Cognitive Effects of Colostral-Val Nonapeptide in Aged Rats

Behavioural Brain Research 118 (2001) 201–208 www.elsevier.com/locate/bbr

Research report Cognitive effects of Colostral-Val nonapeptide in aged rats

Piotr Popik a,*, Zdzislaw Galoch a, Maria Janusz b,Jo´zef Lisowski b, Jerzy Vetulani a

a Institute of Pharmacology, Polish Academy of Sciences, 12 Sme˛tna Street, 31–343 Krako´w, Poland b Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Weigla 12 Street, 53–114 Wrocl*aw, Poland

Received 26 May 2000; received in revised form 8 September 2000; accepted 8 September 2000

Abstract

Colostrinin, a complex of polypeptides derived from sheep colostrum retards the progress of Alzheimer’s disease and facilitates acquisition and retrieval of spatial memory in aged rats. Here we investigated the cognitive effects of colostrinin-derived nonapeptide (Colostral-Val nonapeptide, CVNP) in aged rats that demonstrated learning deficits. Administered for 14 days, CVNP did not affect the acquisition of spatial learning or memory retrieval in the Morris water maze. As a result of reversal learning, placebo treated rats shifted searching behavior and swam less in the area of original platform position and more in the area of recent platform position, suggesting formation of the new spatial map. CVNP treated rats did not change the searching pattern and still investigated the area that contained ‘original’ escape platform, suggesting that CVNP treatment delays the extinction of spatial memory. In another experiment, CVNP administered for 8 days did not influence the acquisition of the active avoidance task, but significantly delayed its extinction. The present findings indicate that colostrinin-derived nonapeptide may delay the extinction of long-term memories. © 2001 Elsevier Science B.V. All rights reserved.

Keywords: Memory; Learning; Alzheimer’s disease; Colostrum; Colostrinin; Proline-rich polypeptide complex; Colostral-Val nonapeptide; Morris water maze; Active avoidance

1. Introduction in the first period of life. One of the immunologically active agents from ovine colostrum was isolated by Investigation of the cognitive effects of peptides has a Janusz et al. [15–17] and was characterized as an 18 long history. In the 1960s, De Wied found that removal kDa polypeptide, built of three 6 kDa subunits, rich in of the posterior lobe of the pituitary in rats impaired proline (22%) and non-polar amino acids. The prepara- the maintenance of avoidance behavior and that this tion was subsequently found to be rather a complex of deficit could be restored by treatment with vasopressin proline-rich polypeptides (PRP). The polypeptide com- [6]. These and other findings resulted in the appearance plex, named colostrinin, was characterized as a new of the ‘neuropeptide’ concept in the 1970s and a hy- cytokine [14]. Recently, the immunoactive properties of pothesis that peptides are implicated in learning and the human analog of ovine colostrinin were described memory processes. This hypothesis was later supported [11,21]. Colostrinin induces maturation and differentia- by numerous findings demonstrating that various pep- tion of murine thymocytes and affects humoral and tides affect learning and memory processes in several cellular immune reactions, both in vivo and in vitro behavioral paradigms (for reviews, see Refs. [7,30]). [15,31]. Colostrum, the early milk produced during puerperal By digestion of colostrinin with chymotrypsin, a non- period, plays an essential role in the survival of new- apeptide fragment of mol.% wt.% 1000, with a sequence borns and is the main source of immunological defense of Val–Glu–Ser–Tyr–Val–Pro–Leu–Phe–Pro was isolated [29]. This Colostral-Val nonapeptide (CVNP), * Corresponding author. Tel.: +48-12-6374630; fax: +48-12- 6374500. both isolated from chymotryptic digest as well as ob- E-mail address: [email protected] (P. Popik). tained by the chemical synthesis [32], showed the full

