Muscle Liver Muscle Liver Muscle Liver Muscle Liver Muscle Liver Muscle Liver Muscle Liver Muscle Liver Muscle

We have cloned the cD)NA encoding glycogen phosphorylase (1,4-a-D-glucan:orthophosphate a-D-glucosyltransferase, EC 2.4.1.1) from human liver. Blot-hybridization analysis using a large fragment of the cDNA to probe mRNA from rabbit brain, muscle, and liver tissues shows preferential hybridization to liver RNA. Determination of the entire qucleotide sequence of the liver message has allowed a comparison with the previously determined rabbit muscle phosphorylase sequence. Despite an amino acid identity of 80%, the two cDNAs exhibit a remarkable divergence in G+C content. In the muscle phosphorylase sequence, 86% of the nucleotides at the third codon position are either deoxyguanosine or deoxycytidine residues, while in the liver homolog the figure is only 60%, resulting in a strikingly different pattern of codon usage throughout most of the sequence. The liver phosphorylase cDNA appears to represent an evolutionary mosaic; the segment encoding the N-terminal 80 amino acids contains >90% G+C at the third codon position. A survey of other published mammalian cDNA sequences reveals that the data for liver and muscle phosphorylases reflects a bias in codon usage patterns in liver and muscle coding sequences in general. Glycogen phosphorylase (1,4-a-D-glucan:orthophosphate aD-glucosyltransferase, EC 2.4.1.1) isozymes play a vital role in mobilization of stored sugar in a variety of mammalian tissues. Three forms of the enzyme have been described that are distinguished by their electrophoretic mobilities on gels and their immunological properties (1-3). The three isozymes are tissue-specific; the brain type (also known as the fetal type) is predominant in adult brain and embryonic tissues, while the liver and muscle types are predominant in adult liver and skeletal muscle tissues, respectively (reviewed in ref. 4). The muscle form is the best characterized; both the primary sequence and the x-ray structure of rabbit muscle phosphorylase are known (5-8). The enzyme functions in muscle to provide glucose required for the energy of contraction. Its physiological role is distinct from the liver isozyme, which is centrally involved in the maintenance of blood glucose homeostasis, and from the brain form, which is associated primarily with provision of an emergency glucose supply during brief periods of anoxia or hypoglycemia (4, 9). Comparisons of the protein and DNA sequences of the phosphorylase isozymes are required to understand the evolutionary and functional relationships among them. Further, such comparisons could ultimately provide insight into how the phosphorylase genes, and perhaps other multigene families, are regulated in a developmental and tissue-specific manner. We have described the cloning and sequencing ofthe rabbit muscle glycogen phosphorylase cDNA and portions of the C-terminal domain from human and rat muscle (6, 7). We report here the cDNA sequence and the deduced amino acid sequence for human liver glycogen phosphorylase. Comparison of liver and muscle phosphorylase sequences reveals extensive amino acid identity between the two isozymes. Remarkably, the nucleotide sequences are found to be less homologous because of a difference in codon usage patterns. Furthermore, a survey of published sequences reveals that the difference between the liver and muscle phosphorylase nucleotide sequences reflects a general tissue-specific codon usage bias in mammalian liver and muscle cDNA sequences. MATERIALS AND METHODS Cloning and Nucleotide Sequencing Strategy for the Human Liver Glycogen Phosphorylase cDNA. A summary of the cloning strategy is shown in Fig. 1. A partial rabbit muscle phosphorylase cDNA (6) encoding amino acids 573-742 (which includes the conserved pyridoxal phosphate cofactor binding site) was labeled with 32P by nick-translation and used to screen a phage Xgtll cDNA library prepared from human liver (courtesy of A. DiLella and S. Woo, Baylor College of Medicine, Houston, TX). The initial screening of 50,000 clones, performed at low stringency [30% (vol/vol) formamide containing 5x NaCl/Cit (1x = 0.15 M NaCl/0.015 M sodium citrate), 5 mM NaH2PO4, 0.2 mg of salmon sperm DNA per ml, and Denhardt's solution (0.02% polyvinylpyrrolidone/0.02% Ficoll/0.02% bovine serum albumin) at 42°C], yielded a single clone of about 750 base pairs (HL-1). Sequencing by the dideoxynucleotide method (10) showed that this fragment was homologous to rabbit muscle phosphorylase over the region of the cDNA encoding amino acids from position 660 to the C terminus. The HL-1 fragment was amplified, purified, radiolabeled, and used to rescreen 100,000 clones from the same library under conditions of increased stringency (40% formamide at 42°C), yielding the overlapping fragments HL-2, HL-3, and HL-4. The HL-4 fragment was subsequently used to screen an additional 100,000 clones, yielding the overlapping fragments HL-5, HL-6, HL-7, and HL-8. When the HL-8 fragment was excised from Xgt11 with the restriction endonuclease EcoRI, a second fragment of -360 nucleotides (HL-9) was observed because of an internal EcoRI site within the cDNA as shown in Fig. 1. HL-9 was used to screen an additional 500,000 clones, yielding only HL-10, which terminates 23 amino acids from the 5' end of the coding region. To clone the last small coding fragment (HL-11), we used HL-9 to screen 80,000 clones from a second, randomly primed human liver cDNA library (courtesy of Jing-hsiung Ou, Hormone Research Institute, University of California, San Francisco). Blot-Hybridization Analysis. Poly(A)+ RNA samples, prepared by the method of Ashley and MacDonald (11), were electrophoresed through a formaldehyde/agarose gel, transferred onto a MSI magna nylon 66 membrane filter, and Abbreviation: kb, kilobase(s). 8132 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. Proc. Natl. Acad. Sci. USA 83 (1986) 8133