Organization of the mouse cerebral cortex: A histochemical study using glycogen phosphorylase

201

Brain Research, 267 (1983) 201-216 Elsevier Biomedical Press

Research Reports

Organization of the Mouse Cerebral Cortex: A Histochemical Study Using Glycogen Phosphorylase MARK NELSON WALLACE

Institute of Physiologr', Universityof Glasgow, GlasgowG12 8QQ ( U.K.) (Accepted October 25th, 1982)

Ker words: phosphorylase a histochemistry - functional activity - mouse cortical columns

The postmortem phosphorylase a activity in cryostat sections taken from the brains of mice killed without any prior period of maintained anaesthesia ('unanaesthetizod' mice) was compared histochemically with the activity in sections taken from mice anaesthetized with pentobarbitone prior to decapitation ('anaesthetized' mice). This study provides evidence for the existence of two separate overlapping modular subdivisions of the somatosensory cortex. The modules observed in the 'unanaesthetized' mice were restricted to layer IV and corresponded to the hollows of the whisker barrels where the thalamic afferents terminate. In contrast, the modules observed in the 'anaesthetized' mice extended from layer I to layer V, and formed a mosaic of cylinders that was out of register with the whisker barrels. These cylinders may correspond to the terminal fields ofcorticocortical afferents. In a few of the 'unanaesthetized' mice bands of high phosphorylase a activity are evident in the visual areas 17 and 18. This suggests that in the mouse, as in higher mammals, the thalamic input to the visual cortex gives rise to a columnar organization. In the 'anaesthetized' mice the mosaic of modules observed in the somatosensory cortex is also present in the auditory and motor areas. The modules in both groups of mice have similar diameters of between 200 and 350 ~m. INTRODUCTION

Many recent anatomical and physiological studies of the mammalian neocortex have provided evidence that it is organized into groups of vertically arranged cells; such groups are called 'columns '36. In the physiological studies the columnar organization is usually demonstrated by some form of specific stimulation such as manipulating a single vibrissa 2° or using a pattern of vertical linesJS. However, evidence of a columnar organization has also been obtained in unstimulated animals in studies using the uptake of ~4C-labelled 2-deoxyglucose4° or the bioluminescent demonstration of adenosine triphosphate (ATP) levels31. This evidence was obtained from both anaesthetized 3~ and unanaesthetized 2~.29 animals where bands of altered metabolic activity were demonstrated. The relationship between these bands of metabolic activity and anatomically or physiologically defined columns is not known. An obvious 0006-8993/83/0000- 0000/$03.00 © 1983 Elsevier Science Publishers

place t~or examining their relationship is in the barrel cortex52 of the mouse, where it has been shown that the thalamic afferents associated with a single whisker48 terminate within a single barrel-like cluster of cells in layer IV and give rise to a functional column stretching through all 6 layers ~6.39. Thus any histological technique allowing the simultaneous demonstration of the bands of altered metabolic activity and the cellular barrels will provide information on the spinal relationship of the metabolic bands and the columns dependent on thalamic input. With the 2-deoxyglucose and ATP bioluminescent methods, radiation is emitted as a sphere, which forms the image of a point as a circle on the photographic emulsion. The diameter of the circle depends on the radius of the sphere and so high resolution can only be obtained by using low energy /3 particles from a tritiated source. However, as these particles can only travel through about 1 #m of tissue, a tritiated source is not suitable for studying features such

202 as metabolic bands which have a low contrast. As some of the barrels in the mouse cortex are only 100/,m in diameter a technique with relatively high resolution is required. The low contrast of the metabolic bands requires a microscopical technique that can extract information from the full depth of a 20 >m section. Therefore a histochemical technique, based on glycogen phosphorylase, has been developed which meets these requirements. EXPERIMENTALTECHNIQUES The brains of 40 mice weighing between 25 and 38 g were used. Originally Parkes (albinb) mice were used, but as these mice often have defective vision and hearing when mature, latterly only C 3 H H E were used; 4 Parkes, 3 CBA and 33 C3HHE mice made up the study group. Two rats anaesthetized with urethane were also used to study the localization of the reaction product. Anaesthesia was induced in 18 mice by the inhalation of ethyl chloride and maintained by the intraperitoneal administration of sodium pentobarbitone (Sagatal) usually at a level such that the t\)repaw withdrawal reflex was abolished. These animals were killed between 5 and 90 min after the induction of anaesthesia. Eighteen mice were transiently stupefied with ethyl chloride vapour before being decapitated while 4 mice were decapitated without the use of any anaesthetic. Two of the mice anaesthetized with pentobarbitone were perfused transcardially with 20 ml of cold phosphorylase a incubating medium 34 from a syringe, and then for 5 rain with cold phosphorylase a incubating medium containing 1% paraformaldehyde from a bottle placed 1.5 m above the bench. However, this procedure did not significantly improve the localization of reaction product and so was not continued. After decapitation the skin was cut away from the top of the head and the skull split along the midline so that the brain could be removed and cut into blocks in a number of different orientations. Usually one half of the brain was cut into blocks immediately while the other hemisphere was cooled in 0.9% saline at 4 °C. Usually in the

