Differentiation (1986) 31 :228--235
Differentiat ion 0 Springer-Verlag 1986
Localization of immunoreactive transthyretin (prealbumin) and of transthyretin mRNA in fetal and adult rat brain Michimasa Kato", Dianne Robert Soprano, Adina Makover, Kuniyo Kato", Joseph Herbert, and DeWitt S. Goodman"" Departments of Medicine and Neurology, Columbia University, College of Physicians and Surgeons, 630 West 168 Street, New York, NY 10032, USA Abstract. We used a combination of immunohistochemical and molecular-biological techniques to investigate the localization of transthyretin (TTR) in the brains of adult and fetal rats. The immunohistochemical studies employed antibodies purified by immunosorbent affinity chrornatography, permitting the specific staining and localization of TTR using the unlabeled peroxidase-antiperoxidase method. TTR mRNA levels were measured by Northern-blot analysis of poly (A+) RNA, followed by hybridization to 32P-labeled TTR cDNA; TTR mRNA was localized in brain tissue sections by in situ hybridization. Immunoreactive TTR was found to be specifically localized in the choroid plexus epithelial cells of adult rat brain. High levels of TTR mRNA were found in poly (A+) RNA samples obtained from the choroid plexus. In addition, the specific localization of TTR mRNA in the epithelial cells of the choroid plexus was demonstrated by in situ hybridization. Neither immunoreactive TTR nor TTR mRNA were found in other regions of adult rat brains. The levels of TTR mRNA in the choroid plexus were at least 30 times higher than those observed in the adult liver. Immunoreactive TTR was observed in the brains of fetal rats on as early as the 11th day of gestation. This immunoreactive TTR was localized in the tela choroidea, the developmental forerunner of the choroid plexus. Immunoreactive TTR was also observed in the fetal choroid plexus as it began to form (14th day of gestation) as well as in the more completely developed choroid plexus (18th day of gestation). Finally, TTR mRNA was detectable in the brain of fetal rats at 16 days of gestation, and in the head of fetal rats at 14 days of gestation. Taken together, these data strongly suggest that the choroid plexus epithelium synthesizes TTR de novo and secretes it into the cerebrospinal fluid, and that TTR synthesis begins in the developing fetal rat brain before the formation of the choroid plexus (in the tela choroidea). The role that TTR plays in the brain and cerebrospinal fluid of the developing and adult rat remains to be determined.
Introduction
Plasma transthyretin (TTR), a protein previously referred to as plasma prealbumin, has been extensively studied and
* **
Present addvem: Dcpartment of Anatomy, Shinshu University, School of Medicine, Matsumoto, Nagano, 390, Japan To whom offprint requests should be sent
characterized during the past two decades. TTR is a 54,980dalton tetrameric protein composed of four identical subunits 15, 141. The amino acid sequence of human TTR was reported by Kanda et al. in 1974 [14], and the complete three-dimensional structure of human TTR was subsequently determined by high-resolution X-ray crystallography [5]. Recently, a number of laboratories have isolated and sequenced the cDNA of both human TTR [18, 23, 241 and rat TTR [7, 261. TTR has high-affinity binding sites for both thyroxine [5,10,201 and plasma retinol-binding protein [16, 20, 291, and hence plays an important role in the plasma transport of both thyroid hormone and retinol (vitamin A). Recent studies from our [22] and from other [8, 191 laboratories have demonstrated that a mutation in TTR appears to be responsible for the genetic disease, familial amyloidotic polyneuropathy (FAP). The same biochemical lesion, with a methionine residue substituted for valine at position 30 of the TTR monomer [with the resulting TTR called TTR(Met3')] has been found in FAP patients of Portuguese [22] and other [8, 191 ethnic origins. Severe autonomic, sensory, and motor neuropathy, associated with deposition of TTR(Met3')-derived amyloid in the nervous system, are prominent features of FAP. It has been suggested [22] that normal TTR may play an important, but as yet unknown, role within the nervous system, and that the mutant TTR(Met3') may be unable to fulfil this function. TTR constitutes a much larger fraction of the total protein in cerebrospinal fluid (CSF) than in serum [ I l , 311. The mechanism responsible for this relative enrichment of TTR in CSF is not understood. It is well established that CSF proteins are predominantly derived from the plasma by the controlled passage of plasma proteins through the choroid plexus. However, the relatively high concentration of TTR in CSF cannot be explained by the permeability of the blood-CSF barrier [6], or by the hydrodynamic radius of TTR as compared to other plasma proteins (e.g. albumin [9]). Recently, it has been demonstrated using immunohistochemical techniques that TTR is localized in the choroid plexus epithelium of human fetuses, neonates, and adults [2, 131. In addition, TTR mRNA has been shown to be localized in specific regions of the rat brain which contain choroid plexus epithelium [24]. These observations suggest that TTR, like certain other proteins [l],may be synthesized in the choroid plexus epithelium and secreted into the CSF.