0166-4328/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0166-4328(00)00323-5 202 P. Popik et al. / Beha6ioural Brain Research 118 (2001) 201–208 spectrum of immunotropic activities of colostrinin. weight 485–500 g), obtained from the colony of the CVNP showed a strong effect on the primary and Institute of Immunology and Experimental Therapy, secondary immune response against SRBC (T-cell de- were housed under standard laboratory conditions pendent antigen) in mice [18]. Whereas colostrinin at (lights on at 06:00 h, lights off at 18:00 h; room concentrations of 1–100 mg/ml induced production of temperature 2391°C) with pelleted food and tap water interferon and tumor necrosis factor a in human pe- available ad libitum. Rats were kept in plastic cages ripheral blood leukocytes as well as in the whole blood measuring 58×37×19 (four to five rats per cage). cell cultures, CVNP was considerably less active as a cytokine inducer. Thus, significant levels of interferon 2.2. Drugs and tumor necrosis factor were induced by 100 mg/ml of CVNP, but lower concentrations were found to be CVNP, the active nonapeptide fragment of ineffective [13]. Similarly, it was also reported that colostrinin (Val–Glu–Ser–Tyr–Val–Pro–Leu–Phe– cytokine-inducing activity of CVNP in cultures of hu- Pro) was synthesized by the solid-phase method [32] man whole blood cells was considerably lower than that and kindly provided by Professor G. Kupryszewski, of colostrinin [33]. Institute of Chemistry, University of Gdansk, Poland. Among the most prominent consequences of normal Physiological saline served as placebo. All injections and pathological aging in humans are cognitive deficits, were done i.p. in a volume of 1 ml/kg. The selection of including learning impairment and delayed amnesia. the dose for current investigation was based on our Perhaps the most severe form of the senile dementia is previous report [23] and immunotropic effects of Alzheimer’s disease, which is associated with numerous colostrinin and CVNP in animals [18]. Since CVNP in pathophysiological alterations of the CNS. Pathophysi- immunological tests appears to be much less potent ological changes in Alzheimer’s patients include, among than colostrinin, and since we previously used doses of others, accumulation of b-amyloid plaques and neurofi- 4–20 mg per rat (:8–40 mg/kg), for the current exper- brillary tangles [12]. In addition to the profound im- iment the dose of 50 mg per rat (100 mg/kg) was pairment in cholinergic transmission, alterations in the selected. CVNP was dissolved in sterile physiological function of cytokines and other factors associated with saline and the dilution was divided into 7-ml aliquots immune reactions have been reported [2,5,9], (see Ref. that were stored at −20°C. A fresh aliquot was used [10] for the recent review). each day. The possibility that immunological factors play a key role in Alzheimer’s disease was investigated in the 2.3. Morris water maze double-blind, placebo controlled clinical experiment. As measured by the Mini Mental State score and the 2.3.1. Apparatus recent memory test, 11 out of 16 patients undergoing A gray metal circular pool (180 cm in diameter, 50 three to six cycles of therapy consisting of oral adminis- cm in height) was filled to a height of 25 cm with tration of colostrinin (100 mg every second day in one lukewarm (22°C) tap water that was changed every day, cycle lasting for 3 weeks) significantly improved their as described previously [22,23]. Curtains, bright lamps cognitive functioning [19]. Recent study from our labo- and other stimuli around and above the pool provided ratory indicate that colostrinin produced a beneficial numerous stable extra-maze cues. effect on spatial learning in aged rats [23]. We found that colostrinin at the dose of 4 mg per rat facilitated 2.3.2. Procedure the acquisition of spatial learning in the Morris water The behavioral procedure consisted of pre-training maze of 13- but not 3-month-old rats. In addition, the (days 1–14), the first ‘transfer test’ (day 15), training to dose of 4 mg per rat improved incidental learning the new platform position (reversal learning, day 16) (habituation test) in 13-month-old rats. The present and the second ‘transfer test’ (day 17). During pre- study was carried out to investigate the possibility that training (one trial per day), rats were injected with colostrinin-derived nonapeptide (Colostral-Val non- placebo or 100 mg/kg of CVNP 30 min before each of apeptide, CVNP) sharing colostrinin’s immunotropic the swimming trials and were required to find the activity and characterized to a greater extent may have platform located in the initial (NE or SW) quadrant. similar effects on learning and memory in rats. Since day 15, no injections were given. During the first ‘transfer test’, rats swam for 1 min in the pool without the platform. On day 16, rats were trained to find the 2. Materials and methods platform positioned in the opposite (SW or NE) quadrant. This was accomplished with four trials, with the 2.1. Animals inter-trial interval of 30 s. The second ‘transfer test’ was carried out similarly to the first, with the platform Male Wistar rats :16–19 months of age (body removed. For one half of the subjects, the platform was P. Popik et al. / Beha6ioural Brain