- Email: [email protected]
Retrograde cell changes in medial septum and diagonal band following fimbria-fornix transection: Quantitative temporal analysis
~twrmmnce Vol. 19. No. 1. pp. 241-255, 1986 Prn~ted in Great Britain
RETROGRADE CELL CHANGES TN MEDIAL SEPTUM AND DIAGONAL BAND FOLLOWING FIMBRIA-FORNIX TRANSECTION: QUANTITATIVE TEMPORAL ANALYSIS F. H. GAGE,* Departments
WICTORIN,$ W. FISCHER,: L. R. WILLIAMS,* S. VARON~and A. BIORKLUND:
K.
of *Neurosciences ZDepartment
and fBiology, IJniversity of California, San Diego, CA 92093, U.S.A. and of Histology, University of Lund, Lund. Sweden
La Jolla.
Abstract-Complete unilateral fimbria-fornix transections. including the overlying cingulate cortex, were administered to female rats. At time points from 1 day to 6 weeks. the sepal-diagonal band region was examined using acetylcholinesterase histochemistry, Cresyl Violet cell staining. and choline acetyitrnnsferase biochemistry. As early as 1 day following the transection a decrease in acetylcholinesterase positive cell body staining was observed in the medial septum; however. no loss of Nissl-stained neurons was measured in Cresyl Violet stained sections until 1 week after the lesion. Maximal loss of acetylcholinesterase”positive cells. as visualized after irreversible acetylcholinesterase inhibition, was measured at I week. and no further change was observed at time points up to 6 weeks after operation. The loss of acetyltransferase-positive cells was greatest in the medial septal area (- 65%) and the vertical limb of the diagonal band (- 55%). Little cell loss was measured in the horizontal limb of the diagonal band. This is consistent with the known projections of these cell bodies. Remaining ac~tyIcholinesterase-positive cell bodies in the medial septum had shrunk by about 20% (measured as the diameter along the major axis). A marked neuronal celf toss (about 50%) was demonstrable in the medial septum and vertical limb of the diagonal band in the C‘resyi Violet-stained sections, too. A pile-up of acetylcholinesterase-stained material was observed in the dorsal-lateral quadrant of the septai area just proximal to the lesion at 1 day following transection. This pile-up occurred in the medial septum and diagonal band area up to 1 week followmg the transection. and had nearly disappeared by 2 weeks post-transection. Choline acetyltransferase biochemical activity. measured in samples of whole septum, decreased significantly at I day but subsequently returned to control levels. By 2 weeks following transection, an increase in acetylcholinesterase-positive stained fibers was observed in the dorsal-lateral quadrant of the septum, ipsilateral to the lesion relative to the contralateral septum. This response, which was interpreted as sprouting from the lesioned axons proximal to the transection. probably accounted for the rise in choline acetyltransferase biochemical activity in the whole septum following the reduction on
the first day.
Neurons of the medial septum and the diagonal band of Broca send fibers via the fimbria-fornix (FF) bundle and the supracallosal striae to terminate in the hippocampal formation. Using standard histological techniques, Daitz and Powell’ and McLardy’* demonstrated that FF transection induces a “severe, rapid and consistent degeneration in the medial septum and diagonal band nuclei” which results in “complete atrophy or shrinkage of all the cells of the ipsilateral medial septal nucleus and partial degeneration of the nucleus of the diagonal band.” A large portion of the cells projecting to the hippocampus from the medial septai and diagonal ~
-__--___
Address correspondence to: F. H. Gage, Department Neurosciences (M-024), University of California,
--
band nuclei is cholinergic.2,3,‘7,23Z” The anterograde decrease in the two cholinergic marker enzymes, choline acetyltransferase (ChAT) and acetylcholinesterase (AChE), in the hippocampus following FF transection is well documented both biochemically and histochemically (e.g. Refs 8, 14 and 24). Also well documented is the AChE-stained “pile-up” in septohippocampal axons proximal to the transe~tion,14 which represents the directionality of the fibers. However, a quantitative assessment of the degenerative changes-cell loss and/or atrophy-of the septal cholinergic neurons is so far lacking. The objective of the present study is to document the time course and extent of cell loss and atrophy in the cholinergic neurons in the septum and diagonal band areas following complete transection of all the known cholinergic fibers leading to the hippocampal formation via the FF and the supracallosal striae. This quantitative morphological time course, together with the documentation of concomitant biochemical changes and changes in cholinergic fiber patterns in the septum. should provide an accurate portrait of
of San
Diego, La Jolla, CA 92093, U.S.A. .4hhreviations:
AChE, acetylcholinesterase; ChAT, choline acetyltransferase; DFP, diisopropyl fluorophosphate; FF, ~rnb~a-fornix; HDB, horizontal limb of the diagonal band of Broca; MS, medial septum; VDB, vertical limb of the diagonal band of Broca. 241
242
I-‘.H. GAOE et al.
the dynamic changes occurring in these neurons foIfowing damage. This info~ation will, moreover, be of particular value as a basis for the assessment of effects of exogenous and endogenous trophic factors or the process of cell death after axotomy in the forebrain cholinergic system. This model of traumainduced cell death and its consequences may also give insights into the mechanisms underlying certain clinical cell death syndromes, such as Alzheimer’s disease. EXPERIMENTAL
PROCEDURES
Animals Adult female Sprague-Dawley rats (ALAB., Stockholm, Sweden) weighing between 200 and 225 g at the beginning of the experiment were used. Surgical procedures The animals were anesthetized with intraperitoneal injections of a ke~mine-xyl~ne mixture (10 mg/kg Keta!ar, Parke-Davis, and 5 mg/kg Rompun, Hoechst). All animals in each of the experimental groups received the same complete unilater~ aspirative lesion of the fimb~a-fo~ix and supracallosal striae (the FF lesion), as described previously* (cf. Figs la and 2A). The lesions were conducted visually with the aid of a surgical microscope. Htktochembtry FF-lesioned animals were taken 1 day and 1, 2,4 and 6 weeks following surgery for histochemical evaluation. Six rats were used at each survival time, except at 4 weeks when the group comprised 5 rats. Four normal unoperated rats were included for comparison. To obtain consistent and complete AChE staining of the septaldiagonal band neurons, the animals were injected intr~~~arly witti the irreversible AChE inhibitor, diisopropyl fluorophosphate (DFP, Sigma) 4 h before being kiiied at a dose of 2 mg/kg (dissolved in peanut oil) according to the method of Butcher.” The rats were perfused via the ascending aorta with 150 ml of ice-cold 4% buffered paraformafdehyde. The brains were postfIxed for 2 h in the same fixative and then left overnight in 10% sucrose at 4°C. Sections were cut at 15pm in a Dittes cryostat and every third section was stained for AChE according to the Koelle thiocholine method,‘*J1 using 10m4M promethazine as an inhibitor of nonspecific esterases and silver nitrate intensification of the reaction product.9 Alternate sections were taken for Cresyl Violet staining. Quantitative morphology AChE-positive and Cresyl Violet-stained cells were counted in three regions defined in the following ways (Fig. 1): Medial septum (MS) = rostrally from the level of the genu of the corpus ca!losum; cauda!ly, to the level of the crossing of the anterior commissure; and latera!!y, 0.75mm from the midline. Vertical limb of the diagonal band of Broca (VDB) = same rostra! and caudaf boundaries as above, with the dorsal boundary defined as the anterior commissure (dashed line in Fig. lc). This landmark WBS chosen to ensure r&ability of a!! counts, rather than for rigid anatomical reasons. The ventral boundary was defined by the medial border of the o!factory tubercle. Horizontal limb of the diagonal band of Broca @IDB) = same rostra! and caudal boundaries, except that the lateral boundary extended out to the lateral tip of the anterior c~mmissutt (Fig. Id). An AChE-positive cell was defined as a dark brown body with a minimum diameter of 12 pm along the major axis. Figure 2 provides representative examp!es of structures defined as cell bodies (arrows). RofG without arrows were not counted, either because staining was too weak or because they were too sma!i (< 12 pm along the major axis). The vast majority of AChE cells fell within the range of
