Localization of lumbar epaxial motoneurons in the rat

Brain Research, 170 (1979) 23-41 © Elsevier/North-Holland Biomedical Press

23

L O C A L I Z A T I O N OF L U M B A R EPAXIAL M O T O N E U R O N S IN T H E R A T

EMILY E. BRINK, JOAN I. MORRELL and DONALD W. PFAFF The Rockefeller University, New York, N. Y. 10021 (U.S.A.)

(Accepted November 2nd, 1978)

SUMMARY In the rat, lateral longissimus and the lumbar transverso-spinalis muscles are lumbar trunk muscles, and participate in the lordosis reflex (female sexual posturing). Medial longissimus, the remaining major lumbar epaxial muscle, is a muscle of the proximal tail-tailbase. To allow an analysis of the motor control of lordosis, motoneurons for these muscles must be localized. Horseradish peroxidase (HRP) was injected into lumbar epaxial muscles of female rats. Following injection into medial longissimus, labeled cells were found posterior to the lumbar enlargement, ventrolaterally in the ventral horn. Following injections into lateral longissimus or into lumbar transverso-spinalis muscles, most labeled cells were found on the medial side of the ventral horn, extending through and anterior to the lumbar enlargement. Injections into lumbar transverso-spinalis muscles at more posterior levels led to more posterior locations of labeled cells. The distributions of the labeled cells agree with previous observations on innervation of the muscles. Additionally, the spinal cord was scanned for sites at which microstimulation produced visible twitches of medial longissimus or transverso-spinalis muscles. Locations of low threshold twitch sites were consistent with conclusions based on the H R P findings.

INTRODUCTION In the rat, the epaxial muscles are important for the performance of lordosis behavior, the hormone-dependent female sexual posturing. Specifically, lateral longissimus and the lumbar transverso-spinalis muscles are anatomically suited for, and are capable of producing, dorsiflexion of the lumbar vertebral column iz. Ablation of these muscles leads to severe deficit in the amount of rump elevation displayed in lordosis n. The elevation of the rump and tailbase during lordosis behavior is a biologically important component of the behavior: this movement exposes the perineal skin and permits the male to intromit, leading to fertilization a4.

24 Vertebral muscles are supplied by the medial cell column of the spinal cord (see ref. 38). In the enlargements, the medial cells are clearly separate from the lateral column of cells4,14,15,17,26,35,36,z8. Outside of the enlargements, there is little distinction between medial and lateral cell columns; motoneurons innervating epaxial muscles lie throughout the ventral horn 38. Little information is available on the location, within the medial cell column, of motoneurons innervating individual epaxial muscles. In the present study, two approaches are used to localize lumbar epaxial motoneurons of the rat. The major approach is neuroanatomical localization using the method of retrograde intraaxonal transport of horseradish peroxidase (see refs. 28-30) injected into muscle. The second approach used is physiological localization by microstimulation1-3,5,6,2a,32,a7,a9: the spinal cord is scanned for sites at which stimulation at low current strengths produces visible muscle twitches. The muscles studied are the lordosis-important lateral longissimus and the lumbar transverso-spinalis muscles. For comparison, the motoneurons of the third major lumbar epaxial muscle, medial longissimus, are also localized. Unlike lateral longissimus and the lumbar transverso-spinalis muscles, medial longissimus is not a trunk muscle, but, rather, a muscle acting upon the proximal tail-tailbase lz. METHODS

Localization by the horseradish peroxidase technique Thirty-four female ovariectomized albino Norway rats weighing 314-352 g were used in the present study. Horseradish peroxidase injection. Rats were pretreated with atropine (0.12 rag, s.c.) and were then anesthetized either with halothane applied via a nose cone, or with Equithesin (a combination of sodium pentobarbital (Nembutal) and chloral hydrate) injected intraperitoneally (0.25 ml/100 g body weight). For horseradish peroxidase injections, 6-20 ~ solutions of horseradish peroxidase (Sigma Type VI, lot numbers: 45C-9530-1, 124C-9610-1, 25C-9570 and 65C9530-1 (animals nos. 34, 35, 36, 37)) were made up in distilled water 31. (No differences associated with HRP lot numbers could be detected above experimental variability.) The lumbar epaxial muscles medial longissimus (ML), lateral longissimus (LL), or muscles of the lumbar transverso-spinalis system (TS) were exposed as described elsewhere 11. Injections of the horseradish peroxidase solutions were made at multiple sites into the muscle under study, using a 1 ml syringe calibrated to 0.01 ml, with a 26or 27-gauge needle. The HRP solution was injected slowly, until the muscle became discolored (turned brown); care was taken to avoid spread to adjacent muscles as indicated by discoloration of those muscles. In some cases, leakage did occur during injections; the leaked HRP was immediately wiped up, and did not produce discoloration of adjacent muscles as observed during injections. Generally, injection into the depths of the muscles was avoided because of poor visualization, and poor separation of muscles (common origin from vertebral processeslZ). The injected muscle was surrounded by gelfoam, the wound closed, and the animal returned to its home cage. Experimental protocols are indicated in Tables I, lI and lll.

