MEMBRANE AND CELLULAR BIOPHYSICS AND BIOCHEMISTRY NEUROREPORT

0959-4965 & Lippincott Williams & Wilkins Vol 11 No 14 28 September 2000 3099

Immunohistochemical study of the hnRNPA2 and B1 in the rat forebrain

Katsuyoshi Mizukami,CA Hiroshi Kamma,2 Masanori Ishikawa and Gideon Dreyfuss2

Department of Psychiatry, Institute of Clinical Medicine and 1Department of Pathology, Institute of Basic Sciences, University ofTsukuba, 1-1-1 Tennoudai, Tsukuba City, Ibaraki 305-8575, Japan; 2Howard Hughes Medical Institute and Department of

Biochemistry and Biophysics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104-6148, USA

CACorresponding Author

Received 8 June 2000; accepted 11 July 2000

Immunohistochemical techniques were employed to examinethe distribution of RNA-binding proteins A2 and B1 in the ratforebrain. Intense A2 and B1 immunolabeling were observed inthe nucleoplasm of the neurons in the cerebral cortices,hippocampal formation, olfactory regions, caudate-putamen aswell as the supraoptic nucleus of hypothalamus. In contrast,within the bed nucleus of the stria terminalis, as well as themedial and lateral habenular nucleus of thalamus, immunoreac-tivity for both proteins was weak. Within the globus pallidus

and thalamic nucleus immunoreactivity for A2 was hardlydetectable despite of intense B1 immunolabeling, while withinthe endopiriform nucleus and lateral and basolateral nucleus ofamygdala intensity of B1 immunolabeling was relatively weakcompared to A2. Our study suggests that the distribution ofA2 and B1 are not constant throughout the forebrain and thisdiversity may re¯ect the post-transcriptional regulation of cell-speci®c gene expression of neuronal cells. NeuroReport11:3099±3102 & 2000 Lippincott Williams & Wilkins.

Key words: Forebrain; hnRNP; Immunocytochemistry; Rat; RNA-binding protein

INTRODUCTIONIt is well known that the heterogeneous nuclear ribonucleo-protein (hnRNP) proteins directly bind to pre-mRNAforming a large multiprotein±RNA complex and haveimportant roles in the splicing and transport of mRNAs[1,2]. The hnRNP A2 and B1 proteins are among the mostabundant of the �20 major hnRNP proteins in the humannucleus [1]. A2 and B1 are produced by alternative splicingfrom a single copy gene [3], and differ from each otheronly by a 12 amino acid insertion in the N-terminal of B1[4]. We produced two monoclonal antibodies against thesetwo proteins (i.e., DP3 and 2B2), and demonstrated tissue-speci®c variations in the amounts of these two proteins[5,6]. Although A2 and B1 proteins are abundantly ex-pressed in the brain, the detailed distribution of theseproteins in the brain is still unknown. In this study,immunohistochemical techniques were employed in orderto examine the detailed distribution of A2 and B1 proteinsin the rat forebrain.

MATERIALS AND METHODSSix male Sprague±Dawley rats weighing �250 g wereanesthetized with 360 mg/kg chloral hydrate and werekilled by transcardial perfusion with 4% paraformaldehydein 100 mM phosphate buffer. Following 24 h post-®xationthe brains were placed in 30% sucrose for 3 days and thensectioned at 40 ìm in the coronal plane and processed forimmunocytochemistry as described [7]. Tissue sectionswere immunocytochemically labeled using the avidin±

biotin-immunolabeled procedure using DP3 and 2B2, bothof which are raised in mice; DP3 represents A2 while 2B2represents B1 [6]. The primary antibodies were diluted 1/500 for DP3 and 1/2000 for 2B2 in Tris±saline containing1% goat serum and 0.25% Triton X-100. Aminoethyl-carbamazole was used as a coloring reagent. The ®naldilutions of the antibodies were determined followingexperiments in which a number of dilutions were tested(i.e. dilution curve). Sections were processed together inorder to control for any variability in the immunocyto-chemical procedure. The adjacent section was stained forNissl substance to delineate the cytoarchitecture describedby Kruger et al. [8] and Paxinos [9].

RESULTSThroughout the brain immunoreactivity for both DP3 and2B2 was localized in the nucleoplasm of neurons andependymal cells and widely distributed even though thelabeling intensity varied depending on regions and celltypes. Within the cerebral cortex, hippocampus, olfactoryregions and caudate-putamen, as well as the supraopticnucleus of the hypothalamus, we observed intensely DP3-labeled neurons (Fig. 1a±c). Within the cerebral cortex,DP3 immunolabeling was prominent in layers II and III, aswell as in a large neuron in layer V, while in the otherlayers immunoreactivity was weak. A lobar difference inlabeling pattern was not remarkable. Within the olfactorycortex, hippocampus, caudate-putamen and the supraopticnucleus, the labeling intensity was variable depending on

neurons, although a number of neurons were intenselylabeled (Fig. 1b,c). In contrast, within the globus pallidus,bed nucleus of the stria terminalis and throughout thethalamic regions, immunoreactivity for DP3 was hardlydetectable (Fig. 1c,d). In the other regions, such as theseptal nucleus, nucleus accumbens, diagonal band, sub-stantia inominata, amygdala complex and the hypothala-mus, neurons demonstrated a moderate intensity of DP3immunolabeling.