Research 118 (2001) 201–208 203 placed in the NE quadrant during pre-training and in quadrant. The time a rat spent in the ‘initial’ quadrant the SW quadrant during training. For the other half, was divided by that sum. This measure allowed the the platforms were placed in opposite quadrants. assessment of the relative preference for the pre-training The swimming trial consisted of manually placing a quadrants (0–100%). The difference between the rela- rat into the water facing the wall of the pool, at one of tive preference for the first ‘transfer test’ and for the two starting positions (SE or NW) around the pool second ‘transfer test’ was used as the index of memory perimeter. At every trial, the starting position was resulting from both pre-training and training. changed in a pseudorandom sequence. The rat was Memory index = Relative preference at pre-training required to find a gray circular wooden platform (15 cm −Relative preference at training=100× of diameter), covered with a metal mesh grid, which was present inside the pool in the middle between its Time in ininitial quadrant at pre-training center and the wall. The upper surface of the platform Time in ininitial+Opposite quadrants at pre-training was submerged 1 cm below the water surface. If a rat Time in ininitial quadrant at training escaped onto the platform, it was permitted to remain − Time in ininitial opposite quadrants at training there for 30 s. If a rat failed to find the platform within + 120 s, it was directed into it by hand and allowed to This method of calculating and presenting data al- remain there for 30 s. After completion of the trial, rats lowed a comparison of the relative efficacy of the were placed in the ‘drying’ cage and heated by infrared ‘opposite’ versus ‘initial’ spatial memory, in that it lamp. reduced the individual variance among subjects that otherwise would mask possible effects of treatment. 2.3.3. Data presentation and statistical analysis The swimming behavior was observed on the TV 2.4. Acti6ea6oidance test screen. It was simultaneously recorded on a PC com- puter with the EYE 1.3 program (J. Dl*ugopolski, Rats were trained in the automatic shuttle-box ap- Krako´w, Poland) and analyzed off-line using paratus (Ugo Basile, Milan, model Reflex Conditioner TRACKER 1.69.63 (P. L*o´j, Krako´w, Poland) and 7501) to avoid electrical foot-shock. The experiment WINTRACK 2.3.00102 (Dr D. Wolfer, Department of consisted of two parts: the acquisition phase and the Anatomy, University of Zu¨rich, Switzerland) software. extinction phase. During acquisition trials, rats were During the training parts of the procedure, the latency injected with placebo or 100 mg/kg of CVNP 30 min (s) and path (m) needed to find the platform was before each of the training sessions. After placing the analyzed. In addition, the ‘efficiency index’, that is, the rat into one of the boxes and after the pre-programmed percent of path traveled with a deviation by 15° or less delay (20–30 s), the buzzer and light (conditioned of the direction of movement from the direction point- stimuli, CS) were turned on. The CS signaled that the ing to the goal platform was analyzed. electrical foot-shock (unconditioned stimulus, UCS) To measure the effects of nonapeptide on the efficacy would be turned on in 4 s. If the rat escaped into the of spatial memory, detailed analysis was performed on other box, it avoided the UCS and such behavior was the swimming behavior recorded during both ‘transfer counted as a successful escape (conditioned avoidance tests’. It involved the measurement of the time (s) spent response, CAR). If the rat did not escape during CS, it by rats swimming in quadrants that contained goal received the UCS (0.2 mA, 3 s). This sequence was platform during pre-training and training. This repeated ten times with a pre-programmed inter-trial analysis was based on a similar one used previously interval (:30 s) each day. The extinction phase was [22]. It was assumed that the pre-training would result carried out essentially the same, with the exception that in the formation of a spatial memory and that treat- no UCS was given. ment with CVNP would affect the acquisition and/or A criterion of 24 CARs out of a possible 30 in 3 retrieval of this memory. It was also assumed that successful days was adapted from Bohus and De subsequent training to the platform placed in the Wied’s work [3]. Rats acquired the active avoidance new position would produce the new spatial map, response until this criterion (8 days). The extinction thereby replacing the original spatial map with the new phase lasted for the next 14 days, i.e. until significant one and that this would be shown as the loss of the decrease of CARs ( 25%) was noted in placebo- preference for the quadrant containing the ‘pre-train- treated rats. ing’ platform. Therefore, to assess the ultimate memory resulting 2.5. Statistics from the training procedure in relation to the memory of pre-training phase, the following calculations were For statistics, the Student’s t-test and two way mixed performed. The time a rat spent in the ‘initial’ quadrant design MANOVA followed by Newman–Keul’s test was added to the time a rat spent in the ‘opposite’ were used. 204 P. Popik et al. / Beha6ioural Brain Research 118 (2001) 201–208