25 Additional attempts were made to reduce possible spread of HRP. First, injection volume and amount of HRP were varied (Tables I, II and III). Secondly, for better exposure of the muscle under study, adjacent muscles were ablated prior to the HRP injection. Muscle ablations were performed as described elsewhere 11. Three rats were prepared for HRP injection into ML by prior ablation of the ipsilateral LL and TS muscles. Five rats were prepared for TS injections by prior ablations of the ipsilateral LL and ML. Four rats were prepared tor LL injections by ablation of ML and TS. Ablations were performed two weeks prior to injections to allow healing of wound edges. Tissue processing. Tissue was processed according to the methods of Graham and Karnovsky18 and Nauta et al. al for the brown reaction product, as modified in this laboratory. Two days were allowed for retrograde transport and accumulation of the HRP 22,29. Rats were re-anesthetized, perfused with 0.9 ~ saline, and then perfused with 1 ~ glutaraldehyde-2 or 3 ~ paraformaldehyde in 0.05 M phosphate buffer (pH 7.4 at 24 °C). Perfusion was often immediately preceded by an injection of heparin (0.1 ml) into the left ventricle. Spinal cord was removed and postfixed for 4-6 h at 4 °C, then washed with phosphate buffer and kept overnight at 4 °C in phosphate buffer containing 30 sucrose. Frozen transverse sections of 100 #m were cut and alternate sections were preincubated in a solution of 3,3'-diaminobenzidene (Sigma) in Tris buffer (30 min). Hydrogen peroxide was added, and the sections were gently agitated in this reaction system for 30 min. The sections were then transferred through several washes of Tris buffer, and of distilled water. Sections were mounted on slides, dried, and lightly counterstained with cresyl violet. Alternate sections were mounted after cutting and frozen for later reaction. Analysis. Sections were examined by bright- and dark-field light microscopy for c~lls containing the brown reaction product. Cells were called positive (labeled HRP cells) if brown granules described a cell-shape (cell body and often proximal processes) and outlined a clear (non-reacted) nucleus2a,a0,al (Fig. 1). Cell location, size, number and strength of staining were noted. All sections containing labeled cells were drawn and all HRP cells charted. To estimate cell size, the maximal and minimal soma diameters were measured with an eyepiece micrometer, and the average calculated (average soma diameterla): no correction for shrinkage was made. For analysis of cell location within the spinal cord, the following definitions were made. The posterior end (cut-off) of the lumbar enlargement was defined as the earliest appearance, as sections were examined in a posterior-to-anterior direction, of the most caudally appearing motoneuron group of the lateral cell column of the enlargement (see Fig. 2, section labeled P). The anterior end (cut-off) of the lumbar enlargement was defined by the rostral disappearance of the remaining distinct cell group (section A in Fig. 3) of the lateral cell column of the enlargement. To further e3tablish anterior-posterior levels of the spinal cord, maps of the spinal cord, vertebrae, and entering dorsal roots were drawn in the present experiments (at the time of cord removal following perfusion) and in related experiments. Reference was also made to previous descriptions of the rat spinal cordX6,25-2L

26 In the lumbar enlargement of the rat, medial and lateral cell columns are clearly separated, and no further definition of medial and lateral is necessary. To define medial and lateral positions within the ventral horn outside of the enlargement, the following procedure was used. The dorsal limit of large ventral horn cells was determined on the charted (transverse) sections (e.g. see Fig. 2). Then, a line equally subdividing medially-laterally the region of large cells was drawn along the long axis of the ventral horn of the charted section. Labeled cells medial to this line were called ventromedial; those lateral to the line were called ventrolateral. The sections charted in Figs. 2-5 were chosen to reflect the average number of cells per section at the particular anterior-posterior level. Thus, the sections reflect relative cell density. Cut nerve experiments. In 8 rats, the small superficial nerve to ML was cut and placed in gelfoam soaked with HRP (3 or 25 ~ in 0.9 ~ saline). In a ninth animal, H R P crystals were stuffed into the cut end of the nerve and the nerve was then placed in HRP-soaked gelfoam: this procedure had produced numerous labeled cells when used on the sciatic nerve in preliminary experiments. No cells meeting criteria for HRP label were found in the first 8 preparations; in the ninth, 2 labeled cells were found. The first 8 preparations may be used as control preparations indicating no interference of endogenous peroxidase or catalase 42.

Localization by microstimulation technique Experiments were performed on ovariectomized female albino Norway rats, weighing 290-370 g. Rats were pretreated with atropine (0.12 mg, s.c.) and were anesthetized with urethane (1.5 g/kg body weight) injected intraperitoneally. ML, the lumbar TS, and extensor caudae medialis (ECM, the caudal continuation of the TS system 12) were exposed for observation. (Since adequate exposure of LL would require damage to that muscle and to ML, investigation of twitch sites for LL was not pursued beyond initial experiments.)A Tla-L2 laminectomy was performed, the dura removed (in later experiments, the pia also), and a map of the spinal cord drawn. The rat was mounted in a stereotaxic apparatus, with head held fixed by earbars, and the abdomen supported. The vertebral column at the level of the laminectomy was fixed by applying vertebral clamps to the spinous processes of T13 and Lz. The cut edges of skin were drawn up to permit formation of an oil pool: the exposed spinal cord and muscles were covered with warm mineral oil. Rectal temperature was monitored and maintained at 36-37 °C by automatic adjustment of a heating pad, and by periodic exposure of the preparation to a heating lamp. Electrocardiogram and skin color were monitored as indicants of the status of the animal. The rat was given repeated doses of atropine, and supplemental urethane as necessary. Stimulation and procedure. Glass micropipettes of outer tip diameter 1-4/~m and resistances of 0.8-5 Mr2 were used to stimulate within the spinal cord. Pipettes were filled with 2 M NaCI saturated with Fast Green FCF dye; in some experiments pipettes filled with 2.5 M NaC1 were used. Typical stimuli were 0.1 or 0.2 msec monophasic negative pulses delivered at 5/sec against an indifferent electrode inserted ipsilaterally