!8-29pm (cf. Fig. 9). AChE-stained cells smaller tfidn 12pm in diameter were without regular cell shape or processes and could not be distinguished by cellular criteria, Cresyl Violet-stained neurons were defined as cells having diameters > I2 pm and containing both a nucleus and Nissl substance. Every 6th of the mounted sections (i.e. every 18th section) was counted. A counting grid (0.75 x 0.75 mm) was used to entirely cover the areas of interest. Figures l&d show the placement of the counting grid. For each animal, both the regions ipsilatera! and contralaterai to the lesion were counted at x f6 objective magnification and the results are presented as a percentage of remaining AChE-positive cells (ipsilateral region/contralateral region x 100). Cell size was estimated with the same counting grid at x40 objective magnification by two independent observers measuring the length along the major axis of 60 randomly chosen AChEpositive cells of each region on both the ipsilaterat and contralateral sides. Size differences between the two sides and differences from normal unoperated rats were again expressed as percentages. All AChE-positive cell number data were corrected for cell shrinkage according to the Abercrombie formula.’ Choline acetyltransferase assay Separate sets of 6 rats with FF lesions were taken at 1day, 1, 2, 4 and 6 weeks following surgery for bi~hemica1 analysis of ChAT activity in the whole septum (media! plus lateral septa! nuclei) according to the micromethod of Fonnum.’ protein content was measured according to Lowry et ~1.‘~Values are expressed as percentages of an unoperated control group and the contralateral side in the operated animals. Statistical analysis To determine whether statistical differences existed between groups on measures of cell size and cell number, a one-factor analysis of variance (ANOVA) was used with post hoc Neuman-Keuls test, for comparisons between group. In the text, the P values presented are derived from the Neuman-Keuls test, but only when the ANOVA test gave a sigt&cant level of P < 0.05. For comparisons of ChAT activity between groups, total septal ChAT activity in an unlesioned control group of animals was compared with the septal ChAT activity from the contralateral septum of each of the lesioned groups using one-way ANOVA. Since no significant differences between groups were observed, these values were combined into a single control group with which the septal CUT activity ipsilateral to the lesion was compared, again using an ANOVA with post hoc Neuman-Keuls. For clarity of presentation, these values were converted to percentages of the combined control group. Statistical analyses, however, refer to the calculations conducted on the actual ChAT va!ues. When making within group comparisons (e.g. comparing actual cell sizes between ipsilateral and contralateral septum), a paired-comparison test for related groups was used, as indicated in the text. RESULTS
in Figs la and ZA) was similar to that used previous1y.s This lesion is known to reduce ChAT activity in the hippocampus by at least 90% and to eliminate essentially all AChEpositive fiber staining in the dorsal two-thirds of the hippocampus. The
FF lesion (illustrated
Acetyichoiinesterase-posirive
cell stair&g
One day following the FF lesion (Figs 3 and 4) a
20% reduction in the number of cells in the MS and VDB was observed on the side ipsilateral to the lesion
Retrograde
cell changes (b)
(d)
Fig. I. Coronal sections from a rat with a fimbria-fornix lesion, indicating approximately where the cells were counted and what constituted the regional boundaries for the purpose of counting. (a) Lesion cavity at the level of the fimbria-fornix; (b) rostra1 medial septal area (MS). first level of cell counts; (c) medial septal area and vertical limb of the diagonal band of Broca (VDB). midlevel of cell counts; (d) medial septal area and horizontal limb of the diagonal band of Broca (HDB). representing the most caudal level of cell counts. Squares indicate actual size of the grid that was used to limit the areas of counting.
related t-test). By 1 week follow(Figs 3 and 5) the cell numbers were reduced by 65% in the MS and by 55% in the VDB (P <’ 0.001). This relative decrease in AChE-positive cells did not change over the subsequent 5 weeks, the longest time period studied (Figs 3 and 6). In the HDB there was no significant cell loss at 1 day, but by 1 week a 20% decrease in AChE-positive cells ipsilateral to the lesion was seen (P < 0.05). This percentage decrease did not change over the subsequent 5 weeks (Fig. 3).
(P < 0.05; ing
the
Student’s
lesion
<‘re.s~Y Violet cell staining To clarify whether the disappearance of AChEstained cells reflected an actual loss of cell bodies from the area, neuronal profiles (defined as Nisslstained cell bodies with a diameter along the major axis of < I2 pm) were counted also in the adjacent Cresyl Violet-stained sections. As illustrated in Table I there was no significant reduction in the total number of neurons in the MS, the VDB or the HDB relative to the side contralateral to the lesion at I day, but at 1 week following the lesion there was a significant loss of neurons in both the MS (40%) and the VDB (20%) (cf. Fig. 5). Two additional animals
were examined at 6 weeks following FF transection. and they both showed an approximate 50% dccreasc in total cell counts as measured with Cresyl Violet staining in both the MS and the VDB (Fig. 6). Cell shrinkage Shrinkage of AChE-positive cells was already apparent 1 day after the FF lesion in the MS and VDB. It amounted to about IO- 15% relative to the contralateral side (Fig. 7) and about 15.20% relative to the normal controls (Fig. 8). The side difference in cell size was no longer apparent at I week, but it reappeared again at 4 and 6 weeks following the lesion (Fig. 7). From Fig. 8, which illustrates the actual cell sizes, it is apparent that the AChE-positive cell bodies in the MS were also significantly reduced in size on the contralateral side at all time points in the lesioned rats. This effect, which was most pronounced at I week, apparently accounted for the transient disappearance of the side difference in cell size at I week post-lesion, as seen in Fig. 7. The AChE-positive cells in the HDB showed significant decrease in size only at I week. and then on both sides of the brain (Fig. 8); this effect did not persist at longer time points. From Fig. 9. which shows the cell
F. H. GAGE et ul.
244
t----r(r
E ----a
z;90-
rY
."....,, ,..... ""._,.. ,___,_,,..,,.'-"'f
5 g
70-
$
60-
0
MS VDB
%o%
_______--
------____
*
g 402 A
T
30-
T
8 20w 2 io-
J,
I
Id
twk
I
I
I
2wks 4wks TIME AFTER UNfFF-LESION
I
I
6wks
Fig. 3. The number of AChE-positive cells in the three regions counted (MS, VDB and HDB) expressed as a percentage of the control side, contralateral to the lesion. Cell counting was conducted in separate groups of animals (symbols indicate means and SEM for n = 6) at each of the time points. Significant decreases (at least P < 0.01, Nmman-Keuls test) in e&l mtmber as compared to the control side are seen at ah time points for MS and VDB. Open symbols indicate unlesioned control animals.
Table 1. Number of ceil bodies after frmbria-fornix lesion in the medial septal nucleus and the vertical limb of the diagonal band of Broca as counted in Cresyi Violet-stained sections Postop. time
n
Cell num$Sr (% of contr. side;
1 day 7 days 6 weeks
4 5 2
101.4 f 4.7 62.4 f 4.2+ 48.3
SEM)
99.2 f 7.7 78.3 f 4.7’ 54.0
*P < 0.01 compared to l-day group (Student’s t-test). Expressed as percent of the control side. The counts were not corrected for cell shrinkage.
size ~s~bution
in a frequency ~sto~am, it is clear that the overall reduction in cell size was due to an increased number of cells on the lesioned side that fell into the smaller range of cell sizes (12-17gm diameter along the major axis).
Septal choline ncetyltransferase activity In whole septum (medial and lateral areas combined, including the diagonal band nuclei) a 25X decrease in ChAT activity was apparent at 1 day after FF transection. This decrease in ChAT activity returned to within the levels of the control side by 1 week and remained at this level at later time points (Fig. 10).
At 1 day post-transection there was a dramatic accumulation of AChE-positive staining in the dorsal lateral septal area and in the cingulate cortex just rostra1 to the lesion (Fig. 1la&. The ovulation appeared to be in the axons which had been transected by the lesion. By 1 week the pile-up was located more proximal to the cell bodies of the septal-diagonal band region (Figs 11b,d). At this time -..
Fig. 4. Photomicrographs from the mediaI septum (a and c) and the vertical limb of the diagonal band (b and d) of AChE-stained sections. (a) and (b) unfesioned control animals; (c) and (d) 1 day follo~ng FF lesion. Fig. 5. AChE (a and b) and Cresyl Violet (c and d) stained sections from the medial septum and the vertical limb of the diagonal band at 1 week after FF lesion. Fig. 6. AChE (a and b) and Cresyl Violet (c and d) stained sections from the medial septum and the vertical limb of the diagonal band at 6 weeks after FF lesion.