2B2-immunolabeled neurons were also distributedthroughout the brain. An intense immunoreactivity for 2B2was observed in the regions in which the intense DP3labeling was present. In these regions, a staining pattern of2B2 was almost comparable to that of DP3, although withinthe caudate-putamen a number of lightly labeled neuronswere observed in addition to intensely labeled neurons (Fig.2a). However, an intense 2B2 immunolabeling was morewidely distributed compared to DP3. For example, in theglobus pallidus (Fig. 2a), most thalamic regions (Fig. 2b),the diagonal band, substantia inominata and the ventrome-dial nucleus of the hypothalamus, the majority of neuronswere intensely 2B2 positive. In contrast to the ventromedialnucleus, within the dorsomedial nucleus a number ofneurons demonstrated a weak immunoreactivity for 2B2,

thus providing us with a clear demarcation between thesetwo regions (Fig. 2c). In addition, within the bed nucleus ofthe stria terminalis, lateral habenular nucleus, lateral andbasolateral nucleus of the amygdala, as well as the endopiri-form nucleus, intensity of 2B2 immunolabeling was weak.Within the medial habenular nucleus the majority of neu-rons appeared unlabeled while some neurons demonstratedan intense immunoreactivity for 2B2 (Fig. 2d). Both DP3and 2B2 immunolabeling were barely detected in glialelements. The labeling intensities of DP3 and 2B2 areclassi®ed into three categories: undetectable or weak (ÿ),moderate (�) and intense (� �) according to the previousreport [6], and the data are summarized in Table 1.

DISCUSSIONSince DP3 and 2B2 immunoreactivity represent A2 and B1expression, respectively [6], the present data demonstratethe comprehensive study of the expression patterns ofhnRNP A2 and B1 proteins in the forebrain. The results ofthis study suggest that expression for A2 and B1 wasvaried in different regions. In some regions such as thecerebral cortex, hippocampus, olfactory regions, caudate-putamen, as well as the supraoptic nucleus, immunoreac-tivity for both A2 and B1 was robust. In contrast, in the

Fig. 1. Photomicrographs showing immunoreactivity for DP3. Within the parietal cortex, intense immunoreactivity for DP3 was observed in layers IIand III, as well as a large neuron in layer V (arrow) while in the other layers immunoreactivity was weak (a). In the hippocampus (b) and caudate-putamen (CP; c) a number of neurons were intensely labeled although the labeling intensity was varied depending on neurons. In contrast, within theglobus pallidus (GP; c) as well as the mediodorsal nucleus of the thalamus, immunoreactivity for DP3 was hardly detectable (d). Bar� 100ìm (a±c);50 ìm (d).

NEUROREPORT K. MIZUKAMI ET AL.

3100 Vol 11 No 14 28 September 2000

bed nucleus of the stria terminalis, as well as in the medialand lateral habenular nucleus, immunoreactivity for thetwo proteins was weak. The amounts of major hnRNPproteins may, in part, re¯ect the activities of transcriptionand RNA processing. Thus, in interpreting our observa-tions, it is possible to speculate that in regions demonstrat-ing an intense labeling for both A2 and B1 thetranscriptional activity is high, while in regions with aweak labeling for both of them the activity is low. How-ever, studies of functional activities in the rat brain bymeasurements of glucose utilization [10,11] and cyto-chrome oxidase [12] argue against this speculation. Accord-ing to these studies the regions with a highly functionalactivity consist of the cerebral cortex, thalamus and lateralhabenular nucleus, while the regions with a low activityconsist of the hippocampus, amygdala, globus pallidus,medial habenular nucleus and hypothalamus [10±12],although a speci®c functional activity remains to be deter-mined in these regions. Therefore, the intensity of A2 andB1 labeling in neurons is probably relevant to factors otherthan transcriptional activity, such as post-transcriptionalactivity.

Within the different regions we observed clear discre-pancies of the expression between A2 and B1. Within the

globus pallidus and most thalamic nuclei, A2 labeling wasmore modest than B1. In contrast, within the endopiriformnucleus, lateral and basolateral nucleus of the amygdala, aswell as the dorsomedial nucleus of hypothalamus, A2labeling was relatively robust compared to B1. Within thecaudate-putamen the number of intensely B1-positive neu-rons was much smaller than intensely A2-positive neurons.These results lead us to an idea that the speci®c roles of A2and B1 are not equivalent in the brain. In previous studieswith various tissues we demonstrated that the amount ofA2 and B1 proteins varied signi®cantly, depending on thecell type, and suggested that cell type-speci®c expressionof hnRNP proteins re¯ects the post-transcriptional regula-tion of cell-speci®c gene expression [5,6]. Taken togetherwith the fact that neurons are a highly differentiated celland their various functions are regulated at a post-tran-scriptional level as well as a transcriptional level, it isplausible to consider that varieties in the expression of A2and B1 proteins in the brain re¯ect the post-transcriptionalregulation of neuron-speci®c gene expression, such asalternative splicing, metabolism and transport of RNAs.

