CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Serial No. 62/398,514, filed September 22, 2016, the entire contents and disclosures of which are hereby incorporated by reference into this application. This application also cites various publications, the entire contents of which are incorporated herein by reference into this application.

FIELD OF THE INVENTION

[0002] This invention relates to a method of preventing, ameliorating and/or treating diseases and disorders associated with the integrated stress response arising from cellular stresses involving the phosphorylated eukaryotic initiation factor 2a (p-eIF2α) pathway. In one embodiment, the present invention provides a method which prevents or alleviates aberrant cell differentiation associated with the p-eIF2α pathway.

BACKGROUND OF THE INVENTION

[0003] Congenital skeletal disorders in human lead to physical disabilities and generate difficulties in education, employment and social life (1). In particular, children with significantly short stature are vulnerable to diverse developmental, social and educational problems (2). These barriers are highly likely to exert a strong influence on quality of life. However, current treatment options in skeletal disorders are extremely limited and may involve controversial surgical procedures such as limb lengthening (3).

[0004] It is well established that mutations in genes encoding extracellular matrix (ECM) components cause skeletal disorders. For example, mutations in genes encoding collagen I cause osteogenesis imperfecta (OI) (4): Metaphyseal chondrodysplasia, type Schmid (MCDS) is associated with heterozygous mutations in the COL10A1 gene (5); Pseudoachondroplasia (PS ACH) and multiple epiphyseal dysplasia result from mutations in COMP, matrilin-3 or collagen IX (6-8). However, it has long been debated whether these disorders arise because of haploinsufficiency of the ECM or because of the intracellular dominant negative impact of these mutant proteins. Recent molecular evidence supports the notion that the underlying pathology is the consequence of retention of such mutant ECM proteins in the endoplasmic reticulum (ER), which induces ER stress and the adaptive unfolded protein response (UPR) (9, 10) . Targeting the UPR is therefore a strategy for the treatment of skeletal disorders associated with ER stress.

[0005] MCDS patients are characterized by short stature, waddling gait, and genu varum (bowing legs). Radiographic analyses also reveal coxa vara (the angle between the proximal head and the shaft of the femur is less than 120°), wider and irregular growth plates, mild spinal syndromes including platyspondyly (flat vertebral bodies), and abnormal vertebral endplates (11). Previously, a transgenic mouse model ( 13del) has been generated to carry an equivalent human mutation ( 13bp deletion) in the CollOal transgene and recapitulates phenotypic features of MCDS (10) and in addition generalized hyperostosis (U.S. Patent No. 6,369,295). The 13 -bp deletion causes a frameshift in the NCI coding domain of collagen X (Fig. 1A), and results in misfolding of this protein in the ER and activation of UPR in 13del-HCs (Fig. IB).

[0006] Significant dwarfism can be observed in mi1c3ed (eFlig. 1C). In human, the average height of MCDS children is approximately 85% of normal at the age of five (12). In mice, th1e3 bdoedl y length is about 90% of normal at plO (equivalent to 5 years old in human), and progresses to 85% of normal after weaning. The 1 m3diecle also feature disproportionate dwarfism, coxa vara (Figs. ID and IE), and other MCDS-like phenotypes (summarized in Fig. IF), The most characteristic skeletal phenotype of MCDS mice is the expansion of the hypertrophic zone in the growth plates of long bones. This phenotype is consistent with a reverted chondrocyte differentiation program, marked by re-expression of premature chondrocyte markers (Ppr, Sox9 and CoI2al) (Fig. 1G) (10). Although chronic ER stress would normally trigger cell apoptosis to eliminate stressed cells (13) , 13deI-HCs survive the ER stress (Fig. 1H) (10).

[0007] The subject invention disclosed a model in which induction of the UPR in hypertrophic chondrocytes (HCs) changes the differentiation program to a less differentiated state that allows them to adapt and survive. As a consequence, the changed differentiation program causes the growth plate abnormalities and consequent skeletal deformities in MCDS (10).

[0008] As discussed in the following sections, the present model places the UPR as the underlying molecular pathology of MCDS and raises the possibility of using the 13del mice as a preclinical animal model to develop potential interventions to alleviate the MCDS phenotype. A prerequisite of such pilot intervention is to demonstrate a direct molecular link between the UPR and the 13del phenotype.

[0009] Sustained activation of UPR has been implicated in the progression of a variety of diseases, including cancer, diabetes, inflammatory disease and neurodegenerative disorders (14). In the past few years, UPR is becoming an attractive target for drug discovery. Bioactive small molecules targeting the UPR pathway have been tested in chondrodysplasia animal models (15, 16). However, the disease phenotypes were not improved (15) or even worsened (16) after the administration of chemical chaperones or ER-stress reducing reagents. These findings suggest that UPR contributes to chondrodysplasia not just through the protein folding pathway, and the pathogenesis of this ER stress/UPR associated disease might involve additional UPR pathway effector(s) that is critical for normal skeletal physiology and homeostasis.

[0010] Upon ER stress, UPR activates three independent ER stress sensors: inositol-requiring 1 (IRE1), PKR-like ER kinase (PERK), and membrane-tethered activating transcription factor 6 (ATF6) (17). Amongst, activation of the PERK signaling pathway is likely to be the first line of defense against ER stress and is a central part of more general integrated stress response (ISR), activated by diverse stress stimuli and is implicated in many diseases including cancer, diabetes, obesity, neurodegeneration, skeletal disorders (18). Recently, ISR has been implicated in intervertebral disc degeneration (19, 20), which is very common in humans, and often causes low back pain (LBP).

[0011] In the past decades, surgical therapy has been undertaken to correct bone deformities in patients with short stature, including achondroplasia, hypochondroplasia, MCDS and other skeletal dysplasias. Current treatment of disc degeneration and LBP is mainly surgical, often involving removal of the disc and spinal fusion (21). Currently, there are no effective pharmacological therapies for abovementioned skeletal diseases. Growth hormone (GH) therapy has been used to treat dwarfism but is clarified to be ineffective for height gain in most congenital skeletal dysplasias, and in some cases with severe spinal deformities, it even results in worsened kyphosis and lordosis (22). Studies have also been reported to test the feasibility of chemical chaperone or ER-stxess reducing reagents in animal models of ER stress related chondrodysplasia. Mouse models of chondrodysplasia have been treated with various ER stress reducing reagents: lithium, valproate, or phenylbutyric acid (PBA) and found to be ineffective (15, 16).

[0012) Integrated stress response (ISR) has a central role in many forms of cellular stress such as oxidative stress, hypoxia, ER stress and its induction is associated with diverse common diseases, such as cancer, diabetes, fibrosis, obesity, neurodegeneration and skeletal disorders (18). With activation of the PERK pathway as an immediate early response of the UPR and the ISR, PERK modulators have the potential in the treatment of cancer and neurodegenerative diseases (78-80). In particular to neurodegenerative diseases, although restoring protein synthesis through pharmacological inhibition of PERK in prion-infected mice using GSK2606414 could be neuroprotective, long term benefit could not be assessed because of side effects of weight loss and pancreatic toxicity (79). Pharmacological inhibition of PERK mediated phosphorylation of tau in a transgenic model of Frontotemporal Dementia was shown to be protective against further neuronal loss (81). These approaches address the postnatal impact of activating the UPR and PERK in neurodegeneration and also highlight the importance of determining the precise mechanism of causality of the UPR under specific scenarios of degeneration versus cell differentiation in development (82).

[0013] While stress responses commonly result in apoptosis, understanding how cells adapt, survive and a molecular understanding on the consequences of inducing the ISR on cell fate and differentiation in vivo is lacking. The majority of molecular mechanistic insights of the impact of the ISR are based on cell based assays not in vivo. Through an in vivo model of human chondrodysplasia, the present invention for the first time provides a mechanistic insight into the question of impacts of the ISR on cell fate and importantly addressed the possibility of preventive therapy.

BRIEF SUMMARY OF THE INVENTION

[0014] The present invention describes a method of preventing, ameliorating and/or treating disorders or diseases associated with the integrated stress response (ISR) involving the phosphorylated eukaryotic initiation factor 2a (p-eIF2α) pathway arising from various cellular stresses such as oxidative stress, hypoxia and ER stress, chronic or prolonged bio-mechanical stress. In one embodiment, the present invention provides a method which prevents or alleviates aberrant cell differentiation that is caused by the activation of the integrated stress response and thereby prevents or alleviates conditions, disorders or diseases resulting therefrom. In one embodiment, ISR-associated diseases subject to the present invention include but are not limited to skeletal disorders including disc degeneration, MCDS and other skeletal dysplasias, cancers, inflammatory diseases, diabetes, fibrosis, obesity and neurodegenerative diseases. In another embodiment, the present invention provides a method of using a p-eIF2α-modulator for the prevention or treatment of conditions, disorders or diseases described herein.

BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figures 1A-1H show dwarfism, skeletal deformities, histological and molecular defects in 13del transgenic mice. Figure 1A is a diagrammatic representation of the 10.4 kb Coll0al-13del transgene used for generating the 13del mice. In Figure IB, the co-localization of mutant COLX ^13del protein (green) with the Con A (red) indicates 13del protein is accumulated within ER (left panel), which lead to the activation of UPR-associated factors in 13del HZ (at/4, chop and Bip) (right panel). In Figure 1C, body lengths of the WT and 13del Iittermates were monitored from birth to 30wk stage, and a consistent reduction of body length in13del mice was observe. In Figure ID, radiographic analysis revealed the skeletal abnormality of 13del mice at 4-week stage. In Figure IE, statistics showed13del mice displayed disproportionate dwarfism at 10- week stage (*Spine here indicates the length of 7 continuous vertebrae consisting of the last lumbar, four sacral and two caudal vertebrae). In Figure IF, 13del transgenic mice display both MCDS-like and non- MCDS features. In Figure 1G, growth plate is abnormally expanded in 13del proximal tibia at 10- day stage. Hypertrophic chondrocytes (HCs) in the lower part of hypertrophic zone (HZ) re- expressed markers more typical of pre-HCs (Sox9, Col2al andPpr). In Figure 1H, TUNEL assay revealed apoptotic cells with yellow nuclei (arrows) were present at the chondro-osseous junction in wild type (left panel). In13del mice (middle and right panel), no cell death was detected in ER stressed HCs, while a few apoptotic nuclei were found near the bone collar.

[0016] Figures 2A-2M show PERK signaling pathway regulates reverted chondrocyte differentiation. Figure 2A shows a schematic PERK signaling pathway in eukaryotes. Figure 2B is a schematic diagram of the rationale for fractionating the WT and 13del p10 growth plates into different chondrocyte populations (left panel), based on the expression patterns of chondrogenic markers (Col2al, Ihh, Ppr, CollOal, MmpI1) and ER stress marker (BiP) (right panel). Figure 2C shows the clustering analyses of differentially expressed genes in chondrocyte subpopulations in plO WT and 13del proximal tibia growth plates. Expression levels were normalized from -0.5 (blue) to 0.5 (yellow). Four major clusters I, Π, III and IV were identified (left panel). The average expression levels (Log ₂ scale) of the genes in different clusters revealed significant expression pattern changes in 13d meilce (right panel). In Figures 2D and 2E, genes in different clusters were functionally categorized using DAVID web tools. The enriched biological processes (Figure 2D) and enriched pathway (Figure 2E) were sequentially shown for Cluster I, II, III and IV. In Figure 2F, PERK signaling and XBP1 ^s regulating ERAD signaling pathway is highly enriched in KEGG pathway of protein processing in endoplasmic reticulum in Cluster I. Red stars indicated the mapped genes in this cluster. Figure 2G illustrate the normalized microarray measurements of major components involved in PERK pathway (Atf4, Chop, At/3, Gadd34 and Eroll) in each zone of WT and 13del growth plates. Figure 2H shows histology (a, a') and gene expression analysis of Atf4 (c, c'), Atfi (d, d'), Chop (e, e'), and Eroll (f, f' ) by radioactive in situ hybridization on the wild type and 13de glrowth plates at p10 stage. Hypertrophic chondrocytes are specifically marked by CollOal (b, b'). ER stressed hypertrophic chondrocytes are marked by Bip (g, g'). Figure 21 refers to the in vivo validation of ATF4 (b, b' c, c'), ATF3 (d, d' e, e'), CHOP (f, f' , g, g') and GADD34 (h, h ⁵ , i, i') by immune-staining on WT and gro13wdtehl plates at plO stage. In Figure 2J, enriched motifs on Cluster I genes were analyzed, using sequences of promoter region (+/-2kb from the TSS) for these genes. Motifs matched to the TFs in the UPR were shown. Figure 2K shows the scheme of XBP1 ^s (spliced form) over- expressing vector, together with FLAG-tag, for generation of FXBPF transgenic mice. Figure 2L shows the results of hematoxylin & Eosin staining (a, a', e, e') and immunochemistry of COL10A1 (b, b', f, f ), FLAG (c, c' g, g') and XBP1 ^s (d, d', h, h') for comparing the growth plate phenotypes between WT and FXBP 1 ^s transgenic mice at plO and 4-week stages. Figure 2M illustrate the measurement of HZ length of WT and FXBP 1 ^s at p 10 stage.

[0017] Figures 3H-3I show ectopic expression of ATF4 in HCs leads to dwarfism and HZ abnormality. Figure 3A are expression profiles of ATF4 on WT growth plates from E14.5 to P10 stages. Figure 3B refers to scheme of Atf4 expression vector. Atf4 cDNA is inserted after the ATG codon in exon 2 of the CollOal-Bac together with an IRES-EGFP cassette. In Figures 3C and 3D, expression specificity of C10-ATF4 transgene was determined by EGFP visualization in developing growth plates (Figure 3C), and validated by expression profiling of Egfp (a, c, e and g) and Atf4 (b, d, f and h) in C10-ATF4 growth plates at different stages (Figure 3D). Figure 3E show the radiographic analysis which revealed the dwarfism and skeletal abnormality of C10- ATF4 mice at 4-month-old. Figure 3F illustrates that body lengths of the WT and C10-ATF4 littermates were monitored from birth to 6-month stage, and a consistent reduction of body length in C10-ATF4 mice was observed. In Figure 3G, abnormal proximal tibial growth plates with expanded HZ, delimited by dotted lines, were observed in C10-ATF4 mice. Figure 3H shows that ectopic expression of ATF4 does not cause cell death. Figure 31 shows that ectopic expression of

Atf4 in HCs was insufficient for stress response induction, indicated by in situ hybridization of ER stress markers (Bip, Αίβ and Chop).