_-
. Fig. 2. (A) AChE-stained cells in the VDB demonstrating the range of cells observed. Arrows indicate profiles counted as AChE-positive neurons. Profiles without arrows were not included, either because they were too small (< 12 pm along major axis), or because they were too weakly stained. (B) Photomicrograph of extent of fimbria-fornix transection. cc, corpus callosum: Iv, lateral ventricle; FL fimbria: f, fornix; sm. stria medullaris: ot. optic tract: ox, optic chiasm. 23s
b
/_ Fig.
s
c
,d Fig. I I. Photomicrograph of “pile-up” in AC’hE-positive stained 4 hr after DFF injections at 1day (a and c) and I week (b and d) denote the position of the photagraphs above. cc. corpus callosum, commlssure; ox. optic cfuasm; ms. medial septum: ~a,
249
fibers in the dorsal lateral septal area following FF lesion. The framed area’; Iv, lateral ventricle: ac, medial anterior IaWral anterior commissure.
Fig. 12. Photomicrograph non-DFP-treated animal
of AChE-positive stained fibers in the dorsal lateral septal area in ‘i at 4 weeks following FF lesion. (A) Low-power magnification, and (B) higher power magnification of the same section.
250
Retrograde
cell changes
MS
c---q . ..
I
I
Id
I
I
lwk
I
I
2wks
.. l
1 Gwks
I
Lwks
TIME AFTER
VDB H/-JB
UNIFF-LESION
Fig. 7. Cell sizes of AChE-positive cell bodies as indexed by cell diameter along the major axes of 60 randomly measured cells in each of the three areas in a total of four animals at each time point. Values are expressed as a percentage of the control side contralateral to the lesion. Si~i~cant decreases in cell sizes were observed at 1 day and 4 and 6 weeks for MS and VDB, at the P <: 0.05 level; Neuman -Kculs lest.
1 WEEK 40 1
AFTER
6 WEEKS
UNIFF-LESION [z1
l
nonieslon lesion
side side
40
AFTER
UNIFF-LESION 0
nonteslon
II
lesron
side side
@L
#
30
zz iz ; 0
l-l 20
10
12-17 18-23 24-29 CELL SIZE, MAJOR AXIS
(vrn)
il,
12-17 18-23 24-29 30-35 CELL SIZE, MAJOR AXIS (pm)
Fig. 8. Size of AChE-positive neurons in the medial septal nucleus as indexed by the diameter along the major axis and expressed in micrometers bm). Values are presented for non-lesioned control animals as well as for the Iesioned and control sides for each of the three brain regions measured at each of the four time points following the FF lesion. Filled stars denote significant differences from normal uniesioned controls (P < 0.05; Neuman-Keuls test). Open stars denote difference between sides (P < 0.05; Student’s related r-test).
F. H. GAGEet al.
HDB
VDB
MS l
lwk
Id
con
Iwk
6wk
con
Id
lwk
TIME AFTER
Lwk
6wk
con
Id
lwk
Lwk
6wk
UNIFF-LESION
Fig. 9. Frequency histogram showing the distribution of cell sizes of AChE-positive neurons in the medial septaf nucleus at 1 week (left) and 6 weeks (right) after FF lesion. For each group a total of 60 cells on each side was measured in 6 rats (i.e. 10 cells per side per rat). l , nonlesion side; a, lesion side.
point it was easy to trace some of the distended AChE-positive fibers from their most distal point to the now atrophied AChE-positive cell bodies. By 2 weeks following the lesion, there were very few
I
1
cm. Id
1
lwk
I
distended fibers with the characteristic “piled-up” appearance. Those that were apparent were in the very medial portion of the medial septal area, or in the vertical limb of the diagonal band.
1
1
4wks 2wks TIME AFTER UNIFF-LESION
I
1
6wks
Fig. 10. Choline acetyltransfwasc activity in whole septum (the cm&&d septal-diagwai band regim) tpthksion,atthesametinM@ntsasin explwscdasa~~oftbe~~,~ Fig. 3. ThecontmIgfo~~xoav e&tr&asw8llasthepoolu¶valueafromthe diihnce (P < 0.05) between controls and contralateral sides of the fke ksion& ipsilaterai septum was observed only at day 1 post-lesion.
Retrograde
cell changes
Fig. 13. Schematic representation of the proposed changes in the three anatomically different types ot cholinergic neurons in the septal-diagonal band region at various time points following FF lesion. For further explanation, see text.
Sprouting
in the septum
The puzzling recovery of biochemically measured ChAT activity in the septum following the initial reduction seen at 1 day was examined further by looking at the pattern of AChE staining in the septum in the absence of DFP pretreatment. Already by 2 weeks following the FF lesion, there was a region of clearly enhanced AChE-positive staining in the dorsal lateral septum ipsilateral to the lesion. This enhanced neuropil staining, which is interpreted as a sprouting response, was clearly different from the “pile-up” that was evident at I day and 1 week post-lesion. While the “pile-up” was characterized by thick, dense accumulations of AChE in the distal portions of the lesioned axon, just proximal to the lesion at I day, and closer to the cell body by 1 week, the sprouting at 4 weeks appeared as fine fiber arborizations similar in density and appearance to the fibers in the intact septum contralateral to the lesion (Fig 12).
DISCUSSION
The present results portray a sequence of degenerative and regenerative events after axotomy in the septo-hippocampal projection system. In Fig. 13 we propose an interpretation of these events based on the assumption that the FF lesion will affect the AChE-positive neurons in the septal-diagonal band region differently depending on the anatomical arrangement of their axonal projections. Cell death The present results confirm the early findings of Daitz and Powell’ and McLardy,‘* demonstrating that there is a rapid and severe loss of cells in the medial septal area and diagonal band of Broca, as indexed by Cresyl Violet-stained cell counts. We further demonstrate that the AChE-positive cells in this region are lost in dramatic proportions. The loss of AChE-positive cells reaches a maximum within 1 week post-transection and no further cell loss is observed over the 6 weeks investigated. That the
disappearance of AChE-stained neurons (in DFPpretreated rats) was due to an actual loss of neurons is supported by the parallel counts in the Crcsyl Violet-stained sections. Interestingly. the loss of AChE-positive cell bodies amounted to about 70% in the MS area, whereas the corresponding figure in the Cresyl Violet-stained sections was about 50%. From the total number of neurons counted in the adjacent sections with the two types of stains, our impression is that about one-third of the medial septal neurons (defined as cell bodies with a diameter > 12pm along the major axis) are AChEpositive in the DFP-treated rats (see also Refs 2, 17 and 23). This means that only about half of the neurons lost in the Cresyl Violetstained sections could be accounted for by the disappearance of cholinergic septal-hippocampal neurons, and thus that noncholinergic neurons arc likely to degenerate as well. In fact. it is known that a significant proportion of the septal-diagonal band neurons projecting to the hippocampus is noncholinergic,* and there is evidence from glutamate decarboxylase immunohistochemistry in combination with retrograde tracing that part of the septohippocampal pathway is GABAergic.“,” It is likely. therefore, that the retrograde cell loss in the septaldiagonal band area after FF lesion is not confined to the cholinergic projection system. It is possible that the loss in cells reported in this paper could be a result of the shrinkage of damaged cells below the 12 bcrn criterion, but we find this unlikely in that there did not appear to be an increase in the number of small AChE positive elements < 12 pm on the lesioned side. It is also possible that the AChE cells were not dying, but rather becoming AChE-negative. We also find this unlikely because a major cell loss (about 50%) was also evident in the Cresyl Violet-stained sections. In Fig. 13 we propose that the observed cc.11loss represents death of neurons whose axons project exclusively through the lesioned area (i.e. along the fimbria-fornix and/or the supracallosal striae) (ceil type C in Fig. 13) whereas cell shrinkage would take place in neurons with branched axons projecting both through the lesioned area and outside this area (cell
