- Email: [email protected]
Developmental changes of heat shock protein 73 in human brain
DEVELOPMENTAL BRAIN RESEARCH Developmental Brain Research 86 (1999 MO-186
Research
Developmental
report
changes of heat shock protein 73 in human brain
Mitsuhiro
Kato
&mrtmentof MentalRctardorinn ami Rbth
&@I
, Masashi
??
Mizuguchi,
Rmeorch, Natimd
Mih~
Sachio Takashima
ofNeumsctmce, Nattmtal Center
of Neuroiogyand Psychtatry,
Km&?&4-l-l Q,vnwh&oshi Tokyo, Japan Aaxpred 24 January1995
The expressionof heat shockprotein (HSP) 73, a constitutiveform of HSP, was evaluated immunohistologicallyin human brains,from embryosto adults.HSP 73 immunoreactivitywasfirst detectedin the embryo at 6 weeks of gestational age (GW) in the ventral horn mlls of the spinal cord and the dorsai root ganglion cells. During the fetal period, the reactivity extended crania@, hemming detectable in the cerebral pyramidal @AIs at 40 GW. GM cells in the spinal cord also showed HSP 73 immunoreactivity, from 22 GW. The time muse of the developmerrtof HSP 73 immunoreactivity was mostly consistent with the
K@twm!s: Heat
shockprotein; ImmunohIstochemistry; Development
1. lntrodtletietl
2. Mate&Is
Heat shock proteins (HSP), or stress proteins, zre produced on exposure to heat stress or to other forms of stress, such as hypoxia-ischemia, trauma, or toxic drugs. These proteins are classified into several subgroups according to molecular weight. One of these subgroups, the HSP 70 family, consists of a constitutive form, HSP 73, and an inducible form, HSP 72. Both HSP 72 and HSP 73 act as molecular chaoerones that
2.1. Hutnan hue
plays a role in cell protulinn under stress conditions, while HSP 73 is expressed constitutively and plays a physiological role in cell metabolism. Developmental changes in the amount of HSP have been investigated in Drosophila and in the mouse [3,6],but not in human subjects. Tlx ~:cser=t study was designed to demonstrate histologically that HSP 73 is expressed in the human fetal nervous system, and to evaluate the possible role of this protein in normal development and in pathological conditions.
’ conupolrdin8 author. Fax (81) (423) 46-1743 0165-3806/%/$09.50 0 1995Elsevier science B.V. AU rishts resersfd SSDI 01653806(95)00025-9
and
methods samph
For immunoblotting, the brain of an adult patient with hypoxic-ischemic encephalopathy (male, 18 years old) and that of a neonate (39 weeks gestational age) were obtained within 12 hrs of postmortem and stored at -WC. For the immunohistochemical study, twenty-five brains of normal control subjects, from 6 weeks of gestation (GW to 50 years of age (from 18 fetuses, 5 children and 2 adults), were investigated. The causes of death were spontaneous abortion, pneumothorax, and omphalocele in the fetuses and neonates; trauma, sudden infant death syndrome, and acute leukemia in the children, and dissecting aneurysm and trauma in the hvo adults, respectively. Autopsy was performed between 1 and 17 h postmortem. The brains were subjected to neuropathological examination and were found to be normal. 2.2. Antibodies A rabbit polyclonal antibody against HSP 70 was purchased from Dako (Carpentia, CA, USA). For the comparison of HSP 73 and HSP 72, a mouse mono-
clonal antibody specific for HSP 72 (Amersham International plc, Amersham, UK) was also obtained. The specificity of these antibodies was confirmed by immunoblotting of homogenates of human brains. For protein extraction, the samples were weighed, rapidly thawed in 10 volumes gf B !xffe: consistmg of 100 mM Tris-HCl (pH 7.61, 1 mM EDTA, 1 mM pheny:methylsulfonyl fluoride, and 1% Triton X-100, and then homogenized in a glass Dounce homogenizer. After being shaken on ice for 30 min, the extracts were clarified by centrifugation at 10,000 X g for 30 min at 4” C. Protein assay was carried out by the method of Bradford, arter which the sample was incubated for 2 min at 95°C in Laemmli’s buffer, conststmg of 1% sodium dodecyl sulfate, 10 mM Tris-HCl, 10% glycerol, 1% mercaptoethanol, and 0.01% Bromophenol blue. The protein samples were separated on a 10% SDS-polyac@amide gel (30 @g/lane) and then electrophoretically transferred to a polyvinylidene difluoride membrane Qmmobilon, Millipore, Bedford, MA, USA). The ‘iembrane was blocked overnight at 4°C with phosphatebuffered saline (PBS) containing 8% skim milk and then incubated for 1 h at room temperature with the primary antibodies diluted 1:lOOO. The blot was then successively incubated with biotinylated goat anti-rabbit or horse anti-mouse IgG (Vector, diluted 1:SOOl and peroxidase-conjugated avidin-biotin complex (Vectastain Elite ABC kit, Vector Labs, Burlingame, CA. USA), for 30 min in both cases, at room temperature. After each step, the blot was washed extensively with PBS containing 0.1% tieen 20. Color development was performed with a mixture of 0.02% diaminobenzidine hydrochloride (Dojin, Osaka, Japan) and 0.006% hydrogen peroxide in 50 mM Tris HCl, pH 7.6.