[0018] Figures 4A-4M show ATF4 modulates chondrocyte reprogramming via directly transactivating Sox9. Figure 4A shows ectopic expression of Atf4 in HCs leads to accumulation of premature chondrocytes in C10-ATF4 HZ, indicated by chondrogenic markers CollOal (a, a', b, b'), Sox9 (c, c', d, d'), Col2al(e, e', f, f ), Ppr (g, g', h, h') and Ihh (i, i', j, j') transcripts. Figure

4B is a presentation of ATF4 ChlP peaks on regulatory region (+/-2kb from TSS) of vital chondrogenic transcriptional factors (SOX, MEF2, RUNX, GLI and FOXA). The expression trend of these factors in WT and 13del chondrocytes were measured by normalized microarray expression profile. Figure 4C shows luciferase activities of Sox9 promoter reporter with different lengths (pSox9-2.7K, pSox9-1.8K and pSox9-0.8K) or ATF4 putative binding sites mutants

(pSox9- 1.8Ml , pSox9-1.8M2 and pSox9-1.8M3) responding to different dosages of ATF4 were measured in ATDC5 cells. In Figure 4D, ChlP-PCR showed the direct binding of ATF4 to the putative motif on the Sox9 promoter in vivo, using the nuclear extracts from El 5.5 WT and C10- ATF4 limbs. An ATF4 ChlP-seq peak (dark triangle) around this region has been reported in ER- stressed MEF cells. Figure 4E is a schematic diagram of generation of C10-ATF4; Sox9 ^c/c and

13del; Sox9 ^c/c mice, by using HC-specific CollOal-Cre. Figure 4F represents the genetic rescue of growth plate abnormalities of C10-ATF4 mice with HC-specific knock-out of Sox9 at plO stage, as shown by histology (A, A'), expression analyses of SOX9 (B, B', C, C), Col2al (D, D') and

CollOal (E, E'). Figures 4G and 4H illustrate the HZ length measurement (Figure 4G) and

CoUal positive cells quantification (Figure 4H) in C10-ATF4; Sox9 ^c/+ and C10-ATF4; Sox9 ^c;c mice. Figure 41 represents genetic rescue of growth plate abnormalities of mice w13itdhe Hl C- specific knock-out of Sox9 at plO stage (n=5), as shown by histology (A, A'), expression analyses of SOX9 (B, B\ C\ C), Col2al (D, D') and CollOal (E, E'). Figures 4J, 4K and 4L respectively illustrate the HZ length measurement and Col2al and Ppr positive cells quantification in 13del and 13del; Sox9 ^c/c mice. Figure 4M shows that ablation of Sox9 in WT HCs does not cause any abnormality at plO stage, as shown by histology (a, a') and expression profiles of Sox9 (b, b'),

Col2al (c, c'), Ppr (d, d ^! ) and CollOal (e, e'). [0019] Figures 5A-5H show the study design. Figures 5A and 5B indicate that the tibia length is further shortened in 13del; Chop ^-/- mice at plO stage. The comparison was performed between13del and 13del: Chop ^-/- littermates. In Figure 5C, exacerbated growth plate abnormalities were observed in 13del mice with global loss of CHOP at plO stage, shown by histology (a, a'), expression analyses of Coll0al (b, b'), SOX9 (c, c' d, d'), Col2al (e, e ^; ) and Ppr (f, f' ), Figures 5D and 5E respectively shows the HZ length measurement, and SOX9, Collal and Ppr positive cells quantification in 13del and 13del; Chop ^7" mice. Figure 5F and 5G show the results of TUNEL assay which revealed increased number of apoptotic cells in 13del: Chop ^-/- H Z. Figure 5H shows normalized microarray measurements of PERK signaling components {Chop, Atft, Gadd34 and Eroll), Chaperone (Bip, Dnajb9, Dnajbll and Calmxin) and ER stress sensors (At/4, Xbpl and Atf6) in WT, 13del and 1: 3 cdheolp ^-/- mice at plO stage.

[0020] Figures 6A-6F show GADD34 inactivation exacerbated the dwarfism while reduced the HZ abnormality in 13del mice. Figures 6A and 6B indicate HZ expansion is reduced in 13dei; GADD34 ^-/- mice. Figures 6C and 6D indicate fewer HCs revert to preHC-like cells in 13del; GADD34 ^-/- mice, marking by Ppr. Figures 6E and 6F indicate the tibia length is further shortened in 13del;GADD34 ^-/- mice at plO.

[0021] Figures 7A-70 show FGF21 , regulated by ATF4 and CHOP, protects the HCs from ER stress. Figure 7A shows the normalized microarray measurement of Fgftl in growt1h3d pelalte comparing with WT. Figure 7B illustrate a significant up-regulation of FGF21 in 13del HCs at plO stage, validated by in situ hybridization (a, b) and western blot. In Figures 7C and 7D, Fgftl is significantly activated in response to Tunicamycin (Tm) in ATDC5, NIH3T3 and MEF cells, at indicated time points. The activation of Bip indicates ER stress is triggered. Figure 7E is a schematic diagram of generation of 13del; Fgf2T ^-/--/- mice. Figure 7F shows the ablation of Fgf21 from WT HCs does not affect cell survival. Figure 7G shows that growth plate of 13del; Fgf21 ^-/- mice exhibited comparable phenotype to 13 mdiecle, as shown by histology (a, a'), expression analysis of CollOal (b, b'), Fgftl (c, c') and Ppr (d, d'). Figures 7H and 71 respectively show the measurement of HZ lengths and quantification of Ppr positive cells in 13del and 13del; Fgf21 - ^/- mice. Figures 7J and 7K illustrate that FGF21 protects the 13del HCs from death in dosage- dependent manner. In Figure 7L, two putative ATF4 binding sites (Al and A2) were predicted in the Fgftl promoter region. Luciferase activities of Fgftl wild type, deletion or mutated ATF4 binding sites promoter reporter in response to Tunicamycm or DMSO treatment. Figure 7M indicates that ectopic expression of Atf4 (a, a') is insufficient for Fgftl (b, b') induction in HCs. In Figure 7N, results of ChlP-PCR showed the binding of ATF4 and CHOP to the ATF4 binding- peak-containing region on the Fgftl promoter under ER stress in NIH3T3 cells. Figure 70 illustrates the normalized microarray measurement of Fgftl in WT, 13del and 13de[;Chop ^-/- chondrocytes.

[0022] Figures 8A-8Q show that small molecule ISRIB, by preventing ATF4 induction under ER stress, ameliorates 13de skl eletal deformities. Figure 8A is a schematic timeline of the ISRIB (2.5mg/kg) or vehicle (0.5% DMSO in 0.9% saline) administration. The mice were administrated by intraperitoneal injection, starting from El 3.5 to p20 stage. The animals were harvested at indicated time-points. Figures 8B and 8C indicated that treatment of ISRIB does not affect the body weight gain or body length growth in wild type mice. In Figure 8D, body lengths of the vehicle-treated (n=20) and ISRIB-treated (n=16) 13del were monitored from birth to p20 stage, and a significantly consistent improvement of body length in ISRIB-treated 13del mice was observed. In Figures 8E and 8F, the radiographic analyses revealed that skeletal deformities of 13del mice were alleviated at p20 stage by ISRIB treatment, including length of tibia, femur and spine (spine here indicated by the length of 7 continuous vertebrae consisting of the last sacral and six tail vertebrae), pelvic bone deformation (Θ1: the angle between ilium and pubis), Coxa Vara (Θ2: the angle between the proximal head and the shaft of the femur) and Genu Varum (Θ3: the angle between proximal head and distal head of tibia). Figures 8G and 8H shows the rescue of growth plate abnormalities in 13del mice by the treatment of ISRIB at plO stage and p20 stage, as shown by histology (a-a") and in vivo expression profiles of SOX9 (b-b" and c-c"), Col2al (d- d") and Ppr, (e-e"). Figures 81 and 8 J respectively show the HZ length measurement and SOX9 ^÷ , Col2al ⁺ and Ppr ⁺ cells quantification of tested animals at indicated time points. Figure 8K suggests that histology of p 10 growth plates were comparable between ISRIB- and vehicle-treated WT mice. Figures 8L and 8M show the rescue of the HZ expansion in caudal intervertebral disc (IVD) by the treatment ISRIB in 13del mice at plO and p20 stage. Figures 8N shows the rescue of the growth plate deformities in caudal IVD by the treatment of ISRIB in 13del mice, indicated by reduced number of SOX9 (a-a", c-c") and Col2al (b-b", d-d") at plO and p20 stages. Figures 80 indicates that at p lO, the transcripts of Atf4 (a-a"), Atfi (b-b"), Chop (c-c"), Eroll (d-d") and Fg/21 (e-e") were down-regulated in HZ from ISRIB-treated 13del mice, while Bip (f- f") was not affected. Figures 8P indicates that at p lO, the protein level of ATF4 (a-b"), ATF3 (c-d"), CHOP (e-f ) and FGF21 (g-h") were down-regulated in HZ from ISRIB-treated 13del mice. Figures 8Q indicates that ISRIB treatment does not induce apoptosis in 13del mice.

[0023] Figures 9A-9B show dwarfism and growth plate deformities in 13del-KI MCDS mice. Figure 9A shows the ten-day-old WT, 13del-KI MCDS heterozygous (Co110al ^13d/+ ) and homozygous (CollOal ^13d/13d ) mice. Figure 9B shows that the growth plate expansion is paralleled to the number of reprogramming HCs in 13del-KI MCDS mice.

[0024] Figures 10A-10F show 13del and 13del-KI MCDS mice display degenerative intervertebral disc (IVD) features. In Figure 10A, the radiographic analysis reveals early onset of IDD in 20-years-old MCDS patient. In Figure 10B, the FAST staining shows swelling of the nucleus pulpous (middle arrow), endplate expansion (upper and lower dashed arrows), and accumulation of chondrocyte-like cells in the inner annulus fibrosus (iAF) in 13del mice, at 4- week stage. In Figure IOC, the radiographic analysis reveals severe intervertebral disc degeneration in tail region (T5/6, T6/7 and T7/8) in 7-, 9-, 12- and 16-month 13del mice. In Figure 10D, histologically, the 13del tail intervertebral disc (IVD) exhibited significant characteristics of disc degeneration, including loss of NP/AF boundary, disc bulging and widening of the AF interlamellar space at both 6-month (upper panel) and 16-month (lower panel) stages. Notably, the cycled region clearly showed the inward bulging of inner AF (iAF) lamellae, and significant decreased volume of vascular canals in subchondral region between spinal growth plate and endplate in 6-month 13del mice. Moreover, at 16-month stage,the 13del disc clearly exhibited (a) the altered NP structure and matrix, (b) the inward bulging of AF lamellae and consequently fissure (boxed regions). In Figure 10E, excessive cell death can be observed in NP of the degenerated tail IVD in 16-month 13de ml ice. In Figure 10F, transcriptionally and transnationally, the essential ER stress sensor BIP was ectopic upregulated in core region of 13del NP at 6-month stage. In Figure 10G, although the transcriptional expression level of Atf4 is not changed, the protein level of ATF4 is significantly upregulated in 6-month N1P3,d ienldicating the contributory regulation of ISR. In Figure 10H, the activation of ATF5, the vital transcription factor of mitochondria- dependent oxidative stress response, can be observed in 6-month 13del NP, indicating the induction of oxidative stress. In Figure 101, in WT control, the peripheral NPC highly expressed Sox9 and the level became much lower in cells within core region. However, as the consequence of induction of stress in 6-month 13del NP, the cell fate of NP cell was affected, indicating by the ectopic expression of Sox9 in cells within the NP core region. In Figure 10 J, OPN, as a major component of NP extracellular matrix, highly expressed in peripheral NPCs at young stage (plO, p20 and 4- month stages), diminished the expression level as maturation (6-month stage) and became absent at elderly stage (16-month stage). However, in 13del NP, not only the persistently upregulation in peripheral NPCs but also the ectopic expression of Opn in NPCs within the NP core region can be observed. Moreover, at elderly 16-month stage, the Opn " cells still can be detected. In Figure 10K, similar to Opn, a-SMA marked the peripheral NPC in WT at young stage (4-month stage) but became absent at 6-month stage, while this marker persistently expressed in 13del peripheral NPCs and ectopic expressed in core NPCs. In Figure 10L, the 13deI-KI mice exhibit similar irregularities in FVD at plO and 4-week stages, including accumulation of chondrocyte-like cells in the inner annulus fibrosus (iAF) (dashed circle in panel f), NP swelling (arrows in the middle of panels a and d) and irregular endplates (upper and lower dashed arrows in panel d).

[0025] Figures 11A-11D show ISRIB ameliorated the IVD syndromes of 13del mice. Figure 11A shows that treatment of ISRIB (2.5mg/kg) eased the IVD abnormalities in 13del lumbar spine, including less expanded endplate and more organized iAF structure. Figure 11B shows that in13del mice, treatment of ISRIB reduced the number of reprogrammed chondrocytes in the growth plates and endplates. Moreover, the ectopic expression of Opn in NP were greatly prevented. In Figure 11C, in 13del lumbar IVD, ATF3, a downstream target of ATF4 is significantly activated in the HCs of the growth plate and endplate as well as in the NP (arrows). With the treatment of ISRIB, no ATF3 expression can be detected and there appeared to be fewer ATF3 expressing HCs.

[0026] Figures 12A-12G show that hypoxia stress is triggered in 13del MCDS mice. Figures 12A and 12B indicate that HIFlct and HIF2α were upregulated in 13del-KI HCs. In Figure 12C, EF5 staining clearly demonstrated hypoxia stress response was triggered in 13del-KI mice. In Figure 12D, GFP/MCDS chimera demonstrated hypoxia stress (EF5 positive cells) is highly correlated with the cells expressing 13del protein. Figure 12E shows that the -expre1s3sdinegl cells (green, upper panel) are under hypoxia stress, marked by EF5 (red, middle panel), in the HCS of the talus cartilage. Figures 12F and 12G show that HIF1α and HIF2α were upregulated in C10-ATF4 HCs.

[0027] Figures 13A-13D demonstrate the putative ATF4 binding regions on Sox9 within the topologically associated domains (TAD), indicating the potential ISR-regulation on Sox9 by enhancers. Figure 13A shows human SOX9 (hSOX9) and mouse Sox9 (mSOX9) are located within the boundary region between 2 sub-TADs and share a highly conserved TAD pattern. Figure 13B and 13C demonstrated the highly conserved CCCT C-binding factor (CTCF) insulator binding region presenting in human and mouse Sox9 gene locus. Figure 13D demonstrated an example of putative ATF4 binding enhancer region on mSox9.

DETAILED DESCRIPTION OF THE INVENTION

[0028] The present invention provides a method of preventing, ameliorating and/or treating conditions, disorders or diseases associated with the integrated stress response (ISR) involving the phosphorylated eukaryotic initiation factor 2a (p-eIF2α) pathway arising from various cellular stresses. As described herein, phosphorylated eukaryotic initiation factor 2a pathways or p-eIF2α pathways include signaling pathways where de -phosphorylated eIF2u or phosphorylated eIF2u is involved, and include signaling pathways which are directly or indirectly affected by the de- phosphorylation or phosphorylation of eIF2α.

ISR-associated skeletal diseases

[0029] The first aspect of this invention is to provide a method of preventing, ameliorating and/or treating a skeletal disorder associated with or caused by the activation of the integrated stress response in a subject. Skeletal disorders subject to the present invention can be arising from cellular stresses such as oxidative stress, hypoxia and ER stress, or caused by chronic or prolonged biomechanical stress. [0030] The integrated stress response (ISR) is an adaptive cell-survival pathway that can be activated when raisfolded proteins trigger endoplasmic reticulum (ER) stress. It is implicated in development and diseases, with many human genetic skeletal deformities being caused by mutations that trigger the ISR.

[0031] In an MCDS transgenic mouse model (13del) which carries a 13 bp deletion in Co12al equivalent to the human mutation, misfolded mutant collagen X induces ER stress. Although the chondrocytes survive, their differentiation is reversed by an unknown mechanism to a more juvenile state characterized by the re-expression of prehypertrophic chondrocyte markers (Ppr,

Sox9 and Co12al), disrupting endochondral ossification, and skeletal dysplasia ensues, A similar effect on hypertrophic chondrocyte differentiation has been described in other mouse models of dwarfism (70). The skeletal defects caused by mutations that induce stress or inactivate key transducers of the stress response in humans and mouse models implicate components of pathways involved in chondrocyte and osteoblast differentiation. However, the relationship between skeletal dysplasia and the ISR remains unclear.

