Differential expression of distinct microtubule-associated proteins ...

Proc. Nati. Acad. Sci. USA Vol. 82, pp. 6006-6009, September 1985

Neurobiology

Differential expression of distinct microtubule-associated proteins during brain development (microtubules/neuronal differentiation/cytoskeleton/monoclonal antibodies/dendrite growth)

BEAT RIEDERER AND ANDREW MATUS Friedrich Miescher Institute, P.O. Box 2543, 4002 Basel, Switzerland

Communicated by Philip Siekevitz, May 6, 1985

trace the appearance of MAPi, -2, and -3 in rat brain from the late embryo to the end of the third postnatal week. We find that each MAP shows a distinct developmental pattern, which includes large changes in the abundance of their components that occur coordinately throughout the brain at particular times during development. In the case of both MAP2 and MAP3, there is a striking transition in the pattern of expression between postnatal days 10 and 20 (P10 and P20).

The levels of three different microtubuleABSTRACT associated proteins (MAP1, -2, and -3) in brain were found to undergo large changes during postnatal development. MAPi was barely detectable at birth but thereafter steadily increased, reaching adult levels by postnatal day 20 (P20). Both MAP2 and MAP3 showed differential expression patterns of their component peptides. At birth, MAP2 was represented by the smaller of two Mr 280,000 peptides (MAP2b) and three antigenically related Mr 70,000 peptides. The larger of the M, 280,000 peptides (MAP2a) first appeared between P10 and P20, and the Mr 70,000 components disappeared at the same time. Of the two MAP3 peptides, the larger (MAP3a) was present in the late embryo, several days before MAP3b appeared. Between P10 and P20, both MAP3 components underwent a striking decrease in abundance (a factor of 10), which correlated with their disappearance from all neuronal compartments except neurofrlament-containing axons. These developmental changes in expression are different and characteristic for each of the three MAPs, yet in each case they are detectable in brain homogenates, indicating that they occur concurrently throughout the brain.

MATERIALS AND METHODS Microtubules were prepared from the brains of albino RA25 strain rats (CIBA-Geigy) by the method of Karr et al. (20). Supernatant fractions were prepared by homogenizing brains in ice-cold buffer A (20) and centrifuging at 105,000 x g for 60 min to pellet solids. Either microtubule or supernatant proteins were separated on 3.6-12% NaDodSO4/polyacrylamide gels, blotted onto nitrocellulose, and stained for either one of the MAP proteins with monoclonal antibodies and peroxidase-coupled second antibody (21). The monoclonal antibodies were as described (10, 12) and were used at 10 pug/ml. Rat brain tissue was fixed by perfusion with formaldehyde and glutaraldehyde as described (12) and cut into 40-.um sections with a Vibratome. The immunoperoxidase staining of these sections with monoclonal antibodies followed the protocol described in ref. 12. Changes in the relative amount of each MAP peptide during development were determined by scanning antibodystained NaDodSO4 gel blots of supernatant or homogenate proteins and measuring the peak height for each band using a reflectance densitometer (Camag, Muttenz, Switzerland). Dilution series of the proteins were separately scanned by immunodot assay (22) to establish that the measurements made with each antibody lay in the linear range of the logarithmic relationship between densitometric peak height of the immunoperoxidase-stained samples and the amount of antigenic protein attached to the nitrocellulose (23). Two series of neonatal supernatants and one homogenate series were measured in this way for MAPi and -3 and three supernatant and one homogenate series were measured for MAP2. None of the MAPs was detectable in the cold pellets so that the supernatants are representative of total brain MAP. All sets gave congruent results with respect to changes in the relative amounts of each MAP during development but, for technical reasons, the absolute amounts differed slightly between sets. Therefore, the results from the supernatant gels were normalized with respect to the day 0 values for MAP 1, MAP2b, and MAP3a, and the means of these normalized values are presented here (Table 1).

