μ-Crystallin as an Intracellular 3,5,3′-Triiodothyronine Holder in Vivo

Abstract

Previously, we identified reduced nicotinamide adenine dinucleotide phosphate-dependent cytosolic T₃ binding protein in rat cytosol. Cytosolic T₃-binding protein is identical to μ-crystallin (CRYM). Recently, CRYM mutations were found in patients with nonsyndromic hereditary deafness. Although it has been established that CRYM plays pivotal roles in reserving and transporting T₃ into the nuclei in vitro and has a clinical impact on hearing ability, the precise functions of CRYM remain to be elucidated in vivo. To further investigate the in vivo functions of CRYM gene products, we have generated mice with targeted disruption of the CRYM gene, which abrogates the production of CRYM. CRYM knockout loses the reduced nicotinamide adenine dinucleotide phosphate-dependent T₃ binding activity in the cytosol of the brain, kidney, heart, and liver. At the euthyroid state, knockout significantly suppresses the serum concentration of T₃ and T₄ despite normal growth, heart rate, and hearing ability. The disruption of the gene does not alter the expression of TSHβ mRNA in the pituitary gland or glutathione-S-transferase α2 and deiodinase 1 mRNAs in either the liver or kidney. When radiolabeled T₃ is injected intravenously, labeled T₃ rapidly enters into and then escapes from the tissues in CRYM-knockout mice. These data suggest that because of rapid T₃ turnover, disruption of the CRYM gene decreases T₃ concentrations in tissues and serum without alteration of peripheral T₃ action in vivo.

THE VARIETY OF effects on growth, development, and metabolism attributable to thyroid hormone suggests that this hormone acts through a fundamental mechanism common to many different tissues, yet is capable of multiple functions (1). The pivotal action of T₃ is initiated through binding to its nuclear receptors in the target tissues (2). After secretion of the hormone from thyroid gland, T₃ travels a long distance to the target. Thyroid hormone moves from outside the plasma membrane into the extranuclear space of the cells through active transporters or by passive diffusion (3). Although many studies demonstrated that cytosolic T₃ binding proteins are present in vitro, little is known about the molecular mechanisms of T₃ retention in cytoplasm, especially, the physiological roles of cytosolic T₃in vivo (4).

In 1986, it was reported that charcoal treatment of rat renal cytosol abolishes the hormone binding activity, whereas the addition of reduced nicotinamide adenine dinucleotide phosphate (NADPH) restores it (5). Two proteins were identified. One was purified from rat kidney (6) and brain (7) and showed a molecular mass of 58 kDa. The other was obtained from rat liver (8), astroglial cells (9), and human kidney (10) and showed a molecular mass of 38 kDa. The latter was identified as μ-crystallin (CRYM), which was initially cloned from kangaroo lens (11, 12). CRYM is a taxon-specific protein that is particularly abundant in kangaroo lens, but not in mouse or other mammals (13). Our previous study demonstrated that CRYM held T₃ in cytoplasm, which increased the T₃ concentration in both whole cells and the nuclei in CRYM-expressing GH3 cells (14). However, expression of CRYM suppressed transactivity mediated by T₃ in the same cell lines.

Recently, it was reported that CRYM is abundant in cochlear and vestibular tissues in the inner ear (15). Immunohistochemical study demonstrated that the protein was preferentially expressed in fibrocyte type 2 where Na-K ATPase is enriched (16). Two patients with nonsyndromic deafness were reported to be associated with point mutations in the CRYM gene (15). In a study assessing NADPH-dependent T₃ binding to the two reported mutations (16), one mutation, T314Y, abolished NADPH-dependent T₃ binding, whereas the other did not influence the binding activity. The proband with T314Y showed severe hearing loss, whereas the patient with X314T demonstrated moderately impaired hearing. It is suggested that the thyroid hormone binding property may relate to the development of hearing impairment in patients with the CRYM mutations.

To clarify the physiological roles of CRYM in vivo, we generated mice in which the CRYM gene was disrupted. We assessed the phenotypes, especially the heart rate, hearing ability, kinetics of the thyroid hormone, serum concentrations of thyroid hormones, and the mRNA expression of the thyroid hormone response genes.

RESULTS

Production of Mice with Inactivated CRYM

To inactivate the CRYM gene, a recombination cassette containing a NeoR gene driven by mouse phosphoglycerate kinase promoter was introduced in the first and second coding exons (Fig. 1A). The mutated allele was introduced into the CRYM locus by homologous recombination in embryonic stem cells. Heterozygous mice were derived from an inbred C57BL6 background. The chimeric mouse was mated to 129Sv mice and then back-crossed six times into the same strain, diluting the C57BL6 background.

