**4. Discussion**

Short kisspeptin analogs are promising candidates in the treatment of infertility and other gynecological conditions [48]. Kisspeptins exert their effect on the reproductive axis via Kiss1r [4], but they also bind to and activate NPFF1 and NPFF2 receptors with lower affinity [11]. In our study, we investigated the behavioral and biological effects of icv. Kp-8 in male rats via performing a battery of behavioral tests (EPM, OF, MB), determining serum corticosterone and LH levels, as well as measuring dopamine release from the VTA and NAc, and GABA release from the NAc.

The 0.1 μg dose of Kp-8 (but not the 1 μg dose) decreased the percentage of open arm entries and open arm time in the EPM, which is characteristic of anxiety-like behavior [51]. It is in accordance with our previous experiments in which a preference for closed arms has been observed following icv. treatment with Kp-13 [37]. Still, it must be noted that only a higher dose of Kp-13 has exerted an anxiogenic action, whereas in the case of Kp-8 an approximately 10-times lower dose was effective. The dose–response curve of Kp-8 shows

a bell-shape (or inverted U-shape), that has been reported in several studies involving neuropeptides [56–58]. This phenomenon, when a lower dose is stimulatory, whereas a higher dose is inhibitory or ineffective, is called hormesis [59]. A review by Calabrese has reported a wide range of explanations for hormetic responses, including receptorial and intracellular mechanisms. For example, the same substance might have a stimulatory effect in a low dose, but an inhibitory effect in a high dose either via the same receptor (often mediated by a so-called 'molecular switch'), or via different receptors to which it has higher and lower affinity, respectively [60]. Based on a review on RF-amides and their receptors, kisspeptins in general can bind to their cognant receptor, Kiss1r, and to NPFF receptors with different affinity [3], the latter of which depends on the length of the peptide: the full length Kp (in rats Kp-52) has a lower affinity to NPFF receptors, whereas the shorter endogenous derivatives' binding affinity to NPFF receptors is higher. Furthermore, Rouméas et al. have performed systemic N-terminus deletions and benzoylations of Kp-10, which has revealed a progressive loss of affinity of the shorter fragments to Kiss1r and a conserved high affinity to NPFF receptors. In contrast, these shorter benzoylated fragments could still act as full agonists on Kiss1r, whereas on NPFF receptors a partial agonistic action has been observed [61]. How the benzoylation affects the affinity profile of these Kp-10 fragments is not known, yet it is possible that the unmodified Kp-8 also has an altered binding profile. In point of fact, agonists of NPFF1 and NPFF2 receptors have been implicated in anxiety [62,63]. Indeed, both Kiss1r and NPFF receptors could mediate the anxiety-like action of Kp-8, and we cannot rule out the possible activation of other receptors, as well.

Kp-8 also induced an elevation in serum corticosterone concentration: the 0.1 μg dose showed a tendency to increase it, whereas the 1 μg dose was significant. Corticosterone elevation is indicative of the activation of the HPA axis, the parvocellular neurons in the hypothalamic paraventricular nucleus (PVN) might have released CRH and AVP, followed by the secretion of ACTH from the pituitary, which consequently triggered the secretion of glucocorticoids from the adrenal cortex. In a study by Rao et al., Kp induced an increase in AVP mRNA expression in PVN-derived cell lines [36], and thus it is possible that Kp-8 activated the axis by increasing AVP release. Moreover, the activity of the HPA axis is modulated by limbic brain regions, including the amygdala [64]—an expression site of both Kp and Kiss1r [65]—which stimulates the HPA axis and regulates the behavioral response to stress [64,66]. The increase in glucocorticoid signaling in itself is also associated with anxiety-like behavior [67]. This result ties well with our previous study in which icv. Kp-13 has also caused an elevation in corticosterone concentration in a higher dose [37].

However, in the first 5 min of computerized open field test, the animals treated with 1 μg of Kp-8 spent more time in the central zone of the arena, which is considered a sign of anxiolysis [68]. As there was no difference in central locomotor activity at any other time point, this result should be interpreted with caution. At the beginning of the OF, the animals are placed in the center of the arena, so it is possible that the increase in central time reflects an initial latency in approaching the periphery rather than a real anxiolytic effect.