2.6. Ethics

All experiments were carried out according to the National Institutes of Health Guide for Care and Use of Laboratory Animals (publication No. 85-23, revised 1985) and were approved by the Internal Bioethics Commission.

3. Results

3.1. Morris water maze

Rats demonstrated rapid acquisition of the spatial information and on day 14, the latencies to escape onto Fig. 2. Presented are mean + or − S.E.M. swimming path lengths the platform were shorter than 10 s. There were no required to find the hidden platform in the Morris water maze of aged rats treated with placebo (open symbols) or CVNP (closed differences between placebo- and CVNP-treated rats in symbols) during the first 14 trials (one trial per day). After the initial the escape latencies, paths or ‘efficacy indexes’ during training, the transfer test has been carried out. During the subsequent pre-training and training swimming trials (Figs. 1–3). four trials (day 16), rats were required to find the platform placed in Table 1 demonstrates that during the first ‘transfer the opposite quadrant (reversal training). Statistical analysis with the test’, all rats were swimming in the vicinity of the area use of repeated measures MANOVA demonstrated significant effect of the trial number F(13,286)=14.93, PB0.0001 but not of the that contained the platform during pre-training trials. treatment F(1,22)=0.28, P\0.05 or interaction F(13,286)=0.48, For example, rats treated with placebo spent on aver- P\0.05. A separate MANOVA performed on the data gained on age 22.4 s out of 60 s in the initial quadrant. Consistent reversal training demonstrated no effect of the trial number with spatial learning and memory deficits demonstrated F(3,66)=1.38, P\0.05, of the treatment F(1,22)=0.00, P\0.05 \ by aged rats [23], the preference for training quadrant and interaction F(3,66)=0.28, P 0.05. was not statistically significant. In addition, there were test’ (Table 1). However, analysis of the data recorded no differences between placebo- and CVNP-treated rats in searching behavior recorded during the first ‘transfer on the second ‘transfer test’ demonstrated differences between treatments. Thus, two-way mixed design MANOVA performed on the relative preference values: TRANSFER TEST [first and second]×TREATMENT [placebo and CVNP] revealed significant effect of the TRANSFER TEST repeated factor: F(1,22)=16.79, PB0.001 and the INTERACTION: F(1,22)=4.33, PB0.05 but not TREATMENT factor: F(1,22)=1.65, P\0.05. Since a significant interaction between TRANSFER TEST and TREATMENT indicates that CVNP-treated rats behaved differently during the transfer tests than controls did, a detailed post-hoc analysis with the use of Newman–Keul’s test was performed. It indicates that the reversal learning resulted in a marked loss of the preference for initial quadrant in placebo treated rats, whose second ‘transfer test’ Fig. 1. Presented are mean + or − S.E.M. latencies to find the hidden platform in the Morris water maze of aged rats treated with relative preferences dropped from 66 to 43%, suggesting placebo (open symbols) or Colostral-Val nonapeptide, CVNP, (closed that the reversal learning resulted in the formation of a symbols) during the first 14 trials (one trial per day). After the initial new spatial memory. This shift was not observed in training, the transfer test has been carried out. During the subsequent CVNP-treated rats, whose first and second relative four trials (day 16), rats were required to find the platform placed in the opposite quadrant (reversal training). Statistical analysis with the preferences were 64.5 and 57%, respectively. Student’s use of repeated measures MANOVA demonstrated significant effect t-test performed on the Memory Index values revealed of the trial number F(13,286)=17.25, PB0.0001, but not of the significant differences between groups: t=2.082, df= treatment F(1,22)=0.27, P\0.05 or interaction F(13,286)=0.65, 22, PB0.05. \ P 0.05. A separate MANOVA performed on the data gained on Examples of most representative swimming paths of reversal training demonstrated no effect of the trial number F(3,66)=1.36, P\0.05, of the treatment F(1,22)=0.01, P\0.05 placebo and nonapeptide treated rats are presented in and interaction F(3,66)=0.38, P\0.05. Fig. 4. P. Popik et al. / Beha6ioural Brain Research 118 (2001) 201–208 205

