The tridecapeptide neurotensin (NT) was isolated from bovine hypothalami on the basis of its ability to induce vasodilatation [1]. NT is synthesized from a precursor protein following excision by prohormone convertases [2]. Since NT is expressed both in the central nervous system and in the gastrointestinal tract [3], the peptide exerts a role of a neurotransmitter/neuromodulator in the central nervous system and a role of a hormone in the periphery.

In the central nervous system, NT is released from neuronal synaptic vesicles [4]. The peptide regulates the dopaminergic system [5], induces an opioid-independent analgesia [6] and a potent hypothermia [7]. A role for NT in schizophrenia was suggested in 1980 [8] and further confirmed by the observation that NT level decreases in the brain of severe schizophrenics [9]. Low NT level can be reversed to a normal level after one month treatment with anti-psychotic drugs [10].

In the periphery, NT is released in the circulation after a meal [11] and stimulates pancreatic secretion in order to facilitate lipid digestion [12]. NT is also involved in the protection of pancreatic beta cells against cytotoxic molecules [13] and in the regulation of insulin secretion [14,15].

The numerous activities of NT are mediated through three identified NT receptors (NTSRs). Two of them, NTSR1 and NTSR2, are classical neuropeptide G protein coupled receptors (GPCR) bearing seven transmembrane domains (TMs) [16,17]. The third one, NTSR3, also called sortilin [18,19], is a type I receptor, non-coupled to G proteins, which belongs to the Vps10p-containing domain receptor family [20,21].

The homeostasis of tissues results from a fine equilibrium between cell differentiation, proliferation and apoptosis. Both proliferation and apoptosis display crucial roles in the growth control and development of normal and neoplastic tissues. Therefore, dysregulation of one of the pathways controlling proliferation or apoptosis leads to carcinogenesis by dysregulation of normal cell cycle [22]. Apoptosis is regulated by multiple intracellular pro- and anti-apoptotic proteins.

The protein B-cell lymphoma/leukemia 2 (Bcl-2) was the first mitochondrial protein shown to block apoptotic programmed cell death [23,24]. Since the discovery of Bcl-2, a number of proteins sharing structural homology and involved in the control of apoptosis has been characterized [25]. Members of the Bcl-2 family are globular proteins structured in α-helices, which possess at least one Bcl-2 homology (BH) domain. However, not all Bcl-2 family members have the same anti-apoptotic function. Bcl-2, Bcl-xL and Mcl-1 are anti-apoptotic; they contain four BH domains and a transmembrane (TM) domain. The proteins Bax and Bak, which also contain four BH domains and a TM domain, are pro-apoptotic. A second group of pro-apoptotic proteins possessing only one BH domain is composed of Bik and Hrk (with a TM domain) and Bad and Bid (without TM domain) [26,27]. Bcl-2, like other anti-apoptotic members, is able to form heterodimers with pro-apoptotic proteins, like Bax. This prevents the permeabilization of the outer mitochondrial membrane and then the release into the cytosol of cytochrome C, responsible for the activation of the caspases cascade [28]. Following cellular stresses, the p53 tumor suppressor, which is another representative of pro-apoptotic proteins, limits cell growth by inducing cell cycle arrest and apoptosis. Since p53 mediates apoptosis through a pathway involving Bax, its activity can be blocked by anti-apoptotic Bcl-2 family members [29]. Numerous studies demonstrated a role of NT in the regulation of apoptosis through the activation of the Bcl-2 family proteins. In the present review, we will focus on the history of the discovery of the anti- and, with a lesser extent, the pro-apoptotic action of NT both in central and peripheral tissues.

The neurotensinergic system is altered in disease-related neurodegeneration. In the temporal lobe of Alzheimer’s disease (AD) patients, the expression levels of both NTSR1 and NTSR2 were profoundly decreased in AD, whereas the expression of NT was slightly reduced, and the level of NTSR3/sortilin did not vary [30]. NT and its receptors are also modulated during age-related neurodegeneration. A study demonstrated that the absence of NTSR3/sortilin prevents age-related neuron degeneration, but not developmental neuronal apoptosis [31]. Also, in the retina, NTSR3/sortilin-deficient mice showed reduced neuronal apoptosis [31]. Although the neurotensinergic system is clearly modified during ageing and neurodegenerative diseases like AD, its functional involvement was not established.

However, it is important to remember that NTSR3/sortilin is able to form heterodimers with neurotrophins receptors, like p75NTR or TrkA and TrkB, to trigger proNGF or proBDNF-induced neuronal cell death (for a review, see [32]). The first demonstration of the involvement of NTSR3/sortilin in neuronal cell death was performed in 2004 in an original work showing that both sortilin and p75NTR form a protein complex, which is crucial for the apoptotic effect of the precursor form of NGF (proNGF) on neurons [33]. Interestingly, this work provides also evidence for a protective role of NT on neurons. The peptide antagonizes the effect of proNGF by competition on the binding site of proNGF on sortilin [33]. Moreover, another work demonstrated that in ageing rodent basal forebrain and sympathetic neurons, both proNGF and sortilin expression were increased, whereas p75NTR levels remained unchanged [34]. In this context, sortilin was involved in the age-related proNGF-mediated neurotoxicity, and the survival of aged neurons was rescued by NT.

