**4. Discussion**

In this study, we generated and characterized mutant mouse strains with selective deletion of TCTP in neural precursor cells mediated by *Nestin*-driven Cre expression. We found (1) a dramatic phenotype of growth retardation, (2) early postnatal death, (3) enhanced neuronal loss and functional defect on Tuj1 and doublecortin-positive neurons, (4) increased apoptosis and decreased proliferation, and (5) decreased expression of mMcl-1, Bcl-xL, hax-1, and Oct4 in *Nestin-cre*-driven TCTP conditional knockout mice. Furthermore, we also demonstrated an obligatory role of TCTP in the maturation process of neuronal progenitor cell development.

In prenatal development, neurogenesis is responsible for populating the growing brain. New neurons are continually made throughout adulthood in predominantly two regions of the brain, the subventricular zone (SVZ) lining the lateral ventricles, where new cells migrate to the olfactory bulb via the rostral migratory stream, and the subgranular zone (SGZ), part of the dentate gyrus of the hippocampus. We observed high levels of TCTP protein expression in those regions in wild-type mice at P0.5. Moreover, loss of TCTP protein in the brain caused decreased BrdU-positive cells in neurogenetic regions, and the size of the lateral ventricle of the mutant telencephalon was significantly larger relative to littermate controls at P1. These results sugges<sup>t</sup> that TCTP regulates prenatal brain development and is involved in new neuron proliferation. Importantly, TCTP has been shown to bind to Mcl-1 and Bcl-xL, antiapoptotic members of the Bcl-2 family [9,36,37]. TCTP is therefore closely associated with apoptotic processes. The TUNEL-positive apoptotic cell-prone nature of the TCTP-deficient brain was first noted at E13.5 and significantly increased at E16.5, with maximal apoptotic cells found at P0.5. We showed that loss of TCTP expression dramatically decreased Mcl-1 and Bcl-xL protein expression then increased cleaved caspase-3 protein levels in apoptotic cells and during neurogenesis on postnatal day 0.5. This phenotype, together with the proliferation defect observed in TCTP mutants, is su fficient to explain why the TCTP-null mice died early during neurogenesis. These results are consistent with the finding from the study of *Xenopus laevis* [21], which demonstrated that TCTP regulates retinal axon growth through the interaction with Mcl-1 and Bcl-xL protein to inhibit caspase-3 activation. Our study demonstrated in a mammalian model system that TCTP is also important for whole brain development and that this function is essential for the perinatal survival of mice. We also found that loss of TCTP in the brain caused a significant decrease in Hax-1 protein expression. Previous studies indicated that Hax-1 was involved in the regulation of apoptosis or programmed cell death [38]. Hax-1 and TCTP have been reported to regulate calcium homeostasis. Homozygous deletion of Hax-1 in mice results in excessive apoptosis of neurons and postnatal death caused by a loss of motor coordination and function, leading to a failure to eat or drink [38]. These phenotypes mimic TCTP mutant mice, suggesting that TCTP might interact with Hax-1 directly or indirectly.

On the other hand, Susini et al. proposed that TCTP could anchor the mitochondria and bind to MCL-1 and Bcl-xL to inhibit the dimerization of Bax and protect from apoptotic cell death by antagonizing the Bax function [33]. However, the antiapoptotic function of TCTP occurs in the mitochondria, where it inhibits Bax-induced damage. Our data did not show a significant di fference in the Bax protein level with whole-cell lysates when TCTP was deleted. An analysis with a purified fraction of mitochondria in the brain tissue is needed to substantiate our results and clarify this issue. There is Mcl-1 pro-survival large isoform (Mcl-1L) and pro-apoptotic short isoform (Mcl-1S) in cells. We detected Mcl-1L but not Mcl-1S. Therefore, a more suitable antibody and well-prepared cell lysates are needed to clarify the changes of Mcl-1L and Mcl-1S in cKO mice. In addition, p53 may contribute to the cell death induced by *TCTP* deletion. Recent evidence has demonstrated that TCTP can facilitate Mdm2-mediated ubiquitination of P53 by competing with and preventing the binding of Numb, an inhibitor of Mdm2, to the Mdm2-P53 complex [39]. Further study is needed to clarify the role of P53 in the defect CNS development of *TCTP* mutant mice. Our study exploring the role of p53 in neurodevelopmental defects of *TCTP* mutant mice with double knockout mice (*NestinCre*/+; *TCTPflox*/*flox*; *p53*−/−) is in progress. Our unpublished data showed that *P53* deficiency caused a more

severe phenotype of *GFAPCre*/+; *TCTPflox*/*flox* knockout mice with *GFAPCre*/+; *TCTPflox*/*flox*; *p53*−/<sup>−</sup> double knockout mice. In addition, Hsu et al. demonstrated that dTCTP functions upstream of mTOR-dS6K and regulates fly cell growth by positively regulating dRheb activity [40]. Our previous publication showed that the mTOR-dS6K pathway contributed to the regulation of TCTP in beta cells (17) but not fibroblasts (15) in a cell-type-specific manner.

In this study, we showed that TCTP is required for the proliferation, di fferentiation, and migration of neuronal progenitor cells. The *Nestin-cre*-derived TCTP disruption specific to neuronal progenitor cells caused the loss of Tuj1- or DCX-negative cells (Figure 5) and cell death in TCTP-negative cells (Figure 8) mediated through a cell autonomous mechanism. In contrast, it has been shown that TCTP could be secreted through the TSAP6 protein, which is also involved in exosome production [8]. Secreted TCTP also promotes liver regeneration and enhances colorectal cancer invasion, suggesting that TCTP protein functions as a cell nonautonomous regulator of cancer cell growth and proliferation [41,42]. Therefore, we could not exclude that the loss of TCTP in neuronal progenitor cells induces cell death through cell autonomous and nonautonomous mechanisms simultaneously.