254
F. H. GAGE et a/.
type B in Fig. 13). A third population of neurons is represented by cells whose projections are entirely outside the lesioned pathways (cell type A). This last group of neurons, which is left undamaged by the FF lesion, would remain unchanged in size and number. This proposal seems to be consistent with the overall projection patterns of the different subnuclei. Thus the HDB, which was virtually unaffected by the FF lesion, is known to have a low proportion of cholinergic neurons that project to the hippocampal formation, whereas this proportion is considerably higher in VDB and, in particular, in the MS where the most severe neuronal cell loss was observed.2~26~23~27 Cell shrinkage
Cell shrinkage was measurable already at 1 day. This shrinkage most likely precedes the death of the choline@ neurons, because cell loss was only apparent by 1 week. The situation at 1 week after lesion is somewhat unclear because, in the l-week animals, the size of the AChE-positive cells was reduced on both sides. This points to a transient reaction of the neurons on the contralateral side, which had passed by 2 weeks after lesion. The cell shrinkage that is subsequently measured ipsilateral to the lesion at 4 and 6 weeks is most probably due to long-term atrophy in surviving but partially axotomized neurons, as indicated in Fig. 13. The histogram presented in Fig. 9 supports the view that not all remaining cholinergic cells undergo this long-term atrophy, but only a subcategory which reduces their average cell size. “Pile up ” Caja12* described the changes that occur in the lesioned cerebral cortex in the axon between the cell body and that part of the axon just proximal to the transection. He observed these changes in four segments, “the terminal bud or ball, varicose segment, hypertrophic segment and normal segment.” Shute and Lewis** also observed changes rostra1 to a transecting FF-lesion, and used these changes as supporting evidence that the cell bodies were located on the septal side of the transection. Using the AChE method they called this hypertrophic state “pile-up,” as the accumulation of AChE in the axon stump gave the appearance of a pile of enzyme. They first saw the “pile-up” several days following transection at which time they felt the enzyme would not have had enough time to be transported to the axon. Thus they speculated that the increase in visible enzyme was due to a local synthesis. We have seen this “pile-up” as early as 1 day following transection, with DFP administered 4 h prior to killing, suggesting that the AChE enzyme transported to the axon stump was synthesized in the cell body and channeled into the axon via the fast axonal transport system, which is known to be the case for the AChE enzyme.16 The “pile-up” is thus most likely the result of the normal transport of materials to the axonal branches of the neurons. The gradual movement of the “pile-up” back to the
region of the cell bodies over the first week following the lesion most likely reflects the progressive dying back of the axon towards the cell. In fact, the length of time that had passed following the FF transection was well correlated with the proximity of the “pileup” of AChE-positive material to the cell body region. Choline acetyltransferase
activity
The decrease in ChAT activity measured at 1 day following the transection was consistent with our measurement of a reduction in AChE-positive cell bodies at day 1. This decrease is most likely due to a reduced expression of the two cholinergic enzymes. At present it is not known whether the decrease in enzyme activity results from transient suppression or a reduced synthesis in the neurons, but our data indicate that this effect is an antecedent to the eventual death of some of these cells. The recovery in ChAT activity measured biochemically in the whole septal area 1 week following the transection may reflect the sprouting response that appears to be occurring in the dorsal lateral quadrant of the septum. Sprouting
By 2 weeks following FF transection there is an apparent sprouting of AChE-positive fibers in the dorsal lateral quadrant of the septal area. Though this apparent cholinergic sprouting response has not been reported previously in the septal region, it is consistent with the electron microscopic observations of Raisman” and Raisman and Field.*’ They demonstrated that, following FF transection, there was an initial degeneration of synaptic terminals in the lateral septum but that the synaptic density returned to normal within about one month. This was interpreted as a reinnervation of the initially vacated synaptic sites through collateral sprouting. In addition, Moore et a1.,19 using the Falck-Hillarp technique6 to visualize catecholamine fibers, observed that after unilateral FF lesion there was a progressive increase in catecholamine fibers in the septal area that reached a peak at about 30 days and persisted for at least 100 days. The apparent sprouting of AChEpositive fibers that we observe in the dorsolateral septum will be discussed in more detail in a future paper (F. H. Gage, unpublished observations). However, it seems clear that this response to the FF transection may well account for the recovery of ChAT activity that was measured biochemically in the whole septal area 1 week following transection. At present it is not known whether this sprouting took place from damaged neurons with remaining intact collaterals in the septum, or from intact neurons that only send local projections to the septum. thank Ulla Jar1and Yrette Jonsson for skillful technical assistance. The study was supported by grants from the Swedish MRC and from the National Institute of Aging (AGO6088). Acknowledgements-We
Retrograde
cell changes
255
REFERENCES I. Abercrombie U. (1946) Estimation of nuclear population from microtome sections. Anat. Rec. 94, 239-247. 2. Amaral D. G. and Kurz J. (1985) An analysis of the origins of the cholinergic and noncholinergic septal projections to the hippocampal formation of the rat. J. cotnp. Neural. 240, 37-59. 3. Armstrong D. M., Saper C. B., Levey A. I., Wainer B. H. and Terry R. D. (1983) Distribution of cholinergic neurons in rat brain: Demonstrated by the immunocytochemical localization of choline acetyltransferase. J. romp. Neural. 216, 53m68. 4. Butcher L. L. (1983) Acetylcholinesterase histochemistry. Handbook qf Chemical Neuroanufomy, Vol. I. Methods 111 ChtTmical Neuroanafomy (eds Bjorklund A. and Hokfelt T.), pp. l-49. 5. Dattz H. M. and Powell T. P. S. (1954) Studies of the connexions of the fornix system. J. Neuro/. Neurosurp. Ps~xhrot. 17, 75-82. 6. Falck B., Hillarp N. A., Thieme G. and Torp A. (1962) Fluorescence of catecholamines and related compounds condensed with formaldehvde. J. Histochem. Cvfochem. 10. 348.-354. 7. Fonnum F. (1970) Topographical and subcellular localization of choline acetyltransferase in the rat hippocampal t-egton. J. Neurochem. 17, 1029-1037. 8. Gage F. H., Bjorklund A. and Stenevi U. (1983) Remnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and noradrenergic afferents. Bruin Res. 268, 27 37. 9. Geneser-Jensen F. A. and Blackstad T. W. (1971) Distribution of acetyl chohnesterase in the hippocampal region of the guinea pig. 2. Zellforsch. 114, 460-48 1. IO. Koelle G. B. (1954) The histochemical localization of chohnesterase in the central nervous system of the rat. J. lwmp. Neural. 100, 21 I-235. II. Koelle G. G. and Friedenwald J. S. (1949) A histochemtcal method for localizing cholinesterase activity. f+oc, Sot e.x[?. Biol. Med., N. Y. 70, 617-622. 12. Kohler C. and Chan-Paley V. (1983) Distribution of gamma aminobutyric acid containing neurons and terminals m the septal area. ilnat. Embryol. 167, 53-65. Ii. Kohler C., Chan-Palay V. and Wu J. Y. (1984) Septal neurons containing giutamic acid decarboxylase immunoreactivity project to the hippocampal region in the rat brain. Anat. Embryol. 169, 41-44. 14. Lewis P. R., Shute C. C. D. and Silver A. (1967) Confirmation from cholineacetvlase of a massive chohnergrc mnervation to the rat hippocampus. J. Physiol., Lond. 191, 215-224. 15. Lowry 0. H., Rosebrouah N. J., Farr A. L. and Randall R. J. (1951) Protein measurement with the folin ohenol reaeent. J. hioj. Chem. 193, 2651274. 16. Lubinska L. (1975) On axoplasmic flow. Int. Rev. Neurobiol. 17, 241.-296. 17. Mcsulam M. M., Mufson E. J., Wainer B. H. and Levey A. 1. (1983) Central cholinergic pathways in the rat: An overview based on an alternative nomenclature (Ch. I -Ch.6). Neuroscience 10, 1185-l 201. 1X. McLardy T. (1955) Observations of the fornix of the monkey. J. camp. Neural. 103, 305. 19. Moore R. Y., Bjorklund A. and Stenevi U. (1971) Plastic changes in the adrenergic innervation of the rat septal area in response to denervation. Bruin Res. 33, 13-35. 20. Raisman G. (1969b) Neuronal plasticity in the septal nuclei of the adult rat. Bruin Res. 14, 2548. II Raisman G. and Field P. M. (1973) A quantitative investigation of the development of collateral reinnervation after partial deafferentation of the septal nuclei. Brain Res. 50, 241-264. 22. Ramon y Cajal S. (1928) Degeneration and Regeneration of the Nerr;ous System. Oxford University Press, London. 73. Rye D. B., Wainer B. H., Mesulam M.-M., Matson E. J. and Saper C. B. (1984b) Cortical projections arising from the basal forebrain: A study of cholinergic and noncholinergic components employing combined retrograde tracing and immunohistochemical localization of choline acetyhransferase. N&roscience-13,v 62%643. 24. Shute C. C. D. and Lewis P. R. (1966) Electron microscoov of cholinereic terminals and acetvlcholinesterase-containine neurons in hippocampal formation of the rat. Z. Zellf&h. mikrosky Anat. 69, 334343.25. Shute C. C. D. and Lewis P. R. (1967) The ascending cholinergic reticular system. Brain: J. Neural. 90, 497-519. 16 Swanson L. W. and Cowan W. M. (1979) The connections of the septal region in the rat. J. camp. Neural. 186.621 656. 27. Woolf N. J.. Eckenstein F. and Butcher L. L. (1984) Cholinergic systems in the rat brain: I. Projections to the limbic telencephalon. Bruin Res. Bull. 13, 751-784.
