Implications of prolonged expression of Fos-related antigens

R

Implicationsof prolonged expressionof Fos-related antigens Keith Il. Pennypacker, Michael K. McMillian

Jau-S. Hong and

The AP-1 transcription

factors are composed of the

Fos and Fos-related

antigens as well as Jun and

related proteins. These factors have been extensively studied in many diverse paradigms

using acute stimuli.

Recent attention has focussed on long-term elevation of Fos-related

antigens in the CNS, and

this is discussed by Keith Pennypacker, and Michael treatment

McMillian.

Jaws.

Hong

Repeated or chronic

elevates Fos-related

antigen levels for

days in many different brain regions. Both direct and indirect stimulation are responsible increase

in Fos-related

for the protracted

antigen-immunoreactive

proteins, which may modulate late onset genes involved in neuroplasticity. these factors in long-lasting

Understanding or permanent

the role of disease

states may provide insight into potential therapeutic strategies to treat chronic CNS disorders.

The AP-1 transcription factors have been extensively studied in many different tissue types in a diverse array of experimental paradigms and are implicated in the modulation of most genes. This family is composed of two groups of proteins, Fos-related antigens and the Junrelated factors’. An AI’-I-DNA binding complex requires dimerization involving a leucine zipper: a Fos-related antigen requires a Jun to form a dimer while Jun proteins can form homo- or heterodimers with other Jun-related proteins. Other leucine zipper-containing transcription factors, such as CAMP responsive element binding protein, also interact with Jun-related factors and can still bind to Al’-1 sites in gene promoterP. Thus, one level of binding specificity is conferred by the various dimer combinations. In the brain, Al’-1 transcription factors are induced by a variety of stimuli. Seizure activity dramatically increases AP-l-DNA binding activity in the hippocampus and other brain regions@, whereas in the striatum and nucleus accumbens, dopamine receptor agonists enhance the expression of these transcription factors, suggesting a role for these factors in drug abuses15. The induction of these DNA binding proteins modifies gene expression in response to a changing environment. It has been proposed that .AP-1 factors initiate genomic programmes for functions such as cell deathlGl9, and learning and memory*O.

0

1995,

Elsevier

Science

Ltd

E

V

I

E

W

Earlier studies primarily focused on the acute response of AP-1 factors. In particular, Fos, the most extensively studied factor, is dramatically increased in response to most stimuli. The expression is usually transient, with levels of Fos-related antigens returning to baseline within several hours. However, recently more attention is being directed to long-term expression of transcription factors that are presumably involved in protracted or permanent changes in the genomic programming of cells. Morgan, Curran and co-workers pioneered the field of Al?-1 transcription factors in the CNS by reporting that c-jos is induced in the hippocampus due to seizure activity4-6.In later reports using an antibody produced against a conserved peptide sequence in Fos-related antigen proteins, the induction of several Fos-immunoreactive protein bands was described in the rodent brain after treatment with seizure-inducing compounds. Some of these Fos-related antigen proteins were cloned including Fos-related antigen 1, Fos-related antigen 2 and FosB2*-21; however, the identities of several long-term Fos-related antigens remain uncertain. The timecourse of expression for these Fos-related antigen proteins differed after treatment with seizureinducing compounds. Excitatory amino acid receptor agonists, especially kainate, prolonged the expression, while the convulsant pentylenetetrazol caused transient inductior@. While the expression of Fos protein generally declined early, using transgenic rodents, Fos expression has been shown recently to be extended for two to three days in dying neurones 17,18. Fos-related antigen proteins with molecular weights of 46 kDa and 35 kDa in particlllar maintain an extended expression. This observation led to the hypothesis that different Fos-related antigen-Jun heterodimer combinations activated a long-term genomic programme that is distinct from the genes modulated by the acute factors. There has been much speculation on the identities of these proteins, but the determination of their exact nature has remained elusive. Recent reports have focused on the protracted expression of Fos-related antigen proteins after repeated or chronic treatments, particularly protein bands in the 35 -37 kDa molecular weight range9,10,12,1~24,~ (see Table 1). The regulation of long-term factors differs from wellcharacterized proteins, such as Fos, which undergo a refractory period during inductionQ6. The prolonged elevation of these factors implies modulation of Al’-I. directed gene expression related to chronic conditions such as brain injury and drug sensitization and tolerance caused by drugs of abuse.