case of the first hemisphere the whole cerebral cortex was dissected away from the underlying structures and flattened, surface down, on a square of stainless steel foil. This piece of brain. as well as blocks oriented to give coronal or sagittal sections, was then placed on pieces of damp filter paper and rapidly frozen in isopentane which had been cooled to - 1 5 0 ° C by liquid N,. The blocks were frozen between 5 and 20 min after decapitation of the mouse when the phosphorylase a levels were no longer subject to rapid fluctuations 5, and were stored in liquid N~ until immediately before they were to be cut. They' were mounted in a drop of diluted Cryotec tissue mountant on a brass chuck which could be adjusted so that the block face was parallel with the edge of the cryostat knife. Sections were cut at 24/,m, picked up on glass coverslips, dried in the air and placed in Columbia jars. The sections were left in the jars containing Meijer's incubating medium for 90 min to demonstrate the presence of phosphorylase a 3~. The glycogen produced in the incubation medium was stained with acidified I, >. The low pH retards the fading of the k and the staining is stable for at least two months. RESULTS The results come from two groups of mice: ( 1) those which, without any period of maintained anaesthesia, were killed either directly or after being rendered insensible with ethyl chloride vapour (these will be referred to as 'unanaesthetized' mice). The results from the alert mice and from those rendered insensible with ethyl chloride were essentially the same and so no distinction will be made between them. (2) those anaesthetized with sodium pentobarbitone (these wilt be referred to as 'anaesthetized' mice). 'Unanaesthetized' rnice

All the layers in sections from the neocortex of the 22 animals in this group show moderate or high levels of phosphorylase a activity. The rank order of activity is as follows: I ~ IV ~ Vb/Vla II/Ili ~ Va = Vlb.

C All figures are of sections demonstrating the postmortem levels of active phosphorylase. Figs. 1-3 are of sections taken from the brains of mice which had not been given any liquid anaesthetic prior to death. Figs. 4~ 7 are of sections taken from mice which had been anaesthetized with sodium pentobarbitone. Fig. 1. A and C: tangential section through l~iyer IV of the barrel cortex showing the pale barrel walls containing a higher density of cells and surrounding the darker barrel core. (Unanaesthetized mouse.) B and D: coronal section through posterior somatosensory cortex showing the higher activity in layers I and Vb and the dark patches corresponding to the barrel cores in between the paler cell walls in layer IV. (Unanaesthetized mouse.)

Somatosensory cortex M o s t o f the p r i m a r y s o m a t o s e n s o r y cortex is c o m p o s e d o f a field o f 200 cellular barrels 5z w h o s e cell d e n s e walls a p p e a r as pale rings in

surface parallel layer IV (Fig. phosphorylase a the p a r t o f the

o r t a n g e n t i a l sections t h r o u g h IA). Relatively high levels o f o c c u r b o t h in cell b o d i e s a n d in n e u r o p i l c o r r e s p o n d i n g to the

204

vl"

-> e, Fig. 2. A and B: coronal section through the visual cortex showing bands of higher activity which are most evident in layers IV and upper VI. (Unanaesthetized mouse.)

205 hollows of the barrels where the thalamic afferents terminate. The neuropil of the barrel walls is only weakly stained and so the barrels appear as pale rings around a darker core. In a coronal section (Fig. 1B) the barrels appear as patches of higher activity located in layer IV and separated from each other by pale strips indicating the barrel walls. There is no consistent evidence for an extension of this patchy staining pattern for more than a short distance into the other layers of the cortex. Visual cortex

In two of the 22 mice the phosphorylase a levels in coronal sections through the visual cortex exhibit two rows of patches of high activity in layers IV and VI (Fig. 2); the patches in layer IV are the more prominent and they sometimes extend into layer V. The two rows of patches are in register with each other. This gives the impression that the visual cortex is organized into bands of alternating high and low activity each of which has a width of about 300/~m. The width is comparable to that of the whisker barrels whose diameters range from 90 to 330/Lm. The spatial extent of visual areas 17 and 18 as well as other architectonic areas was obtained from the cytoarchitectonic map published by Caviness and Frost 9. Assuming the bands are roughly cylindrical in form it is estimated that there are about 45 cylinders of high phosphorylase activity in the visual areas 17 and 18. Cortical architectonics

The major architectonic divisions can be distinguished in surface parallel sections through layer IV where the thalamic afferents terminate and differences between architectonic areas are most obvious. The various areas are more distinctly outlined than in traditional cytoarchitectonic studies where sections are stained to show the Nissl substance present in cell bodies. This is illustrated in Fig. 3, which is a surface parallel section cut at a depth of 600 /~m and passing mainly through layer IV. The primary auditory cortex (area 41) has a pale brown colour different from the purple of the primary visual area (area 17). The primary somatosensory cortex

