- Email: [email protected]
Molecular cloning of rat growth inhibitory factor cDNA and the expression in the central nervous system
Molecular Brain Research, 19 (1993) 188-!94 © 1993 Elsevier Science Publishers B.V. All rights reserved 0169-328x/93/$06.00
188
BRESM 70629
Molecular cloning of rat growth inhibitory factor cDNA and the expression in the central nervous system Hisashi Kobayashi a, Y o k o U c h i d a c, Y a s u o Ihara d, Kazuyuki Nakajima e, Shinichi K o h s a k a e, Tadashi Miyatake a and Shoji Tsuji b a Department of Nearology, Tokyo Medical and Dental University, Tokyo (Japan), b Department of Neurology, Brain Research Institute, Niigata University, Niigata (Japan), c Department of Neuropathology, Tokyo Metropolitan Institute of Gerontology, Tokyo (Japan), d Department of Neuropathology, Institute of Brain Research, Faculty of Medicine, Universityof Tokyo, Tokyo (Japan) and e Department of Neurochemistry, National Institute of Neuroscience, Tokyo (Japan) (Accepted 16 February 1993)
Key words: Growth inhibitory factor; cDNA; Molecular cloning; Alzheimer's disease; Metaiiothionein; Development
Human growth inhibitory factor (GIF) is a new metailothionein-related molecule whose expression is markedly reduced in Alzheimer's disease (AD) brain. We have recently isolated a full-length eDNA for human GIF that has a striking homology to metallothioneins (MTs). As an initial step to better understanding of the biological functions of GIF, we have isolated a full-length cDNA for rat GIF. Comparison of the predicted amino acid sequence with that of human GIF revealed a strikingly high homology (84% identity). Rat GIF cDNA also showed a striking homology to rat MT-I and MT-I! (66% and 62% identities on amino acid sequences, respectively). All cysteine residues were conserved among rat GIF, human GIF and MTs, indicating that cysteine residues play an important function in these molecules. Compared to mammalian MTs, there were 3 bp and 12 bp insertions near the amino-terminal and carboxy-terminal regions in the rat GIF. The rat GIF mRNA was found to be expressed exclusively in the central nervous system, being expressed predominantly in rat cortical astrocytes in primary culture. Although the rat GIF mRNA expression was low during the fetal stage, a dramatic increase of the rat GIF mRNA expression was demonstrated during the postnatal period of 10-17 days. These results indicate that transcriptional regulation of the rat GIF gene is quite different from that of MTs desp;te its strikingly high homology with MTs,
INTRODUCTION
Growth inhibitory factor (GIF) has recently been identified as a new molecule whose expression is drastically decreased in brains of patients with Alzheimer's disease (AD) 26,27. Subsequent purification of the GIF revealed that the GIF is a small protein which is highly homologous to metallothioneins (MTs) 2s. Molecular cloning of a full-length eDNA for human GIF demonstrated that GIF expression is found exclusively in the nervous system and GIF mRNA levels in AD brains are drastically decreased, indicating that down-regulation of the GIF mRNA transcription is the cause of the decrease of GIF activity in AD brains 25. The GIF protein produced by E. coli harboring the GIF eDNA in a prokaryotic expression vector showed growth inhibitory activity on neonatal rat cortical neurons 25.
GIF, however, has been identified only in human brains and the presence of a homologue of human GIF in other species has not been confirmed. If the GIF homologue is present in other species, it would greatly facilitate the study of the biological functions of the GIF. As the first step to better understanding of the biological functions of the GIF, we have undertaken molecular cloning of cDNAs for rat homologue of human GIF. M A T E R I A L S AND M E T H O D S
Molecular cloning of rat GIF cDNA Initially we attempted to isolate both rat and mouse GIF cDNAs. Here, we describe molecular cloning of rat GIF cDNA using a partial mouse cDNA fragment obtained from reverse transcription of mouse brain mRNA.
Correspondence: S. Tsuji, Department of Neurology, Brain Research Institute, Niigata University, I Asahimachi, Niigata 951, Japan. Fax: (81) 25-223-3620.
189 For amplification of mouse GIF eDNA by polymerase chain reaction (PCR) 2°, a pair of unique oligonucleotide primers was designed: primer 488, 5'-ATGGATCCCGAGACCTGCCC; and primer 487, 5'-CTGGCAGCAGCTGCACTTCTC. The sense primer (488) and the antisense primer (487) were designed from the aminoterminus of human GIF (amino acid residues 1-7), and the carboxyterminus of human GIF (amino acid residues 61-68), respectively 25. Oligonucleotides were synthesized using an automated DNA synthesizer (Applied Biosystems Inc., Foster City, CA, USA) according to the manufacturer's instructions. Total RNA was extracted from mouse brains and poly A(+) RNA was isolateds. Single stranded eDNA was synthesized by reverse transcription of 1 /zg of poly(A) + RNA using oligo-dT primer and murine leukemia virus fevet'se transcriptase (Bethesda Research Laboratory, Gaithersburg, MD, USA). The eDNA (200 ng) was amplified by 30 cycles of PCR consisting of denaturation at 94°C for
rat
human
rat human
rat human
Nucleotide sequence analysis Nucleotide sequences were analyzed by the dideoxynucleotide chain terminator method using double-stranded plasmid DNA as a template!22.
61
GCAGCGCATCCGCTTGCCTGGAGGAACTAAGCTACAGTCTCTCGCGGCTGCTGGCCTGGAT :: :::: ::::: : : :: : ::: :: :::::: ::: ::
human
rat
30 s, annealing at 40°C for 30 s and extension at 72°C for 3 min.tg. The PCR products were subsequently subcloned into pBiuescriptKS(+) (pBSMGIF), and the nucleotide sequences were analyzed to confirm the authenticity of the eDNA as coding for mouse GIF. After the authenticity was confirmed, the mouse GIF eDNA insert was used as the probe for screening a rat brain cDNA library (kindly provided by Dr. H. Okayama)TM. Among two dozen eDNA clones, pcD2RGIF2, which contains the longest eDNA insert, was selected for detailed analyses.
CCCAGTTGCTTGGAGAAGCCCGTTCACCGCCTCCAGCTGCTGCTCTCCTCGAC Met ATG :::
Asp GAC :::
ATG Met
GAC Asp
Pro Glu Thr CCT GAG hCC ::: ::: ::: CCT GAG A C C pro Glu Thr
Cys TGC :::
Pro Cys C C C TGT ::: ::
Pro T h r Gly GIy S e r CCT A C T GGT GGT TCC ::: :: ::: :: :::
Cys T h r Cys T G C A C C TGC 109 ::: ::: :::
TGC Cys
C C C TGC Pro Cya
CCT TCT GGT GGC TCC T G C A C C TGC Pro S e t Oly GIy S e t Cys T h r Cys
S e t A s p Lys Cys Lys Cys Lys G l y CyS Lys Cy8 Thr A s n Cys Lys Lys T C G G A C A/hA TGC A A A T G C A A G GGC TGC A A A TGC A C G A A C TGC A A G A A G 157 :: :: ::: :: ::: :: :: ::: ::: ::: :: : ::: ::: ::: GCG GAC Ala Asp
TCC Ser
TGC /lAG T G C GAG G G A TGC /hAA TGC ACC T C C Cys Ly~ Cys Glu Gly Cys Lys Cys Thr S e r
T G C /hAG /IAG Cys Lys Lys
Set AGC
Cys TGC
Cys TGC
:::
:::
:::
:::
AGC Set
TGC Cys
TGC Cys
Ser Cys C y g Pro A l a Gly TCC TGT T G C CCT G C A GGA ::: :: ::: ::: :: : TCC TGC T G C CCT GCG GAG S e t Cys Cys Pro A l a Glu
Cys GIu Lys Cys A l a Lys A s p TGT GAG AAG T G T GCC /hAG GAC :::
:::
:::
:::
::;
TGT GAG A A G T G T GCC /IAG GAC Cyr ~lu Lys Cys A l a Lys A s p
TGT GTG Cys Val
TGC A/hA GGC GGA GAG G C A GCT GAG GCA GAA G C A GAG A AG TGC Cys Lys Gly G l y Glu A l a A l a G I u A l a Glu A l a G I u Lys Cys
Ser AGC
cys TGC
Cys Gln *** TGC CAG T G A G G A C T C C C A C A C A G C C T A T G T G A A T A G T G C T G C G T G T C C C T G G
:::
:::
:::
:::
human
AGC Set
TGC Cys
TGC Cys
CAG T G A G A A G G C A C C C C T C C G T G T G G A O C A C G T G G A G A T A G T G C C A G G T G l n ***
rat
TGGGGCGTGGCTGTTGCCCCCCTCCCTGGCTTCCTGCTCCGCCTCGTGTG!%ATA~TCCCATG
rat
:
::
:
:::
:
:
::
:
:
:
:
:
:
...... ......
