.

.

.

.

.

.

..

Research Program:

The primary interest of my laboratory is the physiological function of the hypermodified nucleoside queuosine (Q), which is widely distributed in the tRNA of bacteria, plants and animals (see Figure 1 for the structure of Q). Queuosine occurs in the first position of the anticodon of tRNAs that recognize NAU and NAC codons (Tyr, Asn, Asp and His). Bacterial synthesis of Q includes a base-for-base exchange of guanine in the unmodified tRNA with 7-methylamino-7 deazaguanine (preQ₁), a reaction catalyzed by tRNA-guanine transglycos- ylase (TGT). The final steps of Q synthesis occur at the polynucleotide level; e.g., the cyclopentenediol moiety attached to the 7-methylamino group. In contrast, eucaryotes synthesize Q by a base-for-base exchange of guanine in the unmodified tRNA with free queuine (the base of Q). Further, eucaryotes cannot synthesize queuine and must obtain it from the diet or gut flora. The sequence of the catalytic subunit of TGT is highly conserved throughout the biosphere (discussed in M. Watanabe et al., 1997).

    Despite many studies, the normal physiological role of Q remains ill defined. Q-modified tRNA recognizes NAU codons better than NAC codons and Q-deficient tRNA has been associated with compromised translation efficiency or fidelity in specific instances. Indirect evidence suggests that in eucaryotes Q may be involved in differentiation, proliferation, cellular signaling, and the oxidative stress response. Q-hypomodification has long been associated with neoplastic cells, especially poor prognosis/malignant tumors. However, TGT-null mutants of E. coli show no obvious phenotype, and starvation for queuine does not noticeably affect growth or development in germ-free mice.

    My laboratory discovered that an established human colon tumor cell line, GC₃, is null for the expression of TGT (Gündüz et al., 1992). This was an unexpected finding because of the absence of a selective phenotype for cells with Q -deficient tRNA. The GC₃ cell line remains the sole published example of an animal cell that does not express TGT. The occurrence of TGT-null expression in GC₃ cells, together with the frequent occurrence of Q-deficiency in the tRNA of tumors, supports my theory that TGT may be encoded by a tumor suppressor gene; i.e., that lesions in Q metabolism favor and, therefore, are selected by tumor progression.

    To pursue the relation of Q hypomodification to neoplasia we set out to clone human TGT, which appears to be a heterodimer of 60-kDa and 44-kDa subunits.

Cloning human TGT: The 60-kDa subunit. For technical reasons, we first cloned and sequenced the rabbit cDNA for the 60-kilodalton subunit of TGT (GenBank # L37420). The sequence of the cDNA for the rabbit TGT 60 kDa subunit enabled us to clone and sequence the cDNA for the human placenta TGT 60 kDa subunit (GenBank # U30888) as well as the homolog from Caenorhabditis elegans (GenBank # U32223). This work was described in Deshpande et al., 1996.

    Subsequent work in my laboratory and by others revealed that the cloned 60-kDa peptide is not the catalytic subunit of TGT. Rather, the 60-kDa peptide is a new member of a large family of deubiquitinating enzymes (ubiquitin C-terminal hydrolases) whose roles in cellular physiology are still being defined. The possibility of a linkage between nucleoside Q and the ubiquitin- dependent proteolytic pathway suggests an attractive model: TGT, functioning as a sensor of Q deficiency (and, perhaps, decreased translational fidelity), may modulate the ubiquitin-dependent degradation of a protein(s) that fosters neoplastic progression if abnormally expressed. Accordingly, mutations in TGT would decouple this Q -tRNA/ubiquitin linkage and contribute to malignancy.

The 44-kDa Catalytic Subunit. More recently, a search of the human EST database for sequences with significant homology to tRNA-guanine transglycosylase from E. coli identified several potentially full-length (1.3-1.4 kb) cDNA clones. Three clones were obtained and their DNA sequences have been determined.

The Future. Availability of the TGT catalytic (44-kDA) subunit will enable us to test its interaction with the 60-kDa subunit cloned previously. If the two proteins interact and modulate each other's activity, a strong case can be made for a direct intersection of Q metabolism and a deubiquitinating activity and the potential role of this phenomenon in neoplastic progression.

Selected Publications

Gündüz, U., M.S. Elliott, P.H. Seubert, J.A. Houghton, P.J. Houghton, R.W. Trewyn, and J.R. Katze. Absence of tRNA-guanine transglycosylase in a human colon adenocarcinoma cell line. Biochimica et Biophysica Acta, 1139 (1992) 229-238.

Deshpande, K.L., P.H. Seubert, D.M. Tilman, W.R. Farkas, and J.R. Katze. Cloning and characterization of cDNA encoding the rabbit tRNA-guanine transglycosylase 60-kilodalton subunit. Archives of Biochemistry and Biophysics, 336 (1996) 1-7

Watanabe, M., M. Matsuo, S. Tanaka, H. Akimoto, S. Asahi, S. Nishimura, J.R. Katze, T. Hashizume, P.F. Crain, J.A. McCloskey, and N. Okada. Biosynthesis of archaeosine, a novel, derivative of 7-deazaguanosine specific for Archael tRNA, proceeds via a pathway involving base replacement on the tRNA polynucleotide chain. Journal of Biological Chemistry, 272 (1997) 20146-20151.