(area 3) is characterized by its more orange colour as well as by the cellular barrels. These are more clearly observed at the higher magnification shown in Fig. 1A. Thus the different regions can be distinguished because of the large variations in both the colour and the intensity of the iodine staining. The colour of the iodine in the section is determined by the unbranched chain length of the glycogen with which it is associated 4t . In the different regions of the brain the colour varies from pale brown to deep purple. This implies that different architectonic regions often have different levels of active phosphorylase and of the associated branching enzyme which determines the unbranched chain length of the synthesized glycogen4a. Although useful in identifying the different architectonic regions, the significance, if any, of the variations in chain length is not known and so the colour variations will not be described further. ',4 naesthetized' mice

There was more diversity in the pattern of phosphorylase a activity in sections from the brains of these mice than in the group of 'unanesthetized' mice. Sections in 6 out of the 18 mice showed a modulation of staining intensity over a large extent of the neocortex. In surface parallel sections these modules are seen as a mosaic of irregular dark patches set against a pale background (Fig. 4). Each patch has a diameter of about 250 ~m. In coronal sections the different laminae when stained are still evident; layers IV and Vb are darkest; the other layers are relatively paler. The results described below come mainly from 3 of these 6 mice whose brains showed a modular organization particularly clearly. Sections from 3 other brains were similar to those from 'unanaesthetized' mice. In coronal sections from the remaining nine mice, there is a low level of staining in layers IV to VI of the cortex compared to layers I-III. Sensori-motor cortex

In the posterior part of the barrel field where the whisker afferents terminate, individual bar-

L

~/

~

P

J

, RS .~-----~-~-

35 36

13

14

E

H

41

40 A

17

B I

I

29b

,

lmm M

Fig. 3. A and B: surface parallel section passing largely through layer IV of the mouse cerebral cortex. (Unanaesthetized mouse.) Some of the main architectonic regions are denoted numerically and others by the following abbreviations: OB~ olfactory bulb; E. entorhinal cortex; RS, rhinal sulcus; H, hippocampus; P, pyriform cortex, M, A and L stand for medial, anterior and lateral. The architectonic areas were identified by reference to the map produced by Caviness and Frost' and their nomenclature has been retained.

B

bI

Fig. 4. A: surface parallel section through layer IV of the mouse cerebral cortex. (Anaesthetized mouse.) Discrete patches of higher activity are seen in the frontal, parietal and temporal regions, In this section the areas of cingulate and visual cortex are missing. B: same section with an outline of the architectonic areas observed in Fig. 3 superimposed for comparison. Scale same as for Fig. 3.

208

rels can be histologically distinguished in layer IV. This area allows the direct comparison of the phosphorylase modules with the modules of thalamic input. In contrast to the dark patches in the 'unanaesthetized' brain, the dark patches of high phosphorylase a activity in these brains are not in register with the barrels (Fig. 5); they are arranged at random with respect to the barrels, some straddling the septal areas and others largely or partly confined to the central part of the barrel. This is direct histochemical evidence for the existence of two separate overlapping modular subdivisions of the cerebral cortex. In coronal sections the modules appear as alternating dark and light bands each with a width of about 250 t~m (Fig. 6A). Again, in contrast to the patches of the 'unanaesthetized' brains which are restricted to layer IV, these bands extend from layer I down to the base of layer V. It is difficult to discern the barrels in coronal sections even when Harris's haematoxytin is used as a counterstain. When it is possible to see the barrels they are found to be arranged at random with respect to the bands of high phosphorylase a activity. Taken together the surface parallel and coronal sections show that the phosphorylase a activity is arranged in irregular cylinders of high activity which are out of phase with the thalamic modules. There were about 100 of these cylinders in the somatosensory areas 3, 1, 3a, 40 and 40a. Auditory cortex

The modules in the auditory areas 41 and 22 were similar to those in the sensori-motor cortex being between 250 and 350 #m in diameter. In coronal sections (Fig. 6B) the modules appear as bands of high activity extending from layer II to the base of layer V. Layer IV of these cylinders is more intensely stained than the other layers. There were about 40 of these cylinders in the auditory areas 41 and 22.

of a surface parallel section with the various regions (drawn in Fig. 4B), it was possible to obtain an idea of which areas had a modular pattern in their phosphorylase a activity. These were the motor areas 6 and 4, somatosensory areas 3, 1, 3a, 40 and 40a and auditory areas 41 and 22. The modular pattern of staining is uniformly distributed over these areas and there are no obvious borders between the modules in the different areas as seen in surface parallel sections. A modular pattern of staining has only been observed once in the cingulate and visual regions and so far not at all in the olfactory region. However this pattern of staining is commonly seen in many parts cf the hippocampal region including the entorhinal cortex 47. The reasons for this variability among the results from the anaesthetized mice are not known even although an attempt was made to discover them. Varying the length of time between decapitation and freezing the brain showed that short times (5 rain) gave a higher staining intensity than longer times (15 min) but there were no consistent indications of an effect on the pattern of staining. The depth or duration of anaesthesia were not critical factors either; modules were observed in sections from animals anaesthetized for between 5 and 60 min and from animals which had been deliberately more lightly or deeply anaesthetized than usual, One possible cause for a modular organization in the phosphorylase a activity is that the modules were formed in areas either around or between arterioles. This was found not to be the case as both the areas of high and low phosphorylase activity were associated with arterioles. The arterioles are not very clear in surface parallel sections (Fig. 5) but are more convincingly demonstrated in coronal sections (Fig. 7) where they can be followed down through the different layers. Cellular localization