G l u Lys Cys GAG A A A TGC 2 4 7
Cys Lys TGC A A A ::: :::
human
GI¥ G I u Glu G l y A l a Lys GGC G/hA GAG GGG GCC A A G :: :: ::: : ! : : : :
Ala GCC
Cys V a l TGT GTT ::: ::
rat
:
:
:
::
:
:
:
::
::
human
GGCTCAGTGCCACCTATGCCTGTGTGIUAGTGTGGCTGGTGTCCCCTTCCCCTGCTGACCTTGG
rat
(~ACAGCATG (/%)n
human
AGGAATGACAATAAATCCCATGAACAGCATG J[
_
205
:::
305 :
367 :
(A )n
_
Fig. 1. The nucleotide sequence of the rat GIF cDNA (pcD2RGIF2) and the deduced amino acid sequence. The nucleotide sequence of pcD2RGIF2 and the deduced amino acid sequence are shown in upper lines. The nucleotide sequence and the deduced amino acid sequence of human GIF (pTKGIF6) are also shown in lower lines for comparison. Colons indicate the identical nucieotides between rat GIF and human GIF cDNAs. Asterisks mark the termination codon. Polyadenylation signal is underlined. Two insertions are boxed.
190 Northern blot analysis Total RNA was extracted from rat brains by a modification of the procedure described by Chirgwin and others 4's. Ten /.tg of total RNA extracted from rat brains was electrophoresed through a 1.5% agarose-0.66 M formaldehyde gel and transferred to nylon membranes 5. Hybridizations were performed as described previously s. Relative radioactivities of the rat GIF and ~-actin were determined with Fuji Bioimage Analyzer BAS2000 using an erasable phosphor imaging plate 1.
Preparation of neurons, astrocytes, microglia and fibroblasts Neurons and fibroblasts were prepared from rat brains as described previously 2m. Astrocytes were isolated by the method of McCarthy and de Veilis12 with minor modifications. In brief, dissociated cells from l-day-old rat brains were cultured for 3-4 weeks in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS), 0.37% NaHCO3, 40 U/ml of penicillin, 40/~g/ml of streptomycin and 4.5 mg/ml of glucose. The culture bottles were shaken for 30 rain. (100 strokes/min) at 37°C and floating cells (microglia) were discarded. The culture bottles were further shaken (120 strokes/min) for 4 h at 37°C in serum-free DMEM supplemented with 10/~g/ml of insulin, 10/~g/ml transferrin, 10/~g/ml of selenium, 100 p.M putrescine, i mg/mi of bovine serum albumin (BSA), 50/~g/ml of ascorbic acid, 2 ng/mi of triiodothyronine and 200 ng/ml of hydrocortisone. After the floating cells were removed, the strongly adhered cells (astrocytes) were detached and dissociated by trypsin treatment and frozen in liquid nitrogen. After 1 week, the astrocytes were regrown to be confluent.
28S -
18S-
GIF
DNA probes To avoid cross-hybridization to MT mRNAs or genes, the unique Y non-coding sequence of the rat GIF cDNA (pcD2RGIF2) was amplified using a primer pair of 551 and 552: primer 551, 5'-ATGCAGCTGCTGCCAGTGAG; and primer 552, 5'-TGTGCATGGGATrTATI'CAC. Human /3-actin genomic DNA was kindly provided by Dr. H. Hamada of University of Tokyo~.
RESULTS AND DISCUSSION
Molecular cloning of rat GIF cDNA We thought that the GIF gene is likely to be conserved through evolution as expected for many other important genes. As we have already cloned a fulllength human GIF eDNA2s, we have devised a primer pair on the basis of human GIF eDNA, carefully avoiding highly homologous regions with MT cDNAs. By reverse transcriptase-polymerase chain reaction (RT-PCR), we have isolated a mouse GIF eDNA fragment (MGIF) whose authenticity as coding for GIF was confirmed by nucleotide sequence analysis. The authenticity of the mouse eDNA as coding for GIF was confirmed by the following findings. (i) There are two insertions, one at the amino terminus and the other at the carboxy terminus, which are at exactly the same positions in the human GIF cDNAs. (ii) The nucleotide sequence of the mouse GIF eDNA shows higher homology to human GIF, compared to the homology to MTs. Using the mouse GIF eDNA fragment (MGIF) as a probe, we screened a rat b,'ain eDNA library under stringent hybridization conditions. Two dozen eDNA clones were isolated, and the nucleotide sequence
Fig, 2. Northern blot analysis of rat brain RNA. Ten /zg of total RNA extracted from a rat brain was electrophoresed through a denaturing 1.5% agarose-0.66 M formaldehyde gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the Y-non-coding region of rat GIF eDNA using a primer pair of 551 and 552.
analyses confirmed the identify of the cDNAs as coding for rat GIF. Among the GIF eDNA clones, pcD2RGIF2, containing the longest eDNA insert was analyzed in detail. Fig. 1 shows the nucleotide sequence of pcD2RGIF2 and the deduced amino acid sequence. The pcD2 RGIF2 eDNA is 376 base pairs (bp) in length, and the open reading frame extends for 198 bp. A polyadenylation signal appears 22 bp upstream of the poly(A) tail. As Northern blotting analysis of mRNAs extracted from a rat brain shows an approximately 400 nucleotide (nt) transcript (Fig. 2), it is most likely that we have isolated a full-length rat GIF eDNA. As shown in Fig. I, the rat GIF eDNA encodes a 66-amino acid protein containing 20 cysteines, II lysines, 5 serines and no aromatic amino acids. The rat GIF eDNA shows a strikingly high homology to human GIF (85% identity on nucleotide sequence and 84% identity on amino acid sequence). Rat GIF eDNA also showed a high homology to rat MT-I and MT-II (66%
191 1
10
20
30
40
50
60
68
Fig. 3. Comparison of amino acid sequences of the rat GIF, human GIF, rat MT-i and rat MT-II. Identical amino-acid residues among rat GIF, human GIF and rat MTs are boxed. Gaps are introduced to obtain an optimal alignment. Conservative cysteines and lysines are indicated in the shaded boxes. Abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; G, Giy; K, Lys; M, Met; N, Asn; P, Pro; Q, Gin; S, Ser; T, Thr; and V, Val.