[0032] The present invention represents the first mechanistic study in a model of human chondrodysplasia associated with ER stress that demonstrates causality and a direct link between the ISR and reprogrammed chondrocyte differentiation. Disclosed herein, ISR signalling reverses hypertrophic chondrocyte differentiation via ATF4-directed transactivation of the transcription factor gene Sox9. By genetic and molecular analyses, the present invention established that the major effect of the ISR is the preferential expression of ATF4 which activates the transcription of a potent transcription factor gene Sox9 (a key regulator of chondrocyte differentiation and proliferation) (Fig.4). The present invention also discloses for the first time the dual action of

CHOP and ATF4 in promoting hypertrophic chondrocyte survival, establishing the critical role of

CHOP in partnership with ATF4 in enabling chondrocyte survival via the transactivation of Fgf21.

The present invention highlights the complex consequences of activating ISR, in part because of the distinct roles of ATF4 in controlling cell differentiation and proliferation depending on cell context.

[0033] The present invention further demonstrates that treatment of mutant 13del mice13 with a small molecule inhibitor of the ISR pathway, ISRIB (trans-N,N'-(Cyclohexane-l ,4- diyl)bis(2-(4-chlorophenoxy)acetamide) which targets the interaction between eukaryotic initiation factor 2 (eIF2) and eukaryotic initiation factor 2B (eIF2B) and thereby suppresses ATF4 induction, prevents the differentiation defects and ameliorates chondrodysplasia in the mice 13del (Fig. 8), and ameliorates the degenerative intervertebral disc (IVD) syndromes of the mice 13del (Fig. 11). The failure of chemical chaperones or ER-stress reducing reagents to rescue chondrodysplasia in mouse models (15) and the benign impact of either absence of (30), or over expressing Xbpls (this study) emphasizes cell context-dependent effects of the different arms of the UPR. Importantly, the present invention identifies a key causative role for the ISR in MCDS and demonstrates that targeting early in the pathway, i.e., at the level of PERK phosphorylation of eIF2α could be an effective therapeutic approach. As disclosed herein, the effect of ISRIB on the aberrant differentiation of ER-stressed HC reveals the contextual complexity of ISRIB action. On the one hand, it antagonizes ATF4, reversing or preventing the de-differentiation. On the other hand, reduced levels of ATF4 and CHOP enfeeble a cell survival mechanism. In the context of the 13 del mutation, the net result improves mouse skeletal development. This may reflect the dominance of de-differentiation in the pathogenesis of the chondrodysplasia. By finding the dose of ISRIB that titrates ATF4-CHOP levels and is protective for de-differentiation without causing death of the stressed chondrocytes, the dualism inherent in the ISR may be exploited therapeutically. In the present invention, a dose of ISRIB is administered to a subject to achieve an optimal level of cell differentiation and cell survival. In one embodiment, 2.5 mg/kg of ISRIB is administered to a subject twice a day, effectively and substantially alleviating the defects. In another embodiment, 5 mg/kg ISRIB is administered to a subject twice a day. In various embodiments, the effective amount of ISRIB is 0.05-0.1, 0.1-1, 1 -5, 5-10, 10-20, 20-25, 25-50 or 50-100 mg/kg per day. In one embodiment, the subject is treated for 1 day or up to 365 days. In various embodiments, the subject is treated for 5, 10, 25, 50, 75, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325 or 350 days.

[0034] In one embodiment, the present invention provides a method of preventing, ameliorating and/or treating a skeletal disorder (including disc degeneration, MCDS and other skeletal dysplasias) associated with integrated stress response involving the p-eIF2α pathway in a subject using a molecule that targets the underlying p-eIF2α-pathway. In one embodiment, molecules to be used in the present invention are modulators which directly or indirectly suppress the translational or transcriptional expression of ATF4 or SOX9. Without limiting the generality of the foregoing, the following illustrates a few embodiments of the present invention.

Chondrodysplasia and congenital dwarfism

[0035] There has been no example whereby the impact on cell fate in a congenital disorder such as dwarfism can be prevented or ameliorated by targeting the ISR in vivo. Therefore, the present invention provides for the first time a feasible approach for the prevention and improvement of congenital dwarfism caused by the activation of ISR.

[0036] In one embodiment, the present invention provides a method of preventing, ameliorating and/or treating a skeletal disorder associated with integrated stress response involving the p-eIF2α pathway in a subject. In one embodiment of the present invention, skeletal disorders include but are not limited to osteogenesis imperfecta (OI), metaphyseal chondrodysplasia, type Schmid (MCDS), pseudo achondroplasia (PSACH), and multiple epiphyseal dysplasia. In another embodiment, skeletal disorders subject to the present method are caused by mutation(s) of extracellular matrix (ECM) proteins accumulating in the endoplasmic reticulum, and mutation(s) of key signaling transducer of ISR. In one embodiment, said skeletal disorders are arising from cellular stresses such as oxidative stress, hypoxia, and ER stress. In another embodiment, said skeletal disorders are caused by chronic or prolonged biomechanical stress.

[0037] In one embodiment, the present invention provides a method of alleviating and/or reversing aberrant differentiation in a chondrocyte through the inhibition of ectopic expression of Sox9/SOX9 (mouse/human). In another embodiment, the present method prevents and/or alleviates conditions, disorders or diseases resulting from an aberrant chondrocyte differentiation. In one embodiment, the present method includes the use of a molecule which inhibits the ectopic expression of Sox9/SOX9. In another embodiment, the present method includes a use of a molecule which inhibits the ectopic expression of ATF4 which subsequently enhances the ectopic expression of Sox9. In one embodiment, the molecule is a modulator that is capable of directly or indirectly inhibiting the transcriptional or translational expression of ATF4.

[0038] In one embodiment, the modulator in the present invention is represented by Formula I:

, wherein each of R1, R2, R3 and R4 is independf

hydrogen, halogen, -OCH3, -OCH ₂ Ph, -C(0)Ph, -CH ₃ , -CF ₃ , -CC1 ₃ , -CN, -S(0)CH ₃ , -OH, -NH2, -COOH, -CONH2, -NO2, -C(0}CH ₃ , -CH(CH ₃ )2, -CCSi(CH ₃ ) ₃ , -CCH, -CH2CCH, -SH, -SO3H, - SO4H, -S0 ₂ NH ₂₎ -NHNH2, -ONH2, -NHC=(0)NHNH _2j -NHC=(0)NH2, -NHSO2H, -NHC=(0)H, -NHOH, -OCH ₃ , -OCF ₃ , -OCHF2, -N ₃ , substituted or unsubstituted alkyl, substituted or unsubstituted heteroalkyl, substituted or unsubstituted cycloalkyl, substituted or unsubstituted heterocycloalkyl, substituted or unsubstituted aryl, or substituted or unsubstituted heteroaryl.

[0039] In one embodiment, the modulator represented by Formula I is IS RIB having the formula

. In one embodiment, the modulator represented by

Formula I includes molecules that are described in WO 2014/144952, the entire contents of which are incorporated herein by reference into this application.

[0040] In another embodiment, the modulator represented by Formula Ϊ is selected from the following molecules: [0041] In one embodiment, the modulator subject to the present invention is represented by

wherein R ¹ is bicycloheteroaryl, including but not limited to

pyrrolopyrimidine, which may be unsubstituted or substituted with groups such as amino and alkyi; R ² is heteroaryl, including but not limited to pyridyl, pyrrolyl and pyrazolyl, which may be unsubstituted or substituted with groups such as halogen, alkyl and trihaloalkyl, and R ³ is hydrogen or halogen. In one embodiment, the modulator is represented by Formula II which includes molecules described in WO201 1/119663, the entire contents of which are incorporated herein by reference into this application.

[0042] In one embodiment, the modulator represented by Formula II is GSK2656157 having the

formula of In various embodiments, the modulator represented

b Formula II is selected from the following molecules:

[0043] In one embodiment, the present method reduces the number of mature hypertrophic chondrocytes reversing to pre-mature chondrocytes as indicated by the decreased transcriptional or translational expression of immature chondrogenic markers such as Sox9, Col2al, Ppr and Ihh,

[0044] In one embodiment, the present invention provides a method of ameliorating one or more conditions of MCDS or congenial dwarfism by modulating the p-eIF2cc -pathway. In one embodiment, the method includes the use of a molecule which inhibits the effects of p-eIF2 on translation regulation or a molecule which inhibits the ectopic translational or transcriptional expression of Sox9 or ATF4 in the defective cells. In one embodiment, the conditions to be ameliorated by the present method include but are not limited to disproportionate dwarfism, short stature or limbs, flaring of metaphyses, genu varum (bowing legs), coxa vara, hip deformation, platyspondyly (flat vertebral bodies), abnormal vertebral endplates, widened growth plate and consequent disc degeneration.

Intervertebral disc degeneration (IPD)

[0045] In one embodiment, the present invention provides a method of preventing or ameliorating, intervertebral disc degeneration (IDD) caused by the activation of the integrated stress response pathway, which might induced by oxidative stress (23), hypoxia (24), ER stress (19), nutrition deprivation (25), accumulation of toxic metabolites (26) and/or excessive mechanical loading (27). In one embodiment, the present method of preventing or ameliorating, intervertebral disc degeneration (IDD) involves the use of one or more modulators or molecules described in the present invention. In 13del mice, the malformed spinal growth plates and endplates causes decreased volume of vascular canals in subchondral region between growth plates and endplates, consequently lowers the oxygen/nutrition importation and toxic metabolites exportation in nucleus pulpous, triggers oxidative stress (indicating by upregulation of ATF5), and changes cell differentiation indicating by ectopic expressing Sox9, Osteopontin (Opn) and α-SMA. Moreover, the ectopic expressing of matrix protein Opn alter the matrix deposition and induced ER stress, indicating by ectopic expression of ER stress sensor Bip and ATF4, and caused cell death at later stage. On the other hand, caged mouse normally used its tail to help its stand up for food and water, which might cause excessive mechanical loading to the tail of 13del with short stature. Consistently, the early onset of disc degeneration in 13del mice was always first observed in tail IVD (level 6- 8), indicating the pathogenesis role of mechanical loading in IDD development. In one embodiment, the present invention uses a molecule that is capable of modulating the activation of ISR or its underlying stresses or causes in intervertebral disc cells for the prevention or treatment of IDD.

ISR-associated diseases involving p-eIF2α pathway

[0046] The second aspect of the present invention is to provide a method of preventing and/or ameliorating aberrant ceil differentiation, or modulating cell fate determination through the modulation of ectopic expression of ATF4 and its potential downstream factors, such as Sox9/SOX9, in a cell. The present invention further provides a method of modulating ISR- associated diseases where aberrant cell differentiation is at least part of the underlying mechanism through the modulation of ectopic expression of ATF4 and its potential downstream factors, such as Sox9/SOX9, in a subject. In one embodiment, the present method of preventing and/or ameliorating aberrant cell differentiation, or modulating cell fate determination involves the use of one or more modulators or molecules described in the present invention.

[0047] The present invention highlights the potential of manipulating levels of the ISR for the treatment of ISR-associated human diseases resulted from various forms of cellular stress. As discussed above, ISR has a central role in many forms of cellular stress such as oxidative stress, hypoxia, ER stress and its induction is associated with diverse common diseases, such as cancer, inflammatory diseases, diabetes, fibrosis, obesity, neurodegeneration and skeletal disorders. The p-eIF2α signaling pathway, being a part of the ISR, could be a target for the prevention or treatment of these ISR-associated diseases. However, while stress responses commonly result in apoptosis, understanding how cells adapt, survive and a molecular understanding on the consequences of inducing the ISR on cell fate and differentiation in vivo is lacking. The majority of molecular mechanistic insights of the impact of the ISR are based on cell based assays not in vivo. The present invention exploited an in vivo model of a congenital developmental disorder (MCDS) in order to provide mechanistic insight into the question of impacts of the ISR on cell fate and importantly also addressed the possibility of preventive therapy.

[0048] As reported herein, over-expression of ATF4 as part of the ISR has far reaching consequences in vivo, it directly activates the expression of Sox9 and thereby reverses the differentiation of a mature chondrocyte to a more immature state. SOX9 is a potent transcription factor with key roles in cell fate determination, not only in chondrocyte differentiation, but also in many other cell types, notably stem cells (e.g. dermal papilla, gonads, intestinal, and neural) and its overexpression or dysfunction results in many diseases including fibrosis and cancer (83). By preventing aberrant cell differentiation, titrated inhibition of the ISR emerges as a rationale therapeutic strategy for treating skeletal disorders or other disorders caused by ISR. The present invention revealed that given the importance of ATF4 to normal development, simply preventing its expression globally may not work therapeutically (Fig. 3A). Rather, the present invention introduces for a novel and viable approach to alter cell differentiation or cell fate by targeting the translational control of ATF4 that leads to its over expression.

[0049] In one embodiment, the present invention provides a method of modulating cell differentiation or cell fate using a molecule that modulates the ectopic expression of ATF4 and its potential downstream factors, such as Sox9/SOX9. In another embodiment, the present invention provides a method of preventing, ameliorating and/or treating an ISR-associated disease where cell differentiation or development is impacted in a subject, the method comprises a step of administering to said subject an effective amount of a molecule that modulates the ectopic expression of ATF4 and its potential downstream factors, such as Sox9/SOX9. In one embodiment, the present invention is used to prevent, ameliorate and/or treat diseases which are associated with or caused by an impacted cell differentiation or development, which include but are not limited to cancer, fibrosis, neurodegeneration and skeletal diseases. Firstly, it has been implied that cancer cells are selected to resist mild and prolonged ER stress by activating pro-survival UPR and PERK signaling pathway induces resistance to cell death elicited by endoplasmic reticulum stress and chemotherapy (28). Notably, gemcitabine resistance in pancreatic ductal adenocarcinoma is enhanced by activating multiple ISR-dependent pathways, including eIF2, Nrf2, Nuprl , BEX2, and Bcl2Al (29). Moreover, increased phosphorylation of eIF2α in chronic myeloid leukemia cells contribute to the disruption of bone marrow niche components by cancer cells and in this way support CML progression (30). Secondly, the persistent presence of ER stress can increase cell death in injured tissues, induction of epithelial-mesenchymal transition (EMT) and promote fibrotic remodeling instead of the restoration of normal tissue architecture (31), and increased oxidative stress is a common pathological feature of fibrosis in a variety of organs, including lung (32), liver (33) and heart (34). Thirdly, both ER stress and oxidative stress have suggested to play important roles in neurodegeneration, such as Parkinson's disease, Alzheimer's and prion disease (35, J 6). Finally, as discussed above, ER stress is associated with chondrodysplasias caused by mutations in ECM protein and mutations in key factors in ISR signaling pathway (37).

p-eIF2q modulators

{0050] Without limiting the generality of the foregoing, the present invention provides a method of preventing, ameliorating and/or treating the conditions, disorders or diseases discussed herein using a molecule which targets the p-eIF2α pathway (a "p-eIF2α modulator"). In one embodiment, the present invention provides a method of preventing, ameliorating and/or treating a condition, a disorder or a disease associated with integrated stress response involving the p-eIF2α pathway in a subject, comprising the step of administering to said subject an effective amount of a p-eIF2α modulator. In another embodiment, tire present invention provides a use of a p-eIF2α modulator for the preparation of a medicament for preventing, ameliorating and/or treating a condition, a disorder or a diseases associated with integrated stress response involving the p-eIF2α pathway. In another embodiment, the present invention provides molecules that are capable of modulating the p-eIF2oc pathway for use in the treatment of diseases or modulation of conditions described herein.