229
We now report the results of a systematic survey of the entire rat brain, which demonstrate the specific immunohistochemical localization of TTR in the choroid plexus epithelium, as well as in the anatomic forerunner of the choroid plexus in the developing fetal rat brain. Using in situ hybridization, high levels of TTR mRNA were also found to be specifically localized in the choroid plexus epithelium. In addition, high levels of TTR mRNA were observed in the developing fetal rat brain well before the completion of the formation of the choroid plexus. Our results provide further evidence that TTR is synthesized de novo in the choroid plexus epithelium, and suggest that TTR may have a significant function in the nervous system both during fetal development and in adults. Methods Experimental animals. For histochemical studies, normal female and male rats (about 90-100 days old) were obtained from Holtzman (Madison, Wis.). These rats were mated (three or four females and one male per cage), and vaginal smears were analyzed each morning and night. The day when sperm was detected in the vaginal smear was designated as being gestational day 0. Pregnant rats were housed individually in pan-type cages with woodchip bedding, and were given a commercially available pelleted diet (Camm Research Institute, Wayne, NJ) and water ad libitum. The estimated values for the day of gestation given in the results are considered to be quite accurate, i.e., within f 0 . 5 days. At various days of gestation (from day 11 to day 21), pregnant rats were exsanguinated by drawing blood from the heart, and the fetuses were removed for the preparation of tissue sections. Adult brains were obtained from male rats weighing 250-30Og. All adult rats as well as fetuses at geslational days 1 6 2 1 were anesthetized with ether before being killed. For studies involving the measurement of level of TTR mRNA in fetal brain tissue, pregnant female rats of the Sprague-Dawley strain were obtained from Camm Breeding Laboratories on the 8th day of gestation and were then maintained as already described. The gestational age of these fetuses is guaranteed to be accurate to within k0.5 days. At various days of gestation (days 14-20), the pregnant rats were anesthetized with ether, and the fetuses were removed. The fetuses were anesthetized with ether, and their heads and brains were removed and used to prepare poly (A+) RNA. RNA from adult liver, adult choroid plexus, and adjacent brain regions was obtained from 26 adult male rats of the Sprague-Dawley strain. Adult rats weighing approximately 250 g were anesthetized with ether, and their liver and brain were removed. Rat brains were dissected in the following manner: coronal slices were placed at the levels of the optic chiasma, the median eminence, and the pontomesencephalic junction, and at the junction of the pons and medulla. The cerebellum was dissected free from the brain stem by sectioning the cerebellar peduncles. Choroid plexus epithelium was removed from the lateral ventricles, the third ventricle, and the fourth ventricle of the brain. We thus obtained tissue and prepared RNA from the choroid plexus, the frontal pole (anterior to the optic chiasma), the medulla, the anterior portion of the cerebellum, and the midbrain (lateral portions of the region between the median eminence and the pontomesencephalic junction).