27 into the skin. Stimulus strength was measured on an oscilloscope as the voltage drop across a 10 Kf~ resistor placed in the return path to ground of the stimulus circuit. The cord was scanned with l0 or 20 pA pulses to locate twitch-producing sites. Thresholds for visible twitch were determined at 25-100 /,m intervals in one downward or upward pass along the electrode tracks. Threshold for visible twitch was determined while observing the muscles through a dissecting microscope; threshold was taken as the current strength which produced discrete, just-visible, twitches to about one half of the pulses delivered. Tracks were marked by dye deposition 41 at the low threshold site or at known points along the track, or were marked by lesion. At the termination of an experiment, the rat was perfused with l0 ~ formalin and appropriate sections of spinal cord were removed for histology. Frozen sections of 50 or 100 # m were cut and stained with cresyl violet. The track marks were identified and the tracks reconstructed. For track reconstruction, a 10 ~ tissue shrinkage factor was allowed (estimated from Baker a and from tracks having multiple dye deposits). In tracks reconstructed from track damage, a best estimate of the track was made and track depth approximated from 0 at the surface; a localization error factor of ~ 100 or 200 #m dorsoventral and/or mediolateral was allowed. Estimation of stimulus spread. Effective current spread may be estimated from the rate of increase of threshold with distance from the minimal threshold site (as stated by Gustafsson and JankowskaZl). For stimulating myelinated axons within short distances (within several hundred/~m, such that the stimulus is effective at only one node), the distance over which the stimulus is effective has been estimated to be linearly related to the stimulus intensity: I (current strength) = AD, where D is distance and A is a constant a. In other estimates of stimulus spread, effective stimulus spread is estimated to be related to the square root of the stimulus intensity: I (current strength) ---- kD z (refs. 2, 5, 7, 39). In the present experiments, stimulus spread was estimated according to both the linear and the square relations, for a number of tracks. In all cases, the calculated spread at the low threshold spots was greater when the data were treated by the square relation. Therefore, the values obtained using the square relation were used. The conversion factor for spread estimation (the k value) was derived for each track and was used to calculate the spread for that track. When insufficient data were available to determine the k value of a track, the average k value from tracks made under the same stimulus parameters was used. For the main stimulus conditions (0.2 msec pulse, 5/sec) k values ranged from 140 to 2440 #A/sq.mm, with an average value of 870/,A/sq.mm (N -----27 tracks). For these conditions, estimates of linear spread gave conversion factors (1/A) of 2.5 to 11 /,m/#A, with an averaged value of 8/zm//zA (N = 18). The k values are comparable to those obtained by Stoney et al.Za; the 1/A values are similar to those obtained by other authorsl,a2,a7. Data presentation. For each muscle under study, sections containing positive tracks were drawn and the low threshold sites identified as described. Circles describing effective current spread at the low threshold sites were drawn. For tracks with poor localization, circles were moved ~: 100 or 200 /zm according to the localization error factor, and the resultant limits of spread fitted by rectangles or squares.

28 RESULTS

HRP experiments HRP-labeled cells were marked by the granularity of the label. The staining ranged in all preparations from excellent to very light. Excellently stained cells were distinguished by dark brown granules and by processes labeled for a distance from the cell body (Fig. 1). Clear nuclei could be seen (Fig. 1). Medial longissimus (ML). Following injection of HRP into medial longissimus (Table I), the majority of HRP-labeled cells formed a continuous ipsilateral string of cells lying posterior to the lumbar enlargement. Typically, labeled cells extended caudally from just posterior to the lumbar enlargement through to at least the $1-$2 dorsal root entry zone levels. The labeled ML cells tended to occupy a ventrolateral

A

29

C Fig. 1. Bright-field photomicrographs of cells labeled for horseradish peroxidase following injection into lumbar epaxial muscles. ( × 570). A: HRP-labeled cells following injectioninto lateral longissimus. B: HRP-labeled cells following injection into lumbar transverso-spinalis muscles. Cells are counterstained with cresyl violet. C: HRP-labeled cells following injection into medial longissimus.

position in the ventral horn. This pattern, which was highly consistent, is illustrated in Fig. 2 for one animal (no. 25). The labeled cells were fairly evenly distributed anteriorly-posteriorly throughout the region in which they were found (Fig. 2). A few HRP-labeled cells were sometimes found medially or laterally in the posterior part of the l u m b a r e n l a r g e m e n t (Table I). These cells were well-separated, typically by 3-4 m m , from the b o d y of labeled cells lying posterior to the enlargement. (The location of the lateral cells was similar to that of lateral cells labeled after some posterior TS injections (Fig. 5).) The facts that these cells were well-separated from the m a i n b o d y of cells, t e n d e d to be more lightly labeled, a n d were i n some experiments TABLE I Medial longissimus injections Expt.

ML 13 ML 15 ML 16 ML 22 ML 25 ML 30 ML 34

Surgical condition

Intact Intact Intact Intact Ablated Ablated Ablated

Amount H R P (rag)

40 30 50 40 20 20 20

Injection volume (ml)

0.75 0.5 0.4 0.3 0.2 0.2 0.2

Cord examined Total Location labeled cells (vert. levels) cells Posterior Enlargement

L~-midiLz Lt-mid L2 Lt-mid L2 Ll-mid Lg L~-upper L2 Ll-upper L~ Ll-upper L2

33 25 20 39 51 28* 73*

VM

VL

M

L

3 6 2 4 0 3 3

29 19 15 27 51 20 50

0 0 0 0 0 5 20

1 0 3 8 0 0 0

* In ML 30 and 34, labeled cells extended to posterior limit of tissue sampled.