In addition, it is not uncommon to have discrepancies between the EPM and OF results. For example, chlordiazepoxide has reduced anxiety-like behavior in the EPM, but has had no significant effect in the OF in Lewis rats [69]. Although the principles of OF and EPM are similar, the two tests seem to load on different factors of anxiety [70]. Moreover, the approach of open arms in the EPM and central locomotion in the OF seem to be independently inherited in rats [71].

Altogether Kp-8 seemed to increase anxiety-like behavior and activate the HPA axis. These results are in accordance with the anxiogenic effect of icv. Kp-13 in rats [37] and the anxiolysis observed in *Kiss1r* KO mice [38]. However, Kp has not influenced anxiety in rats in the study by Rao et al. [36], which might be attributed to the peripheral route of administration and the relatively low dose of Kp used in the experiment. Likewise, intravenous Kp has had no effect on anxiety in human subjects [32]. Apart from the route of administration, another important factor to consider is the species: the Kp system of

zebrafish greatly differs from that of mammals, which might explain the anxiolytic property of kisspeptin observed in the study of Ogawa et al. [39]. Moreover, when compared to a systemic treatment, the regional modulation of neuronal activity can have strikingly different consequences. For example, the selective activation of kisspeptin neurons in the medial posterodorsal amygdala has decreased anxiety [40].

The number of total arm entries in the EPM reflects the general locomotor activity of the animals [72]. Both the 0.1 and 1 μg doses of Kp-8 reduced the number of arm entries, suggesting that Kp-8 might cause hypolocomotion.

The 60-min OF also yielded some remarkable results. When placed in a novel environment, all groups exhibited a pronounced exploratory activity with intense ambulation and a high number of rearings. Following a gradual decline until approximately 30 min, activity returned to a basal level. From that point, differences have started to appear among the groups, as there was a decrease in ambulation and rearing activity, as well as an increase in immobility in the group treated with 1 μg of Kp-8. These results point to a decrease in spontaneous locomotion.

Kp-8 has also significantly reduced the number and time of goal-oriented interaction with marbles in the MB. Although MB has long been considered a test for anxiety-like behavior, now several authors have expressed doubts about it [73]. According to Thomas, the number of buried marbles does not correlate with other anxiety-like traits, namely central time in the open field test and light-dark transitions in the light-dark box test [74]. The utility of the test as a screening tool for anxiety has also been questioned based on the findings that most anxiolytics and antidepressant drugs reduce marble burying behavior secondary to drug-induced hypolocomotion [73]. It has been suggested that digging and burying are species-specific, innate behavioral patterns that are likely triggered by an exploratory drive [75] or by the bedding itself [76]. Nowadays marble burying is regarded as a sign of repetitive, compulsive-like behavior, which is highly dependent on general locomotor activity [52]. Consequently, the reduction in goal-oriented interactions with the marbles is most likely a sign of suppressed locomotion in our study.

These findings contrast with the results reported by our group on the effects of icv. Kp-13, as it has increased exploratory and spontaneous locomotion in rats [37]. In a study by Tolson et al., *Kiss1r* KO female mice have exhibited a decrease in spontaneous locomotion and energy expenditure, but the mutation has had no such effect in male animals [43], pointing to a possibly gender-dependent effect.

Icv. Kp-8 has stimulated LH release in our study, which is a sign of reproductive axis activation, secondary to Kiss1R binding and activation in the hypothalamus. Kiss1R is expressed on hypothalamic GnRH neurons [4], and upon its activation repetitive LH pulses are generated [77]. In our study, the 1 μg dose of Kp-8 caused a significant increase in LH concentration. This is in accordance with literature data as icv. injection of a similar dose of Kp-10 has exerted an LH surge [41]. It must be noted though, that 0.1 μg of Kp-8 did not affect LH release. This is not surprising since Thomson et al. have obtained a similar result when Kp-10 was administered icv in a similarly low dose [41]. Furthermore, in a study by Pheng et el., icv. administered Kp-10 at a similarly low dose was unable to stimulate LH release in male rats, only the full length Kp-52 did. The authors have postulated that slower degradation of Kp-52 might explain their results, but it is also possible that the different binding profiles of Kp-52 and Kp-10 are in the background [78]. This also might explain our result, since Kp-8 similarly to Kp-10 might bind with higher affinity to the NPFF receptors, more specifically to NPFF1 receptor. One of the ligands of NPFF1 receptor is the RF-amide-related peptide 3 (RFRP-3), which has an inhibitory effect on the reproductive axis in adult male rats [79,80]. Thus, it is possible that at a lower dose, the two opposing actions of Kp-8 result in no change in LH concentration. Nevertheless, further studies are required to determine the affinity of Kp-8 to its receptors and the degree of calcium mobilization upon receptor activation. As the hypolocomotor effect of Kp-8 seems to be in contrast with previous studies on kisspeptin and locomotion [37,43], it is likely that this effect is mediated by other mechanisms.

