Ž .Brain Research 804 1998 331–336

Short communication

Time course for kindling-induced changes in the hilar area of the dentategyrus: reactive gliosis as a potential mechanism

Beth Adams a, Ee Von Ling a, Liezanne Vaccarella a, Gwen O. Ivy b, Margaret Fahnestock c,Ronald J. Racine a,)

a Department of Psychology, McMaster UniÕersity, Hamilton, Ontario, Canada L8S 4K1b DiÕision of Life Sciences, UniÕersity of Toronto, Scarborough, Ontario, Canada M1C 1A4

c Department of Biomedical Sciences, McMaster UniÕersity, Hamilton, Ontario, Canada L8S 4K1

Accepted 2 June 1998

Abstract

Recurrent seizure activity induced during kindling has been reported to cause an increase in the hilar area of the dentate gyrus of thehippocampus. To date, very little is known about the mechanism of this increase. This study investigated the time course forkindling-induced changes in the hilar area of the dentate gyrus at seven days, one month, and two months post-kindling. Hilar area of thedentate gyrus was significantly increased by approximately 46% at seven days and remained elevated at one month, but declined back to

Ž .control levels by two months. Glial fibrillary acidic protein GFAP immunostaining was also evaluated at the same time points todetermine whether kindling-induced changes in the hilar area of the dentate gyrus are related to kindling-induced glial cell changes.Increases in hilar GFAP immunostaining by approximately 57% were observed at seven days and at one month post-kindling, but not attwo months post-kindling. These findings indicate that kindling-induced changes in the hilar area of the dentate gyrus and kindling-in-duced glial cell changes follow a similar time course, and that kindling-induced glial cell changes may mediate the observed changes inthe hilar area of the dentate gyrus. q 1998 Elsevier Science B.V. All rights reserved.

Keywords: Kindling; Gliosis; Hippocampus; Hilus; Seizure; Plasticity; Dentate gyrus

Kindling can be defined as the progressive increase inelectrographic and behavioral seizure activity produced byspaced and repeated electrical stimulation of certain fore-

w xbrain structures 6 . It has been well-documented thatkindling produces a variety of changes in the hippocampalregion of the brain, including sprouting of the mossy fiber

Ž . w xpathway i.e., axons of the dentate granule cells 4,12 anda decrease in neuronal density, particularly in the hilus of

w xthe dentate gyrus 3,5 . Kindling also causes another typeof structural change in the hippocampus: an increase in the

w xsize of the hilar region of the dentate gyrus 1,2,14 . Todate, however, potential mechanisms underlying this effecthave not been investigated. From this point onwards, thehilar region of the dentate gyrus will be referred to aseither the hilus or the hilar area.

One possibility is that kindling-induced changes in thew xhilar area may be related to reactive gliosis 2 . Reactive

) Corresponding author. Fax: q1-905-529-6225; E-mail:racine@mcmail.cis.mcmaster.ca

Ž .gliosis is generally characterized by: 1 the proliferationŽ .and hypertrophy of glial cell bodies and processes and 2

the dramatic increases in the levels of glial fibrillary acidicŽ . w xprotein GFAP and GFAP mRNA 13 . Furthermore, in-

creased GFAP immunostaining is considered to be thebiochemical hallmark denoting the transformation of nor-

w xmal glial cells to reactive glial cells 13 .Recent research has demonstrated that kindling up-regu-

lates GFAP mRNA and protein levels in a time-dependentw xmanner 7,13 , and that kindling causes glial cell hypertro-

w xphy and proliferation 8 . It has also been reported thatkindling-induced reactive gliosis can occur in the absence

w xof neuronal loss or degeneration 11 .Ž .The objectives in this study were two-fold: 1 to

determine the time course for kindling-induced changes inŽ .the hilar area in adult rats and 2 to evaluate whether

kindling-induced reactive gliosis could account for kin-dling-induced changes in the hilar area. Hilar area andreactive gliosis were evaluated at seven days, one month,and two months post-kindling compared to non-kindled,implanted controls. Specifically, we were interested in

0006-8993r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.Ž .PII: S0006-8993 98 00605-2

( )B. Adams et al.rBrain Research 804 1998 331–336332

determining whether kindling-induced reactive gliosis andkindling-induced hilar area changes follow a similar timecourse.