Other cortical areas

In sections from the brains of this group it was very difficult to make out the borders of the different architectonic regions. However, by superimposing photographic images over a diagram

The investigation of systems of overlapping modules does not require the identification of individual cells, but phosphorylase histochemistry has the potential for revealing functional

209

B

,~ /" \ /

/ /

\

/

: I

/ / ol

,,,

./ "o,,~/ ~'~

I

O

Fig. 5. A: tangential section through layer IV of the sensori-motor cortex. The patches of higher activity are arranged at random with respect to the barrel walls. (Anaesthetized mouse.) The area in the bottom right hand corner is motor cortex. B: some of the more obvious arterioles are marked in on the diagram and are found both within the patches of higher activity and in the pale areas.

210

i!i Fig. 6. A: coronal section through the sensori-motor cortex of a mouse that had been anaesthetized with sodium pentobarbitone and then perfused with cold incubating medium and paraformaldehyde. Bands of high activity each of about 200 tma are seen throughout the extent of the sensori-motor cortex but not in adjacent regions. AC anterior commissure. B: coronal section through auditory and visual areas of the same brain as shown in A. Bands of high activity are onl\, seen in the auditor\ areas 4 l and 22. These are about 250 ttm in diameter. (Anaesthetized mouse. I

changes at a cellular level. In the cortex, cell bodies and their processes are best seen in sections from 'anaesthetized' brains, particularly those from rats. Rat brains were sectioned as part of another study and it was possible to see layer V pyramidal cells with their apical dendrites ascending towards layer I (Fig. 8A). Cell bodies and their processes were also seen in sections from the brains of 'anaesthetized' mice (Fig. 8B) providing these had been fixed beforehand with 1% paraformaldehyde. In some of the processes the phosphorylase a activity is relatively uniform for up to 1 mm. The cell bodies (large arrows) often have a granular appearance similar to that ofhippocampal pyramidal cells24. DISCUSSION

Method Biochemical studies have provided evidence

that phosphorylase levels are linked to functional activity in the nervous system because total phosphorylase levels are roughly correlated with synaptic density ~5 and brain phosphorylase kinase is activated following stimulation )', during seizures js and following the application of neurotransmitters e,r as well as psychotropic drugg ~. It has been suggested that phosphorylase b kinase activity is particularly important at synapses because, as well as regulating phosphorylase, it can also regulate pyruvate dehydrogenase, an enzyme whose activity may be involved in the long term potentiation of hippocampal synapses 7.s. This and other histochemical studies of phosphorylase indicate that the enzyme occurs in axons and dendrites and so may occur pre-and post-synaptically :4. They also provide clear evidence that a large part of the phosphorylase activity in the brain is located in neuronal cell bodies and processes. This is in contrast to

Fig. 7. A and B: coronal section through somatosensory cortex. Both the bands of high and the intervening areas of low phosphorylase activity are associated with arterioles. These are marked by arrows. (Anaesthetized mouse./

J

j

212

Fig, 8. A: coronal section through sensori-motor cortex of a rat that had been anaesthetized with urethane and then decapitated. Moderately high levels of phosphorylase a activity are seen in the layer V pyramidal cell bodies and their apical dendrites which are ascending to layer I. B: coronal section through sensori-motor cortex of a mouse which bad been anaesthetized with sodium pentobarbitone and then perfused transcardially with a solution containing 1% paraformaldehyde. Again, cell bodies and dendritic processes of neurones in layer V have a moderately high level of phosphorylase a activity.

the suggestion of Edwards et al. t7 that phosphorylase may mainly be involved in glial metabolism. Phosphorylase exists in two forms: as the inactive phosphorylase b and as the active phosphorylase a. Inter-conversion occurs by the addition of a phosphate group to phosphorylase b (a reaction controlled by phosphorylase b kinase), or by the removal of a phosphate group from phosphorylase a ( a reaction controlled by one or more protein phosphates). In the brain as elsewhere, phosphorylase b kinase activity is regulated by Ca 2+ levels37 and cyclic adenosine3',5'-monophosphate dependent kinase 2,~7. Alteration in Ca 2+ levels associated with the action potential in the heart are sufficient to produce a 20% increase in phosphorylase a levels within a few hundred msec 44. Thus phosphorylase b kinase participates in a very sensitive system for the rapid adjustment of phosphorylase activity involving Ca 2+ and cyclic adenosine-3',5'-monophosphate, both of which are known to have a role in synaptic transmission ~9. Protein phosphorylation is also involved in

