 1985; Norregaard and Moskowitz, 1985),

extracranial (McMahon et al., 1985; Norregaard and Moskowitz, 1985),

and visceral (Bessou and Perl, 1960; Lim et al., 1962; Floyd and

Morrison, 1974; Floyd et al., 1976; Guilbaud et al., 1977; Vance and

Bowker, 1983; Barja and Mathison, 1984) blood vessels receive an

extensive afferent innervation. Anatomical studies have also

demonstrated the presence of large, myelinated fibers coursing along

the walls of large veins of the hindlimb and into adjacent nerve

trunks (Woollard, 1926; Hinsey, 1928; Truex, 1936; Polley, 1955).

These observations are further supported by experiments in which

electrical stimulation of carefully isolated segments of the

femoral-saphenous vein elicited (i) compound action potentials

recordable from the femoral nerve and dorsal root fibers entering the

lower lumbar spinal cord (Thompson and Barnes, 1979; Thompson and

Yates, 1986), (ii) evoked field potentials recordable from the lumbar

cord (Thompson and Barnes, 1979; Thompson and Yates, 1986), (iii)

changes in firing patterns recordable from single spinal motoneurons

(Thompson et al., 1982) and (iv) electromyographic activity

recordable from hindlimb muscles (Thompson et al., 1982). The











injection of fluid pulses into the femoral-saphenous vein or

longitudinal stretch of the vein wall similarly elicited field

potentials in the lumbar spinal cord and activity recordable from

ventral roots (Thompson et al., 1983).

The study presented in this chapter was performed in order to

gain insight concerning the projection pattern of these

femoral-saphenous venous afferents to the spinal cord. The

retrograde transport of horseradish peroxidase (HRP) from the

femoral-saphenous vein to the dorsal root ganglia was used to

determine the range of afferent sizes, the relative numbers of these

afferents, and the spinal cord segments to which they project.

Methods and Materials

HRP Application Procedures

Experiments were conducted on kittens (5-6 weeks of age) and

adult cats (weighing 2.5-3.8 kg). Animals were anesthetized using an

intraperitoneal (ip) injection of sodium pentobarbital (Butler, 38

mg/kg) or an intramuscular (im) injection of Ketaset (ketamine

hydrochloride, Bristol, 10 mg/kg) combined with Rompun (xylazine,

Haver-Lockhart, 0.5 mg/kg). Maintenance doses of sodium

pentobarbital (12 mg/kg) or Ketaset (5 mg/kg) were delivered as

necessary to maintain areflexia during the time that tissues were in

contact with HRP.

The skin of either of the hindlimbs was opened ventrally to

expose the femoral-saphenous vein and underlying muscle. In 3

kittens and 9 adult cats, one to three 10-15 mm segments of the vein











were separated from surrounding tissues and placed on rubber dams. A

segment of the proximal femoral vein (overlying the vastus medialis

and adductor femoris muscles) was always prepared for HRP

application; other HRP application sites were located distally and

were spaced 3-6 cm apart. The tissues adjacent to the isolated vein

segments were covered with a thick layer of petroleum jelly to

prevent spread of the tracer. Uptake of HRP was promoted by making

small tears in the outer vein wall with sharp forceps so that fibers

coursing there would be damaged. A 30% solution of Sigma Type VI HRP

in 2% dimethylsulfoxide was applied to the prepared vein segments;

after 3-4 hrs the enzyme was removed from the vein and the

application sites were irrigated with saline. Incisions were closed

using wound clips and the animals were returned to their cages and

allowed to recover.

As a control for these experiments, different HRP application

sites were used in 2 adult cats. In these animals segments of the

femoral-saphenous vein were separated from the adjacent muscle as

described above. However, the HRP solution was applied to the muscle

which lay directly underneath the 10-15 mm segment of vein that had

been dissected away; these tissues were the most likely to be

accidentally exposed to HRP during those experiments in which the

tracer was applied to segments of the femoral-saphenous vein. In one

experiment 2 application sites located 31 mm apart were used; in the

other 3 application sites separated by 20 and 22 mm were used. The

distribution of cells in the dorsal root ganglia labeled by this

procedure was compared with that of cells labeled after the enzyme

was applied to the femoral-saphenous vein.











Histological Procedures

After a recovery period of 50-60 hrs in kittens and 90-96 hrs in

adult cats, the animals were reanesthetized; kittens were

transcardially perfused with 250-500 ml of chilled saline containing

0.1% sodium nitrite and 1.5 units of heparin per ml followed by 2 1

of chilled 1.25% glutaraldehyde/0.6% paraformaldehyde fixative.

After an additional 30 min the fixative was replaced with chilled 0.1

M phosphate buffer (pH 7.4). Adult cats were similarly

transcardially perfused; however, twice the volume of solutions were

used than in the smaller animals.

The spinal cord and lumbosacral dorsal root ganglia were exposed

by dorsal laminectomy. The L3-S1 dorsal root ganglia (on the side

exposed to HRP) and the L6-S1 spinal cord segments were removed and

placed in 0.1 M phosphate buffer (pH 7.4). The tissues were cut into

50 micron sagittal sections using a vibratome (Oxford).

Histochemistry was performed using the chromagen tetramethylbenzidine

according to the procedures of Mesulam (1982). Tissues were

counterstained using neutral red, rapidly dehydrated and cleared with

xylenes. Sections were observed using both bright- and dark-field

illumination. Neuronal cell bodies in the dorsal root ganglia were

noted to be oval in shape (see Figure 2-1). Both the long (major)

and short (minor) axes of labeled cells were measured using an

eyepiece graticule. The mean of these two measurements was taken to

give the average cell diameter. In addition, the dimensions of

unlabeled cells in representative sections were also measured to

indicate the range of sizes of cell bodies present in the ganglia.











Controls for the Diffusion of HRP

In all experiments the ventral horn of the spinal cord was

scrutinized for the presence of labeled motoneuron cell bodies. In

those experiments in which HRP was applied to the muscle underlying

the femoral-saphenous vein numerous densely labeled motoneuron cell

bodies were noted. Thus, the presence or absence of labeled

motoneurons appeared to be an adequate indicator as to whether HRP

had diffused into the muscle underlying the femoral-saphenous vein.

If any labeled cell bodies were noted in the ventral horn of a

preparation in which HRP was applied to isolated segments of the

femoral-saphenous vein, the experimental results were discarded and

not analyzed further.

Results

Five of the 12 experiments (42%) in which HRP was applied to the

femoral-saphenous vein were successful; in these experiments labeled

neuronal cell bodies were present in the dorsal root ganglia and no

labeled motoneuron cell bodies could be detected in the spinal cord.

Examples of cell bodies in the dorsal root ganglia labeled by the

application of HRP to the femoral-saphenous vein are shown in Figure

2-1.

Following the application of HRP to the femoral-saphenous vein,

an average of 289 +/- 402 (mean +/- S.D.)1 labeled cell bodies in the

dorsal root ganglia was counted per experiment; an average of 182




1Other confidence intervals presented here also represent mean +/-
standard deviation.
























Figure 2-1. Examples of cell bodies in the L5 and L6 dorsal
root ganglia labeled by the application of HRP to the
femoral-saphenous vein. Calibration bars represent 50
micrometers. Labeled neuronal cell bodies are denoted by
arrows.















a. S


-:" .' i -i "
.q ,


.. I ,I


S -


-I
At


f i 'i','* .'... '
95$.











+/- 232 cells was counted per vein segment exposed to HRP. The top

portion of Table 2-1 shows the locations of cell bodies in the dorsal

root ganglia labeled by the application of HRP to the vein. In 3

preparations more than 95% of the labeled cell bodies counted were

confined to the L6 dorsal root ganglion; in 1 preparation all of the

labeled cell bodies were confined to the L5 dorsal root ganglion; in

the other preparation 77% of the labeled cell bodies counted were

located in the L5 dorsal root ganglion, and the other 23% were

located in the L6 ganglion. In total 78% of the labeled cells

counted were located in the L6 dorsal root ganglion, 21% were located

in the L5 ganglion and less than 1% were located at other levels.

The numbers and locations of cell bodies in the dorsal root

ganglia labeled by the application of HRP to muscle underlying the

femoral-saphenous vein (bottom portion of Table 2-1) were quite

different from those described above. In one experiment 6285 labeled

cells were counted; in the other 6251 were counted. The cell bodies

labeled by the application of HRP to the muscle underlying the

femoral-saphenous vein had a more widespread distribution than did

the cell bodies labeled by HRP application to the vein itself. In

total 3.9% of the labeled cells counted were located in the L3 dorsal

root ganglion, 7.3% were located in the L4 ganglion, 8.6% were

located in the L5 ganglion, 21.2% were located in the L6 ganglion,

30.5% were located in the L7 ganglion and 28.5% were located in the

Sl ganglion. It is reassuring that most of the cell bodies labeled

by the application of HRP to muscle underlying the femoral-saphenous

vein were located at levels caudal to those containing the cell

bodies labeled by HRP application to the vein itself.


















Table 2-1. Percentage of cell bodies in different dorsal root
ganglia labeled by the application of HRP to the femoral-saphenous
vein (top) or to the tissues underlying the vein (bottom).

HRP Applied to the Femoral-Saphenous Vein


Experi- # of Sites # of Percentage of Labeled Cell Bodies
ment # to which Cell Counted in Each Ganglion
HRP Was Bodies L3 L4 L5 L6 L7 S1
Applied Labeled


1 1 391 0 0 77 23 0 0

2 2 953 0 0 3 96 1 0

3 2 45 0 0 0 100 0 0

4 3 51 0 0 0 100 0 0

5 3 7 0 0 100 0 0 0

SUMMATION -- 1447 0 0 21 78 << 0



HRP Applied to the Muscle Underlying the Femoral-Saphenous Vein



1 2 6251 0 4 10 25 36 25

2 3 6285 8 11 7 18 25 32

SUMMATION --- 12536 4 7 9 21 31 29











Figure 2-2 shows the diameters of cell bodies in the dorsal root

ganglia labeled by the application of HRP to the femoral-saphenous

vein of the kitten; the figure also shows the diameters of unlabeled

cells measured in representative sections to indicate the range of

sizes of cell bodies present in the ganglia. Figure 2-3 shows

similar data collected from adult cats. Data derived from kittens

and adult cats are shown separately because the cell bodies in young

animals were noted to be smaller than those in the adult. Cell

bodies in the kitten labeled by the application of HRP to the

femoral-saphenous vein had an average diameter of 25.4 +/- 8.6

micrometers. The majority (87.5%) of the labeled cells were small in

size (diameter < 35 micrometers); only 1.6% were large (diameter > 50

micrometers). A total of 694 unlabeled cells were measured in the

kitten for comparison with the labeled cells. Unlabeled cells had an

average diameter of 30.5 +/- 11.3 micrometers; this mean was

significantly larger than that for labeled cells by Student's t-test

(p < 0.0001). Seventy-three percent of the unlabeled cells were

small in size (diameter < 35 micrometers); 6.5% were large (diameter

> 50 micrometers).