J
10
20
Jo
40
kw
(GW)
Rg. 2. Development of immuncreactivity for HSP 73. 0 weakly pmmve. R strongly positive; GW, gestational weelu.
2.3. Immunoh~tochemistry Specimens of the spinal cord LSC), dorsal root ganglia (DRG), medulla oblongata, pans, midbrain cerebellum, basal ganglia, thalamus, frontal lobe, and hippocampus were taken after formalin fixation. emhedded in paraffin, and cut into Qpm-thick sections. The s::ciions were deparaffinized and then subjected to microwave irradiation for 9 minutes at 90” C and rinsed in 0.1 M phosphate-buffered saline (PBS). They were then incubated successively in 10% normal goat serum (room temperature, 10 min), the anti-HSP 70 antii (diluted 1500, room temperature, 30 min), biotinylated goat anti-rabbit IgG (diluted 1:5OCl.room temperature. 30 min). and peroxidase-conjugated streptavidii (Nichirei, Tokyo, Japan) (room temperature, 10 min). and then visualized with diaminobenxidine hydrochloride (room temperature, 5 mink The sections were then counterstained with hematoxylin. In negative control experiments, the primary antibody was omitted and replaced with normal rabbit serum.
3. Results
abed
3.1. An&&.
43
Fig. 1. lmmunoblotting of human brain homogenates with anti-HSP 70 and anti-HSP 72 antibodies. Detergent extracts of the cerebellum of a neonate (a), cerebrum of an adult (b and c), 2nd ccrebeihtm of an adult Cd) were blotted and immunostained with anti-HSP 70 (b) and anti-HSP 72 (a, c and d) antibodies. The arrow indicate the band specificallydetected in the brain extracts.The migratton postCoos of molecular weight standards, in kilodaltons, arc indicated on the tight.
z;z++-,- -,
As shown in Fig. 1, a band at 73 kDa was detected for the blots stained with the anti-HSP 70 antibody, while a 72-kDa band was detected for the blots stained with the anti-HSP 72. These results indicated that the anti-HSP 70 antibody recognized HSP 73, but not HSP 72, under the experimental conditions employed. 3.2. Immunohistochtmiwy The cytoplasm of neurons and glial cells showed HSP 73 immunoreactivity of varying intensity, depending on the brain region and age. The developmental changes in HSP 73-positive neurons are summarized in Fig. 2. The intensity of immunoStainitIg was not influ-
Pii 3. DRG cells cxbiit scattered weak immonorcactivityfor HSP 73 at 6 gestational week (GW) (A). No significant staining is detected in negativeconho sectiom :S). Tbe pbo:ographwas taken with a blue filter. Bar - 25 rat.
enced by tbe postmortem delay in fixation or by the cause of death. (Tables 1 and 2). In the SC and DRG, HSP 73 immtmoreactivity was detectable in the neurons as early as 6 GW (Fig. 3A), and in the glial cells after 22 GW (Fig. SA). In the brabt stem, the neurons of the nuclei of the tegmenturn,,such as the dorsal motor nucleus of the vagus, the raphe nucleus, and the locus cendeus, showed positivity from 14 GW (Fig. 4A), i.e., much earlier than the neurons of the pontine nucleus, in which immuno-
reactivity appeared at 33 GW. In contrast, giial cells in the pontine nucleus showed immunoreactivity after 27 GW (Fig. 5B). In the cerebellum, Purkinje cells were weakly immunoreacti-re after 19 GW and glial cells were strongly immunoreactive after 27 GW (Fig. 50. In the cerebrum, large neurons of the globus pallidus and glial cells showed immunoreactivity after 27 GW, pyramidal neurons of the hippocampal CA2 showed immunoreactivity after 32 GW (Fig. 6A,B), ghal cells of the cerebral cortex after 39 GW, and pyramidal neu-
B
Fu 4. Somataof neorons of tbz doraalnockw of tbe vagusshow HSP 73 immonoreactivityat 17 GW (A). No significant staining is detected in oegatiw coatrol wtions (B). Tbc pitoto6rapkwas taken with a blue filter. flat - 25 pm.
Table 1 Tissuesexaminedin fetal cases
-
NO.