Ž .Adult male Long–Evans hooded rats ns20 weighingbetween 300–350 g were used in these experiments. Ratswere maintained on an ad lib feeding schedule, housedindividually, and kept on a 12 h onr12 h off light cycle.Using stereotaxic procedures, rats were anaesthetized with

Ž .sodium pentobarbitol 65 mgrkg and were implantedwith a bipolar electrode made from teflon-coated stainlesssteel wires in the right perforant path. Stereotaxic coordi-nates for the perforant path were 7.6 mm posterior and 4.1mm lateral to bregma, and 3.3 mm ventral to the brainsurface. The electrode was held in place by dental acrylicand three stainless steel screws inserted into the skull.Following a two-week recovery period, rats were randomly

Ž .assigned to either a kindled ns15 or a non-kindledŽ .ns5 group. Non-kindled control rats remained in thecolony for 32 days post-surgery.

Rats in the kindled group received a 1-s train of 1-mspulses at a frequency of 60 Hz and pulse intensity of500–700 mA twice a day for 11 days. Progression ofkindling was monitored behaviourally using Racine’s

w xseizure classification scale 10 and electrophysiologicallyusing an electroencephalogram of the evoked epileptiform

Ž .afterdischarges AD . Each stimulation evoked an AD ofgreater than 5 s. Each rat received a total of 22 afterdis-charges.

Following kindling, rats were randomly assigned to oneŽ . Ž .of three groups: seven-day ns5 , one-month ns5 and

Ž .two-month ns5 groups. Kindled rats remained in thecolony for the assigned time period following the lastkindling stimulation.

As expected, repeated-measures ANOVA confirmed thatbehavioral seizure stage increased as a function of stimula-

Ž .tion number across all groups p-0.05; data not shown ,and there was no significant difference in the behavioral

Ž .progression of kindling between the groups p)0.05 .Also, AD duration increased as a function of stimulation

Žnumber across all three kindled groups p-0.05; data not.shown , but did not differ among the three groups. These

data suggest that there were no differences in either thebehavioral or the electrographic progression of kindlingbetween the three kindled groups. Therefore, any subse-quent differences between the groups cannot be attributedto differences in response to the kindling procedure.

Fig. 1. Schematic of the measurement of the hilar area of the dentate gyrus. Hilar area was outlined by a thick line using the MCID image analysis system.Hilar area was defined by the inner edge of the granule cell layer and the lines connecting the tips of the two granule cell blades to the beginning of thepyramidal cell layer of Ammon’s horn.

( )B. Adams et al.rBrain Research 804 1998 331–336 333

Fig. 2. Mean hilar area of the dentate gyrus as a function of treatmentcondition. A three-way ANOVA with subsequent post-hoc comparisonsshowed that mean hilar area was significantly increased at one week

Žpost-kindling and remained elevated at one month post-kindling p-.0.05 , but declined back to control levels at two months post-kindling

Ž .p-0.05 . Mean hilar area was not significantly different between eitherŽ .the one-week and the one-month groups p)0.05 , or between the

Ž .control and the two-month groups p)0.05 . Values represent meanhilar area expressed in mm2 "S.E.M. ) P -0.05.

Following the assigned time period, rats were perfusedwith 200 ml of 0.1 M phosphate buffered 0.85% salineŽ .PBS; pHs7.4 followed by 4% paraformaldehyde in 0.1

Ž .M phosphate buffer PB at 48C. Brains were post-fixed in4% paraformaldehyde in PB for 24 h and then stored in a

Ž .20% sucrose solution 20 g sucroser100 g 0.1 M PB for24 h at 48C for cryoprotection. Brains were subsequentlyfrozen in isopentane cooled to y408C.