Expression of CRYM Gene

Northern blotting demonstrated a single 2.2-kb band in the lane applied brain total RNA from the wild (CRYM+/+) mice (Fig. 1B, left panel). The signal was attenuated in the lane of the CRYM+/− mice. There was no signal detected in the CRYM−/− mice. Western blotting showed a single 38-kDa band in the CRYM+/+ and CRYM+/− mice, but not in the CRYM−/− mice (Fig. 1B, right panel).

NADPH-Dependent T₃ Binding Activity

NADPH-dependent T₃ binding was assessed in whole brain, heart, kidney, and liver of the CRYM+/+ and CRYM−/− mice. As shown in Fig. 1C, there were no specific bindings to T₃ in the charcoal-treated crude cytosol obtained from any of the tissues we studied (black and shaded bars). When we added NADPH to the charcoal-treated extracts, specific binding to T₃ was observed in all tissues from the CRYM+/+ mice, as previously demonstrated (white and hatched bars) (17). These binding activities were abolished using extracts from the CRYM−/− mice. Scatchard analyses demonstrated that the association constant (K_a) in the CRYM+/+ mice was approximately 2 × 10⁹/m (Fig. 1D). We could not calculate K_a in the CRYM−/− mice because of the lack of specific binding.

Growth Curve, Heart Beats, and Hearing Ability

The mice were intercrossed to produce homozygous mutants. Among 155 pups born, 34 (22%), 87 (56%), and 34 (22%) were wild-type, heterozygous, and homozygous progenies, respectively. These data show that homozygous disruption of the CRYM gene is not deleterious to embryonic development. The growth rate of heterozygous mice was similar to that of wild-type mice (Fig. 2A). Low and high heart rates were shown in wild-type mice fed a low-iodide diet plus 1% propylthiouracil (PTU) and 1 μg/ml T₃ water ad libitum, respectively, as controls. There were no significant differences in the heart rates among CRYM+/+, CRYM+/−, or CRYM−/− mice (Fig. 2B). To assess hearing, auditory-evoked brainstem response (ABR) was measured in CRYM+/+ and CRYM−/− littermates. The responses were apparently normal in CRYM−/− mice (Fig. 2C). The thresholds of CRYM−/− mice were not significantly different from those of CRYM+/+ mice (Fig. 2D).

Serum Concentrations of T₃, T₄ and TSHβ mRNA Expression in the Pituitary Gland

We measured the serum concentrations of T₃ and T₄ in CRYM+/+, CRYM+/−, and CRYM−/− mice. As shown in Fig. 3A, the concentrations of both T₃ and T₄ were significantly lower in the CRYM−/− mice than in CRYM+/+ mice. T₃ and T₄ concentrations were reduced by 13% and 25% in CRYM−/− mice, respectively. In contrast, there was no significant difference in TSHβ mRNA expression between the CRYM+/+ and the CRYM−/− mice. The data from the Northern blotting were consistent with the findings obtained from the quantitated PCR (Fig. 3B). As shown in Fig. 3C, there were no apparent histological differences between CRYM +/+ and CRYM −/− mice in either the pituitary or thyroid gland.

mRNA Levels of T₃-Responsive Genes in the Liver, Kidney, and Heart

To evaluate peripheral activity induced by thyroid hormone in knockout mice, we measured the expressions of Gsta2 and Dio1 mRNAs in the liver and kidney and the expression of Dio1 mRNA in the heart. There were no significant differences in Gsta2 or Dio1 mRNA expression in either the liver or kidney among CRYM+/+, CRYM+/−, and CRYM−/− mice (Fig. 4A). In the heart, the expression of Dio1 in CRYM+/+ mice was not different from that in CRYM−/− mice. The expression of Gsta2 was not performed because the expression was too low to evaluate in this system. Northern blotting also demonstrated similar results to the data using the quantitated PCR (Fig. 4B).

T₃ Kinetics in Wild and Homozygous Mice

To investigate the physiological roles of CRYM in vivo, we injected labeled T₃ iv and measured the radioactivity in various tissues 1, 2, 4, and 8 h post injection. In the CRYM+/+ mice, radioactivity in all tissues studied except for the brain gradually increased for 2 h after injection and then slowly decreased until 8 h post injection (Fig. 5A–D). In the brains of CRYM+/+ mice, it took 4 h to reach the peak radioactivity. However, the radioactivities in all tissues peaked 1 h post injection and then sharply declined in the CRYM−/− mice. There were no significant changes in serum radioactivities between the CRYM+/+ and CRYM−/− until 8 h post injection (Fig. 5E).