The relative sizes of cell bodies in the dorsal root ganglia

labeled by the application of HRP to the femoral-saphenous vein and

unlabeled cell bodies were similar in the adult cat to the relative

sizes in the kitten. Cell bodies in adults labeled by the appli-

cation of HRP to the femoral-saphenous vein had an average diameter

of 32.9 +/- 14.0 micrometers. Most (71.8%) of the labeled cells were

small in size (diameter < 35 micrometers); only 13.2% were large






























(0 4- 4.4




41
0J
0 4







,-I *t
M *0

MOJ
S0a

J0
o< .


O40 4J 0

w 0 0 w
C0



.-1 > cc
-40 4 O 4J
00 1 "-1









rw



0 0 0)
41 m t















0 00
CI 4








1 0 .-*4





Oe C.-4 NO
i-i o a0 "-1











4 *0 r4w
O C0







*rI a <
r* r-4





1 i -l



*4rf 4 *U,
(i> o~ co
PI-l d1 (























LII




a'5








UI
--


3)
Y
.fi


I I I I I
_ Ifl r 3


0
io
3o
fain


I I* In I


e6O4ueOOed





























0 4 10 bO
w@1 00a
bI() 40
r0 0 0
U 0a l
00, UU b (






,CO 0
c-44 00
S0 .C --I 00

E-4 I4 r-I
w-i 0 01






> J U
OM#
C0 C C :
i Q0 0M
000 01

0g 00(D C6







*r4 4c









1 00
L- 0 *r 0
o to a0

,C >f4 bOT





u 4-1 e c
l 0 '>
0N a 0 4-1 >o











$4J 4
4I oe Im
r- t C u





0 O 4.1







000 W0
0U 0 B0

















4 4 4- 14 r. u
4 wcc 0 0H








QU 0 O 4
010 01 -v




1 l 0 4
O rl NU r.
0 1- 4 41


10 (3 0l l 01




4JI a 00 1 (o









10


oI 3 I


I-



I-
-I




Qr


r I I


eOD4Ue3Jod


E
C:
L.
4)
E
0r











(diameter > 50 micrometers). A total of 1255 unlabeled cells were

measured in the adults for comparison with the labeled cells.

Unlabeled cells had an average diameter of 41.7 +/- 19.5 micrometers;

this mean was significantly larger than that for labeled cells by

Student's t-test (p < 0.0001). Forty-five percent of the unlabeled

cells were small in size (diameter < 35 micrometers); 24.1% were

large (diameter > 50 micrometers).

Both large and small cell bodies in the dorsal root ganglia of

cats were labeled by the application of HRP to the femoral-saphenous

vein; however, most of the labeled cells were small. Labeled cell

bodies were smaller, on the average, than other cell bodies located

in the ganglia; in addition, the percentage of large labeled cell

bodies counted in the ganglia was smaller than the percentage of

large unlabeled cell bodies. However, the anterograde transport of

HRP from the dorsal root ganglia to afferent terminals in the spinal

cord was far less extensive than the retrograde transport from the

vein to the ganglia, if it occurred at all. The transganglionic

labeling of intraspinal portions of afferents was never detected

using either bright- or dark-field illumination following the

application of HRP to the femoral-saphenous vein.

Discussion

Previous electrophysiological studies in the cat have shown that

electrical or mechanical stimulation of isolated segments of the

femoral-saphenous vein can elicit potentials recordable from the

lumbar spinal cord (Thompson and Barnes, 1979; Thompson and Yates,

1983, 1984, 1986; Thompson et al., 1982, 1983). This study extends











these findings by providing an anatomical demonstration of the

number, distribution and sizes of the cell bodies from which these

femoral-saphenous venous afferents originate. The selective labeling

of afferents arising in the walls of the femoral-saphenous vein in

these preparations was supported by several lines of evidence. No

labeling of spinal motoneurons could be detected in these studies,

suggesting that HRP had not diffused from the vein to adjacent

muscle. Further evidence suggesting that no spread of tracer

occurred is that the application of HRP to the tissues adjacent to

the femoral-saphenous vein produced labeling of cell bodies found

predominantly in dorsal root ganglia caudal to those containing

labeled cell bodies following application of HRP to the vein itself.

These data also suggest that the afferents which were labeled

terminated in the walls of the femoral-saphenous vein and were not

fibers of passage coursing along the vein but arising in muscle. It

would be very unlikely for all fibers of passage from muscle to be

sensory afferents; if fibers arising in muscle had picked-up HRP in

these experiments, the labeling of motoneuron cell bodies should have

occurred. In addition, if fibers of passage from muscle had

transported HRP, the distribution of labeled cell bodies in the

dorsal root ganglia should have been similar to that following the

application of tracer directly to the muscle. Since the cells

labeled by the application of HRP to the femoral-saphenous vein had a

different distribution, it is likely that they also had a different

origin. Similarly, it is unlikely that fibers of passage arising in

skin were labeled in these studies. The skin overlying the











femoral-saphenous vein in the intact animal is innervated by the

saphenous nerve; this nerve projects to several spinal levels

(Bernhard, 1953; Crouch, 1969). If fibers of passage from skin had

picked-up HRP in these experiments, a more widespread distribution of

labeled cell bodies in the dorsal root ganglia would have been

expected. It is also noteworthy that the cell bodies in the dorsal

root ganglia labeled by the application of HRP to the

femoral-saphenous vein were significantly smaller, on the average,

than other cell bodies in the ganglia. If a general class of

hindlimb afferents, and not a specific population, had taken up HRP

in these studies, the diameters of the labeled cells should have been

identical to those of other cells in the ganglia.

A recent study by Lee et al. (1986) has made it possible to

correlate the diameter of a cell body in a dorsal root ganglion with

the conduction velocity along the peripheral process arising from it.

These investigators made intracellular recordings from single cell

bodies in the dorsal root ganglia of adult cats using microelectrodes

filled with an HRP solution, determined the conduction velocity along

the peripheral process to the cell body, and then injected the tracer

into the neuron so it could later be visualized. They reported that

cell bodies less than 35 micrometers in diameter gave rise to

peripheral processes that carried impulses at less than 2.5 m/sec (C

fibers). Cell bodies greater than 50 micrometers in diameter were

reported to give rise to peripheral processes that carried impulses

at greater than 2.5 m/sec (A fibers). For these neurons there was a

linear correlation between the diameter of the cell body and the











impulse conduction velocity; this relationship was defined by the

following equation:

Conduction Velocity = (0.84 Diameter of Cell Body) 20.3.

However, there was no predictable relationship between the diameters

of intermediate-sized cell bodies (35-50 micrometers) and the

conduction velocities along their peripheral processes.

Most (71.8%) of the cell bodies labeled by the application of

HRP to the femoral-saphenous vein of the adult cat were smaller than

35 micrometers in diameter; these cell bodies would be expected to

give rise to slowly conducting peripheral processes. However, 13.2%

of the cell bodies labeled by the application of HRP to the vein were

larger than 50 micrometers in diameter; these cell bodies should give

rise to rapidly conducting peripheral processes. The estimated

conduction velocities along these processes are shown in Table 2-2;

the table also shows the predicted conduction velocities of other

afferents based upon the diameters of cell bodies that were not

labeled following the application of HRP to the vein. The largest

femoral-saphenous venous afferents were estimated to have conduction

velocities of approximately 60 m/sec. Fifteen percent of the cell

bodies labeled by the application of HRP to the femoral-saphenous

vein were intermediate in size (35-50 micrometers); no estimate could

be made of the conduction velocities along these afferents. Blood

vessels other than the femoral-saphenous vein have previously been

shown to be innervated by thermoreceptors (Fruhstorfer and Lindblom,

1983), chemoreceptors (Moore and Moore, 1933; Lim et al., 1962;

Potter et al., 1962; Besson et al., 1972) and slowly conducting






















Table 2-2. Estimated conduction velocities
processes of large primary femoral-saphenous
other large primary sensory neurons determined
the cell bodies.


Conduction Velocity
Range


along the peripheral
venous afferents and
from the diameters of


Percentage of Cells Falling into the Range
Femoral-Saphenous Other Afferents
Venous Afferents


20-30 m/sec

30-40 m/sec

40-50 m/sec

50-60 m/sec

> 60 m/sec











mechanoreceptors (Bessou and Perl, 1960; Floyd and Morrison, 1974;

Floyd et al., 1976). It is likely that the femoral-saphenous vein is

also innervated by one or more of these receptor types as well as by

rapidly conducting afferents that have not been reported to be

associated with other vessels.

Only a small number of afferents were labeled by the application

of HRP to the femoral-saphenous vein (an average of 182 per vein

segment exposed to the enzyme), and only a small fraction of the

afferents labeled were large. However, electrical activation of

A-alpha/beta femoral-saphenous venous afferents has been reported to

produce large field potentials recordable from the cord (Thompson and

Barnes, 1979; Thompson and Yates, 1986). For the anatomy and

electrophysiology to be in register, it would seem that the

divergence of the inputs from the femoral-saphenous vein to the

spinal cord is substantial. Such divergence could be accomplished by

either the extensive branching of the intraspinal portions of the

afferents or the activation of second-order spinal interneurons that,

in turn, make contacts with a large number of higher-order cells.

Visceral afferent inputs into the lower thoracic spinal cord of the

cat have previously been shown to exhibit a similar widespread

divergence (Cervero et al., 1984).


















CHAPTER III
PROPERTIES OF SPINAL CORD PROCESSING OF FEMORAL-SAPHENOUS VENOUS
AFFERENT INPUT REVEALED BY ANALYSIS OF EVOKED POTENTIALS

Introduction

The study presented in Chapter II suggested that afferents with

both large and small diameters project from the femoral-saphenous

vein to the lumbosacral spinal cord of the cat. In most preparations

the L6 segment was the input cord segment for these primary

afferents. The study discussed in this chapter examined two

characteristics of the interneurons which link inputs along these

afferents with spinal motoneurons. The minimum time required for

activation of spinal motoneurons by the femoral-saphenous venous

afferents was measured to estimate the least number of interneurons

interposed between the primary afferents and motoneurons. In

addition, focal synaptic field potentials were mapped to determine

the location of the interneurons monosynaptically excited by

stimulation of the femoral-saphenous vein.