(Accepted
17 March 1986)
RETROGRADE CELL CHANGES TN MEDIAL SEPTUM AND DIAGONAL BAND FOLLOWING FIMBRIA-FORNIX TRANSECTION: QUANTITATIVE TEMPORAL ANALYSIS F. H. GAGE,* Departments
WICTORIN,$ W. FISCHER,: L. R. WILLIAMS,* S. VARON~and A. BIORKLUND:
K.
of *Neurosciences ZDepartment
and fBiology, IJniversity of California, San Diego, CA 92093, U.S.A. and of Histology, University of Lund, Lund. Sweden
La Jolla.
Abstract-Complete unilateral fimbria-fornix transections. including the overlying cingulate cortex, were administered to female rats. At time points from 1 day to 6 weeks. the sepal-diagonal band region was examined using acetylcholinesterase histochemistry, Cresyl Violet cell staining. and choline acetyitrnnsferase biochemistry. As early as 1 day following the transection a decrease in acetylcholinesterase positive cell body staining was observed in the medial septum; however. no loss of Nissl-stained neurons was measured in Cresyl Violet stained sections until 1 week after the lesion. Maximal loss of acetylcholinesterase”positive cells. as visualized after irreversible acetylcholinesterase inhibition, was measured at I week. and no further change was observed at time points up to 6 weeks after operation. The loss of acetyltransferase-positive cells was greatest in the medial septal area (- 65%) and the vertical limb of the diagonal band (- 55%). Little cell loss was measured in the horizontal limb of the diagonal band. This is consistent with the known projections of these cell bodies. Remaining ac~tyIcholinesterase-positive cell bodies in the medial septum had shrunk by about 20% (measured as the diameter along the major axis). A marked neuronal celf toss (about 50%) was demonstrable in the medial septum and vertical limb of the diagonal band in the C‘resyi Violet-stained sections, too. A pile-up of acetylcholinesterase-stained material was observed in the dorsal-lateral quadrant of the septai area just proximal to the lesion at 1 day following transection. This pile-up occurred in the medial septum and diagonal band area up to 1 week followmg the transection. and had nearly disappeared by 2 weeks post-transection. Choline acetyltransferase biochemical activity. measured in samples of whole septum, decreased significantly at I day but subsequently returned to control levels. By 2 weeks following transection, an increase in acetylcholinesterase-positive stained fibers was observed in the dorsal-lateral quadrant of the septum, ipsilateral to the lesion relative to the contralateral septum. This response, which was interpreted as sprouting from the lesioned axons proximal to the transection. probably accounted for the rise in choline acetyltransferase biochemical activity in the whole septum following the reduction on
the first day.
Neurons of the medial septum and the diagonal band of Broca send fibers via the fimbria-fornix (FF) bundle and the supracallosal striae to terminate in the hippocampal formation. Using standard histological techniques, Daitz and Powell’ and McLardy’* demonstrated that FF transection induces a “severe, rapid and consistent degeneration in the medial septum and diagonal band nuclei” which results in “complete atrophy or shrinkage of all the cells of the ipsilateral medial septal nucleus and partial degeneration of the nucleus of the diagonal band.” A large portion of the cells projecting to the hippocampus from the medial septai and diagonal ~
-__--___
Address correspondence to: F. H. Gage, Department Neurosciences (M-024), University of California,
--
band nuclei is cholinergic.2,3,‘7,23Z” The anterograde decrease in the two cholinergic marker enzymes, choline acetyltransferase (ChAT) and acetylcholinesterase (AChE), in the hippocampus following FF transection is well documented both biochemically and histochemically (e.g. Refs 8, 14 and 24). Also well documented is the AChE-stained “pile-up” in septohippocampal axons proximal to the transe~tion,14 which represents the directionality of the fibers. However, a quantitative assessment of the degenerative changes-cell loss and/or atrophy-of the septal cholinergic neurons is so far lacking. The objective of the present study is to document the time course and extent of cell loss and atrophy in the cholinergic neurons in the septum and diagonal band areas following complete transection of all the known cholinergic fibers leading to the hippocampal formation via the FF and the supracallosal striae. This quantitative morphological time course, together with the documentation of concomitant biochemical changes and changes in cholinergic fiber patterns in the septum. should provide an accurate portrait of
of San
Diego, La Jolla, CA 92093, U.S.A. .4hhreviations:
AChE, acetylcholinesterase; ChAT, choline acetyltransferase; DFP, diisopropyl fluorophosphate; FF, ~rnb~a-fornix; HDB, horizontal limb of the diagonal band of Broca; MS, medial septum; VDB, vertical limb of the diagonal band of Broca. 241
242
I-‘.H. GAOE et al.
the dynamic changes occurring in these neurons foIfowing damage. This info~ation will, moreover, be of particular value as a basis for the assessment of effects of exogenous and endogenous trophic factors or the process of cell death after axotomy in the forebrain cholinergic system. This model of traumainduced cell death and its consequences may also give insights into the mechanisms underlying certain clinical cell death syndromes, such as Alzheimer’s disease. EXPERIMENTAL
PROCEDURES
Animals Adult female Sprague-Dawley rats (ALAB., Stockholm, Sweden) weighing between 200 and 225 g at the beginning of the experiment were used. Surgical procedures The animals were anesthetized with intraperitoneal injections of a ke~mine-xyl~ne mixture (10 mg/kg Keta!ar, Parke-Davis, and 5 mg/kg Rompun, Hoechst). All animals in each of the experimental groups received the same complete unilater~ aspirative lesion of the fimb~a-fo~ix and supracallosal striae (the FF lesion), as described previously* (cf. Figs la and 2A). The lesions were conducted visually with the aid of a surgical microscope. Htktochembtry FF-lesioned animals were taken 1 day and 1, 2,4 and 6 weeks following surgery for histochemical evaluation. Six rats were used at each survival time, except at 4 weeks when the group comprised 5 rats. Four normal unoperated rats were included for comparison. To obtain consistent and complete AChE staining of the septaldiagonal band neurons, the animals were injected intr~~~arly witti the irreversible AChE inhibitor, diisopropyl fluorophosphate (DFP, Sigma) 4 h before being kiiied at a dose of 2 mg/kg (dissolved in peanut oil) according to the method of Butcher.” The rats were perfused via the ascending aorta with 150 ml of ice-cold 4% buffered paraformafdehyde. The brains were postfIxed for 2 h in the same fixative and then left overnight in 10% sucrose at 4°C. Sections were cut at 15pm in a Dittes cryostat and every third section was stained for AChE according to the Koelle thiocholine method,‘*J1 using 10m4M promethazine as an inhibitor of nonspecific esterases and silver nitrate intensification of the reaction product.9 Alternate sections were taken for Cresyl Violet staining. Quantitative morphology AChE-positive and Cresyl Violet-stained cells were counted in three regions defined in the following ways (Fig. 1): Medial septum (MS) = rostrally from the level of the genu of the corpus ca!losum; cauda!ly, to the level of the crossing of the anterior commissure; and latera!!y, 0.75mm from the midline. Vertical limb of the diagonal band of Broca (VDB) = same rostra! and caudaf boundaries as above, with the dorsal boundary defined as the anterior commissure (dashed line in Fig. lc). This landmark WBS chosen to ensure r&ability of a!! counts, rather than for rigid anatomical reasons. The ventral boundary was defined by the medial border of the o!factory tubercle. Horizontal limb of the diagonal band of Broca @IDB) = same rostra! and caudal boundaries, except that the lateral boundary extended out to the lateral tip of the anterior c~mmissutt (Fig. Id). An AChE-positive cell was defined as a dark brown body with a minimum diameter of 12 pm along the major axis. Figure 2 provides representative examp!es of structures defined as cell bodies (arrows). RofG without arrows were not counted, either because staining was too weak or because they were too sma!i (< 12 pm along the major axis). The vast majority of AChE cells fell within the range of