[0051] In one embodiment, the subject is a human including an adult and a child, or an animal.

[0052] In one embodiment, "effective amount" means the amount of a molecule necessary to achieve a desired physiological effect.

[0053] In one embodiment, the present invention provides a method of modulating the p-eIF2α pathway in a cell or a population of cells, the method comprising contacting the cell(s) with an effective amount of a p-eIF2α modulator. [0054] In one embodiment, p-eIF2oc modulators are small molecules, nucleic acids, proteins or other biomolecules. In one embodiment, p-eIF2α modulators are small molecules which are represented by Formula I or II described above. In one embodiment, p-eIF2α modulators are p- eIF2α inhibitors that inhibit one or more downstream molecules or signaling events under the p- eIF2α pathway such as ISRIB and GSK2656157 and their analogs. In another embodiment, p- eIF2α modulators are molecules that activate one or more downstream molecules or signaling events under the p-eIF2α pathway such as Sulubrinal and Guanzbenz and their analogs. In yet another embodiment, p-eIF2α modulators are molecules that alter one or more downstream molecules or signaling events under the p-eIF2cc pathway (for example, those illustrated in Fig. 2), and the p-eIF2α pathway can be part of the cellular stress responses such as oxidative stress, ER stress and hypoxia, or other chronic or prolonged biomechanical stress.

[0055] In one embodiment, said p-eIF2α modulators are capable of targeting eIF2α phosphorylation such as ISRIB and its analogs. In another embodiment, p-eIF2α modulators are molecules which are capable of targeting GADD34-Pplc or promoting the assembly of G ADD 34- Pplc. In yet another embodiment, p-eIF2cc modulators are molecules which are capable of modulating the expression of ATF4 and its potential downstream factors, such as Sox9.

[0056] In one embodiment of the present invention, the effective amount of p-eIF2α modulator such as ISRIB to be given to a subject is 2.5 mg/kg to 20 mg/kg per day. In various embodiments, the effective amount of p-eIF2α modulator is 0.05-0.1, 0.1-1, 1-5, 5-10, 10-20, 20-25, 25-50 or 50-100 mg/kg per day. In one embodiment, the subject is treated for 1 day or up to 365 days. In various embodiments, the subject is treated for 5, 10, 25, 50, 75, 100, 125, 150, 175, 200, 225, 250, 275, 300, 325 or 350 days.

[0057] In one embodiment, two or more p-eIF2αpathway modulators are administered concurrently. In another embodiment where two or more p-eIF2α pathway modulators are to be administered, the second or subsequent p-eIF2α pathway modulators are administered immediately or a certain period after the administration of the previous p-eIF2α pathway modulator.

ATF4-binding site on Sox9 locus and binding enhancers

[0058] In one embodiment, the present invention provides a method of inhibiting the ATF4/ISR mediated activation of transcription of murine SOX9/human SOX9, thereby preventing, alleviating and/or treating conditions resulting from the overexpression of SOX9; the method comprises a step of contacting the cells with, or administering to a subject, a molecule that is capable of blocking the ATF4-binding site on the Sox9/SOX9 locus, or by interfering with molecules that modulate the ATF4-mediated transcription of Sox9//SOX9 (such as a molecule that enhances the binding between ATF4 and Sox9/SOX9 locus, i.e., an ATF4-binding enhancer).

[0059] Sox9 was found to be located within the boundary between two sub-TADs (topologically associated domains) within chromosome 11 (chrl 1 : 110760000-114800000), represented by chrl 1 : 111520000- 112200000 and chrl 1 : 113160000-114160000 respectively (99, 100) (Fig 33A, lower panel). Binding sites for ATF4 in mouse embryonic fibroblasts have been reported (44). While in human genome, SOX9 was found to be located in the boundary between two sub-TADs within chromosome 17 (chrl 7:68055000-72184000), represented by chrl7:68609000-69117000 and chrl7:70514000-71514000 respectively (100, 101 ) (Fig 13A, upper panel). The Sox9/SOX9 TAD pattern is conserved between human and mouse as indicated in Figure 13B.

[0060] The present invention has identified the putative binding site for ATF4 on the Sox9 locus in hypertrophic chondrocytes of mice - a region in chromosome 11 (loci: 112642927-112643074) which covers the promotor region of Sox9 (Fig. 4B). It is thus possible to inhibit the transcription of Sox9 by using molecules that interfere with the entirety or part of the putative ATF4-binding site and thereby modulate conditions resulting from the overexpression of SOX9.

[0061] In one embodiment, the present invention provides a method of inhibiting the transcription of murine Sox9 by using a molecule that blocks or interferes with one or more binding sites for ATF4 on the murine Sox9 locus. In one embodiment, said ATF4 binding site is located within mouse chromosome 11 (loci: 112642927-112643074) having the sequence TGTTGCAA (SEQ ID NO: 1). In another embodiment, said ATF4 binding site is located within the binding sites for ATF4 as reported in Han:

GTCACCCAAACATTTGCTTCCAAAAGACCATTTCTAAGCACTTTTTTTGGAAGCCGG C AGACTCCAGGCGCAGAAGCCCAGCTCCGCTTTGACGAGCAGCTGTTGCAATTTCCA TTGCTGTAAACGCCAGCGAAGTCCCGGGTACCAC) (SEQ ID NO.: 2), the entire contents of which are incorporated herein by reference into this application (44).

[0062] In one embodiment, the present invention further provides a method of inhibiting the transcription of human SOX9 by using a molecule that blocks or interferes with one or more binding sites for ATF4 on the human SOX9 locus. In various embodiments, said ATF4 binding site is located within human chromosome 17 (chrl7:68609000-71514000). In one embodiment, said ATF4 binding site is TGTTGCAA (SEQ ID NO.: 3) (38) which is the consensus sequence of ATF4 binding site on human SOX9 locus.

[0063] In various embodiments, various approaches including those described in Han (44) are used to identify the functional binding sites of ATF4 on Sox9/SOX9 or other related ATF4 potential downstream factors. In one embodiment, ATF4-binding sites on Sox9/SOX9 or other related ATF4 potential downstream factors are mapped by the core Amino Acid Response Element (AARE) sequence TTgCaTCA (SEQ ID: 4), which is the complementary strand of SEQ ID NO.1. [0064] In one embodiment, open chromatin regions in cells expressing Sox9/SOX9 upon induction of the ISR and/or ATF4 over-expression is identified via ATAC-seq (39). This method applies hyperactive Tn5 transposase, which inserts sequencing adapters into accessible regions of chromatin, to mark accessible regions of DNA which are then sequenced. GFP (or other reporter) are inserted into 3' untranslated region of mouse or human locus and targeted so as to provide a readout of SOX9 activity, alternatively the cells derived from Sox9 ^EGFP/+ mice are adopted (40). Cells are subjected to ER stress, hypoxia or other stresses to induce the ISR, or over-expression of ATF4 is induced.

[0065] Three biological replicates for ATAC-seq are generated to identify enhancers that are active and distinct to the So9 ^+ve /SOX9 ^+ve population. Approximately 10,000 FACS sorted EGFP ^+ve and EGFP ^-ve cells are isolated from Sox9 ^EGFP/+ mice and library are prepared via NEBNext High Fidelity 2x PCR Master Mix. The amplified libraries are purified by AMPure beads, quantitated (KAPA biosystems) and sequenced at 10-15 million reads.

[0066] Filtered reads are aligned to the mouse and human reference genomes using BWA and subjected to peak calling using MACS2. The regions with enrichment of transposition events indicating for open chromatin are identified. By comparing Sox9 ^+ve /SOX9 ^+w specific enhancer profiles, we are able to distinguish and capture putative ISR induced and/or ATF4-associated enhancers: a) those for driving Sox9/SOX9 expression under normal non-stressed conditions; and b) those active when the ISR is induced and/or ATF4 is overexpressed. This approach allows us to prioritize amongst the putative enhancers, not only for functional validation but also for generation of a regulatory map of ISR- and/or ATF4-associated Sox9/SOX9 enhancers.

[0067] To overcome variability in expression due to position effects, regions in the mouse genome or human genome that are constitutive ly open and therefore not subject to position effects are used for assaying the enhancer activity (e.g. the TIGRE locus (41, 105)) . Reporter locus are targeted in cell lines and transgenic mice with a vector comprising a minimal promoter (such as hsp68 or the minimal SOX9 promoter which has no activity in cells/transgenic mice) linked by a 2A peptide sequence (42) to a fluorescence reporter (e.g. GFP, RFP, YFP etc.) or other factors (e.g. luciferase).

[0068] In one embodiment, the enhancer interference assay is used for functional validation of enhancer elements by epigenome inhibition in vivo and in vivo, using a nuclease-deficient Cas9 (dCas9)-histone demethylase (43) fusion to inhibit the activity of candidate enhancer(s) by selectively altering the chromatin state of the target enhancer(s). Removal of H3K4mel/me2 modifications from specific active enhancer(s) using targeted catalytically inactive dead-Cas9 (dCas9) fused to the lysine-specific demethylase 1 (KDMIA/LSDI) results in 'inactivation' of enhancer elements and down-regulation of gene expression from the associated loci. The transgene containing dCas9-LSDl is targeted via using CRISPR-Cas9 (44), in which a guide RNA (gRNA) is specifically designed to direct LSD1 to the putative Sox9/SOX9 enhancer(s). In this way, the expression of LSD1 on targeting specific enhancer(s) silences the candidate enhancer(s) by demethylation of histone H3K4me2 and destruction of K27 acetylation (H3K27ac).

[0069] The targeted enhancer(s), resulting in loss of SOX9 driven EGFP expression when the ISR is activated and/or when ATF4 is over expressed, are first identified in vitro. In vivo, the activities of identified ISR- and or ATF4-inducible SOX9 enhancer(s) are assessed by: a) mutating the enhancer(s) in mice using CRISPR-Cas9; and b) targeting the enhancer(s) to the ISR reporter vector described above comprising a minimal hsp68 promoter and testing for its ability to be activated upon ATF4 or ISR induction in double transgenics where ISR is triggered and/or ATF4 is over-expressed.

[0070] On the other hand, a total of 25 putative ATF4 binding enhancer regions were identified in the mouse Sox9-TAD domain (Table 1) by published ATF4 ChlP-seq (45), and Fig I3C demonstrates an example of putative ATF4 binding enhancer region of mSox9. Taken together, these findings strongly suggest that ISR-induced ATF4 may regulate the Sox9 expression by enhancers.

[0071] In one embodiment, the present invention provides a method of inhibiting the transcription of Sox9/SOX9 (such as murine SoxiVHuman SOX9) by using a molecule that blocks or interferes with one or more ATF4 binding enhancers which regulates the transcription of murine Sox9, said ATF4 binding enhancer comprises a sequence selected from the group consisting of SEQ ID NOs. : 5-29.

[0072] In another embodiment, the present invention provides a method of inhibiting the transcription of human SOX9 by using a molecule that blocks or interferes with one or more ATF4 binding enhancers which regulate the transcription of human SOX9, To identify the putative ATF4 binding enhancer regions in the human SOX9-TAD domain, ATF4 ChlP-seq is applied in human fibroblasts, cancer cell lines, and chondrocytes (or any other cell lines where stress induces SOX9 ⁺ ) differentiated from human iPS. The cells are treated with an ER stress-inducer such as tunicamycin (46) to activate the preferential translation of ATF4, and three biological replicates for each cell type are generated. In one embodiment, said ATF4 binding enhancers are located within human chromosome 17 (chrl 7:68609000-71514000).

[0073] In various embodiments, the present invention provides a method of inhibiting the transcription of human SOX9 by using a molecule that blocks or interferes with one or more ATF4 binding enhancers which regulate the transcription of human SOX9, said ATF4 binding enhancer comprises a sequence that could be similar/homologous to the murine sequence selected from the group consisting of SEQ ID Nos:. 5-29. In one embodiment, said ATF4 binding enhancer comprises a sequence corresponding to a sequence which is at least 70%, 75%, 80%, 85%, 90% or 95% homologous to the sequence selected from the group consisting of SEQ ID Nos:. 5-29. In another embodiment, said ATF4 binding enhancer in human comprises a sequence corresponding to and showing high consensus to the murine sequence selected from the group consisting of SEQ ID Nos:. 5-29. It is also possible that there will be human specific ATF4 binding enhancers not present in mouse. These will be detected by the ATAC-seq and ATF4 ChlP-seq approaches described above. Functional validation of the human enhancer activity will be tested by linking putative enhancer(s) to reporter (e.g. Luciferase/fluorescent proteins) constructs and testing for their activation upon inducing the ISR in vitro (mouse or human cell-lines) and in vivo, using trans enic mice in which the ISR is induced, such as in 13del.

Drug screening platform using 13del and 13del-KI transgenic mice

[0074] The third aspect of this invention is to provide a method of screening a candidate molecule for the ability to modulate a skeletal disorder or its phenotypes associated with ISR involving the p-eIF2α pathway using transgenic mice disclosed in the present invention.

[0075] In one embodiment, the present invention provides a method of screening a candidate molecule for the ability to modulate a skeletal disorder or skeletal abnormalities associated with ISR involving the p-eIF2α pathway, comprising the steps of administering said candidate molecule to a transgenic mouse carrying a Co110al transgene or having a DNA sequence encoding the mutated collagen protein Col 10- ,13 adnedl expressing a phenotype lacking hyperostosis. In one embodiment, the transgenic mouse represents a direct equivalent model of human 13del MCDS mutation. In one embodiment, the transgenic mouse is mice13 odre 1l3del-KI mice.

[0076] In one embodiment, the present screening method further comprises a step of detecting any changes in the mouse that indicate improvement of said skeletal disorders or skeletal abnormalities including but are not limited to disproportionate dwarfism, short stature or limbs, flaring of metaphyses, genu varum (bowing legs), coxa vara, hip deformation, platyspondyly (flat vertebral bodies), abnormal vertebral endplates, widen growth plate and disc degeneration.

[0077] In one embodiment, the present screening method comprising the step determining one or more of the following parameters in the mouse treated with said candidate molecule and comparing the results with those from a positive control such as ISRIB and a negative control:

a) the growth curve, including body trunk length and whole body length, during whole period of treatment;

b) limb length, including individual limb bones, after treatment;

c) the spine length and curvature after treatment;

d) the angle between proximal head and distal head of tibia after treatment;

e) pelvic bone orientation (the angle between ilium and pubis) after treatment;

f) the angle between the proximal head and the shaft of the femur after treatment;

g) the heights of growth plates from limb bone and spine;

h) the onset time point of disc degeneration after preventive treatment (administration before IDD observed);

i) the X-ray examination and MicroCT examination of IDD development during the preventive treatment (administration before IDD observed);

j) the X-ray examination and MicroCT examination of IDD development during the therapeutic treatment (administration after IDD observed):

k) the disc histological morphology after or during the preventive (administration before

IDD observed) and therapeutic treatment (administration after IDD observed), including the heights of growth plates, the irregularities of endplates, the volume of vascular canals in subchondral region between spinal growth plate and endplate, the disc distance and the calcification of nucleus pulpous (NP);

1) the expression and transcription level of ectopic expressing factors, such as Sox9 in the affected disc after treatment;

m) the cell number of ectopic expressing factors, such as Sox9 in the affected disc after treatment;

n) the expression and transcriptional level of affected factors, such as CollOal in the affected disc after treatment;

o) the number of affected factors expressing cell, such as Coll0al in the affected disc after treatment; and

p) the number of apoptotic cells in the affected disc after treatment.