and 62% identities on amino acid sequences, respectively). All 20 cysteine residues, which are supposed to be intimately involved in metal binding 28 are conserved among rat and human GIF, and MTs (Fig. 3). The cysteine residues are present in Cys-X-X-Cys, Cys-XCys, or Cys-Cys sequences. Furthermore, lysinc residues juxtaposed to cysteines are also conserved among GIFs and MTs, the rat G!F lacks Arg, His, Ile, and Leu in addition to aromatic amino acids. These data suggest that rat GIF eDNA has structural characteristics common to the MT family. There are 3 bp and 12 bp insertions in the coding sequence at the amino-terminal and carboxy-terminal portions compared to rat MTs, respectively. The insertions have never been observed in vertebrate MTs 7. Although the insertions near the amino terminus are conserved among human, rat and mouse, the insertions near the carboxy-terminus of GIFs, are not conserved among human, mouse and rat GIFs 15'25'2s. Human and mouse GIFs have 6-amino-acid insertions (Glu-AlaAla-Glu-Ala-Glu, and Glu-Gly-Ala-Lys-Ala-Glu, respectively) and rat GIF has a 4-amino-acid insertion (Glu-Gly-Ala-Lys). The amino-terminal portions of the GIF are highly conserved among human, rat and mouse and show the most distinct diversities compared to those of MTs, which raises the possibility that the amino-terminal portion containing the 1 amino acid insertion plays an important role for the growth inhibitory activity. The present study has confirmed the presence of rat homologue of human GIF and elucidated the entire structure. Very recently, nucleotide sequences of genomic clones coding for human and mouse GIFs have been described ~5. As many pseudogenes have been known to be present in MT gene families, the present paper is the first to characterize the presence and the entire structure of GIF mRNA.
The expression of rat GIF is confined to the nervous system Tissue distributions of rat GIF expression were analyzed by Northern blot analysis using the 3' non-coding region of the rat GIF cDNA as a probe to avoid cross-hybridization to MT mRNAs. The rat GIF ex-
pression was detected only in the cerebrum and cerebellum, not in liver, kidney, spleen, lung, heart, or intestine (Fig. 4). It is interesting to note that rat GIF is not expressed in liver or kidney, which are the major sites for MT mRNA expressions. The result showed that the expression of GIF mRNA is highly specific to the nervous system, which is in striking contrast to the tissue distribution of MTs. These tissue distributions of rat GIF mRNA are quite similar compared to those of human. To determine the distribution of the rat GIF mRNA in the adult rat brain, the cDNA was hybridized to a Northern blot containing RNA from different regions of the adult rat brain (Fig. 5). The GIF transcript is expressed widely in the central nervous system.
e - ®
•
28s 18s
GIF
fl-actin Fig. 4. Tissue distribution of rat GIF mRNA. Total RNAs were extracted from various rat tissues including heart, lung, liver, spleen, kidney, intestine, cerebrum and cerebellum. Ten /tg of total RNAs was electrophoresed through a denaturing 1.5%-0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the Y-non-coding region of rat GIF eDNA.
192
.o o- .# .#
0...0,-
28s 18s
-
GIF
I
fl-actin
The developmental profile of the GIF mRNA is quite different from those of MTs, as MT mRNA expression is highest in fetal and neonatal rodent liver compared to basal levels in adult liver7. The results indicate that the transcriptional regulation of rat GIF is quite different from those of MTs despite its strikingly high homology with MTs. Moreover, the distinct developmental change as well as the distinct tissue distributions of the rat GIF mRNA expression (Figs. 4, 6 and 7A,B) strongly suggests that the GIF has physiological functions distinct from those of MTs as proposed by Uchida et al. 26'27'2s. Our present study (Fig. 6) as well as immunohistochemical observations 28 indicates that the GIF mRNA is expressed predominantly and possibly exclusively in astrocytes being highest in the molecular layer in the dentate gyrus, pyramidal cell layer in the hippocampus, and the layers 2 through 6 in the neocortex of normal human brains. Although similar developmental changes are also observed for/3-subunit of S-100, another astro-
Fig. 5. Northern blot analysis of rat GIF mRNA in various regions of an adult rat brain. Total RNAs were extracted from adult rat brain subregions including frontal cortex, occipital cortex, hippocampus, basal ganglia, thalamus, cerebellum and brain stem. Ten/zg of total RNAs was electrophoresed through a denaturing 1.5% - 0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the 3'-non-coding region of the rat GIF cDNA.
Furthermore, to determine what cells in the central nervous system express the GIF mRNA, we analyzed RNA isolated from primary cultures prepared from fetal rat brains. As shown in Fig. 6, rat astrocytes showed much stronger expression of GIF mRNA com. pared to that of neurons or microglia or fibroblasts from rat meninges. The result is concordant with our earlier immunohistochemical observations showing that immunoreactivity against anti-human GIF antibody is exclusively observed in astrocytes in human brains 28.
The developmental change of the rat GIF mRNA expres. sion To characterize t'ne expression of the rat GIF mRNA during postnatal development, RNAs were extracted from rat brains of various ages and analyzed by Northern blot hybridization using the Y-noncoding region of the rat GIF eDNA as a probe. The rat GIF mRNA was detected as early as at an embryonic stage (El8), showed a dramatic increase during postnatal development at 10-17 days, and remained at the same level through 26 months. (Fig. 7A,B).
I
28s
-
18sGIF
/ -actin Fig. 6. Northern blot analysis of rat GIF expression in rat primary cultures. Total RNAs were extracted from the primary cultures enriched in neurons, astrocytes, microglia and fibroblasts. Ten/Lg of total RNAs was electrophoresed through a denaturing 1.5% - 0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the 3'-non-coding region of rat GIF cDNA.
193 #.###
28s-18s--
GIF
/~-actin
mmm
IllO
liii--
1.0"
0.9 0.8 0.7 ,E 0.6 0,5 0.4 0.3 0.2 0.1 0 , fetal 0'D 6[) 10O 17D 280 42D It 6~A 12M 20M 26M (El8) D : days M : months Fig. 7. Developmental changes of rat GIF mRNA expression in the rat brain. A: Northern blot analysis of rat GIF mRNA at various ages. Total brain RNAs were extracted from rat at fetal (El8), 0, 6, 10, 17, 28, 42 days, and 6, 12, 20, 26 months postnatal. Ten pg of total RNAs was electrophoresed through a denaturing 1~5% - 0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the Y-non.coding region of rat GIF eDNA. The same blot was rehybridized with/3-actin genomic DNA probe. B: ratios of the radioactivities of the rat GIF, compared to those of/3-actin were measured with a Fuji Bioimage Analyzer BAS2000 using an erasable phosphor imaging plate.
