Seasonal oscillation of liver-derived hibernation protein complex in the central nervous system of non-hibernating mammals

Abstract

Mammalian hibernation elicits profound changes in whole-body physiology. The liver-derived hibernation protein (HP) complex, consisting of HP-20, HP-25 and HP-27, was shown to oscillate circannually, and this oscillation in the central nervous system (CNS) was suggested to play a role in hibernation. The HP complex has been found in hibernating chipmunks but not in related non-hibernating tree squirrels, leading to the suggestion that hibernation-specific genes may underlie the origin of hibernation. Here, we show that non-hibernating mammals express and regulate the conserved homologous HP complex in a seasonal manner, independent of hibernation. Comparative analyses of cow and chipmunk HPs revealed extensive biochemical and structural conservations. These include liver-specific expression, assembly of distinct heteromeric complexes that circulate in the blood and cerebrospinal fluid, and the striking seasonal oscillation of the HP levels in the blood and CNS. Central administration of recombinant HPs affected food intake in mice, without altering body temperature, physical activity levels or energy expenditure. Our results demonstrate that HP complex is not unique to the hibernators and suggest that the HP-regulated liver-brain circuit may couple seasonal changes in the environment to alterations in physiology.

Keywords: CTRP; Cerebrospinal fluids; Hibernation; Metabolism; Plasma proteins.

Fig. 1.

Identification of chipmunk HP homologs in non-hibernating mammals. (A) Homologs of chipmunk HPs found in other vertebrate species. HP gene loci are found in thirteen-lined ground squirrel (Ictidomys tridecemlineatus; scaffold JH393302), pig (Sus scrofa; contig FP104567), European rabbit (Oryctolagus cuniculus; contig NC_013672), gray mouse lemur (Microcebus murinus; contig NT_165971), greater galago (Otolemur garnettii; contig NT_091708), lesser hedgehog tenrec (Echinops telfairi; contig NT_166802), nine-banded armadillo (Dasypus novemcintus; contig NW_00492307), bottlenose dolphin (Tursiops truncate; contig NW_004200895) and cow (Bos taurus; contig NC_007303) genomes. The transcriptional orientations of HP genes are indicated by the direction of the arrow. ψ, a possible HP-like pseudogene due to nonsense mutation in an exon that encodes the functional C1q domain. (B) Human HP pseudogenes are located on chromosome 22 between the TIMP3 and LARGE genes. Depicted is exon (Ex) 3 of human HP pseudogenes, which encodes the C-terminal globular C1q domain. Nonsense mutations and insertions are indicated with arrows. Exons 1 and 2 of human HP pseudogenes have degenerated beyond recognition.

Fig. 2.

Exon and intron and domain structure of cow HP genes and proteins. Cow HP genes are located on chromosome 5. The exon and intron and protein domain structures of cow HP-20 (A), HP-25 (B) and HP-27 (C) genes. Gray bars indicate exons that code for protein, and white bars indicate exons that code for the 5′ and 3′ UTR of the transcript. Four major domains are found in cow HP-20, HP-25 and HP-27 – a signal peptide (SP), an N-terminus (N-ter) with one conserved cysteine residue, a collagen domain with 12 or 13 G-x-Y repeats, and a C-terminal globular C1q domain. All cysteine residues are indicated with a ball-and-stick. a.a., amino acid residue.

Fig. 3.

Sequence alignment of vertebrate HPs. (A–C) ClustalW alignments of the C1q domain of HP proteins from Siberian chipmunk, cow, pig, gray mouse lemur, greater galago, lesser hedgehog tenrec, European rabbit, bottlenose dolphin and nine-banded armadillo. Identical amino acids are shaded and gaps are indicated by a dashed line. All conserved cysteine residues are indicated with a ball-and-stick. Arrows indicate the highly conserved residues found in all C1q/TNF superfamily members (Shapiro and Scherer, 1998).

Fig. 4.

