Diverse Phenotypes and Specific Transcription Patterns in Twenty Mouse Lines with Ablated LincRNAs

Abstract

In a survey of 20 knockout mouse lines designed to examine the biological functions of large intergenic non-coding RNAs (lincRNAs), we have found a variety of phenotypes, ranging from perinatal lethality to defects associated with premature aging and morphological and functional abnormalities in the lungs, skeleton, and muscle. Each mutant allele carried a lacZ reporter whose expression profile highlighted a wide spectrum of spatiotemporal and tissue-specific transcription patterns in embryos and adults that informed our phenotypic analyses and will serve as a guide for future investigations of these genes. Our study shows that lincRNAs are a new class of encoded molecules that, like proteins, serve essential and important functional roles in embryonic development, physiology, and homeostasis of a broad array of tissues and organs in mammals.

Fig 1. Strategy For Targeted Disruption of the Fendrr Gene Locus.

A partial map of the mouse Fendrr locus is shown encompassing exons 1 to 6. Upon homologous recombination, the targeting BacVec replaced a total of 19.2 kb of the Fendrr genomic sequence with the lacZ-neomycin resistance cassette, introducing a Kozak-ATG sequence. Open boxes indicate noncoding exons. Red and green boxes on the Fendrr genomic locus and in the LacZ-neomycin resistance cassette are homologous sequences used for targeting.

Fig 2. Spatial lacZ Reporter Gene Expression in Mid-gestation Stage LincRNA Gene-Targeted Mouse Embryos.

Heterozygous E12.5 embryos fixed and stained for ß-galactosidase showed a broad range of expression (blue) of the introduced lacZ reporter gene in the developing brain and craniofacial region (Pantr1, Pantr2, Celrr, and Peril, see also S1 Fig), neural tube (Pantr2, Halr1, and Lincppara), dorsal aorta (Celrr), heart (Celrr, Peril, see also S1 Fig), lungs (Fendrr), limb buds (Hottip, HoxA11os, and Mannr), foregut (Hottip, HoxA11os, and Fendrr), posterior region and the tail (Hotair, Hottip, and HoxA11os). Tug1 showed a widespread lacZ expression pattern, whereas expression of other reporter genes was restricted to the epidermis (Eldr), mammary buds (Lincenc1, see also S1 Fig), or the whisker placode (Trp53cor1, see also S1 Fig). Examples shown are representative of at least five genotype-confirmed embryos per lincRNA knockout project.

Fig 3. Temporal lacZ Reporter Gene Expression in Mid-gestation Stage LincRNA Gene-Targeted Mouse Embryos.

Temporal expression patterns for Hottip, HoxA11os, and Celrr in F1 heterozygous embryos from the indicated stages (E9.5-E12.5) showed that expression began early at a restricted site and then extended beyond this initial site at later stages. Celrr expression was confined to the brain at E9.5 and progressed into the spinal cord by E12.5. HoxA11os expression began in the developing tail bud and progressed into the entire caudal region of the embryo, hind limb and forelimb by E12.5. Hottip expression also began in the developing tail bud and was then observed in the developing distal autopods of the forelimb and hind limb by E11.5 and E12.5. Examples shown are representative of at least 5, genotype-confirmed embryos per LincRNA project stained.

Fig 4. LacZ Reporter Expression in Brains of 6–8 Week Old lincRNA Gene-Targeted F0 Generation Heterozygotes.

The brain lacZ expression pattern (blue) for each lincRNA gene is as follows: (A) Crnde, the colliculi (dorsal view, arrow); (B) Pantr1, neocortex, olfactory bulb, basal forebrain, and hypothalamus; (C) Pantr2, neocortex, olfactory bulb, cerebellum, hypothalamus, and basal forebrain; (D) Lincenc1, neocortex, parts of the cerebellum and medial hypothalamus, especially strong patterning in the olfactory projection and olfactory projection areas of the temporal cortex (ventral view, red arrow); (E) Celrr, broadly in gray matter with the exception of the lateral cerebellum and ventral pons; (F) Kantr, possibly in deep cerebellar layers (dorsal view, star); (G) Lincpint, ubiquitously in gray matter, especially intense in the hypothalamus; (H) Lincppara, ubiquitously in gray matter, especially dense in the hypothalamus; (I) Peril, midline of the hypothalamus (ventral view, arrowhead); and (J) Tug1, spinal cord gray matter and light gray matter in most structures except for the neocortex. n = 2, genotype confirmed male mice per lincRNA knockout project.

Fig 5. Increased Expression of Lincpint from Postnatal Day 3 to 8 Weeks of Age.

LacZ reporter expression (blue) at 3 days, 3 weeks, and 8 weeks in F0 heterozygotes shows increasing Lincpint expression with age. (A) At 3 days, ßgalactosidase staining is only observed in portions of the brain, tendons and ligaments of the hind limb, and some bronchioles in the lung (arrow). (B) At 3 weeks, there is increased staining in the brain, hindlimb, atria of the heart, lung, and liver. (C) By 8 weeks of age, the whole brain, skeletal muscle of the hindlimb and chest, atria and myocardium, lung, and liver tissue all exhibit strong ß-galactosidase staining representative of increased Lincpint expression. Examples shown are representative of n>4 mice per group. (D) RT-PCR analysis of GFP in GA (gastrocnemius) muscle and kidney isolated from 8 week-old and 52 week-old wild type (WT), heterozygous (Lincpint-GFP ^+⁄−) and homozygous (Lincpint-GFP ^−⁄−) mice. (E) Comparison of endogenous Lincpint RNA level in GA (gastrocnemius) muscle and kidney of 8 week-old versus 52 week-old wild type (WT) tissues demonstrating increased in Lincpint gene expression. Examples shown are representative of n>4 mice per group.