!8-29pm (cf. Fig. 9). AChE-stained cells smaller tfidn 12pm in diameter were without regular cell shape or processes and could not be distinguished by cellular criteria, Cresyl Violet-stained neurons were defined as cells having diameters > I2 pm and containing both a nucleus and Nissl substance. Every 6th of the mounted sections (i.e. every 18th section) was counted. A counting grid (0.75 x 0.75 mm) was used to entirely cover the areas of interest. Figures l&d show the placement of the counting grid. For each animal, both the regions ipsilatera! and contralaterai to the lesion were counted at x f6 objective magnification and the results are presented as a percentage of remaining AChE-positive cells (ipsilateral region/contralateral region x 100). Cell size was estimated with the same counting grid at x40 objective magnification by two independent observers measuring the length along the major axis of 60 randomly chosen AChEpositive cells of each region on both the ipsilaterat and contralateral sides. Size differences between the two sides and differences from normal unoperated rats were again expressed as percentages. All AChE-positive cell number data were corrected for cell shrinkage according to the Abercrombie formula.’ Choline acetyltransferase assay Separate sets of 6 rats with FF lesions were taken at 1day, 1, 2, 4 and 6 weeks following surgery for bi~hemica1 analysis of ChAT activity in the whole septum (media! plus lateral septa! nuclei) according to the micromethod of Fonnum.’ protein content was measured according to Lowry et ~1.‘~Values are expressed as percentages of an unoperated control group and the contralateral side in the operated animals. Statistical analysis To determine whether statistical differences existed between groups on measures of cell size and cell number, a one-factor analysis of variance (ANOVA) was used with post hoc Neuman-Keuls test, for comparisons between group. In the text, the P values presented are derived from the Neuman-Keuls test, but only when the ANOVA test gave a sigt&cant level of P < 0.05. For comparisons of ChAT activity between groups, total septal ChAT activity in an unlesioned control group of animals was compared with the septal ChAT activity from the contralateral septum of each of the lesioned groups using one-way ANOVA. Since no significant differences between groups were observed, these values were combined into a single control group with which the septal CUT activity ipsilateral to the lesion was compared, again using an ANOVA with post hoc Neuman-Keuls. For clarity of presentation, these values were converted to percentages of the combined control group. Statistical analyses, however, refer to the calculations conducted on the actual ChAT va!ues. When making within group comparisons (e.g. comparing actual cell sizes between ipsilateral and contralateral septum), a paired-comparison test for related groups was used, as indicated in the text. RESULTS
in Figs la and ZA) was similar to that used previous1y.s This lesion is known to reduce ChAT activity in the hippocampus by at least 90% and to eliminate essentially all AChEpositive fiber staining in the dorsal two-thirds of the hippocampus. The
FF lesion (illustrated
Acetyichoiinesterase-posirive
cell stair&g
One day following the FF lesion (Figs 3 and 4) a
20% reduction in the number of cells in the MS and VDB was observed on the side ipsilateral to the lesion
Retrograde
cell changes (b)
(d)
Fig. I. Coronal sections from a rat with a fimbria-fornix lesion, indicating approximately where the cells were counted and what constituted the regional boundaries for the purpose of counting. (a) Lesion cavity at the level of the fimbria-fornix; (b) rostra1 medial septal area (MS). first level of cell counts; (c) medial septal area and vertical limb of the diagonal band of Broca (VDB). midlevel of cell counts; (d) medial septal area and horizontal limb of the diagonal band of Broca (HDB). representing the most caudal level of cell counts. Squares indicate actual size of the grid that was used to limit the areas of counting.
related t-test). By 1 week follow(Figs 3 and 5) the cell numbers were reduced by 65% in the MS and by 55% in the VDB (P <’ 0.001). This relative decrease in AChE-positive cells did not change over the subsequent 5 weeks, the longest time period studied (Figs 3 and 6). In the HDB there was no significant cell loss at 1 day, but by 1 week a 20% decrease in AChE-positive cells ipsilateral to the lesion was seen (P < 0.05). This percentage decrease did not change over the subsequent 5 weeks (Fig. 3).
(P < 0.05; ing
the
Student’s
lesion
<‘re.s~Y Violet cell staining To clarify whether the disappearance of AChEstained cells reflected an actual loss of cell bodies from the area, neuronal profiles (defined as Nisslstained cell bodies with a diameter along the major axis of < I2 pm) were counted also in the adjacent Cresyl Violet-stained sections. As illustrated in Table I there was no significant reduction in the total number of neurons in the MS, the VDB or the HDB relative to the side contralateral to the lesion at I day, but at 1 week following the lesion there was a significant loss of neurons in both the MS (40%) and the VDB (20%) (cf. Fig. 5). Two additional animals
were examined at 6 weeks following FF transection. and they both showed an approximate 50% dccreasc in total cell counts as measured with Cresyl Violet staining in both the MS and the VDB (Fig. 6). Cell shrinkage Shrinkage of AChE-positive cells was already apparent 1 day after the FF lesion in the MS and VDB. It amounted to about IO- 15% relative to the contralateral side (Fig. 7) and about 15.20% relative to the normal controls (Fig. 8). The side difference in cell size was no longer apparent at I week, but it reappeared again at 4 and 6 weeks following the lesion (Fig. 7). From Fig. 8, which illustrates the actual cell sizes, it is apparent that the AChE-positive cell bodies in the MS were also significantly reduced in size on the contralateral side at all time points in the lesioned rats. This effect, which was most pronounced at I week, apparently accounted for the transient disappearance of the side difference in cell size at I week post-lesion, as seen in Fig. 7. The AChE-positive cells in the HDB showed significant decrease in size only at I week. and then on both sides of the brain (Fig. 8); this effect did not persist at longer time points. From Fig. 9. which shows the cell
F. H. GAGE et ul.
244
t----r(r
E ----a
z;90-
rY
."....,, ,..... ""._,.. ,___,_,,..,,.'-"'f
5 g
70-
$
60-
0
MS VDB
%o%
_______--
------____
*
g 402 A
T
30-
T
8 20w 2 io-
J,
I
Id
twk
I
I
I
2wks 4wks TIME AFTER UNfFF-LESION
I
I
6wks
Fig. 3. The number of AChE-positive cells in the three regions counted (MS, VDB and HDB) expressed as a percentage of the control side, contralateral to the lesion. Cell counting was conducted in separate groups of animals (symbols indicate means and SEM for n = 6) at each of the time points. Significant decreases (at least P < 0.01, Nmman-Keuls test) in e&l mtmber as compared to the control side are seen at ah time points for MS and VDB. Open symbols indicate unlesioned control animals.
Table 1. Number of ceil bodies after frmbria-fornix lesion in the medial septal nucleus and the vertical limb of the diagonal band of Broca as counted in Cresyi Violet-stained sections Postop. time
n
Cell num$Sr (% of contr. side;
1 day 7 days 6 weeks
4 5 2
101.4 f 4.7 62.4 f 4.2+ 48.3
SEM)
99.2 f 7.7 78.3 f 4.7’ 54.0
*P < 0.01 compared to l-day group (Student’s t-test). Expressed as percent of the control side. The counts were not corrected for cell shrinkage.
size ~s~bution
in a frequency ~sto~am, it is clear that the overall reduction in cell size was due to an increased number of cells on the lesioned side that fell into the smaller range of cell sizes (12-17gm diameter along the major axis).
Septal choline ncetyltransferase activity In whole septum (medial and lateral areas combined, including the diagonal band nuclei) a 25X decrease in ChAT activity was apparent at 1 day after FF transection. This decrease in ChAT activity returned to within the levels of the control side by 1 week and remained at this level at later time points (Fig. 10).
At 1 day post-transection there was a dramatic accumulation of AChE-positive staining in the dorsal lateral septal area and in the cingulate cortex just rostra1 to the lesion (Fig. 1la&. The ovulation appeared to be in the axons which had been transected by the lesion. By 1 week the pile-up was located more proximal to the cell bodies of the septal-diagonal band region (Figs 11b,d). At this time -..
Fig. 4. Photomicrographs from the mediaI septum (a and c) and the vertical limb of the diagonal band (b and d) of AChE-stained sections. (a) and (b) unfesioned control animals; (c) and (d) 1 day follo~ng FF lesion. Fig. 5. AChE (a and b) and Cresyl Violet (c and d) stained sections from the medial septum and the vertical limb of the diagonal band at 1 week after FF lesion. Fig. 6. AChE (a and b) and Cresyl Violet (c and d) stained sections from the medial septum and the vertical limb of the diagonal band at 6 weeks after FF lesion.
_-
. Fig. 2. (A) AChE-stained cells in the VDB demonstrating the range of cells observed. Arrows indicate profiles counted as AChE-positive neurons. Profiles without arrows were not included, either because they were too small (< 12 pm along major axis), or because they were too weakly stained. (B) Photomicrograph of extent of fimbria-fornix transection. cc, corpus callosum: Iv, lateral ventricle; FL fimbria: f, fornix; sm. stria medullaris: ot. optic tract: ox, optic chiasm. 23s
b
/_ Fig.
s
c
,d Fig. I I. Photomicrograph of “pile-up” in AC’hE-positive stained 4 hr after DFF injections at 1day (a and c) and I week (b and d) denote the position of the photagraphs above. cc. corpus callosum, commlssure; ox. optic cfuasm; ms. medial septum: ~a,
249
fibers in the dorsal lateral septal area following FF lesion. The framed area’; Iv, lateral ventricle: ac, medial anterior IaWral anterior commissure.