K It. Pmqp~cbr. AssIstant Rofessor, Depal7ment of Pharmacology and Therapeutrs. Unwwty of South Flmda. 12901 Bruce B DMwls Boulevard MDC Box 9. Tampa, FL 3X12-4799. USA, J-S. Hong,

Regulation by dopamine receptor agonists The striatum contains dopaminergic terminals originating from the substantia nigra. Dopamine regulates gene expression in striatal neurones, tonically inhibiting genes, such as that for enkephalin through D, receptors, and activating genes for the peptides prodynorphin and tachykinin via D, receptors*+30.Interestingly, both the D, and D, receptor systems are involved in the regulation of

TiPS

- September

1995

(Vol.

16)

Actmg Head, and M. IL McMillim. Senw Fellow. LaL!oratoryof Envlmnmental Neumwence.

National

lnstltute of Enwmnmental Health Sciences, WI Box 12233.111 Alexander Drive. Research Triangle Park. NC 27709. USA

317

R

E

E

I

V

W week of twice daily treatments. Chronic treatment paradigms using cocaine, which indirectly effect dopamine receptor activation, also increase the amount of these Fosrelated antigen proteinW3. Depletion of striatal dopamine by injecting 6-hydroxydopamine into the substantia nigra also increases Fosrelated antigen expressionsi. Immunoreactivity in striatal neurones is observed for at least three months after denervation, and mimicking denervation by administration of D, receptor antagonists similarly increases Fos-related antigen immunoreactivityis. The putative mechanism is by the elimination of tonic inhibition through D, receptors. However, elimination of the input pathways to neurones may also influence Fos-related antigen protein expression, which will be discussed later.

Table 1. Neural tissues that show long-term Fos-related antigen expression Neural tissue

Stimulus

Refs

Cortex

Chronic electroshock Mechanical injury Chronic electroshock Kainate Chronic dehydration Kainate Chronic cocaine administration Chronic apomorphine administration Oenervation Repeated nicotine administration Repeated stress

25 45-48 25 24 IO,36 32 10 9 31 34 35

Hippocampus Hypothalamus Olfactory bulb” Striatum

AdrenaP

aOlfactory bulb and adrenals have high basal levels of Fos-related antigen expression.

I

Fos-related antigen-immunoreactive proteins. Repeated stimulation of D, receptors significantly augments Fos-related antigen expressiong. While Fos is increased within one hour after drug administration, it becomes undetectable within 24h. In contrast, a Fosrelated antigen-immunoreactive protein is increased maximally within 24 h and remains at this level after one

Seizure activity Chemically and electrically induced seizure activity greatly increases the content of Fos-related antigenimmunoreactive proteins, with the 35 kDa Fos-related antigen having a protracted expression. Repeated electroshock, but not acute treatment, similarly increases the levels of Fos-related antigen-immunoreactive proteins in several brain areaszs.After one week, the amounts of these Fos-related antigen proteins decline by 50%. In this

b 1

2

3

4

5

6

1

2

3

4

5

6

-

55kDa

AP-1 DNA binding

Fig.1.Long-term expression of Fos-related antigen after kainate in the rat hippocampus. a:Very little Fos-related antigen immunoreactivity is observed in the dentate gvrus (DG) of control rats (left panel), while Fos-related antigen immunoreactivity is elevated in the DG one week after a single injection of kainate fright panel). Note the loss of the neuronal layers in the CA1 and CA3 regions. b,c: Timecourse of AP-I-DNA binding and Fos-related antigen expression in the rat hippocampus after kainate treatment. Nuclear extracts from rat hippocampi after saline injection (lane I), 4 h (lane 2). 1 day (lane 3). 3 days (lane 4). 7 days (lane 5) or 2 weeks (lane 6) after kainate administration were analysed for AP-I-DNA binding with gel shift assay(b) or with western blots probed with Fos-related antigen antibodies [c). Reproduced with permission from Ref. 24.