Phylogenetic analysis of cow HPs. (A) Phylogenetic tree of cow C1q family members. GenBank accession number for each of the cow proteins: C1q-A chain (NP_001014945), C1q-B chain (NP_001040064), Otolin (NP_001179533), multimerin-1 (Ensembl; Acc:ENSBTAP00000013581), multimerin-2 (XP_589798.3), emilin-1 (XP_594120), emilin-2 (XP_592120.3), CTRP1 (NP_001076878), CTRP2 (XP_589781), CTRP3 (NP_001094608), CTRP4 (XP_001788795), CTRP5 (NP_001092608), CTRP6 (NP_001095342), CTRP7 (NP_001069669), CTRP8 (XP_602976.2), CTRP9 (NP_001069314), CTRP10 (XP_595145.3), CTRP11 (Swiss-Prot;Acc:Q86Z23), CTRP13 (XP_870462.2), CRF (XP_597900.3), adiponectin (NP_777167), caprin-2 (XP_876944.2), cerebellin-1 (Cbln1; XP_876031.2), Cbln2 (NP_001092389), Cbln3 (NP_001073071), Cbln4 (NP_001075006), cow HP-20 (FJ645734), cow HP-25 (FJ645735) and cow HP-27 (FJ645736). (B) Phylogenetic tree of mammalian HPs. The cladogram is based on the sequence alignment of the C1q domain and it is different from the species tree.

Fig. 5.

Liver-specific expression of cow HPs mRNAs. (A) Quantitative real-time PCR analyses of HP-20, HP-25 and HP-27 mRNA expression in cow tissues. Transcript abundance was normalized to 18S rRNA. (B) HEK 293 and HepG2 cell lines were transfected with control pGL3-Basic vector or pGL3-Basic vector containing an HP-20 (−1035 bp) or HP-27 (−1348 bp) promoter region upstream from the transcriptional start site (left panel). Luciferase activity was assayed and values were normalized against total protein content. Values were reported as fold increase of relative luminescence when compared with cells that expressed the control plasmid pGL3-Basic (right panel). Means ± s.e.m. are shown. *P<0.05 and **P<0.005 compared with cells transfected with control pGL3-Basic vector.

Fig. 6.

Cow HPs are secreted multimeric glycoproteins. (A) Immunoblot analyses of the cell lysate and supernatant harvested from transfected HEK 293 cells expressing FLAG-tagged cow HPs using the antibody against FLAG. (B) Immunoblot analyses of recombinant cow HPs treated with (+) or without (−) N-glycanase (PNGase F). (C) Immunoblot analyses of recombinant cow HPs treated with (+) or without (−) reducing agent (β-mercaptoethanol, β-ME). (D) Native gel immunoblot analyses of wild-type (WT) and cysteine mutants of cow HPs. (E) Supernatants from transfected cells expressing different combinations of epitope-tagged cow HPs were subjected to immunoprecipitation (IP) and immunoblot analyses with the indicated antibodies (the first two panels). The bottom three panels indicate the presence of FLAG-, HA- and Myc-tagged input proteins. (F) Diagram showing the globular-domain-only constructs of cow HPs (g-HP). SP, signal peptide. (G) Immunoprecipitation and immunoblot analyses with the indicated antibodies of supernatants harvested from transfected cells that coexpressed different combinations of epitope-tagged cow g-HPs (first two panels). The bottom three panels indicate the presence of FLAG-, HA- and Myc-tagged input proteins.

Fig. 7.

Cow HP levels oscillate in serum and CSF as distinct heteromeric complexes. (A) Immunoblot analysis was performed to demonstrate the specificity of antibodies against HP-25 (top panel) and HP-27 (middle panel). The bottom panel shows the presence of epitope-tagged HPs. (B) Immunoblot analyses revealed the presence (indicated by arrows) of HP-25 and HP-27 in cow, but not mouse, serum. *Non-specific band. (C) Native gel immunoblot analyses revealed the existence of two distinct heteromeric complexes of cow HPs in serum. A representative result from one of the cow samples is shown. (D) Native gel immunoblot analyses of cow HPs in CSF collected at different times of the year. A representative result from one of the cow samples is shown. Serum from the same cow was used in C and D. (E,F) Reducing SDS-PAGE immunoblot analyses of cow HP-25, HP-27, and albumin (loading control) in serum (E) and CSF (F) at different times of the year. A representative result from one of the cow samples is shown. (G,H) Seasonal oscillation of cow HPs (N=3) in CSF (E) and serum (F) as quantified by reducing SDS-PAGE immunoblot analyses. For each separate year, serum or CSF levels of HP-25 and HP-27 were normalized to their peak level (set at 1).

Fig. 8.

Cow HPs act in the CNS to modulate food intake. Vehicle control or recombinant HPs (equal mixture of the three proteins at 0.1 μg μl⁻¹) was injected into the lateral ventricle of cannulated mice (N=6 per group); multiple measurements followed. (A) Food intake over a 24 h period. (B) Locomotor activity during light (12 h) and dark cycles (12 h). (C) Body temperature before (0 h) and after i.c.v. injection of recombinant HPs. (D–F) Indirect calorimetry measurements of (D) and energy expenditure (E). *P<0.05 (between vehicle and recombinant HPs).