Expression and quantitative changes of carbonic anhydrase in developing neurones of rat central nervous system

Int. J. Devl. Neuroscience, Vol. 9, No. 6, pp. 555-561, 1991. Printed in Great Britain.

0736-5748/91 $3.00+0.00 Pergamon Press plc © 1991 ISDN

EXPRESSION AND Q U A N T I T A T I V E C H A N G E S OF C A R B O N I C A N H Y D R A S E IN D E V E L O P I N G N E U R O N E S OF R A T C E N T R A L N E R V O U S SYSTEM ANTAL NOGRADI* a n d ANDRAS MIHALY'I':~ *Department of Anatomy and Developmental Biology, University College London, Gower Street, WC1E 6BT, U.K. tDepartment of Anatomy, Albert Szent-Gy6rgyi Medical University, 6701 Szeged P.O. Box 512, Hungary

(Received 23 May 1991; in revised form 6 August 1991; accepted 9 August 1991) Abstract--Postnatal changes in carbonic anhydrase activity were investigated in the islands of Calleja, which have been previously reported to contain the enzyme. Results obtained with a new modified method of Hansson provided further evidence for the distinction between the medial and lateral islands of Calleja. The enzyme was localized mainly in the nucleus and cytoplasm of granule cells without showing binding to any cytoplasmic organelle. No large neurons of the islands displayed carbonic anhydrase reactivity. The time course and rate of increase of carbonic anhydrase expression were different in the giant island of Calleja and lateral islands and this finding may strengthen the hypothesis regarding the medio-lateral diversity of Calleja's islands. On the other hand, at the end of the maturation process the granule cell complexes showed no significant difference in the proportion of carbonic anhydrase positive neurones. The almost equal rate of appearance of carbonic anhydrase reactive granule cells raises the possibility of a basic common role of both medial and lateral islets.

Key words: carbonic anhydrase, islands of Calleja, postnatal development.

The granule cell islands of the olfactory tubercle (islands of Calleja, IC's, Calleja's granule cell complexes) are present in the central nervous system (CNS) of every investigated animal, including man. t7 They consist not only of granule cells but also associated large and medium-sized neurones. 12 Although generally in most species the IC's are embedded in the laminated corticoid structure of the olfactory tubercle, there is not close association between the islands of Calleja and the olfactory tubercle. The medial islands and the insula magna (ICM) are the structures present in every species, while the lateral islands which may also receive fibers from the lateral olfactory tract vary in size and may even be absent in some anosmatic species, such as the dolphin. 17Histochemical and immunohistochemical investigations presented evidence for the resemblance of IC's with the striatopallidal system. 8,2~ These findings were further supported by developmental studies of Bayer (1985). 2 In contrast to this, Fallon et al. in 1983 found estrogene binding receptors and luteinizing hormone-releasing hormone immunoreactivity in the medial islands and proposed the possible neuroendocrine functions of these granule cell clusters. 7 Carbonic anhydrase (CA; EC 4.2.1.1.) is a metabolic enzyme catalyzing the reaction C O 2 + H 2 0 ~ H C O 3 + H +, first described in the nervous system by van Goor (1940). It plays an important role in removal and transport of carbon dioxide produced by glycolysis, in the intermediary metabolism, in acid secretion by stomach and kidneys and promoting many ion fluxes, but its function is still unknown in some biochemical processes. 6 In the CNS, CA has been localized mainly in oligodendroglial cells 9,22 but it is also present in the choroid plexus epithelial cells, 9"t8 in some axons, 23 in pericytes 13 and recently it has been demonstrated in protoplasmic astrocytes4 and ameboid microglial cells. 2° In the PNS the large and medium-sized dorsal root ganglion (DRG) cells show CA reactivity.13"14 It was originally proposed that CA was localized extra-neuronally, however, some mammalian CNS nerve cell populations were found to contain the enzyme. Aldskogius et al. in 1988 reported that CA is present in neurones of mesencephalic nucleus of the trigeminal nerve, t but these neurones are peripheral ganglion cells, that migrated into the brainstem during embryonic life.15 On the other hand the granule cells of Calleja's islands contain high amount of CA, but the function of CA in these neurons is not defined. Granule cells of IC's were reported to be surrounded by a honeycomb-like membraneous network of astrocyte-like cell processes ~7which Abbreviations: CA, carbonic anhydrase; CNS, central nervous system, CoS, cobalt sulphide; DRG, dorsal root ganglion, IC's, Islands of Calleja; ICM, Insula Calleja Magna. qtAntal N6gr~idi is on leave from the Dept of Anatomy, Albert Szent-Gy6rgyi Medical University.