One possible explanation is the activation of NPFF receptors. Kp-8 has activated human NPFF2 receptors in *Xenopus* oocytes [81], and its N-terminally benzoylated form has shown to fully preserve the affinity of Kp-10 to NPFF1 and NPFF2 receptors [61], which are universally activated by all members of the RF-amide family [3].

It is noteworthy that several members of the RF-amide family have been reported to modulate locomotor activity, pointing to the possible role of NPFF receptors in the regulation of locomotion. Similarly to Kp-8, icv. treatment with RF-amide related peptide 1 (RFRP-1) has reduced total locomotor activity and has also induced anxiety-like behavior and HPA axis activation [63]. Likewise, intra-VTA injection of NPFF has reduced spontaneous locomotion in rats [82]. Interestingly, icv. NPFF has inhibited morphine-induced hyperlocomotion, but has failed to affect the locomotor activity of naïve rats [83]. Although icv. neuropeptide AF (NPAF) has also had an anxiogenic effect, contrary to RFRP-1 and NPFF, it has stimulated spontaneous and exploratory locomotion [56].

Another possible reason for the development of hypolocomotion is the modulation of the mesocorticolimbic dopaminergic system. Based on the expression of kisspeptin in the NAc, as well as the expression of NPFF1 and NPFF2 receptors in the NAc and VTA [16,44], it was reasonable to investigate whether Kp-8 has a direct effect on the VTA-NAc circuitry. The dopaminergic pathway connecting the VTA and NAc has long been implicated in the regulation of locomotion. As a matter of fact, VTA dopaminergic neurons are responsible for the locomotor-enhancing effect of cocaine [84]. Our hypothesis was that Kp-8 might suppress locomotion by directly modulating the activity of VTA dopaminergic neurons. However, in our ex vivo superfusion study, Kp-8 has not affected dopamine release from slices obtained from the VTA and NAc.

As the interaction between Kp and GABA is known from the literature [85,86], it also seemed possible that Kp-8 might directly affect GABA release in NAc. GABAergic neurons in the NAc have been shown to inhibit dopaminergic projections from the VTA [87]. In fact, GABAergic activity can also be connected with the suppression of locomotion, as locomotor activity has increased when GABAA receptor antagonists were injected into the NAc core [46]. In our study, Kp-8 significantly increased GABA release from NAc slices. This result suggests that Kp-8 might directly modulate the activity of GABAergic neurons in NAc, which could contribute to the suppression of locomotor activity. It must be mentioned, however, that ex vivo superfusion measures only the direct effect of Kp-8 on live tissue slices obtained from the NAc, but the complex assessment of the whole VTA-NAc circuitry is beyond the scope of this method. Consequently, further studies (e.g., in vivo microdialysis) are required to confirm these findings on the circuit level.

When considering the receptors involved, it is possible that Kp-8 alters GABA release via NPFF1 or NPFF2 receptors, which are abundantly expressed in the VTA and the NAc, and likely involved in the modulation of both dopaminergic and GABAergic neuronal activity [16]. The role of NPFF receptors is further supported by the results of Cador et al., who have reported a decrease in novelty-induced locomotion upon intra-VTA NPFF treatment. Kiss1r expression, however, has only been detected in the NAc of humans, but not in rodents [88], so it is unlikely that Kp-8 could modulate NAc activity via Kiss1r.