Care was taken to ensure that the frontal pole, medulla, cerebellum, and midbrain tissues were not contaminated with choroid plexus tissue. Preparation and characterization of antibodies. Rabbit antibody against rat TTR and goat antibody against rabbit IgG were purified by immunosorbent affinity chromatography of IgG fractions prepared from the respective specific antisera using purified rat TTR and rabbit IgG, as described previously [15]. A variety of control experiments, described in detail elsewhere [I 51, have established the specificity of these antibodies. Tissue preparation and immunochemical staining. Whole fetuses were fixed with Perfix (Fisher Scientific, Springfield, NJ) with rotation at 4" C overnight (32-15 h). Before fixation, fetuses between the 16th and 21st day of gestation were cut in half in a sagittal plane. Adult male rats were perfused through the common carotid artery (with exit through the right atrium) with Hepes-buffered saline (10 mM Hepes buffer, pH 7.4, 122 mM NaC1,6.6 m M KC1, 1.2 mM CaCI,) for 2 min, followed by perfusion with icecold Perfix for 5 min. Perfusion was performed at a constant flow rate of 20 ml/min. The brains were then removed, each entire brain was cut sagittally into 2- to 3-mm-thick slices, and fixation was continued by immersion for 2 h at 4" C. The fixed pups and the fixed brains from the adult male rats were washed with 95% ethanol three times (8 h each at 4" C ) and embedded in paraffin according to routine procedures. Serial sections (thickness, 4-5 pm) were mounted on glass slides. In one study, some slices of fixed adult rat brains were washed for 24 h at 4" C with 100 m M phosphate buffer, pI3 7.4, and then with the same buffer containing 10% glycerol for 24 h at 4" C. The brain specimens were then dipped in OCT (Tissue-Tek) compound and frozen in acetone which had been prechilled with dry ice. Sections (thickness, 10 pm) were cut using a cryostat and mounted on glass slides. The deparaffinized sections and cryostat sections were subjected to the unlabeled immunohistochemical staining method of Sternberger et al. [25] using rabbit peroxidaseantiperoxidase (Accurate Chemical and Scientific, Westbury, NY), as described in detail previously [15]. After immunohistochemical staining, the deparaffinized sections were studied without additional processing. The cryostat sections were postfixed at room temperature for 30 min in a solution of 2% osmium tetroxide in 100 m M phosphate buffer, pH 7.4, dehydrated in graded ethanol solutions, and then embedded in Epon. Sections (thickness, 1 pm) were cut using an LKB UMIV ultramicrotome and mounted on glass slides. Control experiments were also carried out using a sample of the solution of purified antibody against TTR, which had been absorbed with pure TTR coupled to Sepharose 4B, as described previously [15]. The results obtained with the absorbed and unabsorbed solutions of antibody against TTR were then compared, with regard to the intensity of specific immunohistochemical staining, with serial sections.
RNA isolation and analysis. Total RNA was prepared from all tissue samples according to the method of Tushinski et al. [28]. RNA enriched in poly (A+) RNA was obtained by oligo(dT)-cellulose affinity chromatography as described by Aviv and Leder [3] and was then quantitated spectrophotometrically.
230
Fig. la-c. Localization of immunoreactive TTR in adult rat brain. a Control section stained with absorbed antibody against TTR showing no reaction products. gd, gyrus dentatus; am, nucleus amygdaloideus medialis ; cc, crus cerebri; cp, choroid plexus. b Immunohistochemical localization of TTR in a section close to the one shown in a. lmmunoreactive TTR was detected in the choroid plexus. Blood was completely removed from blood vessels (arrows)by perfusion. c High-magnification view of a portion of the choroid plexus of the lateral ventricle. This section (from an Epon-embedded cryostat section of tissue) shows the apical concentration of the immunohistochemical reaction product in the cytoplasm of epithelial cells. Note the absence of staining in nuclei (n) and microvilli (mu).Nuclei were weakly counterstained with diluted hematoxylin. Bar (in c), 20 pm. a, b x 36; c x 600
The size and the relative amount of TTR mRNA was analyzed using Northern blots. Poly (A +) RNA samples were denatured by heating for 15 min at 55" C in 50% (v/v) formamide, 6.5% (v/v) formaldehyde, and 1X Mops buffer [20 m M morpholinopropanesulfonic acid (pH 7.0), 5 m M Na acetate, and 1 m M EDTA], and electrophoresed in a 1% agarose gel containing 2.2 A4 formaldehyde and 1X Mops as the running buffer [17]. RNA was transferred to nitrocellulose paper essentially as described by Thomas 1271. All filters were hybridized to nick-translated [21] human TTR cDNA isolated by electroelution from polyacrylamide gels after the digestion of pHTTl with EcoRZ, as previously described [24]. Prehybridization, overnight hybridization, and subsequent washing were carried out as described previously [24]. The filters were exposed to Kodak SB5 X-ray film with intensifying screens, and the appropriate exposures of autoradiograms were quantitated by computer-assisted image analysis densitometry (Image Analysis 2000 ; Image Analysis Corporation). In situ hybridization. The in situ hybridization method employed was essentially that described by Gee and Roberts [12]. Rats were anesthetized with ketamine hydrochloride (100 mg/kg body weight), and the systemic circulation was perfused through the heart (with exit from the right atrium) with 20 ml isotonic saline, followed by 300 ml 4% paraformaldehyde in 0.1 M Na phosphate buffer, pH 7.4, at a pressure of 110-140 mm Hg. The brain was removed and placed in 15% sucrose in phosphate-buffered saline for 1 h at 4" C. The whole brain was embedded in OCT compound