DISCUSSION

In this study, we succeeded in disrupting the CRYM gene and omitting the expression of the protein in vivo. NADPH-dependent T₃ binding activity was eliminated in the tissues we studied, suggesting that the CRYM protein is a unique protein among cytosolic proteins that are capable of binding to T₃ in the presence of NADPH. We have reported two proteins of different molecular weights, which possess high affinity to T₃ in the presence of NADPH in rat tissues. One was purified from rat kidney and showed a molecular weight of 4.7S, and mass of 58 kDa (6). The other was isolated from rat liver and showed a molecular mass of 38 kDa (8). The molecular weight was 5.1S, because this protein formed a dimer. Currently, we do not have evidence to explain why NADPH-dependent T₃ binding completely vanished even though one of the proteins should have survived. CRYM protein may be affected by posttranslational modification, such as glycosylation, myristoylation, or other processes. Such posttranslational modification may alter the molecular weight. As a result, there are two molecular weights of the protein originating from a single protein. Furthermore, the antibody may not recognize the modified protein. These proteins have also been reported (18).

On the initial phenotypical assessment, there were no significant differences in terms of growth rate, heart rate, and hearing ability. Hearing is one of the most important functions controlled by thyroid hormone (19, 20). It is reported that CRYM mutations are associated with nonsyndromic deafness. The disruption of the CRYM gene apparently does not alter the hearing function. From these data, the expression of abnormal CRYM proteins may affect the hearing function of the inner ear in patients with CRYM mutations. We suggest that the mutation that abolishes NADPH-dependent T₃ binding causes severe impairment of the hearing as previously reported (16). Taken together, these data indicate that other factors, including thyroid hormone, may also affect hearing loss.

In contrast to normal growth, heart rate, and hearing ability, disruption of the gene decreased the serum concentration of T₃ without alteration of the TSHβ concentration in the pituitary gland. T₄ concentration was also suppressed by the disruption of the gene. It was reported that the binding affinity of CRYM to T₄ is approximately 5–10 times lower than that to T₃in vitro (8). These data suggest that CRYM disruption may suppress the T₄ concentration directly, or through the feedback loop in TSH-releasing factor-TSH-thyroid axis indirectly in vivo. Because of the small change in the thyroid hormone concentration, we note the possibility that a correspondingly small change exists in the amount of TSHβ mRNA that has not been detected.

Gsta2 and Dio1 are reported to be negatively and positively regulated by T₃, respectively (21, 22). Although the concentration of T₃ and T₄ were low in CRYM−/− mice, the expression level of T₃ target genes was normal rather than low in CRYM−/− mice.

A previous study demonstrated that CRYM affected cellular retention of T₃ in living cells (14). This observation presupposes that the disruption of the CRYM gene may alter the kinetics of T₃in vivo. We counted the radioactivity of the tissues after injection of the labeled T₃ iv via the tail vein. Both the influx and efflux of radiolabeled T₃ were increased in all of the tissues we studied in these knockout mice. These findings are consistent with the previous data obtained using the CRYM-expressing cells.

Up to the present, four cytosolic thyroid hormone-binding proteins have been reported, i.e. pyruvate kinase subunits (23), aldehyde dehydrogenase (24), glutathione-S-transferase (25), and CRYM. Although the affinity constants of these proteins to T₃ are 8–34 times lower than that of CRYM, these proteins may have physiological functions for cytoplasmic T₃ retention or transport. MCT8 or other T₃ transporters in plasma membrane also contribute to the regulation of intracellular T₃ concentration (26, 27).

CRYM proteins may have redundant functions. The protein binds not only T₃ but also NADPH (28). The physiological significance of this coupling is not currently known. Because nicotinamide adenine dinucleotide phosphate inhibits T₃ binding, oxidative stress may be another factor for control of T₃ binding to CRYM in cytoplasm (8).

In conclusion, we disrupted NADPH-dependent T₃ binding after the elimination of CRYM expression. Normal growth, heart rate, and hearing were observed. The expressions of TSHβ, Gsta2, and Dio1 were not changed by the disruption, although the serum concentration of thyroid hormone was altered. T₃ rapidly entered into and escaped from the tissues. These data suggest that CRYM affects T₃ kinetics in vivo and that elimination affects the serum concentrations of T₃ and T₄ but not the apparent phenotype and peripheral thyroid hormone action in the euthyroid state. The precise molecular mechanism of CRYM action remains to be elucidated.