The assembly of microtubules is an essential step in the growth of neuronal processes (1-3). Although very little is known about its regulation in the developing nervous system, it is believed that microtubule-associated proteins (MAPs) are involved (4-9). Three MAP species have been shown by immunohistochemistry to be differentially distributed in the brain, differing both in the types of cells in which they occur and in their distribution within the neuronal cytoplasm (10). MAPi is, according to our observations, associated exclusively with neurons in the adult brain and is more abundant in dendrites than in axons (11-13). MAP2 is also neuron specific and is predominantly, and probably exclusively, located in dendrites (12, 14-18). MAP3 occurs in both neurons and glia, but in neurons it is restricted to neurofilament-rich axons (10, 19). In a recent immunohistochemical study, we followed the appearance of these proteins in the developing rat cerebellum and found that each exhibits a characteristic developmental pattern (26). MAPi appears in axons before dendrites, but the balance changes during the first 3 postnatal weeks by the gradual accumulation of MAPi in dendrites and by its simultaneous decrease in axons. MAP2 is exclusively associated with dendrites throughout development. MAP3 appears transitorily in nascent axons (from which it is absent in mature tissue) and later appears in glial cell bodies and processes. These changes in cellular distribution indicate that the expression of these proteins is developmentally regulated. In the present study, we have used monoclonal antibodies to The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Abbreviation: MAP, microtubule-associated protein(s).

6006

Neurobiology: Riederer and Matus

Proc. Natl. Acad. Sci. USA 82 (1985)

Table 1. Changes in the relative amounts of MAP1, -2, and -3 in brain at different postnatal ages P5 PO P10 P15 P20 ad MAP1 1 2.15 4.34 6.45 11.20 7.73; MAP2a 0.57 0.89 0.67 MAP2b 1 0.98 1.11 1.07 1.11 0.88 MAP2 0.15 0.07 (MA, 70,000)* 0.84 0.71 0.63 0.58 MAP3a 1 0.88 0.67 0.10 0.05 0.03 MAP3b 0.63 0.85 1.05 0.12 0.04 0.03 Ages are at 5-day time points from birth (P0) until P20 and for adult (ad). *Values are for the largest of the three M, 70,000 MAP2 peptides.

MAP 1 5 a

iiw~i o

k Yi

~w

e nb 3 5 1015 20 a

~. --.1

---a

6 ow

350

--.280

-.180

FIG. 1. Samples of protein (10 1Lg) from microtubules of 5-day-old (lanes 5) or adult (lanes a) brain run in adjacent slots of NaDodSO4 gels, blotted onto nitrocellulose sheets, and stained with monoclonal antibodies against MAP1, MAP2, or MAP3 as indicated. The apparent molecular weight (x 10-s) for each of the major species is indicated on the right.

to be lower in adult brain (by a factor of >10). We established that, like MAP2 itself, these Mr 70,000 peptides coassemble with microtubules through repeated cycles of microtubule assembly and, like MAP2a and -b, they are resistant to heat treatment (24). We also considered the possibility that the presence of these lower MAP2-like components and the absence, at the same period of development, of the upper MAP2a band might be linked by the presence in neonatal brain of a protease that selectively degrades the MAP2a protein. To determine whether this was so, we exposed adult brain microtubules (which contain both upper bands but almost none of the Mr -70,000 material) to neonatal brain supernatant (which should contain the putative protease). Under conditions where adult brain supernatant induces extensive degradation of both high molecular weight MAP2 components (24), neonatal supernatant produced no breakdown of the MAP2a band and no increase in the Mr 70,000 material (data not shown). In contrast to MAP1 and -2, MAP3 is far more abundant in newborn brain than in the adult (Fig. 2 Right). Only the larger of the two MAP3 peptides (MAP3a) was present in the embryonic brain but by birth the smaller component (MAP3b) had also appeared. Thereafter, the levels of MAP3b steadily increased until by day 10 it was more abundant than MAP3a (Fig. 2, Table 1). After this, there was a sudden decline in the levels of both MAP3 peptides so that by day 15 e nb 3 5 10 1520 a