3.2. Acti6ea6oidance Table 1 Effects of Colostral-Val nonapeptide (CVNP) on spatial memory 9 resulting from pre-training and subsequent training to the platform Fig. 5 shows the mean S.E.M.% of successful es- positioned in the opposite locationa capes (CARs) recorded during acquisition and extinction of the active avoidance test. There were no Treatment [N] differences between placebo and CVNP treated rats during the acquisition phase. Thus, two-way mixed Placebo [11] CVNP [13] design MANOVA performed on the CAR data re- Transfer test c1 vealed significant effect of TRIAL NUMBER repeated ‘initial’ quadrant 92.3Timein 22.191.822.4 factor: F(7,133)=100.8, PB0.001 but not of the ‘opposite’ quadrant 10.991.4 12.291.5Timein cc cc TREATMENT: F(1,19)=0.98, P\0.05 or INTER- Relative preference66.195.1H 64.594.1H ACTION: F(7,133)=0.76, P\0.05. However, in con- Transfer test c2 trast to placebo-treated, CVNP-treated rats ‘initial’ quadrant 13.791.4 19.491.3Timein Time in ‘opposite’ quadrant 18.191.5 14.791.4 demonstrated a significant persistence of CARs during c Relative preference 43.294.42 57.093.57 the extinction phase. Thus, two-way mixed design 9 9 MANOVA demonstrated significant effect of TRIAL 22.95 5.22Memoryindex 7.49 5.21* B NUMBER repeated factor: F(13,247)=18.64, P a Rats were pre-trained for 14 days, one trial per day, to navigate 0.001 and of the TREATMENT: F(1,19)=14.28, PB into the hidden platform placed in the center of ‘initial’ quadrant 0.01 with insignificant INTERACTION: (pre-training). Thirty min before each trial, rats received i.p. injec- F(13,247)=1.23, P\0.05. Results of the post-hoc tions of either placebo or CVNP (100 mg/kg). The efficacy of this analysis with the use of Newman–Keul’s test revealed spatial memory was measured subsequently during the first ‘transfer test’. This was followed by the training phase (1 day, four trials), that CVNP-treated rats presented higher CARs com- during which the platform was positioned in the center of ‘opposite’ pared to placebo-treated controls starting on day 2 of quadrant. On the next day, a second ‘transfer test’ was performed. the extinction phase, which continued until the end of Presented are mean9S.E.M. times/s spent by rats in the ‘initial’ and the experiment. In addition, analysis of the area under ‘opposite’ quadrants during pre-training and training, the resulting relative preferences (%) and memory indexes (%). curves (AUC) with two way mixed design MANOVA c Significant (PB0.05) difference compared to Relative Preference demonstrated significant effect of EXPERIMENT of placebo-treated rats on Transfer Test c2 (Newman–Keul’s test). PHASE: F(1,19)=31.86, PB0.001, TREATMENT: cc Significant (PB0.001) difference compared to Relative Prefer- F(1,19)=14.58, PB0.01 and INTERACTION: ence of placebo-treated rats on Transfer Test c2 (Newman–Keul’s test). * Significant difference compared to placebo-treated rats (PB0.05, Student’s t-test).

F(1,19)=10.57, PB0.01. The AUC of CVNP-treated rats during the extinction phase was higher (PB0.001, Newman–Keul’s test) than that of placebo-treated controls (Fig. 5).

4. Discussion

The present study demonstrates that colostrinin (a Fig. 3. Presented are mean + or − S.E.M. ‘efficiency indexes’ complex of proline-rich polypeptides isolated from displayed by aged rats treated with placebo (open symbols) or CVNP ovine milk)-derived nonapeptide (Colostral-Val non- (closed symbols) during the first 14 trials (one trial per day) in the apeptide, CVNP) with immunomodulatory properties Morris water maze. After the initial training, the transfer test has similar to the parent preparation, may have potent been carried out. During the subsequent four trials (day 16), rats were required to find the platform placed in the opposite quadrant (rever- cognitive effects in rodents. Previous data from this sal training). Statistical analysis with the use of repeated measures laboratory demonstrate that colostrinin improved cog- MANOVA demonstrated significant effect of the trial number nitive functions in the aged, but not young rats [23]. F(13,273)=3.65, PB0.0001 but not of the treatment F(1,21)=0.09, Due to this fact and to the shortage of material, in the \ \ P 0.05 or interaction F(13,273)=0.81, P 0.05. A separate present study we did not investigate if CVNP might MANOVA performed on the data gained on reversal training demonstrated no effect of the trial number F(3,66)=1.38, P\0.05, of the influence learning and memory in young subjects. treatment F(1,22)=0.00, P\0.05 and interaction F(3,66)=1.69, However, CVNP affects learning and memory in a P\0.05. very different way than colostrinin. Thus, in the previ- 206 P. Popik et al. / Beha6ioural Brain Research 118 (2001) 201–208 ous study [23] we found that rats treated with colostrinin during acquisition trials demonstrated shorter swimming latencies, navigated onto the goal platform more precisely and during the ‘transfer test’ searched the training quadrant more intensely than placebo-treated agemates. In fact, administration of colostrinin to aged rats produced swimming patterns observed in the young subjects. In the present study, treatment with CVNP did not influence any of these parameters, suggesting that CVNP did not affect the process of acquisition of spatial memory. Similarly, administration of CVNP did not affect the acquisition of active avoidance response. Instead, rats treated with CVNP did not search for the ‘new’ platform position in the Morris maze and showed prolonged extinction of the active-avoidance task. The difference in cognitive effects between colostrinin and the colostrinin-derived nonapeptide agrees well with the different (and even opposite) effects of peptides on learning and memory due to the modification of their chemical structures (e.g. Fig. 5. Presented are mean + or − S.E.M. conditioned avoidance Ref. [24]). responses (CARs) of rats trained in the active avoidance apparatus. Our initial assumption (see Section 2.3.3) was that Rats were treated with placebo (N=11, open symbols) or CVNP CVNP would facilitate spatial learning and the acquisi- (N=10, closed symbols) only during the acquisition phase (left tion of active avoidance response. At first sight, the curves) which lasted until criterion of 24 out of 30 possible CARs during 3 subsequent days, that is, until day 8. The extinction phase present results may indicate that CVNP-treated rats (right curves) lasted until a significant decrease of CAR was noted in demonstrated a deficit in acquisition of new informa- the placebo-treated rats, i.e. until experimental day 22. Inset: the tion, suggesting impairment of cognitive functions. This same data shown as area under curve (AUC). *Significant difference hypothesis is unlikely in light of the data demonstrating compared to placebo-treated rats (PB0.05–PB0.001, Newman– Keul’s test).