MATERIALS AND METHODS

Generation of the CRYM Knockout Mouse

The CRYM knockout mouse was generated by in Genious Targeting Laboratory (Stony Brook, NY) on a commercial basis. The CRYM gene was cloned from a 129Sv genomic library (Stratagene, La Jolla, CA) and was used to construct targeting vector containing the neomycin resistance gene flanked by two genomic fragments spanning the CRYM gene promoter, Exon 1 and 2 regions. The long fragment of the targeting construct was a 7.6-kb sequence upstream from Exon 1. The short fragment of the targeting construct was 1.2-kb sequence downstream of Exon 2. The targeted allele contained a deletion of the CRYM gene promoter, Exon 1, and Exon 2. IT2 embryonic stem cells were transfected with the targeting vector and incubated in media containing G418. Surviving colonies were analyzed by PCR to identify homologous recombinants. One correctly targeted embryonic stem cell clone was microinjected into C57BL/6J host blastocysts to generate chimeric mice, which were bred with C57BL/6J mice to obtain heterozygous offspring. Mouse genotyping was performed with genomic DNA isolated from toes using sodium dodecyl sulfate/proteinase K solubilization. The wild-type CRYM allele was identified by PCR using primers a (5′-TGCATCCCTGAAGTGGGGTA-3′) and b (5′-GAATGAGCGCAAACGGAATG-3′), which amplified an 800-base product. The knockout allele was identified by PCR using primers a and c (5′-AGGCAGAGGCCACTTGTGTAG-3′), which amplified a 500-base product-containing fragment of the neomycin-resistant cassette. Heterozygous mice were crossed with the 129Sv strain at least six times before they were interbred to generate homozygous knockout mice. The wild-type mice used in this study as controls were littermates of the knockout mice. Procedures carried out in mice were approved by Shinshu University Institutional Animal Care and Use Committee.

Northern and Western Blotting

In Fig. 1, the brains were obtained from 3-wk-old mice for the Northern and Western analyses. Northern hybridization was performed by the standard procedures using random-primed full length CRYM fragment as a probe (29). In Figs. 3B and 4B, the pituitary gland, liver and kidney were obtained from 6-wk-old male mice. Total RNA (10 μg) obtained from six pituitary glands was electrophoresed and hybridized with ³²P-labeled TSHβ or elongation factor 1 (EF1) probe. Total RNA (20 μg) obtained from 50 mg liver or kidney was electrophoresed and hybridized with labeled Dio1, Gsta2, or EF1. A series of the mouse probes were obtained from PCR amplification of the reverse-transcribed poly A RNA extracted from the pituitary gland or liver. The following oligonucleotides were used.

TSHβ sense, 5′-GAACGGAGAGTGGGTCATCACA-3′;

antisense, 5′-AGTAGTTGGTTCTGACAGCCTC-3′;

Dio1 sense, 5′-TGTAGGCAAGGTGCTAATGAC-3′;

antisense, 5′-TGCCGGATGTCCACGTTGTTC-3′;

Gsta2 sense, 5′-GCAGAATGGAGTGCATCAGGT-3′;

antisense, 5′-TCTGCTCTTGAAGGCCTTCAGC-3′;

EF1 sense, 5′-CCATGAAGCTTTGAGTGAAGCTCT-3′;

antisense, 5′-TAGCCTTCTGAGCTTTCTGGGCAG-3′.

Amplified fragment was sequenced and fidelity of the sequences was checked. Western blotting was carried out by using anti-CRYM antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA).

Scatchard Analysis

Scatchard analyses and NADPH-dependent T₃ binding assays were performed as previously described (28). Briefly, tissues were homogenized and centrifuged to isolate the crude cytosolic fraction. The fraction was incubated with 1% charcoal to remove T₃, NADPH, and other small molecules. After spinning down, the supernatant was incubated with [¹²⁵I]T₃ in the presence or absence of NADPH and/or unlabeled T₃ for 20 min on ice. After the brief incubation with charcoal and the following centrifugation, the supernatant and pellet were counted by the γ-counter.

Heart Rate

For the electrocardiogram measurements, mice were anesthetized either with ketamine (100 mg/kg) and xylazine (10 mg/kg) via ip injection as described previously (30). Briefly, mice were positioned prone in a shielded box. The bottom of the shielded box was heated to 37 C using a heating pad to maintain a stable body temperature during the experiment. UAS-108S (Unique Medical, Tokyo, Japan) was used to record bipolar limb leads in standard fashion. For each animal, R-R intervals were measured in a blinded fashion from three consecutive beats in the leads and averaged.