longer term aspects of neuronal regulation, the effects of which may be measurable minutes to hours after they were induced v,~9.33.There is evidence ~,33 to suggest that the protein phosphatases are involved in these intermediate-term aspects. Comparatively little work has been carried out on the function of the phosphatases but there is a suggestion of their being involved in the intermediate-term regulation of phosphorylase activity I~ and at least one hormone affecting glycogen metabolism is thought to act through regulating phosphatase activity by a cellular messenger other than Ca 2+ or cyclic adenosine3',5'-monophosphate ~t. On decapitation the cyclic adenosine-3',5'-monophosphate and Ca -'+ levels in the brain both increase immediately and phosphorylase activity reaches a maximum less than 10 s after death 5. Thereafter the ATP levels fall to low levels32 and, as ATP is required for protein kinases to add phosphate groups to their substrates, after a few minutes the phosphatases alone will control the levels of protein phosphorylation. Thus the postmortem levels of phosphorylase a in the brain may reflect an as-

213 pect of the intermediate-term regulation of metabolism by the protein phosphorylation system, as mediated by the phosphatases, whereas in vivo levels of phosphorylase a will reflect the state of the short-term regulatory system mediated mainly by Ca 2÷ and cyclic adenosine3',5'-monophosphate dependent protein kinases. As it is possible to study the in vivo levels of ATP in the rat brain histochemically3~ it should also be possible to study the in vivo phosphorylase activity histochemically making use of the same method for rapid freezing. This was not attempted in the present study because surface parallel sections were to be studied and so only postmortem levels of phosphorylase a could be observed. As far as the author is aware there has been no previous histochemical study of phosphorylase a activity in relation to intermediate term changes in the functional organization of the brain. However, long-term changes in phosphorylase activity induced by various forms of hypoxia or ischaemia have been demonstrated histochemically24.2s as have the longterm changes produced by whisker damage t3. Other histochemical markers such as succinate dehydrogenasea° or cytochrome oxidase5t have been used in studying long-term changes in the barrel cortex, but as these depend on altered rates of protein synthesis they cannot be used to study the intermediate-term changes underlying the metabolic banding observed in sections from unstimulated animals. Normally phosphorylase breaks down glycogen to glucose- l-phosphate, but to locate the enzyme histochemically this reaction is driven backwards by having a high substrate concentration and a low pH so that glycogen is deposited wherever there is any active phosphorylase present in the section. Phosphorylase requires a certain minimum chain length before it can start adding glucose units but by the time the brain is frozen the levels of brain glycogen are very low32 and in some cells there may not be any intrinsic glycogen for the phosphorylase to act on. Meijer34 overcame the absence of intrinsic glycogen in studying the phosphorylase levels of infarcted cardiac muscle by adding a primer of long unbranched polyglucose chains in the form ofdex-

tran, with a molecular weight of between 40,000 and 80,000. The dextran permeates the section but is not stained by iodine. Thus in the present study the presence of dextran in the incubation medium ensured the demonstration ofphosphorylase activity even in glycogen depleted cells. The author originally had reservations about whether the variations in the phosphorylase a activity did represent changes in physiological activity or were merely postmortem artefacts. However the facts that in the 'unanaesthetized' mice the staining consistently corresponded to the known cytoarchitectonic organization of layer IV, and that in 'anesthetized' mice a corresponding organization was observed in series of both surface parallel and coronal sections from different mice, provide evidence that the histochemical method does have physiological significance. Further evidence to support this conclusion was provided by the histochemical study of phosphorylase a levels associated with the production of an epileptogenic focus46.

Somatosensory cortex The [14C]2-deoxyglucose (2-DG) levels in autoradiographs from control 'unanesthetized' mice show no evidence L6of an extension of the discrete barrel-like organization of layer IV into the other layers. Similarly, in the present work, the phosphorylase a activity in sections from 'unanesthetized' mice showed no evidence of discrete areas of functional activity extending through the various layers of the cortex. In contrast, the brains of 6 'anaesthetized' mice have discrete cylinders of high postmortem phosphorylase a activity, extending through 4 or 5 layers of the cortex, providing strong evidence for a columnar organization. Evidence of a similar arrangement of cylinders in the somatosensory cortex of anaesthetized rats were obtained 3~ by the bioluminescent demonstration of ATP levels. The patches of high phosphorylase a activity observed in sections from the 'unanaesthetized' mice correspond to the barrel hollows, implying that the high phosphorylase a activity is correlated with the presence of thalamic afferent terminals. By contrast, the cylinders of high phosphorylase a activity of the 6 'anaesthetized' mice