30

Fig. 2. Location of HRP-labeled cells in the spinal cord following injection of HRP into medial longissimus. Representative sections from a single, typical animal (no. 25) are presented. No cells appeared anterior or posterior to the sections presented. Each dot represents a single cell; the number of cells per section reflects relative cell density. Broken lines represent the limits of large ventral horn cells, as seen in these sections. Anterior is up in this and subsequent charts. P indicates the posterior cut-off of the lumbar enlargement. Distances in millimeters are with respect to the posterior cut-off.Also, the level of the section is given in terms of dorsal root entry zone, and, in parenthesis, in terms of vertebral level. absent or varied in their location, all suggested that the labeling of these cells was probably due to spread of the HRP, and that the cells did not represent true ML cells. Lateral longissimus (LL). Following extensive injections of H R P into lateral longissimus (Table [I), the majority of HRP-labeled cells formed a continuous medial string of cells reaching from the posterior half of the lumbar enlargement (L4-L5 dorsal root entry level), through and anterior to the enlargement, at least to dorsal root T11-T12 entry levels (Fig. 3). The labeled cells lay ipsilateral to the injected muscle, and tended to be most densely packed at the anterior end of the enlargement and in the several millimeters anterior to the enlargement. Following more restricted injections of LL, labeled cells extended as anteriorly as following the extensive LL injections, but were restricted posteriorly to the anterior half of the enlargement (approximately at the level of the 6 mm section in Fig. 3, at the L3 dorsal root entry level). In all LL animals, a second distinct string of HRP-labeled cells extended rostrally from the anterior end of the lumbar enlargement. In the enlargement, these cells appeared laterally, well-separated from the medial cells (Fig. 3, section labeled 7 ram); anterior to the enlargement these cells were located ventrolaterally. In animals having prior ablations of adjacent muscles (better visualization during injections) and

La-S~ Tla-iliac crest Ll-iliac crest approx. L4 La-L5 L3-L4

LL LL LL LL LL LL

Intact Ablated Ablated Intact Ablated Ablated

Surgical condition

70 80 120 32 37.5 40

Amount H R P (rag)

0.6 0.7 1.0 0.16 0.25 0.2

Injection volume (ml)

* I n all cases, labeled cells e x t e n d e d to anterior limit o f tissue sampled. - , n o t examined.

24 27 31 36 (lft) 35 37 (rt)

Injection level (vert. levels)

Expt.

Lateral Iongissimus injections

T A B L E I1

Total * cells

477 330 237 117 165 243

Cord examined (vert. levels)

T n - m i d L3 m i d Tx0-upper L2 T l o - u p p e r L2 T l l - u p p e r L1 T l l - u p p e r L1 T u - u p p e r Lz

0 0 0 0

12 0 0 2

64 134 97 33 72 138

16 11 21 8 4 28

Location labeled cells Enlargement Posterior "VM VL M L

178 161 92 70 79 53

207 24 27 6 10 22

Anterior VM VL

32

Fig. 3. Location of HR P-labeled cellsfollowing injection ofH RP into lateral longissimus (animal no. 37). Cells continue anterior to the sections presented. A indicates the anterior cut-off of the lumbar enlargement. Other levels and symbols as in Fig. 2. receiving large LL injections, or in animals receiving more restricted injections (with or without prior ablations of adjacent muscle), the numbers of lateral cells were more dramatically decreased than were the numbers of medial cells (Table II: nos. 27, 31 vs no. 24; nos. 36, 35, 37 vs no. 24). (Combining prior ablations with restricted injections did not further decrease lateral cell numbers: Table II, nos. 35 and 37 vs. no. 36). Transverso-spinalis muscles (TS). H R P injections into TS muscles were made at three levels (Table III): posterior TS, from L5 to posterior $1 or anterior $9. vertebral levels of the TS muscles; middle TS, from L5 to anterior-mid $1 TS; and anterior TS, from L3 to anterior L5 TS. Following injections into TS muscles, the large majority of cells lay in the spinal cord ipsilateral to the muscle injected. However, in some cases (TS nos. 17, 19, 21, 26, 32), a few HRP-labeled cells also appeared contralaterally in the lumbar enlargement (contralateral cells included in 'total cells', Table III). Most probably these spuriously appearing contralateral cells were the result of peripheral H R P spread.

23" 37 Oft)** 26* 32**

TS TS TS TS

ant La-ant L5 ant L3-ant L5 mid Ls-ant L5

Ablated Ablated Ablated

Intact Intact Ablated Ablated

Intact Intact Intact Intact Intact

Surgical condition

33.6 27 18

40 28 32.4 30

23 41.5 50 40 40

Amount HRP (rag)

* Anterior limit of lab¢,lext cells not determined. ** Posterior limit of labeled cells not determined.

TS 29* TS 33",** TS 36 (rt)*

post Ls-post Sl ant Ls-ant S2 ant Ls-ant $2 post Ls-post $1 mid Ls-post S1

17 18 19 20 21 **

TS TS TS TS TS

ant L~-ant $1 mid Ls-mid $1 post La-mid SI ant Ls-ant Sx

Injection level (vert. level)

Expt.