that CVNP did not directly affect the acquisition of spatial task or active avoidance response. Instead, CVNP-treated rats were unable to acquire a new aspect of a given task after the treatment was discontinued. These intriguing and unexpected findings may also indicate an enhancement of attention during the acquisition phase or its impairment during the reversal learning. This again, appears unlikely because CVNP did not affect the learning per se and it does not stay in the body long enough to affect reversal learning in Morris water maze or extinction of active avoidance task. From the same reason, even if CVNP produces a stress-like state (purportedly, due to its immunotropic activity), such influence cannot explain delayed effects on the extinction. Another hypothesis worth consider- ation concerns the state-dependent learning. Its contri- bution to the present findings appears also unlikely because CVNP-treated rats demonstrated the better and not the worse memories of these aspects of training that took place under the influence of CVNP. Fig. 4. Swimming paths recorded during the first and second ‘transfer A plausible hypothesis would rather indicate that the tests’ (left and right panels, respectively) of representative rats treated inability to learn a new aspect of the given task is due during pre-training with placebo and CVNP (upper and lower panels, to a better encoding of the original information, result- respectively). The figure shows that in contrast to placebo-treated rat that shifted the searching behavior as a result of training to the new ing not in better performance, but rather in an inhibi- platform position, CVNP-treated rat did not demonstrate such a tion of the extinction of the memory acquired under its shift. influence. P. Popik et al. / Beha6ioural Brain Research 118 (2001) 201–208 207