Further works have shown that proBDNF-induced neuronal apoptosis involves a receptor complex formed also by sortilin and p75NTR. In this system, NT is also able to counteract the proneurotrophin-induced apoptosis by competition for the sortilin binding site [35]. Another neurotrophin precursor (proNT3) uses the same protein complex (sortilin/p75NTR) to trigger inner ear neuron death [36]. In this case, the role of NT has not been investigated.

Sortilin has been shown to participate in light-dependent photoreceptor degeneration in vivo. This work evidenced that the expression of p75NTR, sortilin and proNGF is increased following intense illumination, a process leading to retinal cell death, likely through proNGF-induced neurodegeneration. Pharmacological blockade of sortilin with either NT (10 µM) or the pro-domain of proNGF (100 nM) enhances the survival of cone-progenitor 661W cells subjected to intense light. This demonstrated that NT and the pro-domain of proNGF are able to antagonize, by competition to the binding site of sortilin, the proNGF-induced cell apoptosis [37]. However, the concentration of NT used in this work is largely higher than the physiological concentration of the peptide, which varies from 50 pM in the brain [38] to 100 pM in plasma serum [39]. Taking into account that NT is expressed in amacrine cells of the retina [40] and the three known NT receptors have been detected in human corneal keratocytes [41], NT could serve as a physiological regulator. NT can protect from intense illumination damage, as suggested by the action of the NT analogue JMV449, which at a physiological concentration of 1 or 10 nM decreases keratocyte apoptosis [41].

Finally, an important protective effect of NT through its interaction with the NTSR1 has been demonstrated in a stroke model of focal cerebral ischemia [42]. In this work, the NTSR1 agonist ABS-201, a stable NT analogue originally developed for its ability to cross the blood-brain barrier [43], protected adult mice brain from focal ischemia by decreasing body and brain temperature. ABS-201 treatment immediately or 1h after the ischemia leads to a significant decrease (40%) of infarct volumes. The interaction of ABS-201 with NTSR1 decreases ischemia-dependent caspase-3 activity and increases Bcl-2 expression [42]. These results are crucial, since pharmacological induced hypothermia elicited by NT analogues represented by ABS-201 are promising candidates for treatment of ischemic stroke and possibly for other ischemic or traumatic injuries.

We cannot complete this review without describing some properties of NT that lead to a pro-apoptotic action of the peptide, essentially observed in the central nervous system. In the cultured cortical neurons model, glutamate exerts exocytotoxicity via N-methyl-D-aspartate (NMDA) receptors, leading to an apoptotic cell death [61]. It has been shown that NT significantly enhances glutamatergic signaling, both in vitro and in vivo, by increasing glutamate release [62,63,64,65]. The fact that the cerebral level of NT increases following ischemia [66] could suggest an involvement of the peptide in ischemic brain damage. However, other experiments provide evidence that NT-induced hypothermia improves neurologic outcome and reduces infarct volume after hypoxic ischemia [67]. Indeed, the neuroprotection caused by NT-induced hypothermia in vivo [42,68] is not in agreement with the enhanced pro-apoptotic action of glutamate by NT after oxygen and glucose deprivation of primary cortical cultures [69]. The latter authors hypothesized that the difference in the NT effects may be explained by the influence of temperature. Experiments carried out on primary cultures of neurons are performed in vitro under normothermic conditions, whereas, in experiments achieved in vivo, animals can modulate their temperature. As NT can decrease body and brain temperature, the cytotoxic effect of glutamate is likely partially or even totally blocked or reversed. Further experiments on cultured neurons carried out under reduced temperature, 34 °C instead of 37 °C to mimic NT-induced loss of temperature, should answer this problem.

Although NT has been also shown to stimulate the release of inflammatory cytokines and chemokines from peripheral cancer cells [70] and from central microglial cells [71], the overall action of the peptide, as referenced in this review, is a protective effect (Table 1). This cellular and tissue protection is essentially mediated by a broad spectrum of an anti-apoptotic action of the peptide both in the periphery and in the central nervous system. The role of NT against cell apoptosis is triggered either by its classical GPCR NTSR1 and NTSR2 or by the NTSR3/sortilin, a non-G protein coupled receptor. The NT GPCRs usually mediate their action either directly as demonstrated for the protection of brain infarct against ischemia [42] for NTSR1 or in combination with NTSR3/sortilin, as shown for the protective action of NT on pancreatic beta cells [55]. Inversely, the involvement of NTSR3/sortilin is always made within a protein complex containing other NT receptors or receptors of neurotrophins, like p75NTR. However, when NTSR3/sortilin is associated with another NT receptor, the anti-apoptotic effects of NT are direct through NTSRs activation, as exampled in beta cells [55], and in human colonic cancer cells [72] (Figure 1). By contrast, when NTSR3/sortilin is associated with p75NTR, NT itself does not trigger receptor complex activation, but the peptide antagonizes the pro-apoptotic effect of either proNGF [33] or proBDNF [35] by competition with the pro-domains for the same binding site (Figure 1).

In conclusion, on one hand, the anti-apoptotic action of NT is beneficial when it concerns the protection of cells or tissues against cytotoxic agents (IL-1β, proNGF, proBDNF). On the other hand, the anti-apoptotic effect of NT is deleterious when it concerns the growth of neoplastic cells. The latter case necessitates the blocking of NT receptors, which are overexpressed in tumors. Therefore, the use of NT or NT analogues in therapy remains difficult to set up without taking into account these particular properties.