Transverso-spinalis muscle injections

TABLE III

0 0 --

68 72 62

16 20 36 12 47

Til-upper L~ Tll-upper L2 Tli-upper L1

95 205 132 201

120

0.28 0.18 0.09

1

1 13 0 0

L

99 266 122 50

-

TilL2 Tll-L2 Tit-upper L~ Tll-upper L9

~

0 0 --

21 89 14 9

2 9 17 13 32

10 6 20

70 124 86 37

36 42 68 52 58

0 0 0

0 9 0 0

35 49 72 55 50

Location labeled cells lpsilateral Enlargement Posterior M L VM VL

0.20 0.14 0.27 0.20

I

Total cells

T13-L1 Tlz-L1 TI~-L1 T1~mid L2

T

Cord examined (vert. levels)

0.23 0.32 0.35 0.20 0.20

Injection volume (ml)

0

0 2 0 0 0 10 2

7 29 12 0 58 56 40

m

D

m

5

m

w

Anterior "VM VL

34 Anterior. Following the anterior injections into TS muscles, labeled cells were not seen in the lumbar enlargement until just posterior to, or at, the anterior limit of the enlargement. Cells extended anteriorly at least to Ta1-T12 dorsal root entry levels, where the cells became sparse. The majority of labeled cells lay along the medial side of the ventral horn: medially in the anterior part of the enlargement, and ventromedially anterior to the enlargement (Fig. 4; Table III, animals nos. 29, 33, 36 rt.). In two animals, a few labeled cells were found ventrolaterally anterior to the enlargement (Fig. 4 section at 14 mm). The ventrolateral cells probably resulted from spread of the HRP (see Discussion). Middle. Following middle injections into TS, the majority of cells formed a continuous medial string of cells reaching from the posterior half of the lumbar enlargement (L5 dorsal root entry level) through to (at least) the T13 dorsal root entry level anterior to the enlargement (Table III; animals 23, 37 (left), 36, 32). In all of the animals, a number of additional labeled cells were found posterior to the enlargement (Table III), lying ventrolaterally in the ventral horn and separated from the main body of cells by a gap of 3-4 mm. 16mm TII-TI2 (ant II)

12ram TI3 (post TII)

IOmm

j

A 9mm L2 (post TI2)

7ram L3 (ant-mid TI3)

Fig. 4. Location of HRP-labeledcellsfollowinginjectionof HRP into the La-L~levelof the transversospinalis muscles(anterior TS) (animal no. 33).

35

~// ~

~

L4-L5 (AntLI)

O

•o

I/

\\

Omm postL5 <"°'"

L6 (LI-L2} Fig. 5. Location of HRP-labeled cells followinginjection of HRP into transverso-spinalis muscles at the L5 to posterior S1 or to anterior S~ level (posterior TS) (animal no. 20). Posterior. Following posterior injections into TS, HRP cells were found extending from posterior to the lumbar enlargement through to the anterior limit of the enlargement (Fig. 5). The labeled cells lying posterior to the enlargement were divided between ventromedial and ventrolateral locations (Table III, animals nos. 19, 20, 31), but generally favored a ventromedial position (Table III and Fig. 6, section at 0 mm). A small gap of 1-1.6 mm separated the continuous string of cells that lay (mostly) posterior to the enlargement from the continuous string of medial HRP cells (Fig. 5) in the enlargement. This gap occurred just rostral to the posterior cut-off of the enlargement, and was coincident with a lack of large cells occupying medial positions. In all animals some cells appeared laterally in the posterior half of the enlargement (Fig. 5). Cell size. Average soma diameters (ASD) were measured for 2 ML, 2 LL, and 2 TS preparations. For all preparations, values ranged from about 10 to 60 ffm. The averaged value of the ML samples (N -- 23, 53) was 36 #m. For LL and TS, cell size was determined for cells lying along the medial side of the ventral horn in the

36 enlargement a n d a n t e r i o r to the enlargement. The averaged A S D values were 33 # m for the L L samples (N = 122, 95), a n d 3 8 / z m for the TS samples (N 34, 29). F o r c o m p a r i s o n , A S D were calculated for a small sample o f large l u m b a r e n l a r g e m e n t lateral c o l u m n cells in tissue t a k e n f r o m different p r e p a r a t i o n s a n d subjected only to cresyl violet a n d n o t to H R P staining. T h e a v e r a g e d A S D for these biased samples (N 17, 19) was 41 # m (range: 25-60 #m).

M icrostimulation experiments Transverso-spinalis muscles. Muscles o f the t r a n s v e r s o - s p i n a l i s - e x t e n s o r c a u d a e medialis system were a c t i v a t e d by s t i m u l a t i o n in 24 p e n e t r a t i o n s , yielding 34 low t h r e s h o l d spots (Fig. 6). The range o f m i n i m a l s t i m u l a t i o n used was 2 - 4 4 / ~ A , with 28 spots having thresholds o f less t h a n or equal to 20/~A.

- 2 . 5 tO - 5 r a m S~- S2-S~,

(L2) SI; S I - S3 ECM

Fig. 6. Location of low-threshold microstimulation sites producing twitches in transverso-spinalisextensor caudae medialis muscles. Circles indicate the greatest likely spread of current from the low threshold sites marked by dye deposit; squares and rectangles indicate likely current spread compounded by track localization error as explained in text. Arrows mark low threshold sites for which thresholds were > 20/~A to 44 btA; thresholds for remaining sites are -< 20/~A. Levels are designated as for Figs. 2-6 (HRP localization). Additionally (last designation), the level of the muscle twitch (in terms of vertebral level) is indicated. For example, at the collapsed 4-6 mm section, stimulation produced twitches in the L6-$1 area of the transverso-spinalis muscles. (Extensor caudae medialis (ECM) is the caudal continuation of the lumbar transverso-spinalis muscles (TS).)