555

556

A. N6gr~di and A. Mih~ily

could easily satisfy the metabolic demands of these cells as well as the removal of CO2 produced by these neurons. Despite the recent investigations neither the function of IC's nor the reason for morphological differences between different IC's has been revealed. There are several theories concerning the function of IC's and also the possibility that granule cell clusters have no function has been suggested, though an absence of function of a group of neurones seems unlikely. The aim of this study was to reveal the features and differences of CA activity in different IC's, which may help us to explain the function of these mysterious islets and the function of CA in them. EXPERIMENTAL PROCEDURES Sixteen Wistar rat pups were used aged 10, 12, 14, 16, 18, 20, 22 and 24 days. The day of birth is referred to as postnatal day zero (P0). The pups were anaesthetized with ether and perfused transcardially with ice-cold phosphate-buffered 2.5% glutaraldehyde (pH 7.4). The brains were removed carefully and postfixed overnight in the same fixative containing 10% sucrose. The olfactory tubercle was cut out and rinsed in 0.05 mol/! phosphate buffer containing 0.2 mol/1 sucrose (pH 7.4). Serial sections 15 ~xm thick were cut on a freezing microtome and the sections collected in the same buffer as above. After washing, the sections were mounted on a tubularshaped semipermeable dialysis membrane (pore size approx. 2.5 nm, Union Carbide). The cellophane was filled with the incubation fluid (1.75 mmol/l CoSO4--~-11.7 mmol/l KH2PO4+ 157 mmol/l NaHCO3+53 mmol/1 H2SO4, pH 6.9, after Brown 1980, 3 Hansson, 196711). This arrangement ensured that the sections remained on the outer surface of the membrane, allowing the satisfactory removal of CO2. is The specificity of the staining was controlled by inhibiting CA by addition of 10 - 4 mol/l acetazolamide (SIGMA) to the control incubation medium. Incubation time was 20 min at room temperature. Then the sections were rinsed in phosphate buffered saline solution (0.67 mmol/l NaCI in 0.1 mol/1 phosphate buffer, pH 5.9), and immersed in 0.5% yellow NazS for 2 min, rinsed twice in distilled water and mounted on gelatinized slides. Every fifth section was counterstained with cresyl-violet in order to identify the plane of section.

Electron microscopy Thirty Ixm thick Vibratome sections were incubated in the same way as described above. After the NazS treatment they were rinsed in phosphate buffer (pH 6.0), postfixed in 1% OsO4 at pH 5.2 for 3 min, 23 then washed in phosphate buffer (pH 5.2). They were dehydrated in graded ethanols and embedded into Durcupan. Sections were cut perpendicular to the plane of the embedded section. Semithin sections were stained with toluidine blue. Ultrathin sections stained with lead citrate were examined in a JEOL 100CX electron microscope.

Cell counting Distinct populations of granule cells in ICM and lateral IC's were counted. The number of CAreactive and non-reactive neurones was determined by examining every 3rd section under light microscope. Only cells with visible nucleoli were used (Fig. 4) and no attempt was made to distinguish between those staining heavily or lightly for CA, only whether they stained above the background level. RESULTS During the first ten days of life granule cells of Islands of Calleja (IC's) and Insula Calleja Magna (ICM) did not display any CA staining. Only heavily stained oligodendrocytes and capillaries could be seen in the islands in this period. Capillaries generally were closely attached to granule cell complexes, some of the islands were pierced by smaller capillaries (Fig. 1). After the 10th postnatal day the granule cells of the medial giant island started to show CA reactivity whilst the lateral IC's remained unstained, they showed the first signs of CA activity only from the 13th postnatal day.The CA reactivity was localized within the cells both in the cytoplasm and nucleus (Fig. 4) and also the nucleolus displayed some nonspecific greyish staining. In some cases, the nuclear reaction was stronger than that of the cytoplasm. However, in control sections 10 - 4 M