GW
CRL
Postmortem
1 2 3 4 5 6 7 8 9 10
5 7 8 10 12 14 17 19 21 22
12 18 26 65 80 100
?h ih 4h < 12h 6h ‘,b il2h < 1Zh 5h 4h
11
24
12 13 14 15 16 17 18
26 27 32 33 34 39 40
GW, gestationalwe&
Sagrltalwhole body(DRG, SC, Cerebrum) Saglttal wholebody(DRG. SC) Cerebrum POIIS Pans. Cerebrum DRG. SC. Pans, Cerebrum Medulla. Pans,Cbll, Cerebrum Pans. cbll. Cerebrum Medulla. Pans, Cbll, Cerebrum 210 DRG. SC. Medulla. Pax, Cbll, Cerebrunl Sh DRG. SC 6h DRG. SC 2.50 6h SC. Pans. CbU, Cerebrum < 12h Cerebrum 5h Pans. GUI. Cerebrum Rh Msdulla. Pans, CtAl Sh DRG. SC, Pans,Cl-Al,Cc&rum 17 h Pans. Crreorum ---_. -_CRL, crown-nm’p length (mm): DRG. corsal rapt ganghon;SC. spmal cord
rons of the cerebral cortex after 40 GW (Fig. 7A,Bi. In the postnatal period, HSP 73 immunoreactivity was always positive in the neurons and glial cells, but the
mtensity varied with the case. Ependymal cells were consistently positive for anti-HSP 70 in fetuses to adults. In the negative control experiments, no positive labelling was observed (Figs. 3B, 4B).
4. Discussion HSP 73 is known to be important fcr normal cellular function under normal conditions, and HSP 72 for stress responses under pathological conditions. In anima1 studies, the expression of HSP 72 is induced by hypoxia-&hernia, even in neonates [l], although HSP in histologically normal human 72 ic 21% detectable brain by immunoblotting. The expreskn of HSP 73,-
on the other hand, is constitutive, but is increased by a variety of stresses, as is the expression of HSP 72. Kawagoe et al. [5] recently reported that the vulnerability of CA1 cells after transient
ixhemia
may be related
Table 2 Tissuesexaminedm postnatal cases NO.
1 2 3 4
5 6 7
Specimen
Ase
Postmortem
Specimen
21d 52d 2Y 0.. _I 12Y adult 50~
4h Sh < 12 h 7h <12h 9h Sh
Medulla, Pans, Cerebrum Medulla, Pans. Cbll, Cerebrum SC SC. Pans, Cbll, Cerebrum Cbll. Cerebrum Medulla, Pans. Cbll, Cerebrum Medulla, Pans, Cerebrum
SC, spinal cord; Cbll. cerebellum
to the imbalanced induction of HSP 72 and HSP 73 mRNA. Previous studies in animals have suggested that cognate HSP play roles in embryogenesis and in ceil differentiation [3,6]. In this study, we investigated the developmental expression of HSP 73, the constitutive form of HSP 70, in human brains. In many previous studies. HSP 72 and HSP 73 were not well differentiated. To avoid such confusion, we tkst tested the specificity of the anti-HSP 70 antii we employed by immunoblotting. Although the da% sheet provided by the manufacturer had stated that the antibody rewnizes both HSP 72 and HSP 73, an immunoblot analysis demonstrated that only the latter was detected by this antibody under the experimental conditions we employed. This conclusion was further confirmed by comparing the bands on immuooblots reacted with this anr+ody and those on immunobiots reacted with an
anti-i-lSP 72 monocional antibody that specifically reacts Ah HSP 72 hut not with HSP 73. Our study revealed that HSP 73kmunoreactivity appeared in the human brain as early as 6 G-W, increasing in intensity and extending rostraIward during fetal life. In the normal development of Drosophila, hsc4 transcripts, homologs of human hsp 73, are present in most, if not ail. cells during embryonic develop-
ment, but are abundant in cells active m endocytck, and
in cells undergoing
rapid
growth
shape [6]. In mice, hsc70 is expressed during early development and is then
down-regulated
of embryogenesis [3]. In adulthood, exhibits a high level of hsc70 gene expression [3].These findings, together with ours, suggest that HSP 73 plays an active role in the early development of the nervous system.
towards
the brain
the end
and changes in at high levels
is the only organ that
184
The present study also revealed regional differences in the thne course of munoreactivity in glial stem tegmentum, and HSP 73-immunoreactivity while in the pontine
the appearance of HSP 73 imcells. In the spinal cord, brain cerebelhun, glial cells showed much later than neurons did, nucleus they showed imm~uno-
reactivity earlier than the neurons. In the cerebral cortex, hmnunoreactivity appeared in the &al cells and the neurons during the same period. Mature astrocytes, identified by ghal fibrillary acidic protein (GFAP) immunohistochemistry, are already present at 15 GW in most parts of the normal CNS [S]. In the cerebrum,
F& 5. HSP 73 immunoreactivily of the spinal cord 0, pontinc nuckus (B), and cerebelkm Q at 27 GW. GM cells (arrows) show marked immwwrextivity in alI re&ms, while neshow varkbk imm-activiw. Ventral horn cells 0!H) are stmn&y immunoreactive (A), Pwkinjc ceils fpu) ue weakly immunoreactiw 8, while the iiiarona of the pontine nucleus (B) and the external oranular cells Cgr) of the cerebeUum (0 are not immuaonactfvc at aL A, B; Bar - 25 pm. C; Bar = 50 pm.