Discussion

[0078] In a transgenic mouse model displaying phenotypes reminiscent of congenital dwarfism [Metaphyseal chondrodysplasia, type Schimd (MCDS), MIM 156500] and intervertebral disc changes consistent with early stages of human intervertebral disc degeneration (IDD), it has been shown that synthesis of misfolded collagen X in hypertrophic chondrocytes causes abnormal intracellular retention of secreted proteins and triggers the unfolded protein response (10). Specifically, it has been found that the PERK pathway, which controls protein translation via eIF2α phosphorylation and induction of the transcription factor ATF4, causes the hypertrophic chondrocyte differentiation defect in the growing long bones and spine. In one embodiment of the present invention, ISRIB, a selective modulator of phosphorylated-eukaryotic initiation factor (p- eIF2α) and eIF2B complex, is used to prevent and/or ameliorate the dwarfism and intervertebral disc degeneration caused by induction of the integrated stress response pathway.

[0079] As shown in Fig. 2A, activation of PERK signaling pathway leads to eIF2α phosphorylation, which represses the global protein translation but preferentially facilitates the translation of ATF4 transcripts via bypassing an inhibitory upstream open reading frame (uORF) (47). ATF4 can transactivate Chop and ATF3, and form a heterodimer with ATF3 to modulate the expression of target genes, such as GADD34. CHOP also acts upstream of GADD34, which encodes a regulatory subunit of the protein phosphatase complex that dephosphorylates p-eIF2α and restores protein translation. Thus ATF4, CHOP and GADD34 form a negative feedback loop to ensure transient attenuation of protein synthesis and later recovery of protein translation during ER stress response (13, 47).

[0080] It is observed in the present model that UPR plays a critical role in the pathogenesis of MCDS. The present invention identifies the mechanism(s) underlying the ER stress-associated skeletal defects in the MCDS model. The present invention further discloses a novel approach in preventing or treating ISR-associated diseases, in particular to diseases where aberrant cell differentiation is the underlying cause.

[0081 ] This invention will be better understood by reference to the examples which follow. However, one skilled in the art will readily appreciate that the examples provided are merely for illustrative purposes and are not meant to limit the scope of the invention which is defined by the claims following thereafter.

[0082] Throughout this application, it is to be noted that the transitional term "comprising", which is synonymous with "including", "containing" or "characterized by", is inclusive or open-ended, and does not exclude additional, un-recited elements or method steps.

UPR disrupts transcriptome patterns in the chondrodys plastic growth plate

[0083] The mammalian growth plate comprises four major sub-populations of chondrocytes organized into zones: resting, proliferating (PZ), prehypertrophic (PHZ) and hypertrophic (HZ). These chondrocytes have distinct morphologies and gene expression profiles governed by a precisely tuned gene regulatory network (48) . To investigate the effect of the UPR on transcription, chondrocytes in the proximal tibia growth plates of postnatal day 10 (plO) from wild-type and MCDS 13del mice were fractionated into sub-populations representing proliferating, prehypertrophic and hypertrophic chondrocytes (HC) (Fig. 2B). The wild-type HZ was fractionated into upper and lower zones (UHZ and LHZ) to capture early and late phases of hypertrophy. The 13del HZ was fractionated into three zones: upper, corresponding to early phase of UPR activation, middle (MHZ), where HC adaptation would be initiated, and lower, where HC should be fully adapted.

[0084] k-means clustering was used to categorize the gene expression patterns across different zones in wild- type and 13del growth plates into four clusters. Genes (453) in Cluster I increased expression from PHZ to lower HZ specifically in 13del HC (Fig. 2C). Ontological analyses show these differentially expressed genes are mainly involved in protein processing in the ER and the UPR (Fig. 2D and 2E). Genes in Clusters II (659) and III (314) showed highest expression in wild- type PZ and PHZ followed by progressive down regulation from PHZ to lower HZ, but were upregulated in 13de llower HZ, reflecting UPR induced changes (Fig. 2C). These genes included Sox9, Ppr and Ihh, consistent with the previous report of re-expression of pre-hypertrophic markers (10). Cluster IV genes (680) showed increasing expression from PHZ to lower HZ in wild-type and can be defined as "HC characteristic" genes. Consistent with a change in the HC differentiation state in 13del , these genes were down-regulated in l1o3wdeerl HZ. The concomitant down- regulation of Cluster I stress response genes in 13del lower HZ is consistent with alleviation of the stress in the reprogrammed cells and an adapted state.

PERK signaling is the major contributor to chondrocyte adaptation to ER stress

[0085] The UPR employs three arms of sensors in the ER to mediate cell adaptation and survival under ER stress: PERK, IRElα, and ATF6 family (49-51). Upon ER stress, ATF6 family factors move from the ER to the Golgi, are processed by S I and S2 proteases, and translocate to the nucleus to activate ER quality control genes such as Hspa5 (encodes BiP) and Xbpl (X-box binding protein 1). IREla has kinase and endoribonuclease (RNase) activities. It catalyses the splicing of Xbpl rnRNA, generating the UPR transcription factor XBPP that upregulates genes encoding chaperones and proteins involved in ER-associated protein degradation (ERAD).

[0086] PERK phosphorylates serine 51 in eIF2α, promoting the formation of a p-eIF2α and eIF2B complex, consequently inhibiting the guanine nucleotide exchange activity of eIF2B (52). Inactivation of the eIF2 complex leads to shut down of protein synthesis except for certain proteins, including ATF4, CHOP and other factors with both pro-survival and pro-death functions (47). eIF2α phosphorylation is transient and is reversed by GADD34, the regulatory subunit of eIF2α phosphatase, acting in a negative feedback loop, allowing protein synthesis to restart. When the stress is intense or prolonged, cells fail to adapt and apoptotic cell death is triggered. This PERK- p-eIF2α/ATF4/CHOP modulation of mRNA translation is a central part of the more general ISR (18).

[0087] Contributions of PERK, IRE1 and ATF6 to the HC response to ER stress were investigated. By ontology and pathway analysis of Cluster I, enrichment of genes in the PERK pathway and IRE 1-XBP I ^s regulated ERAD was found, but not for ATF6 signaling (Fig. 2F). Activation of PERK signaling in 13del HC was demonstrated by up-regulation of its key components (Atf4, Atf3, Chop, Eroll and Gadd34) (Fig. 2G) which was validated by in situ hybridization and immuno staining (Fig. 2H and 21). Using Motif enrichment analysis, it was found that the binding motifs of CHOP and ATF4 were highly enriched in Cluster I, but not those for XBP1 or ATF6 (Fig. 2J). By interrogating ATF4 and CHOP ChlP-seq data (45), significant over-representations of ATF4 (odds ratio=2.87, p <0.0001 ) and CHOP (odds ratio=4.33, p <0.0001) binding peaks associated with the genes from Cluster I were observed but not for the other clusters. Cluster I genes are therefore those most likely to be directly regulated by PERK-associated transcription factors.

[0088j Together, these data suggest a prominent contribution of the PERK signaling pathway. To test this notion, Xbpl ^s was ectopically expressed in HC in transgenic mice (Fig. 2K). Over- expression of XBP1 ^s specifically in HC did not affect the growth plate (Fig. 2L and 2M), which is consistent with another MCDS mouse model study that found inactivation oiXbpl in chondrocytes did not alter the severity of dwarfism (53).

ATF4 expression in hypertrophic chondrocytes reprograms differentiation

[0089] Apart from its role in the UPR, ATF4 also regulates chondrocyte differentiation through activating Ihh (54). ATF4 is normally expressed in fetal growth plate chondrocytes, but not in HC by plO (Fig. 3A). Therefore, the chondrocyte differentiation defects in the MCDS model might be directly caused by activation of ATF4 and/or its target Chop.

[0090] To dissect apart the contribution of ATF4 to aberrant HC differentiation in the absence of ER stress, a transgenic mouse model carrying a CollO-Bac-ATF4-IRES~EGFP transgene was generated (hereafter referred to as C10-ATF4) (Fig. 3B), in which ATF4 expression was driven by the highly HC-specific promoter of Coll0al(55, 56). The present invention confirmed HC-specific expression of the C10-ATF4 transgene in the developing growth plates from fetal (El 5.5) to adult (P20) stages (Fig. 3C and 3D). Similar to 13del mice, adult C10-ATF4 transgenic mice were dwarfs, being approximately 20% shorter than wild-type littermates (Fig. 3E and 3F). Histological analyses revealed a greater than three-fold expansion of the HZ in C10-ATF4 mice (Fig. 3G). Although forced expression of ATF4 in fibroblasts was reported to decrease survival (45), cell viability was not affected in C10-ATF4 mice (Fig. 3H). Importantly overexpression of ATF4 in HC in the absence of ER stress, did not induce transcription of the UPR-associated genes Bip and Chop (Fig. 31), although Atfi was slightly upregulated (45). Therefore, activation of ATF4 alone, in the absence of the ER stress response, is sufficient to alter HC differentiation, disturb endochondral ossification and cause skeletal abnormalities similar to those observed in mice. 13del ATF4 reprozrams chondrocyte hypertrophy by directly activating Sox9

[0091] In C10-ATF4 HC, constitutive ATF4 activation down-regulated expression of CollOal, and led to expression of prehypertrophic chondrocyte marker genes Sox9, Col2a, Ppr, and Ihh in the lower portion of the HZ (Fig. 4A). The sequential differentiation process in growth plate chondrocytes is tightly regulated by multiple chondrocyte-specific transcription factors that control expression of cell type-specific genes and secreted growth factors (48, 55, 57-61). The present invention searched published ER stress-associated ATF4 ChlP-Seq data (45) for binding peaks in key chondrogenic transcription factor genes, including members of SOX, RUNX, MEF2,

GLI and FOXA families, and found ATF4 binding peaks in regulatory regions of Sox9, Sox5, Sox6,

Runx2, GU2 and GU3. Amongst these, only the Sox genes were up-regulated in 13del middle and lower region HC (Fig. 4B), suggesting that the Sox family could be the regulatory targets of ATF4.

[0092] SOX9 is highly expressed in immature chondrocytes, transactivates critical cartilaginous matrix genes and regulates chondrocyte proliferation, differentiation and hypertrophy (55, 60-63).

It is required for the expression of SOX5 and SOX6, which cooperate with SOX9 to transactivate

Col2al (61). Two putative C/EBP-ATF4 motifs, named Al and A2, were identified in the Sox9 promoter region covering the ATF4 binding peak. By transfection assays in ATDC5 chondrocytic cells, it was found that ATF4 could transactivate luciferase reporters controlled by motif- containing Sox9 promoter (Fig. 4C). Mutation of Al and A2 respectively reduced or abolished

ATF4 activation of the Sox9 reporters (Fig. 4C). Anti-ATF4 ChlP-PCR assays, using nuclear extracts from E15.5 wild-type and C10-ATF4 limbs, demonstrated that ATF4 binds directly to the putative motifs region on the Sox9 promoter in vivo (Fig. 4D).

[0093] The contribution of ATF4 activation of Sox9 in reverting HC differentiation by conditionally inactivating Sox9 in C10-ATF4 HC was then assessed using HC-specific CollOal- Cre (56) (Fig. 4E). In the absence of Sox9, the expansion of HZ in C 10-ATF4 mice was markedly reduced and there were fewer cells expressing Col2al in the HZ (Fig. 4F-4G). Moreover, conditional inactivation of Sox9 in 13del mice reduced expression of Collal and Ppr in HC and the HZ expansion was greatly reduced (Fig. 4I-4L). Deletion of Sox9 in wild-type HC did not affect chondrocyte hypertrophy (Fig. 4M). Collectively, these data suggest PERK-induced overexpression of ATF4 reverts differentiation in 13del HC by direct activation of Sox9 in HC, thereby perturbing chondrocyte hypertrophy.

CHOP plays an adaptive and pro-survival role in 13del HC

[0094] How do the ER-stressed HC survive? CHOP is another prominent transcription factor activated in the PERK, downstream of p-eIF2α, that regulates protein synthesis via the GADD34 negative feedback loop, which restores protein synthesis and induces oxidative stress via Eroll

(45, 64). Although CHOP is widely considered as a proapoptotic factor (45), it has context- and cell-type specific roles as an adaptive and pro-survival factor in several diseases (65-68). Therefore, the contribution of CHOP in the adaptation of 13del HC was assessed.

[0095] It is found that ablating Chop in 13 mdeicl e exacerbated the skeletal defects and growth plate phenotype. The l3del;Chp ^-/- mice displayed further tibial shortening (Fig. 5A an 5B) with greater (-20%) HZ expansion (Fig. 5C and 5D), and increased number of chondrocytes expressing immature chondrogenic markers SOX9, Col 2a 1 and Ppr in the HZ (Fig. 5C and 5E). Strikingly, in contrast to 13del, there was increased apoptosis in 13del;Chop ^-/- HC, consistent with a pro- survival role for CHOP (Fig. 5F and 5G).

[0096J These results are in contrast to the pro-apoptotic role reported for CHOP in a mouse model of Pseudo achondroplasia caused by expression of misfolded cartilage oligomeric matrix protein in proliferating and hypertrophic chondrocytes, where deleting CHOP reduced apoptosis but exacerbated growth plate chondrocyte disorganization (69, 70). These differences may be due to variation in the responses of proliferating versus hypertrophic chondrocytes and/or the acuteness and duration of the ER stress. The present transcriptome analyses of fractionated 13del;Chop ^-/- growth plates revealed up-regulation of molecular chaperones (Bip, Dnajb9, Dnajbll and

Calnexin) and ER stress sensors Xbpl and Atf4 in the middle and lower HZ (Fig. 5H). By contrast, the PERK signaling pathway was enfeebled, reflected by marked (>3.5 fold) down-regulation of

CHOP targets Αtf3, Gadd34 and Eroll. In the absence of CHOP, recovery of translation mediated by GADD34 would be impaired and sustained shutdown of translation would favour apoptosis.

Therefore, CHOP mediates 13del HC survival and it was important to identify pro-survival/anti- apoptotic factor(s) downstream of CHOP and ATF4.

[0097] On the other hand, ablation of GADD34, the modulator of translation recovery and downstream target of CHOP, from 13del HCs lead to a HZ reduction on day 10 (89% of , 13del n=5) (Fig. 6A and 6B), accompanied by fewer Ppr+ cells in the mid- and lower portions of the HZ

(Fig. 6C and 6D). This finding supports the notion that UPR is interrelated to the MCDS phenotype and attenuation of ER stress via preventing translation recovery is potential to ameliorate the growth plate phenotype. Because GADD34 is important for the reinitiation of protein synthesis which in the context of 13d meleans that the 13del mRNA would be translated and mutant protein would be re-expressed triggering another round of ER stress, inactivating or reducing the activity of GADD34 should prevent another round of insult. However, although the differentiation changes were reduced, the tibia length in 13del;GADD34 ^-/- mice was further shortened (88% of 13del, n=5)

(Fig. 6E and 6F). One possible reason is GADD34 deficient cells may retain eIF2α-directed phosphatase activity, via the activation of a redundancy factor CReP, the constitutive repressor of eIF2α phosphorylation (77).