Recent studies demonstrate that A2 protein binds mye-lin-basic protein in glial cells [13,14] and A2 play a role fortransport of myelin-basic protein to the distal ends of the

Fig. 2. Photomicrographs showing immunoreactivity for 2B2. In the caudate-putamen we observed a number of lightly labeled neurons in addition tointensely labeled neurons (a). In contrast to DP3 immunohistochemistry within the globus pallidus (a) and mediodorsal nucleus of the thalamus (b) 2B2labeling was robust. Within the ventromedial nucleus of the hypothalamus a number of neurons demonstrated an intense immunoreactivity for 2B2while within the dorsomedial nucleus the majority of neurons demonstrated a weak immunoreactivity, thus providing us with a clear demarcationbetween the two regions (c). Within the medial habenular nucleus the majority of neurons appeared unlabeled while some neurons demonstrated anintense immunoreactivity for 2B2 (d). Bar� 100 ìm (a,c); 50 ìm (b,d).

hnRNP A2 AND B1 IN THE RAT FOREBRAIN NEUROREPORT

Vol 11 No 14 28 September 2000 3101

process [14]. In this study, however, neither A2 nor B1immunoreactivity was observed in the glial cells. It isreasonable to suppose that the amount of these proteinswas insuf®cient to be immunohistochemically detected inglial cells. Improvements in sensitivity of immunohisto-chemical methods using tyramide signal ampli®cation(NEN Life Science Pro. Inc, Boston, MA) enable us todetect immunoreactivity in glial elements.

CONCLUSIONOur results suggest that the expression of A2 and B1 arenot constant throughout the brain and this diversity mayre¯ect the post-transcriptional regulation of cell-speci®cgene expression of neurons.

REFERENCES1. Dreyfuss G, Matunis MJ, Pinol-Roma S and Burd CG. Annu Rev Biochem

62, 289ÿ321 (1993).

2. Pinol Roma S and Dreyfuss G. Nature 355, 730ÿ732 (1992).

3. Kozu T, Henrich B and Schafer KP. Genomics 25, 365ÿ371 (1995).

4. Burd CG, Swanson MS, Gorlach M and Dreyfuss G. Proc Natl Acad Sci

USA 86, 9788ÿ9792 (1989).

5. Kamma H, Portman DS and Dreyfuss G. Exp Cell Res 221, 187ÿ196

(1995).

6. Kamma H, Horiguchi H, Wan L et al. Exp Cell Res 246, 399ÿ411 (1999).

7. Mizukami K, Ikonomovic MD, Mishizin A et al. Neuroreport 8,

3379ÿ3383 (1997).

8. Kruger L, Saporta S and Swanson LW. Photographic Atlas of the Rat

Brain. New York: Cambridge University Press; 1995.

9. Paxinos G. The Rat Nervous System, 2nd edn. San Diego: Academic Press;

1995.

10. Sokoloff L. J Cerebr Blood Flow Metab 1, 7ÿ36 (1981).

11. Wree A, Goller HJ and Beck T. J Neurosci Methods 58, 143ÿ149 (1995).

12. Hevner RF, Liu S and Wong-Riley MTT. Neuroscience 65, 313ÿ342

(1995).

13. Carson JH, Kwon S and Barbarese E. Curr Opin Neurobiol 8, 607ÿ612

(1998).

14. Hoek KS, Kidd GJ, Carson JH and Smith R. Biochemistry 37, 7091ÿ7029

(1998).

Table 1. Immunohistochemical examination of A2 and B1 in the ratforebrain

Region A2(DP3) B1(2B2)

Cerebral cortex �� ��Olfactory regions

Piriform cortex �� ��Olfactory tubercle �� ��

Hippocampal formationDentate gyrus �� ��CA ®elds �� ��Subiculum �� ��

Septal nucleus � �Nucleus accumbens � �Caudate-putamen �� ��Globus pallidus ÿ ��Diagonal band � ��Substantia innominata � ��Endopiriform nucleus � ÿBed of nucleus stria terminalis ÿ ÿThalamus

Medial habenular nucleus ÿ ÿLateral habenular nucleus ÿ ÿAnterior nucleus ÿ �Mediodorsal nucleus ÿ ��Central medial nucleus ÿ ��Ventrobasal nucleus ÿ ��Ventrolateral nucleus ÿ ��Lateral dorsal nucleus ÿ ��Reticular nucleus ÿ �Zona incerta ÿ �

AmygdalaMedial nucleus � �Central nucleus � �Cortical nucleus � �Lateral nucleus � ÿBasolateral nucleus � ÿBasomedial nucleus � �

HypothalamusAnterior hypothalamic nucleus � �Supraoptic nucleus �� ��Dorsomedial nucleus � ÿVentromedial nucleus � ��Lateral hypothalamic area � �Paraventricular nucleus � �

NEUROREPORT K. MIZUKAMI ET AL.

3102 Vol 11 No 14 28 September 2000