and frozen in liquid nitrogen. Sections (thickness, >I0 pm) through the lateral ventricle were prepared using a cryostat, thaw mounted on gelatin-coated slides, and stored at -70" C. For hybridization, slides were removed from a freezer (- 70" C), covered immediately with 5 pg/ml proteinase K, and incubated for 10 min at room temperature. Tissue sections were washed in 0.2 x SSC (1 x SSC= 150 m M NaCl and 15 m M Na citrate, pH 7.0) and then prehybridized at 30" C for 90min in 5 x SSC, 1X Denhardt's solution (0.02% each of Ficoll, polyvinylpyrollidone, and bovine serum albumin) and 50% formamide. Following prehybridization, tissue sections were hybridized overnight at 30" C in the same buffer as that used for prehybridization, with the addition of 2 M O x lo3 cpm nick-translated [21], 3Hlabeled TTR cDNA. After hybridization, the slides were washed twice (10 min each) in 2 x SSC at room temperature and then overnight in 0.5 x SSC. The washed sections were dehydrated in a graded series of alcohols containing 0.3 A4 ammonium acetate and then dried under vacuum. The grains were visualized by autoradiography after 3 weeks of exposure. The tissue sections were then stained with hematoxylin and eosin, and examined by polarized light epiluminescence with bright-field illumination. Results Immunoreuctive TTR in adult rat bruin
A complete and systematic study of serial sections of the entire rat brain was carried out. Figure l a and b shows
231
A B C D E F
- 28s
- 18s
interesting to note that, when the brain had not been perfused with Hepes-buffered saline, immunoreactive TTR was always found in the blood vessels. This observation indicates that the concentration of TTR in adult rat plasma is high enough to be detected using our immunostaining procedures. As indicated by arrows in Fig. 1b, plasma TTR was not detected in the blood vessels of well-perfused brains. A higher-magnification micrograph of the choroid plexus of the lateral ventricle is shown in Fig. 1 c; this micrograph was obtained using a l-pm-thick section from an Epon-embedded cryostat section of brain, which was studied to demonstrate the intracellular localization of TTR. In choroid plexus epithelial cells, the immune reaction product tended to accumulate in the apical portion of the cytoplasm (ventricular side), indicating the specific localization of TTR in the apical portion of the epithelial cells. Specific immunostaining was never observed in the nuclei or microvilli (see Fig. 1 c). Epithelial-cell apical blebs were not observed in the microscopic sections studied. TTR mRNA in adult rat brain
Fig. 2. TTR mRNA in the choroid plexus epithelium; 2 pg poly ( A + ) RNA was electrophoresed in a 1% agarose/formaldehyde gel, transferred to nitrocellulosc paper, and hybridized to 32P-labeled human TTR cDNA. A , frontal pole; B, cerebellum; C, liver; D , medulla; E, choroid plexus; F, midbrain
the choroid plexus of the lateral ventricle, which is surrounded by the gyrus dentatus. nucleus amygdaloideus, and crus cerebri. Specific immunostaining was never seen in any portion of the brain when the section was stained with antibody against TTR that had been absorbed with pure TTR (Fig. 1a). However, as shown in Fig. 1b, a specific and intense immunoreaction for TTR was detected in the choroid plexus of adult rat brain when it was stained with the antibody against TTR. Specific immunostaining for TTR was not observed in any other portions of the rat brain. It is
To determine whether the immunoreactive TTR present in the choroid plexus epithelium is synthesized de novo in these cells, poly (A+) RNA was isolated from the choroid plexus and from several parts of the brain that do not contain choroid plexus. Analysis of these poly (A+) RNA samples by Northern blotting and hybridization to 32Plabeled TTR cDNA demonstrated that the choroid plexus contains a high concentration of TTR mRNA (Fig. 2). In these studies (Fig. 2; see also Fig. 6), the single band that hybridized with the TRR cDNA was of the size (approximately 0.7 kilobases) expected for TTR mRNA. The level of TTR mRNA in the choroid plexus was found to be at least 30 times greater than that in the liver, when expressed as relative amounts in poly ( A + ) RNA. TTR mRNA was not detected in any of the several other regions of the brain examined (Fig. 2), including the frontal pole, the anterior portion of the cerebellum, the medulla, and the midbrain, none of which contain choroid plexus epithelium.