The field potential mapping techniques utilized in this study

were first described by Gasser and Graham (1933); the intracord

mapping of potential fields has since been performed in a number of

studies (Campbell, 1945; Austin and McCouch, 1955; Howland et al.,

1955; Coombs et al., 1956; Fernandez de Molina and Gray, 1957; Willis

et al., 1973; Fu et al., 1974; Beall et al., 1977; Foreman et al.,

1979) and has proven to be an effective preliminary method of










describing the spatial location of the interneuronal pools, activated

by a particular stimulus. The most prominent field potential

recordable from the dorsal horn is the result of ionic movements at

excitatory axo-dendritic and axo-somatic synapses (Willis, 1980).

When the synapses are activated, positive ionic current leaves the

extracellular space at the synapses (sinks) and reappears along the

axons of the interneurons (sources). Because the majority of dorsal

horn interneurons ventral to the substantial gelatinosa have ventrally

projecting axons (Matsushita, 1969, 1970), the dorsal horn takes on a

net negative charge and the ventral horn takes on a net positive

charge. The negativity of shortest latency is focused in a region

which contains a high density of interneurons monosynaptically

excited by primary afferents (Yates et al., 1982).

Methods and Materials

Surgical Preparation of Animals

Data were recorded from 25 adult cats of either sex.

Intraspinal potential fields were recorded from 7 of these animals;

the minimum times between the arrival of the primary afferent volleys

along muscle, cutaneous and venous afferents and detectable

discharges in the ventral roots were determined in the other 18.

Animals were initially anesthetized with 3% Fluothane (halothane,

Ayerst laboratories) vaporized in a mixture of nitrous oxide and

oxygen. A tracheostomy was performed and either of the common

carotid arteries was cannulated; the other common carotid artery was

ligated. The tracheostomy allowed for artificial ventilation with a











Harvard respiration pump; cannulation of the carotid allowed for

measurement of arterial blood pressure. The left cephalic vein was

also cannulated to allow the intravenous administration of drugs and

fluids. Animals were rendered decerebrate at the midcollicular level

by transaction of the brainstem and aspiration of the portions of the

brain rostral to this level. Following decerebration, anesthesia was

discontinued; this occurred at least 4 hours prior to the beginning

of the recording session. The spinal cord was also transected at

T10. Animals were immobilized in a spinal unit (David Kopf);

paralysis was induced through intravenous injection of Flaxedil

(gallamine triethiodide, Davis-Geck, 2 mg/kg). Flaxedil was

readministered when any muscle contractions were noted. Spinal cord

segments L3-S1 were exposed by laminectomy, the dura mater was opened

and pinned away from the cord dorsum, and the exposed tissues were

covered with warmed mineral oil. Body and oil pool temperatures were

maintained at 37 degrees C using a heating pad and a heat lamp.

The skin of either of the hindlimbs was opened dorsally to

expose a segment of the femoral-saphenous vein. A 10-15 mm segment

of the vein was separated from surrounding connective tissue and

gently placed upon a silver bipolar stimulating electrode. A plastic

membrane was placed beneath the stimulated vein segment to prevent

accidental contact of the electrode with surrounding tissues. In

those experiments in which the minimum time between the arrival of

the primary afferent volley and a detectable discharge in the ventral

roots was determined, a segment of one or more of the following

nerves was also prepared, in a manner similar to that used to prepare











the vein, for bipolar electrical stimulation: the sural nerve,

posterior tibial nerve, gastrocnemius nerve, deep peroneal nerve,

plantaris nerve or deep posterior nerve (branch of the tibial

innervating foot musculature). Bipolar stimuli were sometimes also

applied to the skin overlying the femoral-saphenous vein. In

addition, the L6 and L7 ventral roots were cut at dural entry and

were placed across bipolar hook recording electrodes.

Stimulation Procedures

Bipolar electrical stimulation was used in these experiments to

excite the femoral-saphenous venous afferents. A number of previous

studies have also utilized electrical stimulation to activate the

venous afferents (Thompson and Barnes, 1979; Thompson and Yates,

1983, 1984, 1986; Thompson et al., 1982); these studies provided

considerable evidence that this stimulation only excited afferent

fibers arising in the vein wall. Application of ligatures to

portions of the vein immediately proximal and distal to the

stimulating electrode abolished the evoked potentials elicited by

passing current along that vein segment (Thompson and Yates, 1986).

This result was apparently due to the blockade of impulse conduction

in afferents coursing along the vein wall (Gelfan and Tarlov, 1956).

Stimulation of isolated segments of the femoral-saphenous vein

elicited a unique spinal reflex; the stimulation produced the

simultaneous excitation of hindlimb flexor and extensor muscles

acting on the thigh, crus and foot (Thompson et al., 1982).

Activation of cutaneous afferents produces the reciprocal excitation

and inhibition of flexors and extensors; activation of large muscle











afferents also elicits the reciprocal excitation and inhibition of

hindlimb muscles (see Chapter I). It is thus likely that the

afferents which were excited by the stimulation of isolated segments

of the femoral-saphenous vein belonged to a unique class arising in

the vein wall and were not fibers of passage coursing along the vein

but arising in muscle or skin. The studies presented in the last

chapter involved the application of HRP to isolated vein segments;

the procedures used to isolate portions of the femoral-saphenous vein

for exposure to this tracer and for electrical stimulation are

similar. The results presented in Chapter II suggested that only

afferents arising in the vein were labeled by the application of HRP

to the isolated vein segments; thus, muscle and cutaneous afferents

apparently were not coursing along these segments. It is clear that

electrical stimulation is a technique which allows for the selective

activation of femoral-saphenous venous afferent fibers.

During experiments in which intracord field potentials were

mapped, the prepared segment of the femoral-saphenous vein was

stimulated using single-shock square-wave pulses of current, 0.2 msec

in duration, repeated at a frequency of 1 Hz. Stimuli were delivered

by a digital stimulator (Frederick Haer Pulsar 6i). Stimuli 3 times

threshold for eliciting field potentials recordable from the cord

dorsum were used to excite the afferent fibers. This stimulus

intensity was chosen because previous studies had shown it sufficient

to elicit compound action potentials recordable from the dorsal roots

with A-beta components of maximal amplitude. The A-delta and C

fibers shown in Chapter II to also arise from the femoral-saphenous











vein were not excited by this stimulus intensity (Thompson and Yates,

1986); thus, a very synchronous spinal input was elicited. During

the experiments in which the minimal times required for activation of

spinal motoneurons were measured, the femoral-saphenous vein was

stimulated at an intensity 2-3 times that necessary to evoke field

potentials recordable from the cord dorsum. The hindlimb skin and

the sural nerve were stimulated at twice cord dorsum threshold.

Stimuli applied to muscle nerves were graded to evoke only

monosynaptic reflexes recordable from the ventral roots or to elicit

both monosynaptic and polysynaptic reflexes.

Recording Procedures

Intraspinal potential fields. Intraspinal field potentials

were recorded using tungsten microelectrodes insulated with epoxy or

glass except for a 10-20 micrometer tip which had initial impedances

of 0.5-4.0 megohms. Potentials were recorded at 0.2 mm intervals

from the cord dorsum to 4.4 mm in depth; electrodes were positioned

using a hydraulic microdrive (Narishige). Five tracks were conducted

across the L6 cord segment. Penetrations were optimally made at

midline and 0.5, 1.0, 1.5 and 2.0 mm lateral to midline; slight

alterations in the choice of penetration site were often necessary to

avoid puncturing large blood vessels. Potentials were referenced to

animal ground; the reference electrode was positioned in the wound

margin adjacent to the cord. Potentials were led into A.C.

preamplifiers (Grass P511) with a time constant of 250 msec and

subsequently into a tape recorder (Crown-Vetter model A) which

contained preamplifiers with a bandwidth of D.C. to 15 kHz. The











potentials were displayed on an oscilloscope and were averaged by an

Ortec computer. Records of averaged potentials were made with an X-Y

plotter (Hewlett-Packard).

A radiofrequency lesion was placed at the bottom of each track

to facilitate the histological localization of the track. The depths

of lesions were also analyzed as a control for tissue shrinkage by

fixative and other factors (such as cord dimpling) which might cause

depths measured from histological material to deviate from those

measurements read from the vernier of the hydraulic microdrive.

Cord dorsum potentials. Cord dorsum potentials were

recorded from L6 using silver ball electrodes and were referenced to

animal ground. The potentials were led into A.C. preamplifiers, etc,

as for potentials recorded from microelectrodes. Cord dorsum

potentials were recorded in these experiments for two reasons. The

measurement of cord dorsum potentials provided a means to determine

the latency at which input along the fastest-conducting afferents

arrived at the cord; several studies have shown that the initial

triphasic spike of the cord dorsum potential is a compound action

potential propagating through large primary afferent fibers (Austin

and McCouch, 1955; Howland et al., 1955). In addition, the amplitude

of cord dorsum potentials is known to vary in proportion to spinal

cord excitability (Bernhard and Koll, 1953; Gelfan and Tarlov, 1956;

Molt et al., 1978). Thus, the measurement of cord dorsum potentials

provided a control to assure that the level of spinal cord

excitability remained constant throughout the recording session and

that little cord damage was produced by electrode penetrations in











those experiments in which intraspinal field potentials were mapped.

Data recorded from animals in which the cord dorsum potentials varied

in amplitude during the recording session were not analyzed.

Compound action potentials. Compound action potentials were

recorded from the L6 and L7 ventral roots during those experiments in

which the minimal time required for an afferent input to activate

motoneurons was determined. The bipolar recording electrodes were

positioned in close proximity to the spinal cord in these

experiments. Compound action potentials were led into A.C.

preamplifiers, etc, as described above.

Histological Procedures

At the conclusion of intraspinal field potential mappings,

animals were perfused through the heart with 1 1 of heparinized

saline followed by 1.5 1 of 1.25% glutaraldehyde/1% paraformaldehyde

fixative. The L6 segment was removed, quick-embedded in

gelatin-albumin, cut into 40 micrometer sections with a freezing

microtome and stained with cresyl violet. Most electrode tracks

could easily be reconstructed from histological material; only data

recorded along reconstructed tracks were considered for analysis. If

fewer than 4 tracks were reconstructed in a particular animal, data

recorded from that animal were not analyzed. Identification of the

laminae in which the recording sites were located was done through

microscopic examination of histological material and by comparison

with published accounts of laminar organization (Rexed, 1952, 1954).












Results

Figure 3-1 shows the results of experiments in which minimal

times required for activation of spinal motoneurons by inputs from

muscle, skin, and the femoral-saphenous vein were examined. A

determination of this minimal time for femoral-saphenous venous

afferent input is illustrated in part "A" of the figure. This

latency represents a combination of conduction time along the

intraspinal portions of primary afferents, synaptic delay times, and

conduction time along the processes of spinal neurons. However, a

comparison of the central delay following venous afferent stimulation

with the delay following stimulation of cutaneous and muscle

afferents, the synaptology of which is well established (Brown,

1981), provides an estimate of the minimal number of synapses

interposed between the primary afferents and alpha-motoneurons.