It is difficult to assess whether such an effect may be At present, we are unable to propose the mechanism regarded as a ‘cognitive improvement’. From the thera- of action of CVNP. It cannot be excluded that its peutical perspective, a pharmacological treatment sup- immunomodulatory effects [13,18,33] may be of impor- posed to improve cognition would likely facilitate the tance. For instance, in a rodent model of AIDS, Sei et acquisition of new information. On the other hand, al. [28] demonstrated that spatial learning is impaired in potential users of a ‘memory pill’ may look for the mice suffering profound deficits in immunological re- treatment with which the information acquired under sponses. Converging lines of evidence indicate the im- its influence will last longer and be recalled with higher portance of immunological factors in the pathogenesis accuracy. It might be hypothesized that this is the way of Alzheimer’s disease [10]. Since CVNP induces inter- CVNP influenced learning and memory of aged rats in feron g and is regarded as a cytokine, these factors may the present study. Earlier data demonstrate that pep- play a role in the mechanism of its promnestic effects. tides, like vasopressin and ACTH, produce somewhat Findings demonstrating that interferon g inhibits pro- b similar effects, e.g. Ref. [3]. Further experiments are duction of the Alzheimer’s amyloid precursor protein required to elucidate if CVNP may, like vasopressin, [26,27] are in favor of this idea. However, whether or ACTH and related neuropeptides produce similar, not the immunomodulatory effects of colostrinin- beneficial effects on the acquisition, retrieval and ex- derived nonapeptide are of importance for the improve- tinction of memory. ment of cognitive functions in rats, remains to be We found similar effects of CVNP on learning and established. memory in two paradigms including spatial learning in the Morris water maze [20] as well as the active avoidance task [3]. Although these tests differ profoundly in Acknowledgements the quality of information to be acquired (spatial information versus an association between US and CS), the This study was supported in part by statutory activi- nature of the negative reinforcement (escape from water ties of the Institute of Pharmacology, Polish Academy and escape from the electrified grid) appears similar. of Sciences, Krako´w and of the Institute of Immunol- The active avoidance paradigm is regarded as a rapid ogy and Experimental Therapy, Polish Academy of and effective screening test, but the low ‘predictive Sciences, Wrocl*aw. We thank Dr David Wolfer from validity’ and high component of anxiety compromises the Department of Anatomy, University of Zu¨rich, its use. On the other hand, the Morris water maze is Switzerland, for enabling us to use Wintrack 2.3.00102 regarded as a test very sensitive to a number of treat- to calculate Morris water maze data. The excellent ments producing deficits in spatial learning, including technical assistance of Zofia Magalas is appreciated. brain lesions and a variety of amnestic agents (see Ref. [4] for review). This test measures spatial memory, which is highly dependent on the presence of extra- References maze cues. Both tests, however, are dependent upon a potent motivation and involve explicit, long-term mem- [1] Bartus RT, Dean RL. Animal models for age-related memory ory, and both do not involve food or water deprivation. disturbances. In: Animal Models of Dementia. New York: Alan Researchers studying memory-improving drugs lack R. Liss, 1987:69–79. both a reliable model of human dementia and a stan- [2] Bauer J, Ganter U, Strauss S, Stadmuller G, Frommberger U, Bauer H, Volk B, Berger M. The participation of interleukin-6 in dard, recognizable memory-facilitating reference com- the pathogenesis of Alzheimer’s disease. Res Immunol pound. The lack of these two basic tools in the learning 1992;143:650–6. and memory research is a serious obstacle in drug [3] Bohus B, De Wied D. Inhibitory and facilitatory effect of two development. This prompts a development of a variety related peptides on extinction of avoidance behavior. Science 1966;153:318–20. of animal models of dementia, of which none to date [4] Brandeis R, Brandys Y, Yehuda S. The use of the Morris water seems to be fully satisfactory (for a review see Ref. maze in the study of memory and learning. Int J Neurosci [25]). Nonetheless, there is common agreement that the 1989;48:29–69. best animal model of the normal and pathological [5] Cacabelos R, Alvarez XA, Franco-Maside A, Fernandes-Novoa aging in humans is the aging animal itself. Aged ro- L, Caamano J. Serum tumor necrosis factor (TNF) in Alzheimer’s disease and multi-infarct dementia. Methods Find dents show deficits in a number of learning and mem- Exp Clin Pharmacol 1994;16:29–35. ory paradigms, including avoidance behavior [1] and [6] De Wied D. The influence of the posterior and intermediate lobe spatial memory [8,25]. Thus, our attempt to investigate of the pituitary and pituitary peptides on the maintenance of a learning and memory effects of CVNP in aged rats conditioned avoidance response in rats. Int J Neuropharmacol 1965;4:157–67. seems to correspond well with the currently available [7] De Wied D, Diamant M, Fodor M. Central nervous system and most broadly accepted animal models of human effects of the neurohypophyseal hormones and related peptides. dementia. Front Neuroendocrinol 1993;14:251–302. 208 P. Popik et al. / Beha6ioural Brain Research 118 (2001) 201–208