cyte specific gene, the developmental change of rat GIF is slightly different from those of GFAP 9'z°'24. While only a trace amount of GFAP mRNA was detected before birth, and the GFAP mRNA undergoes a biphasic evolution with an increase between birth and day 15 (during the period of intensive astroglial proliferation) followed by a decrease in aged rat brains 24. In contrast to the GFAP mRNA, substantial amount of the rat GIF mRNA was detected even at fetal stage (El8, see Fig. 7A,B). The rat GIF mRNA showed monophasic increase at days 10-17 and the decrease of the GIF mRNA in the aged rat brains was not observed (Fig. 7A,B). The distinct profiles of the
rat GIF mRNA expressions during the development suggest that transcriptional regulation of the rat GIF gene is different from that of the GFAP gene. The results further suggest that there are subsets of astrocytes with different gene expressions of GIF and GFAP. On the basis of morphology and distributions, two major classes of astrocytes have been distinguished. While fibrous astrocytes have many glial filaments and are located mainly in white matter, protoplasmic astrocytes have fewer glial filaments and are found mainly in gray matter ~6. With use of a monoclonal antibody (A2B5) and tetanus toxin b'~ading, two types of astrocytes (type 1 and type 2) have been defined. Type 1 astrocytes are found in cultures of both developing white and gray matter a,,zd type 2 astrocytes are found in cultures of developing white matter, not of gray matter. Whereas type 1 astrocytes develop before birth from a type 1 precursor cell, type 2 astrocytes develop after the first postnatal week 14'~8. The majority of astrocytes in gray matter have a type 1 phenotype, whereas the majority of astrocytes in white matter have a type 2 phenotype 13'!7. The immunohistochemical observation that anti-human GIF antibody stains astrocytes in layers 2 through 6, not astrocytes in white matter raises the possibility that the GIF expression is present predominantly in type 1 or protoplasmic astrocytes. Immunohistochemical studies on the proliferation of reactive astrocytes following brain injury also indicates the presence of two types of reactive astrocytes. Astrocytes doubly labeled with [3H]thymidine and GFAP were found only in the molecular layer of the ipsilateral cortex and the white matter adjacent to the lesion. By contrast, none of the astrocytes in the layers 2 through 6 were double labeled 2a. These data suggest that there are also subsets of reactive astrocytes with distinct regional localizations and gene expressions. The GIF expression, therefore, may also be useful for delineating various astrocytes. The GIF molecule was initially discovered as a molecule responsible for an apparently increased neurotrophic activities of Alzheimer's brain ~'2a. The fact that the GIF expression is found exclusively in astrocytes of gray matter of normal human brains and the GIF expression in the astrocytes of the gray matter is dramatically decreased in Alzheimer's disease suggests that the astrocytes closely associated with neuronal perikarya and dendrites must exert significant influence on neurons. Although the understanding of the mechanisms of the growth inhibitory activity and other potential functions of GIF is still preliminary, the availability of a full-length rat eDNA for rat GIF should make it possible to investigate its functions in detail.
194 Acknowledgments. This work was supported in part by a Grant-inAid for Scientific Research (B) and Grant-in-Aid for Scientific Research on Prior,.'ty Areas No. 04268102 from the Ministry of Education, Science and Culture of Japan.
15 16
REFERENCES 17 1 Amemiya, Y. and Miyahara, J., Imaging plate illuminates many fields, Nature, 336 (1988) 89-90. 2 Chen, E.Y. and Seeburg, P.H., Supercoil sequencing: a fast and simple method for sequencing plasmid DNA, DNA, 4 (1985) 165-170. 3 Chen, C. and Okayama, H., High-efficiency transformation of mammalian cells by plasmid DNA, Mol. Cell. BioL, 7 (1987) 2745-2752. 4 Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J., Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease, Biochemistry, 18 (1979)5294-5299. 5 Fourney, R.M., Miyakoshi, J., Day, R.S. III and Paterson, M.C., Northern blotting: efficient RNA staining and transfer, Focus, 10 (1988) 5-7. 6 Hamada, H., Petrino, M.G. and Kakunaga, T., A novel repeated element with Z-DNA-forming potential is widely found in evolutionarily diverse eukaryotic genomes, Proc. Natl. Acad. ScL USA, 79 (1982) 6465-6469. 7 Hamer, D.H., Metallothionein, Annu. Rec. Biochem., 55 (1986) 913-951. 8 Kobayashi, H., Sakimura, K., Kuwano, R., Sato, S., Ikuta, F., Takahashi, Y., Miyatake, T. and Tsuji, S., Stability of messenger RNA in postmortem human brains and construction of human brain cDNA libraries, J. Mol. NeuroscL, 2 (1990) 29-34. 9 Kuwano, R., Usui, H., Maeda, T., Fukui, T., Yamanari, N., Ohtsuka, E., lkehara, M. and Takahashi, Y., Molecular cloning and the complete nucleotide sequence of eDNA to mRNA for S-100 protein of rat brain, Nucleic Acids Res., 12 (1984) 7455-7465. 10 Landry, C.F., Ivy, G.O., Dunn, R.J., Marks, A. and Brown, I.R,, Expression of the gene encoding the/3-subunit of S-100 protein in th: developing rat brain analyzed by in situ hybridization, MoL Bra~ Res., 6 (1989) 251-262. 11 Maniatis, T., Fritsh, E.F. and Sambrook, J., Molecular Cloning, Cold Spring Harbor Laboratory, NY, 1982. 12 McCarthy, K.D. and De Vellis, J, Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell. Biol., 85 (1980) 890-902. 13 Miller, R.H., and Raft, M.C., Fibrous and protoplasmic astrocytes are biochemically and developmentally distinct, J. Neurosci., 4 (1984) 585-592. 14 Miller, R.H., David, S., Patel, R., Abney, E.R. and Raft, M.C., A quantitative immunohistochemical study of macroglial cell devel-
18 19 20
21
22 23
24
25
26
27
28
opment in the rat optic nerve: in vivo evidence for two distinct astrocyte lineages, Det,. Biol., 111 (1985)35-41. Palmiter, R.D., Findley, S.D., Whitmore, T.E. and Durnam, D.M., MT-III, a brain-specil~c member of the metallothionein gene family, Proc. Natl. Acad. Sci. USA, 89 (1992) 6333-6337. Peters, A., Palay, S.L. and Webster, H. de F., The fine structure of the Nerrous System: The Neurons and Supporting Cells, Saunders, Philadelphia, 1976, pp. 233-244. Raft, M.C., Abney, E.R., Cohen, J., Lindsay, R. and Noble, M., Two types of astrocytes in cultures of developing and growth characteristics, J. Neurosci., 3 (1983) 1289-1300. Raft, M.C., Abney, E.R. and Miller, R.H., Two glial cell lineages diverge prenatally in rat optic nerve, Dev. Biol., 106 (1984) 53-60. Rappolee, D.A., Mark, D., Banda, M.J. and Werb, Z., Wound macrophages express TGF-alpha and other growth factors in vivo: analysis by mRNA phenotyping, Science, 241 (1988) 708-712. Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullins, K.B. and Erlich, H.A., Primer directed enzymatic amplification of DNA with a thermostable DNA polymerase, Science, 239 (1988) 487-491. Saitoh, S., lijima, N., lkeda, M., Nakajima, K., Kimura, M., Katsuki, M., Mori, T. and Kohsaka, S., De novo production of alpha 2-macroglobulin in cultured astroglia from rat brain, Mol. Brain Res., 12 (1992) 155-161. Sanger, F., Niclen, S. and Coulson, A.R., DNA sequencing with chainterminating inhibitors, Proc. Natl. Acad. Sci. USA, 74 (1977) 5463-5467. Takamiya, Y., Kohsaka, S., Toya, S., Otani, M. and Tsukada, Y., Immunohistochemical studies on the proliferation of reactive astrocytes and the expression of cytoskeletal proteins following brain injury in rats, Dec. Brain Res., 38 (1988) 201-210. Tardy, M., Fages, C., Rioi, H., Leprince, G., Rataboul, P., Charriere-Bertrand, C. and Nunez, J., Developmental expression of the glial fibrillary acidic protein mRNA in the central nervous system and in cultured astrocytes, J. Neurochem., 52 (1989) 162167. Tsuji, S., Kobayashi, H., Uchida, Y., lhara, Y. and Miyatake, T., Molecular cloning of human growth inhibitory factor eDNA and its down-regulation in AIzheimer's disease, EMBO J., 11 (1992) 4843 -4850. Uchida, Y., lhara, Y. and Tomonaga, M., Alzheimer's disease brain extract stimulates the survival of cerebral cortical neurons from neonatal rats, Biochem. Biophys. Res. Commun., 150 (1988) 1263- ! 267. Uchida, Y. and Tomonaga, M., Neurotrophic action of AIzheimer's disease brain extract is due to the loss of inhibitory factors for survival and neurite formation of cerebral cortical neurons, Brain Res., 481 (1989) 190-193. Uchida, Y., Takio, K., Titani, K., lhara, Y. and Tomonaga, M., The growth inhibitory factor that is deficient in the Alzheimer's disease brain is a 68 amino acid metallothionein-like protein, Neuron, 7 (1991) 337-347.