Fig. 3. In situ hybridization to localize TTR mRNA in the choroid plexus epithelial cells of the adult rat brain. Photomicrograph of the choroid plexus in a section through the lateral ventricle that had been hybridized with 3H-labeled human TTR cDNA and exposed for 3 weeks. The photograph was taken using a combination of polarized light epiluminescence and bright-field illumination, so that the silver grains appear white. Note the high density of the grains over the cytoplasm of the choroid plexus epithelial cells. x 640
232
Fig. 4a-c. Immunohistochernical localization of TTR in the fetal brain at the 11th day of gestation. a A sagittal section stained with hematoxylin and eosin. 4" V , fourth ventricle; sa, subarachnoid space; tc, tela choroidea, b Immunohistochemical localization of TTR in a section close to the one shown in a. Dense reaction products are visible in the tela choroidea (arrow). TTR in cerebrospinal fluid is also present in the fourth ventricle and in the subarachnoid space. c High magnification of the tela choroidea shown in b. Note the intense immunohistochemical staining for TTR in the cytoplasm of the epithelial cells. No counterstaining. a, b x30; c x 850
In situ hybridization of TTR mRNA in adult rat choroid plexus
In order to demonstrate and define further the localization of TTR mRNA in the choroid plexus, in situ hybridization studies were performed. Adult rat brain sections containing the lateral ventricles were hybridized with 3H-labeled human TTR cDNA. Figure 3 shows the 3Hgrains visualized in the choroid plexus after 3 weeks of exposure. The grains were intensely localized in the epithelial cells of the choroid plexus. Only a very low density of grains (background level) was observed in the ependymal lining of the ventricles, the brain cells, and the tissue adjacent to and near the choroid plexus (not shown). Immunoreuctive TTR in fetal rat brain
The cerebral ventricles of fetal rats are formed between the 10th and 12th days of gestation. At this time, choroid plexuses are not yet present in any of the ventricles [4]. However, the tela choroidea, which consists of a single layer of ependymal cells of the roof plate and is the structural forerunner of the choroid plexus, is recognizable in the dorsal wall of the fourth ventricle of the fetal brain on the 11th day of gestation (Fig. 4a). Figure 4 b shows the localization of immunoreactive TTR on a section close to the section shown in Fig. 4a. Intense immunostaining for TTR was found in the tela choroidea. Specific immunostaining for TTR was also observed in the fourth ventricle and in the subarachnoid spaces; both of these spaces are known to be filled with CSF. A high-magnification micrograph of the tela choroidea is shown in Fig. 4c; intense and granu-
lated reaction products were observed throughout the cytoplasm of ependymal cells. The choroid plexuses appear in the ventricles between the 13th and 15th days of gestation [4]. As shown in Fig. 5a, some portions of the tela choroidea begin to differentiate morphologically (to form plexus) during the 14th day of gestation. The choroid plexus exhibits its typical structure by the 18th day of gestation. while the tissue volume of the tela choroidea diminishes (Fig. Sc). Figure 5b and d shows the localization of immunoreactive TTR in sections adjacent to those shown in Fig. 5a and c, respectively. Intense immunostaining for TTR was observed in the tela choroidea and developing choroid plexus at both gestational ages (14th and 18th days). Within the epithelium of the tela choroidea and choroid plexus, the intensity of the immunostaining for TTR tended to be higher in the apical portion than in the basal portion of the cells from as early as the 13th day of gestation. This gradient of staining intensity was observed throughout the development of the choroid plexus and at all later periods.