Postnatal e x p r e s s i o n o f C A in Calleja's islands

Figs 1-4. 1, Photomicrograph of a giant island at the 16th postnatal day. Numerous granule cells exert CA reactivity. Note the blood vessel (V) closely associated to the island and the unstained medium-sized cells (arrows). Scale bar = 100 pan. 2, CA activity in ICM from a 20-day-old rat. The scattered lines represent the vertical borders of the island. Note that the vast majority of granule cells is CA reactive. Scale bar = 100 p.m. 3, High magnification of a ring-like IC, processed for CA histochemistry. Note the closely attached CA reactive granule cells (arrowheads) and the neuron-sparse core of the island (asterisk) showing only moderate activity. Scale bar = 20 tim. 4, Photomicrograph of a lateral IC from a 18-day-old rat. Arrows represent the cell nucleoli, which showed nonspecific staining in every cell. Whilst the CA reactive cells can be easily observed, not all the unstained granule cells can be seen, because their nucleoli are not focused. Scale bar = 20 p.m.

557

558

A. N6gr~idi a n d A. Mih~ily

Fig. 5. Electron micrograph of a mature IC. The nuclei (N) of some CA reactive neurons contain the CoS deposits. The cytoplasm of granule cells denoted with small arrowheads also displays CA staining. Note the CoS reaction product in myelin sheath (large arrowhead). Since the ultrathin section was cut perpendicular to the plane of the sections, the reaction product is mainly localized near the edge of the section. Scale bar = 10 p,m.

Postnatal expression of CA in CaUeja's islands

559

acetazolamide completely abolished both the nuclear and cytoplasmic CA reaction, only the nucleoli maintained their nonspecific staining. The cobalt sulphide (COS) reaction product was equally distributed in the cytoplasm and because of the faint cytoplasmic CA staining we could not see the enzyme binding to any cytoplasmic organelle (Fig. 5). Some of the myelinated fibers contained CoS deposits in the myelin sheath. No neurons of the surrounding tissues showed CA staining. Though the lateral IC's displayed initially only low CA activity (Fig. 4), at the same time some ring-like IC's with a neuron-spare core were heavily stained already at the beginning of CA activation of lateral granule cell islands. Almost every cell of these special IC's was CA-reactive while the neuropil in the center displayed only moderate staining at the background level (Fig. 3). The cells seemed to be closely attached to each other. The relatively early appearance of CA in the Insula Magna of Calleja (ICM) was followed by a slow but continuous increase in the expression of CA (Figs 2 and 7). The appearance of CA activity in the lateral IC's was markedly different: it appeared to be more rapid, though started several days later than that of the ICM, and this can be seen in the steeper gradient of Fig. 6. However, both granule cell populations reached similar levels of CA activity, since from the 22nd postnatal day the percentages of CA reactive granule cells were approximately the same (Table 1). From this age on no significant difference could be observed in the ratios of CA reactive and nonreactive granule cells between lateral IC's and ICM. DISCUSSION The CA activity of CNS neurons is a unique feature of Calleja's granule cell complexes. As previously described, this extraordinary enzyme activity in granule cells may suggest a special function and/or property of these neurons.19 Although the presence of the histochemical reaction

'~ ~g

80 70

oN

g-~m "5 ~ 40 50 I °

20

G

1'6 l'a

~o ~2

~4

postnatal days

Fig. 6.

ICM

9O •->

::~

80

70

~u ~ 60' ~6o

50

g~

40

8

20

1:4

1'6

1'8

2'o

;2

~4

postnatal days

Fig. 7. Figs 6 & 7. Time course of CA histochemical activity in granule cell populations of lateral IC's (Fig. 6) and ICM (Fig. 7). The percentage of stained granule cells in each group were averaged for each postnatal day point. Standard errors are encompassed by vertical lines at each data point.

560

A. N6gr~di and A. Mih~ily Table 1. Percentage of CA reactive granule cells in ICM and lateral IC's during the late phase of postnatal development Postnatal days 14