-2a

3a__ . 3b-

_

*

-

-.70

MAPMP1MAP 1 2

MAP3 5 a

w

RESULTS Comparing microtubules from 5-day-old brain to those from adult revealed that the MAP1 content was distinctly lower in microtubules of juvenile tissue (Fig. 1). MAP2 levels were also higher in adult than in juvenile brain, but the change was less striking than for MAP1. MAP3 showed the opposite pattern, being much more abundant in young brain microtubules than in those from mature rats. Developmental Time Course of MAP Peptides in Brain. To determine whether these changes in MAP levels in microtubules reflected their changing abundance in brain, we used the monoclonal antibodies to selectively label each individual MAP in brain homogenates and supernatant fractions, obtaining essentially the same results in each case. Data for the supernatant fractions are given here (Fig. 2, Table 1). Anti-MAPl-stained blots (Fig. 2 Left) showed a steady accumulation of this protein from very low levels in the embryonic and newborn brain to a peak, reached at around P20 when it was >10-fold more abundant than at postnatal day 0 (P0) (Table 1). The lower molecular weight anti-MAPi cross-reactive peptides, which are proteolytic fragments of the native Mr 350,000 band (24), appeared concurrently with the higher molecular weight MAP1 protein. MAP2 presented the most complicated pattern of the proteins studied here. As previously reported (15, 25), the smaller of the two Mr 280,000 peptides (MAP2b) appeared first, and it was present at constant levels between birth and P20. The larger MAP2a component first appeared during the second postnatal week, and it continued to increase until P20 (Fig. 2 Center). In addition, we also observed complimentary changes in lower molecular weight anti-MAP2 reactive components (Mr, -70,000), which appeared as a major band and two smaller bands. These were present at comparable levels from 19-day embryo to P10, after which they rapidly declined e nb 3 5 10 15 20 a

MAP2 5 a

6007

MAP 3

FIG. 2. The time course of appearance of MAP1, -2, and -3 revealed by staining NaDodSO4 gel blots of supernatant fraction proteins from rat brains of different ages with monoclonal antibodies. Congruent results were obtained with brain homogenate (data not shown). All three blots are identical, with each slot containing 40 ,g of total protein. They differ only in the antibody with which they were stained. MAP1, MAP2a and -2b, the Mr 70,000 MAP2 components (70), and MAP3a and -3b are indicated. Lane numbers give the postnatal age in days; e, embryonic day 19; nb, newborn; a, adult.

6008

Neurobiology: Riederer and Matus

1.

W

S

(1985)

Proc. NatL Acad. Sci. USA 82

t

M

F '''

i

''

(t;I' {'e't;L'ssi'h''$" 1 '4l

W-l

- l'

.

klp~s

A

d

. .; ~~

sx

*

* tsFF

;oair-

+

^

^~~~~~~~~~~~~~~~~~~~~~~~~

*S. -byi

I

o

]

-

%

z .

s

a

g*

14 0

I' a

I

.Al

20d

I

i

FIG. 3. Immunohistochemical localization of MAP1, -2, and -3 (top, middle, and bottom row, respectively) at P5, P10, and P20 (left, center, and right columns). All the micrographs are from 40-Mtm thick sections, which in each category were cut from a single tissue block and processed together in common reagents except for the use of one or another specific monoclonal antibody against each MAP species. (X85.)

they were 1/10th as abundant, a situation that persisted in the adult brain. We considered the possibility that the absence of the MAP3b band from embryonic brain might be the result of a protease that selectively degrades this peptide. However,

neonatal brain supernatant failed to induce degradation of this protein (data not shown). Cytological Features of MAP Development. We used immunoperoxidase histochemistry to examine the pattern of ap-