TTR mRNA in fetal rat brain To investigate whether the immunoreactive TTR observed in the tela choroidea and the choroid plexus of developing rat brains results from de novo synthesis, we analyzed poly (A +) RNA samples obtained from brains of rats at various gestational ages for the presence of TTR mRNA. Figure 6 shows that TTR mRNA is present in developing rat brain at 16, 18, and 20 days of gestation. TTR mRNA was also observed in the heads of fetuses at the 14th day of gestation
233
Fig. 5a-d. Immunohistochemical localization of TTR in fetal brains at 14 (a, b) and 18 (c, d) days of gestation. a, c Sections stained with hernatoxylin and eosin, corresponding to the adjacent serial sections shown in b and d, respectively. 4"V. fourth ventricle; cp, choroid plexus; tc, tela choroidea; sa, subarachnoid space; fL, foramen of Luschka. b, d Localization of immunoreactive TTR in the tela choroidea and choroid plexus of fetal brains at 14 (b) and 18 (d) days of gestation. Specific immune reaction products were also seen in the ventricles and in subarachnoid spaces at both gestational days. No counterstaining. x 36
(Fig. 6, lanes G and H). The relative amount of TTR mRNA present in fetal brain averaged about 40% of that found in the adult brain.
Discussion The concentration of TTR in the CSF of man is approximately 12 times greater than that which would be expected from the CSF concentration of other plasma proteins [9]. This relatively high concentration of TTR in CSF cannot be accounted for by passive diffusion of TTR from the plasma. Thus, CSF TTR must be actively concentrated or synthesized de novo in the central nervous system [30]. The present study combined immunohistochemistry and molecular biology to investigate TTR in the brain of adult and fetal rats. In the immunohistochemical study, a complete and systematic survey of serial sections of the entire adult rat brain was carried out. The results obtained clearly demonstrate that the TTR present in the rat brain is specifically localized in the choroid plexus epithelial cells. Immu-
noreactive TTR was not observed anywhere else in adult rat brains. Moreover, TTR was found to be concentrated in the apical portion of choroid plexus epithelial cells. Similar findings in humans have been reported by Aleshire et al. [2]. Thesc observations suggest that TTR is secreted by the epithelial cells into the CSF. Future studies are needed to address this issue more definitively, and to explore the possible mechanism(s) of such TTR secretion into the CSF. The apocrine secretion of protein into the CSF by the choroid epithelium has been suggested by Agnew et al. [l]. In the present study, however, we did not observe the blebcovered cells that have been suggested as being involved in such a process. In addition to finding immunoreactive TTR in the choroid plexus, we also found, by Northern-blot analysis, that a high concentration of TTR mRNA is present in t h s tissue. This finding was confirmed and extended by our in situ hybridization study, which showed the specific localization of large quantities of TTR mRNA in the epithelial cells of the choroid plexus. The finding of high levels of
234
A B C D E F G H
- 28s - 18s
6.TTRm N in developing fetal rat brains; 10 pg poly ( +) RNA from brain tissues and both 1 pg and 10 pg poly (A +) RNA from the liver were electrophoresed in a 1 % agarose/formaldehyde gel, transferred to a nitrocellulose filter, and hybridized to 32Plabeled human TTR cDNA. Lanes A-G were exposed for 3 h; lane H is an 18 h exposure of lane G. A , adult liver (10 pg); B, adult liver (1 pg); C, adult brain; D,20th-day-gestation brain; E, 18th-day-gestation brain; F, 16th-day-gestation brain; C , 14th-daygestation head (3-h exposure); H , 14th-day-gestation head (1 8-h exposure) I.,.