The minimal times between the arrival of the primary afferent

volley and a detectable discharge in the ventral roots are indicated

in part "B" of Figure 3-1. The abscissa shows the afferent type

stimulated. Data regarding muscle afferents represent the pooling of

delays measured following stimulation of the following nerves:

posterior tibial, gastrocnemius, deep peroneal, deep posterior or

plantaris. The latencies from triphasic spike to onset of both

monosynaptic and polysynaptic reflexes recorded from the ventral root

are shown. Data regarding cutaneous afferents represent the pooling

of delays measured following stimulation of the sural nerve or the

skin overlying the femoral-saphenous vein. Latencies reported for

muscle and cutaneous input reflect the processing time by the input





















Figure 3-1. Minimal spinal cord processing time for inputs from
muscle, skin and the femoral-saphenous vein. (A) The method
used to determine minimal spinal cord processing time. (B) A
comparison of these latencies. Abbreviations: CD, cord dorsum
potential; VR, potential recorded from the ventral roots; TPS,
triphasic spike.






















MINIMAL PROCESSING
TIME


mmm
V. A. MUSCLE
L6- L6 (polysyn)


CUTAN. MUSCLE
(monosyn)


CD


2.5 msec


VR


6w



s5.



-4-


z
-3-
VI
u
U
0
a,
L2



-J


0 =


V.A.
L6- L7











cord segment for the stimulated afferents (L7 for the stimulated

muscle nerves and sural nerve and L6 for the skin overlying the

femoral-saphenous vein). Two central delay times following

stimulation of the femoral-saphenous vein are reported: the latency

separating the L6 triphasic spike and activity recorded from L6

ventral roots as well as the latency separating the L6 triphasic

spike and activity recorded from the L7 ventral roots. Thick bars

indicate mean minimal spinal cord processing time; thin bars indicate

the standard error. Stars indicate values shown to differ

significantly from all others by an analysis of variance procedure in

combination with either a Waller-Duncan K-ratio T test or Duncan's

multiple range test (alpha-0.05). The delay separating the triphasic

spike elicited by muscle afferent stimulation and the monosynaptic

reflex recorded from the ventral root was significantly shorter than

all others; the delay separating the L6 triphasic spike elicited by

femoral-saphenous venous afferent stimulation and ventral root

activity recorded from L7 was significantly longer than all others.

Figure 3-2 shows a series of waveforms recorded following

femoral-saphenous venous afferent stimulation at various depths along

a typical microelectrode track through L6. The potentials were

recorded as a series of negativities near the cord dorsum or in the

dorsal gray matter waveformss A-D). As the microelectrode tip

advanced into the ventral horn, the negative waves decreased in

amplitude and eventually reversed in polarity waveformss E and F).

Figure 3-3 shows the focus of the short latency negative waves

elicited by stimulation of the femoral-saphenous vein. The dots


































0
'-I

0




*C
14
















0


0





I
co
0

0o
U 44


4)
r -
0)













4r1









-i
0
u *








0co
o >










to3




4 0
rZ4 44
0 u~



br-14-
B 0
T









U
E
a
In
IE
Io EE
N=!,






























ad






MO x
bOW


04)

co
4 J0



.4
0
S<4

4o



44


00


01
0





4O
0 4






4O
C 4

0 0
1 00
o 0
e







0
u 0)
0 1
U-1 i-
n)
d-
0)
.c


* A>
I


4 0
bo x















u
0














CO
CV


I1












indicate points where maximal-amplitude potentials were recordable;

the lines surrounding the dots indicate the area in which potentials

80% or more of maximal could be recorded; the arrows at the top of

both diagrams indicate the location and orientation of electrode

tracks used to collect data; the time over each diagram indicates

latency from the onset of the triphasic spike of the cord dorsum

potential. In most experiments (5 of 7), the initial negativities

were focused in the neck of the dorsal horn; this region corresponds

to Rexed's lamina V (Rexed, 1952, 1954). In 4 of 7 experiments, the

initial negativities were also focused, along the medial-to-lateral

axis, roughly in the center of the gray matter; however, foci both

medial and lateral to this point were also noted. It is likely that

the inconsistencies in results observed in a few experiments were due

more to limitations in the field potential mapping technique than to

physiological variability between animals. If a microelectrode

failed to pass directly through a potential generator during a

recording session, the spatial location of that potential generator

could appear displaced from its actual location.

Discussion

The latency separating evoked activity recordable from the

dorsal and ventral roots of L6, the input segment for

femoral-saphenous venous afferents, following stimulation of the

femoral-saphenous vein (4.4 msec) was approximately the same as the

latency noted following stimulation of cutaneous afferents (4.3

msec). Minimal spinal cord processing time, estimated in the same

way for inputs from muscle which reach alpha-motoneurons through











polysynaptic circuits, was only slightly shorter (3.8 msec). It is

known that the fastest circuits which link cutaneous inputs and

muscle inputs with interneurons and then motoneurons are di- and tri-

synaptic (Lloyd, 1943b, 1943c). These data are consistent with a di-

or tri- synaptic circuit being the shortest pathway between primary

femoral-saphenous venous afferents and alpha-motoneurons.

Following electrical stimulation of the femoral-saphenous vein,

negative waves of the cord dorsum potential can be recorded in spinal

animals for 20 msec or more (Thompson and Yates, 1986). These waves

reflect the activation of dorsal horn interneurons (Coombs et al.,

1956; Fernandez de Molina and Gray, 1957; Willis et al., 1973; Fu et

al., 1974; Beall et al., 1977; Foreman et al., 1979; Willis, 1980;

Yates et al., 1982). In addition, maximal ventral root action

potentials elicited in spinal animals by femoral-saphenous venous

afferent stimulation have a duration of 20-25 msec (Thompson et al.,

1982). Thus, although relatively short di- and tri- synaptic

circuits interconnect venous afferents and motoneurons, much higher

order circuits also appear to exist.

The minimal latency separating L6 dorsal root activity and L7

ventral root activity elicited by FVA stimulation was significantly

longer (by at least 1.7 msec) than other processing times estimated.

Increased intraspinal conduction distances can account for only a

portion of this longer latency; the conduction time along primary

femoral-saphenous venous afferents from L6 to L7 is less than 1 msec

(Thompson and Yates, 1986). It is likely that at least one

additional synapse is interposed between femoral-saphenous venous











afferents and L7 motoneurons than is interposed between the afferents

and L6 motoneurons.

A distribution of intracord potential fields similar to that

produced by stimulation of the femoral-saphenous vein (negativities

in the dorsal horn and positivities in the ventral horn) is also

elicited by stimulation of other afferent types. Previous studies

have shown that the negative waves of shortest latency in the dorsal

horn can be attributed to monosynaptic excitation of interneurons by

primary afferents (Coombs et al., 1956; Fernandez de Molina and Gray,

1957; Willis et al., 1973; Fu et al., 1974; Beall et al., 1977;

Foreman et al., 1979; Willis, 1980; Yates et al., 1982). Since the

largest negative waves of shortest latency were recorded from Rexed's

lamina V, it appears that most of the first interneurons excited by

the large venous afferents are found in this region.


















CHAPTER IV
PHYSIOLOGICAL PROPERTIES OF SINGLE SPINAL CORD NEURONS
ACTIVATED BY STIMULATION OF THE FEMORAL-SAPHENOUS VEIN

Introduction

The data presented in Chapter III suggested that at least one or

two interneurons are interposed between large (A-beta) primary

femoral-saphenous venous afferents and motoneurons. It was also

concluded that the first interneurons activated by these large

afferents are located in Rexed's lamina V. The study discussed in

this chapter examined the responses of single spinal cord neurons

following stimulation of the femoral-saphenous vein. In addition,

the patterns of convergence of inputs from muscle and skin on the

venous afferent-activated interneurons were examined. This study was

conducted to show whether venous afferent inputs were highly

processed by spinal interneurons and, if so, whether these neurons

were modality-specific.

Methods and Materials

Surgical Preparation of Animals

Experiments were performed on 19 unanesthetized, decerebrate

cats with spinal cords transected at T12 and 5 intact cats

anesthetized using alpha-chloralose. The two preparations were used

because a comparison of the data obtained from each type would

suggest whether an elicited input had strong or weak effects on the

excitabilities of the neurons being studied. It is known that the











excitabilities of spinal interneurons in intact animals anesthetized

using alpha-chloralose are lower than in decerebrate-spinal

preparations. The lower excitabilities in these anesthetized

preparations are due to the influences of descending supraspinal

inputs to the spinal interneurons as well as to direct effects of the

anesthetic on the neurons (Alvord and Fuortes, 1954; Balls and

Monroe, 1964; Brown, 1981). If spinal neurons are driven in both

decerebrate and anesthetized preparations by a particular input, it

is likely that the input has powerful effects on these neurons.

However, if spinal neurons are activated by the input in

decerebrate-spinal preparations, but not anesthetized preparations,

it is likely that the input has only weak influences on the

excitabilities of these neurons.

Anesthesia was induced in the group of animals which would not

be decerebrated using the intravenous injection of alpha-chloralose

(Merck, 75 mg/kg). Maintenance doses of the drug (25 mg/kg) were

routinely delivered every four hours. However, the electrocardiogram

and arterial blood pressure were continuously monitored during the

experiment; if either the heart rate or blood pressure increased,

indicating that the animal was recovering from anesthesia, additional

alpha-chloralose was administered until both of these measurements

stabilized at their original level. The other animals were rendered

decerebrate using the procedures described in Chapter III.

Most of the surgical procedures used to prepare cats for single

unit recordings were similar to those described in Chapter III for


___











preparing animals for the recording of field potentials. Several

additional procedures were also used to provide better mechanical

stability for recording. A dural hammock was formed by placing

stitches through the dura and then by suspending small weights from

the threads. In this way the cord was lifted away from underlying

vertebral bodies and pulsating blood vessels. In many animals the

bladder was catheterized so that it would remain empty; this helped

to stabilize the blood pressure and prevented blood pressure surges

due to a full bladder. A bilateral pneumothorax was also used to

minimize movements of the spinal cord during respiration. In

addition to these procedures for increasing mechanical stability,

holes were made in the arachnoid and pia overlying the L6 spinal cord

using sharp forceps. This procedure permitted the insertion of

microelectrodes into the cord without breaking the tips.

Stimulation Procedures

Bipolar electrical stimulation was used to activate

femoral-saphenous venous afferents, cutaneous afferents coursing in

the sural nerve and muscle afferents running in the posterior tibial,

deep posterior, lateral gastrocnemius and hamstring nerves.

Square-wave pulses of current, 0.2 msec in duration, were used to

excite afferent fibers. While data were being collected, stimulus

frequencies of 1 Hz were used. However, stimulation frequencies up

to 100 Hz were used to differentiate afferent responses from

responses generated by interneurons (see below). Throughout the time

microelectrode penetrations were being done, the femoral-saphenous

vein was continuously stimulated at an intensity 3 times that












necessary to evoke a field potential recordable from the cord dorsum.