Measurement of Serum T₃ and T₄

Peripheral blood (400 μl) was corrected from the 6-wk-old male mouse. Serum T₃ and T₄ were measured by Chemilumi ACS-T3, T4 (Bayer Med, Tokyo, Japan). Six individual mice from each genotype group were used in the study.

ABR

The chimeric mouse was mated to 129Sv mice and then back-crossed six times into the same strain, diluting the C57BL6 background. Male mice (6 wk old) were used in this study. Mice were anesthetized with ketamine (100 mg/kg)/xylazine (10 mg/kg) by ip injection, and the ABR tests were performed with a heating pad to maintain the body temperature in a soundproof room as previously described (31). Click stimuli was performed at synthesized durations and specified amplitudes using a digital signal processing platform (Tucker-Davis Technologies, Alachua, FL), and analyzed with PowerLab systems (ADInstruments, Colorado Springs, CO) as described elsewhere (32). Click stimuli were 0.1-msec clicks, composed of a square pulse of 0.1 msec duration. Stainless steel needle electrodes were placed at the vertex and ventrolateral to the left and right ears for the recording, and a tweeter modified with a coupler was inserted into the external canal to deliver acoustic stimuli. ABR waveforms were recorded in 5- to 10-decibel (dB) intervals down from a maximum amplitude of 85 dB until no waveform could be visualized.

Measurement of TSHβ, Gsta2, and Dio1

The contents of TSHβ, Gsta2, and Dio1 were measured as follows. After iv injection with ice-cold PBS, pituitary glands, livers, and kidneys were removed and later washed with RNA (QIAGEN, Chatsworth, CA). Total RNA was prepared by using RNeasy kit (QIAGEN). After reverse transcription with avian myeloblastosis virus reverse transcriptase, samples were applied with Quantitect SYBR Green PCR kit (QIAGEN) or TaqMan universal PCR Master Mix to ABI PRISM 7300 Sequence detection system according to a manufacturer’s protocol. Mouse TSHβ, β-actin, Dio1, and Gsta2, were measured with the following oligonucleotides.

TSHβ sense, 5′-GCTCGGGTTGTTCAAAGCATGA-3′

antisense, 5′-CCCACAAGCAAGAGCAAAAAGCA-3′

β-actin sense, 5′-TCTCCAGAGGCACCATTGAAATTCT-3′

antisense, 5′-CGCTGGCTCCCACCTTGTCT-3′

Dio1 sense, 5′-CCAGTTCAAGAGACTCGTAGATGAC-3′

antisense, 5′-GCGTGAGCTTCTTCAATGTAAATGA-3′

FAM-TCCACAGCCCATTTC-3′

Gsta2 sense, 5′-AAGACTACCTTGTGGGCAACAG-3′

antisense, 5′-CTGGCATCAAGCTCTTCAACATAGA-3′

FAM-CCTGCTGGAACTTC-3′

The amount of 18s rRNA was measured as described in the manufacturer’s protocol using commercially available probes (Applied Biosystems, Foster City, CA).

Tracer Experiments

Mice were kept under a 12-h light, 12-h dark cycle and provided food and water ad libitum. Three or four CRYM+/+ and three CRYM−/− male littermates were used for each group of treatment. Each mouse was injected iv with 0.2 μCi (91 pmol) of [¹²⁵I]T₃ dissolved into 50 μl of saline. At 1, 2, 4, and 8 h after injection, the mice were killed and tissues were removed, after which the blood samples were drawn. Each tissue and blood sample was weighed and then evaluated for radioactivity with γ-counter.

Statistics

In Figs. 2D and 5, P values were calculated by an ANOVA unpaired t test using Statview 4.1 software (Abacus Concepts, Inc., Berkely, CA). In Figs. 3 and 4, statistical significance was determined by an ANOVA followed by the Bonferroni multiple comparison test. P values ≥ 0.05 were considered not significant.

Acknowledgments

We thank Ms. Izumi Kinoshita for her technical assistance. We are also grateful to Ms. Tomoko Nishizawa for preparation of the tissues.

This work was supported in part by a Grant-in-Aid for Scientific Research no. 17590957 from the Ministry of Education, Science, Sports, and Culture, Japan.

Disclosure Statement: The authors have nothing to disclose.

Abbreviations:

 
• ABR,

  Auditory-evoked brainstem response;

 
• CRYM,

  μ-crystallin;

 
• dB,

  decibel;

 
• Dio1,

  deiodinase 1;

 
• EF1,

  elongation factor 1;

 
• Gsta2,

  glutathione-S-transferase α2;

 
• NADPH,

  reduced nicotinamide adenine dinucleotide phosphate;

 
• PTU,

  propylthiouracil.