both immunoreactive TTR and TTR mRNA in the choroid plexus strongly suggests that the TTR in CSF is largely derived from the de novo synthesis of TTR in choroid plexus epithelial cells. Interestingly, the relative amount of TTR mRNA in poly ( A +) RNA was found to be at least 30 times greater in the choroid plexus than in the liver. Dickson et al. [qhave recently observed that TTR mRNA levels are 25 times higher in rat choroid plexus than in rat liver. Our observations and those of Dickson et al. suggest either that the transcription rate of the TTR gene is greater in the choroid plexus or that TTR m R N A is more stable in the choroid plexus. To determine which of these mechanisms functions in the choroid plexus will require measuring and comparing the transcription rates of the TTR gene in the liver and choroid plexus. We next applied comparable immunohistochemical and molecular-biological techniques to study the brains of fetal rats at different stages of gestational development. The aim of these studies was to obtain information about the developmental age at which the onset of TTR synthesis occurs in the rat brain and choroid plexus. We found that, as early as the 11th day of gestation (the earliest gestation time examined), immunoreactive TTR was localized and clearly apparent in the developmental forerunner of the choroid plexus, the tela choroidea. During subsequent fetal development (14th and 18th days of gestation), immunohistochemical staining for TTR was observed in both the tela choroidea and the choroid plexus (as the latter developed anatomically). These data suggest that the synthesis of TTR begins in the structural forerunner of the choroid plexus and continues as the choroid plexus develops. In support
of this conclusion, TTR mRNA was detected in the fetal heads at the 14th day of gestation, and in the brains of fetal rats at 16-20 days of gestation. More specific localization of the specific cells in the fetal rat brain that contain TTR mRNA will require further in situ hybridization studies. From the information at hand, however, we presume that TTR mRNA is present in the same cells that show immunohistochemical staining for TTR, i.e., choroid epithelial cells. It is interesting that the expression of the TTR gene in fetal brain begins before the morphological differentiation of the choroid plexus. Moreover, the expression of the TTR gene in the tela choroidea begins before TTR mRNA can be detected in fetal liver (unpublished observations from our laboratory). Studies directed toward the elucidation of the gene sequences involved in the control of the expression of the TTR gene in the brain and liver would clearly be of value. The function (or functions) that TTR has in the developing and mature nervous system is not known. The apparent commencement of TTR synthesis in the brain from at least the 11th day of gestation (before TTR synthesis is apparent in the liver) onward suggests that TTR may play an essential role in fetal brain development. Both retinol (vitamin A) and thyroxine are known to be essential for the normal development of the central nervous system, and TTR is known to have important roles in the plasma transport of both thyroid hormone and retinol. Thus, TTR may play a role in the metabolism of retinol and/or thyroxine in the fetal brain. In addition, however, TTR may have a different, as yet undiscovered, function in the developing and adult nervous system. This last possibility is suggested by the results of studies of patients with FAP, since the mutant, TTR(Met3'), exhibits normal binding of both thyroxine and retinol-binding protein [22], yet its presence ultimately leads to severe neuropathy. Future studies are needed to clarify the functional role(s) of TTR in the nervous system. Acknowledgements. We thank Dr. E. Zimmerman for his help and advice in the dissection of adult brain, and Dr. J. Roberts, Dr. J. Wilcox, Dr. J. Chow, and Ms, I. Lug0 for their help and advice in the in situ hybridization experiments. We thank Ms. Margo Wyatt for her expert technical assistance. This work was supported by the National Institutes of Health (grants AM 05968 and HL 21006; SCOR).
References 1. Agnew WF, Alvarez RB, Yuen TGH, Crews AK (1980) Protein synthesis and transport by the rat choroid plexus and ependyma. Cell Tissue Res 208:261-281 2. Aleshire SL, Bradley CA, Richardson LD, Par1 FF (1983) Localization of human prealbumin in choroid plexus epithelium. J Histochem Cytochem 31 :608612 3. Aviv H, Leder P (1972) Purification of biologically active globin messenger RNA by chromatography on oligothymidylic acidcellulose. Proc Natl Acad Sci USA 69: 1408-1412 4. Beaudoin AR (1980) Embryology and teratology. In: Baker HJ, Lindsey JR, Weisbroth SH (eds) The laboratory rat, vol 2. Academic Press, New York, pp 75-1 01 5. Blake CCF, Geisow MJ, Oatley SJ, RCrat B, RCrat C (1978) Structure of prealbumin: Secondary, tertiary and yu?ternary interactions, determined by Fourier refinement at 1.8 A. J Mol Biol 121:339-356 6. Davson H (1970) Physiology of the cerebrospinal fluid. Churchill, London, pp 273-279 7. Dickson PW, Howlett GJ, Schreiber G (1985) Rat transthyretin
