Probable role of reverse transcription in learning: Correlation between hippocampal RNA-dependent DNA synthesis and learning ability in rats

Neuroscience Letters, 36 (1983) 317-322

317

Elsevier Scientific Publishers Ireland LtcL

PROBABLE ROLE OF REVERSE TRANSCRIPTION IN LEARNING: CORRELATION BETWEEN HIPPOCAMPAL RNA-DEPENDENT DNA SYNTHESIS AND I.EARNING ABILITY IN RATS*

R.i. S A L G A N I K , H. PARVEZ, V.P. TOMSONS ,rod i . A . SHUMSKAYA

Institute of Cytology and Genetics, Siberian Branch of U.S.S.R. Academy of Sciences, 630090 Novosibirsk (U.S.S.R.) and Neuropharmacolog), Unit, University of Paris XI, I~t 440, 91405 Orsay (France) (Received June 22nd. 1982; Revised version received and accepted February 21st, 1983)

Key words: reverse transcription - hippocampus - learning - rat - DNA

The activities of RNA-dependent DNA polymerase and DNA-dependent DNA polymerase were measured in hippocampus of fast and slow learning Wistar rats. The RNA-dependent DNA polymerase activity in the hippocampus of fast learning rats exceeds two-fold that in the slow learning ones, while the rates of the DNA-dependeat DNA polymerase activities are similar. A significant increase in RNAdependent DNA polymerase only was found in the hippocampus of rats 20 min after training for the conditioned food response before the trace consolidation registered 40 rain after the training session. The data obtained are consistent with the suggestion that reverse transcriplion plays an important role in memory consolidation,

Evidence is accumulating that learning processesare accompanied by the induction of DNA-dependent synthesisof RNA moleculescoding obviously for neuronal proteins, particularly in the hippocampus [4, 14-16, 18]. The hippocampus is regarded as being involved in the transfer of the learning information from shortto long-term storage. It was recently demonstrated by us that the induction of transcription and translation in newborn animals by certain hormones or substrates results in stable changes of the inducible enzyme activities in target cells for a long period of life [12, 13]. Thus, repeated treatment of newborn rats with cortisol led to a stable enhancement of liver tyrosine andnotransferase activity for many months. Early postnatal treatment of rats with galactose resulted in a stable decrease in erythrocyte galactose-l-phosphate uridyltransferase and .increase in glucose-6phosphate dehydrogenase. This phenomenon, which was referred to as enzyme imprinting, may be regarded as extraneuronal cell memory. Recently it was reported that treatment of animals with inducers increases not only RNA synthesis, but also enhances RNA-dependent synthesis of DNA 112]. if under the effect of an inducer there occurs an increase in the synthesis of certain * Dedicated to Dr, Das on the occasion of his 50th birthday.

0304-394018310000-000015 03.00 © 1983 Elsevier Scientif;¢ Publishers Ireland Ltd.