Neurobiology: Riederer and Matus of these three MAP proteins in developing brain. The results show a strong correspondence between the cytological pattern of appearance and the changing levels of expression for each of the three MAPs. In the cerebral cortex (Fig. 3), anti-MAP1 staining is barely detectable at postnatal day 5, has increased by day 10, and shows a further increase by day 20, at which stage the staining pattern is that of the adult. At all stages, only neurons exhibit the antigen. As in the cerebellum (26), MAP2 in the cerebral cortex was exclusively associated with neuronal dendrites throughout development (Fig. 3, middle row). During the most active phases of dendrite growth, the MAP2 antigen is concentrated in the distal branches of the apical dendrites, in a fashion similar to that observable in developing Purkinje cells (10, 26), whereas in the mature cortex MAP2 is apparently evenly distributed throughout the dendritic tree. MAP3 shows the most dramatic developmental change. In the newborn cortex and subsequently up until day 10, it is prominently expressed in neurons where, like MAP2, it is especially concentrated in the terminal branches of apical dendrites (Fig. 3, bottom row). Thereafter, neuronal antiMAP3 staining declines rapidly, so that by day 20 the glial staining that characterizes the adult is the major feature. There is still a trace of neuronal MAP3 present at this stage, but it is completely absent from the adult cortical neuropil (19).

pearance

DISCUSSION Each of the microtubule proteins dealt with here is associated with a different set of cellular compartments in the adult brain; MAP2 with dendrites, MAP1 with axons and dendrites, and MAP3 with axons and glia (10). Their changing patterns of abundance in the neonatal rat brain reflect the morphological changes occurring during this period of rapid development, but not in a simple manner. For example, MAP1, which is most concentrated in neuronal dendrites in the adult brain, makes its major increase in abundance between P10 and P20, when dendrite growth is already well advanced. This is also the time at which the larger MAP2 component (MAP2a) first appears and accumulates (see also refs. 15 and 25). Although it has been shown that the phosphorylation state of MAP2b alters during development, neither the kinetics of phosphorylation nor the change in electrophoretic migration pattern it induces (25) are sufficient to account for the appearance of MAP2a. We also observe lower molecular weight (Mr -70,000) peptides antigenically related to but not derived from the high molecular weight MAP2 components, which are abundant in embryonic and newborn brain and disappear between days 10 and 20. All these MAP2-related proteins selectively coassemble with microtubules and thus also appear to possess a tubulinbinding site as a common feature. Their number and large differences in size make it unlikely that they represent post-translational modifications of a single basic gene product but, rather, suggest that the levels at which they are synthesized are regulated during development. Between postnatal days 10 and 15, MAP3 levels plunge, and this coincides with its disappearance from neurons of the cerebral cortex. During this period, it also disappears from cerebellar granule cells in which it occurs at high levels in the cell bodies and axons during early development (26). In the adult brain, it is a prominent feature of glia and the only neuronal compartments in which it remains present are axons rich in neurofilaments (10, 19). Both our earlier cytological observations and the biochemical results reported here are thus consistent with the involvement of MAP3 in early development of neuronal compartments from which it is absent in mature tissue. Of the two MAP3 peptides, only the

Proc. Natl. Acad. Sci. USA 82 (1985)