235 (prealbumin). Molecular cloning, nucleotide sequence, and gene expression in liver and brain. J Biol Chem 260: 8214-8219 8. Dwulet FE, Benson MD (1984) Primary structure of an amyloid prealbumin and its plasma precursor in a heredofamilial polyneuropathy of Swedish origin. Proc Natl Acad Sci USA 81 :694-698 9. Felgenhauer K (1974) Protein size and cerebrospinal fluid composition. Klin Wochenschr 52: 1158-1164 10. Ferguson RN, Edelhoch H, Saroff HA, Robbins J (1975) Negative cooperativity in the binding of thyroxine to human serum prealbumin. Biochemistry 14:282-289 3 1. Fishman RA (1980) Cerebrospinal fluid in diseases of the nervous system. W.B. Saunders, Toronto, pp 193-194 12. Gee CE, Roberts JL (1983) Laboratory methods. In situ hybridization histochemistry: A technique for the study of gene expression in single cells. DNA 2: 157-163 13. Jacobsen M, Jacobsen GK, Clausen PP, Saunders NR, Mollgard K (1982) Intracellular plasma proteins in human fetal choroid plexus during development. 11. The distribution of prealbumin, albumin, alpha fetoprotein, transferrin, IgG, IgA, IgM and alpha-1-antitrypsin. Dev Brain Res 3 :251-262 14. Kanda Y, Goodman DS, Canfield RE, Morgan FJ (1974) The amino acid sequence of human plasma prealbumin. J Biol Chem 249 :6796-6805 15. Kato M, Kato K, Goodman DS (1984) Immunocytochemical studies on the localization of plasma and of cellular retinolbinding proteins and of transthyretin (prealbumn) in rat liver and kidney. J Cell Biol98: 1696-1704 16. Kopelman M, Cogan U, Mokady S, Shinitzky M (1976) The interaction between retinol-binding proteins and prealbumins studied by fluorescence polarization. Biochim Biophys Acta 439:449460 17. Lehrach H, Diamond D, Wozney JM, Boedtker H (1977) RNA molecular weight determinations by gel electrophoresis under denaturing conditions, a critical reexamnation. Biochemistry 16:47434751 18. Mita S, Maeda S, Shimada K, Araki S (1984) Cloning and sequence analysis of cDNA for human prealbumin. Biochem Riophys Res Commun 124:558-564 19. Nakazato M, Kangawa K, Minamino N, Tawara S, Matsuo H, Araki S (1984) Identification of a prealbumin variant in the serum of a Japanese patient with familial amyloidotic polyneuropathy. Biochem Biophys Res Commun 122:712-718 20. Raz A, Shiratori T, Goodman DS (1970) Studies on the protein-protein and protein-ligand interactions involved in retinol transport in plasma. J Biol Chem 245: 1903-1912
21. Rigby PWJ, Dieckmann M, Rhodes C, Berg P (1977) Labeling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J Mol Biol 113:237-251 22. Saraiva MJM, Birken S , Costa PP, Goodman DS (1984) Amyloid fibril protein in familial amyloidotic polyneuropathy, Portuguese type. Definition of molecular abnormality in transthyretin (prealbumin). J Clin Invest 74 : 104-1 19 23. Sasaki H, Sakaki Y, Matsuo H, Goto I, Kuroiwa Y, Sahashi I, Takahashi A, Shinoda T, Isobe T, Takagi Y (1984) Diagnosis of familial amyloidotic polyneuropathy by recombinant DNA techniques. Biochem Biophys Res Commun 125:63G642 24. Soprano DR, Herbert J, Soprano KJ, Schon EA, Goodman DS (1 985) Demonstration of transthyretin mRNA in the brain and other extrahepatic tissues in the rat. J Biol Chem 260:11793-11798 25. Sternberger LA, Hardy PH, Cuculis JJ, Meyer HG (1970) The unlabeled antibody enzyme method of immunohistochemistry. Preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochetes. J Histochem Cytochem 18:315-333 26. Sundelin J, Melhus H, Das S , Eriksson U, Lind P, Traglrdh L, Peterson PA, Rask L (1985) The primary structure of rabbit and rat prealbumin and a comparison with the tertiary structure of human prealbumin. J Biol Chem 260:64814487 27. Thomas P (1980) Hybridization of denatured RNA and small DNA fragments transferred to nitrocellulose. Proc Natl Acad Sci USA 77 :5201-5205 28. Tushinski RJ, Sussman PM, Yu LY, Bancroft FC (1977) Pregrowth hormone messenger RNA . Glucocorticoid induction and identification in rat pituitary cells. Proc Natl Acad Sci USA 74~2357-2361 29. Van Jaarsveld PP, Edelhoch H, Goodman DS, Robbins J (1973) The interaction of human plasma retinol-binding protein with prealbumin. J Biol Chem 248 :4698-4705 30. Weisner B, Kauerz U (1983) The influence of the choroid plexus on the concentration of prealbumin in CSF. J Neurol Sci 61 ~27-35 31. Weisner B, Rothing HJ (1983) The concentration of prealbumin in CSF, indicator of CSF circulation disorders. Eur Neurol 22: 96-105
Received November 1985/Acceptedin revised form March 28,1986