classes of RNA and a simultaneous rise in reverse transcriptase activity, predominantly synthesized RNA molecules may serve as preferential templates for the synthesis of DNA copies. The DNA copies, in ~urn, might be integrated into the target cell genome, thereby ensuring stable changes of relevant gene expression. These mechanisms may underlie ~ C imprinting, It ~ reaso~ble to ~ u m e that similar m e c h a n i s ~ may form a background for neuronal memory' Establishment of meaningful correlations between learning ability, RNA synthesis and the intensity of reverse transcription in the brain cells as well as between changes in RNA synthesis and reverse transcription during learning, would support this suggestion. The aim of this investigation was to verify this assumption experimentally. Two groups of Wistar rats were obtained earlier by continuous selection for either fast learning (FL) or slow learning ($L) !16]. The age of the rats was 5 months. The animals were ~crificed by decapitation. The hippocampi were removed rapidly and homogenized in 0.025 M potassium phosphate buffer, pH 6.8. The homogenates were centriqtged at 20,000 g for 20 rain to obtain a postmitochondrial supernatant, and the latter was centrifuged at 100,000 g for 60 min. The 100,000 g supernatant served as a source of RNA-dependent DNA polymerase (RDP) and DNA-dependent DNA polymerase (DDP). The measurements of RDP were based on the incorporation of [•HldATP or [t4CIdTTP into an acid-insoluble product for 30 min at 37°C using as templates endogenous RNA or oligo(dT)te-t~ poly(A)-mRNA in the presence of a high dose of actinomycin-D. The incubation mixture contained, in a final volume of 50 #1:5 nmol of each of dCTP, dGDP, dTTP, 0.5 pCi [~H]dATP (spec. act. 20 Ci/mmol), 60 #g/ml of actinomycin-D, 100 mM KCI, I mM MnCI2, 40 mM Tris-HCI, pH 7.8, I mM dithiothreitol, 0.1 mM ATP, 0.01°70 Triton X-100, and 50 pg of hippocampal protein (100,000 g supernatant), oligo-(dT)te~ t~ poly(A)mRNA (when added) I ~g. When [t4C]dTTP was used (spec. act. 0.23 Ci/mmol) unlabeled dTTP in the incubation mixture was substituted by dATP. For estimation of DDP the incubation mixture contained, in a final volume of 50 #1:5 nmol of each of dATP, dGDP and dCTP, 0.5/ACi [t4C]dTTP, 10 mM MgCI,, 50 mM Tris-HCI, pH 7.4, I mM dithiothreitoi, 0.1 mM ATP, 0.01~/0 Triton X-100, 5.6 nmol of thymus DNA activated by DNAase as described previously [I], and 50 pg of rat hippocampai protein (100,000 g supernatant). The acid-precipitable radioactivity was determined as indicated above. It is well established that mRNA can serve as template for RDP but not fgr DDP. As shown in Fig. I, the addition of oligo(dT)t2-t.~ poly(A)-mRNA to the incubation mixture for the RDP estimation produces an appreciable rise in DNA sythesis. The data demonstrate that there is RDP activity in rat hippocampns as determined under the conditions employed. The measurements of RDP were performed when DDP was inhibited by the high concentration of actinomysin-D and high ionic strength. ATP, present in the incubation medium at high concentration, would have inhibited the terminal deoxynucleotidyitransferase, if the latter were present in the brain tissue [15].

319

I

I

I co.

,o,

.dl

-llo| Ill

p ¢~ o . o o 5

Fig. I. Effect of the addition of oligo (dT)l,-t~ poly(A)-mRNA to the itJ',tbation mixture for RDP estimation containing hippocampal protein (100,000 g supernatant) as a source of the enzyme or, the DNA synthesis, n -- number of experiments.

in subsequent experiments, DNA synthesis on endogenous RNA templates was studied as an overall measure of RDP in the tissue examined. The activities of hippocampal deoxyribonuclease and ribonuclease were determined by the release of acid-soluble fragments from DNA or RNA added to the incubation mixture for RDP and DDP activities estimation. The learning of the food.procuring conditioned response was performed as described previously [14]. in separate experiments, the rats were given electroconvulsive shock as reported before [171 with a time lapse of 10, 20, 40 or 120 rain after the conditioned response was acquired in order to determine the period of consolidation. Unlabeled dcoxynucleoside triphosphates and oligo(dT)j2-ts POly(A)-mRNA were obtained from the Novosibirsk pilot plant; [3H]dATP and [t4CIdTTP were supplied by Amersham; rat liver poly(A)-mgNA was a gift of Dr. N.P. Mertvetsov. The results of this investigation show that the baseline RDP in the hippocampus of FL rats exceeded by 2-fold that in the SL group, whereas the measurements of DDP were similar (Fig. 2a). It was shown previously that the baseline synthesis of hippocampal RNA in the FL is also higher than in the $L rats [14, 16]. Transfer of the memory trace from short- to long-term storage, its consolidation, occurs during a limited period of time after the acquisition of a conditioned response in trained animals. An electroconvulsive shock given promptly after the response has been acquired 'erases' the tasks learned. The FL rats were trained for the conditioned foodprocuring response. The rats were shocked immediately, 10, 20, 40 or 120 rain after