3

1 8

Tip.5 -September

1995 (Vol. 16)

R model, two Fos-related antigen-immunoreactive bands at 35 and 37kDa can be discerned, and the former is the longer lasting of the two. Kainate causes reliable and recurrent seizures for several hours. As a result of these repeated convulsive episodes, c-fos and Fos-related antigen proteins are induced for at least six hours in the rat hippocampus7. At three days, only the 35 kDa Fos-related antigen can be detected (Fig. l), and it continues to be expressed for at least three months32. Pentobarbital pretreatment effectively blocks induction of these factorP; however, treatment with pentobarbital several days after kainate administration has no effect on levels of the Fos-related antigen@. Subconvulsant doses of kainate increase the 35 kDa Fos-related antigen without inducing any other Fos-related antigen proteins:. The initial induction is linked to stimulation by seizure activity but long-term expression is correlated with long-lasting, or possibly permanent, changes in the brain. A second action of kainate is to induce neuronal death in several brain regions, including entorhinal cortex and CA3, hilar and CA1 regions of the hippocampus33. The neurones of the dentate gyrus lose their input from entorhinal cortex and CA3 region as well as their targets

E

v

I

E

w

in the hilar and CA3 regions. Fos-related antigen protein(s) of the 35 kDa band is increased in the hippocampus and olfactory bulb for at least three months after kainate administration, whereas it is only transiently expressed in other brain region@. This lotig-lasting, perhaps permanent, elevation of Fos-related antigen immunoreactivity in the neurones of the dentate gyrus appears to be connected to neuronal loss, particularly of inhibitory tone from the hilar neurones (Fig. 2). Thus, kainate is inducing this protein initially as a result of seizures, and the prolonged Fos-related antigen expression correlates with removal of either input or output (or both).

Other chronic treatment paradigms The adrenal medulla contains high basal levels of 35 and 40 kDa Fos-related antigen protein+. Since mRNAs for known Fos-related antigens were not detected by northern blot analysis, the 35 and 40 kDa proteins may be novel factors. Stimulation by acetylcholine released from the splanchnic nerve may be responsible for the high basal Fos-related antigen levels in the adrenal medulla, but further experiments are necessary to test whether loss of stimulation affects these transcription factors. Repeated

direct stimulation

indirect stimulation

1 neurochemical

1 disinhibition 2 disruption of circuit

2 seizure

kainate receptor (+)

Fig. 2. tong-term Fos-related antigen expression is stimulated by both direct and indirect stimulation. Left side shows prolonged receptor activation leading to protracted elevation of Fos-related antigen-immunoreactive proteins increasing AP-1 stimulated gene expression. Right side shows that Indirect stimulation also leads to increases in long-term Fos-related antigens. Loss of afferent innervation, target or inhibitory interneurones may all play a role In the prolonged elevation of these proteins. These proteins form AP-1 complexes with Jun or JunCl to promote the expression of genes related to neuronal plasticity. such as GAP-43.

TiPS - September

1995 (Vol.

16)

319

R

E

V

I

E

W

nicotine injection, as well as repeated stress, further augments the levels of these proteins, suggesting AP-1 involvement in the modulation of genes necessary for the organism to deal with persistent environmental challenges34,35. The hypothalamus contains the neural centre that regulates water balance. Chronic dehydration induces Fos-related antigen immunoreactivity in the supraoptic neuronesx, with the immunoreactive protein bands having molecular weights in the range of 35-37 kDa (Ref. 10). Thus, it is possible that chronic stimuli induce these transcription factors to adapt to long-lasting biochemical and physiological changes.

long-lasting Fos-related antigen immunoreactivity in the striatum for at least three monthssi. The DNA binding activity of AP-1 is increased after kainate treatment, 6-hydroxydopamine or neuronal damage which correlates with an increase in the 35 kDa Fos-related antigen. Immunoreactivity to the Fos-related antigens is detected in these AI’-l-DNA binding complexes, indicating that these long-term 35 kDa proteins are active components involved in gene regulation. Thus, a correlation exists between brain regions undergoing structural changes and elevated Fos-related antigen levels, suggesting that the 35 kDa Fos-related antigen is a component of AI’-1modulation of the genes necessary for neuronal plasticity.