[8] Frick KM, Baxter MG, Markowska AL, Olton DS, Price DL. [22] Popik P. Facilitation of memory retrieval by anti-addictive alka- Age-related spatial reference and working memory deficits as- loid, ibogaine. Life Sci 1996;59:379–85. sessed in the water maze. Neurobiol Aging 1995;16:149–60. [23] Popik P, Bobula B, Janusz M, Lisowski J, Vetulani J. [9] Fundenberg HH, Singh VK. Alzheimer’s ‘syndrome’ prognosis Colostrinin, a polypeptide isolated from early milk facilitates of subsets with different etiology and preliminary effects of learning and memory in rats. Pharmacol Biochem Behav immunotherapy. Drug Dev Res 1988;15:165–9. 1999;64:183–9. [10] Gahtan E, Overmier JB. Inflammatory pathogenesis in [24] Popik P, Vetulani J, VanRee JM. Facilitation and attenuation of Alzheimer’s disease: biological mechanisms and cognitive sequeli. the rat’s social recognition by different oxytocin-related peptides. Neurosci Biobehav Rev 1999;23:615–33. Eur J Pharmacol 1996;308:113–6. [11] Georgiades J, Gelder F, Inglot AD. Isolation and preliminary [25] Porsolt RD, Roux S, Wettstein JG. Animal models of dementia. characterization of a new cytokine present in human colostrum. Drug Dev Res 1995;35:214–29. Its similarity to ovine Colostrinin. Eur Cytokine Netw [26] Ringheim GE, Szczepanik AM, Burgher KL, Petko W, Heroux 1996;7:511. JA, Cavalieri F. Transcriptional inhibition of the beta amyloid [12] Hyman BT, Van Hoesen GW, Damasio AR. Memory-related neural systems in Alzheimer’s disease: An anatomic study. Neu- precursor protein by interferon gamma. Biochim Biophys Res rology 1990;40:1721–30. Comm 1996;224:246–51. [13] Inglot AD, Janusz M, Lisowski J. Colostrinine: a proline-rich [27] Schmitt TL, Steiner E, Klinger P, Sztankay H, Grubeck-Loeben- polypeptide from ovine colostrum is a modest cytokine inducer stein B. The production of an amyloidogenic metabolite of the in human leukocytes. Arch Immunol Ther Exp 1996;44:215–24. Alzheimer amyloid beta precursor protein (APP) in thyroid cells [14] Janusz M, Inglot AD, Lisowski J, Piasecki E, Krukowska K, is stimulated by interleukin 1 beta, but inhibited by interferon Georgiades J. Colostrinin identified as new cytokine. Eur Cy- gamma. J Clin Endocrinol Metab 1996;81:1666–9. tokine Netw 1996;7:512. [28] Sei Y, Arora PK, Skolnick P, Paul IA. Spatial learning impair- [15] Janusz M, Lisowski J. Proline-rich polypeptide (PRP) — an ment in a murine model of AIDS. FASEB J 1992;6:3008– immunomodulatory peptide from ovine colostrum. Arch Im- 113. munol Ther Exp 1993;41:275–9. [29] Staroscik K, Janusz M, Zimecki M, Wieczorek Z, Lisowski J. [16] Janusz M, Lisowski J, Franek F. Isolation and characterization Immunologically active nonapeptide fragment of a proline-rich of proline-rich polypeptide from ovine colostrum. FEBS Letts polypeptide from ovine colostrum: Amino acid sequence and 1974;49:276–9. immunoregulatory properties. Mol Immunol 1983;20:1277–82. [17] Janusz M, Staroscik K, Zimecki M, Wieczorek Z, Lisowski J. [30] VanRee JM, Bohus B, Versteeg DHG, De Wied D. Neurohypo- Chemical and physical characterization of proline-rich polypep- physeal principles and memory processes. Biochem Pharmacol tide from sheep colostrum. Biochem J 1981;199:9–15. 1978;27:1793–800. [18] Janusz M, Wieczorek Z, Spiegel K, Kubik A, Szewczuk Z, [31] Wieczorek Z, Zimecki M, Janusz M, Staroscik K, Lisowski J. Siemion I, Lisowski J. Immunoregulatory properties of synthetic Proline-rich polypeptide from ovine colostrum: Its effect on skin peptides, fragments of a proline-rich polypeptide (PRP) from permeability and on the immune response. Immunology ovine colostrum. Mol Immunol 1987;24:1029–31. 1979;36:875–81. [19] Leszek J, Inglot AD, Janusz M, Lisowski J, Krukowska K, [32] Wirkus-Romanowska I, Miecznikowska H, Zablocka A, Rybka Georgiades J. Colostrinin; a proline-rich polypeptide (PRP) complex isolated from ovine colostrum for treatment of Alzheimer’s K, Fortuna W, Miedzybrodzki R, Szymaniec S, Janusz M, disease. A double-blind, placebo-controlled study. Arch Im- Lisowski J, Kupryszewski G. Fragments of a proline-rich munol Ther Exp 1999;47:377–85. polypeptide settled on a hexapeptide carcass constructed of [20] Morris RG. Spatial localisation does not depend on the presence glycine and L-lysine residues. Synthesis and biological properties. of local cues. Learn Motiv 1981;12:239–60. Pol J Chem 2000;74:219–25. [21] Piasecki E, Inglot AD, Winiarska M, Krukowska K, Janusz M, [33] Zablocka A, Janusz M, Rybka K, Miecznikowska H, Lisowski J. Coincidence between spontaneous release of inter- Kupryszewski G, Wirkus-Romanowska I, Lisowski J. Cytokine feron and tumor necrosis factor by colostral leukocytes and the induction by proline-rich polypeptide complex and analogues of production of colostrinin by human mammary gland after nor- its active nonapeptide. Folia Histochem Cytobiol 1999;1(Suppl. mal delivery. Arch Immunol Ther Exp 1997;45:109–17. 37):47.