Fig. 6. HSP 73 immunoreactivity of the hippocampus at 32 GW. A. CA2 neuron has HSP 70 prsdve granules in its prilaryoo. CA1 neumn does not exhibit HSP 70 immunoreactivrty Bar = 25 pm
GFAP-positive astrocytes increase in number in the deep white matter from 32 GW [9]. Thus, the dcvelopment of glial HSP 73 appears to be one aspect of the differentiation and maturation of elial cells. In regard to pathological condi;ons. a ~OLUI led bv Smith reported that early maternal hype
B: ia megs
bryos, pretreatment with mild hyperthermia induces thermotolerance to the teratogens that cause neural tube defects [lo]. These fmdmgs, together with ours, suggest thar constitutive ar,d indu&le HSP play a protective role in the CNS during the embryonal and early fetal periods. Our study also demonstrated regional differences in HSP 73 immunoreactivity around 30 GW; in the hippocampus, pyramidal cells of CA2 showed earlier and
Fig. 7. HSP 73 immunoreactivity of neurons in the third layer of the cerebral cortex 21 days post parium 0, and of g&xl cclb in the white maner (B) at 39 GW. Both all types show HSP 73 immunonactivity in their soa. but Ihc lablling is more intense in the Jia! 4s. Bar - 25 rm.
186
more remarkableexpression than those of CA1 and the subiculum;in the pons, the pontine nuclei exhibited immunoreactivity much later than the other brain stem nuclei, It is noteworthythat the regions with low levels of HSP 73 at this stage are vulnerableto perinatal hypoxic-ischemic insults,which insultsoften cause a distinct lesion known as pontosubiculu necrosis. This coincidence again raises the possibility that HSP 73 protects neurons from hypoxic-ischemicinsults.In human infants,the relationshipbetween the vulnerability of neurons to various insultsand the expressionof the HSP 70 family requires further research, in which immunohiitochemical localiition of the HSP 70 family in pathologicalbrain tissue uri!Jbe important.
[ll Ferricm, D.M.. Solwano, HO., Sin, RP. and Sharp, F.R., Hypoxia-ischemia iadue+s heat shock protein-liie WSP72)immunonxtkity inneonatal nt brain, Dcu. Brain Ru., 53 (1990) 145-156. [21 Fisher, N.L and Smith, D.W., Occipital encephakxele and early gcstadonal hypcrthwmia, Pediatria, 68 (1981) 480-483.
131 Giebel, LB., LJwmiczak, B.P. and Bautz, E&F.. Developmental regulation of a constitutively expressed mouse mRNA encoding a 72-kDa heat shock-lie protein, Deu. Bid., 125 (1988) u)o-207. (41 Hartl, F., Hknian, R. and Langer, T., Molecular chaperones in rmtein fold& the art of avoiding sticky situations, Trwrds &x&m. SC& 19 (1994) 20-25. [51 Rawa8w, J., Abe, R. and Rogure, K., Regional difference of HSP70 and HSC70 heat shack mRNA inductions in rat hip pocampus after transient global irhemia, Newosc~ Mr., 153 (1993) 165-168. [61 Perk& LA., Doctor, IS., Rang, 2, Stinson, L, Penimon, N. and Cm& EA., Molecular and developmental characterization of the heat shock cagnate 4 gene of Drosophila melanogaster, Md CeU. Biof., 10 (1990) 3232-3238. ID Pket, H., Graham, J.M.J. aad Smith, D.W., Central “CTVOUS system and facial defects associated with maternal hyperthermia at four to 14 weeks’ gestation, Pediattfcs, 67 (1981) 785-789. [81 Rwsmann, U. and Gamhctti, P., Astrccytes in the developing human brain. An immunohistochemical study, ACM Neuqhzrthd @era, 70 (1986) 3a8-313. [9] T&a&ii S. and Becker, LE., Developmental changes of g&al fibriUa~ acidic protein in cerebral white matter, Arch. New& 40 m33) 14-18. [lOI Walsh. D.A., Klein, N.W., Hightower, LE. and Edwards. M.J., Heat shack and thennotolerance during early rat embryo developmcnt, Terurdogy, 36 (1987) 181-191.