37 Spots producing twitches of TS-ECM muscles were located anteriorly-posteriorly from the La level through to the S2-Sa level (levels in terms of entering dorsal roots) of the spinal cord (Fig. 6). Roughly, stimulation (at strengths just above threshold) at more anterior spinal cord sites resulted in twitches at more anterior levels of the muscle system (Fig. 6). In the enlargement, TS-ECM stimulation sites localized to the gray matter were located medially within the ventral horn (sections at 2 mm, 6-7 mm of Fig. 6). Most stimulation sites posterior to the enlargement either contained large areas of white matter, or lay exclusively in the white. However, a few sites lay entirely in gray matter and may indicate TS-ECM motoneurons. Posterior to the lumbar enlargement, the majority of TS-ECM stimulation spots also produced twitching in ML, at the same or at different thresholds. Twitching of the leg was also seen at some sites; these sites lay in the lateral or ventrolateral white matter within the first millimeter posterior to the enlargement. In the enlargement, many stimulation sites, including all the sites lying in white matter, also produced leg twitches. One of the very discrete spots (circles) within the gray matter produced also leg twitches; the rest of these spots produced only TS twitches. Medial longissimus. Nine penetrations yielded 12 low threshold sites for activation of ML (Fig. 7). Minimal thresholds ranged from 2 to 36 #A, with 11 sites having thresholds less than or equal to 20/~A. (The 36/~A threshold spot is the largest square of the --2.5 to --5 mm level in Fig. 7.) Sites producing ML twitches lay exclusively posterior to the lumbar enlargement. No ML twitches were produced in the TS tracks in the enlargement. The majority of ML twitch sites (all but the 3 circular spots of Fig. 7) also produced TSECM twitches. No leg twitches were encountered in the ML penetrations.

/ 0 to -2.5mm L6 ( mid - post L I )

-2.5 to - 5 m m SI -$2-S 3 {L 2)

Fig. 7. Locationof low threshold sites producingtwitchesof medial longissimus.Symbolsand levels as for Fig. 6.

38 DISCUSSION Comment on methods For any of the H R P preparations, the majority of labeled cells formed a continuous body of cells (3 or less sections separating labeled cells). However, in many preparations some labeled cells also appeared variously outside the main body of the labeled cells. Attempts were made to eliminate these latter cells, thought to be due to peripheral spread of the HRP. Neither varying HRP amount and injection volume nor ablating adjacent muscles was completely effective in reducing numbers of labeled cells lying outside the main body of labeled cells without also reducing the numbers of labeled cells lying within the main body. Indeed, similar injections often resulted in quite different cell totals. Only in the case of extensive LL injections were prior ablations helpful (Table II). It should be remarked that the rats were mobile within hours after the H R P injections: it is possible that movement of the animals promoted leakage of H R P from the injected muscles. In the microstimulation experiments, interpretation of results requires estimation of the extent of stimulus spread (variable for each experimental situation) 7 and is further complicated in that the effects of the stimulus may be both direct and transsynaptic21, 39. In the present experiments, the endpoint used (visible twitch) did not permit distinction between direct versus transsynaptic activation of neurons: the low threshold areas described in these experiments may indicate location of motoneurons, or of regions invaded by the motoneuronal processes, or may represent location of cells or fibers synapsing onto the motoneurons. The data may be used to confirm HRP localization of motoneurons in so far as the low-threshold sites include regions in which H R P cells are found. Location o f HRP-labeled cells Following H R P injection into the proximal tail-tailbase muscle lz ML, labeled cells appeared ventrolaterally in the ventral horn, posterior to the lumbar enlargement. The anterior-posterior location of the cells, within dorsal root entry zones L6-$2, correlates well with description of ML innervation by branches of the dorsal rami of spinal nerves L6, $1 and $212. In contrast, the majority of labeled cells found after injection of either LL or TS (lumbar trunk muscles) 12 lay along the medial side of the ventral horn, largely in and anterior to the lumbar enlargement. Both LL and TS H R P cells tended to be more spread out mediolaterally anterior to the enlargement than in the enlargement. Since, within each of these two (LL, TS) types of injections, cell distributions were consistent whether or not adjacent muscle had been ablated, it is unlikely that the similarity of TS and LL cell distributions was due simply to peripheral H R P spread to LL in TS injections, or the reverse. The anterior-posterior extent of the HRP-labeled LL cells (in and anterior to the lumbar enlargement) is in line with the observation that LL is innervated along its length by branches of the dorsal rami of lumbar and lower thoracic spinal nerves 12. H R P injection into different levels of TS muscles led to different anterior-posteri-

39 or locations of labeled cells. HRP cells found after posterior TS injections extended caudal to the lumbar enlargement, and here also favored a medial location, again in contrast to the ventrolateral location of ML cells. The posterior TS injections included some of extensor caudae medialis (a proximal tail muscle of the TS systemlZ): it is likely that the labeled TS cells caudal to the lumbar enlargement are ECM cells. These labeled cells caudal to the enlargement were separated by a small gap from the labeled cells lying medially in the enlargement: such a gap in the medial cell column at this level has been described in rhesus monkey 88 and in man 14. The anterior-posterior extent of the HRP-labeled TS cells agree with the observation that the dorsal ramus branch supplying TS muscles typically sends branches into that muscle system at three vertebral levelsl2: the TS system at any one level may receive nerve branches from a number of dorsal rami. Other HRP-labeled cells appeared primarily in three locations: (1) lateral in the posterior part of the lumbar enlargement (after some TS or ML injections), (2) lateral-ventrolateral in the anterior part of the enlargement and anteriorly (after LL and some TS injections), or (3) ventrolaterally posterior to the enlargement (following some LL or TS injections). The inconsistent occurrence of these other labeled cells, and the inconsistency of their number relative to the number of labeled cells constituting the main body, suggested that these other labeled cells did not represent motoneurons innervating the injected muscle, but probably resulted from peripheral spread of the HRP. The first of these locations resembles the location of gluteal motoneurons in the catZ5: since rat gluteal muscles lie adjacent to the epaxial muscles at the level of the ilium 2°, it is possible that HRP may have spread to the gluteal muscles. The third location is similar to that of ML labeled cells: these labeled cells may reflect leakage to ML or to its caudal continuation, extensor caudae lateralis. Possible identity of the second group of other labeled cells, the anterior lateral-ventrolateral group, is unclear. The mean average soma diameters (ASD) of rat ML, TS, or LL HRP-labeled cells ranged from 33 to 38/,m (uncorrected for tissue shrinkage). For large lateral ceils of the rat lumbar enlargement, a mean ASD of 41/,m was obtained. These estimates of average cell size fall within range of previous estimates made on large ventral horn cells of rat lumbar cord: 31/zm (uncorrected, but shrinkage estimated at 17-22 ~o)25, 30/~m (L4 dorsal root level) 1°, roughly 35/*m (L4, corrected for 20 ~o shrinkage) 19, roughly 3 0 / , m (Ls-Ls, corrected for 20 ~o shrinkage) 19, and 4 5 / , m (uncorrected) 4°. The populations of HRP-labeled cells found following injection of ML or of TS or of LL muscles are apparently not distinguishable from each other on the basis of cell size. Also, the HRP-labeled cells to these epaxial muscles are not clearly distinct in size from rat large lumbar ventral horn cells (mainly motoneurons 25,an) in general. Although some small-sized (10-20 #m) labeled cells were seen, clear secondary modes were not apparent when cell size histograms were cast. This is probably simply due to sample size: it is most likely that samples were too small to permit clear differentiation of such secondary populations of smaller-sized (possibly gamma 13) cells.