Brain injury

Identity and function The identity of these long-term Fos-related antigenimmunoreactive proteins has been in question. These proteins share homology through the peptide sequence of c-fos used to produce antibodies with Fos and other known Fos-related antigens. These factors may represent a new class of transcription factor. A truncated form of FosB has been identified as one of these long-term Fos-related antigens in several of these modelslo. This protein has the gene activation domain spliced out of the mRNA, resulting in an inhibitory facto+-55. An Al’-I-DNA binding complex containing this factor will have inhibitory effects on gene transcription. However, at least four Fos-related antigen-immunoreactive proteins have been identified with long-lasting expression, so the possibility exists that not all these proteins are related to FosB (Ref. 10). The AI’-l-DNA binding complexes observed in these models contain 35 kDa Fos-related antigen-immunoreactivity. The Fos-related antigen proteins are components of DNA binding activity that recognize AI’-1 sites in gene promoters. JunD and Jun are constitutively produced in some tissues and immunoreactivity to these proteins is also detected in these DNA binding complexes, suggesting that the Fos-related antigen proteins expressed longterm form dimers with a juntranscription factor, as do the acutely expressed Fos-related antigen proteinsQ4. The long-term Fos-related antigens in AI’-1 complexes would appear to be targeting genes for cellular adaptation in response to chronic or permanent environmental changes. The long-term Al’-l-DNA binding activity, at least in the repeated electroshock model, has a different affinity for Al’-1 sites relative to the acutely induced AP-1 binding25. Either a different set of genes are modulated, or genes are differentially modulated (or both), by the long-term Al’-1 complex as compared to the short-term AI’-l-DNA binding activity.

The neurones of the dentate gyrus express high levels of Fos-related antigen proteins for prolonged periods after kainate. However, the kainate-induced neurotoxicity increases Fos-related antigen immunoreactivity in only a few astroglia24, although Fos-positive astroglia have been reported in other brain injury paradigms37 (during this time, the dentate neurones have lost their target and input pathways%). To compensate for this loss of target, the dentate neurones sprout mossy-fibre collaterals, which form granule cell-granule cell synapseW9. These observations suggest that Fos-related antigen transcription factors are modulating genes concerned with neurite outgrowth and synaptogenesis. Indeed, genes for proteins such as growth-associated protein 43 (GAP43), mitogen-activated protein 2 (MAP-2) and tuu, are expressed in dentate neurones during the remodelling of the mossy-fibre pathway”-43, and GAP43 is known to have an AI’-1 element in its promoter which appears important for transcription 4. Cortical injury also induces longlasting Fos-related antigen immunoreactivity that is dependent on activation of NMDA receptor@-&. These Fos-related antigen proteins may be expressed to modulate genes related to neuronal plasticity; however, the newly formed dentate granule cell circuits are excitatory and could be self-stimulatory in raising basal Fos-related antigen immunoreactivity38,39. The olfactory bulb contains the highest basal levels of 35 kDa Fos-related antigen relative to other brain regions. The synapses from the receptor neurones of the nasal epithelium to the olfactory bulb are constantly being remodelled because of the continuous turnover of receptor neurones, the only neurones in the CNS that regenerate49,w.The granule neurones within the olfactory bulb appear to be expressing nuclear Fos-related antigen immunoreactivity 32.The inception of these cells occurs postnatal, and they proliferate through the majority of adult lifesi. The high basal level of Fos-related antigen protein may be stimulated by the perpetual reinnervation and synaptogenesis. Analogous to the hippocampus, the olfactory bulb contains increased levels of Fos-related antigen proteins for at least three months after kainate administration, possibly the result of neuronal damage52. Additionally, removal of the dopamine input induces