Horizontal serial 30-mm sections were cut using asliding microtome and section depth was determined using

w xan atlas, The Rat Brain in Stereotaxic Coordinates 9 . Toensure that brain sections included in the data analysiswere from comparable levels, we selected six pairs of

Ž .adjacent sections 12 sectionsrbrain of the hippocampalarea at 4.1–7.1 mm ventral to bregma and 600 mm apart tobe processed for immunocytochemistry.

Sections were incubated overnight with monoclonalŽanti-GFAP clone G-A-5; Boehringer Mannheim, Laval,

.Quebec, Canada; 1:200 at 48C. After washing in 0.1 Mphosphate buffer, sections were incubated with biotinyl-

Žated anti-mouse IgG BA-2000; Dimension Laboratories,.Mississauga, Ontario, Canada; 1:200 for 1 h at room

Ž .Fig. 3. Photomicrographs 445 mm=285 mm of GFAP immunostainingin the hilar region at 200= magnification. Representative examples of

Ž .GFAP immunostained sections in a control rat A , in a rat at seven daysŽ . Ž .post-kindling B , in a rat at one month post-kindling C and in a rat at

Ž .two months post-kindling D . Note that reactive gliosis is evident in boththe one-week and one-month groups, but is not apparent in either thecontrol or the two-month groups.

temperature. After washing in PB, sections were thenŽincubated with Vectastain ABC reagent Vector Laborato-

.ries for 45 min at room temperature. Finally, after wash-

( )B. Adams et al.rBrain Research 804 1998 331–336334

Ž . Ž . Ž .Fig. 4. MCID images of GFAP immunostained hilar fields 200 mm=160 mm before A and after B density thresholding, as indicated by the bluecolour overlay. Note that after density thresholding, only blue target regions contribute to the proportional area measurement of GFAP immunostaining foreach hilar field.

( )B. Adams et al.rBrain Research 804 1998 331–336 335

ing in PB, sections were pre-incubated for 5 min with 70Ž .mg of diaminobenzidine Sigma in 100 ml of 0.1 M PB

and then incubated in this solution for 60–90 s by adding30 ml of 30% H O , until the desired staining intensity2 2

developed. To ensure comparable levels of immuno-staining, tissue from all groups was always batch-processed. Following immunostaining, tissue sections weremounted on chrom alum-coated slides. One section from

Ž .each adjacent pair of sections six sectionsrbrain was alsocounterstained with Cresyl violet for the determination ofthe hilar area. Slides were then coded and all subsequentanalyses were conducted by an observer who was unawareof the treatment of the animal to ensure objectivity in thedata analysis.

For the evaluation of the hilar area, horizontal sectionsimmunostained with GFAP and counterstained with Cresylviolet were examined at 50= magnification by creating adigitized image with a Micro Computer Imaging DeviceŽ . ŽMCID Brock University, St. Catherine’s, Ontario,

. ŽCanada attached to a light microscope Zeiss Axioskop,.Oberkochen, Germany with a high-resolution charge-cou-

Ž . Ž .ple device CCD camera MTI CCD 72 . Hilar area wasdefined by the inner edge of the granule cell layer and thelines connecting the tips of the two granule cell blades tothe beginning of the pyramidal cell layer of Ammon’s hornw x Ž .2 Fig. 1 .

Ž Ž .. Ž ŽA 4= 2=6 ANOVA group= brain hemisphere=..level was conducted to evaluate the hilar area. There was

Ž . Ža main effect for group: F 3,16 s22.22; p-0.001 Fig..2 . Post-hoc Tukey tests showed that mean hilar area was

significantly increased by approximately 47% at one weekŽ .post-kindling p-0.01 and remained elevated at one

Ž .month post-kindling p-0.01 , but declined back to con-Ž .trol levels by two months post-kindling p-0.01 . There

were no significant differences between the one-week andŽ .one-month kindled groups p)0.05 or between the con-

Ž .trol group and the two-month kindled group p-0.05 .These findings suggest that kindling-induced increases inthe hilar area are not permanent.