214 were not in register with the barrels. Therefore these cylinders are unrelated to the thalamic afferents. Experiments investigating the effects of anaesthetics such as pentobarbitone; have been interpreted as showing that, during anaesthesia, the somatosensory modulation system inhibits the specific thalamic input to the primary sensory areas of the cortex. Thus, as suggested below, a non-thalamic input may be dominant during anaesthesia. Even within the specialized barrel neuropil, only 20% of the synapses are formed by thalamic afferents 5°. The other major input to the cortex is from corticocortical fibres zT. As there are few callosal fibres terminating in the barrel cortex 53, a large proportion of the input to the barrel cor,tex must be from ipsilateral corticocortical ,fibres. Such fibres arising from the mouse barrel cortex have been shown 49 to terminate in columns with a diameter of about 250 t~m. Szentago:thaP 2 has suggested that generally within the neocortex corticocortical fibres terminate in 250 ~m wide columns in an array that is out of register with the thalamic afferents. Therefore the cylinders of high phosphorylase a activity in the somatosensory cortex and perhaps elsewhere may correspond to the columnar terminal fields ofcorticocortical afferents. The hypothesis of two overlapping sets of columns is consistent with the results of degeneration studies in the rat I. Large lesions placed in the upper 3 layers of the barrel cortex produce a uniform degeneration of fibres within layer V. If there were only one set of columns in the barrel cortex then there should be a patchy degeneration in layer V that would be in register with the barrels. However if there are two sets of overlapping columns in the barrel cortex, each with its vertically arranged intracortical connections~ then the layer V degeneration produced by a layer III lesion should be fairly uniform, as observed. As phosphorylase a levels in the brain are related to functional activity, it is necessary to be cautious about assuming that a physiological marker would be able to demonstrate columns which have only been defined on anatomical grounds. However, evidence from rodent ~,2°3~

and m o n k e f 7 somatosensory cortex and cat visual cortex 3~' show that at least in some cases physiologically and anatomically defined columns are the same. Visual cortex

In the visual cortex of a few 'unanaesthetized' mice there were rows of patches of high phosphorylase a activity in layers IV and VI. As the patches were in register, this area of cortex appears to have a columnar organization. This is in accord with results from alert unstimulated monkeys where the levels of [HC]2-DG in the primary visual cortex indicate a clear columnar organization 2L2~. In the monkey these bands of high and low activity may represent orientation columns 22 because these have been demonstrated in the monkey 23 and cat 3~ with the [~4C]2DG technique after 45 min of the appropriate stimulation. In the mouse Drager H found that about 45% of the cells in the primary visual cortex are orientation sensitive, but there was little evidence of these cells being organized into columns. This was not surprising because Towe4~ has shown on theoretical grounds that the small columns found in the mouse brain cannot be demonstrated on electrophysiological grounds alone. Therefore one cannot rule out the possibility of there being columns related to line orientation in the mouse cortex. Most of the rodent primary visual cortex has a monocular input and even in the small lateral area which receives a binocular input Drager ~4 was unable to find any ocular dominance columns. Thus the bands of high phosphorylase a activity which were found largely in parts of the cortex with a monocular input cannot correspond to ocular dominance columns. If, as in the barrel cortex of 'unanaesthetized' mice, the patches of high phosphorylase a activity correspond to the thalamic input then, as the thalamic input to areas 18a and b is from the lateral thalamic nuclei and not the lateral geniculate9, the finding of modules in area 18 implies that these thalamic nuclei and not just the primary sensory relay nuclei may give rise to afferents which terminate in a columnar fashion.

215 Other cortical areas

In 3 of the 'anaesthetized' mice cylinders of high phosphorylase staining were observed in the motor and auditory areas. These cylinders were similar to the ones found in the somatosensory cortex and may also correspond to anatomically defined corticocortical columns 49. There is physiological evidence for a columnar organization of the motor cortex in cats, monkeys4 and rats ~2, as well as direct anatomical evidence for columns produced by corticocortical fibres in the mouse motor cortex 49. Similarly in the auditory cortex there is physiological evidence for a columnar organization in higher mammals 26 and anatomical evidence from the rat for calloREFERENCES 1 Akers, R. M. and Killackey, H. P., Organization ofcorticocortical connections in the parietal cortex of the rat, J. comp. Neurol., 181 (1978) 513-538. 2 Anchors, J. M. and Garcia-Rill, E., Dopamine, a modulator of carbohydrate metabolism in the caudate nucleus, Brain Research, 133 (1977) 183-189. 3 Angel, A., Effect of anaesthetics on nervous pathways. In T. C. Gray, John F. Nunn and J. E. Utting (Eds.), General Anaesthesia, Vol. 1,4th edn., Butterworths, 1980, pp. 11%139. 4 Asanuma, H., Recent developments in the study of the columnar arrangement ofneurones within the motor cortex, Physiol. Rev., 55 (1975) 143-156. 5 Breckenridge, B. M. and Norman, J. H., Glycogen phosphorylase in brain, J. Neurochem., 9 (1962) 383-392. 6 Breckenridge, B. M. and Norman, J. H., The conversion of phosphorylase b to phosphorylase a in brain, J. Neurochem., 12 (1965) 51-57. 7 Browning, M., Bennett, W. and Lynch, G., Phosphorylase kinase phosphorylates a brain protein which is influenced by repetitive synaptic activation, Nature (Lond.), 278 (1979) 273-275. 8 Browning, M., Bennett, W. F., Kelly, P. and Lynch, G., Evidence that the 40,000 Mr phosphoprotein influenced by high frequency synaptic stimulation is the alpha subunit of pyruvate dehydrogenase, Brain Research, 218 (1981) 25~266. 9 Caviness, V. S. Jr. and Frost, D. O., Tangential organization of thalamic projections to the neocortex in the mouse, J. comp. NeuroL, 194 (1980) 335- 367. 10 Cipolloni, P. B. and Peters, A., The bilaminar and banded distribution of the callosal terminals in the posterior neocortex of the rat, Brain Research, 176 (1979) 3347. 11 Cohen, P., Protein phosphorylation and the co-ordinatd control of intermediary metabolism. In P. Cohen (Ed.),