As discussed in Chapter III, this stimulus intensity provides a

maximal A-beta volley to the spinal cord but does not excite the

A-delta and C afferents. High current intensities were not used in

an attempt to recruit the small-diameter venous afferents due to

concerns regarding stimulus spread. Nerves were stimulated at

intensities up to 10 times the threshold for eliciting a cord dorsum

potential; the minimal current intensity for affecting the

excitability of a unit was also determined.

Recording Procedures

Intracellular and extracellular single unit recordings were done

using glass micropipette electrodes filled with 3 M KC1 or 8% HRP in

tris-buffered 0.15 M KC1 (pH 7.3). Electrodes filled with 3 M KC1

had initial impedances of 15-30 megohms; electrodes filled with the

HRP solution had initial impedances of 30-90 megohms.

Microelectrodes were positioned using a hydraulic microdrive

(Narishige). The L6 dorsal columns ipsilateral to the stimulated

vein and nerve segments were usually impaled by the electrodes;

however, a few penetrations were also made through the L6 dorsal root

entry zone. The microdrive was angled sharply towards the midline

during these latter penetrations. When electrodes filled with 3 M

KC1 were used to record responses, only units located at 700-1500

micrometers depth were studied. At these depths, the electrode tips

should have remained in the dorsal horn or in the most superficial

portion of lamina VII (Rexed, 1952, 1954); thus, no motoneurons

should have been encountered. In addition, the units were tested for











their responses to high frequency stimulation; it is known that

action potentials recorded from primary afferents show constant

latency with high stimulation frequencies, whereas the latencies of

responses recorded from interneurons are altered by high frequency

stimulation (Yates et al., 1982). If responses recorded from a unit

were invariant in latency with 100 Hz stimulation, these responses

were not analyzed. The use of electrodes filled with the HRP

solution allowed recording sites to be marked either by the

intracellular staining of the characterized neuron or by the

extracellular iontophoresis of the tracer; the procedures used to

reconstruct recording sites are discussed in Chapter V. Only data

recorded from units located outside of lamina IX were analyzed.

All potentials were referenced to animal ground; the reference

electrode was located in the wound margin adjacent to L6. Potentials

were led into a microelectrode pre-amplifier (Winston model 1090A)

and subsequently into an A.C. amplifier (Grass P511) with a time

constant of 250 msec, tape recorder (Vetter model D) with a 10 db

bandwidth of 0-2500 Hz and oscilloscope. The direct output of the

microelectrode pre-amplifier was also led into an oscilloscope and

into the tape recorder. Analysis of data was facilitated by the time

histogram unit of an Ortec computer (model 4621); the latter unit

generated poststimulus time histograms from the data. Hardcopy

records of histograms were generated using an X-Y plotter

(Hewlett-Packard).

At the conclusion of single unit recordings from 11 animals,

small bundles of the L7 dorsal rootlets were cut proximally and











placed across a silver bipolar recording electrode. The sural and

posterior tibial nerves were then stimulated using the same current

intensities applied to activate interneurons during the single unit

recordings. Potentials were led into an A.C. amplifier, tape

recorder and oscilloscope. Potentials were also averaged using the

Ortec computer. In addition, the conduction distances from the nerve

stimulation sites to the dorsal root recording site were measured.

These data were used to determine the conduction velocities of the

afferents activated by the minimal current intensities necessary to

alter the excitabilities of interneurons.

Results

Responses of Interneurons Following Stimulation of the Femoral-Saph-
enous Venous Afferents

Data were recorded from 81 neurons activated by stimulation of

the femoral-saphenous vein in decerebrate cats; detailed analyses

were made of the data recorded from 66 units. While only the latter

data were useful in quantitative determinations of such parameters as

the onset latency of evoked activity in neurons, etc., all of the

data were useful in determining the patterns of convergence of

femoral-saphenous venous afferent input with inputs from muscle and

skin. In anesthetized animals, data were recorded from 14 units; the

responses of all of these neurons following femoral venous afferent

stimulation were characterized in detail.

Figure 4-1 illustrates the types of responses recorded from

interneurons following femoral-saphenous venous afferent stimulation

at an intensity 3 times that necessary to evoke field potentials































4-1 0 .





S000
0 0 (0

0 0)
C 0)m 0
W) 0 bO A


.0 0m
I 0( w
w bO 4
0 0

0
1 00 41 0





W Ce
44 W W0
0)0 V4J






o 41 -.4 >
C.44
S0)






d00 0
44 0) )-1








01..

Cm
0 w
4-4 A 4) T
.41 0





O0 0
OW




.(0 0 04


O* (
- 0 C



01 e4- m
ao.










0a) (00














1.5 mV 30 spikes
C
+0
em
c

- 1








2.5 mV 20 spikes











recordable from the cord dorsum. A burst of action potentials was

recorded from 42 out of 66 (64%) of the venous afferent-driven

neurons following stimulation of the femoral-saphenous vein in

decerebrate animals. The other 24 neurons (36%) responded to vein

stimulation with a combination of one or more bursts of action

potentials and a period during which the firing rate of the unit was

depressed below the spontaneous rate. The neurons studied in cats

anesthetized using alpha-chloralose responded in similar ways to

femoral-saphenous venous afferent stimulation; half of the 14 neurons

characterized were excited by vein stimulation, and the other half

were both excited and inhibited.

The quantitative aspects of the response properties of these

neurons activated by stimulation of the femoral-saphenous vein are

summarized in Table 4-1. The mean spontaneous firing rates of the

neurons were similar in decerebrate and anesthetized animals (15.5

+/- 16.7 spikes per second in the former and 14.3 +/- 20.3 spikes per

second in the latter). Clearly, most of the units examined had

substantial ongoing activity in the absence of stimulation; however,

a number of units showed no recordable spontaneous firing. Twelve of

the units (18%) characterized in decerebrate cats were silent in the

absence of stimulation; 5 of the units (36%) in anesthetized animals

exhibited no spontaneous firing. The occurrence of these silent

units is important, since a unit which does not fire spontaneously

will be classified as only being excited by elicited inputs, whether

or not it receives inhibitory inputs. It is only possible to show

that the firing rate of a unit decreases below the spontaneous level












Table 4-1. Responses of spinal interneurons to electrical stimulation
of the femoral-saphenous venous afferents, posterior tibial nerve and
sural nerve. All confidence intervals represent mean +/- standard
deviation. Onset latencies were measured from the onset of the
triphasic spike of the cord dorsum potential. All means except those
for spontaneous firing rate are in msec; mean spontaneous firing rates
are in spikes per sec.


Decerebrate Animals
'Burst' 'Burst + Inhi- Both Types
Units bition' Units Combined

Spontaneous 13.5 +/- 17.2 19.1 +/- 15.6 15.5 +/- 16.7
firing rate (n = 42) (n = 24) (n = 66)
Onset latency of
activity from 13.5 +/- 19.6a 5.6 +/- 4.9a 10.6 +/- 16.3
vein stimula- (n = 42) (n = 24) (n = 66)
tion
Duration unit
was excited by 34.0 +/- 65.0 59.7 +/- 82.7 43.3 +/- 72.4
vein stimula- (n = 42) (n = 24) (n = 66)
tion
Duration unit
was inhibited --------- 41.3 +/- 26.4 ---
by vein stimu- (n = 24)
lation
Total duration
excitability
was altered by 34.0 +/- 65.0a 101 +/- 88.6a 58.4 +/- 80.6
vein stimula- (n = 42) (n = 24) (n = 66)
tion
Total duration
excitability
was altered by 102 +/- 186 89.8 +/- 73.7 95.8 +/- 138
tibial nerve (n = 11) (n = 11) (n = 22)
stimulation
Total duration
excitability
was altered by 42.6 +/- 48.0 45.0 +/- 30.4 43.7 +/- 40.0
sural nerve (n = 10) (n = 8) (n = 18)
stimulation


aValues for 'burst' and 'burst + inhibition' units were significantly
different by Student's t-test (p < 0.05).



















Table 4-1-continued.


Anesthetized Animals
'Burst' 'Burst + Inhi- Both Types
Units bition' Units Combined


4.0 +/- 9.6
(n = 7)

6.8 +/- 6.2
(n 7)


17.4 +/- 19.8
(n = 7)


17.4 +/- 19.8
(n = 7)



18.6
(n 1)



5.1 +/- 5.9
(n = 3)


24.6 +/- 23.4
(n = 7)

4.0 +/- 4.2
(n = 7)


18.5 +/- 20.8
(n = 7)


43.6 +/- 63.3
(n = 7)


62.1 +/- 66.4
(n = 7)



75.1 +/- 81.1
(n 4)



81.5 +/- 68.6
(n = 5)


14.3 +/- 20.3
(n = 14)

5.4 +/- 5.3
(n = 14)


17.9 +/- 19.4
(n = 14)


38.5 +/- 51.9
(n 14)



63.8 +/- 64.6
(n = 5)



52.9 +/- 65.3
(n = 8)











following stimulation if the unit fires spontaneously. Thus, some of

the units with a low spontaneous rate, classified in this study as

being only excited by venous afferent stimulation, may indeed also

have been inhibited by this input.

Table 4-1 also shows the onset latencies of the responses

elicited by venous afferent stimulation and the durations that the

excitabilities of the units were altered by this stimulation. The

variances in these values were enormous. The changes in the

excitabilities of neurons elicited by stimulation of the

femoral-saphenous vein had an average onset latency of 10.6 +/- 16.3

msec following the onset of the triphasic spike of the cord dorsum

potential in decerebrate animals and 5.4 +/- 5.3 msec in anesthetized

animals. These means were shown to be significantly different by

Student's t-test (p < 0.039). The triphasic spike is a compound

action potential propagating through the largest active primary

afferent fibers (Austin and McCouch, 1955; Howland et al., 1955;

Yates et al., 1982); thus, the onset latency measured from the

triphasic spike represented the onset latency of activity following

the arrival of femoral-saphenous venous afferent input at the cord.

In decerebrate cats, there was also a significant difference (p <

0.016) in the onset latency of elicited activity in the units excited

by venous afferent stimulation and the units both excited and

inhibited by this input; the mean latency was 13.5 +/- 19.6 msec in

the former and 5.6 +/- 4.9 msec in the latter. A similar significant

difference in the onset latency of activity elicited by venous

afferent stimulation was not shown between units of the two types in

anesthetized cats.











The excitabilities of the units receiving inputs from the

femoral-saphenous vein were modulated for long durations following

vein stimulation. The mean duration the excitabilities of the

neurons were altered by vein stimulation was 58.4 +/- 80.6 msec in

decerebrate animals and 38.5 +/- 51.9 msec in anesthetized animals.