Fig. 3 shows representative examples of GFAP im-Ž .munostained sections in a control rat A , in rats at seven

Ž . Ž .days B , one month post-kindling C and two monthsŽ .post-kindling D . These sections clearly show increases in

glial cell size in the seven-day and one-month groups.Given that increased GFAP immunostaining is considered

w xto reflect reactive gliosis 13 , hilar GFAP immunostainingwas used to quantify reactive gliosis in the present study.Horizontal sections immunostained with GFAP were ex-amined at 400= magnification using MCID. Hilar GFAP

Žimmunostaining was evaluated in a hilar field 0.2 mm=.0.48 mm starting at the hilar end of the CA3rCA4 for

each brain section, using MCID’s target detection feature.This feature permits image components to be separatedinto valid targets and background based on the opticaldensity of the target. Glial cell bodies and processes wereregarded as valid targets in the present study. A target

acceptance criteria was established using a segmentationrange between the upper and lower density thresholds ofthe target. That is, pixels lying within the segmentationrange were regarded as valid targets, whereas pixels lyingoutside of the range were ignored as background. Thesegmentation range was set by manually decreasing thethresholding value until a blue overlay display, designatingthe thresholded area, completely occupied the glial cell

Ž .bodies and processes i.e., the target , but not the back-ground. Fig. 4 shows MCID images of GFAP immunos-

Ž . Ž .tained hilar fields before A and after B density thresh-Žolding. The mean proportional area i.e., the proportion of.the field that is occupied by the target of GFAP immuno-

staining was calculated for each hilar field per section.Ž Ž .. Ž ŽA 4= 2=6 ANOVA group= brain hemisphere=..level was conducted to evaluate the mean proportional

area of hilar GFAP immunostaining. There was a mainŽ . Ž .effect for group: F 3,16 s44.24; p-0.0001 Fig. 5 .

Post-hoc Tukey tests showed that the mean proportionalarea of GFAP immunostaining was significantly increasedby approximately 57% at seven days post-kindling, re-mained elevated at one month post-kindling, but declinedback to control levels at two months post-kindling. Therewere no differences between the seven-day and one-month

Ž .post-kindling groups p)0.05 , and there were no differ-ences between the control and the two-month post-kindling

Ž .groups p)0.05 . These findings are similar to thoseŽ .obtained by Hansen et al. 1991 , who demonstrated that

GFAP protein levels remained elevated in most limbic

Fig. 5. Mean proportional area of hilar GFAP immunostaining as afunction of treatment condition. A three-way ANOVA with subsequentpost-hoc comparisons showed that mean area GFAP immunostaining as aproportion of the total field was significantly elevated at one week

Žpost-kindling and remained elevated at one month post-kindling p-.0.05 , but decreased to control levels by two months post-kindling

Ž .p-0.05 . Mean proportional area of GFAP immunostaining was notsignificantly different between either the one-week and the one-month

Ž .groups p)0.05 , or between the control and the two-month groupsŽ .p)0.05 . ) P -0.05.

( )B. Adams et al.rBrain Research 804 1998 331–336336

structures at one week post-kindling, but declined to con-w xtrol levels at two months post-kindling 7 . These data

suggest that kindling-induced up-regulation of GFAP istransient.

The findings in the present study provide the firstevidence that kindling-induced changes in the hilar area ofthe dentate gyrus and kindling-induced changes in GFAPimmunostaining levels follow a similar time course. Inaddition, these findings provide support for the hypothesisthat kindling-induced hilar area changes are mediated bykindling-induced glial cell changes. To date, however, thesignal which leads to increased GFAP expression follow-ing kindling is unclear, and the role of reactive gliosis in

w xkindling remains to be elucidated 11 . Additional researchis required not only to define the relevance of glial cellmechanisms in the kindling process, but also to elucidatethe role that kindling-induced reactive gliosis plays inmediating kindling-induced hilar area changes in the den-tate gyrus.