Furthermore, the contribution of altered metabolism and thermoregulation should not be ruled out in the background of altered locomotor activity. Kisspeptin's stimulatory effect on locomotion seemed to be coupled with metabolic effects in the literature. Icv. Kp-13 has induced hyperthermia [37], and *Kiss1r* KO has resulted in obesity, increased adiposity and impaired glucose tolerance in female mice [43]. Kp could also be involved in hypothalamic appetite regulation by exciting proopiomelanocortin (POMC) neurons and inhibiting neuropeptide Y/Agouti-related peptide (NPY/AgRP) neurons, resulting in an anorexigenic effect [89]. Although only a few studies have addressed the metabolic effects of other RF-amides, they have usually revealed significant results. Icv. NPFF has reduced food intake in food-deprived rats [90] and also had a hypothermic effect in mice [91]. Moreover, the stimulation of central NPFF1 and NPFF2 receptors have evoked hypothermia and hyperthermia, respectively [92].

Alternatively, Kp-8 might modulate the activity of other, locomotion-related systems differently than the naturally occurring kisspeptins, resulting in an opposing effect of locomotion. For example, central Kp-10 treatment has stimulated vasopressin release in rats [93], and vasopressin has induced hyperlocomotion by acting on V1a receptors on hypothalamic orexin/hypocretin neurons in mice [94]. Furthermore, kisspeptin has been shown to induce BDNF expression in the hippocampus [95] and the lack of active BDNF in tissue plasminogen activator deficient mice has been associated with a decrease in nocturnal wheel running activity [96]. It is a question of future research to investigate whether Kp-8 could modulate vasopressin release and BDNF secretion in a similar or different fashion as other kisspeptins.

**Author Contributions:** Conceptualization, G.S., G.T., and K.C.; data curation, K.E.I. and K.C.; formal analysis, Z.B. (Zsolt Bagosi); funding acquisition, G.S.; investigation, K.E.I., É.B., and Z.B. (Zsolt Bozsó); methodology, É.B., Z.B. (Zsolt Bagosi), and Z.B. (Zsolt Bozsó); project administration, K.E.I.; resources, G.T.; supervision, G.S. and K.C.; visualization, K.E.I.; writing—original draft, K.E.I.; writing—review & editing, K.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Hungarian Government and the European Union through the EFOP-3.6.2-16-2017-00006 grant.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the University of Szeged Ethical Committee for the Protection of Animals in Research. Permission for the experiments (number: X./1207/2018, date: 6 July 2018.) has been granted by the Government Office of Csongrád County Directorate of Food Chain Safety and Animal Health.

**Data Availability Statement:** Data is contained within the article.

**Acknowledgments:** We thank Ágnes Pál, Gusztáv Kiss and Veronika Romhányi for the excellent technical support during the experiments.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **Appendix A**

Following solid-phase peptide synthesis, Kp-8 was analyzed using HPLC and ESI-MS, the results of which are shown in Figures A1 and A2, respectively.

**Figure A1.** HPLC trace of Kisspeptin-8. Column: Phenomenex Luna C18, 5 μ, 100 Å, 4.6 mm × 250 mm, flow rate: 1 mL/min, wavelength: 220 nm, A eluent: 0.1% TFA in water, B eluent: 0.1% TFA/80% ACN/water, gradient: 30–55% eluent B in eluent A over 25 min.

**Figure A2.** The ESI-MS trace of Kp-8 peptide.

#### **Appendix B**

During the computerized open field test, data were collected for 60 min, in 5-min timeframes. The cumulative results obtained after 60 min of data collection can be seen in Figure A3. There was no significant difference in any of the parameters measured: total distance of ambulation (Figure A3a, F (2, 34) = 1.691, *p* = 0.1994), total time of ambulation (Figure A3b, F (2, 34) = 1.728, *p* = 0.1928), time spent immobile (Figure A3c, F (2, 34) = 1.274, *p* = 0.2927), number of rearings (Figure A3d, F (2, 34) = 1.522, *p* = 0.2328), percentage of central ambulation distance (Figure A3e, F (2, 34) = 0.6885, *p* = 0.5092), and percentage of central ambulation time (Figure A3f, F (2, 34) = 0.7265, *p* = 0.4910).

**Figure A3.** Cumulative data of 60-min open field test: (**a**) total distance of ambulation, (**b**) total time of ambulation, (**c**) total time spent immobile, (**d**) number of rearings, (**e**) percentage of distance travelled in the central zone, (**f**) percentage of time spent in the central zone, *n* = 12–13.