A previous study showed that decreased levels of Oct-4 resulted in a failure to form the inner cell mass and cell number, lost pluripotency, and di fferentiated into troph-ectoderm [43,44]. Therefore, the level of Oct-4 expression in mice is vital for regulating pluripotency and early cell di fferentiation, including the brain, spinal cord, and nervous system, since one of its main functions is to keep embryonic stem cells from di fferentiating [44]. Oct-4 is also involved in the self-renewal of undi fferentiated neuronal stem cells by forming a heterodimer with Sox2 followed by binding to DNA [45]. Low Oct-4 expression sustains self-renewal but is deficient in di fferentiation [46].

Our in vitro results showed that a deficiency of TCTP, specifically in neuronal progenitor cells, led to decreased survival of neuron progenitor cells and reduced Oct-4 protein expression in vivo, suggesting that disruption of TCTP reduces the capability of Oct-4 to regulate the self-renewal of neuron progenitor cells. Our results sugges<sup>t</sup> that TCTP deficiency in neuronal progenitor cells leads to defects in transcription factor signaling, particularly impaired Oct-4 activation. This result is consistent with the finding that TCTP activates transcription of Oct-4 in *Xenopus laevis* oocytes [35] and kidney-derived stem cells [47] but not in mouse embryonic carcinoma P19 cells and J1 embryonic stem cells in vitro [48]. These results imply that Oct-4 may co-interact with TCTP protein and facilitate neuron cell di fferentiation. Furthermore, TCTP may mediate Oct-4-stimulated dependent or independent signals in neural stem cells (NSCs). The deficiency of TCTP causes the inhibition of NSC self-renewal with unidentified signaling pathways.

On the other hand, our previous study indicated that homozygous deletion of TCTP reduced cyclins D2 and E2 in the cell cycle G1 and S phases in developing embryos [15]. This result is consistent with selective deletion of TCTP in the brain, suggesting that TCTP plays a role in embryonic and brain cell proliferation via a mechanism that is linked to the regulation of the cell cycle machinery. More experiments will be required to address the molecular mechanisms involved in this interesting finding.

There was little or no milk in the stomachs of TCTP-deficient mice a few hours after birth. Previous studies indicated that the inability to feed resulted in neonatal death caused by the absence of nourishment but also because the liquid derived from milk is essential for homeostatic processes in newborns [49]. Our data show that TCTP deficiency triggering neonatal death was not rescued by oral administration of saline containing 15% glucose (Supplementary Table S1). Our data show that there was no di fference in the cardiac function detected by an echocardiogram with a VisualSonics Vevo 660TM high-resolution imaging system between TCTP-deficient mice (*Nestin-cre*; *TCTPf*/*f*) and littermate controls at E16.5 (unpublished results). Thus, the phenotype of early perinatal lethality of TCTP mutant mice was not caused by the heart problem. We also could not exclude the possibility that respiration failure may exist from the TCTP deficiency-induced defect in the respiration center in the pons and medulla. Several reports indicated that analysis of several neonatal lethal mutant mice revealed that nonfeeding or normal fasting newborns died 12 to 24 h after birth [49–51]. The TCTP mutant mice died between 24 and 36 h, suggesting that the inability to suckle milk was one possible

cause of neonatal death. Feeding problems may be due to rejection by the mother. The major cause of the inability to feed in TCTP mutant mice might be associated with neuromuscular dysfunction.

Loss of TCTP in the developing brain might contribute to interference with normal physiological functions, such as the ability to suckle milk, olfactory sensitivity, movement coordination, and muscular function. Suckling is a complex process that consists of many structures of the brain and nerves, as well as the muscles required to extract milk [52]. According to our observation, TCTP-deficient mice can find and attach to a female mouse nipple, but we are not convinced whether these mice have a suckling response. The suckling response includes nipple attachment, suckling with rhythmic movement of the jaw and tongue, and the stretch response [53]. The interactions between motor and sensory neuronal pathways are linked to the central nervous system through the brain stem trigeminal complex. The data from Supplementary Figure S3 do not show a significant di fference in the brain stem trigeminal ganglion between control and cKO mice detected by hematoxylin and eosin staining. To further confirm that the neurodevelopmental defects and phenotype of TCTP-cKO mice were due to the TCTP deficiency, *NestinCre*/+; *TCTPf*/+ mice will be crossed with hTCTP transgenic mice to produce double transgenic mice (*NestinCre*/+; *TCTPf*/+; *R26R-hTCTP*), which do not express the mouse TCTP protein but overexpress human TCTP, specifically in neuron progenitor cells. The phenotypes of the double transgenic mice will be compared with TCTP-cKO mice and *R26R-hTCTP* transgenic mice to check the rescue e fficiency of hTCTP overexpression. Cortical neurons [25] and cortical progenitor cells [26] will be cultured for further study to explore whether disruption of TCTP sensitizes the neurons to apoptosis induced by DNA damage. On the other hand, Mcl-1 is a key regulator of apoptosis during CNS development [18]. *Nestin-cre*-derived knockout of TCTP decreased the expression of mMcl-1 (Figure 8C). Human Mcl-1 overexpression might rescue the early neonatal death of TCTP-cKO mutant mice.

Our results show new findings on the requirement for TCTP in the development and maintenance of neurons within the CNS. In vivo and in vitro results demonstrated that TCTP is required for neurogenesis and survival of the neuron maturation process. Additionally, TCTP regulates apoptotic cell death. Most importantly, our results implicate TCTP as a key regulatory molecule, involved in expanding the neural precursor pool and maintaining neuronal survival in the newborn mouse brain.