40 Location o f low threshold microstimulation twitch sites T S - E C M or M L twitch sites lying within the gray m a t t e r o f the ventral h o r n o v e r l a p p e d the l o c a t i o n o f labeled cells f o u n d following h o r s e r a d i s h p e r o x i d a s e injection o f the different muscles. Particularly, T S - E C M twitch sites extended f r o m p o s t e r i o r to the l u m b a r e n l a r g e m e n t t h r o u g h the e n l a r g e m e n t ; there is a p r e d i c t a b l e

a n t e r i o r - p o s t e r i o r c o r r e l a t i o n between spinal c o r d level o f H R P - l a b e l e d cells or o f t w i t c h - p r o d u c i n g sites a n d level o f muscle injected or twitching. M L twitch sites, like M L H R P cells, were located p o s t e r i o r to the enlargement, up to j u s t behind the p o s t e r i o r limit o f the enlargement. W i t h respect to m e d i o l a t e r a l localization, T S - E C M twitch sites in the e n l a r g e m e n t were located medially, like the m a j o r i t y o f H R P cells. P o s t e r i o r to the enlargement, m e d i o l a t e r a l localization could not be d e t e r m i n e d f r o m the twitch site locations. Twitch sites in white m a t t e r m o s t likely r e p r e s e n t T S - E C M or M L m o t o n e u r o n axons passing t h r o u g h to the ventral roots. It is also possible that sites in the white m a t t e r reflect a c t i v a t i o n o f descending tracts (for e x a m p l e , lateral vestibulospinal or reticulospinal tracts 3~) t h a t possibly synapse o n t o T S - E C M or M L m o t o n e u r o n s located m o r e posteriorly. ACKNOWLEDGEMENTS The a u t h o r s wish to a c k n o w l e d g e the technical assistance o f Arlene Ballin, and the assistance o f G a b r i e l e Z u m m e r in p r e p a r a t i o n o f the manuscript. P h o t o m i c r o g r a p h s were t a k e n by Lee G r e e n b e r g e r .

REFERENCES 1 Abzug, C., Maeda, M., Peterson, B. W. and Wilson, V. J., Cervical branching of lumbar vestibulospinal axons, J. Physiol. (Lond.), 243 (1974) 499-522. 2 Akaike, T., Fanardjian, V. V., Ito, M., Kumada, M. and Nakajima, H., Electrophysiologicalanalysis of the vestibulospinal reflex pathway of rabbit. 1. Classification of tract cells, Exp. Brain Res., 17 (1973) 477-496. 3 Andersen, P., Hagan, P. J., Phillips, C. G., Powell, F. R. S. and Powell, T. P. S., Mapping by microstimulation of overlapping projections from area 4 to motor units of the baboon's hand, Proc. roy. Soc. B, 188 (1975) 31-60. 4 Angulo y Gonz~tlez, A. W., The motor nuclei in the cervical cord of the albino rat at birth, J. comp. Neurol., 43 (1927) 115-142. 5 Asanuma, H. and Sakata, H., Functional organization of a cortical efferent system examined with focal depth stimulation in cats, J. Neurophysiol., 30 (1967) 35-54. 6 Asanuma, H., Stoney, S. D., Jr. and Abzug, C., Relationship between afferent input and motor output in cat motorsensory cortex, J. Neurophysiol., 31 (1968) 670-681. 7 Bagshaw, E. V. and Evans, M. H., Measurement of current spread from microelectrodes when stimulating within the nervous system, Exp. Brain Res., 25 (1976) 391-400. 8 Baker, J. R., Cytological Technique, 3rd edn., John Wiley, New York, 1950, 31 pp. 9 Bean, C. P., (Appendix) A theory of microstimulation of myelinated fibers, J. Physiol. (Lond.), 243 (1974) 514-520. 10 Bradley, K. and Somjen, G. G., Accomodation in motoneurones of the rat and the cat, J. Physiol. (Lond.), 156 (1961) 75-92. 11 Brink, E. E., Modianos, D. T. and Pfaff, D. W., Ablations of lumbar epaxial musculature: Effects on lordosis behavior of female rats, Brain, Behav. Evol., in press. 12 Brink, E. E. and Pfaff, D. W., Vertebral muscles of the back and tail of the albino rat (Rattus norvegicus albinus) , Brain, Behav. Evol., in press. 13 Burke, R. E., Strick, P. L., Kanda, K., Kim, C. C. and Walmsley, B., Anatomy of medial gastrocnemius and soleus motor nuclei in cat spinal cord, J. Neurophysiol., 40 (1977) 667-680.