Research
Developmental
report
changes of heat shock protein 73 in human brain
Mitsuhiro
Kato
&mrtmentof MentalRctardorinn ami Rbth
&@I
, Masashi
??
Mizuguchi,
Rmeorch, Natimd
Mih~
Sachio Takashima
ofNeumsctmce, Nattmtal Center
of Neuroiogyand Psychtatry,
Km&?&4-l-l Q,vnwh&oshi Tokyo, Japan Aaxpred 24 January1995
The expressionof heat shockprotein (HSP) 73, a constitutiveform of HSP, was evaluated immunohistologicallyin human brains,from embryosto adults.HSP 73 immunoreactivitywasfirst detectedin the embryo at 6 weeks of gestational age (GW) in the ventral horn mlls of the spinal cord and the dorsai root ganglion cells. During the fetal period, the reactivity extended crania@, hemming detectable in the cerebral pyramidal @AIs at 40 GW. GM cells in the spinal cord also showed HSP 73 immunoreactivity, from 22 GW. The time muse of the developmerrtof HSP 73 immunoreactivity was mostly consistent with the
K@twm!s: Heat
shockprotein; ImmunohIstochemistry; Development
1. lntrodtletietl
2. Mate&Is
Heat shock proteins (HSP), or stress proteins, zre produced on exposure to heat stress or to other forms of stress, such as hypoxia-ischemia, trauma, or toxic drugs. These proteins are classified into several subgroups according to molecular weight. One of these subgroups, the HSP 70 family, consists of a constitutive form, HSP 73, and an inducible form, HSP 72. Both HSP 72 and HSP 73 act as molecular chaoerones that
2.1. Hutnan hue
plays a role in cell protulinn under stress conditions, while HSP 73 is expressed constitutively and plays a physiological role in cell metabolism. Developmental changes in the amount of HSP have been investigated in Drosophila and in the mouse [3,6],but not in human subjects. Tlx ~:cser=t study was designed to demonstrate histologically that HSP 73 is expressed in the human fetal nervous system, and to evaluate the possible role of this protein in normal development and in pathological conditions.
’ conupolrdin8 author. Fax (81) (423) 46-1743 0165-3806/%/$09.50 0 1995Elsevier science B.V. AU rishts resersfd SSDI 01653806(95)00025-9
and
methods samph
For immunoblotting, the brain of an adult patient with hypoxic-ischemic encephalopathy (male, 18 years old) and that of a neonate (39 weeks gestational age) were obtained within 12 hrs of postmortem and stored at -WC. For the immunohistochemical study, twenty-five brains of normal control subjects, from 6 weeks of gestation (GW to 50 years of age (from 18 fetuses, 5 children and 2 adults), were investigated. The causes of death were spontaneous abortion, pneumothorax, and omphalocele in the fetuses and neonates; trauma, sudden infant death syndrome, and acute leukemia in the children, and dissecting aneurysm and trauma in the hvo adults, respectively. Autopsy was performed between 1 and 17 h postmortem. The brains were subjected to neuropathological examination and were found to be normal. 2.2. Antibodies A rabbit polyclonal antibody against HSP 70 was purchased from Dako (Carpentia, CA, USA). For the comparison of HSP 73 and HSP 72, a mouse mono-
clonal antibody specific for HSP 72 (Amersham International plc, Amersham, UK) was also obtained. The specificity of these antibodies was confirmed by immunoblotting of homogenates of human brains. For protein extraction, the samples were weighed, rapidly thawed in 10 volumes gf B !xffe: consistmg of 100 mM Tris-HCl (pH 7.61, 1 mM EDTA, 1 mM pheny:methylsulfonyl fluoride, and 1% Triton X-100, and then homogenized in a glass Dounce homogenizer. After being shaken on ice for 30 min, the extracts were clarified by centrifugation at 10,000 X g for 30 min at 4” C. Protein assay was carried out by the method of Bradford, arter which the sample was incubated for 2 min at 95°C in Laemmli’s buffer, conststmg of 1% sodium dodecyl sulfate, 10 mM Tris-HCl, 10% glycerol, 1% mercaptoethanol, and 0.01% Bromophenol blue. The protein samples were separated on a 10% SDS-polyac@amide gel (30 @g/lane) and then electrophoretically transferred to a polyvinylidene difluoride membrane Qmmobilon, Millipore, Bedford, MA, USA). The ‘iembrane was blocked overnight at 4°C with phosphatebuffered saline (PBS) containing 8% skim milk and then incubated for 1 h at room temperature with the primary antibodies diluted 1:lOOO. The blot was then successively incubated with biotinylated goat anti-rabbit or horse anti-mouse IgG (Vector, diluted 1:SOOl and peroxidase-conjugated avidin-biotin complex (Vectastain Elite ABC kit, Vector Labs, Burlingame, CA. USA), for 30 min in both cases, at room temperature. After each step, the blot was washed extensively with PBS containing 0.1% tieen 20. Color development was performed with a mixture of 0.02% diaminobenzidine hydrochloride (Dojin, Osaka, Japan) and 0.006% hydrogen peroxide in 50 mM Tris HCl, pH 7.6.