ATF4 and CHOP mediate chondrocyte survival by activating: Fgftl

[0098] CHOP acts not only downstream of ATF4, but also as its interacting partner in modulating ER stress targets (45). To elucidate the pro-survival role of the PERK signaling pathway in 13del HC, target genes of CHOP and ATF4 in Cluster I were searched. Fgf21, a reported target of ATF4 (72), was found to be the most up-regulated gene in 13del HC (Fig. 7A), which was confirmed by in situ hybridization and miraunoblotting (Fig. 7B). Fg/21 was similarly reported to be activated in ER-stressed chondrocytes (73).

[0099] FGF21 is a hormone with roles in glucose and lipid metabolism (74) and plays an survival role in the response to diverse stressful conditions, such as amino acid deprivation, mitochondrial stress and ER stress associated with diseases such as diabetes, cardiovascular disease (75-77). Fgftl expression was found to be greatly increased (>100 fold) in response to treatment with the ER stress inducer tunicamycin in fibroblasts (NIH3T3 and MEF cells) and ATDC5 cells (Fig. 7C and 7D).

[0100] The present invention assessed whether FGF21 had a survival role in H1C3de bly genetically ablating the gene (Fig. 7E). Fgftl null mice have normal growth plates and HC viability (Fig. 7F). The hypertrophic zone expansion in 13del;Fgf27 ^-/- mice was comparable to that of13del mice, and the reverted differentiation process was not affected (Fig. 7G-7I). However, increased apoptosis was found in the HZ of Fgftl -deficient 13del mice. This protective effect of FGF21 against apoptosis is dose dependent (Fig. 7J and 7K).

[0101] The present invention next tested the functional relevance of a reported C/EBP-ATF4 binding motif in the. Fgftl promoter (78) that coincided with an ATF4 peak. By transactivation assays (Fig. 7L), it was found that deleting (pFgf21-Luc3 and pFgf21-Luc4) or mutating (pFgf21- Ml and pFgf21-M3) the ATF4 binding motif abolished the transactivation ability of ATF4. However, Fgftl was not induced in C10-ATF4 mice (Fig. 7M), suggesting ATF4 is necessary but not sufficient for Fgftl induction. The need for another factor is supported by ChIP assays where we found both ATF4 and CHOP bind to the ATF4 motif-containing peak region in the cells under ER stress (Fig. 7N), even though no CHOP binding peak was detected in the Fgftl promoter. Expression of Fgftl was down-regulated by approximately 40% in 13 del; Chop ^-/- HC despite the upregulation of Atf4 (Fig. 70), consistent with a requirement for ATF4-CHOP cooperation to induce Fgftl under ER stress. Thus, the present results show that ATF4 and CHOP work together to induce directly Fgf21 expression in 13del HC for their survival.

ISRIB inhibition of the PERK signaling can ameliorate 13del skeletal deformities

[0102] In the UPR, PERK phosphorylation of serine 51 in eIF2α is the critical upstream controlling point that triggers the p-eIF2α /ATF4/CHOP signaling pathway (18). The present data show that genetically ablating the key transcription factors in the PERK signaling as a strategy for rescuing the aberrant chondrocyte differentiation is imperfect, because of effects on cell survival In addition, addressing the effects of transcription factor over- expression and cell-type specificity is required because ATF4 is essential for normal development. Therefore, it is necessary to identify a suitable entry point in the pathway which can be manipulated for protection or rescue from the effects of ER stress, without interfering with normal developmental function.

[0103] As summarized in Table 2, small molecules targeting PERK signaling pathway have been reported in neurodegenerative disorders and cancer therapy (14). Recently, a small molecule, Integrated Stress Response InhiBitor (ISRIB) has been reported to render cells insensitive to eIF2α phosphorylation by targeting the interaction between eIF2 and eIF2B, and its activity is independent of eIF2a phosphorylation (79, 80). ISRIB shows acceptable pharmacokinetic properties and no overall toxicity in mice, and has been reported to show significant neurotrophic effects in mice (79, 81).

Table 2. Pharmaceutically Targeting PERK Signaling Pathway in Disease

[0104] The potential of ISRIB to modify the chondrodysplasia phenotype was tested by treating 13del and wild-type littermates with ISRIB or vehicle twice daily by intraperitoneal injection from E13.5 (onset of expression of 13del transgene) to postnatal day 20 (p20) (Fig. 8A). In wild-type mice, ISRIB had no adverse effects on weight gain or body growth (Fig. 8B and 8C). However, ISRIB markedly reduced the dwarfism of 13del mice from new born to juvenile stages (Fig. 8D). Radiographic analyses revealed treatment with ISRIB ameliorated the skeletal deformities at p20 (Fig. 8E and F), including the length of tibia/femur and spine; tibia bowing (genu varum: the angle between proximal head and distal head of tibia); pelvic bone orientation (the angle between ilium and pubis), and coxa vara (narrowed angle between the proximal head and the shaft of the femur)

(Fig. 8E).

[0105] Moreover, it was found that the HZ expansion in the limb growth plates of ISRIB-treated 13del mice was greatly reduced and the number of SOX9 ⁺ , Col2al ^÷ and Ppr ⁺ cells in the HZ at pl O and p20 was diminished (Fig. 8G-8J). ISRIB had no observable effect on the limb growth plates in wild-type mice (Fig. 8K). ISRIB treatment in 13del mice also reduced the deformities in other growth plates such as the axial skeleton, with reduced HZ expansion and decreased number of Sox9 ⁺ and Col2al ⁺ premature cells in tail intervertebral disc growth plates (Fig. 8L-8N). As expected, ISRIB specifically reduced the amount of ATF4 and CHOP protein, and inhibited p- eIF2α/ATF4/CHOP signaling transduction, marked by the down-regulation of the transcripts as well as the protein level of their downstream targets (ATF3. ERO l i and FGF21) (Fig. 80 and 8P). Importantly, inhibition of p-eIF2α/ATF4/CHOP by ISRIB did not induce apoptosis (Fig. 8Q) in 13de! HC. Thus, without any obvious adverse effect, ISRIB corrected the molecular, histological, and skeletal defects in 13del mice.

The 13 del-knockin MCDS mice

[0106] These mice are similar to the transgenic 13del mice in terms of the Coll Oal 13del mutation except that in these mice the 13del deletion was introduced into the endogenous Coll Oal allele by homologous recombination in mouse embryonic stem (ES) cells. This mouse, referred to here as 13dei-KI, therefore represents a direct equivalent model of the human 13del MCDS mutation and is another mouse model for the UPR triggered by expression of 13del- collagen X (Fig. 9A). Comparison between 13 del and 13del-MCDS mice

[0107] The I 3del and I 3del-KI mice express the same collagen X mutation. The gross phenotype of dwarfism of the 13de!-MCDS was similar to that of 13del except for the absence of hyperostosis and the degree of expansion of the 13del-KI HZ was not as severe as for 13del. The 13del-KI mice

recapitulated all the skeletal pheno types of MCDS, including disproportionate dwarfism, and

skeletal abnormalities including flaring of the metaphysis, coxa vara, deformities in the pelvic

bones, in heterozygotes and homozygotes with the latter being more affected. Both mouse models

display expanded HZ and accumulation of premature HCs (Fig. 9B). The bone (generalized

hyperostosis) phenotype in 13del is caused by activation of ER stress in osteocytes due to ectopic

expression of the transgene (106). The differences in severity in the degrees of expansion of the

HZ, being more severe in the 13d theal n in 13del-KI mice, could be due to differences in the

relative level of expression of wild-type and 13del collagen X protein in the two mutants. In

addition, the duration of expression of 13del is shorter, being reduced by the 4 weeks while it is

still expressed in the 13del-KI at that stage. The difference in duration can be attributed to the

difference in regulatory region driving the mutant in the transgenic 13del compared to the 13del KI which is controlled by the full complement of Col 10a J regulatory elements. Additional

differences are in the profile of differentiation defects in the HCs, In mice1 H3dCesl re-enter the

cell-cycle but they remain in Gl-S phase and do not undergo apoptosis but in the 13del-KI two

rounds of hypertrophy occur with an intervening cell cycle re-entry and exit.

Intervertebral disc degeneration (IDD) in MCDS mice

[0108] It is noted that some cases of MCDS display spinal abnormalities including in vertebral

bodies and end plate irregularities (82). S. Ikegawa at Center for Integrative Medical Sciences,

RIKEN, Tokyo, has examined the MRI of the spine of a 20-year-old male MCDS patient and found

evidence for signal intensity loss ("dark disc") and irregularities in the end plate (Fig.1 OA),

consistent with the notion that endplate irregularities are associated with IDD (83).

[0109] The expanded and irregular endplates can be observed in 13del mice1 at early plO and p20

stage (Fig. 8L-8N). Concomitantly, the nucleus pulposus (NP) were swollen in appearance and

there were irregularities in the inner AF (iAF) and endplate boundaries, with chondrocyte-like cells

present at NP-AF and endplate boundaries (Fig. 10B). As a consequence, the tail intervertebral

disc (IVD) of 13del mice exhibited significant characteristics of disc degeneration at adult stages

(Fig. IOC), including altered NP structure and matrix, loss of NP/AF boundary, disc bulging,

widening of the AF interlamellar space and the inward bulging of AF lamellae and consequently

fissure (Fig. 10D). Interestingly, excessive cell death can be observed in 13del degenerated disc at

16-month stage, consistent with human IDD studies (84) (Fig. 10E). It is notable that volume of

vascular canals in subchondral region between spinal growth plate and endplate significantly

decreased in 13del disc (Fig. 10D), indicating the importation of oxygen and/or nutrition from

endplate to NP and exportation of metabolites from NP to endplate might be lowered, consequently

inducing integrated stress response. Consistently, the transcriptional and translational upregulation of BIP, the essential ER stress sensor, can be observed in core region of 13del degenerated NP (Fig. 10F). Concomitantly, the protein level of ATF4 is upregulated, while the transcriptional expression level of ATF4 is not changed (Fig. 10G). Strikingly, ATF5, the key sensor and signal transducer of mitochondria-related oxidative stress is also significantly upregulated in 13del degenerated NP Taken together, these data strongly suggest the induction of multiple stresses in 13del NP (Fig. 10H).

(0110] Interestingly, the elevated stress response in 13del is accompanied by significant cell fate change, implied by the ectopic expression of Sox9 (Fig. 101). On the other hand, disrupted matrix deposition in NP could be an important characteristic of degenerative changes. In 13del lumbar NP, the ectopic upregulation of Opn was observed at from plO to 16-month stages (Fig. 10J). Moreover, the WT peripheral NP cells were found to be a-SMA ⁺ at young stage (4-month), but became absent of it at 6 -month stage. However, in 13del degenerated NP, the peripheral NP cells persistently expressed a-SMA at 6-month stage and ectopic expression of this factor in core region of NP can also be observed (Fig. 10K).

[011 II Spinal changes were also observed in 13del-KI mice, including shortened vertebral bodies throughout the spine, disc space narrowing and similar irregularities of iAF (4-weeks-old mice) (Fig. 10L).

[0112] These findings suggest 13del and 13del-MCDS mice could be used to model changes in the IVD from activation of ER stress in hypertrophic chondrocytes of the cartilage endplate. ISRIB prevents the molecular changes in the 13del IVD

[0113] As abovementioned, ISRIB treatment in 13del mice reduced the deformities in growth plates of axial skeleton, with reduced HZ expansion and decreased number of Sox9* and Collar premature cells in tail intervertebral disc growth plates (Fig. 8L-8N). In addition, after 20-days ISRIB treatment, the iAF was more regular in ISRIB-treated mice1,3 wdeitlh fewer chondrocyte- like cells present (Fig. 11 A).

[0114] Disrupted matrix deposition in NP could be an important characteristic of degenerative changes. In 13de llumbar NP, the ectopic upregulation of Opn was observed at plO, p20 and 16- month stages (Fig. 1 IB and 11 C). Strikingly, 20-days treatment of ISRIB greatly reduced the ectopic NP expression of Opn (Fig. 11C), which may prevent the consequent NP degeneration in later stage.

[0115] Multiple stresses can activate the integrated stress response, in which ATF4 directly transactivates another important transcription factor ATF3. In 13del lumbar IVD, ATF3 is significantly activated not only in HCs in the growth plate and endplate, but also in the NP. Notably, the activation of ATF3 can only be observed in L3-L6, the region mostly affected by spine bending caused by lumbar lordosis. This finding strongly suggests the correlation between spine alignment and the onset of IDD. After the ISRIB treatment, no Atfl ⁺ cell can be detected in L3-L6 NP, nor growth plates or endplates (Fig. 11D).

Hypoxia stress in 13 del MCDS mice

[0116] The ISR is activated by many cellular stresses including oxidative, nutritional and hypoxic. HIF pathway activation can be a consequence of UPR, and PERK pathway is also at the heart of hypoxia stress signaling pathway (18). In 13del HCs, HIFla and its associated or downstream factors were upregulated (Table 3).

[0117] In 13del-KI mice, HIFla and HIF2α immunohistochemistry staining showed a strong accumulation of HIFs proteins in hypertrophic chondrocytes when compared with wild-type. In postnatal 10-day-old growth plate, HIFla is detected in proliferating chondrocytes and resting chondrocytes, but not in hypertrophic chondrocytes. In the 13del-KI littermates, HIFla is not only accumulated in proliferating chondrocytes, but also in the hypertrophic chondrocytes from the upper to lower hypertrophic zones (Fig. 12A). HIF2α is predominantly found in pre-hypertrophic chondrocytes in wild-type 10-day-old mice but not much in hypertrophic chondrocytes. However, it can be clearly detected in both pre-hypertrophic chondrocytes and hypertrophic chondrocytes in 13del-KI littermates (Fig. 12B).

[0118] EF5 assay, which detects hypoxic cells in vivo system, was then performed in P10 littermates of WT and 13del-KI mice. The 13del immunofluorescence staining shows that almost all the hypertrophic chondrocytes are expressing mutant collagen type X in 13del-KI growth plate (Fig. 12C). Few EF5 bindings were detected in the lower hypertrophic zone in 13del-MCDS tibia proximal growth plate, nor in the whole hypertrophic zone in wild-type. However, the upper and middle hypertrophic zones of the growth plate in 13del-KI mice bound EF5 significantly more than the hypertrophic zone in wild-type (Fig. 12D). These data indicate that chondrocytes on the upper and middle hypertrophic Eones of 13del-KI are under lower oxygen tension compared to wild-type.

[0119] To assess the relative contribution of intrinsic and extrinsic response in the activation of the hypoxia response, EF5 was administrated by intraperitoneal injection into a P9 GFP/13del-KI chimera mice and sacrificed 4h later. In the distal tibia growth plate, columns of 13del positive hypertrophic chondrocytes can be detected with the 13del antibody indicating these are derived from clonal expansion of MCDS proliferating chondrocytes that are differentiated to hypertrophic chondrocytes. EF5 positive cells are highly correlated with the cells expressing protein (1F3idge. l 12D). This strong correlation of 13 edxepl ressing cells and EF5 staining positive cells are also observed in the hypertrophic chondrocytes of the talus cartilage (Fig. 12 E). ER stress is highly correlated with EF5 staining, suggesting the "hypoxic" signal is an intrinsic effect of 13deI-KI cells.