188
BRESM 70629
Molecular cloning of rat growth inhibitory factor cDNA and the expression in the central nervous system Hisashi Kobayashi a, Y o k o U c h i d a c, Y a s u o Ihara d, Kazuyuki Nakajima e, Shinichi K o h s a k a e, Tadashi Miyatake a and Shoji Tsuji b a Department of Nearology, Tokyo Medical and Dental University, Tokyo (Japan), b Department of Neurology, Brain Research Institute, Niigata University, Niigata (Japan), c Department of Neuropathology, Tokyo Metropolitan Institute of Gerontology, Tokyo (Japan), d Department of Neuropathology, Institute of Brain Research, Faculty of Medicine, Universityof Tokyo, Tokyo (Japan) and e Department of Neurochemistry, National Institute of Neuroscience, Tokyo (Japan) (Accepted 16 February 1993)
Key words: Growth inhibitory factor; cDNA; Molecular cloning; Alzheimer's disease; Metaiiothionein; Development
Human growth inhibitory factor (GIF) is a new metailothionein-related molecule whose expression is markedly reduced in Alzheimer's disease (AD) brain. We have recently isolated a full-length eDNA for human GIF that has a striking homology to metallothioneins (MTs). As an initial step to better understanding of the biological functions of GIF, we have isolated a full-length cDNA for rat GIF. Comparison of the predicted amino acid sequence with that of human GIF revealed a strikingly high homology (84% identity). Rat GIF cDNA also showed a striking homology to rat MT-I and MT-I! (66% and 62% identities on amino acid sequences, respectively). All cysteine residues were conserved among rat GIF, human GIF and MTs, indicating that cysteine residues play an important function in these molecules. Compared to mammalian MTs, there were 3 bp and 12 bp insertions near the amino-terminal and carboxy-terminal regions in the rat GIF. The rat GIF mRNA was found to be expressed exclusively in the central nervous system, being expressed predominantly in rat cortical astrocytes in primary culture. Although the rat GIF mRNA expression was low during the fetal stage, a dramatic increase of the rat GIF mRNA expression was demonstrated during the postnatal period of 10-17 days. These results indicate that transcriptional regulation of the rat GIF gene is quite different from that of MTs desp;te its strikingly high homology with MTs,
INTRODUCTION
Growth inhibitory factor (GIF) has recently been identified as a new molecule whose expression is drastically decreased in brains of patients with Alzheimer's disease (AD) 26,27. Subsequent purification of the GIF revealed that the GIF is a small protein which is highly homologous to metallothioneins (MTs) 2s. Molecular cloning of a full-length eDNA for human GIF demonstrated that GIF expression is found exclusively in the nervous system and GIF mRNA levels in AD brains are drastically decreased, indicating that down-regulation of the GIF mRNA transcription is the cause of the decrease of GIF activity in AD brains 25. The GIF protein produced by E. coli harboring the GIF eDNA in a prokaryotic expression vector showed growth inhibitory activity on neonatal rat cortical neurons 25.
GIF, however, has been identified only in human brains and the presence of a homologue of human GIF in other species has not been confirmed. If the GIF homologue is present in other species, it would greatly facilitate the study of the biological functions of the GIF. As the first step to better understanding of the biological functions of the GIF, we have undertaken molecular cloning of cDNAs for rat homologue of human GIF. M A T E R I A L S AND M E T H O D S
Molecular cloning of rat GIF cDNA Initially we attempted to isolate both rat and mouse GIF cDNAs. Here, we describe molecular cloning of rat GIF cDNA using a partial mouse cDNA fragment obtained from reverse transcription of mouse brain mRNA.
Correspondence: S. Tsuji, Department of Neurology, Brain Research Institute, Niigata University, I Asahimachi, Niigata 951, Japan. Fax: (81) 25-223-3620.
189 For amplification of mouse GIF eDNA by polymerase chain reaction (PCR) 2°, a pair of unique oligonucleotide primers was designed: primer 488, 5'-ATGGATCCCGAGACCTGCCC; and primer 487, 5'-CTGGCAGCAGCTGCACTTCTC. The sense primer (488) and the antisense primer (487) were designed from the aminoterminus of human GIF (amino acid residues 1-7), and the carboxyterminus of human GIF (amino acid residues 61-68), respectively 25. Oligonucleotides were synthesized using an automated DNA synthesizer (Applied Biosystems Inc., Foster City, CA, USA) according to the manufacturer's instructions. Total RNA was extracted from mouse brains and poly A(+) RNA was isolateds. Single stranded eDNA was synthesized by reverse transcription of 1 /zg of poly(A) + RNA using oligo-dT primer and murine leukemia virus fevet'se transcriptase (Bethesda Research Laboratory, Gaithersburg, MD, USA). The eDNA (200 ng) was amplified by 30 cycles of PCR consisting of denaturation at 94°C for
rat
human
rat human
rat human
Nucleotide sequence analysis Nucleotide sequences were analyzed by the dideoxynucleotide chain terminator method using double-stranded plasmid DNA as a template!22.
61
GCAGCGCATCCGCTTGCCTGGAGGAACTAAGCTACAGTCTCTCGCGGCTGCTGGCCTGGAT :: :::: ::::: : : :: : ::: :: :::::: ::: ::
human
rat
30 s, annealing at 40°C for 30 s and extension at 72°C for 3 min.tg. The PCR products were subsequently subcloned into pBiuescriptKS(+) (pBSMGIF), and the nucleotide sequences were analyzed to confirm the authenticity of the eDNA as coding for mouse GIF. After the authenticity was confirmed, the mouse GIF eDNA insert was used as the probe for screening a rat brain cDNA library (kindly provided by Dr. H. Okayama)TM. Among two dozen eDNA clones, pcD2RGIF2, which contains the longest eDNA insert, was selected for detailed analyses.
CCCAGTTGCTTGGAGAAGCCCGTTCACCGCCTCCAGCTGCTGCTCTCCTCGAC Met ATG :::
Asp GAC :::
ATG Met
GAC Asp
Pro Glu Thr CCT GAG hCC ::: ::: ::: CCT GAG A C C pro Glu Thr
Cys TGC :::
Pro Cys C C C TGT ::: ::
Pro T h r Gly GIy S e r CCT A C T GGT GGT TCC ::: :: ::: :: :::
Cys T h r Cys T G C A C C TGC 109 ::: ::: :::
TGC Cys
C C C TGC Pro Cya
CCT TCT GGT GGC TCC T G C A C C TGC Pro S e t Oly GIy S e t Cys T h r Cys
S e t A s p Lys Cys Lys Cys Lys G l y CyS Lys Cy8 Thr A s n Cys Lys Lys T C G G A C A/hA TGC A A A T G C A A G GGC TGC A A A TGC A C G A A C TGC A A G A A G 157 :: :: ::: :: ::: :: :: ::: ::: ::: :: : ::: ::: ::: GCG GAC Ala Asp
TCC Ser
TGC /lAG T G C GAG G G A TGC /hAA TGC ACC T C C Cys Ly~ Cys Glu Gly Cys Lys Cys Thr S e r
T G C /hAG /IAG Cys Lys Lys
Set AGC
Cys TGC
Cys TGC
:::
:::
:::
:::
AGC Set
TGC Cys
TGC Cys
Ser Cys C y g Pro A l a Gly TCC TGT T G C CCT G C A GGA ::: :: ::: ::: :: : TCC TGC T G C CCT GCG GAG S e t Cys Cys Pro A l a Glu
Cys GIu Lys Cys A l a Lys A s p TGT GAG AAG T G T GCC /hAG GAC :::
:::
:::
:::
::;
TGT GAG A A G T G T GCC /IAG GAC Cyr ~lu Lys Cys A l a Lys A s p
TGT GTG Cys Val
TGC A/hA GGC GGA GAG G C A GCT GAG GCA GAA G C A GAG A AG TGC Cys Lys Gly G l y Glu A l a A l a G I u A l a Glu A l a G I u Lys Cys
Ser AGC
cys TGC
Cys Gln *** TGC CAG T G A G G A C T C C C A C A C A G C C T A T G T G A A T A G T G C T G C G T G T C C C T G G
:::
:::
:::
:::
human
AGC Set
TGC Cys
TGC Cys
CAG T G A G A A G G C A C C C C T C C G T G T G G A O C A C G T G G A G A T A G T G C C A G G T G l n ***
rat
TGGGGCGTGGCTGTTGCCCCCCTCCCTGGCTTCCTGCTCCGCCTCGTGTG!%ATA~TCCCATG
rat
:
::
:
:::
:
:
::
:
:
:
:
:
:
...... ......