J
10
20
Jo
40
kw
(GW)
Rg. 2. Development of immuncreactivity for HSP 73. 0 weakly pmmve. R strongly positive; GW, gestational weelu.
2.3. Immunoh~tochemistry Specimens of the spinal cord LSC), dorsal root ganglia (DRG), medulla oblongata, pans, midbrain cerebellum, basal ganglia, thalamus, frontal lobe, and hippocampus were taken after formalin fixation. emhedded in paraffin, and cut into Qpm-thick sections. The s::ciions were deparaffinized and then subjected to microwave irradiation for 9 minutes at 90” C and rinsed in 0.1 M phosphate-buffered saline (PBS). They were then incubated successively in 10% normal goat serum (room temperature, 10 min), the anti-HSP 70 antii (diluted 1500, room temperature, 30 min), biotinylated goat anti-rabbit IgG (diluted 1:5OCl.room temperature. 30 min). and peroxidase-conjugated streptavidii (Nichirei, Tokyo, Japan) (room temperature, 10 min). and then visualized with diaminobenxidine hydrochloride (room temperature, 5 mink The sections were then counterstained with hematoxylin. In negative control experiments, the primary antibody was omitted and replaced with normal rabbit serum.
3. Results
abed
3.1. An&&.
43
Fig. 1. lmmunoblotting of human brain homogenates with anti-HSP 70 and anti-HSP 72 antibodies. Detergent extracts of the cerebellum of a neonate (a), cerebrum of an adult (b and c), 2nd ccrebeihtm of an adult Cd) were blotted and immunostained with anti-HSP 70 (b) and anti-HSP 72 (a, c and d) antibodies. The arrow indicate the band specificallydetected in the brain extracts.The migratton postCoos of molecular weight standards, in kilodaltons, arc indicated on the tight.
z;z++-,- -,
As shown in Fig. 1, a band at 73 kDa was detected for the blots stained with the anti-HSP 70 antibody, while a 72-kDa band was detected for the blots stained with the anti-HSP 72. These results indicated that the anti-HSP 70 antibody recognized HSP 73, but not HSP 72, under the experimental conditions employed. 3.2. Immunohistochtmiwy The cytoplasm of neurons and glial cells showed HSP 73 immunoreactivity of varying intensity, depending on the brain region and age. The developmental changes in HSP 73-positive neurons are summarized in Fig. 2. The intensity of immunoStainitIg was not influ-
Pii 3. DRG cells cxbiit scattered weak immonorcactivityfor HSP 73 at 6 gestational week (GW) (A). No significant staining is detected in negativeconho sectiom :S). Tbe pbo:ographwas taken with a blue filter. Bar - 25 rat.
enced by tbe postmortem delay in fixation or by the cause of death. (Tables 1 and 2). In the SC and DRG, HSP 73 immtmoreactivity was detectable in the neurons as early as 6 GW (Fig. 3A), and in the glial cells after 22 GW (Fig. SA). In the brabt stem, the neurons of the nuclei of the tegmenturn,,such as the dorsal motor nucleus of the vagus, the raphe nucleus, and the locus cendeus, showed positivity from 14 GW (Fig. 4A), i.e., much earlier than the neurons of the pontine nucleus, in which immuno-
reactivity appeared at 33 GW. In contrast, giial cells in the pontine nucleus showed immunoreactivity after 27 GW (Fig. 5B). In the cerebellum, Purkinje cells were weakly immunoreacti-re after 19 GW and glial cells were strongly immunoreactive after 27 GW (Fig. 50. In the cerebrum, large neurons of the globus pallidus and glial cells showed immunoreactivity after 27 GW, pyramidal neurons of the hippocampal CA2 showed immunoreactivity after 32 GW (Fig. 6A,B), ghal cells of the cerebral cortex after 39 GW, and pyramidal neu-
B
Fu 4. Somataof neorons of tbz doraalnockw of tbe vagusshow HSP 73 immonoreactivityat 17 GW (A). No significant staining is detected in oegatiw coatrol wtions (B). Tbc pitoto6rapkwas taken with a blue filter. flat - 25 pm.
Table 1 Tissuesexaminedin fetal cases
-
NO.