[0120] As abovementioned, ATF4 is regulated at the translational level in ISR under anoxic and hypoxic conditions, mediates in part by the unfolded protein response and is an important regulator of cell fate. Therefore, ATF4 could be the essential link between ER stress and hypoxic stress. In C10-ATF4 mice, HIFla can be detected in proliferating chondrocytes as well as the hypertrophic chondrocytes in the expanded hypertrophic zone in C10-ATF4 transgenic mice, while HIFla can be only detected in proliferating chondrocytes but not in hypertrophic chondrocytes in wild-type mice as previously shown (Fig. 12F). Furthermore, HIF2α was also detected in the same region of the expanded hypertrophic chondrocytes in C10-ATF4 transgenic mice (Fig. 12G), suggesting both HIF proteins are responding similarly to the ectopic expression of ATF4. These findings suggest that up-regulation of ATF4 as part of the UPR could be involved in the regulation of HIF proteins in hypertrophic chondrocytes under ER stress.

Identified putative ATF4 binding region on Sox9 topologically associated domains (TAD)

[0121] Chromosome conformation capture methods (capture Hi-C and 4C-seq methods) have been used to identify SOX9 subchromosomal structures of higher-order chromatin interactions called topologically associated domains (TADs) (85), which are separated by boundary regions that have comparatively high levels of transcriptional repressor CCCTC-binding factor (CTCF) (86). The TADs subdivide the SOX9 genome into discrete regulatory units, to which the majority of observed interactions between promoters and enhancers are restricted (87, 88).

[0122] Both human SOX9 (hSOX9) and mouse Sox9 (mSox9) are located within the boundary region between 2 sub-TAD domains in human and mouse genome respectively, and share a highly conserved TAD pattern (Fig. 13 A) in embryonic stem cells. Consistent to that, the CTCF insulator binding regions were found in both human SOX9 and mouse Sox9 gene locus (Fig. 13B and C). On the other hand, a total of 25 putative ATF4 binding enhancer regions were identified in the mouse Sox9-TAD domain (Table 1 ) by published ATF4 ChlP-seq (45), and Fig 13D demonstrates an example of putative ATF4 binding enhancer region of mSox9. Taken together, these findings strongly suggest that ISR-induced ATF4 may regulate the Sox9 expression by enhancers.

Methods Mice breeding

[0123] The 13del transgenic mice were maintained in Fl (C57BL/6 x CBA) background. The 13del-KI mice were maintained in C57BL/6 background. The Chop-null mice and Fgf21 -null mice were reported previously (13, 89). The Sox9-fIox mice was a gift from Prof. Andreas Schedl' lab (Institute of Biology Valrose, France) (60). The C 10-ATF4 transgenic mice were generated by injecting a BAC vector (Col l0al-ATF4-IRES-EGFP) into the Fl zygotes and maintained in FL Mice were genotyped by PCR using primers (5'-CAGATCAGTGATGGGCTATG-3' (SEQ ID NO: 30) and 5'-GAACCACCTGGAGAAGGCAGATT-3' (SEQ ID NO: 31 ). Animal care and experiments performed were in accordance with the protocols approved by the Committee on the Use of Live Animals in Teaching and Research of the University of Hong Kong.

Generation of wtGFP/13del-KI chimeric mice

[0124] GFP (homozygous) and 13del-KI homozygous (MM) morula-stage embryos (2.5 dpc) were collected, zona pellucida was then removed by acid tyrode and the embryos were washed and aggregated in 1 : 1 ratio at 37°C overnight in an aggregation plate. Successful aggregated embryos were transferred to a pseudopregnant ICR foster mother.

EF5 distribution analysis

[0125] To study EF-5 distribution, P 10 mice were injected with 10 mM EF5 at 1% of body weight and staining was performed as described previously (90, 91).

RNA preparation and Microarray analysis

[0126] The proximal part of tibia was embedded in O.C.T. compound (ebsciences, USA) for cryosection. Transverse sections (5μπι thick) were cut and pooled into fractions consisting of 10 sections per fraction to ensure separation of each cell type in the growth plates before lysed in Trizol® reagent (Invitrogen). Total RNA were extracted and hybridized to Mouse Genome 430 2.0 Gene Chip (Affymetrix). Gene expression data for each sample in triplicate were normalized using Robust Multi-chip Average (RMA) algorithm in R Bioconductor package. The k-Means Clustering algorithm (92, 93) was used to identify the distinct expression patterns of genes in WT and 13del growth plates. The Gene Ontology analysis was performed for each cluster of genes by using the Gene Ontology database (94) and the David Web Tools (95).

HOMER Motif discovery

[0127] The DNA binding motif enrichment analysis was performed by using HOMER software package (96). The DNA sequences flanking the genes' transcription start sites 2kb up- and downstream were extracted from the mouse reference genome assembly (mm9). The HOMER, the TRANSFAC (97) and the ISMARA (98) transcription factor databases were integrated to create the TF binding motif library for screening. The DNA sequences of the interrogated gene sets were compared with those extracted from the remainder gene sets to identify the differentially enriched DNA binding motifs and the TFs.

ChlP-sequencing data analysis

[0128] The ATF4 and CHOP ChlP-sequencing datasets (45) were downloaded from the GEO database (GSE35681). The DNA sequences were aligned to the mm9 mouse reference genome assembly with Bowtie program (99). The analysis of coverage signal intensity and peak detection were performed by using Picard toolkit of Broad Institute (MIT) (https://tldrlegal.com/license/mit- license). The binding peaks located within lOkb up or downstream of the TSS in each target gene were identified for statistical analysis in each cluster.

Histological and immunofluorescence analyses

[0129] Limbs were fixed in 4% PFA, follwed by demineralization in 0.5M EDTA (pH 8.0) prior to embedding in paraffin. Slides were stained with Alcian Blue for cartilage matrix and Fast Red for nuclei. Immunofluorescence was performed using antibodies against ATF4 (sc-200, Santa Cruz), ATF3 (HPA001562, Sigma), CHOP (sc-575, Santa Cruz), GADD34 (sc-825, Santa Cruz), FGF21 (42189, AIS) and Sox9 (AB5535, Millipore).

FAST Staining

10130] FAST staining refers to a multidye staining procedure using fast green, Alcian blue, Safranin-O, and tartrazine and was performed as described previously (100).

In-Situ hybridization

[0131] In-situ hybridization was performed as previously described (101), using [ ³⁵ S]UTP-labeled ribopobes for CollOal, CoUal, Bip, Bdel(lO), Ihh (from A. McMahon), Sox9 (102) and the PTHrP receptor (Ppr) (from H. Kronenberg). The probes for Atf4, Atfi, Chop, Eroll and Fg/21 were mouse cDNA fragments, generated by RT-PCR from growth plate total RNA. The primers used are as follows: Aft, 5'- GAGGTGGCCAAGCACTTGAAA (SEQ ID NO: 32) and 5'- GAACCACCTGGAGAAGGCAGATT (SEQ ID NO: 33); Atfi, 5'- GCTTCCCCAGTGGAGCCAAT (SEQ ID NO : 34) and 5 '- CCACCTCTGCTTAGCTCTGCAAT (SEQ ID NO: 35); Chop, 5'- ATGAGGATCTGCAGGAGGTCCTGTC (SEQ ID NO: 36) and 5'- GATGCCCACTGTTCATGCTTGGT (SEQ ID NO: 37); Eroll, 5'- AAGACTACAAAAGCTTCTTG (SEQ ID NO: 38) and 5 ' -A AG A ATTCTC ATCGA AGTGCA A (SEQ ID NO: 39); and Fg/21, 5'- CAGGGGTCATTCAAATCCTG (SEQ ID NO: 40) and 5'- AGGAATCCTGCTTGGTCTTG (SEQ ID NO: 41).

TUNEL assay

[0132] Apoptotic cells in the growth plate of examined animals were detected by in situ terminal deoxynucleotidyltransferase deoxyuridine triphosphate nick end labeling (TUNEL) assay using the In Situ Cell Death Detection Kit (Roche) following the manufacturer's instructions.

Chromatin imrmmoprecipitation (ChIP) assay [0133] The protocol used for ChlP was adapted from the instructions of ChIP Assay Kit (Millipore). Culture cells or limbs dissected from El 5.5 WT and C10-ATF4 embryos were homogenized and crosslinked. DNA was sonicated and immunoprecipitated with rabbit anti-ATF4 (sc-200, Santa Cruz Biotechnology) or rabbit anti-CHOP (s-575, Santa Cruz Biotechnology) antibody. The pull-down DNA was purified and analyzed by PCR.

Protein extraction and immunoblot analysis

[0134] Cartilages isolated from the mice were pulverized in liquid nitrogen and then lysed with RIPA buffer. The lysate was subjected to SDS-PAGE under reducing conditions and probed with FGF21 and beta-actin antibody.

Dual luciferase reporter assay

[0135] Luciferase assays were conducted using a dual luciferase reporter assay kit (Promega), according to the manufacturer's protocol. Different promoter fragments of Sox.9 or Fg/21 were cloned into pGL3-basic vector (Promega) to drive the expression of firefly luciferase. ATDC5 or NIH3T3 cells were plated at 2x10 ⁴ cells/well in 24-well plates. After 18-hours incubation, the cells were transfected with tested constructs with Renilla luciferase vector, which served as an internal control. Data presented are ratios of Luc/Renilla activity from at least three different experiments and each experiment was performed in triplicate for each DNA sample.

Quantitative PCR

[0136] Quantitative PCR was performed using SYBR-Green master mixture according to the manufacturer's instruction (Takara). Appropriate amounts of cDNA (or DNA) and primers were mixed with distilled water up to ΙΟμΙ and combined with equal amount of SYBR-Green master mixture. The reaction was run on the StepOne Real Time PCR system (Applied Biosystems, A&B).

The Ct (cycle threshold) is defined as the cycle number required for the fluorescent signal to cross the threshold. The relative expression levels of target genes are calculated by normalizing to the expression level of GAPDH using delta-delta-Ct (Relative expression level = 2 ^Λ - (Ct _target - Ct _Gapdh )).

Melting curve was also measured to detect the specificity of the primers.

ISRJB treatments

[0137] ISRIB (SML0843, Sigma) was dissolved in DMSO to make a 5mg/ml stock and stored at 4-degree. Animals were intraperitoneally injected with ISRIB ((103, 104) (2.5 mg/kg, diluted in 0.9% saline) or vehicle (5% DMSO in saline) from El 3.5 till p20 stage. The animals were collected at plO and p20 stages for further analysis.

Radiography of mouse skeleton

[0138] Mice were anesthetized before radiography using digital Faxitron system (UltraFocus) at 20kVA for 5 second exposure.

Statistical analyses [0139] No statistical methods were used to predetermine sample size. Statistical analyses used are detailed in the figure legends. Unpaired two tailed Student's t-test was used to establish statistical significance. For growth analysis, two tailed Mann- Whitney U-test was used. P < 0.05 was considered statistically significant.

Data availability

[0140] All primary microarray data are deposited into Gene Expression Omnibus (GEO) website

(Accession Number GSE99306).

References:

1. P. T. Siegel, R. Clopper, B. Stabler, Psychological impact of significantly short stature. Acta paediatrica Scandinavica. Supplement 377 ', 14- 18; discussion 19 (1991).

2. S. Thompson, T. Shakespeare, M. J. Wright, Medical and social aspects of the life course for adults with a skeletal dysplasia: a review of current knowledge. Disability and rehabilitation 30, 1-12 (2008).

3. C. W, Lie, W. Chow, Limb lengthening in short-stature patients using monolateral and circular external fixators. Hong Kong medical journal = Xianggang yi xue za zhi / Hong Kong Academy of Medicine 15, 280-284 (2009).

4. T. S. Lisse et al, ER stress-mediated apoptosis in a new mouse model of osteogenesis imperfecta. PLoS genetics 4, e7 (2008).

5. B. R. Olsen, Mutations in collagen genes resulting in metaphyseal and epiphyseal dysplasias.

Bone 17, S45-S49 (1995).

6. F. Suleman et al, A novel form of chondrocyte stress is triggered by a COMP mutation causing pseudoachondropiasia. Hum Mutat 33, 218-23 1 (2012).

7. K. A. Pirog, Y. Katakura, A. Mironov, M. D. Briggs, Mild myopathy is associated with COMP but not MATN3 mutations in mouse models of genetic skeletal diseases. PloS one 8, e82412 (2013).

8. M. D. Briggs, K. L. Chapman, Pseudoachondropiasia and multiple epiphyseal dysplasia:

Mutation review, molecular interactions, and genotype to phenotype correlations. Human Mutation 19, 465-478 (2002).

9. K. Y. Tsang, D. Chan, J. F. Bateman, K. S. Cheah, In vivo cellular adaptation to ER stress: survival strategies with double-edged consequences. Journal of cell science 123, 2145-2154 (2010).

10. K. Y. Tsang et al. , Surviving endoplasmic reticulum stress is coupled to altered chondrocyte differentiation and function. PLoS Biol 5, e44 (2007).

1 1. O. Makitie et ai, Schmid type of metaphyseal chondrodysplasia and COL10A1 mutations- findings in 10 patients. Am J Med Genet A 137A, 241 -248 (2005).

12. M. Ridanpaa et al., Genetic changes in the RNA components of RNase MRP and RNase P in Schmid metaphyseal chondrodysplasia. Journal of Medical Genetics 40, 741 -746 (2003).

13. H. Zinszner et al, CHOP is implicated in programmed cell death in response to impaired function of the endoplasmic reticulum. Genes & development 12, 982-995 (1998).

14. C. Hetz, E. Chevet, H. P. Harding, Targeting the unfolded protein response in disease. Nat Rev Drug Discov 12, 703-719 (2013).

15. S. Nundlall et al , An unfolded protein response is the initial cellular response to the expression of mutant matrilin-3 in a mouse model of multiple epiphyseal dysplasia. Cell Stress & Chaperones 15, 835-849 (2010).

16. K. L. Posey et al., Chondrocyte-specific pathology during skeletal growth and therapeutics in a murine model of pseudoachondropiasia. (2014).

17. D. T. Rutkowski, R. S . Hegde, Regulation of basal cellular physiology by the homeostatic unfolded protein response. (2010).

18. K. Pakos-Zebrucka et al, The integrated stress response. EMBO reports 17, 1374-1395 (2016).

19. C.-Q. Zhao, Y.-H. Zhang, S.-D. Jiang, L.-S. Jiang, L.-Y. Dai, Both endoplasmic reticulum and mitochondria are involved in disc cell apoptosis and intervertebral disc degeneration in rats. ,Age 32, 161-177 (2010).

20. D. Xu et al, Hydrogen sulfide protects against endoplasmic reticulum stress and mitochondrial injury in nucleus pulposus cells and ameliorates intervertebral disc degeneration. Pharmacological Research 117, 357-369 (2017).

21. T. Oosterhuis et al , Early rehabilitation after lumbar disc surgery is not effective or cost- effective compared to no referral: a randomised trial and economic evaluation. Journal of Physiotherapy 63, 144-153.

22. H. Kanazawa et al, Efficacy of growth hormone therapy for patients with skeletal dysplasia.

Journal of Bone and Mineral Metabolism 21, 307-310 (2003).

23. S. Suzuki et al, Excessive reactive oxygen species are therapeutic targets for intervertebral disc degeneration. Arthritis Research & Therapy 17, 316 (2015).

24. A. Hiyama et al., Hypoxia Activates the Notch Signaling Pathway in Cells of the Intervertebral Disc: Implications in Degenerative Disc Disease. Arthritis and rheumatism 63, 1355-1364 (2011).