Fig 6. Aging-associated Phenotypes in Lincpint Knockout Mice.

(A) Lincpint ^−⁄− and Lincpint ^+⁄− male mice exhibit a significantly slower growth rate than their wild type (WT) littermates and begin to show significant weight loss near 6 months of age. Data are plotted as the mean +/− SEM, n > 9 mice for each group. Significance was assessed by a one-way ANOVA (*, P < 0.05; **, P < 0.005; ***, P < 0.001). (B) Kaplan-Meier analysis of homozygous with heterozygous and WT mice. Lincpint ^−⁄− male mice exhibit a significant reduction in survival compare to Lincpint ^+⁄− and wild type littermates. Data are plotted as percent survival over 1 year observation. (C) Ventral and dorsal skin sections in Lincpint ^−⁄− mice compared with Lincpint ^+⁄− and WT littermates. (D, E, F, and G) MicroCT evaluation of body composition at 12-, 26- and 52-weeks of age. (D, E) Male Lincpint ^−⁄− and Lincpint ^+⁄− mice exhibit a significant reduction in body fat as early as 26-week of age. Female Lincpint ^−⁄− mice have reduced body fat at an older age noticeably at 52-week of age (***, P < 0.001, one-way ANOVA). (F, G) A significant reduction in femur bone mineral density (BMD) observed in both males and females Lincpint ^−⁄− compared with their Lincpint ^+⁄− and WT littermates (*, P < 0.05; ***, P < 0.001, one-way ANOVA). (H) MicroCT images depict pronounced lordokyphosis (curvature of the spinal column) seen in older male and female Lincpint ^−⁄− mice compared with WT littermates. (I) Approximately 70% (6/9 males and 7/10 females) of Lincpint ^−⁄− mice have lordokyphosis by 12 weeks of age, compared with 0–20% of Lincpint ^+⁄− (1/12 males and 2/10 females) and WT (1/10 males and 0/11 females) littermates. By 26 weeks of age the proportion of Lincpint ^−⁄− mice with lordokyphosis increased to nearly 90% (7/8 males and 8/9 females) and appeared in approximately 60% (8/12 males and 6/10 females) of Lincpint ^+⁄− mice, compared with less than 20% (2/10 males and 2/11 females) of WT littermates. n ≥ 9 mice per group for all observations reported.

Fig 7. Abnormal Lung Morphology in Fendrr Knockout Mice at E13.5.

(A) LacZ reporter gene expression at E12.5 in Fendrr ^−⁄− embryos in the frontonasal region (FN) of the face, the aorta gonad mesonephros (AGM) region, and the respiratory tract, including the lungs (L) and trachea (T). (B) Dissection of lungs at E13.5 revealed an abnormal, disorganized, globular phenotype in the lobes of Fendrr ^−⁄− embryos compared with Fendrr ^+⁄−.

Fig 8. Homeotic Transformation Observed in the 4th Caudal Vertebra of Hotair ^−⁄− mice.

(A) Visualization of the sacral and caudal region of the mouse skeleton by microCT reveals a homeotic transformation in Hotair ^−⁄− mice of the 4th caudal vertebra to a structure similar to that of the 3rd caudal vertebra. (B) Dorsal, lateral and ventral comparison of WT and Hotair ^−⁄− 4th caudal vertebra reveals a structural abnormality in homozygotes indicative of a homeotic transformation.

Fig 9. Abnormal Hindlimb Posture, Reduced Grip Strength, and Muscle Wasting in Hottip ^−⁄− Mice.

(A) Hottip ^−⁄− mice demonstrated unusual hindlimb clasping posture when suspended by the tail. (B) Cage endurance testing revealed that Hottip ^−⁄− mice have a reduced ability to remain suspended from an inverted wire cage top, n = 5 mice for each group. (C) The right and left TA (tibialis anterior), GA (gastrocnemius) and Quad (quadriceps) muscles from WT, Hottip ^+⁄− and Hottip ^−⁄− mice were weighed. Muscle weights are normalized to body weight and calculated to include both right/left muscle weights. Data are means +/−SEM, n = 6 mice for each group. A significant decrease in muscle weight was observed only in the GA of Hottip ^−⁄− animal in both males and females (male data not shown). Asterisks indicate a significant difference in the Hottip ^−⁄− GA muscle weights compared to all other control groups (P< 0.01). (D) Comparison of GA muscle fiber numbers in WT, Hottip ^+⁄− and Hottip ^−⁄−. A significant reduction of fiber count was observed in Hottip ^−⁄−. Significance assessed by one-way ANOVA (P < .0001 (E) Comparison of mean cross-sectional area of muscle fibers. Cross sections taken from the GA muscle were stained with an antibody against laminin and measured. There is no noticeable size difference between Hottip ^−⁄− and control skeletal muscles, n = 6 mice per group for all muscle analyses.