In decerebrate animals there was also a significant difference in the

duration that units excited by venous afferent stimulation and

neurons both excited and inhibited by this stimulation were affected

by inputs from the femoral-saphenous vein (p < 0.0008).

Convergence of Inputs from the Femoral-Saphenous Vein with Inputs from
Muscle and Skin on Single Spinal Neurons

The effects of stimulation of the posterior tibial nerve, the

largest muscle nerve of the hindlimb, and the sural nerve, the

largest cutaneous nerve of the hindlimb, on the excitabilities of

units receiving input from the femoral-saphenous vein were also

studied. The posterior tibial nerve innervates a number of muscles

of the crus and foot, including the triceps surae, tibialis

posterior, flexor hallucis longus, lumbrical muscles and interosseus

muscles. The sural nerve innervates skin overlying the foot and

ankle (Jefferson, 1954; Crouch, 1969). In decerebrate animals, 78

out of 80 (98%) of the venous afferent-activated units tested could

also be activated by stimulation of the tibial nerve at an intensity

10 times threshold for eliciting a cord dorsum potential; 58 out of

78 (74%) of the tested units were also driven by sural nerve

stimulation at an intensity 10 times that necessary to evoke a cord

dorsum potential; 56 out of 77 (73%) of the neurons were activated by











stimulation of both nerves. In anesthetized animals, 9 out of 13

(69%) of the venous afferent-activated units examined could also be

activated by stimulation of the tibial nerve; 10 out of 13 (77%) of

the tested units were also driven by sural nerve stimulation; 7 out

of 13 (54%) could be activated by stimulation of both nerves.

Figure 4-2 shows poststimulus time histograms generated from

data recorded from a single unit following stimulation of the

femoral-saphenous vein, the sural nerve and the tibial nerve. The

activity elicited by tibial nerve stimulation had an average onset

latency of 6.6 +/- 10.3 msec following the triphasic spike produced

by stimulation of this nerve in decerebrate animals and 3.9 +/- 2.8

msec in anesthetized animals. The activity elicited by sural nerve

stimulation had an average onset latency of 8.2 +/- 5.4 msec

following the triphasic spike produced by stimulation of this nerve

in decerebrate animals and 4.4 +/- 4.9 msec in anesthetized animals.

The venous afferent-activated neurons were excited, inhibited or both

excited and inhibited by stimulation of the sural and tibial nerves.

Most of these units (37 out of 54 or 69%) studied in decerebrate

animals were both excited and inhibited by stimulation of the tibial

nerve; 16 units (30%) were excited by tibial nerve stimulation and 1

neuron (2%) was inhibited by stimulation of this nerve. In

contrast, most of these units (22 out of 37 or 59%) were only excited

by sural nerve stimulation in decerebrate cats. Only 13 units (35%)

were both excited and inhibited by sural nerve stimulation; 2 cells

(5%) were inhibited by the stimulation of this nerve. In

anesthetized animals, 40% of the tested units were excited by tibial
























Figure 4-2. Poststimulus time histograms showing the effects of
stimulation of the femoral-saphenous vein, posterior tibial nerve
and sural nerve on the excitability of a single spinal neuron. The
histograms were each generated from 64 consecutive sweeps. The
stimulus was delivered at the onset of the traces.


I







FVA


100 msec


SURAL
kr*#NyM


spikes
spikes


TIBIAL











nerve stimulation, 40% were both excited and inhibited, and 20% were

inhibited. In the same animals, half of the units examined were

excited by sural nerve stimulation, and the other half were both

excited and inhibited by the stimulation of this nerve. There were

no obvious simple relationships between how a neuron responded to

stimulation of the femoral-saphenous vein and how it responded to

stimulation of the tibial and sural nerves. The excitabilities of

the neurons were modulated for long durations following stimulation

of the sural and tibial nerves; these durations are shown in Table

4-1.

To better determine the patterns of convergence of muscle

afferent input on femoral-saphenous venous afferent-activated spinal

interneurons, the effects of stimulation of 3 muscle nerves--the

hamstring nerve, the lateral gastrocnemius nerve and the deep

posterior nerve--were tested on 17 units in decerebrate cats. The

nerves were stimulated using current intensities 10 times that

necessary to evoke a field potential recordable from the cord dorsum.

The hamstring nerve innervates 3 flexors of the thigh, the

semimembranosus, the semitendinosus and the biceps femoris; the

lateral gastrocnemius nerve innervates 2 extensors located in the

crus, the lateral gastrocnemius and the soleus; the deep posterior

nerve innervates a number of muscles located in the foot (Jefferson,

1954; Crouch, 1969). All of the 17 venous afferent-activated units

examined could also be driven by stimulation of the deep posterior or

the lateral gastrocnemius nerve; 12 of the units could also be driven

by stimulation of the hamstring nerve. It was clear that muscle











inputs from many parts of the hindlimb and from both flexors and

extensors converged on single interneurons also receiving inputs from

the femoral-saphenous vein.

Following recordings from many interneurons, small bundles of

the L7 dorsal rootlets were cut proximally and placed across a silver

bipolar recording electrode. The tibial and sural nerves were then

exposed to the same current intensities which were necessary to alter

the excitabilities of interneurons. Compound action potentials,

recorded from the dorsal rootlets, elicited by these stimulus

intensities were analyzed to determine the conduction velocities of

the excited fibers. In most cases, low stimulus intensities elicited

A-alpha and A-beta volleys, whereas higher stimulus intensities also

elicited A-delta volleys; the results of a typical study are shown in

Figure 4-3. This figure shows the effects of different stimulus

intensities applied to the sural nerve on the amplitudes of 3

components of the evoked dorsal root action potential. In

decerebrate cats, 18 out of 29 units (62%) for which an analysis was

done were driven by stimulus intensities applied to the tibial nerve

which only elicited A-alpha and A-beta components in the compound

action potentials recorded from the dorsal roots. The other 11 units

(38%) were activated by current intensities that produced A-delta

volleys as well as A-alpha and A-beta volleys. Sixteen out of the 17

units (94%) for which an analysis was done were driven by stimulus

intensities applied to the sural nerve which only elicited A-alpha

and A-beta components in the compound action potentials recorded from

the dorsal roots. Only 1 unit was driven by current intensities that
































-C

0 1



4-i 4-

1-4 C
0 4-
w 0)
-r a
41 w
r-l





SC

1-4

Ai
"4 $0





ca
(04


4C -i




0
M (

-H 0




Sri



4,4
H M
Im











*0
0l







0
*O C
4 0)


0 cc

0
> *

-4 41


T 0


4J O-
SC


u

0


4 1 *


000





0 0

4-1 4
So



0





co
0 a







0 4>
a 0




0




0



0 WT







w 0
CCOO

0 C


ri e

4- C &


I

















U
Qa

E -
c NOV M


O\ I-










"dS3I "XVlN %
^ ^z
^ z)











elicited A-delta as well as A-alpha and A-beta volleys. The results

in anesthetized animals were quite similar. Six out of 7 units (86%)

for which an analysis was done were driven by stimulus intensities

applied to the tibial nerve which only elicited A-alpha and A-beta

components in the compound action potentials recorded from the dorsal

roots. The other unit was activated by current intensities that

produced A-delta volleys as well as A-alpha and A-beta volleys. Two

out of the 3 units for which an analysis was done were driven by

stimulus intensities applied to the sural nerve which only elicited

A-alpha and A-beta components in the compound action potentials

recorded from the dorsal roots. The other unit was driven by current

intensities that elicited A-delta as well as A-alpha and A-beta

volleys. However, at minimal stimulus intensities necessary to evoke

A-delta dorsal root volleys, the A-alpha and A-beta volleys were

often still increasing in amplitude. This was especially true for

dorsal root action potentials elicited by stimulation of the tibial

nerve. Thus, in most cases in which both A-delta volleys and volleys

along larger afferents were elicited by stimulus intensities

minimally necessary to activate a neuron, it was impossible to

conclude which fiber group carried the input which activated the

unit. Nonetheless, it was clear that large muscle and cutaneous

afferents were included in the population which carried inputs to the

spinal cord that altered the excitabilities of neurons also driven by

inputs from the femoral-saphenous vein.











Discussion

Venous Afferent Effects on the Excitability of Spinal Neurons

The response properties of the neurons driven by stimulation of

the femoral-saphenous vein were similar to those reported by others

for units activated by electrical stimulation of cutaneous and muscle

afferents. These previous studies have shown that most spinal cord

neurons, especially those in decerebrate preparations, have a

substantial spontaneous firing rate. However, units which do not

fire in the absence of stimulation have also been reported (LeBars et

al., 1975; Light and Durkovic, 1984). In addition, the onset latency

and duration of activity elicited in spinal neurons by venous

afferent stimulation are included in the range of values reported

following electrical stimulation of large fibers in the hindlimb

nerves (Feldman, 1975).

Neurons were either excited or both excited and inhibited by

stimulation of the femoral-saphenous vein. In this study, as well as

in previous studies, spinal interneurons were also shown to be

excited or both excited and inhibited by electrical stimulation of

muscle and cutaneous afferents (Hongo et al., 1966; Hillman and Wall,

1969; Wagman and Price, 1969; Feldman, 1975). Units both excited and

inhibited by venous afferent stimulation are likely to receive inputs

through several separate pathways, some responsible for the

excitation, others for the inhibition. The excitation could either

be due to direct inputs from primary afferents to the unit or to

inputs from higher-order spinal neurons driven by venous afferent

stimulation. The inhibition is also likely due to inputs from











venous afferent-driven neurons to the unit; however, the neurons

responsible for the inhibition are likely to be different from the

neurons that are responsible for the excitation. An alternate

possibility is that the inhibition is due to the venous

afferent-induced presynaptic inhibition of the tonically active

afferents responsible for the high spontaneous firing rate observed

in these neurons. However, the presynaptic inhibition of primary

spinal afferents has a time course of 200 msec or more (Eccles et

al., 1954, 1962, 1963a, 1963b; Eccles and Krnjevic, 1959; Wall,

1962). The period of inhibition produced by venous afferent

stimulation in most units had a time course of 50 msec or less; in

many units the duration was less than 20 msec. Thus, it is likely

that the inhibition recorded from most of the units following venous

afferent stimulation is due to postsynaptic inhibition, and not

primary afferent depolarization. However, in the few units that

were inhibited for long durations following venous afferent

stimulation, it is possible that primary afferent depolarization also

contributed to the reduced firing rate. Data recorded from one such

unit are shown in Figure 4-1.