Fig. 12. Photomicrograph non-DFP-treated animal
of AChE-positive stained fibers in the dorsal lateral septal area in ‘i at 4 weeks following FF lesion. (A) Low-power magnification, and (B) higher power magnification of the same section.
250
Retrograde
cell changes
MS
c---q . ..
I
I
Id
I
I
lwk
I
I
2wks
.. l
1 Gwks
I
Lwks
TIME AFTER
VDB H/-JB
UNIFF-LESION
Fig. 7. Cell sizes of AChE-positive cell bodies as indexed by cell diameter along the major axes of 60 randomly measured cells in each of the three areas in a total of four animals at each time point. Values are expressed as a percentage of the control side contralateral to the lesion. Si~i~cant decreases in cell sizes were observed at 1 day and 4 and 6 weeks for MS and VDB, at the P <: 0.05 level; Neuman -Kculs lest.
1 WEEK 40 1
AFTER
6 WEEKS
UNIFF-LESION [z1
l
nonieslon lesion
side side
40
AFTER
UNIFF-LESION 0
nonteslon
II
lesron
side side
@L
#
30
zz iz ; 0
l-l 20
10
12-17 18-23 24-29 CELL SIZE, MAJOR AXIS
(vrn)
il,
12-17 18-23 24-29 30-35 CELL SIZE, MAJOR AXIS (pm)
Fig. 8. Size of AChE-positive neurons in the medial septal nucleus as indexed by the diameter along the major axis and expressed in micrometers bm). Values are presented for non-lesioned control animals as well as for the Iesioned and control sides for each of the three brain regions measured at each of the four time points following the FF lesion. Filled stars denote significant differences from normal uniesioned controls (P < 0.05; Neuman-Keuls test). Open stars denote difference between sides (P < 0.05; Student’s related r-test).
F. H. GAGEet al.
HDB
VDB
MS l
lwk
Id
con
Iwk
6wk
con
Id
lwk
TIME AFTER
Lwk
6wk
con
Id
lwk
Lwk
6wk
UNIFF-LESION
Fig. 9. Frequency histogram showing the distribution of cell sizes of AChE-positive neurons in the medial septaf nucleus at 1 week (left) and 6 weeks (right) after FF lesion. For each group a total of 60 cells on each side was measured in 6 rats (i.e. 10 cells per side per rat). l , nonlesion side; a, lesion side.
point it was easy to trace some of the distended AChE-positive fibers from their most distal point to the now atrophied AChE-positive cell bodies. By 2 weeks following the lesion, there were very few
I
1
cm. Id
1
lwk
I
distended fibers with the characteristic “piled-up” appearance. Those that were apparent were in the very medial portion of the medial septal area, or in the vertical limb of the diagonal band.
1
1
4wks 2wks TIME AFTER UNIFF-LESION
I
1
6wks
Fig. 10. Choline acetyltransfwasc activity in whole septum (the cm&&d septal-diagwai band regim) tpthksion,atthesametinM@ntsasin explwscdasa~~oftbe~~,~ Fig. 3. ThecontmIgfo~~xoav e&tr&asw8llasthepoolu¶valueafromthe diihnce (P < 0.05) between controls and contralateral sides of the fke ksion& ipsilaterai septum was observed only at day 1 post-lesion.
Retrograde
cell changes
Fig. 13. Schematic representation of the proposed changes in the three anatomically different types ot cholinergic neurons in the septal-diagonal band region at various time points following FF lesion. For further explanation, see text.
Sprouting
in the septum
The puzzling recovery of biochemically measured ChAT activity in the septum following the initial reduction seen at 1 day was examined further by looking at the pattern of AChE staining in the septum in the absence of DFP pretreatment. Already by 2 weeks following the FF lesion, there was a region of clearly enhanced AChE-positive staining in the dorsal lateral septum ipsilateral to the lesion. This enhanced neuropil staining, which is interpreted as a sprouting response, was clearly different from the “pile-up” that was evident at I day and 1 week post-lesion. While the “pile-up” was characterized by thick, dense accumulations of AChE in the distal portions of the lesioned axon, just proximal to the lesion at I day, and closer to the cell body by 1 week, the sprouting at 4 weeks appeared as fine fiber arborizations similar in density and appearance to the fibers in the intact septum contralateral to the lesion (Fig 12).
DISCUSSION
The present results portray a sequence of degenerative and regenerative events after axotomy in the septo-hippocampal projection system. In Fig. 13 we propose an interpretation of these events based on the assumption that the FF lesion will affect the AChE-positive neurons in the septal-diagonal band region differently depending on the anatomical arrangement of their axonal projections. Cell death The present results confirm the early findings of Daitz and Powell’ and McLardy,‘* demonstrating that there is a rapid and severe loss of cells in the medial septal area and diagonal band of Broca, as indexed by Cresyl Violet-stained cell counts. We further demonstrate that the AChE-positive cells in this region are lost in dramatic proportions. The loss of AChE-positive cells reaches a maximum within 1 week post-transection and no further cell loss is observed over the 6 weeks investigated. That the
disappearance of AChE-stained neurons (in DFPpretreated rats) was due to an actual loss of neurons is supported by the parallel counts in the Crcsyl Violet-stained sections. Interestingly. the loss of AChE-positive cell bodies amounted to about 70% in the MS area, whereas the corresponding figure in the Cresyl Violet-stained sections was about 50%. From the total number of neurons counted in the adjacent sections with the two types of stains, our impression is that about one-third of the medial septal neurons (defined as cell bodies with a diameter > 12pm along the major axis) are AChEpositive in the DFP-treated rats (see also Refs 2, 17 and 23). This means that only about half of the neurons lost in the Cresyl Violetstained sections could be accounted for by the disappearance of cholinergic septal-hippocampal neurons, and thus that noncholinergic neurons arc likely to degenerate as well. In fact. it is known that a significant proportion of the septal-diagonal band neurons projecting to the hippocampus is noncholinergic,* and there is evidence from glutamate decarboxylase immunohistochemistry in combination with retrograde tracing that part of the septohippocampal pathway is GABAergic.“,” It is likely. therefore, that the retrograde cell loss in the septaldiagonal band area after FF lesion is not confined to the cholinergic projection system. It is possible that the loss in cells reported in this paper could be a result of the shrinkage of damaged cells below the 12 bcrn criterion, but we find this unlikely in that there did not appear to be an increase in the number of small AChE positive elements < 12 pm on the lesioned side. It is also possible that the AChE cells were not dying, but rather becoming AChE-negative. We also find this unlikely because a major cell loss (about 50%) was also evident in the Cresyl Violet-stained sections. In Fig. 13 we propose that the observed cc.11loss represents death of neurons whose axons project exclusively through the lesioned area (i.e. along the fimbria-fornix and/or the supracallosal striae) (ceil type C in Fig. 13) whereas cell shrinkage would take place in neurons with branched axons projecting both through the lesioned area and outside this area (cell