G l u Lys Cys GAG A A A TGC 2 4 7
Cys Lys TGC A A A ::: :::
human
GI¥ G I u Glu G l y A l a Lys GGC G/hA GAG GGG GCC A A G :: :: ::: : ! : : : :
Ala GCC
Cys V a l TGT GTT ::: ::
rat
:
:
:
::
:
:
:
::
::
human
GGCTCAGTGCCACCTATGCCTGTGTGIUAGTGTGGCTGGTGTCCCCTTCCCCTGCTGACCTTGG
rat
(~ACAGCATG (/%)n
human
AGGAATGACAATAAATCCCATGAACAGCATG J[
_
205
:::
305 :
367 :
(A )n
_
Fig. 1. The nucleotide sequence of the rat GIF cDNA (pcD2RGIF2) and the deduced amino acid sequence. The nucleotide sequence of pcD2RGIF2 and the deduced amino acid sequence are shown in upper lines. The nucleotide sequence and the deduced amino acid sequence of human GIF (pTKGIF6) are also shown in lower lines for comparison. Colons indicate the identical nucieotides between rat GIF and human GIF cDNAs. Asterisks mark the termination codon. Polyadenylation signal is underlined. Two insertions are boxed.
190 Northern blot analysis Total RNA was extracted from rat brains by a modification of the procedure described by Chirgwin and others 4's. Ten /.tg of total RNA extracted from rat brains was electrophoresed through a 1.5% agarose-0.66 M formaldehyde gel and transferred to nylon membranes 5. Hybridizations were performed as described previously s. Relative radioactivities of the rat GIF and ~-actin were determined with Fuji Bioimage Analyzer BAS2000 using an erasable phosphor imaging plate 1.
Preparation of neurons, astrocytes, microglia and fibroblasts Neurons and fibroblasts were prepared from rat brains as described previously 2m. Astrocytes were isolated by the method of McCarthy and de Veilis12 with minor modifications. In brief, dissociated cells from l-day-old rat brains were cultured for 3-4 weeks in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS), 0.37% NaHCO3, 40 U/ml of penicillin, 40/~g/ml of streptomycin and 4.5 mg/ml of glucose. The culture bottles were shaken for 30 rain. (100 strokes/min) at 37°C and floating cells (microglia) were discarded. The culture bottles were further shaken (120 strokes/min) for 4 h at 37°C in serum-free DMEM supplemented with 10/~g/ml of insulin, 10/~g/ml transferrin, 10/~g/ml of selenium, 100 p.M putrescine, i mg/mi of bovine serum albumin (BSA), 50/~g/ml of ascorbic acid, 2 ng/mi of triiodothyronine and 200 ng/ml of hydrocortisone. After the floating cells were removed, the strongly adhered cells (astrocytes) were detached and dissociated by trypsin treatment and frozen in liquid nitrogen. After 1 week, the astrocytes were regrown to be confluent.
28S -
18S-
GIF
DNA probes To avoid cross-hybridization to MT mRNAs or genes, the unique Y non-coding sequence of the rat GIF cDNA (pcD2RGIF2) was amplified using a primer pair of 551 and 552: primer 551, 5'-ATGCAGCTGCTGCCAGTGAG; and primer 552, 5'-TGTGCATGGGATrTATI'CAC. Human /3-actin genomic DNA was kindly provided by Dr. H. Hamada of University of Tokyo~.
RESULTS AND DISCUSSION
Molecular cloning of rat GIF cDNA We thought that the GIF gene is likely to be conserved through evolution as expected for many other important genes. As we have already cloned a fulllength human GIF eDNA2s, we have devised a primer pair on the basis of human GIF eDNA, carefully avoiding highly homologous regions with MT cDNAs. By reverse transcriptase-polymerase chain reaction (RT-PCR), we have isolated a mouse GIF eDNA fragment (MGIF) whose authenticity as coding for GIF was confirmed by nucleotide sequence analysis. The authenticity of the mouse eDNA as coding for GIF was confirmed by the following findings. (i) There are two insertions, one at the amino terminus and the other at the carboxy terminus, which are at exactly the same positions in the human GIF cDNAs. (ii) The nucleotide sequence of the mouse GIF eDNA shows higher homology to human GIF, compared to the homology to MTs. Using the mouse GIF eDNA fragment (MGIF) as a probe, we screened a rat b,'ain eDNA library under stringent hybridization conditions. Two dozen eDNA clones were isolated, and the nucleotide sequence
Fig, 2. Northern blot analysis of rat brain RNA. Ten /zg of total RNA extracted from a rat brain was electrophoresed through a denaturing 1.5% agarose-0.66 M formaldehyde gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the Y-non-coding region of rat GIF eDNA using a primer pair of 551 and 552.
analyses confirmed the identify of the cDNAs as coding for rat GIF. Among the GIF eDNA clones, pcD2RGIF2, containing the longest eDNA insert was analyzed in detail. Fig. 1 shows the nucleotide sequence of pcD2RGIF2 and the deduced amino acid sequence. The pcD2 RGIF2 eDNA is 376 base pairs (bp) in length, and the open reading frame extends for 198 bp. A polyadenylation signal appears 22 bp upstream of the poly(A) tail. As Northern blotting analysis of mRNAs extracted from a rat brain shows an approximately 400 nucleotide (nt) transcript (Fig. 2), it is most likely that we have isolated a full-length rat GIF eDNA. As shown in Fig. I, the rat GIF eDNA encodes a 66-amino acid protein containing 20 cysteines, II lysines, 5 serines and no aromatic amino acids. The rat GIF eDNA shows a strikingly high homology to human GIF (85% identity on nucleotide sequence and 84% identity on amino acid sequence). Rat GIF eDNA also showed a high homology to rat MT-I and MT-II (66%
191 1
10
20
30
40
50
60
68
Fig. 3. Comparison of amino acid sequences of the rat GIF, human GIF, rat MT-i and rat MT-II. Identical amino-acid residues among rat GIF, human GIF and rat MTs are boxed. Gaps are introduced to obtain an optimal alignment. Conservative cysteines and lysines are indicated in the shaded boxes. Abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; G, Giy; K, Lys; M, Met; N, Asn; P, Pro; Q, Gin; S, Ser; T, Thr; and V, Val.
and 62% identities on amino acid sequences, respectively). All 20 cysteine residues, which are supposed to be intimately involved in metal binding 28 are conserved among rat and human GIF, and MTs (Fig. 3). The cysteine residues are present in Cys-X-X-Cys, Cys-XCys, or Cys-Cys sequences. Furthermore, lysinc residues juxtaposed to cysteines are also conserved among GIFs and MTs, the rat G!F lacks Arg, His, Ile, and Leu in addition to aromatic amino acids. These data suggest that rat GIF eDNA has structural characteristics common to the MT family. There are 3 bp and 12 bp insertions in the coding sequence at the amino-terminal and carboxy-terminal portions compared to rat MTs, respectively. The insertions have never been observed in vertebrate MTs 7. Although the insertions near the amino terminus are conserved among human, rat and mouse, the insertions near the carboxy-terminus of GIFs, are not conserved among human, mouse and rat GIFs 15'25'2s. Human and mouse GIFs have 6-amino-acid insertions (Glu-AlaAla-Glu-Ala-Glu, and Glu-Gly-Ala-Lys-Ala-Glu, respectively) and rat GIF has a 4-amino-acid insertion (Glu-Gly-Ala-Lys). The amino-terminal portions of the GIF are highly conserved among human, rat and mouse and show the most distinct diversities compared to those of MTs, which raises the possibility that the amino-terminal portion containing the 1 amino acid insertion plays an important role for the growth inhibitory activity. The present study has confirmed the presence of rat homologue of human GIF and elucidated the entire structure. Very recently, nucleotide sequences of genomic clones coding for human and mouse GIFs have been described ~5. As many pseudogenes have been known to be present in MT gene families, the present paper is the first to characterize the presence and the entire structure of GIF mRNA.