Since units

afferent inputs

of pathways, the

these units appe

manner following

that venous affe

the excitabilitie


that were both excited and inhibited by venous

were likely to receive these inputs through a number

Svenous afferent influences on the excitabilities of

*ar to be powerful. Units that responded in this

venous afferent stimulation were common, suggesting

*rent inputs, in general, have a powerful impact on

*s of spinal cord neurons. Units both excited and











inhibited by venous afferent inputs also probably are involved with

the integration and processing of these inputs, although the purpose

of this integration cannot be determined in studies using electrical

stimulation. In contrast, units that are only excited by an afferent

input are more likely to have a role of relaying the input without

significantly processing it. These latter units may fire for a long

duration after receiving the afferent input, however, thereby

prolonging its effects in the nervous system.

Stimulation of the femoral-saphenous vein had prolonged effects

on the excitabilities of the units examined in both decerebrate and

anesthetized animals. This is further evidence that the inputs

carried by the large femoral-saphenous venous afferents have a potent

influence on the excitabilities of spinal neurons.

Convergence of Inputs on Single Neurons

Stimulation of A-beta femoral-saphenous venous afferents

activated many spinal interneurons that could also be driven by

electrical stimulation of muscle and cutaneous afferents. Previous

studies of the properties of single spinal interneurons that received

noxious inputs from the internal environment showed that widespread

convergence of inputs from muscle and skin occurred on these units as

well (Pomeranz et al., 1968; Hancock et al., 1973, 1975; Gokin et

al., 1977; Guilbaud et al., 1977; Foreman and Weber, 1980; Milne et

al., 1981; Blair et al., 1981, 1984; Cervero, 1982; McMahon and

Morrison, 1982; Rucker and Holloway, 1982; Takahashi and Yokota,

1983; Rucker et al., 1984). Pomeranz et al. (1968, p. 528), in fact,

"failed to detect cells which responded only to visceral afferents"











in the thoracic spinal cord. Thus, spinal neurons which processed

venous afferent inputs were similar to those that process noxious

inputs from the viscera, in that both types of neurons received

widespread convergent inputs from muscle and skin.

Comparison of Responses in Decerebrate and Anesthetized Animals

The responses of spinal neurons following venous afferent

stimulation differed somewhat in decerebrate and in anesthetized

preparations. These differences could be explained by a generalized

lower neuronal excitability in the latter animals. The onset latency

of the activity elicited by venous afferent stimulation was

significantly shorter in anesthetized cats than in decerebrate cats.

Inputs to neurons mediated through multi-synaptic connections are

more likely to be affected by a lowered state of excitability than

inputs mediated through connections comprised of only a few synapses.

Thus, neurons activated at long latency would be expected to be more

affected by anesthesia than neurons activated at short latency. The

mean onset latency of elicited activity in the neurons in

anesthetized animals would accordingly be expected to be shorter than

in decerebrate animals.

However, many of the properties of the venous afferent-activated

units were similar in both anesthetized and decerebrate preparations.

In both groups of animals the convergence of venous, muscle and

cutaneous inputs on single neurons was common. This suggests that

muscle and cutaneous inputs have powerful influences on the

excitabilities of these units. It was also common for units to be

both excited and inhibited following venous afferent stimulation in








69

the two types of preparations. As discussed above, it is likely that

these units which responded to inputs from the femoral-saphenous vein

in a complex manner are involved with the integration and processing

of the inputs. Thus, it appears that venous afferent inputs are

highly processed in the spinal cord by interneurons.


















CHAPTER V
THE VARIABILITY IN THE RESPONSES OF NEURONS LOCATED IN DIFFERENT
LAMINAE FOLLOWING STIMULATION OF THE FEMORAL-SAPHENOUS VEIN

Introduction

The data presented in Chapter IV suggested that spinal neurons

respond in a wide variety of ways following the electrical activation

of A-beta femoral-saphenous venous afferent fibers. Some neurons

were excited by stimulation of the femoral-saphenous vein; others

were both excited and inhibited. Some units were driven within a few

milliseconds following stimulation of the vein; others were driven

only at long latency. The excitabilities of some neurons were

affected only a few milliseconds by vein stimulation; the

excitabilities of others were modulated for over 100 msec. In

addition, most neurons driven by stimulation of the femoral-saphenous

vein could also be activated by stimulation of muscle and cutaneous

afferents; however, a few units were not activated by the stimulation

of hindlimb nerves. It would be of interest to determine whether

the neurons that responded in similar ways following stimulation of

the femoral-saphenous vein had similar locations in the spinal cord.

The study discussed in this chapter examined this possibility.

Methods and Materials

Experiments were performed on 12 unanesthetized, decerebrate

cats with spinal cords transected at T12 and 5 intact cats

anesthetized using alpha-chloralose. The responses recorded from











these animals were described in Chapter IV; this chapter will

describe the locations of the units with different response patterns

following stimulation of the femoral-saphenous venous afferents,

tibial nerve and sural nerve. Surgical procedures used to prepare

animals for recordings and the techniques used to record single unit

responses were described in detail in Chapter IV. The data described

in this chapter were recorded using microelectrodes filled with an 8%

solution of HRP in tris-buffered 0.15 M KC1 (pH 7.3).

After a neuron had been characterized extracellularly, an

attempt was made to penetrate the element studied. If the attempt

was successful, HRP was iontophoretically injected with 1-3 nA

positive-going rectangular pulses, 500 msec in duration, at 1 Hz.

Current was injected through the electrode for 7.5-15 min and was

delivered through the companion bridge unit (Winston BR-1) to the

microelectrode pre-amplifier. This bridge unit had been

custom-modified by the manufacturer to make it capable of presenting

high voltages. If the characterized neuron could not be impaled, HRP

was iontophoresed extracellularly using 5 nA pulses for 20-30 min. A

blank electrode was inserted into the opposite side of the cord

parallel to each track which yielded data. Once positioned, the

blank electrode was broken using scissors so that only 1-2 mm

protruded above the surface of the cord. These electrode tips were

left in place through the fixation process; they produced easily

observed tracks in the tissue which were used to help identify

sections from which data were recorded. A rostrocaudal distance of

at least 1 mm separated sites of HRP iontophoresis.











After at least an hour following the last marking attempt, the

animal was perfused through the carotid artery with 1 1 of

heparinized saline containing 0.1% sodium nitrite followed by 2 1 of

2.5% glutaraldehyde/1.2% paraformaldehyde fixative. The L6 cord was

removed and stored overnight in the fixative. The tissue was cut

into sections 50 micrometers thick using a vibratome (Oxford). The

sections were then incubated for 30 min at room temperature in a

solution comprised of 0.05% 3, 3' diaminobenzidine tetrahydrochloride

and 0.1% hydrogen peroxide in 0.1 M tris buffer at pH 7.6. The

sections were then rinsed in 0.1 M phosphate buffer (pH 7.4), mounted

onto slides, dried, counterstained using 1% neutral red, dehydrated,

cleared in xylene and coverslipped. Two-dimensional camera lucida

reconstructions were made of stained neurons and extracellular

recording sites.

Pugh and Stern (1984) have shown that HRP can be very reliably

used to label extracellular single unit recording sites. Examples of

histologically reconstructed recording sites are shown in Figure 5-1.

Injection site diameters ranged from 50 to 250 micrometers. In

addition to the visible brown reaction product in the extracellular

space, internalization of HRP by neurons provided positive

identification of the center of the injection site which was readily

distinguishable from damaged vascular or neural elements mechanically

produced by the electrode.

Results

Fifty sites at which responses were recorded from venous

afferent-activated neurons were reconstructed; 36 of these




























wtw 0 0 0 0 0


r-I 3 00 ,0 !
Q-l 0 U f 0
0 a) A a a

u0. 0t r-4 Ic cO
0 C 4 r-l ,Cd 0
1-40 00 < 1 w
m .o A i-I




4 0 w r4 0 0i

4 10 a 41 4-1 r
t- A0 0x u

3 34 -I 0 .0 0I


) n a) 03 0: t
Sr-I 0 -H r -I
1.4 ( W 4 0 0



O0 0 1.4 0 C 0
,C 'O co j O





01 r-4 U C l
4j rl Oa 4u 3 4 r-
w-11 0 0 3 r-
M 2 U 0 W 0 Q

S0 0 w .I w
1 o A 0 0 4 OD
bO4 ,o E-4 Co a J
"1 0 cw4 .- 00
0 H 0 4-1 00
-H 0 -B 3 0
O3 T- 4-I 0 4 .0 O
m C -r14

0J W (1 40 4 0



0 1#4 0 0 00
00 *o Q --. 00 0q
0 u 10 4 -1
u -4 4 0.3 u 000
41C 0 *CO4
u l 0 0i w
0 a) w o A u w c0
0 ,C *4 )





O0 0 0 0 0
0- 4 *re 4o
a4 0 0 ) a)0 ,0
4 0 P -1 4U CO
aj -41 *w 4 41 0 (4
bO0 44JI O r -4



4 I 04- co i 10 0 o 0


i-i a *r



S*r 0 O ia 0 (0 ( 0

















'
w.


s


%A II


**
9.


dit
4e
5'


4*


Ie


S** 0


6
I


! ET-
9 .


i

J7\1


* .


'C


.if


1
1)











reconstructed recording sites were in decerebrate animals and the

other 14 were in anesthetized animals. The locations of the

recording sites are shown in Figure 5-2. In decerebrate cats,

neuronal activity elicited by stimulation of the femoral-saphenous

vein was recorded from Rexed's laminae IV-VIII and X; in anesthetized

cats, venous afferent-elicited neuronal activity was recorded from

laminae V and VI and from the dorsal portion of lamina VII. Most of

the recording sites were located either in lamina V (11 out of 50) or

in the most superficial portion of lamina VII, dorsal to the

dorsalmost border of lamina VIII (15 out of 50).

Units located in the different laminae had similar spontaneous

firing rates; these values are shown in Table 5-1. However, the

units which were excited by venous afferent stimulation and the units

both excited and inhibited by vein stimulation were highest in

density in different regions of spinal gray matter. The locations of

the units of the two types are shown in Figure 5-3. Most of the

units in the dorsal horn (12 out of 20 or 60%) were both excited and

inhibited by stimulation of the vein. In contrast, most of the units

in the ventral horn (22 out of 30 or 73%) responded to venous

afferent stimulation with only a burst of action potentials. Only 2

units located deeper than the dorsalmost border of lamina VIII were

both excited and inhibited by venous afferent stimulation; clearly,

units of this latter type were rare deep in the ventral horn.

An investigation was also done to determine the locations of the

neurons driven at monosynaptic latencies by femoral-saphenous venous

afferents. The cord dorsum potential elicited by venous afferent































0




0





*o











r.
0




4-









o
0

















r i












o
4 J
0















0
0
0















-r
0


PO CO










u ia

00





cu I
3 a
b0'l
Tlu-
O






























AJ=

(.)
0
LU -3x



















Table 5-1. Spontaneous firing rates of neurons located in different
laminae. Confidence intervals represent mean +/- standard deviation.
All spontaneous firing rates are in spikes per second.