Molecular Aspects of Cellular Regulation, Iiol. 1. Recently Discovered Systems of Enzyme Regulation by Reversible Phosphorylation, Elsevier/North-Holland, Amsterdam, 1980, pp. 255-268.

sal afferents terminating in discrete strips ~°. ACKNOWLEDGEMENTS

I wish to thank Dr. N. C. Spurway for teaching me about histochemistry and pointing out that phosphorylase a is a marker of immediate metabolic history in muscle, and my supervisor Dr. O. Holmes for much help and encouragement. The cryostat system was partially funded by SERC Grant B187818 to Dr. N. C. Spurway. M.N.W. was supported by a Glasgow University Faculty of Science Postgraduate Research Scholarship.

12 Collins, R. C., Kennedy, C., Sokoloff, L. and Plum, F., Metabolic anatomy of focal motor seizures, Arch. Neurol., 33 (1976) 536-542. 13 Dietrich, W. D., Durham, D., Lowry, O. H. and Woolsey, T. A., Quantitative histochemical effects of whisker damage on single identified cortical barrels in the adult mouse, J. Neurosci., 1 ( 1981) 929- 935. 14 Drager, U. C., Observations on monocular deprivation in mice, J. Neurophysiol., 41 (1978) 28-42. 15 Drummond, G. I. and Bellward, G., Studies on phosphorylase b kinase from neural tissues, J. Neurochem., 17 (1970) 475-482. 16 Durham, D. and Woolsey, T. A., Barrels and columnar cortical organization; evidence from 2-deoxyglucose (2DG- experiments, Brain Research, 137 (1977) 169- 174. 17 Edwards, C., Nahorski, S. R. and Rogers, K. J., In vivo changes of cerebral cyclic adenosine-3',5'-monophosphate induced by biogenic amines: association with phosphorylase activation, J. Neurochem., 22 (1974) 565572. 18 Folbergrova, J., Changes in glycogen phosphorylase activity and glycogen levels of mouse cerebral cortex during convulsions induced by homocysteine, J. Neurochem., 24 (1975) 15-20. 19 Greengard, P., Cyclic Nucleotides, Phosphorylated Proteins, and Neuronal Function, Raven Press, New York, 1978. 20 Hand, P. J., The 2-deoxyglucose method. In L. Heimer and M. J. Robards (Eds.), Methods in Contemporary Neu-

roanatom)2: The Tracing of Central Nervous Pathways, Plenum, New York, 1981. 21 Hendrickson, A. E. and Wilson, J. R., A difference in [Inc]deoxyglucose autoradiographic patterns in striate cortex between Macaca and Saimiri monkeys following monocular stimulation, Brain Research, 170 (1979) 353358. 22 Hubel, D. H. and Wiesel, T. N., Receptive fields and functional architecture of monkey striate cortex, J. Physiol. (Lond.), 195 (1968) 215-243. 23 Hubel, D. H., Wiesel, T. N. and Stryker, M. P., Anatomical demonstration of orientation columns in Macaque monkey, J. comp. Neurol., 177 (1978) 361- 379.