% ICM IC's

39.85-+1.15 8.17-+3.4

%

%

18

20 %

22 %

42.7-+2.0 29.3+-1.7

48.7_+2.9 38.9-+3.84

68.4_+3.4 5,1.4+1.12

72.8_+2.(I 71.0-+l.22

16

24 % 84.65-+2.15 82.4-+2.6

product in the nucleus raises the possibility of an artifact, the appearance of CoS deposits may be explained with the easy diffusion of the enzyme during fixation. Nevertheless, this phenomenon was observed by other authors. ~J Since the reaction product was localised mainly in the nucleus and the cytoplasm was only faintly stained without any peculiar staining of cytoplasmic organelles, the cytoplasmic reaction was considered to be cytosolic. However, recent publications presented evidence for the diversity of medial and lateral islands. The lateral islands show some connection with the lateral olfactory tract, and they may receive olfactory inputs, but the medial ones exist even in anosmatic animals, such as the dolphin and are well developed in the microsmatic man. 17 Medial islands of the cat display substance P-like immunoreactive granule cells, 24 and these were suggested to excite the dendrites of neurons located in the hilus of the olfactory tubercle, while no such immunoreactivity has been observed in the lateral islands. Fallon et al. in 1983 reported the [3H] estrogene binding and luteinizing hormone-releasing hormone ( L H - R H ) immunoreactivity of medial islands 7 and suggested the possible neuroendocrine function of medial islands. If this is the case it is not surprising that the development of CA expression in the medial and lateral islets is different, although they show the same CA activity and the same ratios of CA reactive granule cells by the end of neuronal maturation, and the function of CA is not directly associated with any neuroendocrine process or olfactory mechanism. On the other hand, high concentration of CA is required in the early neural maturation especially in the rapid growth phase of neurones, m Although the time course of CA expression in lateral IC's is similar to the activation of soluble CA in brain, the reason for the earlier appearance and particularly slow increase of activity in the ICM is not clear. Hosoya (1973) reported first the presence of gap junctions and reciprocal synapses between granule cell somata and dendrites and also between their somata. 12These ceils have the tendency to form rows or rings of closely apposed perikarya and this may explain the synchronous discharge of granule cells. 21 These special synaptic organizations form the morphological base of electronic coupling between granule cells and probably function as an amplifier, which may maintain and enhance activity. However, even this suggested 'electric amplifier' f u n c t i o n - which may not explain completely the real role of I C ' s - requires continuous and high energy supply. The presence of capillaries in close contact with IC's and the honeycomb-like astrocytic compartments for granule cells may support this opinion. Given the relationship between CA and the promotion of elimination of carbon dioxide in metabolically active cells, 6"1° the presence of CA seems to be essential in these neurones. On the other hand, the CA activity in neurones may not entirely be related to energy metabolism of the cell, but also to some other functions fulfilled by C A ? Recently Carr et al. (1989) reported, that most, but not all CA positive D R G cells exhibited dense or at least moderate cytochrome oxidase reactivity, casting doubts that all C A positive D R G neurones are tonically active, s because cytochrome oxidase activity is closely coupled to energy demands of neurones and therefore it is a well established indicator of neuronal electrical activity. They have suggested that CA in D R G neurones not only facilitates carbon dioxide transport but also participates in the regulation of transmembrane fluxes of chloride ions and/or maintains the appropriate distribution of this ion. Since most large and medium sized D R G neurones have muscle afferents, in which chloride ion fluxes may generate receptor potentials, it may be a reasonable explanation for the presence of CA in these neurones. However, no similar alternative processes are known in the granule cells of IC's. Our results further strengthened the recently proposed distinction between medial and lateral IC's. Although these latero-medial differences are striking they do not provide satisfactory

Postnatal expression of CA in Calleja's islands

561

explanation for the function of Calleja's granule cell complexes. Nevertheless, independent of differences in species, morphology, histology and immunohistochemistry, there is a common denominator in the case of all IC's, which is the equal CA reactivity both in medial and lateral islands. Since both the CA activity of Calleja's island neurones and the electrotonic coupling between these granule cells are unique features of this CNS structure, we suggest a close relationship between these phenomena. Notwithstanding the possibility that medial and lateral IC's may have different functions, the role of IC's and that of CA in granule cells remains to be determined. Acknowledgements--The authors are indebted to Prof. Gerta Vrbov~i for stimulating discussions and constructive criticism. We greatly appreciate the contribution of Mrs Katalin Lakatos and Mr Istv~in Farag6.