GW
CRL
Postmortem
1 2 3 4 5 6 7 8 9 10
5 7 8 10 12 14 17 19 21 22
12 18 26 65 80 100
?h ih 4h < 12h 6h ‘,b il2h < 1Zh 5h 4h
11
24
12 13 14 15 16 17 18
26 27 32 33 34 39 40
GW, gestationalwe&
Sagrltalwhole body(DRG, SC, Cerebrum) Saglttal wholebody(DRG. SC) Cerebrum POIIS Pans. Cerebrum DRG. SC. Pans, Cerebrum Medulla. Pans,Cbll, Cerebrum Pans. cbll. Cerebrum Medulla. Pans, Cbll, Cerebrum 210 DRG. SC. Medulla. Pax, Cbll, Cerebrunl Sh DRG. SC 6h DRG. SC 2.50 6h SC. Pans. CbU, Cerebrum < 12h Cerebrum 5h Pans. GUI. Cerebrum Rh Msdulla. Pans, CtAl Sh DRG. SC, Pans,Cl-Al,Cc&rum 17 h Pans. Crreorum ---_. -_CRL, crown-nm’p length (mm): DRG. corsal rapt ganghon;SC. spmal cord
rons of the cerebral cortex after 40 GW (Fig. 7A,Bi. In the postnatal period, HSP 73 immunoreactivity was always positive in the neurons and glial cells, but the
mtensity varied with the case. Ependymal cells were consistently positive for anti-HSP 70 in fetuses to adults. In the negative control experiments, no positive labelling was observed (Figs. 3B, 4B).
4. Discussion HSP 73 is known to be important fcr normal cellular function under normal conditions, and HSP 72 for stress responses under pathological conditions. In anima1 studies, the expression of HSP 72 is induced by hypoxia-&hernia, even in neonates [l], although HSP in histologically normal human 72 ic 21% detectable brain by immunoblotting. The expreskn of HSP 73,-
on the other hand, is constitutive, but is increased by a variety of stresses, as is the expression of HSP 72. Kawagoe et al. [5] recently reported that the vulnerability of CA1 cells after transient
ixhemia
may be related
Table 2 Tissuesexaminedm postnatal cases NO.
1 2 3 4
5 6 7
Specimen
Ase
Postmortem
Specimen
21d 52d 2Y 0.. _I 12Y adult 50~
4h Sh < 12 h 7h <12h 9h Sh
Medulla, Pans, Cerebrum Medulla, Pans. Cbll, Cerebrum SC SC. Pans, Cbll, Cerebrum Cbll. Cerebrum Medulla, Pans. Cbll, Cerebrum Medulla, Pans, Cerebrum
SC, spinal cord; Cbll. cerebellum
to the imbalanced induction of HSP 72 and HSP 73 mRNA. Previous studies in animals have suggested that cognate HSP play roles in embryogenesis and in ceil differentiation [3,6]. In this study, we investigated the developmental expression of HSP 73, the constitutive form of HSP 70, in human brains. In many previous studies. HSP 72 and HSP 73 were not well differentiated. To avoid such confusion, we tkst tested the specificity of the anti-HSP 70 antii we employed by immunoblotting. Although the da% sheet provided by the manufacturer had stated that the antibody rewnizes both HSP 72 and HSP 73, an immunoblot analysis demonstrated that only the latter was detected by this antibody under the experimental conditions we employed. This conclusion was further confirmed by comparing the bands on immuooblots reacted with this anr+ody and those on immunobiots reacted with an
anti-i-lSP 72 monocional antibody that specifically reacts Ah HSP 72 hut not with HSP 73. Our study revealed that HSP 73kmunoreactivity appeared in the human brain as early as 6 G-W, increasing in intensity and extending rostraIward during fetal life. In the normal development of Drosophila, hsc4 transcripts, homologs of human hsp 73, are present in most, if not ail. cells during embryonic develop-
ment, but are abundant in cells active m endocytck, and
in cells undergoing
rapid
growth
shape [6]. In mice, hsc70 is expressed during early development and is then
down-regulated
of embryogenesis [3]. In adulthood, exhibits a high level of hsc70 gene expression [3].These findings, together with ours, suggest that HSP 73 plays an active role in the early development of the nervous system.
towards
the brain
the end
and changes in at high levels
is the only organ that
184
The present study also revealed regional differences in the thne course of munoreactivity in glial stem tegmentum, and HSP 73-immunoreactivity while in the pontine
the appearance of HSP 73 imcells. In the spinal cord, brain cerebelhun, glial cells showed much later than neurons did, nucleus they showed imm~uno-
reactivity earlier than the neurons. In the cerebral cortex, hmnunoreactivity appeared in the &al cells and the neurons during the same period. Mature astrocytes, identified by ghal fibrillary acidic protein (GFAP) immunohistochemistry, are already present at 15 GW in most parts of the normal CNS [S]. In the cerebrum,
F& 5. HSP 73 immunoreactivily of the spinal cord 0, pontinc nuckus (B), and cerebelkm Q at 27 GW. GM cells (arrows) show marked immwwrextivity in alI re&ms, while neshow varkbk imm-activiw. Ventral horn cells 0!H) are stmn&y immunoreactive (A), Pwkinjc ceils fpu) ue weakly immunoreactiw 8, while the iiiarona of the pontine nucleus (B) and the external oranular cells Cgr) of the cerebeUum (0 are not immuaonactfvc at aL A, B; Bar - 25 pm. C; Bar = 50 pm.