216 24 Ibrahim, M. Z. M., Pascoe, E. and Necco Khavat, M. Y.. Histochemical evidence for phosphorylasc, branching enzyme and glycogen synthetase activities in rat brain, ./. neurol, Sci.,19(1973) 117 131. 25 Ibrahim, M. Z. M., Pascoe, E., Alam, S. and Miquel, J,, Glycogen and phosphorylase activity in rat brain during recovery from several forms of hypoxia, Amer..1. Path., 60 (1970) 403- 420. 26 Imig, T. J. and Adrian, H. O.. Binaural columns in the primary field (A I) of cat auditory cortex, Brain Research, 138(1977)241 257. 27 Jones, E. G., Anatomy of cerebral cortex: columnar input output organization. In F. O. Schmitt, F.G. Worden, G. Adelman and S. G. Dennis (Eds.), The Organication o f the Cerebral Cortex. M.I.T. Press, London, 1981. pp. 199 235. 28 Keller, M. and Tanka, D., Preservation of the iodine glycogen complex of the phosphorylase reaction, Acta histochem., 22 (1965) 73 76. 29 Kennedy, C., Gillin, J. C., Mendelson, W., Suda, S., Miyaoka, M., Ito, M., Nakamura, R. K., Storch, F. I.. Pettigrew, K., Mishkin, M. and Sokolofl\ k., Local cerebral glucose utilization in non-rapid eye movement sleep, Nature (Lond.). 297 (1982) 325- 327. 30 Killackey, H. P. and Belford, G. R., The formation of afferent patterns in the somatosensory cortex of the neonatal rat, J. comp. Neurol., 183 (1979) 285-304. 31 Kogure, K. and Alonso, O. F., A pictorial representation of endogenous brain ATP by a bioluminescent method, Brain Research. 154 (1978) 273 284. 32 Lowry, O. H., Passonneau, J. V., Hasselberger, F. X. and Schulz, D. W., Effect of ischaemia on known substrates and cofactors of the glycolytic pathway in brain. J. biol. Chem., 239(1964) 18 30. 33 Mcllwain, H., Extended roles in the brain for secondmessenger systems, Neuroscience, 2 (1977) 357 372. 34 Meijer, A. E. F. H., Improved histochemical method for the demonstration of the activity of (,-glucan phosphorylase. I The use of glucosyl acceptor dextran, Histochemie, 12(1968)244 252. 35 Moffett, D. F. and LaManna, J. C., Contributions ofglycolysis and oxidative metabolism to recovery from electrical pulses in the isolated toad brain, Brain Research. 152(1978)365 368. 36 Mountcastle, V. B., An organizing principle for cerebral function: the unit module and the distributed system. In G. M. Edelman and V. B. Mountcastle (Eds.), The Mindful Brain. Cortical Organization and the Group-Selective Theory o f Higher Brain Function, M.I.T. Press, 1978, pp. 7 50. 37 Ozawa, E., Activation of phosphorylase kinase from brain by small amounts of Ca -+, J. Neurochem., 20 (1973) 1487 1488. 38 Schoppmann, A. and Stryker, M. P., Physiological evidence that the 2-deoxyglucose method reveals orientation columns in cat visual cortex, Nature (l, ond.), 293 (1981)574 576.

39 Simons, D. J. and Wool>c\. I . A., Functional ~)rganization in mouse barrel cortex. Brain Reseat Rosters. M.H., Patlak, C. S.. Pettigrew. K, I).. Sakurada, O. and Shinohara, M., The [laC]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure and normal values in the conscious and anaesthetized albino rat, ,I. ,\eurochem.. 28 (1977) 897 916. 41 Swanson, M. A., Studies on the structure of polysaccharides. IV. Relation of the iodine color to the structure. ,1. biol. Chem., 172 (1948) 825 837. 42 Szentagothai, J., The Ferrier Lecture. 1977. The neuron network of the cerebral cortex: a functional interpretalion, Proc. roy. Soc. 13, 201 (I978)219 248. 43 Takeuchi. T,, Histochemistr'r ofintracellular polysaccharide synthesis, Ann. Histochim.. 7 (1962) 61 80. 44 Thompson, C, I., Rubio, R. and Berne, R, M.. Changes m adenosine and glycogen phosphorylase activity during the cardiac cycle, Amer..I. Phvviol., 238 (198()) H389 398. 45 Towe. A. L., Notes on the hypothesis of columnar organization in somatosensory cerebral cortex. Brain. Behav. Evol., II (1975)16 47. 46 Wallace. M. N. and Holmes. O.. Early histochemical changes in epileptogenesis in the rat. Electroenceph, olin. Neuropl~vsiol., 50 (1980) 231P. 47 Wallace. M. N., Modular organization of phosphorylase a activity in the rodent hippocampal region, ,1, Physiol. (Lond.), 329 (1982) 32P. 48 Welker, C., Receptive fields of barrels in the somatosensorv neocortex of the rat, J. comp. Neurol.. 166 (1976)

17~ 190. 49 White, E. L. and De Amieis, R. A., Afferent and efferent projections of the region in mouse Sm I cortex which contains the posteromedial barrel subfield. J. comp. Neurol., 175 (1977)455 482. 50 White. E, L., Identified neurones in mouse Sm I cortex which are poslsynaptic to thalamoeortical axon terminals: a combined Golgi-electron microscopic and degeneration study, J. comp. Neurol., 181 (1978) 627 662. 51 Wong-Riley, M. T. T. and Welt, C., Histochemical changes in cytochrome oxidase of cortical barrels after vibrissal removal in neonatal and adult mice, Proc. nat. Acad. Sci. U.S.A.. 77 (1980) 2333-2337. 52 Woolsey, T.A. and Van der Loos, H., The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex: the description of a cortical field composed of discrete cytoarchitectonie units, Brain Research, 17 (1970) 205 242. 53 Yorke, C. H., Jr. and Caviness, V. S., Jr., Interhemispheric neocortica[ connections of the corpus callosum in the normal mouse: a study based on anterograde and retrograde methods, J. comp. NeuroL. 164 (1975) 233 246.