320

2b

2a RDP

ODP

10

|

c

12001

10

¢~

4001

s, ,,,,

10

0[ -

6

8 "~" ,-< 0 001

:]

5 .

c

0 70 40 120

17

6

4

6

IlPlll

3

Fig. 2. a: the hipl~campal RDP and DI)P activities in the fast learning (FI.) and slow learning (St) rats. n ~ number of e~periment.~, b: hippoeampal RI)P activity in the FL rats after training se~sion~ of 20. •~0 Or ! ~0 r a m . I1 : n u l n ~ ' r ot" e x p e r i m e n t s I x ' r f o r m e d f o r each g r o u p .

the training session; memory trace consolidation was registered 40 min after training. Analysis of hippocampal RDP activities showed that promptly after the training session there were no changes in RDP activity. A significant increase in the RDP activity was found 20 rain after the training session (Fig. 2b), and it declined 40 rain alter it. There were no modifications in hippocampal DDP in rats after training. Previous data demonstrated that, under training conditions, there was an increase of hippocampal RNA synthesis which in the FL rats exceeds largely man that of SL rats [14, 16]. Special experiments were performed to measure ribonuclease and deoxyribonuclease activities concomitantly with assays of hippocampal RDP and DDP activities, it was found that the FL and SL rats did not differ statistically in total hippocampal activities. Hence, differences in the hippocampal RDP were not due to more intense RNA and DNA degradation. The results obtained are consistent with the hypothesis that reverse transcription may be of importance in memory processes. Quite conceivably, the induction of the synthesis of RNA molecules, which code for 'meme,, proteins', and the subsequent copying of these RNA molecules by reverse transcriptase, may culminate in the integration of the copied DNA into the genome. The amplification of certain genes would ensure the enhanced synthesis of the proteins needed to maintain constantly new neuronal circuits conserv;.ng the memory trace.

321

[3H]Thymidine incorporation into brain DNA is enhanced during learning [9, 10]. In contrast, electroconvulsive shock induces a reversible inhibition of DNA synthesis in rat brain [6]. There occurs a rapid turnover of neuronal DNA [8]. Since DNA synthesis in neurons in not concerned with the DNA replication needed for cell division, these results are compatible with the assumption that neuronal newly synthesized DNA may be a product of reverse transcription. The origin and nature of enzymic activities ensuring reverse transcription in hippocampal cells are unknown and beyond the scope of present investigation. The endogenous retroviruses are a possible source of reverse transcription enzymes in normal tissues; their appearance is known to be controlled by the physiological conditions [7, 1I1. Furthermore, there are data indicating that animal cells may possess their own RNA-dependent RNA polymerase different from retroviral enzymes [2]. It cannot be excluded that such enzymes may also be present in neurons which are concerned with the transfer of learned information into the genome which can eventually serve as memory store. The role of DNA as a memory storage molecule was first suggested by Gaito in 1963 [5]. However, the notion that ~.hromosomal DNA is highly conservative was widely held for a great number of years. These findings add some more new information that there occurs non-random amplification and recombination of DNA in animal cells and may lead to the revival of the abandoned idea. In conclusion, it can be stated that the process of learning is accompanied by the induction of the DNA-dependent synthesis of RNA molecules conceivably programming neuronal 'memory' proteins. The copying of these preferentially synthesized RNA molecules by reverse transcription may culminate in the integration of the copied DNA into the genome of neuron. The amplification of particular genes would ensure the enhanced synthesis of the proteins needed to maintain constantly new neuronal circuits conserving the memory trace. The authors wish to express their sincere thanks to all the members of the Institute of Cytology and Genetics (Laboratory of Molecular Genetic',) for their encouraging comments and suggestions. The kind help of Dr. Roger V~ertolotti of CNRS, Institute of Molecular Genetics, Gif-sur-Yvette, France, in helping to prepare the bibliography is acknowledged with thanks. ! Aposhian, H.V. and Kornberg, A., Enzymatic synthesis of deoxyribonucleic acid. J. biol. Chem., 237 (1962) 519-525. 2 Bauer, G. and Hoffschneider, P.H., An RNA-dependent DNA polymerase, different from the known viral reverse transcriptase, in the chicken system. Proc. nat. Acad. Sci. U.S.A., 73 (1976) 3025-3029. 3 Bollum, F.J., Thermal conversion of nonpriming deoxyribonucleic acid to primer, J. biol. Chem., 234 (1959) 2733-2734. 4 Cupeilo, A. and Hyden, H., Studies on RN:\ metabolism in the nerve cells of hippocampus during training in rats, Exp. Brain Res., 31 (1978) 143-152. 5 Gaito, J., DNA and ~.NA as memory molecules, Physiol. Rev., (1963) 471.