3 2 0

TiPS - September

1995 (Vol. 16)

Concluding remarks The prolonged expression of Fos-related antigen proteins is observed in a number of experimental models, in which significant and long-lasting biochemical changes have occurred. These transcription factors are components of AP-l-DNA binding activity which adjust the genomic

x programme in response to these changes. The extended expression occurs either by direct stimulation by repeated treatment, or indirect stimulation by denervation leading to disinhibition. The levels of Fos-related antigen proteins decline by 50% one week after removal of stimulationl0~~, while in brain injury models, the content remains elevated for several monthW2. Although direct and indirect stimulation appear to be two distinct phenomena, both are related to the remodelling of synapses, Genes related to processes involved with neuronal plasticity may be targets for long-term Fos-related antigen-containing AP-1 complexes and these factors may play a role in chronic or permanent disorders, such as brain injury, epilepsy and drug abuse. These factors appear to be good markers for synaptogenesis and other neuroplastic events; however, further understanding of their regulation may provide insights into genomic programmes relating to neuroplasticity. In the future, therapeutic agents could be devised to selectively increase these genes to help treat a variety of CNS disorders where increases in neural outgrowths are required, possibly such as Alzheimer’s disease.

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Selected references

5

10 11 12 13 14 15 16

Morgan, J. I. and Curran, T. (1991) Annu. Rev. Neurosci. le421-451 Hai, T. and Curran, T. (1991) Proc. Nutl Acud Sci. USA 88,3720-3724 Ryseck, R-P. and Bravo, R. (1991) Oncogene 6,533-542 Morgan, J. I., Cohen, D. R., Hempstead, J. L. and Curran, T. (1987) Science 237,192-197 Sonnenberg, J. L., Macgregor-Leon, P. F., Curran, T. and Morgan, J. I. (1989) Neuron 3,359-365 Sonnenberg, J. L. et al. (1989) 1. Neurosci. Res. 24,72-80 Pennypacker, K. R. et al. (1993) J. Neurochem. 60,204-211 Kaminska, B. et al. (1994) Eur. J. Neurosci. 6, 1558-1566 Pennypacker, K. R., Zhang, W. Q,, Ye, H. and Hong, J. S. (1992) Mol. Brain Res. 15, 151-155 Hope, B. T. et al. (1994) Neuron 13,1235-1244 Graybiel, A. M., Moratalla, R. and Robertson, H. A. (1990) Proc. N&J Acad. Sci. USA 87,6912-6916 Bronstein, D. M., Ye, H., Pennypacker, K. R., Hudson, P. M. and Hong, J. S. (1994) Mol. Bruin Res. 23,191-203 Rosen, J. B., Chuang E. and Iadarola, M. J. (1994) Mol. Bruin Rcs. 25, X8-172 Robertson, G. S., Herrera, D. G., Dragunow, M. and Robertson, H. A. (1991) Eur. J. PhrmucoJ. 150,99-100 Dragunow, M., Robertson, G. S., FauII, R. L., Robertson, H. A. and Jansen, K. (1990) Neuroscience 37,287-294 Kaminska, B., Lukasiuk, K. and Kaczmarek, L. (1994) Actu Neurobiol. Exp. Wursz. 54,65-72