25. L. J. Smith, N. L. Nerurkar, K.-S. Choi, B. D. Harfe, D. M. Elliott, Degeneration and regeneration of the intervertebral disc: lessons from development. Disease Models & Mechanisms 4, 31 -41 (2011 ) .

26. A. G. Hadjipavlou, M. N. Tzermiadianos, N. Bogduk, M. R. Zindrick, The pathophysiology of disc degeneration. Journal of Bone &amp;amp; Joint Surgery, British Volume 90-B, 1261 (2008).

27. H. S. An, K. Masuda, N. Inoue, Intervertebral disc degeneration: biological and biomechanical factors. Journal of Orthopaedic Science 11, 541 -552 (2006).

28. I. C. Salaroglio et al., PERK induces resistance to cell death elicited by endoplasmic reticulum stress and chemotherapy. Molecular Cancer 16, 91 (2017).

29. L. R. Palam, J. Gore, K. E. Craven, J. L. Wilson, M. Korc, Integrated stress response is critical for gemcitabine resistance in pancreatic ductal adenocarcinoma. Cell Death & Disease d, el913 (2015).

30. P. Podszywalow-Bartnicka et al, Increased phosphorylation of eIF2α in chronic myeloid leukemia cells stimulates secretion of matrix modifying enzymes. Oncotarget 7, 79706- 79721 (2016).

31. Y. Y. Ho, D. Lagares, A. M. Tager, M. Kapoor, Fibrosis[mdash]a lethal component of systemic sclerosis. Nat Rev Rheumatol 10, 390-402 (2014).

32. P. Cheresh, S.-J. Kim, S. Tulasiram, D. W. Kamp, Oxidative Stress and Pulmonary Fibrosis.

Biochimica et biophysica acta 1832, 1028-1040 (2013).

33. R. M. Liu, K. A. Gaston Pravia, Oxidative stress and glutathione in TGF-β -mediated fibrogenesis. Free Radical Biology and Medicine 48, 1 -15 (2010).

34. I. Cucoranu et al, NAD(P)H Oxidase 4 Mediates Transforming Growth Factor-β1 -Induced Differentiation of Cardiac Fibroblasts Into Myofibroblasts. Circulation Research 97, 900 (2005).

35. G. H. Kim, J. E. Kim, S. J. Rhie, S. Yoon, The Role of Oxidative Stress in Neurodegenerative Diseases. Experimental Neurobiology 24, 325-340 (2015).

36. D. Lindholm, H. Wootz, L. Korhonen, ER stress and neurodegenerative diseases. Cell Death Differ 13, 385-392 (2006).

37. S. E. Patterson, C. N. Dealy, Mechanisms and models of endoplasmic reticulum stress in chondrodysplasia. Developmental Dynamics 243, 875-893 (2014).

38. Stela S. Palii, Michelle M. Thiaville, Y.-X. Pan, C. Zhong, Michael S. Kilberg, Characterization of the amino acid response element within the human sodium-coupled neutral amino acid transporter 2 (SNAT2) System A transporter gene. Biochemical Journal 395, 517-527 (2006).

39. J. Buenrostro, B. Wu, H. Chang, W. Greenleaf, ATAC-seq: A Method for Assaying Chromatin Accessibility Genome- Wide. Current protocols in molecular biology / edited by Frederick M. Ausubel ... fet al.] 109, 21.29.21-21.29.29 (2015).

40. L. Nel-Themaat et αί , Morphometric analysis of testis cord formation in Sox9-EGFP mice.

Developmental Dynamics 238, 1100-1110 (2009).

41. L. Madisen et al. , Transgenic mice for intersectional targeting of neural sensors and effectors with high specificity and performance. Neuron 85, 942-958 (2015).

42. E. Provost, J. Rhee, S. D. Leach, Viral 2A peptides allow expression of multiple proteins from a single ORF in transgenic zebrafish embryos, genesis 45, 625-629 (2007).

43. N. A. Kearns et al, Functional annotation of native enhancers with a Cas9-histone demethylase fusion. Nat Meth 12, 401-403 (2015).

44. L. Cong et al, Multiplex Genome Engineering Using CRISPR/Cas Systems. Science (New York, N Y.) 339, 819-823 (2013).

45. J. Han et al. , ER-stress-induced transcriptional regulation increases protein synthesis leading to cell death. Nat Cell Biol 15, 481 -490 (2013).

46. C. M. Oslowski, F. Urano, Measuring ER stress and the unfolded protein response using mammalian tissue culture system. Methods in enzymology 490, 71-92 (2011 ).

47. H. P. Harding et al , Regulated translation initiation controls stress-induced gene expression in mammalian cells. Molecular celt 6, 1099-1108 (2000).

48. H. Hojo, A. P. McMahon, S. Ohba, An Emerging Regulatory Landscape for Skeletal Development. Trends Genet 32, 774-787 (2016).

49. G. S. Hotamisligil, R. J. Davis, Cell Signaling and Stress Responses. Cold Spring Harb Perspect Biol 8, (2016) .

50. P. Walter, D. Ron, The Unfolded Protein Response: From Stress Pathway to Homeostatic Regulation. Science 334, 1081 (2011).

51. K. Horiuchi, T. Tohmonda, H. Morioka, The unfolded protein response in skeletal development and homeostasis. Cell Mol Life Sci 73, 2851-2869 (2016).

52. A. Sudhakar et al, Phosphorylation of Serine 51 in Initiation Factor 2a (eIF2α) Promotes Complex Formation between eIF2α(P) and eIF2B and Causes Inhibition in the Guanine Nucleotide Exchange Activity of eIF2B. Biochemistry 39, 12929-12938 (2000).

53. T. L. Cameron et al, XBP1 -Independent UPR Pathways Suppress C/EBP-beta Mediated Chondrocyte Differentiation in ER-Stress Related Skeletal Disease. PLoS genetics 11, el005505 (2015).

54. W. Wang et al, Atf4 regulates chondrocyte proliferation and differentiation during endochondral ossification by activating Ihh transcription. Development (Cambridge, England) 136, 4143-4153 (2009).

55. V. Y. L. Leung et al, SOX9 Governs Differentiation Stage-Specific Gene Expression in Growth Plate Chondrocytes via Direct Concomitant Transactivation and Repression. PLoS genetics 7, el002356 (2011).

56. L. Yang, K. Y. Tsang, H. C. Tang, D. Chan, K. S. Cheah, Hypertrophic chondrocytes can become osteoblasts and osteocytes in endochondral bone formation. Proc Natl Acad Sci U S

A \\\, 12097-12102 (2014).

57. S. Strieker, R. Fundele, A. Vortkamp, S. Mundlos, Role of Runx Genes in Chondrocyte Differentiation. Developmental Biology 245, 95-108 (2002).

58. L. Koziel, M. Wuelling, S. Schneider, A. Vortkamp, Gli3 acts as a repressor downstream of Ihh in regulating two distinct steps of chondrocyte differentiation. Development (Cambridge,

England) 132, 5249 (2005).

59. A. Ionescu et al, FoxA family members are crucial regulators of the hypertrophic chondrocyte differentiation program. Developmental Cell 22, 927-939 (2012).

60. H. Akiyama, M.-C. Chaboissier, J. F. Martin, A. Schedl, B. de Crombrugghe, The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes & development 16, 2813-2828 (2002).

61. C. F. Liu, W. E. Samsa, G. Zhou, V. Lefebvre, Transcriptional control of chondrocyte specification and differentiation. Semin Cell Dev Biol 62, 34-49 (2017).

62. D. M. Bell et al, SOX9 directly regulates the type-11 collagen gene. Nat Genet 16, 174-178 (1997).

63. P. Dy et al , Sox9 directs hypertrophic maturation and blocks osteoblast differentiation of growth plate chondrocytes. Dev Cell 22, 597-609 (2012).

64. S. J. Marciniak et al, CHOP induces death by promoting protein synthesis and oxidation in the stressed endoplasmic reticulum. Genes & development 18, 3066-3077 (2004).

65. M. Pennuto et al, ABLATION OF THE UPR-MEDIATOR CHOP RESTORES MOTOR FUNCTION AND REDUCES DEMYELINATION IN CHARCOT MARIE TOOTH IB MICE. Neuron 57, 393-405 (2008).

66. C. M. Southwood, J. Garbern, W. Jiang, A. Gow, The Unfolded Protein Response Modulates Disease Severity in Pelizaeus-Merzbacher Disease. Neuron 36, 585-596 (2002).

67. M. Lu et al, Opposing unfolded-protein-response signals converge on death receptor 5 to control apoptosis. Science 345, 98-101 (2014).

68. J. A. Moreno et al, Sustained translational repression by eIF2a!pha-P mediates prion neurodegeneration. Nature 485, 507-511 (2012).

69. K. A. Pirog et al , Abnormal chondrocyte apoptosis in the cartilage growth plate is influenced by genetic background and deletion of CHOP in a targeted mouse model of pseudoachondroplasia. PloS one 9, e 85145 (2014).

70. K. L. Posey et al. , Chop (Ddit3) is essential for D469del-COMP retention and cell death in chondrocytes in an inducible transgenic mouse model of pseudoachondroplasia. Am J Pathol

180, 727-737 (2012).

71. C. Jousse et al. , Inhibition of a constitutive translation initiation factor 2a phosphatase, CReP, promotes survival of stressed cells. The Journal of Cell Biology 163, 767-775 (2003). 72. A. L. De Sousa-Coelho, P. F. Marrero, D. Haro, Activating transcription factor 4-dependent induction of FGF21 during amino acid deprivation. Biochem J 443, 165-171 (2012).

73. T. L. Cameron et al, Transcriptional Profiling of Chondrodysplasia Growth Plate Cartilage

Reveals Adaptive ER-Stress Networks That Allow Survival but Disrupt Hypertrophy. PloS one 6, e24600 (2011).

74. A. Kharitonenkov et al, FGF-21 as a novel metabolic regulator. The Journal of Clinical Investigation 115, 1627- 1635 (2005).

75. M. A. Gomez-Samano et al, Fibroblast growth factor 21 and its novel association with oxidative stress. Redox Biol 11, 335-341 (2017).

76. K. H. Kim, M. S. Lee, FGF21 as a mediator of adaptive responses to stress and metabolic benefits of anti-diabetic drugs. J Endocrinol 226, Rl-16 (2015).

77. A. Salminen, K. Kaarniranta, A. Kauppinen, Regulation of longevity by FGF21 : Interaction between energy metabolism and stress responses. Ageing Res Rev 37, 79-93 (2017).

78. K. H. Kim et al. , Autophagy deficiency leads to protection from obesity and insulin resistance by inducing Fgf21 as a mitokine. Nat Med 19, 83-92 (2013).

79. Y. Sekine et al, Mutations in a translation initiation factor identify the target of a memory- enhancing compound. Science 348, 1027 (2015).

80. C. Sidrauski et al, Pharmacological dimerization and activation of the exchange factor eIF2B antagonizes the integrated stress response. eLife 4, e07314 (2015).

81. G. V. Di Prisco et al, Translational control of mGluR-dependent long-term depression and object-place learning by eIF2alpha. Nature neuroscience 17, 1073-1082 (2014).

82. R. Savarirayan, V. Cormier-Daire, R. S. Lachman, D. L. Rimoin, Schmid type metaphyseal chondrodysplasia: a spondylometaphyseal dysplasia identical to the "Japanese" type. Pediatric Radiology 30, 460-463 (2000).

83. D. Samartzis et al, Classification of Schmorl's nodes of the lumbar spine and association with disc degeneration: a large-scale population-based MRI study. Osteoarthritis and

Cartilage 24, 1753-1760.

84. H. Wang et al, Role of death receptor, mitochondrial and endoplasmic reticulum pathways in different stages of degenerative human lumbar disc. Apoptosis 16, 990 (2011). 85. M. Franke et al, Formation of new chromatin domains determines pathogenicity of genomic duplications. Nature 538, 265-269 (2016).

86. S. Cuddapah et al, Global analysis of the insulator binding protein CTCF in chromatin barrier regions reveals demarcation of active and repressive domains. Genome Research 19, 24-32 (2009).

87. Dario G. Lupianez et al , Disruptions of Topological Chromatin Domains Cause Pathogenic Rewiring of Gene-Enhancer Interactions. Cell 161, 1012-1025.

88. O. Symmons et al, Functional and topological characteristics of mammalian regulatory domains. Genome Res 24, 390-400 (2014).

89. Y. Hotta et al, Fibroblast Growth Factor 21 Regulates Lipolysis in White Adipose Tissue But Is Not Required for Ketogenesis and Triglyceride Clearance in Liver. Endocrinology 150, 4625-4633 (2009).

90. H. E. Ryan, J. Lo, R. S. Johnson, HIF-lo. is required for solid tumor formation and embryonic vascularization. The EMBO Journal 17, 3005 (1998).

91. E. Schipani et al, Hypoxia in cartilage: HIF-Ια is essential for chondrocyte growth arrest and survival. Genes & development 15, 2865-2876 (2001).

92. J. Quackenbush, Computational analysis of raicroarray data. Nature reviews. Genetics 2, 418-427 (2001).

93. A. K. Jain, Data clustering: 50 years beyond K-means. Pattern Recognition Letters 31, 651- 666. (2010).

94. M. Ashburner et al, Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25, 25-29 (2000).

95. G. Dennis, Jr. et al, DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol 4, P3 (2003).

96. S. Heinz et al, Simple combinations of lineage-determining transcription factors prime cis- regulatory elements required for macrophage and B cell identities. Molecular cell 38, 576- 589 (2010).

97. E. Wingender, P. Dietze, H. Karas, R. Knuppel, TRANSFAC: a database on transcription factors and their DNA binding sites. Nucleic Acids Res 24, 238-241 (1996),

98. P. J. Balwierz et al, ISMARA: automated modeling of genomic signals as a democracy of regulatory motifs. Genome Res 24, 869-884 (2014).

99. B. Langmead, C. Trapnell, M. Pop, S. L. Salzberg, Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biology 10, R25-R25 (2009).

100. V. Y. L. Leung, W. C. W. Chan, S.-C. Hung, K. M. C. Cheung, D. Chan, Matrix Remodeling During Intervertebral Disc Growth and Degeneration Detected by Multichromatic FAST

Staining. Journal of Histochemistry and Cytochemistry 57, 249-256 (2009).

101. A. W. K. Wai et al, Disrupted expression of matrix genes in the growth plate of the mouse cartilage matrix deficiency (cmd) mutant. Developmental Genetics 22, 349-358 (1998).

102. L.-J. Ng et al , SOX9 Binds DNA, Activates Transcription, and Coexpresses with Type II Collagen during Chondrogenesis in the Mouse. Developmental Biology 183, 108-121 (1997).

103. C, Sidrauski et al, Pharmacological brake-release of mRNA translation enhances cognitive memory. Elife 2, e00498 (2013).

104. G. V. Di Prisco et al, Translational control of mGluR-dependent long-term depression and object-place learning by eIF2[alpha]. Nature neuroscience 17, 1073-1082 (2014).

105. H, Zeng et al, An inducible and reversible mouse genetic rescue system. PLoS Genetics 4(5):el000069. doi: 10.1371/joumal.pgen. l000069 (2008).

106. W.C.W. Chan et al, Activating the unfolded protein response in osteocytes causes hyperostosis consistent with craniodiaphyseal dysplasia. Human Molecular Genetics doi: 10.1093/hmg/ddx339 (2017).