Acknowledgements

This work was supported by grants from the Neuro-Ž . Žscience Network Centers of Excellence NCE R.J.R. and

. Ž .M.F. and the Medical Research Council of Canada MRCŽ .R.J.R. and M.F. . B.A. was supported by a studentshipfrom the Savoy Foundation and a supplement from theNCE.

References

w x1 B. Adams, M. Sazgar, P. Osehobo, C.E.E.M. Van der Zee, J.Diamond, M. Fahnestock, R.J. Racine, Nerve growth factor acceler-

ates seizure development, enhances mossy fiber sprouting and atten-uates seizure-induced decreases in neuronal density in the kindling

Ž . Ž .model of epilepsy, J. Neurosci. 17 14 1997 5288–5296.w x2 E.H. Bertram, E.W. Lothman, Morphometric effects of intermittent

kindled seizures and limbic status epilepticus in the dentate gyrus ofŽ .the rat, Brain Res. 603 1993 25–31.

w x3 J.E. Cavazos, T.P. Sutula, Progressive neuronal loss induced bykindling: a possible mechanism for mossy fiber synaptic reorganiza-

Ž .tion and hippocampal sclerosis, Brain Res. 527 1990 1–6.w x4 J.E. Cavazos, G. Golarai, T. Sutula, Mossy fiber sprouting reorgani-

zation induced by kindling: time course, development, progressionŽ .and permanence, J. Neurosci. 11 1991 2795–2803.

w x5 J.E. Cavazos, I. Das, T.P. Sutula, Neuronal loss induced in limbicpathways by kindling: evidence for induction of hippocampal sclero-

Ž .sis by repeated brief seizures, J. Neurosci. 14 1994 3106–3121.w x6 G. Goddard, D. McIntyre, C. Leech, A permanent change in brain

function resulting from daily electrical stimulation, Exp. Neurol. 245Ž .1969 745–761.

w x7 A. Hansen, O. Steen Jorgensen, T.G. Bolwig, D.I. Barry, Hippocam-pal kindling in the rat is associated with time-dependent increases inthe concentration of glial fibrillary acidic protein, J. Neurochem. 57Ž . Ž .5 1991 1716–1720.

w x8 M. Khurgel, R.J. Racine, G.O. Ivy, Kindling causes changes in theŽ .composition of the astrocytic skeleton, Brain Res. 592 1992 338–

342.w x9 G. Paxinos, C. Watson, The Rat Brain in Stereotaxic Coordinates,

2nd edn., Academic Press, Sydney.w x10 R.J. Racine, Modification of seizure activity by electrical stimula-

tion: II. Motor seizure, Electroencephalogr. Clin. Neurophysiol. 32Ž .1972 281–294.

w x11 O. Steward, E.R. Torre, R. Tomasulo, E. Lothman, Neuronal activityŽ .up-regulates astroglial gene expression, PNAS 88 1991 6819–6823.

w x12 T. Sutula, X.X. He, J. Cavazos, G. Scott, Synaptic reorganization inthe hippocampus induced by abnormal functional activity, Science

Ž .239 1988 1147–1150.w x13 E.R. Torre, E. Lothman, O. Steward, Glial response to neuronal

activity: GFAP mRNA and protein levels are transiently increased inŽ .the hippocampus after seizures, Brain Res. 631 1993 256–264.

w x14 Y. Watanabe, R.S. Johnson, L.S. Butler, D.K. Binder, B.M. Spiegel-man, V.M. Papaioannou, J.O. McNamara, Null mutation of c-fosimpairs structural and functional plasticities in the kindling model of

Ž .epilepsy, J. Neurosci. 16 1996 3827–3836.