REFERENCES 1. Aldskogius H., Arvidsson J. and Hanson (1988) Carbonic anhydrase enzyme histochemistry of cranial nerve primary sensory afferent neurons in the rat. Histochemistry 88, 151-154. 2. Bayer S. A. (1985) Neurogenesis in the olfactory tubercle and islands of Calleja in the rat. Int. J. Devl. Neurosci. 3, 135-147. 3. Brown D. (1980) Carbonic anhydrase localization in mounted cryostat sections. Stain Technol. 55, 115-118. 4. Cammer W. and Tansey F. A. (1988) Carbonic anhydrase immunostaining in astrocytes in the rat cerebral cortex. J. Neurochem. 50, 319-322. 5. Cart P. A., Yamamoto T., Staines W. A., Whittaker M. E. and Nagy J. I. (1989) Quantitative histochemical analysis of cytochrome oxidase in rat dorsal root ganglion and its co-localization with carbonic anhydrase. Neuroscience 33, 351-362. 6. Deutsch H. F. (1987) Carbonic anhydrases. Int. J. Biochem. 19, 101-113. 7. Fallon J. H., Loughlin S. E. and Ribak C. E. (1983) The islands of Calleja complex of the rat basal forebrain III. Histochemical evidence for a striatopallidal system. J. Comp. Neurol. 218, 91-120. 8. Fallon J. H., Riley J., Sipe J. and Moore R. Y. (1978) The islands of Calleja: Organization and connections. J. Comp. Neurol. 181,375-396. 9. Giacobini E. (1961) Localization of carbonic anhydrase in the nervous system. Science 134, 1524-1525. 10. Giacobini E. (1987)Carbonic anhydrase: The first marker of glial development. Curt. Top. Devel. Biol. 21,207-215. 11. Hansson H. P. J. (1967) Histochemical detection of carbonic anhydrase activity. Histochemie 11, 112-128. 12. Hosoya Y. (1973) Electron microscopic observations of the granule cells (Calleja's islands) in the olfactory tubercle of rats. Brain Res. 54. 330-334. 13. Kazimierczak J., Sommer E. W., Philippe E. and Droz B. (1986) Carbonic anhydrase activity in primary sensory neurons I. Requirements for the cytochemical localization in the dorsal root ganglion of chicken and mouse by light and electron microscopy. Cell. Tiss. Res. 245. 487-495. 14. Korhonen L. K. and Hyyppa (1967) Histochemical localization of carbonic anhydrase activity in the spinal and coeliac ganglia of the rat. Histochemie 26, 75-79. 15. Lieberman A. R. (1976) Sensory ganglia. In The Peripheral Nerve (ed. Landon D. N.), pp. 188-278, Chapman & Hall, London. 16. Meyer G. and Wahle P. (1986) The olfactory tubercle of the cat I. Morphological components. Exp. Brain Res. 62, 515-527. 17. Meyer G., Gonzalez-Hernandez T., Carillo-Padilla F. and Ferres-Torres R. (1989) Aggregations of granule cells in the basal forebrain (Islands of Calleja): Golgi and cytoarchitectonic study in different mammals, including man. J. Comp. Neurol. 284, 405-428. 18. Mihfily A., Kirzily E., N6gr,'idi A. and Bencsik K. (1988) The semipermeable membrane technique in carbonic anhydrase histochemistry of the nervous system: a new modification of the method of Hansson. Histochemistry 88, 485-487. 19. N6gr,'ldi A., Kir~tly E. and Mih,'ily A. (1989) Neuronal carbonic anhydrase activity in the central nervous system of the rat: light and electron histochemical investigations of the islands of Calleja. Acta Histochem. 85, 187-193. 20. N6gr~ldi A. and Mih~lly A. (1990) Light microscopic histochemistry of the postnatal development and localization of carbonic anhydrase activity in glial and neuronal cell types of the rat central nervous system. Histochemistry 94, 441447. 21. Ribak C. E. and Fallon J. H. (1982) The islands of Calleja complex of rat basal forebrain l.Light and electron microscopic observations. J. Comp. Neurol. 205, 207-218. 22. Ridderstrale Y. and Hanson M. (1985) Histochemical study of the distribution of carbonic anhydrase in the cat brain. Acta Physiol. Scand. 124, 557-564. 23. Riley D. A., Ellis S. and Bain J. L. W. (1984) Ultrastructural cytochemical localization of carbonic anhydrase activity in rat peripheral sensory and motor nerves, dorsal root ganglia and dorsal column nuclei. Neuroscience 13, 189-206. 24. Wahle P. and Meyer G. (1986) The olfactory tubercle of the cat II. Immunohistochemical compartmentation. Exp. Brain. Res. 62, 528--540.