The expression of rat GIF is confined to the nervous system Tissue distributions of rat GIF expression were analyzed by Northern blot analysis using the 3' non-coding region of the rat GIF cDNA as a probe to avoid cross-hybridization to MT mRNAs. The rat GIF ex-
pression was detected only in the cerebrum and cerebellum, not in liver, kidney, spleen, lung, heart, or intestine (Fig. 4). It is interesting to note that rat GIF is not expressed in liver or kidney, which are the major sites for MT mRNA expressions. The result showed that the expression of GIF mRNA is highly specific to the nervous system, which is in striking contrast to the tissue distribution of MTs. These tissue distributions of rat GIF mRNA are quite similar compared to those of human. To determine the distribution of the rat GIF mRNA in the adult rat brain, the cDNA was hybridized to a Northern blot containing RNA from different regions of the adult rat brain (Fig. 5). The GIF transcript is expressed widely in the central nervous system.
e - ®
•
28s 18s
GIF
fl-actin Fig. 4. Tissue distribution of rat GIF mRNA. Total RNAs were extracted from various rat tissues including heart, lung, liver, spleen, kidney, intestine, cerebrum and cerebellum. Ten /tg of total RNAs was electrophoresed through a denaturing 1.5%-0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the Y-non-coding region of rat GIF eDNA.
192
.o o- .# .#
0...0,-
28s 18s
-
GIF
I
fl-actin
The developmental profile of the GIF mRNA is quite different from those of MTs, as MT mRNA expression is highest in fetal and neonatal rodent liver compared to basal levels in adult liver7. The results indicate that the transcriptional regulation of rat GIF is quite different from those of MTs despite its strikingly high homology with MTs. Moreover, the distinct developmental change as well as the distinct tissue distributions of the rat GIF mRNA expression (Figs. 4, 6 and 7A,B) strongly suggests that the GIF has physiological functions distinct from those of MTs as proposed by Uchida et al. 26'27'2s. Our present study (Fig. 6) as well as immunohistochemical observations 28 indicates that the GIF mRNA is expressed predominantly and possibly exclusively in astrocytes being highest in the molecular layer in the dentate gyrus, pyramidal cell layer in the hippocampus, and the layers 2 through 6 in the neocortex of normal human brains. Although similar developmental changes are also observed for/3-subunit of S-100, another astro-
Fig. 5. Northern blot analysis of rat GIF mRNA in various regions of an adult rat brain. Total RNAs were extracted from adult rat brain subregions including frontal cortex, occipital cortex, hippocampus, basal ganglia, thalamus, cerebellum and brain stem. Ten/zg of total RNAs was electrophoresed through a denaturing 1.5% - 0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the 3'-non-coding region of the rat GIF cDNA.
Furthermore, to determine what cells in the central nervous system express the GIF mRNA, we analyzed RNA isolated from primary cultures prepared from fetal rat brains. As shown in Fig. 6, rat astrocytes showed much stronger expression of GIF mRNA com. pared to that of neurons or microglia or fibroblasts from rat meninges. The result is concordant with our earlier immunohistochemical observations showing that immunoreactivity against anti-human GIF antibody is exclusively observed in astrocytes in human brains 28.
The developmental change of the rat GIF mRNA expres. sion To characterize t'ne expression of the rat GIF mRNA during postnatal development, RNAs were extracted from rat brains of various ages and analyzed by Northern blot hybridization using the Y-noncoding region of the rat GIF eDNA as a probe. The rat GIF mRNA was detected as early as at an embryonic stage (El8), showed a dramatic increase during postnatal development at 10-17 days, and remained at the same level through 26 months. (Fig. 7A,B).
I
28s
-
18sGIF
/ -actin Fig. 6. Northern blot analysis of rat GIF expression in rat primary cultures. Total RNAs were extracted from the primary cultures enriched in neurons, astrocytes, microglia and fibroblasts. Ten/Lg of total RNAs was electrophoresed through a denaturing 1.5% - 0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the 3'-non-coding region of rat GIF cDNA.
193 #.###
28s-18s--
GIF
/~-actin
mmm
IllO
liii--
1.0"
0.9 0.8 0.7 ,E 0.6 0,5 0.4 0.3 0.2 0.1 0 , fetal 0'D 6[) 10O 17D 280 42D It 6~A 12M 20M 26M (El8) D : days M : months Fig. 7. Developmental changes of rat GIF mRNA expression in the rat brain. A: Northern blot analysis of rat GIF mRNA at various ages. Total brain RNAs were extracted from rat at fetal (El8), 0, 6, 10, 17, 28, 42 days, and 6, 12, 20, 26 months postnatal. Ten pg of total RNAs was electrophoresed through a denaturing 1~5% - 0.66 M formaldehyde agarose gel, transferred to a nylon membrane and hybridized with a probe prepared by PCR amplification of the Y-non.coding region of rat GIF eDNA. The same blot was rehybridized with/3-actin genomic DNA probe. B: ratios of the radioactivities of the rat GIF, compared to those of/3-actin were measured with a Fuji Bioimage Analyzer BAS2000 using an erasable phosphor imaging plate.
cyte specific gene, the developmental change of rat GIF is slightly different from those of GFAP 9'z°'24. While only a trace amount of GFAP mRNA was detected before birth, and the GFAP mRNA undergoes a biphasic evolution with an increase between birth and day 15 (during the period of intensive astroglial proliferation) followed by a decrease in aged rat brains 24. In contrast to the GFAP mRNA, substantial amount of the rat GIF mRNA was detected even at fetal stage (El8, see Fig. 7A,B). The rat GIF mRNA showed monophasic increase at days 10-17 and the decrease of the GIF mRNA in the aged rat brains was not observed (Fig. 7A,B). The distinct profiles of the
rat GIF mRNA expressions during the development suggest that transcriptional regulation of the rat GIF gene is different from that of the GFAP gene. The results further suggest that there are subsets of astrocytes with different gene expressions of GIF and GFAP. On the basis of morphology and distributions, two major classes of astrocytes have been distinguished. While fibrous astrocytes have many glial filaments and are located mainly in white matter, protoplasmic astrocytes have fewer glial filaments and are found mainly in gray matter ~6. With use of a monoclonal antibody (A2B5) and tetanus toxin b'~ading, two types of astrocytes (type 1 and type 2) have been defined. Type 1 astrocytes are found in cultures of both developing white and gray matter a,,zd type 2 astrocytes are found in cultures of developing white matter, not of gray matter. Whereas type 1 astrocytes develop before birth from a type 1 precursor cell, type 2 astrocytes develop after the first postnatal week 14'~8. The majority of astrocytes in gray matter have a type 1 phenotype, whereas the majority of astrocytes in white matter have a type 2 phenotype 13'!7. The immunohistochemical observation that anti-human GIF antibody stains astrocytes in layers 2 through 6, not astrocytes in white matter raises the possibility that the GIF expression is present predominantly in type 1 or protoplasmic astrocytes. Immunohistochemical studies on the proliferation of reactive astrocytes following brain injury also indicates the presence of two types of reactive astrocytes. Astrocytes doubly labeled with [3H]thymidine and GFAP were found only in the molecular layer of the ipsilateral cortex and the white matter adjacent to the lesion. By contrast, none of the astrocytes in the layers 2 through 6 were double labeled 2a. These data suggest that there are also subsets of reactive astrocytes with distinct regional localizations and gene expressions. The GIF expression, therefore, may also be useful for delineating various astrocytes. The GIF molecule was initially discovered as a molecule responsible for an apparently increased neurotrophic activities of Alzheimer's brain ~'2a. The fact that the GIF expression is found exclusively in astrocytes of gray matter of normal human brains and the GIF expression in the astrocytes of the gray matter is dramatically decreased in Alzheimer's disease suggests that the astrocytes closely associated with neuronal perikarya and dendrites must exert significant influence on neurons. Although the understanding of the mechanisms of the growth inhibitory activity and other potential functions of GIF is still preliminary, the availability of a full-length rat eDNA for rat GIF should make it possible to investigate its functions in detail.