6~ A., A~escia, P. and Rmilr,hno, B., Effect of ciearmhock on thyroid(he incorporation into rm ~ I ~ A , J. Naeodu~., .~1 (1978) 9L~987o 7 Liebmman, D., Hoffman-Liebaman, K. and Sacb, L., ~ o f ~ type C vires a drains normal myeloJd ~ diffc~liadon, Vbolow, 107 0980) 121-1~1.. S lhnnronc-Capano, C., D'Onofrio, G. and Gindi, a, A., DNA remover in rat cerebral cortex, J. Nemrockm., in press. 9 Rein(s. S.. A u t ~ study of JH-thymidinc incorporation imo brain DNA during learning, my o,, a m . my,.. 4 ,972) 39,-397.

10 ~ S ;

and ~ b l e ~ II~W, ~ g

of brain DNA by 3 H ~

during ~ .

P~I.

~ . ~ s , . 4 (1972) 3 ~ 1 , I I RingoM, G,M.. Yamamo/o. K,R,, Bishop, J.M. and Varmus, H~E,, G l u c o c o ~ stimulated accumulatinn of mouse mammary tumor virus RNA: increased rate of symhcsisof viral RNA, Proc. nat. Acad. Sci. U.S.A, 74 (1977) 21179-21183. 12 $albanik, R.I., C~ymmova, I,M., MatEd, A,L., Mammkova, N.M. and 5olovyova, N.A., Enzymic "impriming" as the resuh of ¢mly postnatal administration of ¢m~m¢ inducers, Expcrimnia, 36 (1980) 43--44. 13 Salganik, R.I.. Tomsons, V.P. and Drcvick, V.F., increase of the RNA depend.,mt DNA p o l y m m activity in rat liver under genetic induction of adaptive enzymes, Dokl. Akad. Nauk. SSSR, 254 (1980) 1482-1486. 14 Shumskaya. I.A., Belaycv, A.I. and Korochkin, L.i., Anal)sis of the hippocampal RNA in rats with genetically determined differmt abilities fm learning, Zh. vyssh, herr. Deyat. Pavlova, 29 (1979) 269-274. 15 Shumskaya. I.A. and Korochkin, L.I., Studies on the intensity of RNA s~nlhesis in rat hippocampus during training, Zh. vyssh, ncrv. Dcyat. Pavlova, 25 (1975) 778--782. 16 $humskaya, I.A., Korochkin, L.I. and Marchenko, !.!., Studies on biochemical and genetical mechanisms of learning. II1. Selection of rats for high and slow rates of acquisition of food procuring motor conditioned rdlex, Genetica, 15 (1979) ~27-$J4 (in Russian). 17 Tcnchcva. C.S. and P~,'zner, LZ., Influence of dectroconvulsiv¢ shock on the content of RNA and proteins in hippocampal neurons in rats of diffucnt age, Physiol. Zh., 4 (1974))7. 11,1Uphou~. L.L.. Maclnnes. J.W. and Schlcsinger, K.. Role of RNA and protein in memory storage: a review, Bchav. (;enel.. 4 (1974) 29-Sl.