Fig. 6. HSP 73 immunoreactivity of the hippocampus at 32 GW. A. CA2 neuron has HSP 70 prsdve granules in its prilaryoo. CA1 neumn does not exhibit HSP 70 immunoreactivrty Bar = 25 pm
GFAP-positive astrocytes increase in number in the deep white matter from 32 GW [9]. Thus, the dcvelopment of glial HSP 73 appears to be one aspect of the differentiation and maturation of elial cells. In regard to pathological condi;ons. a ~OLUI led bv Smith reported that early maternal hype
B: ia megs
bryos, pretreatment with mild hyperthermia induces thermotolerance to the teratogens that cause neural tube defects [lo]. These fmdmgs, together with ours, suggest thar constitutive ar,d indu&le HSP play a protective role in the CNS during the embryonal and early fetal periods. Our study also demonstrated regional differences in HSP 73 immunoreactivity around 30 GW; in the hippocampus, pyramidal cells of CA2 showed earlier and
Fig. 7. HSP 73 immunoreactivity of neurons in the third layer of the cerebral cortex 21 days post parium 0, and of g&xl cclb in the white maner (B) at 39 GW. Both all types show HSP 73 immunonactivity in their soa. but Ihc lablling is more intense in the Jia! 4s. Bar - 25 rm.
186
more remarkableexpression than those of CA1 and the subiculum;in the pons, the pontine nuclei exhibited immunoreactivity much later than the other brain stem nuclei, It is noteworthythat the regions with low levels of HSP 73 at this stage are vulnerableto perinatal hypoxic-ischemic insults,which insultsoften cause a distinct lesion known as pontosubiculu necrosis. This coincidence again raises the possibility that HSP 73 protects neurons from hypoxic-ischemicinsults.In human infants,the relationshipbetween the vulnerability of neurons to various insultsand the expressionof the HSP 70 family requires further research, in which immunohiitochemical localiition of the HSP 70 family in pathologicalbrain tissue uri!Jbe important.
[ll Ferricm, D.M.. Solwano, HO., Sin, RP. and Sharp, F.R., Hypoxia-ischemia iadue+s heat shock protein-liie WSP72)immunonxtkity inneonatal nt brain, Dcu. Brain Ru., 53 (1990) 145-156. [21 Fisher, N.L and Smith, D.W., Occipital encephakxele and early gcstadonal hypcrthwmia, Pediatria, 68 (1981) 480-483.
131 Giebel, LB., LJwmiczak, B.P. and Bautz, E&F.. Developmental regulation of a constitutively expressed mouse mRNA encoding a 72-kDa heat shock-lie protein, Deu. Bid., 125 (1988) u)o-207. (41 Hartl, F., Hknian, R. and Langer, T., Molecular chaperones in rmtein fold& the art of avoiding sticky situations, Trwrds &x&m. SC& 19 (1994) 20-25. [51 Rawa8w, J., Abe, R. and Rogure, K., Regional difference of HSP70 and HSC70 heat shack mRNA inductions in rat hip pocampus after transient global irhemia, Newosc~ Mr., 153 (1993) 165-168. [61 Perk& LA., Doctor, IS., Rang, 2, Stinson, L, Penimon, N. and Cm& EA., Molecular and developmental characterization of the heat shock cagnate 4 gene of Drosophila melanogaster, Md CeU. Biof., 10 (1990) 3232-3238. ID Pket, H., Graham, J.M.J. aad Smith, D.W., Central “CTVOUS system and facial defects associated with maternal hyperthermia at four to 14 weeks’ gestation, Pediattfcs, 67 (1981) 785-789. [81 Rwsmann, U. and Gamhctti, P., Astrccytes in the developing human brain. An immunohistochemical study, ACM Neuqhzrthd @era, 70 (1986) 3a8-313. [9] T&a&ii S. and Becker, LE., Developmental changes of g&al fibriUa~ acidic protein in cerebral white matter, Arch. New& 40 m33) 14-18. [lOI Walsh. D.A., Klein, N.W., Hightower, LE. and Edwards. M.J., Heat shack and thennotolerance during early rat embryo developmcnt, Terurdogy, 36 (1987) 181-191.