194 Acknowledgments. This work was supported in part by a Grant-inAid for Scientific Research (B) and Grant-in-Aid for Scientific Research on Prior,.'ty Areas No. 04268102 from the Ministry of Education, Science and Culture of Japan.
15 16
REFERENCES 17 1 Amemiya, Y. and Miyahara, J., Imaging plate illuminates many fields, Nature, 336 (1988) 89-90. 2 Chen, E.Y. and Seeburg, P.H., Supercoil sequencing: a fast and simple method for sequencing plasmid DNA, DNA, 4 (1985) 165-170. 3 Chen, C. and Okayama, H., High-efficiency transformation of mammalian cells by plasmid DNA, Mol. Cell. BioL, 7 (1987) 2745-2752. 4 Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter, W.J., Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease, Biochemistry, 18 (1979)5294-5299. 5 Fourney, R.M., Miyakoshi, J., Day, R.S. III and Paterson, M.C., Northern blotting: efficient RNA staining and transfer, Focus, 10 (1988) 5-7. 6 Hamada, H., Petrino, M.G. and Kakunaga, T., A novel repeated element with Z-DNA-forming potential is widely found in evolutionarily diverse eukaryotic genomes, Proc. Natl. Acad. ScL USA, 79 (1982) 6465-6469. 7 Hamer, D.H., Metallothionein, Annu. Rec. Biochem., 55 (1986) 913-951. 8 Kobayashi, H., Sakimura, K., Kuwano, R., Sato, S., Ikuta, F., Takahashi, Y., Miyatake, T. and Tsuji, S., Stability of messenger RNA in postmortem human brains and construction of human brain cDNA libraries, J. Mol. NeuroscL, 2 (1990) 29-34. 9 Kuwano, R., Usui, H., Maeda, T., Fukui, T., Yamanari, N., Ohtsuka, E., lkehara, M. and Takahashi, Y., Molecular cloning and the complete nucleotide sequence of eDNA to mRNA for S-100 protein of rat brain, Nucleic Acids Res., 12 (1984) 7455-7465. 10 Landry, C.F., Ivy, G.O., Dunn, R.J., Marks, A. and Brown, I.R,, Expression of the gene encoding the/3-subunit of S-100 protein in th: developing rat brain analyzed by in situ hybridization, MoL Bra~ Res., 6 (1989) 251-262. 11 Maniatis, T., Fritsh, E.F. and Sambrook, J., Molecular Cloning, Cold Spring Harbor Laboratory, NY, 1982. 12 McCarthy, K.D. and De Vellis, J, Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue, J. Cell. Biol., 85 (1980) 890-902. 13 Miller, R.H., and Raft, M.C., Fibrous and protoplasmic astrocytes are biochemically and developmentally distinct, J. Neurosci., 4 (1984) 585-592. 14 Miller, R.H., David, S., Patel, R., Abney, E.R. and Raft, M.C., A quantitative immunohistochemical study of macroglial cell devel-
18 19 20
21
22 23
24
25
26
27
28
opment in the rat optic nerve: in vivo evidence for two distinct astrocyte lineages, Det,. Biol., 111 (1985)35-41. Palmiter, R.D., Findley, S.D., Whitmore, T.E. and Durnam, D.M., MT-III, a brain-specil~c member of the metallothionein gene family, Proc. Natl. Acad. Sci. USA, 89 (1992) 6333-6337. Peters, A., Palay, S.L. and Webster, H. de F., The fine structure of the Nerrous System: The Neurons and Supporting Cells, Saunders, Philadelphia, 1976, pp. 233-244. Raft, M.C., Abney, E.R., Cohen, J., Lindsay, R. and Noble, M., Two types of astrocytes in cultures of developing and growth characteristics, J. Neurosci., 3 (1983) 1289-1300. Raft, M.C., Abney, E.R. and Miller, R.H., Two glial cell lineages diverge prenatally in rat optic nerve, Dev. Biol., 106 (1984) 53-60. Rappolee, D.A., Mark, D., Banda, M.J. and Werb, Z., Wound macrophages express TGF-alpha and other growth factors in vivo: analysis by mRNA phenotyping, Science, 241 (1988) 708-712. Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullins, K.B. and Erlich, H.A., Primer directed enzymatic amplification of DNA with a thermostable DNA polymerase, Science, 239 (1988) 487-491. Saitoh, S., lijima, N., lkeda, M., Nakajima, K., Kimura, M., Katsuki, M., Mori, T. and Kohsaka, S., De novo production of alpha 2-macroglobulin in cultured astroglia from rat brain, Mol. Brain Res., 12 (1992) 155-161. Sanger, F., Niclen, S. and Coulson, A.R., DNA sequencing with chainterminating inhibitors, Proc. Natl. Acad. Sci. USA, 74 (1977) 5463-5467. Takamiya, Y., Kohsaka, S., Toya, S., Otani, M. and Tsukada, Y., Immunohistochemical studies on the proliferation of reactive astrocytes and the expression of cytoskeletal proteins following brain injury in rats, Dec. Brain Res., 38 (1988) 201-210. Tardy, M., Fages, C., Rioi, H., Leprince, G., Rataboul, P., Charriere-Bertrand, C. and Nunez, J., Developmental expression of the glial fibrillary acidic protein mRNA in the central nervous system and in cultured astrocytes, J. Neurochem., 52 (1989) 162167. Tsuji, S., Kobayashi, H., Uchida, Y., lhara, Y. and Miyatake, T., Molecular cloning of human growth inhibitory factor eDNA and its down-regulation in AIzheimer's disease, EMBO J., 11 (1992) 4843 -4850. Uchida, Y., lhara, Y. and Tomonaga, M., Alzheimer's disease brain extract stimulates the survival of cerebral cortical neurons from neonatal rats, Biochem. Biophys. Res. Commun., 150 (1988) 1263- ! 267. Uchida, Y. and Tomonaga, M., Neurotrophic action of AIzheimer's disease brain extract is due to the loss of inhibitory factors for survival and neurite formation of cerebral cortical neurons, Brain Res., 481 (1989) 190-193. Uchida, Y., Takio, K., Titani, K., lhara, Y. and Tomonaga, M., The growth inhibitory factor that is deficient in the Alzheimer's disease brain is a 68 amino acid metallothionein-like protein, Neuron, 7 (1991) 337-347.