254
F. H. GAGE et a/.
type B in Fig. 13). A third population of neurons is represented by cells whose projections are entirely outside the lesioned pathways (cell type A). This last group of neurons, which is left undamaged by the FF lesion, would remain unchanged in size and number. This proposal seems to be consistent with the overall projection patterns of the different subnuclei. Thus the HDB, which was virtually unaffected by the FF lesion, is known to have a low proportion of cholinergic neurons that project to the hippocampal formation, whereas this proportion is considerably higher in VDB and, in particular, in the MS where the most severe neuronal cell loss was observed.2~26~23~27 Cell shrinkage
Cell shrinkage was measurable already at 1 day. This shrinkage most likely precedes the death of the choline@ neurons, because cell loss was only apparent by 1 week. The situation at 1 week after lesion is somewhat unclear because, in the l-week animals, the size of the AChE-positive cells was reduced on both sides. This points to a transient reaction of the neurons on the contralateral side, which had passed by 2 weeks after lesion. The cell shrinkage that is subsequently measured ipsilateral to the lesion at 4 and 6 weeks is most probably due to long-term atrophy in surviving but partially axotomized neurons, as indicated in Fig. 13. The histogram presented in Fig. 9 supports the view that not all remaining cholinergic cells undergo this long-term atrophy, but only a subcategory which reduces their average cell size. “Pile up ” Caja12* described the changes that occur in the lesioned cerebral cortex in the axon between the cell body and that part of the axon just proximal to the transection. He observed these changes in four segments, “the terminal bud or ball, varicose segment, hypertrophic segment and normal segment.” Shute and Lewis** also observed changes rostra1 to a transecting FF-lesion, and used these changes as supporting evidence that the cell bodies were located on the septal side of the transection. Using the AChE method they called this hypertrophic state “pile-up,” as the accumulation of AChE in the axon stump gave the appearance of a pile of enzyme. They first saw the “pile-up” several days following transection at which time they felt the enzyme would not have had enough time to be transported to the axon. Thus they speculated that the increase in visible enzyme was due to a local synthesis. We have seen this “pile-up” as early as 1 day following transection, with DFP administered 4 h prior to killing, suggesting that the AChE enzyme transported to the axon stump was synthesized in the cell body and channeled into the axon via the fast axonal transport system, which is known to be the case for the AChE enzyme.16 The “pile-up” is thus most likely the result of the normal transport of materials to the axonal branches of the neurons. The gradual movement of the “pile-up” back to the
region of the cell bodies over the first week following the lesion most likely reflects the progressive dying back of the axon towards the cell. In fact, the length of time that had passed following the FF transection was well correlated with the proximity of the “pileup” of AChE-positive material to the cell body region. Choline acetyltransferase
activity
The decrease in ChAT activity measured at 1 day following the transection was consistent with our measurement of a reduction in AChE-positive cell bodies at day 1. This decrease is most likely due to a reduced expression of the two cholinergic enzymes. At present it is not known whether the decrease in enzyme activity results from transient suppression or a reduced synthesis in the neurons, but our data indicate that this effect is an antecedent to the eventual death of some of these cells. The recovery in ChAT activity measured biochemically in the whole septal area 1 week following the transection may reflect the sprouting response that appears to be occurring in the dorsal lateral quadrant of the septum. Sprouting
By 2 weeks following FF transection there is an apparent sprouting of AChE-positive fibers in the dorsal lateral quadrant of the septal area. Though this apparent cholinergic sprouting response has not been reported previously in the septal region, it is consistent with the electron microscopic observations of Raisman” and Raisman and Field.*’ They demonstrated that, following FF transection, there was an initial degeneration of synaptic terminals in the lateral septum but that the synaptic density returned to normal within about one month. This was interpreted as a reinnervation of the initially vacated synaptic sites through collateral sprouting. In addition, Moore et a1.,19 using the Falck-Hillarp technique6 to visualize catecholamine fibers, observed that after unilateral FF lesion there was a progressive increase in catecholamine fibers in the septal area that reached a peak at about 30 days and persisted for at least 100 days. The apparent sprouting of AChEpositive fibers that we observe in the dorsolateral septum will be discussed in more detail in a future paper (F. H. Gage, unpublished observations). However, it seems clear that this response to the FF transection may well account for the recovery of ChAT activity that was measured biochemically in the whole septal area 1 week following transection. At present it is not known whether this sprouting took place from damaged neurons with remaining intact collaterals in the septum, or from intact neurons that only send local projections to the septum. thank Ulla Jar1and Yrette Jonsson for skillful technical assistance. The study was supported by grants from the Swedish MRC and from the National Institute of Aging (AGO6088). Acknowledgements-We
Retrograde
cell changes
255
REFERENCES I. Abercrombie U. (1946) Estimation of nuclear population from microtome sections. Anat. Rec. 94, 239-247. 2. Amaral D. G. and Kurz J. (1985) An analysis of the origins of the cholinergic and noncholinergic septal projections to the hippocampal formation of the rat. J. cotnp. Neural. 240, 37-59. 3. Armstrong D. M., Saper C. B., Levey A. I., Wainer B. H. and Terry R. D. (1983) Distribution of cholinergic neurons in rat brain: Demonstrated by the immunocytochemical localization of choline acetyltransferase. J. romp. Neural. 216, 53m68. 4. Butcher L. L. (1983) Acetylcholinesterase histochemistry. Handbook qf Chemical Neuroanufomy, Vol. I. Methods 111 ChtTmical Neuroanafomy (eds Bjorklund A. and Hokfelt T.), pp. l-49. 5. Dattz H. M. and Powell T. P. S. (1954) Studies of the connexions of the fornix system. J. Neuro/. Neurosurp. Ps~xhrot. 17, 75-82. 6. Falck B., Hillarp N. A., Thieme G. and Torp A. (1962) Fluorescence of catecholamines and related compounds condensed with formaldehvde. J. Histochem. Cvfochem. 10. 348.-354. 7. Fonnum F. (1970) Topographical and subcellular localization of choline acetyltransferase in the rat hippocampal t-egton. J. Neurochem. 17, 1029-1037. 8. Gage F. H., Bjorklund A. and Stenevi U. (1983) Remnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and noradrenergic afferents. Bruin Res. 268, 27 37. 9. Geneser-Jensen F. A. and Blackstad T. W. (1971) Distribution of acetyl chohnesterase in the hippocampal region of the guinea pig. 2. Zellforsch. 114, 460-48 1. IO. Koelle G. B. (1954) The histochemical localization of chohnesterase in the central nervous system of the rat. J. lwmp. Neural. 100, 21 I-235. II. Koelle G. G. and Friedenwald J. S. (1949) A histochemtcal method for localizing cholinesterase activity. f+oc, Sot e.x[?. Biol. Med., N. Y. 70, 617-622. 12. Kohler C. and Chan-Paley V. (1983) Distribution of gamma aminobutyric acid containing neurons and terminals m the septal area. ilnat. Embryol. 167, 53-65. Ii. Kohler C., Chan-Palay V. and Wu J. Y. (1984) Septal neurons containing giutamic acid decarboxylase immunoreactivity project to the hippocampal region in the rat brain. Anat. Embryol. 169, 41-44. 14. Lewis P. R., Shute C. C. D. and Silver A. (1967) Confirmation from cholineacetvlase of a massive chohnergrc mnervation to the rat hippocampus. J. Physiol., Lond. 191, 215-224. 15. Lowry 0. H., Rosebrouah N. J., Farr A. L. and Randall R. J. (1951) Protein measurement with the folin ohenol reaeent. J. hioj. Chem. 193, 2651274. 16. Lubinska L. (1975) On axoplasmic flow. Int. Rev. Neurobiol. 17, 241.-296. 17. Mcsulam M. M., Mufson E. J., Wainer B. H. and Levey A. 1. (1983) Central cholinergic pathways in the rat: An overview based on an alternative nomenclature (Ch. I -Ch.6). Neuroscience 10, 1185-l 201. 1X. McLardy T. (1955) Observations of the fornix of the monkey. J. camp. Neural. 103, 305. 19. Moore R. Y., Bjorklund A. and Stenevi U. (1971) Plastic changes in the adrenergic innervation of the rat septal area in response to denervation. Bruin Res. 33, 13-35. 20. Raisman G. (1969b) Neuronal plasticity in the septal nuclei of the adult rat. Bruin Res. 14, 2548. II Raisman G. and Field P. M. (1973) A quantitative investigation of the development of collateral reinnervation after partial deafferentation of the septal nuclei. Brain Res. 50, 241-264. 22. Ramon y Cajal S. (1928) Degeneration and Regeneration of the Nerr;ous System. Oxford University Press, London. 73. Rye D. B., Wainer B. H., Mesulam M.-M., Matson E. J. and Saper C. B. (1984b) Cortical projections arising from the basal forebrain: A study of cholinergic and noncholinergic components employing combined retrograde tracing and immunohistochemical localization of choline acetyhransferase. N&roscience-13,v 62%643. 24. Shute C. C. D. and Lewis P. R. (1966) Electron microscoov of cholinereic terminals and acetvlcholinesterase-containine neurons in hippocampal formation of the rat. Z. Zellf&h. mikrosky Anat. 69, 334343.25. Shute C. C. D. and Lewis P. R. (1967) The ascending cholinergic reticular system. Brain: J. Neural. 90, 497-519. 16 Swanson L. W. and Cowan W. M. (1979) The connections of the septal region in the rat. J. camp. Neural. 186.621 656. 27. Woolf N. J.. Eckenstein F. and Butcher L. L. (1984) Cholinergic systems in the rat brain: I. Projections to the limbic telencephalon. Bruin Res. Bull. 13, 751-784.
(Accepted
17 March 1986)