41 14 Elliott, H. C., Studies on the motor cells of the spinal cord. I. Distribution in the normal human cord, Amer. J. Anat., 70 (1942) 95-117. 15 Eiliott, H. C., Studies on the motor cells of the spinal cord. IV. Distribution in experimental animals, J. comp. Neurol., 81 0944) 97-103. 16 Geldred, J. B. and Chopin, S. F., The vertebral level of origin of spinal nerves in the rat, Anat. Rec., 188 (1977) 45-48. 17 Goering, J. H., An experimental analysis of the motor-cell columns in the cervical enlargement of the spinal cord in the albino rat, J. comp. Neural., 46 (1928) 125-151. 18 Graham, R. C. and Karnovsky, M. J., The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique, J. Histoehem. Cytochem., 14 (1966) 291-302. 19 Granit, R., Kernell, D. and Smith, R. S., Delayed depolarization and the repetitive response to intracellular stimulation of mammalian motoneurones, J. Physiol. (Lond.), 168 (1963) 890-910. 20 Greene, E. C., Anatomy of the Rat, Hafner, New York, 1935 (reprinted 1959). 21 Gustafsson, B. and Jankowska, E., Direct and indirect activation of nerve cells by electrical pulses applied extracellularly, J. Physiol. (Lond.), 258 (1976) 33-61. 22 Hansson, H. A., Uptake and intracellular bidirectional transport of horseradish peroxidase in retinal ganglion cells, Exp. Eye Res., 16 (1973) 377-388. 23 Hedreen, J. C. and McGrath, S., Observations on labeling of neuronal cell bodies, axons, and terminals after injection of horseradish peroxidase into rat brain, J. comp. NeuroL, 176 (1977) 225-246. 24 Jankowska, E. and Roberts, W. J., An electrophysiological demonstration of the axonal projections of single spinal interneurones in the cat, J. Physiol. (Lond.), 222 (1972) 597-622. 25 Janzen, R. W. C., Speckmann, E.-J. and Caspers, H., Distribution of large ventral horn cells in the lumbar cord of the rat, Cell Tiss. Res., 151 (1974) 159-170. 26 Kaizawa, J. and Takahashi, I., Motor cell columns in rat lumbar spinal cord, Tohoku J. exp. Med., 101 (1970) 25-34. 27 Kow, L.-M. and Pfaff, D. W., Dorsal root recording relevant for mating reflexes in female rats: Identification of receptive fields and effects of peripheral denervation, J. NeurobioL, 6 (1975) 23-37. 28 Kristensson, K. and Olsson, Y., Retrograde axonal transport of protein, Brain Research, 29 (1971) 363-365. 29 Kristensson, K., Olsson, Y. and Sjostrand, J., Axonal uptake and retrograde transport ofexogenous proteins in the hypoglossal nerve, Brain Research, 32 (1971) 399-406. 30 LaVail, J. H. and LaVail, M. M., The retrograde intraaxonal transport of horseradish peroxidase in the chick visual system: a light and electron microscope study, J. comp. NeuroL, 157 (1974) 303-358. 31 Nauta, H. J. W., Pritz, M. B. and Lasek, R. J., Afferents to the rat caudoputamen studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method, Brain Research, 67 (1974) 219-238. 32 Peterson, B. W., Maunz, R. A., Pitts, N. G. and Mackel, R. G., Patterns of projection and branching of reticulospinal neurons, Exp. Brain Res., 23 (1975) 333-351. 33 Petras, J. M., Cortical, tectal, and tegmental connections in the spinal cord of the cat, Brain Research, 6 (1967) 275-324. 34 Pfaff, D. W. and Lewis, C., Film analyses of lordosis in female rats, Horm. Behav., 5 (1974) 317-335. 35 Romanes, G. J., The motor cell columns of the lumbo-sacral spinal cord of the cat, J. comp. Neural., 94 (1951) 313-358. 36 Romanes, G. J., The motor pools of the spinal cord. In J. C. Eccles and J. P. Schad~ (Eds.), Organization of the Spinal Cord, Progr. Brain Res., Vol. 11, Elsevier, Amsterdam, 1964, pp. 93-119. 37 Shinoda, Y., Arnold, A. P. and Asanuma, H., Spinal branching of corticospinal axons in the cat, Exp. Brain Res., 26 (1976) 215-234. 38 Sprague, J. M . , A s t u d y o f motor cell localization in the spinal cord of the rhesus monkey, Amer.J. Anat., 82 (1948) 1-26. 39 Stoney, S. D., Jr., Thompson, W. D. and Asanuma, H., Excitation of pyramidal tract cells by intracortical microstimulation: Effective extent of stimulating currents, J. Neurophysiol., 31 (1968) 659-669. 40 Tamaki, K., On the sectional areas and volume of the largest motor cells and their nuclei in the lumbar cord of the albino rat according to sex, Anat. Rec., 56 (1933) 229-240. 41 Thomas, R. C. and Wilson, V. J., Precise localization of Renshaw cells with a new marking technique, Nature (Lond.), 206 (1965) 211-213. 42 Wong-Riley, G. T. T., Endogenous peroxidatic activity in brain stem neurons as demonstrated by their staining with diaminobenzidene in normal squirrel monkeys, Brain Research, 108 (1976) 257-277.