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

E

V

I

E

W

Kasof, G. M. et al. (1995) J. Neurosci. 15,4238-4249 Smeyne, R. J. et al. (1993) Nature 363,166169 Estus, S. et al. (1994) J. Cell Biol. 127, 1717-1727 Kaczmarek, L. (1993) J. Neurosci. Res. 34,377-381 Cohen, D. R. and Curran, T. (1988) Mol. Cell. BioJ. 8,20632069 Nishina, H., Sato, H., Suzuki, T., Sato, M. and Iba, H. (1990) Proc. NutJ Acud. Sci. USA 87,3619-3623 Zerial, M. et al. (1989) EMBO J. 8,805-813 Pennypacker, K. R., Thai, L., Hong, J. S. and McMillian, M. K. (1994) J. Neurosci. 7,39984006 Hope, B. T., Kelz, M. B., Duman, R. S. and Nestler, E. J. (1994) J. Neurosci. 14,4318-4328 Winston, S. M., Hayward, M. D., Nestler, E. J. and Duman, R. S. (1990) J. Neurochem. 54,1920-1925 Gerfen, C. R. et al. (1990) Science 7, 1429-1432 Gerfen, C. R., McGintv, I. F. and Young, W. S., III (1991) r. Neurosn. 11,1016-1031 ’ Li, S. J. et al. (1988) J. Pharmucol. Exp. Ther. 246,4@-408 Li, S. et al. (19901 Bruin Res. 8,219-225 Dragunow; M., Leah, J. D. and Faull, R. L. (1991) Mol. Bruin Rn. 10, 355-358 Pennypacker, K. R., Lennard, D. E., Hudson, P. M., Hong J. S. and McMiIlian, M. K. Mol. Bruin Res. (in press) Jarrad, L. E. and Meldrum, 8. S. (1993) Neuroputhol. AppJ. Ncurobiol. 19,381-389 Pennypacker, K. R., Hong J. S., Douglass, J. and McMillian M. K. (1992) 1. BioJ. Chem. 267,20148-20152 &r&ova, B., Devlin, D., Kvetnansky, R., Kopin, I. J. and Sabban, E. L. (1993) J. Neurochem. 61,77&779 Lafarga, M. et al. (1992) Neuroscience 50,867-875 Dragunow, M., de Castro, D. and, FauIl, R. I. (1990) Brain Res. 527, 41-54. Okazaki, M. M. and Nadler, J. V. (1995) J. Comp. NeuroJ. 352,515-534 Tauck, D. L. and Nadler, J. V. (1985) J. Neurosci. 5,10161022 Schreyer, D. J. and Skene, J. H. (1991) J. Neurosci. 11,3738-3751 Benowitz, L. I., Rodriguez, W. R. and Neve, R. L. (1990) Mol. Bruin Res. 8, 17-23 Lin, L. H., Bock, S., Carpenter, K., Rose, M. and Norden, J. J. (1992) Mol. Bruin Res. 14, 147-153 Pollard, H., Khrestchatisky, M., Moreau, J., Ben-Ari, Y. and Represa. A. (1994) Neuroscience 61,773-787 Nedivi, E., Basi, G. S., Akey, I. V. and Skene, J. H. (1992) 1. Nei~rosci. 12,691-704 Herrera, D. G., Figueiredo, B. F. and Cuello, A. C. (1993) Dev. Brain Res. 76, 79-85 Herrera, D. G. and Robertson, H. A. (1990) Neuroscience 35,273-281 Sharp, J, W., Sagar, S. M., Hisanaga, K., Jasper, P. and Sharp, F. R. (1990) Erp. NeuroJ. 109,323-332 Dragunow, M., Faull, R. L. and Jansen, K. L. (1990)Ncurosci. Lett. 109, 128-133 Oakley, B. and Riddle, D. R. (1992) Exp. Neural. 115,50-54 Costanzo, R. M. (1994) Cibu Found. Symp. 160,233-242 Brunjes, P. C. (1994) Bruin Res. Rev. 19,146-160 Alter, C. A. and Baudry, M. (1990) Exp. NeuroJ. 109,333-341 Dobrzanski, P. et al. (1991) Mol. CeJJ. BioJ. 11,54705478 Nakabeppu, Y. and Nathans, D. (1991) Cell 64,751-759 Mumberg, D., LucibeIlo, F. C., Schuermann, M. and Muller, R. (1991) Genes Dev. 5,1212-1223

1 Postscript

is a new section that publishes news (such as scientific awards and appointments) from national pharmacological societies and other learned societies. News relating to research and teaching initiatives will also be announced. Announcements are free of charge and should be addressed to TiPS Postscript, Elsevier Trends Journals, 68 Hills Road, Cambridge, UK CB2 1LA.

Postscript

TiPS -September

1995 (Vol.16) 3 2 3