6009

upper component appears prior to birth, suggesting further differentiated aspects in the involvement of this protein in brain maturation. However, it is not the case that one of the MAP3 peptides is expressed in neurons and the other in glia, because (i) both MAP3a and MAP3b fall in abundance concurrently with the enormous drop in neuronal expression and (ii) both neuronal and glial cell lines express both MAP3 peptides (19). The results presented here illustrate two significant features of the relationship of microtubular proteins to neuronal development in the brain. First, it is clear that large developmentally regulated changes occur in the expression of different MAP species. Some of these, such as MAP1 and MAP2a, can be categorized as "late," whereas others such as MAP3 and the M, 70,000 MAP2-related material are "early." Second, these changes are detectable in brain homogenates, so they must be occurring simultaneously in many cells throughout the brain. Third, they take place at particular times and, for MAP2 and MAP3, they are particularly striking between P10 and P20. This period thus seems to constitute a transition in microtubule composition. There is no obvious cytological event to which these changes can be related, but the degree of change suggests that a subtle yet profound change takes place in the developing brain during this short period. 1. Daniels, M. P. (1972) J. Cell Biol. 53, 164-176. 2. Seeds, N. W., Gilman, A. G., Amano, T. & Nirenberg, M. W. (1970) Proc. Natl. Acad. Sci. USA 66, 160-167. 3. Yamada, K. M., Spooner, B. S. & Wessels, N. K. (1970) Proc. Natl. Acad. Sci. USA 66, 1206-1212. 4. Dentler, W. L., Granett, S. & Rosenbaum, J. L. (1975) J. Cell Biol. 65, 237-241. 5. Fellous, A., Francon, J., Lennon, A.-M. & Nunez, J. (1977) Eur. J. Biochem. 78, 167-174. 6. Herzog, W. & Weber, K. (1978) Eur. J. Biochem. 92, 1-8. 7. Murphy, D. B. & Borisy, G. G. (1975) Proc. Natl. Acad. Sci. USA 72, 2696-2700. 8. Sloboda, R. D., Rudolph, S. A., Rosenbaum, J. L. & Greengard, P. (1975) Proc. Natl. Acad. Sci. USA 72, 117-181. 9. Weingarten, M. D., Lockwood, H., Hwo, S.-Y. & Kirschner, M. W. (1975) Proc. Natl. Acad. Sci. USA 72, 1858-1862. 10. Matus, A., Huber, G. & Bernhardt, R. (1983) Cold Spring Harbor Symp. Quant. Biol. 48, 775-782. 11. Bloom, G. S., Schonfeld, T. A. & Vallee, R. B. (1984) J. Cell Biol. 98, 320-330. 12. Huber, G. & Matus, A. (1984) J. Neurosci. 4, 151-160. 13. Huber, G. & Matus, A. (1984) J. Cell Biol. 98, 777-781. 14. Bernhardt, R. & Matus, A. (1984) J. Comp. Neurol. 226, 203-219. 15. Burgoyne, R. D. & Cumming, R. (1984) Neuroscience 11, 157-167. 16. Caceres, A., Binder, L. I., Payne, M. R., Bender, P., Rebhun, L. & Stewart, L. (1984) J. Neurosci. 4, 394-410. 17. De Camilli, P., Miller, P. E., Navone, T., Theurkauf, W. E. & Vallee, R. B. (1984) Neuroscience 11, 819-846. 18. Wiche, G., Briones, E., Hirt, H., Krepler, R., Artlieb, U. & Derk, H. (1983) EMBO J. 2, 1915-1920. 19. Huber, G., Alaimo-Beuret, D. & Matus, A. (1985) J. Cell Biol. 100, 496-507. 20. Karr, T. L., White, H. D. & Purich, D. L. (1979) J. Biol. Chem. 254, 6107-6111. 21. Matus, A., Pehling, G., Ackermann, M. & Maeder, J. (1980) J. Cell Biol. 87, 346-359. 22. Hawkes, R., Niday, E. & Gordon, J. (1982) Anal. Biochem. 123, 143-147. 23. Towbin, H. & Gordon, J. (1984) J. Immunol. Methods 72, 313-340. 24. Matus, A. & Riederer, B. (1985) Ann. N.Y. Acad. Sci., in press. 25. Binder, L. I., Frankfurter, A., Kim, H., Caceres, A., Payne, M. R. & Rebhun, L. I. (1984) Proc. Natl. Acad. Sci. USA 81, 5613-5617. 26. Bernhardt, R., Huber, G. & Matus, A. (1985) J. Neurosci. 5, 977-991.