Lamina of Mean Spontaneous Range of Spontaneous
Rexed n Firing Rate Firing Rates


IV 2 17.5 +/- 21.1 2.5 32.4

V 11 8.8 +/- 11.7 0.0 33.3

VI 7 15.6 +/- 12.9 1.1 37.2

VII 22 20.8 +/- 21.0 0.0 60.0

VIII 5 9.9 +/- 9.6 0.0 20.8

X 2 5.0 +/- 7.1 0.0 10.0
































>1


*a


*0

-14









C
J-l
(0
0a

,4









0
e 4*u



41


ca


1-1
0
#0

0)



41




0
a o

m m

001

I0
0) (0













au














C
O



CIO


















I
c)
GQ


u-I
LL.
N

I--
I
CO


-o












LU





u.
LI

(a)
U

LU

I











stimulation was used to estimate this latency at which units were

first activated by inputs from the vein. The cord dorsum potential

is comprised of an initial triphasic spike, a series of negative

waves and a slow positive wave. The triphasic spike is a compound

action potential reflecting activity in the largest afferent fibers

excited by the stimulus, the negative waves are generated by dorsal

horn interneurons, and the positive wave is produced by the

presynaptic inhibition of afferent fibers (Austin and McCouch, 1955;

Howland et al., 1955; Eccles et al., 1963a, 1963b; Willis, 1980;

Yates et al., 1982). Thus, the latency of the first negative wave of

the cord dorsum potential corresponds to the latency at which

interneurons are being directly excited by large primary afferent

fibers. Neurons excited at the same latency as that of the peak of

the first negative wave of the cord dorsum potential, or at shorter

latencies, were assumed to receive monosynaptic, or at least

relatively direct, inputs from from primary femoral-saphenous venous

afferents. Figure 5-4 illustrates these procedures used to determine

which units received direct inputs from the venous afferents. The

average latency of the peak of the first negative wave of the cord

dorsum potential was 2.67 +/- 0.71 msec from the onset of the

triphasic spike in decerebrate animals and 2.06 +/- 0.85 msec in

anesthetized animals.

Based on these criteria, 8 of the 50 units with recording sites

reconstructed were classified as being directly excited by the

primary venous afferents; 4 of these units were in anesthetized cats,

and the other 4 were in decerebrate cats. The locations of these



























4J 0 41 w w 0 41 J >J 4 0
or4 m U 0 ( w e
E-4 W 1-4J 4 4 0 0


0 -4 0 10 34 m
> A =, .( 4 wc.,
-H 4 41 o 4 0 0, 0
4JC 0 -4 4 0 0 W ) blD




SL0 4J 0 ., 4J
0 04 ) 0 i b O
S,44 0 0 0 C 1 C
S( 4 c. 0 r -,

CL 0 M o o o a rH.

0 CO > o >1.M
0 0 0 *0.
C O 0 .4U O .-I t A
3 0 *0 A O w 0 404 00
> 0 0 4 0 b0
-0 K -) 4 0 .4) 00 W 0

0004 40.04-4

0 C 0 0 0 0 ,C 10
S0 41 u wi to 0".40 0


0 rU a- W -H M 4 0
3 %. 0 > 0 r- 41 L
AL01 J 0 C .
W0 .C 4-1 4 -H bO
I q 1: U > v 0. TOa w 4 (0
)< -1 U -H tH "0 -r4 C W S
*B q : W 0 0 0 f>



wo e m.0 o 01.
S0 1 I 4 4- 0, uT

u Z C > 3 0. *r0

44 V 94W 00 4J
00 00 4 01

09) il 3 M o f a .3 ,0


0) 4J l 0 "o0

4 0 4 01 U 4 CO 10
o Q 4. A 0 41o .0
w 0 rW 0 C:
O O w I 0 0B 4

4 J C r9. "-) GM I l C

4 0 OO 000 b0W a) 0 0
4 1 M0 -H
00 C6 00 4 0

to i0 bo M t 1i) c o
W 4 41 41 ca 0 4J1 0) -g -
) C 4l WU) 0) O O bCI rtv
Hal tO 0 > 4C D CL 4) O
wo 0CUM o0 c &j 0 *w -41







83









> > >







40 0V lOspik



E
(I






e1!dds

410 4 2 so'd

>1
4-j
n"U"js Z) 4


E
C


t-5
-oc
015
Umo










units are shown in part "A" of Figure 5-5; most (6 out of 8 or 75%)

were located in lamina V, 1 was located in lamina VI and the other

was located much more ventrally in lamina VII. These units driven at

short latency by venous afferent stimulation in decerebrate animals

had a mean spontaneous firing rate of 8.2 +/- 7.2 spikes per second

and were activated at a mean latency of 1.8 +/- 0.72 msec following

the triphasic spike of the cord dorsum potential. Two of the units

studied in decerebrate animals were excited by venous afferent

stimulation, and the other two were both excited and inhibited. The

excitabilities of these units were modulated for 32.6 +/- 44.2 msec.

The spontaneous firing rates of the units driven by venous afferent

stimulation at short latency in anesthetized animals were much lower

than those in decerebrate cats; the mean rate was only 0.24 +/- 0.28

spikes per second. Two of the units in anesthetized animals were

silent in the absence of stimulation; it was impossible to determine

whether these units were both excited and inhibited by venous

afferent stimulation or only excited. It is only possible to show

that the firing rate of a unit decreases below the spontaneous level

following stimulation if the unit fires spontaneously. The other two

units in anesthetized cats were both excited and inhibited by vein

stimulation. These four units were activated at a mean latency of

1.06 +/- 0.55 msec following the triphasic spike of the cord dorsum

potential; the excitabilities of the units were modulated by venous

afferents for a mean duration of 20.4 +/- 12.8 msec.

The locations of the units driven at longer latencies by the

stimulation of the femoral-saphenous vein are shown in parts "B" to




























-l 4- 4J 4)
0 0W

U

f*i 08 0
po 0u we
e- ca 14

Q1 Ur-4
9 0
04 4-
4-4 H U)
1 0 *r U

41
4 4-4 CC



w 000
0 41 04-
> 0ca 04
a a 4

-H0 0 4.10


0 .






I o





4 -4 >

cOr
0'








*r I .
rlC 0



O0








44-4 0 -H
0 3 0 04-
i > 41
04 4J 0)









Vi4 0 0
> rT a)
i-l C
1


















U
0
4V
E




O


0




C,
V


E
CO



0











"E" of Figure 5-5. Within 3 msec following the triphasic spike of

the cord dorsum potential, but at latencies longer than that of the

peak of the first negative wave of the cord dorsum potential, a few

units located in laminae VI and VII were activated by vein

stimulation. At slightly longer latencies, many units located in the

ventral horn were activated. However, units in the dorsal horn were

also activated at these longer latencies. The onset latency of

neuronal activity produced by venous afferent stimulation ranged to

over 12 msec following the arrival of the afferent volley at the

cord; neurons activated at latencies longer than 12 msec following

the triphasic spike were located both in the dorsal and ventral horn.

Figure 5-6 shows the durations that the excitabilities of units

were modulated following stimulation of the femoral-saphenous vein;

neurons were activated from less than 10 msec to over 100 msec. The

excitabilities of units located in the dorsal horn and in the

superficial portion of lamina VII were modulated for both long and

short durations. However, most of the units located deeper than the

dorsalmost border of lamina VIII were only activated for short

durations (less than 50 msec). The excitabilities of units located

in the dorsal horn and superficial portions of lamina VII of

decerebrate cats were modulated for a mean duration of 74.8 +/- 86.1

msec following vein stimulation, whereas units located deeper were

activated for only an average duration of 26.2 +/- 44.5 msec. These

means were shown to be significantly different by Student's t-test (p

< 0.035).



















Figure 5-6. Locations of units activated for different durations
following the stimulation of the femoral-saphenous vein.








*-DECEREBRATE o-ANESTHETIZED
<10 msec 10-49 msec


50-100 msec











As discussed in the last chapter, most of the venous

afferent-driven units could also be activated by stimulation of the

tibial and sural nerves; however, a few units were not driven by the

stimulation of hindlimb nerves. Table 5-2 shows the locations of the

venous afferent-driven neurons that also received convergent inputs

from muscle or skin. In anesthetized animals most of the units which

did not receive convergent inputs from muscle or skin were located in

lamina V, the region which contains most of the neurons driven at

shortest latency by stimulation of the femoral-saphenous vein.

Accordingly, 2 of the 3 units in anesthetized animals that were not

driven by stimulation of the sural nerve were activated at

monosynaptic latencies by stimulation of the femoral-saphenous vein.

The other unit was driven by vein stimulation at a fairly short

latency (3.6 msec following the onset of the triphasic spike of the

cord dorsum potential). Two of the 4 units in anesthetized animals

that were not driven by stimulation of the tibial nerve were

activated at monosynaptic latencies by stimulation of the

femoral-saphenous vein. The other two units were activated at 3.6

and 3.8 msec following the triphasic spike of the cord dorsum

potential. Only one unit studied in decerebrate animals, located in

lamina VII, could neither be activated by stimulation of the sural

nerve or the tibial nerve; this unit was driven at monosynaptic

latencies by stimulation of the femoral-saphenous vein. However, the

other four units characterized in decerebrate cats that were not

driven by sural nerve stimulation were activated by venous afferent

stimulation only at long latency (11.5 +/- 1.9 msec following the































I 0 0
I 0o
r-4


I Lf C4 10 I


o)o
I 0 0
I -< 0
r-l


(Y) I
0 I1
oo I


r-l
a



4.1






o
1-4
C41
Od


O-4

4..
u 0
u
(a


-H


U 0
U W
r-4 141







a
0>

> 0

co



O.10
00


:a




OW
r< $
0w
-C1
*
w I





0
H 0
U
0 0



. 02


< U
hc) t-
E-40


C oN 0

C 0) 0




i-E 4 l Z3

0H
H z



0) wU .0
4-loC
N *







4.1 C U
w0 0--





4000


O' 1.-1



S41







0) 0
41 *






w 0
1 00
1- r. 4





0 De
u wd











9C0
UO
owii


0 0 0
0 0 0
r- rl -


C 0o
r-l


NC. 1.0 UL, -4
-4


> H
S> > >


CC
U>

PQ


(D
Om
Z E-


S>

U >

PQ



0 3
am
Zg -


W
iS
I0c
ZE-4
4-I

0>



0) 14



a H
3Wa


I if) C14 '


0 r- 0 0 0
Q 00 m 0
i-E




c-I \ Ln 0 a-0
(n9Q~-4




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E9HW3OHLP_GJ5G8G INGEST_TIME 2012-08-13T17:45:19Z PACKAGE AA00011760_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES